Exposure to Mycobacterium avium can modulate ... - CiteSeerX

Exposure to Mycobacterium avium can modulate established immunity against Mycobacterium tuberculosis infection generated by Mycobacterium bovis BCG vaccination David K. Flaherty,* Bridget Vesosky,* Gillian L. Beamer,*,† Paul Stromberg,† and Joanne Turner*,1 Departments of *Internal Medicine, Division of Infectious Diseases, Center for Microbial Interface Biology, and † Veterinary Biosciences, The Ohio State University, Columbus, Ohio, USA

Abstract: Mycobacterium bovis bacille Calmette Guerin (BCG), the current vaccine against infection with Mycobacterium tuberculosis, offers a variable, protective efficacy in man. It has been suggested that exposure to environmental mycobacteria can interfere with the generation of BCGspecific immunity. We hypothesized that exposure to environmental mycobacteria following BCG vaccination would interfere with established BCG immunity and reduce protective efficacy, thus modeling the guidelines for BCG vaccination within the first year of life. Mice were vaccinated with BCG and subsequently given repeated oral doses of live Mycobacterium avium to model exposure to environmental mycobacteria. The protective efficacy of BCG with and without subsequent exposure to M. avium was determined following an aerogenic challenge with M. tuberculosis. Exposure of BCGvaccinated mice to M. avium led to a persistent increase in the number of activated T cells within the brachial lymph nodes but similar T cell activation profiles in the lungs following infection with M. tuberculosis. The capacity of BCG-vaccinated mice to reduce the bacterial load following infection with M. tuberculosis was impaired in mice that had been exposed to M. avium. Our data suggest that exposure to environmental mycobacteria can negatively impact the protection afforded by BCG. These findings are relevant for the development of a vaccine administered in regions with elevated levels of environmental mycobacteria. J. Leukoc. Biol. 80: 1262–1271; 2006. Key Words: environmental mycobacteria 䡠 interference 䡠 tuberculosis 䡠 murine

INTRODUCTION Tuberculosis remains one of the most significant sources of human morbidity and mortality caused by an infectious agent, with an estimated 16 million individuals diagnosed with active disease and almost 2 million deaths from tuberculosis each year [1, 2]. Perhaps more remarkable is that the incidence of tuberculosis remains elevated in countries that have had a 1262

Journal of Leukocyte Biology Volume 80, December 2006

rigorous vaccination program in place for decades [3]. The current vaccine against Mycobacterium tuberculosis infection, Mycobacterium bovis bacille Calmette Guerin (BCG), is administered to over 150 countries worldwide [4]. Despite such a thorough global vaccination campaign, protective efficacy of BCG can range anywhere from below zero to over 80% [5]. A more fundamental understanding of why the efficacy of BCG can be so variable among different study populations could better serve vaccine design by leading to more efficacious vaccination strategies, which would reduce the global incidence of tuberculosis significantly. Several hypotheses have been proposed to explain why BCG efficacy has been so variable [6, 7], and one of the most widely supported and experimentally tested has been that of interference by environmental mycobacteria [8 –11]. Support for this theory has come from data showing that exposure to environmental mycobacteria is prevalent in those countries where BCG efficacy is at its lowest [12]. Animal models have been used to determine how environmental mycobacteria can interfere with BCG efficacy, whereby animals have been given environmental mycobacteria prior to BCG vaccination and subsequently challenged with virulent M. tuberculosis. These studies have led to two independent models. The first model suggests that exposure to environmental mycobacteria can mask the effect of BCG by inducing a prior level of protection, making the protective efficacy of BCG indistinguishable from that measured in the nonvaccinated population [11, 13]. The second model advocates that immunity generated against environmental mycobacteria can cross-react with BCG antigens sufficiently to drive early clearance of BCG from the host [8]. In doing so, BCG-specific, protective immunity, namely one that can protect against M. tuberculosis infection, is not generated. The current World Health Organization (WHO) guidelines recommend that BCG is given within the first year of life [4], and in most developing countries, it is recommended that BCG is administered at birth [14]. In countries monitored by WHO, the success rate of administering BCG vaccination at birth is estimated at between 40% and 95%, and the majority of

1 Correspondence: The Ohio State University, 420 West 12th Avenue, Columbus, OH 43210, USA. E-mail: [email protected] Received June 21, 2006; revised July 28, 2006; accepted August 14, 2006; doi: 10.1189/jlb.0606407.

0741-5400/06/0080-1262 © Society for Leukocyte Biology

countries exceed 80% coverage [14]. Under these conditions, it is unlikely that significant exposure to environmental mycobacteria has occurred prior to vaccination. Based on this rationale, we designed experiments to determine how exposure to environmental mycobacteria after BCG vaccination could impact the generation of an established, protective immune response engendered by BCG. In addition, to more closely model environmental exposure in man, we used a live, lowdose oral delivery of environmental mycobacteria. Our studies demonstrate that exposure of BCG-vaccinated mice to live, environmental mycobacteria (Mycobacterium avium) had a negative impact on the protective efficacy normally afforded by BCG in response to infection with M. tuberculosis. These studies have particular relevance to the design and administration of vaccines in countries where exposure to environmental mycobacteria is prevalent.

MATERIALS AND METHODS Mice Specific, pathogen-free, female, 6- to 8-week-old C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, ME; Experiment 1) or Charles River Laboratories (Wilmington, MA; Experiment 2). Mice were acclimatized for at least 1 week prior to manipulation and supplied with sterilized water and chow ad libitum. Mice were maintained in micro-isolator cages for the entire experimental protocol. Prior to infection with M. tuberculosis, mice were moved to a Biosafety Level 3 facility and maintained as described above. The Ohio State University Institutional Laboratory Animal Care and Use Committee (Columbus) approved all experimental procedures.

Experimental groups Mice were divided into four groups: un-manipulated, BCG-vaccinated, M. avium-treated, or BCG-vaccinated and M. avium-treated. All treatment groups were subsequently infected with M. tuberculosis.

