PhD thesis

FACULTY OF HEALTH AND MEDICAL SCIENCES UNIVERSITY OF COPENHAGEN

PhD thesis Niels Peter Hell Knudsen Faculty of Health and Medical Sciences University of Copenhagen Department of Infectious Disease Immunology Statens Serum Institut

Comparative analysis of tuberculosis vaccines containing type VII secretion system protein substrates and the importance of ESX-1 secreted proteins for Mtb virulence and pathology

Academic advisors: Allan Randrup Thomsen , Professor, MD University of Copenhagen Claus Aagaard, PhD Statens Serum Institut Else Marie Agger, PhD Statens Serum Institut Submitted: January 2017

Comparative analysis of tuberculosis vaccines containing type VII secretion system protein substrates and the importance of ESX-1 secreted proteins for Mtb virulence and pathology

Niels Peter Hell Knudsen PhD thesis 2017

This thesis has been submitted to the Graduate School of the Faculty of Health and Medical Sciences, University of Copenhagen

Comparative analysis of tuberculosis vaccines containing type VII secretion system protein substrates and the importance of ESX-1 secreted proteins for Mtb virulence and pathology Niels Peter Hell Knudsen Submitted January 30th 2017

Department of Infectious Disease Immunology Center for Vaccine Research Statens Serum Institut, Denmark

Principal supervisor: Allan Randrup Thomsen, Professor Department of Immunology and Microbiology University of Copenhagen, Denmark

Co-supervisors: Claus Aagaard, PhD Department of Infectious Disease Immunology Statens Serum Institut, Denmark

Else Marie Agger, PhD Department of Infectious Disease Immunology Statens Serum Institut, Denmark

Assessment committee: Charlotte Menne Bonefeld, Associate Professor (Head of committee) Department of Immunology and Microbiology University of Copenhagen, Denmark

Gregers Jungersen National Veterinary Institute Technical University of Denmark, Denmark

Andrea Cooper, Professor Department of Infection, Immunity and Inflammation University of Leicester, United Kingdom

Preface This thesis has been submitted to the Graduate School at the Faculty of Health and Medical Sciences, University of Copenhagen to meet the requirement for obtaining a PhD degree. The studies presented here were primarily conducted at the Department of Infectious Disease Immunology, Statens Serum Institut, Copenhagen, Denmark under the supervision of Claus Aagaard, Else Marie Agger and Allan Randrup Thomsen. The work was supported by a scholarship from the Danish Council for Independent Research (DFF – 1333-00194).

The overall objectives of this thesis were to: 1. Evaluate the efficacy of a novel vaccine strategy targeting the virulence-associated type seven secretion systems of Mycobacterium tuberculosis 2. Assess the contribution of adjuvants in subunit vaccinations in a head-to-head study of clinical human adjuvants against three different disease targets 3. Determine the cause behind the lack of protection by H56 against the clinical isolate DK9897 isolated from a patient with extrapulmonary TB 4. Characterize the host-pathogen interactions in the absence of ESX-1/EsxA secretion by infection with DK9897

The main outcomes are presented in the following four full-length articles: • Niels Peter H. Knudsen, Sara Nørskov-Lauritsen, Gregory M Dolganov, Gary K Schoolnik, Thomas Lindenstrøm, Peter Andersen, Else Marie Agger & Claus Aagaard; “Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostics”, Proceedings of the National Academy of Sciences 111 (3), 1096-1101, 2014. • Niels Peter H. Knudsen, Anja Olsen, Cecilia Buonsanti, Frank Follmann, Yuan Zhang, Rhea N. Coler, Christopher B. Fox, Andreas Meinke, Ugo D´Oro, Daniele Casini, Alessandra Bonci, Rolf Billeskov, Ennio De Gregorio, Rino Rappuoli, Ali M. Harandi, Peter Andersen & Else Marie Agger; “Different human vaccine adjuvants promote distinct antigen-independent immunological signatures tailored to different pathogens”, Scientific reports 6, 2016. • Helena Strand Clemmensen*, Niels Peter H. Knudsen*, Erik Michael Rasmussen, Jessica Winkler, Ida Rosenkrands, Ahmad Ahmad, Troels Lillebaek, David R. Sherman, Peter Andersen, Claus Aagaard; “An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology”, Submitted to Scientific reports Dec 2016. *Co-first authors • Niels Peter Hell Knudsen, Helena Strand Clemmensen, Erik Michael Rasmussen, Ahmad Ahmad, Troels Lillebaek, Peter Andersen, Claus Aagaard; “A clinical isolate able of causing chronic disease in the absence of ESX1 secretion exhibits alternative immune pattern in murine challenge model”, Full-length manuscript ready for submission.

Acknowledgements First, I would like to express my gratitude towards my supervisors Claus Aagaard, Else Marie Agger and Allan Randrup Thomsen for giving me the opportunity to follow through with this PhD project under their excellent supervision and support.

To all of my colleagues at the Department of Infectious Disease Immunology, Statens Serum Institute, I want to express a deep appreciation for the supportive and enthusiastic working environment, with plenty of productive discussions and exchange of ideas over the years. I gratefully acknowledge Thomas Lindenstrøm and Rolf Billeskov for all the valuable scientific input. I also want to thank my fellow students Signe Tandrup Scmidt, Thomas Blauenfeldt, Jonathan Filskov, P. Nissen, Line Lindebo Holm, Helena Strand Clemmensen and Rasmus Skaarup Mortensen for creating a work environment that was both fruitful and fun! Special mention to Malene Aaby Neustrup for making the long hours in the office during the last month bearable. I am grateful for all the assistance for my experiments, with a special thanks to Sandra Isling, Linda Christensen, Camilla Haumann Rasmussen and Merete Henriksen. A big thanks goes to Rune Fledelius Jensen for always being ready with a giant hug when it was needed. But most of all, a big thanks goes out to Josh Woodworth for restoring my faith in religion by exemplifying that the existence of a god on earth is not only possible but is happening right now! All hail Josh! (Told you I was gonna keep it! Thanks man!). Last, but not least, the expert technical help of all the animal caretakers at the SSI is greatly appreciated, with a special thanks to Anniezette Lander and Flemming Kim Petersen.

Finally, I am grateful for all of my friends and family who have been a great support throughout the whole period. The biggest thanks goes out to my dear girlfriend Luna without whom this PhD would have never been possible. Thanks for all your love and support when it was needed the most.

Niels Peter Hell Knudsen January 30th 2017

Table of content SUMMARY - ENGLISH ........................................................................................................ I SUMMARY - DANISH ........................................................................................................ III ABBREVIATIONS .............................................................................................................. V INTRODUCTION ................................................................................................................. 1 General background of TB .............................................................................................................................................. 1 TB epidemiology ........................................................................................................................................................... 1 History of TB ................................................................................................................................................................. 2 Transmission and Progression ....................................................................................................................................... 4 Clinical manifestation .................................................................................................................................................... 4 Diagnostic tools ............................................................................................................................................................. 5 Chemotherapy and drug resistance ................................................................................................................................ 6 Mycobacterium tuberculosis .......................................................................................................................................... 7 Type VII secretion systems ............................................................................................................................................ 8 Immunology of TB .......................................................................................................................................................... 10 A brief overview .......................................................................................................................................................... 10 Innate immunity ........................................................................................................................................................... 11 The granuloma ............................................................................................................................................................. 12 Adaptive immunity ...................................................................................................................................................... 14 Mtb Immune evasion ................................................................................................................................................... 19 The murine TB challenge model .................................................................................................................................. 20 Current TB vaccine and future candidates .................................................................................................................. 21 BCG vaccine ................................................................................................................................................................ 21 Novel TB vaccines in development ............................................................................................................................. 22

RESULTS .......................................................................................................................... 25 Manuscript I.................................................................................................................................................................... 25 Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostic ............... 25 Manuscript II .................................................................................................................................................................. 26 Different human vaccine adjuvants promote distinct antigen-independent immunological signatures tailored to different pathogens....................................................................................................................................................... 26 Manuscript III................................................................................................................................................................. 26

An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology. ............................................................................................................................... 26 Manuscript IV ................................................................................................................................................................. 27 A clinical isolate able of causing chronic disease in the absence of ESX-1 secretion exhibits alternative immune pattern in murine challenge model ............................................................................................................................... 27

UNPUBLISHED RESULTS ............................................................................................... 29 Functional difference between EsxA and EsxH specific T cells .................................................................................. 29 Introduction.................................................................................................................................................................. 29 Results.......................................................................................................................................................................... 30 Discussion .................................................................................................................................................................... 32 EsxA based vaccine unable to protect against DK9897 while EsxH hybrid does...................................................... 34 Introduction.................................................................................................................................................................. 34 Results.......................................................................................................................................................................... 34 Discussion .................................................................................................................................................................... 35 Materials and Methods................................................................................................................................................... 37

DISCUSSION .................................................................................................................... 39 CONCLUDING REMARKS AND PERSPECTIVES .......................................................... 45 REFERENCES .................................................................................................................. 46 APPENDIX I ...................................................................................................................... 61 Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostics

APPENDIX II ..................................................................................................................... 75 Different human vaccine adjuvants promote distinct antigen-independent immunological signatures tailored to different pathogens

APPENDIX III .................................................................................................................... 93 An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology

APPENDIX IV .................................................................................................................. 125 A clinical isolate able of causing chronic disease in the absence of ESX-1 secretion exhibits alternative immune pattern in murine challenge model

Summary - English

Summary - English Tuberculosis (TB) is an infectious lung disease caused by an infection with Mycobacterium tuberculosis (Mtb). TB is the leading cause of death worldwide, on average claiming the life of an infected individual every 18 seconds. TB has resulted in the loss of nearly 1 billion lives over the past few centuries, killing around 1.8 million people every year. Mtb has coevolved with humans for tens of thousands of years, adapting to persist in the human host by carefully manipulating and evading the immune system. BCG, the only currently available TB vaccine, has been available since 1921, and despite being the most widely used vaccine to date, it has failed to put an end to the global TB pandemic. The BCG vaccine has shown large variations in protective efficacy which fades over time, lasting on average 10-15 years. Revaccination with BCG has no positive preventive effect, meaning that new and improved vaccines are urgently needed.

Many vaccines in the TB clinical pipeline are based on a narrow subset of Mtb antigens (e.g. Ag85A/B). The current work focuses on targeting substrates from the type seven secretion systems also known as the ESX systems in Mtb. The ESX substrates have been linked to several mechanisms mediating the virulence of the Mtb bacteria. In Manuscript I, we describe the process of rationally designing a novel TB vaccine. We found that, due to close homology between the substrates secreted through the same system, ESX-1 – ESX-5, targeting one substrate should ideally cover the remaining substrates from the same system. We evaluated pairs from each secretion system and with exception of EsxU-T, we found these to be protective against challenge with Mtb in mice. Considering factors like expression, immunogenicity and individual protection, we created a novel subunit fusion, H65, based on antigens secreted through ESX-2, -3 and -5. We found that the immunization with the H65 fusion in combination with CAF01 was able to induce immune responses towards the individual components of the fusion protein, EsxD-EsxC-EsxG-EsxH-EsxW-EsxV. The novel fusion protein was able to protect on par with H56, the phase II clinical trial TB fusion protein and BCG.

The majority of the work presented herein is based on vaccination with the adjuvant CAF01. There are however several other adjuvant formulations in the TB clinical pipeline, in combination with different fusion proteins. In Manuscript II, we describe the head-to-head comparison of five human vaccine adjuvants, all of which are on their way through clinical trials or part of already licensed vaccines. We evaluated the adjuvants Alum, MF59®, GLA-SE, IC31® and CAF01 in combination with antigens from Mtb, influenza and chlamydia. We found that each of the adjuvants induced a i

Summary - English

unique immunological signature, independent of which antigen with which it was combined. Of the tested adjuvants, GLA-SE, IC31® and CAF01 were all able to induce potent Th1 CD4 immune responses, specific for the vaccine target H56 (Ag85B-EsxA-Rv2660c). The responses were dominated by triple-positive IFN-γ+IL-2+TNF-α+ and double producing IL-2+TNF-α+ T cells, which are usually considered a key trait of central memory T cells. There was however no correlation between the magnitude of any of these responses to the degree of protection against TB.

The majority of studies, evaluating the mechanisms of Mtb or assessing the protection of a TB vaccine, have been performed using lab-adapted strains like Mtb Erdman or H37Rv. In order to address the growing concern that results found with these strains are not transferrable to infections with “real” Mtb strains, the study of more recent clinical isolates has emerged. In Manuscript III, we describe the characterization of a novel Mtb clinical isolate, Mtb DK9897, isolated from a patient with extrapulmonary TB. We identified the strain due to an unusual resistance towards H56, a vaccine currently in phase II clinical trials, while also being increasingly susceptible to BCG vaccination. Using in vitro cultures, we found that DK9897 was unable to secrete EsxA, a component of H56, and thus unable to be recognized by an EsxA specific CD4 T cell response. We demonstrate that the lack of EsxA secretion is likely caused by a mutation in the gene encoding EccCa1 and that reintroduction of this gene re-established EsxA secretion of DK9897. In Manuscript IV, we evaluate the consequences of the lack of ESX-1/EsxA secretion by DK9897 on infection in mice. In an aerosol infection, DK9897 exhibits delayed onset of exponential growth and diminished growth rates, which results in delayed onset of adaptive immunity. The overall in vivo attenuation due to the lack of EsxA causes decreased inflammation in the lungs, which also result in reduced recruitment of immune cells. However, in an intravenous (IV) infection model, DK9897 exhibits comparable systemic virulence as the EsxA secreting Mtb Erdman. The IV model results in an almost immediate priming of adaptive immunity, indicating that the suppressive/virulence functions mediated by EsxA are likely lung specific.

ii

Summary - Danish

Summary - Danish Tuberkulose (TB) er en smitsom lungesygdom forårsaget af en infektion med Mycobacterium tuberculosis (Mtb). TB er den hyppigste dødsårsag på verdensplan og dræber i gennemsnit et inficeret individ hver 18. sekund. TB har forårsaget næsten 1 milliard dødsfald over de seneste århundreder og er skyld i omkring 1,8 millioner dødsfald på årsbasis. Mtb har igennem titusinder af år udviklet sig side om side med dens primære vært, mennesket, og har derigennem udviklet mekanismer hvormed den kan manipulere og undvige immunsystemet. BCG er på nuværende tidspunkt den eneste tilgængelige vaccine mod TB og til trods for at denne har været tilgængelig siden 1921, og er den mest bruge vaccine nogensinde, har BCG ikke været i stand til at stoppe den globale TB pandemi. BCG-vaccinen har udvist store variationer i den beskyttende effekt, som derudover også svinder med tiden, med en varighed på omkring 10-15 år i gennemsnit. Revaccination med BCG har ingen yderligere præventiv effekt, hvilket betyder at der er et stort behov for nye og forbedrede vacciner.

Flere kandidater i den kliniske udvikling af TB vacciner indeholder Mtb antigener som går igen i flere af de andre kandidater (fx Ag85A/B). Studierne beskrevet i denne opgave fokuserer hovedsageligt på substrater fra de såkaldte type syv secerneringssystemer, også kendt som ESX systemer i Mtb. Flere af substraterne secerneret igennem ESX er blevet koblet med mekanismer der medierer virulensen af Mtb bakterien. I Manuskript I beskriver vi rationalet bag udviklingen af en ny TB-vaccine. På grund af tæt homologi mellem substrat parrene udskilt gennem det samme system, ESX-1 - ESX-5, argumenterede vi for at inklusion af en enkelt af disse par derfor også kunne inducere genkendelse af de resterede fra samme system. Vi evaluerede substrat par fra hver secerneringssystem og, med undtagelse af EsxU-T, fandt vi at disse var i stand til at beskytte mod infektion med Mtb i mus. På baggrund af faktorer som gene ekspression, immungenisitet og individuelt beskyttelse fremstillede vi fusionsproteinet, H65, baseret på antigener secerneret gennem ESX-2, -3 og -5. Vaccination med H65, i kombination med CAF01, inducerede immune responser rettede mod de enkelte komponenter i fusionsproteinet, EsxD-EsxC-EsxG-EsxH-EsxW-EsxV. Det nye vaccine protein var i stand til at beskytte på niveau med H56, et TB fusionsprotein i fase II klinisk afprøvning, og BCG.

Størstedelen af det arbejde præsenteret i denne rapport er baseret på immuniseringer med adjuvanset CAF01. Foruden CAF01, er der dog flere andre adjuvansformuleringer, kombineret med forskellige fusionsproteiner, i det kliniske afprøvning af nye TB vacciner. I Manuskript II, beskriver vi et headto-head sammenlignings studie af fem vaccine adjuvanser som enten er på vej igennem klinisk afprøvning eller er del af godkendte vacciner. Vi sammenlignede adjuvanserne Alum, MF59®, GLAiii

Summary - Danish

SE, IC31® og CAF01 i kombination med antigener fra Mtb, influenza og klamydia. De forskellige adjuvanser inducerede immunologiske signaturer der var unik for den enkelte kandidat og er desuden uafhængig af hvilket antigen de blev kombineret med. Af de testede adjuvanser var GLA-SE, IC31® og CAF01 i stand til at inducere et potent Th1 CD4 T celle respons specifikt for vaccinationsantigenet H56 (Ag85B-EsxA-Rv2660c). Responset var domineret af triple- (IFN-γ+IL-2+TNF-α+) og dobbeltpositive (IL-2+TNF-α+) T celler som tidligere er blevet associeret med en centrale hukommelses T celler. Der var dog ingen korrelation imellem de målte frekvenser af T celler og den inducerede beskyttelse mod TB.

De fleste studier der undersøger Mtb’s infektionsmekanismer eller evaluerer den beskyttende effekt af en ny TB-vaccine er blevet udført ved hjælp af laboratorie-tilpassede stammer som fx Mtb Erdman eller H37Rv. For at imødegå den stigende bekymring om at resultater fundet med disse stammer ikke kan overføres til infektioner med "rigtige" Mtb-stammer er der opstået en stigende tendens til at undersøge nyligt isolerede kliniske stammer. I Manuskript III beskriver vi karakteriseringen af et ny klinisk Mtb stamme, Mtb DK9897, isoleret fra en patient med ekstrapulmonal TB. Vi identificerede stammen på grund af en usædvanlig modstand mod H56 vaccinen, på trods af øget grad af modtagelighed for BCG-vaccination. Ved anvendelse af in vitro celle kulturer fandt vi ud af at DK9897 ikke er i stand til at udskille EsxA, en vigtig komponent i H56, og således ikke genkendes af EsxA specifikke T celler. Vi viser, at den manglende EsxA secernering sandsynligvis skyldes en mutation i genet der koder for EccCa1 og at genindførelse af dette gen genetablerer EsxA udskillelsen af DK9897. I Manuskript IV karakteriserer vi konsekvenserne af den manglende ESX-1/EsxA secernering af DK9897 på interaktionerne i en infektion i mus. Ved en aerosol infektion udviser DK9897 en forsinket indtrædelse af eksponentiel vækst og reduceret vækstrate, hvilket resulterer i forsinket indtræden af den adaptive immunitet. Den overordnede svækkelse af DK9897, på grund af den manglende EsxA udskillelse, forårsager et fald i inflammation i lungerne, som derved resulterer i reduceret rekruttering af immunceller til lungerne. Hvis vi derimod anvender en intravenøs (IV) infektionsmodel, udviser DK9897 sammenlignelige systemisk virulens som Mtb Erdman. IV infektionen resultater i en næsten øjeblikkelig indtræden af adaptiv immunitet, hvilket indikerer, at de undertrykkende/virulensfunktioner medieret af EsxA sandsynligvis er specifikke for lungerne.

iv

Abbreviations

Abbreviations ADITEC

Advanced Immunization Technologies

MDR

Multi Drug Resistant

AIDS

Acquired Immune Deficiency Syndrome

MMP

Matrix Metalloproteinases

APC

Antigen Presenting Cell

Mtb

Mycobacterium Tuberculosis

BCG

Bacillus Calmette Guerin

MTBC

Mycobacterium Tuberculosis Complex

CAF01

Cationic Adjuvant Formulation 01

NHP

Non Human Primate

CFU

Colony forming units

NK

Natural Killer Cells

DC

Dendritic cells

NOS2

Nitric Oxide Synthase 2

dLN

Draining Lymph Node

NTM

Nontuberculous Mycobacteria

ESAT-6

Early secreted antigenic target of 6 kDa

PE

Proline(P)-Glutamic acid(E)

Esp

ESX Secretion-Associated Protein

PPD

Purified Protein Derivative tuberculin

ESX

ESAT-6 Secretion System

PPE

Proline(P)-Proline(P)-glutamic acid(E)

EsxA

ESAT-6

PRR

Pathogen Recognition Receptors

EsxB

CFP-10

RD1

Region of Difference 1

EsxH

TB10.4

RNI

Reactive Nitrogen Intermediates

H28

HyVac28 (Ag85B-EsxH-Rv2660c)

ROI

Reactive Oxygen Intermediate

H56

Hybrid56 (Ag85B-EsxA-Rv2660c)

SSI

Statens Serum Institut

H64

EsxA-EspD-EspC-EspF-EspR-PE35

T7SS

Type VII Secretion Systems

H65

EsxD-EsxC-EsxG-EsxH-EsxW-EsxV

TB

Tuberculosis

HIV

Human Immunodeficiency Virus

TDR

Totally Drug Resistant

IFN

Interferon

Th1

T Helper type 1

IGRA

Interferon Gamma Release Assay

TLR

Toll-Like Receptor

IL

Interleukin

TNF

Tumor Necrosis Factor

INOS

Inducible Nitric Oxide Synthase

Tregs

T Regulatory cells

IP-10

Interferon Gamma-Induced Protein 10

TST

Tuberculin Skin Test

KLRG1

Killer cell Lectin-like Receptor G1

WHO

World Health Organisation

LAM

Lipoarabinomannam

WxG100

ESAT-6 family proteins

LTBI

Latent Tuberculosis Infection

XDR

Extensively Drug Resistant

M.bovis

Mycobacterium bovis

v

Introduction

Introduction General background of TB TB epidemiology

Tuberculosis (TB) is an infectious respiratory disease caused by an infection with the intracellular pathogen Mycobacterium tuberculosis (Mtb). TB is now the leading cause of death due to a single infectious microbe and was in 2015 responsible for 1.8 million deaths[1], and has taken the lives of nearly 1 billion individuals over the past few centuries[2]. Although TB incidence and mortality rates have been declining for the past decade, there were still 10.4 million new cases of TB in 2015, of which 95% occur in low and middle-income countries along with 99% of TB deaths (Fig. 1, [2]). South-East Asia and Western Pacific Regions account for 58% of the global TB incidences, while Sub-Saharan Africa, with the highest burden relative to population, accounts for 28% [1].

Estimated new TB cases (all forms) per 100 000 population per year 0–9.9 10–19 20–49 50–124 125–299 300–499 ≥500 No data Not applicable

Figure 1 - Estimated global TB-incidence rates 2014, adapted from [1]

TB is transmitted via airborne droplets containing Mtb, released from the lungs of an infected individual with active TB by coughing or sneezing. Once infected, only a relatively small fraction of individuals, about 5%-10%, will develop active TB over a lifetime[3]. However, considering the large estimated pool of 2-3 million infected individuals, TB is a major global health concern[1]. This large pool of infected individuals and high yearly mortality rates persists in spite of TB medical advances in both its prevention and treatment[4]. A vaccine against TB, Bacillus 1

Introduction

Calmette–Guérin (BCG) vaccine, has been available since 1921[5], and the first antibiotic drug against Mtb (Streptomycin) has been available since 1944[6].

While BCG is capable of protecting against severe forms of TB, and thereby reduce mortality, the efficacy of BCG is still widely debated and the induced protection appears to wane in adulthood, thus failing to control the Mtb epidemic[5]. And while the discovery of Streptomycin and other antibiotic agents might have led people of the post WWII era to believe that bacterial infections was a concern of the past, the widespread use of antibiotics has led to the development of drug resistance[7]. In 2015, an estimated 480,000 of TB incidences were caused by multi-drug resistant strains (MDR-TB), resistant to the two most powerful anti-TB drugs isoniazid and rifampicin[1]. Of these, 9% were categorized as extensively drug resistant tuberculosis (XDR-TB), which are also resistant to any fluoroquinolone and at least one of three injectable second-line drugs (i.e., amikacin, kanamycin, or capreomycin)[1]. In addition to being significantly harder to treat, the average cost of treating MDR-TB and XDR-TB are drastically increased compared to their drug susceptible counterparts. In the EU, the average cost for treating a case of drug-susceptible TB is €10.282, while the prices for treating MDR-TB and XDR-TB averages €57.213 and €170.744, respectively, making these virtually untreatable in most developing countries[8].

History of TB

TB is an ancient disease and appears to have co-evolved and plagued humankind throughout known history. Based on geological plate drifting, it has been speculated that the genus Mycobacterium might have originated as far as 150 million years ago[9], while the common ancestor of Mtb is estimated to have arisen around 15,000 – 20,000[10]. This timeframe fits with the hypothesis that Mtb might have arisen by zoonosis from the closely related M. bovis around the time its natural host, cattle, was domesticated around 8,000 – 10,000 years ago[11]. However, while the origin of the Mtb species is still speculative, more concrete evidence of ancient TB can be documented in 3000-year-old Andean Mummies[12], in 5000-year-old Egyptian mummies[13] and even be detected in bone samples with typical tuberculosis lesions from a woman and an infant submerged in the Eastern Mediterranean 9000 years ago[14].

One of the earliest descriptions of TB is by Hippocrates in classical Greece, where it was called phthisis[15]. The disease was described as attacking between the age of 18 and 35, and usually being fatal[16]. Throughout the centuries, the disease keeps surging in waves of TB epidemics, 2

Introduction

with death rates reaching between 800 - 1,000 per 100,000 per year in the early 1800s[16]. Around this time TB was known as ‘the white plague’, and was responsible for up to 25% of deaths in Europe, most likely facilitated by a high population density, combined with poor sanitary conditions[17].

Up until this point, TB was still generally believed to be of hereditary nature, even though its infectious nature had been suggested by Aristotle in the time of Hippocrates[18]. It took up until 1865 for the French surgeon Jean-Antoine Villemin to demonstrate the contagious nature of TB, by infecting a rabbit with the liquid from a tuberculous cavity removed at the autopsy of a patient who succumbed to TB[15]. In the footsteps of this milestone, it was the German physician and microbiologist, Robert Koch, who in 1885 confirmed the infectious nature and determined the etiologic agent of TB, namely Mycobacterium tuberculosis, for which he received the Nobel Prize in Physiology or Medicine in 1905[16].

Furthermore, Koch coined the term tuberculin, a substance he isolated from sterilized tubercle bacilli, which he initially tested as a therapeutic agent. This was later used for the development of the first TB diagnostic test, also known as the tuberculin skin test (TST)[16]. Koch’s discovery also led Albert Calmette and Camille Guérin on the path of attenuative sub-culturing of tubercle bacilli, and after 230 subcultures of a M. bovis strain, the first anti-tuberculosis vaccine was ready for human testing in 1921[5]. With the ability to culture Mtb, also came the ability to test newly developed anti-TB agents. As mentioned, streptomycin was the first anti-TB drug tested in humans in 1944, and the subsequent decade saw the emergence of several new drugs, e.g. isoniazid (1951), pyrazinamide (1952), ethionamide (1956), rifampin (1957), and ethambutol (1962), some of which are still in use today[19].

Due to the availability of anti-TB chemotherapeutics and widespread implementation of the Directly Observed Treatment Short course (DOTS) in the 1980s, it was believed that the global TB incidences would decline[20]. However, due to the severe co-mortality with the HIV/AIDS epidemic, combined with the failure to comply with the lengthy DOTS treatment, TB is resurging and is again a public health concern[21]. This led TB to be declared a global health emergency by the WHO in 1993[4].

3

Introduction

Transmission and Progression

TB is transmitted by aerosolized droplets containing Mtb, expelled from the lungs when a TB patient exhales, coughs or sneezes, and subsequently infects a new host through the airways. A healthy individual needs to inhale only minute amounts of Mtb to be infected[22]. However, compared to some other infectious diseases, TB is not considered highly contagious, since only about one in three close contacts, and fewer that 15% of remote contacts, are likely to become infected[23]. In general, close, frequent and/or prolonged contact is needed to contract the disease[23]. Once infected with Mtb, there is still a relatively small chance of developing TB disease (Fig. 2, [24]). An estimated 5% of infected individuals will develop symptoms within the first 12-24 months, also known as primary TB, while the majority are able to control the initial infection[24]. While the human immune system is able to contain Mtb and prevent disease in most cases, the pathogen is rarely eradicated, but kept under control for the remainder of the patients lifetime, classified as having a latent TB infection (LTBI)[23]. In about 5-10% of LTBI cases, reactivation will occur and the patient will subsequently develop active TB and become contagious. Among the most common causes of reactivation are HIV-infections, treatment with immunosuppressive agents, malnutrition, drug and/or alcohol abuse.

Bacterial load

Primary progression (5%) Reactivation (5–10%)

Latent infection (~90%) Elimination? Innate Immunity Adaptive immunity Figure 2 - The disease progression following a Mtb infection[24]

Clinical manifestation

Upon infection with Mtb, the pathogen is taken up by macrophages in the lungs, where they are able to persist in the internalized compartments. In an effort to contain the invading pathogen, an early immune response results in an influx of neutrophils, inflammatory monocytes, macrophages 4

Introduction

and dendritic cells (DC) to the site of infection. However, unable to eradicate the intruder, the cells attempt to surround and encapsulate the pathogen-infected macrophages and form the hallmark of a Mtb infection, the granuloma[25]. The solid granuloma formation successfully contains the Mtb infected macrophages in 95% of cases and avoids the development of primary TB, but the afflicted individuals now harbour a latent infection. Although Mtb is rarely fully eradicated, the LTBI individuals are asymptomatic and non-contagious, and continue silently battling the infection in a delicate balance between pathogen eradication and minimizing host pathology[26].

As mentioned, 5-10% of LTBI individuals will reactivate over their lifetime. Reactivation marks the event where the balance between pathogen and the immune system is interrupted and containment of the pathogens is no longer possible[25]. This is manifested by the center of the granulomas becoming necrotic and liquefied, forming a medium where Mtb are able to multiply extracellularly and uncontrollably[27]. As dead tissue accumulates, the necrotic area expands leading to gross pathology in the surrounding lung tissue and form cavities harbouring large numbers of extracellular Mtb, that are now readily expelled while coughing or sneezing[28, 29].

Although TB is predominantly a disease of the lungs, with pulmonary TB accounting for 70% of all TB cases, the pathogen can migrate to various areas of the body and cause disease in other organs (extrapulmonary TB)[30]. Pulmonary TB will usually manifest itself with a persistent bloody cough, respiratory insufficiency, fatigue, weight loss, lack of appetite, chills, fever and night sweats[31]. The infection can disseminate to, e.g., the lymphatic system (TB lymphadenitis), the circulatory system (miliary TB), the kidneys (genito-urinary TB), as well as bones and joints including the spine (Pott’s disease), and may even cause inflammation of the protective membranes covering the brain and spinal cord in the central nervous system (TB meningitis). Based on which location is afflicted, TB will manifest itself with different symptoms, making the diagnosis of TB very difficult[32].

Diagnostic tools

The first diagnostic test for identifying TB patients was the tuberculin skin test (TST), which was initially introduced by Robert Koch in the 1890s as a failed treatment[33]. The TST, a.k.a. Mantoux test, was developed by Von Pirquet and Mantoux in 1907-1908 and was followed by the standardization of tuberculin manufacturing by Florence Seibert in 1931. The subsequent use of the TST with the resultant purified protein derivative of tuberculin (PPD) became widely accepted 5

Introduction

and is still widely used for diagnosing TB[33]. The TST relies on a delayed-type hypersensitivity (DTH) reaction elicited by the immune system in response to the intradermal injection of PPD, a sterile solution of purified protein precipitated from a filtrate of tubercle bacillus[34]. The diagnosis is based on the degree of swelling and redness developed at the site of intradermal PPD injection. Unfortunately, the TST is incompatible with the BCG vaccine and has a relatively low specificity due to a large amount of false-positive TST reactions from exposure to nontuberculous mycobacteria[35].

Recently, this shortcoming has been overcome by introduction of a more specific diagnosis test; the Interferon Gamma Release Assays (IGRAs). The IGRAs are blood-based assays, where the amount of gamma interferon (IFN-γ) release from peripheral blood mononuclear cells (PBMCs) in response to exposure of Mtb specific antigens, reveals the previous exposure to the pathogen[36]. Currently approved IGRAs include QuantiFERON® TB Gold test, and the TSPOT® TB test, which are both based on recognition of the hallmark Mtb antigens EsxA(ESAT6) and EsxB(CFP-10) that are not shared with the BCG vaccine[37]. While the more advanced and specific IGRAs have been available for more than a decade, the increased specificity comes with an increase in cost compared to older diagnostics tests[38]. With over 90% of the global TB burden occurring in low- and middle-income countries, the diagnosis still mainly relies on the less expensive, but less accurate sputum smear microscopy and chest radiology[39].

Chemotherapy and drug resistance

In the pre-antibiotic era, a common treatment for TB included collapsing of the lung either by injecting gas into the intrapleural space (pneumothorax), by crushing the phrenic nerves responsible for inflating and deflating the lung (phrenic paralysis) or by surgically removing several ribs (thoracoplasty)[40]. It was believed that by collapsing the afflicted part of the lung, it would have a chance to repair itself and eradicate the mycobacteria by cutting off oxygen supply[41]. Prior to the lung-collapse therapy, the patients were isolated in TB sanatoriums for treatment with e.g. bed rest, better nutrition, sunbathing and moderate exercise, which provided moderate success[41].

With the introduction of the first antibiotics effective against Mtb in 1944, the treatment of TB was revolutionized, but soon after followed the reports of resistance development[42]. In a postWWII clinical trial from 1948, bacteria resistant to high concentrations of streptomycin were 6

Introduction

isolated after only 2 months of treatment in patients who were still sputum positive[43]. However, combining streptomycin with the para-aminosalicylic acid (PAS) turned out to yield superior therapeutic activity and resulted in less streptomycin resistance development, compared with streptomycin alone[44]. The addition of isoniazid, creating the triple therapy, further reduced the rates of resistance and increased the effectiveness of treatment, although 24 months of treatment was still recommended[45].

TB treatment was continuously improved and adjusted over the following decades, and today a treatment for presumed drug susceptible pulmonary TB includes a two month treatment with isoniazid, rifampicin, pyrazinamide and ethambutol, followed by four months with only isoniazid and rifampicin[46]. However, reports of drug resistant TB continue to emerge, with resistance towards the most common first line drugs becoming more and more frequent. There are two main types of drug resistant TB: multi drug resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB). MDR-TB is caused by a strain of Mtb resistant to the two frontline drugs isoniazid and rifampicin, while XDR-TB is at least MDR-TB, but also resistant to one of the fluoroquinolones, as well as to at least one of the second line injectable TB drugs[47]. A 2007 study reported two Italian cases from 2003, that where resistant to all drugs with known anti TB activity[48] and since then several other cases of totally drug-resistant Mtb bacilli (TDR-TB) have been reported[49-51].

Factors like insufficient chemotherapy, low drug quality, poor compliance to prescribed treatment, treatment failure and prior treatment all contribute to the development of drug resistance[52]. Of these, prior treatment of TB is a major risk factor for development of MDR-TB[53]. The poor compliance is often triggered by factors like alcohol consumption, psychiatric illness, adverse drug reactions, drug addiction, homelessness and inability to afford treatment[54, 55]. Furthermore, a lot of new cases of MDR-TB occur due to errors in the TB management, like the use of a single drug to treat TB, the addition of a single drug to a failing regimen or the failure to identify preexisting resistance[52, 53].

Mycobacterium tuberculosis

Mtb is part of the genus Mycobacterium, comprising over 140 different species, which are subdivided into the Mycobacterium tuberculosis complex (MTBC), M. leprae, and the nontuberculous mycobacteria (NTM), based on their ability to cause disease in humans[56]. 7

Introduction

Besides Mtb, the MTBC also includes M. aficanum, M, microti, M. bovis, and even the strongly attenuated M. bovis BCG, which all differ in phenotype and host species[57]. Mycobacteria are rod-shaped, aerobic, Gram-positive bacteria of about 1.5 to 4 μm in length, with a long generation time of about 15-20 hours[58]. Furthermore, the Mtb cell envelope is composed of a double lipid cell membrane, cover by a thick waxy surface consisting of a mixture of polysaccharides and lipids, characterized by a high content of mycolic acid[59]. The slow growth and thick cell wall are contributing factor to the resistance to antibiotics, longevity of treatment and clinical latency[58].

Type VII secretion systems

Incorporated in the cell membrane, mycobacteria poses a unique protein secretion system also known as the Type VII Secretion Systems (T7SSs)[60]. Mtb possesses five paralogue T7SSs, ESX-1 – ESX-5, of which at least three are essential for mycobacterial virulence and/or viability[60]. The first of these to be discovered was the ESX-1 system, which is named for its ability to secrete the immunogenic antigen EsxA (ESAT-6, 6-kDa Early Secretory Antigenic Target). Part of the ESX-1 locus is missing in the BCG vaccine (ΔRD1), which is the main reason behind the attenuation of this strain[61]. EsxA was first identified as a major part of a short-term culture filtrate of Mtb, and initially discovered based on its immunological memory recall in Mtbinfected mice[62, 63]. The EsxA protein is part of the WxG100 family of homologous proteins which share a Trp(W)-Xaa(x)-Gly(G) amino acid motif and are around 100 residues in size[64]. There are 23 WxG100 members in the Mtb genome, the proteins of which are named EsxAEsxW[65]. Most of the WxG100 family are secreted in pairs, where one member contains the highly conserved YxxxD/E motif, signaling for ESX secretion[66]. In addition to EsxA, ESX-1 is responsible for secretion of its pair EsxB, PE–PPE and various Esp proteins[67-70], although the mechanism is still not well understood.

