ABSTRACT
Human anelloviruses are ubiquitous members of the commensal virome, and blood Torquetenovirus (TTV) load has emerged as a candidate marker of host immune status. Their behavior during acute respiratory viral infections across blood and respiratory compartments remains poorly defined. To assess TTV DNA load by quantitative PCR, anellovirus species distribution by next‐generation sequencing, soluble immune markers, and blood‐respiratory genomic overlap of anelloviruses in hospitalized patients with acute respiratory viral infection. Study design: A single‐center observational study was performed on leftover specimens from patients admitted to the University Hospital of Varese, Italy, between September 2024 and March 2025. We collected 37 respiratory specimens and 9 plasma samples. Respiratory viruses were identified by multiplex real‐time PCR. TTV DNA was quantified in blood and respiratory samples by an in‐house universal TaqMan assay targeting the untranslated region; this assay provides a total TTV DNA measurement. Anellovirus species composition was characterized separately by amplicon‐based next‐generation sequencing. Plasma cytokines and adhesion molecules were measured by automated microfluidic immunoassay. In five paired blood‐respiratory cases, reconstructed anellovirus genomes were compared at the whole‐genome level. Total TTV load differed by infecting virus, with significantly higher levels in human rhinovirus (HRV) than influenza (FLU) or respiratory syncytial virus (RSV) in both blood and respiratory samples. IFN‐β was higher in FLU than RSV, whereas ICAM‐1 was higher in HRV than RSV. Age had a stronger effect than sex: children aged 0‐10 years had higher blood and respiratory TTV loads and higher IFN‐α, IFN‐β, and ICAM‐1, whereas adults aged 34–90 years showed greater representation of TTV9 and TTV24. No significant association was found between the number of detected anellovirus species and TTV DNA load. Whole‐genome comparison showed partial overlap of blood‐ and respiratory‐derived anelloviruses without compartment‐specific clustering. In acute respiratory viral infections, TTV DNA load and age‐related differences in anellovirus community composition provided information distinct from simple anellovirus species counts. HRV infection was associated with the highest total TTV burden, and blood and respiratory compartments shared part of the same anellovirus population.
Keywords: anellovirus, anellovirus species, compartmentalization, cytokines, respiratory viral infection, rhinovirus, torque teno virus
1. Introduction
Anelloviruses (AV) are small circular single‐stranded DNA viruses that infect most humans early in life and usually persist chronically without a proven causal disease association [1, 2]. Their sequence diversity is extreme, and each individual carries a highly personal anellome that can remain partly stable over time despite continuous lineage turnover [3, 4]. Yet it remains poorly understood which anellovirus species populate different body compartments and whether they contribute directly to pathogenesis. For this reason, blood TTV load has attracted interest less as a marker of viral pathogenicity than as a possible readout of immune competence or immune disturbance [5, 6, 7]. Recent aging‐focused studies further support this view by linking higher TTV viremia to chronic inflammation, immunosenescence, and age‐related disease risk [8, 9].
The respiratory tract also harbors anelloviruses. Targeted and metagenomic studies have shown frequent detection of TTV or related anelloviruses in nasopharyngeal or respiratory samples from children with acute respiratory illness [10, 11, 12]. In these studies, respiratory anelloviruses detection has been associated more with host context and coinfection than with a single disease‐specific pattern. In pediatric respiratory samples, betatorqueviruses may predominate, and respiratory anellomes appear dynamic, age‐dependent, and individually structured [10, 11]. More broadly, work on the circulating virome in aging suggests that both total viremia and anelloviruses‐community structure can carry biologic information about host immune state [8, 13].
Acute respiratory viruses trigger distinct immune profiles. RSV severity has been linked to imbalanced inflammatory and interferon responses [14, 15, 16]. Rhinoviruses can induce heterogeneous innate responses, including differences in IFN‐β, IL‐6, IL‐8, and adhesion‐related pathways [17, 18]. Recent upper‐respiratory tract profiling across multiple viral pathogens also showed that interferon and proinflammatory patterns vary with pathogen, age, and viral load, with HRV showing a distinct interferon profile compared with influenza A and RSV [19]. Whether these infection‐specific immune states are accompanied by distinct TTV burdens in blood and respiratory compartments remains unclear.
The present study examined TTV DNA load, anellovirus species distribution, and selected plasma immune markers in paired blood and respiratory samples from hospitalized patients with respiratory virus‐positive infection. A secondary aim was to assess whether anellovirus genomes reconstructed from blood and respiratory specimens show evidence of compartment‐specific clustering.
