Characterization of arenaviruses using a family-specific primer set for RT-PCR amplification and...

Virus Research 49 (1997) 79–89

Characterization of arenaviruses using a family-specific primerset for RT-PCR amplification and RFLP analysis

Its potential use for detection of uncharacterized arenaviruses

M.E. Lozano a,b, D.M. Posik a,b, C.G. Albarino a,b, G. Schujman a,P.D. Ghiringhelli a,b, G. Calderon c, M. Sabattini c, V. Romanowski a,b,*a Instituto de Bioquımica y Biologıa Molecular, Depto. de Ciencias Biologicas, Facultad de Ciencias Exactas,

Uni6ersidad Nacional de La Plata, Calles 47 y 115, 1900 La Plata, Buenos Aires, Argentinab Departamento de Ciencia y Tecnologıa. Centro de Estudios e In6estigaciones. Uni6ersidad Nacional de Quilmes,

Roque Saenz Pena 180, 1876 Bernal, Argentinac Instituto Nacional de Estudios sobre Enfermedades Virales Humanas, Monteagudo 2510, 2700 Pergamino, Argentina

Received 7 November 1996; received in revised form 29 January 1997; accepted 29 January 1997

Abstract

Arenaviruses are enveloped viruses with a genome composed of two ssRNA species, designated L and S. Thearenaviruses were divided in two major groups (Old World and New World), based on serological properties andgenetic data, as well as geographic distribution. A sequence alignment analysis of all reported arenavirus S RNAsyielded 17 conserved regions in addition to a reported conserved region at the end of both RNAs. The consensussequences of these regions were used to design generalized primers suitable for RT-PCR amplification of a set ofoverlapping nucleotide sequence fragments comprising the complete S RNA of any arenavirus. A restriction analysis(RFLP) was designed to rapidly typify the amplified fragments. This RT-PCR-RFLP approach was tested with OldWorld (LCM) and New World (Junın and Tacaribe) arenaviruses. Furthermore, using this procedure the whole SRNA of a novel arenavirus isolate obtained from a rodent trapped in central Argentina, was amplified andcharacterized. Partial nucleotide sequence data were used for phylogenetic analyses that showed the relationshipsbetween this arenavirus and the rest of the members of the family. This relatively simple methodology will be usefulboth in basic studies and epidemiological survey programs. © 1997 Elsevier Science B.V.

Keywords: Arenaviridae; RFLP-RT-PCR; Conserved nucleotide sequences; Phylogenetic analyses

* Corresponding author. Tel.: +54 21250497 (ext. 32); fax: +54 21259223; e-mail: [email protected]

0168-1702/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved.

PII S 0 1 6 8 -1702 (97 )01458 -5

M.E. Lozano et al. / Virus Research 49 (1997) 79–8980

1. Introduction

Arena6iridae is a family composed of a grow-ing number of enveloped, single stranded RNAviruses with at least 18 recognized membersaround the world. All members of the familyArena6iridae were subdivided into two groupsbased on geographical site of isolation, serologi-cal crossreactivities and genetic data.

The prototype of the family, lymphocyticchoriomeningitis (LCM) virus, is the only mem-ber with a worldwide distribution, whilst allother described arenaviruses are known to be-long to a restricted geographical area. LCM isa member of the Old World arenavirus group,which also includes Ippy, Lassa, Mobala andMopeia viruses. The other arenaviruses wereincluded in the New World arenavirus group,that comprises Amapari, Flexal, Guanarito,Junın, Latino, Machupo, Parana, Pichinde,Tacaribe and Tamiami viruses (McCormick,1990; Salas et al., 1991). More recently, threenew members of the New World group wererecognized, Sabia in Brazil, Oliveros in Ar-gentina and Whitewater Arroyo in USA (Lisieuxet al., 1994; Bowen et al., 1996a; Fulhorst et al.,1996).

All arenaviruses, except Tacaribe, have a ro-dent host and Lassa, Machupo, Junın, Guanar-ito and Sabia are known to be highlypathogenic for humans. In particular, the SouthAmerican viruses produce hemorrhagic feverscalled Argentine, Bolivian and Venezuelan hem-orrhagic fever, respectively, three endemoepi-demic diseases with cardiovascular, renal,immunological and neurological alterations.

The arenavirus genome is composed of twosingle stranded RNA species, designated L (forlarge, ca. 7 kb) and S (for small, ca. 3.5 kb).The complete nucleotide sequence of the LRNA is known only for LCM (Salvato and Shi-momaye, 1989; Salvato et al., 1989), Tacaribe(Iapalucci et al., 1989a,b) and Pichinde viruses(Harnish, personal communication). Conversely,the nucleotide sequence of the S RNA is knownfor Pichinde (Auperin et al., 1984), LCM (Ro-manowski et al., 1985), Tacaribe (Franze-Fer-

