Evaluation of a novel nested PCR for the routine diagnosis of feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV)

Abstract

Laboratory diagnosis of feline leukemia virus (FeLV) and feline immunodeficiency virus (FIV) usually involves both viruses, as the clinical signs are similar and coinfection may occur. Serological methods may not represent an accurate diagnosis: maternal antibodies or cross-reactions may give false positive results to FIV, and false negative results may occur in latent FeLV status, or in certain FIV infection stages. A nested polymerase chain reaction (PCR) technique was designed to detect FeLV, FIV and feline endogenous retrovirus simultaneously. The detection of endogenous sequences was considered indicative of successful DNA extraction. The technique was used to diagnose FIV and FeLV in the blood cells of 179 cats. The κ value with the serological data was 0.69 for FeLV and 0.87 for FIV. The joint detection of FeLV and FIV by this novel nested PCR is sensitive, specific, fast and convenient, and its applicability for clinical diagnosis is promising, as the direct evidence of the presence of the virus is more realistic than the indirect data provided by the serological detection.

Feline leukemia virus (FeLV) is an exogenous retrovirus which has a worldwide distribution and affects domestic cats (Felis catus) and sporadically, wild cats (Lutz et al 1992, Daniels et al 1999, Leutenegger et al 1999). By comparison of the sequences that code for the envelope proteins (env), it has been possible to identify three FeLV subgroups: FeLV-A only replicates in cat cells and is present in all the infected cats, alone or in combination with FeLV-B and/or FeLV-C. FeLV-B is thought to have arisen from the recombination between FeLV-A and feline endogenous sequences (Bechtel et al 1998) and FeLV-C from mutations of FeLV-A and -B (Jarrett 1992). Thus, FeLV is very closely related to feline endogenous sequences.

Feline immunodeficiency virus (FIV) belongs to the genus Lentivirus, and it is, in many aspects, very similar to human immunodeficiency virus type 1 (HIV-1), to the extent that it has been proposed as the best animal model to study the pathogenesis and therapy for the human virus (Bendinelli et al 1995). Cats infected by FIV have four times more probability of being infected by FeLV, and double infected cats show more severe clinical signs than in the case of monoviral infection (Courchamp et al 1997).

FeLV-associated diseases are typically divided into proliferative diseases (lymphoma or leukemia), or degenerative diseases (Rojko and Hardy 1994). Within these degenerative diseases, the feline acquired immunodeficiency syndrome induced by some variants of FeLV (FeLV-FAIDS (feline acquired immunodeficiency syndrome)) is usually fatal, and very similar to human AIDS (Hardy and Essex 1986). Immunodeficiency is produced because FeLV replicates in the cells of the immune system, producing a dramatic decrease in the populations of lymphocytes and granulocytes. More cats die from FeLV-induced immunodeficiency than from proliferative diseases (Hardy and Essex 1986).

The clinical course of feline immunodeficiency associated with FIV has been divided into five stages (Bendinelli et al 1995, Hartmann 1998). Stage 1 is characterized by non-specific clinical signs (fever, depression, lymphadenitis). Stage 2 may last 1–5 years, and is asymptomatic. Stage 3 may last from months to years, and is also characterized by non-specific signs. In stage 4 (ARC, AIDS-related complex), there are secondary infections (but no opportunistic infections). Its duration ranges from several months to 1 year. Stage 5 (FAIDS), which lasts for various months, is characterized by the onset of opportunistic infections (Bendinelli et al 1995).

As clinical signs are quite non-specific in the infections with FIV and FeLV, laboratory diagnosis is always required. In addition, the long periods in which no signs are observed, make routine tests advisable in order to determine the infection status of the animals, as well as to implement the adequate prevention strategy. Feline leukemia and immunodeficiency are typically diagnosed by the detection of FeLV antigens and antibodies against FIV in peripheral blood, usually by an enzyme-linked immunosorbent assay (ELISA) test. However, false positives may occur, especially in very young animals, when maternal anti-FIV antibodies are still present, or due to cross-reactions (Barr 1996). In addition, a killed whole virus FIV vaccine has been made available to practitioners. Vaccinated cats seroconvert by ELISA and Western blot, making presently available diagnostic tests, which rely on antibody detection, useless in cats after vaccination (Andersen and Tyrrell 2004). Furthermore, false negatives may also occur, as FIV-infected cats do not have antibodies during a period that ranges from 2 to 4 weeks up to a year post-infection (Swango 1991), or during the terminal stages of infection, or in leukopenia. Similarly, FeLV antigen may not be detected in latent FeLV, when there is no viremia, but the viral genome is integrated in bone marrow or other sites (Pedersen et al 1989, Innis et al 1990, Barr 1996, Herring et al 2001).

