Introduction

Bovine leukemia virus (BLV) belongs to the family Retroviridae and genus Deltaretrovirus. BLV infection is characterised by lymphoproliferative disorders in cattle, with persistent lymphocytosis and development of lymphosarcoma in 1% to 5% of infected animals older than four years [18]. This virus is distributed worldwide and has had considerable economic impact on the cattle industry. [4, 6, 9, 21]. BLV infection has been reported with a high seroprevalence at some state farms in Turkey [3, 4]. Although no eradication programme exists at the national level, a study [4] on the control and eradication programme carried out on state farms has been conducted; monitoring of the herds from this study is ongoing (unpublished data).

The complete genome of BLV consists of a sequence of 8714 nucleotides and can be anatomised in the form of 5’ LTR-gag-pol-env-px-LTR 3’ [7, 18, 22]. The gag gene is highly conserved and consists of a sequence of 1178 nucleotides. It is apparent that there are two different peptide sequences preferred for p15 and p24. The BLV env gene (consisting of a sequence of 1493 nucleotides) encodes one polyprotein precursor (gp 72), which is cleaved to gp 51 (an immunodominant protein), the target for neutralising antibodies, and gp 30 (a transmembrane protein), which is responsible for enabling the virus to enter cells [15]. Several authors [1, 15, 21] have analysed nucleotide sequences from env and gag genes in BLV-positive samples in different geographical areas and reported that different BLV variants were found in different geographical regions. No study has molecularly characterised BLVs from Turkish samples. The presence and seroprevalence of the infection has been reported in several studies [2–4, 25].

In this study, we attempt to partially molecularly characterise env and gag genetic sequences from BLVs obtained from eleven naturally infected animals from four herds located in three different geographic regions. This design should help build understanding of the epidemiology of BLV infection in Turkey.

Materials and methods

Samples and history

Blood samples from cattle from four different herds (one of which included cattle imported from Uruguay) located in three geographic regions in Turkey were tested for antibodies to BLV using a commercially available enzyme-linked immunosorbent assay (ELISA) (Institut Pourquier, Montpellier, France) that identified antibodies against BLV gp51. Seropositivity rates ranged from 0.3% to 6.0% in these herds (unpublished data). The next step was to analyse two samples for each of three randomly selected herds, and five samples for the imported cattle. These herds were tested using a BLV-specific nested polymerase chain reaction (PCR) assay; the gag and env genes of BLV in these herds were molecularly characterised. For this purpose, the expected 444-bp and 380-bp fragments were amplified for all blood samples (n=11); these fragments corresponded to part of the env and gag genes, respectively, of BLV. Amplicons were sequenced for molecular characterisation of these field viruses.

Extraction of proviral DNA

Proviral DNA was extracted from whole blood and serum samples using a High Pure Viral Nucleic Acid Kit (Roche Applied Science, catalog no. 11858874001, Mannheim, Germany) according to the manufacturer’s recommendations.

DNA amplifications for nested PCR

The following primer sets were used in this work:

  • For env gene amplifications, env5032 and env5608 were used as outer primers (600-bp fragment); env5099 and env5521 were used as inner primers (444-bp fragment from proviral BLV) [17, 22], and PCR was done as described elsewhere [8].

  • For gag gene amplications, the pair of amplification primers gag3 and gag4 was used (380-bp fragment) [22], and PCR was done as described elsewhere [23].

In order to visualise nested PCR products, 3 μL of each amplified product was run on a 1.0% agarose gel, followed by ethidium bromide staining.

Sequencing and phylogeny

The nested PCR products were purified using a High Pure PCR Product Purification Kit (Roche Applied Sciences, Germany) and sequenced using an automatic sequence analyser (CEQ 8000; Beckman Coulter, Brea, CA, United States).

Sequence editing and multiple alignments were performed with the publicly available Bioedit software package version 7.0.9.0 [12]. Phylogenetic analysis (neighbour-joining) with bootstrap analysis (1000 replicates, 111 random seeds) and Kimura 2-parameter correction was conducted using the MEGA (Molecular Evolutionary Genetics Analysis) software package version 4.0 [24].

Results

The nested PCR results for BLV based on gag and env genes obtained from blood samples in 11 animals that were BLV-positive by ELISA revealed DNA products of the expected sizes (380 bp for the gag gene and 444 bp for the env gene). The partial nucleotide sequences of the BLV env and gag genes from Turkish cattle were compared to the corresponding sequences from other countries with BLV samples deposited in GenBank (Figs. 1 and 2). Because gag gene sequences of all of the BLV strains that were used in phylogenetic analysis of the env genes were not available in the GenBank database, we could not compare the sequences of the env and gag genes of the same BLVs. The percentages of nucleic acid identity and divergence for the gag and env genes were estimated. Amino acid sequences deduced for the env and gag genes were also compared and are shown in Fig. 3 and 4, respectively.

