Phylogenetic Analysis of Feline Immunodeficiency Virus in Feral and Companion Domestic Cats of New Zealand

Abstract

Nested PCR was used to amplify envelope V3-V6 gene fragments of feline immunodeficiency virus (FIV) from New Zealand cats. Phylogenetic analyses established that subtypes A and C predominate among New Zealand cats, with clear evidence of intersubtype recombination. In addition, 17 sequences were identified that were distinct from all known FIV clades, and we tentatively suggest these belong to a novel subtype.

Feline immunodeficiency virus (FIV) is a lentivirus that infects the domestic cat (Felis catus), causing progressive immunodeficiency analogous to AIDS in humans (21). Five distinct FIV subtypes have been identified, based on sequence diversity within the V3 to V5 region of the envelope gene (13, 20, 29). A recent study identified a distinct group of FIV isolates from Texas that possibly represent a new subtype (32). In addition, intersubtype recombination has been detected in natural populations (1, 4, 23).

The only previously published study on FIV in New Zealand cats examined the infection status of 250 domestic cats of different sex, age, breed, and health condition (30). In the present study, New Zealand cats were sampled from April 2003 to January 2006. Lymph nodes were dissected from 334 New Zealand feral cats and 28 New Zealand stray cats obtained from seven locations throughout both the North and South Islands of New Zealand. Feral cats are defined as unowned cats that inhabit rural areas, including wooded countryside and parkland, whereas stray cats inhabit urban areas. Blood samples from 48 FIV-symptomatic domestic New Zealand cats were also obtained from a local veterinary diagnostic facility, and a further 29 seropositive New Zealand domestic cat blood samples were sourced from veterinarians. Genomic DNA was extracted from lymph nodes by using QIAamp DNA Minikit (QIAGEN) and from blood samples by using QIAamp DNA blood minikit (QIAGEN).

Nested PCR was used to amplify 858 bp of the FIV envelope gene within the V3 to V6 region. Briefly, a 25-μl reaction containing 0.4 μM concentrations of each primer, 0.2 mM concentrations of each deoxynucleoside triphosphate, and 0.625 U of Platinum Taq DNA polymerase was used. In the first round, 2 μl of genomic DNA was added. In the second round, 1 μl from the first round tube was added. The reactions were run on a Biometra T1 thermal cycler. The first-round primers, VE1R and VE1S, amplified a 1,230-bp fragment containing the annealing sites for the second-round primers, VE2R and VE2S (17). Thermal cycling parameters used were as follows: 3 min at 94°C, followed by 5 cycles of 1 min at 94°C, 1 min at 57°C, and 2 min at 72°C, followed in turn by 31 cycles of 15 s at 94°C, 45 s at 57°C, and 1 min at 72°C, followed finally by extension for 2 min at 72°C. PCR products were visualized by electrophoresis on a 1.6% agarose gel at 100 V.

Direct sequencing of PCR products was performed by using a BigDye terminator version 3.1 ready reaction cycle sequencing kit and an Applied Biosystems genetic analyzer AB13730. Sequences were edited in Sequencher v4.1 (Genecodes Corp.). A multiple alignment was constructed by using CLUSTALX v1.81 (11, 12). In Se-Al v2.0a11 Carbon (A. Rambaut, University of Oxford) a 33-bp ambiguously aligned region of V5 was deleted to exclude it from subsequent analyses. The best model of evolution for our data was determined by using Modeltest v3.7 (22). This model, TVM+I+G, allowing for variable rates across sites and a proportion of invariant sites, was used to construct a neighbor-joining (NJ) tree (27) with PAUP* v4.0b10 (31). A nonparametric bootstrap (6) with 1,000 replicates was conducted.

Novel sequences from New Zealand cats were submitted to GenBank (accession numbers EF153955 to EF154083).

FIV subtypes A and C predominate within infected New Zealand cats (Fig. 1). The principal subtype is C, comprising 48% (±4.4%) of all sequences (Table 1) . Although there is some clustering of sequences from the same location, such as that seen from Great Barrier Island in subtype A, there is no obvious geographical pattern to the subtypes found throughout NZ. Australia, the country closest to New Zealand, has only subtypes A and, more rarely, B (14). Thus, the high prevalence of FIV-C in New Zealand cats was unexpected. FIV subtype A sequences from three infected Australian cats were included in Fig. 1 to determine whether Australia is a potential country of origin of New Zealand FIV-A. These Australian sequences group within the New Zealand subtype A clade, as did reference FIV-A sequences from other countries; thus, there is no firm evidence that Australia is the origin of New Zealand FIV-A.