8

Introduction

EspB

EspB

PE35 EspR

PPE68 Heptamer EspB

Nutrition?

EspG

EsxA

EsxB

EspD

EspA

EspC

EspF EspE

EspC

EspK

Free lipids Mycomembrane Mycolic acids Arabinogalactan Peptidoglycan Periplasmic space

EccE1

EccD1

MycP1

EccB1

Plasma membrane

EccCb1

Bacterial cytosol

EspD ATP ADP

EccCa1 c

c

EccA1

ATP ADP

c

EsxA EspE

EsxB

PE/PPE

c

EspG

EspA/EspC

EspF c

EspB

EspK

EspD EspC EspA Gene transcription Rv3614c Rv3615c Rv3616c

+

EspR

Figure 3 - Model of the ESX-1 secretion system based on review of the current literature.

The ESX-1 system is composed of several conserved components EccB1, EccCa1, EccCb1, EccD1 and EccE1, each with a transmembrane region spanning the mycobacterial plasma membrane(Fig. 3, [71]). The EccC translocase is driven by ATP and consists of two subunits, EccCa1 and EccCb1, which are assembled once EccCb1 binds its target substrate, in this case the EsxA/EsxB[72] or EspB/EspK[67]. EccCb1 binds EsxB via the carboxy‑terminal signal sequence (marked as C in the figure)[73]. EsxB functions as a chaperone for EsxA secretion, which is a major ESX-1 virulence factor[74]. The secretion of EsxA/EsxB is co-dependent with the secretion of EspC/EspA and the C terminus of EspC targets for the interaction with the cytosolic ATPase EccA1[70, 75]. EspC polymerizes during secretion, indicating that EccA1 and EspA might function as cytosolic chaperones[76]. The polymerization of EspC results in the formation of a surface-exposed filamentous structure that spans the entire cell envelope[76], possibly serving as a channel responsible for transporting ESX-1 substrates. Likewise, EspB has been shown to adopt a PE/PPE-like fold that results in a heptameric oligomerization once secreted[77], a fold that has been suggested to form membrane channels allowing nutrient uptake[78, 79]. EspD stabilizes EspA and EspC[80] and has furthermore been found to interact with EccE1 [81]. Like EccA, EspG 9

Introduction

functions as a chaperone for recruitment of heterodimers formed by a Pro-Glu (PE) family protein and a Pro-Pro-Glu (PPE) family protein[66, 82]. EspG has been found to be cell wall-associated after translocation[83]. EspR is a transcription factor of the rv3614c-rv3616c operon, and thus secretion of EspR functions as a negative regulator of ESX-1 secretion[84]. The M. marinum analogue of EspE, Mh3864, was identified as an ESX-1 substrate that remains partially cell wallassociated after translocation at the bacterial poles[69]. EspF was found in the culture medium of Mtb cultures[85] and as specifically secreted by ESX-1 in M. marinum[70].

Immunology of TB A brief overview

Once inhaled as respiratory droplets and deposited in the alveoli sacs, the intruding Mtb are encountered and phagocytosed by the alveolar macrophages. The ingested foreign pathogen can then reside within phagosomes, where Mtb is able to survive and proliferate. Activation of the antigen presenting cells (APCs), macrophages and DCs, will cause the influx of monocyte-derived macrophages, neutrophils, NK cells and γδ-T cells. The influx is driven by the local secretion of chemokines and cytokines, and the incoming cells attempt to aid the elimination of the intruder by secreting inflammatory molecules themselves. When a certain threshold is reached, the pathogen is carried to the draining lymph nodes, where antigen-bearing dendritic cells (DCs) prime T and B cells. Once the naïve T cells have been primed, they migrate from the lymph nodes and are recruited to the site of infection as activated effector T cells, mainly type 1 T helper cells (Th1). Th1 CD4 T cells are capable of secreting IFN-γ, which stimulates the activation of infected macrophages to become efficient effector cells, hence contributing to control of the mycobacterial infection. The effector T cells will target the infected cells in the lung, and over time cells progressively assemble in a compact, organized aggregate of mature macrophages surrounded by fibroblasts, interspersed with neutrophils, DCs, NK cells, B cells, CD4+ and CD8+ T cells, forming the characteristic TB granuloma. The granuloma is a very dynamic structure, where dying cells are continuously being replenished by cellular recruitment, and thus forms the primary foci for the interaction between pathogen and host immune system. Unable to eradicate the internalized bacilli without causing excessive pathology to the surrounding tissue, the two opposing forces enter an equilibrium, forming the basis of a LTBI. Once the delicate balance is upset in either direction, the equilibrium is interrupted and active TB will ensue.

10

Introduction

Innate immunity

The innate immune response plays an important role in the protection against TB as it provides the first line of defence encountered by the invading pathogen. However, since Mtb has evolved strategies to manipulate various immune factors, allowing intracellular survival and replication, this also makes the innate immune system a prerequisite for mycobacterial pathogenesis and persistence[86]. Once Mtb is inhaled as respiratory droplets and deposited in the terminal alveoli, it is presumed to first encounter and be engulfed by alveolar macrophages[87]. The uptake of Mtb is facilitated by the recognition of pathogen associated molecular patterns (PAMPs) by specific pathogen recognition receptors (PRRs), which is a necessary first step to trigger the innate immune response[88]. Mtb is recognized by a variety of different host receptors, and the combination of which receptors are triggered will determine the fate of the ingested bacilli[61]. Mtb are recognized by Toll-like receptors (TLRs), nucleotide-binding oligomerization domain- (NOD-) like receptors (NLRs), and C-type lectins (CLRs). The C-type lectins include the mannose receptor (CD207), the dendritic cell-specific intercellular adhesion molecule grabbing nonintegrin (DC-SIGN) and Dectin-1[61].

Once triggered, the alveolar macrophages start producing inflammatory cytokines, TNF-α, IL-6 and IL-12, and chemokines, CXCL8 and monocyte chemotactic protein-1 (MCP-1), which results in the attraction of monocytes, macrophages, neutrophils and lymphocytes to the focal site of infection[89]. While the influx of innate immune cells does little to restrict bacterial growth, it conversely provides a reservoir of fresh hosts cells for the multiplying Mtb to infect[89]. The recruited phagocytes engulf the bacteria and trap them in intracellular vesicular compartments known as phagosomes[89]. Following a maturation process of the phagosome, the bacillicontaining compartment is marked for acidification by fusion with lysosomes, forming a phagolysosome[89]. Mtb has however developed mechanisms to prevent phagosomal maturation, which prevents phagosomes and lysosomes from fusing in resting macrophages[90]. This enables bacteria to persist and proliferate within the immature phagosomal compartment[90].

Neutrophils are among the first cells to respond to the inflammatory stimuli and an Mtb infection is usually accompanied by a massive influx of neutrophils[91-93]. Neutrophils are professional phagocytes and are, like macrophages, capable of engulfing pathogens and rapidly killing ingested microbes by phagolysosomal degradation. The neutrophils contain an arsenal of antimicrobial effector molecules including defensins, lysozyme, lactoferrin and cathelicidin, which are transferred into the Mtb-containing phagosome upon fusion with granules[94, 95]. Besides the 11

Introduction

direct intracellular killing, neutrophils are capable of inducing macrophage activation by secreting pro-inflammatory cytokines (e.g. IFN-γ, TNF-α) and induce attraction of other immune cells by secreting chemokines (e.g. IP-10, MCP-1, MIP-1α/β)[93]. However, while neutrophils constitute a potent population of antimicrobial effector cells, they are also a key mediator of immunopathology in human TB, due to the release of effector molecules that indiscriminately damages bacterial as well as host cells[86].

Natural killer (NK) cells are innate lymphocytes which are a first line of defence against infection[96]. The NK cells are granular lymphocytes which possess potent cytolytic and cytokineproducing effector functions[97]. NK cells are not MHC-restricted and are thus able to act early in the infection[98]. NK cells are activated by various Mtb cell wall components, such as mycolic acid via the natural cytotoxicity receptor (NCR) NKp44 on NK cells[98]. NK cells are capable of restricting Mtb by secreting GM-CSF, IL-12, TNF-α, IL-22, and IFN-γ, thus aiding the activation of Mtb infected macrophages[86]. The granules of NK cells contain perforin, granulysin and granzymes, a combination of molecules capable of inducing apoptosis in infected host cells[99, 100].

The granuloma

After the first encounter between Mtb and the alveolar macrophages, lymphocytes are recruited to the site of infection and form a lymphoid inflammatory structure around the infected macrophages known as a granuloma, the hallmark of a TB infection[101]. A granuloma is a compact, organized immunological structure consisting of immune cells, macrophages, dendritic cells (DCs), neutrophils, NK cells, foamy macrophages, T and B cells, and multi-nucleated giant cells[25, 102]. The structure organizes around a mass of infected macrophages, and has for a long time been considered a host protective measure combating the infecting pathogen by creating a microenvironment deprived of nutrition (starvation)[103] and oxygen (hypoxia)[104]. The structure has however more recently been suggested as advantageous to the bacilli[105]. The close aggregation of immune cells provides ideal conditions for the microbes, released from necrotic cells, to readily infect the recruited bystanders[106]. Furthermore, Mtb is able to dampen the immune response within the granuloma through the induction of IL-10, which suppresses the activation of T cells and macrophages[107]. The induction of regulatory IL-10 also inhibit excessive tissue destruction, and thus benefits the host[108]. Overall, the balance of mechanisms

12

Introduction

at play in the granuloma is indicative of the delicate host-pathogen interaction for the fitness of both organisms[109]. In humans, as a consequence of the excessive cell death following Mtb infection, cell debris accumulates at the center of the compact granulomas resulting in the formation of central necrotic area rich in lipids, referred to as the caseum[25, 110]. This is not the case for granulomas in most mice strains, which will generally form loose non-necrotic aggregates of immune cells[25], although, mice strains capable of inducing tight necrotic aggregates similar to human TB have been identified[111-113]. Common for both men and mice is however the strong dependency on TNF-α for maintaining granuloma integrity, which when absent results in disorganized granulomas and impaired control of infection[114]. Control of the infection is maintained as long as the solid necrotic granulomas are maintained, which is stabilized by the onset of the adaptive immune response(Fig. 4, [115]). If the equilibrium breaks down, the caseous center will liquefy, resulting in cavity formation and extracellular bacterial growth of Mtb, practically without restriction from the immune system. The lesion will increase in size and cause massive pathology in the surrounding tissue, eventually breaching the alveolar barrier and readily releasing Mtb into the airways and facilitating transmission to other individuals[25, 115]. Mycobacterium tuberculosis host-to-host transmission

Mycobacterium tuberculosis

Reactivation and dissemination in 10% of infected individuals

Initial infection

Macrophage

Innate immune phase

T cell B cell

Eradication?

Neutrophil Caseating granuloma

Innate lymphocyte Calcified granulomas

T cell immunity? Innate factors?

Necrotic cell

Unknown Factors

Adaptive immune phase. Containment of infection in 90% of individuals Mycobacterium tuberculosiscontrol?

Figure 4 – Model of granuloma formation and disease progression in TB[116]

13

Introduction

While neutrophils, γδ-T cells and NK cells perform the effector functions of the innate immune system, the dendritic cells (DCs) perform the necessary task of bridging the innate and adaptive immune responses[86]. DCs are the primary professional cells responsible for transport of Mtb to the draining lymph nodes (dLN)[117]. However, while airway DCs that sample the mucosal environment readily migrate to the draining node, the DCs within the alveolar tissue are in a regulated environment, with the permissive alveolar macrophages, which may limit their ability to respond to infection and migrate to the lymph node[118, 119]. While the resistance to migration and activation is crucial for the maintenance of the alveolar space, the delay in dissemination of bacteria due to infection of the permissive alveolar DCs and macrophages, results in delayed adaptive immune response[118, 120]. Furthermore, Mtb is able to impair the DC cytokine secretion, maturation, migration, and antigen presentation, further delaying the onset of the adaptive immune response[121]. The optimal priming of CD4 T cells requires the release of Mtb antigens from infected transport DCs, which is taken up by uninfected resident lymph node DCs[122].

Adaptive immunity

One of the key characteristics of the Mtb specific adaptive immune response is the long delay in onset and the continuous need for persistence in an effort to maintain the latent infection[123]. In humans, a measureable TST response is not detectable until approximately 42 days after Mtb exposure and infection[89], while in mice a delay of about 11-14 days occurs before the appearance of antigen-specific T cell responses following aerosol infection[124]. In comparison, an influenza virus takes about 20 hours to reach the lymph nodes[125]. The delayed response is a result of the carefully orchestrated symphony of evasions strategies utilized by Mtb to avoid detection and ensure persistence (discussed later). However, once sufficient numbers of Mtb bacilli and Mtb antigens have been shuttled to the dLN by DCs, the activation of an adaptive immune response is triggered[126, 127]. Here the DCs present Mtb antigens to T cells, along with costimulatory activation ligands and polarizing cytokines, in order to promote efficient development of effector T-cells (Teff)[86]. The adaptive immune response against Mtb involves the priming of a panel of different immune cells, each with a specific role in containing the invading pathogen (Fig. 5, [128]).

14

Introduction

Mφ

Lysis

IFNγ

CD8 CTL

Lysis

IFNγ

Latent infection

Protection CD8 TM

CD8 TM

CD8 T

M. tuberculosis PRR

Macrophage

Cross-priming

Exhaustion

CD8

Mφ

PNG

Mφ

MHC I

T Mφ

Mφ

Protection

Solid granuloma

T DC MHC II CD4

M. tuberculosis

MHC II

IL17

IFNγ

Th17

Th1

Th1 TM

CD4

PRR

Caseous granuloma

CD4 T DC

Active disease

Dendritic cell MHC I

Direct presentation

IL2 IFNγ TNF

M Th2

Treg

IL4

IL10 TGFβ

Suppression

CD8

T

B

Figure 5 - Overview of the cell mediated immune response against TB[128]

CD4 Th1 T-Cells A protective immune response against an Mtb infection requires the priming of adaptive T helper cells (CD4 T cells) via the MHC class II pathway. These have been shown to mediate protection, contribute to inflammation, and regulate immune responses[129]. Mice deficient in CD4 T cell production or the MHC class II presentation pathway have an increased susceptibility to TB[130, 131], while the adoptive transfer of Mtb-experienced CD4 T cells can provide TB protection in a naïve host[132]. Furthermore, IFN-γ has long been established as a key functional cytokine in a Mtb infection, since mice lacking IFN-γ exhibit increased dissemination and impaired control of bacterial growth[133, 134]. Additionally, a partial or complete IFN-γ receptor deficiency in humans can lead to disseminated nontuberculous mycobacterial (NTM) infections or BCGosis[135]. Taken together, these findings led to the assumption that a robust IFN-γ response would be a strong correlate of protection for testing the immunogenicity of TB vaccine candidates whereby the IFN-γ producing CD4 T helper type 1 (Th1) is the main mediator of this protection[136].

15

Introduction

The induction of a Th1 immune response is initiated by the production of the polarizing cytokine IL-12, which is secreted by the activated APCs in response to triggers by Mtb components, like lipoproteins and LAM[137]. The activated Th1 lymphocytes migrate to the site of infection where they proliferate and partake in forming the early stage granuloma [101]. The Th1 effector T cells secrete IFN-γ and TNF-α, which synergize to activate infected macrophages and promote the killing of bacteria via activation of nitric oxide synthase 2 (NOS2) and the resulting reactive nitrogen intermediates RNI-production(also discussed later)[138, 139]. As mentioned previously, IFN-γ knockout mice quickly succumb to a TB infection, which is associated with very low production of RNIs[133, 134], while in the absence of CD4 T cells, intact levels of IFN-γ and NOS2 are insufficient for maintaining control over an Mtb infection[130, 140]. This suggests that while CD4 T cells and IFN-γ are essential for TB protection, CD4 Th1 cells might also have an IFN-γ–independent mechanisms of controlling an Mtb infection[141]. This could in part be due to their ability to secrete TNF-α, which plays a key role in initiating and maintain granuloma integrity[142]. But while the CD4 Th1 T cells might have IFN-γ–independent mechanisms, the lack of CD4 T cells produced IFN-γ causes exacerbated bacterial growth and results in mice succumbing more rapidly to the infection[140]. The importance of a Th1 response has also been validated in humans by the observation of exacerbated mycobacterial disease in humans with mutations in the IFN-γ and IL-12 signalling pathways[143, 144] and in patients given anti-TNF to treat rheumatoid arthritis or Crohn’s disease[145, 146].

While the importance of IFN-γ and TNF-α have been known for a while, more recent studies indicate that a specific combination of these cytokines might be important for protection against TB[147]. In a TB vaccine setting, ‘triple positive’ effector memory that are IFN-γ+TNF-α+IL-2+ and ‘double positive’ central memory TNF-α+IL-2+ T cells have been shown to correlate with enhanced protection in mice[148, 149] and in cows with bovine TB[150]. In humans, the exacerbated immune response associated with active tuberculosis is dominated by single positive TNF-α+ CD4 T cells (that do not produce IL-2 or IFN-γ), compared to the triple positive T cells found with latent controllers of the Mtb infection[151].

Furthermore, it has been demonstrated that lung resident CD4 T cells, characterized by the expression of CXCR3+ signalling for homing to the lung parenchyma, confers increased protection against Mtb, compared to their CXCR1- CX3CR1+ counterparts, which are stuck in the lung vasculature[152]. The lung parenchyma localized CXCR3+ T cells are further distinguished by their expression of PD1, a marker associated with a high proliferative potential and moderate 16

Introduction

cytokine expression[153]. The vasculature-homing CX3CR1+ T cells have an increased expression of the killer lectin-like receptor G1 (KLRG1), which has been linked to the loss of IL2 expression, reduced proliferative potential, and impaired protection against a Mtb infection[154]. In this regard, the lung parenchymal cells are considered highly protective against Mtb, while their vasculature-residing counterparts are deemed terminally differentiated and only weakly protective[152].

Th17 and Regulatory T Cells In recent years, the Th17 T cell subset has been gaining increasing interest in the field of TB immunology and is now considered as part of the main protective effector CD4+ T tells during a Mtb infection[129]. Th17/IL-17 induces the recruitment of neutrophils and the production of chemokines and cytokines, which promotes the local inflammatory responses[129]. Previously, Th17 T cells were mainly associated with the induction of inflammation in models of autoimmune diseases, where excess of IL-17 mediates significant immunopathology[155]. In more recent studies, IL-17 has been reported necessary for protection against certain intracellular pathogens[156, 157] and that it is required to mount a Th1 T cell response in an acute infection with M. Bovis BCG[158]. Interestingly, it turns out that the M. Bovis BCG vaccination requires Th17 T cells to prime an efficient Th1-cell response, utilizing the IL-17 ability to counteract the mycobacterial-induced IL-10, thus inhibiting priming of the regulatory T cell (Tregs) subset[159]. It has recently been reported that Th17-produced IL-17 is not required for the primary immunity against infection with less virulent strains of Mtb like the clinical isolate CDC1551 or the labadapted strain Mtb H37Rv, while Th17 produced IL-17 was required for the protective immunity against the hypervirulent Beijing isolate HN878[160]. In a recent study, γδ-T cells were found to be the major source of IL‐17 in the lungs of Mtb‐infected mice, while Th17-produced IL-17 was important for maintenance of the Th1 T cells during chronic infection[161]. Furthermore, it has been shown that the Th1/Th17 inducing adjuvant CAF01 has increased protective efficacy in a head-to-head comparison with other Th1 inducing adjuvants (Manuscript II[149]). However, just like with the autoimmune diseases, there is a delicate balance between protective and pathogenic immune responses in the TB-infected lung[162] and increased levels of IL-17 have been associated with pathology in a bovine model of tuberculosis[163].

In order to maintain control over an Mtb infection the immune response must be tightly regulated[164]. The response needs to be both sufficiently strong to limit bacterial growth and dissemination, while at the same time limiting excessive inflammation, which could cause host 17

Introduction

pathology[165]. The role of dampening an excessive immune response is maintained, in part, by the CD4+FoxP3+ Tregs[165]. Like Th17, the Tregs were initially identified in studies of autoimmune disease, but rather as key components in preventing their development[166], and have since then proven to play a critical role in microbial infections, including Mtb[167-169]. During initial infection, Mtb induces the production of the immune suppressive cytokine IL-10, which potentiates the differentiation of Tregs[170], delaying the onset of the adaptive immunity and in the meantime allowing Mtb to replicate relatively unrestricted[171]. Several studies have shown that there is a direct correlation between severity of the infection and the number of Tregs[172174]. Furthermore, the loss of control associated with caveating granulomas, those harbouring highest bacterial burdens, have been linked with significantly increased numbers of Tregs[175], which was also the case for patients with disseminated miliary TB[176]. Tregs and Th17 function as opposing forces in the fight against TB and maintaining the right balance between the two subset is critical to the immunopathogenesis of active tuberculosis[177].

CD8 T cells While the importance of CD4 T cells has been evident for a long time, the contribution of CD8 T cells to the protection against TB is a more recent finding. CD8 T cells are primed by presentation of antigens on MHC class I molecules, a process that in TB requires either cross-presentation of antigens released from apoptotic cells and taken up by bystander DCs[178] or the direct translocation of Mtb, or its antigens, from the phagosome to the cytosol[179]. Depletion of CD8 T cells during the acute stage of the infection has no effect on the bacterial burden, while depletion during the chronic stage of infection has been associated with a loss of the ability to limit bacterial growth[180, 181]. In humans, development of active TB is accompanied by a diminished frequency of Mtb specific CD8 T cells compared to LTBI individuals, a difference that however disappears after about 4 months of treatment[182]. CD8 T cells are, like their CD4 counterparts, able to secrete IFN-γ, TNF-α and IL-2, and thus able to aid protection against TB by directly activating macrophages for increased intracellular killing[182, 183]. However, the protective CD8 T cell response has also been directly linked to their cytolytic function and dependent on their ability to secrete perforin[184].

B cells Since Mtb is mainly an intracellular pathogen the effects of B cells and antibodies have been thought inferior to the cell mediated immune response and have thus been largely overlooked. Several serum transfer studies were performed in the early 1900s, with most of the studies claiming 18

Introduction

some benefit from serum administration[185]. However, due to the lack of a consistently effective serum formulation and absence of proper trial controls, the antibody-mediated immunity to Mtb remains uncertain[185]. Nevertheless, the involvement of B cell-mediated immune responses have recently been revised and several independent groups have shown evidence of B cell/antibody involvement in protection against TB (Reviewed in [186]). Studies have shown that passive transfer of antibodies increases protection against mycobacterial infection[187, 188] and that induction of high antibody titre is associated with reduced susceptibility[189]. Furthermore, deficiencies in antibody production have been associated with increased susceptibility in both mice[190, 191], NHP[192] and humans[193, 194]. Although the cell mediated immunity remains the main correlate of protection, there is evidence to suggest that antibodies/B cells may contribute, at least in part, to immunity.

Mtb Immune evasion

One of the key factors behind the success of Mtb is its ability to survive in a hostile environment in the presence of a primed host adaptive immune response. Mtb has evolved several different immune-evasion mechanisms that help prevent eradication and ensure persistence in the host. Several of these mechanisms are linked to virulence factors secreted via the T7SS’s; ESX-1 secretion has been shown essential for the induction of cytolysis[195], cell-to-cell spread[196] and escape of Mtb bacilli from the phagosome to the cytosol[197]. Furthermore, Mtb has been shown to modulate the balance between necrosis and apoptosis via secretion of e.g. necrosis-inducing EsxA[198] and PPE68[199], while secretion of PE13 promotes apoptosis[200]. A necrotic cell will be accompanied by cell lysis resulting in the release of viable Mtb, while apoptosis will maintain intact cellular membranes, thus benefiting compartmentalization and mycobacterial containment [123, 201]. Mtb is also able to induce the production of matrix metalloproteinases (MMPs) in the surrounding tissue, via the secretion of EsxA[202, 203]. The MMPs cause degradation of the lung extracellular matrix, thus driving inflammation, promoting disease transmission and pathology[203]. The induction of MMP9 in the surrounding host epithelium has been suggested as the molecular mechanism with which Mtb induces the hallmark feature of a TB infection, the granuloma[202]. Furthermore, Mtb is able to delay the onset of a host immune response by preventing phagosome maturation[204, 205], controlling antigen presentation via secretion of EsxH[206], modulating the balance between protective IL-1β induction and type I interferons (IFNα/β) that are beneficial to the pathogen[207], and disrupting immune modulation by interfering with TLR2 signalling[208, 209]. These evasion mechanisms reduce epitope 19

Introduction

presentation on the surface of infected cells and subsequently affect the adaptive immune response via delayed recruitment of T cells to the site of infection and suboptimal T cell activation of infected cells [210, 211].

Inside the infected cell, Mtb also resists direct bactericidal attack. Once fully activated, the macrophages are able to perform a respiratory burst, where phagocyte oxidase (NOX2) and inducible nitric oxide synthase (iNOS), generate reactive oxygen species (ROS) and reactive nitrogen intermediates (RNI), that are released into the phagolysosome[90]. Mtb is however able to survive and persist within macrophages despite the toxic effects of ROS and RNI[212]. ROS are partly neutralized by performing their degradative actions on the thick layer of mycolic acid, and merely crosslinking and degrading the outer layers of the envelope[213]. Mtb is furthermore able to secrete various antioxidant enzymes, like superoxide dismutase (SodA, SodC) or catalase (KatG), and a secretory redox buffer ergothionine (ERG), which combined yields an excellent anatomical barrier to and detoxification of exogenous oxidants[213].

The murine TB challenge model

Several animal models have been used to study Mtb infections including mice[214], guinea pigs[215], rabbits[216], non-human primates[217], cattle[218], zebrafish[219], rats[220], ferrets[221], mini pigs[222], fruit flies[223], and even amoebas[224]. Mice are however the most commonly used species in TB research, due to their small size, cost-efficiency and reproducibility, combined with the availability of a broad set of immunological reagents for studying qualitative and quantitative aspects of the immune responses[225]. Despite being the origin of several breakthroughs in the study of TB, there are some key differences between human and mouse TB that need to be considered when interpreting the obtained results. As mentioned earlier, human TB can manifest itself in several different forms, with a spectrum of different symptoms, but with the most frequent outcome being a latent TB infection (LTBI) with low bacterial numbers[226]. Infection in mice is notably different in several aspects, whereby a low dose aerosol challenge results in a chronic, progressive infection of the lungs with a high bacterial burden[227, 228]. Furthermore, the mouse model has been criticized for not forming the stratified granuloma structure observed in human TB, nor reproducing human pathology[110]. Additionally, the two most commonly used inbred laboratory mouse strains, C57BL/6 and BALB/c, fail to develop necrotizing lesions[214, 229, 230]. However, despite some discrepancies between mice and men, multiple immunological mechanisms found in the mice, like the importance of the CD4 T cells, 20

Introduction

IFN-γ, TNF-α and TLR signalling, have found to translate into human TB[230]. Like in humans, murine infection with Mtb primarily targets the lungs causing a range of pathologies. Furthermore, the availability of unsurpassed genetic resources that include hundreds of inbred, congenic, recombinant, mutant and genetically engineered strains, and abundance of immunological reagents and methodologies allow in-depth analysis of practically any aspect of TB pathogenesis in mice, whereas their assortment is scarce for other animal species. Combining this with the fact that the BCG vaccine shows excellent protection against an aerosol challenge with Mtb in mice has made the mouse model indispensable for the assessment of TB vaccine candidates.

In the current work, all experiments have been carried out in inbred mice, using either a prophylactic or post-exposure Mtb challenge model[231]. As the names suggest, in the prophylactic model, the vaccines are given prior to aerosol Mtb challenge, whereas the postexposure model use mice that are vaccinated after aerosol Mtb challenge[232]. Whereas the prophylactic model is designed to mimic protection in previously Mtb negative individuals, the post-exposure model is intended to address vaccine-induced protection against reactivation in the large number of LTBI patients. As the number of bacteria in human LTBI is low, yet the burden in chronic Mtb-infected mice is high[233], the incomplete treatment with antibiotics is used to artificially induce a low burden infection to mimic LTBI and assess post-exposure vaccination effect on spontaneous reactivation[231, 232].

Current TB vaccine and future candidates BCG vaccine

Currently, the only licensed vaccine against TB is the Bacillus Calmette-Guérin vaccine (BCG), based on an attenuated M. bovis strain (the virulent bovine equivalent of Mtb for humans). M. bovis, originally isolated in 1902, was over a period of 11 years passaged 230 times under attenuating in vitro conditions on a potato derived medium containing glycerine and ox bile, eventually being unable to induce TB in lab animals[5]. Compared to the virulent Mtb strain H37Rv, the attenuated M. bovis BCG was found to have lost several genomic regions, known as regions of difference (RD1-RD16), of which RD1 was determined as a main contributor to the loss of virulence[234]. The first human administration of the attenuated M. bovis BCG was given in 1921[5], and it has since then become the most widely used vaccine in the world with more than 4 billion administrations[235]. Although BCG is able to effectively protect against the early manifestations of TB in children[236-238], estimates of efficacy in adults range from 0–80%[239]. 21

Introduction

In a systematic review of 132 clinical studies, the protection varied from substantial in a UK trial to completely absent in an South India trial, and was in general found to be greater in sites further from the equator[240]. The correlation between efficacy and latitude is generally attributed to the exposure to environmental mycobacteria closer to the equator, causing sensitization and thus loss of protective efficacy[241]. Furthermore, BCG only exhibits a limited ability to prevent infection with Mtb[242] and the protection declines over time, with protection against pulmonary and extrapulmonary TB lasting for up to 10 years[235]. However, despite being a highly cost effective vaccine, with an estimated cost of $2-3 per BCG inoculation[243] and the ability to prevent 73% of TB meningitis and 77% of military disease in children[240], BCG alone has failed to sufficiently reduce TB morbidity and mortality[244]. In order to achieve the End TB vision of reducing TB morbidity by 90% and TB mortality by 95% by 2035, new and improved interventions are needed[245]. Among these interventions are more accurate diagnostics, shortened and less-toxic drug treatments and, most importantly, more efficient vaccines. Novel TB vaccines in development

Given the limitations presented by the BCG vaccine, many novel TB vaccine candidates have been developed, of which 13 have reached various stages in the clinical trial pipeline(Fig. 6). The candidates have been designed either to replace BCG with a superior immunity-priming vaccine or administered as post-exposure supplementary vaccines that boost the existing response, building on the widespread use of BCG vaccination or prior exposure to Mtb[128]. A third strategy is the use of vaccines administered in adjunct with standard TB drug treatment, in order to improve the existing immune responses and prevent the development of recurrent disease[128]. The vaccines are further subcategorized based on their adjuvanting strategy: Protein subunits (either viral vectored or adjuvanted recombinant protein) and the mycobacterial derived (either attenuated whole cell or extract from culture)[246].

Of the current 13 candidates, eight are subunit vaccines, based of one or more Mtb antigens that are considered protective. Typically, a set of antigens with various characteristics are combined, in order to target the pathogen at various stages of disease progression[231] or targeting different virulence factor families(Manuscript I[247],[248]). A large focus has been put on immunodominant antigens, ideally recognized in both Mtb infected patients and experimental animals[249]. While this approach seems intuitive, it has resulted in a lack of diversity in the antigens represented in the TB clinical pipeline, where the predominant antigens are from the Ag85A/85B complex and/or the ESX family, including EsxA(ESAT-6) and EsxH(TB10.4)[245]. 22

Introduction

To remedy this shortcoming, new screening tools are being utilized to identify novel antigenic targets for vaccination, like in silico immunogenicity screening of the Mtb genome[250] and in vivo transcriptomic analysis of Mtb infected murine lungs[251], both of which resulted in the discovery of novel antigens recognized in LTBI patients.

Figure 6 – The global clinical pipeline of TB vaccine candidates as of October 2016 as reported by the vaccine sponsors. Image courtesy of Aeras.

Protein subunit vaccines are recombinantly produced, meaning that novel antigens can be combined or integrated into fusions relatively easy, and furthermore be produced in highly purified forms, thus generating very safe vaccine materials. However, with the high purity also comes an inherent lack of immunostimulatory properties, thus requiring subunit vaccines to be formulated with adjuvants or expressed by a recombinant viral vectors in order to increase immunogenicity. Each adjuvant is characterized by the induction of a signature immunological profile, implicating that while the design of the subunit protein is an essential first step, choosing the right “flavour” of adjuvant might be equally important. Currently, there are only a few adjuvants licensed for human use, e.g. Alum and MF59®, which are mainly used in combination with different viral targets[252]. The licensed adjuvants are capable of inducing potent antibody responses, while their ability to induce cell-mediated immunity is very limited. Novel adjuvant platforms, like IC31® and CAF01, are designed to induce both cell mediated and antibody responses, in theory providing both broader and longer lasting immunity. A head to head comparison of the mentioned adjuvants can be found in Manuscript II[149]. 23

Results

Results In this section, I will present an overview of the manuscripts upon which this thesis is based. Manuscript I and II have been published, manuscript III has been submitted and manuscript IV is ready for submission. The manuscripts are attached as appendices I-IV. Manuscript I Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostic

Purpose: To introduce a new vaccine approach, eliminating the use of EsxA, EsxB or Ag85B, but targeting antigens from different virulence associated secretion systems and assess the efficacy of this novel vaccine (H65) against TB disease. Results: We evaluated substrate pairs from each of the five ESX paralogues of Mtb for homology to other WxG100 family proteins, transcription during infection, immunogenicity and protection in different mice strains. We found a high homology between substrates secreted via the same ESX system, thus we argued that targeting one would ideally target most of the others from the same system as well. The substrates were found to be immunogenic when used as vaccines in combination with CAF01, although at different levels depending on the substrate and the mouse strain. All of the tested substrate pairs were protective, with the exception of EsxU-T, which showed no protection in either of the two mouse strains tested. Furthermore, there was no correlation between immunogenicity and protection. In vivo expression analysis revealed that EsxG-H and EsxB-A are highly expressed throughout the infection, whereas EsxW-V and EsxDC constitute small, but significant fractions of the transcriptome. Based on the immunogenicity, expression pattern, and protective efficacy, the ESX dimer substrates EsxD-EsxC, EsxG-EsxH, and EsxW-EsxV were combined into a subunit fusion. The novel subunit fusion, H65, was found to be highly immunogenic and induce robust CD4 T cell responses against the individual components. H65 was able to protect as well as H56 in CB6F1 mice, while protecting on par with BCG in B6C3F1. The improved protection was associated with increased frequencies of polyfunctional H65-specific CD4 T cells, high in triple positive (IFN-γ+TNF-α+IL-2+) and double positive (TNF-α+IL-2+), which have been associated with high memory potential. Furthermore, in an in silico based binding study we found that H65 had a broad predicted global coverage, with a minimum of seven predicted epitopes in each of the 34 HLA-DRB1 alleles common in TB highburden populations. 25

Results

Manuscript II Different human vaccine adjuvants promote distinct antigen-independent immunological signatures tailored to different pathogens

Purpose: To perform a head-to-head comparison of five different adjuvants, either licensed or in various stages of clinical trials, combined with three different vaccine targets from three disease models, tuberculosis, chlamydia and influenza in an effort to streamline the choice of clinical vaccine adjuvant for novel vaccine constructs. Results: We performed a direct comparison of three adjuvants currently in clinical trials, GLASE, IC31® and CAF01, and two adjuvant already licensed for human use, MF59® and Alum. Overall, we found that the adjuvants induced immunological signatures independent of the disease-specific antigenic targets and that there was good agreement between the murine data in this study and the previously published human studies. Alum was able to increase IgG1 antibody titres, but unable to improve cell mediated immunity. MF59® induced IgG1 and antigen specific IL-5 responses. GLA-SE was able to improve both IgG1 and IgG2a antibody titres, while also being a mild Th1 inducer. IC31® induces an IgG2a antibody response along with a strong Th1 CD4 response. CAF01 only induces a moderate increase in antibody titres, but promotes the strongest Th1 profile of any of the tested adjuvants, along with a Th17 T cell response. The induced profiles led MF59® and GLA-SE to protect against influenza, measured by induction of influenza HA inhibition titres, while CAF01, GLA-SE and IC31® enhanced protection to TB and chlamydia, measured by reduced bacterial burdens in the respective disease models. Manuscript III An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology.

Purpose: To characterize an Mtb clinical isolate capable of causing extrapulmonary tuberculosis, despite the absence of an EsxA specific T cells response and with an unusual resistance to H56 vaccination. Results: The Mtb DK9897 isolate was initially identified as a clinical isolate with attenuated in vivo growth, but with an unusual resistance to the phase II clinical-trial vaccine H56. We found that while the in vitro growth was comparable to Mtb Erdman and that the in vivo infection with Mtb DK9897 did not induce an EsxA specific CD4 T cell response. On further investigation, we found that Mtb DK9897 is able to produce EsxA, but that the protein is retained within the bacteria. Using whole genome sequencing we found that the EccCa1 (Rv3870) gene, essential for ESX-1 26

Results

secretion, has a frameshift mutation, resulting in a truncated version of the protein. Complementation of Mtb DK9897 with a plasmid borne version of the intact Rv3870 gene resulted in complete restoration of the EsxA function and an increase in virulence. Furthermore, we found that the lack of EsxA secretion resulted in lower levels of cytokine expression at the site of infection, less gross pathology and reduced ability to escape the confinement of the phagosome.

