Glycoproteins in circulating immune complexes are biomarkers of patients with Indian PKDL: A study from endemic districts of West Bengal, India

Background Post Kala Azar Dermal Leishmaniasis (PKDL) occurs as dermal consequence of previous Visceral Leishmaniasis (VL) infection and serves as an important reservoir for transmission of VL. Diagnosis of PKDL is often challenging for its symptomatic resemblance to other co-endemic diseases like Leprosy or Vitiligo. Parasitological examination by slit-skin smear and culture are the standard methods but lack high sensitivity. Thus, for efficient control of VL, reliable diagnostic and prognostic assay of PKDL are required. Objective Previously, glycoproteins (9-OAcSA) have been reported as promising biomarkers of Indian VL patients. However, till date, the status of glycans in Indian PKDL patients remains unexplored. Accordingly, in this study, the glyco-profile of PKDL Circulating Immune Complexes (CICs) as compared to other cross diseases like Vitiligo and Leprosyhas been investigated. Further, a novel Glyco CIC assay has been developed for efficient Indian PKDL patient diagnosis. Methods/principal finding In the present study, 90 PKDL patients were enrolled from 3 VL endemic districts of West Bengal during 2015–16. Glycosylation profile of isolated CICs from sera of PKDL patients were initially analyzed through gradient SDS gel electrophoresis followed by PAS silver double staining, which revealed the presence of several glycan rich PKDL specific proteins of varying molecular weights. To further characterize the glyco-profile of acid dissociated affinity purified immuno-reactive antigens present in the CICs, glycosylation was demonstrated in these purified CIC antigens by DIG glycan differentiation kit with or without glycosidase as well as neuraminidase treatment. Diagnostic evaluation of the newly developed colorimetric Glyco CIC assay through Receiver Operating Characteristic (ROC) curve analysis revealed excellent (0.99) AUC value as compared to other conventional serodiagnostic assays like PEG CIC, Parasite ELISA (IgG and IgM). Additionally, longitudinal monitoring of 18 PKDL patients further revealed its good prognostic utility. Conclusion These results highlight the glycosylation status of CICs among Indian PKDL patients present in all the studied endemic districts of West Bengal. These PKDL biomarkers were completely absent in cross diseases like Vitiligo and Leprosy. Further, the newly developed Glyco CIC assay had an improved sensitivity of 95.6%, specificity of 99.3%, NPV of 97.1% and PPV of 98.9%.


Objective
Previously, glycoproteins (9-OAcSA) have been reported as promising biomarkers of Indian VL patients. However, till date, the status of glycans in Indian PKDL patients remains unexplored. Accordingly, in this study, the glyco-profile of PKDL Circulating Immune Complexes (CICs) as compared to other cross diseases like Vitiligo and Leprosyhas been investigated. Further, a novel Glyco CIC assay has been developed for efficient Indian PKDL patient diagnosis.

Methods/principal finding
In the present study, 90 PKDL patients were enrolled from 3 VL endemic districts of West Bengal during 2015-16. Glycosylation profile of isolated CICs from sera of PKDL patients were initially analyzed through gradient SDS gel electrophoresis followed by PAS silver double staining, which revealed the presence of several glycan rich PKDL specific proteins of varying molecular weights. To further characterize the glyco-profile of acid dissociated affinity purified immuno-reactive antigens present in the CICs, glycosylation was demonstrated in these purified CIC antigens by DIG glycan differentiation kit with or without glycosidase as PLOS