Molecular diagnostic methods like polymerase chain reaction (PCR), are becoming more popular due to their advantages over serological methods. The detection of proviral DNA in peripheral blood monocytic cells allows identification of the virus independently of the presence of antibodies or viremia. The PCR technique is extremely sensitive, and the method theoretically allows the detection of a single DNA molecule in a background of 10⁵ cells (Saiki et al 1988). If the amplification is first performed using an outer pair of primers followed by a second amplification using a pair of inner nested primers, both the specificity and the sensitivity of the test are increased (Kemp et al 1989). PCR has been used in several experimental studies for the detection of FeLV and FIV in peripheral blood monocytic cells (Bendinelli et al 1995, Miyazawa and Jarrett 1997), and in fresh and formalin-fixed tissues (Ellis et al 1996).

Due to the possibility of double FIV-FeLV infections, and to the advantages of a joint detection of both retroviruses, the aim of the present work was to design and optimize a nested PCR using a pair of outer primers common to both viruses and two sets of inner primers specific for each one in the same reaction tube. This technique was also evaluated for the routine diagnosis in 179 field samples by comparing both PCR and serological (ELISA) results.

Materials and Methods

Cell lines

Cell lines used included the feline lines Ho6 (persistently infected by FIV) and FL-74 (persistently infected by FeLV). Cell line CRFK (ATCC CCL94) and CC81 (Fischinger et al 1974), were used as positive endogenous retrovirus controls. Cell lines U937 (ATCC CRL1593), J774 A.1 (ATCC TIB 67), THP-1 (ATCC TIB 202), HL-60 (ATCC CCL 240) BGM, PM2, BHK-21 (ATCC CCL 10), MDBK (NBL-1) (ATCC CCL 22), MDCK (NBL-2) (ATCC CCL 34), Vero (ATCC CCL 81), RK-13 (ATCC CCL 37), and Caco2 (ATCC HTB 37) were also used as negative controls. The specificity of the assay was checked using cells infected by other retroviruses: FLK-BLV and BLV-bat₂ (both infected by bovine leukemia virus), ovine fibroblasts infected by maedi-visna virus (MVV), and Cf₂Th (ATCC CRL 1430) (persistently infected with bovine immunodeficiency virus, BIV). All cells were grown in RPMI 1640 (BioMedia, Boussens, France) and 10% fetal bovine serum (FBS, Gibco, Rockville, MD, USA), to which penicillin (100 U/ml), streptomycin (0.1 mg/ml), antimycotic (1%) and anti-PPLO (Gibco) had been added.

DNA extraction

Cells were centrifuged and resuspended in 500 μl of DNA extraction buffer (200 mM Tris–HCl pH 7.5, 250 mM NaCl, 25 mM EDTA, 0.5% SDS), vortexed, and incubated for 15 min in ice. Samples were centrifuged at 4000 rpm for 8 min in a Microfuge (Henle, Wehingen, Germany), and the supernatant was transferred to a clean Eppendorf tube. DNA was extracted with 500 μl of phenol–chloroform–isoamyl alcohol (25:24:1). After centrifuging at 14,000 rpm for 3 min, the aqueous phase was again transferred to a clean Eppendorf tube, and one volume of chloroform–isoamyl alcohol (24:1) was added. Then, the samples were centrifuged as before. DNA was precipitated with one volume of isopropanol, centrifuged at 14,000 rpm for 3 min and washed with ethanol. Finally, the air-dried samples were resuspended in 50 μl of double distilled water (ddH₂O).

Primers

The pol gene was targeted for primers' design, as it is a very conserved region in retroviruses. The primers (Table 1) were chosen after carefully comparing at least 10 sequences published in GenBank of each FIV, FeLV and endogenous retroviruses, using the program OMIGA v.3.0 (Oxford Molecular Company). The first step of the nested PCR pursued the amplification of sequences of FeLV, FIV and endogenous retroviruses which had annealed to primers common to all three. The second step of the nested PCR sought the amplification of differential sequences between the three viruses. Diagnosis was based on the different molecular weight of each fragment, allowing for the joint detection of these viruses.