Manuscript IV A clinical isolate able of causing chronic disease in the absence of ESX-1 secretion exhibits alternative immune pattern in murine challenge model

Purpose: To elaborate how the lack of a functional ESX-1 system and associated EsxA secretion, influences the host-pathogen interaction during in vivo infection with DK9897. Results: When used in an aerosol based challenge model, infection with DK9897 was associated with an initial lag phase in onset of exponential replication, attenuated growth rates and delayed onset of adaptive immunity. The overall attenuation during in vivo growth resulted in decreased induction of cytokines (IFN-γ, IL-12) and chemokines (MCP-1, KC/GRO, IP-10), which resulted in attenuated recruitment of immune cells to lungs compared to Mtb Erdman. In contrast, when administered intravenously, the DK9897 infection demonstrates systemic virulence comparable to Mtb Erdman with similar bacterial burdens and induction of adaptive immunity. Pathogen specific CD4 T cells were detectable in both spleens and lungs after only 11 days of infections with both DK9897 and Erdman. Despite the increased fitness observed with IV challenge, the DK9897 exhibited attenuated dissemination to the lungs, comparable to the delay observed with an aerosol infection. Furthermore, utilizing an intravenous staining technique in combination with in vitro antigen restimulation, we found that despite being induced earlier in infection, the Mtb specific T cells are unable to enter the lung parenchyma until a certain level of mycobacteria had been reached. The lowered bacterial burdens of Mtb DK9897 in the lungs resulted in a diminished recruitment of EsxH specific CD4 T cells to the parenchyma in compared to Erdman infected animals.

27

Unpublished Results

Unpublished Results In this section, I present and discuss a small selection of additional data generated during my studies, which relates to the overall aim of the thesis. The materials and methods used for these studies are listed in the end of this section. Functional difference between EsxA and EsxH specific T cells Introduction

As mentioned, Mtb is capable of manipulating the host immune response and evading eradication by a broad spectrum of different virulence factors, many of which are secreted by the ESX systems. Two of the best characterized ESX substrates are EsxA (ESAT-6) secreted by ESX-1 and EsxH (TB10.4) secreted by ESX-3[253]. The two antigens are structurally similar[254], highly expressed throughout infection[255], strongly recognized by the immune system[256], and secreted as part of dimer complexes[257]. The proteins have been evaluated as vaccine candidate antigens in novel fusion proteins that differ only in the incorporation of either EsxA or EsxH. The fusions were evaluated in the preventive and post-exposure TB challenge models described earlier[232]. While all hybrids, in combination with CAF01, gave rise to protection in the preventive model, only the EsxA-containing constructs gave rise to protection in the post exposure setting[232]. As expression of EsxA and EsxH have not been demonstrated to be vastly different[247], we hypothesized that the lack of protection in the post-exposure model could be attributed to functional differences in the EsxA vs EsxH specific T cells after infection and vaccination. In brief, previous work has indicated that KLRG1 expression is a marker of a terminally differentiated CD4 T cell subset with poor entry into the parenchyma, reduced proliferative potential and weak protection against TB, while PD-1 expression is a marker for a highly activated T cell population with efficient entry into the parenchyma and increased protections against Mtb[152, 162]. Mtb-specific T cells found in the lung parenchyma are more protective than their counterparts stuck in the lung vasculature[152]. We hypothesized that EsxH specific T cells are more functionally exhausted at the post-exposure vaccination time point and thus less susceptible to vaccination with the CAF01/hybrid vaccines. T-bet is a transcription factor directing Th1 CD4 T cell commitment[258]. ICOS is a costimulatory molecule that is expressed on activated T cells[259]. CXCR3 is a chemokine receptor regulating T cell trafficking activated by e.g. CXCL10 (IP-10)[260]. CCR7 and CD62L are markers for memory T cells[147]. TIM3 is an inhibitory marker like PD-1 and the two are often co-expressed[261]. 29

Unpublished Results

Results

In an effort to address potential differences of the EsxA- and EsxH-specific T cells, we vaccinated groups of mice with hybrid fusions H56 (Ag85b-EsxA-Rv2660c) and H28 (Ag85b-EsxHRv2660c) in the post-exposure and preventive setting. Using flow cytometry we assessed the functional phenotype of vaccine-specific epitopes using the EsxA- and EsxH-specific class II tetramers (NIH) combined with an intravascular CD45 staining technique[262]. The IV staining revealed that there was no difference in recruitment of EsxA or EsxH specific CD4 T cells during the course of infection in both settings (data not shown). Furthermore, we analysed the tetramer specific CD4 T cells for expression of KLRG1, PD-1, Tbet, CD103, CXCR3, CD62L, CCR7, ICOS and Tim3. Comparing the EsxA-specific CD4 T cells from H56 vaccinated animals and EsxH-specific CD4 T cells from H28 vaccinated animals there is no difference in any of the analysed markers (Fig. 7). There was low constitutive expression of KLRG1, PD-1 and CD103, while more than 50% of tetramer positive CD4 T cells were CXCR3+ for the duration of the infection. A similar pattern was observed with the hybrid fusions in the preventive model (data not shown).

Figure 7 – Post exposure vaccination, Hybrid. CB6F1 mice were aerosolly challenged with Mtb, rested for six weeks, and treated for six weeks with antibiotics. Animals were vaccinated with Hybrid 56 (H56) or H28 in CAF01 at week 10, 13 and 16 after infection. The X-axis indicate weeks after final vaccination. Lung cells were stained for flow cytometry and tetramer positive T cells were identified as follows: Singlets > Lymphocytes > DUMP- CD3+ > CD4+ > CD44highTetramer+. From this, the fraction of KLRG1+, PD1+, CD103+ and CXCR3 were determined. Mean + SEM are indicated (n=4).

However, comparing the EsxA- and EsxH-specific CD4 T cells isolated from saline vaccinated control animals, we observed different expression levels for several of the assessed surface 30

Unpublished Results

markers (Fig. 8). Initially in the post exposure model the expression KLRG1 is elevated in EsxA specific T cells, with around 40% KLRG1+, this however drops to around 20% and continues to fall throughout the infection. In contrast, the EsxH specific T cells initially have a lower fraction of around 28% KLRG1+, which slowly increases during infection and peaks at around 32% at week five of Mtb infection. In addition, EsxA specific CD4 T cells constitutively express higher levels of PD1, CD103 and CXCR3 compared to EsxH (Fig. 8). The same pattern is observed in the preventive setting with saline vaccinated animals (data not shown).

Figure 8 - Post exposure vaccination, Saline. CB6F1 mice were aerosolly challenged with Mtb, rested for six weeks, and treated for six weeks with antibiotics. Animals were vaccinated with PBS at week 10, 13 and 16 after infection. The X-axis indicate weeks after final vaccination. Lung cells were stained for flow cytometry and tetramer positive T cells were identified as follows: Singlets > Lymphocytes > DUMP- CD3+ > CD4+ > CD44highTetramer+. From this, the fraction of KLRG1+, PD1+, CD103+ and CXCR3 were determined. Mean + SEM are indicated (n=4).

When directly comparing the EsxA- and EsxH- specific T cell responses in the saline group to the hybrid vaccinated mice, it becomes evident that while Mtb infection induces a differential EsxAand EsxH- specific profile, hybrid vaccination with CAF01 equalizes the phenotype (Fig. 9A). Independent of the levels induced by Mtb alone, the hybrid/CAF01 vaccinations result in comparable expression of KLRG1, PD1, CD103 and CXCR3 on the vaccine-specific T cells (Fig. 9A), as well as Tbet, CD62L, CCR7, ICOS and Tim3 (data not shown). Furthermore, hybrid vaccinations induces an antigen specific immune response dominated by triple positive IFN-γ+IL2+TNF-α+ T cells constituting ~60% of the cells, but also with considerable population of double 31

Unpublished Results

positive IL-2+TNF-α and TNF-α single positives, independent of the vaccine antigen used (Fig. 9B).

%+ve of TET+ CD4

%+ve of TET+ CD4

A

B

Saline

H56

H28

EsxA

EsxH

Figure 9 – Post exposure vaccination, Saline and Hybrid. CB6F1 mice were aerosolly challenged with Mtb, rested for six weeks, and treated for six weeks with antibiotics. Animals were vaccinated with Saline, Hybrid 56 (H56) or H28 in CAF01 at week 10, 13 and 16 after infection. Organs were harvested five weeks after final vaccination. A) Lung cells were stained for flow cytometry and tetramer positive T cells were identified as follows: Singlets > Lymphocytes > DUMP- CD3+ > CD4+ > CD44highTetramer+. From this, the fraction of KLRG1+, PD1+, CD103+ and CXCR3 were determined. Saline indicates either EsxA or EsxH specific responses in the Saline vaccinated mice, whereas Hybrid referrers to vaccination with either H56 or H28, for EsxA and EsxH specific T cell responses, respectively. Mean + SEM are indicated (n=4). B) Animals were assayed using IV staining in combination with in vitro restimulation with EsxA and EsxH, and then evaluated for their production of TNF-α, IFN-γ and/or IL-2. The pies show only the IV- populations. Pies represent means of 4 animals.

Discussion

As previously described, the quality and quantity of the CD4 T cell immune response is detrimental in the protection against Mtb. Surprisingly, we found no clear differences in the phenotype of EsxA- and EsxH- specific CD4 T cells that could explain the disparities in post-exposure protection by H56 and H28 previously observed by Hoang et al.[232]. In the work by Hoang et al., vaccinations with H56 and H28 were able to alter the phenotype of the EsxA- and EsxHspecific CD4 T cells, respectively, but in line with what was observed in Manuscript II, the phenotypes induced by vaccinations seem to be dictated by the adjuvant and not by the antigen target. However, when EsxA and EsxH responses are primed and driven solely by the infection, EsxH specific CD4 T cells express higher levels of KLRG1 and decreased PD1, indicating a more terminally exhausted phenotype[152]. This correlates well with the cytokine profiles induced 32

Unpublished Results

during in vitro stimulation with recombinant EsxA and EsxH. Comparing the antigen specific responses in the saline group, the EsxH specific CD4 T cells have a larger fraction of double positive IFN-γ+TNF-α+ and IFN-γ single positive T cells, which combined make up roughly 70% of the total response. These phenotypes have been associated with terminal differentiation and exhaustion of the Th1 CD4 T cell response[152]. In contrast, the EsxA response is dominated by triple-positive IFN-γ+IL-2+TNF-α+, and along with the TNF-α+ single positive, these make up roughly 50% of the total response. Furthermore, in agreement with the lack of difference in the surface markers of EsxA and EsxH tetramer specific T cells, the cytokine profiles for EsxA in H56 vaccinated mice and EsxH in H28 are virtually indistinguishable. These are comparable to what was previously observed for vaccination with CAF01 in the post exposure[232] and preventive model[231], as well as what was seen in Manuscript II. When designing the experiment we expected to find EsxH specific T cells that were functionally more exhausted than their EsxA specific counterparts and thus less susceptible to the priming effects of the Hybrid/CAF01 vaccination. This was however not the case, since the vaccinations were able to equalize the phenotypes, unaffected by their initial starting point.

Despite the lack of difference between the EsxA and EsxH tetramer specific T cells in the current data set, the given information does not necessarily dismiss the original hypothesis. Since the protective efficacy in the post exposure model is usually assessed 20 weeks after the final vaccination, the week 10 time point measured in this experiment so far, might not provide a sufficient window to determine potential differences. Since we saw that EsxH specific T cells are more terminally differentiated in the saline control group, this indicates that the infection-driven immune pressure on this subset is greater compared to EsxA specific T cells. Given the same starting profile that CAF01 vaccinations provide, the increased pressure on the EsxH population could with time cause the loss of Mtb control and result in increased bacterial burdens at week 20 post vaccination. Furthermore, the infectious load assessment at the 20 week time point is essential to determine if the original finding that “H56 protects and H28 does not” is reproduced in this experiment.

33

Unpublished Results

EsxA based vaccine unable to protect against DK9897 while EsxH hybrid does Introduction

In Manuscript III, we demonstrate that a clinical isolate, DK9897, capable of causing extrapulmonary TB in the source patient is unable to secrete EsxA and is unusually resistant towards H56 vaccination. In Manuscript IV, we demonstrate that part of the attenuation of DK9897 is associated with the aerosol inoculation route, and when administered intravenously, the virulence is more comparable to Mtb Erdman. In an effort to determine if the altered virulence observed with IV administration had an effect on the resistance towards vaccination we decided to assess the two novel vaccine candidates, H64 and H65, against DK9897 challenge. H64 is a fusion protein based on ESX-1 secreted proteins, EsxA-EspD-EspC-EspF-EspR-PE35, and H65 is based on non-EsxA, WxG100 family proteins, EsxD-EsxC-EsxG-EsxH-EsxW-EsxV (Manuscript I).

Results

Groups of mice were vaccinated with either H65 or H64 fusion proteins in CAF01 adjuvant, or BCG. Six weeks after the third vaccination all mice were challenged with Mtb strain Erdman or Mtb strain R98-97, either aerosolly (Fig. 10A) or intravenously (Fig. 10B). As expected, BCG yields the best protection in both cases, and not unexpected, H65 is also able to significantly reduce the bacterial, significantly better than H64. Three weeks before the vaccine efficacy assessment, the vaccine specific T cell responses in Mtb DK989 challenged mice were investigated by flow cytometry in lung-derived lymphocyte cultures. In the aerosol model, H64 and H65 drive robust EsxA and EsxH specific CD4 T cell responses, respectively, each dominated by triple positive IFN-γ+IL-2+TNF-α+, double positive IL-2+TNF-α+ and single positive (TNF-α) CD4 T cells (Fig. 10C). In the intravenous challenge model, only a moderate EsxA CD4 T cell response is maintained three weeks into the infection, while strong EsxH specific CD4 T cell responses are detected in all vaccine groups (Fig. 10D). In the IV challenge model, the non-EsxH inducing vaccines are dominated by double positive (IFN-γ+TNF-α+) and single positive (IFN-y+), while the EsxH primed groups have higher fractions of positive (IFN-γ+IL-2+TNF-α+), double positive (IL-2+TNF-α+) and single positive (TNF-α+).

34

Unpublished Results

Figure 10 - Esx-family based vaccine is capable of protecting against Mtb DK9897 challenge. Groups of mice were vaccinated once with BCG or three times with either saline, H64 (EsxA-EspD-EspC-EspF-EspR-PE35) or H65 (EsxD-EsxC-EsxG-EsxHEsxW-EsxV) recombinant protein formulated in CAF01 adjuvant. Mice were rested for six weeks and subsequently aerosolly or intravenously challenged with ether Mtb Erdman or Mtb DK9897. After six weeks of infection the animals were sacrificed and the bacterial burden was assessed by bacterial enumeration. A) Aerosol challenge with Mtb DK9897. B) Intravenous challenge with Mtb DK9897. A+B: Mean and SEMs are indicated (n=8). C)+D) From infections corresponding to figure C and D, the polyfunctionality of vaccine primed and infection driven CD4 T cells were assessed by flow cytometry. Measuring the frequencies of the seven possible combinations of TNF-α, IFN-γ and/or IL-2 producing T cells after stimulation with EsxA or EsxH in the aerosol (C) or intravenous challenge (D) (mean of 3 mice).

Discussion

Since several of the vaccine candidates currently in clinical trials are based on EsxA, the lack of an immune response against this key antigen in a clinical isolate, capable of causing extrapulmonary TB, makes the Mtb DK9897 strain a unique challenge for our lead candidates. As previously published, while BCG and H56 are capable of protecting against virulent strains like Mtb Erdman, H56 only shows moderate protection against aerosol challenge with Mtb DK9897, while the protective efficacy of BCG is high compared to Mtb Erdman (see Manuscript III). The 35

Unpublished Results

H64 fusion protein, based on antigens secreted through the ESX-1 secretion system, including EsxA, protects on par with BCG against Mtb Erdman, but shows no significant protection against DK9897, lacking ESX-1 secretion. The H65 fusion protein on the other hand targets antigens secreted via ESX-2(EsxD-EsxC), ESX-3(EsxG-EsxH) and ESX-5(EsxW-EsxV)(see Manuscript I). In Manuscript III, we demonstrated that while Mtb DK9897 has a dysfunctional ESX-1 system, the paralogue ESX-3 system, responsible for secreting EsxH appears to be intact and working. As expected, with H65 targeting an intact secretion systems capable of excreting antigenic targets, resulted in excellent protection against aerosol and IV challenge with Mtb DK9897. From the cytokine profiles measured after three weeks of infection, it is evident that H64 and H65 induce very comparable qualities of effector T cells towards their respective antigens EsxA and EsxH, high in both memory and effector potential[147]. However, the lack of ESX-1 secretion by Mtb DK9897 infection renders H64 unable to protect significantly. There is however a slight reduction in the bacterial burden of H64 vaccinated animals challenged aerosolly with DK9897 compared to saline controls, which is not observed in the IV challenge. We argue that this could be facilitated by vaccine-induced antibodies or by an innate mechanism triggered by vaccination. It could however also be mediated by an immune response towards one of the non-EsxA components leaking out of the defective ESX-1 system or one of the other ESX paralogues. While lack of EsxA secretion was confirmed(Manuscript III), none of the other H64 components were confirmed absent in the culture filtrates of DK9897. This could be addressed by assessing in vitro immune recalls towards the individual components of H64 during a DK9897 infection.

As mentioned, we observed an increased virulence of Mtb DK9897 when administered intravenously compared to an aerosol challenge. Comparing the cytokine profiles of EsxA- and EsxH-specific CD4 T cells recruited to the lungs, there is a stark difference in the CD4 T cell profiles between the two inoculation routes. Where the profiles observed in the aerosol challenge are high in memory potential, the cells recruited in the IV infection are high in terminally differentiated/exhausted phenotypes. The degree of exhaustion is increased in unprotected groups and suggests an increased virulence of DK9897 in the intravenous challenge model. Overall, even though Mtb DK9897 is an attenuated isolate and part of a small cluster of isolates (see Manuscript III), the characterization of this isolate illustrates how small phenotypical differences can have detrimental outcomes on vaccine efficacy. Antigens selected for vaccines need to be important for virulence, strongly antigenic, available for presentation throughout infection and conserved across different strains across the world. 36

Unpublished Results

Materials and Methods Animals Six to eight week old female CB6F1 mice (Envigo, Scandinavia) were rested for one week prior to initiation of experiments. The handling of mice was conducted in compliance with the European Community Directive 2010/63 for the care and use of laboratory animals. Mice were housed under specific pathogen–free conditions in animal facilities at Statens Serum Institut (Copenhagen, DK), provided with radiation-sterilized food and water ad libitum. Infected animals were housed in biosafety level III facilities. All techniques and procedures have been refined to provide for maximum comfort and minimal stress to the animals. In agreement with the Danish Animal Welfare Act, all experimental protocols involving animals were reviewed prior to the start of the experiment by an independent ethical review board at Statens Serum Institut and approved to be in accordance with our license for animal experiments issued by The Animal Experiments Inspectorate (License no 2014-15-2934-01065 and 2012-15-2934-00272) under the Ministry on Environment and Food of Denmark.

Immunization Mice were immunized subcutaneously (s.c.) at the base of the tail either once with BCG (BCG Danish 1331 (SSI, DK)) or three times at two-week (preventive) or three-week (post-exposure) intervals with 5µg H56, H28, H65 or H64 in CAF01 (SSI, DK). Hybrid fusion were recombinantly produced, as described previously: H56[149], H65[247] (Also applies to H64) and H28[263].

Mycobacterial challenge Mtb DK9897 and Mtb Erdman (ATCC 35801) were grown at 37°C for 3–4 weeks either on solid medium (Middlebrook 7H11) or in liquid medium (Middlebrook 7H9) supplemented with 10% oleic acid–albumin–dextrose–catalase (OADC). The handling of strains was done in biosafety level III facilities at either Statens Serum Institut, Denmark. Six weeks after the third immunization animals were infected by the aerosol route[264] or by intravenous injection of 200 uL 2.5×105CFU/mL, with either Mtb Erdman or Mtb DK9897. Six weeks after infection, mice were sacrificed and organs homogenized in PBS for bacterial enumeration [264].

37

Unpublished Results

Lymphocyte cultures and flow cytometry Lymphocytes were isolated from lungs as described previously[265]. Cultures were adjusted to 12×106 cells/well in a total volume of 200 µl/well for flow cytometry, and stimulated with antigens at a final concentration of 2 µg/ml. Lymphocytes for IC-FACS were stimulated at 37 °C in the presence of recombinant antigen (2 µg/ml) for 1 hour, and subsequently incubated for 6 hours after adding 10 µg/ml Brefeldin A (Sigma-Aldrich) and stained with antibodies[266]. The following antibodies were used for surface staining: anti-CD4-APC (RMA4-5, BD), anti-CD44-FITC (IM7, eBioscience). For intracellular staining, the following antibodies were applied: anti-IFN-γ-PE-Cy7 (XMG1.2; eBiosciences), anti-TNF-α-PE (MP6-XT22; eBiosciences), anti-IL-2-APC-Cy7 (JES65h4; BD). Alternatively, cells were stained with tetramers and surface stain: I-Ab:ESAT-64-17-PE, I-Ad:TB10.470-84-PE, and I-Ab: hCLIP-PE, I-Ad:hCLIP-PE-negative controls were provided by the NIH tetramer facility (Atlanta, USA). Cells were stained with tetramers for 30 min at 37 °C before surface staining at 4 °C with cocktails of anti-CD3-BV650 (1:100, 17A2; Biolegend), anti-CD4BV510 (RM4-5, Biolegend), anti-CD44-AF700 (IM7; Biolegend), anti-CD8-APC-Cy7 (53-6.7, BD), anti-CD90.2-PB (53-2.1, Biolegend), anti-CCR7-AF780 (4B12, eBiosciences)(stained at 37°C), anti-CXCR5-BV785 (L138D7, Biolegend)(stained at 37°C), anti CD62L-APC (MEL-14, BD), anti-CXCR3-BV421 (CXCR3-173, Biolegend), anti-TIM3-PE-Cy7 (RMT3-23, Biolegend), anti-KLRG1-APC-Cy7 (2F1/KLRG1, Biolegend), anti-PD1-BV785 (29F.1A12, Biolegend), antiICOS-PE-Cy7 (C398.4A, Biolegend), anti-CD103-PE-Cy7 (2E7, Biolegend), anti-Tbet-APC (4B10, Biolegend), anti-CD45.2-FITC (104, BD)(IV stain), DUMP: anti-CD11b-PerCP-Cy5.5 (M1/70, Biolegend), anti-CD11c-PerCP-Cy5.5 (N418, Biolegend), anti-F4/80-PerCP-Cy5.5 (BM8, Biolegend), anti-B220-PerCP-Cy5.5 (RA3-6B2, Biolegend) and anti-NK1.1-PerCPCy5.5(PK136, Biolegend). Responses were analyzed using a LSRFortessa flow cytometer (BD) and FlowJo v.10.2 (Tree Star).

Statistical analysis Prism 7 software (GraphPad v7.02) was used for all statistical analyses. Bacterial numbers were log-transformed before being analyzed using one way ANOVA with Tukey’s multiple comparisons test.

38

Discussion

Discussion This thesis demonstrates that a clinical isolate, Mtb DK9897, isolated from a patient with extrapulmonary TB, exhibits attenuated in vivo growth during aerosol challenge in mice and shows increased sensitivity towards BCG primed immunity, but demonstrates unusual resistance towards vaccination with H56. We determined that this resistance is caused by a mutation in a component of the ESX-1 system, rendering DK9897 unable to secrete EsxA. We show that a vaccination targeting the intact ESX-1 paralogues, namely the H65 vaccine, results in efficient protection against DK9897. Furthermore, we found that H65 in combination with CAF01 is able to protect on par with H56 and BCG against challenge with Mtb Erdman. Finally, we were able to show that while the choice of antigen is important for efficiently targeting the pathogen, the choice of adjuvant determines the type of immune response, independent of the vaccine antigen.

The concept of vaccination encompasses the priming of the immune system with a pathogen specific antigen in order to develop adaptive immunity against the pathogen. Upon subsequent encounters, the early recognition will prevent the development of disease. However, in order to be protective, the primed immune response needs to fulfil several criteria: 1) The immune response needs to target a relevant antigen, i.e. the target should ideally be essential for pathogen growth, minimizing the possibility of mutations. 2) The antigen needs to be available for recognition during infection, i.e. choosing an antigen that is downregulated early after onset of infection or selecting an antigen only secreted under certain conditions do not make a good vaccine targets on their own. 3) The induced immune response needs to match the natural immune response induced against the pathogen, i.e. induction of an antibody response is going to have little effect against an intracellular pathogen, while a pure cell mediated immune response is going to be worthless against e.g. influenza. In addition to these, several other factors like age group, population diversity, genetic variations in the pathogen, compatibility with other diseases (e.g. HIV/AIDS) and prior exposure all play a major part in vaccine efficacy.

As mentioned, the TB vaccines currently in the clinical pipeline are largely based on a small subset of antigens that are immunodominant during natural Mtb infection, with the majority of vaccines including either Ag85A and/or Ag85B. Although the method of delivering these antigens ranges from overexpression in recombinant BCG, genetically engineered viral vector or as a protein fusion combined with an adjuvant[65], this panel represents a very narrow selection of antigens. Combining this with the failure of the MVA85A vaccine, a modified Vaccinia Ankara virus 39

Discussion

expressing Ag85A, to boost the protective efficacy against TB in infants previously vaccinated with BCG[267], raises additional concern about the lack of antigen diversity in the pipeline. The study presented in Manuscript I describes the rational design of a novel fusion protein, H65, combining several Mtb-secreted antigens in order to target several ESX secretion systems. Of the five ESX paralogues, ESX-1, -3 and -5 have been found functional, mediating various Mtb functions by secreting effector proteins[67, 79, 268], whereas no information about function or active secretion by ESX-2 or -4 have been published. ESX-1, -3 and -5 have all been associated with mediation of virulence[206, 269], while ESX-3 is also required for iron transport and thus viability of Mtb[79]. Despite the large effort put into the study of the ESX secretion system, associating several Mtb specific attributes to the various systems, there is still a large gap in linking specific functions to certain substrates. However, since we found a large degree of homology between substrates secreted through the same secretion system, H65 should, at least in theory, be able to prime for recognition of most of the ESX secreted WxG100 proteins. This, combined with the high in silico predicted population coverage, makes H65 a promising TB vaccine candidate.

In a direct comparison, the H65 fusion protein protected as well as the phase II clinical construct H56 (Ag85B-EsxA-Rv2660c) against TB challenge in mice. The H56 construct combines antigens expressed at various stages of Mtb infection in a multistage vaccine strategy, providing protection in mice[231] and non-human primates[270]. However, in Manuscript III we demonstrate that while H56 vaccination provides efficient protection against Mtb Erdman challenge, the H56 immunity provides only a moderate reduction in bacterial burden when challenged with the clinical isolate DK9897. Surprisingly, DK9897 exhibits attenuated growth during in vivo infection in mice and is remarkably susceptible to BCG induced immunity. We found that DK9897 was unable to secrete EsxA due to a defect in the ESX-1 secretion mechanism and thus unable to prime an EsxA specific immune response, accounting for the large drop in H56 efficacy. Despite the lack of EsxA secretion, the isolate was able to cause extrapulmonary TB in the source patient. Previously, outbreaks of Mtb have been associated with the emergence of clinical isolates with altered virulence, as discussed in Manuscript IV. The Mtb CDC1551 isolate was the cause of an outbreak associated with high close contact infection rates and was found to induce elevated host immune responses during infection, however with no increased virulence in mice[271]. Around the same time, the HN878 isolate was the cause of several outbreaks, which was found to induce decreased immune responses in murine challenge studies[272]. HN878 however caused significantly higher mortality and pathology, although reaching similar bacterial burdens as CDC1551[273]. The HN878 strain is part of the W-Beijing lineage, a family of strains which represent 50% of Mtb 40

Discussion

isolates in East Asia, and about 13% of all Mtb isolates worldwide[160]. It has been suggested that part of this increased prevalence is caused by the BCG vaccine being less effective against WBeijing[160]. Despite being hyper- and hypoimmunogenic, the two isolates were both able to cause separate outbreaks of TB. The contrast in the two isolates illustrates how small different in the host-pathogen interactions can tip the scale in favour of Mtb.

As demonstrated in Manuscript IV, the aerosol infection with DK9897 results in attenuated in vivo growth in mice, exhibiting a lag phase in the onset of exponential growth and diminished doubling rates. We argue that the overall attenuation is caused by the suggested role EsxA has in penetrating the alveolar wall[274] and the many virulence functions of ESX-1/EsxA secretion, essential for establishing successful infection (Discussed in Manuscript III and IV). Considering the lack of this key virulence factor and the absence of lung pathology, it is puzzling how the DK9897 strain was transmitted and able to establish infection in the first place. Yet, seeing as the strain was able to cause extrapulmonary TB in the source patient and considering the recently speculated theory that TB is a lymphatic disease with a pulmonary gateway[275], we decided to assess DK9897 in an intravenous challenge model. We found that while arrival of bacterial to the lungs was still delayed compared to Mtb Erdman, the ability to establish infection in the spleen and replication rates were equalized between strains. This indicates that, while ESX-1 is important for dissemination to the lungs, it is less critical in establishing infection in other tissues. Furthermore, while the induction of an Mtb specific immune response in the aerosol challenge was delayed for up to three and four weeks for Erdman and DK9897, respectively, both strains induced a response in less than 11 days in the IV challenge model. This is in agreement with previous studies, where IV inoculation can generate robust T-cell activation within 1–3 days post-infection[276]. This suggests that the immune evasion tactics developed by Mtb are very specific to manipulating the lung environment. In the lymphocentric model, the lungs function as an entry and exit point, meaning that the attenuated ability of DK9897 to enter the lungs and lack of immunopathology would likely result in reduced disease transmission, thus probably meaning an evolutionary dead end for DK9897. Considering the fact that the index case for DK9897 was elderly and was not part of major outbreak of TB, this suggests that while DK9897 is interesting from an academic perspective, the clinical implications are probably relatively minor. However, since the protection of the clinical-trial TB vaccine H56 was suboptimal against this isolate, while the vaccination with H65 was protective in both the aerosol and IV model, the study of this strain highlights the problem of potentially releasing a vaccine candidate with a too narrow selection of antigens. Scientist developing vaccines against the human papillomavirus (HPV), responsible for causing cervical cancers, 41

Discussion

currently face this issue. While vaccines contain antigens against several prevalent subtypes of HPV, it leaves the vaccine recipient unprotected against dozens of other high-risk subtypes responsible for cervical cancers[277]. While the vaccines might prevent the spread of the chosen subtypes, the selective immune pressure will most likely result in the surge of previously less frequent strains. Ideally, this could be avoided by including antigens conserved across different disease subtypes. However, as a counter argument for this proposal, the genome sequencing of 21 strains of Mtb, representative for the global diversity, showed that 491 experimentally confirmed human T cell epitopes were evolutionary conserved[278]. This suggests a strong pressure on the maintenance of the corresponding antigens and could be interpreted to mean that these epitopes present an advantage to the pathogen instead of the host. This could imply that future vaccine development might benefit from focusing on proteins that vary among different TB strains.

The vaccines used in these studies were all formulated in the cationic adjuvant formulation 01 (CAF01), a liposomal adjuvant system capable of inducing persisting CD4 T cell responses in both mice[256] and humans [279]. CAF01 induces a strong Th1 biased cell mediated immune response as well as vaccine specific IgG1/2 antibody responses[280]. There are however several other clinical adjuvants included in the TB vaccine pipeline like GLA-SE, IC31® and AS01. All of these are based on different delivery strategies combined with immunostimulatory molecules. GLA-SE is a stable oil emulsion utilizing the TLR4 stimulating glucopyranosyl lipid[281]. IC31® is a combination of immunostimulatory antibacterial peptides and the TLR9 stimulating ODN1a[282]. AS01 and CAF01 are liposome-based adjuvants utilising the TLR4 stimulating MPL[283] and the MINCLE-receptor stimulating TDB[284], respectively. The adjuvants have all individually been reported to induce Th1 CD4 T cells responses and mediate protection against TB challenge. However, differences in the used vaccine targets, number of vaccinations, challenge strain and dose, make the comparison between these virtually impossible based on the published data. Due to an international collaborative research programme, Advanced Immunization Technologies (ADITEC), we were able to perform a head-to-head comparison of several adjuvants in the clinical pipeline, GLA-SE, CAF01 and IC31®, described in Manuscript II. Furthermore, this comparison was extended to include two already licensed adjuvant systems, MF59® and Alum, which were all assessed across three different disease targets, TB, Chlamydia and Influenza. We found that the adjuvants were able to induce immunological profiles that were independent of the antigenic vaccine-target used and correlated well with what had previously been publish in humans, as discussed in Manuscript II. CAF01, GLA-SE and IC31® were all able to induce vaccine specific responses dominated by triple-positive IFN-γ+IL-2+TNF-α+, double positive IL-2+TNF-α and 42

Discussion

TNF-α single positive CD4 T cell. It has previously been published that these populations indicate a T cell population high in memory potential[147] and that these are important for protection against TB[265, 285]. Consistent with this we found that all three adjuvants were able to induce protection against TB, but also that there was no correlation between the magnitudes of either of these populations to the level of protection. This indicates, that while the aforementioned T cell populations are important for protection, they are not a direct correlate of TB protection.

Finally, while the head-to-head comparison described above was performed with the H56 subunit fusion as the comparative target, similar studies comparing different TB specific antigenic targets would be equally, if not more, interesting. As illustrated with the difference in protective efficacy between Erdman and DK9897, and the two fusion proteins H56 and H65, the selection of the targeted antigens might be as important as the choice of adjuvant. In a direct comparison between fusion proteins only differing in the content of either EsxA or EsxH, the fusions containing EsxA were able to protect in a preventive and post-exposure setting, while the EsxH based fusion only protected in the preventive model[232]. In an attempt to determine the cause of discrepancy between two such closely related antigens, we assessed the phenotypes of the CD4 T cell populations specific for each antigen, based on surface markers recently published by Sakai et al, 2014[152]. By assessing markers specific for the parenchymal and intravascular residing T cells, we hoped to identify a difference between the two vaccinated EsxA- and EsxH-specific T cell populations. Following hybrid vaccination with either H56 (containing EsxA) or H28 (containing EsxH), we were unable to identify dissimilarities between the two populations. The T cell populations were very similar in expression of every investigated marker and recruited similarly to the parenchyma as identified by IV staining[262]. Furthermore, we observed a similar profile of cytokine production between the EsxA- or EsxH-specific T cell populations, which correlates well with the facts that both H56 and H28 were administered in CAF01. However, in the absence of the CAF01 vaccination (i.e. saline vaccinated controls), there were phenotypical differences between the two populations. EsxH-specific T cells exhibited increased KLRG1 expression and diminished PD-1, CD103 and CXCR3, all of which indicate a response leaning towards the intravascular phenotype with reduced protection against TB[152]. Furthermore, in the unvaccinated animals, the cytokine profile for EsxH specific CD4 T cells has a higher proportion of IFN-γ single positives, which is also a sign of exhaustion/terminal differential[147]. We observed the same trend in the preventive setting, but as with the post exposure experiment, the differences were abrogated by vaccination. Taken together, this could indicate that, while the H56 and H28 vaccines prime a similar response with CAF01, the increased immune pressure on the 43

Discussion

EsxH specific T cells exerted by the infection will eventually cause the host to lose control. While functional differences between EsxA and EsxH relating to bacterial pathophysiology could explain the discrepancy in protection, the far more trivial, but likely, explanation is the duration between final vaccination and assessment of protective efficacy. In the post exposure model, the efficacy is usually assessed 20 weeks after vaccination, where the mice are at least 42 weeks old, whereas the preventive model is usually evaluated after 6-12 weeks of infection, at an age of 22-28 weeks. The H56/CAF01 vaccine has been proven to induce protection even after 24 weeks of infection in the preventive model[231], while H28 has never been assessed at such a late time point. These points could be addressed with an extended preventive challenge model where animals have been age matched to the corresponding time points in the post exposure study.

44

Concluding Remarks and Perspectives

Concluding Remarks and Perspectives Despite being an ancient disease, Tuberculosis (TB) remains a major cause of morbidity and mortality worldwide. Mycobacterium tuberculosis (Mtb) has coevolved with human kind over millennia and has developed the ability to survive in the hostile human environment by the use of numerous immune evasion mechanisms. Despite the large number of vaccine candidates undergoing various stages of clinical development, the current TB clinical pipeline represents a very narrow set of Mtb antigens. In light of the failure of MVA85A[267] and the fact that human Mtb T cell epitopes seem to be hyperconserved[278], there is an urgent need for an increased diversity in the antigen portfolio. The current work illustrates how targeting specific virulence factors, in this case the ESX secretion systems, could be used to develop a vaccine capable of protecting against Mtb. Due to the ease of generating recombinant fusion proteins, the strategy could be adapted to include novel virulence factors as they are discovered. In addition, the study of DK9897, demonstrates how important it is to diversify the screening tools used during preclinical development. All the current candidates are selected based on experiments performed in mice challenged with lab-adapted Mtb strains, like Mtb Erdman or H37Rv. Illustrated by the loss of H56 protection due to the lack of EsxA secretion from DK9897, new candidates should ideally be evaluated by a broader spectrum of Mtb strains, preferably recently isolated clinical strains. Furthermore, since mice are the most commonly used model for TB vaccine studies, the use of inbred mice should be diversified to include several different strains, potentially also including inbred hybrids and diversity outbred mice[286]. Also, there is a large focus on generating novel antigen targets for TB vaccines, but as illustrated by the adjuvant comparison study in the current thesis, the subunit vaccines are equally dependent on the adjuvants to direct and shape the proper immune response. As demonstrated by the direct comparison, different adjuvants have unique immunological signatures making them ideal for different disease targets. The side-by-side comparison of competing formulations should therefore be encourage, as well as the open sharing of scientific findings. Fruitful collaborative research programmes, like the EU funded ADITEC initiative, should therefore be an inspiration to stakeholders and investors worldwide.

