The detection of bovine leukemia virus proviral DNA by PCR–ELISA

Journal of Virological Methods 99 (2002) 33 – 40 www.elsevier.com/locate/jviromet

The detection of bovine leukemia virus proviral DNA by PCR–ELISA M. Rola, J. Kuzmak * National Veterinary Research Institute, Al.Partyzantow 57, 24 -100 Pulawy, Poland Received 23 April 2001; received in revised form 27 July 2001; accepted 30 July 2001

Abstract A sensitive non-radioactive microplate hybridization assay for the detection of proviral DNA of bovine leukemia virus (BLV)-specific polymerase chain reaction (PCR) product is described. The PCR products are labeled by adding digoxigenin-dUTP to the nested PCR reaction and are captured by a microtitre plate coated with oligonucleotide probe, which is complementary to the inner region of the amplification product. Captured products are reacted with an anti-DIG Fab fragment conjugated to peroxidase, and detected using a colorimetric reaction. The PCR-enzyme linked immunosorbent assay (ELISA), detecting as low as 10 − 4 ng of proviral DNA in a background of 1 mg of BLV-negative DNA, was up to 100-fold more sensitive than ethidium bromide staining, and showed equal sensitivity to Southern blot hybridization. Using this method it was possible to monitor the presence of proviral DNA in four sheep infected experimentally with BLV, over a 10 months postinfection period, as well as in 29 cattle infected naturally. The test is rapid and highly sensitive and is a useful additional tool for the detection of BLV-infected animals. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Bovine leukemia virus; Proviral DNA; PCR–ELISA detection

1. Introduction Bovine leukemia virus (BLV), an exogenous retrovirus, is the causative agent of enzootic bovine leukosis. Under natural conditions, the disease occurs only in cattle, but sheep are also susceptible to experimental infection. The virus infects preferentially and transforms B lymphocytes, but has also been found in T cells, * Corresponding author. Tel.: + 48-81-8188-63051; fax: + 48-81-88-62595. E-mail address: [email protected] (J. Kuzmak).

monocytes and granulocytes (Heeney et al., 1992; Schwartz et al., 1994). Infection by BLV may remain silent clinically as an aleukemic form but about one-third of infected cattle develop persistent lymphocytosis and 5–10% develop lymphoid tumors (Ghysdael et al., 1984). Screening for antibodies has been the primary means of detecting the presence of the BLV infection. The agar-gel immunodiffusion (AGID) and enzyme linked immunosorbent assay (ELISA) are used widely for routine detection of antibodies against BLV (Johnson and Kaneene, 1992). The general use of these methods is hampered frequently by the dis-

0166-0934/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 0 9 3 4 ( 0 1 ) 0 0 3 8 4 - 6

34

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

covery that BLV infected cattle can be found with low, transient or without BLV-antibody titers (Cockerell and Rovnak, 1988; Eaves et al., 1994). It is therefore important to determine the BLV status by direct detection of the virus. Since the virus remains latent in vivo, procedures for its direct detection are more difficult and consequently not useful for routine diagnosis of BLV infection. It has been shown that amplification and detection of BLV proviral DNA or BLV RNA sequences by the polymerase chain reaction (PCR) provides a sensitive means of direct diagnosis of BLV infection (Sherman et al., 1992; Klintevall et al., 1994). The majority of the PCR assays are based on a single PCR, however, it was shown that in nested PCR less than eight genome copies of the provirus were detected in the background of 2 million negative lymphocytes (Ballagi-Pordany et al., 1992). One of the problems that still restricts the diagnostic use of PCR is the difficulty in detecting the specific PCR products. The detection is based largely on ethidium bromide-stained gel electrophoresis but this technique has a low level of DNA detection and is often complicated by the presence of other amplified DNA (Reichel et al., 1998; Naif et al., 1990). To circumvent these problems and to increase the sensitivity several methods for detecting hybrids between DNA probes and amplified products have been described (Naif et al., 1990; Kelly et al., 1993; Kuzmak et al., 1993). But these techniques, based on the blotting process, require considerably more laboratory work. In this study, a procedure was developed which allows rapid and quantitative detection of BLV proviral DNA with PCR followed by PCR – ELISA. A nested PCR was carried out to incorporate into the amplified target DNA digoxigenin (DIG)-labeled dUTP and the amplified product was then detected by an ELISA using oligonucleotide-coated microtiter plates and anti-DIG peroxidase conjugate consecutively. In particular, the use of such a system has been well documented in the diagnosis of infection with human and bovine retroviruses (Mallet et al., 1993; Baron et al., 1998). In this study, the PCR– ELISA was applied

to detect the proviral BLV-DNA in field cattle as well as in sheep infected experimentally and the results were compared with the analysis of PCR products by gel electrophoresis and Southern blot hybridization.