Table 1

Sequence and location of FeLV and FIV oligonucleotide primers

	Sequence (5′→3′)	Polarity	Location^a	Product size (bp)
FF1	AMCCRTTATTRGGRAGAG	Sense	FIV, 2278–2295	FIV, 1325
			FeLV, 2681–2697
FF2	CAMAGYAGCATGGATRTM	Antisense	FIV, 3582–3602	FeLV, 490
FeLV, 3152–3170
FI5	CAATGGCCATTAAATGAA	Sense	2403–2424	1138
FI4	AGAGAGGCCTGGAATCAAAT	Antisense	3519–3540
FE7	GAAAGTACACAAAAACAGGAG	Sense	2827–2846	306
FE4	CTTAAGTCCTGCACTGG	Antisense	3115–3132

Based on PubMed. Nucleotide Query Accession No. AF052723 (FeLV), M25381, and M25729 (FIV).

Reagents

PCR was performed in 0.2 ml thin wall PCR tubes in a total volume of 25 μl. Each tube contained 2.5 μl reaction Buffer 10× (Biotools, Madrid, Spain) (75 mM Tris–HCl (pH 9.0), 2 mM MgCl₂, 50 mM KCl, 20 mM (NH₄)₂SO₄), 0.5 μl of deoxynucleotide mix (dNTPs, stock solution 10 mM), 1 unit of Ultratools Taq polymerase (Biotools), 1 μl of each primer solution (10 μM), 3 μl of DNA sample in the first step, or 4 μl of product of first amplification diluted 1/10 for the second PCR. The volume was made up to 25 μl with ddH₂O.

Amplification

First and second amplification reactions composing the nested PCR protocol were conducted separately in a thermal cycler (MJ Research, Inc., Watertown, MA, USA). Different temperatures of annealing were compared, in the range between 45 and 65°C, using an Eppendorf Mastercycler Gradient (Eppendorf, Hamburg, Germany). The extension time in both PCR reactions was optimized, comparing 1, 2, and 3 min for the first PCR reaction and 30 s, 1 min, and 90 s for the second PCR reaction. The number of cycles was optimized: 25, 30, and 35 cycles were tested for each amplification step.

Product analysis

The amplification products were analyzed in 1.3% agarose gel electrophoresis, with ethidium bromide, using both the 100 bp and 1 kb DNA ladders (Biotools) as reference markers. Electrophoresis was performed in TAE buffer (40 mM Tris-acetate, 0.5 M EDTA), for 1 h at 90 V. Results were visualized under UV light transillumination. To confirm that amplification products corresponded to the target sequences of FeLV, FIV and endogenous retroviruses, purification and later sequencing were done in some cases. Geneclean (Bio101, Carlsbad, CA, USA) and QIAquick Spin (Qiagen, Hilden, Germany) kits were used for purification. Sequencing was undertaken in the Sequencing Laboratory of the Pharmacy School of Complutense University (Madrid). Sequencing data were analyzed with BLAST 2.0.10 (http://www.ncbi.nlm.nih.gov/blast/) software, which confirmed the results.

Sampling

Once optimized, the PCR technique was applied to 179 pet cats from Madrid metropolitan area. The international guiding principles for biomedical research involving animals were always followed. A blood sample (0.5–1.5 ml) was extracted from every cat by venepuncture using EDTA or heparin as an anticoagulant, and was centrifuged at 10,000 rpm for 5 min to separate the plasma for serological testing. Erythrocyte lysis buffer (10 mM KHCO₃, 460 mM NH₄Cl, 0.1 mM EDTA) was added to the 200–500 μl cellular fraction and samples were transferred to Eppendorf tubes. When the red blood cells were evidently lysed, as judged from the change in the color of the mixture and the different viscosity, samples were centrifuged at 4000 rpm for 5 min in the Microfuge, and the supernatant was discarded. The pellet obtained was washed with PBS and centrifuged at 8000 rpm for 3 min, a process which was repeated at least three times, till a clean white pellet was obtained. DNA was extracted from the field samples and tested as described above. After purification, DNA was resuspended in 40 μl of ddH₂O and 3 μl were used for PCR analysis. Scrupulous laboratorial procedures were followed in order to minimize cross-contamination and false positives: DNA extraction, first PCR, second PCR, and amplicon analysis were all conducted in different rooms. Moreover, PCRs were prepared in different biosafety cabinets.

Serological testing

The commercial kit Snap Combo Plus (Idexx, Inc., Westbrook, ME, USA) was used to detect p27CA from FeLV and antibodies against gp40SU from FIV in the 179 plasma samples. According to the literature, the sensitivity and specificity of this type of ELISA ranges between 95 and 100% for FeLV (Hawks et al 1991, Jacobson and Lopez 1991, Swango 1991). For FIV, sensitivity is between 93 and 100%, and specificity varies between 98 and 99.6% (Barr et al 1991, Fierro et al 1995).