FIG. 1.

Midpoint-rooted neighbor-joining tree with map of New Zealand showing regions and feral cat prevalence data. Subtypes are shown along the right side of the tree. F represents sequences from a recent study in Texas (32) with accession numbers U02422, AY139094, AY139095, AY139096, AY139097, AY139098, AY139099, AY139100, AY139101, AY139102, AY139103, and AY139104. U represents sequences from New Zealand that do not group with a known subtype. Three subtype A Australian sequences are included (AUS1, AUS2, and AUS3). The reference sequences downloaded from GenBank include subtype A: JN-BR1 (D67052, Japan), Petaluma (M25381, United States), PTH-BM3 (AB010401, Japan), SAP03 (AB010404, Japan), Sendai1 (D37813, Japan); subtype B: Aomori2 (D37817, Japan), RP1-A (AJ304988, Portugal), Sendai2 (D37814, Japan), TM2 (M59418, Japan), and TY1 (D67064, Japan); subtype C: CABCpady02C (U02392, Canada), CABCpbar01C (U02393, Canada), DEBAfred (U57020, Germany), TI-2 (AB016026, Taiwan), TI-3 (AB016027, Taiwan), and TI-4 (AB016028, Taiwan); subtype D: Fukuoka (D37815, Japan), MC8 (D67062, Japan), MY2 (D67063, Japan), Shizuoka (D37811, Japan), and TSU104 (AB021111, Japan); subtype E: LP3 (D84496, Argentina) and LP20 (D84498, Argentina). Numbers are bootstrap values based on 1,000 replicates.

TABLE 1.

