INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Infection of cattle with bovine leukemia virus (BLV), a member of the BLV-human T-cell leukemia virus (HTLV) group of retroviruses, results in persistent lifelong infection, the mechanism of which is still poorly understood. The main target of BLV infection is the B lymphocyte expressing surface immunoglobulin M (IgM) (7). The clinical manifestations of this persistent infection are polyclonal expansion of B cells in many of the infected animals and lymphosarcoma in a small percentage of infected animals (4, 17). Available data suggest two possible mechanisms for this expansion: (i) the activity of the BLV transactivating protein Tax, and (ii) interactions of other BLV proteins with cellular proteins. It is indeed well documented that although BLV does not possess a typical oncogene in its genome, BLV Tax can behave as such (30, 31). Transforming properties of Tax are better documented in HTLV biology where it has been shown to transactivate a variety of genes or long terminal repeat (LTR) sequences, transcriptional enhancers, oncogenes, and interleukins (3). Another possible mechanism is the direct interaction of viral proteins other than Tax with lymphocyte signaling pathways, resulting in an increased rate of proliferation and/or reduced B-cell apoptosis. Both phenomena are documented, although detailed interactions and proteins involved at each step are still not known (8, 9, 13).

BLV gp30, the transmembrane component of envelope glycoprotein, can participate in signaling interactions (1, 2, 6, 29). BLV gp30 has a long cytoplasmic tail with several motifs, including an immunoreceptor tyrosine-based activation motif (ITAM), an immunoreceptor tyrosine-based inhibition motif (ITIM), and an upstream proline-rich motif (Src homology [SH] 3 recognition site motif) (5, 25). SH2 and SH3 motifs are found in a diverse collection of cellular proteins and are involved in downstream signaling events of receptors for growth factors, cytokines, hormones, antigens, and extracellular matrices in the control of cell growth, differentiation, migration, and death (20). ITAMs and ITIMs are recognition sites for the SH2 motif and are shared among a number of signaling proteins associated with the B-cell and T-cell antigen receptors (25) and in several viruses that infect B cells (5). Proline-rich sequences, especially with the sequence PX₂PX_4-5P, where X represents any amino acid, are recognition sites for the SH3 motif (24). The proline-rich motif in proximity to an ITAM is found not only in BLV gp30 but also in proteins of other viruses that infect B cells, including the herpesviruses Epstein-Barr virus LMP2A and herpesvirus papio LMP2A and the orbiviruses African horsesickness virus VP7 and epizootic hemorrhagic disease virus VP7 (5). The presence of this motif in unrelated viral groups led us to hypothesize that the proline-rich motif is essential for viral survival and replication in B cells.

To test the significance of the proline-rich PX₂PX_4-5P motif in BLV gp30, we changed three of the prolines to alanines in an infectious molecular clone of BLV. The influence of the proline-rich motif on viral load and pathogenicity was studied in sheep.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Provirus mutants. The source of BLV provirus was the infectious molecular clone pBLV344H, as previously described (32). In this clone, the proline-rich motif has the sequence PHFPEISFPPK. In other isolates of BLV, there is an L instead of F, or a T or A instead of the third P (18). Mutations were performed using PCR and resulted in changing three prolines (positions 471, 474, and 479 according to Rice et al. [26]) to alanines. Briefly, two pairs of oligonucleotides were used in mutagenesis: flanking primers OL and OR, and internal primers OC1 and OC2 that contained altered bases encoding the P A mutations. The first round of amplification consisted of two separate reactions, using two primer sets. The first primer pair consisted of upstream OL (5'-ATC AAC AAT GGA TGA CAA CAT-3') and downstream OC1 (5'-CGA AGG AGA TTT CAG CGA AGT GGG CAG CCT GC-3'). The second pair was upstream OC2 (5'-GCC CAC TTC GCT GAA ATC TCC TTC GCC CCT AAA C-3') and downstream OR (5'-GAG GGT GGA ATA AAA AGA AAG-3'). Underlined bases designate those changed to cause the P A mutations. After the first round of amplification using two sets of primers (OL plus OC1 and OR plus OC2), the products of each reaction were mixed and used as template in a second round of amplification with only the flanking primers OL and OR. This resulted in a 2-kb amplicon that contained the desired PA mutations. The new amino acid sequence, starting with amino acid 471 (26) is AHFAEISFAPK. We took advantage of the presence of NcoI and XbaI restriction sites in the amplicon, and after cutting of the NcoI-XbaI fragment from the whole amplicon we cloned it into the original pBLV344H. The resultant construct is designated pppBLV344H. Correctness of the construct was verified by sequencing. For the transfection experiments, plasmids were purified using the Qiagen Plasmid Mega kit.

