JCM Accepted Manuscript Posted Online 3 February 2016 J. Clin. Microbiol. doi:10.1128/JCM.02620-15 Copyright © 2016 Vallur et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.
1
Accurate sero-detection of asymptomatic Leishmania donovani infection using defined
2
antigens
3
Aarthy C Vallur1, Caroline Reinhart1, Raodoh Mohamath1, Yasuyuki Goto1,*, Prakash Ghosh2,
4
Dinesh Mondal2, Malcolm S Duthie1 and Steven G Reed1, #
5
1
6
Diseases Research, Laboratory Sciences Division, Dhaka, Bangladesh
7
Running title: Novel antigens for detecting L. donovani infection
8
*
9
Life Sciences, University of Tokyo, Tokyo, Japan
Infectious Disease Research Institute, Seattle, USA,
2
International Center for Diarrhoeal
Present address: Laboratory of Molecular Immunology, Graduate School of Agricultural and
10
#
11
Steven G Reed
12
e-mail:
[email protected]
13
Key words: Leishmania, kala azar, antigen, asymptomatic, protein, diagnosis, antibody,
14
surveillance, elimination
15
Abbreviations: VL: visceral leishmaniasis, Ld/ L. donovani: Leishmania donovani, DAT: Direct
16
Agglutination Test, EC: endemic control, ELISA: enzyme-linked immunosorbent assay, NEC:
17
non-endemic control, OD: other diseases, PBS: phosphate-buffered saline, PBS-T: PBS-Tween,
18
PCR: Polymerase Chain Reaction, PKDL: Post-Kala azar dermal leishmaniasis, RDT: Rapid
19
Diagnostic Test, TR: tandem repeat, SLA: Soluble lysate antigen
Address correspondence and reprint requests to
1
20
ABSTRACT
21
Infection with Leishmania donovani is typically asymptomatic, but a significant number of
22
individuals may progress to visceral leishmaniasis (VL), a deadly disease that threatens 200
23
million people in endemic areas. While diagnosis of acute VL has been simplified by the use of
24
cost-effective confirmatory serological tests, similar standardized tools are not widely available
25
for detecting asymptomatic infection which can be 4-20 times more prevalent than active
26
disease. A simple and accurate serological test capable of detecting asymptomatic L. donovani
27
infection will be useful for surveillance programs targeting VL control and elimination. To
28
address this unmet need, we evaluated recombinant antigens for their ability to detect serum
29
antibodies in 104 asymptomatic L. donovani infected individuals (qualified as positive for L.
30
donovani-specific antibodies by direct agglutination test; DAT) from the VL hyperendemic
31
Mymensingh district of Bangladesh. The novel proteins rKR95 and rTR18 possessed the
32
greatest potential and detected 69% of DAT positive individuals, with rKR95 being more robust
33
in reactivity. Agreement in results with individuals with high DAT responses, who are more
34
likely to progress to VL disease, was 74%. When considered along with rK39, a gold standard
35
antigen used to confirm clinical diagnosis of VL but which is now becoming widely used for
36
surveillance, rKR95 and rTR18 conferred a sensitivity of 84% based on a theoretical combined
37
estimate. Our data indicate that incorporating rKR95 and rTR18 with rK39 in serological tests
38
amenable to rapid or high-throughput screening could enable simple and accurate detection of
39
asymptomatic infection. Such tests will be important tools to measure L. donovani infection
40
rates, a primary goal in surveillance and a critical measurement with which to assess elimination
41
programs.
2
42
Introduction
43
Visceral Leishmaniasis (VL or Kala azar) is one of the deadliest and most neglected
44
tropical diseases in the world. It affects the poorest among mostly rural endemic populations,
45
with around 300,000 new cases every year (1, 2). Of these, about 90% occur in the Indian sub-
46
continent, Brazil and East Africa, while VL is an emerging threat in the Mediterranean basin (1).
47
In the Indian sub-continent, VL is caused by infection with Leishmania donovani.
