12
Molecular epidemiology of the HHV-8 K1 gene from Moroccan patients with Kaposi's sarcoma Renan Duprez a, ,1 , Oumkaltoum Hbid a,b,1 , Philippe Afonso a , Hélène Quach c , Lamia Belloul d , Noura Fajali e , Nadia Ismaili d , Hakima Benomar b , El Hassane Tahri f,g , Michel Huerre h , Luis Quintana-Murci c , Antoine Gessain a, a Unité dEpidémiologie et Physiopathologie des Virus Oncogènes, Département de Virologie, Institut Pasteur, 25-28 rue du Dr. Roux, 75724, Paris, Cedex 15, France b Service dTAnatomie-Pathologie, Institut Pasteur, Maroc, France c CNRS FRE2849, Unité de Prévention et Thérapie Moléculaires des Maladies Humaines, Institut Pasteur, France d Service de Dermatologie, Avicenne Rabat, Maroc, France e Service de Dermatologie, CHU Casablanca, Maroc, France f Service de Maladies Infectieuses, CHU Casablanca, Maroc, France g Laboratoire de Physiologie, Faculté de science Ben Msik, Casablanca, Maroc, France h Unité de Recherche et dTExpertise en Histotechnologie et Pathologie, Institut Pasteur, France Received 2 March 2006; accepted 14 April 2006 Available online 21 June 2006 Abstract The genetic variability of the human herpesvirus 8 (HHV-8) strains circulating in the populations living in the Maghreb region, an endemic area for HHV-8 and associated Kaposi's sarcoma, remains largely unknown. We have thus analyzed the genetic variation of the complete K1 gene of HHV-8 in a series of 35 viral strains, originating from 28 Moroccan patients with classic, AIDS-associated or iatrogenic Kaposi's sarcoma lesions. All but one of the 35 strains belonged to the large C molecular subtype. Furthermore, high genetic diversity within the C subtype was observed in the 35 sequenced HHV-8 K1 genes, with strains belonging to several and distinct subgroups highly supported from a phylogenetically viewpoint (e.g., C3, C7, Cand C5). Considering these newly identified Moroccan viral strains in the context of 189 complete K1 genes, we were able to characterized, using the Simplot program, two main groups of recombinant chimeric K1 genes, either intertypic (C5) or intratypic (C7). In addition, the genetic characterization of the host maternal gene pool, through the analyses of mtDNA variation, did not provide evidence for any association between a particular human ethno-geographic background (i.e., North African vs. sub-Saharan African vs. West Eurasian linages) and any HHV-8 strain because both Cand Cstrains were randomly distributed among the different patients' population backgrounds. © 2006 Elsevier Inc. All rights reserved. Keywords: Oncogenic herpesviruses; Human herpesvirus 8; K1 gene; Genetic variability; Kaposi's sarcoma; Molecular epidemiology; Recombinant; Morocco Introduction Human herpesvirus 8 (HHV-8), also known as Kaposi's sarcoma-associated herpesvirus (KSHV), is a γ2-herpesvirus that was first identified in a biopsy tumor from an AIDS-related Kaposi's sarcoma (KS) (Chang et al., 1994). This virus is now considered as the etiological agent of all clinico-epidemiolog- ical forms of KS (Boshoff et al., 1995; Chuck et al., 1996; Dupin et al., 1995; Huang et al., 1995; Moore and Chang, 1995, 2001; Whitby et al., 1995), a tumor of mixed cellularity, and frequently occurring during HIV-1 infection (epidemic KS, AIDS-KS) or in transplant recipients (iatrogenic KS, I-KS). More rarely, such tumor occurs among non-HIV-infected individuals, either predominantly in aged men of Mediterranean and Middle East origin (classic KS, C-KS) or in inhabitants from East and Central Africa (endemic KS, E-KS) (Beral, 1991; Virology 353 (2006) 121 132 www.elsevier.com/locate/yviro Corresponding authors. Fax: +33 1 40 61 34 65. E-mail addresses: [email protected] (R. Duprez), [email protected] (O. Hbid), [email protected] (P. Afonso), [email protected] (H. Quach), [email protected] (M. Huerre), [email protected] (L. Quintana-Murci), [email protected] (A. Gessain). 1 Both authors similarly contributed to this work. 0042-6822/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2006.04.026

Molecular epidemiology of the HHV-8 K1 gene from … · Molecular epidemiology of the HHV-8 K1 gene from Moroccan patients with Kaposi's sarcoma ... Unité de Prévention et Thérapie

  • Upload
    lamnga

  • View
    213

  • Download
    0

Embed Size (px)

Citation preview

6) 121–132www.elsevier.com/locate/yviro

Virology 353 (200

Molecular epidemiology of the HHV-8 K1 gene from Moroccanpatients with Kaposi's sarcoma

Renan Duprez a,⁎,1, Oumkaltoum Hbid a,b,1, Philippe Afonso a, Hélène Quach c, Lamia Belloul d,Noura Fajali e, Nadia Ismaili d, Hakima Benomar b, El Hassane Tahri f,g, Michel Huerre h,

Luis Quintana-Murci c, Antoine Gessain a,⁎

a Unité d’Epidémiologie et Physiopathologie des Virus Oncogènes, Département de Virologie, Institut Pasteur,25-28 rue du Dr. Roux, 75724, Paris, Cedex 15, France

b Service dTAnatomie-Pathologie, Institut Pasteur, Maroc, Francec CNRS FRE2849, Unité de Prévention et Thérapie Moléculaires des Maladies Humaines, Institut Pasteur, France

d Service de Dermatologie, Avicenne Rabat, Maroc, Francee Service de Dermatologie, CHU Casablanca, Maroc, France

f Service de Maladies Infectieuses, CHU Casablanca, Maroc, Franceg Laboratoire de Physiologie, Faculté de science Ben M’sik, Casablanca, Maroc, France

h Unité de Recherche et dTExpertise en Histotechnologie et Pathologie, Institut Pasteur, France

Received 2 March 2006; accepted 14 April 2006Available online 21 June 2006

Abstract

The genetic variability of the human herpesvirus 8 (HHV-8) strains circulating in the populations living in the Maghreb region, an endemicarea for HHV-8 and associated Kaposi's sarcoma, remains largely unknown. We have thus analyzed the genetic variation of the complete K1 geneof HHV-8 in a series of 35 viral strains, originating from 28 Moroccan patients with classic, AIDS-associated or iatrogenic Kaposi's sarcomalesions. All but one of the 35 strains belonged to the large C molecular subtype. Furthermore, high genetic diversity within the C subtype wasobserved in the 35 sequenced HHV-8 K1 genes, with strains belonging to several and distinct subgroups highly supported from a phylogeneticallyviewpoint (e.g., C3, C7, C″ and C5). Considering these newly identified Moroccan viral strains in the context of 189 complete K1 genes, we wereable to characterized, using the Simplot program, two main groups of recombinant chimeric K1 genes, either intertypic (C5) or intratypic (C7). Inaddition, the genetic characterization of the host maternal gene pool, through the analyses of mtDNA variation, did not provide evidence for anyassociation between a particular human ethno-geographic background (i.e., North African vs. sub-Saharan African vs. West Eurasian linages) andany HHV-8 strain because both C′ and C″ strains were randomly distributed among the different patients' population backgrounds.© 2006 Elsevier Inc. All rights reserved.

Keywords: Oncogenic herpesviruses; Human herpesvirus 8; K1 gene; Genetic variability; Kaposi's sarcoma; Molecular epidemiology; Recombinant; Morocco

Introduction

Human herpesvirus 8 (HHV-8), also known as Kaposi'ssarcoma-associated herpesvirus (KSHV), is a γ2-herpesvirusthat was first identified in a biopsy tumor from an AIDS-related

⁎ Corresponding authors. Fax: +33 1 40 61 34 65.E-mail addresses: [email protected] (R. Duprez),

[email protected] (O. Hbid), [email protected] (P. Afonso),[email protected] (H. Quach), [email protected] (M. Huerre),[email protected] (L. Quintana-Murci), [email protected] (A. Gessain).1 Both authors similarly contributed to this work.

