Search for a command to run...
Human anelloviruses (hAVs), first identified in 1997, are small, single-stranded, negative-sense circular DNA viruses of the Anelloviridae family [1, 2]. The three major genera—torque teno virus (TTV), torque teno mini virus (TTMV), and torque teno midi virus (TTMDV)—are ubiquitous in humans and generally considered symbiotic. Despite conserved genomic architecture, their genome sizes differ: TTV 3.5–3.9 kb, TTMV 2.7–2.9 kb, and TTMDV ~3.2 kb [2]. Although associations with human diseases have been proposed, the pathogenic mechanisms of hAVs remain largely unknown [3, 4]. In November 2021, Astolfi et al. reported that TTMV drives a subtype of atypical acute promyelocytic leukemia (APL) through genomic integration and fusion with the retinoic acid receptor α (RARA) gene, forming the TTMV::RARA fusion [5]. To date, this remains the only definitive molecular evidence linking anelloviruses to human oncogenesis. Since then, additional cases of TTMV::RARA-APL have been reported, including our recent analysis of 29 cases [6-9]. Here, we further report that TTV, another Anelloviridae genus, can also integrate into the human genome to form TTV::RARA fusion, giving rise to TTV::RARA-APL. While sharing features with TTMV::RARA-APL, this entity displays distinct clinical and molecular characteristics, representing a novel subtype within the APL spectrum. Following its initial description, the Lu Daopei Molecular Medicine team incorporated TTMV::RARA analysis into the study framework for PML::RARA-negative atypical APL and led a nationwide collaborative effort [9]. Given the shared genomic features across Anelloviridae genera, the identification of TTMV::RARA facilitated the recognition of TTV::RARA. Over the past 3 years, we identified 22 TTMV::RARA-APL cases [9] and 2 TTV::RARA-APL cases. The study was approved by the Ethics Committee of Hebei Yanda Lu Daopei Hospital, and written informed consent was obtained from all participants. Clinical, laboratory, and genomic data of these two TTV::RARA-APL cases were collected. Whole-genome sequencing (WGS), whole-transcriptome sequencing (WTS), and genetic profiling of diagnostic bone marrow (BM) samples were performed as previously described [9]. Sequencing reads were aligned to the GRCh37/hg19 human reference genome. Reference TTMV/TTV strains, representing defined genera, were obtained from the International Committee on Taxonomy of Viruses (Table S1) [2]. All cellular and protein assays followed established protocols [9] and are detailed in the Supporting Information Methods. Clinical and laboratory data are summarized in Tables S2 and S3. Both patients were female, aged 56 and 79 years. In contrast, the 29 reported TTMV::RARA-APL cases were predominantly pediatric (ages 2–49 years; median 8 years) [9], reflecting a significant age-of-onset difference (p = 0.002; Table S4). Despite higher prevalence and viral load of TTV compared to TTMV in the general population [2-4], TTV::RARA-APL remains rarer, suggesting additional restrictions for TTV-driven leukemogenesis. Case 1 presented with dizziness and pancytopenia, while Case 2 presented with fever and thrombocytopenia. Both exhibited laboratory evidence of consumptive coagulopathy. BM morphology revealed abnormal promyelocytes with Auer rods (Figure 1A,B), and flow cytometry demonstrated APL-like immunophenotypes in both cases. Notably, Auer rods are uncommon in TTMV::RARA-APL [9]. Given their APL-like features, both patients promptly received all-trans retinoic acid (ATRA) plus arsenic trioxide (ATO) therapy according to the standard-risk APL treatment protocol (Figure 1C) [10]. In Case 1, BM evaluation at Week 3 showed 6% abnormal promyelocytes. Despite negative PML::RARA results by reverse transcription PCR (RT-PCR) and fluorescence in situ hybridization (FISH), therapy was continued based on clinical response. The patient achieved complete remission by Week 5, which was sustained at the 18-month follow-up. In Case 2, white blood cell count rose to 41.7 × 109/L by Day 5. Given negative