Statistical methods

The association test χ² was used to calculate P, with an α value of 0.05. To calculate concordance level, κ value was used with a confidence level of 95%. Confidence intervals were calculated with an α value of 0.05.

Results

Optimization of PCR parameters

Of the different combinations of primers, the best results were obtained using FF1/FF2 for the first reaction and FI4/FI5 (specific for FIV) and FE4/FE7 (specific for FeLV) for the second amplification reaction. These primers rendered products of 1325 bp for FIV and 490 bp for FeLV in the first PCR step, and 1138 bp for FIV, 306 bp for FeLV and 257 bp for the endogenous retroviruses in the second PCR step (Fig. 1).

Fig 1

Optimization of the number of cycles for the first and second PCR reactions. For the first PCR reaction, the number of cycles is read across and for the second it is read down. The positions of the bands of 1138 bp corresponding to FIV, 306 bp to FeLV and 257 bp to the feline endogenous retroviruses are shown. Lanes 1, 10⁵ Ho6 (FIV infected) cells; lanes 2, 10⁵ FL-74 (FeLV infected) cells; lanes 3, 10⁵ CRFK (feline endogenous retroviruses control) cells; lanes 4, 10⁷ U937 (negative control of DNA) cells; and lanes 5, double distilled water.

The number of cycles for each amplification reaction was optimized, comparing the results obtained with 25, 30, and 35 cycles for each PCR reaction. Results are shown in Fig. 1. Even though a clear detection of endogenous band was not achieved in all the samples, the best results were considered to be 35 cycles for the first PCR reaction, and 25 for the second (Fig. 1), as the possibilities of seeing a non-specific band in the negative control decreased with these conditions, and the bands were the sharpest. Likewise, the temperature for annealing was also optimized. Few differences were seen in the range of temperatures 45–60°C (data not shown). However, 51°C was chosen, as it was closer to the Tm of the primers. Lastly, extension time was also optimized (Fig. 2). The optimal conditions for PCR were established as follows: (a) First PCR reaction using FF1 (10 μM), 1 μl and FF2 (10 μM), 1 μl: 35 cycles of 94°C, 1 min; 51°C, 1 min; 72°C, 3 min; preceded by 94°C, 7 min, and concluded with 72°C, 10 min. (b) Second PCR reaction using FI4 (10 μM), 1 μl; FI5 (10 μM), 1 μl; FE4 (10 μM), 1 μl; FE7 (10 μM), 1 μl: similar to the first step but including 25 cycles and extending at 72°C for 90 s.

Fig 2

Optimization of the time of extension at 72°C for the first and second PCR reactions. For the first PCR reaction, the time is read across and for the second it is read down. The positions of the bands of 1138 bp corresponding to FIV, 306 bp to FeLV and 257 bp to the feline endogenous retroviruses are shown. Lanes 1, 10⁶ FL-74 (FeLV infected) cells; lanes 2, 10³ FL-74 cells; lanes 3, 10⁶ Ho6 (FIV infected) cells; lanes 4, 10³ Ho6 cells; lanes 5, 10⁶ CRFK (feline endogenous sequences control) cells; and lanes 6, double distilled water.

PCR specificity and sensitivity

Sensitivity was determined when all the different parameters had been optimized. FL-74 cells were decimally diluted using a background of 10⁶ THP-1 cells. Ho6 cells were also decimally diluted using a background of 10⁶ U937 cells. From the results shown in Fig. 3 it may be seen that the technique was able to detect 100 FeLV infected cells and one FIV cell. Regarding specificity, no amplification products other than the endogenous band were observed in the feline uninfected control CRFK (Figs 1 and 2). No non-specific products were detected in any of the 16 cell lines used as negative controls (including lines infected with BLV, BIV, and MVV). Although we did not perform specific sensitivity assays with both FeLV and FIV together, no cross-reactions were observed in the joint detection of FeLV, FIV and endogenous retroviruses neither when Ho6 and FL-74 cells were mixed, nor in double positive field samples (data not shown).

