Identification and Characterization of Hepatitis B Virus Immune Escape Mutants in Kenya ()
1. Introduction
Hepatitis B Virus (HBV) infections affect about 400 million people globally about 1.4 million HBV deaths annually with more than 200,000 of these deaths coming from Africa. Kenya has a prevalence rate of >8.0% HBsAg. Several HBV mutants with clinical implications have been identified and documented worldwide, indicating the potential for spreading and developing their own epidemiology [1] [2]. They include the HBV polymerase gene mutants, S genes mutants, pre-core/core, and X mutants [1] [2]. However, mutational occurrence within or around the major “antigenic alpha” determinant region (121 - 149 aa) of HbsAg S gene, causes conformational changes. This affects HBV antigenicity and induces protective antibodies that make the virus go undetected by the rapid HBsAg test in some patients [3] [4]. Additionally, mutations occur in the S region, which permits the virus to evade/escape immune surveillance such as the body’s own immune response, the immunological response caused by vaccination, and anti-HBV immunoglobulin therapy [5] [6]. Thus, these immune escape mutations provide false-negative results by commercial HBsAg assays (occult hepatitis B), failure to vaccine-induced immunity and anti-HBV immunoglobulin therapy which is linked to HBV reactivation, increased mutant viral replication, immunosuppression antiviral treatments failure and hepatic damage [5]. These highlights show the clinical importance of early, cost-effective, accurate, and sensitive detection methods of the viral pathogens, genotypes and mutations that confer resistance to therapy and immune escape profiles [7]. This is key for the effective prevention and clinical management of this disease [2]. The findings can be used to guide effective disease transmission prevention, enables practicing physicians to identify those patients at risk of disease progression and guide the appropriate/suitable first-line anti-viral treatment for patients. It can also monitor the efficacy and modification of antiviral therapy in patients to avoid the clinical effects of resistance [7] [8] [9] [10] [11]. Thus, ensuring proper clinical management of infected patients with first-line potent antiviral therapy [8] [9]. This can increase a patient’s survival, safeguard the public’s health, and achieve long-term treatment success [4] [5] [9].
2. Materials and Methods
2.1. Retrieval of Kenyan-Sourced HBsAgGene Sequences from in NCBI and HBV Databases and Their Analysis
From 547 HBV sequences available from Kenya in NCBI and HBV databases, 120 full large Surface (S) gene sequences containing the “a” determinant region of the gene were retrieved. These sequences were translated to the protein sequences using EMBOSS Transeq (https://www.ebi.ac.uk/Tools/st/) and aligned with the referenced amino acid sequence from HBV database in Bioedit software version 7.2.5. Mutations or amino acids exchange in each sequence was identified, and recorded in an excel table. The clinical significance of mutations was flagged out from the search of peer-reviewed journals using Pubmed search engine and later submitted to geno2pheno HBV online software (https://hbv.geno2pheno.org/index.php) version 2.0 for further confirmation of drug resistance and immune escape mutation identification.
2.2. Clinical Samples Collection and Processing
A total of forty-six HBV-positive blood donor samples without personal identifiers/information were obtained from the National Blood Transfusion Centers in Kenya. Hepatitis B virus template DNA was extracted from 400 μl positive HbsAg serum samples with an automatic ExiPrep Dx Viral DNA kit (Bioneer, Korea) on an ExiPrep 48 Dx internal system according to the manufacturer’s instructions, and 50 μl of elute recovered and stored at −20˚C till use.