0042-6822/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.virol.2006.04.026

Kaposi's sarcoma (KS) (Chang et al., 1994). This virus is nowconsidered as the etiological agent of all clinico-epidemiolog-ical forms of KS (Boshoff et al., 1995; Chuck et al., 1996;Dupin et al., 1995; Huang et al., 1995; Moore and Chang, 1995,2001; Whitby et al., 1995), a tumor of mixed cellularity, andfrequently occurring during HIV-1 infection (epidemic KS,AIDS-KS) or in transplant recipients (iatrogenic KS, I-KS).More rarely, such tumor occurs among non-HIV-infectedindividuals, either predominantly in aged men of Mediterraneanand Middle East origin (classic KS, C-KS) or in inhabitantsfrom East and Central Africa (endemic KS, E-KS) (Beral, 1991;

122 R. Duprez et al. / Virology 353 (2006) 121–132

Hengge et al., 2002). Furthermore, HHV-8 has been also clearlyassociated with primary effusion lymphoma (Cesarman et al.,1995; Nador et al., 1996) and with some aggressive cases ofmulticentric Castleman's disease (Gessain et al., 1996; Soulier

et al., 1995), as well as some other rare associated lymphomas(Chadburn et al., 2004).

Epidemiological surveys indicated that HHV-8 was not awidespread ubiquitous virus but that its presence was mainly

123R. Duprez et al. / Virology 353 (2006) 121–132

restricted to areas of high endemicity for classic or endemicforms of KS (Boshoff and Weiss, 2001; de-The et al., 1999;Dukers and Rezza, 2003; Gessain et al., 1999; Mbulaiteye et al.,2003; Plancoulaine et al., 2004; Schulz, 1998).

Initial studies conducted on the genetic variability andpolymorphism of HHV-8 focused on 2 small gene fragmentsof the ORF 26 (minor capsid gene) and ORF 75 (tegumentregion gene) from KS specimens and demonstrated theexistence of 3 distinct molecular patterns with however avery low overall nucleotide variability (<5%) among these 3genotypes (Alagiozoglou et al., 2000; Fouchard et al., 2000;Kasolo et al., 1998; Zong et al., 1997). Molecular epidemi-ological studies exploiting the significantly higher geneticvariability of the K1 gene were more recently published(Biggar et al., 2000; Cook et al., 1999; Kadyrova et al., 2003;Kakoola et al., 2001; Kasolo et al., 1998; Kazanji et al., 2005;Lacoste et al., 2000a, 2000b; Meng et al., 1999, 2001;Nicholas et al., 1998; Whitby et al., 2004; Zong et al., 1999,2002). These studies identified five K1 subtypes (called A, B,C, D and E), which exhibit, at least for most of them, a cleargeographic clustering. Thus, HHV-8 B strains predominate inAfrica and are phylogenetically more distant from groups Aand C (which are found in Europe and the European-descentpopulations from the US) than A and C from each other,whereas the few subtype D strains are restricted to Pacificpopulations. The last molecular clade (subtype E) comprisesonly few related divergent strains, all found in remotepopulations of Amerindians from Brazil, the French Guianaand Ecuador.

Several countries of the Mediterranean basin are endemicfor HHV-8 and classic Kaposi's sarcoma. Studies on the K1genetic variability mainly concern Italy, Greece, Israel andeven Saudi Arabia (Cook et al., 1999; Davidovici et al.,2001; Poole et al., 1999; Zong et al., 2002). However, dataon the Maghreb region, considered also as a high endemicarea for KS (Hbid et al., 2005; Mseddi et al., 2005), remainscarce or even lacking. Indeed, to our knowledge, only onestudy (Davidovici et al., 2001), which reported epidemio-logical and molecular data on HHV-8 and KS from Jewishpopulations of different origin, provided some features onthe K1 genetic variability of few patients with KSoriginating from the Maghreb.

Presence of HHV-8 recombinant genomes have beenpreviously reported including some chimeric K1 gene (Cooket al., 1999; Zong et al., 2002), but only few data are availableon their characterization including especially the site ofrecombination in the two parental strains.

The goals of the present molecular epidemiological studywere to gain new insights into the genetic variability and

Fig. 1. Phylogenetic analysis of K1 nucleotide sequences. Phylogenetic comparison wK1 gene of 189HHV-8 isolates including, in red, the 35 new complete K1 sequences geusing GTRmodel (I = 0.1448, G = 0.7207). Horizontal branch lengths are drawn to scanode indicate the percentage of bootstrap samples (of 1000) in which the cluster to tindicate subtypes, groups and subgroups. D subtype was used as the outgroup. A, B, CC2, C3, C5 andC7 to subgroupswithin the subtypes described by Zong et al. (1999). Tbased on the (nearly) entire extra-cellular K1 domain (aa 28–238). Not all samples hcolour in this figure legend, the reader is referred to the web version of this article.)

evolution of HHV-8 in the Maghreb region, focusing ourstudy in Morocco. This Northwest African country, whosecurrent population is made of two main ethnic groups (i.e.,Arab and Berber-speaking population), is an endemic areafor classical KS and HHV-8 (Hbid et al., 2005). To achievethese goals, we specifically focused our work on thefollowing:

(1) The study of the complete K1 gene sequence and thegenotyping of the K14.1 locus of HHV-8 viral isolatesfrom a series of Moroccan patients suffering fromdifferent KS types including classic, AIDS-associated orpost-transplant types.

(2) The search and characterization of possible viral recom-binant forms of the K1 region in such HHV-8.

(3) The description of the population genetic background ofthe patients' cohort by analyzing genetic variation in thematernally inherited mitochondrial DNA (mtDNA).

Results

Epidemiological and clinical aspects of the patients

High molecular weight DNA was obtained from the 35skin biopsies, originating from 28 different Moroccanpatients with a KS (Table 1). Amplification of the K1gene was successful for the 35 DNA samples and thecomplete K1 sequence (870bp) was then obtained from allthe studied specimens, after cloning and sequencing. Themain clinical and epidemiological features (age, sex andtype of KS) of the 28 patients as well as the results of theK1 gene sequencing and of the K14.1 genotyping are listedin Table 1. This series included 17 patients (14 men and 3women) with a C-KS (mean age: 64 years, range 23–92),10 patients (5 men and 5 women) with an AIDS-KS (meanage: 37 years, range 29–45) and one man, of 41 years old,with an I-KS. In five cases, two biopsies were performed attwo different sites in the same patient (Table 1). The greatmajority of the patients were considered (by the clinicians incharge of them), as of Arab origin and none was ofEuropean or sub-Saharan African origin.

K1 genetic variability

Overall variabilityAmong the 35 sequences obtained, 28 were unique and

different one from the other. Concerning the remaining 7sequences, in one case, 3 sequences (D11/D6/D16) were foundidentical whereas the 4 others formed two pairs of identical

as performed on 867-bp fragments corresponding to the best alignment of entirenerated in this work. The phylogenywas derived by the neighbor-joiningmethodle, with the bar indicating 0.02 nucleotide replacement per site. Numbers on eachhe right is supported. Only bootstrap values of ≥75 are given. Bars on the rightand D correspond to 4 of the 5 HHV-8 subtypes and A1, A2, A3, A4, A5, and C1,heA′, Aʺ, C′ andCʺ grouping reflects the Cook et al. (1999) classification schemeave been labeled due to space constraints. (For interpretation of the references to

Table 1Main clinical and epidemiological features of the patients with Kaposi'ssarcoma analyzed in this study

Case Age Gender KS form K1 K14.1 Accession no.

