The study of the human virome in health and disease has gained growing attention over recent years (Webb et al., 2020; Dodi et al., 2021). Viruses belonging to Anelloviridae family are the most abundant eukaryotic viruses in the virome and may be detected in a variety of samples, such as blood, plasma, urine or saliva (Kaczorowska and van der Hoek, 2020; Arze et al., 2021). Anelloviruses are non-enveloped viruses with small circular replication-associated protein-encoding single-stranded DNA genomes (Biagini, 2009; Kaczorowska and van der Hoek, 2020), which lack attributable pathogenic roles (“orphan viruses”) (Focosi et al., 2016; Rezahosseini et al., 2019). Once primary infection occurs at early stages of life, anelloviruses remain in different body compartments and fluids—including peripheral blood mononuclear cells, feces, semen, throat swabs, umbilical cord blood, lungs, kidneys or cerebrospinal fluid—under the control of the immune system, resulting in a prevalence as high as 90% in the adult population (Redondo et al., 2022a). The precise underlying mechanisms on how this immune control is carry out largely remain to be determined, although a major role has been proposed for the cellular arm. Belonging to the Alphatorquevirus genus and discovered in 1997 (Nishizawa et al., 1997), Torque teno virus (TTV) has been proven by us and others to serve as a convenient surrogate marker of the overall status of immunosuppression after solid organ (SOT) and allogeneic hematopoietic stem cell transplantation (HSCT) (Fernandez-Ruiz et al., 2019; Rezahosseini et al., 2019; Mouton et al., 2020; Redondo et al., 2022a; Jaksch et al., 2022).

The innate immunity acts as a frontline defense against viruses through an orchestrated response, that is, triggered upon recognition of viral motifs by pathogen recognition receptors (PRRs) present in macrophages and dendritic cells (Takeuchi and Akira, 2010). The rationale for the use of TTV DNAemia as a biomarker of immune competence after SOT lies on the assumption that the viral kinetics is mainly dictated by the T-cell-mediated immunity (Redondo et al., 2022a; Jaksch et al., 2022). Indeed, various studies have shown a direct correlation between TTV DNA loads and calcineurin inhibitors trough levels (Gorzer et al., 2014; Jaksch et al., 2018). The role played by the innate immune arm in the setting of ongoing immunosuppression remains largely unknown, as is the potential impact of polymorphisms in genes coding for PRRs (such as toll-like receptors [TLRs]), interleukins (IL) or interferons (IFNs) (Prasetyo et al., 2015; Ramzi et al., 2019; Ramzi et al., 2021). Evidence of an individual genetic susceptibility to TTV regardless of the amount of immunosuppressive therapy would question the reliability of viral replication as clinical biomarker in the SOT population.

We aimed to investigate the association between 14 single genetic polymorphisms (SNPs) in different genes mainly involved in the orchestration of innate immune responses (Table 1) and TTV DNA levels at baseline and various points during the first post-transplant year in a well characterized cohort of kidney transplant (KT) recipients (Fernandez-Ruiz et al., 2019). The selection of these SNPs was dictated by previous research showing a potential impact on the susceptibility to viral infections. In the case of TLR3 (rs3775291), various pieces of evidence have shown an effect on the incidence of infection by cytomegalovirus (CMV) or BK polyomavirus (BKPyV), two relevant viral pathogens in the KT scenario, but also tick-borne encephalitis, chikungunya or hepatitis B virus (HBV) (Kindberg et al., 2011; Geng et al., 2016; Studzinska et al., 2017; Fischer et al., 2018; Bucardo et al., 2021; Redondo et al., 2022b). We have previously reported that certain SNPs in TLR9 (rs5743836, rs352139) modulate the risk of CMV infection in two independent cohorts of KT recipients (Fernandez-Ruiz et al., 2015; Redondo et al., 2022b). Regarding SNPs located in the CD209 gene, rs735240 appears to increase the incidence of CMV infection in seropositive KT recipients not receiving antiviral prophylaxis (Fernandez-Ruiz et al., 2015), whereas rs4804803 has been correlated with an increased susceptibility to dengue virus (Vargas-Castillo et al., 2018) and, more recently, BKPyV (Redondo et al., 2022c). We analyzed the SNPs located in IFNL4 (rs12979860, rs8099917) due to its well-established relevance in other viral infections, including CMV (Fernandez-Ruiz et al., 2015) and hepatitis C virus (HCV) (Ge et al., 2009; Thomas et al., 2009). Finally, we aimed to validate in the SOT population the associations reported by other authors between CTLA4 (rs5742909, rs231775), TNF (rs1800629), IL10 (rs1800872, rs1878672), IL12B (rs3212227) and IL17 (rs2275913) SNPs and the kinetics of TTV DNAemia after HSCT (Ramzi et al., 2019; Ramzi et al., 2021).