In vitro activity. To determine if the mutated virus was functional in vitro, approximately 2 × 10⁵ canine osteosarcoma cells (D17) were transfected by calcium phosphate-precipitated plasmid DNA (ProFection Mammalian Transfection System; Promega). Cells were cotransfected with pBLVLTR-CAT and either pBLV344H or pppBLV344H. The cells were then washed and cultivated in minimal essential medium (Gibco) supplemented with 10% heat-inactivated fetal calf serum. After 48 h the cells were harvested and washed with phosphate-buffered saline (PBS) and one-half of the cells were lysed by three cycles of freeze-thaw. After centrifugation, chloramphenicol acetyltransferase (CAT) activity was determined from the supernatants (as described in reference 32). The other half of the transfected cells were lysed by one cycle of freeze-thaw and used for p24 antigen titration by an enzyme-linked immunosorbent assay (ELISA) procedure as described previously (21, 22). Briefly, 96-well microtiter plates were precoated with the anti-p24 monoclonal antibody (MAb) 4'G9. The antigen mixtures to be tested were then added to the wells. After washing of the plates, the p24 antigen was revealed by colorimetric assay using two antibodies (2'C1 and 4'F5) conjugated with peroxidase.

In vivo transfection of sheep. Before starting the experiments, the animals were adapted to the housing and feeding conditions in the experimental herd in Pulawy, Poland. During this period, sheep were treated with parasiticides, and absence of parasites was confirmed.

Serological examination. Serologic response to BLV proteins was evaluated by agar gel immunodiffusion (AGID) assay, which detects both gp51 and p24 antibodies (Dr. Bommeli AG, Liebefeld Switzerland). Antibodies to p24 were determined by ELISA (ELISA BLV kit; Bioveta, Ivanovice na Hane, Czech Republic). Additionally, serial dilutions were prepared to measure the p24 antibody titer using the same ELISA kit. Sera from pppBLV-injected sheep with no detectable p24 antibodies were tested again to confirm results by using a second ELISA kit that also detects p24 antibodies (Institut Pourquier, Montpellier, France).

BLV p24 titration in PBMC cultures. Peripheral blood mononuclear cells (PBMCs) were isolated from the blood samples using Histopaque (Sigma) density gradient centrifugation and cultured for 48 h at a concentration of 3 × 10⁶/ml of Eagle medium supplemented with 10% heat-inactivated calf serum, L-glutamine, gentamicin, and amphotericin B. Cell-free supernatants were prepared by centrifugation for 10 min at 900 × g. The p24 major Gag antigen was then titrated from the culture supernatants by ELISA (21, 22).