D1 23 M Classic C7 P DQ394034D2 65 F Classic C3 P DQ394035D3 72 M Classic C3 P DQ394036D4-1 65 F Classic C5 P DQ394037D4-2 C5 P DQ394038D5-1 92 M Classic C3 P DQ394039D5-2 C3 P DQ394040D6 58 M Classic C7 P DQ394041D11 60 M Classic C7 P DQ394046D7 56 M Classic C3 P DQ394042D14-1 C3 P DQ394049D14-2 C3 P DQ394050

D9 59 M Classic C7 – DQ394043D10-1 63 M Classic C3 P DQ394044D10-2 C3 P DQ394045

D12 67 M Classic C3 P DQ394047D13 73 M Classic C2 M DQ394048D15 60 M Classic C3 P DQ394051D16 63 M Classic C7 P DQ394052D17 80 M Classic C3 P DQ394053D18 58 M Classic C7 P DQ394054D19 70 F Classic C1 P DQ394055D20 60 M Classic C7 P DQ394056I1 31 F Epidemic A M DQ394057I4 34 M Epidemic C2 P DQ394058I6 37 M Epidemic C3 P DQ394059I7 31 F Epidemic* C3 P DQ394060I8-1 34 F Epidemic C7 P DQ394061I8-2 C7 P DQ394062

I9 29 F Epidemic C3 P DQ394063I10 45 M Epidemic C5 P DQ394064I11 38 M Epidemic* C7 – DQ394065I12 44 M Epidemic C2 P DQ394066I13 42 F Epidemic C2 P DQ394067N1 41 M Iatrogenic C5 P DQ394068

Specimen identification, age, gender, type of Kaposi's sarcoma (KS) and resultsof the molecular typing of K1 and K14.1 loci for the 35 studied samples arepresented. D7/D14 and D6/D11 samples originated each from the same patientat a different time. Curly brace indicates samples originating from the samepatient. K1 sequences were determined by sequencing of the complete gene.Groups within C subtype are based on the classification of Cook et al. (1999)and Zong et al. (2002). The K14.1 M or P subtype was determined by a PCRassay as described in Materials and methods. “–” indicates that there was noamplification product. “*” indicates a pathologically uninvolved skin biopsy(absence of detectable spindle cells) of an AIDS-KS patient.

fff

f

f

f

124 R. Duprez et al. / Virology 353 (2006) 121–132

sequences (I11/D20 and D15/D12). Pairs of sequences ob-tained fromDNAs from different biopsy sites of the same patient(D4-1/D4-2, D5-1/D5-2, D6/D11, D7/D14-1/D14-2 and 18-1/18-2) exhibited between themselves less than 5 nucleotidedifferences (more than 99.4% of nucleotide similarity). Over-all, the 35 new sequences exhibited among themselves from 0%to 6.5% (I12 vs. N1) nucleotide divergence and from 0%to 14.4% (I8-2 vs. N1) amino acid divergence among pairwisecomparisons.

Phylogenetic studiesBecause tree-building algorithms rely on different

assumptions, we used two different methods – neighborjoining (NJ) and maximum parsimony (MP) – to increase

the robustness of the derived tree topologies. Furthermore,nucleotide and amino acid phylogenetic analyses wereperformed using two slightly different sets of sequences.The first one comprised the great majority of the completeK1 nucleotide sequences (around 870 bp) available inGenBank. The second set comprised the same translatedsequences plus few others (as SKS1, SKS2, BCP1, OKS3,etc.), representing different prototype strains, which are onlyavailable in a published paper (Zong et al., 1999) but not inGenBank.

The four previously described main molecular subtypes(A, B, C and D) were clearly identified on the basis of boththe consistent topology exhibited by the NJ (Fig. 1) and MP(data not shown) and the high bootstrap values obtained.Furthermore, some of the main molecular subgroups ofstrains previously individualized within A and C molecularsubtypes by Zong et al. (2002) and Cook et al. (1999) werealso identified in our study (e.g., A′, C″, C5, A5 clusters).However, this was not the case for several of the previouslyidentified subgroups within the A subtype (with theexception of the A5), which did not form monophyleticgroups and/or were not supported by high bootstrap valuesin both NJ and MP methods (Fig. 1). Interestingly enough,the C5 subgroup was observed to cluster in an intermediateposition between A and C subtypes (Figs. 1 and 2).

Among the 35 entire K1 new sequences obtained in thisstudy, all but one (I1 strain) were located in the C subtype.This unique non-C strain, originating from an AIDS-KSpatient, belonged to the A subtype and was closely relatedto some strains from Russian patients with KS (K1/E24 andK1/E32) (Fig. 1). The remaining 34 sequences clustered in4 different clades of the C subtype (Fig. 1). Thus, foursequences (N1, D4-1, D4-2 and I10) clearly belonged to the“intermediate” C5 subtype. Fifteen other new sequences (12from C-KS and 3 from E-KS) were located within a largecluster, comprising the BC2 prototype, and part of the C3subgroup (C′). Four other new sequences (D13, I4, I12 andI13) clustered with EKS1/2 in the C2 subgroup whereas theD19 strain clustered with ASM72 in the C1 subgroup. TheC1 and C2 were both clearly located in the highlysupported C″ cluster. Interestingly, 10 new sequencesclustered with BBG-1 sequence and few others and formeda highly supported monophyletic cluster. As BBG-1 wasproposed as the single C7 variant (Zong et al., 2002), wenamed this large monophyletic cluster the subtype C7 (Figs.1 and 2). Subtype-associated diseases (E-KS versus C-KS)was not observed within the C subtype of K1 gene in thisstudy.

Existence of recombination events in K1 geneThe intermediate position of the C5 cluster observed in

both of our analyses together with the fact that SKS9, a C5strain located in this cluster has been considered as arecombinant in a previous study (Zong et al., 2002) promptedus to study all remaining C5 sequences. Thus, in order todetermine the C5 parental origin, we used the Simplotsoftware and the boot-scanning method, in which the

Fig. 2. Phylogenetic analysis of K1 amino acid sequences. Phylogenetic tree was performed on 283 amino acid long K1 sequences by the neighbor-joining methodusing distances calculated with Kimura two-parameter method. This tree focuses in the amino acid sequences from the C group. The new Moroccan sequences areshown in red. The sequences available from GenBank and originating from patients of the Middle East and North Africa are shown in green. Horizontal branch lengthsare drawn to scale, with the bar indicating 0.02 amino acid replacement per site. Numbers on each node indicate the percentage of bootstrap samples (of 1000) in whichthe cluster to the right is supported. Only bootstrap values of ≥70 are given. Bars on the right indicate subtypes, groups and subgroups. C1, C2, C3, C5 and C7subgroups are defined by Zong et al. (1999). The C′ and C″ grouping reflects the Cook et al. (1999) classification scheme based on the (nearly) entire extracellular K1domain (aa 28–238). Not all samples have been labeled due to space constraints. (For interpretation of the references to colour in this figure legend, the reader isreferred to the web version of this article.)

125R. Duprez et al. / Virology 353 (2006) 121–132

phylogenetic relationship of the sequences is calculated usingbootstrap resampling. As seen in Fig. 3, the C5 cluster wouldresult from a recombination event between an A3 (A″) strain

and a C3 (C′) strain at a nucleotide position located around330 of the K1 gene. In order to confirm that the 2 studiedfragments can be used to reconstruct a phylogenetic tree,

Fig. 3. Analyzes of recombinant events of the C5 group. (A) C5 subgroup was compared with the Simplot programwith different groups (C′, C″, A′, A3, A5, B and D).The analysis used a 250-bp long window and a step of 20-bp long. The x values reflect the genome position at the midpoint of the analyze windows and the y valuesreflect the bootstrap value calculated from the windows. The vertical bar and the number indicate the site of probable recombination event. VR1 and VR2 regions areindicated. Triangles correspond to likelihood mapping results for each alignment fragment. As the central areas include less than 20% of quartets, we considered thateach alignment fragment is reliable for phylogenetic inference. (B/C) Phylogenetic trees corresponding to the first 330 nucleotides and the 537 last nucleotides,respectively. The technique is the same that for Fig. 1. The query group has been colored in blue. The 2 other groups of interest are colored with the same code used onthe Simplot output (A). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

126 R. Duprez et al. / Virology 353 (2006) 121–132

evolutionary information contained in each data set wasestimated by maximum likelihood; as the likelihood mappingshowed less than 20% of dots in the central area, separatephylogenetic analyses of the 2 fragments (1–330 and 331–867) was performed and these analyzes verified; indeed, theclustering of the C5 subgroup with A3 (A″) and C3 (C′)(Figs. 3B and C).

