Detection of new Mycobacterium leprae subtype in Bangladesh by genomic

4 Maria Tió-Coma, Charlotte Avanzi, Els M. Verhard, Louise Pierneef, Anouk van Hooij, 5 Andrej Benjak, Johan Chandra Roy, Marufa Khatun, Khorshed Alam, Paul Corstjens, 6 Stewart T. Cole, Jan Hendrik Richardus, and Annemieke Geluk 7 8 From the Department of Infectious Diseases, Leiden University Medical Center, Leiden, The 9 Netherlands; Global Health Institute, Ecole Polytechnique Fédérale de Lausanne, Lausanne, 10 Switzerland; Rural Health Program, The Leprosy Mission International Bangladesh, Nilphamari, 11 Bangladesh; Department of Cell and Chemical Biology, Leiden University Medical Center, Leiden, 12 The Netherlands; Institut Pasteur, Paris, France; Department of Public Health, Erasmus MC, 13 University Medical Center Rotterdam, Rotterdam, The Netherlands. 14 § Current laboratory: Mycobacteria Research Laboratories, Colorado State University, Fort Collins, 15 CO, USA (Charlotte Avanzi); Department for BioMedical Research, University of Bern, Bern, 16 Switzerland (Andrej Benjak). 17 18 RUNNING TITLE: M. leprae genotypes in Bangladesh 19

Leprosy is a complex infectious disease often resulting in severe, life-long disabilities and still poses a 5 serious health threat in low-and middle income countries (WHO, 2019). Despite the very limited M. 6 leprae genome variability (Singh and Cole, 2011), the disease presents with characteristically 7 different clinico-pathological forms (Ridley and Jopling, 1966) due to genetically dependent 8 differences in the immune response to the pathogen, resulting in the WHO classification from 9 paucibacillary (PB) to multibacillary (MB) leprosy (Kumar et al., 2017). Notwithstanding the efficacy 10 of multidrug therapy (MDT), approximately 210,000 new cases are still annually diagnosed and this 11 incidence rate has been stable over the last decade (WHO, 2019). Aerosol transmission via respiratory 12 routes is generally assumed to be the most probable way of bacterial dissemination (Bratschi et al.,   22 causing it to be an obligate intracellular pathogen which cannot be cultured in axenic media that 23 requires support of a host to survive. This poses major limitations to obtain sufficient bacterial DNA  22 2018). However, these methods were never applied on challenging samples such as slit skin smears 23 (SSS) and nasal swabs (NS) containing a low amount of bacterial DNA compared to skin lesions of 24 patients. 25 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  the potential role of asymptomatic carriers, we further explored the diversity and transmission of M. 22 leprae in four districts of the northwest of Bangladesh. We collected SSS and NS of 31 leprosy 23 patients with a high bacterial load as well as 279 of their household contacts and characterized M. 24 leprae DNA by WGS or Sanger sequencing. The resulting genotypes were correlated to the subjects' 25 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. ; https://doi.org/10.1101/2020.03.05.20031450 doi: medRxiv preprint GIS location. Additionally, this is the first study to examine M. leprae DNA detection in comparison 1 to anti-PGL-I IgM levels in plasma measured by up-converting phosphor lateral flow assays (UCP-2 LFAs).