We included 221 KT recipients, whose demographics, clinical characteristics and patient and graft outcomes are detailed in Table 2. Samples from all the patients were successfully genotyped for the 14 SNPs considered. The median number of assessments for plasma TTV DNA per patient was 5 (IQR: 4–5). The majority of recipients had detectable TTV DNAemia (i.e. above the LLoD) at every time point, ranging from 96.3% (180/187) at baseline to 99.4% (176/177) at post-transplant month 6. The genotypic frequencies of candidate SNPs are shown in Supplementary Table S1. The observed genotype frequency distributions did not deviate from those expected according to the Hardy-Weinberg equilibrium except for TLR9 (rs5743836) and IFNL3 (rs12979860).

First, the effect of studied polymorphisms on plasma TTV DNAemia at discrete time points was investigated. In particular, we explored the impact of the minor alleles in each SNP in both dominant (heterozygous and homozygous) and recessive (homozygous only) models. Across the 14 SNPs considered, we did not find significant differences in TTV DNA levels at any of the monitoring points (Table 3).

Next, we analyzed if there was any association between candidate SNPs and the presence of undectectable plasma TTV DNAemia at baseline (before the initiation of immunosuppressive therapy). Seven (3.7%) patients had pre-transplant TTV DNA levels below the LLoD. We observed that carriers of the minor C allele of the IL10 (rs1878672) SNP in the homozygous state (CC) were more likely to have undetectable baseline TTV DNAemia compared to recipients bearing the reference G allele (GG/GC) [12.5% (3/24) versus 2.5% (4/163), respectively; nominal p-value = 0.046]. There were also significant differences within the TLR3 (rs3775291) SNP, since all the 7 patients with undetectable TTV DNAemia harbored the minor T allele either in the heterozygous or the homozygous state [7.1% (7/99) versus 0.0% (0/88) for CT/TT and CC carriers; nominal p-value = 0.015]. Finally, the minor allele of CD209 (rs4804803) in the homozygous state was also associated with undetectable TTV DNAemia before transplantation [37.5% (5/8) versus 2.2% (4/179) for GG and AA/AG carriers; nominal p-value = 0.0017]. Nevertheless, it should be noted that only the latter association was below the Bonferroni-corrected p-value threshold for statistical significance (which was settled at 0.00178) (Table 4). We further assessed whether the impact of the CD209 (rs4804803) SNP remained significant after adjusting for recipient demographics and pre-transplant clinical characteristics also associated with undetectable TTV DNAemia at baseline (Supplementary Table S2). In a logistic regression model that included recipient age and previous renal replacement therapy as covariates, the presence of the minor G allele of CD209 (rs4804803) in the homozygous state was still significantly associated with pre-transplant TTV DNA levels below the LLoD (adjusted OR: 36.96; 95% CI: 4.72–289.67; p-value = 0.001).

In order to better characterize the genetic determinants of post-transplant TTV viral kinetics, we compared peak TTV DNA levels through different time intervals according to candidate SNPs. The only apparent correlation was observed for TNF (rs1800629), with carriers of the minor allele either in the heterozygous or homozygous state showing lower peak levels during the first 3 post-transplant months (4.2 ± 1.5 versus 5.0 ± 1.8 log₁₀ copies/mL for GA/AA and GG carriers; nominal p-value = 0.008) (Supplementary Table S3). This comparison, however, did not attain the Bonferroni-corrected significance level (settled at 0.00059). Accordingly, recipients bearing the minor A allele of this SNP also showed a non-significant trend—by applying the Bonferroni correction—towards a lower AUC for plasma TTV DNAemia through month 6 (5.8 ± 1.7 versus 6.8 ± 1.7 log₁₀ copies/mL for GA/AA and GG carriers; nominal p-value = 0.007) (Supplementary Table S4). Finally, no significant differences at the Bonferroni-adjusted α level were found in increments (Δ) in TTV DNA levels from baseline to day 15 and months 1, 3, 6 and 12 after transplantation either (Supplementary Table S5).

There is increasing evidence on the usefulness of TTV as a surrogate marker of the immune status in a variety of clinical scenarios (Martin-Lopez et al., 2020; Honorato et al., 2021; Studenic et al., 2021), in particular SOT (Fernandez-Ruiz et al., 2019; Redondo et al., 2022a; Eldar-Yedidia et al., 2022; Jaksch et al., 2022), under the rationale that the T-cell-mediated immunity plays an instrumental role in controlling viral replication. Nevertheless, the relative contribution of the innate system—and its genetic determinants—has not been characterized so far. To our knowledge only three previous works have analyzed the impact of genetic polymorphisms on TTV replication in HSCT recipients (with two studies from the same group) and people living with human immunodeficiency virus (HIV) (Prasetyo et al., 2015; Ramzi et al., 2019; Ramzi et al., 2021). Ramzi et al. (2019); Ramzi et al. (2021) found a correlation between SNPs in IL10, CTLA4 and TNF genes and TTV infection in allogeneic HSCT recipients. In detail, the heterozygote genotypes of IL10 rs1800872 (−592C/A) and CTLA4 rs231775 (+49 A/G) were associated with a higher prevalence of TTV DNAemia, whereas the A allele of TNF rs1800629 (−308G/A) had a protective effect. On the other hand, Prasetyo et al. reported a correlation between the APOBEC3B deletion polymorphism status and TTV, HBV and HCV infection among HIV patients (Prasetyo et al., 2015; Ramzi et al., 2019; Ramzi et al., 2021). The present investigation is the first to evaluate to what extent the kinetics of TTV DNA levels following KT are influenced by SNPs in genes coding for PRRs (TLR3, TLR9 and CD209), ILs and cytokines (IL-12B, IL-17, IL-10, TNF), IFN-λ3 (IL-28B) and the costimulatory receptor CTLA-4). These candidate SNPs were chosen on the basis of prior studies performed in the HSCT population (Ramzi et al., 2019; Ramzi et al., 2021) or due to their well-established involvement in other viral infections in SOT recipients (Redondo et al., 2022d).