48
Importantly, VL in the Indian subcontinent has some unique features that appear to make it a
49
candidate for elimination (3, 4). The presence of a single vector, anthroponotic transmission and
50
a defined, rural area of endemicity have empowered the governments of India, Nepal and
51
Bangladesh to propose the Kala-azar elimination program, which aims to reduce the incidence
52
of VL to less than 1 case per 10,000 individuals by 2015 (5). Early case detection, treatment
53
and integrated vector control are three of six pillars of the program (5-7). Of these, the practical
54
difficulties plaguing early case detection have been a roadblock in advancing toward the
55
elimination program objectives (8, 9). Recognizing the potential of the rk39-based rapid
56
diagnostic tests (RDT), the elimination program has made it readily available to clinics and
57
health care workers in endemic regions, thereby enhancing detection of acute VL cases. A
58
major challenge for the elimination campaign is the large population of asymptomatic L.
59
donovani infected individuals residing in endemic regions. Such individuals harbor low levels of
60
infection but display no symptoms of VL. Although clearance of the parasites appears to be the
61
typical outcome, these individuals can potentially transmit parasites to others and are at an
62
elevated risk of developing VL (10, 11). Early case detection hinges on identifying
63
asymptomatics and monitoring them periodically. Thus, surveillance, detection, and monitoring
64
of asymptomatic L. donovani infection are absolutely necessary to realize Kala-azar elimination.
3
65
Existing diagnostic tools are not entirely suitable for detection of asymptomatic infection.
66
Diagnostic methods such as microscopy of splenic aspirates are unethical in asymptomatic
67
individuals and unsuitable for large population surveillance. The rk39 RDT recommended for
68
confirming VL disease in the Indian sub-continent is not fully reliable to screen for asymptomatic
69
L. donovani infection (12). At present, ELISA against rK39 and the Direct Agglutination Test
70
(DAT) are commonly used in large surveillance studies for asymptomatic infection in the Indian
71
sub-continent (13-16). Although DAT is effective for detecting L. donovani infection because it
72
offers a broader antigen panel, the use of freeze-dried promastigotes as the detecting antigen
73
can render it susceptible to lot-to-lot variations (13, 17). In addition, DAT is labor intensive,
74
markedly lower in throughput than a standard ELISA and most importantly, being a visual test, it
75
is extremely difficult to set uniform standards for widespread use.
76
Detection of asymptomatic L. donovani infection is an urgent need within VL control
77
programs. Given that asymptomatic L. donovani infections are more common than VL, there is a
78
need for simple and standardized tools to provide sensitive, specific and quantitative results
79
whilst also facilitating high-throughput screening within endemic regions. In an attempt to
80
develop a recombinant antigen-based serological test with these properties for use in
81
surveillance for asymptomatic infection, we evaluated several Leishmania antigens in an ELISA
82
on sera from likely asymptomatic L. donovani infected individuals in Bangladesh. Through this
83
process we selected two recombinant antigens that complemented rk39 ELISA and were
84
comparable to DAT in detecting asymptomatic L. donovani infection. We discuss our results in
85
terms of a dedicated serological tool to screen endemic areas for asymptomatic L. donovani
86
infections in a conventional ELISA or a rapid test format.
87
Methods and Materials:
88
Samples
4
89 90 91
All samples were collected following approval from the respective Ethics Committee and after obtaining individual consent forms. Blood was obtained and sera/DNA prepared from individuals with no past history of VL
92
or PKDL residing in the VL hyper-endemic region in Harirampur Union, Trishal sub-district,
93
Mymensingh district, Bangladesh as described before (10). Initial consent was obtained from
94
the head of household to screen household members, then individual written consent was
95
obtained from participants prior to study enrollment. Sera from clinically confirmed VL patients
96
were included as positive controls. Sera from 46 healthy individuals in the US who had no
97
history of travel outside the US (purchased from Equitech, TX), were used as non-endemic
98
controls (NEC) to establish cut-offs for sensitivity. In addition to the NEC, sera from healthy
99
endemic controls (EC) from the Mymensingh district were used. To measure cross-reactivity
100
with other diseases (OD), sera from patients with non-VL febrile illnesses from a non-endemic
101
region (the Philippines) were used. These individuals were defined with no previous history of
102
VL due to negative responses in DAT and rK39 RDT.