2. Matherials and Methods
2.1. Study Population
A total of 46 samples, comprising 37 respiratory specimens and 9 plasma samples, were collected from 41 unique patients admitted to the Respiratory Physiopathology Unit, intensive care unit, neonatal intensive care unit, and Pediatric Ward of the University Hospital of Varese, Italy, between September 2024 and March 2025. Five patients had paired respiratory and plasma samples. Leftover materials were used for the study and, according to local policy, no institutional review board approval was required. A study inclusion criterion is that patients test positive for at least one respiratory virus (FLU: 19 patients, 15 respiratory and 5 blood samples; HRV: 5 patients, 5 respiratory and 1 blood sample; RSV: 16 patients, 16 respiratory and 2 blood samples; Human Para‐Influenza Virus: 1 patient, 1 respiratory and 1 blood sample).
2.2. Respiratory Virus Testing and Plasma Biomarkers
Respiratory pathogen testing was performed after nucleic acid extraction using the Seegene Allplex Respiratory panel, a one‐step real‐time PCR assay covering influenza A and B, RSV A and B, adenovirus, enterovirus, parainfluenza viruses 1‐4, metapneumovirus, bocavirus, rhinovirus, seasonal coronaviruses, SARS‐CoV‐2, and selected bacterial respiratory pathogens.
Peripheral venous blood was collected at the time of respiratory sampling. Routine biochemical data, including C‐reactive protein, prothrombin time, activated partial thromboplastin time, and fibrinogen, were available in a subpopulation as part of standard care. Plasma was separated from EDTA blood by centrifugation at 1000–2000g for 10 min, aliquoted, and stored at −20°C or below until analysis. Hemolyzed, icteric, or lipemic samples were excluded.
Plasma IL‐6, IL‐8, IFN‐α, TNF‐α, IFN‐γ, IL‐1β, IFN‐β, ICAM‐1, and VCAM‐1 were measured in triplicate using ELLA microfluidic immunoassays according to the manufacturer's instructions. Intra‐assay and interassay coefficients of variation were below 11%.
2.3. TTV DNA Detection and Quantification
Viral DNA was extracted using the QIAamp DNA Blood Mini Kit. TTV DNA presence and load were assessed using a single‐step in‐house universal TaqMan PCR targeting a highly conserved untranslated‐region segment, as previously described [20]. This assay quantifies total TTV DNA load without resolving individual TTV species. Throughout the manuscript, “TTV load” refers exclusively to this qPCR‐derived measurement and should not be interpreted as a measure of total anellovirus abundance. The lower detection limit was 1.0 log copies/mL in blood. Respiratory sample results were expressed as log TTV DNA/ng extracted DNA.
2.4. NGS‐Based Genomic Characterization of Anellovirus Species
Individual human anellovirus species were identified using a whole‐genome, amplicon‐based next‐generation sequencing approach, according to previously published protocols [21]. Briefly, rolling circle amplification (RCA) was applied to preferentially enrich circular anellovirus DNA. The amplified material was then quantified using the Qubit dsDNA Quantification Assay Kit, following the manufacturer's instructions (Thermo Fisher Scientific), and diluted to serve as template for universal anellovirus inverse PCR (UAvI PCR). UAvI PCR was performed to generate genome‐length amplicons representative of human anelloviruses. The resulting PCR products were purified and quantified before library preparation. Sequencing libraries were prepared with the Illumina DNA Prep kit, starting from 150 ng of input DNA. The pooled libraries were quantified using Qubit, while fragment size distribution was assessed with the 4200 TapeStation system (Agilent). After denaturation, libraries were sequenced on Illumina platforms with paired‐end sequencing layout. Raw paired‐end sequencing reads were processed using an in‐house bioinformatic workflow for anellovirus genome reconstruction and species‐level characterization. Briefly, paired‐end reads were quality‐filtered and adapter‐trimmed, and human‐derived reads were removed before downstream analysis. Host‐cleaned reads were then analyzed through a combined de novo assembly and reference‐guided strategy. De novo contigs were generated using multiple assemblers (SPAdes, Megahit and Haploflow), filtered according to the expected anellovirus genome size range, dereplicated to reduce redundancy, and validated by remapping cleaned reads to candidate contigs. Consensus sequences were then generated from validated alignments. Candidate anellovirus genomes were classified by sequence similarity against a curated anellovirus reference database. ORF prediction was performed to identify coding regions, with particular attention to ORF1, which was used to support species‐level assignment. Sequences not meeting species‐level classification criteria were retained as unclassified anelloviruses. For readability, species are hereafter referred to by their established virus names (e.g., TTV) rather than ICTV binomial names (e.g., Alphatorquevirus homin). For abundance estimation, host‐cleaned reads were mapped back to reconstructed ORF1, complete coding sequences, or assembled genomes, according to sequence availability. Species‐level abundance was expressed as reads per million non‐human reads (RPM).