Fig 3

Assay of sensitivity. Samples were amplified using 35 cycles for the first PCR reaction and 25 for the second. Extension was performed at 72°C for 3 min in the first PCR reaction, and 90 s in the second PCR reaction. The position of the bands of 1138 bp corresponding to FIV, 306 bp to FeLV and 257 bp to the feline endogenous retroviruses is shown. Lane 1, 10⁶ FL-74 (FeLV infected) cells; lane 2, 10⁵ FL-74 cells; lane 3, 10⁴ FL-74 cells; lane 4, 10³ FL-74 cells; lane 5, 10² FL-74 cells; lane 6, 10 FL-74 cells; lane 7, one FL-74 cell; lane 8, FeLV DNA control; lane 9, CRFK (endogenous retroviruses control); lane 10, double distilled water; lane 11, 100 kb DNA ladder; lane 12, 10⁶ Ho6 (FIV infected) cells; lane 13, 10⁵ Ho6 cells; lane 14, 10⁴ Ho6 cells; lane 15, 10³ Ho6 cells; lane 16, 10² Ho6 cells; lane 17, 10 Ho6 cells; lane 18, 1 Ho6 cell; lane 19, FIV DNA control; and lane 20, double distilled water.

Field samples

The design of this nested PCR for the joint detection of FeLV and FIV included the amplification of endogenous feline sequences (that are present in the DNA of all cats), as internal control of PCR reaction. This endogenous band did not show consistently in all the samples in which FeLV and/or FIV bands were detected, but it did appear in all FeLV/FIV negative samples. Thus, the risk of false negative results due to the absence of DNA in samples (faulty extractions, leukopenia) or to PCR inhibitors, was avoided. Samples showing discordant PCR/serological results were retested again and the results were confirmed. Non-specific products were seen only in very few samples (around 2%). Furthermore, the bands were very faint and did not interfere with the detection of the target sequences. These unspecific bands may have arisen from the amplification of related viruses such as feline foamy virus (FeFV). FeLV proviral genome was detected in 64 of 179 samples (35.7%); of these 64 positive cases, 16 (25%) were ELISA negative (Table 2). Due to the design of the nested PCR, there were two types of FeLV positive cases: (a) unique detection of the FeLV band (65.6%), or (b) joint detection of the FeLV and the endogenous retroviruses bands (34.4%). Most of the cases (95.2%) of the first type were ELISA positive, while only 36.4% of the second type was ELISA positive. There is a statistical association (P<0.05) between the detection of the FeLV band solely and the ELISA positive result. Of the remaining 115 cases which were negative to the detection of FeLV proviral genome, six (5.2%) were ELISA positive to this virus. Regarding FIV, the proviral sequence was amplified in 20 of 179 samples (11.2%). All of these positive cases were also ELISA positive (Table 2). However, there were five seropositive samples that resulted PCR negative (FIV was detected by PCR in 80% of seropositive cats). On the other hand, the proviral genome was not detected in any seronegative case (ELISA negative). Unlike FeLV positive samples, there was not any significant association between the serological data and the joint detection of endogenous retroviruses in FIV positive cases.

Table 2

ELISA and PCR results of the 179 cats tested for FeLV and FIV

	FIV		FeLV
PCR +	PCR −		PCR +	PCR −
ELISA +	20	5	25	48	6	54
ELISA −	0	154	154	16	109	125
	20	159	179	64	115	179

Concordance

For FeLV diagnosis, the concordance value κ between PCR and ELISA was 0.69 (0.55–0.84). For FIV diagnosis, κ value between PCR and ELISA was 0.87 (0.76–0.98). These κ values are considered substantial, and almost perfect, respectively (Herring et al 2001).

Discussion

A simple, reliable, fast and reproducible nested PCR technique for the joint diagnosis of FIV and FeLV has been developed and successfully applied to field samples. The novelty of this method resides in the use of an outer pair of primers common to the feline retroviruses, and two inner pairs simultaneously for the differentiation of FeLV and FIV. So, unlike other previously published PCR methods, this method allows the simultaneous detection of FIV and FeLV. The assay has been tested in a large number of field samples. This should be very useful, as the clinical signs which accompany either infection are very non-specific and most clinicians demand the diagnosis of both retroviruses in any given cat. This method is highly sensitive, as well as specific, as no bands were detected in both non-infected feline cells and other non-feline cells infected with other retroviruses. The pol region was chosen on the basis that it is a very conserved region in the retroviral genome. Several inner primers were studied; the ones chosen gave the best results, as respect sensitivity and specificity.

In the present study, DNA was extracted using the phenol–chloroform method. However, the time necessary to provide a result may be decreased by using commercial kits for DNA extraction. The analysis of the PCR amplicons on 1.3% agarose gels allowed the differentiation of small bands close in size, such as the one for endogenous retroviruses (257 bp), FeLV (306 bp), and the possible detection of the related feline foamy virus (FeFV, 360 bp), even though no evident band of this size was ever seen.

