Identification of hepatitis B virus genotype I in Thailand

The rare hepatitis B virus genotype I (HBV‐I) classification includes complex A/G/C/U recombinants identified amongst the individuals from China, India, Laos, and Vietnam. Herein we report the first HBV‐I specimen from Thailand, with detectable HBsAg despite a 10–amino‐acid truncation. This HBV‐I genome has a similar recombinant pattern to reference strains, including a C region that branches basal to references, suggesting a premodern era recombination event gave rise to HBV‐I. With an average sequence divergence from other genotypes ranging from 7.66% (standard deviation [SD], 0.42%; C) to 14.27% (SD, 0.31%; H), this new genome supports the HBV‐I classification as a unique genotype.


| INTRODUCTION
Despite introduction of a vaccine for hepatitis B virus (HBV) in 1982, the global burden of HBV remains high, with an estimated 257 million chronic infections globally, of which more than 90% remained undiagnosed in 2015. 1 Accurate diagnosis of HBV is a critical initial step in caring for patients with HBV and preventing new transmissions; however, the diversity of HBV sequences presents a challenge for diagnostic tests. 2 It has been estimated that HBV was first introduced into the human population approximately 33 600 years ago, 3 and since then it has evolved to nine unique genotypes (A-I) and a growing list of more than 30 subgenotypes. [4][5][6] This diversity is largely driven by the error-prone viral reverse transcription polymerase, which lacks proofreading abilities and intergenotypic recombination. 6,7 HBV genotypes and subgenotypes are unevenly distributed, with each HBV genotype predominating in unique regions of the world. 4 In particular, HBV genotype I (HBV-I) has been found in China, [8][9][10][11] India, 12,13 Laos, 14 and Vietnam, [15][16][17] with 39 unique sequences confirmed cases in Asia to date. 5 In addition to these cases, HBV-I strains have also been sequenced from Vietnamese immigrants living in Canada 18 and France, 19 indicating that a more widespread distribution may exist for HBV-I outside of the Asia. Although HBV-I sequences branch together in a phylogenetic tree, recombination analysis has determined that HBV-I sequences are complex recombinants of genotypes A, C, G, and unclassifiable sequences with exact breakpoints varying between individual strains. 7,8,12,[14][15][16]18,20 Classification of these strains as HBV-I or A/C/G/U recombinants is a topic of debate 7,8,21,22 since the classical definition of a unique HBV genotype is at least 8% intergenotypic divergence, yet HBV-I strains are differ from the closest relative, genotype C, by only 7 to 7.9%. 8,12,14,15,18,22 Thus, in additional, putative HBV-I genome sequences will be required to determine the appropriate classification for these unique sequences.
HBV coinfection with human immunodeficiency virus (HIV) puts patients at higher risk for severe liver disease, making prevention of such infections a priority. 1 The recent evaluation of HIV pre-exposure prophylaxis (PrEP) regimens by several groups has led to the successful implementation of PrEP programs amongst high-risk populations. 23,24 Emerging evidence supports the safe use of PrEP amongst HBV carriers. 25 In a recent assessment of the safety of PrEP for patients with active HBV infection, serial blood samples were collected from 12 participants who had chronic HBV infections at the time of enrollment to monitor their infections with and without PrEP treatment during the iPrEx study. 25 HBV sequences obtained from these individuals indicated that resistance mutations did not develop amongst the participants on PrEP, supporting the safe use of PrEP for chronic HBV patients. 25 The samples were collected from diverse geographical regions (United States, Thailand, Peru, Ecuador, and South Africa), although the HBV genotypes for these infections were unknown. To evaluate the HBV diversity encompassed within the iPrEx study, we characterized the sequences generated from serial blood draws of the chronic HBV participants and identified an HBV-I infection among them, which we characterized in depth to expand the known diversity of the genotype I classification.

| Samples
The sample cohort has been described previously, 25 and a complete diagnostic profile of the HBV-I infection time course can be found in Figure 1C of Solomon et al. 25 Sequences were generated from leftover cryopreserved plasma collected at the screening visit (day 12), 0, 79, 112, 251, and 336 days timepoints. All specimens were collected with informed consent from donors as approved by the University of California San Francisco IRB.

| Sequencing
Subgenomic sequences were generated for the surface antigen/reverse transcriptase (SRT) region as previously described. 26 The HBV-I genome was amplified in two overlapping fragments using Adventage

| Serology testing
The HBV-I specimen collected at the study entry day 0 was tested at a 1:500 dilution using ARCHITECT HBsAg Qualitative II, and the ARCHITECT CORE assays, and neat using the ARCHITECT HBeAg assay in combination with the ARCHITECT HBeAg Quantitative Calibrators (Abbott Diagnostics, Abbott Park, IL,)

| DISCUSSION
The unusual complex recombinants classified as HBV-I include a growing number of sequences from individuals born in Asia, where HBV is highly endemic. 4 The HBV-I genotype has been defined by strains having 7% to 8% nucleotide divergence to the closest relative genotype C 8,12,14,15,18,22 and the new recombinant genome sequence presented here, 9736409, meets this criteria with a mean nucleotide divergence of 7.66% (SD, 0.42%) to genotype C reference sequences.
Therefore, this sequence is classified as genotype I, making it the first HBV-I strain identified from Thailand. With a HBsAg serotype of adw2, this specimen meets the criteria for the I1 subgenotype. 28 Like other sequences classified as genotype I, 9736409 is a complex U/C recombinant with a similar recombination pattern and breakpoints (1500-2900) as other strains from China and Vietnam ( Figure 2). 8,12,14,15,18,22 Although other HBV-I strains also have A-and G-like regions [13][14][15][16][17] (Vietnam refs, Laos, India), these segments could not be classified in the 9736409 genome due to low bootstrap support ( Figure 2). Notably, genotype G sequences are quite rare in HOLZMAYER ET AL.

| 719
Asia, with all reported strains coming primarily from the United States, France, and Germany. 4 However, a genotype G/C recombinant (CU400) has been reported in Thailand, 29 indicating that genotype G strains might be present in the region. Yet the 9736409 strain is unlikely to be closely related to the CU400 strain since the C-regions of the two sequences have different breakpoints and the two strains only share 89% sequence identity. 29 Consistent with previous HBV-I genome analyses, 8,13,16 the region of the 9736409 genome classified as genotype C branched basal to reference strains ( Figure 1C). This suggests that the recombination events that resulted in HBV-I did not occur recently, although in additional, analysis and sequences will be required to explore the origins of HBV-I further. Likewise, the unclassifiable region from 1 to specimen collected upon enrollment in the iPrEx study. 25 Furthermore, the specimen was HBeAg negative as expected due to the precore amino acid 28 stop codon (G1896A). As HBV sequence diversity continues to expand, persistent global HBV surveillance remains essential to ensuring diagnostic tests can detect all strains regardless of their sequence or geographical location.

ACKNOWLEDGMENTS
The authors are grateful to the entire iPrEx study team and participants for their contributions to this study. In particular, the authors acknowledge Pongpun Saokhieo and Dr Suwat Chariyalertsak from the iPrEx team in Thailand for collecting the HBV-I specimen. The authors also thank Dr Catherine Brennan for initiating this project and Barbara Harris for technical input.