2.5. Species‐Level Anellovirus Community Analysis
To explore differences in species‐level anellovirus composition according to infection condition, an abundance matrix was generated using RPM values for each detected anellovirus species. RPM values were log10(RPM+1)‐transformed to reduce the influence of highly abundant species, and Bray‐Curtis dissimilarities were calculated on the transformed abundance matrix. Species‐level community structure was visualized by principal coordinate analysis (PCoA). Total TTV load was used only as a descriptive continuous variable for point coloring in the ordination plot (log TTV DNA copies/mL of plasma for blood samples and log TTV DNA copies/ng of extracted DNA for respiratory samples). This variable was not included in the Bray‐Curtis dissimilarity calculation, PCoA ordination, or PERMANOVA model. Mixed HRV/RSV infections were grouped into a single category. Differences in overall species‐level composition across infection conditions were tested by PERMANOVA using 9999 permutations. Homogeneity of multivariate dispersion among infection‐condition groups was assessed using PERMDISP based on distances to group centroids.
2.6. Whole‐Genome Comparison of Anelloviruses
Paired blood and respiratory swab samples collected at the same time point from five patients underwent whole‐genome comparison. Reconstructed anellovirus genomes were aligned with MAFFT. A pairwise genomic dissimilarity matrix based on nucleotide differences was computed in R and used for hierarchical clustering with average linkage. Heatmap visualization was performed with ClustVis using the precomputed dissimilarity matrix without row centering or scaling [22].
2.7. Statistical Analysis
Continuous variables were summarized as median and interquartile range (IQR). Group comparisons were performed with Kruskal–Wallis or Mann–Whitney tests, as appropriate. Prevalence comparison was assessed using the Fisher test. Species‐overlap patterns were explored using Jaccard and cosine similarity indices. Correlations between the number of detected anellovirus species and total TTV load were assessed by non‐parametric correlation analysis (Spearman's rank correlation). A two‐sided p value < 0.05 was considered statistically significant.
3. Results
3.1. Differences According to Infecting Respiratory Virus
No significant differences were observed in the abundance of individual anellovirus species across FLU, HRV, and RSV groups. In contrast, total TTV load differed significantly by infecting virus in both compartments. In blood, mean TTV load were 5.6 (4.85−6.85) log/mL in FLU (n = 15), 8.9 (7.05–9.3) log/mL in HRV (n = 5), and 6.4 (5.33–6.87) log/mL in RSV (n = 16), with significantly higher values in HRV than in FLU (p = 0.02) as well as a similar trend observed with RSV (p = 0.07) (Figure 1A). The same pattern was observed in respiratory samples, where TTV load were 4.6 (3.1–5.15) log/ng DNA in FLU, 7.8 (6.9–8.5) log/ng DNA in HRV, and 4.9 (1– 6.05) log/ng DNA in RSV, and HRV‐positive cases again showed the highest total TTV burden (vs. FLU p = 0.01; vs. RSV p = 0.03) (Figure 1B). Among soluble markers, IFN‐β differed across virus groups and was higher in FLU than in RSV (0.54 (0.006–1.64) pg/mL vs. 0 (0–0.07) pg/mL; p = 0.02). ICAM‐1 also differed significantly, with higher concentrations in HRV than in RSV (505,134 (327,883–637,468) pg/mL vs. 183,110 (112,706– 363,896) pg/mL; p = 0.03) (Figure 1C). These findings are in line with prior evidence that RSV and HRV are characterized by distinct interferon and inflammatory programs [14, 16, 18, 19]. Anellovirus species‐level prevalence differences were limited to TTV18 (FLU: 6%, HRV: 0%, RSV: 44%; FLU vs. RSV p = 0.035, HRV vs. RSV p = 0.05) and TTV24 (FLU: 26%, HRV: 0%, RSV: 50%; HRV vs RSV p = 0.04). The Kruskal–Wallis test was used to perform the aforementioned analyses. Similarity analyses based on both Jaccard (FLU vs. RSV: 0.67; FLU vs. HRV: 0.3; HRV vs. RSV: 0.35) and cosine (FLU vs. RSV: 0.16; FLU vs. HRV: 0.08; HRV vs. RSV: 0.07) indices showed low overlap among the three virus‐defined groups (Table 1). To further evaluate whether the overall anellovirus species composition differed according to infection condition, Bray‐Curtis dissimilarities calculated from log10(RPM + 1)‐transformed species‐level abundance data were visualized by PCoA. PCoA1 and PCoA2 explained 19.3% and 16.2% of the total variation, respectively. The ordination showed broad overlap among infection‐condition groups, with only partial structuring of samples (Figure 2). PERMANOVA performed on the same Bray‐Curtis dissimilarity matrix did not show a statistically significant association between infection condition and species‐level anellovirus composition, although a trend was observed (R 2 = 0.101; F = 1.38; p = 0.0854). PERMDISP was not significant (F = 0.51; p = 0.6773), indicating no evidence of marked heterogeneity of multivariate dispersion among groups.
