Sequencing and characterization of human bocavirus genomes from patients diagnosed in Southern France between 2017 and 2022

The diversity and evolution of the genomes of human bocavirus (HBoV), which causes respiratory diseases, have been scarcely studied. Here, we aimed to obtain and characterize HBoV genomes from patients's nasopharyngeal samples collected between 2017 and 2022 period (5 years and 7 months). Next‐generation sequencing (NGS) used Illumina technology after having implemented using GEMI an in‐house multiplex PCR amplification strategy. Genomes were assembled and analyzed with CLC Genomics, Mafft, BioEdit, MeV, Nextclade, MEGA, and iTol. A total of 213 genomes were obtained. Phylogeny classified them all as of Bocavirus 1 (HBoV1) species. Five HBoV1 genotypic clusters determined by hierarchical clustering analysis of 27 variable genome positions were scattered over the study period although with differences in yearly prevalence. A total of 167 amino acid substitutions were detected. Besides, coinfection was observed for 52% of the samples, rhinoviruses then adenoviruses (HAdVs) being the most common viruses. Principal component analysis showed that HBoV1 genotypic cluster α tended to be correlated with HAdV co‐infection. Subsequent HAdV typing for HBoV1‐positive samples and negative controls demonstrated that HAdVC species predominated but HAdVB was that significantly HBoV1‐associated. Overall, we described here the first HBoV1 genomes sequenced for France. HBoV1 and HAdVB association deserves further investigation.

1][12][13][14] Respiratory infections caused by HBoV1 are mostly acute, but chronic infection has also been reported, as for the case of a 29-year-old immunocompetent patient. 15HBoV was also detected in some children with Kawasaki disease. 16,17HBoV was commonly found in coinfection with other respiratory viruses that co-circulate with them, such as respiratory syncytial virus, rhinoviruses, adenoviruses (HAdVs), human parainfluenza viruses, influenza A virus, human coronavirus OC43, or SARS-CoV-2. 18,19As a matter of fact, the pathogenicity of HBoV has long been debated due to their high rate of co-infection with other respiratory viruses.
Nonetheless, some studies showed that they can cause respiratory infection in absence of an association with any other known respiratory virus or bacteria. 20,21oV infections are distributed worldwide, with a seasonal distribution that changes according to the country.In temperate countries, they are mainly detected during winter and spring.
Their prevalence was estimated to be 6.3% globally, and was reported to range between 6.1% and 8.5% in France. 1,21,22As observed with the SARS-CoV-2 epidemics, understanding viral epidemics linked to different genotypes requires the availability of thousands of genome sequences to characterize the most comprehensively mutations and genetic evolution.However, there were only 309 complete HBoV genomes available in the NCBI Genbank nucleotide sequence database (https://www.ncbi.nlm.nih.gov/genbank/) 23

| Primer design for PCR amplification of overlapping regions covering the whole viral genome
As viral genome NGS from clinical samples most often requires enrichment of viral nucleic acid, we implemented an "Artic-like" strategy, first used for Zika virus 24 then used at a large scale for SARS-CoV-2 genome NGS (https://artic.network/).All HBoV1 genomes available in GenBank as of 05/06/2023 were retrieved.Then, a sequence set clean-up was carried out by removing very short genomes and those with gaps, while keeping those with a minimal length of 5200 nucleotides, corresponding to ≥94% of the length of reference genome GenBank Accession no.NC_075120.1.
Genome alignment was thereafter performed using the Mafft software. 25PCR primers were designed on the basis of the alignment using the GEMI software 26 that detects the most conserved genome regions and propose primer sequences.In addition, this tool allows selecting parameters such as amplicon sizes (here, the 300-1700 nucleotide range was selected), primer sizes (here, the 19-22 nucleotide range was selected), primer hybridization temperatures (here, 58°C was selected), or numbers of degenerated nucleotides in the primers to handle intergenome diversity (here, up to 2 degenerated primers were allowed).Primers' sequences are available in Supplementary Data.

