Multiple recombinant events in human T-cell Leukemia virus Type 1: complete sequences of recombinant African strains

Abstract

Africa is the largest endemic area for HTLV-1, with many molecular genotypes. We previously demonstrated that some strains from North Africa (a-NA clade) originated from a recombinant event between Senegalese and West African strains. A series of 52 new HTLV-1 strains from 13 North and West African countries were sequenced in the LTR region and/or a env gene fragment. Four samples from French Guyanese of African origin were also added. Furthermore, 7 complete sequences from different genotypes were characterized. Phylogenetic analyses showed that most of the new African strains belong to the Cosmopolitan a-genotype. Ten new strains from the a-NA clade were found in Morocco, Western Sahara, Mali, Guinea, Côte d'Ivoire and Ghana. A new a-G-Rec clade, which arose from a distinct recombination event between Senegalese and West African strains, was identified in Guinea and Ghana. The complete sequences suggest that recombination occur in the LTR as well as the env/pol region of the genome, thus a-NA and a-G-Rec strains have a mosaic profile with genetic segments from either a-WA or a-Sen strains. Our work demonstrates that recombination in HTLV-1 may not be as rare an event as previously proposed.

Keywords: Africa; HTLV-1; HTLV-1 complete genome; molecular epidemiology; recombination; reverse transcription.

Figure 1.

Geographical distribution of HTLV-1 strains in North and West Africa. The 52 HTLV-1 strains characterized were from Morocco (1), West Sahara (1), Mauritania (2), Mali (6), Senegal (3), Guinea (6), Sierra Leone (2), Côte d’Ivoire (14), Burkina Faso (2), Ghana (4), Togo (2), Benin (2), and Nigeria (7). The four strains from French Guiana are not represented on this map.

Figure 2.

Phylogenetic analysis of African LTR sequences. A- Phylogenetic comparison was performed on 772-nucleotide-long LTR alignment of African isolates, including the 52 sequences generated in this study (in red). The Melanesian sequence Mel5 was used as outgroup. The phylogenetic tree was derived by the neighbor- joining method using the GTR model (gamma = 0.5017). Horizontal branch lengths are drawn to scale, with the bar indicating 0.01 nucleotide replacement per site. Numbers on each node indicate the percentage of bootstrap samples (of 1,000 replicates) in which the cluster to the right is supported. Next to each sequence, three letters symbolize the country of origin of the infected individual (mostly IOC country codes): ALG - Algeria, ANG - Angola, BEN - Benin, BUR - Burkina Faso, CAM - Cameroon, CAR - Central African Republic, CHA - Chad, CIV - Côte d’Ivoire, COM - Comores, CPV - Cape Verde, DRC - Democratic Republic of Congo, FRG - French Guiana, GAB - Gabon, GAM - Gambia, GBS - Guinea-Bissau, GHA - Ghana, GUI - Guinea, MAR - Morocco, MLI - Mali, MTN - Mauritania, NGR - Nigeria, RSA – South Africa, SEN - Senegal, SLE - Sierra Leone, SWZ - Swaziland, TOG - Togo, UGA - Uganda, ZAM - Zambia, ZIM - Zimbabwe. B- Phylogenetic comparison was performed on 772-nucleotide-long LTR alignment of African isolates, including the 52 sequences generated in this study (in red). The Melanesian sequence Mel5 was used as outgroup. The consensus phylogenetic tree was constructed using a Bayesian approach based upon the GTR substitution model. The MCMC analysis was performed with 4 chains that ran for 2,000,000 cycles. Horizontal branch lengths are drawn to scale, with the bar indicating 0.01 nucleotide replacement per site. Numbers on each node indicate the posterior probabilities of the branches (in percentage).

Figure 3.

Boot-scanning for a-NA and a-G-Rec and Phylogenetic analysis of U3 and RU5 LTR segments. The a-NA (panel A) and the a-G-Rec (panel B) subgroups were compared by boot-scanning (Simplot program) to different clades (a-TC, a-WA, a-Sen, b, and c). The analysis used a 200- bp-long window and a 20 bp-long step, and the Kimura 2p model. The x values reflect the genome position at the midpoint of the analyzed windows, and the y values reflect the bootstrap value calculated from the windows (for 1,000 replicates). Phylogenetic trees corresponding to the first 371 nucleotides (Panel C) and the 401 last nucleotides (Panel D), respectively, were derived from the Maximum Likelihood method. Values correspond to the approximate likelihood-ratio test for each group. The groups of interest are coloured as follows: red, green, dark green, blue and grey sequences belong to a-Sen, a-NA, a-G-Rec, a-WA, and a-TC, respectively.

Figure 4.

Evidencing the mosaic profile of a-NA and a-G-Rec and Phylogenetic analyses of env and gag genes. A- Phylogenetic comparison was performed on 522-nucleotide-long env gene fragments of African isolates. The Melanesian sequence Mel5 was used as outgroup. The phylogenetic tree was derived by the Neighbor-Joining method using the Tamura Nei model (gamma = 0.3043). Horizontal branch lengths are drawn to scale, with the bar indicating 0.01 nucleotide replacement per site. Numbers on each node indicate the percentage of bootstrap samples (of 1,000) in which the cluster to the right is supported. Phylogenetic topologies were similar using different methods, i.e. Maximum likelihood and a Bayesian approach (data not shown). B-C- The a-NA (Panel A) and the a-G-Rec (Panel B) subgroups were compared by boot- scanning (Simplot program) to different clades (a-TC, a-Jap, a-WA, a-Sen, b, and c). The analysis used a 800-bp-long window and a 80-bp-long step, and the Kimura 2p model. The x values reflect the genome position at the midpoint of the analyzed windows, and the y values reflect the bootstrap value calculated from the windows (for 1,000 replicates). D- Phylogenetic comparison was performed on 2,094-nucleotide-long gag fragments (obtained from complete genomes). Six Australo-Melanesian HTLV-1c complete sequences were used as outgroup. The phylogenetic tree was derived by the neighbor-joining method using the Tamura Nei model (gamma = 0.8793; i = 0,5391). Horizontal branch lengths are drawn to scale, with the bar indicating 0.01 nucleotide replacement per site. Numbers on each node indicate the percentage of bootstrap samples (of 1,000) in which the cluster to the right is supported. Phylogenetic topologies were similar using different methods, i.e. Maximum likelihood and a Bayesian approach (data not shown). The groups of interest are coloured as follows: red, green, dark green, blue, and grey sequences belong to a-Sen, a-NA, a-G-Rec, a-WA, and a-TC respectively. New sequences are in bold.