Fig. 1
figure 1

Phylogenetic tree based on nucleotide sequences of the env gene of Turkish BLV strains and BLV strains from various countries. The accession numbers of the env gene region sequences (obtained from the GenBank database) used in this study are shown in the phylogenetic tree. Symbols are shown for identical strains from the same herd. Bar = number of base substitutions per site

Full size image
Fig. 2
figure 2

Phylogenetic tree based on nucleotide sequences of the gag gene of Turkish BLV strains and BLV strains from various countries. The accession numbers of the gag gene region sequences (obtained from the GenBank database) used in this study are shown in the phylogenetic tree. The Turkish BLV strains obtained from the same herds are shown with identical symbols in the phylogenetic tree. Bar = number of base substitutions per site

Full size image
Fig. 3
figure 3

Multiple alignment of the deduced amino acid sequences of env genes of Turkish BLV strains with those of BLV strains from various countries. Dots denote amino acid identity with a BLV complete genome (K02120). Stars indicate stop codons

Full size image
Fig. 4
figure 4

Multiple alignment of deduced amino acid sequences of gag genes of Turkish BLV strains with those of BLV strains from various countries. Dots denote amino acid identity with a BLV complete genome (K02120). Stars indicate stop codons

Full size image

GenBank accession numbers FJ009173 to FJ009178 and FJ009179 to FJ009184 were assigned to env and gag genes of BLV field strains (n=6) from cattle representing 3 different domestically bred herds. For the cattle imported from Uruguay, GenBank accession numbers JF894789 to JF894793 and JF894794 to JF894798 were assigned to env and gag genes of BLV field strains (n=5).

Discussion

At present, four complete BLV genomes have been reported worldwide: Belgian isolates [19, 20], Japanese isolates [22], Australian isolates [6] and recently, Argentine isolates [7]. However, partial sequencing of different genes, including env and gag, has been reported in many studies [5, 16, 21] on the molecular characterisation of BLVs from different countries.

In this paper, Turkish BLVs were classified by phylogenetic analysis based on gag and env regions of the viral genome. The proportion of the nucleotides that were similar for the env gene sequences in the BLVs reported in this study ranged from 82.0% to 98.9%, while the same index for the env gene sequences of TR-5044 and others ranged from 82% to 83.9%. Similarly, the divergence rates between the TR-5044 sequence and the sequences of BLV deposited in GenBank from cattle from other countries ranged from 80.6% to 86.6%; these rates were the lowest between the Turkish BLVs and other BLVs deposited in GenBank. On the other hand, nucleotide sequence similarities in the gag gene sequences of BLVs reported in this study ranged from 98.7% to 100%, while those of the env gene sequences were between 82% and 83.7%. For the gag gene sequences, the highest degree of similarity (100%) was found with the Bolv385-Iran-strain (unpublished data).

Comparison of amino acid sequences from the gag and env nucleotide sequences showed some amino acid differences among Turkish BLVs, and also among the Turkish BLVs and BLVs deposited in GenBank from different countries. Analysis of the amino acid composition of env genes of BLV variants used in this study showed some highly conserved regions, as has been reported in other studies [6, 15]. Howewer, two BLVs (FJ009175 and FJ009176) from the same herds showed a higher level of amino acid changes for env genes than did other Turkish BLVs (Fig. 3). Data obtained from comparisons of nucleotide and amino acid sequences of env genes in particular demonstrated the presence of BLV variants in different herds from different geographic regions.

It is known that a single point mutation can appreciably change the amino acid sequence, that conformation changes in enveloped proteins (important for BLV’s ability to bind to cellular receptors) can cause a decrease of antibody-binding capacity, and that some mutant BLV strains are resistant to antibodies. Therefore, detecting BLV-infected cattle by serological tests alone is not sufficient, as some BLV strains will not cause the production of the antibodies detected by these tests. [7, 13, 15]. PCR is the most sensitive and specific assay for the diagnosis of several viruses, including BLV. On the other hand, reports indicate that disagreement between different diagnostic techniques exists on the basis of the multiple genetic backgrounds of field strains, the sensitivity and specificity of the diagnostic techniques, and the diagnostic purpose (antigen, antibody or nucleic acid detection) of the technique [8, 16, 21]. In this study, because the blood samples detected as positive by ELISA were analysed for the purpose of molecular characterisation of BLV strains, it is not possible to speculate about the sensitivity or specificity of these methods for detection of BLV infection. Turkish BLVs detected by PCR have yielded positive results in ELISA tests, although they have genetic/antigenic differencies in the env gene, especially TR-5044. Additionally, it is important to remember that the variable sequences of BLVs, especially in the env gene, are encoded within a single polyprotein precursor (gp 72), which is cleaved to gp 51 (an immunodominant protein). This fact is important for vaccine development.

Data on genetic characterisation showed interaction between the BLV field strain and animal importation [14, 16]. Licursi et al. [14] have reported that most Japanese BLVs are similar to those originating in the United States and Australia, while most Argentine BLVs are related to the European type. These authors also noted that Argentine BLV might have come mainly from Europe a long time ago and developed individually. In Turkey, BLV infection was first detected in imported, highly productive, pedigreed Holstein and Swedish milking cows on a state farm [10, 11]. Many studies have reported the prevalence of BLV in Turkey [2, 25]. In some of these studies, the prevalence of infection has been extremely high on state farms [2–4], probably due to the movement of animals between farms. Additionally, BLV infections have been detected serologically at some private farms [25]. From our data, obtained from an imported herd and three domestic herds, and drawn from different geographic locations, it may be said that our general idea that BLV has been present in Turkey for years, and that the importation of animals has caused the importation of dissimilar types of BLV (as shown by dissimilarity in env and gag genetic sequences), concurs with these other studies [14, 16].

The data presented here emphasize the need for molecular characterisation of BLVs circulating in Turkey. Further studies will depend on the detection and molecular characterisation of the BLVs in cattle from different origins and with different specialties or purposes (local, imported from abroad, beef and milk cows, and so forth), and these should shed light on BLV transmission between herds/countries and on the molecular epidemiology of BLV infection in Turkey.