PML::RARA results and apparent clinical deterioration, treatment was discontinued, and the patient died 25 days postdiagnosis. Integrated viral and fusion sequences are summarized and annotated in Supporting Information Result 1. WGS analysis identified TTV-derived integration at RARA exon 2 (Case 1, 2415 bp) and intron 2 (Case 2, 1138 bp) (Figure S1A,B), unlike the intron 2-only pattern observed in TTMV::RARA-APL [9]. Integration near the 3′ end of exon 2 is also functionally plausible for generating an in-frame fusion with RARA exon 3. Both integration sites exhibited microdeletions and virus–host microhomology sequences (Figure S1C), consistent with microhomology-mediated recombination observed in TTMV::RARA-APL [9]. WTS analysis of Case 2 revealed abundant TTV::RARA transcripts containing an 11-bp insertion from RARA intron 2 (Figure S1D). WTS was not feasible for Case 1 due to limited RNA; the fusion transcript was confirmed by RT-PCR and Sanger sequencing (Figure S1E, Supporting Information Result 2). In both cases, the fusion transcripts comprised the 5′ region of TTV open reading frame 2 (ORF2) fused in-frame to RARA exon 3, with intervening sequences derived from RARA intron 2 or exon 2, mirroring the structure of TTMV::RARA fusions. WTS clustering placed Case 2 with TTMV::RARA-APL [9] and PML::RARA-APL [12, 13], revealing a shared transcriptional profile (Figure S1F). Targeted sequencing [9] identified WT1 c.640C>T/p.Q214*, GATA2 c.1187G>A/p.R396Q, ARID1B c.1223_1234del12/p.G408_G411del mutations in Case 1; and WT1 c.1074delA/p.Q358Hfs*22, KANSL1 c.1273C>T/p.Q425*, and STAT5A c.1226C>T/p.T409M mutations in Case 2 (Supporting Information Result 3). The mutational landscape resembled that of TTMV::RARA-APL [9]. Notably, both cases harbored truncating WT1 mutations, warranting further investigation. Mapped integration and expression regions against the TTV reference genome are shown in Figure 1D. Building on prior findings in TTMV::RARA-APL [9] and known hAVs genomic features, we delineated the characteristics of TTV::RARA integration and fusion transcripts (Figure 1E). Similar to TTMV::RARA-APL, integrated TTV fragments contained a canonical TATA-box and regulatory elements consistent with RNA polymerase II-mediated transcription. The ~660-bp core integration region encompasses the 5′ portion of ORF2 and the transcriptional regulatory region. Unlike TTMV, the transcriptional regulatory region of TTV extends beyond the coordinate origin and spans the GC-rich region of the viral genome, consistent with differences in their genomic architecture. Fusion transcript splicing utilized a derived donor site from one of three sources: native RARA intron 2 (Case 1, Figure S1A), integrated TTV sequence, or adjacent RARA intron segments (Case 2, Figure S1B). Phylogenetic analysis of ORF2 proteins from the two TTV::RARA-APL cases and reference strains placed them closest to Alphatorquevirus homin29 (Figure 1F). To evaluate the epidemiologic distribution, 3181 TTV strains from febrile Tanzanian children [11] were mapped onto the phylogenetic tree, revealing a 6% combined prevalence of strains most homologous to these two cases. Although TTV species are typically classified based on ORF1 nucleotide identity with a threshold of ≥ 69% [2], the core integration region in TTV::RARA-APL involves the 5′ region of ORF2 and the transcriptional regulatory region. Accordingly, we performed ORF2-based nucleotide analysis (Figure 1G), showing that APL-associated strains are closely related to, or fall within, Alphatorquevirus homin29 taxonomically. Combined with the phylogenetic results, these findings indicate that leukemogenic potential is confined to specific TTV subsets. Transcription factor prediction [9, 14] identified PU.1—but not RUNX1—binding motifs in the integrated transcriptional regulatory regions (Figure 1H), in contrast to TTMV::RARA-APL. PU.1 binding was further supported by AlphaFold 3 modeling [15], dual-luciferase, and electrophoretic mobility shift assays (Figure S2A–C). These findings indicate distinct regulatory