Hepatitis B Virus in Haiti and Africa | CDC EID

Volume 15, Number 8–August 2009
Research
Slave Trade and Hepatitis B Virus Genotypes and Subgenotypes in Haiti and Africa
Iris E. Andernach, Claudine Nolte, Jean W. Pape, and Claude P. Muller
Author affiliations: Institute of Immunology, Luxembourg, Luxembourg (I.E. Andernach, C.P. Muller); Groupe d'Etude du Sarcome de Kaposi et des Infections Opportunistes, Port-au-Prince, Haiti (C. Nolte, J.W. Pape); and Cornell University, Ithaca, New York, USA (J.W. Pape)

Suggested citation for this article

Abstract
In Haiti, >90% of the population descended from African slaves. Of 7,147 Haitian pregnant women sampled, 44% of hepatitis B virus (HBV) infections were caused by genotype A1, which today is found mainly in eastern Africa. Twenty percent belong to a rare subgenotype, A5, which has been found only in the former Bight of Benin, a former primary slave trading post. Haitian A subgenotypes appear to have separated early from the African subgenotypes; the most prevalent genotype and subgenotype in West Africa today (E and A3, respectively) are rare in Haiti. This difference indicates that the dominant subgenotypes in Africa emerged in the general population only after the slave trade and explains the low genetic diversity of genotype E. The high prevalence of HBV genotype E in much of Africa further suggests that HBV hyperendemicity is a recent phenomenon, probably resulting from extensive use of unsafe needles.

Because of a viral polymerase that lacks proofreading activity (1), hepatitis B virus (HBV) has evolved into at least 8 recognized genotypes, A–H (2–4), and a potential new genotype (tentatively designated genotype I) found mainly in Laos (5,6) but also in Vietnam (7). Except for genotypes E, G, and H, genotypes can be further divided into a variety of subgenotypes, sometimes with more or less geographic distribution. Genotype D strains are found almost worldwide (8), but subgenotype D1 occurs mostly in the Mediterranean and Middle East but also in Europe. D2 has been reported in India, Japan, Europe, and the United States; D3, mainly in South Africa and Brazil but also in Rwanda, Costa Rica, the United States, and Europe (8–11); and D4, in Australia, South Africa, Somalia, Rwanda, and Oceania (8–10).

In sub-Saharan Africa, genotypes E and A predominate. East of the E/A divide (9), subgenotype A1 is dominant in countries along the eastern coast from South Africa to the Horn of Africa (12). Although genotype A has been found on every continent, its genetic diversity is higher in Africa (4% over the complete genome) than in the rest of the world (3%). Five subtypes of HBV/A (A1–A5) have been proposed in Africa (13), whereas essentially only A2 and, to a lesser extent, A1 have been reported from other continents (14). Therefore, some researchers have suggested that genotype A has emerged in Africa (15) and, after a long evolution, has been introduced to other continents. However, despite the high genetic diversity of HBV/A in West Africa, this genotype is rare there. In contrast, genotype E has been found only in Africa, with some rare exceptions on other continents in persons with a link to Africans. Genotype E is found almost exclusively throughout the vast expanses of a crescent from Senegal in the west (16) to the Central African Republic in the east (17) and Namibia in the south (13). In comparison to HBV/A, the conspicuously low genetic diversity of HBV/E suggests its short natural history in Africa (18) and relatively recent introduction into the general population there (18). However, the recent presence of HBV/E in Africa contrasts sharply with its current high prevalence and extensive geographic distribution there. The wide spread of genotype E also seems difficult to reconcile with a long natural history of genotype A in Africa (18).

In Haiti, where >90% of the population descends directly from African slaves (19), we investigated the phylogeny of HBV to learn which genotypes may have been prevalent in Africa several centuries ago. The conspicuous absence of genotype E in Haiti suggests recent and rapid spread of genotype E in Africa during the past 200 years, probably as the result of public health interventions.

Materials and Methods
Serum samples were collected in 2006 after informed consent as part of a national survey to evaluate prevalences of human immunodeficiency virus infection, hepatitis B, and serologic syphilis among pregnant women at their first prenatal medical visit in 19 clinics throughout Haiti. Women were tested for hepatitis B surface antigen (HBsAg) by using the Murex HBsAg Kit (Abbott Laboratories, Ottiginies, Belgium). DNA was extracted from HBsAg–positive samples by using the QIAGEN DNA Blood Mini kit (QIAGEN, Venlo, the Netherlands) according to the manufacturer's protocol. The complete HBV genome was amplified in 4 overlapping fragments (preS, S, X, and C) as described previously (20). Phylogenetic analysis and distance calculations were performed by using MEGA v.4 (21) with the neighbor-joining method of the Kimura 2-parameter model with 1,000 bootstrap replicates. Genotyping was performed by analyzing the complete genome or at least 1 of the 3 fragments of preS, S, or C genes. Subgenotyping was done on the full-length genome or on at least 2 complete fragments preS, S, or C. Sequences were submitted to EMBL/GenBank/DDBJ under accession nos. FJ692502–FJ692553 (Haiti S-fragment sequences), FJ692557–FJ692613 (Haiti complete genome sequences), and FJ692554–FJ692556 (Nigeria complete A5 sequences).