Introduction
Leishmaniasis is considered as one of the major parasitic disease, affecting individuals, worldwide. Residents mostly belonging to the tropical countries, with poor living conditions, serve as an easy prey for the vector sand flies, harboring the parasite of Leishmania sp. Currently 350 million people are at risk, worldwide, with India being one of the endemic hotspots, along with Sudan, Nepal, Brazil, Bangladesh and Ethiopia [1,2]. In India, Bihar alone contributes to 90% of Leishmaniasis cases [3]. Moreover, Indian Leishmaniasis are profoundly manifested in two forms-namely (i)Visceral Leishmaniasis (VL); where patients develop characteristics of high fever, hepatosplenomegaly, anemia, loss of appetite and (ii) Post Kala Azar Dermal Leishmaniasis (PKDL); a dermal consequence, which occurs mostly in patients with previous VL history (10-15%), having superficial manifestations as painless dermal lesions of macular, nodular or polymorphic types [3][4][5]. PKDL patients are often reluctant to seek medical care as the disease results in no fatal consequences other than cosmetic complicacies, which is more of a social stigma rather than a grave problem. Considering such neglected scenario, the ongoing VL elimination program in India is targeted to reduce VL incidences to less than 1 per 10,000 population in block and sub block levels, where the upsurge of PKDL cases are posing threat to this holistic approach of elimination. Moreover, accurate case diagnosis of PKDL patients still stands as the foremost challenge for successful elimination of VL, since unattended or misdiagnosed cases of PKDL serve as a silent reservoir of Leishmania donovani, in VL endemic zones and play active role in the anthroponotic transmission of VL [1]. Currently, several modern tools and techniques are available for PKDL diagnosis. Parasite identification by skin slit smear is considered gold standard, but has sensitivity of only 58% for PKDL case detection as highlighted from studies in Bihar [6]. On the other hand, rk39 strip test lacks efficacy in macular PKDL case detection with only 73% sensitivity, in Indian scenario [7,8]. In addition, the Leishmanin Skin Test (LST) for determining parasite load or culture isolation from skin biopsies is also not reliable, as it has only 54% sensitivity. Therefore, in the absence of proper diagnostics, there is likelihood that a hypopigmented macular form of PKDL can be misdiagnosed as a Vitiligo case [9][10][11]. Histopathological studies with patients' skin biopsies too have very low (7-33%) sensitivity towards macular case detection [9,12]. Further, Direct Agglutination Test (DAT), which is otherwise efficient for VL diagnosis fails in PKDL case detection [13][14][15]. On the other hand, high throughput confirmatory assay like microscopy and PCR/qPCR requires highly trained individuals, and has limitations in field settings [16]. Although various analytical tools for both VL and PKDL diagnosis are available, however, very few prognostic markers or biomarkers to monitor treatment response of VL/ PKDL have been reported. Such biomarkers would be very useful in successfully diagnosing a cured VL/treated PKDL from an active VL/PKDL case. Since PKDL patients serve as silent parasitic reservoir of Leishmania donovani in VL affected regions and actively participates in the anthroponotic VL transmission [1], hence there exists an urgent need for identification of novel biomarkers and the development of an accurate diagnostic and prognostic assay.
Several studies have previously confirmed the diagnostic utility of 9-O acetylated sialoglycoproteins among patients with Indian VL [17][18][19][20][21]. Moreover, 9-O acetylated sialoglycoproteins were reported to have high sensitivity and specificity for VL case detection. Till date, very little effort towards development of glycan based diagnostics for PKDL case detection has been done. Consecutively, studies from our group have recently demonstrated increased titres of IgG1 containing Immune Complexes (ICs) in Indian Leishmaniasis patients with the development of a novel ELISA for serodiagnosis of both VL and PKDL [22]. With the aim to characterize novel PKDL glycan biomarkers, the present study investigates the glycosylation status of CICs of PKDL patients, followed by the development of glycan based assay for diagnosis and prognostic evaluation of Indian PKDL patients.
In the current study, we enrolled study subjects through the ongoing VL elimination programme, in endemic districts of West Bengal. Various factors like demography, ethnicity, VL history, drug and immune profile were evaluated for 90 primarily confirmed PKDL patients during the period of 2015-2016. In parallel, CICs from sera of PKDL patients along with Endemic and Non endemic controls as well as disease controls like Vitiligo and Leprosy were prepared and analyzed for development of a novel glycan based assay; the Glyco CIC assay. The assay utilizes the ability of Periodic Acid Solution (PAS) to specifically oxidize the serum CICs' hydroxyl (-OH) groups within the carbohydrates to form aldehydes, which upon addition with Schiff reagent forms a purple color adduct, which can easily be read in a colorimeter [23]. The efficacy of this assay was further tested for its ability to distinguish PKDL positive patients from other patients with similar skin diseases like Vitiligo and Leprosy. Longitudinal monitoring of Glyco CIC titres was analyzed, to evaluate the prognostic efficiency of the assay. Further, efforts were directed towards the characterization of the novel glycan biomarkers, using SDS and gradient gel electrophoresis followed by PAS silver double staining procedures. Subsequently, Lectin blot was also performed to ascertain the glyco-profile of these newly identified proteins.