nandez et al., 1987), Lassa (Auperin andMcCormick, 1989; Clegg et al., 1990), Junın(Ghiringhelli et al., 1991), Mopeia (Wilson andClegg, 1991), Oliveros (Bowen et al., 1996a)and, more recently, for Sabia (Gonzalez et al.,1996). In addition, the 3% half of Machupo SRNA has been determined (Griffiths et al.,1992). The L RNA, encodes the RNA poly-merase (L) and a zinc-finger-like protein (Z) andthe S RNA codes for the viral nucleocapsidprotein (N) and the precursor of the envelopeglycoproteins (GPC). The open reading framesof both RNA species are arranged in oppositeorientations (ambisense; Bishop, 1990), and areseparated by a non-coding intergenic region thatfolds into a stable secondary structure (Ro-manowski, 1993; Salvato, 1993). The N and Lproteins are translated from antigenomic (or vi-ral complementary) sense mRNA species thatare encoded by the 3% half of the viral S or LRNA, respectively. The GPC and Z are trans-lated from a genomic (or viral) sense mRNAcorresponding to the 5% half of the S or LRNA, respectively (Salvato, 1993).

All arenaviruses that were sequenced share a19 nucleotide long sequence at the 3% end oftheir S RNAs (Auperin et al., 1982). In addi-tion, this sequence is almost fully conserved inthe L RNAs and bears a high degree of comple-mentarity to the 5% termini of both S and LRNA. We used the reported nucleotide se-quences to identify additional conserved regionsin order to design generalized amplificationprimers. These primers were designed to amplifyby RT-PCR a set of overlapping cDNA frag-ments comprising the whole S RNA from allarenaviruses. Moreover, we achieved a restric-tion analysis (RFLP) to typify the amplifiedfragment and eventually to identify novelviruses.

In view of its characteristics (quick procedure,simplicity and relatively low cost) this RT-PCR-RFLP procedure could be useful as a first stepin the molecular characterization of the are-naviruses for screening projects such as in epi-demiological studies when large numbers of fieldsamples are being analyzed.

M.E. Lozano et al. / Virus Research 49 (1997) 79–89 81

2. Methods

2.1. Chemicals

All reagents were P.A. or molecular biologygrade, supplied by Merck (Darmstadt, Germany),Sigma (St. Louis, MO) or Carlo Erba (Milano,Italy). The enzymes were from stratagene (La Jolla,USA), Promega (Madison, USA) or New EnglandBiolabs (Beverly, USA). The tissue culture mediawere from Gibco (Grand Island, USA) and serawere from GEN (Buenos Aires, Argentina).

2.2. Virus and cell culture

Lymphocytic choriomeningitis virus (WEstrain), Tacaribe virus, Pichinde virus and Junınvirus, MC2 strain (Ghiringhelli et al., 1991), werepropagated in BHK21 C13 cells. BHK cells weregrown to 50% confluency and infected with theviruses at a multiplicity of approximately 1 PFUper cell. After virus adsorption for 1 h at 37°C,infected cells were washed with phosphate-bufferedsaline (PBS) and maintained at 37°C in mediumcontaining 2% fetal bovine serum. The viruses wererecovered from the supernatant media of infectedmonolayers and purified by ultracentrifugation(Ghiringhelli et al., 1991). Virus titers were deter-mined by plaque assay on Vero E6 cell monolayersas described elsewhere (Rustici, 1984).

BHK21 C13 (ATCC, CRL8544) and Vero E6(ATCC, CRL1586, C1008) cells were grown inDulbecco’s minimal essential medium supple-mented with 10% fetal calf serum and 2 mML-glutamine, and were maintained in the samemedium containing 2% fetal bovine serum.

Vero cells (ATCC, CCL 81) were grown inMinimum Essential Medium with Earle’s saltssupplemented with non-essential amino acids, 1%penicillin, 1% streptomycin, 1% L-glutamine and10% heat-inactivated fetal bovine serum and weremaintained in the same medium containing 2%fetal bovine serum.

2.3. Virus isolation from rodents

Rodents of different species (Calomys musculi-nus, Bolomys sp., Akodon azarae, Mus musculus,

Oligoryzomys fla6escens) were captured and differ-ent tissues (cerebrum, lung, blood) were extractedand kept frozen for subsequent analysis. An are-navirus was detected in a tissue sample from aBolomys sp., captured at Maciel, Province of SantaFe, Argentina. This virus (Strain PAn 18400,Calderon, et al., unpublished results), was isolatedin Vero CCL 81 cells. After virus adsorption for 1h at 37°C, infected cells were maintained at 37°Cin medium containing 2% fetal bovine serum.Infected-Vero CCL 81 cells were harvested 6 dayspost-infection.

The virus was characterized by immunofluores-cence at the Instituto Nacional de EnfermedadesVirales Humanas (INEVH, Pergamino, Ar-gentina).

