Phylogenetic analysis of Portuguese Feline Immunodeficiency

Virus sequences reveals high genetic diversity

Ana Duarte *, Luis Tavares

Centro de Investigacao Interdisciplinar em Sanidade Animal (CIISA), Faculdade de Medicina Veterinaria,

Avenida da Universidade Tecnica, 1300-477 Lisboa, Portugal

Received 27 July 2005; received in revised form 26 October 2005; accepted 15 November 2005

Abstract

Feline Immunodeficiency Virus (FIV) is a Lentivirus responsible for an immunodeficiency like disease in domestic cats.

Based on the genetic diversity of the V3–V5 region of env gene FIV is divided in five phylogenetic subtypes (A, B, C, D and E)

with a world-wide distribution. To understand the subtype diversity of FIV in Portugal a serological survey was conducted during

1 year in the Veterinary Faculty Hospital, Lisbon, Portugal to identify seropositive animals. Two viral genomic regions were

amplified by a nested PCR, sequenced and the phylogenetic relationships between 24 new Portuguese FIV sequences and other

previously published FIV isolates were assessed. The introduction of these sequences induced a subclustering in subtype B

including most of the new Portuguese sequences. Moreover, a new cluster emerged, with two highly divergent new sequences

that might represent a new subtype. The study of these new FIV isolates showed the presence in Portugal of a unique viral

population subclustering within subtype B and of sequences clearly divergent from the five known subtypes, providing a

contribution for the understanding of FIV’s genetic diversity.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Feline Immunodeficiency Virus; Subtypes; Phylogenetic analysis

www.elsevier.com/locate/vetmic

Veterinary Microbiology 114 (2006) 25–33

1. Introduction

Feline Immunodeficiency Virus (FIV) is a Lenti-

virus of the Family Retroviridae that causes an

immunodeficiency like disease in domestic cats,

similar to AIDS in humans (Pedersen et al., 1987).

* Corresponding author. Tel.: +351 21 3652800;

fax: +351 21 3652882.

E-mail address: anaduarte@fmv.utl.pt (A. Duarte).

0378-1135/$ – see front matter # 2005 Elsevier B.V. All rights reserved

doi:10.1016/j.vetmic.2005.11.056

Due to the common biological characteristics with

HIV, FIV has been used as a valuable model to

understand HIV and other Lentivirus pathogenesis as

well as for the development of vaccine strategies.

Based on the genetic diversity of the V3–V5 region

of the env gene FIV is currently divided in five

phylogenetic subtypes (A, B, C, D and E) with a

world-wide distribution. Subtype A is found in the

USA, Australia and Europe (Greene et al., 1993;

Rigby et al., 1993; Sodora et al., 1994; Bachmann

et al., 1997), subtype B in Japan, Europe and USA

.

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–3326

(Kakinuma et al., 1995; Bachmann et al., 1997; Pistello

et al., 1997) and subtype C in Canada, Europe, Taiwan

and Vietnam (Sodora et al., 1994; Kakinuma et al.,

1995; Inada et al., 1997; Nakamura et al., 2003).

Subtypes D and E are found in Japan, Vietnam

(Kakinuma et al., 1995; Nakamura et al., 2003) and

Argentina (Pecoraro et al., 1996). Subtype A and B

include most of the known FIV isolates. Furthermore,

several recombinant sequences have been identified and

classified between subtypes A and B, subtypes B and D

and subtypes A and C (Reggeti and Bienzle, 2004).

A dual-subtype FIV vaccine including subtype A

and D is currently available in the United States.

However, due to the continuing evolutionary pattern of

FIV, the study of the molecular epidemiology of this

virus in different geographical areas is of utmost

importance to assess the introduction of such a vaccine

in other countries, and to establish diagnostic strategies

based on the detection of viral nucleic acids, influenced

by the genetic diversity of local subtypes.

Five FIV sequences of Portuguese isolates have

been previously reported, grouping in a subcluster

within subtype B (Duarte et al., 2002). In this study,

we present 24 new sequences of Portuguese FIV

isolates, and propose the division of subtype B in four

different clusters with high values of genetic

divergence between them. The phylogenetic analysis

of our Portuguese sequences also revealed a new

cluster including two sequences that might represent

an additional sixth subtype.