2. Material and methods

2.1. Animals and sample collection Four male sheep, Kent breed, were inoculated intravenously with peripheral blood mononuclear cells (PBMCs; 5× 105 cells) from a lymphocytotic cow persistently infected with BLV. Two sheep inoculated with PBMCs from BLV-free cattle were kept as controls. Blood samples were collected biweekly for the first 3 months and monthly thereafter up to 10 months after inoculation (a.i.). The study also involved 29 cattle randomly selected from five herds with a high incidence of BLV infection. PBMCs from both sheep and cattle were harvested as by the manufacturer’s protocol on Histopaque 1.077 (Sigma), washed twice with phosphate buffer saline (PBS) and stored at − 20 °C, until DNA extraction.

2.2. DNA extraction A pellet of PBMCs (2×106 cells) was suspended in a lysis buffer (10 mmol/l Tris –HCl, pH 8.0; 1 mmol/l EDTA, 100 mg/ml proteinase K) and the mixture was incubated at 37 °C overnight and proteinase K was inactivated at 95 °C for 10 min. DNA was extracted by the phenol–chloroform/CIA method and the concentration was evaluated by measuring the optical density (OD) at 260 nm. Samples were stored at − 20 °C until 1 mg of DNA was subjected to PCR analysis.

2.3. Oligonucleotide PCR primers and probe The first amplification was done by using the primers spanning the entire en6 gene region encoding for envelope glycoprotein gp51 4920(5%TGGAGA TGC TCC CTG T C CCT-3%) 4940, and 5703 (5%-CTCC TACCCGGGTCAGACGT3%) 5723. The second amplification was performed

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

to amplify a 427 bp fragment using the primers described originally by Ballagi-Pordany et al. (1992), 5035 (5%-GTGCCAAGTCTC CCAGATACA-3%) 5055 and 5442 (5%-TATAGCACAGTCTGG GAAGGC-3%) 5462. The oligonucleotide probe, 5%-CATAAGGGCATCGGGGCTCGCAATCA TATG-3%, specific for an internal region of 427 bp PCR product described by the same investigators, was synthesized with a phosphorylated ten thymidine linker at the 5% end. All nucleotide numbers originate from a previous study (Sagata et al., 1985).

2.4. Polymerase chain reaction Amplification was carried out in 50 ml containing 10 mM Tris–HCl, pH 8.8, 1.5 mM MgCl2, 150 mM KCl and 0.1% Triton X-100, 0.2 mM each dNTP (Amersham), 2U of Taq polymerase (Finnzymes), 1 mM of each primer and 1 mg of DNA template. Thermal cycling was carried out using a PEC Thermal Cycler 480 as follows, 45 s at 94 °C, 1 min at 65 °C, 2 min at 72 °C; after the last cycle, the samples were incubated at 72 °C for an additional 10 min. The first PCR was carried out in 35 cycles and after this the reaction products were diluted 100-fold and subjected to the second amplification which was undertaken under the same conditions as stated above except 0.2 mM of primers, 0.19 mM dTTP and 0.01 mM digoxigenin-dUTP (Boehringer Mannheim) concentration, respectively. The annealing step for the second amplification was performed at 55 °C for 1 min. Aliquots (10 ml) were analyzed by the microtitre plate hybridization assay described below. For electrophoretic analysis, 10 ml aliquots of PCR products were run in 1.4% agarose gel in a TBE buffer; molecular weight markers (100 bp DNA Ladder, Gibco BRL) were included in each run. Gels were stained in 0.5 mg/ml ethidium bromide for 10 min, and then analyzed under ultra violet (UV) light using a Gel-Doc 1000 system (Bio-Rad).

35

(Kuzmak et al., 1993) on the second amplification which was carried out without the incorporation of DIG-dUTP. PCR products were probed with BLV-427 bp amplicon after its labeling with digoxigenin-dUTP (Boehringer Mannheim).