2.4. RNA isolation

RNA was prepared from pelleted virions andinfected cell samples using the method of Chom-czynski and Sacchi (1987) and from tissue samplesusing a modification of that method (Lozano et al.,1993a). The pelleted virions were disrupted by with1×GTC solution (4 M guanidinium isothio-cyanate, 0.5% sarkosyl, 25 mM sodium citrate).BHK cells were harvested 72 h post infection,washed with PBS and lysed with 1×GTC. Tissuesamples were homogenized with 1 volume of 2×GTC and transported at room temperature to theInstituto de Bioquımica y Biologıa Molecular(IBBM, La Plata, Argentina), 400 km away fromthe INEVH, to proceed with the RNA purification.

These GTC containing lysates were extractedwith acid phenol/Cl3CH/isoamyl alcohol, precipi-tated with one volume of isopropanol, reprecipi-tated with ethanol and dissolved in 50 m l of H2O.The RNA solution was stored at −70°C.

2.5. RT-PCR

RNA samples were heated to 95°C for 5 min inthe presence of ARS1, ARS8V and ARS6Cprimers (2 mM, final concentration; see Fig. 1 andTable 1) and denatured with methyl mercury(II)hydroxide (10 mM, final concentration) for 5 minbefore initiating reverse transcription. After inac-


Fig. 1. Design of arenaviruses generalized primers. The long open bar represents the arenavirus S RNA, numbers indicating relativepositions in the PILEUP comparison file. The black boxes correspond to non-coding sequences. The short lines above the S RNAindicate the positions of the conserved regions. The positions of the generalized primers (ARS series, for arenavirus S RNA) areindicated by arrows below the S RNA, with their designation composed of a number and a letter indicating their polarity (V forviral and C for viral-complementary). The lengths of the primers are not drawn to scale, but their 3% ends correspond to the arrowheads. The open bars with grey boxes at their ends represent a set of RT-PCR amplification fragments of different lengths obtainedwhen using the appropriate combinations of generalized primers on different arenaviruses.

tivation of the denaturant with 14 mM 2-mercap-toethanol, the first-strand cDNA synthesis wascarried out in a total volume of 10 m l containing5 m l of the total RNA, 60 mM KCl, 25 mMTris–HCl (pH 8.0), 10 mM MgCl2, 1 mMdithiotrehitol, 0.5 mM (each) deoxynucleosidetriphosphates (dNTPs), 7.5 U RNasin (Promega)and 4 U AMV reverse transcriptase (Promega).The reaction mixture was incubated for 1 h at42°C. The first-strand cDNA reaction was ethanolprecipitated with 100 mM sodium acetate and 2.5mg of linear polyacrylamide and dissolved in 10 m lof H2O.

PCR amplifications were carried out in 10 m lfinal volume, containing 1 m l of the cDNA reac-tion, 0.125 units of Taq DNA polymerase(Promega), 0.5 mM each primer, 200 mM eachdNTP, 50 mM KCl, 10 mM Tris–HCl pH 8.3, 1.5mM MgCl2 and 0.01% gelatin. The mixture wasoverlaid with mineral oil. The PCR cycle progres-sion for the ARS1/ARS16V and ARS6C/ARS3Vprimer pairs was as follows: 2 min at 92°C, fol-lowed by 35–40 cycles of 30 s at 92°C (denatura-tion), 30 s at 55°C (annealing) and 1 min at 72°C(extension) in a Hybaid thermal reactor (Tedding-ton, UK). The PCR cycle progression for theARS16C/ARS8V and ARS3C/ARS1 primer pairs

was similar, except that 30 s at 50°C in theannealing and 3 min at 72°C in the extension wereused. A final incubation for 5 min at 72°C wasperformed to allow completion of unfinishedDNA strands.

A 4 m l aliquot of the PCR reaction mixture wasloaded on a 1.5% agarose gel in TAE buffer (40mM Tris–acetate, 1 mM EDTA, pH 8.0), elec-trophoresed at 4 V/cm for 45–60 min, stainedwith 0.5 mg/ml ethidium bromide and pho-tographed with Polaroid 667 film.

2.6. RFLP

cDNA amplification fragments from differentarenaviruses were size-fractionated by gel elec-trophoresis and purified by NaI and glass powderadsorption and elution (Bio101, La Jolla, CA).The eluted fragments were digested with HinfIrestriction enzyme. Alternatively, the RT-PCRproduct was digested, without prior purification,adjusting the MgCl2 concentration, according tothe optimum buffer for restriction enzyme diges-tion. The digested products were resolved by gelelectrophoresis in 4% agarose in TAE buffer,stained with ethidium bromide and photographedwith Polaroid 667 film.


2.7. Cloning and sequencing

The amplification fragment from the uncharac-terized sample was inserted by ligation into amodified pBS KS+ vector (Bluescript KS+,Stratagene, La Jolla, CA). The pBS KS+ vectorwas modified by digestion with HincII and treat-

ment with Taq DNA pol in the presence of dTTPto extend the 3% ends. The resulting overhang-Tannealed to the untemplated A at the 3% ends ofthe PCR fragment in the presence of T4 DNAligase. The ligation reactions were used to trans-form E. coli DH5 a F% cells. Selection was doneon LB-agar plates containing 100 mg/ml ampi-cillin, IPTG (30 m l of 100 mM stock per 10 cmPetri dish) and X-gal (50 m l of 2% stock inN,N %-dimethylformamide per 10 cm Petri dish)(Sigma, St. Louis, MO). The nucleotide sequencesof five independent clones were determined by thedideoxy method of Sanger et al. (1977) using amodified T7 DNA polymerase (Sequenase™ 2.0,USB, Cleveland, OH) and a [32P]dATP in the re-action mixture.