2.3. Amplification of the HBV S Gene
Amplifications of the HBV S gene were run by nested PCR using Veriti PCR System. First round of the PCR was performed using specific outer primer pair 230F (5’-TCACAATACCGCAGAGTCT-3’) and 800R primers (5’AACAGCGGTATAAAGGGACT-3’) targeting position 1 to 571 bp of S gene. The PCR was performed at 40 cycles using the following thermocycling parameters, 94˚C for 5 min; followed by 45 cycles at 94˚C for 4 min; 94˚C for 45 seconds, annealing at 55˚C for 1 min, extension at 72˚C for 2 min and final extension at 72˚C for 7 minutes. The Second round nested PCR was conducted using specific Inner primers pair P7(+) (5’-GTGGTGGACTTCTCTCAATTTTC-3’) and P8(−) primers (5’-CGGTATAAAGGGACTCAAGAT-3’) targeting position 1 to 541 bp of S gene. Under the same thermocycling conditions as of the 1st PCR except that annealing at 50˚C with 20 μl master mix and 5 μl of the first-round product as per the method of Lindh et al. (2007). All amplified PCR products were examined on a 2% agarose gels stained with Gel Red and viewed using a UV transilluminator. A digital image of the gel was taken using a gel documentation system. The products were then purified, prepared for sequencing and directly sequenced targeting position 1 to 571 bp of S gene. Using the same PCR primer pair and an automated sanger technic Applied Biosystems ABI3500xl sequencer (Macrogen, Netherlands).
2.4. Sequence, Genotyping, and Mutations Analysis of S Gene Sequences
All Forward and reverse sequences obtained by the sanger technique were analyzed, edited and assembled using Finch TV (http://www.geospiza.com/Products/finchtv.shtml), and Bioedit sequence alignment editor software version 7.2.5 [12] and then subjected to NCBI nucleotide BLAST for quality check. Of the 46 sequences, only 17 were of good quality, qualified for genotyping and mutation analysis. Genotypes and subgenotypes of the 120 retrieved sequences from the database and 17 assembled HBV sequences obtained from blood donors were determined using geno2pheno HBV online software (https://hbv.geno2pheno.org/index.php) version 2.0 [13]. Mutation analysis of sequences from donors was carried our as described previously at the beginning of this section.
2.5. Phylogenetic Analysis
A phylogenetic analysis was performed using MEGA version 11 [14]. The analysis was done using the Maximum-likelihood statistical method, Tamura-Nei model, and the bootstrap method of 1000 replicates.
2.6. Ethical Approval
Jomo Kenyatta (JKUAT) Institutional Ethical Review Committee approved this study protocol (Ref. No. (IERC) (JKU/IERC/02316/0508)).
2.7. Data Analysis
All the statistical analyses were performed using SPSS version 19.0 and GraphPad software for P-Value.
3. Results
3.1. In-Silico Identification of the Predominant and Clinically Significant Mutations of HBsAgGene in Kenya from NCBI/HBVdb Databases
A total of one hundred (120) sequences containing complete Hepatitis B Surface Antigen gene sequences from Kenya were retrieved from NCBI database/HBVdb databases and analyzed. Out of these sequences, 79 different types of mutations were identified of which 15 mutations were of clinical importance with the occurrence frequency of at least 5% were identified (Table 1).
![]()
Table 1. Predominant and clinically important mutations of HBsAggene in Kenya as retrieved from NCBI database/HBVdb databases.
Majority (97.5%, n = 77) of the identified mutations were immune escape mutants and they could reduce HBsAg detection with the serological assays. Out 51 (66.2%) were reported as immune escape mutants and they could reduce HBsAg detection with the serological assays alone. Further, 24.6% (n = 19) of mutations were associated with immune escape/reduced HBsAg detection signals and high probability of hepatocellular carcinoma (HCC). Five (6.3%) of mutations were associated hepatocellular carcinoma, two were associated with immune escape/Drug (tenofovir) resistance while one (1.3%) each were associated with immunosuppression/HBV reactivation and Low serum HBV DNA respectively (Figure 1).
![]()
Figure 1. Frequency of clinical significance associated with mutations on HBsAggene (overall).
3.2. Frequency of Substitution at HBsAgGene and Its Correlation with Amino Acid Usage in Reference HBV Virus [NC_00397] HBsAgGene
The frequency of substitutions in HBsAggene at the specific loci from 79 samples and their correlation with amino acid preferential usage in Reference HBV virus [NC_00397] HBsAggene was determined (Figure 2). The ratio of amino acid usage in Reference HBV virus [NC_00397] HBsAg gene and the Frequency of substitution at HBsAggene (Z/B) was highest for Alanine (Z/B = 11.00), followed by Lysine (Z/B = 7.00), Serine (Z/B = 3.52), Asparagine (Z/B = 3.40) and Valine (Z/B = 2.60) (Figure 2). Although technically a positive correlation, the relationship between amino acid preferential usage and the frequencies of mutations was weak (R = 0.4136, P-Value = 0.000151, at p < 0.05). This implies that the mutations at this amino acid positions were highly biased in HBsAg gene (Table 2).
