Chronic stage of samples were collected from a BLV naturally-infected Holstein cow

Blood samples were collected at three time points from a Holstein dairy cow naturally infected by BLV (first: 2017.05.23, second: 2017.11.02, last time: 2018.02.15 Leukemia onset). Table 1 shows the number of White blood cells (WBC) and lymphocytes and BLV proviral load (PVL) in stages I, II, and III, respectively. The lymphocyte count in Stage I was significantly higher than that in Normal cattle, while that in Stage II was subsequently reduced to one-third of the normal value. When Stage III was finally followed to develop lymphoma and death, the lymphocyte count drastically increased. PVL were analyzed by BLV-CoCoMo-qPCR-2 and were 64,554, 4463, and 83,163 copies per 105 cells in Stages I, II, and III, respectively, in the same manner as lymphocyte counts. Thus, the three stages from the same cattle were classified according to the EC-leukosis key [16]: stage I, BLV-infected but clinically normal cattle with lymphocytosis; Stage II, BLV-infected but clinically normal cattle with a decreased lymphocyte count; and Stage III, BLV-infected with lymphoma (Table 1).

Table 1 Clinical information used in this study Results of BLV proviral DNA-capture-seq

We clarified the BLV integration sites from Stages I–III using BLV proviral DNA-capture-seq, which is a target-enrichment high-throughput sequencing system for characterization of BLV integration sites [14]. The DNAs molecules were fragmented to an average length of approximately 600 bp to generate DNA libraries. Because virus-host chimeric reads generally contain the BLV long terminal repeat (LTR) sequence and BLV LTR contains 531 nucleotides, the BLV probes used in this study (Additional file 1) would correctly capture the virus-host chimeric fragments. After enrichment of the BLV genome, the prepared library was analyzed using an Illumina MiSeq sequencing system. Paired-end reads were first aligned against the BLV reference FLK-BLV (EF600696), mapped to the cow host reference genome (Bos_taurus_UMD_3.1/bosTau6), and visualized using Integrative genomics viewer (IGV) (Additional file 2). We obtained 4,399,364, 3,730,443, and 4,343,678 reads from Stage I, II, and III DNA, respectively. Of these reads, 3,544,097 (80.5%), 3,042,346 (81.6%), and 3,499,444 (80.6%) were aligned to the BLV provirus LTR, gag, or tax regions. These data indicate that target enrichment high-throughput sequencing was successful in all samples.

Confirmation of the Integration Sites

BLV was integrated at 16, 9, and 2 distinct integration sites into the host genome in Stages I, II, and III, respectively (Fig. 1A). We observed 6 or 7 bp of short duplicated host genome sequences generated during integration around the BLV provirus integration in all 24 integrated sites (red in Fig. 1A), as visualized using IGV. Furthermore, we determined the strand orientation of the BLV genome to the host genome for all 24 integrated sites (Fig. 1A). To confirm the integration sites detected in the IGV profile, we selected three major integration sites in Stage I [Chromosome (Chr) 8], Stage II (Chr 8 and Chr 1), and Stage III (Chr 1 and Chr 17), and subjected them to PCR using primers set for the host genome and BLV genome (Additional file 1).

Fig. 1

Change of the BLV integration sites during the course of disease progression in a BLV naturally-infected cow. A Summary of the integration sites detected in Stage I, Stage II, and Stage III, respectively. 1Chr, Chromosome. 2D.P, Duplicated Nucleotide. 3L, left read; R, right read; red color, duplicated nucleotide sequence; upper case, nucleotide sequence of the cattle; lower case, nucleotide sequence of the BLV; bold letter, different sequence compared to the reference sequence; –, deleted sequence compared to the reference sequence. 4CAPN7, calpain 7; ACER2, alkaline ceramidase 2; DCAF12, DDB1 and CUL4 associated factor 12; TMEM231, transmembrane protein 231; NPR3, natriuretic peptide receptor 3; FHIT, fragile histidine triad diadenosine triphosphatase; ASPHD1, aspartate beta-hydroxylase domain containing 1; SGMS1, sphingomyelin synthase 1; DNMBP, dynamin binding protein; POGK, pogo transposable element derived with KRAB domain; FAM78B, family with sequence similarity 78 member B; SF3B5, splicing factor 3b subunit 5; UTRN, utrophin; NDUFAF6, NADH:ubiquinone oxidoreductase complex assembly factor 6; TP53INP1, tumor protein p53 inducible nuclear protein 1; PGR, progesterone receptor; CCDC82, coiled-coil domain containing 82; COA6, cytochrome c oxidase assembly factor 6; TOMM20, translocase of outer mitochondrial membrane 20; NBEAL1, neurobeachin like 1; CHEK2, checkpoint kinase 2; CLRN1, clarin 1; SIAH2, siah E3 ubiquitin protein ligase 2; REEP6, receptor accessory protein 6; MBD3, methyl-CpG binding domain protein 3; ABCA1, ATP binding cassette subfamily A member 1; FSD1L, fibronectin type III and SPRY domain containing 1 like; TREM1, triggering receptor expressed on myeloid cells 1; FOXP4, forkhead box P4. 5Low Complexity, Low complexity repeats; SINE, Short interspersed nuclear elements; DNA, DNA repeat elements; LTR, Long terminal repeat elements. B Pie charts showing the relative abundance of each sample at various disease progression from a BLV-infected cow in Stages I, II, and III, respectively. The numbers of integration site represent each of the detected clone in Fig. 1 were described. State of clonal expansion of cells carrying the integration site and massive depletion of cells carrying the integration site, and clinical stage of disease progression of Stages I, II, and III from one BLV-infected cow, respectively, indicated