2.8. Data analysis

Nucleotide sequences of the S RNA from are-naviruses were obtained from GENBank (Na-tional Institute of Health, Bethesda, Maryland,USA). The accession numbers were: LCM WEstrain, M22138; LCM Armstrong strain, M20869;Lassa Nigeria strain, X52400; Lassa Josiah strain,J04324; Mopeia, M33879; Pichinde, K02734;Tacaribe, M20304 and M65834; Junın MC2strain, D10072; Junın XJ strain, U70803 andU70804; Oliveros virus, U34248 and SabiaU41071.

The sequence information was processed andanalyzed on a MicroVAX 3100 (Digital) com-puter using the package from Genetics ComputerGroup (GCG, Sequence Analysis Software Pack-age, version 7.1, University of Wisconsin,Madison, WI). Sequence alignments were doneusing the PILEUP program with the defaultparameters, and a consensus sequence was ob-tained with the PRETTY program. Restrictionmaps were generated with MAPSORT program.

Dendrograms were obtained as an output op-tion of the PILEUP program. Furthermore, thephylogenetic analysis of nucleotide and aminoacid sequences were done on a personal computer,with the programs DNAPARS and PROTPARS,respectively, both from the PHYLIP package(Phylogeny Inference Package, version 3.5, JosephFelsenstein, University of Washington, USA).

Table 1Arenavirus generalized primers

Name LocationbSequence (5%�3%)a

CGCACCGGGGATCCTAGGCARS1 13647c

CATGACKMTGAATTYTGTG-ARS3V 1117ACA

ARS3C TCAAAKAGCCKCAGCRTGT- 1154C

ARS4C 1172ATAGCATTYTTGTTGWAATCAAYTACACMAARTTTTGGT- 1297ARS5VARAGAGRCAGGGGARAACT-ARS6V 1465CC

ARS6C 1487AATGGAGTGYTRCCCTGCCTARS7V ATACCAACTCAYAGRCACAT 1558ARS7C ATGTGYCTRTGAGTTGG 1577

1918ARS8V ATRCAGTCWAKCAGTGCAC-A

ARS8C1 1931CTGMTGGACTGCATMATGT-T

ARS8C2 GTTGMCCCTCACTGTGC 1949ARS9V 2134CTGATGTCATCWGAMCCTT-

G2181ARS10C GCTSCCTCMGRARATGGT

ARS13V TCMACWATHGTGTTTTCCC- 2546A

ARS14C AKAGRAAYCCTTATGAAAA 2650GGCATWGANCCAAACTGA- 3007ARS16VTT

ARS16C CAAYTKAACAATCARTTTGG 30353071ARS17C GGTGTTGTRAGRGTKTGGG-

A

a The following symbols were used for mixed bases: Y, C/T; R,A/G; S, C/G; M, A/C; W, A/T; K, G/T; H, A/C/T; N,A/C/G/T.b The number indicates the location of the 5% termini of theprimer in the multiple sequence alignment file obtained usingthe PILEUP program (GCG package) with the defaultparameters (Gap Weight, 5; GapLengthWeight, 0.3) on the SRNA sequences (see design of generalized primers).c Primer ARS1 is fully complementary to the 3%-most-terminalnucleotides of the viral-sense arenavirus S RNA and yields aheteroduplex with the 3% terminus of the viral-complementarysense arenavirus S RNA.


3. Results

3.1. Design of generalized arena6irus primers

All members of the arena6iridae family have aconserved sequence at the 3% termini of their SRNAs (Auperin et al., 1982). An oligonucleotidethat hybridizes to this region was used previouslyin our laboratory in the synthesis of cDNA and inRT-PCR amplifications of Junın virus samples(Ghiringhelli et al., 1991; Lozano et al., 1993b).

Several well conserved regions were identified inthe alignment of the S RNAs of the arenavirusesLCM (WE and Armstrong strains), Lassa (Nige-ria and Josiah strains), Mopeia, Pichinde,Tacaribe, Junın (MC2 and XJ strains), Oliverosand Sabia. Nineteen of these regions were used todesign primers to RT-PCR amplify different ge-nomic fragments from all members of the Are-naviridae (Fig. 1).

These primers were designed, with at least, fiveconsecutive strictly conserved bases at their 3% ends.In the remaining 15–17-base stretch, we used thefollowing criteria: firstly, the primer should notdiffer more than 20% with respect to any listedsequences and, secondly, degenerations were al-lowed to enhance the annealing of each primer toany particular arenavirus S RNA. In general, weprivileged primers with minor differences with thealigned sequences over the strict consensus.