Study area and population
PKDL study subjects were enrolled through the ongoing VL Elimination programme (2015-2016) at block as well as sub-block levels in the state of West Bengal, India. The study subjects recruited were mainly from endemic districts of West Bengal namely Malda (

Ethics statement
Patients' blood samples were collected after explaining the benefit and objective of the study to the patient and his/her family members by the doctors from School of Tropical Medicine, Kolkata. It was difficult to obtain written informed consents from all suspected participants who consented to take part in this study, in those cases verbal consents were obtained for rk39 RDT testing and documented through research notes. However, written informed consents were taken from all of the confirmed PKDL patients (from parents/legal guardian in case of child patients) and were assured that their identities will be kept confidential. Further, they could withdraw their names from the study at any time they wish.
After proper explanation, the volunteered participants were selected for the study. Upon agreement, the medical team members collected demographic data like sex, age, mode and duration of treatment, past history of VL/PKDL. Primary screening was done based on lesionsover trunk, peripheral parts like limbs and face (macular, nodular or polymorphic). After primary screening, 2ml of venous blood was collected in heparinized vials. Further, rK39 test was performed by trained field assistants. The ethical considerations at the study sites and outdoor and Indoor Departments of Dermatology and Tropical Medicine, School of Tropical Medicine were reviewed and approved by the Clinical Research Ethics Committee of School of Tropical Medicine, Kolkata(CREC-STM).

Diagnosis
Suspected PKDL patients based on dermal lesions (macular, nodular or polymorphic) were carefully examined, individually by the clinicians. Venous blood (2ml) was collected in heparinized vials from consented patients. rK39 strip test (Kala Azar TM Rapid Test InBios International, Seattle, WA, USA) was performed by trained field assistants according to the standardized protocol. Confirmed PKDL patients were identified based on rK39 positivity. Continuous cold chain was maintained to deliver the blood samples at School of Tropical Medicine, Kolkata, where further experiments were performed.

Glyco CIC assay
CICs were prepared based on protocols described by Raja et al, 2006 [26]. Briefly, for measuring the levels of glycosylated CICs, initially CICs were prepared with patient serum sample (10μl) diluted in Borate buffered saline (BBS) at 1:10 ratio, which was mixed with 12% PEG 6000 (Polyethylene glycol) solution in 1:1 ratio and incubated at 4˚C for 45 minutes. This mixture was further centrifuged at 10,000g for 10 minutes, at room temperature (RT) to isolate the CICs. Further, acid dissociation of antigen and antibody present in CICs was performed, using glycine-HCl buffer as described by Gupta and Tan, 1981 with slight modification [27]. Concisely, the isolated CICs was dissolved in 200μl glycine-HCl, pH 1 and incubated at 47˚C for 30 minutes, to dissociate the antibody-antigen complex. Consequently, the acid dissociated CICs were neutralized with 2M NaOH. Quantification of glycosylated CICs was performed according to protocol described by Schömig et al, 2016 [28]. Briefly, acid dissociated CICs were treated with 100μl of acidified 0.06% PAS (HIMEDIA, India) in 7% acetic acid solution (SRL, India) and incubation of the reaction mixture was performed in dark at 37˚C for 90 minutes. Consequently, 100μl of Schiff reagent (HIMEDIA, India) was added to the above mixture and incubated for another 40 minutes in dark at RT, for color development. Absorbance reading was measured in colorimeter (Systronics, India) at 550nm and further validated on Smart-Spec TM Plus spectrophotometer platform (Bio-Rad, USA). Mucin from porcine stomach (PSM) (Sigma-Aldrich) was used as standard in the range of 0.125 mg mL -1 to 1 mg mL -1 .

Characterization of patient CICs by SDS PAGE analysis and gradient gel electrophoresis
CICs were isolated from sera of PKDL patients by PEG precipitation and analyzed using SDS PAGE (Laemmli, 1970). Next, the gels were stained with silver staining according to the method of Merril and Harrington (1984). Further, gels were subjected to a combination of PAS silver double staining, as described by Moller et al, 1995 [29], to identify PKDL disease specific glycoproteins. Briefly, after electrophoresis, gels were fixed with 10% (v/v) Trichloroacetic acid (TCA) in ddH 2 O and subsequently oxidized with PAS for 40-45 minutes at RT. Further, gel was treated with acetic acid solution (1% in ddH 2 O) followed by subsequent treatment with Schiff fuchsin and Sodium metabisulphite reduction, till the bands turn magenta. Silver staining was done by subsequent wash of the PAS stained gel with acetic acid, water and dichromate fixation followed by silver nitrate treatment to develop the gel with sodium carbonate in formaldehyde solution for 40-45 minutes. The combined PAS silver staining technique enabled the visualization of intensified disease specific glycoprotein bands [29][30].
In addition, protein bands were resolved in denaturing gradient gel electrophoresis. Briefly, 5%-15% of SDS PAGE gradient gel was prepared and 50μg of CICs sample were loaded for electrophoresis, which was performed at a fixed voltage of 110V for 1.5 hrs. Protein bands in the gel were visualized by coomassie brilliant blue stain [31].