| PCR amplification of overlapping regions covering the whole genomes and NGS
The previously designed PCR primers were mixed into two pools using the mutliplex PCR "Artic-like" principle.PCR amplification using the two previously pools of designed PCR primers was performed separately for each pool, using the SuperScript III One-Step RT-PCR System with Platinum Taq High Fidelity (SS-III-HF kit; Invitrogen, Life Technologies).Primer concentrations in the pools were between 10 and 12.5 pmol/µL (Supplementary Table 1).Per PCR reaction, a total of 12.5 µL of 2X mix was added to 8.05 µL of water, 0.7 µL of enzyme, 0.75 µL of each PCR primer pool, and 3 µL of nucleic acid extract.The PCR program included an initial denaturation at 95°C for 5 min, then 39 cycles that included a denaturation step at 95°C for 15 s, an hybridization step at 58°C for 45 s, and an elongation step at 70°C for 90 s.These 39 cycles were followed by a final elongation step at 70°C for 5 min.Purification of the PCR products was carried out on a NucleoFast 96-well plate (Macherey Nagel ref 743100.50)   followed by an elution of the PCR amplicons with 40 µL of pure water.Subsequently, the two pools of PCR products were mixed at equal volume.
To set up our system, we used the Oxford Nanopore Technology with the Ligation Sequencing Kit (SQK-LSK109) library that was sequenced on a SpotON flow cell Mk I, R9.4.1 on a GridION instrument, according to the manufacturer's instructions (Oxford Nanopore Technologies).After our "Artic-like" system was developed to be used on the studied nasopharyngeal samples, HBoV genomes were sequenced using the Illumina technology on a NovaSeq 6000 instrument or on a MiSeq instrument using the COVIDSeq protocol (Illumina Inc.) and replacing the COVID-19 ARTIC PCR primers with the PCR primers designed in the present work.The NovaSeq.6000 was loaded using the classic Illumina loading procedure in a SP flow cell according to the NovaSeq-XP workflow with a reading of 2 × 50, according to a previously described procedure. 27For the Miseq instrument, procedures of loading and sequencing with a reading of 2 × 250 were performed as recommended in the MiSeq sequencing guide (Illumina Inc.) and following a previously described procedure. 27nomes obtained were submitted to the GenBank sequence database (https://www.ncbi.nlm.nih.gov/genbank/;submission no.2793582; GenBank Accession no PP296703 -PP296915).

| Bioinformatic analyses of viral genomes
A pipeline was set up using the CLC Genomics Workbench v.7 software, consisting of NGS reads trimming and subsequent mapping using as reference the genome Accession no.LC651179.1 available in GenBank.Trimming was carried out with a minimum number of 15 and a maximum of 1000 nucleotides in the reads.For mapping, the parameters used were a minimum coverage of sequencing reads on the reference genome of 90% with a minimum identity of 90%.The number of mismatches, deletions, and insertions allowed was 2, 3, and 3, respectively.Ambiguity symbols "N" were in the parts where there was ambiguity.After read mapping, the consensus genome sequences with a minimum of 3 reads as NGS depth were recovered in fasta format.Sequence alignments were then performed using the Mafft software 25 with default parameters and attempting to align regions with gaps for further analysis.The BioEdit software (https://bioedit.software.informer.com/)was used to visualize the alignments and create the sequence identity matrices.The MEGA v.11 28 and iTOL 29 softwares were used to reconstruct and visualize phylogenetic trees.
The phylogenetic tree based on HBoV genomes was constructed using the Neighbor-Joining method, the Kimura 2-parameter model with 1000 replicates and a number of 100 threads.Other parameters used were inclusion of transition and tranversion as substitutions, uniform rates among sites, pairwise deletion for gaps/missing date, and selection of all codon positions and of noncoding sites.
The Nextclade tool (https://clades.nextstrain.org/) 30was adapted to HBoV1 and allowed the report of nucleotide and amino acid mutations harbored by HBoV genomes.We chose as reference the genome GenBank Accession no.NC_075120.1,which was obtained from a sample collected in 2005 and which was updated in March 2023.Finally, to delineate HBoV1 genotypic clusters, a hierarchical clustering was performed based on presence/absence patterns identified using the BioEdit software within a set of 27 mutations each present in at least 10% of the whole genome set, using the Multiple Experiment Viewer (MeV) software, 31,32 with default parameters.A principal component analysis (PCA) was then carried out using the Xlstat Microsoft software to see if there was any correlation between a HBoV1 genotypic cluster and a particular co-infection.