2. Materials and methods

2.1. Serological survey

Seropositive cats were identified by immunoblot-

ting in a serological survey conducted during 1 year in

the Veterinary Faculty Hospital of Lisbon, Portugal.

FIV Petaluma strain was produced in FL4 cells.

The virus was precipitated in the presence of PEG

6000, and purified by ultracentrifugation. The viral

proteins were separated by SDS-PAGE in a 12% gel

and transferred to a nitrocellulose membrane. Animals

that reacted against the capsid protein (p24) were

considered positives.

Information regarding the 24 samples obtained is

summarized in Table 1.

2.2. gag and env PCR

Genomic DNA was extracted from whole blood

(100–1000 ml) (Wizard Genomic DNA Purification

Kit Promega), eluted in 50–100 ml of H2O and stored

at �20 8C.

A 329 bp region located in the CA amino terminus

in the gag gene (Hohdatsu et al., 1998; Kurosawa

et al., 1999) and a 550 bp fragment including the V3–

V5 region of the env gene (Sodora et al., 1994;

Bachmann et al., 1997) were amplified from the

genomic DNA by nested PCR.

Primer sequences and PCR cycling conditions are

summarized in Table 2.

Primers GagO1 and GagO2 were used in the first

round amplification of gag gene. The second round

was performed with primers GagI1 and GagI2

yielding a product of 329 bp.

Primers Env1/Env2 and Hdfor/Hdrev were used to

amplify the V3–V5 region of env gene. The second

round primers yield a product of 550 bp (Bachmann

et al., 1997). In the second round PCR the extension

period during the 30 cycles was 72 8C for 1 min.

The PCR reactions were performed in 50 ml total

volume with FideliTaq PCR Master Mix (2�)

(USB1), 100 pmol of each primer and 100 ng of

genomic DNA.

The 329 bp gag amplicons were purified and

directly cloned into p-Gem T easy vector (Promega)

according to the manufacturer instructions.

2.3. Heteroduplex mobility assay (HMA)

In order to confirm its genetic homogeneity the env

550 bp fragment was subjected to a HMA, before

cloning (Delwart et al., 1993; Bachmann et al., 1997).

Briefly 10 ml of env second round amplicons and 1 ml

of 10� annealing buffer (1 M NaCl, 100 mM Tris pH

7.8, 20 mM EDTA) were combined. The mixture was

heated to 94 8C for 5 mn, rapidly cooled on ice, 1 ml of

loading buffer (50% glycerol, 0.01% of xyleno

cyanol) was added and loaded in a 8% polyacrylamide

gel (29.2:0.8, acrylamide:bisacrylamide) poured

into plates of 15 cm � 15 cm. The electrophoresis

was carried at a constant 150 V for 5 h in a vertical

gel apparatus (Hoefer Scientific Instruments) in 1�TBE (88 mM Tris–borate, 89 mM borate, 2 mM

EDTA).

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–33 27

Table 1

Sample identification, origin and subtype assignment in the gag and env partial genes