45

References

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14.

15. 16. 17. 18. 19.

20. 21. 22. 23.

WHO, Global tuberculosis report 2016. 2016: World Health Organization. Evans, T.G., L. Schrager, and J. Thole, Status of vaccine research and development of vaccines for tuberculosis. Vaccine, 2016. 34(26): p. 2911-4. Lin, P.L. and J.L. Flynn, Understanding latent tuberculosis: a moving target. J Immunol, 2010. 185(1): p. 15-22. Grange, J.M. and A. Zumla, The global emergency of tuberculosis: what is the cause? The journal of the Royal Society for the Promotion of Health, 2002. 122(2): p. 78-81. Luca, S. and T. Mihaescu, History of BCG Vaccine. Maedica (Buchar), 2013. 8(1): p. 538. Woodruff, H.B., Selman A. Waksman, winner of the 1952 Nobel Prize for physiology or medicine. Appl Environ Microbiol, 2014. 80(1): p. 2-8. Millard, J., C. Ugarte-Gil, and D.A. Moore, Multidrug resistant tuberculosis. BMJ, 2015. 350: p. h882. Diel, R., et al., Costs of tuberculosis disease in the European Union: a systematic analysis and cost calculation. European Respiratory Journal, 2014. 43(2): p. 554-565. Hayman, J., Mycobacterium ulcerans: an infection from Jurassic time? The Lancet, 1984. 324(8410): p. 1015-1016. Kapur, V., T.S. Whittam, and J.M. Musser, Is Mycobacterium tuberculosis 15,000 years old? Journal of Infectious Diseases, 1994. 170(5): p. 1348-1349. Webb, G., Tuberculosis, Hoeber. New York, 1936. Konomi, N., et al., Detection of mycobacterial DNA in Andean mummies. Journal of Clinical Microbiology, 2002. 40(12): p. 4738-4740. Zink, A.R., et al., Characterization of Mycobacterium tuberculosis complex DNAs from Egyptian mummies by spoligotyping. Journal of Clinical Microbiology, 2003. 41(1): p. 359-367. Hershkovitz, I., et al., Detection and molecular characterization of 9000-year-old Mycobacterium tuberculosis from a Neolithic settlement in the Eastern Mediterranean. PloS one, 2008. 3(10): p. e3426. Frith, J., History of tuberculosis. Part 1-phthisis, consumption and the white plague. Journal of Military and Veterans Health, 2014. 22(2): p. 29. Daniel, T.M., The history of tuberculosis. Respir Med, 2006. 100(11): p. 1862-70. Lawn, S.D. and A.I. Zumla, Tuberculosis. Lancet, 2011. 378(9785): p. 57-72. Sepkowitz, K.A., Tuberculosis and the health care worker: a historical perspective. Ann Intern Med, 1994. 120(1): p. 71-9. Makam, P., et al., 2-(2-Hydrazinyl) thiazole derivatives: Design, synthesis and in vitro antimycobacterial studies. European journal of medicinal chemistry, 2013. 69: p. 564576. Zaman, K., Tuberculosis: a global health problem. Journal of health, population and nutrition, 2010: p. 111-113. Porter, J. and J. Ogden, Ethics of directly observed therapy for the control of infectious diseases. Bulletin de l'Institut Pasteur, 1997. 95(3): p. 117-127. Organization, W.H., Tuberculosis fact sheet no 104, revised August 2002. Geneva, Switzerland, 2002. Olendorf, D., C. Jeryan, and K. Boyden, The Gale Encyclopedia of Medicine: GM. Vol. 3. 1999: Gale Cengage.

46

References

24.

25. 26. 27. 28.

29. 30. 31. 32. 33.

34. 35.

36.

37. 38.

39.

40. 41. 42. 43. 44.

Harding, C.V. and W.H. Boom, Regulation of antigen presentation by Mycobacterium tuberculosis: a role for Toll-like receptors. Nature Reviews Microbiology, 2010. 8(4): p. 296-307. Ramakrishnan, L., Revisiting the role of the granuloma in tuberculosis. Nature Reviews Immunology, 2012. 12(5): p. 352-366. Ndlovu, H. and M.J. Marakalala, Granulomas and inflammation: Host-Directed Therapies for Tuberculosis. Frontiers in immunology, 2016. 7. Vogt, P.K. and M.J. Mahan, Bacterial infection: close encounters at the host pathogen interface. Vol. 225. 2012: Springer Science & Business Media. Reece, S.T. and S.H. Kaufmann, Floating between the poles of pathology and protection: can we pin down the granuloma in tuberculosis? Current opinion in microbiology, 2012. 15(1): p. 63-70. Ulrichs, T. and S.H. Kaufmann, New insights into the function of granulomas in human tuberculosis. The Journal of pathology, 2006. 208(2): p. 261-269. O'Garra, A., et al., The immune response in tuberculosis. Annual review of immunology, 2013. 31: p. 475-527. Ducati, R.G., et al., The resumption of consumption: a review on tuberculosis. Memórias do Instituto Oswaldo Cruz, 2006. 101(7): p. 697-714. Kumar, S.V., et al., A systematic review of different type of tuberculosis. Eur Rev Med Pharmacol Sci, 2010. 14(10): p. 831-843. Reece, S.T., et al., Skin test performed with highly purified Mycobacterium tuberculosis recombinant protein triggers tuberculin shock in infected guinea pigs. Infection and immunity, 2005. 73(6): p. 3301-3306. Crowle, A.J., Immunizing constituents of the tubercle bacillus. Bacteriological reviews, 1958. 22(3): p. 183. Burl, S., et al., The tuberculin skin test (TST) is affected by recent BCG vaccination but not by exposure to non-tuberculosis mycobacteria (NTM) during early life. PloS one, 2010. 5(8): p. e12287. Nienhaus, A., A. Schablon, and R. Diel, Interferon-gamma release assay for the diagnosis of latent TB infection–analysis of discordant results, when compared to the tuberculin skin test. PLoS One, 2008. 3(7): p. e2665. Connell, T.G., et al., A three-way comparison of tuberculin skin testing, QuantiFERONTB gold and T-SPOT. TB in children. PloS one, 2008. 3(7): p. e2624. De Keyser, E., F. De Keyser, and F. De Baets, Tuberculin skin test versus interferongamma release assays for the diagnosis of tuberculosis infection. Acta clinica Belgica, 2014. 69(5): p. 358-366. Siddiqi, K., M.-L. Lambert, and J. Walley, Clinical diagnosis of smear-negative pulmonary tuberculosis in low-income countries: the current evidence. The Lancet infectious diseases, 2003. 3(5): p. 288-296. Gaensler, E.A., The Surgery for Pulmonary Tuberculosis 1, 2. American Review of Respiratory Disease, 1982. 125(3P2): p. 73-84. Riva, M.A., From milk to rifampicin and back again: history of failures and successes in the treatment for tuberculosis. The Journal of antibiotics, 2014. 67(9): p. 661-665. Davies, J. and D. Davies, Origins and evolution of antibiotic resistance. Microbiology and Molecular Biology Reviews, 2010. 74(3): p. 417-433. Marshall, G., et al., Streptomycin treatment of pulmonary tuberculosis: a Medical Research Council investigation. BMJ, 1948. 2(4582): p. 769-782. FOX, W. and I. Sutherland, A Five-Year Assessment of Patients in a Controlled Trial of Streptomycin, Para-Aminosalicylic Acid, and Streptomycin plus Para-Aminosalicyiic Acid, in Pulmonary Tuberculosis. Report to the Tuberculosis Chemotherapy Trials 47

References

45. 46. 47. 48. 49.

50. 51. 52.

53. 54.

55.

56. 57. 58. 59. 60.

61. 62. 63.

64.

Committee of the Medical Research Council. Quarterly Journal of Medicine, 1956. 25(98): p. 221-43. Fox, W., Various combinations of isoniazid with streptomycin or with PAS in the treatment of pulmonary tuberculosis. Group, 1955. 1: p. 1. TBFACTS.org. TB Treatment – Curing TB, failure, relapse & recurrence. 2016 18-102016]; Available from: http://www.tbfacts.org/tb-treatment/. Olaru, I.D., et al., Novel drugs against tuberculosis: a clinician's perspective. European Respiratory Journal, 2015. 45(4): p. 1119-1131. Migliori, G., et al., First tuberculosis cases in Italy resistant to all tested drugs. Euro surveill, 2007. 12(5): p. E070517. Velayati, A.A., et al., Emergence of new forms of totally drug-resistant tuberculosis bacilli: super extensively drug-resistant tuberculosis or totally drug-resistant strains in Iran. Chest Journal, 2009. 136(2): p. 420-425. Coetzee, G., et al., Emergence and spread of extensively and totally drug-resistant tuberculosis, South Africa. 2015. Udwadia, Z.F., Totally drug resistant-tuberculosis in India: The bad just got worse. The Journal of Association of Chest Physicians, 2016. 4(2): p. 41. Caminero, J., Multidrug-resistant tuberculosis: epidemiology, risk factors and case finding [State of the art series. Drug-resistant tuberculosis. Edited by CY. Chiang. Number 4 in the series]. The International Journal of Tuberculosis and Lung Disease, 2010. 14(4): p. 382-390. Akl, M. and A. Mahalli, Drug resistant tuberculosis: Risk factors and resourcesutilization at a chest disease clinic, Alexandria, Egypt. J Am Sci, 2012. 8: p. 16-22. Farley, J.E., et al., Outcomes of multi-drug resistant tuberculosis (MDR-TB) among a cohort of South African patients with high HIV prevalence. PloS one, 2011. 6(7): p. e20436. Baghaei, P., et al., Risk factors associated with multidrug-resistant tuberculosis. TANAFFOS-Journal of Respiratory Disease, Thoracic Surgery, Intensive Care and Tuberculosis, 2009. 8(3): p. 17-21. Jagielski, T., et al., Current methods in the molecular typing of Mycobacterium tuberculosis and other mycobacteria. BioMed research international, 2014. 2014. Vasconcellos, S.E., et al., Distinct genotypic profiles of the two major clades of Mycobacterium africanum. BMC Infect Dis, 2010. 10: p. 80. Kieser, K.J. and E.J. Rubin, How sisters grow apart: mycobacterial growth and division. Nature Reviews Microbiology, 2014. 12(8): p. 550-562. Ahmad, S., Pathogenesis, immunology, and diagnosis of latent Mycobacterium tuberculosis infection. Clinical and Developmental Immunology, 2010. 2011. Houben, E.N., K.V. Korotkov, and W. Bitter, Take five—Type VII secretion systems of Mycobacteria. Biochimica et Biophysica Acta (BBA)-Molecular Cell Research, 2014. 1843(8): p. 1707-1716. Philips, J.A. and J.D. Ernst, Tuberculosis pathogenesis and immunity. Annual Review of Pathology: Mechanisms of Disease, 2012. 7: p. 353-384. Andersen, P. and I. Heron, Specificity of a protective memory immune response against Mycobacterium tuberculosis. Infection and immunity, 1993. 61(3): p. 844-851. Sørensen, A.L., et al., Purification and characterization of a low-molecular-mass T-cell antigen secreted by Mycobacterium tuberculosis. Infection and immunity, 1995. 63(5): p. 1710-1717. Pallen, M.J., The ESAT-6/WXG100 superfamily–and a new Gram-positive secretion system? Trends in microbiology, 2002. 10(5): p. 209-212. 48

References

65.

66.

67. 68.

69. 70.

71. 72. 73. 74.

75.

76. 77. 78. 79.

80. 81. 82.

83.

84.

Uplekar, S., et al., Comparative genomics of Esx genes from clinical isolates of Mycobacterium tuberculosis provides evidence for gene conversion and epitope variation. Infection and immunity, 2011. 79(10): p. 4042-4049. Daleke, M.H., et al., General secretion signal for the mycobacterial type VII secretion pathway. Proceedings of the National Academy of Sciences, 2012. 109(28): p. 1134211347. McLaughlin, B., et al., A mycobacterium ESX-1–secreted virulence factor with unique requirements for export. PLoS Pathog, 2007. 3(8): p. e105. Xu, J., et al., A unique Mycobacterium ESX‐1 protein co‐secretes with CFP‐10/ESAT‐6 and is necessary for inhibiting phagosome maturation. Molecular microbiology, 2007. 66(3): p. 787-800. Carlsson, F., et al., Polar localization of virulence-related Esx-1 secretion in mycobacteria. PLoS Pathog, 2009. 5(1): p. e1000285. DiGiuseppe Champion, P.A., et al., ESX‐1 secreted virulence factors are recognized by multiple cytosolic AAA ATPases in pathogenic mycobacteria. Molecular microbiology, 2009. 73(5): p. 950-962. Houben, E.N., et al., Composition of the type VII secretion system membrane complex. Molecular microbiology, 2012. 86(2): p. 472-484. Champion, P.A.D., et al., C-terminal signal sequence promotes virulence factor secretion in Mycobacterium tuberculosis. Science, 2006. 313(5793): p. 1632-1636. Rosenberg, O.S., et al., Substrates control multimerization and activation of the multidomain ATPase motor of type VII secretion. Cell, 2015. 161(3): p. 501-512. de Jonge, M.I., et al., ESAT-6 from Mycobacterium tuberculosis dissociates from its putative chaperone CFP-10 under acidic conditions and exhibits membrane-lysing activity. Journal of bacteriology, 2007. 189(16): p. 6028-6034. Das, C., T.S. Ghosh, and S.S. Mande, Computational analysis of the ESX-1 region of Mycobacterium tuberculosis: insights into the mechanism of type VII secretion system. PLoS One, 2011. 6(11): p. e27980. Lou, Y., et al., EspC forms a filamentous structure in the cell envelope of Mycobacterium tuberculosis and impacts ESX‐1 secretion. Molecular Microbiology, 2016. Solomonson, M., et al., Structure of EspB from the ESX-1 Type VII secretion system and insights into its export mechanism. Structure, 2015. 23(3): p. 571-583. Ates, L.S., et al., Essential role of the ESX-5 secretion system in outer membrane permeability of pathogenic mycobacteria. PLoS Genet, 2015. 11(5): p. e1005190. Tufariello, J.M., et al., Separable roles for Mycobacterium tuberculosis ESX-3 effectors in iron acquisition and virulence. Proceedings of the National Academy of Sciences, 2016. 113(3): p. E348-E357. Chen, J.M., et al., EspD is critical for the virulence-mediating ESX-1 secretion system in Mycobacterium tuberculosis. Journal of bacteriology, 2012. 194(4): p. 884-893. MacGurn, J.A., et al., A non‐RD1 gene cluster is required for Snm secretion in Mycobacterium tuberculosis. Molecular microbiology, 2005. 57(6): p. 1653-1663. Ekiert, D.C. and J.S. Cox, Structure of a PE–PPE–EspG complex from Mycobacterium tuberculosis reveals molecular specificity of ESX protein secretion. Proceedings of the National Academy of Sciences, 2014. 111(41): p. 14758-14763. Bottai, D., et al., ESAT-6 secretion-independent impact of ESX-1 genes espF and espG1 on virulence of Mycobacterium tuberculosis. Journal of Infectious Diseases, 2011. 203(8): p. 1155-1164. Raghavan, S., et al., Secreted transcription factor controls Mycobacterium tuberculosis virulence. Nature, 2008. 454(7205): p. 717-721. 49

References

85.

86.

87.

88. 89. 90. 91.

92.

93. 94. 95. 96. 97. 98.

99. 100. 101.

102. 103.

104. 105.

Bahk, Y.Y., et al., Antigens secreted from Mycobacterium tuberculosis: identification by proteomics approach and test for diagnostic marker. Proteomics, 2004. 4(11): p. 32993307. Sia, J.K., M. Georgieva, and J. Rengarajan, Innate immune defenses in human tuberculosis: an overview of the interactions between Mycobacterium tuberculosis and innate immune cells. Journal of immunology research, 2015. 2015. Rothchild, A., et al., Determining transcriptional networks regulating the response of alveolar macrophages to Mycobacterium tuberculosis. The Journal of Immunology, 2016. 196(1 Supplement): p. 65.6-65.6. Akira, S., S. Uematsu, and O. Takeuchi, Pathogen recognition and innate immunity. Cell, 2006. 124(4): p. 783-801. Ernst, J.D., The immunological life cycle of tuberculosis. Nature Reviews Immunology, 2012. 12(8): p. 581-591. Ehrt, S. and D. Schnappinger, Mycobacterial survival strategies in the phagosome: defence against host stresses. Cellular microbiology, 2009. 11(8): p. 1170-1178. Condos, R., et al., Local immune responses correlate with presentation and outcome in tuberculosis. American journal of respiratory and critical care medicine, 1998. 157(3): p. 729-735. Lasco, T.M., et al., Rapid accumulation of eosinophils in lung lesions in guinea pigs infected with Mycobacterium tuberculosis. Infection and immunity, 2004. 72(2): p. 11471149. Eum, S.-Y., et al., Neutrophils are the predominant infected phagocytic cells in the airways of patients with active pulmonary TB. CHEST Journal, 2010. 137(1): p. 122-128. Appelberg, R., Neutrophils and intracellular pathogens: beyond phagocytosis and killing. Trends in microbiology, 2007. 15(2): p. 87-92. Segal, A.W., How neutrophils kill microbes. Annual review of immunology, 2005. 23: p. 197. Culley, F.J., Natural killer cells in infection and inflammation of the lung. Immunology, 2009. 128(2): p. 151-163. Vivier, E., et al., Functions of natural killer cells. Nature immunology, 2008. 9(5): p. 503-510. Esin, S., et al., Interaction of Mycobacterium tuberculosis Cell Wall Components with the Human Natural Killer Cell Receptors NKp44 and Toll‐Like Receptor 2. Scandinavian journal of immunology, 2013. 77(6): p. 460-469. Allen, M., et al., Mechanisms of control of Mycobacterium tuberculosis by NK cells: role of glutathione. Frontiers in immunology, 2015. 6. Lu, C.-C., et al., NK cells kill mycobacteria directly by releasing perforin and granulysin. Journal of leukocyte biology, 2014. 96(6): p. 1119-1129. Sasindran, S.J. and J.B. Torrelles, Mycobacterium tuberculosis infection and inflammation: what is beneficial for the host and for the bacterium? Frontiers in microbiology, 2011. 2: p. 2. Ndlovu, H. and M.J. Marakalala, Granulomas and inflammation: Host-Directed Therapies for Tuberculosis. Frontiers in Immunology, 2016. 7: p. 434. Betts, J.C., et al., Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Molecular microbiology, 2002. 43(3): p. 717-731. Barry, C.E., et al., The spectrum of latent tuberculosis: rethinking the biology and intervention strategies. Nature Reviews Microbiology, 2009. 7(12): p. 845-855. Davis, J.M. and L. Ramakrishnan, The role of the granuloma in expansion and dissemination of early tuberculous infection. Cell, 2009. 136(1): p. 37-49. 50

References

106. 107. 108. 109. 110. 111.

112.

113.

114.

115.

116. 117. 118. 119. 120.

121.

122. 123. 124.

125.

Cambier, C., S. Falkow, and L. Ramakrishnan, Host evasion and exploitation schemes of Mycobacterium tuberculosis. Cell, 2014. 159(7): p. 1497-1509. Fogel, N., Tuberculosis: A disease without boundaries. Tuberculosis, 2015. 95(5): p. 527-531. Couper, K.N., D.G. Blount, and E.M. Riley, IL-10: the master regulator of immunity to infection. The Journal of Immunology, 2008. 180(9): p. 5771-5777. Ehlers, S. and U.E. Schaible, The granuloma in tuberculosis: dynamics of a host– pathogen collusion. Frontiers in immunology, 2013. 3: p. 411. Russell, D.G., Who puts the tubercle in tuberculosis? Nature Reviews Microbiology, 2007. 5(1): p. 39-47. Pichugin, A.V., et al., Dominant role of the sst1 locus in pathogenesis of necrotizing lung granulomas during chronic tuberculosis infection and reactivation in genetically resistant hosts. The American journal of pathology, 2009. 174(6): p. 2190-2201. Reece, S.T., et al., Serine protease activity contributes to control of Mycobacterium tuberculosis in hypoxic lung granulomas in mice. The Journal of clinical investigation, 2010. 120(9): p. 3365-3376. Heuts, F., et al., CD4+ cell-dependent granuloma formation in humanized mice infected with mycobacteria. Proceedings of the National Academy of Sciences, 2013. 110(16): p. 6482-6487. Clay, H., H.E. Volkman, and L. Ramakrishnan, Tumor necrosis factor signaling mediates resistance to mycobacteria by inhibiting bacterial growth and macrophage death. Immunity, 2008. 29(2): p. 283-294. Lenaerts, A., C.E. Barry, and V. Dartois, Heterogeneity in tuberculosis pathology, microenvironments and therapeutic responses. Immunological reviews, 2015. 264(1): p. 288-307. Nunes-Alves, C., et al., In search of a new paradigm for protective immunity to TB. Nature Reviews Microbiology, 2014. 12(4): p. 289-299. Dheda, K., et al., The immunology of tuberculosis: from bench to bedside. Respirology, 2010. 15(3): p. 433-450. Cooper, A.M., Cell mediated immune responses in tuberculosis. Annual review of immunology, 2009. 27: p. 393. Korbel, D.S., B.E. Schneider, and U.E. Schaible, Innate immunity in tuberculosis: myths and truth. Microbes and Infection, 2008. 10(9): p. 995-1004. Wolf, A.J., et al., Mycobacterium tuberculosis infects dendritic cells with high frequency and impairs their function in vivo. The Journal of Immunology, 2007. 179(4): p. 25092519. Madan-Lala, R., et al., Mycobacterium tuberculosis impairs dendritic cell functions through the serine hydrolase Hip1. The Journal of Immunology, 2014. 192(9): p. 42634272. Srivastava, S. and J.D. Ernst, Cell-to-cell transfer of M. tuberculosis antigens optimizes CD4 T cell priming. Cell host & microbe, 2014. 15(6): p. 741-752. Bozzano, F., F. Marras, and A. De Maria, Immunology of tuberculosis. Mediterranean journal of hematology and infectious diseases, 2014. 6(1): p. 2014027. Chackerian, A.A., et al., Dissemination of Mycobacterium tuberculosis is influenced by host factors and precedes the initiation of T-cell immunity. Infection and immunity, 2002. 70(8): p. 4501-4509. Ho, A.W., et al., Lung CD103+ dendritic cells efficiently transport influenza virus to the lymph node and load viral antigen onto MHC class I for presentation to CD8 T cells. The Journal of Immunology, 2011. 187(11): p. 6011-6021. 51

References

126.

127.

128. 129. 130.

131.

132.

133.

134. 135. 136. 137.

138.

139.

140.

141.

142. 143. 144.

Reiley, W.W., et al., ESAT-6-specific CD4 T cell responses to aerosol Mycobacterium tuberculosis infection are initiated in the mediastinal lymph nodes. Proceedings of the National Academy of Sciences, 2008. 105(31): p. 10961-10966. Wolf, A.J., et al., Initiation of the adaptive immune response to Mycobacterium tuberculosis depends on antigen production in the local lymph node, not the lungs. The Journal of experimental medicine, 2008. 205(1): p. 105-115. Kaufmann, S.H., G. Hussey, and P.-H. Lambert, New vaccines for tuberculosis. The Lancet, 2010. 375(9731): p. 2110-2119. Lyadova, I. and A. Panteleev, Th1 and Th17 cells in tuberculosis: protection, pathology, and biomarkers. Mediators of inflammation, 2015. 2015. Caruso, A.M., et al., Mice deficient in CD4 T cells have only transiently diminished levels of IFN-γ, yet succumb to tuberculosis. The Journal of Immunology, 1999. 162(9): p. 5407-5416. Ladel, C.H., S. Daugelat, and S.H. Kaufmann, Immune response to Mycobacterium bovis bacille Calmette Guerin infection in major histocompatibility complex class I‐and II‐ deficient knock‐out mice: contribution of CD4 and CD8 T cells to acquired resistance. European journal of immunology, 1995. 25(2): p. 377-384. Orme, I. and F.M. Collins, Protection against Mycobacterium tuberculosis infection by adoptive immunotherapy. Requirement for T cell-deficient recipients. The Journal of experimental medicine, 1983. 158(1): p. 74-83. Flynn, J.L., et al., An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection. The Journal of experimental medicine, 1993. 178(6): p. 22492254. Cooper, A.M., et al., Disseminated tuberculosis in interferon gamma gene-disrupted mice. The Journal of experimental medicine, 1993. 178(6): p. 2243-2247. Seneviratne, S.L., et al., Disseminated Mycobacterium tuberculosis infection due to interferon γ deficiency. Response to replacement therapy. Thorax, 2007. 62(1): p. 97-99. Doherty, T.M. and P. Andersen, Vaccines for tuberculosis: novel concepts and recent progress. Clinical microbiology reviews, 2005. 18(4): p. 687-702. Bafica, A., et al., TLR9 regulates Th1 responses and cooperates with TLR2 in mediating optimal resistance to Mycobacterium tuberculosis. The Journal of experimental medicine, 2005. 202(12): p. 1715-1724. Saito, S. and M. Nakano, Nitric oxide production by peritoneal macrophages of Mycobacterium bovis BCG-infected or non-infected mice: regulatory role of T lymphocytes and cytokines. Journal of leukocyte biology, 1996. 59(6): p. 908-915. Chan, J., et al., Killing of virulent Mycobacterium tuberculosis by reactive nitrogen intermediates produced by activated murine macrophages. The Journal of experimental medicine, 1992. 175(4): p. 1111-1122. Green, A.M., R. DiFazio, and J.L. Flynn, IFN-γ from CD4 T cells is essential for host survival and enhances CD8 T cell function during Mycobacterium tuberculosis infection. The Journal of Immunology, 2013. 190(1): p. 270-277. Gallegos, A.M., et al., A gamma interferon independent mechanism of CD4 T cell mediated control of M. tuberculosis infection in vivo. PLoS Pathog, 2011. 7(5): p. e1002052. Mayer‐Barber, K.D. and A. Sher, Cytokine and lipid mediator networks in tuberculosis. Immunological reviews, 2015. 264(1): p. 264-275. Altare, F., et al., Interleukin-12 receptor β1 deficiency in a patient with abdominal tuberculosis. Journal of Infectious Diseases, 2001. 184(2): p. 231-236. Casanova, J.-L. and L. Abel, Genetic dissection of immunity to mycobacteria: the human model. Annual review of immunology, 2002. 20(1): p. 581-620. 52

References

145. 146. 147. 148.

149.

150.

151.

152.

153.

154.

155.

156. 157.

158.

159.

160. 161.

162.

Harris, J. and J. Keane, How tumour necrosis factor blockers interfere with tuberculosis immunity. Clinical & Experimental Immunology, 2010. 161(1): p. 1-9. Keane, J., et al., Tuberculosis associated with infliximab, a tumor necrosis factor α– neutralizing agent. New England Journal of Medicine, 2001. 345(15): p. 1098-1104. Seder, R.A., P.A. Darrah, and M. Roederer, T-cell quality in memory and protection: implications for vaccine design. Nature Reviews Immunology, 2008. 8(4): p. 247-258. Lindenstrøm, T., et al., Tuberculosis subunit vaccination provides long-term protective immunity characterized by multifunctional CD4 memory T cells. The Journal of Immunology, 2009. 182(12): p. 8047-8055. Knudsen, N.P.H., et al., Different human vaccine adjuvants promote distinct antigenindependent immunological signatures tailored to different pathogens. Scientific reports, 2016. 6. Maggioli, M.F., et al., Increased TNF-α/IFN-γ/IL-2 and Decreased TNF-α/IFN-γ Production by Central Memory T Cells Are Associated with Protective Responses against Bovine Tuberculosis Following BCG Vaccination. Frontiers in Immunology, 2016. 7. Harari, A., et al., Dominant TNF-{alpha} Mycobacterium Tuberculosis-specific CD4 Tcell responses discriminate between latent infection and active disease. The Journal of Immunology, 2011. 186(Meeting Abstracts 1): p. 99.10. Sakai, S., K.D. Mayer-Barber, and D.L. Barber, Defining features of protective CD4 T cell responses to Mycobacterium tuberculosis. Current opinion in immunology, 2014. 29: p. 137-142. Reiley, W.W., et al., Distinct functions of antigen-specific CD4 T cells during murine Mycobacterium tuberculosis infection. Proceedings of the National Academy of Sciences, 2010. 107(45): p. 19408-19413. Lindenstrøm, T., et al., Control of chronic Mycobacterium tuberculosis infection by CD4 KLRG1− IL-2–secreting central memory cells. The Journal of Immunology, 2013. 190(12): p. 6311-6319. Langrish, C.L., et al., IL-23 drives a pathogenic T cell population that induces autoimmune inflammation. The Journal of experimental medicine, 2005. 201(2): p. 233240. Lin, Y., et al., Interleukin-17 is required for T helper 1 cell immunity and host resistance to the intracellular pathogen Francisella tularensis. Immunity, 2009. 31(5): p. 799-810. Bai, H., et al., IL-17/Th17 promotes type 1 T cell immunity against pulmonary intracellular bacterial infection through modulating dendritic cell function. The journal of immunology, 2009. 183(9): p. 5886-5895. Umemura, M., et al., IL-17-mediated regulation of innate and acquired immune response against pulmonary Mycobacterium bovis bacille Calmette-Guerin infection. The Journal of Immunology, 2007. 178(6): p. 3786-3796. Gopal, R., et al., IL‐23‐dependent IL‐17 drives Th1‐cell responses following Mycobacterium bovis BCG vaccination. European journal of immunology, 2012. 42(2): p. 364-373. Gopal, R., et al., Unexpected role for IL-17 in protective immunity against hypervirulent Mycobacterium tuberculosis HN878 infection. PLoS Pathog, 2014. 10(5): p. e1004099. Umemura, M., et al., Involvement of IL‐17A‐producing TCR γδ T cells in late protective immunity against pulmonary Mycobacterium tuberculosis infection. Immunity, Inflammation and Disease, 2016. Orme, I.M., R.T. Robinson, and A.M. Cooper, The balance between protective and pathogenic immune responses in the TB-infected lung. Nature immunology, 2015. 16(1): p. 57-63. 53

References

163. 164.

165. 166. 167. 168. 169.

170. 171.

172.

173.

174. 175.

176.

177. 178. 179. 180.

181.

182.

Blanco, F.C., et al., Increased IL-17 expression is associated with pathology in a bovine model of tuberculosis. Tuberculosis, 2011. 91(1): p. 57-63. Richert-Spuhler, L.E. and J.M. Lund, Chapter Ten-The Immune Fulcrum: Regulatory T Cells Tip the Balance Between Pro-and Anti-inflammatory Outcomes upon Infection. Progress in molecular biology and translational science, 2015. 136: p. 217-243. Larson, R.P., S. Shafiani, and K.B. Urdahl, Foxp3+ regulatory T cells in tuberculosis, in The New Paradigm of Immunity to Tuberculosis. 2013, Springer. p. 165-180. Cvetanovich, G.L. and D.A. Hafler, Human regulatory T cells in autoimmune diseases. Current opinion in immunology, 2010. 22(6): p. 753-760. Kursar, M., et al., Cutting Edge: Regulatory T cells prevent efficient clearance of Mycobacterium tuberculosis. The Journal of Immunology, 2007. 178(5): p. 2661-2665. Scott-Browne, J.P., et al., Expansion and function of Foxp3-expressing T regulatory cells during tuberculosis. The Journal of experimental medicine, 2007. 204(9): p. 2159-2169. Quinn, K.M., et al., Inactivation of CD4+ CD25+ regulatory T cells during early mycobacterial infection increases cytokine production but does not affect pathogen load. Immunology and cell biology, 2006. 84(5): p. 467-474. Hsu, P., et al., IL-10 potentiates differentiation of human induced regulatory T cells via STAT3 and Foxo1. The Journal of Immunology, 2015. 195(8): p. 3665-3674. Shafiani, S., et al., Pathogen-specific regulatory T cells delay the arrival of effector T cells in the lung during early tuberculosis. The Journal of experimental medicine, 2010. 207(7): p. 1409-1420. Guyot-Revol, V., et al., Regulatory T cells are expanded in blood and disease sites in patients with tuberculosis. American journal of respiratory and critical care medicine, 2006. 173(7): p. 803-810. Chen, X., et al., CD4+ CD25+ FoxP3+ regulatory T cells suppress Mycobacterium tuberculosis immunity in patients with active disease. Clinical immunology, 2007. 123(1): p. 50-59. Jacobo, R.G., et al., Analysis of Th1, Th17 and regulatory T cells in tuberculosis case contacts. Cellular immunology, 2014. 289(1): p. 167-173. Welsh, K.J., et al., Immunopathology of postprimary tuberculosis: increased Tregulatory cells and DEC-205-positive foamy macrophages in cavitary lesions. Clinical and Developmental Immunology, 2010. 2011. Sharma, P.K., et al., FoxP3+ regulatory T cells suppress effector T-cell function at pathologic site in miliary tuberculosis. American journal of respiratory and critical care medicine, 2009. 179(11): p. 1061-1070. Luo, J., et al., Imbalance of Th17 and Treg in peripheral blood mononuclear cells of active tuberculosis patients. The Brazilian Journal of Infectious Diseases, 2016. Winau, F., et al., Apoptotic vesicles crossprime CD8 T cells and protect against tuberculosis. Immunity, 2006. 24(1): p. 105-117. van der Wel, N., et al., M. tuberculosis and M. leprae translocate from the phagolysosome to the cytosol in myeloid cells. Cell, 2007. 129(7): p. 1287-1298. van Pinxteren, L.A., et al., Control of latent Mycobacterium tuberculosis infection is dependent on CD8 T cells. European journal of immunology, 2000. 30(12): p. 36893698. Turner, J., et al., CD8-and CD95/95L-dependent mechanisms of resistance in mice with chronic pulmonary tuberculosis. American journal of respiratory cell and molecular biology, 2001. 24(2): p. 203-209. Caccamo, N., et al., Analysis of Mycobacterium tuberculosis-specific CD8 T-cells in patients with active tuberculosis and in individuals with latent infection. PloS one, 2009. 4(5): p. e5528. 54

References

183. 184.

185.

186.

187. 188.

189.

190.

191.

192.

193. 194.

195.

196.

197.

198.

199.

200.

Pollock, K.M., et al., T-cell immunophenotyping distinguishes active from latent tuberculosis. Journal of Infectious Diseases, 2013. 208(6): p. 952-968. Woodworth, J.S., Y. Wu, and S.M. Behar, Mycobacterium tuberculosis-specific CD8+ T cells require perforin to kill target cells and provide protection in vivo. The Journal of Immunology, 2008. 181(12): p. 8595-8603. Glatman-Freedman, A. and A. Casadevall, Serum therapy for tuberculosis revisited: reappraisal of the role of antibody-mediated immunity againstMycobacterium tuberculosis. Clinical microbiology reviews, 1998. 11(3): p. 514-532. Achkar, J.M., J. Chan, and A. Casadevall, Role of B cells and antibodies in acquired immunity against Mycobacterium tuberculosis. Cold Spring Harbor perspectives in medicine, 2014: p. a018432. Balu, S., et al., A novel human IgA monoclonal antibody protects against tuberculosis. The Journal of Immunology, 2011. 186(5): p. 3113-3119. López, Y., et al., Induction of a protective response with an IgA monoclonal antibody against Mycobacterium tuberculosis 16kDa protein in a model of progressive pulmonary infection. International Journal of Medical Microbiology, 2009. 299(6): p. 447-452. Glatman-Freedman, A., et al., Antigenic evidence of prevalence and diversity of Mycobacterium tuberculosis arabinomannan. Journal of clinical microbiology, 2004. 42(7): p. 3225-3231. Maglione, P.J., et al., Fcγ receptors regulate immune activation and susceptibility during Mycobacterium tuberculosis infection. The Journal of Immunology, 2008. 180(5): p. 3329-3338. Rodríguez, A., et al., Role of IgA in the defense against respiratory infections: IgA deficient mice exhibited increased susceptibility to intranasal infection with Mycobacterium bovis BCG. Vaccine, 2005. 23(20): p. 2565-2572. Kunnath-Velayudhan, S., et al., Proteome-scale antibody responses and outcome of Mycobacterium tuberculosis infection in nonhuman primates and in tuberculosis patients. Journal of Infectious Diseases, 2012. 206(5): p. 697-705. Dayal, R., et al., Serological diagnosis of tuberculosis. The Indian Journal of Pediatrics, 2008. 75(12): p. 1219-1221. Gupta, S., R. Bhatia, and K. Datta, Serological diagnosis of childhood tuberculosis by estimation of mycobacterial antigen 60-specific immunoglobulins in the serum. Tubercle and Lung Disease, 1997. 78(1): p. 21-27. Hsu, T., et al., The primary mechanism of attenuation of bacillus Calmette–Guerin is a loss of secreted lytic function required for invasion of lung interstitial tissue. Proceedings of the National Academy of Sciences, 2003. 100(21): p. 12420-12425. Gao, L.Y., et al., A mycobacterial virulence gene cluster extending RD1 is required for cytolysis, bacterial spreading and ESAT‐6 secretion. Molecular microbiology, 2004. 53(6): p. 1677-1693. Koo, I.C., et al., ESX‐1‐dependent cytolysis in lysosome secretion and inflammasome activation during mycobacterial infection. Cellular microbiology, 2008. 10(9): p. 18661878. Francis, R., R. Butler, and G. Stewart, Mycobacterium tuberculosis ESAT-6 is a leukocidin causing Ca2+ influx, necrosis and neutrophil extracellular trap formation. Cell death & disease, 2014. 5(10): p. e1474. Danelishvili, L., J. Everman, and L.E. Bermudez, Mycobacterium tuberculosis PPE68 and Rv2626c genes contribute to the host cell necrosis and bacterial escape from macrophages. Virulence, 2016. 7(1): p. 23-32. Li, H., et al., Mycobacterium tuberculosis PE13 (Rv1195) manipulates the host cell fate via p38-ERK-NF-κB axis and apoptosis. Apoptosis, 2016: p. 1-14. 55

References

201. 202. 203. 204.