103
Initial sample characterization
104
Initial serum characterization was conducted using the Direct Agglutination Test (DAT)
105
(KIT Biomedical, Amsterdam) performed according to manufacturer’s instructions at icddr,b,
106
Dhaka, Bangladesh. Based on a DAT titer >1600 in these evaluations, 104 sera samples were
107
designated as being DAT positive and are referred to as asymptomatic L. donovani infected
108
individuals in this study. The 104 individuals were also tested on finger prick blood for positivity
109
with the rK39-RDT (KalaAzar DetectTM, InBios International, Seattle) by investigators from
110
icddr,b, Dhaka, Bangladesh. Healthy endemic controls were defined as DAT and rK39-RDT
111
negative individuals from the same area as the asymptomatic individuals. Aliquots of the sera
112
samples were shipped to IDRI and archived.
5
113
Recombinant proteins
114
Several tandem repeat (TR) proteins previously identified in a bioinformatics screen and
115
validated as being reactive with VL patient sera were evaluated (18). Kinesin-like protein was
116
identified by a LC-MS/MS analysis of sera and urine samples from VL patients performed using
117
standard procedures at ITSI Biosciences (Johnstown, PA). Peptide hits were then screened for
118
Leishmania specificity against L. donovani and L. infantum databases using Proteome
119
Discoverer 1.2 and the SEQUEST algorithm. Peptides specific to VL patient samples but not
120
healthy US samples were considered as Leishmania-specific peptides. L. donovani kinesin-
121
related protein, henceforth referred to as rKR95 (GI.112293604) was identified by 5 peptide hits
122
in 4 VL sera and 1 peptide hit in 1 VL urine sample and was chosen for further validation after
123
confirming lack of homology to human counterparts (Table 1). The construct encoding the 742
124
amino acids sequence that contains all six peptides was synthesized and cloned into pET17a
125
with an N-terminal 6XHis tag (Gene Dynamics, Tigard, OR). The sequence verified expression
126
construct was expressed from E. coli strain BL21 (DE3) pLys (Invitrogen) and purified as
127
described previously (18).The purity and integrity of each protein was confirmed by SDS-PAGE.
128
rK39 antigen was made as described previously(19).
129
Ld SLA was made
by lysing 1×107 cultured L. donovani/MHOM/80/IN/DD8
130
promastigotes in the presence of 1X protease inhibitors (HALT protease inhibitor cocktail,
131
Thermofisher) and 5mM EDTA by repeated freeze thawing followed by sonication and
132
separating insoluble material by centrifuging for 45 minutes at 15,000 rpm.
133
Antibody reactivity by ELISA
134
All ELISAs were done at IDRI. High binding 384-well plates (Corning, NY) were coated
135
with 1µg/mL recombinant protein or 2.5µg/mL Ld SLA in the ELISA Coating buffer
136
(eBioSciences, Inc., San Diego, CA) at room temperature for 4 hours and washed with PBS6
137
Tween (PBS-T) before being blocked overnight at 4oC with PBS-T with 1% BSA. Plates were
138
washed as before. Serum diluted 1:200 in PBS-T with 0.1% BSA was then added to each well
139
and plates were incubated at room temperature for 2 hours on a shaking platform. After washing
140
as before, horseradish peroxidase-conjugated anti-human IgG (heavy and light chains)
141
(Invitrogen) diluted 1:8000 in PBS-T with 0.1% BSA was added to each well and incubated for
142
1h at room temperature. Plates were washed again and developed with SureBlue™ TMB
143
Microwell Peroxidase Substrate (KPL, Inc., Gaithersburg, MD). 1N H2SO4 was then added
144
to each well to stop the reaction. The optical density of each well was immediately read at both
145
450nm and 570nm. The final, background-adjusted signal for each sample was calculated by
146
subtracting the 570nm reading from the 450nm reading. Cut-off for sero-positivity on each plate
147
was defined as the mean signal of normal sera added 2 standard deviations. Signal to noise
148
ratios were calculated as ELISA signal over cut-off for any antigen. Ratios over 1 were
149
considered positive.
150
Sequence alignments
151
The BLASTn suite (NCBI) using the cloned sequences as queries against Leishmania
152
and non-Leishmania genomes or cDNA sequences when genomes were not available.
153
Statistical analysis
154
One-way ANOVA was used to determine the statistical significance between non-
155
parametric data sets to assess the performance of antigens, with p