Cultivation of mycobacterial stocks M. bovis BCG Pasteur [American Type Culture Collection (ATCC), #35734] and M. tuberculosis Erdman (ATCC #35801) were obtained from ATCC (Manassas, VA), and M. avium Strain 104 was a generous gift from Dr. Andrea Cooper (Trudeau Institute, Saranac Lake, NY). Mycobacterial stocks were grown in Proskauer-Beck liquid medium (M. avium Strain 104) or ProskauerBeck liquid medium containing 0.05% Tween 80 (M. bovis BCG Pasteur and M. tuberculosis Erdman) to mid-log phase and frozen in aliquots at – 80°C.

BCG vaccination Mice were injected s.c. in the scruff of the neck with 100 ␮l M. bovis BCG Pasteur (1⫻106 CFU), diluted in sterile PBS. Mice were housed without further experimental manipulation for 6 weeks.

M. avium gavage

40-min exposure. The bacterial burden was assessed at various time-points postinfection by culturing serial dilutions of organ homogenates onto Middlebrook 7H11 agar (Becton Dickinson, San Jose, CA), supplemented with OADC (Becton Dickinson). Colonies were enumerated after 21 days incubation at 37°C. Data are expressed as the log10 value of the mean number of CFU recovered per organ (n⫽4 –5 mice). Lung homogenate was also plated onto supplemented 7H11 media containing clarithromycin (Abbott Laboratories, Abbott Park, IL, 2 ␮g/ml) and/or 2-thiophenecarboxylic acid hydrazide (TCH; Sigma Chemical Co., St. Louis, MO, 2 ␮g/ml) to exclude M. avium or BCG growth, respectively.

Lung cell isolation Mice were killed, and the lungs were cleared of blood via perfusion through the pulmonary artery with 10 ml PBS containing 50 U/ml heparin (Sigma Chemical Co.). Lungs were removed from the thoracic cavity and placed into 2 ml cold DMEM (500 ml, Mediatech, Herndon, VA), supplemented with 10% heatinactivated FBS (Atlas Biologicals, Ft. Collins, CO), 1% HEPES buffer (1 M, Sigma Chemical Co.), 1% L-glutamine (200 nM, Sigma Chemical Co.), 10 ml 100⫻ MEM nonessential amino acid solution (Sigma Chemcial Co.), 5 ml penicillin/streptomycin solution (50,000 U penicillin, 50 mg streptomycin, Sigma Chemical Co.), and 0.1% 2-ME (50 mM, Sigma Chemical Co.; complete DMEM). The lungs were diced with sterile razor blades and incubated for 30 min at 37°C with 4 ml complete DMEM containing collagenase XI (0.7 mg/ml, Sigma Chemical Co.) and type IV bovine pancreatic DNase (30 ␮g/ml, Sigma Chemical Co.). Complete DMEM (10 ml) was added to stop the enzymatic activity, and digested lungs were dispersed gently through a nylon screen to obtain a single-cell suspension. Residual RBC were lysed using ACK lysis buffer (0.15 M NH4Cl, 1 mM KHCO3), and cells were suspended at a working concentration in complete DMEM.

Lymph node (LN) isolation LN were harvested 2 weeks following the second and fifth gavage with M. avium. Brachial (draining LN for BCG vaccination), inguinal (nondraining LN for gavage or vaccination), axillary (nondraining LN for gavage or vaccination), and mesenteric (draining LN for gavage) LN were harvested from individual mice. LN were dispersed gently through a nylon screen to obtain a single-cell suspension in complete DMEM.

Flow cytometry Lung or LN cell suspensions were adjusted to 1 ⫻ 107 cells/ml in FACS buffer (deficient-RPMI, Irvine Scientific, Santa Ana, CA), supplemented with 0.1% sodium azide (Sigma Chemical Co.), and incubated on ice for 30 min. Cells (1⫻106) were labeled with 0.5 ␮g-specific fluorescent-labeled antibody for 20 min at 4°C in the dark followed by two washes with 200 ␮l FACS buffer. Samples were read on a Becton Dickinson LSRII flow cytometer, and data were analyzed using FACSDiva software (Becton Dickinson). Lymphocytes were gated according to their forward- and side-scatter profiles, and CD4 or CD8 T cells were identified by the presence of specific, fluorescent-labeled antibody in combination with CD3ε. Appropriate isotype controls were included in each experiment and used to set gates for analysis. Cell surface markers analyzed were FITC-labeled CD4 (GK1.5), PE-labeled CD45RB (Clone 16A), PerCPCy5.5-labeled CD3ε (Clone 145-2C11), allophycocyanin (APC)-labeled CD44 (Clone IM7), PE-Cy7-labeled CD95 (Clone Jo2), and APC-Cy7-labeled CD8a (Clone 53-6.7).

In vitro cell culture

Mice were administered 102 CFU of live M. avium (Strain 104) in 100 ␮l sterile PBS by gavage once every 2 weeks for a total of eight individual doses. Following the final dose, mice were rested for 6 weeks prior to any experimental manipulation. Eight gavages were delivered to model multiple exposures to environmental mycobacteria. The extended time-scale of the experimental design prevented the delivery of additional gavage.

M. tuberculosis infection Mice were infected aerogenically with a low dose of M. tuberculosis Erdman using the Glas-Col (Terre Haute, IN) inhalation exposure system. Briefly, the nebulizer compartment was filled with a suspension of bacteria calculated to deliver between 50 and 100 viable bacteria into the lung per mouse during a

Following infection with M. tuberculosis, single-cell suspensions were prepared from mice in each experimental group (un-manipulated, BCG, M. avium, and BCG/M. avium) at Days 14, 21, and 35 postinfection. Lung cells were diluted to a final concentration of 1 ⫻ 106 cells/ml in complete DMEM and cultured in 96-well tissue-culture plates with Con A (Sigma Chemical Co.), OVA (Sigma Chemical Co.), or M. tuberculosis culture-filtrate protein [CFP; provided by Dr. John Belisle (Colorado State University, Fort Collins, CO) under National Institutes of Health (NIH), National Institute of Allergy and Infectious Diseases (NIAID), Contract No. HHSN 266200400091C] in a humidified incubator at 37°C and 5% CO2. All antigens were used at a final concentration of 10 ␮g/ml. After 3 days incubation, cultures were frozen at –70°C for analysis by ELISA.