mechanisms: TTMV::RARA is RUNX1-driven, whereas TTV::RARA is PU.1-driven, with both transcription factors being active in myeloid progenitors. The TTV::RARA fusion protein combines the N-terminus of TTV-ORF2 with the DNA- and ligand-binding domains of RARA. Alignment of ORF2 N-termini revealed two residues, I41 and P46, that form a distinctive IX4P motif, shared across all TTV::RARA- and TTMV::RARA-APL cases (Figure 2A). This motif is highly variable in TTV/TTMV reference strains, suggesting its functional relevance in leukemogenesis. The restriction of this motif, consistent with phylogenetic and homology analyses, underscores the confinement of leukemogenic strains and may primarily account for the rarity of anellovirus-associated APL. Analytical ultracentrifugation assays confirmed that TTV-ORF2 assembles into homotetramers (Figure 2B), whereas TTMV-ORF2 forms homotrimers [9]. As observed in TTMV::RARA [9], TTV-ORF2 localized to the cytoplasm, whereas TTV::RARA accumulated in the nucleus (Figure S3A). Co-immunoprecipitation confirmed that TTV::RARA formed both homodimers and heterodimers with retinoid X receptor α (RXRA), co-localizing with RXRA in the nucleus (Figure S3B, C). Two TTV-ORF2-IX4P motif mutants were tested: I41L remained cytoplasmic, whereas P46G exhibited mixed nuclear–cytoplasmic distribution (Figure S3D). ATRA treatment induced marked degradation of TTV::RARA by Western blot, whereas ATO had minimal effect (Figure 2C). Luciferase assays demonstrated attenuated yet dose-dependent ATRA-induced transcriptional activation for TTV::RARA, TTMV::RARA, and PML::RARA relative to wild-type RARA, with TTV::RARA exhibiting the highest responsiveness among fusion proteins (Figure 2D). Clinically, Case 1 achieved durable remission with ATRA plus ATO. Case 2 developed leukocytosis 5 days after treatment initiation, possibly reflecting differentiation induction. Together, the clinical courses and in vitro data strongly indicate that TTV::RARA-APL responds favorably to ATRA-based therapy. The markedly divergent clinical courses of the two cases further underscore the importance of timely and accurate molecular diagnosis. This study reports the first identification and comprehensive characterization of TTV::RARA-APL. While sharing features with TTMV::RARA-APL, these cases differ in incidence, age of onset, and virologic, molecular, and clinical characteristics (Table 1), establishing TTV::RARA-APL as a novel APL subtype with favorable ATRA sensitivity. Our findings expand the spectrum of anellovirus-associated APL to include both TTV::RARA-APL and TTMV::RARA-APL. Importantly, identification of the signature ORF2-IX4P motif shared across anellovirus-associated APL highlights a potential target for mechanistic and therapeutic investigation. This work also broadens the paradigm of virus-mediated malignancies. H.L. and X.Z. conceptualized and designed the study and wrote the manuscript. J.Z. and Z.Z. provided valuable discussion and direction. Q.M, H.C., Y.C., H.W. and G.O. provided study materials or recruited patient. X.Z., Y.Z., P.C., Xiaoli M., J.C., X.C., J.F., Q.C., H.S., Xiujuan M. and F.W. were responsible for laboratory experiments and bioinformatic analysis. All authors reviewed the results and approved the final version of the manuscript. The authors acknowledge the patients and their families and the referring physicians who made this study possible. The authors thank Na Wei and Shuang Sun for their contributions to the mechanism diagram artwork. The authors also thank Siyuan Liu for his help in bioinformatic analysis and discussions. This work was supported by grants from the Natural Science Foundation of China (82570228) and Hebei Provincial Medical Science Research Projects (20261157 and 20261158). This study was approved by the Ethics Committee of Hebei Yanda Lu Daopei Hospital. The authors declare no conflicts of interest. For original data, please contact [email protected] on reasonable request. Data S1: ajh70276-sup-0001-Supinfo.docx. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.