Study population

802 anti-HCV positive plasma samples were collected from 5 Chinese blood centers including: Chongqing Blood Center, Chongqing; Urumqi Blood Center, Urumqi, Xinjiang; Luoyang Blood Center, Luoyang, Henan; Mianyang Red Cross Blood Center, Mianyang, Sichuan; Guangxi Blood Center, Liuzhou, Guangxi. Samples were collected with EDTA anticoagulation tube, stored at -20°C, and then transported to the Institute of Blood Transfusion, Chinese Academy of Medical Sciences through the cold chain.

The study population was divided into different age groups: 18–20, 21–30, 31–40, 41–50, >50 years of age; Ethnicity group: Han, Hui, Uygur, Zhuang; Occupation groups: student, commercial, worker, peasant, working at home, other; Educational levels: ≤primary school, middle or high school, >high school, other, missing.

HCV RNA extraction and reverse-transcription polymerase chain reaction (RT-PCR)

The required 8 primers including outer and inner amplification were designed after reference [14]. At present, three most commonly used HCV coding regions to genotyping are Core, E1 and NS5B [15–17], which represents high accuracy and sensitivity, however, sometimes different HCV subtypes can be obtained with the above three regions for the same HCV genome, suggesting the presence of mixed infection or gene recombination [18,19]. In our study, coding regions of Core and E1 were separately used to reconstruct independent HCV GTs/subtypes phylogenetic trees in order to make more accurate genotyping, and the final amplified HCV products based on Core and E1 gene fragments were 401bp and 592bp, respectively.

HCV RNA was extracted from 140μL of each plasma sample by the principle of Membrane adsorption using the QIAmp Viral RNA Mini kit (QIAGEN, German) following the instructions strictly, and then it was reverse transcribed into DNA fragment with the help of antisense outer primers of Core and E1 using the Transcriptor First Strand cDNA Synthesis Kit (Roche, Switzerland). The nested PCR based on Core and E1 gene fragment was performed using 5μL of the RT-PCR product with 2×TSINGKE MASTER MIX (TSINGKE Biotech Co Ltd, Chengdu, China), and primers as specified previously [20]. The final PCR products were confirmed by 2% agarose gel electrophoresis and visualization with ethidium bromide staining, and amplified products were sent to TSINGKE Biotech Co Ltd (Chengdu, China) for purifying and sequencing with the sense and antisense inner primers.

HCV genotyping, phylogenetic and evolutionary analysis

All the sequences of the Core and E1 regions were checked and spliced using Mega 7.0, and the spliced HCV sequences were aligned with HCV reference sequences (S1 Table). Phylogenetic analysis was performed using the neighbor-joining and the Kimura two-parameter method to get the accurate HCV genotyping. Node supports were evaluated with 1000 bootstrap replicates. HCV GT/subtype was identified when it was included in a monophyletic cluster with classified reference strains.

Arlequin v3.0 was used to perform population genetics data analysis including calculating the haplotype frequencies, estimating the past population expansion of species by Tajima's D and Fu' s Fs value, and if the values are negative and P <0.05 or 0.01, representing the population is considered to have undergone expansion. BEAST v1.10.4 was used to get the Bayesian skyline plot (BSP) to analyze growth history of different HCV subtypes. Uniform prior was selected and the evolutionary rates of Core and E1 fragment for each dominant subtypes were shown in S2 Table.

Statistical analysis

Statistical data analyses were performed with SPSS Statistics 17.0. The Pearson Chi-Square test was used to compare the ratio of risk factors between HCV GT/subtypes and demographic characteristics (age, gender, ethnic, occupation, educational level), and P value < 0.05 was considered to be statistically significant.

HCV genotyping and phylogenetic analysis

Among the 802 anti-HCV positive plasma samples, 419 and 293 were successfully amplified based on Core gene fragment and E1 gene fragment correspondingly. The determined genotype among samples that were successfully amplified by both Core and E1 fragment were consistent, and 36 subtypes, that were successfully amplified by E1 fragment, were not amplified by Core fragment. Based on the above, 455 of 802 samples based on Core and (or) E1 fragment were successfully amplified. The successfully amplified samples were then sequenced and used to construct phylogenetic trees (Fig 1). In detail, the predominant HCV subtype was 1b (58.68%, 267/455), followed by subtype 2a (17.80%, 81/455). Other HCV subtypes including la, 3a, 3b, 6a, 6e and 6n were also found among these blood donors, and the proportion of the above subtypes were 0.66% (3/455), 5.05%(23/455), 3.52% (16/455), 12.31% (56/455), 0.88% (4/455) and 0.66% (3/455), respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Phylogenetic tree constructed using HCV Core and E1 gene sequences from blood donors in China.