Flaherty et al. M. avium interferes with established BCG efficacy

1263

Quantification of IFN-␥ by ELISA Cell culture supernatants were assayed for the presence of IFN-␥ by ELISA. Antibodies were purchased from BD Biosciences (San Jose, CA). Briefly, the primary antibody (Clone R4-6A2) was incubated overnight in 96-well roundbottom Maxisorp plates (Nunc, Roskilde, Denmark) in 0.1 M sodium bicarbonate-coating buffer. Excess antibody was washed away using PBS containing 0.5% Tween 20 (PBS-T). The wells were blocked with 3% BSA (Sigma Chemical Co.) in PBS-T. The samples were dispensed in duplicate into the wells. The presence of cytokine was detected by the addition of a secondary biotinylated antibody (Clone XMG1.2), followed by streptavidin peroxidase (Invitrogen, Carlsbad, CA) and 3,3⬘,5,5,-tetramethylbenzidine substrate (Invitrogen).

Histology The right caudal lung lobe was isolated from each individual mouse and inflated with and stored in 10% formal-buffered saline. Lung tissue was processed, sectioned, and stained with H&E for light microscopy with lobe orientation designed to allow for maximum surface area of each lobe to be seen. Sections were examined by a board-certified veterinary pathologist without prior knowledge of the experimental groups and graded according to severity, granuloma size and number, and presence of pyknotic debris indicative of necrosis.

Statistical analysis Statistical significance was determined using Prism 4 software (GraphPad Software, San Diego, CA). The unpaired, two-tailed Student’s t-test was used

for two group comparisons. Multiple comparisons were analyzed using one-way ANOVA with Tukey’s post-test. Statistical significance was reported as *, P ⬍ 0.05; **, P ⬍ 0.01; or ***, P ⬍ 0.001.

RESULTS M. avium exposure alters the pattern of T cell activation in the draining LN of mice that have received M. bovis BCG C57BL/6 mice were vaccinated with M. bovis BCG and rested for 6 weeks prior to receiving a low-dose (102 CFU) gavage of live M. avium strain 104. Gavage with M. avium was repeated every 2 weeks for a period of 4 months (eight gavages in total). Two weeks after the second and fifth gavages with M. avium, groups of mice were killed, and LN tissue was removed for analysis by flow cytometry. T cell activation was assessed by expression of CD44 on the cell surface. Exposure to M. avium did not increase the number of activated CD4 or CD8 T cells within the mesenteric (Fig. 1, A and B), brachial (Fig. 1, C and D), inguinal (data not shown), or axillary (data not shown) LN when compared with unmanipulated mice. In contrast, BCG vaccination produced an increase in the number of acti-

Fig. 1. T cell activation within the draining LN in response to M. avium. Mice were s.c.-vaccinated with BCG or un-manipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. Two weeks following the second and fifth gavage, the mesenteric (A, B) and brachial (C, D) LN were isolated, and cells were analyzed by flow cytometry for CD3, CD4, and CD44 (A, C) or CD3, CD8, and CD44 (B, D). Data are expressed as the mean and SEM for five mice and are representative of two independent experiments.

1264


http://www.jleukbio.org

vated CD4 (Fig. 1C) and CD8 (Fig. 1D) T cells within the brachial LN. Increased numbers of CD44hi T cells remained detectable in the brachial LN up to 8 weeks after the initial vaccination (determined at the time-point of the second gavage) but declined over time (determined at the time-point of the fifth gavage). Gavage of BCG-vaccinated mice with M. avium led to a persistently increased number of activated CD4 and CD8 T cells in the brachial LN at the fifth gavage time-point (Fig. 1, C and D). Similar results for CD95 expression were observed (data not shown). As exposure to M. avium alone had no influence on T cell activation, the data suggest that the sustained number of activated CD4 and CD8 T cells within the brachial LN of mice was related directly to the influence that M. avium exposure had on an established, BCG-specific immunity.

Exposure to M. avium reduces the protective efficacy of BCG upon challenge with M. tuberculosis Mice from each of the four experimental groups (un-manipulated, BCG, BCG/M. avium, or M. avium only) were infected with M. tuberculosis via the respiratory route using a Glas-col inhalation device. The initial infectious innoculum was deter-

mined from four individual mice 1 day post infection (1.49 Log CFU⫾0.13 and 1.66 Log CFU⫾0.12 for Experiments 1 and 2, respectively). At 14, 21, and 35 days postinfection, five mice from each experimental group were killed, and the bacterial load was determined in the lung. BCG vaccination reduced the bacterial load within the lung greater than 1 Log, and such a reduction was maintained throughout the first 35 days of infection (Fig. 2A). Gavage of M. avium provided limited, early protection in the lungs against an infection with M. tuberculosis; however, this protection was not sustained or consistently reproducible between experiments (Fig. 2B). In contrast, gavage of BCG-vaccinated mice with M. avium interfered with BCG-protective efficacy, resulting in a significantly increased number of M. tuberculosis CFU within the lungs when compared with BCG-vaccinated mice after 35 days infection with M. tuberculosis (Fig. 2, C and D). CFU determination was performed using plates that had been supplemented with clarythromycin or TCH to exclude the detection of M. avium or BCG, respectively, which may be present within the lung. The data in Figure 2 are indicative of M. tuberculosis CFU only. Lung homogenates were analyzed further for the presence of M. avium or BCG in the lung using subtraction analysis from untreated 7H11 plates. Occasional

Fig. 2. Control of M. tuberculosis infection within the lungs. Mice were s.c.-vaccinated with BCG or un-manipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. After eight gavages with M. avium, mice were infected with M. tuberculosis by aerosol. M. tuberculosis CFU in the lung was determined by plating lung homogenate onto supplemented 7H11 agar plates and counting colonies after 21 days incubation at 37°C. Data are expressed as the mean and SEM for five mice and are representative of two independent experiments. Data are expressed comparing the control group with BCG (A), to M. avium (B), or to BCG/M. avium (C) or a comparison of all four groups at Day 35 post infection (D). Statistical significance was determined using a one-way ANOVA with Tukey’s post-test. *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001.


1265

CFU of BCG were found within the lungs of M. tuberculosisinfected mice; however, for the majority of animals, BCG CFU were below the level of detection (⬍1 Log, data not shown). Therefore, it was not possible to determine whether infection with M. avium had any influence on BCG persistence. In contrast, M. avium could be detected readily within the lungs of M. tuberculosis-infected mice (Fig. 3). Approximately 3 Log CFU M. avium was found consistently within the lungs of M. avium-gavaged mice up to 35 days post infection with M. tuberculosis, regardless of whether they had received prior vaccination with BCG or not. The quantity of M. avium did not change during early infection with M. tuberculosis; however, by Day 130 postinfection with M. tuberculosis, M. avium was no longer detectable within the lung (data not shown).