205. 206. 207.

208.

209.

210.

211.

212.

213.

214.

215. 216.

217. 218. 219.

Martin, C.J., et al., Efferocytosis is an innate antibacterial mechanism. Cell host & microbe, 2012. 12(3): p. 289-300. Volkman, H.E., et al., Tuberculous granuloma induction via interaction of a bacterial secreted protein with host epithelium. Science, 2010. 327(5964): p. 466-469. Brilha, S., et al., ESAT-6 Drives MMP-10 Gene Expression and Secretion in Tuberculosis. American Journal of Respiratory Cell and Molecular Biology, 2016(ja). MacGurn, J.A. and J.S. Cox, A genetic screen for Mycobacterium tuberculosis mutants defective for phagosome maturation arrest identifies components of the ESX-1 secretion system. Infect Immun, 2007. 75(6): p. 2668-78. Tan, T., et al., The ESAT-6/CFP-10 secretion system of Mycobacterium marinum modulates phagosome maturation. Cell Microbiol, 2006. 8(9): p. 1417-29. Portal-Celhay, C., et al., Mycobacterium tuberculosis EsxH inhibits ESCRT-dependent CD4+ T-cell activation. Nat Microbiol, 2016. 2: p. 16232. Wassermann, R., et al., Mycobacterium tuberculosis differentially activates cGAS-and inflammasome-dependent intracellular immune responses through ESX-1. Cell host & microbe, 2015. 17(6): p. 799-810. Stanley, S.A., et al., The Type I IFN response to infection with Mycobacterium tuberculosis requires ESX-1-mediated secretion and contributes to pathogenesis. J Immunol, 2007. 178(5): p. 3143-52. Pathak, S.K., et al., Direct extracellular interaction between the early secreted antigen ESAT-6 of Mycobacterium tuberculosis and TLR2 inhibits TLR signaling in macrophages. Nat Immunol, 2007. 8(6): p. 610-8. Reiley, W.W., et al., ESAT-6-specific CD4 T cell responses to aerosol Mycobacterium tuberculosis infection are initiated in the mediastinal lymph nodes. Proc Natl Acad Sci U S A, 2008. 105(31): p. 10961-6. Wolf, A.J., et al., Initiation of the adaptive immune response to Mycobacterium tuberculosis depends on antigen production in the local lymph node, not the lungs. J Exp Med, 2008. 205(1): p. 105-15. Tyagi, P., et al., Mycobacterium tuberculosis has diminished capacity to counteract redox stress induced by elevated levels of endogenous superoxide. Free Radical Biology and Medicine, 2015. 84: p. 344-354. Yuan, Y., et al., Identification of a gene involved in the biosynthesis of cyclopropanated mycolic acids in Mycobacterium tuberculosis. Proceedings of the National Academy of Sciences, 1995. 92(14): p. 6630-6634. Beamer, G.L. and J. Turner, Murine models of susceptibility to tuberculosis. ARCHIVUM IMMUNOLOGIAE ET THERAPIAE EXPERIMENTALIS-ENGLISH EDITION-, 2005. 53(6): p. 469. Clark, S., Y. Hall, and A. Williams, Animal models of tuberculosis: guinea pigs. Cold Spring Harbor perspectives in medicine, 2015. 5(5): p. a018572. Dannenberg, A. and F. Collins, Progressive pulmonary tuberculosis is not due to increasing numbers of viable bacilli in rabbits, mice and guinea pigs, but is due to a continuous host response to mycobacterial products. Tuberculosis, 2001. 81(3): p. 229242. Scanga, C.A. and J.L. Flynn, Modeling tuberculosis in nonhuman primates. Cold Spring Harbor perspectives in medicine, 2014: p. a018564. Waters, W., et al., Tuberculosis immunity: opportunities from studies with cattle. Clinical and Developmental Immunology, 2010. 2011. Ramakrishnan, L. The zebrafish guide to tuberculosis immunity and treatment. in Cold Spring Harbor symposia on quantitative biology. 2013. Cold Spring Harbor Laboratory Press. 56

References

220. 221. 222. 223.

224. 225. 226. 227.

228.

229. 230. 231. 232. 233. 234.

235.

236.

237.

238.

239.

Dong, H., et al., Establishment of rat model of silicotuberculosis and its pathological characteristic. Pathogens and global health, 2014. McCallan, L., et al., A New experimental infection model in ferrets based on aerosolised Mycobacterium bovis. Veterinary medicine international, 2011. 2011. Gil, O., et al., Granuloma encapsulation is a key factor for containing tuberculosis infection in minipigs. PLoS One, 2010. 5(4): p. e10030. Dionne, M.S., N. Ghori, and D.S. Schneider, Drosophila melanogaster is a genetically tractable model host for Mycobacterium marinum. Infection and immunity, 2003. 71(6): p. 3540-3550. Hagedorn, M., et al., Infection by tubercular mycobacteria is spread by nonlytic ejection from their amoeba hosts. Science, 2009. 323(5922): p. 1729-1733. Gupta, U. and V. Katoch, Animal models of tuberculosis. Tuberculosis, 2005. 85(5): p. 277-293. Andersen, P., Tuberculosis vaccines—an update. Nature Reviews Microbiology, 2007. 5(7): p. 484-487. Behrends, J., et al., IL-22 is mainly produced by IFNγ-secreting cells but is dispensable for host protection against Mycobacterium tuberculosis infection. PloS one, 2013. 8(2): p. e57379. Kramnik, I. and G. Beamer. Mouse models of human TB pathology: roles in the analysis of necrosis and the development of host-directed therapies. in Seminars in immunopathology. 2016. Springer. Bharath, S. and V. Balasubramanian, A color atlas of comparative pathology of pulmonary tuberculosis. 2011. Orme, I.M., The mouse as a useful model of tuberculosis. Tuberculosis, 2003. 83(1): p. 112-115. Aagaard, C., et al., A multistage tuberculosis vaccine that confers efficient protection before and after exposure. Nature medicine, 2011. 17(2): p. 189-194. Hoang, T., et al., ESAT-6 (EsxA) and TB10. 4 (EsxH) based vaccines for pre-and postexposure tuberculosis vaccination. PLoS One, 2013. 8(12): p. e80579. Flynn, J.L., et al., Effects of aminoguanidine on latent murine tuberculosis. The Journal of Immunology, 1998. 160(4): p. 1796-1803. Pym, A.S., et al., Loss of RD1 contributed to the attenuation of the live tuberculosis vaccines Mycobacterium bovis BCG and Mycobacterium microti. Molecular microbiology, 2002. 46(3): p. 709-717. Dye, C., Making wider use of the world's most widely used vaccine: Bacille Calmette– Guérin revaccination reconsidered. Journal of the Royal Society Interface, 2013. 10(87): p. 20130365. Zodpey, S., et al., Effectiveness of Bacillus Calmette Guerin (BCG) vaccination in the prevention of childhood pulmonary tuberculosis: a case control study in Nagpur, India. 1998. Sterne, J., L. Rodrigues, and I. Guedes, Does the efficacy of BCG decline with time since vaccination? The international journal of tuberculosis and lung disease, 1998. 2(3): p. 200-207. Lanckriet, C., et al., Efficacy of BCG vaccination of the newborn: evaluation by a followup study of contacts in Bangui. International journal of epidemiology, 1995. 24(5): p. 1042-1049. Andersen, P., Vaccine strategies against latent tuberculosis infection. Trends in microbiology, 2007. 15(1): p. 7-13.

57

References

240.

241. 242. 243.

244. 245.

246. 247.

248.

249.

250.

251.

252. 253. 254.

255. 256.

257.

258.

Abubakar, I., et al., Systematic review and meta-analysis of the current evidence on the duration of protection by bacillus Calmette–Guerin vaccination against tuberculosis. 2013. von Reyn, C.F., BCG, latitude, and environmental mycobacteria. Clinical Infectious Diseases, 2014. 59(4): p. 607-608. Roy, A., et al., Effect of BCG vaccination against Mycobacterium tuberculosis infection in children: systematic review and meta-analysis. 2014. Trunz, B.B., P. Fine, and C. Dye, Effect of BCG vaccination on childhood tuberculous meningitis and miliary tuberculosis worldwide: a meta-analysis and assessment of costeffectiveness. The Lancet, 2006. 367(9517): p. 1173-1180. Andersen, P. and T.M. Doherty, The success and failure of BCG—implications for a novel tuberculosis vaccine. Nature Reviews Microbiology, 2005. 3(8): p. 656-662. Fletcher, H.A. and L. Schrager, TB vaccine development and the End TB Strategy: importance and current status. Transactions of The Royal Society of Tropical Medicine and Hygiene, 2016. 110(4): p. 212-218. Kaufmann, S.H., et al., Novel approaches to TB vaccine development. International Journal of Infectious Diseases, 2016. 27: p. 30Z. Knudsen, N.P.H., et al., Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostics. Proceedings of the National Academy of Sciences, 2014. 111(3): p. 1096-1101. Bertholet, S., et al., A defined tuberculosis vaccine candidate boosts BCG and protects against multidrug-resistant Mycobacterium tuberculosis. Science translational medicine, 2010. 2(53): p. 53ra74-53ra74. Louise Vinther Skjøt, E.M.A., Peter Andersen, Rikke, Antigen discovery and tuberculosis vaccine development in the post-genomic era. Scandinavian journal of infectious diseases, 2001. 33(1): p. 79-83. Arlehamn, C.S.L., et al., Memory T cells in latent Mycobacterium tuberculosis infection are directed against three antigenic islands and largely contained in a CXCR3+ CCR6+ Th1 subset. PLoS Pathog, 2013. 9(1): p. e1003130. Coppola, M., et al., New Genome-Wide Algorithm Identifies Novel In-Vivo Expressed Mycobacterium Tuberculosis Antigens Inducing Human T-Cell Responses with Classical and Unconventional Cytokine Profiles. Sci Rep, 2016. 6: p. 37793. Reed, S.G., M.T. Orr, and C.B. Fox, Key roles of adjuvants in modern vaccines. Nature medicine, 2013. 19(12): p. 1597-1608. Ligon, L.S., J.D. Hayden, and M. Braunstein, The ins and outs of Mycobacterium tuberculosis protein export. Tuberculosis, 2012. 92(2): p. 121-132. Ilghari, D., et al., Solution Structure of the Mycobacterium tuberculosis EsxG· EsxH Complex FUNCTIONAL IMPLICATIONS AND COMPARISONS WITH OTHER M. TUBERCULOSIS Esx FAMILY COMPLEXES. Journal of Biological Chemistry, 2011. 286(34): p. 29993-30002. Hoang, T., et al., Differential effect of ESAT-6 (EsxA) and TB10.4 (EsxH) based vaccines for pre- and post-exposure vaccination. Submitted for Plos Pathogen, 2013. Lindenstrom, T., et al., Tuberculosis subunit vaccination provides long-term protective immunity characterized by multifunctional CD4 memory T cells. J Immunol, 2009. 182(12): p. 8047-55. Callahan, B., et al., Conservation of structure and protein-protein interactions mediated by the secreted mycobacterial proteins EsxA, EsxB, and EspA. J Bacteriol, 2010. 192(1): p. 326-35. Lazarevic, V., L.H. Glimcher, and G.M. Lord, T-bet: a bridge between innate and adaptive immunity. Nature Reviews Immunology, 2013. 13(11): p. 777-789. 58

References

259. 260. 261. 262. 263.

264. 265. 266.

267.

268. 269.

270.

271.

272.

273.

274. 275. 276. 277. 278.

Hutloff, A., et al., ICOS is an inducible T-cell co-stimulator structurally and functionally related to CD28. Nature, 1999. 402: p. 21-24. Groom, J.R. and A.D. Luster, CXCR3 in T cell function. Experimental cell research, 2011. 317(5): p. 620-631. Jayaraman, P., et al., TIM3 mediates T cell exhaustion during Mycobacterium tuberculosis infection. PLoS Pathog, 2016. 12(3): p. e1005490. Anderson, K.G., et al., Intravascular staining for discrimination of vascular and tissue leukocytes. Nature protocols, 2014. 9(1): p. 209-222. Billeskov, R., et al., Comparing adjuvanted h28 and modified vaccinia virus ankara expressing h28 in a mouse and a non-human primate tuberculosis model. PloS one, 2013. 8(8): p. e72185. Aagaard, C., et al., A multistage tuberculosis vaccine that confers efficient protection before and after exposure. Nat Med, 2011. 17(2): p. 189-94. Lindenstrom, T., et al., Control of chronic mycobacterium tuberculosis infection by CD4 KLRG1- IL-2-secreting central memory cells. J Immunol, 2013. 190(12): p. 6311-9. Woodworth, J., et al., Subunit vaccine H56/CAF01 induces a population of circulating CD4 T cells that traffic into the Mycobacterium tuberculosis-infected lung. Mucosal Immunology, 2016. Tameris, M.D., et al., Safety and efficacy of MVA85A, a new tuberculosis vaccine, in infants previously vaccinated with BCG: a randomised, placebo-controlled phase 2b trial. The Lancet, 2013. 381(9871): p. 1021-1028. Shah, S. and V. Briken, Modular Organization of the ESX-5 Secretion System in Mycobacterium tuberculosis. Frontiers in cellular and infection microbiology, 2016. 6. Ates, L.S., et al., The ESX-5 system of pathogenic mycobacteria is involved in capsule integrity and virulence through its substrate PPE10. PLoS Pathog, 2016. 12(6): p. e1005696. Lin, P.L., et al., The multistage vaccine H56 boosts the effects of BCG to protect cynomolgus macaques against active tuberculosis and reactivation of latent Mycobacterium tuberculosis infection. The Journal of clinical investigation, 2012. 122(1): p. 303-314. Manca, C., et al., Mycobacterium tuberculosis CDC1551 induces a more vigorous host response in vivo and in vitro, but is not more virulent than other clinical isolates. The journal of immunology, 1999. 162(11): p. 6740-6746. Sreevatsan, S., et al., Restricted structural gene polymorphism in the Mycobacterium tuberculosis complex indicates evolutionarily recent global dissemination. Proceedings of the National Academy of Sciences, 1997. 94(18): p. 9869-9874. Manca, C., et al., Virulence of a Mycobacterium tuberculosis clinical isolate in mice is determined by failure to induce Th1 type immunity and is associated with induction of IFN-α/β. Proceedings of the National Academy of Sciences, 2001. 98(10): p. 5752-5757. Kinhikar, A.G., et al., Potential role for ESAT6 in dissemination of M. tuberculosis via human lung epithelial cells. Molecular microbiology, 2010. 75(1): p. 92-106. Behr, M.A. and W.R. Waters, Is tuberculosis a lymphatic disease with a pulmonary portal? The Lancet Infectious diseases, 2014. 14(3): p. 250-255. Winslow, G.M., et al., Early T‐cell responses in tuberculosis immunity. Immunological reviews, 2008. 225(1): p. 284-299. Ma, B., et al., Emerging human papillomavirus vaccines. Expert opinion on emerging drugs, 2012. 17(4): p. 469-492. Comas, I., et al., Human T cell epitopes of Mycobacterium tuberculosis are evolutionarily hyperconserved. Nature genetics, 2010. 42(6): p. 498-503. 59

References

279.

280. 281.

282.

283. 284.

285. 286.

van Dissel, J.T., et al., A novel liposomal adjuvant system, CAF01, promotes long-lived Mycobacterium tuberculosis-specific T-cell responses in human. Vaccine, 2014. 32(52): p. 7098-107. Christensen, D., et al., Cationic liposomes as vaccine adjuvants. Expert review of vaccines, 2007. 6(5): p. 785-796. Cauwelaert, N.D., et al., The TLR4 Agonist Vaccine Adjuvant, GLA-SE, Requires Canonical and Atypical Mechanisms of Action for TH 1 Induction. PloS one, 2016. 11(1): p. e0146372. Olafsdottir, T., et al., IC31®, a Two‐Component Novel Adjuvant Mixed with a Conjugate Vaccine Enhances Protective Immunity against Pneumococcal Disease in Neonatal Mice. Scandinavian journal of immunology, 2009. 69(3): p. 194-202. Didierlaurent, A.M., et al., Adjuvant system AS01: helping to overcome the challenges of modern vaccines. Expert Review of Vaccines, 2017. 16(1): p. 55-63. Schoenen, H., et al., Cutting edge: Mincle is essential for recognition and adjuvanticity of the mycobacterial cord factor and its synthetic analog trehalose-dibehenate. J Immunol, 2010. 184(6): p. 2756-60. Darrah, P.A., et al., Multifunctional TH1 cells define a correlate of vaccine-mediated protection against Leishmania major. Nature medicine, 2007. 13(7): p. 843-850. Schmidt, C.W., Diversity outbred: a new generation of mouse model. 2015, National Institute of Environmental Health Science.

60

Appendix I

Appendix I Manuscript I

Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostics

Niels Peter H. Knudsen Sara Nørskov-Lauritsen Gregory M Dolganov Gary K Schoolnik Thomas Lindenstrøm Peter Andersen Else Marie Agger Claus Aagaard

Published in Proceedings of the National Academy of Sciences 111 (3), 1096-1101, 2014.

61

Appendix I

Tuberculosis vaccine with high predicted population coverage and compatibility with modern diagnostics Niels Peter H. Knudsena, Sara Nørskov-Lauritsena,1, Gregory M. Dolganovb, Gary K. Schoolnikb, Thomas Lindenstrøma, Peter Andersena, Else Marie Aggera,2, and Claus Aagaarda,2 a Department of Infectious Disease Immunology, Statens Serum Institut, DK 2300 Copenhagen, Denmark; and bDepartment of Microbiology and Immunology, Stanford University School of Medicine, CA 94304

Edited by Barry R. Bloom, Harvard School of Public Health, Boston, MA, and approved December 4, 2013 (received for review August 7, 2013)

A central goal in vaccine research is the identification of relevant antigens. The Mycobacterium tuberculosis chromosome encodes 23 early secretory antigenic target (ESAT-6) family members that mostly are localized as gene pairs. In proximity to five of the gene pairs are ESX secretion systems involved in the secretion of the ESAT-6 family proteins. Here, we performed a detailed and systematic investigation of the vaccine potential of five possible Esx dimer substrates, one for each of the five ESX systems. On the basis of gene transcription during infection, immunogenicity, and protective capacity in a mouse aerosol challenge model, we identified the ESX dimer substrates EsxD-EsxC, ExsG-EsxH, and ExsW-EsxV as the most promising vaccine candidates and combined them in a fusion protein, H65. Vaccination with H65 gave protection at the level of bacillus Calmette–Guérin, and the fusion protein exhibited high predicted population coverage in high endemic regions. H65 thus constitutes a promising vaccine candidate devoid of antigen 85 and fully compatible with current ESAT-6 and culture filtrate protein 10-based diagnostics. T-cell immunology

| gene expression | tuberculosis

D

espite the availability of a number of antibiotics and the extensive use of the live vaccine bacillus Calmette–Guérin (BCG), tuberculosis (TB) remains a major global health problem, with an estimated incidence rate of 9 million new cases per year (1). BCG vaccination protects against severe progressive TB in children but is not able to prevent reactivation of pulmonary disease in adult life (2). As a consequence, there is a need for a vaccine to supplement or replace BCG. There are currently several TB vaccines in clinical trials, which include attenuated Mycobacterium tuberculosis (M.tb) strains, recombinant BCG strains, viral vectored vaccines, and proteinbased subunit vaccines administered together with an adjuvant (3). Apart from the whole-cell vaccines, most are based on the same very limited number of M.tb antigens (Ags), with Ag85A and Ag85B in particular being present in many candidate vaccines (4). Recently, a TB efficacy study based on a recombinant strain of modified vaccinia Ankara virus expressing Ag85A (MVA85A) was unblinded. MVA85A was given to infants as a BCG booster but did not statistically improve the protective efficacy of the bacillus. In vivo gene expression data have further questioned the use of Ag85 proteins as vaccine targets, as the expression of all Ag85 genes after an initial peak during the first weeks of infection are dramatically reduced; in mice, this correlates with the onset of the adaptive immune response (5). Adoptive transfer studies of Ag85B-specific T cells have shown that during the early phase of infection, ∼10% of the transferred T cells produced IFN-γ in vivo, whereas this percentage declined to EsxW-EsxV. No IFN-γ release could be detected after restimulation with either EsxD-EsxC or a mixture of the linker sequences. In the adjuvant control group, there was no detectable response toward any of the proteins/peptides. The vaccinated groups of animals were challenged with M.tb, and after 6 wk of infection, their lungs were analyzed for vaccine-induced protection. All vaccinated animals had reduced bacterial burdens compared with the adjuvant control group (Fig. 3B). H65 vaccination resulted in a CFU reduction of ∼0.85 log10 (P < 0.01) relative to the adjuvant control, whereas BCG vaccination gave a 1.39 log10 (P < 0.001) reduction. The level of protection obtained with H65 was identical to the protection obtained after vaccination with the H56 fusion protein (Ag85B-ESAT6-Rv2660c). As a snapshot of the in vivo immune activity at the site of infection, the IFN-γ levels were measured in lung homogenates from individual animals (Fig. S3). The highest IFN-γ level was found in the adjuvant control group, and only the H65-vaccinated group had a statistically lower level (P < 0.05). Because the recognition of the Esx dimer proteins during infection is broader (Fig. 2C) and their protective efficacy closer to 1098 | www.pnas.org/cgi/doi/10.1073/pnas.1314973111

lated from perfused lungs and stimulated in vitro with either of the three vaccine-containing ESX dimers or EsxB-EsxA to determine the frequency of cytokine-producing Esx dimer-specific CD4+ T cells (Fig. 4A). In both the H65-vaccinated and adjuvant control groups, there was a strong recruitment of cytokineproducing Esx dimer-specific CD4+ T cells to the lungs. The strong recruitment supports the theory that all four dimers were expressed during infection and confirmed their immunogenicity. In addition to responses against the vaccine proteins EsxD-EsxC, EsxG-EsxH, and EsxW-EsxV, the infection also induced a strong response against EsxB-EsxA, particularly in the CAF01 vaccinated group (Fig. 4A). In a more detailed analysis, we compared the differentiation status of the Esx-specific CD4+ T cells not only between but also within the vaccination groups. In the infected adjuvant control animals, we found primarily IFNγ-producing CD4+ T cells that were of an effector phenotype (IFN-γ+, TNF-α+, or IFN-γ+) or, to a minor degree, memory T cells (IL-2+, TNF+, IFN-γ+), regardless of their Esx protein specificity (Fig. 4B). However, in H65-vaccinated/infected animals, the T-cell functionality depended on whether the cells were vaccine-specific or not. The functionality of the nonvaccinespecific T cells (recognizing EsxB-EsxA) were, despite the lower CFU counts in these animals, practically identical to the effector and memory T-cell populations found in the infected adjuvant control animals. In contrast, the vaccine-specific CD4+ T-cell population was more diverse in terms of functionality. It consisted preferentially of memory cells (IL-2+, TNF+, IFN-γ+), a solid frequency of effector cells (IFN-γ+, TNF-α+, or IFN-γ+), and a smaller but distinct fraction of less-differentiated CD4+ T cells that were IL-2+, TNF+ double-positive or IL-2+/TNF+ single-positive. The H65 Fusion Protein Has Broad Human Coverage. Despite proven

success in animal studies, the large host genetic variability on a global scale could compromise the potential efficacy of the vaccine in humans. We therefore used in silico epitope-binding predictions to estimate the population coverage of the H65 vaccine and the individual protein components in countries with a high burden of TB. Because CD4+ T-cell-mediated immunity is essential to combat M.tb infection and class 2 HLA proteins are responsible for stimulating CD4+ T cells, we examined the class 2 HLA-DRB1 diversity (22).The DRB1 allele was chosen because DR alleles bind the vast majority of known M.tb epitopes (23), and among the DR alleles, the DRB1 surface expression is five times greater than for other DR alleles (24). Finally, epitope prediction programs for DRB1 alleles are more frequently available than for other class 2 HLA alleles. Binding predictions for epitopes in the H65 fusion protein were generated for 34 HLA-DBR1 alleles, representing the three most common HLADRB1 alleles in the 22 countries that, according to the World Health Organization, have the highest burden of TB (25). The predictions were generated using four methods: CombLib, SMM-align, NN-align, and/or NetMHCIIpan. Wherever possible, we used a consensus method approach based on two or Knudsen et al.

64

Fig. 2. Expression dynamic and Ag recognition during infection. (A) The relative expression of the genes encoding the five secretion dimers and Ag85A, Ag85B, and Ag85C was measured in M.tb (strain Erdman) in vitro cultures (0 d) and in lungs of infected CB6F1 mice (n = 4 per time). The relative expression for each of the 5 dimer operons was calculated by [(RGCNgene1+ RGCNgene2)/∑(RGCNtotal)]*100, and for the fbp genes by [(RGCNfbpA + RGCNfbpB + RGCNfbpC)/ ∑(RGCNtotal)]*100. RGCN = RNA gene copy number. RGCNtotal is the sum of the 13 genes investigated (10 esx and 3 fbp genes). For more information, see Table S2. We found no detectable expression of esxU-T under the conditions tested. The Esx and Ag85B specificity of T cells recruited to the lung was measured in nonvaccinated CB6F1 (B) or B6C3F1 (C) mice 21 d after infection with M.tb (strain Erdman) (n = 3 per mice strain).

three of the methods, but for many of the alleles, only the NetMHCIIpan method was available (Table S3). The median number of vaccine epitopes predicted to bind any of DRB1 allele ranged from 7 to 67 for the H65 fusion protein and from 0 to 14 for the individual proteins, clearly supporting the importance of including multiple proteins in vaccines. Among the individual proteins, the highly expressed, ESX-3-secreted EsxG holds the largest total number of predicted human CD4 T-cell epitopes for the 34 alleles, followed by its secretion partner EsxH and the much lower expressed EsxD from the ESX-2 system. The other substrate for the ESX-2 system, EsxC, holds the lowest number of predicted epitopes among the single proteins, even less than what was predicted for the two 20-amino acid long linkers combined. The combination of low expression and relatively few human epitopes may suggest that in a potential future improvement of the H65 vaccine, EsxC could be replaced with another Ag. Discussion The mycosyltransferases Ag85A and Ag85B are frequently used Ags in TB vaccines currently in clinical trials (3). Both are primarily expressed during the initial stage of infection, where M.tb divides rapidly and thus has a need for synthesis and assembly of cell wall components. As the growth rate decreases, there is less need for mycolyl synthesis and the expression of Ag85 proteins is reduced to a low level. Thus, this study is a step toward identifying vaccine targets that have a more constitutive expression profile and then testing the protective efficacy of a TB vaccine on the basis of a multiple of these targets. The reported density of B- and T-cell epitopes in mycobacterial proteins is at least five to six times higher for extracellular proteins than for cytoplasmic, membrane, or cell wall proteins (23). It is therefore logical to look for new vaccine targets among the secreted proteins. Protein secretion is essential for all bacteria to interact with their environment, and mycobacteria have, during coevolution with their host cell, acquired specialized protein secretory pathways that deliver effector proteins to the host cell. These include the ESAT-6 secretion system (ESX) that secretes proteins lacking a classical signal sequence (26). The genome of M.tb encodes five ESX regions (ESX-1 to ESX-5) Knudsen et al.

arranged in conserved clusters. The ESX systems secrete high quantities of proteins with different biological functions, and as we know several of the ESX substrates are strong vaccine Ags, we speculate that a combination of substrates from different ESX systems could generate an efficient vaccine. On the basis of phylogenetic relationships, we identified the most likely Esx dimer substrate for each of the five ESX systems. From our protection and recognition studies using the Esx dimer substrates, it is clear that the substrates for four of the ESX systems are protective. Only the ESX-4 substrate EsxU-EsxT had no protective efficacy against a M.tb challenge, nor was it recognized during infection. This could be because of the use of inbred mice strains and lack of T-cell epitopes, but vaccination with EsxU-EsxT did induce strong T-cell responses, suggesting that there are indeed potential T-cell epitopes represented in this protein. Gene expression studies up until 20 wk of infection showed no expression of the esxT or esxU genes, suggesting that the absence of protection 6 wk after infection is a result of the lack of expression of EsxT and EsxU. The ESX-2 substrate EsxD-EsxC is constitutively expressed at a low but seemingly sufficient level for protection against M.tb, even though there was no detectable T-cell response in the lungs after 6 wk of infection. EsxD and EsxC have neither been identified in the culture supernatants of in vitro cultures nor do they contain the ESX signature tag, and it is therefore questionable whether the proteins are actually being secreted by the ESX-2 system. Within the first 3 wk of infection, esxA/B and esxG/H were the strongest expressed esx genes. However, as a consequence of the encoded proteins’ biological functions, the relative expression of the esxA/B and esxG/H genes changes radically during this period. The ESX-1 substrates EsxB and EsxA are virulence factors that are used for escaping the phagosome and possibly facilitating cell-to-cell spread during the stages of infection in which

Fig. 3. H65-vaccine-specific responses reduce bacteria load. (A) Splenocytes were isolated from H65-vaccinated animals (n = 3) 3 wk after vaccination and stimulated in vitro with single proteins, after which secreted IFN-γ was measured in the supernatants. Linkers were a 1:1:1 mix of the three sequences included in the fusion (Fig. S2B). Bacterial load in groups of vaccinated CB6F1 mice (B) were measured in the lungs of individual mice 6 wk after challenge with M.tb (strain Erdman); means ± SEM are indicated (n = 6 per group). **P < 0.01, ***P < 0.001, one-way ANOVA, Tukeys multiple comparison test. The frequency of H65-specific CD4+ T cells producing IFN-γ, TNF-α, or IL-2 was measured in splenocytes by flow cytometry 3 wk after immunization with CAF01 (gray bars) or H65 (black bars) in B6C3F1 mice after in vitro stimulation with single proteins (C). The bacterial burden in vaccinated B6C3F1 mice (D) was measured by enumerating the bacteria in the lung of individual animals 6 wk after challenge with M.tb (strain Erdman) (n = 6 per group). *P < 0.05, one-way ANOVA, Tukeys multiple comparison test. For both mice strains, the tests have been done twice with similar results.

PNAS | January 21, 2014 | vol. 111 | no. 3 | 1099

65

MEDICAL SCIENCES

Appendix I

Appendix I

Fig. 4. Vaccine-primed T cells maintains a memory phenotype. (A) The frequency of dimer-specific CD4+ T cells producing IFN-γ, TNF-α, or IL-2 after in vitro stimulation with dimer substrates. Cells were isolated from perfused lungs 6 wk after M.tb (strain Erdman) infection from B6C3F1 mice immunized with CAF01 adjuvant alone or in combination with H65 (n = 3 per group). (B) The polyfunctionality of H65 vaccine epitopes (EsxD-EsxC + EsxGEsxH + EsxW-EsxV) was compared with nonvaccine epitopes (EsxB-EsxA) in infected (labeled CAF01) and H65 vaccinated/infected groups of animals (labeled H65) by multicolor flow cytometry. Only cytokine expression combinations with frequencies above 0.05% were included.

M.tb is multiplying (e.g., during the early infection or reactivation from latency). Both EsxB and EsxA are strongly recognized in infected individuals and are therefore exploited in TB diagnostic tests. The ESX-3 substrates EsxG and EsxH are described to be involved in the acquisition of divalent metal ions such as iron and zinc (27). When grown in vitro in rich medium, there are plenty of metal ions available in an accessible form. Thus, there is little need for these proteins, and the gene expression is relatively low. Under normal physiological conditions, the concentration of free iron in body fluids is very low (10−18 M). As a consequence, intracellular mycobacteria are competing with the host for iron and zinc and to sequester metals for survival M.tb strongly upregulates the expression of EsxG and EsxH. In M.tb-infected guinea pigs, iron has been shown to accumulate within the primary lesions, but most of it accumulates as extracellular ferric iron, and it is unknown whether M.tb can exploit this source (28). ESX-5 is the most recently evolved ESX cluster and is only present in the group of slow-growing pathogenic species of mycobacteria (29). ESX-5 is responsible for the secretion of several proteins involved in virulence/pathogenicity (26) and is a major modulator of the host immune response and a key virulence determinant of M.tb. The inactivation of ESX-5 results in severe attenuation of the mutant strains, which are unable to replicate even in immune-deficient mice (30). To target a broad range of relevant epitopes and maximize the likelihood of one or more of the vaccine targets being expressed at any time during an infection, we included the ESX-2, ESX-3, and ESX-5 substrates (EsxC, EsxD, EsxG, EsxH, EsxV, and EsxW) in a six-protein fusion (H65) and evaluated its protective efficacy in two different inbred mice strains. Although fulfilling all criteria, EsxA and EsxB were excluded because of their common use as Ags in diagnostic tests. In CB6F1 mice, H65 vaccination induced an EsxG-EsxH- and EsxW-EsxV-specific immune response but did not protect to the same degree as Mycobacterium bovis BCG 6 wk after infection. Reflecting the reduced bacteria load, the overall in vivo IFN-γ level was significantly reduced in lungs of H65-vaccinated animals. Reduction of T-cell activation is likely to extend the time to exhaustion, a problem associated with chronic pathogens such as M.tb (31). In B6C3F1 mice, there is a broader vaccine specific immune response, and the protective efficacy of H65 vaccination was comparable to the protection observed with BCG. The H65 vaccinated and adjuvant control animals from this study were used for multiparameter flow-cytometry analysis to compare the quality of the T-cell response at the site of infection. This analysis demonstrates that the CD4 T cells maintained at the site of infection in H65-vaccinated animals has a larger fraction of polyfunctional CD4 T cells compared with the T cells that infiltrate the lungs of nonvaccinated animals. It has previously been shown that in mice vaccinated with BCG, the T cells 1100 | www.pnas.org/cgi/doi/10.1073/pnas.1314973111

are primarily effector or effector-memory T cells 6 mo after vaccination (32). In contrast we have shown that 1 y after vaccination with a fusion protein (H1) formulated in the same adjuvant used in this study, the T cells are primarily central memory T cells (33). In HIV-infected individuals, polyfunctional CD4 T cells are a characteristic feature observed in HIV controllers, and an inverse correlation has been shown with viral load, whereas noncontrollers elicit responses dominated by IFN-γ single-positive CD4 T cells (34). However, in this study, it was not clear whether the high degree of polyfunctional CD4 T cells caused improved control of HIV or whether the improved quality of the T-cell response was an effect of a lower viral load. We have the possibility to address this question by simultaneously monitoring the quality of the T-cell responses to Ags that are present in the H65 vaccine and comparing these to responses to Ags that are absent from the vaccine but promoted by the infection. EsxBEsxA serves as a marker of a purely infection-promoted response, whereas the Esx molecules in the H65 vaccine are markers for a vaccine-primed and infection-expanded response. In H65-vaccinated animals, there is a striking difference in the quality of the EsxB-EsxA- and vaccine-specific T cells at the site of infection that is not seen in nonvaccinated animals. Because the quality of the EsxB-EsxA-specific response in the H65-vaccinated animals is identical to the EsxA-EsxB response in the adjuvant control group, it demonstrates that the polyfunctional quality of the response is maintained selectively for the vaccine-promoted part of the mycobacteria-specific response. Thus, T-cell polyfunctionality is not a consequence of efficient bacterial containment but, rather, a vaccine-related phenomenon. In mice, vaccine-induced multifunctional CD4 T cells have been shown to be superior to their single-positive counterparts in terms of protection against HIV, M.tb, and Leishmania major infections (34–36). The improved effector/memory CD4 T-cell balance in H65-vaccinated animals may have implications for their long-term ability to suppress the infection; this is currently under investigation (37). Although successful in mice, we did find variations in the protective efficacy of H65 in the two mice strains tested. This underlines the importance of the genetic background and that there is no guarantee that the protective efficacy for any vaccine tested in animals is transferable to humans, given the large genetic variability worldwide. From the literature, it is clear that the generation of a strong CD4+ T-cell response is crucial for protection against M.tb (13). We therefore assessed the potential effect of host genetic diversity on the protective coverage of the H65 vaccine, using open-source epitope binding prediction programs evaluating the binding of vaccine epitopes to class 2 HLA (DRB1) alleles. Binding prediction is a cost-effective method for assessment of TB vaccine candidates that can give information regarding a worldwide use that could not be obtained even from several clinical trials. We focused on epitope-binding predictions for high-frequency HLA alleles among TB high-burden populations, as these are the groups in which a vaccine is most needed. The H65 vaccine contains many predicted epitopes for the tested alleles and is therefore likely to be broadly recognized in the populations with the highest burden of TB. The lowest number of predicted H65 vaccine epitopes was seven for the alleles tested, which is clearly above the cutoff of four epitopes that has previously been used to identify alleles of concern (38). The presence of human T-cell epitopes in a TB vaccine is a prerequisite for human recognition, but they also represent an interesting caveat. Comparative analyses of 21 M.tb genomes revealed that the known human T-cell epitopes, including epitopes in Ag85, CFP10, and ESAT-6, are evolutionary hyperconserved (39). It has been speculated that the bacteria might benefit from a strong T-cell response because it could lead to necrosis, cavitation, and escape in humans (40). The implications for vaccine design are not clear, but the presence of conserved and highly immunodominant human T-cell epitopes in vaccines deserves further attention. Among the H65 Ags, only EsxG and EsxH epitopes were included in the comparative study, and EsxH was specifically shown to harbor a relatively high number of Knudsen et al.