PCR analysis of proviral sequences in blood samples. DNA for PCR was isolated from Histopaque-purified sheep PBMCs using the Genomic DNA Prep Plus kit (DNA-Gdansk II s.c., Gdask, Poland). The composition of the reaction mixture at a volume of 50 µl was PrimeZyme polymerase buffer (Biometra Ltd, Goettingen, Germany), 1.5 mM Mg²⁺, 0.5 µM concentrations of each primer, 1 mM deoxynucleoside triphosphates, and 2 U of PrimeZyme polymerase (Biometra). DNA (0.5 µg) was added to the reaction mixture and the total volume was covered with 2 drops of mineral oil. The amplification process was performed in a programmed thermal cycler (UnoII; Biometra Ltd). To examine proviral load, semiquantitative PCR was performed on the PBMCs. The number of PCR cycles was restricted to 25 in order to eliminate the "plateau effect" and to allow comparison between amplification of abundant and scarce BLV sequences. The reaction was started with denaturation at 94°C for 4 min followed by 25 cycles of 40-s denaturation at 94°C, 40-s primer hybridization at 65°C, and 1-min elongation at 72°C. The amplification was finished with a 5-min elongation. Two oligonucleotides were used: MCF-1 (5'-GCGAGAAACCATTCATTCTG-3') and MCR-2 (5'-CAAGAAGAGGCTTGTGATGG-3'). Amplification products of the BLV DNA were detected electrophoretically in a 2% agarose gel. The specificity of the PCR was confirmed by Southern blot hybridization of amplified DNA using a molecular probe (SacI insert of previously cloned provirus DNA) (23) followed by autoradiography. DNA samples from an FLK cell line infected persistently with BLV and from a persistently lymphocytotic, ELISA-positive cow were the positive controls. To more sensitively test for the presence of proviral DNA, a nested PCR was used. Initial template DNA (0.5 µg), approximately 5,000 cell-equivalents, was initially amplified with upstream primer 5'-ATCAACAATGGATGACAACAT and downstream primer 5'-GAGGGTGGAATAAAAAGAAAG. Denaturation was at 94°C for 4 min, followed by 30 cycles of 1-min denaturation at 94°C, 30-s primer hybridization at 57°C, and 1-min elongation at 72°C and concluding with a 7-min elongation. One-tenth of the initial amplification was used as a template for the second PCR, using primers MCF-1 and MCR-2 as described above. In the second reaction, 30 cycles were performed as above, except with a 40-s denaturation at 94°C and 40-s primer hybridization at 65°C. Amplification products were detected electrophoretically in a 2% agarose gel. As a control of sensitivity, known dilutions of plasmid DNA (pBLV344H) were prepared in water.

Flow cytometry. Cytometric analysis was performed using a FACStrak flow cytometer (Becton Dickinson Immunocytometry Systems), and the percentages of B-cell and T-cell subpopulations were recorded using Simulset and PC Lysis programs. Leukocytes were gated with MAbs directed against ovine antigens CD2 (MUC2A), CD4 (GC50A1), CD8 (CACT80C), B-B2 (BAQ44A), and WC1-N2 (BAQ4A) (VMRD Inc., Pullman, Wash.). WC1-N2 is a determinant on WC1⁺ T cells, the predominant population of peripheral blood T cells in sheep. Debris was excluded from the analysis by the conventional scatter gating method.

Statistical analysis. Analysis of variance was conducted by using the Statgraphics Plus statistical analysis package, version 2.0. The results are presented as the mean ± 1 standard deviation of the absolute number and percentage of each lymphocyte subpopulation. The statistical significance of the observed differences in the numbers and percentages of lymphocyte subsets between three experimental groups as well as between estimated time points within each group was evaluated by Student's t-test. A P value of <0.05 was considered statistically significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of mutated provirus. We constructed a mutant of bovine leukemia provirus by introducing three PA point mutations within the proline-rich motif upstream of the gp30 ITAM. The location of the mutations in the BLV gp30 and the mutated sequence are presented in Fig. 1. Before in vivo injection, wtBLV was compared with pppBLV in vitro by transient cotransfection of either wtBLV or pppBLV, together with pLTRCAT in the canine D17 osteosarcoma cell line. Expression of the transactivator Tax, as determined by activation of BLV LTR-CAT, and expression of the capsid p24 protein, as determined by ELISA, did not reveal significant differences (p > 0.2) between the proline-mutated and wild-type provirus (data not shown).

View larger version (27K): [in this window] [in a new window]   FIG. 1.   Proline mutations. Location of proline-rich motifs (PX2PX4P) inside of BLV envelope protein. The diagram presents the external (gp51) and transmembrane (gp30) glycoproteins anchored in the cell membrane. Proline-rich motifs are located in the cytoplasmic portion of gp30.