T cell activation profiles within the lung during infection with M. tuberculosis As exposure to M. avium altered the number of activated T cells within the brachial LN of BCG-vaccinated mice and also interfered with the protective efficacy of BCG in the lung, we sought to determine whether T cell activation was also altered in the lungs during infection with M. tuberculosis. In response to infection with M. tuberculosis, significantly more activated CD4 T cells were observed within the lungs of BCG-vaccinated mice when compared with non-vaccinated mice (Fig. 4, A and C). By Day 35 postinfection, the number of activated CD4 T cells declined in the BCG-vaccinated group, demonstrating accelerated kinetics of T cell activation normally observed in BCG-vaccinated animals [15]. Similar results were observed for CD8 T cells (Fig. 4, B and D).

Despite our observations in the brachial LN, mice, which had been BCG-vaccinated and subsequently exposed to live M. avium by gavage, were able to generate equivalent numbers of activated CD4 and CD8 T cells within the lungs as mice that had received BCG alone (Fig. 4, A–D). At 35 days postinfection, we did observe a moderate but continued activation of CD8 T cells within the lung of mice that had received BCG and M. avium, but these numbers were always equivalent or lower than that seen in the un-manipulated group (Fig. 4, B and D). For clarity, mice, which received M. avium alone, have been omitted from the graphs; the T cell activation profiles were identical to un-manipulated mice. Of particular interest, however, was that in BCG/M. avium-gavaged mice, we consistently observed more resident-activated CD4 T cells within the lung prior to infection with M. tuberculosis (Fig. 4, A and C). For CD4 CD95⫹ T cells, these differences reached significance (P⬍0.05 compared with un-manipulated or BCG-vaccinated mice), whereas a trend for more CD4 CD44hiCD45RBlo T cells was observed (P⬍0.08 compared with un-manipulated or BCG-vaccinated mice). As M. avium could be detected readily within the lungs of mice prior to infection with M. tuberculosis (Fig. 3), we attributed such increased CD4 T cell activation to a M. avium-specific response. A moderate increase in activated CD4 T cells was also observed in mice that received M. avium alone (data not shown). Lung cells from each experimental group (unmanipulated, BCG, M. avium, and BCG/M. avium) were isolated after 14, 21, or 35 days infection with M. tuberculosis and cultured in vitro with M. tuberculosis CFP to investigate whether exposure to M. avium could alter the generation of antigen-specific immunity. At each time-point tested, all four experimental groups were capable of producing equivalent IFN-␥ in response to CFP (data not shown). IFN-␥ production was absent in OVA cultures and present at equivalent amounts in all four groups in response to Con A (data not shown).

Impact of M. avium exposure on the long-term control of an infection with M. tuberculosis

Fig. 3. M. avium in the lungs of M. tuberculosis-infected mice, which were s.c.-vaccinated with BCG or unmanipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. After eight gavages with M. avium, mice were infected with M. tuberculosis by aerosol. M. avium CFU in the lung were determined by plating lung homogenate onto supplemented 7H11 agar plates and counting colonies after 21 days incubation at 37°C. M. avium was distinguished from M. tuberculosis and BCG by subtraction from plates supplemented with clarythromycin or TCH. Data are expressed as the mean and SEM for five mice and are representative of two independent experiments. Statistical significance was determined using a one-way ANOVA with Tukey’s post-test; no significant differences were observed.

1266


BCG efficacy in man is determined by the ability of vaccination to prevent tuberculosis disease and not by short-term reduction of bacterial burden within the lungs. In this regard, we determined whether long-term control of an infection with M. tuberculosis could be compromised by exposure to environmental mycobacteria. Mice were treated with M. avium using an identical experimental protocol as described above, and CFU were determined in the lung at 60 (data not shown) and 130 days post infection with M. tuberculosis (Fig. 5). As anticipated, the protective efficacy of BCG gradually waned over time [16], and less than 1 Log protection was observed by Day 130 postinfection (Fig. 5). At all time-points tested, however, BCG vaccination provided highly significant levels (P⬍0.001) of protection relative to the un-manipulated group. In contrast, BCG-vaccinated mice, which had received gavage with M. avium, no longer afforded a significant level of protection against infection with M. tuberculosis when compared with the un-manipulated group (P⬎0.05 at Day 130). Gavage with M. avium in the absence of BCG vaccination had no impact on the long-term control of M. tuberculosis in the lung.


Fig. 4. T cell activation within the lungs of M. tuberculosis-infected mice, which were s.c.-vaccinated with BCG or unmanipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. After eight gavages with M. avium, mice were infected with M. tuberculosis by aerosol. Lung cells were isolated and analyzed by flow cytometry for CD3, CD4, CD44, and CD45RB (A); CD95 (C); CD3, CD8, CD44, and CD45RB (B); or CD95 (D). Data are expressed as the mean and SEM for five mice and are representative of two independent experiments. Statistical significance was determined using a one-way ANOVA with Tukey’s post-test. *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001.

Influence of M. avium exposure on granuloma formation during infection with M. tuberculosis

DISCUSSION

Microscopic examination of lung sections throughout the experimental infection showed that BCG vaccination reduced the severity and number of granulomas within the lungs in comparison with un-manipulated mice (Table 1). Exposure to M. avium alone had little influence on granuloma formation or integrity when compared with un-manipulated mice. Mice, which had been vaccinated with BCG and subsequently gavaged with M. avium, showed subtle differences in granuloma formation when compared with BCG-vaccinated or un-manipulated mice, with an increased number of epithelioid macrophages, polymorphonuclear cells (neutrophils), and pyknotic debris, indicative of necrosis, by Day 35. These differences became more pronounced after 130 days infection with M. tuberculosis (Fig. 6) with marked pyknotic debris, an increased number of neutrophils within the lung, and increased cholesterol cleft formation.