Similarly, because of its possible “intermediate” position, theC7 cluster, containing many new Moroccan strains, was alsoanalyzed using the same methodology. The Simplot softwareallowed us to determine that all the strains of this cluster resultfrom a recombinant event, which occurred between a C3 (C′)strain and a C1/C2 (C″) strain, at a nucleotide position locatedaround 585 of the K1 gene (at the beginning of the VR2 region)(Fig. 4).

Regions of variations and functional relevanceThe K1 genes of all the 35 studied strains were very

conserved in the N-terminal region, with the predicted signalpeptide sequence, and in the C-terminal region, including thepredicted hydrophobic transmembrane domain and the smallcytoplasmic domain. In contrast, the central region, containingthe predicted extracellular domain, was more variable with, aspreviously shown, two hypervariable domains, aa 51–92(VR1) and aa 191–228 (VR2). The cystein residues werehighly conserved with only sporadic mutations in two strains

(I12 and I13). In contrast, some of the 9 predicted N-linkedglycosylation sites (NXS/NXT) presented slightly morevariability. The ITAM motif, implicated in intracellularpathways leading to proliferation, appeared to be highlyconserved between all sequences, except in I1/N1. A specificdeletion of 5 aa (positions 201–205) was found in all the Csubtype strains as compared to the sole A strain (I1). Adeletion of 9 aa (positions 200–201 and 206–213), ascompared to all the other C strains, occurred in two sequences(I10 and N1). In addition, an insertion of one aa, at nucleotideposition 65 (already observed in some strains (Lacoste et al.,2000a, 2000b), was present in the sequence I1.

Subtype characterization of the K14.1 geneMolecular subtype characterization of the right hand side of

the HHV-8 genome, which comprises the K14.1 and K15 genes,was determined for 33 of the 35 samples. The large majority ofsamples (31/33) gave a 362-bp PCR product specific of the Psubtype (Table 1). In 2 cases, no amplification product wasobtained in our PCR studies suggesting a low viral load in these2 samples.

Genetic affinities of the studied human population

The sequence of the hypervariable region I (HVS-I) ofthe human mtDNA allowed us to infer the broad geographic

Fig. 3 (continued).

127R. Duprez et al. / Virology 353 (2006) 121–132

origin of most human samples as well as and to get insightsinto the maternal genetic composition of the populationunder study. The human mtDNA gene pool of the 25Moroccan studied samples, for which sufficient DNA wasavailable after HHV-8 analyzes, is constituted of three majorlineage-clusters of different geographic origins (i.e., WestEurasian, North African and sub-Saharan African) present atvarying frequencies. The West Eurasian component, whichencompasses lineages typically observed in the Middle East,Europe and North Africa, was mainly represented bylineages within the R macro-haplogroup and amounted to68%. The sub-Saharan African fraction reached an overallfrequency of 24% and was represented by the characteristichaplogroups L1 and L2 as well as L3–L3b. Finally, theautochthonous North African component represented byhaplogroup U6, which is thought to have originated inNorth Africa ∼40,000 years ago (Macaulay et al., 1999),was observed at a frequency of 8%.

Discussion

In this report, we have analyzed the genetic variation andpolymorphism of the complete K1 gene of HHV-8 of a series of35 viral strains, originating from 28 Moroccan patients withclassic, AIDS-associated or iatrogenic Kaposi's sarcomalesions. Morocco is a country of Northwest Africa and, togetherwith Algeria and Tunisia, belongs to the Maghreb region. Thisgeographic region is considered as an endemic area for HHV-8 and classic Kaposi's sarcoma. However, the few reports onsuch a topic in this region mainly concern series of clinical and/or pathological description of KS (Ben Tekaya et al., 2001;Benslimane et al., 1987; Chadli et al., 1976; Chakib et al., 2003;Kassimi et al., 2003; Mseddi et al., 2005; Ravisse, 1984). Thus,the genetic variability of the HHV-8 strains circulating in thepopulations of Maghreb remains largely unknown.

In our work, based on the complete sequence of the K1gene of HHV-8, we have shown that all but one of the 35

Fig. 4. Analysis of recombinant events of the C7 group. C7 subgroup was compared with the Simplot programwith different groups (C′, C″, A′, A3, A5, B and D). Theanalysis used a 250-bp long window and a step of 20-bp long. The x values reflect the genome position at the midpoint of the analyze windows and the y values reflectthe bootstrap value calculated from the windows. The vertical bar and the number indicate the site of probable recombination event. Triangles correspond to likelihoodmapping results for each alignment fragment.

128 R. Duprez et al. / Virology 353 (2006) 121–132

strains belong to the large C molecular subtype. Furthermore,we have also demonstrated that there exist a large geneticdiversity of these viruses within the C subtype with strainsbelonging to several and distinct subgroups highly supportedphylogenetically. Indeed, at least 4 distinct clusters (C3, C7,C″ and C5) contain some new strains from our Moroccanseries. The unique non-C strain of our series originates froman AIDS-KS patient and belongs to the A subtype (group A').This strain is closely related to some HHV-8 strains fromRussian patients with KS (Lacoste et al., 2000b). Lastly, ourresults as well as previously reported data from differentgeographical HHV-8 endemic areas have failed to observeany association of a specific viral subgroup with a specificform of KS (classic or AIDS associated) (Lacoste et al.,2000b; Zong et al., 2002).

In the Davidovici et al. (2001) work, 12 HHV-8 strains fromKS patients living in Israel (of Sephardi Jewish origin) butoriginating from Morocco, Algeria, Tunisia and Egypt wereanalyzed and all but two were found to be of the molecularsubtype C. The two others belong to the A group. Unfortu-nately, these 12 partial K1 sequences are not available inGenBank making impossible a precise comparison with ourfindings. However, visual comparative analyses of publishedphylogenetic trees comprising these sequences (Davidovici etal., 2001; Zong et al., 2002) raise the possibility that some ofthem may be very related to some strains reported in the presentstudy.

Based in part on the findings of some C″ (mainly C2)viruses among North African Sephardi Jews together with theobservation that most of the sequences originating from SaudiArabian patients belong to the C″ group (also mainly C2),Zong et al. hypothesized that C″ strains may be more ancientbranches than the C′ viruses. They thus suggested thatpossibly the C″ branch arose first in the Middle Eastern/NorthAfrican area, whereas the more recent C′ strains might berepresentative of European and North Asian migrations (Zonget al., 2002). In our study, if we consider that C7 subgroupbelongs to C″ group as suggested by Zong et al., we foundapproximately half of the Moroccan sequences as belonging tothe C″ group whereas the other half as belonging to the C′group. Several human population genetics studies have shownthat the present-day Moroccan population has mainlyoriginated from two major ethnic components, includingArab and Berber populations. It was thus tempting tohypothesize that most of the C″ strain presence may witnessthe Arabic/Middle Easter influence in North Africa whereasmost of the C′ stain may testify the more autochthonousBerber component. In this context, it has been unequivocallyshown a genetic link between Berber and European popula-tions that is attributable to European hunter–gathererspopulations who, in the end of the last glacial period 15,000years ago, moved towards the south and, by crossing the Straitof Gibraltar, contributed to the modern North African genepool (Achilli et al., 2005).