We named these generalized primers with theprefix ‘ARS’ (indicating its common presence inthe S RNA of the arenavirus family), followed bya number identifying the conserved region (1through 19, Fig. 1) and the polarity (‘V’, for viralor genomic and ‘C’ for viral complementary orantigenomic sense). The nucleotide sequences ofthe generalized primers are indicated in Table 1.As shown in Fig. 1 by open rectangles (below theS RNA) these primers could be used to obtain atleast seven overlapping amplification fragmentsthat span the whole S RNA. Six of the products(600 bp average length) may be sequenced with-out subcloning. Due to the low degree of conser-vation at the 5% region of GPC gene, theamplification fragment resulting from an RT-PCRwith the closest generalized primers, would have alength of 1170 bp.

Fig. 2. RT-PCR/RFLP arenavirus characterization. ExpectedHinfI restriction pattern of the RT-PCR amplified fragmentusing the ARS1/ARS16V primers on different arenaviruses SRNAs. The upper two boxes above each lane indicate thename of each virus and the length of the correspondingamplified fragment. Below, the boxed numbers represent therestriction fragments with their expected sizes inside, as if theywere resolved by gel electrophoresis. JUN-MC2, MC2 strainof Junın virus; JUN-XJ, XJ strain of Junın virus; MAC,Machupo; TAC, Tacaribe; OLV, Oliveros; PIC, Pichinde;LCM-WE; WE strain of LCM virus; LCM-Arm, Armstrongstrain of LCM virus; LAS-Jos, Josiah strain of Lassa virus;LAS-Nig; Nigeria strain of Lassa virus; MOP, Mopeia.

3.2. Characterization of arena6irus byRT-PCR/RFLP

The restriction maps of the amplification frag-ments shown in Fig. 1, were calculated for differ-ent enzymes. The HinfI restriction map of theARS1/ARS16V amplification fragment allowed usto distinguish at least those arenaviruses for whichthe S RNA sequence is available (Fig. 2). Wetested this pair of primers on several viral RNAsamples. These primers define an RT-PCR am-plification fragment of approximately 600 bpcomprising the 3% non-coding region of the ge-nomic S RNA and a gene segment coding for thefirst 170 amino acids of the N protein (Fig. 1).The amplified fragments from Junın, LCM andTacaribe arenaviruses (Fig. 3a, lanes J, L and T)were gel purified and digested with HinfI (Fig. 3b,lanes J, L and T). The amplified fragment and the


Fig. 3. RT-PCR and RFLP analysis of arenaviruses S RNAs. Agarose gel electrophoresis of the products obtained by (a) RT-PCRamplification using the ARS1/ARS16V primer pair on total cellular RNA from infected BHK21 cells and (b) HinfI digestions ofthese fragments (RFLP). M, marker; J, Junın virus MC2 strain; L, LCM virus WE strain; T, Tacaribe virus; P, Pampa virus.

HinfI restriction patterns were in agreement withthe expected sizes (Fig. 2, lanes JUN-MC2, LCM-WE and TAC).

We have also used the generalized primer pairsARS16C/ARS8V, ARS8C/ARS6V, ARS6C/ARS3V and ARS3C/ARS1 to RT-PCR amplifythe whole S RNA of Junın, LCM and Tacaribevirus (not shown). In the amplification reactionthe cycling parameters of the hybridization tem-perature and the extension time were optimized inorder to obtain the expected product for each pairof primers.

3.3. Characterization of a no6el isolate

A rodent capture program is being conductedby the INEVH in order to pursue epidemiologicalstudies for Argentine Hemorrhagic Fever (AHF),the endemo-epidemic disease caused by Junınvirus. The immunochemical screening revealed thepresence of an arenavirus in a tissue sample froma rodent (Bolomys sp.) captured in the province ofSanta Fe, 50 km outside the AHF endemic area.This virus was isolated by passage on Vero CCL81 cells and immunologically characterized, ex-hibiting a stronger cross-reactivity with Oliveros

than with Junın and LCM viruses (Calderon,1996).

Total RNA from Vero infected cells was pre-pared for RT-PCR/RFLP analysis as describedabove. The ARS1/ARS16V RT-PCR amplifiedfragment shown in Fig. 3a, lane P had an appar-ent molecular weight of 600 bp, which is similarto that from other arenaviruses. The RT-PCRamplification of RNA samples from different tis-sues of the same rodent (brain, lung and blood)yielded identical results. The HinfI restriction pat-tern of the ARS1/ARS16V amplified fragment,was different from that of the other viruses ana-lyzed (Fig. 3b) and from the calculated pattern ofthe rest of the arenaviruses with a known S RNAsequence (Fig. 2). Thus, these results are sugges-tive of the presence of a novel isolate, which was,provisionally, named Pampa.

In view of this, the amplification fragment wascloned and sequenced. This sequence informationwas employed in a pairwise comparison with thehomologous region from the total set of the re-ported arenavirus sequences. In particular, Pampavirus showed the closest homology with Oliveros(i.e. nucleotide sequence identity of 84%), as ex-pected from the immunological data (Calderon,1996).