Phylogenetic analysis was performed using the neighbor-joining and the Kimura two-parameter method to get the accurate HCV genotyping, and the node supports were evaluated with 1000 bootstrap replicates. The colored triangles and vertical lines represent different HCV subtypes. The upper bar refers to the Core region and the lower bar refers to the E1 region. The colored rectangular area represents the number of HCV subtypes in five regions including Chongqing (CQ), Guangxi (GX), Luoyang (LY), Mianyang (MY) and Urumqi (WS).

https://doi.org/10.1371/journal.pone.0235612.g001

The distribution of HCV GT among blood donors in China showed significant geographical difference (P <0.05) (S3 Table). The predominant HCV subtypes in different regions were as followed: 1b (59.72%, 43/72) and 2a (34.72%, 25/72) in Luoyang, 1b (46.21%, 61/132) and 6a (18.94%, 25/132) in Chongqing, 1b (80.77%, 63/78) and 2a (10.26%, 8/78) in Mianyang, 1b (67.57%, 75/111) and 2a (22.52%, 25/111) in Urumqi, and 1b (41.94, 26/62) and 6a (35.48%, 22/62) in Guangxi.

HCV GT and demographic characteristics

Table 1 showed the association between some HCV subtypes (1b, 2a, 3a, 3b and 6a) and gender, age, ethnicity, education status and occupation. There were significant differences between HCV GT and age and ethnicity (P < 0.05), while no statistical differences were observed between HCV GT and gender, occupation and education level (P value > 0.05). In the age groups of 21–30, 41–50 and >50, HCV subtypes 1b and 2a accounted for a higher proportion, while in the age groups of ≤20 and 31–40, lb and 6a were more common. In Han population, HCV lb and 2a were the prevalent subtypes, accounting for 72.6% (312/430) of the whole ethnic groups, while HCV 6a was more common in Zhuang population.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. The association between HCV genotypes/subtypes and demographic characteristics.

https://doi.org/10.1371/journal.pone.0235612.t001

HCV population dynamics

The dominant subtypes were 1b, 2a, and 6a. The subtypes determined by Core gene were 251, 72 and 51, respectively and the corresponding haplotypes of Core 1b, 2a and 6a were 178, 55 and 44; The sequences of 1b, 2a, and 6a determined by E1 gene were 186, 42 and 31, respectively and the corresponding haplotypes of E1 1b, 2a and 6a were 183, 42 and 31 (Table 2). The neutral test values of Tajima's D and Fu's Fs of each HCV subtypes were negative, the P value corresponding to Fu's Fs of each subtype was < 0.01, and the P value corresponding to Tajima's D of subtype 1b was < 0.05; The Tajima's D and Fu's Fs values of HCV 6a were both negative and P value is less than 0.01 (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. HCV population genetic characteristics in China.

https://doi.org/10.1371/journal.pone.0235612.t002

The spatio-temporal changes estimated in the HCV population size were illustrated as results of BSP shown in Fig 2A and 2B. For HCV-1b the Core and E1 genes showed mutually rapid population growth between 1992 and 1998. While the Core gene for HCV-1b showed another period of continuous population growth until 2005. As for HCV-2a, both Core and E1, the genes mutually showed population growth after 2000. HCV-6a was estimated as a new member of the blood donor population that appeared around 1995 and the sudden population expansion occurred between 2002 and 2010, based on the calculation of BSP of the Core and E1 genes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Population genetic characteristics of HCV with different genotypes among blood donors in China.

The solid black part is the median of the estimated changes in genetic diversity. Other colored parts is the 95% confidence interval. The horizontal axis is the year. The vertical axis is Neτ (the effective population size of the experimental sample). a: The BSP for HCV 1b(blue), 2a(light blue) and 6a(green) based Core gene. b: The BSP for HCV 1b(orange), 2a(light orange) and 6a(yellow) based E1 gene.

https://doi.org/10.1371/journal.pone.0235612.g002

Regional differences

In this study, HCV GT/subtype distribution in groups of blood donors showed significant geographical variations that are different from findings in previous publications.