Figure 1.
(A, B) Total TTV levels measured in blood (A) and respiratory samples (B) in patient cohort sorted by infecting virus (FLU, HRV or RSV). (C) IFN‐β and ICAM‐1 levels measured in blood samples of patient cohort sorted by infecting virus. Statistical analysis: Kruskal–Wallis test.
Table 1.
Total TTV load in blood and respiratory samples according to infecting respiratory virus.
| Variable | FLU (n = 15) | HRV (n = 5) | RSV (n = 16) | Significant comparison |
|---|---|---|---|---|
| Blood TTV load, median (IQR) | 5.6 (4.85–6.85) | 8.9 (7.05–9.3) | 6.4 (5.33–6.87) | FLU vs. HRV, p = 0.02; RSV vs. HRV, p = 0.07 |
| Respiratory TTV load, median (IQR) | 4.6 (3.1–5.15) | 7.8 (6.9–8.5) | 4.9 (1–6.05) | FLU vs. HRV, p = 0.01; RSV vs. HRV, p = 0.03 |
| IFN‐β, median (IQR) | 0.54 (0.006–1.64) | 0.18 (0–1.82) | 0 (0–0.07) | FLU vs. RSV, p = 0.02 |
| ICAM‐1, median (IQR) | 256,886 (198,289 – 414,618) | 505,134 (327,883–637,468) | 183,110 (112,706–363,896) | RSV vs. HRV, p = 0.03 |
| Species prevalence | — | — | — | TTV18: FLU vs. RSV, p = 0.035; HRV vs. RSV, p = 0.05; TTV24: RSV vs. HRV, p = 0.040 |
Figure 2.
Principal coordinate analysis (PCoA) based on Bray‐Curtis dissimilarities calculated from log10(RPM+1)‐transformed species‐level abundance data. Each point represents one sample and is colored according to matrix‐specific total TTV load. Ellipses represent 68% normal ellipses by infection condition.
3.2. Age‐Related Patterns
Age showed a clearer association with TTV loadthan infecting virus. The cohort was stratified into 0–10 (n = 14) and 34–90 (n = 22) years age groups to compare the pediatric and adult populations: the intermediate group (ages 11–33) was not represented, as no patients falling within this age range were enrolled because there weren't positive respiratory samples. First thing we observed was that children had significantly higher total TTV load in blood and respiratory samples (Blood 0–10: 6.9 (6–7.95) log/mL, 34–90: 5.4 (4.85–6.83) log/mL, p = 0.017; respiratory samples 0–10: 5.8 (3.6–6.9) log/ng DNA, 34–90: 4.6 (1.45–5.1) log/ng DNA, p = 0.037) (Figure 3A,B). Children also had higher IFN‐α (0–10: 5.95 (1.23–215.4) pg/mL, 34–90: 1.05 (0.8–18.2) pg/mL; p = 0.046), IFN‐β (0–10: 0.44 (0.01–1.64) pg/mL, 34–90: 0 (0–0.16) pg/mL; p = 0.012), and ICAM‐1 (0–10: 422, 473 (204, 884–547, 154) pg/mL, 34–90: 253, 939 (139, 220–323, 338) pg/mL; p = 0.012) concentrations, whereas adults showed higher representation of TTV9 (0–10: 0 (0–0)%, 34–90: 0 (0–4.55)%, p = 0.038) and TTV24 (0–10: 0 (0–0)%, 34–90: 0.12 (0–14.42)%, p = 0.002) (Figure 3C,D).
Figure 3.
(A, B) Total TTV levels measured in blood (A) and respiratory samples (B) in patient cohort sorted by age groups (0–10 and 34–90). (C) IFN‐α, IFN‐β and ICAM‐1 levels measured in blood samples of patient cohort sorted by age groups. (D) TTV9 and TTV 24 percentage levels measured in the blood of patients sorted by age group. (E) Jaccard and cosine indexes evaluated in children and adults. Statistical analysis: Mann–Whitney test.