3.4. Phylogenetic analysis

The nucleotide and the predicted amino acidsequences of the ARS1/ARS16V amplified frag-ment were aligned to obtain a multiple alignmentfile and a dendrogram (Fig. 4). The constructionof dendrograms is based on the comparison be-tween all the possible pairs of aligned sequences,establishing global similarity values. Identicalcluster relationships were observed when the nu-cleotide or predicted amino acid sequences wereanalyzed. In these dendrograms, it was observedthat Old World arenaviruses, Mopeia, Lassa andLCM grouped independently of New World

arenaviruses Junın, Tacaribe, Oliveros, Pampa,Pichinde and Machupo viruses, as noted previ-ously (Bowen et al., 1996a; Clegg, 1993). Also, itwas observed that Pampa and Oliveros viruses aregrouped together (Fig. 4).

In order to confirm the results obtained withthe procedure described above, we used the cladis-tic approach to generate a phylogenetic tree. Thisstrategy provided a classification that reflects thegenealogical relationships and the establishmentof a hypothetical reconstruction of the phylogeny.The DNAPARS and the PROTPARS programsfrom the PHYLIP package yielded the same clus-ter relationship as those as obtained in the abovementioned dendrograms.

4. Discussion

The arenaviruses share morphological andmolecular properties. The latter include a com-mon sequence present at both termini of theirviral S RNAs. In this paper we described a num-ber of other genomic regions that exhibit con-served nucleotide sequences (Fig. 1). From theanalysis of conserved regions we designed general-ized primers suitable for RT-PCR amplification,cloning and sequencing the complete S RNA ofall arenaviruses. Furthermore, we described anRT-PCR/RFLP (Fig. 2) designed to rapidly char-acterize the amplified arenavirus cDNAs a testthat could be useful in epidemiological studies.

Subsequently, a virus immunologically relatedto Oliveros virus was isolated from a rodent(Bolomys sp.), of the same species from whichOliveros virus was isolated. This virus was charac-terized using the RT-PCR/RFLP approach de-scribed here and the HinfI restriction pattern ofits ARS1/ARS16V amplification fragment dif-fered from other characterized arenaviruses, in-cluding Oliveros (Figs. 2 and 3b). We named thenovel isolate Pampa virus, provisionally, after thePampas, the phytogeographic region where it wasisolated. Recently, we characterized another novelarenavirus isolate using the same approach (datanot shown).

The relationship between virus species and theirhost rodents have been studied for several hanta-

Fig. 4. Dendrogram of the Arena6iridae family. The sequencesof the region flanked by the ARS1 and ARS16V primers indifferent arenavirus (obtained from GeneBank, and in thispaper) were aligned and a dendrogram was constructed usingan output option of the PILEUP program. Distances alongthe horizontal axis is proportional to the differences betweensequences, the scale bar indicate 10% of differences. The virusnames are as in Fig. 2; PAM, Pampa virus.


viruses (a genus of Bunya6iridae). Mills (1996),reported that each species of virus infects a uniquespecies of rodent and that the virus phylogenyappears to mirror their host taxonomy. This resultis consistent with the hypothesis of coevolution ofhantavirus and their rodent hosts, associated witha low rate of interspecies transmission of viruses.

Since our data indicate that Oliveros andPampa viruses are related by sequence (nucleotidesequence identity, 84%; amino acid identity,93.7%; amino acid similarity, 96.5%) it is notsurprising that they were isolated from the samegenus of rodents. However, the possibility thatthey were actually different viruses cannot beruled out. For arenaviruses, Bowen et al. (1996b)showed some evidence that supports the hypothe-sis of coevolution between arenaviruses and theirrodent hosts. They grouped the New World are-naviruses into three different lineages, named A,B and C, using a segment of the carboxy-terminalregion of the N gene. In particular, Pichinde,Parana and Flexal (members of the lineage A)have hosts of the genus Oryzomys, while Junınand Machupo (members of the lineage B) havehosts of the genus Calomys. However, they alsoreported that the same arenavirus could infectvarious species of rodents, i.e. Junın virus caninfect Calomys musculinus, Calomys laucha,Akodon azarae, Bolomys obscurus and Mus mus-culus. Also, the same species of rodent can beinfected by two different arenaviruses, i.e.Machupo and Latino viruses can infect Calomyscallosus. These facts would be consistent with thecoevolution hypothesis for arenaviruses and theirhosts, provided that a high rate of interspeciestransmission of virus could also take place.

In an attempt to establish general rules todefine viral species Ward (1993) proposed thatspecies of the same genus must possess the samegene set and should show a moderate sequenceidentity between most of their genes (35–85%).Moreover, it was proposed for viral strains thatthe sequence identity of all coding regions shouldbe relatively high (85–90%). Furthermore, in theclassification of the picornavirus family, Rueckert(1990), has shown that reported serotypes(strains) exhibit more than 85% nucleotide se-quence identity (for the P1 region). According to

the same author, species and genera are found toexhibit 65–85% and 45–65% nucleotide sequenceidentity, respectively.