129R. Duprez et al. / Virology 353 (2006) 121–132

The mosaic picture of different ethno-geographic componentsobserved in the maternal gene pool (represented by mtDNAvariation) of our samples perfectly matches the lineage–clusterfrequencies previously reported in Arab and Berber populationsfrom Northwest Africa (Brakez et al., 2001; Cherni et al., 2005;Fadhlaoui-Zid et al., 2004). Most of these studies, based onmtDNA diversity, point to present-day North African populationsas (i) exhibiting a small fraction of a North African lineage whodeveloped in situ (i.e., the U6 haplogroup); (ii) sharing a commongenetic stock with West Eurasians (including the MiddleEasterners); and (iii) having incorporated sub-Saharan Africanlineages via trans-Saharan gene flow. Thus, our data showed thatthematernal gene pool of theMoroccan population here studied isin perfect agreement with the expected haplogroup diversity inNorth Africa. Therefore, this observation excludes particulardemographic events (i.e., bottleneck, genetic isolation, etc.)having shaped the genetic diversity of our cohort. In addition, ourresults do not provide evidence for any association between aparticular maternal host genetic background (i.e., North Africanvs. sub-Saharan African vs. West Eurasian lineages) and anyHHV-8 strain because both the C′ and C″ strains are randomlydistributed among the different patients' population backgrounds.Furthermore, it is interesting to note the absence, in our series, ofHHV-8 strains of the B subtype (which are restricted to sub-Saharan Africa) as well as among the 12 strains above quoted.This observation is surprising in light of the well-documented andancient interactions between North African and sub-SaharanAfrican populations, the latter being a very high endemic area forHHV-8 infection (Plancoulaine et al., 2004; Rezza et al., 2001).Thus, the absence of HHV-8 strains of the B subtype in ourpopulation sample is in strong contrast with the geneticarchitecture of the host's population, which exhibits a strongsub-Saharan African maternal component, as attested by the24% presence of lineages that ultimately find their originssouthwards of the Sahara Dessert. The observed contrast is ofinterest in view of the somehow “analogous” mode of trans-mission, at least in part, of the HHV-8 (mainly mother to childand between sibs) and the mtDNA (mother to offspring). Theseresults suggests therefore that the sub-SaharanAfrican gene flowcurrently observed in Moroccan populations was not accompa-nied by a concomitant transmission of HHV-8 B subtypes andhighlights a more complex dynamics of host-pathogen co-dispersals and transmission.

In our study, we have also characterized some recombinantchimeric K1 genes. The first group consists of the C5 strainsand the second group comprises all the C7 strains. The presenceof chimeric K1 gene is not without precedent as, at least twoteams, have previously demonstrated the presence of somerecombinant strains in this genomic region (Cook et al., 1999;Zong et al., 2002). Zong et al. have thus suggested that C6 andC4 groups were chimeric K1 genes with recombinationoccurring between the A and C branches for the former orbetween C″ and C′ branches for the latter. Cook et al. have alsosuggested that the sequences of their C″ subgroup possessed amembrane proximal region (VR2) related to the group Aisolates. However, they considered that these C″ sequences arenot true A/C recombinant but rather favored an independent

evolutive selection of these groups of isolates. The conclusionsof both these studies were mainly based on sequencecomparison and separate phylogenetic analyses.

Here, we have also used systematically a boot scanningwith the Simplot program to search and characterize possiblerecombinants in the series of the 35 novel Moroccan HHV-8 strains in a context of 189 complete K1 genes. This allows usto better define the two parental strains of the chimeric K1genes and gain new insights into the precise site ofrecombination. Thus, using such analysis, the C5 group,which comprises now 5 sequences (the SKS9 strain and 4 newones generated in this study), appears to be a true recombinantbetween two parental strains (an A3 and a C3), with the bestestimation of the recombination site located around nucleotideposition 330 of the protein. It is interesting to note that Zong etal. rather considered that SKS9 strain (the only C5 strain thatwas available at the time of their work) was a genuineevolutionary intermediate between A and C subgroups. Usingthe same analysis, it was also clear that all the C7 strainsoriginated from two different parental strains (C3 and C1/C2).Concerning the C6 and C4 strains, defined as chimeric K1gene by Zong et al. (2002). It was impossible to uses suchprogram because these sequences are unfortunately notavailable in GenBank.

The presence of intertypic and intratypic recombinantgenomes assumes dual infection in at least a single host(Beyari et al., 2003). The modes and timing of suchrecombinant events remain largely unknown and should bedetermined by further studies using especially novel method-ological approaches.

This study emphasizes also that the current classification ofHHV-8 strains should still considered as temporary and mostprobably will be modified with the description of new strainsfrom different geographical areas. Thus, as an example, the K1-43/Ber strain initially described either as a possible recombinantor a genuine specific evolutionary strain may be now consideredas the prototype strain of the new F type recently described(Kajumbula et al., 2006). Furthermore, as already suggested byCook et al. (1999) and Lacoste et al. (2000b), consensus on thecriteria used to define a subtype and a subgroup is absolutelyrequired.

Lastly, we consider that studies combining both viralphylogeny and human genetics will permit to gain new insightsinto the origin, evolution and modes of dissemination of thispeculiar human herpes oncovirus but also will open newavenues of research aiming to provide new insights of humanmigrations from a pathogen's perspective (Kajumbula et al.,2006).

Materials and methods

PatientTs samples

Thirty-five skin biopsies from patients with KS wereincluded in this study. The main epidemiological features ofthe patients, all originating from Morocco, are summarized inTable 1. In 33 cases, the studied samples corresponded to a

130 R. Duprez et al. / Virology 353 (2006) 121–132

tumor biopsy from AIDS-associated, post-transplant or classicKS patients and in 2 cases (I7 and I11) to clinically tumoral butpathologically uninvolved (absence of detectable spindle cells),skin of patients with epidemic KS. For 6 patients (D4, D5, D6/D11, D7/14, D10 and I8), two biopsies, at different sites, wereobtained. DNA from frozen tumor lesions was extracted aspreviously described (Lacoste et al., 2000a). The detailedclinical history and pathological findings of this series of KSpatients have been previously reported (Hbid et al., 2005).

HHV-8 molecular analysis

K1 gene amplificationFor all the DNA extracted from the frozen biopsies, the 870-

bp K1 coding region, from nucleotide positions 105–974, asnumbered according to the strain BC1 (Russo et al., 1996), wasamplified directly by polymerase chain reaction (PCR), usingthe Thermo-Start DNA Polymerase (ABgene, Epsom, UK), as a1073-bp PCR product as previously described (Lacoste et al.,2000a).

K14.1/K15 genes molecular characterizationMolecular subtype characterization of the right hand side of

the HHV-8 genome was determined as previously described(Poole et al., 1999). The PCR strategy used a triple primer set ofK14.1 covering the divergent junction of the two subtypes ofK15 genes.

Stringent precautions against PCR contamination weretaken. The amplification mixes were made in a special roomphysically separated from the laboratory and at least twonegative controls (mix, water or cellular DNA prepared fromHHV-8-negative samples) were included. PCR cycling condi-tions, cloning and sequencing were performed as previouslydescribed (Lacoste et al., 2000a).

Phylogenetic analyzesFor the phylogenetic analyses on nucleotide, multiple

sequence alignment was performed with the DAMBEprogram (v4.2.13) on the basis of a previous amino acidalignment created from the original sequences. The finalalignment was submitted to the Modeltest program (v3.6) toselect the best evolutionary model, according to the AkaikeInformation Criterion, to apply for the phylogenetic analyses.The phylogeny was derived by both the neighbor-joining(NJ) and maximum parsimony (MP) method, performed inthe PAUP program (v4.0b10) and the reliability of theinferred tree was evaluated by bootstrap analysis on 1000replicates.

For the analyses on amino acid, alignment was performedwith DAMBE program, Neighbor-joining method was per-formed in the Phylip 3.63 package with protdist and neighborprogram. The reliability of the inferred tree was evaluated bybootstrap analysis on 1000 replicates.

The 35 new entire ORF K1 sequences determined hereinhave been deposited in the National Center for BiotechnologyInformation database. GenBank accession numbers are listed inTable 1 and are DQ394034 through DQ394068.