3 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. in the districts where this study was performed was 0.9 per 10,000 and the new case detection rate 1.18 8 per 10,000 (Rural health program, the leprosy mission Bangladesh, yearly district activity report 2018). 9 For M. leprae detection and characterization, SSS from 2-3 sites of the earlobe and NS (tip wrapped 10 with traditional fiber, CLASSIQSwabs, Copan, Brescia, Italy) were collected and stored in 1 ml 70%  18 about the study objectives, the samples and their right to refuse to take part or withdraw without 19 consequences for their treatment. All subjects gave informed consent before enrollment and treatment 20 was provided according to national guidelines. 21 DNA isolation from slit skin smears and nasal swabs 22 DNA was isolated using DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA) as per manufacturer's 23 instructions with minor modifications. Briefly, tubes containing 1 ml 70% ethanol and SSS were 24 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. ; https://doi.org/10.1101/2020.03.05.20031450 doi: medRxiv preprint amplified using the following profile: 2 min at 50ºC and 10 min at 95ºC followed by 40 cycles of 15 s 1 at 95ºC and 1 min at 60ºC with a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems). 2 Presence of M. leprae DNA was considered if a sample was positive for RLEP qPCR with a cycle 3 threshold (Ct) lower than 37.5 or was positive for RLEP PCR at least in two out of three indecently 4 performed PCRs to avoid false positives. 5 Library preparation and enrichment 6 A total of 60 DNA extracts were selected for sequencing, including 30 from SSS and 30 from NS 7 ( Figure S1, Supplementary Data 1). At least one sample from each index leprosy patient was selected 8 as well as RLEP positive samples of HHC or patients who were household contacts of the index case 9 (selection based on Ct value and household overlap). A maximum of 1µg of DNA in a final volume of 10 50µL was mechanically fragmented to 300 bp using the S220 Focused-ultrasonicator (Covaris) 11 following the manufacturer's recommendations and cleaned-up using a 1.8x ratio of AMPure beads. Up 12 to 1µg of fragmented DNA was used to prepare indexed libraries using the Kapa Hyperprep kit 13 (Roche) and the Kapa dual-indexed adapter kit as previously described (Benjak et al., 2018) followed 14 by two rounds of amplification. All libraries were quantified using the Qubit fluorimeter (Thermo 15 Fisher Scientific, Waltham, MA), and the fragment size distribution was assessed using a fragment 16 analyzer. 17 Libraries were target enriched for the M. leprae genome using a custom MYbaits Whole Genome 18 Enrichment kit (ArborBioscence) as previously described (Honap et al., 2018). Briefly, biotinylated 19 RNA baits were prepared using DNA from M. leprae Br4923. A total of 1500 ng of each amplified 20 libraries was used for enrichment. Each library was pooled prior to enrichment with another library 21 with similar qPCR Ct value. Enrichment was conducted according to the MYbaits protocol with the 22 hybridization being carried out at 65 °C for 24 hours. After elution, all pools were amplified using the 23 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. ; https://doi.org/10.1101/2020.03.05.20031450 doi: medRxiv preprint Kapa amplification kit with universal P5 and P7 primers (Roche). All amplification reactions were 1 cleaned up using the AMPure beads (1X ratio).  16 Genotyping and antimicrobial resistance by Sanger sequencing 17 To further characterize the M. leprae strains for which the whole genome sequence was not obtained, 18 specific primers were designed to perform Sanger sequencing based on unique SNPs (Table S3 and S4) 19 of each index case strain. Additionally, Sanger sequencing was performed after amplifying several loci 20 (Table S2)  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. ; mixes. DNA was denatured for 2 minutes at 95ºC, followed by 45 cycles of 30 s at 95ºC, 30 s at 50-58 1 ºC and 30 s at 72 ºC and a final extension cycle of 10 min at 72ºC. PCR products were resolved by 2 agarose gel electrophoresis as explained above. PCR products showing a band were purified prior to 3 sequencing using the Wizard SV Gel and PCR Clean-Up System (Promega). Sequencing was 4 performed on the ABI3730xl system (Applied Biosystems) using the BigDye Terminator Cycle 5 Sequencing Kit (Thermo Fisher Scientific). Sequences were analyzed using Bioedit v7.0.5.3. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