2.6. PCR – ELISA The detection and quantification of DIG-labeled BLV amplified products was done using the PCR –ELISA system (Nunc A/S, Roskilde, Denmark) according to the supplier’s protocol. Briefly, NucleoLink™ strips were coated with 10 pmol of oligonucleotide probe for 5 h at 50 °C. Then the strips were washed three times, soaked for 5 min and washed again with TBST buffer (100 mM Tris–HCl pH 7.5, 150 mM NaCl and 0.1% Tween 20). Washing steps were repeated with deionized sterile water to remove salt residues. The empty, coated strips were stored at 4 °C in a polyethylene bag for further use. Ten microliters of DIG-labeled nested PCR product was mixed with 10 ml of denaturing solution containing 0.1 mmol/l NaOH. The hybridization was carried out by adding 80 ml of hybridization solution for 2 h at 50 °C. The plates were washed with 250 ml 0.5× SSC and 0.1% Tween 20. DIGlabeled amplified products were detected by the addition of 100 ml of anti-DIG-AP Fab fragment (Boehringer), diluted to 1:5000 in TBST buffer with 1% of Blocking Reagent (Boehringer), followed by incubation at 37 °C for 1 h. The strips were washed again three times with TBST buffer and incubated with 100 ml of para-nitrophenylene phosphate (p-NPP) substrate solution for 30 min at room temperature. The OD was measured at 405 nm in the ELISA reader (Dynatech MR 5000) and the results were expressed as net OD, after the substrate blank was subtracted for each sample.

3. Results

2.5. Southern blot hybridization

3.1. Sensiti6ity of PCR–ELISA

Confirmatory Southern blot hybridization was carried out according to our previous report

Because the concentration of proviral BLVDNA in the positive samples was unknown the

36

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

sensitivity of the PCR– ELISA was tested in the amplification experiments by 10-fold dilutions of plasmid DNA containing the entire BLV genome (kindly provided by Dr R. Kettmann, Gembloux, Belgium) in the presence of 1 mg background DNA from BLV-negative cattle. PCR products (10 ml) were then detected by the ELISA. As shown in Fig. 1, the limit of detection of this assay is about 10 − 4 ng of BLV-DNA. This detection level was reached when 10 ml of PCR products were analyzed by Southern blot hybridization but not when analyzed by ethidium bromide staining, which allows the detection of BLV-DNA at concentrations as low as 10 − 2 ng of DNA per lane (data not shown). In the range between 10 − 4 and 10 − 1 ng of specific DNA, there was a linear correlation; template concentration higher than 10 − 1 ng of DNA resulted in saturation of oligonucleotide probe.

3.2. Detection of pro6iral DNA of BLV in experimentally infected sheep PCR –ELISA was then applied in an experiment involving the infection of four sheep, monitored for the presence of BLV-DNA at different intervals after infection (Table 1). Each amplified product was analyzed by PCR– ELISA as well as by ethidium bromide staining and Southern hy-

Fig. 1. Relation between different amounts of plasmid BLVDNA and OD value. Each plotted value has been derived from three independent replicates.

bridization. The mean OD obtained after the amplification of samples collected at intervals over a period of 40 weeks a.i. from non-infected control animals (0.068+ 0.021) was used as a basis for setting the cut-off limit 0.131 for the positive signal (mean+ 3 S.D.). The results of spectrophotometric readings clearly shows that all animals were negative before inoculation and became positive within 2 weeks a.i. However, the load of provirus was not stable. The OD value of three DNA preparations declined below the calculated cut-off, at 16, 28 and 32 weeks a.i., respectively. These samples also yielded negative results by Southern blot and gel electrophoresis. All positive results of PCR–ELISA agreed fully with those of Southern blot and gel electrophoresis.