Recombinant searchThe recombinant search was performed by boot scanning

with the Simplot program v3.5.1 (Lole et al., 1999). This pro-gram compares the cluster of sequences of interest (query) toseveral others. Phylogenetic relationships of these clusters areestimated for successive overlapping subregions (window 250bp; step 20 bp). For each subregion, the bootstrap value of thequery and the references are calculated and recorded (accordingto Kimura 2-parameters model; with 500 replicates). Bootstrapvalues are then plotted along the genome on an x/y plot, so that xvalues reflect the genome position at the midpoint of the analyzewindows and the y values reflect the bootstrap value calculatedfrom the windows. We can also determine recombinationbreakpoints and verify this by phylogenetic analyses of the 2fragments.

Evolutionary information contained in each alignmentfragment was confirmed by likelihood-mapping analyses (vonHaeseler and Strimmer, 2003). The method evaluates randomquartets drawn from N sequences and evaluates, for eachquartet, maximum likelihood values for each three possibletrees. The result is plotted inside an equilateral triangle, whereeach corner represents a specific tree topology. Seven mainareas in the triangle support different evolutionary information.The center represents sets where trees are poorly supported,named starlike area. From the biological standpoint, alikelihood mapping showing more than 20% of dots in thestarlike area suggests that the data are noisy and therefore notreliable for phylogenetic inference. The likelihood mapping wasperformed with the Tree-Puzzle v5.2 program.

Mitochondrial DNA analysesMitochondrial DNA hypervariable region I (HVS-I) was

amplified using the primers L15997 (5′-CACCATTAGCACC-CAAAGC-3′) and H16401 (5′-TGATTTCCGGAG-GATGGTG-3′). The amplification products were purified withthe QIAquick PCR purification kit (Qiagen). The sequencingreaction was carried out using the Big Dye Terminator cycle-sequencing ready reaction kit (AB Applied Biosystems) andsequencing products were run in an ABI 377 (AB AppliedBiosystems) automatic sequencer. The nucleotide positionsconsidered for the analyses were 16024–16383. Haplogroupassignation was performed using HVRI sequence motifs(Brakez et al., 2001; Richards et al., 2000; Salas et al., 2002).

Acknowledgments

Renan Duprez and Hbid Oumkaltoum were supported by theLa Ligue Nationale contre le Cancer, by the la Caisse Nationaled'Assurance Maladie et Maternité des Travailleurs Non Salariésdes Professions Non Agricoles and by a grant from theInternational Network of Pasteur Institute, respectively. We alsothank the International Network of the Pasteur Institute andAssociated Institutes, the Cancéropôle Ile-de-France, theINSERM and the Association de Recherche sur le Cancer forfinancial support. Luis Quintana-Murci and Hélène Quach aresupported by an Agence Nationale de la Recherche (ANR)research grant (ANR-05-JCJC-0124-01).

131R. Duprez et al. / Virology 353 (2006) 121–132

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at doi:10.1016/j.virol.2006.04.026.

References

Achilli, A., Rengo, C., Battaglia, V., Pala, M., Olivieri, A., Fornarino, S.,Magri, C., Scozzari, R., Babudri, N., Santachiara-Benerecetti, A.S.,Bandelt, H.J., Semino, O., Torroni, A., 2005. Saami and Berbers-anunexpected mitochondrial DNA link. Am. J. Hum. Genet. 76 (5),883–886.

Alagiozoglou, L., Sitas, F., Morris, L., 2000. Phylogenetic analysis of humanherpesvirus-8 in South Africa and identification of a novel subgroup. J. Gen.Virol. 81 (Pt. 8), 2029–2038.

Ben Tekaya, N., Zeglaoui, F., Ben Abdallah, T., Ben Chaabane, T., Fazaa, B.,Mokhtar, I., Kamoun, M.R., 2001. Kaposi's sarcoma in Tunisia. Review of91 cases. Tunis. Med. 79 (8–9), 429–433.

Benslimane, A., Essaket, S., Zahni, A., Sebti, A., Lakhdar, H., Veysseyre, C.,Sekkat, A., Carraz, M., Rollier, R., 1987. Immunovirologic profile ofKaposi's sarcoma in Morocco. Presse Med. 16 (18), 917.

Beral, V., 1991. Epidemiology of Kaposi's sarcoma. Cancer Surv. 10, 5–22.Beyari, M.M., Hodgson, T.A., Cook, R.D., Kondowe, W., Molyneux, E.M.,

Scully, C.M., Teo, C.G., Porter, S.R., 2003. Multiple human herpesvirus-8 infection. J. Infect. Dis. 188 (5), 678–689.

Biggar, R.J., Whitby, D., Marshall, V., Linhares, A.C., Black, F., 2000. Humanherpesvirus 8 in Brazilian Amerindians: a hyperendemic population with anew subtype. J. Infect. Dis. 181 (5), 1562–1568.

Boshoff, C., Weiss, R.A., 2001. Epidemiology and pathogenesis of Kaposi'ssarcoma-associated herpesvirus. Philos. Trans. R. Soc. London, Ser. B Biol.Sci. 356 (1408), 517–534.

Boshoff, C., Schulz, T.F., Kennedy, M.M., Graham, A.K., Fisher, C., Thomas,A., McGee, J.O., Weiss, R.A., O'Leary, J.J., 1995. Kaposi's sarcoma-associated herpesvirus infects endothelial and spindle cells. Nat. Med. 1(12), 1274–1278.

Brakez, Z., Bosch, E., Izaabel, H., Akhayat, O., Comas, D., Bertranpetit, J.,Calafell, F., 2001. Human mitochondrial DNA sequence variation in theMoroccan population of the Souss area. Ann. Hum. Biol. 28 (3), 295–307.

Cesarman, E., Chang, Y., Moore, P.S., Said, J.W., Knowles, D.M., 1995.Kaposi's sarcoma-associated herpesvirus-like DNA sequences in AIDS-related body-cavity-based lymphomas. N. Engl. J. Med. 332 (18),1186–1191.

Chadburn, A., Hyjek, E., Mathew, S., Cesarman, E., Said, J., Knowles, D.M.,2004. KSHV-positive solid lymphomas represent an extra-cavitary variant ofprimary effusion lymphoma. Am. J. Surg. Pathol. 28 (11), 1401–1416.

Chadli, A., Rethers, L., Landreat, A., Habanec, B., 1976. La physionomie ducancer en Tunisie. Etude de 7959 cancers primitifs. Arch. Inst. Pasteur Tunis4, 317–423.

Chakib, A., Hliwa, W., Marih, L., Himmich, H., 2003. Kaposi's sarcoma duringHIV infection in Morocco (apropos of 50 cases). Bull. Soc. Pathol. Exot. 96(2), 86–89.

Chang, Y., Cesarman, E., Pessin, M.S., Lee, F., Culpepper, J., Knowles, D.M.,Moore, P.S., 1994. Identification of herpesvirus-like DNA sequences inAIDS-associated Kaposi's sarcoma. Science 266 (5192), 1865–1869.

Cherni, L., Loueslati, B.Y., Pereira, L., Ennafaa, H., Amorim, A., El Gaaied,A.B., 2005. Female gene pools of Berber and Arab neighboringcommunities in central Tunisia: microstructure of mtDNA variation inNorth Africa. Hum. Biol. 77 (1), 61–70.

Chuck, S., Grant, R.M., Katongole-Mbidde, E., Conant, M., Ganem, D., 1996.Frequent presence of a novel herpesvirus genome in lesions of humanimmunodeficiency virus-negative Kaposi's sarcoma. J. Infect. Dis. 173 (1),248–251.

Cook, P.M., Whitby, D., Calabro, M.L., Luppi, M., Kakoola, D.N., Hjalgrim, H.,Ariyoshi, K., Ensoli, B., Davison, A.J., Schulz, T.F., 1999. Variability andevolution of Kaposi's sarcoma-associated herpesvirus in Europe and Africa.International Collaborative Group. AIDS 13 (10), 1165–1176.