If the current taxonomy of the Arena6iridae, asobtained by nucleotide sequence comparison,were analyzed using the same rules that were usedfor other virus families, the nucleotide sequenceidentity between Pampa and Oliveros virus(84.6%), would fall near the proposed cut offfigure (85%) between species and strains. A clearerdefinition of the taxonomic relationship of Pampaand Oliveros, will have to await more comprehen-sive nucleotide sequence data. Indeed, if we usedthe proposed cut off percentage between speciesand strains for Arena6iridae, the relationship be-tween the Nigeria and Josiah strains of Lassavirus should be reevaluated. The nucleotide se-quence identity between these two strains (73.9%)was similar to that obtained between related butwell defined viral species, i.e. Machupo and JunınMC2 virus exhibited a 73.9% of nucleotide se-quence homology if the same region is considered.

Finally, to our knowledge this paper is the firstdescription of a conserved primer set to RT-PCRamplify fragments of the S RNA from all of themembers of the family Arena6iridae. The use ofthis set of primers in RT-PCR would help re-searchers to detect already known as well as yetundescribed arenaviruses in field and/or clinicalsamples and the RFLP assay would allow theirpreliminary characterization.

Acknowledgements

This work was supported by grants from Comi-sion de Investigaciones Cientıficas de la Provinciade Buenos Aires (CIC BA) and Universidad Na-cional de Quilmes. DP has a fellowship from CICBA and VR is a research career member of theConsejo Nacional de Investigaciones Cientıficas yTecnicas (CONICET), Argentina.

References

Auperin, D.D., Compans, R.W. and Bishop, D.H.L. (1982)Nucleotide sequence conservation at the 3% termini of the


virion RNA species of New World and Old World are-naviruses. Virology 121, 200–203.

Auperin, D.D., Romanowski, V., Galinski, M.S. and Bishop,D.H.L. (1984) Sequencing studies of Pichinde arenavirus SRNA indicate a novel coding strategy, an ambisense viralS RNA. J. Virol. 52, 987–904.

Auperin, D.D. and McCormick, J.B. (1989) Nucleotide se-quence of the Lassa virus (Josiah strain) S genome RNAand amino acid sequence comparison of the N and GPCproteins to other arenaviruses. Virology 168, 421–425.

Bishop, D.H.L. (1990) Arenaviruses and their replication. In:B.N. Fields, D.M. Knipe et al. (Eds) Virology, 2nd edition.Raven Press, New York, pp. 1231–1243.

Bowen, M.D., Peters, C.J., Mills, J.M. and Nichol, S.T.(1996a) Oliveros virus: a novel arenavirus from Argentina.Virology 217, 362–366.

Bowen, M.D., Peters, C.J. and Nichol, S. (1996b) The phy-logeny of New World (Tacaribe complex) arenaviruses.Virology 219, 285–290.

Calderon, G. (1996) Aislamiento, distribucion y caracterısticasde nuevos Arenavirus en la Argentina. Presented at thesymposia of Diversos Arenavirus de Argentina. V Con-greso Argentino de Virologıa. Tandil, Argentina.

Clegg, J.C.S., Wilson, S.M. and Oram, J.D. (1990) Nucleotidesequence of the S RNA of Lassa virus (Nigerian strain)and comparative analysis of arenavirus gene products.Virus Res. 18, 151–164.

Clegg, J.C.S. (1993) Molecular phylogeny of the arenavirusesand guide to published sequence data. In: M.S. Salvato(Ed) The Arenaviridae, In: R. Wagner and H. FrankelConrat (Ed.), Serie: The Viruses. Plenum, New York, pp.175–187.

Chomczynski, P. and Sacchi, N. (1987) Single step method ofRNA isolation by acid guanidinium thiocyanate-phenol-chlorophorm extraction. Anal. Biochem. 162, 156–159.

Franze Fernandez, M.T., Zetina, C., Iapalucci, S., Lucero,M.A., Boissou, C., Lopez, R., Rey, O., Daheli, M., Cohen,G. and Zakin, M. (1987) Molecular structure and earlyevents in the replication of Tacaribe Arenavirus S RNA.Virus Res. 7, 309–324.

Fulhorst, C.F., Bowen, M.D., Ksiazek, T.G., Rollin, P.E.,Nichol, S.T., Kosoy, M.Y. and Peters, C.J. (1996) Isola-tion and characterization of Whitewater Arroyo virus, anovel North American Arenavirus. Virology, 224, 114–120.

Ghiringhelli, P.D., Rivera-Pomar, R.V., Lozano, M.E., Grau,O. and Romanowski, V. (1991) Molecular organization ofJunın virus S RNA: complete nucleotide sequence, rela-tionship with the other members of Arenaviridae andunusual secondary structures. J. Gen. Virol. 72, 2129–2141.

Gonzalez, J.P., Bowen, M.D., Nichol, S.T. and Rico-Hesse, R.(1996) Genetic characterization and phylogeny of Sabiavirus, an emergent pathogen in Brazil. Virology 221, 318–324.