The prevalence of FeLV infected cats found by this novel PCR (35.7%) was strikingly higher when compared to the prevalence of FeLV positive cats found in seroepidemiological studies performed in Spain and other European countries like Italy and Germany (Bandecchi et al 1992, Fuchs et al 1994, Arjona et al 2000). In these studies, the prevalence of FeLV positive cats by ELISA ranged between 13 and 18%.

In our study, in 25% of the samples proviral FeLV was detected but not p27 antigen (PCR positive/ELISA negative). Assuming that the false negative ELISA results are exceptional (Barr 1996), this result may be due to the detection of latent FeLV cases in which there is no viremia. On the other hand, six cats out of 115 PCR negative samples were positive by ELISA. Different possibilities might explain this finding: (a) early stages of infection in which there is an initial viremia but the viral genome is yet to integrate in the cellular genome (Pedersen et al 1989, Innis et al 1990, Barr 1996); (b) localized infections (rather than systemic), where an efficient immune response takes place before the virus reaches the bone marrow, and where the provirus is hidden in tissues; and (c) false positive ELISA results. Additional testing (eg, virus isolation) might have been valuable in determining the true status of discordant subjects. However, in line with our findings, Tandon et al (2005) have also shown that proviral detection assays for FeLV are more sensitive than ELISA and virus isolation in the early phase of infection.

Due to the design of the nested PCR, there were two types of FeLV positive cases: unique detection of FeLV band, or joint detection of FeLV and endogenous bands. The detection of a single intense FeLV band had a high correlation with the ELISA positive result (which means viremia). Furthermore, 63.6% of the cases where two bands were seen, corresponding to FeLV and endogenous retroviruses, were ELISA negative. During the viremic phase the virus increases its multiplication and expression, so the amount of infected lymphocytes may be higher. As there is more proviral DNA to be detected, primers could have more affinity to annealing with FeLV sequences, depleting reagents for the detection of the endogenous sequences.

FIV proviral genome was detected in 80% of seropositive cats. As it is highly improbable that an infected cat rejects the infection (Yamamoto et al 1989), it seems that this discordance could be due to false positive ELISA results, as this nested PCR had shown high sensitivity. Research at New York State Diagnostic Laboratory (Cornell, NY) showed that there were around 20% of false positive results (when confirmed by Western Blot) using ELISA kits for in-office diagnosis (Barr 1996). Another possible reason for PCR negative/ELISA positive discordance, is the detection of maternal antibodies or the use of FIV vaccine. As maternally derived antibodies may persist for 6 months or longer (Barr et al 1991), PCR is highly recommended for kitten testing. Even though the PCR developed here targeted pol for being highly conserved among retroviruses, the fact that FIV presents a high diversity of strains, any of which could escape detection of this technique has to be considered. There was no FIV proviral detection in seronegative cases. That shows, assuming a high concordance between infection and seropositivity (Pedersen et al 1989), that this nested PCR technique has a high specificity for FIV detection. However, a PCR positive result in seronegative cases may occur in cats which have not seroconverted, or in advanced clinical stages in which the debilitation of the immune system hampers the production of antibodies. There is a window of 4–8 weeks in which the cat is infected but does not test positive by ELISA (Barr 1996). Currently available PCR tests for FIV have been shown to vary significantly in diagnostic accuracy between different laboratories (Bienzle et al 2004, Crawford et al 2005). As a nested PCR inherently improves sensitivity and specificity (Kemp et al 1989), we believe that the novel PCR approach presented in this study may significantly enhance the performance of PCR-based diagnostic assays.

The diagnosis by PCR shows the following advantages compared to the ELISA testing. First, the PCR technique allows the detection of latent FeLV cases, which may reactivate to the viremic status (Rojko et al 1982, Rojko and Olsen 1984), with the following consequences: development of FeLV-related diseases, contagion to other cats, and maternal transmission of the virus. Second, the PCR technique allows the diagnosis of FIV infection in cats which have not seroconverted, and avoids the risk of false positive results of maternally derived antibodies.

In conclusion, this nested PCR has been shown as an effective diagnostic technique. As it is advantageous in some circumstances for the diagnosis of FeLV and FIV, its routine use should be considered for the detection of these retroviruses. The joint detection of FeLV, FIV, and endogenous sequences in the same PCR reaction is an important technologic innovation that, in addition to conferring a high security in the diagnosis, allows time and money saving.