Affinity purification of CIC antigens
To further characterize the CIC antigens, affinity purification of CIC antigens was performed using acid dissociated CICs as described by Gupta and Tan [27] with slight modification. Briefly, PKDL serum samples were pooled down (n = 20), precipitated with PEG and further treated with glycine-HCl as described earlier. CICs were affinity purified using equilibrated Protein A Sepharose 4B column (Invitrogen, USA) at RT, for 45 minutes. The antigen fraction free from antibody was thus collected. The CIC antigen fraction was subsequently dialyzed using Dialysis tubing (Sigma-Aldrich) with three PBS changes of 400 ml for 24 hours at 4˚C.
Immunoglobulin fraction bound to Protein A was further incubated for another 90 minutes followed by washing in PBS (0.1 M, pH 7.2). Subsequently, the immunoglobulin was eluted with citrate buffer (0.1M, pH 3.0). Protein concentrations were measured by Lowry method [32].

Deglycosylation of affinity purified CIC antigens
To characterize and reconfirm the presence of sugars in the affinity purified CIC antigens of PKDL patients, samples were deglycosylated using Arthrobacter ureafaciens neuraminidase (Sigma-Aldrich) treatment according to manufacturer's instruction. Briefly, 50μg of affinity purified CIC antigens were treated with 2 sigma units of neuraminidase in reaction buffer and incubated at 37˚C for 3 hours. The reaction mixture was heated at 100˚C for 5 minutes, followed by treatment with 5 units of N-Glycosidase F (Roche, USA) in reaction buffer [20mM sodium phosphate, 10mM EDTA, 0.5% (w/v) CHAPS and 0.05% (w/v) SDS] and incubated for 18 hours at 37˚C. Reaction was ultimately stopped by heating at 100˚C for 5 minutes [33].

Lectin blot of deglycosylated affinity purified CIC antigens
Glycans present in affinity purified CIC antigens were analyzed by DIG Glycan differentiation kit (Roche, USA) according to manufacturer's instruction. Briefly, 15μg of affinity purified CIC antigens (both enzyme treated and untreated) samples were run on 7.5% SDS PAGE gel, transferred to PVDF membrane followed by Lectin blot using DIG Glycan differentiation kit. Membrane was blot dried and analyzed [34].

Immunoblot of affinity purified CIC antigens resolved by SDS PAGE
Affinity purified CIC antigens (15μg) was electrophoresed in SDS PAGE as described earlier.

Statistics
Demographic analysis of the study population was performed using two tailed Chi-Square test. Statistical analysis on ELISA of IgG, IgM and IgG subclasses (IgG1, IgG2, IgG3 and IgG4), PEG index, PEG CIC and Glyco CIC assay data (2-tailed unpaired t-test for unpaired samples, with 95% confidence interval and One way ANOVA), ROC curve and longitudinal analysis (paired t-test for paired samples) was performed using Graph Pad Prism version 6.01for Windows (Graph Pad Software, San Diego, California, USA).Diagnostic performances of the assays were evaluated using MedCalc online calculator. The sensitivity and specificity for each assay were evaluated by using the equations: sensitivity = True positive cases/ (True positive cases+False negative cases)× 100% and specificity = True negative cases/ (True negative cases+ False positive cases)× 100%.Positive predictive values (PPV) and Negative predictive values (NPV) were determined using the equations: PPV = True positive cases/(True positive cases+ False positive cases) × 100% and NPV = True negative cases/(True negative cases+ False negative cases) × 100%.Receiver Operating Characteristic (ROC) curve have been constructed, summarizing the sensitivities and specificities of each assay.

Study population
In the present study, a total of 90 PKDL cases were recruited among which,56.7% (n = 51) were male and 43.3% (n = 39) were female patients. Further, among these 90 PKDL cases, the overall inclusive median age was 20 years (range = 5-60 years). The median age of females (20 years; range = 2-56 years) was comparable to that of males (20 years; range = 5-60 years). Additionally, eighteen PKDL cases were longitudinally monitored, after administration of Liposomal Amphotericin B/Miltefosine.
In the present study, the overall median time of manifestation of PKDL after VL treatment was 60 months (range = 12-300 months). PKDL lesions developed within 2 years in 1.2%, within 5 years in 15.5% and after five years or more in 83.3% of the cases, after apparent cure from VL. Among the enrolled PKDL cases, most of them were rk39 positive (97.8%, n = 88), however two cases recorded as rk39 negative, were confirmed parasitologically. Majority of cases with history of VL had been treated for VL with SSG (84.5%, n = 76), while the remaining were treated with Miltefosine (14.4%, n = 13) and Liposomal Amphotericin B (1.1%, n = 1).