Mapping of the locations of Alanine, Lysine, Serine, Asparagine, and Valine on the Reference HBV virus [NC_00397] HBsAggene showed that there some large loci where these amino acids were not clustered (Figure 3).
The most predominant mutations of clinical importance in the in-silico analysis were S207N (n = 53), followed by mutation A194V (n = 40). The two mutations are associated with immune escape while the latter is also associated with Drug resistance especially that of tenofovir.
Fifteen mutations in the HBsAggene from donors’ samples were of clinical significance with 13 (86.7%) of them being associated with immune escape and reduced chances of being detected by the diagnostics kits that function by detection of HBsAg such as ELISA and immunochromatographic assays. One mutation, T114S, was associated with immunosuppression and HBV reactivation while another, I68T, was associated with Hepatocellular carcinoma (HCC) (Figure 4).
![]()
Figure 2. Frequency of substitution at HBsAg gene and its correlation with amino acid usage in Reference HBV virus [NC_00397] HBsAg gene.
![]()
Figure 3. Frequency of substitution at HBsAg gene and its correlation with amino acid usage in Reference HBV virus [NC_00397] HBsAg gene.
![]()
Figure 4. Frequencies of clinical significance associated with mutations on the HBsAg gene.
![]()
Table 2. Frequency of substitution at HBsAggene and its correlation with amino acid usage in Reference HBV virus [NC_00397] HBsAggene.
The most dominant genotype was found to be Genotype A in both in-silico analysis and blood donors’ sequences (Figure 5).
3.3. Phylogenetic Analysis of Sequenced Samples from Blood Donors in Kenya
Phylogenetic analysis of sequenced portion of HbsAg gene (Figure 6) showed that HBV PGENE 10 and PGENE2F clustered with the standard Genotype A NC_003977.1 references while HBV10, HBV7 and HBV13 clustered with the sub-genotype A1 AY128092.1, X51970.1 reference sequences from Africa. Sequences from samples HBV26, HBV46, HBV48, HBV38, HBV39, HBV27,
![]()
Figure 5. Frequencies of various genotypes HBV from analysis of HBsAg gene.
![]()
Figure 6. A phylogenetic tree constructed from HBV 17 sequences obtained from the 46 HBV blood donors samples obtained from KEMRI with 6 HBV reference sequences.
HBV28, PC, HBV66 clustered with sub-Genotype A1 AY233274.1 from Africa. This shows the close relationships of the above donor sequences between references AY233274.1 and AY128092.1, X51970.1. Nevertheless, HBV33, HBV56, HBV37, and HBV29, clustered with clustered with sub-Genotype A1 AY233274.1 from Africa and subgenotype A1 AB241114.1. This show that all these sequences clade for HBsAg region analysis. Further genetic analysis of the A genotype showed that HBV subgenotype A1 (94.20%) predominated Phylogenetic analysis.
The phylogenetic analysis shows that the evolutionary history was inferred by using the Maximum Likelihood method and Tamura-Nei model. The bootstrap consensus tree inferred from 1000 replicates is taken to represent the evolutionary history of the taxa analyzed. Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test 1000 replicates) are shown next to the branches [3]. Initial tree(s) for the heuristic or experimental search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Tamura-Nei model, and then selecting the topology with superior log likelihood value. This analysis involved 19 nucleotide sequences. Codon positions included were 1st + 2nd + 3rd + Noncoding. There were a total of 1203 positions in the final dataset. Evolutionary analyses were conducted in MEGA11.