Identification Origin gag env

Subtype Accession number Subtype Accession number

355_02CaP Canecas Subcluster B DQ072532 Subcluster B DQ072556

23_02LisP Lisbon Subtype B DQ072535 Subtype B DQ072559

43_02LisP Lisbon Subtype A DQ072536 Subtype A DQ072560

82_02LisP Lisbon Subtype A DQ072539 Subtype A DQ072563

120_02LisP Lisbon Subcluster B DQ072540 Subcluster B DQ072564

135_02LisP Lisbon Subcluster B DQ072541 Subcluster B DQ072565

150_02LisP Lisbon Subcluster B DQ072542 Subtype A DQ072566

151_02LisP Lisbon Putative F DQ072543 Putative F DQ072567

194_02LisP Lisbon Subcluster B DQ072548 Putative F DQ072572

206_02LisP Lisbon Subcluster B DQ072549 Subcluster B DQ072573

235_02LisP Lisbon Subcluster B DQ072551 Subtype A DQ072575

268_02LisP Lisbon Subcluster B DQ072552 Subcluster B DQ072576

300_02LisP Lisbon Subcluster B DQ072554 Subtype A DQ072576

350_02LisP Lisbon Subtype A DQ072555 Subcluster B DQ072579

164_02UZP Cattery Subtype B DQ072544 Subtype A DQ072568

165_02UZP Cattery Subcluster B DQ072545 Subtype B DQ072569

190_02UZP Cattery Subcluster B DQ072546 Subcluster B DQ072570

192_02UZP Cattery Putative F DQ072547 Subcluster B DQ072571

217_02UZP Cattery Subtype A DQ072550 Subtype A DQ072574

4_02MonP Montijo Subcluster B DQ072533 Subtype B DQ072557

62_02MonP Montijo Subcluster B DQ072537 Subcluster B DQ072561

63_02MonP Montijo Subtype A DQ072538 Subtype A DQ072562

14_02PalP Palmela Subcluster B DQ072534 Subcluster B DQ072558

296_02CaP Canecas Subcluster B DQ072553 Subcluster B DQ072576

The Accession number of each nucleotide sequence is indicated.

2.4. Phylogenetic analysis

The nucleotide sequences representing each geno-

mic region were sequenced and aligned with

previously characterized subtyped sequences of FIV

isolates, retrieved from their EMBL/Genbank/DDBJ

Table 2

Primers used for the amplification of the CA amino terminus in the gag gen

(Bachmann et al., 1997)

Primer Sequence (50–30)

PCR cycling conditions: 95 8C/2 min; 35 cycles of 94 8C/1 min, 50 8C/45

GagO1 AATATGACTGTACTA

GagO2 TTTTCTTCTAGAGTA

GagI1 TATTCAAACAGTAAA

GagI2 CTGCTTGTTGTTCTT

PCR cycling conditions: 5 cycles of 94 8C/1 min, 50 8C/1 min, 72 8C/2 m

72 8C/5 min

Env1 GCTCAGGTAGTATGG

Env2 ACTTCATCATTCCTC

Hdfor ATACCAAAATGTGGA

Hdrev CAAGACCAATTTCCA

aAll the nucleotide positions are in reference to FIV Petaluma (Accessio

accession numbers. The alignments were created

using ClustalW and manually corrected to maximise

genetic similarities.

The phylogenetic relationship between sequences

of each genomic region was inferred by two methods.

Genetic distance between pairs of sequences was

e (Hohdatsu et al., 1998) and the V3–V5 region within the env gene

Location (nt)a

s, 72 8C/1 min; 72 8C/5 min

CTGC 917–936

CTTTCTGG 1650–1628

TGGAG 1036–1055

GAGTT 1364–1345

in; 72 8C/5 min; 30 cycles of 94 8C/15 s, 55 8C/45 s, 72 8C/2 min;

AGACT 6788–6807

CTCTT 8836–8817

TGGTG 7316–7335

GCAAT 7866–7847

n number M25381).

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–3328

Fig. 1. Unrooted phylogenetic trees from the 329 nt gag alignment. (A) Neighbour joining tree using Kimura’s two parameter model with a

transition/transversion of 2.07 estimated from the data set. Each scale bar represents 0.10 of genetic distance. Bootstrap values are shown at the

branch points. (B) Maximum likelihood tree (quartet puzzling using the substitution model HKY85). Genbank Accession numbers: FIV Wo

(L06136), PPR (M36968), Petaluma (M25381), Sendai 1 (D37820); Sendai 2 (D37821), Aomori 2 (D37824), Aomori 1 (D37823), Yokohama

(D37819), Usil2489_7B (U11820), Italy M2 (Y13866), Italy M3 (Y13867), TI1 (AB027298), TI4 (AB027301), Fukuoka (D37818), Shizuoka

(D37822), Lp3 (AB027302), Lp20 (AB027303), Lp24 (AB027304), RP! (AJ304955), TXTG (AY139111), TX132 (AY139112), TX200

(AY139110).

calculated by the Kimura’s Two Parameter Method

(DNADIST, Phylip Package) (Kimura, 1980; Felsen-

stein, 1993) using values of transition/transversion and

nucleotide frequencies calculated from the data sets

(TREE-PUZZLE, Strimmer and von Haeseler, 1996).

Maximum likelihood was performed with the sub-

stitution model of HKY85, assuming two rates of

variability (one variable and one invariable) along the

alignment (TREE-PUZZLE). The trees were con-

structed by neighbour joining (Saitou and Nei, 1987)

and the branching order reliability was evaluated by

bootstrap analysis of 1000 replicates (Felsenstein,

1985).