Footnotes

Acknowledgements

The authors wish to thank the veterinary practitioners who have participated in this study. In this sense, we are especially grateful to Gustavo Sanchez Visconti and the personnel of the Laboratorio de Analisis Veterinarios, as well as the staff of the Hospital Clinico Veterinario for their kind collaboration. The assistance of Prof Dr Jose Antonio Ruiz-Santa-Quiteria in the epidemiological study, and of David Bruhn for his editorial assistance are gratefully acknowledged. Ho6 and FL-74 were kind gifts of Dr Horzinek (University of Utrecht); U937, J774, THP-1, HL-60 were a gift from Dr Essex (University of Harvard); BGM and PM2 were donated by Dr Lopez Carrascosa (Universidad Autonoma de Madrid); CC81 and FLK-BLV were a gift from Dr Burny (Universite Livre de Bruxelles); BLV-bat₂ were generously provided by Dr Radke (U.C. Davis); ovine fibroblasts and MVV were provided by Dr Amorena (Universidad de Zaragoza); Cf₂Th was a gift from Dr Levy (Universite dAlfort). The Spanish funding CAM-08.2/0011/98/98 and DGES-PM98-0077/98 supported this work.

References

Andersen

P.R.

Tyrrell

Feline immunodeficiency virus diagnosis after vaccination, Animal Health and Research Reviews 5, 2004, 327–330.

Arjona

Escolar

Soto

Barquero

Martín

Gomez-Lucia

Seroepidemiological survey of infection by feline leukemia virus and feline immunodeficiency virus in Madrid. Correlation with some clinical aspects, Journal of Clinical Mirobiology 38, 2000, 3448–3449.

Bandecchi

Matteucci

Baldinotti

Guidi

Abramo

Tozzini

Bendinelli

Prevalence of feline immunodeficiency virus and other retroviral infections in sick cats in Italy, Veterinary Immunology and Immunopathology 31, 1992, 337–345.

Barr

M.C.

FIV, FeLV, and FIPV: interpretation and misinterpretation of serological test results, Seminars in Veterinary Medicine and Surgery (Small Animals) 11, 1996, 144–153.

Barr

M.C.

Pough

M.B.

Jacobson

R.H.

Scott

F.W.

Comparison and interpretation of diagnostic tests for feline immunodeficiency virus infection, Journal of the American Veterinary Medical Association 199, 1991, 1377–1381.

Bechtel

M.K.

Mathes

L.E.

Hayes

K.A.

Phipps

A.J.

Roy-Burman

In vivo evolution and selection of recombinant feline leukemia virus species, Virus Research 54, 1998, 71–86.

Bendinelli

Pistello

Lombardi

Poli

Garzelli

Matteucci

Ceccherini-Nelli

Malvaldi

Tozzini

Feline immunodeficiency virus: an interesting model for AIDS studies and an important cat pathogen, Clinical Microbiology Reviews 8, 1995, 87–112.

Bienzle

Reggeti

Wen

Little

Hobson

Kruth

The variability of serological and molecular diagnosis of feline immunodeficiency virus infection, Canadian Veterinary Journal 45, 2004, 753–757.

Courchamp

Supo

Fromont

Bouloux

Dynamics of two feline retroviruses (FIV and FeLV) within one population of cats, Proceedings of the Royal Society of London Series B. Biological Sciences 264, 1997, 785–794.

10.

Crawford

P.C.

Slater

M.R.

Levy

J.K.

Accuracy of polymerase chain reaction assays for diagnosis of feline immunodeficiency virus infection in cats, Journal of the American Veterinary Medical Association 226, 2005, 1503–1507.

11.

Daniels

M.J.

Golder

M.C.

Jarrett

MacDonald

D.W.

Feline viruses in wildcats from Scotland, Journal of Wildlife Diseases 35, 1999, 121–124.

12.

Ellis

J.A.

Jackson

M.L.

Bartsch

R.C.

McGill

L.C.

Martin

K.M.

Trask

B.R.

Haines

D.M.

Use of immunohistochemistry and polymerase chain reaction for detection of oncornaviruses in formalin-fixed, paraffin-embedded fibrosarcomas from cats, Journal of the American Veterinary Medical Association 209, 1996, 767–771.

13.

Fierro

G.M.

Bundesen

Martin

Jones

Beetson

Robinson

W.F.

Evaluation of an autologous red cell agglutination test, VetRED FIV, for the presence of FIV antibody in cats, Veterinary Immunology and Immunopathology 46, 1995, 93–101.

14.

Fischinger

P.J.

Blevins

C.S.

Nomura

Simple quantitative assay for both xenotropic murine leukemia and ecotropic feline leukemia viruses, Journal of Virology 14, 1974, 177–179.

15.