3.4. Analysis of Mutations/Substitutions on the HBV S-Region of the Sequenced Samples at the Same Amino Acid Loci of the Most Two Clinically Important Dominant Mutants amongst HBV Sequences in Kenya
Analysis of HBsAg of amino acid loci of the most common clinically important mutations amongst HBV sequences in Kenya demonstrated twenty (22) types of mutations/substitutions (Table 3). Ten (45.5%) mutations (sA194Y, sS207, sA194S, sS207I, sA159I, sP46A, sA194T, sS207I, sP46R, and sT143P] according to current knowledge, had not been published before. Four (20%) of the mutations were clinically significant. These included sS207R, sT143S, sC76F and sK122R.
Out of ten samples that showed amino acid substitutions, six had more than two substitutions. Comparison of mutations of the HBsAg from the blood donors’ samples with those that were determined in silico from the Hepatitis B virus database from Kenya show a marked difference between them (Table 3).
Only two (13.3%) samples showed the same mutations of sK122R and sT143S for both in-silicoanalysis and actual sequenced samples (Figure 7). The two mutations were of clinical importance as they are associated with immune escape and occult infection as they are within the “a determinant” region of HBsAg.
Comparison of mutations on 14 most common mutations of clinical importance that were identified in the in-silicoanalysis of mutations from HBsAg Analysis in Kenya, those identified in the sequenced samples and the mutations in same loci of HBsAggene in the past studies as identified by [22] showed
![]()
Table 3. Analysis of mutations/substitutions on the HBV S-region of the sequenced samples at the same amino acid loci of the most two clinically important dominant mutants amongst HBV sequences in Kenya.
NB: *Common substitutions.
![]()
Figure 7. Concordance between HBsAg analyses in silico on blood donor sequenced samples.
general lack of consistency of mutations in a given loci (Table 4). Mutation sK122R was the most common mutation (n = 4) in these specific loci followed by mutation sT143S (n = 3), and sA194V, sT114S, sY161F, and sA159V (n = 2 each). The frequency (n = 2) of occurrence of mutation sK122R and sT143S for the HBsAg sequenced in this study was the same with that noted by Kiptoon [24].
![]()
Table 4. Comparison of HBsAg gene mutations from various studies in Kenya.
When mutations on HBsAg gene position 1 - 225 were mapped for sequences that were obtained in the in-silicoanalysis and compared with a comparable mapping that was done by [26] on the same gene and region it was found that the regions of mutations clustering and conserved regions were in concordance (Figure 8).
![]()
Figure 8. Frequency of mutations at specific loci of HBsAg gene [1 - 226 positions] from Lauder et al. (1993) study [A] and the in-silico analysis of HBV sequences in Kenya [B] comparing the mutation clustering between the two studies. The regions of clustering are comparable.
4. Discussion
In-Silico Identification of the Predominant and Clinically Significant Mutations of HBsAgGene in Kenya from NCBI/HBVdb Databases
From 100 HBsAggene sequences extracted from the NCBI, and HBV database from Kenya, 79 different types of mutations were identified after in-silico analysis. The most common mutations of clinical significance (with occurrence frequency > 5%) were: S207N, A194V, A194V, K122R, T114S, V168A, N131T, Y161F7, N40S, I68T, T143S, A159G, S45T5/A5, P46T5, and C76Y5 starting from the highest to the lowest [15] [17]. Majority (64.6%, n = 51) of the mutation were associated with immune escape and reduced HBsAg detection signals. This was followed by 24.1% (n = 19) of mutations that were both associated with immune escape/reduced HBsAg detection signals and high probability of hepatocellular carcinoma (HCC). Five (6.3%) of mutations were associated hepatocellular carcinoma, with one being associated with immune escape/drug (Tenofovir) resistance while one (1.3%) each were associated with immunosuppression/HBV reactivation and Low serum HBV DNA respectively [15]. Mutations on the “a” determinant region of HBsAggene are associated with escape from vaccine-induced immunity have been reported by various studies and they include T116N, P120S/E, I/T126A/N/I/S, Q129H/R, M133L, K141E, P142S, D144A/E, and G145R/A S143L. [15] [17] [19]. Of these mutations only T116N was identified from the in-silico analysis of the 100 sequences in this study. Although studies in Sub-Saharan Africa, have identified G145R as the commonest mutation within the HBsAg immunodorminant “a” determinant the in-silico and sample analysis in this study together with other studies [23] [24] [25] carried out in Kenya have not found this mutation.