3.3. Detection of pro6iral DNA of BLV in field cattle DNA from PBMCs of 29 field cattle was amplified and BLV-DNA was determined by PCR – ELISA. As described above, each amplified product was analyzed by Southern blot and ethidium bromide staining and the results of this analysis are shown in Table 2. The cut-off value (0.120) was determined as the mean+ 3 S.D. for 20 DNA samples, isolated from PBMC of BLVfree cattle. Under this estimation BLV genome was detected in 22 samples whereas seven were considered as negative. Six showed readings greater than the background level, but as already described a sample was considered positive if the OD reading was greater than the background level plus 3 S.D. This finding should not be misinterpreted as an adverse result, since a clearcut negative signal was obtained for these samples when tested by confirmatory Southern hybridization assay. Therefore, these results demonstrate a 100% correlation with the results obtained by PCR –ELISA and by Southern blot hybridization. It should be noted that the highest OD value of PCR –ELISA, ranging between 1.513 and 2.100, agreed fully with the strong positive results by ethidium bromide staining. However, several samples which were doubtful or negative by ethidium bromide staining were clearly shown positive by PCR –ELISA and by Southern blot.

Sheep

6 12 15 16 Negative control 0.068 (0.021)b Cut-off 0.131c

Weeks a.i. 0

2

4

6

8

12

16

20

24

28

32

36

40

0.074 a 0.059 0.036 0.067

2.023 2.032 1.418 1.039

0.597 1.012 0.560 0.570

0.372 0.863 0.278 0.752

0.263 1.111 0.589 0.832

0.943 0.621 0.409 1.004

0.105 1.152 0.885 0.384

1.354 1.260 0.711 0.203

1.119 1.415 0.286 1.159

1.486 1.964 0.089 1.123

1.246 1.794 0.364 0.104

0.854 1.071 0.480 0.520

0.582 0.730 0.620 0.420

Italic areas presents DNA samples negative by PCR–ELISA which were also negative by Southern blot and ethidium bromide staining, respectively. a Each value calculated from two replicates. b Mean (and S.D.) for 26 samples taken between 0 and 40 weeks a.i. from two non-infected sheep. c Mean+3 S.D. of negative control.

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

Table 1 Estimation of OD value of proviral DNA using the PCR–ELISA in PBMCs obtained from sheep experimentally infected with BLV

37

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

38

Table 2 Detection of proviral DNA by different methods in PBMCs obtained from cattle naturally infected with BLV Sample

Southern blot hybridization

Agarose gel

PCR–ELISA (OD)

2 36 42 45 57 228 180 31 67 127 164 52 80 118 10 11 4 17 16 22 44 12 111 222 229 216 110 74 121 Negative control 0.044 (0.020)b Cut-off 0.120c

+ + + + + + + + + + + + + + + + + + Weak + + + − − − − − − −

+++a +++ +++ +++ +++ +++ ++ ++ ++ ++ ++ + + + +/− +/− +/− +/− − − − − − − − − − − −

+ + + + + + + + + + + + + + + + + + + + + + − − − − − − −

(1.979) (1.896) (1.513) (2.087) (1.953) (2.100) (0.900) (0.820) (0.693) (0.750) (0.924) (0.345) (0.420) (0.425) (0.239) (0.278) (0.323) (0.187) (0.153) (0.230) (0.160) (0.120) (0.110) (0.107) (0.054) (0.045) (0.053) (0.050) (0.021)

a

+++, ++, + strong positive, positive and weakly positive reaction. Mean (and S.D.), n = 20. c Mean+3 S.D. of negative control. b

4. Discussion A sensitive and quantitative PCR– ELISA method was developed in order to improve the diagnosis of infection with BLV. The method was based on a nested PCR reaction using DIG-labeled dUTP resulting in DIG-labeled amplicons that were detected by hybridization with captured oligonucleotide probe and anti-DIG peroxidase conjugate. The method was shown to be up to 100-fold more sensitive than ethidium bromide staining and capable of detecting about 10 − 4 ng of proviral DNA. The analysis of DNA samples

from animals infected experimentally and naturally by PCR–ELISA and by PCR plus Southern blot hybridization indicate that both techniques had an equal sensitivity level. Yet the detection of PCR products by ELISA offers numerous advantages over hybridization systems based on Southern blot (Naif et al., 1990; Ballagi-Pordany et al., 1992; Kuzmak et al., 1993) or liquid hybridization (Sherman et al., 1992). This system is simpler, more rapid to conduct and in micrititre format gives the capacity for a larger number of samples to be analyzed simultaneously. The results are measured using an ELISA reader which removes