Davidovici, B., Karakis, I., Bourboulia, D., Ariad, S., Zong, J., Benharroch, D.,Dupin, N., Weiss, R., Hayward, G., Sarov, B., Boshoff, C., 2001.Seroepidemiology and molecular epidemiology of Kaposi's sarcoma-associated herpesvirus among Jewish population groups in Israel. J. Natl.Cancer Inst. 93 (3), 194–202.

de-The, G., Bestetti, G., van Beveren, M., Gessain, A., 1999. Prevalence ofhuman herpesvirus 8 infection before the acquired immunodeficiencydisease syndrome-related epidemic of Kaposi's sarcoma in East Africa. J.Natl. Cancer Inst. 91 (21), 1888–1889.

Dukers, N.H., Rezza, G., 2003. Human herpesvirus 8 epidemiology: what we doand do not know. AIDS 17 (12), 1717–1730.

Dupin, N., Grandadam, M., Calvez, V., Gorin, I., Aubin, J.T., Havard, S., Lamy,F., Leibowitch, M., Huraux, J.M., Escande, J.P., Agut, H., 1995.Herpesvirus-like DNA sequences in patients with Mediterranean Kaposi'ssarcoma. Lancet 345 (8952), 761–762.

Fadhlaoui-Zid, K., Plaza, S., Calafell, F., Ben Amor, M., Comas, D., BennamarEl gaaied, A., 2004. Mitochondrial DNA heterogeneity in Tunisian Berbers.Ann. Hum. Genet. 68 (Pt. 3), 222–233.

Fouchard, N., Lacoste, V., Couppie, P., Develoux, M., Mauclere, P., Michel, P.,Herve, V., Pradinaud, R., Bestetti, G., Huerre, M., Tekaia, F., de The, G.,Gessain, A., 2000. Detection and genetic polymorphism of human herpesvirus type 8 in endemic or epidemic Kaposi's sarcoma from West andCentral Africa, and South America. Int. J. Cancer 85 (2), 166–170.

Gessain, A., Sudaka, A., Briere, J., Fouchard, N., Nicola, M.A., Rio, B.,Arborio, M., Troussard, X., Audouin, J., Diebold, J., de The, G., 1996.Kaposi sarcoma-associated herpes-like virus (human herpesvirus type 8)DNA sequences in multicentric Castleman's disease: is there any relevantassociation in non-human immunodeficiency virus-infected patients? Blood87 (1), 414–416.

Gessain, A., Mauclere, P., van Beveren, M., Plancoulaine, S., Ayouba, A.,Essame-Oyono, J.L., Martin, P.M., de The, G., 1999. Human herpesvirus8 primary infection occurs during childhood in Cameroon, Central Africa.Int. J. Cancer 81 (2), 189–192.

Hbid, O., Belloul, L., Fajali, N., Ismaili, N., Duprez, R., Tanguy, M., Benomar,H., Tahri, E.H., Gessain, A., Huerre, M., 2005. Kaposi's sarcoma inMorocco: a pathological study with immunostaining for human herpesvirus-8 LNA-1. Pathology 37 (4), 285–292.

Hengge, U.R., Ruzicka, T., Tyring, S.K., Stuschke, M., Roggendorf, M.,Schwartz, R.A., Seeber, S., 2002. Update on Kaposi's sarcoma and otherHHV8 associated diseases. Part 2: pathogenesis, Castleman's disease, andpleural effusion lymphoma. Lancet Infect. Dis. 2 (6), 344–352.

Huang, Y.Q., Li, J.J., Kaplan, M.H., Poiesz, B., Katabira, E., Zhang, W.C.,Feiner, D., Friedman-Kien, A.E., 1995. Human herpesvirus-like nucleic acidin various forms of Kaposi's sarcoma. Lancet 345 (8952), 759–761.

Kadyrova, E., Lacoste, V., Duprez, R., Pozharissky, K., Molochkov, V., Huerre,M., Gurtsevitch, V., Gessain, A., 2003. Molecular epidemiology of Kaposi'ssarcoma-associated herpesvirus/human herpesvirus 8 strains from Russianpatients with classic, posttransplant, and AIDS-associated Kaposi'ssarcoma. J. Med. Virol. 71 (4), 548–556.

Kajumbula, H., Wallace, R.G., Zong, J.C., Hokello, J., Sussman, N., Simms, S.,Rockwell, R.F., Pozos, R., Hayward, G.S., Boto, W., 2006. UgandanKaposi's sarcoma-associated herpesvirus phylogeny: evidence for cross-ethnic transmission of viral subtypes. Intervirology 49 (3), 133–143.

Kakoola, D.N., Sheldon, J., Byabazaire, N., Bowden, R.J., Katongole-Mbidde,E., Schulz, T.F., Davison, A.J., 2001. Recombination in human herpesvirus-8 strains from Uganda and evolution of the K15 gene. J. Gen. Virol. 82 (Pt.10), 2393–2404.

Kasolo, F.C., Monze, M., Obel, N., Anderson, R.A., French, C., Gompels, U.A.,1998. Sequence analyses of human herpesvirus-8 strains from both Africanhuman immunodeficiency virus-negative and -positive childhood endemicKaposi's sarcoma show a close relationship with strains identified in febrilechildren and high variation in the K1 glycoprotein. J. Gen. Virol. 79 (Pt. 12),3055–3065.

Kassimi, B.E., Benchemsi, N., Mikou, O., Ouazzani, E.E., Lakhdar, H., 2003.Kaposi's sarcoma and anti-herpes virus 8 antibodies in Morocco. Méd. Mal.Infect. 33, 226–228.

Kazanji, M., Dussart, P., Duprez, R., Tortevoye, P., Pauliquen, J.-F.,Vandekerhove, J., Couppié, P., Morvan, J., Talarmin, A., Gessain, A.,

132 R. Duprez et al. / Virology 353 (2006) 121–132

2005. Serological and molecular evidence that human herpesvirus 8 isendemic among Amerindians in French Guiana. J. Infect. Dis. 192,1525–1529.

Lacoste, V., Judde, J.G., Briere, J., Tulliez, M., Garin, B., Kassa-Kelembho, E.,Morvan, J., Couppie, P., Clyti, E., Forteza Vila, J., Rio, B., Delmer, A.,Mauclere, P., Gessain, A., 2000a. Molecular epidemiology of humanherpesvirus 8 in Africa: both B and A5 K1 genotypes, as well as the M and Pgenotypes of K14.1/K15 loci, are frequent and widespread. Virology 278(1), 60–74.

Lacoste, V., Kadyrova, E., Chistiakova, I., Gurtsevitch, V., Judde, J.G., Gessain,A., 2000b. Molecular characterization of Kaposi's sarcoma-associatedherpesvirus/human herpesvirus-8 strains from Russia. J. Gen. Virol. 81 (Pt.5), 1217–1222.

Lole, K.S., Bollinger, R.C., Paranjape, R.S., Gadkari, D., Kulkarni, S.S., Novak,N.G., Ingersoll, R., Sheppard, H.W., Ray, S.C., 1999. Full-length humanimmunodeficiency virus type 1 genomes from subtype C-infectedseroconverters in India, with evidence of intersubtype recombination. J.Virol. 73 (1), 152–160.

Macaulay, V., Richards, M., Hickey, E., Vega, E., Cruciani, F., Guida, V.,Scozzari, R., Bonne-Tamir, B., Sykes, B., Torroni, A., 1999. The emergingtree of West Eurasian mtDNAs: a synthesis of control-region sequences andRFLPs. Am. J. Hum. Genet. 64 (1), 232–249.

Mbulaiteye, S.M., Pfeiffer, R.M., Whitby, D., Brubaker, G.R., Shao, J., Biggar,R.J., 2003. Human herpesvirus 8 infection within families in rural Tanzania.J. Infect. Dis. 187 (11), 1780–1785.

Meng, Y.X., Spira, T.J., Bhat, G.J., Birch, C.J., Druce, J.D., Edlin, B.R.,Edwards, R., Gunthel, C., Newton, R., Stamey, F.R., Wood, C., Pellett, P.E.,1999. Individuals from North America, Australasia, and Africa are infectedwith four different genotypes of human herpesvirus 8. Virology 261 (1),106–119.