FIV-infected NZ cat samples, showing cat lifestyle, location, and FIV subtype

Sample	Lifestyle	Locationa	Subtypeb		Sample	Lifestyle	Locationa	Subtypeb
190	Domestic	Auckland	A		EV09	Stray	Auckland	C
192	Domestic	Canterbury	A		WSPCA05	Stray	Northland	C
256	Domestic	Hawke's Bay	A		WSPCA15	Stray	Northland	C
PN1	Domestic	Wellington	A		BP07	Feral	Canterbury	C
PN2	Domestic	Wellington	A		BP08	Feral	Canterbury	C
PN6	Domestic	Manawatu-Wanganui	A		BS03	Feral	Hawke's Bay	C
PN10	Domestic	Wellington	A		BS11	Feral	Hawke's Bay	C
PN13	Domestic	Taranaki	A		BS13	Feral	Hawke's Bay	C
PN14	Domestic	Manawatu-Wanganui	A		BS14	Feral	Hawke's Bay	C
PN16	Domestic	Taranaki	A		BS16	Feral	Hawke's Bay	C
PN19	Domestic	Wellington	A		BS44	Feral	Hawke's Bay	C
WST01	Domestic	West Coast	A		GBI11	Feral	Auckland	C
WST02	Domestic	West Coast	A		GBI25	Feral	Auckland	C
WST04	Domestic	NA	A		GBI31	Feral	Auckland	C
WST05	Domestic	NA	A		GBI43	Feral	Auckland	C
WSPCA11	Stray	Northland	A		GBI46	Feral	Auckland	C
BP01	Feral	Canterbury	A		GBI47	Feral	Auckland	C
BP09	Feral	Canterbury	A		MF07	Feral	Otago	C
BP12	Feral	Canterbury	A		MF09	Feral	Otago	C
BP18	Feral	Canterbury	A		MF11	Feral	Otago	C
BP26	Feral	Canterbury	A		MF12	Feral	Otago	C
BP28	Feral	Canterbury	A		MF16	Feral	Otago	C
GBI04	Feral	Auckland	A		TKP05	Feral	Northland	C
GBI08	Feral	Auckland	A		TKP07	Feral	Northland	C
GBI14	Feral	Auckland	A		TKP08	Feral	Northland	C
GBI37	Feral	Auckland	A		TKP18	Feral	Northland	C
GBI49	Feral	Auckland	A		TKP43	Feral	Northland	C
GBI72	Feral	Auckland	A		TKP52	Feral	Northland	C
GBI84	Feral	Auckland	A		TKP54	Feral	Northland	C
MF01	Feral	Otago	A		TKP60	Feral	Northland	C
MF02	Feral	Otago	A		TKP64	Feral	Northland	C
MF04	Feral	Otago	A		TKP87	Feral	Northland	C
MF05	Feral	Otago	A		TKP93	Feral	Northland	C
MF33	Feral	Otago	A		TKP95	Feral	Northland	C
MF34	Feral	Otago	A		TKP104	Feral	Northland	C
MF37	Feral	Otago	A		TKP105	Feral	Northland	C
MF40	Feral	Otago	A		168	Domestic	West Coast	PR
MF42	Feral	Otago	A		197	Domestic	Nelson	PR
TKP74	Feral	Northland	A		214	Domestic	Waikato	PR
164	Domestic	Wellington	C		258	Domestic	Bay of Plenty	PR
177	Domestic	Southland	C		259	Domestic	Taranaki	PR
193	Domestic	Canterbury	C		260	Domestic	Bay of Plenty	O
229	Domestic	Auckland	C		PN17	Domestic	Wellington	PR
240	Domestic	NA	C		PN21	Domestic	Manawatu-Wanganui	PR
253	Domestic	Waikato	C		PN22	Domestic	Manawatu-Wanganu	O
266	Domestic	Taranaki	C		PN23	Domestic	Wellington	PR
282	Domestic	NA	C		MF14	Feral	Otago	PR
298	Domestic	NA	C		EV01	Stray	Auckland	U
B	Domestic	Auckland	C		WSPCA01	Stray	Northland	U
BD1	Domestic	Auckland	C		BS08	Feral	Hawke's Bay	U
NZVP01	Domestic	Wellington	C		MF19	Feral	Otago	U
NZVP02	Domestic	Wellington	C		MF21	Feral	Otago	U
NZVP03	Domestic	NA	C		MF29	Feral	Otago	U
PN3	Domestic	Hawke's Bay	C		TKP02	Feral	Northland	U
PN4	Domestic	NA	C		TKP14	Feral	Northland	U
PN5	Domestic	Wellington	C		TKP15	Feral	Northland	U
PN7	Domestic	Wellington	C		TKP17	Feral	Northland	U
PN8	Domestic	Manawatu-Wanganui	C		TKP20	Feral	Northland	U
PN9	Domestic	Manawatu-Wanganui	C		TKP21	Feral	Northland	U
PN11	Domestic	Wellington	C		TKP22	Feral	Northland	U
PN12	Domestic	Manawatu-Wanganui	C		TKP57	Feral	Northland	U
PN15	Domestic	Wellington	C		TKP73	Feral	Northland	U
PN20	Domestic	Wellington	C		TKP88	Feral	Northland	U
WST03	Domestic	West Coast	C		TKP94	Feral	Northland	U
EV07	Stray	Auckland	C

^aNA, not available. Locations by region are shown on the map in Fig. 1.

^bU, unknown subtype; O, outlier but not labeled putative recombinant since not significant as determined by KH test; PR, putative recombinant.

A group of 17 New Zealand sequences, 11 of which are from feral cats from one location, did not group with any known subtype on the phylogenetic tree. Fifteen sequences from this group are very closely related (91.0% similarity), suggesting their recent spread in New Zealand cats. All 17 sequences were analyzed by using a recombination identification program (RIP) (28; http://hivweb.lanl.gov/RIP/RIPsubmit.html) to test for recombination. None of these sequences showed any significant similarity to any previously described subtype (Fig. 2a), nor did they appear to be recombinants of known subtypes. Two lymph node samples, one that yielded FIV sequence of subtype C and another of the “unknown” subtype, were amplified and sequenced independently at the School of Veterinary Science, University of Queensland, in a blind, independent check for contamination. These sequences were almost identical (99%) to those we obtained, thus confirming the novelty of the “unknown” group. We note that the suggested requirements for designating a new human immunodeficiency virus subtype requires three genome sequences from epidemiologically independent infected individuals (24). While FIV subtype designation has been based on V3-V5 env sequence (29), we are currently sequencing the pol and gag genes to provide further sequence data to validate the “unknown” group as a novel NZ-specific subtype.