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

any element of subjective interpretation as can occur with agarose gel detection. A further advantage is that no hazardous components such as ethidium bromide or radioactive isotopes are used. The main feature of PCR– ELISA is the feasibility of quantitative estimation of proviral DNA. The data from sheep infected experimentally, monitored for the presence of provirus, clearly shows that the negative results obtained by PCR were attributable to the temporal decrease of BLV DNA available for amplification. This could explain the failure of PCR assay performed with animals positive serologically (Eaves et al., 1994; Jacobs et al., 1992; Reichel et al., 1998). The test presented above shows a 10-fold higher sensitivity than PCR– ELISA that includes microwells coated with double-stranded-DNA-binding protein GCN4 and only a single PCR amplification (Naif et al., 1992). The other PCRbased quantitative ELISA assay has been reported to detect BLV DNA at the same format as our test with a detection limit of 6×102 molecules of template (Rasmussen et al., 1994). This level of sensitivity allowed the estimation of positive DNA samples isolated from bovine lymphoid tumor or lymph nodes of experimentally infected sheep. It was found that those tissues are known to contain relatively high proviral load (Kettmann et al., 1978). In contrast, PBMCs from seropositive, asymptomatic and hematologically normal cattle contain one to three proviral copies and the frequency of BLV infected cells was found to be less than 10% (Cockerell and Rovnak, 1988; Mirsky et al., 1996). Furthermore, some studies suggest that, in field asymptomatic cattle, the proportion of PBMCs infected with BLV can vary by as much as 10-fold (Molly et al., 1994; Cowley et al., 1992). Therefore, in the PCR– ELISA described above, an effort was made to adopt this test for detection provirus in the animals infected at an early stage of disease. The data from the analysis of the negative sera of sheep and cattle showed that 50% of the calculated cut-off level corresponds to three times the S.D. of background variation, providing the test with high specificity. This was also illustrated by the discovery that none of the samples negative by PCR–

39

ELISA was negative by Southern blot hybridization. Nevertheless, two samples (cattle 111, 222) were found with an OD value below the cut-off and well above from the OD of negative control, but there are few reasons to believe that these samples, which were not positive by Southern blot, were false negative. In conclusion, determination of BLV proviral load in field samples by the PCR–ELISA assay described here could be a practically useful PCR technique for the detection of BLV, and could be of great value in the eradication programme. Such an approach could be seen as an equivalent to PCR followed by hybridization and applied as a confirmatory test (Enzootic Bovine Leukosis, 1996 OIE Manual).

Acknowledgements This work was supported by the Polish Committee of Scientific Research grant, 5PO6K00611. The authors are very grateful to Barbara Furtak for her technical support.

References Ballagi-Pordany, A., Klintevall, K., Merza, M., Klingeborn, B., Belak, S., 1992. Direct detection of bovine leukemia virus infection: practical applicability of a double polymerase chain reaction. J. Vet. Med. 39, 69 – 77. Baron, T., Betemps, D., Mallet, F., Cheynet, V., Levy, D., Belli, P., 1998. Detection of bovine immunodeficiency-like virus infection in experimentally infected calves. Arch. Virol. 143, 181 – 189. Cockerell, G., Rovnak, J., 1988. The correlation between the direct and indirect detection of bovine leukemia virus infection in cattle. Leuk. Res. 12, 465 – 469. Cowley, J., Molly, J., Dimmock, C., Walker, P., Bruyeres, A., Ward, W., 1992. Infectvity of bovine leukaemia virus infected cattle: an ELISA for detecting antigens expressed in vitro cultured lymphocytes. Vet. Microbiol. 30, 137 – 150. Eaves, F.W., Molloy, J., Dimmock, C., Eaves, L., 1994. A field evaluation of the polymerase chain reaction procedure for the detection of bovine leukaemia virus proviral DNA in cattle. Vet. Microbiol. 39, 313 – 321. Enzootic Bovine Leukosis, 1996. Manual of Standards for Diagnostic Tests and Vaccines. OIE, Paris, pp. 276 – 280. Ghysdael, J., Bruck, C., Kettmann, R., Burny, A., 1984. Bovine leukemia virus. In: Current Topics in Microbiology