66

Appendix I

amino acid substitutions across the strains tested and was therefore not associated with this potential problem. To summarize, our study presents a recombinant subunit vaccine, based on six secreted M.tb Ags not including the popular Ag85A/B, ESAT-6, or CFP10. The H65 vaccine is capable of protecting against infection with M.tb at a level comparable with M. bovis BCG and can be used without interfering with ESAT-6- and CFP10-based diagnostics. HLA allele binding predictions suggest that it contains ample CD4 epitopes for worldwide coverage.

Cytokine Secretion Assays. Peripheral blood mononuclear cells, splenocytes, or lung mononuclear cells (2 × 105 per well) were cultured in the presence of 2 μg/mL Ag at 37 °C for 72 h, after which cytokine secretion was tested using ELISA.

Materials and Methods

Expression of M.tb Genes. We determined the gene expression profile of 13 M. tb genes in in vitro cultures and during in vivo infection by real-time PCR, using total RNA isolated from lungs of individual mice.

Animals, Immunizations, and Infection. Groups of 6–8-wk-old female mice were immunized three times s.c. on days 1, 14, and 28, with CAF01 emulsified with 5 μg protein to a final volume of 200 μL. Ten weeks after the first immunization, the animals were challenged with ∼100 CFU M.tb strain Erdman per mouse.

Epitope Binding Predictions. MHC-II binding predictions were done using IEDB Analysis Resources server for 34 HLA-DBR1 alleles. Binding affinities of IC50 values less than 500 nM were selected as cutoff. For more details on material methods see SI Materials and Methods.

CFU Measurements. Lung homogenates of individual mice were plated as threefold serial dilutions on Middlebrook 7H11 Bacto agar and enumerated after 3 wk incubation at 37 °C.

ACKNOWLEDGMENTS. We thank Vivi Andersen, Merete Henriksen, and Linda Christensen for excellent technical assistance. The research leading to these results has received funding from the European Union’s Seventh Framework Programme (FP7/2007-2013) under Grant Agreement 280873 ADITEC and Grant Agreement 241745 NEWTBVAC.

1. World Health Organization (2013) Global Tuberculosis Report 2013 (World Health Organization, Geneva). Available at www.who.int/tb/publications/global_report/en/. Accessed December 17, 2013. 2. Rodrigues LC, Mangtani P, Abubakar I (2011) How does the level of BCG vaccine protection against tuberculosis fall over time? BMJ 343:d5974. 3. Kaufmann SH (2012) Tuberculosis vaccine development: Strength lies in tenacity. Trends Immunol 33(7):373–379. 4. Rowland R, McShane H (2011) Tuberculosis vaccines in clinical trials. Expert Rev Vaccines 10(5):645–658. 5. Rogerson BJ, et al. (2006) Expression levels of Mycobacterium tuberculosis antigenencoding genes versus production levels of antigen-specific T cells during stationary level lung infection in mice. Immunology 118(2):195–201. 6. Bold TD, Banaei N, Wolf AJ, Ernst JD (2011) Suboptimal activation of antigen-specific CD4+ effector cells enables persistence of M. tuberculosis in vivo. PLoS Pathog 7(5): e1002063. 7. Egen JG, et al. (2011) Intravital imaging reveals limited antigen presentation and T cell effector function in mycobacterial granulomas. Immunity 34(5):807–819. 8. Kassa D, et al. (2012) Analysis of immune responses against a wide range of Mycobacterium tuberculosis antigens in patients with active pulmonary tuberculosis. Clin Vaccine Immunol 19(12):1907–1915. 9. Andersen P, Andersen AB, Sørensen AL, Nagai S (1995) Recall of long-lived immunity to Mycobacterium tuberculosis infection in mice. J Immunol 154(7):3359–3372. 10. Elhay MJ, Oettinger T, Andersen P (1998) Delayed-type hypersensitivity responses to ESAT-6 and MPT64 from Mycobacterium tuberculosis in the guinea pig. Infect Immun 66(7):3454–3456. 11. Pollock JM, Andersen P (1997) Predominant recognition of the ESAT-6 protein in the first phase of interferon with Mycobacterium bovis in cattle. Infect Immun 65(7): 2587–2592. 12. Ravn P, et al. (1999) Human T cell responses to the ESAT-6 antigen from Mycobacterium tuberculosis. J Infect Dis 179(3):637–645. 13. Kaufmann SH, Hussey G, Lambert PH (2010) New vaccines for tuberculosis. Lancet 375(9731):2110–2119. 14. Uplekar S, Heym B, Friocourt V, Rougemont J, Cole ST (2011) Comparative genomics of Esx genes from clinical isolates of Mycobacterium tuberculosis provides evidence for gene conversion and epitope variation. Infect Immun 79(10):4042–4049. 15. Bertholet S, et al. (2008) Identification of human T cell antigens for the development of vaccines against Mycobacterium tuberculosis. J Immunol 181(11):7948–7957. 16. Bertholet S, et al. (2010) A defined tuberculosis vaccine candidate boosts BCG and protects against multidrug-resistant Mycobacterium tuberculosis. Sci Transl Med 2(53):53ra74. 17. Abdallah AM, et al. (2007) Type VII secretion—mycobacteria show the way. Nat Rev Microbiol 5(11):883–891. 18. Fortune SM, et al. (2005) Mutually dependent secretion of proteins required for mycobacterial virulence. Proc Natl Acad Sci USA 102(30):10676–10681. 19. Daleke MH, et al. (2012) Specific chaperones for the type VII protein secretion pathway. J Biol Chem 287(38):31939–31947. 20. Skjøt RL, et al. (2000) Comparative evaluation of low-molecular-mass proteins from Mycobacterium tuberculosis identifies members of the ESAT-6 family as immunodominant T-cell antigens. Infect Immun 68(1):214–220.

21. Gideon HP, et al. (2010) Hypoxia induces an immunodominant target of tuberculosis specific T cells absent from common BCG vaccines. PLoS Pathog 6(12):e1001237. 22. Hoft DF (2008) Tuberculosis vaccine development: Goals, immunological design, and evaluation. Lancet 372(9633):164–175. 23. Ali H, Zeynudin A, Mekonnen A, Abera S, Ali S (2012) Smear Posetive Pulmonary Tuberculosis (PTB) Prevalence Amongst Patients at Agaro Teaching Health Center, South West Ethiopia. Ethiop J Health Sci 22(1):71–76. 24. Batra S, et al. (2012) Childhood tuberculosis in household contacts of newly diagnosed TB patients. PLoS ONE 7(7):e40880. 25. Banerji D (2012) The World Health Organization and public health research and practice in tuberculosis in India. Int J Health Serv 42(2):341–357. 26. Daku M, Gibbs A, Heymann J (2012) Representations of MDR and XDR-TB in South African newspapers. Soc Sci Med 75(2):410–418. 27. Ilghari D, et al. (2011) Solution structure of the Mycobacterium tuberculosis EsxG·EsxH complex: Functional implications and comparisons with other M. tuberculosis Esx family complexes. J Biol Chem 286(34):29993–30002. 28. Basaraba RJ, et al. (2008) Increased expression of host iron-binding proteins precedes iron accumulation and calcification of primary lung lesions in experimental tuberculosis in the guinea pig. Tuberculosis (Edinb) 88(1):69–79. 29. Raviglione M, et al. (2012) Scaling up interventions to achieve global tuberculosis control: Progress and new developments. Lancet 379(9829):1902–1913. 30. Millet JP, et al. (2013) Factors that influence current tuberculosis epidemiology. Eur Spine J 22(Suppl 4):539–548. 31. Reiley WW, et al. (2010) Distinct functions of antigen-specific CD4 T cells during murine Mycobacterium tuberculosis infection. Proc Natl Acad Sci USA 107(45): 19408–19413. 32. Henao-Tamayo MI, et al. (2010) Phenotypic definition of effector and memory Tlymphocyte subsets in mice chronically infected with Mycobacterium tuberculosis. Clin Vaccine Immunol 17(4):618–625. 33. Lindenstrøm T, et al. (2009) Tuberculosis subunit vaccination provides long-term protective immunity characterized by multifunctional CD4 memory T cells. J Immunol 182(12):8047–8055. 34. Kannanganat S, Ibegbu C, Chennareddi L, Robinson HL, Amara RR (2007) Multiplecytokine-producing antiviral CD4 T cells are functionally superior to single-cytokineproducing cells. J Virol 81(16):8468–8476. 35. Beveridge NE, et al. (2007) Immunisation with BCG and recombinant MVA85A induces long-lasting, polyfunctional Mycobacterium tuberculosis-specific CD4+ memory T lymphocyte populations. Eur J Immunol 37(11):3089–3100. 36. Darrah PA, et al. (2007) Multifunctional TH1 cells define a correlate of vaccine-mediated protection against Leishmania major. Nat Med 13(7):843–850. 37. Seder RA, Darrah PA, Roederer M (2008) T-cell quality in memory and protection: Implications for vaccine design. Nat Rev Immunol 8(4):247–258. 38. Davila J, McNamara LA, Yang Z (2012) Comparison of the predicted population coverage of tuberculosis vaccine candidates Ag85B-ESAT-6, Ag85B-TB10.4, and Mtb72f via a bioinformatics approach. PLoS ONE 7(7):e40882. 39. Comas I, et al. (2010) Human T cell epitopes of Mycobacterium tuberculosis are evolutionarily hyperconserved. Nat Genet 42(6):498–503. 40. Orme IM (2013) A new unifying theory of the pathogenesis of tuberculosis. Tuberculosis (Edinb), 10.1016/j.tube.2013.07.004.

Knudsen et al.

PNAS | January 21, 2014 | vol. 111 | no. 3 | 1101

67

MEDICAL SCIENCES

Recombinant Proteins. The five heterodimers and the H65 fusion were recombinantly expressed in E. coli BL21 AI and purified from inclusion bodies by a three-step process.

Flow Cytometry. Splenocytes or lung mononuclear cells (2 × 106 cells per well) were stimulated in vitro in the presence of recombinant Ag (2 μg/mL) for 1 h, and subsequently incubated for 5–6 h in the presence of 10 μg/mL brefeldin A. After overnight storage at 4 °C, cells were stained with antibodies against CD4, CD44, IFN-γ, TNF-α, or IL-2 and analyzed, using a flow cytometer.

Appendix I

Supporting Information Knudsen et al. 10.1073/pnas.1314973111 Materials and Methods Animals. Six- to 8-wk-old female CB6F1 (BALB/c x C57BL/6) [H-2d/b] or B6C3F1 (C57BL/6 × C3H) [H-2b/k] mice (Harlan) were rested for 1 wk before initiation of experiments. All experiments were done according to Danish Ministry of Justice and Animal Protection Committees and in compliance with European Community Directive 86/609. Mice were housed in animal facilities at Statens Serum Institut, provided with radiation-sterilized food (Harlan) and water ad libitum, and handled in accordance with the Danish Ministry of Justice and Animal Protection Committee regulations by authorized personnel. Infected animals were housed in a biosafety level 3 facility in cages contained within laminar flow safety enclosures (Scantainer). Recombinant Proteins. All DNA constructs used in this study were codon-optimized for expression in Escherichia coli and made by chemical synthesis followed by insertion into the pJexpress 411 vector (DNA2.0). In the five heterodimers, the single Esx proteins are connected by the thrombin-cleavable 9-mer linker GLVPRGSTG, and in the H65 fusion, the three heterodimers are in addition connected by the 20-amino acid linkers LIGAHPRALNVVKFGGAAFL and LGFGAGRLRGLFTNPGSWRI from Rv1986. After transformation into E. coli BL21 AI (Invitrogen), protein expression was induced with 0.2% arabinose and the proteins purified from inclusion bodies by a three-step process, as previously described (1), resulting in a very high purity of the final products (>99%). Using the NanoOrange Protein Quantitation Kit (Invitrogen), the protein concentrations were found to be between 0.1 and 0.5 mg/mL and the yield to be between 1 and 13.5 mg purified protein from 3 L of culture. The identity of all purified proteins were confirmed by mass spectrometry analysis (matrix-assisted laser desorption/ionization– time-of-flight). Immunizations and Infections. Mice were immunized s.c. in the neck

or at the base of the tail three times at 2-wk intervals. Cationic adjuvant formulation 01 [CAF01, 75 μg DDA (dimethyldioctadecylammonium)/25 μg TDB (trehalose 6,6′-dibehenate)] was emulsified with 5 μg recombinant antigen to a final volume of 200 μL for each injection. Negative control mice received three equivalent doses of CAF01, and positive control mice received a single dose of 5 × 104 CFU bacillus Calmette–Guérin Danish 1331 (Statens Serum Institut) in the first round of immunization. Ten weeks after the first immunization, the animals were challenged with Mycobacterium tuberculosis (M.tb) strain Erdman. Using a Biaera exposure system controlled by the AeroMP aerosol management system, virulent mycobacteria suspended in PBS Tween 20 (0.05%) were aerosolized and delivered via the respiratory route at ∼100 CFU per mouse. Isolation of Cells and CFU Measurements. Blood samples from six mice were pooled within immunization groups before peripheral blood mononuclear cells were isolated by density-gradient centrifugation, using Lympholyte Mammal (Cedar-Lane Laboratories). Splenocytes and lung mononuclear cells were isolated from individual/pooled animals by forcing cells through a 70-μm nylon cell strainer (BD Pharmingen). For CFU measurements, lung homogenates were prepared in PBS Tween 80 (0.05%) from individual mice and plated at threefold serial dilutions on Middlebrook 7H11 Bacto agar. After 3 wk of incubation at 37 °C, the CFUs were enumerated.

Cytokine Secretion Assays. Peripheral blood mononuclear cells, splenocytes, or lung mononuclear cells (2 × 105 per well) were cultured in round-bottomed 96-well plates in 200 μl complete RPMI media [RPMI 1640 supplemented with 1 mM L-glutamine, 50 μM 2-mercaptoehanol, 1% pyruvate, 1% penicillinstreptomycin, 1% Hepes, and 10% (vol/vol) FCS; Gibco Invitrogen], and 2 μg of antigen at 37 °C in a humidified incubator under 95% air, 5% CO2. Culture supernatants were harvested from lymphocyte cultures after 72 h of in vitro antigen stimulation and tested in triplicates. For detection of secreted IFN-γ, 96-well Maxisorb microtiter plates (Nunc) were coated with 1 μg/ mL monoclonal rat anti-murine IFN-γ (clone R4-6A2; BD Pharmingen). Free binding sites were blocked with 2% (wt/vol) milk powder in PBS. IFN-γ was detected with a 0.1 μg/mL biotinlabeled rat anti-murine antibody (clone XMG1.2; BD Pharmingen) and 0.35 μg/mL horseradish peroxidase-conjugated streptavidin (Zymed). The enzyme reaction was developed with 3.3′, 5.5′-tetramethylbenzidine, hydrogen peroxide (TMB plus; Kementec) and stopped with 0.2 M H2SO4. rIFNγ (BD Pharmingen) was used as a standard. Plates were read at 450 nm with an ELISA-reader and analyzed with KC4 3.03 Rev 4 software (BioTek). Flow Cytometry. Splenocytes or lung mononuclear cells (2 × 106

cells per well) were stimulated in vitro in V-bottom 96-well plates at 37 °C in 200 μl complete media containing anti-CD49d (1 μg/mL) and anti-CD28 (1 μg/mL) antibodies in the presence of recombinant antigen (2 μg/mL) for 1 h, and subsequently incubated for 5–6 h in the presence of 10 μg/mL brefeldin A (Sigma-Aldrich). After overnight storage at 4 °C, cells were washed in FACS buffer (PBS containing 0.1% sodium azide and 1% FCS) and subsequently stained for 30 min at 4 °C for surface markers with mAbs, as indicated, using 1/100 dilutions of antiCD4-allophycocyanin-Cy7 (clone GK1.5) and anti-CD44-FITC (clone IM7) (all BD Pharmingen). Cells were then washed in FACS buffer, permeabilized using the Cytofix/Cytoperm kit (BD Pharmingen) according to the manufacturer’s instructions, and stained intracellularly for 30 min at 4 °C in dilutions of 1/100, using anti-IFN-γ-PE-Cy7 (clone XMG1.2; eBioscience), antiTNF-PE (MP6-XT22; BD Pharmingen), or anti-IL-2-allophycocyanin (clone JES6-5h4; BD Pharmingen) mAbs. Cells were subsequently washed with BD Perm/Wash buffer (BD Pharmingen), resuspended in FACS buffer, and analyzed using a FACSCanto flow cytometer (BD Pharmingen) and FlowJo software v.8.8.7 (Tree Star). The relative proportions of cells producing different combinations were determined using PESTLE and SPICE v.5.22 software. In Vitro and in Vivo Expression of Selected M.tb Genes. We de-

termined the gene expression profile of 13 selected M.tb genes in in vitro cultures and followed their in vivo expression in lungs of CB6F1 mice infected via the aerosol route with M.tb strain Erdman. The expression profile was determined by reversetranscribing isolated total RNA and amplifying cDNA as previously described (2), followed by quantification in individual real-time PCR reactions. The 13 genes represented Esx substrates for all five ESX secretion systems plus antigens 85A, 85B, and 85C.

Epitope-Binding Predictions. The Immune Epitope Database Analysis Resources server (3), one of the most accurate prediction servers available (4), was used to perform MHC class 2

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

1 of 6

68

Appendix I

binding predictions for H65, each of its six protein components, and the two linkers included in H65. This was done for 34 HLA-DBR1 alleles representing the high-frequency HLADRB1 alleles among tuberculosis (TB) high-burden populations (5). The Immune Epitope Database documents a large number of peptides tested for binding to various MHC class 2 allelic variants, which are used to predict binding affinities for related peptides (6). Binding affinities are given as IC50 values (half maximal inhibitory concentration) in units of nanomoles. Experimental data have classified peptides with IC50 < 1,000 nM as binders and >1,000 nM as nonbinders (7). For conservative reasons, 500 nM was selected as cutoff (7). Epitope binding predictions were generated using the prediction algorithms CombLib (6), SMM (8), NN (9), or NetMHCIIpan (10). Where possible, a consensus approach of

multiple algorithms was used to identify the predicted epitopes (Table S2). All of the algorithms provide prediction output in the form of IC50 values. The generated prediction results were compiled in Excel 2010 (Microsoft). The various algorithms often generated multiple positive hits of various lengths, centered on the same binding core. In cases like this, the smallest core fragment was identified and counted only once to avoid overestimation.

1. Aagaard C, et al. (2009) Protection and polyfunctional T cells induced by Ag85BTB10.4/IC31 against Mycobacterium tuberculosis is highly dependent on the antigen dose. PLoS One 4(6):e5930. 2. Aagaard C, et al. (2011) A multistage tuberculosis vaccine that confers efficient protection before and after exposure. Nat Med 17(2):189–194. 3. Zhang Q, et al. (2008) Immune epitope database analysis resource (IEDB-AR). Nucleic Acids Res 36(Web Server issue, suppl 2):W513–W518. 4. Lin HH, Zhang GL, Tongchusak S, Reinherz EL, Brusic V (2008) Evaluation of MHC-II peptide binding prediction servers: Applications for vaccine research. BMC Bioinformatics 9 (Suppl 12):S22. 5. Davila J, McNamara LA, Yang Z (2012) Comparison of the predicted population coverage of tuberculosis vaccine candidates Ag85B-ESAT-6, Ag85B-TB10.4, and Mtb72f via a bioinformatics approach. PLoS ONE 7(7):e40882.

6. Wang P, et al. (2010) Peptide binding predictions for HLA DR, DP and DQ molecules. BMC Bioinformatics 11:568. 7. Wang P, et al. (2008) A systematic assessment of MHC class II peptide binding predictions and evaluation of a consensus approach. PLOS Comput Biol 4(4): e1000048. 8. Nielsen M, Lundegaard C, Lund O (2007) Prediction of MHC class II binding affinity using SMM-align, a novel stabilization matrix alignment method. BMC Bioinformatics 8(1):238. 9. Nielsen M, Lund O (2009) NN-align. An artificial neural network-based alignment algorithm for MHC class II peptide binding prediction. BMC Bioinformatics 10:296. 10. Nielsen M, et al. (2008) Quantitative predictions of peptide binding to any HLA-DR molecule of known sequence: NetMHCIIpan. PLOS Comput Biol 4(7):e1000107.

Statistical Analysis. Prism 5 software (GraphPad) was used for

all statistical analyses. CFU data were log-transformed before analyses. One-way ANOVA combined with Tukeys multiple comparison test was used for comparing between multiple groups. Statistical significant differences are marked by asterisks in figures and explained in the figure legends.

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

2 of 6

69

Appendix I

Fig. S1. Phylogenic relationship among substrates for the five ESX secretion systems in Mycobacterium tuberculosis. (A) Cladogram based on multiple sequence alignment of ESX substrate dimers using Clustaw W2. (B) Homology among five closely related ESX5 substrates. Lowest homology is in bold. One equals 100% identity. Total amino acid number for each substrate dimer is given in parenthesis. (C) Homologs to the esxG-H genes and possible substrates for the ESX3 secretion system. One locus encodes only one of the proteins in the dimer substrate (esxQ) with modest homology to the other substrates.

Fig. S2. Dimer fusions of Esx proteins. (A) Quality control of the purified protein dimers. Coomassie stained SDS gel. Lane 1, EsxB-EsxA; lane 2, EsxD-EsxC; lane 3, EsxG-EsxH; lane 4, EsxU-EsxT; lane 5, EsxW-EsxV. A molecular weight standard was included for size confirmation. All protein identities were confirmed by mass spectroscopy. (B) Schematic representation of H65 showing the six subunits separated by linkers. The individual proteins in the secretion dimers are linked (L, white box) with the same nine-amino acid linker (GLVPRGSTG). The three heterodimers are connected by two nonidentical 20-amino acid-long linkers (L, black box). (C) Quality control of the purified H65 fusion protein. H65 fusion protein in SDS-PAGE gels that have been stained with Coomassie blue (1) or transferred to membranes and developed using anti-His (2) or anti-Escherichia coli (2) antibodies.

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

3 of 6

70

Appendix I

Fig. S3. In vivo INF-γ in lungs of infected mice. Groups of mice were vaccinated with H65 or bacillus Calmette–Guérin or injected with CAF01 adjuvant. Six weeks after the third vaccination, all animals were infected with M.tb, and after 6 wk of infection, IFN-γ was measured by ELISA in supernatants from homogenized lungs.

Table S1. Characteristics of the 23 ESAT-6 family proteins T7S system

H37Rv*

Protein

ESX1 ESX1 ESX2 ESX2 ESX3 ESX3 ESX3 ESX3 ESX3 ESX4 ESX4 ESX5 ESX5 ESX5 ESX5 ESX5 ESX5 ESX5 ESX5 ESX5 ESX5 None None

Rv3874 Rv3875 Rv3890c Rv3891c Rv0287 Rv0288 Rv3017c Rv3019c Rv3020c Rv3444c Rv3445c Rv1037c Rv1038c Rv1197 Rv1198 Rv1792 Rv1793 Rv2346c Rv2347c Rv3619c Rv3620c Rv3904c Rv3905c

EsxB EsxA EsxC EsxD EsxG EsxH EsxQ EsxR EsxS EsxT EsxU EsxI EsxJ EsxK EsxL EsxM EsxN EsxO EsxP EsxV EsxW EsxE EsxF

Common name

kDa

pI

Motif†

Identified in culture filtrate‡

CFP10 ESAT-6 ES6_11 — TB9.8 TB10.4 TB12.9 TB10.3 PE28 — — Mtb9.9D ES6_2 ES6_3 Mtb9.9C TB11.0 Mtb9.9A Mtb9.9E ES6_7 ES6_1 ES6_10 ES6_12 ES6_13

10.8 9.9 9.9 11.2 9.8 10.4 12.9 10.3 9.8 11.1 11.4 9.8 11.0 11.0 9.9 10.9 9.9 10.0 11.0 9.8 11.0 9.6 10.5

4.31 4.19 4.17 4.43 6.51 4.35 8.06 4.13 6.68 6.34 6.68 4.48 5.02 5.02 4.83 5.02 4.56 4.56 5.02 4.48 5.02 5.45 4.47

YSRAD — No No YVAAD — No — YVAAD No No — YEQQE YEQQE — YEQQE — — YEQQE — YEQQE — YQHNE

Yes Yes No No Yes Yes No No No No No Yes Yes Yes Yes Yes Yes Yes No No Yes No No

*Proteins used in this study are in bold. † The amino acid pattern YXXXD/E has been identified as an ESX secretion marker present in one of the dimer partners, directing secretion of both proteins. Dash indicated a motif found in a predicted secretion partner. ‡ For references, see the TubercuList Webpage (http://tuberculist.epfl.ch/).

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

4 of 6

71

Appendix I

Table S2. Gene expression of selected M. tuberculosis genes Gene expression, RGCN* Gene ‡

esxW esxV‡ esxB esxA esxG esxH esxD esxC esxT esxU fbpB fbpC fbpA

H37RV identity

Protein size†

Rv3620c Rv3619c Rv3874 Rv3875 Rv0287 Rv0288 Rv3891c Rv3890c Rv3444c Rv3445c Rv1886c Rv0129c Rv3804c

98 94 100 95 97 96 107 95 100 105 325 340 338

In vivo, t = 17

In vitro

31975 39502 48802 53961 102456 274157 915779 581589 254912 25261 232947 23816 18301 22531 35725 14063 39 322 261 223 24656 1770 50348 2418 29833 12543 RGCNtotal :

RGCNgene1+ RGCNgene2 In vivo, t = 17 80777§

In vitro 93463

Percentage of total In vivo, t = 17

In vitro

4,6{

8,9

1018235

855746

58,3

81,3

487859

49077

27,9

4,7

54026

36594

3,1

3,5

300

545

0,0

0,1

104837 1746034

16731 1052156

6,0

1,6

*RNA gene copy number (RGCN) measured by RT-PCR. † Number of amino acids. ‡ Due to high sequence homology the five potential ESX-5 substrates cannot be distinguished by gene expression analysis. § Calculated as 31975 + 48802. { Calculated as (80777/1746034) × 100.

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

5 of 6

72

Appendix I

Table S3. Epitope-binding predictions for 34 HLA-DRB1 alleles common in TB high-burden populations Esx proteins DRB1

D

C

G

H

W

*01:01 *01:02 *03:01 *03:02 *04:01 *04:03 *04:04 *04:05 *04:11 *07:01 *08:01 *08:02 *08:03 *08:04 *08:07 *09:01 *10:01 *11:01 *12:01 *12:02 *13:01 *13:02 *13:03 *14:01 *14:02 *14:03 *14:04 *14:05 *14:13 *15:01 *15:02 *15:03 *15:04 *16:02

14 10 1 1 3 0 4 2 3 5 7 0 8 7 0 1 8 1 1 8 5 3 11 2 5 1 4 5 10 0 1 3 2 5

7 7 1 2 3 0 3 1 1 1 1 1 2 3 0 0 6 1 0 2 3 1 9 0 3 0 0 1 8 0 0 0 0 0

17 11 0 3 7 1 6 1 2 5 8 3 7 8 5 3 9 2 2 7 7 2 12 2 6 4 4 6 10 2 4 4 4 8

12 8 1 3 6 4 6 3 3 2 4 1 9 5 2 3 7 1 0 6 4 1 10 2 4 2 2 5 9 2 2 2 2 4

5 8 2 3 5 1 5 2 4 2 3 1 6 4 2 1 5 1 0 4 4 2 10 3 5 2 4 5 8 1 1 3 3 6

Sum:

141

67

182

137

121

L1*

L2†

H65

7 7 2 0 3 0 1 2 5 1 5 1 5 5 0 1 5 1 1 5 5 1 8 1 3 2 3 4 7 1 1 3 3 6

1 1 1 0 0 0 0 0 0 0 1 0 1 1 0 0 1 0 1 1 1 0 2 0 1 1 0 1 1 0 0 1 0 1

4 4 4 2 0 1 0 1 2 3 3 0 3 3 2 1 3 0 4 3 3 0 4 1 3 2 3 3 4 1 2 3 4 3

67 56 12 14 27 7 25 12 20 19 32 7 41 36 11 10 44 7 9 36 32 10 66 11 30 14 20 30 57 7 11 19 18 33

105

18

79

850

V

The prediction algorithms CombLib, SMM-align, NN-align, and/or NetMHCIIpan were used. *Combined results for the three nine-amino acid linkers including 5 amino acids upstream and downstream. † Combined results for the two 20-amino acid linkers including 5 amino acid upstream and downstream.

Knudsen et al. www.pnas.org/cgi/content/short/1314973111

6 of 6

73

Appendix II

Appendix II Manuscript II

Different human vaccine adjuvants promote distinct antigen-independent immunological signatures tailored to different pathogens

Niels Peter H. Knudsen Anja Olsen Cecilia Buonsanti Frank Follmann Yuan Zhang Rhea N. Coler Christopher B. Fox Andreas Meinke Ugo D´Oro Daniele Casini Alessandra Bonci Rolf Billeskov Ennio De Gregorio Rino Rappuoli Ali M. Harandi Peter Andersen Else Marie Agger

Published in Scientific reports 6, 2016.

75

Appendix II

™™™Ǥƒ–—”‡Ǥ…‘Ȁ•…‹‡–‹ˆ‹…”‡’‘”–•



R‡…‡‹˜‡†ǣͷͺ…–‘„‡”͸Ͷͷͻ A……‡’–‡†ǣͷͻ‡…‡„‡”͸Ͷͷͻ P—„Ž‹•Š‡†ǣ͸ͷ ƒ—ƒ”›͸Ͷͷͼ

‹ơ‡”‡–Š—ƒ˜ƒ……‹‡ƒ†Œ—˜ƒ–• ’”‘‘–‡†‹•–‹…–ƒ–‹‰‡Ǧ ‹†‡’‡†‡–‹—‘Ž‘‰‹…ƒŽ •‹‰ƒ–—”‡•–ƒ‹Ž‘”‡†–‘†‹ơ‡”‡– ’ƒ–Š‘‰‡• ‹‡Ž•‡–‡”Ǥ—†•‡ͷǡŒƒŽ•‡ͷǡ‡…‹Ž‹ƒ—‘•ƒ–‹͸ǡ ”ƒ ‘ŽŽƒͷǡ—ƒŠƒ‰͹ǡ Š‡ƒǤ‘Ž‡”ͺǡŠ”‹•–‘’Š‡”Ǥ ‘šͺǡ†”‡ƒ•‡‹‡ͻǡ‰‘Ʋ”‘͸ǡƒ‹‡Ž‡ƒ•‹‹͸ǡ Ž‡••ƒ†”ƒ‘…‹͸ǡ‘Žˆ‹ŽŽ‡•‘˜ͷǡ‹‘‡ ”‡‰‘”‹‘͸ǡ‹‘Rƒ’’—‘Ž‹͸ǡŽ‹Ǥƒ”ƒ†‹͹ǡ ‡–‡”†‡”•‡ͷƬŽ•‡ƒ”‹‡‰‰‡”ͷ Š‡ƒŒ‘”‹–›‘ˆ˜ƒ……‹‡…ƒ†‹†ƒ–‡•‹…Ž‹‹…ƒŽ†‡˜‡Ž‘’‡–ƒ”‡Š‹‰ŠŽ›’—”‹Ƥ‡†’”‘–‡‹•ƒ†’‡’–‹†‡• ”‡Ž›‹‰‘ƒ†Œ—˜ƒ–•–‘‡Šƒ…‡ƒ†Ȁ‘”†‹”‡…–‹—‡”‡•’‘•‡•Ǥ‡•’‹–‡–Š‡ƒ…‘™Ž‡†‰‡†‡‡† ˆ‘”‘˜‡Žƒ†Œ—˜ƒ–•ǡ–Š‡”‡ƒ”‡•–‹ŽŽ˜‡”›ˆ‡™ƒ†Œ—˜ƒ–•‹Ž‹…‡•‡†Š—ƒ˜ƒ……‹‡•Ǥ˜ƒ•–—„‡” ‘ˆƒ†Œ—˜ƒ–•Šƒ˜‡„‡‡–‡•–‡†’”‡Ǧ…Ž‹‹…ƒŽŽ›—•‹‰†‹ơ‡”‡–‡š’‡”‹‡–ƒŽ…‘†‹–‹‘•ǡ”‡†‡”‹‰‹– ‹’‘••‹„Ž‡–‘†‹”‡…–Ž›…‘’ƒ”‡–Š‡‹”ƒ…–‹˜‹–›Ǥ‡’‡”ˆ‘”‡†ƒŠ‡ƒ†Ǧ–‘ǦŠ‡ƒ†…‘’ƒ”‹•‘‘ˆƤ˜‡†‹ơ‡”‡– ƒ†Œ—˜ƒ–•Ž—ǡ ͻͿ ǡ Ǧǡ ͹ͷ ƒ† Ͷͷ‹‹…‡ƒ†…‘„‹‡†–Š‡•‡™‹–Šƒ–‹‰‡•ˆ”‘ M. tuberculosisǡ‹ƪ—‡œƒǡƒ†…ŠŽƒ›†‹ƒ–‘–‡•–‹—‡Ǧ’”‘ƤŽ‡•ƒ†‡ƥ…ƒ…›‹‹ˆ‡…–‹‘‘†‡Ž•—•‹‰ •–ƒ†ƒ”†‹œ‡†’”‘–‘…‘Ž•Ǥ‡‰ƒ”†Ž‡••‘ˆƒ–‹‰‡ǡ‡ƒ…Šƒ†Œ—˜ƒ–Šƒ†ƒ—‹“—‡‹—‘Ž‘‰‹…ƒŽ•‹‰ƒ–—”‡ •—‰‰‡•–‹‰–Šƒ––Š‡ƒ†Œ—˜ƒ–•Šƒ˜‡’‘–‡–‹ƒŽˆ‘”†‹ơ‡”‡–†‹•‡ƒ•‡–ƒ”‰‡–•ǤŽ—‹…”‡ƒ•‡†ƒ–‹„‘†› –‹–‡”•Ǣ ͻͿ ‹†—…‡†•–”‘‰ƒ–‹„‘†›ƒ† Ǧͻ”‡•’‘•‡•Ǣ Ǧ‹†—…‡†ƒ–‹„‘†‹‡•ƒ†ŠͷǢ Ͷͷ •Š‘™‡†ƒ‹š‡†ŠͷȀŠͷͽ’”‘ƤŽ‡ƒ† ͹ͷ ‹†—…‡†•–”‘‰Šͷ”‡•’‘•‡•Ǥ ͻͿ ƒ† Ǧ™‡”‡ •–”‘‰‹†—…‡”•‘ˆ‹ƪ—‡œƒ –‹–‡”•™Š‹Ž‡ Ͷͷǡ Ǧƒ† ͹ͷ ‡Šƒ…‡†’”‘–‡…–‹‘–‘ƒ† …ŠŽƒ›†‹ƒǤ ’‘”–ƒ–Ž›ǡ–Š‹•‹•–Š‡Ƥ”•–‡š–‡•‹˜‡ƒ––‡’––‘…ƒ–‡‰‘”‹œ‡…Ž‹‹…ƒŽǦ‰”ƒ†‡ƒ†Œ—˜ƒ–•„ƒ•‡† ‘–Š‡‹”‹—‡’”‘ƤŽ‡•ƒ†’”‘–‡…–‹˜‡‡ƥ…ƒ…›–‘‹ˆ‘”ƒ”ƒ–‹‘ƒŽ†‡˜‡Ž‘’‡–‘ˆ‡š–‰‡‡”ƒ–‹‘ ˜ƒ……‹‡•ˆ‘”Š—ƒ—•‡Ǥ

®

®

®

®

®

®

Beyond doubt, the advent of sequencing, rendering microbial genomes readily accessible, has been of utmost importance for vaccinology and has allowed for highly rational identification of vaccine antigens. Many novel vaccines currently in development are thus based on proteins or peptides predicted by computer databases or by screening antigen libraries1,2. Although these advances in the post-genomic era have enabled the design of highly pure, safe and simple vaccines, other challenges have emerged in parallel, including the inherent lack of immunostimulatory properties of proteins and peptides. Vaccine adjuvants are therefore considered key components in modern vaccinology since they provide the necessary help of enhancing the immune responses. In addition, adjuvants have many other favorable features including the ability to overcome immune senescence in elderly3, prolonging memory of the vaccines4, broadening the antibody repertoire5 and providing means for dose-sparing6. Since the adjuvants currently licensed for human-use almost exclusively induce antibody responses, there is a need for continued adjuvant research and development with the aim of expanding the repertoire of human adjuvants with the ability to induce cell-mediated immunity (CMI). ͷ

Department of Infectious Disease Immunology, Statens Serum Institut, Copenhagen, Denmark. ͸Novartis Vaccines and Diagnostics s.r.l (a GSK Company), Siena, Italy. ͹Department of Microbiology and Immunology, University of Gothenburg, Gothenburg, Sweden. ͺInfectious Disease Research Institute, Seattle, WA, USA. ͻValneva Austria GmbH, Vienna, Austria. Correspondence and requests for materials should be addressed to N.P.H.K. (email: NPK@ ssi.dk) or E.M.A. (email: [email protected])

1

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

76

Appendix II

www.nature.com/scientificreports/

Early work in adjuvant discovery has been highly empirical and little in depth mechanistic insights is available for the adjuvants currently used in human vaccines. The scarce mechanistic insight is illustrated by the fact that although aluminum salt (commonly referred to as Alum) adjuvants have been used in a range of different vaccines in humans since the 1920s, their mechanism of action remains incompletely understood7. The growing insights into innate immunology and the development of a range of novel technical tools have provided a unique opportunity to take adjuvant research and development to a new level, avoiding previous empirical trial-and-error approaches. The recent clinical testing of several novel exploratory adjuvants with the ability to generate CMI responses8,9 clearly indicates that it is plausible to expand the range of adjuvants available for human use. In addition to improving the vaccine technology, there is also an urgent need for a more rational preclinical development of adjuvants and better tools for predicting the protective potential of an adjuvant for a given antigen and/or disease target. However, many studies aimed at investigating the potential of novel adjuvants, are carried out with a range of different protocols, different antigens and in many cases with a lack of appropriate comparators e.g. Alum. Studies are also often carried out with non-clinically relevant model antigens, e.g. ovalbumin (OVA), or without animal challenge experiments, which could otherwise provide information on the efficacy. Furthermore, the problems with gaining access to adjuvants currently in clinical development make it difficult to perform comparisons with adjuvants for which we already know the immunological profile in humans. All of these parameters render it impossible to obtain an overview of the potency of different adjuvants and appreciate whether it is worthwhile to pursue further clinical development of a novel adjuvant, hence leaving the adjuvant field fragmented and dispersed10. This could potentially undermine valuable information about candidate adjuvants that could otherwise lead to identification of novel viable adjuvants for clinical development. Furthermore, this may lead to overlooking adjuvants with very distinct immunological profiles, which are markedly different from those already filling the clinical pipeline. Herein, we performed a detailed comparison of five different clinical adjuvants in three different disease models (tuberculosis (TB), chlamydia, and influenza) using a standardized protocol that allowed comparison of results across different antigens and adjuvants. The main objectives of this study were to dissect and compare the immunological profile of these adjuvants and to test the efficacy in very different disease models with different requirements for protection; TB, which primarily requires a Th1 profile11, chlamydia for which an antibody response with a mixed Th1/Th17 profile is critical12, and influenza where measurement of antibody responses are accepted as a correlate of protection13. The proprietary adjuvants included in this study were obtained from partners involved in the EU-funded large collaborative project Advanced Immunization Technologies (ADITEC), and have the common denominator of already being advanced to clinical development (IC31 , CAF01, GLA-SE) or included in registered human vaccines (MF59 and Alum). This provided the possibility of comparing the mouse data reported herein to human data obtained from the clinical studies. The five adjuvants gave rise to very distinct immunological signatures irrespective of the target antigen tested, suggesting that the adjuvants when used together with antigens have potential in induction of protection against highly different diseases. These results on categorization of different clinical-grade adjuvants with proven excellent safety based on their immune profiles and their ability to mount protective immunity, lays the foundation for development of next-generation vaccines for human use.