Our data demonstrate that the capacity of BCG to protect against an infection with M. tuberculosis can be interfered with by exposure to environmental mycobacteria (M. avium). In contrast to other studies, in which environmental mycobacteria had been delivered prior to BCG vaccination [8, 17], we are the first to deliver environmental mycobacteria after BCG. In doing so, we sought to model the current WHO BCG vaccination recommendations, whereby BCG is administered between birth and 1 year of life [4], before substantial exposure to environmental mycobacteria is likely to have occurred. The loss of protective efficacy, normally afforded by BCG, was associated with a continued activation of CD4 and CD8 T cells within the brachial LN, an accumulation of activated CD4 T cells within the lungs of mice prior to M. tuberculosis exposure, increased pyknotic debris, and a basal level of M. avium bacilli within the lungs throughout the entire infection period.


1267

Fig. 5. Long-term control of M. tuberculosis infection in the lung. Mice were s.c.-vaccinated with BCG or un-manipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. After eight gavages with M. avium, mice were infected with M. tuberculosis by aerosol. After 130 days infection, M. tuberculosis CFU in the lung were determined by plating lung homogenate onto supplemented 7H11 agar plates and counting colonies after 21 days incubation at 37°C. Data are expressed as the mean and SEM for five mice. Statistical significance was determined using a one-way ANOVA with Tukey’s post-test. **, P ⬍ 0.01. NS, Not significant.

In addition to delivering M. avium after BCG vaccination, we delivered multiple doses of live M. avium directly into the gastrointestinal tract by gavage. This contrasts other studies, whereby environmental mycobacteria were delivered by i.p., s.c., or i.v. injection [8, 11, 17]. Our rationale for such a strategy was to mimic exposure to environmental mycobacteria in man more closely. M. avium and other environmental mycobacteria are readily found in water and soil [18] and are most likely contracted by ingestion and infection through the intestinal tract. M. avium was also delivered in multiple, live, TABLE 1.

low-dose exposures. Again, this strategy was implemented to model what is believed to happen in man in areas where environmental mycobacteria are prevalent. It is expected that exposure to environmental mycobacteria would be repeated throughout life and would most likely include live mycobacteria, which could actively colonize the host. We did not deliver sufficient M. avium to cause disease but alternatively chose to address whether a subclinical infection with M. avium could interfere with immunity. Similarly, M. avium and BCG were not eradicated with antibiotics before challenge with M. tuberculosis to model live, concurrent infection with multiple species of mycobacteria. We distinguished between each species in the lung homogenates using supplemented media containing TCH or clarythromycin to exclude the growth of BCG and M. avium, respectively. Finally, we chose to use a single strain of M. avium, which although not necessarily representative of a natural infection, provided us with a more controlled model to work with in a system that has multiple variables. M. avium 104 was chosen for our experiments, as it is a clinical isolate, therefore demonstrating pathogenicity in man. In addition, the genome for this strain is currently being sequenced (www.tigr.org). M. avium 104 is virulent in mice, and upon aerosol infection, CFU increase progressively in the lung for ⬃21 days, after which, they level off and persist at a relatively stable bacterial load for up to 250 days post infection. Persistence is not associated with mortality, and lung granulomas show no signs of necrosis or caseation [19]. M. avium 104 is therefore capable of establishing a persistent infection in mice, yet does not cause significant morbidity and mortality, which has been reported in other M. avium strains [20]. In an i.v. model, M. avium 104 was more virulent than strain 100 but equivalent to strain 101, and CFU were detected in the spleen, liver, and lung [21]. Similarly, M. avium 104 has been shown to invade the gut epithelium and disseminate equally well as strain 101 if administered orally [22]. In contrast to other published studies, the delivery of M. avium after BCG vaccination avoided the generation of immunity against environmental mycobacteria, which could subse-

Analysis of Microscopic Lung Lesions from M. tuberculosis-Infected Mice

Groupa

DPIb

Severityc

Granulomasd

Epithelioid M⌽se

PMN necrosis f

Cholesterol clefts

UN BCG MA BCG/MA UN BCG MA BCG/MA UN BCG MA BCG/MA

21 21 21 21 35 35 35 35 130 130 130 130

⫹⫹ ⫹ ⫹ ⫹ ⫹⫹ ⫹ ⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹

⫹ (8) ⫹ (4) ⫹ (5) ⫹ (4) ⫹⫹ (12) ⫹ (3) ⫹⫹ (10) ⫹ (5) ⫹⫹⫹⫹ (34) ⫹⫹ (14) ⫹⫹⫹ (21) ⫹⫹⫹ (25)

⫹⫹ ⫹ ⫹ ⫹ ⫹⫹ (E) ⫹/– (E) ⫹⫹ (E) ⫹ (E) ⫹⫹⫹⫹ (V) ⫹⫹⫹ (V) ⫹⫹⫹ (V) ⫹⫹⫹ (V)

⫹/– ⫹/– ⫹ ⫹/– ⫹ ⫹/– ⫹/– ⫹ ⫹⫹ ⫹⫹ ⫹ ⫹⫹⫹

0 0 0 0 0 0 0 0 ⫹⫹ ⫹ ⫹⫹ ⫹⫹

a

Experimental groups. UN, Unmanipulated; MA, M. avium. b DPI, Days post infection with M. tuberculosis. c Graded on low-power magnification, estimating total lung involvement and designated mild, moderate, or marked (⫹, ⫹⫹, ⫹⫹⫹, respectively). d Graded from small, individual to larger, confluent granulomas (⫹, ⫹⫹, ⫹⫹⫹, ⫹⫹⫹⫹, respectively). Numbers in parentheses represent actual number of granulomas counted per section. e Scored as an estimate of the total number of macrophages (⫹, ⫹⫹, ⫹⫹⫹, ⫹⫹⫹⫹). Qualitative score is in parentheses: E, Epithelioid macrophages; V, highly vacuolated macrophages (M⌽s); no designation, relatively normal pulmonary-alveolar macrophages. f PMN, Polymorphonuclear cells.