Fuchs

Binzel

Lonsdorfer

Epidemiology of FeLV and FIV infection in the Federal Republic of Germany, Tierarztliche Praxis 22, 1994, 273–277.

16.

Hardy

W.D.

Jr. Essex

FeLV-acquired immune immunodeficiency syndrome (FAIDS). A model for human AIDS, Progress in Allergy, 1986, 353–373.

17.

Hartmann

Feline immunodeficiency virus infection: an overview, Veterinary Journal 155, 1998, 123–137.

18.

Hawks

D.M.

Legendre

A.M.

Rohrbach

B.W.

Comparison of four test kits for feline leukemia virus antigen, Journal of the American Veterinary Medical Association 46, 1991, 93–101.

19.

Herring

E.S.

Troy

G.C.

Toth

T.E.

Forrester

S.D.

Weigt

L.A.

Herring

I.P.

Detection of feline leukaemia virus in blood and bone marrow of cats with varying suspicion of latent infection, Journal of Feline Medicine and Surgery 3, 2001, 133–141.

20.

Innis

M.A.

Gelfand

D.H.

Sninsky

J.J.

White

T.J.

PCR Protocols, a Guide to Methods and Applications, 1990, Academic Press: San Diego, CA, pp. 3–482

21.

Jacobson

R.H.

Lopez

N.A.

Comparative study of diagnostic testing for feline leukemia virus infection, Journal of the American Veterinary Medical Association 199, 1991, 1389–1391.

22.

Jarrett

Pathogenicity of feline leukemia virus is commonly associated with variant viruses, Leukemia, (suppl 3), 1992, 135S–154S.

23.

Kemp

D.J.

Smith

D.B.

Foote

S.J.

Samaras

Peterson

G.M.

Colorimetric detection of specific DNA segments amplified by polymerase chain reactions, Proceedings of the National Academy of Sciences of the United States of America 86, 1989, 2423–2427.

24.

Leutenegger

C.M.

Hofmann-Lehmann

Riols

Liberek

Worel

Lups

Fehr

Hartmann

Weilenmann

Lutz

Viral infections in free-living populations of the European wild cat, Journal of Wildlife Diseases 35, 1999, 678–696.

25.

Lutz

Isenbugel

Lehmann

Sabapara

R.H.

Wolfensberger

Retrovirus infections in non-domestic felids, serological studies and attempts to isolate a lentivirus, Veterinary Immunology and Immunopathology 35, 1992, 215–224.

26.

Miyazawa

Jarrett

Feline leukemia virus proviral DNA detected by polymerase chain reaction in antigenaemic but non-viremic (discordant) cats, Archives of Virology 142, 1997, 323–332.

27.

Pedersen

N.C.

Yamamoto

J.K.

Ishida

Hansen

Feline immunodeficiency virus infection, Veterinary Immunology and Immunopathology 21, 1989, 111–129.

28.

Rojko

J.L.

Hardy

W.D.

Feline leukemia virus. Sherding

R.G.

The Cat – Diseases and Clinical Management, 2nd edn, 1994, Churchill Livingstone: New York, 263–432.

29.

Rojko

J.L.

Hoover

E.A.

Quackenbush

S.L.

Olsen

R.G.

Reactivation of latent feline leukemia virus infection, Nature 298, 1982, 385–388.

30.

Rojko

J.L.

Olsen

R.G.

The immunology of feline leukemia virus, Veterinary Immunology and Immunopathology 6, 1984, 107–165.

31.

Saiki

R.K.

Gelfand

D.H.

Stoffel

S.J.

Higuchi

Horn

T.B.

Mullis

K.B.

Erhlich

H.A.

Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science 239, 1988, 487–491.

32.

Swango

L.J.

Evaluation of feline leukemia virus diagnostic tests available for in-office use by veterinarians, Journal of the American Veterinary Medical Association 199, 1991, 1386–1389.

33.

Tandon

Cattori

Gomes-Keller

M.A.

Meli

M.L.

Golder

M.C.

Lutz

Hofmann-Lehmann

Quantitation of feline leukaemia virus viral and proviral loads by TaqMan real-time polymerase chain reaction, Journal of Virological Methods 130, 2005, 124–132.

34.

Yamamoto

J.K.

Hansen

E.W.

Morishita

T.Y.

Okuda

Sawa

T.R.

Nakamura

R.M.

Pedersen

N.C.

Epidemiologic and clinical aspects of feline immunodeficiency virus infection in cats from the continental United States and Canada and possible mode of transmission, Journal of the American Veterinary Medical Association 194, 1989, 213–220.