3. Results

3.1. Phylogenetic analysis of partial gag

sequences

Genetic distances and maximum likelihood were

used to assess the phylogenetic relationships of 24 new

Portuguese FIV sequences comprehending 329 nt of

gag gene. Both methods showed the same distribution

pattern for this genomic region (Fig. 1A and B).

The trees revealed six main clusters, supported by

reliable bootstrap values. Visual inspection suggested

that cluster B is divided in three branches, one of them

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–33 29

including the Argentine isolates (Lp3, Lp20 and Lp24)

previously classified as subtype E. Regarding the

distribution of the new Portuguese sequences, fifteen

grouped together with RP1, a previously reported

sequence, in a second branch of subtype B (Fig. 1A and

B). The third branch includes prototype B sequences

and two Portuguese sequences 164_01UZP and

23_02LisP. Sequence 23_02LisP although included

in subtype B, is positioned outside the main B branch,

but without significant bootstrap value.

Five Portuguese isolates were included within

subtype A (350_02LisP, 82_02LisP, 63_02MonP,

43_02LisP, 217_02UZP). None of the Portuguese

sequences analysed clustered with subtypes C and E.

Sequences 192_02UZP and 151_02LisP, grouped

together in an individual branch, supported by a

bootstrap value of 99.9%.

No individual clustering of sequences obtained

from animals of the same household or in the same

area was evident (Table 1).

To investigate the reliability of the phylogenetic

information obtained through the analysis of the

329 nt alignment, likelihood mapping was performed

using TREE-PUZZLE, indicating that 89.8% of the

analysed quartets were fully resolved assuring there-

fore its utility to assess phylogenetic relationships

between viral sequences.

Furthermore, this region was considered useful for

rapid subtyping by a RFLP assay (Hohdatsu et al.,

1998; Kurosawa et al., 1999; Duarte et al., 2002).

Based on the nucleotide sequence we confirmed the

presence of targets for six restriction endonucleases,

previously described (Table 3).

Sequence 151_02LisP revealed an extra target for

PstI endonuclease (229 bp), and sequence 192_02UZP

had an additional PvuII target (287 bp).

Table 3

Nucleotide position of restriction endonucleases sites in the 329 bp

fragment included in the gag gene, used for rapid subtype assignment

(Hohdatsu et al., 1998; Kurosawa et al., 1999; Duarte et al., 2002)

BamHI HincII PstI PvuII TaqI XbaI

A 187 bp,

286 bp

B 227 bp

B-Portuguese 222 bp 286 bp 287 bp

C 286 bp 287 bp 41 bp

D 44 bp 286 bp 240 bp

E 286 bp 287 bp 41 bp

Sequence 4_02MonP included in the Portuguese

subcluster did not share the same restriction pattern. A

unique HincII target (308 bp) positioned in a different

site was observed. Sequence 164_02UZP revealed

subtype B pattern plus an additional XbaI target

(200 bp). The unique restriction pattern observed was

in accordance with their positioning in the gag tree.

The restriction pattern of the other sequences

showed no divergence from the subtype assignment in

the phylogenetic gag tree.

3.2. Phylogenetic analysis of env V3–V5 region

Regarding the V3–V5 region of the env gene, two

phylogenetic trees were constructed by neighbour

joining, using genetic distances and maximum like-

lihood. Both trees showed the same rearrangement

observed in the gag tree (Fig. 2A and B), however

subtype A was divided in two subclusters and subtype

B in four subclusters with high bootstrap values to

support them. Additionally, several Portuguese

sequences change their positioning in the env tree.

The following sequences included in the Portu-

guese B subcluster in the gag tree behaved differently.

Sequences 235_02LisP, 300_02LisP and 150_02LisP

were included in subtype A. Sequence 165_02UZP

grouped in subtype B and sequence 194_02LisP

grouped in an individual branch together with

151_02LisP. Sequence 150_02LisP moved to a branch

within subtype A in the env tree.

Sequence 164_02UZP classified as subtype B in

the gag tree, was located in subtype A. Sequence

350_02LisP was included in subtype A in the gag tree

and moved to the Portuguese B subcluster in the env

tree. Sequence 192_02UZP clustered with sequence

151_02LisP in the gag tree also moved to Portuguese

B subcluster.

Moreover a subclustering of subtype A was

observed, including sequences 150_02LisP and

164_01UZP grouping in a single branch with high

reliability (100% of bootstrap).