1268



Fig. 6. Granuloma composition. Mice were s.c.-vaccinated with BCG or un-manipulated and after 6 weeks, were given M. avium via gavage every 2 weeks for a period of 16 weeks. After eight gavages with M. avium, mice were infected with M. tuberculosis by aerosol. After 130 days infection with M. tuberculosis, lung tissue was fixed in neutral-buffered formalin, sectioned, and stained with H&E for light microscopy. BCG-vaccinated mice (A, C) had granulomas consisting of dense lymphocyte accumulations associated with foamy macrophages. BCG-vaccinated mice, which received M. avium (B, D), had increased numbers of epithelioid macrophages and polymorphonuclear cells, and pyknotic debris and cholesterol deposition indicated tissue necrosis. (A, B) 10⫻ original magnification; (C, D) 40⫻ original magnification. Arrows indicate tissue debris and cholesterol clefts.

quently eradicate BCG rapidly from the host [8], a mechanism that is thought to contribute to the loss of BCG efficacy. In addition, there was no imprinting of environmental mycobacterial immunity onto the host, which could potentially mask immunity generated by BCG [11]. In our studies, BCG-specific immunity was fully intact prior to administration of M. avium, and yet, we found that exposure to an environmental mycobacterial species could interfere with established BCG immunity. How this interference occurs is currently unknown, but several different hypotheses can be postulated. One potential mechanism is that M. avium infection induced the deletion of antigen-specific cells by apoptosis. During infection with M. avium, T cells up-regulate the expression of CD95 on their surface and can be stimulated to undergo apoptosis by cross-linking of Fas [23–25]. We detected sustained expression of CD95 on the surface of CD4 and CD8 T cells within the LN of BCG-vaccinated, M. avium-infected mice (data not shown; paralleled increased CD44 expression). In addition, more CD8 CD95⫹ T cells persisted in the lungs of M. tuberculosis-infected mice, which had received BCG and M. avium, suggesting that these cells may be targeted for apoptosis. CD8 CD95⫹ T cells could, however, be a persistent population of cells, as it has also been shown that M. avium

infection can lead to persistence of CD8 T cells in vivo [24]. It remains uncertain whether the loss of protective efficacy in the lungs of BCG-vaccinated mice is a result of M. avium-induced apoptosis of antigen-specific cells. In our model, BCG-vaccinated, M. avium-exposed mice were equally capable of recruiting or expanding activated T cells within the lung throughout the first 35 days of an infection with M. tuberculosis and could generate an antigen-specific response against CFP. These data argue against apoptosis as a mechanism of M. avium-induced interference. An alternative mechanism for interference is that like other environmental mycobacteria [26], M. avium infection could alter macrophage/dendritic cell responses. In addition to the induction of pro-inflammatory cytokines, M. avium has been shown to influence the production of IL-10 in vitro [23, 27–29]. Furthermore, M. avium infection can suppress IL-12 production in vitro and in vivo [30], reduce the secretion of IL-8 and RANTES [31], and suppress T cell-macrophage interactions by interfering with accessory molecule expression on antigenpresenting cells [32]. All these mechanisms could impact control of an infection with M. tuberculosis. BCG infection can also alter macrophage function [33–35]; however, it is unclear at this time how this might interact with a subsequent exposure


1269

to M. avium or M. tuberculosis. It is more likely that an ongoing infection with M. avium within the lung would impact the generation of M. tuberculosis-specific immunity directly. We did not detect any obvious alterations in T cell immunity between mice that received BCG or BCG/M. avium and therefore, have no indication that M. avium could suppress local immunity at the time-points tested. Indeed, if M. avium were suppressing immunity directly, then we would anticipate that mice receiving M. avium alone would show increased susceptibility to infection with M. tuberculosis. This was not seen, suggesting that the loss of protection against infection with M. tuberculosis appears to be related directly to the influence of M. avium exposure on the generation or maintenance of BCGspecific responses. The increased number and persistence of activated CD4 and CD8 T cells within the brachial LN of BCG-vaccinated mice, which had received M. avium, suggest that ongoing exposure to M. avium may sequester protective T cells to the draining LN. This mechanism would prevent migration of antigen-specific T cells from the LN to the lung in response to infection with M. tuberculosis, although we could detect antigen-specific responses in the lungs at all time-points tested. Alternatively, a continual immune challenge in the form of M. avium exposure may lead to immune exhaustion as cells undergo multiple rounds of stimulation and proliferation. Such an event might explain why BCG-vaccinated mice, which received M. avium, could initially control an infection in the lung as well as BCG-vaccinated animals, yet this protective response waned within the lungs over time. Indeed, we could culture M. avium from the lungs of mice between 14 and 35 days post infection with M. tuberculosis. We also observed increased necrosis in the lung tissue of mice, which received BCG and M. avium. Such an increased mycobacterial burden in the lung could lead to increased local immune stimulation and ultimately contribute to immune exhaustion, tissue damage, and increased growth of M. tuberculosis within the lung. It is interesting that we found ⬃3 Log CFU of M. avium within the lungs of M. tuberculosis-infected mice, regardless of whether they had received prior vaccination with BCG or not. In contrast to Torrelles et al. [19], M. avium 104 did not exceed 3 Log in the lung in our experiments; however, we can attribute this to the concurrent infection with M. tuberculosis or the alternative method of delivery of M. avium used in each study (aerosol vs. oral). During early time-points, the quantity of M. avium did not increase or decrease during infection with M. tuberculosis, suggesting that M. avium may have located to an intracellular niche within the lung, which was refractory to the local immune response. Petrofsky and Bermudez [36] showed that delivery of M. avium via gavage established infection in the mesenteric LN and terminal ileum of mice; however, they also reported that M. avium could be found in the lung. Such detection may reflect tropism of M. avium specifically to this organ; however, although unlikely, we cannot rule out that we inadvertently delivered M. avium directly into the lung during gavage. Although we did not observe a dramatic interference with BCG efficacy in our model of M. avium exposure, our studies demonstrate proof of the concept that exposure to environmental mycobacteria can negatively impact established, BCG-spe1270