Griffiths, C.M., Wilson, S.M. and Clegg, J.C.S. (1992) Se-quence of the nucleocapsid protein gene of Machupo virus:

close relationship with another South American pathogenicarenavirus, Junın. Arch. Virol. 124, 371–374.

Iapalucci, S., Lopez, R., Rey, O., Lopez, N., Franze-Fernan-dez, M.T., Cohen, G.N., Lucero, M., Ochoa, A. andZakin, M.M. (1989a) Tacaribe virus L gene encodes aprotein of 2210 amino acid residues. Virology 170, 40–47.

Iapalucci, S., Lopez, N., Rey, O., Zakin, M.M., Cohen, G.N.and Franze-Fernandez, M.T. (1989b) The 5% region ofTacaribe virus L RNA encodes a protein with a potentialmetal binding domain. Virology 173, 357–361.

Lisieux, T. Coimbra, T.L.M., Nassar, E.S. Burattini, M.N., deSouza, L.T.M., Ferreira, I.B., Rocco, I.M., Travassos daRosa, A.P.A., Vasconcelos, P.F.C., Pinheiro, F.P., LeDuc,J.W., Rico-Hesse, R., Gonzalez, J.P., Jahrling, P.B. andTesh, R.B. (1994) New arenavirus isolated in Brazil.Lancet 343, 391–392.

Lozano, M.E., Grau, O. and Romanowski, V. (1993a) Awhole blood RNA isolation method for reliable RT-PCRamplifications. Trends Genet. 9, 296.

Lozano, M.E., Ghiringhelli, P.D., Romanowski, V. and GrauO. (1993b) Nucleic acids amplification assay for the earlyand rapid detection of Junın virus in whole blood samples.Virus Res. 27, 37–53.

McCormick, J.B. (1990) Arenaviruses. In: B.N. Fields, D.M.Knipe, et al. (Eds) Virology, 2nd edition. Raven Press,New York, pp. 1245–1267.

Mills, J. (1996) Relaciones evolutivas entre los Arenavirus ysus reservorios. Presented at the symposia of DiversosArenavirus de Argentina. V Congreso Argentino de Vi-rologıa. Tandil, Argentina.

Romanowski, V., Matsuura, Y. and Bishop, D.H.L. (1985)Complete sequence of the S RNA of lymphocytic chori-omeningitis virus (WE strain) compared to that of Pichindearenavirus. Virus Res. 3, 101–114.

Romanowski, V. (1993) Genetic organization of Junın virus,the etiological agent of argentine hemorrhagic fever. In:M.S. Salvato (Ed), The Arenaviridae, In: R. Wagner andH. Frankel Conrat (Eds), Serie: The Viruses. Plenum, NewYork, pp. 51–83.

Rueckert R.R. (1990) Picornaviridae and their replication. In:B.N. Fields, D.M. Knipe et al. (Eds), Virology, 2nd edi-tion. Raven Press, New York.

Rustici, S.M. (1984) Desarrollo in vitro del virus Junın,proteınas intracelulares. Ph.D. Thesis. Faculty of Sciences,National University of La Plata, Argentina.

Salas, R., Manzione, N., Tesh, R.B., Rico-Hesse, R., Shope,R., Betancourt, A., Godoy, O., Bruzual, R., Pacheco, M.,Ramos, B., Taibo, M.E., Garcıa Tamayo, J., Jaimes, E.,Vasquez, C., Araoz, F. and Quarales, J. (1991) Venezuelanhemorrhagic fever. Lancet 338, 1033–1036.

Salvato, M., Shimomaye, E. and Oldstone, M.B.A. (1989) Theprimary structure of the lymphocytic choriomeningitisvirus L gene encodes a putative RNA polymerase. Virol-ogy 169, 377–384.

Salvato, M.S. and Shimomaye, E.M. (1989) The completesequence of lymphocytic choriomeningitis virus reveals aunique RNA structure and a gene for a zinc finger protein.Virology 173, 1–10.


Salvato, M.S. (1993) Molecular biology of the prototype are-navirus, Lymphocytic choriomeningitis virus. In: M.S. Sal-vato (Ed.), The Arenaviridae, In: R. Wagner and H.Frankel Conrat (Eds.), Serie: The Viruses. Plenum, NewYork, pp. 133–156.

Sanger, F., Nicklen, S and Coulson, A.R. (1977) DNA se-quencing with chain terminating inhibitors. Proc. Natl.

Acad Sci. 74, 5463–5466.Ward, C.W. (1993) Progress towards a higher taxonomy of

viruses. Res. Virol. 144, 419–453.Wilson, S.M. and Clegg, J.C.S. (1991) Sequence analysis of the

S RNA of the African Arenavirus Mopeia: an unusualsecondary structure feature in the intergenic region. Virol-ogy 180, 543–552.

. .

Characterization of arenaviruses using a family-specific primer set for RT-PCR amplification and...

Documents

Transcript of Characterization of arenaviruses using a family-specific primer set for RT-PCR amplification and...