FIG. 2.

RIP graphical outputs of TKP21, a sequence from the “unknown” group (a); 214 (b) and 197 (c), two outlier sequences. The plots represent the distance between the query sequence and the subtype A (blue) or subtype C (red) reference sequence, over a sliding window of size 200 bp. In panels b and c, there is a clear point (the putative recombinant crossover point) at which the query sequences cease to be closer to A and become closer to C. In panel a, all five known subtypes are compared to the query sequence, with no significant similarity, although at the very end of the sequenced region there is some similarity to subtype C.

Eleven other sequences that did not group with any subtype were tested for recombination by using RIP (Fig. 2b and c). A Kishino-Hasegawa (KH) test (9, 10, 15) implemented in PAUP* was used to further investigate recombination. Thirteen reference sequences were used in each KH test: five subtype A sequences, five subtype C sequences, and one sequence of each subtype B, D, and E. The results from this test (Table 2) show that, conservatively, 9 of the 11 samples can be assigned as putative recombinants of subtypes A and C.

TABLE 2.

Kishino-Hasegawa test results of the eleven sequences ambiguously positioned on the phylogenetic tree

Sequencea	Diff −lnLb	Pc
Firstd
168†	67.635	0.000*
197	85.935	0.000*
214	28.094	0.006*
258†	108.153	0.000*
259†	68.407	0.000*
260	2.656	0.344
MF14	29.355	0.005*
PN17	72.269	0.000*
PN21†	35.537	0.002*
PN22	62.599	0.000*
PN23†	66.752	0.000*
Seconde
168	31.741	0.006*
197	51.234	0.000*
214	117.730	0.000*
258	31.743	0.002*
259	25.720	0.035*
260	40.658	0.000*
MF14	116.522	0.000*
PN17	23.612	0.047*
PN21	61.732	0.000*
PN22	20.601	0.092
PN23	27.922	0.019*

^a†, Samples for which endpoint dilutions were performed.

^bThe difference in log-likelihood scores between the best tree for that subalignment against the best tree for the alternative subalignment.

^cThe probability of obtaining that difference in log-likelihoods by chance alone, given that the true trees are identical. *, P values significant at the 0.05 level. Note that, to be conservative, we only identify a sequence as a putative recombinant if both subalignments produce trees that are mutually (statistically) different.

^dThe subalignment of sequences using the query sequence before the putative crossover point identified by RIP.

^eThe subalignment of sequences using the query sequence after the putative crossover point identified by RIP.

To exclude PCR-mediated recombination as a potential source for these putative recombinants (16, 18), we used endpoint dilution (25, 26) on representative samples from which recombinants were detected. Because of limited sample availability, we were only able to perform these tests on five of the nine putative recombinants (Table 2). A fivefold serial dilution was used, and the PCR results were entered into QUALITY (http://ubik.microbiol.washington.edu/computing/quality/jquality.htm) to determine the proviral copy number. Samples were then diluted to an endpoint and amplified to give at least five sequences. In all cases, the endpoint sequences were at least 98% similar to the respective consensus sequences and to each other. Consequently, these recombinants are confirmed as authentic viral sequences, and there is no evidence of dual infection.

Of the 334 feral cats tested, 72 (21.5% ± 2.25%) were positive for FIV proviral DNA. This is generally higher than other feral cat populations globally (2, 8, 19, 33; however, see also reference 3). The FIV prevalence in feral cats from the five New Zealand locations ranges from 11 to 36%, with no obvious geographic pattern (Fig. 1).

More than two-thirds of the FIV-infected feral cats were male, a statistically significant difference (χ² = 14.16, df = 1, P ≤ 0.001), and adult male cats had the highest prevalence of 51% (n = 70). This trend is expected because mature male cats demonstrate higher levels of social aggression, caused by competition over mates and territories (5, 7). Thus, mature male cats tend to have a greater number of biting encounters, the typical mode of transmission of FIV (34).

This study is the first to use phylogenetics to analyze FIV in New Zealand cats. Two subtypes and nine putative A/C recombinant strains are identified in New Zealand cats. A novel clade was detected, possibly representing a New Zealand-specific subtype.