The concordance of mutations in “a” determinant region of HBsAggene among various studies is minimal. In a study done in Central Africa Republic seven mutations (sL127I, sY100C, sA128V, sG130S, sM133T, sF134I, and S140T) within the “a” determinant region were identified of only one mutation (sG130S) was similarly identified in this study [27]. Avellón and Echevarria analyzed amino-acid positions 112 - 157 in the HBV “a” determinant region and established the following mutations: M133T (2.2%); Q129H, M133I, F/Y134N (1.8%); F/Y134L, G145A (1.5%), and P120T (1.1%) and G145R (0.4%) [28]. None of these mutations were captured in the current study. Following these findings, it is very important to continuously monitor the mutations of HBV in each specific geographical region as the general findings from other geographical regions may not be fully relied on to inform a virus mutation states. The demonstrated evidence of mutations on the “a” determinant region of HBsAggene calls for the need for identification and development of other biomarkers to produce better serology-dependent tests like Enzyme-Immunosorbents Assays (ELISA) and immunochromatographic tests. Otherwise, the future could be in development of multiplex PCR tests that will detect various loci of HBV gene [29] Similarly, the indication of development of vaccine escape HBV mutant points to the future trend of increased use of multivalent HBV vaccine that will be able to capture escape mutants. Already, a recombinant HBsAg of the Wild-Type and the G145R Escape Mutant vaccine is in the development pipeline [30] though it could end up not being useful in countries like Kenya where this study has demonstrated the low prevalence of G145R Escape Mutant.
The in-silico analysis identified the following mutations that were associated with hepatocellular carcinoma (HCC): I68T, Q30R, G44E2/X1, L49R1/X2/P4/H1, Q54H, S64C, P67Q2/L1, Q101R, L109M/P, and F134S1/Y4/L1. The analysis also identified mutations A194V40/X1 with potential to lead to drug resistance for drugs like Tenofovir. The presence of these mutations points to the need of carrying out HBV sequence analysis regularly for immune escape, drug resistance, and detection of diagnostic markers. The frequency of specific amino acid mutations at HBsAggene from the 79 sequences demonstrated significant bias (R = 0.4136, P-Value = 0.000151, at p < 0.05). The preference of substitution at HBsAggene was highest for Alanine, followed by Lysine, Serine, Asparagine and Valine. When mutations on HBsAggene position 1 - 225 in the in-silico analysis were compared with a similar mapping that was done by Lauder et al. [26] on the same gene and region it was found that the regions of mutations clustering and conserved regions were in concordance. It is noteworthy that the mutation clusters at these positions of HBsAggene have not been over a period of 30 years.
Analysis of mutations on the HBsAggene of 17 sequenced samples in this study at the same amino acid loci as those determined by the in-silico study demonstrated twenty (22) types of mutations/substitutions. Ten (45.5%) mutations (sA194Y, sS207, sA194S, sS207I, sA159I, sP46A, sA194T, sS207I, sP46R, and sT143P) had not been published before, according to current knowledge. Four (20%) of the mutations were clinically significant. These included sS207R, sT143S, sC76F and sK122R.
5. Conclusion
This study, like other related studies in Kenya, did not identify G145R which is the commonest mutation within the HBsAg immunodorminant “a” determinant that has been identified in other regions of the world with closely associated with immune escape. The concordance of mutations in “a” determinant region of HBsAggene among various studies is minimal. The study identified new mutations (sA194Y, sS207, sA194S, sS207I, sP46A, sA194T, sS207I, sP46R, and sT143P) that had not been published before, according to current knowledge. Four (20%) of the mutations were clinically significant. These included sS207R, sT143S, sC76F and sK122R.
Acknowledgements
This research is supported by the African Union Commission under the Pan African University of Basic Sciences, Technology and Innovation (PAUSTI). The Kenya Medical Research Institute provided the laboratory facility that facilitated this study. I would like to acknowledgements the KEMRI Innovation & Technology Transfer Department staff including Samuel Omari.