Meng, Y.X., Sata, T., Stamey, F.R., Voevodin, A., Katano, H., Koizumi, H.,Deleon, M., De Cristofano, M.A., Galimberti, R., Pellett, P.E., 2001.Molecular characterization of strains of Human herpesvirus 8 from Japan,Argentina and Kuwait. J. Gen. Virol. 82 (Pt. 3), 499–506.

Moore, P.S., Chang, Y., 1995. Kaposi's sarcoma findings. Science 270 (5233),15.

Moore, P.S., Chang, Y., 2001. Molecular virology of Kaposi's sarcoma-associated herpesvirus. Philos. Trans. R. Soc. London, Ser. B Biol. Sci. 356(1408), 499–516.

Mseddi, M., Abdelmaksoud, W., Slah, M., Taha, J.M., Bouassida, S., Boudaya,S., Turki, H., Zahaf, A., 2005. Kaposi's sarcoma. A series of 65 cases. Tunis.Med. 83 (9), 528–531.

Nador, R.G., Cesarman, E., Chadburn, A., Dawson, D.B., Ansari, M.Q., Sald, J.,Knowles, D.M., 1996. Primary effusion lymphoma: a distinct clinicopath-ologic entity associated with the Kaposi's sarcoma-associated herpes virus.Blood 88 (2), 645–656.

Nicholas, J., Zong, J.C., Alcendor, D.J., Ciufo, D.M., Poole, L.J., Sarisky, R.T.,Chiou, C.J., Zhang, X., Wan, X., Guo, H.G., Reitz, M.S., Hayward, G.S.,1998. Novel organizational features, captured cellular genes, and strainvariability within the genome of KSHV/HHV8. J. Natl. Cancer Inst.Monogr. (23), 79–88.

Plancoulaine, S., Abel, L., Tregouet, D., Duprez, R., van Beveren, M.,Tortevoye, P., Froment, A., Gessain, A., 2004. Respective roles ofserological status and blood specific antihuman herpesvirus 8 antibodylevels in human herpesvirus 8 intrafamilial transmission in a highly endemicarea. Cancer Res. 64 (23), 8782–8787.

Poole, L.J., Zong, J.C., Ciufo, D.M., Alcendor, D.J., Cannon, J.S., Ambinder,R., Orenstein, J.M., Reitz, M.S., Hayward, G.S., 1999. Comparison ofgenetic variability at multiple loci across the genomes of the major subtypesof Kaposi's sarcoma-associated herpesvirus reveals evidence for recombi-nation and for two distinct types of open reading frame K15 alleles at theright-hand end. J. Virol. 73 (8), 6646–6660.

Ravisse, P., 1984. Kaposi's sarcoma overseas (according to the OverseasPasteur Institutes). Bull. Soc. Pathol. Exot. Filiales 77 (4 Pt. 2),560–565.

Rezza, G., Danaya, R.T., Wagner, T.M., Sarmati, L., Owen, I.L., Monini, P.,Andreoni, M., Suligoi, B., Ensoli, B., Pozio, E., 2001. Human herpesvirus-8 and other viral infections, Papua New Guinea. Emerging Infect. Dis. 7 (5),893–895.

Richards, M., Macaulay, V., Hickey, E., Vega, E., Sykes, B., Guida, V., Rengo,C., Sellitto, D., Cruciani, F., Kivisild, T., Villems, R., Thomas, M., Rychkov,S., Rychkov, O., Rychkov, Y., Golge, M., Dimitrov, D., Hill, E., Bradley, D.,Romano, V., Cali, F., Vona, G., Demaine, A., Papiha, S., Triantaphyllidis, C.,Stefanescu, G., Hatina, J., Belledi, M., Di Rienzo, A., Novelletto, A.,Oppenheim, A., Norby, S., Al-Zaheri, N., Santachiara-Benerecetti, S.,Scozari, R., Torroni, A., Bandelt, H.J., 2000. Tracing European founderlineages in the Near Eastern mtDNA pool. Am. J. Hum. Genet. 67 (5),1251–1276.

Russo, J.J., Bohenzky, R.A., Chien, M.C., Chen, J., Yan, M., Maddalena, D.,Parry, J.P., Peruzzi, D., Edelman, I.S., Chang, Y., Moore, P.S., 1996.Nucleotide sequence of the Kaposi sarcoma-associated herpesvirus (HHV8).Proc. Natl. Acad. Sci. U.S.A. 93 (25), 14862–14867.

Salas, A., Richards, M., De la Fe, T., Lareu, M.V., Sobrino, B., Sanchez-Diz, P.,Macaulay, V., Carracedo, A., 2002. The making of the African mtDNAlandscape. Am. J. Hum. Genet. 71 (5), 1082–1111.

Schulz, T.F., 1998. Kaposi's sarcoma-associated herpesvirus (human herpesvi-rus-8). J. Gen. Virol. 79 (Pt. 7), 1573–1591.

Soulier, J., Grollet, L., Oksenhendler, E., Cacoub, P., Cazals-Hatem, D., Babinet,P., d'Agay, M.F., Clauvel, J.P., Raphael, M., Degos, L., Sigaux, F., 1995.Kaposi's sarcoma-associated herpesvirus-like DNA sequences in multi-centric Castleman's disease. Blood 86 (4), 1276–1280.

von Haeseler, A., Strimmer, K., 2003. Phylogeny inference based onmaximum-likelihood methods with TREE-PUZZLE. In: Salemi, M.,Vandamme, A. (Eds.), The Phylogenetic Handbook. A Practical Approachto DNA and Protein Phylogeny. Cambridge Univ. Press, Cambridge, pp.137–159.

Whitby, D., Howard, M.R., Tenant-Flowers, M., Brink, N.S., Copas, A.,Boshoff, C., Hatzioannou, T., Suggett, F.E., Aldam, D.M., Denton, A.S.,Miller, R.F., Weller, I.V.D., Weiss, R.A., Tedder, R.S., Schulz, T.F., 1995.Detection of Kaposi sarcoma associated herpesvirus in peripheral blood ofHIV-infected individuals and progression to Kaposi's sarcoma. Lancet 346(8978), 799–802.

Whitby, D., Marshall, V.A., Bagni, R.K., Wang, C.D., Gamache, C.J., Guzman,J.R., Kron, M., Ebbesen, P., Biggar, R.J., 2004. Genotypic characterizationof Kaposi's sarcoma-associated herpesvirus in asymptomatic infectedsubjects from isolated populations. J. Gen. Virol. 85 (Pt. 1), 155–163.

Zong, J.C., Metroka, C., Reitz, M.S., Nicholas, J., Hayward, G.S., 1997. Strainvariability among Kaposi sarcoma-associated herpesvirus (human herpes-virus 8) genomes: evidence that a large cohort of United States AIDSpatients may have been infected by a single common isolate. J. Virol. 71 (3),2505–2511.

Zong, J.C., Ciufo, D.M., Alcendor, D.J., Wan, X., Nicholas, J., Browning, P.J.,Rady, P.L., Tyring, S.K., Orenstein, J.M., Rabkin, C.S., Su, I.J., Powell,K.F., Croxson, M., Foreman, K.E., Nickoloff, B.J., Alkan, S., Hayward,G.S., 1999. High-level variability in the ORF-K1 membrane protein gene atthe left end of the Kaposi's sarcoma-associated herpesvirus genome definesfour major virus subtypes and multiple variants or clades in different humanpopulations. J. Virol. 73 (5), 4156–4170.

Zong, J., Ciufo, D.M., Viscidi, R., Alagiozoglou, L., Tyring, S., Rady, P.,Orenstein, J., Boto, W., Kalumbuja, H., Romano, N., Melbye, M., Kang,G.H., Boshoff, C., Hayward, G.S., 2002. Genotypic analysis at multipleloci across Kaposi's sarcoma herpesvirus (KSHV) DNA molecules:clustering patterns, novel variants and chimerism. J. Clin. Virol. 23 (3),119–148.