Three main clusters with 100–95% of bootstrap

emerged in parallel with subtype B, assuming similar

values of genetic distance between them, namely

subtype E, three Texas isolates (Weaver et al., 2004)

and sequences 4_02MonP and 23_02LisP.

Sequence 151_02LisP group with 194_02LisP in a

unique branch also observed in the gag tree.

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–3330

Fig. 2. Unrooted phylogenetic trees from the 550 nt env alignment. (A) Neighbour joining tree using Kimura’s two parameter model with a

transition/transversion of 2.16 estimated from the data set. (B) Maximum likelihood tree (quartet puzzling using the substitution model HKY85).

Genbank Accession numbers: FIV Wo (L06312), PPR (M36968), Petaluma (M25381), Sendai 1 (D37813); Sendai 2 (D37814), Aomori 2

(D37817), Aomori 1 (D37816), Yokohama (D37812), Usil2489_7B (U11820), Italy M2 (X69501), Italy M3 (X69502), TI1 (AB016025), TI4

(AB016028), Fukuoka (D37815), Shizuoka (D37811), Lp3 (D84496), Lp20 (D84498), Lp24 (D84500), RP1 (AJ304986), TXTG (AY139101),

TX132 (AY138099), TX200 (AY139096).

Due to the discordance in the subtype assignment in

gag and env trees, observed for several Portuguese

sequences, a bootscanning was performed. SIMPLOT

program was used in order to test for possible

recombination points along the alignment (Lole

et al., 1999). A window of 100 nt with 40 nt overlaps

was used to calculate similarity values, using maximum

likelihood with 100 replicates. Petaluma (subtype A),

Usil2489_7B (subtype B), TI1 (subtype C), Fukuoka

(subtype D) and Lp3 (subtype E) were used as reference

subtypes and each discordant sequence was tested

individually (235_02LisP, 300_02LisP, 150_02LisP,

165_02UZP, 194_02Lis, 192_02UZP, 164_02UZP,

4_02MonP and 350_02LisP). However, no recombi-

nant sequences were detected (data not shown).

To further clarify the subtype assignment of

different sequences in the env tree, genetic distance

between and among subtypes was calculated

by maximum likelihood (Table 4) (Sodora et al.,

1994).

The genetic distance within subtypes range from

0.034 to 0.10 (Table 4). The minimal value between

subtypes did not exceed 0.208, but between sub-

clusters of subtype B, higher values were observed

(Table 4). The observed divergences are in accordance

with previously reported data (Sodora et al., 1994;

Bachmann et al., 1997). The genetic distance between

sequences 194_02LisP/151_02LisP and the major

subtypes was always superior to 0.24.

The value observed between sequences

150_02LisP/164_01UZP and subtypes A, B and the

Portuguese main group was unexpected, due to the

sequences positioning in the env tree (Fig. 2A and B).

The visual inspection and the bootstrap value

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–33 31

Fig. 3. Neighbour joining tree using Dayhoff PAM001 matrix based

on the predicted amino-acid sequences of FIV isolates.

Tab

le4

Aver

age

gen

etic

div

erg

ence

bet

wee

nan

dam

on

gsu

bty

pes

inth

een

vg

ene

Su

bty

pe

AB

Su

bB

CD

EF

A0

.10

(0.1

4–0

.02)

0.2

7(0

.35

–0

.19)

0.3

0(0

.35

–0

.26)

0.2

8(0

.32

–0

.25)

0.3

1(0

.28

–0

.35)

0.2

6(0

.30

–0

.23)

0.2

7(0

33

–0

.25

)

B0

.09

(0.1

7–0

.05)

0.1

6(0

.24

–0

.13)

0.2

1(0

.26

–0

.12)

0.2

6(0

.31

–0

.23)

0.2

1(0

.26

–0

.12)

0.2

6(0

.29

–0

.22)

Su

bB

0.0

6(0

.09

–0

.004

)0

.19

(0.2

0–

0.1

6)

0.2

7(0

.29

–0

.25)

0.2

1(0

.22

–0

.19)

0.2

7(0

.29

–0

.24)

E0

.05

(0.0

6–

0.0

4)

0.3

0(0

.31

–0

.28)

0.2

1(0

.22

–0

.20)

0.2

6(0

.28

–0

.25)

C0

.03

0.2

6(0

.28

–0

.25)

0.2

8(0

.28

–0

.27)

D0

.06

0.2

4(0

.25

–0

.22)

F0

.21

15

0_

02

Lis

P/1

64

_0

1U

ZP

0.2

3(0

.31

–0

.09)

0.2

1(0

.26

–0

.18)

0.1

4(0

.16

–0

.13)

0.2

9(0

.29

–0

.28)

0.2

4(0

.25

–0

.23)

0.2

2(0

.23

–0

.20)

0.2

9(0

.32

–0

.26)

The

max

imum

and

min

imum

val

ues

are

indic

ated

inea

chce

ll.

suggested an additional branch in subtype A but this

was not supported by genetic distance.