Comparison of anti-leishmanial antibody titres between PKDL patients and controls
Leishmania infection mounts humoral responses against Leishmania antigen in the form of immunoglobulins of isotypes IgG, IgM and IgG subclasses, which are important serological markers in the host. To assess the status of primary immunoglobulin responses, present study investigated IgG, IgM and IgG subclass antibody titres in the sera of PKDL patients with respect to EC, non PKDL disease controls like Leprosy and Vitiligo as well asNEC. Parameters involving sera dilutions and antigen concentrations were standardized to set optimal ELISA condition for the serological assay. Anti-leishmanial antibody titres of IgG, IgG1, IgG2, IgG3 and IgG4 in sera of PKDL patients showed enhanced responses. Reactivity of Leishmania antigen was also investigated against antibody Isotype of IgM present in the sera samples. The performances of the sensitivities and the specificities of the ELISA with sera from patients with PKDL were also evaluated. The reactivity of the IgG, IgM, and IgG subclass antibodies from the sera of patients with PKDL (n = 90) were significantly higher than those of the sera from the controls, including NEC, EC, and patients with Leprosy and other skin disease like Vitiligo (p<0.0001). The 'mean ± S.E.M' of anti-leishmanial IgG antibody titres in sera of PKDL patients were 0.718 ± 0.032 (95% CI: 0.65-0.78) in comparison to the controls, which were 0.25 ± 0.007 (95% CI: 0.23-0.26) for patients with Leprosy, 0.23 ± 0.005 (95% CI: 0.22-0.24) for patients with Vitiligo, 0.23 ± 0.004 (95% CI: 0.22-0.24) for EC and 0.23 ± 0.004 (95% CI: 0.22-0.24) for NEC (Fig 1B). There is a sensitivity and specificity of 91.1% and 98.5%, respectively for IgG ELISA. Thus, among the 90 PKDL cases, 8 false negative cases were reported by IgG ELISA assay with 2 false positive cases among control cases. The cut-off was determined as 0.350. Through the assessment of ROC curve, AUC value of 0.95 was obtained, showing good performance of IgG ELISA for discriminating PKDL cases from controls (Fig 2B).
To differentiate PKDL serologically from the control populations, the Leishmania antigen specific isotype IgG titres were analyzed. High levels of anti-leishmanial IgG antibodies were observed in the sera of PKDL patients with respect to EC (p<0.0001) as well as disease controls (p<0.0001).

Glyco CIC assay
To quantify the glycosylated CICs in PEG precipitated sera, the Glyco CIC assay was employed. The assay was performed using only 10 μl of sera. The measurement was taken at 550 nm wavelength. Parameters such as hours of PAS and Schiff reagent treatment were optimized to set standardized assay time for the rapidity and reproducibility of the diagnosis. Finally, all sera samples were separated, from finger pricked amount of whole blood, for examination using Glyco CIC assay. The 'mean ± S.E.M' of Glyco CIC assay of PKDL patients was 0.458± 0.03 (95% CI: 0.39-0.51) as compared to controls, which was 0.051± 0.001 (95% CI: 0.04-0.05) for patients with Leprosy, 0.046± 0.001 (95% CI: 0.044-0.048) for patients with Vitiligo, 0.069± 0.002 (95% CI: 0.06-0.07) for EC and 0.062± 0.001 (95% CI: 0.05-0.06) for NEC ( Fig 1D). The test cut-off point was determined as 0.091. Sera from PKDL patients demonstrated significantly stronger (p<0.0001) recognition of glycosylated CICs than any of the control groups. Thus, complete agreement (κ = 0.944) was observed between the Glyco CIC assay and the reference rk39 test for diagnosis of PKDL. The overall sensitivity of the assay was 95.6% and specificity was 99.3%. There were only 4 false negatives and 1 false positive from 90 PKDL cases. ROC curve for Glyco CIC assay shows AUC value of 0.99, indicating excellent diagnostic efficacies (Fig 2D). Additionally, two PKDL cases which were negative to rK39 tests, but were confirmed parasitologically, were also found positive through Glyco CIC assay.