cific immunity against infection with M. tuberculosis. M. avium is just one species of mycobacteria found in the environment [37, 38], which itself exists as many serovars [39]. The impact of one single serovar (serovar 1; ref. [19]) of one species of environmental mycobacteria (M. avium) may only provide a proportion of the interference normally seen in man. It could be hypothesized that the addition of other mycobacterial species, which are found in the environment, such as Mycobacterium fortuitum, Mycobacterium intracellulare, and Mycobacterium kansasii [40], could further enhance the interference that we observed. In support of this theory, Brandt et al. [8] showed that M. avium, M. fortuitum, and Mycobacterium vaccae could interfere with BCG efficacy if it were administered prior to BCG vaccination. In contrast, however, Demangel et al. [17] sensitized mice with individual species of the same mycobacteria and found marginal interference from M. avium and some degree of protection from M. fortuitum and M. vaccae. Further studies are necessary to determine whether other environmental mycobacteria, administered live or dead, can interfere with BCG efficacy. In summary, we demonstrate that exposure to M. avium by the gastrointestinal route can interfere with an established, protective immune response generated by BCG vaccination. These findings have particular relevance to BCG vaccination programs implemented in areas where environmental mycobacteria are prevalent and may explain why BCG efficacy is low in such locations. In addition, in countries where BCG vaccination has been successful in preventing childhood tuberculosis, there are also reports of sensitization to environmental mycobacteria [41, 42]. Such widespread exposure to environmental mycobacteria could potentially lead to waning of BCG efficacy over time, as we demonstrate here in mice, and account for the loss of protective efficacy of BCG against M. tuberculosis infection in the adult population.

ACKNOWLEDGMENTS This publication was made possible by Grant No. R21 AI54841, awarded by NIAID. We thank the Department of Veterinary Biosciences Histopathology Laboratory for preparation of tissue sections. CFP was received as a part of NIH, NIAID, Contract No. HHSN 266200400091C, entitled “Tuberculosis Vaccine Testing and Research Materials,” which was awarded to Colorado State University.

REFERENCES 1. Corbett, E. L., Watt, C. J., Walker, N., Maher, D., Williams, B. G., Raviglione, M. C., Dye, C. (2003) The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch. Intern. Med. 163, 1009 –1021. 2. Kochi, A. (1991) The global tuberculosis situation and the new control strategy of the World Health Organization. Tubercle 72, 1– 6. 3. World Health Organization (2006) Global tuberculosis surveillance, planning, financing. WHO, Geneva. 1–250. 4. World Health Organization (2005) WHO Vaccine-Preventable Diseases: Monitoring System. WHO, Geneva. 1–333. 5. Fine, P. E. M. (1994) Immunities in and to tuberculosis: implications for pathogenesis and vaccination. In Tuberculosis: Back to the Future (J. D. H.


6. 7. 8.

9. 10. 11. 12. 13.

14. 15.

16. 17.

18. 19.

20. 21.

22. 23.

24.

25. 26.

Porter, K. P. W. J. McAdam, eds.), Australia, John Wiley & Sons Ltd., 53–71. Fine, P. E. (2001) BCG: the challenge continues. Scand. J. Infect. Dis. 33, 243–245. Smith, D., Wiegeshaus, E., Balasubramanian, V. (2000) An analysis of some hypotheses related to the Chingelput bacille Calmette-Guerin trial. Clin. Infect. Dis. 31 (Suppl. 3), S77–S80. Brandt, L., Feino Cunha, J., Weinreich Olsen, A., Chilima, B., Hirsch, P., Appelberg, R., Andersen, P. (2002) Failure of the Mycobacterium bovis BCG vaccine: some species of environmental mycobacteria block multiplication of BCG and induction of protective immunity to tuberculosis. Infect. Immun. 70, 672– 678. de Lisle, G. W., Wards, B. J., Buddle, B. M., Collins, D. M. (2005) The efficacy of live tuberculosis vaccines after presensitization with Mycobacterium avium. Tuberculosis (Edinb.) 85, 73–79. Buddle, B. M., Wards, B. J., Aldwell, F. E., Collins, D. M., de Lisle, G. W. (2002) Influence of sensitization to environmental mycobacteria on subsequent vaccination against bovine tuberculosis. Vaccine 20, 1126 –1133. Palmer, C. E., Long, M. W. (1966) Effects of infection with atypical mycobacteria on BCG vaccination and tuberculosis. Am. Rev. Respir. Dis. 94, 553–568. Fine, P. E. (1989) The BCG story: lessons from the past and implications for the future. Rev. Infect. Dis. 11 (Suppl. 2), S353–S359. Howard, C. J., Kwong, L. S., Villarreal-Ramos, B., Sopp, P., Hope, J. C. (2002) Exposure to Mycobacterium avium primes the immune system of calves for vaccination with Mycobacterium bovis BCG. Clin. Exp. Immunol. 130, 190 –195. World Health Organization (2006) Immunization Surveillance, Assessment and Monitoring. Vaccine-preventable diseases: tuberculosis. WHO, Geneva. 1. Kipnis, A., Irwin, S., Izzo, A. A., Basaraba, R. J., Orme, I. M. (2005) Memory T lymphocytes generated by Mycobacterium bovis BCG vaccination reside within a CD4 CD44lo CD62 ligandhi population. Infect. Immun. 73, 7759 –7764. Brooks, J. V., Frank, A. A., Keen, M. A., Belisle, J. T., Orme, I. M. (2001) Boosting vaccine for tuberculosis. Infect. Immun. 69, 2714 –2717. Demangel, C., Garnier, T., Rosenkrands, I., Cole, S. T. (2005) Differential effects of prior exposure to environmental mycobacteria on vaccination with Mycobacterium bovis BCG or a recombinant BCG strain expressing RD1 antigens. Infect. Immun. 73, 2190 –2196. Primm, T. P., Lucero, C. A., Falkinham III, J. O. (2004) Health impacts of environmental mycobacteria. Clin. Microbiol. Rev. 17, 98 –106. Torrelles, J. B., Ellis, D., Osborne, T., Hoefer, A., Orme, I. M., Chatterjee, D., Brennan, P. J., Cooper, A. M. (2002) Characterization of virulence, colony morphotype and the glycopeptidolipid of Mycobacterium avium strain 104. Tuberculosis (Edinb.) 82, 293–300. Saunders, B. M., Frank, A. A., Cooper, A. M., Orme, I. M. (1998) Role of ␥ ␦ T cells in immunopathology of pulmonary Mycobacterium avium infection in mice. Infect. Immun. 66, 5508 –5514. Saunders, B. M., Dane, A., Briscoe, H., Britton, W. J. (2002) Characterization of immune responses during infection with Mycobacterium avium strains 100, 101 and the recently sequenced 104. Immunol. Cell Biol. 80, 544 –549. Sangari, F. J., Goodman, J., Petrofsky, M., Kolonoski, P., Bermudez, L. E. (2001) Mycobacterium avium invades the intestinal mucosa primarily by interacting with enterocytes. Infect. Immun. 69, 1515–1520. Bermudez, L. E., Parker, A., Petrofsky, M. (1999) Apoptosis of Mycobacterium avium-infected macrophages is mediated by both tumor necrosis factor (TNF) and Fas, and involves the activation of caspases. Clin. Exp. Immunol. 116, 94 –99. Zhong, J., Gilbertson, B., Cheers, C. (2003) Apoptosis of CD4⫹ and CD8⫹ T cells during experimental infection with Mycobacterium avium is controlled by Fas/FasL and Bcl-2-sensitive pathways, respectively. Immunol. Cell Biol. 81, 480 – 486. Roger, P. M., Bermudez, L. E. (2001) Infection of mice with Mycobacterium avium primes CD8⫹ lymphocytes for apoptosis upon exposure to macrophages. Clin. Immunol. 99, 378 –386. Martino, A., Sacchi, A., Volpe, E., Agrati, C., De Santis, R., Pucillo, L. P., Colizzi, V., Vendetti, S. (2005) Non-pathogenic Mycobacterium smegmatis