To clarify this observation one additional tree was

calculated (Fig. 3) based on the alignment (ClustalW)

of the predicted amino-acid sequence of the isolates,

using Dayhoff PAM001 matrix of distances. The

sequences distribution were similar to the one

obtained with the nucleotide alignment. Nevertheless,

the branch including sequences 150_02LisP and

164_01UZP emerged outside subtype A much closer

to the B Portuguese subcluster. This observation is in

accordance with the genetic relationship between

these sequences. Sequences 194_02LisP and

151_02LisP maintained the individual clustering,

although with lower bootstrap consistency.

4. Discussion

In the last few years, an effort has been made to

understand the molecular epidemiology of FIV, due to

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–3332

the implications of viral genetic diversity on vaccine

development and diagnosis of FIV infection (Sodora

et al., 1994; Bachmann et al., 1997; Steinrigl and

Klein, 2004; Reggeti and Bienzle, 2004).

FIV has been divided in five subtypes (A, B, C, D

and E) based on the phylogenetic analysis of two main

genomic regions, the V3–V5 regions of the SU

glycoprotein in the env gene and CA amino terminus

in the gag gene (Rigby et al., 1993; Sodora et al., 1994;

Bachmann et al., 1997; Pistello et al., 1997; Hohdatsu

et al., 1998).

The V3–V5 env region is responsible for cell

tropism, cytopathogenity and effective host immuno

response (Vahlenkamp et al., 1999). The observed

nucleotide variability of different isolates strongly

suggests a positive evolutionary pressure in this region

(Sodora et al., 1994; Steinrigl and Klein, 2004) and the

presence of recombination hot spots. Through the

analysis of the variation pattern within gag gene the

region coding the CA amino terminus was considered

useful for the genetic subtyping of FIV (Rigby et al.,

1993; Pistello et al., 1997). This region was also tested

for rapid subtyping by a RFLP assay (Cammarota

et al., 1996; Hohdatsu et al., 1998).

Based on two genomic regions, we analysed 24 new

Portuguese FIV sequences and found a higher viral

diversity with the introduction of these sequences. Most

of them grouped within a subcluster of subtype B along

with previously reported Portuguese sequences (Duarte

et al., 2002). We also found Portuguese sequences

belonging to subtype A and B, confirming the presence

of these subtypes in Europe reported by other authors

(Rigby et al., 1993; Pistello et al., 1997; Bachmann

et al., 1997; Steinrigl and Klein, 2004).

In several sequences the subtype assignment

differed with the genomic region used for phyloge-

netic analysis. In the V3–V5 region, all amplicons

were cloned after confirming the absence of viral

quasispecies within the same animal by HMA. No

evidence of recombination was observed between the

suspected sequences in the same genomic region. This

rearrangement may be due to the occurrence of

recombinant events between different viral genes.

However, to clarify this statement, the analysis of full

length genes or full FIV genomes would be necessary.

In both genomic regions we also described two

sequences (151_02LisP and 194_02LisP/192_UZP)

grouping in an individual cluster supported by high

bootstrap values and by different phylogenetic

methodologies. Sequence 151_02LisP also revealed

a unique restriction pattern in the partial gag region

studied. Considering the proposed use of this region

for rapid subtyping by RFLP of new FIV sequences,

this finding was consistent with the individual

branching observed in the gag tree (Cammarota

et al., 1996; Hohdatsu et al., 1998). The assignment of

a new subtype should be supported by a higher number

of isolates and by the analysis of a larger genomic

region, similarly to what has been done for HIV.