At the prevalence level, TTV24 was not detected in the younger group and was present in 50% of the older group (p = 0.002). TTV1 (0–10: 0%, 34–90: 27%; p = 0.06) and TTV9 (0–10: 7%, 34–90: 36%; p = 0.068) showed non‐significant trends in the same direction. These findings fit prior work showing that anellovirus colonization begins early, that childhood respiratory anellomes differ from adult profiles, and that lineage composition shifts with age [1, 11].
Similarity analyses also supported age structuring. The between‐group Jaccard index indicated only moderate overlap (Jaccard index = 0.45), and cosine similarity was low (cosine index: 0.12). Within‐group Jaccard (0–10: 0 (0–0), 34–90: 0.11 (0–0.2); p < 0.001) and cosine (0–10: 0 (0–0), 34–90: 0.004 (0–0.29); p < 0.001) values differed significantly between children and adults, suggesting that age influences not only which species are present but also the internal structure of anellovirus communities (Figure 3E) (Table 2). The Mann–Whitney test was used to perform the aforementioned analyses.
Table 2.
Age‐stratified differences in total TTV load, IFN‐α, IFN‐β, and ICAM‐1.
| Variable | 0–10 years (n = 14) | 34–90 years (n = 22) | p value |
|---|---|---|---|
| Blood TTV load, median (IQR) | 6.9 (6–7.95) | 5.4 (4.85–6.83) | 0.017 |
| Respiratory TTV load, median (IQR) | 5.8 (3.6–6.9) | 4.6 (1.45–5.1) | 0.037 |
| IFN‐α, median (IQR) | 5.95 (1.23–215.4) | 1.05 (0.8–18.2) | 0.046 |
| IFN‐β, median (IQR) | 0.44 (0.01–1.64) | 0 (0–0.16) | 0.012 |
| ICAM‐1, median (IQR) | 422,473 (204,884–547,154) | 253,939 (139,220–323,338) | 0.012 |
| TTV9 relative percentage, median (IQR) | 0 (0–0) | 0 (0–4.55) | 0.038 |
| TTV24 relative percentage, median (IQR) | 0 (0–0) | 0.12 (0–14.42) | 0.002 |
| TTV24 prevalence | 0% | 50% | 0.002 |
3.3. Sex and Relationship Between Anellovirus Species Richness and TTV Load
No statistically significant differences in anellovirus species levels or prevalence were found between males and females. Similarity analyses by sex were likewise unremarkable.
The number of anellovirus species detected in each patient did not correlate significantly with TTV load in blood (r = 0.11, p = 0.46) or respiratory samples (r = −0.08, p = 0.64, Spearman's rank correlation) (Figure 4A,B). This suggests that anellovirus species richness and TTV load capture different dimensions of the anellome, consistent with the known individuality of anellovirus communities [3, 4]. This distinction is also consistent with recent aging data showing that species diversity and total viremia can track partly different aspects of immune dysregulation [13].
Figure 4.
Correlation between TTV load and number of anellovirus species detected in blood (A) and respiratory samples (B). Spearman's r and p values are reported in the image. Statistical analysis: Spearman's rank correlation.
3.4. Blood‐Respiratory Genome Overlap
Anellovirus genomes were detected in both blood and respiratory samples from all five patients included in the genomic comparison. Overall, 19 reconstructed genomes were identified in blood and 36 in respiratory samples, of which 11 were shared between the two compartments. In several patients, the same anellovirus species was recovered from both blood and respiratory specimens (Table 3).
Table 3.
Reconstructed anellovirus genomes identified in blood and respiratory samples, including genomes shared between the two compartments.
| Genomic comparison metric | Value |
|---|---|
| Patients with paired genome analysis | 5 |
| Reconstructed genomes in blood | 19 |
| Reconstructed genomes in respiratory samples | 36 |
| Shared genomes across compartments | 11 |
Whole‐genome comparison did not reveal clear compartment‐specific segregation. Hierarchical clustering based on pairwise genomic dissimilarity showed that reconstructed genomes clustered mainly according to viral species rather than sampling sites. Consistently, heatmap visualization confirmed the intermixing of blood‐ and respiratory‐derived genomes within multiple clusters, rather than a separation between the two anatomical compartments (Figure 5).
Figure 5.
Whole‐genome dissimilarity heatmap of anellovirus genomes detected in blood and respiratory samples. The color scale ranges from blue, indicating low dissimilarity/high genomic similarity, to red, indicating high dissimilarity. Annotation bars indicate sample matrix and viral species.