Initial infectivity. Seroconversion as measured by AGID assay and ELISA occurred in all of the animals transfected with wild-type or proline-mutated provirus 4 to 7 weeks after plasmid injection (Table 1). Negative control animals injected only with DOTAP remained seronegative. The five sheep transfected with wild-type provirus (wtBLV) seroconverted at a mean of 5.6 ± 1.1 and 4.8 ± 0.8 weeks, as measured by AGID and ELISA, respectively, while the five sheep transfected with the proline-mutated provirus (pppBLV) seroconverted at a mean of 5.8 ± 1.3 and 6.0 ± 1.6 weeks, as similarly measured. The differences between the time of seroconversion of sheep transfected with wtBLV versus pppBLV were not significant, regardless of the type of assay used. Similarly, there were no significant differences in the time to onset of p24 antigen expression between wtBLV- and pppBLV-infected sheep (5.4 ± 3.3 and 7.7 ± 2.3 weeks, respectively) as measured in supernatant of cultured PBMCs. In the wtBLV-transfected group of sheep, the shortest period to onset of p24 antigen expression was 3 months, while in the pppBLV-transfected group of animals it was 5 months.

Long-term serologic response. Despite an initial similarity in infectivity, as determined by the time of seroconversion, the dynamics of serum titers varied considerably after infection (Fig. 2, Table 1). Two of the five pppBLV-transfected sheep (animals 5 and 7) were serologically positive by AGID and ELISA for 4 and 2 months, respectively, but then became seronegative again. The other three pppBLV-transfected sheep (animals 8, 9, and 10) remained seropositive for the duration of the study. Twelve months from the time of transfection, the mean titer in the wtBLV-transfected group was four times higher than the mean titer for the three seropositive animals in the pppBLV-transfected group (273,060 versus 68,260). The difference between the group of animals (sheep 5 and 7) producing the low p24 antibody titers and the remaining pppBLV-infected sheep (numbers 8, 9, and 10) was also evident when the lymphocyte phenotype was compared (see below).

View larger version (34K): [in this window] [in a new window]   FIG. 2.   Comparison of mean p24 antibody titers in sera of four groups of sheep: animals transfected with wtBLV, the three animals transfected with pppBLV that were long-term antibody responders (numbers 8, 9, and 10) (pppBLV lr), the animals transfected with pppBLV that were only transient antibody responders (numbers 5 and 7) (pppBLV tr), and the negative control animals. Titers were determined by ELISA (ELISA BLV kit, Bioveta).

p24 expression in cultured PBMCs. Comparison of BLV p24 expression in supernatant from cultured PBMCs from the wtBLV and pppBLV groups revealed similar tendencies. There was a much higher mean p24 expression level from the wtBLV-transfected animals compared to the three long-term-seropositive, pppBLV-transfected animals (Fig. 3). Significantly, the two sheep that became seronegative also failed to express p24 protein after culture of PBMCs.

View larger version (51K): [in this window] [in a new window]   FIG. 3.   Titers of BLVp24 antigen in 2-day PBMCs cultures from four groups of sheep. The designations of the experimental groups are as for Fig. 2. PBMCs were collected at 3-month intervals, purified on a Histopaque gradient (Sigma), and cultured for 48 h in RPMI 1640 (Gibco BRL), and the supernatant was used for p24 ELISA titer determinations (optical density [O.D.]).