Results
Genotypes and Subgenotypes
In 7,147 blood samples of pregnant Haitian women, HBsAg prevalence was 5%, ranging from 1.0% to 8.5%, depending on the sampling clinic. Of 320 HBsAg–positive samples available, 247 (77.2%) were positive for at least 1 of the 4 overlapping PCR fragments (Table 1). Interpretable sequences from at least 1 of the 4 PCR fragments were obtained from 213 viruses. A total of 179 of these strains could be clearly assigned to a genotype by analyzing the complete genome or at least 1 of the 3 fragments of preS, S, or C genes. Of the 213 strains, 31 showed signs of mixed infection or recombination, and 3 strains were considered outliers because they could not be genotyped.

Phylogenetic analysis of the above 179 genotypeable strains (excluding mixed, recombinant, and untypeable strains) showed that 128 (71.5%) viruses belonged to genotype A; 40 (22.4%), to genotype D; and 11 (6.1%), to genotype E (Table 2). Genotype A strains were attributable to subgenotypes A1 (n = 77 [43.0%]) and A5 (n = 35 [19.6%]). Genotype D strains belonged to D4 (16.2%) and D3 (3.9%). Fifteen viruses of genotype A and 4 of genotype D could not be further subgenotyped (Table 2) because only partial gene sequences or single preS, S, or C fragments were obtained. In all of the above strains, genotypes of the different fragments agreed with each other. In addition, 31 viruses were suspected mixed genotype infections or recombinants; they were not included in the above analysis and are discussed later.

Phylogenetic Analysis
Subgenotype A1

Figure 1

Figure 1. Phylogenetic analysis of selected sequences clustering with subgenotype A5, based on the complete genome...


Figure 2

Figure 2. Phylogenetic analysis of selected sequences clustering with subgenotype D4 (A) or D3 (B), based on the S fragment, including potential mixed or recombinant strains (*)...


Figure 3

Figure 3. Distribution of hepatitis B virus A subgenotypes and D4 (only in Rwanda) in Africa and their potential routes of spread toward Haiti (color-coded arrows)...


Appendix Figure 1

Appendix Figure 1. Phylogenetic analysis of selected sequences clustering with subgenotype A1, based on the complete genome. Diamonds indicate Haiti sequences...


Appendix Figure 2

Appendix Figure 2. Phylogenetic analysis of selected sequences clustering with genotype E, based on the S fragment, including potential mixed or recombinant strains (*)...

Phylogenetic analysis of A1 complete genome sequences showed that Haiti strains form several clusters (not necessarily supported by bootstrap values) within available full-length A1 strains (Appendix Figure 1) from South Africa and other eastern African countries, as well as from the Philippines. Haiti's complete genome A1 strains showed a mean genetic diversity of 1.45% (maximum diversity of 3.86%) that rose to a mean genetic diversity of 2.49% (maximum 6.61% between FJ692589 and U87742) when all available A1 strains (mean 2.87%, maximum 7.64% between AY161140 and U87742) were included.

Subgenotype A5

A5 originally was proposed on the basis of the preS and preC/C gene fragments of 3 Nigerian strains (20). We present here the full-length sequences of the latter strains (accession nos. FJ692554–FJ692556) and compare them with all full-length A sequences from Haiti. Twenty-one sequences clustered with the only available A5 sequences from Nigeria. The overall mean intrasubgenotype diversity of A5 is 1.42% (maximum genetic diversity 2.89%). The mean intersubgenotype distance of the proposed A5 subgenotype was above the approximately 4% proposed for a new subgenotype (3,22) for subgenotypes A1 (4.1%), A2 (4.8%), and A3 (5.1%); it was 3.8% when compared with the previously proposed subgenotype A4 (20). Nevertheless, A5 strains from Haiti and Nigeria form 2 distinct phylogenetic subgroups within A5, supported by high bootstrap values (99%; Figure 1). These subgroups are separated by a mean, minimal, and maximum genetic distance of 2.28%, 1.71%, and 2.89%, respectively.

abrir aquí para acceder al documento CDC completo del que sólo se reproduce el 32% ya que además de extenso y minucioso, contiene numerosas figuras:
Hepatitis B Virus in Haiti and Africa | CDC EID