In Guangxi, two studies performed in 1990s showed that the prevalent GTs were 1b and 1a, 3b and 1b [21,22], respectively. However, our study together with reports in recent years, suggest that HCV 1b, 6a has become the dominant subtype over time in Guangxi, which is consistent with a meta-analysis reported [12]. In addition, we also found that the proportion of HCV-6a increased compared with other previous reports in Guangxi while the proportion of HCV-2a were decreasing [21–26]. In early years of 1990s, HCV-2b has been detected in Guangxi [21,22], and since then, this subtype hasn’t never been found in later studies including this one. HCV-6n was detected which was never been reported before in Guangxi. Probably due to the special geographical environment of Guangxi and its proximity to Hong Kong, Macao, Vietnam, Thailand and other countries and regions, such as the HCV transmission of rare genotypes in these countries and regions; coupled with the wide communication leading to HCV genotype in Guangxi extremely rich and complicated.

In Mianyang (Sichuan province), Luoyang (Henan province) and Urumqi (Xinjiang province), the prevalent genotypes were both HCV 1b and 2a. According to the previous reports, the major two HCV subtypes in Sichuan were 1b and 3b or 3a [27–30]. In our study, however, HCV-1b and 2a became the predominant subtypes in Mianyang with the prevalence rate of 1b up to 80%, and the prevalence of 2a was significantly higher than 3a and 3b. In Urumqi, the prevalence of two major subtypes was consistent with three previous reports [31–33]. With limited previous studies about HCV genotypes, the changing trend of each HCV GT in the above two regions can’t be easily found over time. In Luoyang, the prevalent HCV subtypes were consistent with most of previous studies reported, while HCV-2a exhibits a trend of slight increase.

In Chongqing, the distribution of predominant HCV subtypes showed high gene abundance that HCV-1b and 2a [34–36], 1b and 3b [37,38], 2a and 3a [39] have ever been reported as the major HCV subtypes. Comparing with a 2003–2004 study performed in blood donors of Chongqing [39], the proportion of HCV 6a has increased significantly, which was consistent with pooled results of a meta-analysis [12].

The distribution of HCV subtypes had undergone some changes over time in some places of China, just as what we have found in Guangxi, Chongqing and Mianyang. The emergence of new HCV subtypes and changing trends of HCV GTs may be ascribed to: 1) Increasing population mobility: With the increase of domestic or international trade and communication, the interaction between people will naturally increase, which may lead to the spread of HCV; 2) The changing transmission routes: In recent years, there was an increase in HCV IDUs, possibly due to common syringes contaminated by blood, which may represent an increasing of HCV subtypes 3 and 6 [8]. In our study, the increasing proportion of HCV-6a in Guangxi and Chongqing may be a consequence of IDU related spreading in these two areas [25,38]; 3) The improved sensitivity and specificity of testing methods and the sample size. Most notably, a multi-center large sample research still may need to be performed to reveal more meaningful results showing HCV changing trends.

Demographic association

Reports about the association between HCV GTs and host demographics in China were inconsistent [8,40]. Subtype 1b was higher in male than in female in some reports [41]. The possible reason may be that gender difference exhibits different immune response against HCV GTs and the life style of male population is complex and thus the chances of infecting HCV are relatively high. Previous studies found that in the older group (≥ 40), HCV-1b and 2a were the predominant subtypes, while in the younger group, the proportion of HCV-6 and 3 showed an increasing trend with drug use or high-risk sexual behavior by young people [42,43].

In our study, significant statistical result was found between HCV GTs and age. HCV subtypes 1b and 2a are more common in the age groups of 21–30 and >41, while lb and 6a are more common in the younger groups of 18–20 and 31–40, which was consistent with early reports, while higher proportion of 1b and 2a was found for the group of 21-30.We speculate that it may take years from most HCV infected individuals to develop clinical symptoms [44], thus HCV lb and 2a still remained the main subtypes, and in addition, sampling deviation should also be taken into account.

In Han population, lb and 2a accounted for 72.6% of the whole ethnic groups, while HCV 6a is more common in Zhuang population, and the possible reasons are: 1) The lifestyle of different nationalities are diverse, and HCV transmission modes are also diverse [45]; 2) different ethnicities show different immune responses to HCV infection with different subtypes [46]. In addition, there was no significant association between HCV GTs and occupation, educational level in our study, which was similar to the previous reports [11,47]. The sample size in this study is large and representative; however, the exact mechanism of the relationship between HCV GT distribution and demographic characteristics needs more investigation.

Population dynamics

BSP analyses were performed using the sequences determined in this study and it showed that HCV 1b, 2a and 6a have experienced different rates of population expansion at different times in the past.