Glyco characterization of CIC
To characterize the CICs among different forms of Leishmaniasis as compared to other cross diseases and healthy controls, initially, PEG precipitated CICs from sera of PKDL, VL and Cutaneous Leishmaniasis (CL) patients, cross disease control (Leprosy, Vitiligo) and NEC were analyzed using one dimensional SDS PAGE, which revealed a differential banding pattern among different types of disease manifestations (Fig 3A). Gradient SDS gel electrophoresis of PKDL PEG precipitated CICs indicated the exclusive presence of ten bands of molecular weight (MW) 162, 109, 70, 61, 45, 41, 37, 31.6, 30, 23 kDa (Fig 3B Lane 1,4). Importantly, these bands were either absent or present with decreased intensity in CICs of treated PKDL patients (Fig 3B Lane 2, 3, 5 respectively). Also, CICs of VL patients and Leishmania membrane protein (L. donovani strain Ag83) specifically indicated the presence of 109 kDa band, which was also present in CICs of PKDL patients (Fig 3A Lane 2, 4 respectively). When SDS gel were subjected to silver staining alone, CICs from PKDL patients indicated presence of extra band of molecular weight (MW) 55 kDa along with 162, 109, 70, 61, 45, 41, 37, 31.6, 30, 23 kDa bands (Fig 3C Lane 6,7). For further characterization of specific protein bands in CICs of PKDL patients, gels were further subjected to PAS silver double staining, which revealed seven Glycoprotein bands of MW 162, 109, 70, 61, 55, 45 and 37 kDa (Fig 3D Lane 2-7). CICs of disease control and NEC indicated total absence of the corresponding bands (Fig 3D Lane 1, 8).
To characterize the antigens, PEG precipitated CICs from 20 PKDL patients were pooled down, acid dissociated and affinity purified using Protein A Sepharose 4B. The immunogenic nature of the affinity purified CIC antigens was deciphered through immunoblot analysis using purified anti-leishmanial antibody (Fig 4A). Protein bands of molecular weight 162, 109, 70, 61, 55, 45 and 37 kDa (Fig 4A Lane 1) in the affinity purified CIC antigens corresponded to similar molecular weight protein bands present in L. donovani antigen (Fig 4A Lane 2).
To further validate the presence of glycosylation in CIC antigens, lectin blot was performed (Fig 4B Lane 1, Lane 3, Lane 4) with lectin SNA (binding specificity to SAα2-6 Gal), GNA (binding specificity to Man) and DSA (binding specificity to Gal1-4GlcNAc) respectively, using the DIG Glycan differentiation kit, which revealed the presence of several protein bands of molecular weight (MW) ranging from 20-260 kDa (Fig 4B Lane 1). Additionally, to reconfirm the glycan nature of the CIC antigens, CICs were deglycosylated, using enzyme antibody based ELISA failed to differentiate follow-up from baseline cases in both drug responsive and unresponsive groups, where the mean titre showed insignificant differences for both drug unresponsive (p = 0.95) and drug responsive (p = 0.99) cases (Fig 5E and 5F).