4. Discussion
This study shows that in hospitalized patients with respiratory virus‐positive infection, TTV load differs by infecting virus and age, whereas single anellovirus species prevalence is comparatively stable. The clearest virus‐related signal was the markedly higher TTV load observed in HRV‐positive patients in both blood and respiratory samples. By contrast, anellovirus species‐level differences between FLU, HRV, and RSV were sparse, and similarity indices did not support clear virus‐specific anellome groupings. Consistently, multivariate analysis of species‐level abundance data showed broad overlap among infection‐condition groups. Although a partial structuring was visible in the PCoA, PERMANOVA did not reach statistical significance after exclusion of parainfluenza‐positive samples, supporting the interpretation that infecting virus was associated mainly with quantitative differences in TTV load rather than with a robust virus‐specific restructuring of the species‐level anellome.
These results favor a quantitative rather than purely compositional interpretation of TTV behavior during acute respiratory infection. TTV load is increasingly regarded as an indirect marker of host immune status rather than a disease‐specific pathogen signal [5, 6, 7]. In critical illness and severe viral infection, higher or dynamically changing anellovirus loads have been linked to immune disturbance, subsequent infectious complications, or inflammatory context [6, 7, 23]. In aging cohorts, higher TTV viremia has also been associated with chronic low‐grade inflammation and ischemic heart disease risk, reinforcing the idea that circulating TTV load reflects host immune and inflammatory tone rather than direct viral pathogenicity [8, 9]. In the present cohort, HRV was the viral group with the highest TTV burden and higher ICAM‐1 than RSV. Rhinoviruses are known to trigger heterogeneous antiviral and inflammatory responses, including IFN‐β‐related and adhesion‐related pathways [17, 18]. One plausible interpretation is that the host milieu accompanying HRV infection was more permissive to TTV replication or reflected a different balance between mucosal inflammation and systemic immune control.
Age had an even stronger effect than infecting viruses. Children aged 0–10 years showed higher total TTV load in both compartments and higher IFN‐α, IFN‐β, and ICAM‐1, whereas adults aged 34–90 years showed greater representation of selected species, especially TTV9 and TTV24. This aligns with a current understanding of early life anellovirus ecology. Anelloviruses colonize humans soon after birth, but the structure of the anellome evolves with age, and childhood respiratory communities differ from adult ones in both composition and dynamics [1, 11]. Recent aging‐focused work also suggests that species‐level structure becomes more informative in later life, with higher TTV species diversity linked to immune dysregulation, altered T‐cell homeostasis, and inflammaging‐related phenotypes [8, 13]. The present data suggest that during acute respiratory infection, pediatric patients may carry a larger overall quantitative burden even when adults show enrichment of lineages.
The absence of significant sex‐related differences indicates that sex is a weaker determinant of TTV ecology in this clinical setting than age or the current infecting virus. Similarly, the lack of correlation between numbers of detected anellovirus species and TTV load suggests that richness should not be used as a surrogate for burden. Large‐scale genomic studies have shown that anellovirus communities are highly individualized and shaped by persistence, turnover, and recombination, so abundance and diversity do not necessarily move together [3, 4].
The whole‐genome comparison adds a compartmental dimension. Blood and respiratory samples shared part of the anellovirus repertoire, but clustering was driven mainly by viral species rather than by anatomical origin. This argues against strict compartmental segregation and supports a model of partial overlap between systemic and respiratory anellovirus populations. Respiratory detection of anellovirus has been reported repeatedly, especially in children with acute respiratory illness [10, 11, 12]. The present genome‐level analysis strengthens that observation by showing that overlap is not limited to PCR positivity alone.
The study has limitations. It was conducted at a single center and relied on leftover clinical material. Furthermore, the sample size should be increased, even though studies regarding anellovirus are usually conducted with a similar sample unit. The genome comparison was limited to five paired cases, and clinical interpretation remains constrained by the absence of longitudinal follow‐up. Nonetheless, the paired design and the combined quantitative, immunologic, and genomic approach provide a useful framework for studying host‐anellovirus interactions in respiratory infection.
In conclusion, TTV load and anellovirus composition provided complementary information in patients with acute respiratory viral infection. HRV infection was associated with the highest TTV load, while age was associated with both TTV load and the distribution of selected anellovirus species. Blood and respiratory compartments also shared part of the same anellovirus population. These findings support further evaluation of TTV load and broader anellome profiling as adjunct markers of host‐virus interaction in clinical respiratory virology.