40

M. Rola, J. Kuzmak / Journal of Virological Methods 99 (2002) 33–40

and Immunology, vol. 112. Springer, Berlin, Heidelberg, pp. 1– 19. Heeney, J., Valli, P., Jacobs, R., Valli, V., 1992. Evidence for bovine leukemia virus infection of peripheral blood monocytes and limited antigen expression in bovine lymphoid tissue. Lab. Invest. 66, 608 –617. Jacobs, R., Song, Z., Poon, H., Heeney, J., Taylor, A., Jefferson, B., Vernau, W., Valli, V., 1992. Proviral detection and serology in bovine leukemia virus-exposed normal cattle and cattle with lymphoma. Can. J. Vet. Res. 56, 339 –348. Johnson, R., Kaneene, J.B., 1992. Bovine leukaemia virus and enzootic bovine leukosis. Vet. Bull. 62, 286 –312. Kelly, E., Jackson, M., Marsolais, G., Morrey, J., Callan, R., 1993. Early detection of bovine leukemia virus in cattle by use of the polymerase chain reaction. Am. J. Vet. Res. 54, 205 – 209. Kettmann, R., Dechamps, J., Cleuter, Y., Ghysdael, J., Mammerickx, M., 1978. Distribution of bovine leukemia virus proviral DNA sequences in tissue of animals with enzootic bovine leukosis. Leuk. Res. 2, 23 – 32. Klintevall, K., Ballagi-Pordany, A., Naslund, K., Belak, S., 1994. Bovine leukemia virus: rapid detection of proviral DNA by nested PCR in blood and organs of experimentally infected calves. Vet. Microbiol. 42, 191 –204. Kuzmak, J., Grundboeck, J., Kozaczynska, B., 1993. Detection of the bovine leukaemia virus proviral DNA in cattle by the polymerase chain reaction. Med. Vet. 49, 312 –315. Mallet, F., Hebrard, C., Brand, D., Chapuis, E., Cros, P., Allibert, P., Besnier, J.M., Barin, F., Mandrand, B., 1993. Enzyme-linked oligosorbent assay for detection of polymerase chain reaction-amplified human immunodeficiency virus type 1. J. Clin. Microbiol. 31, 1444 –11449. Mirsky, M., Olmstead, C., Da, Y., Lewin, H., 1996. The prevalence of proviral leukemia virus in peripheral blood mononuclear cells at two subclinical stages of infection. J. Virol. 70, 2178 – 2183.

Molly, J., Dimmock, C., Eaves, F., Bruyeres, A., Cowley, J., Ward, W., 1994. Control of bovine leukaemia virus transmission by selective culling of infected cattle on the basis of viral antigen expression in lymphocyte cultures. Vet. Microbiol. 39, 323 – 333. Naif, H., Brandon, R., Daniel, R., Lavin, M., 1990. Bovine leukaemia proviral DNA detection in cattle using the polymerase chain reaction. Vet. Microbiol. 25, 117 – 129. Naif, H., Daniel, R., Cougle, W., Lavin, M., 1992. Early detection of bovine leukemia virus by using an enzymelinked assay for polymerase chain reaction-amplified proviral DNA in experimentally infected cattle. J. Clin. Microbiol. 30, 675 – 679. Rasmussen, S.R., Rasmussen, H.B., Larsen, M.R., Hoff-Jorgensen, R., Cano, R.J., 1994. Combined polymerase chain reaction – hybridization microplate assay used to detect bovine leukemia virus and Salmonella. Clin. Chem. 40, 200 – 205. Reichel, M.P., Tham, K.M., Barnes, S., Kittelberger, R., 1998. Evaluation of alternative methods for detection of bovine leukaemia virus in cattle. New Zealand Vet. J. 46, 140 – 146. Sagata, N., Yasunaga, T., Tsuzuku-Kawamura, J., Ohishi, K., Ogawa, Y., Ikawa, Y., 1985. Complete nucleotide sequence of the genome of bovine leukemia virus: its evolutionary relationship to the other retroviruses. Proc. Natl. Acad. Sci. USA 82, 677 – 681. Schwartz, I., Bensaid, A., Polack, B., Perrin, B., Berthelemy, M., Levy, D., 1994. In vivo leukocyte tropism of bovine leukemia virus in sheep and cattle. J. Virol. 68, 4589 – 4596. Sherman, M., Ehrlich, G., Ferrer, J., Sninsky, J., Zandomeni, R., Dock, N., Poiesz, B., 1992. Amplification and analysis of specific DNA and RNA sequences of bovine leukemia virus from infected cows by polymerase chain reaction. J. Clin. Microbiol. 30, 185 – 191.