M. leprae detection in patients and healthy household contacts
2 At diagnosis of the index cases and recruitment of contacts into this study 250 household contacts had 3 no signs or symptoms of leprosy or other diseases (HHC), whereas 22 household contacts were 4 diagnosed as PB and seven as MB patients (Table S1, Supplementary Data 1). 5 Presence of M. leprae DNA was determined by RLEP PCR or qPCR in SSS and NS of leprosy patients 6 and HHC ( (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. following genotypes were found for these 21 subjects: 1A (n=5), 1B (n=4), 1C (n=3) and 1D (n=9). 4 Interestingly, the four newly sequenced 1B genotype strains do not cluster with the two previously 5 described 1B strains from Yemen and Martinique ( Figure 2). Instead, they form a new cluster in the 6 phylogenetic tree located between genotypes the 1A and 1B, which we refer to as 1B-Bangladesh 7 (Figure 2, blue, Supplementary Data 1). Using Sanger sequencing, the M. leprae strain for eight 8 additional individuals were determined as 1A (n=4) or 1D (n=4). Three subjects carried genotype 1 but 9 subtype could not be established (Supplementary Data 1). 10 The SNP used to differentiate genotype 1C (A61425G; Met90Thr, mutated in genotypes 1D and 2-4) is 24 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. The most prevalent M. leprae genotype in the studied area of Bangladesh is 1D, found in 55% of the 2 individuals (n=16, Table 1, Supplementary Data 1), followed by 1A in 31% (n=9), and 1B-Bangladesh 3 in 14% (n=4). Genotype 1D is the most widely distributed throughout the whole area studied (Figure 3, 4 blue and purple), whilst genotypes 1A and the here identified genotype 1B-Bangladesh are only 5 observed in the eastern area (green and orange respectively). The latter genotype was found in 4 6 individuals: two from the same household and two unrelated subjects residing 56, 51 and 11 km from 7 each other. However, due to privacy regulations on patient information to third parties it could not be 8 established whether subjects in different households had had contact with any of the others. 9 In a total of four households the same M. leprae genotype was detected in two individuals 10 (Supplementary Data 1). In the first household, both subjects were MB patients and WGS showed no 11 genetic variation at all between both patients' genomes (RB001 and RB003, 1B-Bangladesh genotype, 12 Supplementary Data 2). In the second household with two MB patients, the M. leprae whole genome 13 was only obtained from the index case but the same genotype, 1A, and a strain-specific SNP of the 14 index case (Table S3 and S4) was also identified by Sanger sequencing in the other patient (RB182 and 15 RB266). In the last two households, the genotype of strains from both MB index cases' were 16 determined by WGS (RB030, genotype 1D) and, by Sanger sequencing (RB065, genotype 1D-esxA), 17 while the M. leprae genotype 1 was located in the NS of both HHC but no further subtyping was 18 possible. 19

20
M. leprae whole genomes of six patients were successfully recovered from both SSS and NS. Genomic 21 comparison showed no differences between DNA from SSS and NS for two patients: RB001-RN001 22 (genotype 1B-Bangladesh) and RB048-RN059 (genotype 1D-esxA, Supplementary Data 2, Figure 2). 23 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. In a third patient (RB073-RN084, genotype 1A), both strains were identical except that in the NS strain 1 13% of 32 reads in ml1512 harbored a T1824441C (Gly56Asp) ( Table 2) In press). 16 The patient with the M. leprae strains that were the most genetically different between the NS and SSS 17 carried the genotype 1B-Bangladesh (RB069 and RN165). The NS strain had a mixed population in 18 glpQ (25% of 257 reads C9231T, Leu34Phe) and ml1752 (16% of 307 reads C2121552T, Val226Ile). 19 These genes encode a glycerophosphoryl diester phosphodiesterase and a conserved hypothetical  For 11 patients a whole genome sequence was recovered only from SSS but Sanger sequencing was 22 successfully performed to identify the subtype in NS. The same subtype observed in SSS was also 23 found in the NS of these 11 patients. Moreover, unique M. leprae SNPs identified in the genomes of 24 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. ; https://doi.org/10.1101/2020.03.05.20031450 doi: medRxiv preprint the SSS (Table S3 and S4) were also detected in seven of the genomes of the NS of these patients 1 (Supplementary Data 1).