Discussion
One of the earliest records of Indian PKDL cases acting as the sole reservoir for transmission of the fatal parasitic disease VL, was made by Addy & Nandy in 1992 [37]. They demonstrated the outbreak of the disease in two foci, one of these foci was in Malda district, in northern West Bengal and the other was located in the south of West Bengal. This study highlighted the important role played by PKDL cases in anthroponotic transmission of VL in West Bengal, India. Since then, there have been very few reports on PKDL epidemiology or surveillance from this part of India, until recently however, Ganguly et al 2015 [38] reported a large number of PKDL cases from endemic areas of Malda district. Subsequently, Sahaet al 2017 [39] have also reported a huge number of asymptomatic cases from these endemic areas. Thus, the accelerated VL elimination programme has resulted in increased case detection. The ongoing VL elimination survey in West Bengal has recorded a large number of cases in eleven endemic districts. Active survey in West Bengal has resulted in identification of 248 new cases in 2015 and 240 new cases in 2016 [40]. However, one of the major challenges faced by this programme is accurate and reliable diagnosis of PKDL cases, as a large percentage of the patient population have no history of VL and are often rk39 negative. The problem is further compounded with the fact that our available diagnostic arsenal fails to distinguish between fresh and cured cases. Further, the treatment of PKDL is also difficult and is yet not standardized, often the regime is prolonged and frequently patients discontinue therapy and thus, continue to serve as reservoirs of infection. Additionally, the drug responsiveness of PKDL patients also varies, where cure is early in some patients, whereas others are late responders. Thus, an appropriate tool for therapeutic monitoring is much needed, which would suffice the current need of the hour for successful implementation of the elimination programme. Government of India action plan encompassed major commitment towards elimination of VL by 2017 through memorandum of understanding signed by five countries [41]. Unfortunately, even after the active drive, India and Bangladesh still have a prevalence of around 8 to 16 cases per 10,000 populations in some of the endemic districts at block levels [42]. From the diagnostic point of view, parasite identification in slit skin smears and culture methods are confirmatory, but lack sensitivity, being only 40%-60% sensitive in case of nodular lesions of PKDL [43,44], whereas PCR demands non field set up and highly skilled personnel under stringent laboratory conditions, though being 93.8% sensitive [45]. Serological methods like rK39 dipstick tests, DAT, ELISA and Leishmanin tests although bears 91%-100% sensitivity and 0-100% in terms of differential diagnosis of Leprosy, Vitiligo and other diseases in the endemic zones, but have zonal variability of diagnostic efficacies. Serological diagnosis could only establish detection of 20% of PKDL cases, lacking reports of VL history and they seldom distinguish treated cases from early PKDL cases as these are all antibody based tests having persistence of antibody titres even after cure [46]. PKDL is a chronic disease. In India, it usually develops after several years of cure from VL. The causative parasite i.e., Leishmania donovani which persists in the skin comes in contact with high levels of antibody, but are not removed due to defective clearance and thus drive the development of ICs in circulation [47]. Formation of ICs is a normal part of the immune defense to remove pathogen. More recently, Jamal et al, 2017 [48] have demonstrated a 36 fold higher ICs in human VL subjects. In our present study, we observed almost 5 fold higher ICs in human PKDL subjects. Aberrant glycosylation of proteins in ICs often play important role in pathogenesis of various disease conditions like SLE, Rheumatoid Arthritis and Cancer [49][50][51][52]. Recently, CICs has also been identified to play active role in Leishmania infection [22,53]. In Rheumatoid Arthritis, there have been implications of glycosylation alterations in ICs progression to disease manifestation. Overwhelming data supports the role of glycosylation of IgG in human diseases involving human Rheumatoid Arthritis [54][55]. Recent studies have also demonstrated that glycosylation of protein antigens can immensely affect adaptive immune responses [56,57]. Taken together, glycoproteins have been utilized as diagnostic markers in innumerable number of diseases as they are known to possess disease specificity and are very reliable [58].
Preliminary studies from our laboratory have indicated that PKDL CICs are glycosylated [22]. Based on the above strong background literature, we explored the potential of glycated CICs towards PKDL disease diagnosis and prognosis. We developed a glycan based assay which quantifies the amount of glycosylated CICs in PKDL patients and thus helps in the diagnosis of these patients. The study was conducted on PKDL subjects collected from endemic districts of West Bengal. Our data further, revealed the presence of specific glycoproteins among PKDL positive individuals with various degrees of dermal manifestations such as Macular lesions, Papular, Mixed and Polymorphic lesions including Nodular types of lesions. These findings are consistent with our earlier report of identification and characterization of CICs in patients with VL and PKDL [22]. The enrolled PKDL patients in our study were mainly from north of West Bengal covering districts like Malda. Districts like Birbhum and Murshidabad were also surveyed, though the disease burden was comparatively less there. Most of the cases registered were macular PKDL (64.4%). Diagnosis of these cases are particularly difficult as the clinical presentations bear very close resemblance to cross diseases like Vitiligo or Leprosy and have scanty parasite load. The rK39 is undoubtedly a useful test, especially in a field setting, but it has low sensitivity