The IGV profile mapped to the bovine Chr 17 genome is shown in Fig. 2A. In the case of Chr 17, the integration site was detected in Stages II and III only. Six nucleotide duplications of the host sequence (CATTTC) were detected in the IGV profile. Figure 2B shows a schematic diagram to detect both the Chr 17/BLV-3′LTR chimeric fragment and the BLV-5′LTR/Chr 17 chimeric fragment. Figure 2C, D show the results of PCR amplification on Chr 17. Following PCR analysis, we obtained PCR products of 181 bp and 207 bp in Stages II and III, but not in Stage I (Fig. 2C, D). After sequencing both the 5′ and 3′ sites, we confirmed six nucleotides of the host sequence (CATTTC) (Fig. 2C, D). In addition, to confirm the integration sites with greater accuracy, we performed inverse PCR using Stage III DNA and obtained the same results as the IGV profile, successfully identified one integration site on Chr 17, and observed six nucleotides of the host sequence (CATTTC) directly flanking the 5′ and 3′ ends of the BLV provirus (data not shown).

Fig. 2

Visualization of the integration site identified on Chr 17. A Visualization of NGS reads detected on IGV profile. B Schematic diagram to detect both Chr 17/BLV-3′LTR chimeric fragment and BLV-5′LTR/Chr 17 chimeric fragment. Arrows indicate the position of primers. C Confirmation of the Chr 17/BLV-3′LTR chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the Chr 17/BLV-3′LTR chimeric read. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water substituted for DNA template). Six nucleotides duplication of host sequence (CATTTC) was detected by Sanger sequence. D Confirmation of the BLV- 5′LTR/Chr 17 chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the BLV- 5′LTR/Chr 17 chimeric fragment. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water used as the substitute for DNA template). 6 nucleotides duplication of host sequence (CATTTC) colored in red was detected by Sanger sequence

The IGV profile mapped to the bovine Chr 8 genome is shown in Fig. 3A. In the case of Chr 8, the integration site was detected in Stages I and II, but not in Stage III. Seven nucleotides of the host sequence (AGAATTT), directly flanking the 5′ and 3′ ends of the BLV provirus, were detected in the IGV profile. Figure 3B shows a schematic diagram to detect both the Chr 8/BLV-3′LTR chimeric fragment and the BLV- 5′LTR/Chr 8 chimeric fragment. Figure 3C, D show the results of PCR amplification on Chr. 8. Following PCR analysis, we obtained PCR products of 195 bp and 220 bp in Stage I and Stage II, but not in Stage III (Fig. 3C, D). After sequencing both the 5′ and 3′ sites, we confirmed six nucleotides (GAATTT) (Fig. 3C, D).

Fig. 3

Visualization of the integration site identified on Chr 8. A Visualization of NGS reads detected on IGV profile. B Schematic diagram to detect both Chr 8/BLV-3′LTR chimeric fragment and BLV-5′LTR/Chr 8 chimeric fragment. Arrows indicate the position of primers. C Confirmation of Chr 8/BLV- 3′LTR chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the Chr 8/BLV-3′LTR chimeric fragment. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water substituted for DNA template). Six nucleotides duplication of host sequence (GAATTT) was detected by Sanger sequence. D Confirmation of BLV-5′LTR/Chr 8 chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the BLV- 5′LTR/Chr 8 chimeric fragment. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water substituted for DNA template). 6 nucleotides duplication of host sequence (GAATTT) colored in red was detected by Sanger sequence

The IGV profile mapped to the bovine Chr 1 genome is shown in Fig. 4A. In the case of Chr 1, the integration site was detected in Stages II and III, but not in Stage I, 7 nucleotides duplication of the host sequence (AGTAAAA) was also detected in the IGV profile. Figure 4B shows a schematic diagram to detect both the Chr 1/BLV-5′LTR chimeric fragment and the BLV-3′LTR/Chr 1 chimeric fragment. Following PCR analysis, we obtained PCR products of 156 bp and 193 bp in Stages II and Stage III, but not in Stage I (Fig. 4C, D). After sequencing both the 5′ and 3′ sites, we confirmed six nucleotides (GTAAAA) (Fig. 4C, D).