2 Combining host and pathogen detection 3 Anti-PGL-I IgM levels were determined in plasma of 308 subjects. All MB patients with BI 2-6 (n=33) 4 showed high levels for anti-PGL-I IgM (Table 3) 8 Of the four contacts who developed leprosy within the first year after sample collection, two were 9 positive for anti-PGL-I IgM whilst negative for RLEP PCRs 10 and 12 months before diagnosis. Since 10 the two other subjects had a positive RLEP PCR in SSS or NS 5 or 8 months before diagnosis, it can be 11 concluded that all of the new cases showed positivity either for host-or pathogen-associated 12 diagnostics 5-12 months before developing disease. 13 Individual anti-PGL-I levels were compared to RLEP Ct values in SSS and NS samples (Figure 4), 14 showing an expected negative correlation between anti-PGL-I ratio and Ct value since both values are 15 associated with BI. A subtle difference can be observed in the correlation between anti-PGL-I IgM 16 levels and RLEP Ct if the qPCR was performed on either SSS or NS DNA, with a coefficient of 17 determination (R 2 ) 0.73 and 0.69 respectively. 18 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.   (n=28 and   16 n=104, respectively versus n=250 HHC in this study); ii) we conducted a more stringent approach by 17 testing the samples in three independent PCRs; and iii) the epidemiology and incidence of MB cases in 18 India and Brazil differ from the studied area in Bangladesh where MB leprosy cases occur less 19 frequently than PB and also usually display a low BI (Richardus et al., 2017). (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  3 However, this difference was not significant. In line with that study, we found that M. leprae DNA 4 detection was slightly higher (25% vs 18% in NS and 25% vs 12% in SSS) in contacts who developed 5 disease compared to those who did not. Additionally, we determined anti-PGL-I IgM levels, which 6 correlated well with Ct qPCR values. Notwithstanding this correlation, serology provided added value: 7 when positivity in any of the three techniques was considered (NS PCR, SSS PCR or anti PGL-I), all of 8 the contacts (n=4) who developed leprosy within the first year after sample collection, were identified. 9 In agreement with this, a combination of host and pathogen markers was previously integrated in a 10 machine learning model using qPCR and serological data (antibodies against LID-1 or ND-O-LID) 11 (Gama et al., 2019) to identify prospective leprosy patients among contacts leading to an increased 12 sensitivity in diagnosis, particularly in PB leprosy. It is of note that in our study, three of the four 13 contacts who developed leprosy were genetically related to the index cases in their households,  This new genotype is thus far restricted to Bangladesh and two of the four individuals carrying this 25 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. does not form an independent subtype but actually belongs to subtype 1D. SNP61425 used to 8 distinguish genotypes 1A-C is located at esxA encoding the virulence factor ESAT-6 (Monot et al., 9 2009). The Esx protein family also revealed high diversity in the more pathogenic mycobacterium, M.  16 mutations resulting in antimicrobial resistance and resistance to up to two different drugs was detected. 17 In our study, which is the first investigating M. leprae drug resistance in Bangladesh, we detected no 18 resistance by WGS, however, a partial missense mutation in the codon for Ser456 of the rpoB gene 19 potentially leading to rifampicin resistance (n=1) was observed by Sanger sequencing. This could be 20 the result of a mixed infection or an emerging mutation of the M. leprae strain occurring in the patient. 21 Silent mutations in the rpoB gene were detected in several locations, which indicates that mutations do 22 occur, and this may eventually lead to missense mutations conferring antimicrobial resistance. 23 However, drug resistance is not only induced by genetic mutations in drug targets, efflux systems 24 resulting in antimicrobial resistance have also been described for M. leprae (Machado et al., 2018). 25 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  This could be an indication of in-host evolution in the nasal mucosa, mixed infection or mixed 16 colonization. Thus, it may imply that colonization occurred with two different strains causing a co- 17 infection or that one is present, likely from a later colonization, but does not cause the disease. 18 The presence of mixed infections emphasises once more the importance of monitoring asymptomatic 19 carriers, who may contribute to the spread of the pathogen. Therefore, providing PEP only to the 20 (close) contacts of leprosy patients might not be sufficient to stop transmission. Instead, an approach 21 including the entire community but targeting only individuals testing positive for M. leprae DNA or 22 host immune markers associated to M. leprae infection, would represent a preferred strategy for PEP. 23 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  8 Consortium. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  2011). This SNP is also found in strains from the genotype 3I and 2E (Figure 2, green). 1* are samples 9 with genotype 1 for which the subtype could not be determined due to DNA concentration limit.
10 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  Genomic differences between M. leprae genomes obtained from slit skin smears (SSS) and nasal swabs 2 (NS) of the same MB patient. Percentage of mutated reads and total number of reads aligned at the 3 position of the mutation. No differences were found between the SSS and NS genomes of two patients: 4 RB001-RN001 and RB048-RN059. 5 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020. 14 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted April 18, 2020.  Figure S1. Samples analysed by whole genome sequencing.