Semiquantitative PCR to compare viral loads. To examine proviral load, semiquantitative PCR was performed on PBMCs. The number of PCR cycles was restricted to 25, and dilutions of infected bovine PBMCs were used as a positive control. A 337-bp fragment from the env gene that is present in both the wild-type and proline-mutated proviral BLV was amplified. To increase sensitivity of detection, the PCR products were analyzed by Southern blotting using a BLV-specific probe (23), followed by nonradioactive detection using the ECL system (Amersham Pharmacia Biotech). As expected, the 337-bp fragment was present in the PCR of all wtBLV-transfected sheep throughout the whole experiment. At the same time, however, the proviral sequence was weakly and inconsistently detected in three of the five pppBLV-transfected sheep (numbers 5, 7, and 8) (Fig. 4). However, pppBLV-transfected sheep numbers 9 and 10 had proviral loads equivalent to those of the wtBLV-transfected animals. After initial infection, confirmed both by the presence of specific antibodies and proviral sequences in PBMCs, we were unable to detect BLV DNA in pppBLV-transfected sheep number 5 and 7 after 6 months posttransfection (Fig. 5). These results correlate well with the serological results for these two sheep (Table 1). The sheep were serologically positive for 4 and 2 months, respectively, but then became seronegative. The five negative control (uninfected) sheep had no detectable BLV DNA, confirming that there was no transmission among the groups of sheep.

View larger version (56K): [in this window] [in a new window]   FIG. 4.   Semiquantitative PCR analysis of proviral loads. PBMCs were collected and purified at 3-month intervals, starting 3 months after transfection. DNA was consistently extracted from constant amounts of 3 × 106 PBMCs, and 0.5 µg was amplified by 25 PCR cycles. Amplicons were electrophoresed and analyzed by Southern blot hybridization with a BLV probe. The semiquantitative analysis was supported by amplification of serial dilutions (1:10 and 1:1,000) of DNA from a naturally BLV-infected cow.

View larger version (55K): [in this window] [in a new window]   FIG. 5.   (a) Nested PCR of DNA from two pppBLV-transfected seronegative sheep (animals 5 and 7) obtained 38, 54, and 62 weeks after transfection. As a control of the sensitivity of the test, the results of nested PCR of serial dilutions of the parental plasmid DNA (pBLV344H) used to transfect the sheep as well as of BLV-positive bovine DNA are included. (b) The results of PCR of samples from the same sheep using primers for the goat -actin gene as a control for the integrity of the DNA samples. The primers do not amplify bovine -actin. Lanes 1 to 8 are as in panel a. C, PCR of sample with goat -actin plasmid, used as a positive control.

Leukocyte phenotype. The wtBLV group had a statistically significant increase in total numbers of lymphocytes and numbers of B cells compared with both the pppBLV and uninfected control groups. No significant differences were found between the pppBLV group and the uninfected negative control group. In the wtBLV-transfected group, the mean total lymphocyte count was (5.5 ± 0.3) × 10⁹/liter, whereas in the pppBLV-transfected group, the mean total lymphocyte count was (4.1 ± 0.3) × 10⁹/liter [(4.4 ± 0.5) × 10⁹/liter and (3.8 ± 0.4) × 10⁹/liter for pppBLVtr and pppBLVlr, respectively], and the negative control group had a mean total lymphocyte count of (4.1 ± 0.3) × 10⁹/liter (Table 2). The mean B-cell count was particularly increased in the wtBLV animals and was (2.4 ± 0.2) × 10⁹/liter (46% of the total lymphocytes), while the mean B-cell count of the pppBLV-transfected animals was (1.1 ± 0.2) × 10⁹/liter (27% of the total lymphocytes) [(1.0 ± 0.7) × 10⁹/liter and (1.2 ± 0.6) × 10⁹/liter for pppBLVtr and pppBLVlr, respectively], and the mean B-cell count in the negative control group was (1.2 ± 0.2) × 10⁹/liter (28% of the total lymphocytes). Simultaneous lack of significant differences in absolute T-cell numbers among groups and decreased T-cell percentage in the wtBLV group confirmed that the elevation in total numbers of PBMCs in the wtBLV group was due to increased numbers of B cells. No significant differences were found in the absolute numbers of CD4⁺ T cells, CD8⁺ T cells, or WC1-N2⁺ cells. Therefore, the mean percentage of T cells, CD4⁺ T cells, and CD8⁺ T cells decreased (34 versus 48%, 20 versus 26%, and 11 versus 18%, respectively) in the wtBLV-transfected group compared with the pppBLV-transfected group. In contrast to the wtBLV-infected group of sheep, no differences between any lymphocyte subpopulations in the overall pppBLV group and the negative control group were evidenced (Table 2). Phenotyping of cells within the pppBLV-infected group of sheep revealed differences between the low-titer-producing group (sheep numbers 5 and 7) and high-titer group (numbers 8, 9, and 10). In particular, significant differences were observed between percentage of B-B2 and WC1-N2 antigen-bearing cells. The percentage of B-B2 B cells was lower (mean, 22 versus 30%) in the low-titer-producing group (sheep 5 and 7) than in the high-titer group (sheep numbers 8, 9, and 10). Simultaneously, the low-titer-producing group of sheep showed an increase in percentage of WC1-N2 lymphocytes compared with that of the high-titer group (15 versus 11%) (data not shown).