We found no clear association between any of the SNP genotypes considered and various parameters reflecting viral kinetics after transplantation, such as mean and peak DNA levels or AUCs for plasma TTV DNAemia during discrete periods, or absolute increments from baseline. The significant differences observed for the minor A allele of TNF (rs1800629) in terms of lower peak DNA levels and AUC through months 3 and 6 were not consistent across the entire post-transplant monitoring period and did not survive correction for multiple comparisons. Interestingly, Ramzi et al. (2010) Reported that the A allele of the TNF (rs1800629) SNP was associated with undetectable TTV DNAemia in a single-center cohort of HSCT recipients (OR: 0.46; 95% CI: 0.22–0.96; p-value = 0.025), although the timing for monitoring was unclear and no correction for multiple testing was performed. In line with our results, the same group observed no apparent impact of genotypes of CTLA4 (rs5742909) on the incidence of TTV infection after HSCT (Ramzi et al., 2019; Ramzi et al., 2021).

In addition to the longitudinal post-transplant monitoring of TTV replication, we have specifically investigated the associations between candidate SNPs and the presence of TTV infection at the baseline assessment, before immunosuppressive therapy was initiated. The cross-sectional comparison at this time point would reveal the potential role of genetic predisposition to TTV among ESRD patients in the absence of iatrogenic immunosuppression. In contrast to the negative results observed for the post-transplant period, we found that the minor G allele of CD209 (rs4804803) in the homozygous state exerted a protective effect even after the Bonferroni correction, and that this association with undetectable TTV DNAemia at baseline remained significant after adjusting for clinical covariates. Although caution must be exercised due to the low number of patients with pre-transplant TTV DNA levels below the LLoD, this finding is in accordance with a recent study by our group showing a protective effect against BK polyomavirus viremia linked to the G allele of CD209 (rs4804803) after KT (Redondo et al., 2022c). In addition, the minor G allele has been also associated with a lower susceptibility to tuberculosis (Vannberg et al., 2008) and severe dengue (Sakuntabhai et al., 2005). The CD209 gene codes for DC-SIGN, a transmembrane PRR belonging to the CLR family. It has been described that the presence of the G allele negatively affects gene transcription, thus downregulating the synthesis of DC-SIGN in dendritic cells (Sakuntabhai et al., 2005). In addition, DC-SIGN acts as the cell receptor for many viruses through its high affinity binding of mannose-containing carbohydrates expressed by viral glycoproteins (Lin et al., 2003). In view of the non-enveloped structure of anelloviruses, the mechanistic explanation for the association found between the CD209 (rs4804803) SNP and baseline TTV DNAemia remains to be determined and demands further investigation in healthy subjects (e.g., blood donors).

Despite its large sample size, frequent TTV DNA monitoring and comprehensive set of SNPs screened, some limitations to our study should be acknowledged. As previously described (Redondo et al., 2022a; Jaksch et al., 2022), the vast majority of recipients had TTV replication early after transplantation. Thus, associations between genetic polymorphisms and undetectable TTV DNAemia (below the LLoD of the PCR assay) were only analyzed at baseline. Since the relatively high number of SNPs imposed stringent thresholds for statistical significance, false-negative results due to insufficient statistical power cannot be excluded, particularly for those SNPs—such as CTLA4 (rs5742909) or TNF (rs1800629)—with very low absolute numbers of patients bearing the corresponding minor alleles.

In conclusion, the G allele of CD209 (rs4804803) in the homozygous state would play a protective role against TTV in non-immunocompromised patients listed for KT, whereas no significant associations have been found during the post-transplant period for any of the studied SNPs. Thus, the present results support the conception that variations in plasma TTV DNA levels after KT are mainly driven by the effect of immunosuppressive therapy rather than by underlying genetic predisposition, reinforcing its clinical usefulness as a surrogate marker of immunosuppression. Post-transplant TTV replication kinetics seems to be mainly under the control of the adaptive immune responses, with no meaningful effect of SNPs in genes orchestrating innate arm.