Fig. 4

Visualization of the integration site identified on Chr 1. A Visualization of NGS reads detected on IGV profile. B Schematic diagram to detect both Chr 1/BLV-5′LTR chimeric fragment and BLV-3′LTR/Chr 1 chimeric fragment. Arrows indicate the position of primers. C Confirmation of the Chr 1/BLV-5′LTR chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the Chr 1/BLV-5′LTR chimeric read. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water substituted for DNA template). Six nucleotides duplication of host sequence (GTAAAA) was detected by Sanger sequence. D Confirmation of the BLV-3′LTR/Chr 1 chimeric fragment using conventional PCR and Sanger sequence. Arrow indicates the PCR product of the BLV-3′LTR/Chr 1 chimeric fragment. M, 100 bp DNA Ladder marker (MIXELL Inc, Tokyo, Japan); 1, Stage I; 2, Stage II; 3, Stage III; C, no-template DNA-negative control (water substituted for DNA template). 6 nucleotides duplication of host sequence (GTAAAA) colored in red was detected by Sanger sequence

Clonal expansion and massive depletion of cells carrying distinct BLV Integration Sites

Figure 1A shows a summary of BLV integration sites in Stages I, II, and III, and Fig. 1B shows their pie charts.

In Stage I, 16 distinct proviral integration sites were detected. The integration sites in Stage I were the most abundant among the three stages. Therefore, Stage I was characterized as a polyclonal stage (Fig. 1B). These integration sites were mapped to 12 autosomes and X-chromosomes. Multiple integration sites were mapped to three autosomal sites (chr 8, 14, and 26). Interestingly, 9 (56%) out of 16 distinct integration sites were located within introns of reference sequence (Refseq) genes, such as calpain 7 (CAPN7), alkaline ceramidase 2 (ACER2), DDB1 and CUL4 associated factor 12 (DCAF12), transmembrane protein 231 (TMEM231), natriuretic peptide receptor 3 (NPR3), fragile histidine triad diadenosine triphosphatase (FHIT), aspartate beta-hydroxylase domain containing 1 (ASPHD1), sphingomyelin synthase 1 (SGMS1), and dynamin binding protein (DNMBP). The remaining five (31%) and two (13%) sites were located in intergenic regions and within repetitive sequences, such as low-complexity repeats and short interspersed nuclear elements (SINE). In 15 out of 16 integration sites, the total number of reads was 20 or less, and they subsequently disappeared in Stage II in the later 6 months. In contrast, only one clone in which BLV integration occurred within the intron of ACER2 gene on Chr 8 and the total number of 60 reads was able to survive until Stage II; however, the read depth of this integration decreased in Stage II and finally disappeared in Stage III.

In Stage II, we identified nine distinct proviral integration sites. Among them, eight integration sites that were not identified in Stage I were newly detected, unlike one site that was inserted in the intron of ACER2 already identified in Stage I. Contrastingly, 15 integration sites were present in Stage I and were not detected in Stage II, indicating a massive depletion of 15 BLV-infected clones that occurred for 163 days during Stages I and II. Collectively, the lymphocyte count and PVL in Stage II were considerably reduced to one-fourth and one-fifteenth of that in Stage I, respectively. Therefore, stage II clonality was characterized as polyclonal towards the oligoclonal stage, as shown in Fig. 1B. Multiple integration sites were mapped to an autosomal chromosome (chr 8). In seven (78%) out of nine integration sites, the total number of reads was 20 or less and disappeared in Stage III. Among these nine distinct integration sites, three (33%) were located within the introns of Refseq genes, such as neurobeachin-like 1 (NBEAL1) and checkpoint kinase 2 (CHEK2), including ACER2 gene. The remaining four (44%) and two (22%) sites were located in intergenic regions and within repetitive sequences, such as DNA repeat elements (DNA) and LTR, respectively. The most common integration site was Chr 1; BLV was integrated into the intergenic region downstream of clarin 1 (CLRN1) and upstream of siah E3 ubiquitin protein ligase 2 (SIAH2). The read depth of this integration was 774, which rapidly increased in Stage III. Another integration site within the intron of CKEK2 on Chr 17 was detected in Stage II.

As shown in Table 1, at Stage III, the test subject was diagnosed with lymphoma. Seven integration sites that were present in Stage II, could not be detected in Stage III, indicating that seven BLV-infected clones were eliminated for approximately 3 months during Stage II and Stage III. In contrast, we identified only two distinct proviral integration sites that were generated first at Stage II, indicating that the surviving clones carrying the integration site on Chr 1 and in the intron of CHEK2 evaded massive clone selection from Stage II to Stage III and provoked polyclonal expansion. The read depth of the two integrations rapidly increased as the total number of reads reached 2845 in Stage III. Another integration site within the intron of CHEK2 on Chr 17 was detected in stage II. Although the total number of reads of the integration site within the intron of CHEK2 on Chr 17 was only nine in Stage II, it increased by 78-fold in Stage III.