Nevertheless, we propose that sequence 151_02LisP

might represent the prototype of a new subtype termed

subtype F. Sequences 194_02LisP and 192_02UZP

with different subtype assignment in both genomic

regions may be representatives of mosaic virus

between the Portuguese B subcluster and the putative

new subtype, present in Portugal. Sequence

151_02LisP 194_02LisP, and 192_02UZP have no

epidemiological linkage between them. However, we

found high subtype diversity in the same geographical

area (Table 1). The presence of different circulating

subtypes within the same population might contribute

to recombinant events.

The overall conclusion of this study is the

increasing viral diversity within and between subtypes

and the need for the assignment of subclusters within

subtype B, represented by the main Portuguese

sequences and the Texas isolates, already proposed

by Weaver et al. (2004). The genetic heterogeneity

within subtype B, stated by other authors (Sodora

et al., 1994; Bachmann et al., 1997), may justify its

future division in subtypes or sub groups according to

genetic divergence and branch reliability.

The observation of a restricted FIV group in Lisbon

(Portugal) enhances the need for a more comprehen-

sive epidemiological survey in different areas. This

information would contribute for the correct evalua-

tion of a FIV vaccine in Portugal and development of

new diagnostic methodologies.

Acknowledgments

This work was sponsored by the Research

Interdisciplinary Center in Animal Health, Veterinary

Faculty of Lisbon. FL4 cells were gently provided by

Dr. Janet Yamamoto.

A. Duarte, L. Tavares / Veterinary Microbiology 114 (2006) 25–33 33

We are grateful to our colleagues from the Hospital

of the Veterinary Faculty of Lisbon for their

collaboration in biologic sample collections. The

availability of the Bioinformatics programs was made

possible through the Portuguese EmbNet Node (Pen)

at the Gulbenkian Institute of Science, Portugal. We

would also like to thank Dr. Isabel Marques (Science

Gulbenkian Institute) for the revision of this work.

References

Bachmann, M.H., Mathiason-Rubard, C., Learn, G.H., Rodrigo,

A.G., Sodora, D.L., Mazzetti, P., Hoover, E.A., Mullins, J.I.,

1997. Genetic diversity of Feline Immunodeficiency Virus: dual

infection, recombination, and distinct evolutionary rates among

envelope sequence clades. J. Virol. 71, 4241–4253.

Cammarota, G., Da Prato, L., Nicolleti, E., Matteucci, D., Bend-

inelli, M., Pistello, M., 1996. Quantitation of feline immunode-

ficiency proviruses in doubly infected cats using competitive

PCR and a fluorescence-based RFLP. J. Virol. Methods 62, 21–

31.

Delwart, E.L., Shpaer, E.G., Louwagie, J., McCutchan, F.E., Grez,

M., Rubsamen-Waigmann, H., Mullins, J.I., 1993. Genetic

relationships determined by a DNA heteroduplex mobility

assay: analysis of HIV env genes. Science 262, 1257–1261.

Duarte, A., Marques, M.I., Tavares, L., Fevereiro, M., 2002. Phy-

logenetic analysis of five Portuguese strains of FIV. Arch. Virol.

147, 1061–1070.

Felsenstein, J., 1985. Confidence limits on phylogenies. An

approach using the bootstrap. Evolution 39, 783–791.

Felsenstein, J., 1993. PHYLIP. Phylogeny Inference Package (Ver-

sion 3.5) University of Washington, Seattle.

Greene, W.K., Meers, Fierro, G., Carnegie, P.R., Robinson, W.F.,

1993. Extensive sequence variation of feline immunodeficiency

virus env genes in isolates from naturally infected cats. Arch.

Virol. 133, 51–62.

Hohdatsu, T., Motokawa, K., Usami, M., Amioka, M., Okada, S.,

Koyama, H., 1998. Genetic subtyping and epidemiological study

of feline immunodeficiency virus by nested polymerase chain

reaction-restriction fragment length polymorphism analysis of

the gag gene. J. Virol. Methods 70, 107–111.

Inada, G., Miyazawa, T., Inoshima, Y., Kohmoto, M., Ikeda, Y., Liu,

C.-H., Lin, J.A., Kuo, T.F., Mikami, T., 1997. Phylogenetic

analysis of feline immunodeficiency virus isolated from cats

in Taiwan. Arch. Virol. 142, 1459–1467.

Kakinuma, S., Motokawa, K., Hohdatsu, T., Yamamoto, J.K.,

Koyama, H., Hashimoto, H., 1995. Nucleotide sequence of

Feline Immunodeficiency Virus: classification of Japonese iso-

lates into two subtypes which are distinct from non-Japonese

subtypes. J. Virol. 69, 3639–3646.