®

®

ƒ–‡”‹ƒŽ•ƒ†‡–Š‘†•

‹ƒŽ•Ǥ Female CB6F1 (TB and influenza) and B6C3F1 (chlamydia), 6–8 weeks old, were ordered from Harlan Laboratories (The Netherlands, TB and chlamydia) or Charles River (Italy, influenza). TB- and chlamydia-infected mice were housed in animal facilities at Statens Serum Institut, Denmark, while influenza experiments were performed at Novartis Vaccines and Diagnostics s.r.l., Italy. Mice were provided standard food and water ad libitum. TB-infected animals were housed in a biosafety level 3 facility in cages contained within laminar flow safety enclosures (Scantainer, Scanbur). –Š‹…•–ƒ–‡‡–Ǥ The use of mice was conducted in accordance with the regulations set forward by the respective national animal protection committees and in accordance with European Community Directive 86/609 and the U.S. Association for Laboratory Animal Care recommendations for the care and use of laboratory animals. All the techniques/procedures have been refined to provide for maximum comfort/minimal stress to the animals. Experiments performed at Statens Serum Institut have been approved by the governmental Animal Experiments Inspectorate under licenses 2014-15-2934-01065 (TB) and 2013-15-2934-00978 (chlamydia), while experiments performed at Novartis Vaccines and Diagnostics s.r.l. have been approved by the Ministry of Health under license AEC201111.

–‹‰‡•Ǥ H56. H56 (Ag85B-ESAT6-Rv2660c) was recombinantly produced in E. coli K12 containing the H56 expression vector under IPTG induction, after which the recombinant H56 was harvested as inclusion bodies from disrupted host cells by centrifugation. The inclusion bodies were dissolved in 8 M Urea, 10 mM cysteine, pH 6 and filtered through 0.22 μ m filters. The recombinant protein was purified using a two-stage ion exchange scheme (Sartobind Q SingleSep and S SingleSep), at pH 4.9 and 6.0 respectively) after which the buffer was switched to 20 mM glycine, pH 8.8 using a tangential flow filtration system (Merck Millipore). Protein purity was assessed by SDS-PAGE followed by coomassie staining and western blot with anti-H56 (Polyclonal, rabbit serum) and anti-E. coli antibodies (DAKO) to detect contaminants. Furthermore, the protein was validated for residual DNA, endotoxins and bioburden in accordance with internal GMP standards.

®

CT681. A synthetic DNA construct containing ompA (CT681 – AA23-AA349, leaving out the signal peptide) was codon-optimized for expression in E. coli followed by insertion into the pJexpress 411 vector (DNA2.0),

2

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

77

Appendix II

www.nature.com/scientificreports/

Figure 1. Study design for the animal experiments.

including a N-terminal histidine tag. Following protein expression, the protein was initially purified by metal chelate affinity chromatography as described in Follmann et al. 200814 and subsequently purified by IEC using HiTrap Q Sepharose HP (GE Healthcare). Purified rMOMP was refolded by a stepwise removal of buffer containing urea ending up in 20 mM Tris, 150 mM NaCl, 10% glycerol pH 8 which yielded soluble protein. Protein purity was assessed by SDS-PAGE followed by coomassie staining and western blot with anti-penta-His (Qiagen) and anti-E. coli antibodies to detect contaminants (DAKO). The Limulus Amebocyte Lysate test was used to determine the amount of endotoxin in the rMOMP preparation and was below 10 EU/mg. Hemagglutinin (HA). Influenza A/California/2009 (H1N1) subunit vaccine was produced by Novartis Vaccines. The vaccine was derived from embryonated eggs applying production procedures followed for clinical grade influenza vaccines and its major constituent is hemagglutinin (HA) from the indicated influenza strain.

†Œ—˜ƒ–•Ǥ CAF01 from Statens Serum Institut (Copenhagen, Denmark), IC31® from Valneva Austria

GmbH (previously Intercell AG, Vienna, Austria), GLA-SE from the Infectious Diseases Research Institute (Seattle, WA, United States), MF59 from Novartis Vaccines (Siena, Italy), Aluminium hydroxide (Al(OH)3) (2% Alhydrogel), from Brenntag Biosector (Frederikssund, Denmark) was used in the TB and Chlamydia experiment, while the Aluminium hydroxide suspension used in the Influenza experiment was produced at Novartis Vaccines and Diagnostics (Marburg, Germany).

®

—‹œƒ–‹‘Ǥ Mice were immunized subcutaneously (s.c.) at the base of the tail three times at three-week intervals. Adjuvants were mixed with 5 μ g recombinant antigen or 1 μ g of HA by vortexing to a final volume of 200 μ l for each injection. CAF01 (dose 250 μ g/50 μ g (DDA/TDB)/200 μ l)15, IC31 (dose 100 nmol/4 nmol (KLK/ ODN1a)/200 μ l)16, GLA-SE (dose 5 μ g GLA and 2% v/v squalene in 200 μ l)17, Alhydrogel 2.0% (dose 500 μ g aluminum content /200 μ l)18, MF59 (dose of 100 μ l 4.3% w/v squalene, 0.5% w/v Tween 80, 0.5% w/v Span 85 mixed 1:1 with PBS)19. Control mice received three injections of 5 μ g recombinant antigen or 1 μ g HA in 200 μ L PBS, respectively, while negative controls received three injections of sterile PBS (pH 7.4). The study designs for the three disease models are shown in Fig. 1.

®

®

 ˆ‘”ƒ–‹‰‡Ǧ•’‡…‹Ƥ…•‡”—ƒ–‹„‘†‹‡•Ǥ Maxisorb Plates (Nunc) were coated with 0.05 μ g/well H56 or MOMP in PBS overnight at 4 °C. Individual mouse sera from at least four mice per group were analyzed in duplicate, ten- (IgG1) or fivefold (IgG2a) dilution series, at least 8 times in PBS with 2% BSA, starting with a 20-fold dilution. HRP-conjugated secondary antibodies (rabbit anti-mouse IgG1 and IgG2a; Zymed) was diluted 1:2000 in PBS with 1% BSA. After 1 h of incubation, antigen-specific antibodies were detected using TMB substrate as described by the manufacturer (Kem-En-Tec Diagnostics). The absorbance values were plotted as a function of the reciprocal dilution of serum samples. Reciprocal plasma dilutions corresponding to a cut-off of 0.2 OD450 were computed using Excel (v.15.0.4693.1000). HA specific serum antibodies were determined by a two-step fully automated rapid ELISA (Hamilton Starlet System, Switzerland) with individual sera to titrate total HA specific IgG, IgG1 and IgG2 as described previously20. ›’Š‘…›–‡…—Ž–—”‡•ǡ Ǧͷͽ ƒ† Ž‘™…›–‘‡–”›ˆ‘”ƒ†ͻͼǤ Peripheral blood mononuclear cells (PMBC) and splenocytes were isolated as described previously21,22. Cultures were adjusted to 2 × 105 cells/well (MSD/ELISA) or 1–2 × 106 cells/well in a total volume of 200 μ l/well (IC-FACS) and stimulated with antigens at a final concentration of 2 μ g/ml, whereas Con A was used at a concentration of 1 μ g/ml as a positive control for cell viability. Culture supernatants were harvested after 72 h of incubation for the investigation of Multiplex cytokine assay and IL-17 ELISA (as previously described23). Splenocytes were stimulated at 37 °C in the presence of recombinant antigen (2 μ g/ml) for 1 hour, and subsequently incubated for 5 hours after adding 10 μ g/ ml BFA (Sigma-Aldrich) and stained as previously described22. Responses were analyzed using a FACSCanto flow cytometer (BD Biosciences) and FlowJo v.10.0.7 (Tree Star Inc.).

–”ƒǦ…‡ŽŽ—Žƒ”ˆŽ‘™…›–‘‡–”›•–ƒ‹‹‰ˆ‘”•’Ž‡‘…›–‡••–‹—Žƒ–‡†in vitro™‹–Šƒ–‹Ǧ ‰‡Ǥ Spleens were harvested and filtered through a 70 μ m nylon mesh (BD Biosciences) and stimulated with 1 μ g/ml HA O/N and subsequently incubated for 4 hours with BFA (2.5 μ g/ml) as described previously24. Briefly,

3

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

78

Appendix II

www.nature.com/scientificreports/ Adjuvant

Category

Immune profile*

Stage of development

References

Insoluble aluminum salts

TH2, antibody

In several licensed products

62,63

Squalene oil-in-water emulsion

(TH1), TH2, antibody

Licensed for use in influenza vaccine in EU

64–66

CAF01

Cationic liposomes formulated with a Mincle agonist

TH1, TH17

Phase 1

45,67

GLA-SE

Squalene emulsion combined with a TLR4 agonist

TH1, antibody

Phase 2

9,68

Cationic antimicrobial polypeptides combined with a TLR9 agonist

TH1

Phase 2

8,69

Alum MF59

®

®

IC31

Table 1.

*Immune profile reported in the literature, see references.

cells were stained with combinations of the following antibodies: CD4-V500, CD44-V450, CD3-PerCP-Cy5.5, CD8-Texas Red , IL-4 and IL-13 Alexa Fluor 488, IL-2 APC, TNF-Alexa Fluor 700, IFN-γ PE, IL-17 PE Cy7 and LIVE/DEAD Fixable Aqua. Flow cytometry was performed on FACS LSRII instruments using DIVA software (BD Biosciences) and data were then analyzed using Flowjo software (Tree Star Inc.).

®

®

®

®

—Ž–‹’Ž‡š…›–‘‹‡ƒ••ƒ›Ǥ The proinflammatory panel 1 (Mouse) 7-plex cytokine assay (IFN-γ, IL-2, IL-4, IL-5, IL-6, IL-10, TNF-α ) or the Th1/Th2 panel (mouse) 9-plex cytokine assay (IFN-γ , IL-1b, IL-2, IL-4, IL-5, IL-10, TNF-α , IL-12) or the mouse IL-17 assay (all from Meso Scale Discovery) were performed according to the manufacturer’s instructions. The plates were read on the Sector Imager 2400 system or 6000 system (Meso Scale Discovery) and calculation of cytokine concentrations in unknown samples was determined by 4-parameter logistic non-linear regression analysis of the standard curve.

›…‘„ƒ…–‡”‹ƒŽ …ŠƒŽŽ‡‰‡Ǥ Six weeks after the third immunization animals were infected with Mycobacterium tuberculosis Erdman (ATCC) by the aerosol route as described previously15. Six weeks after infection, mice were sacrificed and organs homogenized in PBS for bacterial enumeration as described previously in Ref. 15. ŠŽƒ›†‹ƒ…ŠƒŽŽ‡‰‡Ǥ Six weeks after the final vaccination the mice were challenged intra-vaginally with 4 × 105 inclusion forming units (IFU) Chlamydia trachomatis Serovar D (UW-3/Cx, ATCC VR-885) purchased from the American Type Culture Collection (ATCC) essentially as described previously15. The mice oestrus cycles were synchronized ten and 3 days before challenge by subcutaneous injection of 2.5 mg Medroxyprogesteronacetat (Depro-Provera; Pfizer). Vaginal swabs were obtained at day 3 and 10 after infection and analyzed as previously described18. Inclusions were enumerated by visual inspection using a fluorescence microscopy. ƪ—‡œƒ‹Š‹„‹–‹‘ƒ••ƒ›Ǥ Serum samples were pre-treated with neuraminidase (Biogenetics) over-

night at 37 °C, and HI was tested on individual sera. Briefly, 25 μ L of two-fold serially diluted samples were incubated with 25 μ L of strain-specific influenza antigen (whole virus inactivated, containing four hemagglutinating units) for 60 min at room temperature. A 0.5% (v/v) suspension of adult turkey red blood cells was added and incubated for 60 min. Outcomes were determined by visual inspection: a red dot indicated a positive reaction (inhibition), and a diffuse patch indicated a negative reaction (hemagglutination). All sera were tested in duplicates. The titer was defined as the reciprocal of the serum dilution at which the last complete agglutination inhibition occurred.

–ƒ–‹•–‹…ƒŽƒƒŽ›•‹•Ǥ Prism 6 software (GraphPad v6.05) was used for all statistical analyses. TB CFU, Chlamydia IFU, IgG and HA inhibition titers were log-transformed before analysis. TB CFU were analyzed using one way ANOVA with Dunnett’s multiple comparisons test. Chlamydia IFU, IgG and HA inhibition titers were analyzed using Kruskal-Wallis test followed by Dunn’s post-test. Statistically significant differences are marked by asterisks in figures and explained in the figure legends.

‡•—Ž–•

’ƒ‡Ž‘ˆ†‹˜‡”•‡ƒ†Œ—˜ƒ–•ˆ‘”…‘’ƒ”ƒ–‹˜‡–‡•–‹‰Ǥ As part of the EU-funded large collaborative project Advanced Immunization Technologies (ADITEC), several partners contributed with their proprietary adjuvants to create a panel of highly diverse adjuvants already in clinic (Table 1). The panel included squalenebased oil-in-water emulsions (MF59 and GLA-SE – the latter combined with a synthetic TLR4 agonist), cationic peptide KLK combined with the TLR9-stimulatory oligodeoxynucleotide ODN1a (IC31 ), cationic liposomes formed of dimethyldioctadecylammonium (DDA) stabilized with synthetic mycobacterial cord factor (trehalosedibehenate) signaling through the C-type lectin receptor Mincle (CAF01), and conventional Alum delivered in the form of a 500 μ g aluminum-content dose of diluted 2.0% Alhydrogel. The adjuvants were combined with antigens from three pathogens; H56 is an engineered fusion of three Mycobacterium tuberculosis (M.tb.) antigens, Ag85B, ESAT-6 and Rv2660c, MOMP is the recombinantly produced major outer membrane protein of Chlamydia trachomatis (C.t.), and HA is the purified egg-derived hemagglutinin (HA) from A/California/7/2009 (H1N1) influenza strain.

®

®

4

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

79

Appendix II

www.nature.com/scientificreports/

Figure 2. Vaccine-specific humoral immune-responses after vaccination. Groups of mice were immunized 3 times s.c. with H56, MOMP or HA formulated in Alum, MF59 CAF01, GLA-SE or IC31 with 3 week intervals (H56 with Alum was not tested in this assay). Two weeks after each vaccination serum samples from individual mice (n = 4 CB6F1 in TB, n = 12 B6C3F1 in chlamydia, n = 8 CB6F1 in influenza) were analyzed for antigen-specific IgG1 and IgG2a antibodies. Sera from vaccinated animals were analyzed in serial dilutions. H56 and MOMP IgG1/2a titers were determined at a cutoff of 0.2 OD450, while HA-specific IgG1/2a titers were determined from an internal standard. (a) IgG1 and IgG2a titers two weeks after 1st, 2nd and 3rd vaccination (Post 1, 2 and 3). Each point represents the geometric mean + 95% CI of 4, 12 and 8 mice, for H56, MOMP and HA, respectively. Adjuvanted groups were compared to the no adjuvant group at each time-point using the Kruskal-Wallis test followed by Dunn’s post-test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. (b) The ratio of IgG1/IgG2a titers after the second and third vaccination are shown. Bars represent the geometric means of 4, 12 and 8 mice, for H56, MOMP and HA, respectively. The bars are plotted from the median ratios within each antigen group (H56 10, MOMP 25, HA 256). Post 1 ratios have been omitted due to low titer responses resulting in large variations in ratios. The H56 and MOMP results are representatives of two independent experiments with similar results.

®

®

‘’ƒ”ƒ–‹˜‡‡˜ƒŽ—ƒ–‹‘‘ˆ–Š‡Š—‘”ƒŽ‹—‡”‡•’‘•‡‹†—…‡†„›–Š‡†‹ˆˆ‡”‡–ƒ†Œ—Ǧ ˜ƒ–•Ǥ We first compared the ability of the five different adjuvants to induce humoral immune responses using the three model antigens. Groups of mice were vaccinated three times with H56, MOMP, or HA in combination with either CAF01, IC31 , GLA-SE, MF59 or Alum (H56 with Alum was not tested in this assay). Three vaccinations were chosen as previous optimization studies have shown this immunization regimen to induce stronger immune responses compared to two vaccinations using CAF01 (unpublished data) and GLA-SE25. Adding a fourth dose did not improve responses (unpublished data). Two weeks after each vaccination, the mice were partially bled and the vaccine-specific IgG1 and IgG2a antibody responses were measured in the serum using ELISA (Fig. 2). The kinetics of specific antibody induction after one, two or three vaccinations showed that GLA-SE in general induced the highest and earliest IgG2a antibody response, with a clear and detectable response after only a single vaccination (Fig. 2a). After the second vaccination, MF59 induced the highest IgG1 antibody response and was clearly superior to the other adjuvants in terms of generating HA-specific IgG1 titers in mice. Compared to vaccination with antigen H56 and MOMP alone (No Adj), there was a clear effect of adding adjuvants on the induction of IgG2a titers in particular after the second and third vaccination. In contrast, the benefit

®

®

®

5

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

80

Appendix II

www.nature.com/scientificreports/

Figure 3. H56, MOMP and HA specific cellular immune response induced by vaccination with different adjuvants. Groups of mice (n = 4 CB6F1 in TB and influenza, B6C3F1 in chlamydia) were vaccinated three times with H56, MOMP or HA formulated in Alum, MF59 CAF01, GLA-SE or IC31 . Two weeks after the third vaccination splenocytes from individual mice were isolated and stimulated with H56, MOMP or HA for 72 hours. The level of cytokines released to the culture supernatants were measured. (a) Pie-charts showing the relative contribution of antigen-specific IL-17, IL-5 and IFN-γ . Means of four mice, measured in triplicates and the background from samples stimulated with medium without antigen have been subtracted. (b) The levels of IFN-γ , TNF-α , IL-10, IL-5 and IL-17 released to the supernatant are shown as stacked bars, representing the means of four individual mice measured in triplicates and background from samples stimulated with medium without antigen have been subtracted. The H56 and MOMP results are representatives of two independent experiments.

®

®

of adding adjuvants on the IgG1 titers was less evident and three vaccinations with recombinant H56 or MOMP alone showed comparable antibody titers as the adjuvanted groups, with the exception of MOMP combined with MF59 or Alum (P > 0.0001 and P > 0.01 compared to No Adj). HA titers followed a slightly different pattern with a more clear benefit of adding adjuvants on IgG1 titers, while the second and third vaccinations, even with adjuvants, had very limited boosting effect on the IgG2a antibody responses after the initial priming. This is in line with previously published data where adjuvanted influenza vaccines are prone towards IgG1 induction while inducing little IgG2a26. Common for all three antigens, after the second vaccination and particularly the third vaccination, an IgG1 bias was observed with the Th2-biased adjuvants Alum and MF59 (Fig. 2b), while the Th1-biased adjuvants GLA-SE and IC31 skewed the balance towards IgG2a. CAF01 had a slight bias towards IgG2a induction but with an overall more balanced IgG1/IgG2a ratio. When administered unadjuvanted, all three antigens induced IgG1 biased responses, which is in agreement with previously published data27–29.

®

®

®

‘’ƒ”ƒ–‹˜‡‡˜ƒŽ—ƒ–‹‘‘ˆ–Š‡…‡ŽŽ—Žƒ”‹—‡”‡•’‘•‡•‹†—…‡†„›–Š‡†‹ˆˆ‡”‡–ƒ†Œ—Ǧ ˜ƒ–•Ǥ To evaluate CMI responses, splenocytes were isolated two weeks after the third vaccination and re-stimulated with the homologous recombinant protein. The levels of secreted cytokines released to the supernatant after 72 hours of in vitro antigen stimulation were measured using MSD technology. Depicting the overall profiles as the relative cytokine contribution demonstrated a breakdown into different classes of adjuvants (Fig. 3a). Thus, Alum and MF59 exhibited a Th2 bias whereas GLA-SE and IC31 induced a clear Th1-biased response with a strong IFN-γ component and negligible levels of IL-5. CAF01 induced a mixed Th1/Th17 phenotype and was the only adjuvant with a clear IL-17 response for all three antigens. Considering the absolute amount of cytokine secretion showed a more diverse pattern and magnitude depending on the type of antigen used (Fig. 3b). Thus, CAF01 was the only adjuvant consistently giving rise to high levels of cytokine secretion, primarily IFN-γ , with all three antigens whereas IC31 gave highest response in combination with H56,

®

®

®

6

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

81

Appendix II

www.nature.com/scientificreports/ HA inhibition titers POST 1 Adjuvant

POST 2

POST 3

GMT

95% CI

GMT

95% CI

GMT

< 10

—

< 10

—

< 10

—

No Adj

62

[19–203]

1050

[434–2543]

1076

[548–2113]

Alum

538*

[227–1274]

1810

[1327–2468]

2915

[1636–5198]

MF59

®

320

[133–768]

19612****

[10087–38177]

8611****

[5890–12582]

CAF01

336

[117–800]

2690

[2143–3964]

2436

[2213–2716]

GLA-SE

269

[179–405]

5346*

[3242–8822]

5263*

[2078–13333]

381

[145–1001]

2792

[1765–4416]

2560

[1878–3491]

PBS

®

IC31

95% CI

Table 2.

intermediate levels with MOMP and low amounts in combination with HA. GLA-SE had a prominent adjuvant effect for H56 whereas it showed a modest effect in adjuvanting HA and MOMP. The phenotype and cytokine subsets were further assessed by flow cytometry using the frequency of IFN-γ , IL-17 and IL-4/IL-13 for HA (Fig. S1) and the IFN-γ , IL-2, TNF-α distribution for H56 and MOMP (Fig. S2). Common for all adjuvants was the induction of CD4 T cell responses whereas no CD8 T cell responses were observed (results not shown). All groups receiving HA showed a measurable IL-4/IL-13 population, suggesting a Th2-bias response to this antigen in CB6F1 mice. This Th2 response was increased in combination with all adjuvants, albeit groups receiving HA in combination with the Th1 adjuvants, GLA-SE, IC31 , and CAF01 also had a population of IFN-γ producing CD4 T cells. CAF01 induced the overall highest recall response upon restimulation with HA and an additional IL-17 producing component. Overall, the phenotypes determined by flow cytometry were coherent with those determined by MSD. In the H56/MOMP analysis, IC31 induced the most vigorous response in terms of frequency of cytokine-producing CD4 T cells in particular when combined with H56. The response was dominated by triple-positive IFN-γ /IL-2/TNF-α T cells constituting ~50% of the cells, but also a considerable population of IL-2/TNF-α double producing cells usually considered to be the key trait of central memory T cells22. GLA-SE and CAF01 induced comparable T cell profiles with triple-positive, TNF-α /IL-2 double positive and TNF-α single producers. Furthermore, IFN-γ /TNF-α double producers were observed with CAF01, GLA-SE and IC31 , while IC31 was the only adjuvant that induced a small but noticeable population of IFN-γ single producers. IFN-γ /TNF-α and IFN-γ single producers are generally considered as more differentiated/effector subpopulations30. MF59 induced a strong Th2 bias response, based on the IL-5 secretion (see Fig. 3a,b), as well as distinct populations of TNF-α single producing CD4 T cells and IL-2/TNF-α co-producers.

®

®

®

® ®

In vitro‡˜ƒŽ—ƒ–‹‘‘ˆ’”‘–‡…–‹˜‡‡ƥ…ƒ…›ƒ‰ƒ‹•–‹ƪ—‡œƒǤ In order to assess the adjuvants ability to induce protective immunity against influenza, the hemagglutination inhibition (HAI) titers were measured two weeks after each vaccination (Table 2). The HAI titer is a surrogate marker of influenza-specific neutralizing antibodies widely used for in vitro correlate of protection13. MF59 adjuvanted HA induced the highest levels of HAI titers, significantly higher than that of HA alone after the second vaccination (p < 0.0001, Kruskal-Wallis followed by Dunn’s post-test compared to no adjuvant). GLA-SE also elicited a significantly elevated HAI titer after the second vaccination (p < 0.05) while IC31 , CAF01 and Alum did not induce significantly higher HAI titers compared to that of antigen alone.

®

®

”‘–‡…–‹˜‡‡ƥ…ƒ…›‹ƒƒ‡”‘•‘Ž…ŠƒŽŽ‡‰‡‘†‡ŽǤ The potential of different adjuvants in induction of protective immunity to M.tb. infection was then studied. Six weeks after the third vaccination, H56 vaccinated mice were challenged with M.tb. Erdman through the aerosol route. Two weeks after challenge, PBMCs were isolated and stimulated with H56 for 72 hours and analyzed for vaccine-specific secretion of IFN-γ , IL-10, IL-2, IL-4, IL-5, IL-6, TNF-α and IL-17 (Fig. 4a). The post-infection H56 responses measured in the PBMC’s (post) were compared to the pre-infection responses in the splenocytes (pre) with the differences in target organ kept in mind. While infection alone induces limited responses at this early time point after challenge, all groups previously primed with an adjuvanted vaccine were able to mount significant responses. H56 in combination with CAF01, IC31 and GLA-SE all primed IFN-γ , IL-6 and TNF-α responses prior to infection and all of these cytokine responses remained two weeks after challenge. The strongest amplification of IFN-γ and IL-6 responses was observed in the GLA-SE vaccinated group. MF59 maintained the same profile with no IFN-γ secretion after infection. CAF01 was the only group where an IL-17 response was observed after immunization and this response was strongly elevated after infection. In contrast, the IL-17 response was induced only after infection in the IC31 and GLA-SE groups. H56-specific IL-10 was seen in all groups independent of their Th-profile and was not augmented by the infection. Six weeks after infection, the mice were sacrificed and the numbers of viable colony-forming units (CFU) in the lungs were enumerated in order to assess protective efficacies. (Fig. 4b). The highest levels of protection were seen upon vaccination with H56 in CAF01/GLA-SE (p < 0.0001) and IC31 (p < 0.001), whereas MF59 gave rise to a relatively modest, but significant, reduction of 0.27 log10 CFU (p < 0.05). As with the related fusion protein H1 (Ag85B-ESAT-6), H56 combined with Alum did not mount protective immunity to TB challenge with comparable levels of bacterial numbers as the H56 alone group15. There was no significant difference between the PBS control group and H56 antigen alone (No adj).

®

®

®

®

®

7

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

82

Appendix II

www.nature.com/scientificreports/

Figure 4. H56 specific immune responses and protective efficacy against challenge with Mycobacterium tuberculosis. Groups of mice (CB6F1) were immunized three times s.c. with 5 μ g of H56 formulated in Alum, MF59 CAF01, GLA-SE or IC31 with 3 week intervals. Six weeks after the third vaccination, mice received an aerosol challenge with M.tb. Erdman (~50 CFU/mouse). Two weeks after challenge, PBMC’s were isolated from a pool of 12 individual mice and stimulated with recombinant H56 (2 μ g/ml) for 72 hours. Culture supernatants were analyzed for released cytokines (TNF-α , IFN-γ , IL-2, IL-4, IL-5, IL-6, IL-10, IL-17). (a) Grey bars indicate the levels of cytokines released from PBMCs after infection (Post) and the black bars represent the levels of cytokines measured in splenocytes prior to infection (Pre). Black bars represent mean + SEM of four mice measured in triplicates (Pre), while grey bars are means of triplicates measured of PBMC’s pooled from 12 mice within groups (Post). (b) After six weeks of infection, mice were euthanized and the bacterial loads in the lungs (CFUs) of individual mice was assessed. Each bar represents the mean + SEM Log10 CFU levels and represents 14–16 mice (Alum n = 8). The results are pooled from two experiments with same overall results. Adjuvanted groups were compared to the “no adjuvant” group using one way ANOVA with Dunnett’s multiple comparisons test. *P < 0.05; ***P < 0.001; ****P < 0.0001.

®

®

”‘–‡…–‹˜‡‡ƥ…ƒ…›‹ƒ—”‹‡…ŠŽƒ›†‹ƒ…ŠƒŽŽ‡‰‡‘†‡ŽǤ Next, we sought to evaluate the efficacy of different vaccine formulations in induction of protective immunity to a vaginal C.t. challenge. Groups of mice, vaccinated with MOMP singly or in combination with different adjuvants, were challenged vaginally with C.t. six weeks after the last vaccination. Three weeks after challenge, PBMCs were stimulated with recombinant MOMP for 72 hours for analysis of IFN-γ , IL-10, IL-2, IL-4, IL-5, IL-6, TNF-α and IL-17 in the supernatant. The post-infection PBMC responses (post) as well as the pre-infection splenocyte responses were examined (pre) (Fig. 5a). Like for H56, the cytokines IFN-γ , IL-6 and TNF-α were among the dominating cytokines in the CAF01, GLA-SE, and IC31 groups but only IFN-γ seemed to be selectively increased upon chlamydia infection. The CAF01-induced IL-17 response was shown to be clearly boosted by the infection, although not to the same extent as that observed with the TB infection and further, no IL-17 responses are seen in any other groups. Vaginal swabs were collected at day 3 and 10 post challenge and protection was assessed by isolating C.t. from vaginal swabs and comparing the number of IFUs recovered at the indicated time points (Fig. 5b). CAF01, IC31 and GLA-SE were able to significantly reduce the bacterial shedding at day 3 and 10 post challenge, whereas MOMP adjuvanted with MF59 or Alum was not able to induce significant protection at any time-point. The un-adjuvanted MOMP vaccinated group did not show any significant reduction in their bacterial loads on day 3 after challenge, albeit a significant reduction was seen at day 10 post challenge.

®

®

®

‹•…—••‹‘ Herein, we report for the first time a comprehensive head-to-head comparison of immunogenicity and protective efficacy of five different clinically tested/practiced adjuvants; Alum, MF59 , CAF01, GLA-SE and IC31 . The adjuvant panel is highly diverse, but with the common denominator of having been tested in human subjects

®

®

8

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

83

Appendix II

www.nature.com/scientificreports/

Figure 5. MOMP specific immune responses and protective efficacy against challenge with Clamydia trachomatis. Groups of mice (B6C3F1) were immunized three times s.c. with 5 μ g of MOMP formulated in Alum, MF59 CAF01, GLA-SE or IC31 with 3 week intervals. Six weeks after the third vaccination, mice received a vaginal challenge with C. trachomatis. Three weeks after challenge, PBMC’s were isolated from individual mice (n = 12), pooled within groups and stimulated with recombinant MOMP for 72 hours. Culture supernatants were analyzed for released cytokines (TNF-α , IFN-γ , IL-2, IL-4, IL-5, IL-6, IL-10, IL-17). (a) The levels of cytokines released from PBMCs after infection (post) were compared to the levels of cytokines prior to infection, measured in splenocytes (pre). Black bars represent mean + SEM of four mice measured in triplicates (Pre), while grey bars are means of triplicates measured of PBMC’s pooled from 12 mice within groups (Post). (b) On day 3 and 10 after infection, mice were vaginally swabbed and the vaginal bacterial load (IFUs) of 28 individual mice (Naïve n = 56) was assessed. Each point represents the median Log10 ± interquatile range. Adjuvanted groups were compared to the naïve controls at each time-point using the Kruskal-Wallis test followed by Dunn’s post-test. *P < 0.05; ***P < 0.001; ****P < 0.0001.