Author Contributions
Conceptualization: Roberto Ferrarese, Fabio Filippini, Pietro Giorgio Spezia and Federica Novazzi. Methodology: Angelo Paolo Genoni, Fabio Flippini, Gaia Zambon, Giorgia Baffioni. Formal analysis: Roberto Ferrarese, Fabio Filippini, Federica Novazzi, Pietro Giorgio Spezia, Daniele Focosi. Investigation: Angelo Paolo Genoni, Gaia Zambon, Giorgia Baffioni. Resources: Serena Delbue, and Federica Novazzi. Data curation: Roberto Ferrarese, Fabio Filippini, Pietro Giorgio Spezia and Federica Novazzi. Writing – original draft preparation: Roberto Ferrarese, Pietro Giorgio Spezia, Federica Novazzi. Review and editing: Fabrizio Maggi, Daniele Focosi, Claudia Minosse, Pietro Giorgio Spezia, Nicasio Mancini, Lucia Signorini, Serena Delbue, and Federica Novazzi. Visualization: Roberto Ferrarese, Fabrizio Maggi and Federica Novazzi. Supervision: Fabrizio Maggi, Serena Delbue, Lucia Signorini and Federica Novazzi. Project administration: Fabrizio Maggi, Lucia Signorini, Federica Novazzi. Funding acquisition: Fabrizio Maggi, Lucia Signorini and Federica Novazzi. All authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
This specific study was supported by funds allocated to the National Institute for Infectious Diseases “Lazzaro Spallanzani”, IRCCS, 00149, Rome (Italy), from the Italian Ministry of Health (Ricerca Corrente—Linea 1 on emerging and re‐emerging infections). The authors also thank the European Union (Next Generation EU)for their financial support. This work was funded by the European Union (Next Generation EU) under the PRIN PNRR P20225P4T7 grant, entitled “Study of the role of inflammatory mediators and endothelial dysfunctions in the pathogenesis of respiratory diseases.”
Ferrarese R., Spezia P. G., Filippini F., et al., “Blood and Respiratory TTV Load, Anellovirus Species Distribution, and Compartment Overlap in Acute Respiratory Viral Infections,” Journal of Medical Virology 98 (2026): e71199. 10.1002/jmv.71199.
Roberto Ferrarese and Pietro Giorgio Spezia contributed equally to this work.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
References
- 1. Kaczorowska J., Cicilionytė A., Timmerman A. L., et al., “Early‐Life Colonization by Anelloviruses in Infants,” Viruses 14, no. 5 (April 2022): 865, 10.3390/v14050865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Kaczorowska J. and van der Hoek L., “Human Anelloviruses: Diverse, Omnipresent and Commensal Members of the Virome,” FEMS Microbiology Reviews 44, no. 3 (May 2020): 305–313, 10.1093/femsre/fuaa007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Arze C. A., Springer S., Dudas G., et al., “Global Genome Analysis Reveals a Vast and Dynamic Anellovirus Landscape Within the Human Virome,” Cell Host & Microbe 29, no. 8 (August 2021): 1305–1315.e6, 10.1016/j.chom.2021.07.001. [DOI] [PubMed] [Google Scholar]
- 4. Kaczorowska J., Deijs M., Klein M., et al., “Diversity and Long‐Term Dynamics of Human Blood Anelloviruses,” Journal of Virology 96, no. 11 (June 2022): e00109‐22, 10.1128/jvi.00109-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Focosi D., Spezia P. G., and Maggi F., “The Emerging Tool of the Human Anellovirome,” Viruses 16, no. 6 (June 2024): 990, 10.3390/v16060990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Forqué L., Albert E., Giménez E., et al., “Monitoring of Torque Teno Virus Dnaemia in Critically Ill COVID‐19 Patients: May It Help to Predict Clinical Outcomes?,” Journal of Clinical Virology 148 (March 2022): 105082, 10.1016/j.jcv.2022.105082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thijssen M., Devos T., Meyfroidt G., Van Ranst M., and Pourkarim M. R., “Exploring the Relationship Between Anellovirus Load and Clinical Variables in Hospitalized COVID‐19 Patients: Implications for Immune Activation and Inflammation,” IJID Regions 9 (December 2023): 49–54, 10.1016/j.ijregi.2023.