Lack of reversions and/or transmission of pppBLV mutant. The sequence of the viral proline-rich motif was examined 7 months after transfection to verify that in vivo reversions or sheep-to-sheep transmission had not occurred (Fig. 6). DNA from PBMCs was extracted, amplified by PCR, and sequenced. The sequence of gp30 in the five pppBLV-transfected animals was consistent with the original mutated and transfected sequence. At the mutated sites, peaks were homogeneous, consistent with a pure population of mutated virus. To verify that there were no reversions at later times after transfection, the proviral gp30 sequence of sheep 9 and 10 was determined at 12 and 24 months posttransfection. The sequences were identical with that of the original mutant pppBLV used for transfection.

View larger version (32K): [in this window] [in a new window]   FIG. 6.   Sequence of provirus DNA amplified from samples collected 7 months after transfection of experimental sheep

Progression to neoplasia. One wtBLV-transfected sheep (number 33) died of unrelated causes at 10 months posttransfection. This animal became listless and recumbent and was euthanatized, but a specific cause of death could not be determined. All four of the remaining wtBLV-transfected sheep died of leukemia and lymphoma within 31 months posttransfection. Sheep numbers 4 and 11 at the time of death showed marked lymphocytosis (~5 × 10¹¹ to 7 × 10¹¹/liter) as well as solid lymphomas in many organs, including lymph nodes, heart, kidney, and peritoneum. A third sheep (number 6) had a single spike of leukocytosis of up to 10¹²/liter that later subsided, and this sheep later developed solid lymphomas. The fourth sheep (number 32) also developed solid lymphomas and died. None of the pppBLV-transfected sheep or negative control sheep had developed leukemia or lymphoma 34 months after transfection.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the cytoplasmic tail of the BLV transmembrane protein, gp30, a proline-rich motif is located 12 amino acids upstream of the ITAM. Because of the observation that proline-rich motifs with nearly identical spacing of three prolines (PX₍₂₎PX_(4-5)P) are found in close proximity to ITAMs in five other viruses, including four that infect lymphocytes (5), we hypothesized that the proline-rich motif is necessary for a key step in the viral life cycle. To test this hypothesis in vivo, we mutated the first two conserved prolines and a third proline of an infectious molecular clone and transfected the plasmid into sheep.

Initial serologic responses of the pppBLV- and wtBLV-transfected animals were similar, suggesting that the mutation did not cause such dramatic changes in the virus that infectivity was significantly inhibited. Following initial infection, however, sheep transfected with pppBLV failed to maintain high levels of virus. In three of five sheep, semiquantitative PCR showed little or no proviral load in PBMCs, as compared with the wtBLV-transfected animals. In two pppBLV-transfected animals, proviral load in PBMCs was similar to that in the wtBLV-transfected animals. However, p24 expression in cultured PBMCs was less than in the wtBLV-transfected animals, and the serologic titers were lower as well in all of the pppBLV-transfected animals. Consistent with the reduction in proviral load and viral expression, the B-cell population in the pppBLV-transfected animals was similar to the uninfected control animals, unlike the expanded B-cell population of the wtBLV-transfected animals. One wtBLV-transfected animal died of unrelated causes, and all four of the remaining wtBLV-transfected animals developed leukemia and lymphoma. The pppBLV-transfected animals have not developed disease.