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39. 40.

41.

42.

induces the differentiation of human monocytes directly into fully mature dendritic cells. J. Clin. Immunol. 25, 365–375. Bermudez, L. E., Champsi, J. (1993) Infection with Mycobacterium avium induces production of interleukin-10 (IL-10), and administration of antiIL-10 antibody is associated with enhanced resistance to infection in mice. Infect. Immun. 61, 3093–3097. Balcewicz-Sablinska, M. K., Gan, H., Remold, H. G. (1999) Interleukin 10 produced by macrophages inoculated with Mycobacterium avium attenuates mycobacteria-induced apoptosis by reduction of TNF-␣ activity. J. Infect. Dis. 180, 1230 –1237. Silva, R. A., Pais, T. F., Appelberg, R. (2001) Blocking the receptor for IL-10 improves antimycobacterial chemotherapy and vaccination. J. Immunol. 167, 1535–1541. Wagner, D., Sangari, F. J., Kim, S., Petrofsky, M., Bermudez, L. E. (2002) Mycobacterium avium infection of macrophages results in progressive suppression of interleukin-12 production in vitro and in vivo. J. Leukoc. Biol. 71, 80 – 88. Sangari, F. J., Petrofsky, M., Bermudez, L. E. (1999) Mycobacterium avium infection of epithelial cells results in inhibition or delay in the release of interleukin-8 and RANTES. Infect. Immun. 67, 5069 –5075. Tsuyuguchi, I., Kawasumi, H., Takashima, T., Tsuyuguchi, T., Kishimoto, S. (1990) Mycobacterium avium-Mycobacterium intracellulare complexinduced suppression of T-cell proliferation in vitro by regulation of monocyte accessory cell activity. Infect. Immun. 58, 1369 –1378. Gagliardi, M. C., Teloni, R., Giannoni, F., Pardini, M., Sargentini, V., Brunori, L., Fattorini, L., Nisini, R. (2005) Mycobacterium bovis Bacillus Calmette-Guerin infects DC-SIGN-dendritic cell and causes the inhibition of IL-12 and the enhancement of IL-10 production. J. Leukoc. Biol. 78, 106 –113. Hubeau, C., Singer, M., Lagranderie, M., Marchal, G., Vargaftig, B. (2003) Extended freeze-dried Mycobacterium bovis Bacillus Calmette-Guerin induces the release of interleukin-12 but not tumor necrosis factor-␣ by alveolar macrophages, both in vitro and in vivo. Clin. Exp. Allergy 33, 386 –393. Matsumoto, H., Suzuki, K., Tsuyuguchi, K., Tanaka, E., Amitani, R., Maeda, A., Yamamoto, K., Sasada, M., Kuze, F. (1997) Interleukin-12 gene expression in human monocyte-derived macrophages stimulated with Mycobacterium bovis BCG: cytokine regulation and effect of NK cells. Infect. Immun. 65, 4405– 4410. Petrofsky, M., Bermudez, L. E. (2005) CD4⫹ T cells but not CD8⫹ or ␥␦⫹ lymphocytes are required for host protection against Mycobacterium avium infection and dissemination through the intestinal route. Infect. Immun. 73, 2621–2627. Weir, R. E., Fine, P. E., Nazareth, B., Floyd, S., Black, G. F., King, E., Stanley, C., Bliss, L., Branson, K., Dockrell, H. M. (2003) Interferon-␥ and skin test responses of schoolchildren in Southeast England to purified protein derivatives from Mycobacterium tuberculosis and other species of mycobacteria. Clin. Exp. Immunol. 134, 285–294. Fine, P. E., Floyd, S., Stanford, J. L., Nkhosa, P., Kasunga, A., Chaguluka, S., Warndorff, D. K., Jenkins, P. A., Yates, M., Ponnighaus, J. M. (2001) Environmental mycobacteria in Northern Malawi: implications for the epidemiology of tuberculosis and leprosy. Epidemiol. Infect. 126, 379 – 387. Inderlied, C. B., Kemper, C. A., Bermudez, L. E. (1993) The Mycobacterium avium complex. Clin. Microbiol. Rev. 6, 266 –310. Black, G. F., Dockrell, H. M., Crampin, A. C., Floyd, S., Weir, R. E., Bliss, L., Sichali, L., Mwaungulu, L., Kanyongoloka, H., Ngwira, B., Warndorff, D. K., Fine, P. E. (2001) Patterns and implications of naturally acquired immune responses to environmental and tuberculous mycobacterial antigens in Northern Malawi. J. Infect. Dis. 184, 322–329. Kroger, L., Korppi, M., Pelkonen, J., Pietikainen, M., Katila, M. L. (2000) Development of tuberculin reactivity and sensitization to M. scrofulaceum and M. fortuitum in children BCG-vaccinated at birth. Eur. Respir. J. 15, 382–387. Huang, J. H., Kao, P. N., Adi, V., Ruoss, S. J. (1999) Mycobacterium avium-intracellulare pulmonary infection in HIV-negative patients without preexisting lung disease: diagnostic and management limitations. Chest 115, 1033–1040.


1271