for Asian macular PKDL [59]. In the present study, we employed the Glyco CIC assay to identify both rK39 positive and negative PKDL samples. The assay could successfully identify two parasitologically confirmed rK39 negative cases. For evaluating the diagnostic potential of our newly developed Glyco CIC assay, we compared it with other available serodiagnostic assays, namely PEG CIC assay, conventional parasite ELISA (anti-leishmanial IgM and IgG). The Glyco CIC assay yields superior sensitivity of 95.6% and specificity of 99.3%, whereas the NPV was 97.1% and PPV was 98.9%. On the other hand PEG CIC assay yielded a mere sensitivity of only 81.1% and specificity of 97.1%. Available and mostly practiced serodiagnostic assay like ELISA which employs Leishmania antigen for detection of serum anti-leishmanial antibodies level yielded a sensitivity of 91.1% and specificity of 98.5% for IgG and the corresponding sensitivity and specificity for IgM were 28.9% and 99.3%, respectively (Table 3).
Thus, Glyco CIC assay had remarkable sensitivity and specificity limits, though it failed to diagnose 2 macular and 2 polymorphic PKDL cases from Malda district. Interestingly, glycated CICs as quantified by our new assay was significantly high for PKDL patients in all the studied districts namely Malda (p<0.0001), Birbhum (p<0.0001) and Murshidabad (p<0.0001). Additionally, the PKDL CICs titre OD as evaluated through Glyco CIC assay scored 5.03 fold the OD cut off value. We further statistically evaluated our findings based on ROC curve for Glyco CIC assay compared with available serodiagnostic assay and obtained AUC value of 0.99 for Glyco CIC assay, 0.96 for PEG CIC at 450 nm, 0.95 and 0.87 for ELISA based antibody titres of IgG and IgM, respectively (Fig 2). Thus, taken together, Glyco CIC assay had an impressive diagnostic efficacy as compared to the other available assays.
As our study demonstrated high titres of glycosylated CICs in PKDL patients, we then systematically characterized the glycosylation profile of PKDL CICs. SDS and gradient gel analysis revealed the presence often major protein bands of molecular weight (MW) 162 kDa, 109 kDa, 70 kDa, 61 kDa, 45 kDa, 41 kDa, 37 kDa, 31.6 kDa, 30 kDa and 23 kDa exclusively in PKDL patients. Among them, distinct seven bands were found to be glycoprotein in nature as demonstrated by PAS silver double stain and lectin blot (Fig 3D, Fig 4B). Notably, the 162 kDa, 109 kDa, 70 kDa MW CIC protein bands have similar MW as glycoproteins found on amastigotes and promastigotes of Leishmania donovani [60], whereas other protein bands like 61 kDa, 55 kDa, 45 kDa and 37 kDa had MW similar to those found in Leishmania antigen [48]. All the above mentioned seven protein bands were present in immunoblot (Fig 4A). These glycoprotein biomarkers were present in all the patients enrolled from the different districts of West Bengal. These bands were totally absent in other cross diseases (Leprosy and Vitiligo) and Healthy controls (EC and NEC). Further, longitudinal monitoring of PKDL patients revealed the disappearance (162 kDa, 109 kDa, 70 kDa, 45 kDa, 37 kDa) /decreased (61 kDa) band intensity in drug responsive patients, whereas these bands persisted in the drug unresponsive group (Fig 5B). This thus, highlighted the prognostic application of these newly identified PKDL glycoprotein biomarkers.
Taken together, our study demonstrated that the newly developed Glyco CIC assay could successfully differentiate between PKDL drug responsive vs. drug un-responsive patients. Thus, among the 18 patients, who were monitored longitudinally, eleven PKDL patients were recognized as clinically cured and had no dermal lesions after treatment with Miltefosine. These patients showed almost 3 fold reduced PSM equivalent CICs concentration (Fig 5). Whereas,7 such patients, who were kept under LAMB regime, did not show regression of the dermal lesions, and showed almost 2.03 fold increased PSM equivalent CICs concentration in their follow up, reflecting persistence of the infection (Fig 5). PSM has been used as a standard for our newly developed Glyco CIC assay, as it is a rich source of naturally occurring glycoproteins with high carbohydrate content. The standard has helped us to quantitate the amount of glycosylated CICs present in the sample. Glyco CIC assay may thus, serves as an efficient prognostic evaluator of PKDL patients. In contrast, the conventional parasite ELISA (IgG) failed to distinguish between these groups (Fig 5). The titre value 'mean ± S.E.M' value of 1881 ± 462.3 μg/ml was obtained for drug unresponsive cases at presentation with respect to follow up mean titre value of 1970 ± 1555 μg/ml. Similarly, for drug responsive cases 'mean ± S.E.M' value of 1879 ± 692.9μg/ml at presentation with respect to follow up mean titre value of 1881 ± 835.4 μg/ml was observed (Fig 5E).
Glyco CIC assay utilizes the glycation status of CICs for identification and prognostic evaluation of PKDL patients. The assay is efficient and very cheap. Although PCR assay for identification of parasite load in PKDL patient stands as gold standard, but it requires central laboratory facilities. Also, it requires high end instruments and expertise. Glyco CIC assay on the other hand, provides an easy alternative for diagnostic and prognostic evaluation of PKDL patients. It does not involve much expertise and high end instruments. Also, the assay reagents are cost effective and can be used for large sample sizes. Our findings highlight the efficacy of Glyco CIC assay over available invasive methods like punch biopsy and tissue aspiration. The assay has potential for working both as a diagnostic and prognostic evaluator of Indian PKDL patients. It may also be used as an alternative to highly invasive tissue aspiration techniques particularly in primary health care centers. The assay requires very less blood sample, hence finds advantage over painful invasive techniques. Thus, the assay can be a potent tool for efficient case detection and could go hand in hand with the VL management and elimination program.