It is of interest that sheep transfected with pppBLV reacted differently. Two of the pppBLV-transfected animals were initially infected, seroconverted, and had PBMCs with detectable provirus by PCR. However, after 26 weeks these animals reverted to seronegativity, and no provirus could be detected in the PBMCs by semiquantitative PCR or by nested PCR. A third pppBLV-transfected animal remained seropositive and proviral load was low, as determined by semiquantitative PCR. In contrast, the fourth and fifth animals in this group had semiquantitative PCR levels similar to the wtBLV-transfected animals. Various mechanisms might explain these findings. Although the gp30 sequence in the pppBLV-transfected animals was identical to the original plasmid, indicating lack of reversion, it is possible that there was a compensatory mutation elsewhere in the viral genome that permitted increased viral growth. It is also possible that the host sheep differed genetically, thus facilitating viral growth in some of the animals. Alternatively, it may be that stochastic events in some mutant-transfected animals shortly after initial transfection lead to greater infection. Regardless of the mechanism, the pppBLV-transfected animals with higher proviral loads, as determined by semiquantitative PCR, did have lower serologic titers than the wtBLV-transfected animals. Moreover, p24 protein expression in the supernatant of cultured cells was lower and the animals were free of tumor development, compared with the wtBLV-transfected animals.

Previously, it has been shown that the ITAM of gp30 is essential for in vivo viral infectivity and maintenance of viral load (29). This study shows, in addition, that the PX₂PX_4-5P proline-rich motif is also important for maintenance of proviral load in vivo. It is intriguing that in viruses with both proline-rich motifs and ITAMs, the two motifs are in close proximity. It is not known what role the relative positions of these two motifs plays, but it is possible that BLV gp30 and the other viral proteins with both motifs act as a signaling scaffold by bringing signaling molecules with SH3 motifs together with molecules with SH2 motifs (12, 16).

Although this in vivo study demonstrates the importance of the proline-rich motif, the mechanisms of action of this motif have not been determined. There is considerable evidence that proline-rich motifs with PX₂PX_4-5P spacing are involved in signal transduction and are recognition sites of the SH3 motifs, which are common in a wide variety of intracellular signaling molecules, including the Src family of tyrosine kinases, Fyn, Lyn, and Hck, Vav, and others (11, 16, 24, 27). A variant of the proline-rich motif contains PY sequences and interacts with WW domains on signaling molecules (16). The PY motif is found in proximity to many of the viruses with ITAMs, but it is not present in BLV gp30 (5).

Functions of the proline-rich motif other than interactions with SH3 domains of signaling molecules also may be utilized in retroviruses. In human immunodeficiency virus type 1 (HIV-1) and other retroviruses, proline-rich motifs are essential in a variety of late processes in the viral life cycle, including efficient release of particles from the cell surface (14), viral maturation (10), and incorporation of the Pol proteins, reverse transcriptase (RT) and integrase, into the virion (10). In HIV-2, a proline-rich motif of the Vpx protein is necessary for nuclear localization of the preintegration complex (19). In other viruses, the P protein of Borna disease virus has two nuclear localization signals consisting of proline-rich motifs (28), and a proline-rich PPPY motif of vesicular stomatitis virus is necessary for a late step of virus release (15). Further studies will be needed to investigate if there are specific signaling defects in SH3-mediated signaling pathways or in other aspects of the viral life cycle in pppBLV-infected lymphocytes.

BLV is evolutionarily and biologically similar to HTLV-1 and -2. The value of the BLV model system is that viral mutations can be generated and tested in experimental animals. This approach can examine the in vivo significance of different putative viral signaling motifs and their interactions with host signal pathways. Identification of motifs that cause expansion of B-cell populations can lead to understanding of specific host signaling pathways that are altered and may be significant for cancer research and targeted drug development.