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Badillo‐Pazmay G. V., Fortunato C., Cianfruglia L., et al., “The Gut and Circulating Virome: Emerging Players in Aging and Longevity,” Frontiers in Aging 6 (February 2026): 1731621, 10.3389/fragi.2025.1731621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Giacconi R., Piacenza F., Maggi F., et al., “Association Between TTV Viremia, Chronic Inflammation, and Ischemic Heart Disease Risk: Insights From MARK‐AGE and Report‐Age Projects,” Journals of Gerontology, Series A: Biological Sciences and Medical Sciences 79, no. 11 (November 2024): glae228, 10.1093/gerona/glae228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Bal A., Destras G., Sabatier M., et al., “Metagenomic Analysis Reveals High Abundance of Torque Teno Mini Virus in the Respiratory Tract of Children With Acute Respiratory Illness,” Viruses 14, no. 5 (May 2022): 955, 10.3390/v14050955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Cao L., Ma Y., Wan Z., et al., “Longitudinal Anellome Dynamics in the Upper Respiratory Tract of Children With Acute Respiratory Tract Infections,” Virus Evolution 9, no. 2 (September 2023): vead045, 10.1093/ve/vead045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. del Rosal T., García‐García L., Casas I., et al., “Torque Teno Virus in Nasopharyngeal Aspirate of Children With Viral Respiratory Infections,” Pediatric Infectious Disease Journal 42, no. 3 (March 2023): 184–188, 10.1097/INF.0000000000003796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Novazzi F., Spezia P. G., Ferrante F. D., et al., “TTV Species Diversity as a Novel Biomarker of Immune Dysregulation in Aging,” Journal of Medical Virology 97, no. 11 (November 2025): 1–14, 10.1002/jmv.70656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Pita‐Martínez C., Goez‐Sanz C., Virseda‐Berdices A., et al., “Reduced Antiviral Gene Expression and Elevated CXCL8 Expression in Peripheral Blood Are Associated With Severe Hypoxemia in RSV‐Infected Children,” Frontiers in Immunology 15 (September 2024): 1438630, 10.3389/fimmu.2024.1438630. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Taveras J., Garcia‐Maurino C., Moore‐Clingenpeel M., et al., “Type III Interferons, Viral Loads, Age, and Disease Severity in Young Children With Respiratory Syncytial Virus Infection,” Journal of Infectious Diseases 227, no. 1 (December 2022): 61–70, 10.1093/infdis/jiac404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Vázquez Y., González L., Noguera L., et al., “Cytokines in the Respiratory Airway as Biomarkers of Severity and Prognosis for Respiratory Syncytial Virus Infection: An Update,” Frontiers in Immunology 10 (June 2019): 1154, 10.3389/fimmu.2019.01154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Chałubiński M., Szulc A., Pawełczyk M., et al., “Human Rhinovirus 16 Induces Antiviral and Inflammatory Response in the Human Vascular Endothelium,” APMIS 129, no. 3 (March 2021): 143–151, 10.1111/apm.13103. [DOI] [PubMed] [Google Scholar]
- 18. Kim J. H., Jang J. Y., and Jang Y. J., “Human Rhinovirus Serotypes Induces Different Immune Responses,” Virology Journal 18, no. 1 (November 2021): 232, 10.1186/s12985-021-01701-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ferrarese R., Novazzi F., Arcari G., et al., “Expression Patterns of Interferons and Proinflammatory Cytokines in the Upper Respiratory Tract of Patients Infected by Different Viral Pathogens: Correlation With Age and Viral Load,” Biomolecules 15, no. 11 (November 2025): 1545, 10.3390/biom15111545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Maggi F., Pifferi M., Fornai C., et al., “TT Virus in the Nasal Secretions of Children With Acute Respiratory Diseases: Relations to Viremia and Disease Severity,” Journal of Virology 77, no. 4 (February 2003): 2418–2425, 10.1128/JVI.77.4.2418-2425.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Spezia P. G., Matsudaira K., Filippini F., et al., “Viral Load of Torquetenovirus Correlates With Sano's Score and Levels of Total Bilirubin and Aspartate Aminotransferase in Kawasaki Disease,” Scientific Reports 13, no. 1 (October 2023): 18033, 10.1038/s41598-023-45327-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Metsalu T. and Vilo J., “ClustVis: A Web Tool for Visualizing Clustering of Multivariate Data Using Principal Component Analysis and Heatmap,” Nucleic Acids Research 43, no. W1 (July 2015): W566–W570, 10.1093/nar/gkv468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Castain L., Petrier M., Bulteau S., et al., “Association of Dynamics of Anellovirus Loads With Hospital‐Acquired Pneumonia in Patients With Brain Injury During the Intensive Care Unit Stay,” Journal of Infectious Diseases 230, no. 5 (November 2024): 1139–1146, 10.1093/infdis/jiae110. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.