Kimura, M., 1980. A simple model for estimating evolutionary rates

of base substitutions through comparative studies of nucleotide

sequences. J. Mol. E 16, 111–120.

Kurosawa, K., Ieda, Y., Miyazawa, T., Izumiya, Y., Nishimura, Y.,

Nakamura, K., Sato, E., Mikami, T., Kai, C., Takahashi, E.,

1999. Development of restriction fragment-length polymorph-

ism method to differentiate five subtypes of Feline Immunode-

ficiency Virus. Microbiol. Immunol. 43, 817–820.

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 human immunodeficiency virus type 1 gen-

omes from subtype C-infected seroconverters in India, with

evidence of intersubtype recombination. J. Virol. 73, 152–160.

Nakamura, K., Suzuki, Y., Ikeo, K., Ikeda, Y., Sato, E., Nguyen,

N.T.P., Gojobori, T., Mikami, T., Miyazawa, T., 2003. Phyloge-

netic analysis of Vietnamese isolates of feline immunodeficiency

virus: genetic diversity of subtype C. Arch. Virol. 148, 783–791.

Pecoraro, M.R., Tomonaga, K., Miyazawa, T., Kawaguchi, Y.,

Sujita, S., Tohya, Y., Kai, C., Etcheverrigaray, M.E., Mikami,

T., 1996. Genetic diversity of Argentine isolates of feline

immunodeficiency virus. J. Gen. Virol. 77, 2031–2035.

Pedersen, N.C., Ho, E.W., Brown, M.L., Yamamoto, J.K., 1987.

Isolation of a T-lymphotropic Virus from domestic cats with an

immunodeficiency-like syndrome. Science 235, 790–793.

Pistello, M., Cammarota, G., Nicoletti, E., Matteucci, D., Curcio,

M., Del Mauro, D., Bendinelli, M., 1997. Analysis of the

genetic diversity and phylogenetic relationship of Italian isolates

of feline immunodeficiency virus indicates a high prevalence

and heterogeneity of subtype B. J. Gen. Virol. 78, 2247–2257.

Rigby, M.A., Holmes, E.C., Pistello, M., Mackay, A., Leigh Brown,

A.J., Neil, J.C., 1993. Evolution of structural proteins of feline

immunodeficiency virus: molecular epidemiology and evidence

of selection for change. J. Gen. Virol. 74, 425–436.

Reggeti, F., Bienzle, D., 2004. Feline immunodeficiency virus

subtypes A, B and C and intersubtype recombinants in Ontario,

Canada. J. Gen. Virol. 85, 1843–1852.

Saitou, N., Nei, M., 1987. The neighbour-joining method. A new

method to reconstructing phylogenetic trees. Mol. Biol. E 4,

406–426.

Sodora, D.L., Shpaer, E.G., Kitchell, B.E., Dow, S.W., Hoover, E.A.,

Mullins, J.I., 1994. Identification of three of Feline Immunode-

ficiency Virus (FIV) env genes subtypes and comparison of FIV

and Human Immunodeficiency Virus Type 1 evolutionary pat-

terns. J. Virol. 68, 2230–2238.

Steinrigl, A., Klein, D., 2004. Phylogenetic analysis of feline

immunodeficiency virus in Central Europe: a prerequisite for

vaccination and molecular diagnostic. J. Gen. Virol. 84, 1301–

1307.

Strimmer, K., von Haeseler, A., 1996. Quartet puzzling: a quartet

maximum likelihood method for reconstructing tree topologies.

Mol. Biol. E 13, 964–969.

Vahlenkamp, T.W., De Ronde, A., Schuurman, N.N.M.P., van Vliet,

A.L.W., van Drunen, J., Horzinek, M.C., Egberink, H.F., 1999.

Envelope gene sequences encoding variable regions 3 and 4 are

involved in macrophage tropism of feline immunodeficiency

virus. J. Gen. Virol. 80, 2639–2646.

Weaver, E.A., Collisson, E.W., Slater, M., Zhu, G., 2004. Phylo-

genetic analysis of Texas isolates indicate an evolving subtype of

the clade B Feline Immunodeficiency Viruses. J. Virol. 78,

2158–2163.