®

®

offering a unique opportunity to compare the mouse data from the current study with already available human data. Alum was the first adjuvant to be used in human vaccines and since the 1920s, billions of doses have been administered primarily with the purpose of adsorbing antigen and inducing antibodies. From numerous human trials, it is known that Alum is a poor inducer of cellular immune responses. In line with the human data and as reported previously in several animal studies, the single characteristic of Alum in our study was its ability to enhance antibody titers. MF59 has been used in licensed vaccines in Europe since 1997 and primarily induces antibodies and an increased CD4 T cell response with an unbiased Th-profile in humans31–33. Overall, this is in agreement with our findings in mice showing strong antibody responses but also induction of a CD4 response with a Th2 profile along with CD4 T cells co-producing TNF-α /IL-2. GLA-SE, CAF01, and IC31 are in the early stages of clinical development but it is clear from the emerging data that they are able to induce strong and long-lived CMI responses in humans and for CAF01 and IC31 phenotypic profiles with a remarkable resemblance between mouse and man. MF59 was originally developed as a delivery system for immunomodulators but pre-clinical studies suggested that this emulsion had potential on its own with influenza antigens34. MF59 was the most effective adjuvant in an influenza vaccine assessed as increased IgG1 and protective HI titers and indeed numerous clinical trials in humans show the same overall profile and also a clear efficacy of MF59 in influenza vaccines35 (reviewed by Reed et al. 201336). However, in humans with no pre-existing immune response e.g. against H5N1, two vaccinations seem to be required in order to enhance IgG titers (Galli et al. 2009, Fig. 237). Since the discovery of the adjuvant activity of aluminum compounds, there have been several attempts to use different aluminum forms for augmenting the effect of influenza vaccines but with very limited success38 (reviewed by Tetsutani et al. 201239). In a clinical study directly comparing the effect of Alum and MF59 , there was an 8–9 fold increase in frequency of

®

®

®

®

®

®

®

9

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

84

Appendix II

www.nature.com/scientificreports/ subjects reaching the predefined endpoint (HAI titers ≥ 40), when vaccinated with equivalent doses of subvirion influenza A/H5N1 in MF59 compared to Alum40. The superiority of MF59 over Alum in induction of influenza specific antibody responses was also observed in our study with a 2.5–10 fold increase in HAI for MF59 compared to Alum and in a recent study aimed at investigating the efficacy of emulsion adjuvants in influenza subunit vaccines24. It is well-known that the antibodies raised by current inactivated influenza vaccines are mainly strain-specific antibodies directed against highly variable surface proteins resulting in protection against antigenically matching virus strains. With the threat of novel pandemic influenza variants and the continuous emergence of drifted strains, there has been considerable interest in the development of influenza vaccines with broader coverage. Several clinical trials have shown that MF59 can induce responses of increased breadth compared to un-adjuvanted influenza vaccines, which mainly relates to the antibody repertoire, whereas there is limited information on the ability of MF59 to induce CMI responses to the vaccine (reviewed in O´Hagan et al. 201141). Thus, it has been suggested that the potency of emulsion-based adjuvants could be improved by combining them with additional immunomodulators capable of inducing CD8 and/or CD4 T cell responses, with the ability to induce influenza cross-protection42,43. Accordingly, it has been shown in preclinical studies that GLA-SE containing the TLR4-ligand GLA, and not the stable emulsion (SE) alone, was capable of inducing protection against infection with a heterologous virus strain44. As in the studies herein, GLA-SE was found to enhance IFN-γ secretion and induce a shift from IgG1 to IgG2 production. In our study, GLA-SE also elicited significantly elevated HAI titers, which is in line with recent human data where GLA-SE induced HAI titers ≥ 1:40 in more than 66% of the subjects45. In the direct comparison of the HAI titers against the homologous virus strains, MF59 gave rise to stronger responses but it is plausible that changing the experimental set-up e.g. assessing survival upon challenge with a heterologous strain would offer advantages to GLA-SE. The adjuvants IC31 and CAF01 were developed with the purpose of generating Th1 responses and both induce lower HAI titers in direct comparison to emulsion-based adjuvants. Paralleling the design of GLA-SE, it may be possible to benefit from their Th1-inducing ability e.g. by combining their individual immunomodulators into a stable emulsion formulation. Although the major purpose would be to augment the induction of neutralizing antibodies, it is important to maintain their ability to generate memory T cells that not only recognize conserved viral proteins, and thereby provide cross-protective responses, but also provide long-lived protection, which is of particular relevance for prophylactic vaccines. In this context, CAF01 in combination with trivalent inactivated vaccines (TIV) has previously been shown to promote significant T cell responses in the ferret model, which is considered to be the most accepted model of influenza disease, and importantly provided higher levels of protection compared to TIV alone and even TIV in a squalene emulsion46. A detailed characterization of the murine immune response afforded by a trivalent influenza vaccine administered in IC31 showed a very pronounced induction of IFN-γ that was maintained even 200 days after a single vaccination, at least in part, due to the formation of a vaccine depot at the injection site47,48. CAF01-based TB vaccines were also capable of inducing long-term protective responses with a high frequency of vaccine-induced CD4 T-cells retrieved more than one year after vaccination in mice4. The T-cells had a strong proliferative potential presumably facilitated by the specific physicochemical properties of the delivery vehicle (DDA) that supports a slow-release of antigen from the site of vaccination49. The immunogenicity results from human trials have so far been comparable for IC31 and CAF01, both of which were capable of inducing very persisting T cell responses as late as 2½ (IC31 8) and 3 years (CAF019) after vaccination. Assessing the phenotype of the vaccine-induced CD4 T-cells at this late-stage post vaccination showed that more than 50% of the cells co-secreted TNF-α and IL-2; a phenotype which has been associated with central memory T cells (Tcm)22. In our studies, all three adjuvants with the ability to induce highly significant protection against TB, induced phenotypic profiles very similar to what was observed previously in humans with a substantial proportion of IL-2/ TNF-α co-producers, most pronounced in the CAF01 group where app. 50% of the cells were of this phenotype. Indeed, of the relatively few Th1 cells induced by MF59 , the majority were IL-2/TNF-α co-producers, which were reported to represent the main population induced by MF59 adjuvanted influenza vaccine in children33, and this could explain the modest, albeit significant, protection induced by a MF59 -adjuvanted TB vaccine. There is an increasing amount of evidence that Tcm’s may play an important role in protection against TB. Most recently, the Kaufmann lab demonstrated that the protection afforded by a recombinant BCG primarily resided in the Tcm population50, but also in humans this phenotype was shown to be associated with responses in individuals capable of controlling infection51. Whether this profile directly mediates the superior TB disease control in humans, including vaccine-promoted protection, remains to be investigated. Whereas there is concordance between murine and human data in regards to effector and memory T cell phenotypes, the specific induction of IL-17 producing CD4 T cells seen upon vaccination with the CAF01 adjuvant in this and several other studies is not directly mirrored in humans. In a recent phase I study, vaccination with the TB fusion-protein H1 in CAF01 was unable to induce high levels of IL-17 following two immunizations in healthy individuals9. In this study, we compared the post vaccination splenic responses to post infection PBMC’s, and although cytokine levels from these two different compartments are not directly comparable, some key differences between the adjuvants became evident from the relative comparison between groups. The vaccine-induced IL-17 primed by CAF01 vaccination was strongly boosted upon M.tb. infection (Fig. 4a). In groups vaccinated with GLA-SE or IC31 , no apparent IL-17 responses were detected after vaccination in the splenocytes, but upon M.tb. infection both exhibited measurable IL-17 responses in the PBMC’s. It is possible that CAF01-based vaccines are also able to induce IL-17 in humans but at a very low and un-detectable level and that the response only becomes apparent after infection. This boosting was not observed with MF59 , H56 alone, or in the saline control group. The immunomodulator in CAF01, TDB, was shown to signal through the C-type lectin receptor Mincle in mice, which in turn elicits a unique Th1/Th17-biased immune response through the Syk-Card9 signaling pathway52. Mincle is expressed on a variety of cell types e.g. neutrophils, dendritic cells, and macrophages and

®

®

®

®

®

®

®

®

®

®

®

®

®

®

®

10

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

85

Appendix II

www.nature.com/scientificreports/ a human variant of the Mincle receptor has also been identified53. The recent study showing that TDB is capable of binding to human Mincle and activate APCs in vitro54 indicates that Mincle-activation may also be important for induction of immune response in humans. However, the use of more sensitive techniques and studying more immunologically competent compartments other than the PBMCs might be needed to detect Th17 responses in humans. It is well-established that CMI responses, and in particular IFN-γ producing CD4 T-cells, play a major role in protection against chlamydia infection55,56. In agreement with this, vaccination with all three Th1 adjuvants (IC31 , GLA-SE, CAF01) reduced bacterial shedding in the genital tract at both day 3 and 10 (Fig. 5b). Testing different adjuvants, Dr. Robert Brunham’s lab demonstrated a superior effect of an adjuvant-formulation giving rise to a combined IFN-γ /IL-17 response57, however in our studies we did not see a superior effect of the IL-17 inducing CAF01 compared to the sole Th1 inducing adjuvant GLA-SE. The Th2 inducing adjuvant MF59 in combination with MOMP had no protective effect, which is in line with previous reports showing reduced migration to the genital mucosa of a chlamydia-specific Th2 clone compared to its Th1 counterpart58. It has been shown that antibodies play an important role for providing early clearance of infection using the same chlamydia model as in this study59 as well as against reinfection60. Somewhat surprising, we could observe that three vaccinations with un-adjuvanted MOMP induced an antibody response comparable to those of adjuvanted MOMP, which may explain the significant protection observed at day 10 in the un-adjuvanted MOMP group. While it is generally accepted that a combination of humoral and cell mediated immunity is protective against chlamydia and that protection when using HI titers as influenza readout is mediated by a humoral response, the contribution of antibodies to protection against M.tb. is still debated. The comparison of the IgG1/2a titers (Fig. 2a) and TB bacterial burdens (Fig. 4b) indicates that protection against TB does not depend on antibody induction. While this fits with the general notion that a CMI Th1 response is crucial for TB protection, it does not elaborate on the possibility of a synergetic effect between cell mediated and humoral immunity. In particular, the induction of IgA titers by vaccination might be relevant as this isotype previously has been associated with protection against TB61. Currently, there are no vaccines building on the concept of inducing humoral immune responses in the clinical TB pipeline but this specific topic has generated interest from the funding bodies and it is possible that we will see new types of TB vaccines in the coming years. It should be noted that the head-to-head evaluation of a wide range of different adjuvants for their ability to enhance immunogenicity of different antigens is a challenging task. In this regard, it is highly satisfying to observe that each individual adjuvant produced similar antigen-specific immunological profiles for three unique vaccine antigens. Thus, the distribution of T-cell phenotypes are the same whether using e.g. a TB antigen or the chlamydia MOMP protein, MF59 induces IL-5 independent of the antigen of choice, and GLA-SE generates the highest IgG2a titers with antigens produced in E. coli as well as with an antigen produced by embryonated eggs. Despite this remarkable consistency in adjuvant activity across different pathogen targets, there are, however, some key limitations in the current study that need to be taken into account; In order to perform this head-to-head comparison a standardized protocol was employed across the three models. This included unifying the number of vaccinations, amount of antigen, dose of adjuvant, timing between vaccinations, administration route and resting period before challenge and subsequent determination of protection. Nevertheless, all of these factors would presumably have a different optimum for different combinations of adjuvants and antigens, implying that some of these factors might not have been optimal for the tested adjuvants, hence preventing the adjuvants to exert their full adjuvanticity in the head-to-head comparison. However, due to the consistency in the immunological profiles of the different adjuvants across antigens, which are in line with the previously published human data, we are confident that these results could inform rational selection of suitable adjuvants to mount desired protective immune responses. To our knowledge, this is the first report of a comprehensive non-clinical comparison of clinically tested/ practiced adjuvants from different stakeholders, including academia and the private sector, and it is our hope that results obtained from this study will provide a platform for how adjuvants should be compared and thereby help accelerating adjuvant research and development.

®

®

®

‡ˆ‡”‡…‡• 1. Giefing, C. et al. Discovery of a novel class of highly conserved vaccine antigens using genomic scale antigenic fingerprinting of pneumococcus with human antibodies. J Exp Med. 205, 117–131 (2008). 2. Rappuoli, R., Pizza, M., Del Giudice, G. & De Gregorio, E. Vaccines, new opportunities for a new society. Proc Natl Acad Sci USA 111, 12288–12293 (2014). 3. Dorrington, M. G. & Bowdish, D. M. Immunosenescence and novel vaccination strategies for the elderly. Front Immunol. 4, 171 (2013). 4. Lindenstrom, T. et al. Tuberculosis subunit vaccination provides long-term protective immunity characterized by multifunctional CD4 memory T cells. J Immunol. 182, 8047–8055 (2009). 5. Khurana, S. et al. Vaccines with MF59 adjuvant expand the antibody repertoire to target protective sites of pandemic avian H5N1 influenza virus. Sci Transl Med. 2, 15ra15 (2010). 6. Dietrich, J., Andreasen, L. V., Andersen, P. & Agger, E. M. Inducing dose sparing with inactivated polio virus formulated in adjuvant CAF01. PLoS One. 9, e100879 (2014). 7. Marrack, P., McKee, A. S. & Munks, M. W. Towards an understanding of the adjuvant action of aluminium. Nat Rev Immunol. 9, 287–293 (2009). 8. van Dissel, J. T. et al. Ag85B-ESAT-6 adjuvanted with IC31 promotes strong and long-lived Mycobacterium tuberculosis specific T cell responses in naive human volunteers. Vaccine. 28, 3571–3581 (2010). 9. van Dissel, J. T. et al. A novel liposomal adjuvant system, CAF01, promotes long-lived Mycobacterium tuberculosis-specific T-cell responses in human. Vaccine. 32, 7098–7107 (2014). 10. Harandi, A. M., Davies, G. & Olesen, O. F. Vaccine adjuvants: scientific challenges and strategic initiatives. Expert Rev Vaccines. 8, 293–298 (2009). 11. Redford, P. S. et al. Enhanced protection to Mycobacterium tuberculosis infection in IL-10-deficient mice is accompanied by early and enhanced Th1 responses in the lung. Eur J Immunol. 40, 2200–2210 (2010).

11

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

86

Appendix II

www.nature.com/scientificreports/ 12. Peng, Y. et al. Interleukin-22 promotes T helper 1 (Th1)/Th17 immunity in chlamydial lung infection. Mol Med. 20, 109–119 (2014). 13. Ohmit, S. E., Petrie, J. G., Cross, R. T., Johnson, E. & Monto, A. S. Influenza hemagglutination-inhibition antibody titer as a correlate of vaccine-induced protection. J Infect Dis. 204, 1879–1885 (2011). 14. Follmann, F. et al. Antigenic profiling of a Chlamydia trachomatis gene-expression library. J Infect Dis. 197, 897–905 (2008). 15. Agger, E. M. et al. Cationic liposomes formulated with synthetic mycobacterial cordfactor (CAF01): a versatile adjuvant for vaccines with different immunological requirements. PLoS One. 3, e3116 (2008). 16. Aagaard, C. et al. Protection and polyfunctional T cells induced by Ag85B-TB10.4/IC31 against Mycobacterium tuberculosis is highly dependent on the antigen dose. PLoS One. 4, e5930 (2009). 17. Windish, H. P. et al. Protection of mice from Mycobacterium tuberculosis by ID87/GLA-SE, a novel tuberculosis subunit vaccine candidate. Vaccine. 29, 7842–7848 (2011). 18. Davidsen, J. et al. Characterization of cationic liposomes based on dimethyldioctadecylammonium and synthetic cord factor from M. tuberculosis (trehalose 6,6′ -dibehenate)-a novel adjuvant inducing both strong CMI and antibody responses. Biochim Biophys Acta. 1718, 22–31 (2005). 19. Calabro, S. et al. The adjuvant effect of MF59 is due to the oil-in-water emulsion formulation, none of the individual components induce a comparable adjuvant effect. Vaccine. 31, 3363–3369 (2013). 20. Hekele, A. et al. Rapidly produced SAM((R)) vaccine against H7N9 influenza is immunogenic in mice. Emerg Microbes Infect. 2, e52 (2013). 21. Hoang, T. et al. ESAT-6 (EsxA) and TB10.4 (EsxH) based vaccines for pre- and post-exposure tuberculosis vaccination. PLoS One. 8, e80579 (2013). 22. Lindenstrom, T., Knudsen, N. P., Agger, E. M. & Andersen, P. Control of chronic mycobacterium tuberculosis infection by CD4 KLRG1- IL-2-secreting central memory cells. J Immunol. 190, 6311–6319 (2013). 23. Lindenstrom, T. et al. Vaccine-induced th17 cells are maintained long-term postvaccination as a distinct and phenotypically stable memory subset. Infect Immun. 80, 3533–3544 (2012). 24. Caproni, E. et al. MF59 and Pam3CSK4 boost adaptive responses to influenza subunit vaccine through an IFN type I-independent mechanism of action. J Immunol. 188, 3088–3098 (2012). 25. Orr, M. T. et al. Immune subdominant antigens as vaccine candidates against Mycobacterium tuberculosis. J Immunol. 193, 2911–2918 (2014). 26. Geeraedts, F. et al. Whole inactivated virus influenza vaccine is superior to subunit vaccine in inducing immune responses and secretion of proinflammatory cytokines by DCs. Influenza Other Respir Viruses. 2, 41–51 (2008). 27. Daifalla, N. S., Bayih, A. G. & Gedamu, L. Immunogenicity of Leishmania donovani iron superoxide dismutase B1 and peroxidoxin 4 in BALB/c mice: the contribution of Toll-like receptor agonists as adjuvant. Exp Parasitol. 129, 292–298 (2011). 28. O’Meara, C. P. et al. Immunization with a MOMP-based vaccine protects mice against a pulmonary Chlamydia challenge and identifies a disconnection between infection and pathology. PLoS One. 8, e61962 (2013). 29. Goff, P. H. et al. Adjuvants and immunization strategies to induce influenza virus hemagglutinin stalk antibodies. PLoS One. 8, e79194 (2013). 30. Seder, R. A., Darrah, P. A. & Roederer, M. T-cell quality in memory and protection: implications for vaccine design. Nat Rev Immunol. 8, 247–258 (2008). 31. Fabbiani, M. et al. HIV-infected patients show impaired cellular immune response to influenza vaccination compared to healthy subjects. Vaccine. 31, 2914–2918 (2013). 32. Galli, G. et al. Adjuvanted H5N1 vaccine induces early CD4+ T cell response that predicts long-term persistence of protective antibody levels. Proc Natl Acad Sci USA. 106, 3877–3882 (2009). 33. Zedda, L. et al. Dissecting the immune response to MF59-adjuvanted and nonadjuvanted seasonal influenza vaccines in children less than three years of age. Pediatr Infect Dis J. 34, 73–78 (2015). 34. Ott, G., Barchfeld, G. L. & Van Nest, G. Enhancement of humoral response against human influenza vaccine with the simple submicron oil/water emulsion adjuvant MF59. Vaccine. 13, 1557–1562 (1995). 35. Vesikari, T. et al. Oil-in-water emulsion adjuvant with influenza vaccine in young children. N Engl J Med. 365, 1406–1416 (2011). 36. Reed, S. G., Orr, M. T. & Fox, C. B. Key roles of adjuvants in modern vaccines. Nat Med. 19, 1597–1608 (2013). 37. Galli, G. et al. Fast rise of broadly cross-reactive antibodies after boosting long-lived human memory B cells primed by an MF59 adjuvanted prepandemic vaccine. Proc Natl Acad Sci USA. 106, 7962–7967 (2009). 38. Davenport, F. M., Hennessy, A. V. & Askin, F. B. Lack of adjuvant effect of AlPO4 on purified influenza virus hemagglutinins in man. J Immunol. 100, 1139–1140 (1968). 39. Tetsutani, K. & Ishii, K. J. Adjuvants in influenza vaccines. Vaccine. 30, 7658–7661 (2012). 40. Bernstein, D. I. et al. Effects of adjuvants on the safety and immunogenicity of an avian influenza H5N1 vaccine in adults. J Infect Dis. 197, 667–675 (2008). 41. O’Hagan, D. T., Rappuoli, R., De Gregorio, E., Tsai, T. & Del Giudice, G. MF59 adjuvant: the best insurance against influenza strain diversity. Expert Rev Vaccines. 10, 447–462 (2011). 42. Sridhar, S. et al. Cellular immune correlates of protection against symptomatic pandemic influenza. Nat Med. 19, 1305–1312 (2013). 43. Wilkinson, T. M. et al. Preexisting influenza-specific CD4+ T cells correlate with disease protection against influenza challenge in humans. Nat Med. 18, 274–280 (2012). 44. Clegg, C. H. et al. Adjuvant solution for pandemic influenza vaccine production. Proc Natl Acad Sci USA. 109, 17585–17590 (2012). 45. Treanor, J. J. et al. Evaluation of safety and immunogenicity of recombinant influenza hemagglutinin (H5/Indonesia/05/2005) formulated with and without a stable oil-in-water emulsion containing glucopyranosyl-lipid A (SE+ GLA) adjuvant. Vaccine. 31, 5760–5765 (2013). 46. Martel, C. J. et al. CAF01 potentiates immune responses and efficacy of an inactivated influenza vaccine in ferrets. PLoS One. 6, e22891 (2011). 47. Riedl, K., Riedl, R., von Gabain, A., Nagy, E. & Lingnau, K. The novel adjuvant IC31 strongly improves influenza vaccine-specific cellular and humoral immune responses in young adult and aged mice. Vaccine. 26, 3461–3468 (2008). 48. Schellack, C. et al. IC31, a novel adjuvant signaling via TLR9, induces potent cellular and humoral immune responses. Vaccine. 24, 5461–5472 (2006). 49. Christensen, D. et al. A cationic vaccine adjuvant based on a saturated quaternary ammonium lipid have different in vivo distribution kinetics and display a distinct CD4 T cell-inducing capacity compared to its unsaturated analog. J Control Release. 160, 468–476 (2012). 50. Vogelzang, A. et al. Central memory CD4+ T cells are responsible for the recombinant Bacillus Calmette-Guerin DeltaureC::hly vaccine’s superior protection against tuberculosis. J Infect Dis. 210, 1928–1937 (2014). 51. Day, C. L. et al. Functional capacity of Mycobacterium tuberculosis-specific T cell responses in humans is associated with mycobacterial load. J Immunol. 187, 2222–2232 (2011). 52. Schoenen, H. et al. Cutting edge: Mincle is essential for recognition and adjuvanticity of the mycobacterial cord factor and its synthetic analog trehalose-dibehenate. J Immunol. 184, 2756–2760 (2010). 53. Flornes, L. M. et al. Identification of lectin-like receptors expressed by antigen presenting cells and neutrophils and their mapping to a novel gene complex. Immunogenetics. 56, 506–517 (2004).

12

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

87

Appendix II

www.nature.com/scientificreports/ 54. Ostrop, J. et al. Contribution of MINCLE-SYK Signaling to Activation of Primary Human APCs by Mycobacterial Cord Factor and the Novel Adjuvant TDB. J Immunol. 195, 2417–2428 (2015). 55. Brunham, R. C. et al. The epidemiology of Chlamydia trachomatis within a sexually transmitted diseases core group. J Infect Dis. 173, 950–956 (1996). 56. Igietseme, J. U. et al. Resolution of murine chlamydial genital infection by the adoptive transfer of a biovar-specific, Th1 lymphocyte clone. Reg Immunol. 5, 317–324 (1993). 57. Yu, H. et al. Chlamydia muridarum T-cell antigens formulated with the adjuvant DDA/TDB induce immunity against infection that correlates with a high frequency of gamma interferon (IFN-gamma)/tumor necrosis factor alpha and IFN-gamma/interleukin-17 double-positive CD4+ T cells. Infect Immun. 78, 2272–2282 (2010). 58. Hawkins, R. A., Rank, R. G. & Kelly, K. A. A Chlamydia trachomatis-specific Th2 clone does not provide protection against a genital infection and displays reduced trafficking to the infected genital mucosa. Infect Immun. 70, 5132–5139 (2002). 59. Olsen, A. W., Follmann, F., Erneholm, K., Rosenkrands, I. & Andersen, P. Protection Against Chlamydia trachomatis Infection and Upper Genital Tract Pathological Changes by Vaccine-Promoted Neutralizing Antibodies Directed to the VD4 of the Major Outer Membrane Protein. J Infect Dis. 212, 978–989 (2015). 60. Morrison, R. P., Feilzer, K. & Tumas, D. B. Gene knockout mice establish a primary protective role for major histocompatibility complex class II-restricted responses in Chlamydia trachomatis genital tract infection. Infect Immun. 63, 4661–4668 (1995). 61. Balu, S. et al. A novel human IgA monoclonal antibody protects against tuberculosis. J Immunol. 186, 3113–3119 (2011). 62. Petrovsky, N. & Aguilar, J. C. Vaccine adjuvants: current state and future trends. Immunol Cell Biol. 82, 488–496 (2004). 63. Centers for Disease Control and Prevention. Vaccine excipient and media summary in Epidemiology and prevention of vaccinepreventable diseases. 13th edn (eds J. Hamborsky, A. Kroger & S. Wolfe) App. B-7 (Centers for Disease Control and Prevention, 2015). 64. Podda, A. & Del Giudice, G. MF59-adjuvanted vaccines: increased immunogenicity with an optimal safety profile. Expert Rev Vaccines. 2, 197–203 (2003). 65. El Sahly, H. MF59 ; as a vaccine adjuvant: a review of safety and immunogenicity. Expert Rev Vaccines. 9, 1135–1141 (2010). 66. O’Hagan, D. T., Ott, G. S., De Gregorio, E. & Seubert, A. The mechanism of action of MF59 - an innately attractive adjuvant formulation. Vaccine. 30, 4341–4348 (2012). 67. Coler, R. N. et al. A synthetic adjuvant to enhance and expand immune responses to influenza vaccines. PLoS One. 5, e13677 (2010). 68. Werninghaus, K. et al. Adjuvanticity of a synthetic cord factor analogue for subunit Mycobacterium tuberculosis vaccination requires FcRgamma-Syk-Card9-dependent innate immune activation. J Exp Med. 206, 89–97 (2009). 69. Agger, E. M. et al. Protective immunity to tuberculosis with Ag85B-ESAT-6 in a synthetic cationic adjuvant system IC31. Vaccine. 24, 5452–5460 (2006).

™

…‘™Ž‡†‰‡‡–• We appreciate the excellent technical assistance provided by Sharmila Subratheepam, Patricia Grenés, Linda Christensen, Merete Henriksen and the staff at the experimental animal facilities at Statens Serum Institut. This work was supported by the European Commission through contract FP7-HEALTH-2011.1.4-4-280873 (ADITEC).

—–Š‘”‘–”‹„—–‹‘• A.O., R.B., C.B., F.F., R.C., A.M., U.D., E.D.G., R.R., A.H., P.A., E.A. and N.P.K. contributed to the design of the experiments, N.P.K. performed and evaluated the tuberculosis experiments supervised by P.A. and E.A. R.B. performed parts of the experiments related to T.B. A.O. performed and evaluated the chlamydia associated experiments supervised by F.F. C.B., U.D., D.C. and A.B. performed and evaluated the influenza experiments supervised by E.D.G. and R.R. A.M., R.C., C.F., U.D. and Y.Z. contributed reagents/materials/analysis tools and expertise. N.P.K. and E.A. wrote the manuscript. A.H. contributed to the study design and writing the manuscript. F.F., A.O., E.D.G., U.D., C.B., C.F. and R.C. contributed to the manuscript. All authors reviewed the manuscript.

††‹–‹‘ƒŽ ˆ‘”ƒ–‹‘ Supplementary information accompanies this paper at http://www.nature.com/srep Competing financial interests: PA is coinventor on a patent application covering the use of H56 and CAF01 in vaccines. All rights have been assigned to Statens Serum Institut, a Danish nonprofit governmental institute. NPK, RB, EA, PA, FF and AO are employees at Statens Serum Institut. RC and CF are employees at the Infectious Disease Research Institute. CB, AB, DC, UD, EDG and RR are employees of GSK vaccines. AM is an employee of Valneva Austria GmbH. How to cite this article: Knudsen, N. P. H. et al. Different human vaccine adjuvants promote distinct antigenindependent immunological signatures tailored to different pathogens. Sci. Rep. 6, 19570; doi: 10.1038/ srep19570 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

13

SCIENTIFIC REPORTSȁͼǣͷͿͻͽͶȁ ǣͷͶǤͷͶ͹;Ȁ•”‡’ͷͿͻͽͶ

88

Appendix II

Figure 1S. icFACS analysis of vaccine responses measured by cytokine expression in HA specific CD4 T-cells. Mice were vaccinated 3 times s.c. with 1 μg of HA formulated in Alum, MF59® CAF01, GLA-SE or IC31® with 3 week intervals. Splenocytes from four individual mice, isolated two weeks after the third immunization, were stimulated with 1 μg/ml of recombinant HA O/N + 4h in the presence of BFA. Cells were gated as follows: live > singlets > lymphocytes > CD3+ > CD4+ versus CD8+. Cytokine-producing cells (IFN-γ, IL-17 or IL-4/IL-13) within the CD4+CD44high population were measured. Bar charts show the frequency of each cytokine subset being CD44high out of the total CD4 T cell population. Bars represent the mean + SEM of four individual mice. The background from non-stimulated samples have been subtracted.

Figure 2S. An icFACS analysis of vaccine responses measured by cytokine co-expression in H56 or MOMP-specific CD4 T-cells. Mice were vaccinated 3 times s.c. with 5 μg of H56 or MOMP formulated in Alum, MF59® CAF01, GLA-SE or IC31® with 3 week intervals. Splenocytes from four individual mice, isolated two weeks after the third immunization, were stimulated with 2 μg/ml of recombinant H56 or MOMP for 1h + 5h in the presence of BFA. Cells were gated as follows: singlets > lymphocytes > CD4+ versus CD8+. Cytokine-producing cells (IFN-γ, TNF-α, and IL-2) within the CD4+CD44high population were divided into seven distinct subpopulations based on Boolean gating and are shown as bar and pie charts. Bar charts show the frequency of each cytokine subset being CD44high out of the total CD4 T cell population. Bars represent the mean + SEM of four individual mice. Pie charts show the relative contribution of each cytokine subpopulation to the Ag-responsive CD4 T cell population. The background from non-stimulated samples have been subtracted. The experiment was repeated with similar results.

89

Appendix II

Supplementary Figure S1

90

Appendix II

Supplementary Figure S2

91

Appendix III

Appendix III Manuscript III

An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology

Helena Strand Clemmensen* Niels Peter H. Knudsen* Erik Michael Rasmussen Jessica Winkler Ida Rosenkrands Ahmad Ahmad Troels Lillebaek David R. Sherman Peter Andersen Claus Aagaard *Co-first authors

Submitted to Scientific reports Dec 2016

93

www.nature.com/scientificreports

OPEN

received: 17 January 2017 accepted: 22 March 2017 Published: 24 April 2017

An attenuated Mycobacterium tuberculosis clinical strain with a defect in ESX-1 secretion induces minimal host immune responses and pathology Helena Strand Clemmensen1,*, Niels Peter Hell Knudsen1,*, Erik Michael Rasmussen2, Jessica Winkler3, Ida Rosenkrands1, Ahmad Ahmad1, Troels Lillebaek2, David R. Sherman3, Peter Lawætz Andersen1 & Claus Aagaard1 Although Mycobacterium tuberculosis (M.tb) DK9897 is an attenuated strain, it was isolated from a patient with extrapulmonary tuberculosis and vaccination with a subunit vaccine (H56) induced poor protection against it. Both attenuation and lack of protection are because M.tb DK9897 cannot secrete the EsxA virulence factor nor induce a host response against it. Genome sequencing identified a frameshift mutation in the eccCa1 gene. Since the encoded EccCa1 protein provides energy for ESX-1 secretion, it suggested a defect in the ESX-1 type VII secretion system. Genetic complementation with a plasmid carrying the M.tb H37Rv sequence of eccCa1-eccCb1-pe35 re-established EsxA secretion, host specific EsxA T-cell responses, and increased strain virulence. The ESX-1 secretion defect prevents several virulence factors from being functional during infection and therefore attenuates M.tb. It precludes specific T-cell responses against strong antigens and we found very little in vivo cytokine production, gross pathology or granuloma formation in lungs from M.tb DK9897 infected animals. This coincides with M.tb DK9897 being unable to disrupt the phagosome membrane and make contact to the cytosol. The facultative intracellular pathogen Mycobacterium tuberculosis (M.tb) is one of the most devastating human pathogens in the world. Although the incidence rate of tuberculosis (TB) is on a steady decline, TB remains one of the major public health problems, with 9 million new cases and the deaths of around 1.5 million people each year1. Pathogenic mycobacteria have the ability to resist killing by phagocytic cells of the host immune system. Phagocytic cells degrade invading microbes by engulfment inside a vacuole or phagosome that progressively acidifies and accumulates hydrolytic properties. Mycobacterium marinum and Mycobacterium avium are opportunistic pathogens that can survive and proliferate in host cells because they are capable of blocking accumulation of the vacuolar H+-ATPase on the mycobacterial vacuole and prevent delivery of the lysosomal protease cathepsin D2–4. M.tb inhibits phagosomal maturation and the endosomal-lysosomal degradation pathway and can, therefore, manipulate antigen presentation5. Very recent data have shown that EsxH (TB10.4) and EsxL, substrates of the ESX-3 and ESX-5 type VII secretion systems respectively, are part of this control. EsxH prevents the ability of antigen presenting cells to activate CD4 T cells by inhibiting the endosomal sorting complex required for transport (ESCRT) machinery and EsxL inhibits major histocompatibility complex class II (MHC-II) expression by enhancing the methylation of a transactivator loci6,7. All these defense mechanisms reduce epitope presentation on the surface of infected cells and subsequently affect the adaptive immune response in terms of delayed recruitment of T cells to the site of infection and suboptimal T cell activation of infected cells8,9. In addition, virulent 1

Department of Infectious Disease Immunology, Statens Serum Institut, DK-2300, Copenhagen, Denmark. International Reference Laboratory of Mycobacteriology, Statens Serum Institut, DK-2300, Copenhagen, Denmark. 3 Center for Infectious Disease Research, Seattle, Washington, 98109, USA. *These authors contributed equally to this work. Correspondence and requests for materials should be addressed to C.A. (email: [email protected]) 2

Scientific Reports | 7:46666 | DOI: 10.1038/srep46666

1

www.nature.com/scientificreports/ M.tb also exploits the ESX-1 type VII secretion system to secrete virulence factors that are involved in survival and spreading of the pathogen via interactions with the host cells10,11. Comparative analysis of genomes from attenuated M. bovis BCG strains and pathogenic mycobacterial species identified the main chromosomal ESX-1 locus, containing region of difference 1 (RD1) genes, and showed that this region encodes the immunodominant T cell antigens EsxA (ESAT-6) and EsxB (CFP-10)12–14. RD1 gene complementation not only re-established the expression and secretion of EsxA and EsxB but also increased the virulence of M. bovis BCG15. Deleting single genes in the M.tb ESX-1 locus, encoding core components of the ESX-1 apparatus, blocked EsxA and EsxB secretion and attenuated the bacillus in cellular and animal models of infection16. After synthesis, EsxA and EsxB form a heterodimer inside the mycobacterial cytoplasm. EsxB has a dual function as a chaperone and secretion partner, holding the sequence needed for secretion of the dimer via ESX-1. Once secreted, the heterodimer dissociates at low pH in the acidic environment of the phagosome. EsxA has been reported to be involved in numerous biological processes relevant for virulence including; initiation of granuloma formation17, phagosome maturation18,19, apoptosis through caspase activation20 and induction of membrane damage and phagosomal disruption21. Two most recent studies demonstrate that EsxA is not directly responsible for membrane lysis, rather this activity is attributed to ESX-1 in concert with phthiocerol dimycocerosates (DIMs) and is contact dependent, which results in gross membrane disruptions rather than pore formation22,23. ESX-1 has also been shown to be involved in host cell immune modulation24,25. The isolation of an M.tb strain unable to secrete EsxA from a Danish patient with extrapulmonary TB was unexpected because of its importance as a virulence factor for M.tb. Here we report on the initial characterization of this clinical isolate and the responses it induces in the host during infection.

Results

M.tb DK9897 belongs to a lineage with few members.  Since M.tb strains from different lineages can induce variable host responses in macrophages, cell lines and mouse models26–28 the genetic diversity among lineages could potentially influence the protective efficacy of TB vaccines. We, therefore, set out to test the ability of the H56 vaccine29 to protect against aerosol infection with clinical M.tb isolates. H56 is a fusion protein of the M.tb proteins Ag85B, EsxA, and Rv2660c. The M.tb DK9897 isolate was one of six clinical isolates selected from the M.tb strain collection at the International Reference Laboratory of Mycobacteriology harboring >​ten thousands of clinical isolates cultured from individuals infected with mycobacteria. In our selection, we prioritized lineage coverage and sequence diversity but for safety reasons, we only included strains that were susceptible to standard anti-tuberculous treatment. M.tb DK9897 was originally isolated in February of 1998 from the cervical pus of a 92-year-old woman with tuberculous lymphadenitis. The isolate was susceptible to isoniazid, rifampicin, ethambutol, pyrazinamide and streptomycin. To investigate if M.tb DK9897 was part of a larger subgroup of mycobacterial isolates we genotyped the M.tb DK9897 strain along with the laboratory-adapted strains M.tb Erdman and M.tb H37Rv and an isolate belonging to the large Beijing family, M.tb DK9417. One quick and reliable marker commonly used for M.tb genotyping is the mycobacterial interspersed repetitive units (MIRU), located in variable number tandem repeats (VNTR) found at multiple loci scattered throughout the genome. The MIRU-VNTR genotyping data (Supplementary Table S1) was uploaded to the MIRU-VNTRplus database and a phylogenetical analysis was performed using a neighbor-joining algorithm and categorical distance coefficient using our four 24-locus MIRU-VNTR typing data and all isolates available in the MIRU-VNTR+​database as input. The results show that DK9897 does not belong to any of the established lineages but is a member of a new lineage with very few members that cluster between the M.tb Erdman and M.tb H37Rv (Fig. 1). The H56 subunit vaccine induced poor protection against M.tb DK9897.  Groups of mice were vaccinated with either H56 fusion protein in CAF01 adjuvant or BCG. Three weeks after the third H56 vaccination, the vaccine-specific T cell responses were investigated by flow cytometry in splenocytes. CD4 T cells that after stimulation with a pool of the three single antigens in H56 produced at least one of the cytokines IFN-γ​, tumor necrosis factor-alpha (TNF-α​or IL-2 were taken as being vaccine specific. Overall, the H56 vaccine-induced immune responses, in terms of T cell polyfunctionality and frequency of vaccine-specific cells, was at a similar level to what we have seen before30. On average 58.073 ( +​/− ​8704) of the spleen CD4 T cells recognized the vaccine (Fig. 2a). More than 60% of these were IL-2 producing T cells with high proliferative capacity and potential to become memory cells (Fig. 2b). The strongest T cell response was against Ag85B followed by EsxA and then a weak recognition of Rv2660c (Fig. 2c). Thus, we had a robust high-quality T cell response in the H56 vaccinated animals that previously have resulted in efficient protection against M.tb30. The expression of the Rv2660c gene has been questioned since no Rv2660c gene expression was found in cultures starved for nutrition’s nor in mice infected with M.tb for up to 31 days with M.tb, instead strong expression of a small RNA on the opposite strand was found31. Yet, in two studies investigating human immune responses, more than 60% of persons with LTBI had T cell responses against Rv2660c and half of the TB patients responded specifically against a peptide pool covering the Rv2660c sequence32,33. Furthermore, immune mice and non-human primates have clear recall responses to Rv2660c emphasizing the availability of this antigen for recognition by primed T cells30,34. Thus, it is possible that the Rv2660c gene expression is tightly controlled and transcription only takes place under certain conditions such as e.g. immune stress. Six weeks after the third vaccination all mice were either aerosol challenged with M.tb Erdman or M.tb DK9897. The challenge dose was approximately 100 viable bacilli per animal for all infections. To compare M.tb strain virulence and vaccine protective efficacy the number of mycobacteria was determined in individual lungs from M.tb Erdman (Fig. 2d) and M.tb DK9897 infected (Fig. 2e) animals. Combined results from two independent experiments are shown for both strains (n =​ 12–16/grp). Within the six-week infection period, M.tb Erdman reaches an average of approximately one million bacteria per lung (Log10 =​  6.05  +​/−​  0.19 Scientific Reports | 7:46666 | DOI: 10.1038/srep46666

2

www.nature.com/scientificreports/

L5

L6

L1

L4

L4 L2

L3

DK9417 Erdman H37Rv DK9897 Figure 1.  The unique M.tb DK9897 isolate is related to laboratory-adapted strains. Phylogenetic distribution of M.tb isolates based on MIRU-VNTR typing. A radial tree was constructed based on the neighbor-joining algorithm and categorical distance coefficient using the 24-locus MIRU-VNTR typing data for DK9897 and all isolates available in the MIRU-VNTR+​database as input.

SEM) whereas M.tb DK9897 grew to a level approximately 10-fold lower with less than 100.000 colony forming bacteria (CFU) per lung (Log10 =​  4.93  +​/−​ 0.10 SEM) in non-vaccinated animals. In terms of protection, M.bovis BCG and H56 both protected quite effectively against the M.tb Erdman challenge with no statistical difference in protection. However, in M.tb DK9897 challenged animals H56 vaccination only reduced the CFU load 5-fold compared to non-vaccinated animals whereas the M.bovis BCG vaccinated group had a 50-fold reduction in bacterial number, and a significant difference (p