| Viral coinfection analysis and HAdV genotyping
By performing a hierarchical clustering analysis with Mev, we found a correlation between one of the genotypic clusters and the HAdV.We hence investigated if this correlation was linked to a particular HAdV species or not.For nasopharyngeal samples for which a coinfection with HAdV was diagnosed, HAdV typing was performed using a previously described in-house qPCR system specific to each type of HAdV 33 with the Light Cycler 480 Probes Master reagent on a Light Cycler 480 thermal cycler (Roche), according to the manufacturer's protocol.Specific qPCR for the different HAdV species C serotypes (1, 2, 5 and 6) were performed using previously described primers and probe systems. 34

| Primer design for PCR amplification of overlapping regions covering the whole genomes
We retrieved 100 HBoV1 genomes from GenBank.Their alignment with the GEMI software allowed obtaining 80 proposed primer pairs as output, of which 16 were selected as their PCR products allowed covering the whole HBoV1 genomes with overlaps between contiguous amplicons.These 16 primer pairs were tested, and 8 were selected for pooling as they provided spots on agarose gels with sufficient intensity and at the expected nucleotide size.These eight primers were optimized by increasing their concentration when the bands on the agarose gel were of low intensity.Selected concentrations varied between 10 and 12.5 pmol/µL.Nanopore technology on a Gridion instrument was used to carry out the NGS tests.All samples were then sequenced using Illumina technology.

| NGS
For the NGS step, 9% of our samples were sequenced on a Miseq instrument and 91% on a Novaseq.6000 instrument.A total of 601 nucleic acid extracts obtained from HBoV1 DNA-positive nasopharyngeal samples collected between February 2017 and August 2022 were tested.Sequencing of these samples was attempted using our "artic-like" strategy, and 213 genome sequences were obtained with at least 90% coverage of the reference genome GenBank Accession no.LC651179.1 and a depth of NGS reads of at least 3X.These genomes were obtained from nasopharyngeal samples collected between February 06, 2017 and July 07, 2022 (67 months).
The total number of reads per sample mapped to the reference genome ranged from 5675 to 9 917 033, with an average of 3 569 789 ± 2 943 862 reads.These 213 selected genomes exhibited a coverage ranging from 91% to 99% of the reference genome, with an average of 98% ± 1%.The cycle threshold values (Ct) of the qPCR used for diagnostic of HBoV1 infection for the 213 samples analyzed ranged from 8 to 34 with an average of 22 ± 5, while that for the samples from which the obtained genomes displayed a coverage <90%, and which were not retained, ranged from 17 to 38 with an average of 33.1 ± 2.8.
The HBoV1 sequences obtained in this study have been deposited on GenBank (GenBank accession no PP296703 -PP296915).As a comparison, we were also interested in recording mutations present in the HBoV1 complete genomes available in GenBank, which were between 4800 and 5600 nucleotide in size.A total of 308 genomes were included in this analysis.Four exhibited no mutation compared with reference NC_075120.1.Regarding the 304 remaining genomes, the total number of nucleotide substitutions ranged from 10 to 68, with an average of 23 ± 8 substitutions, and the total number of amino acid substitutions ranged from 5 to 58 with an average of 10 ± 6 substitutions.Seven genomes exhibited between 1 and 6 amino acid deletions.No amino acid insertions were found in these samples.Overall, 596 different amino acid mutations were harbored by any of these 308 genomes, 80 being present in at least 3 genomes.Of these 596 amino acid substitutions, 69 were in common with those found in the genomes obtained in the present study.Amino acid substitutions, either in common between our data and those of GenBank or not in common between these two sets of genomes, affected all genes.We plotted in Figure 2 all 165 amino acid mutations present in at least two genomes (as mutations harbored by only one genome was too large for a legible representation) (Figure 2).

| HBoV1 classification into genotypic clusters
The 213 HBoV genomes obtained in the present study were from samples whose collection dates were scattered over the whole 2017-2022 period (Figure 3).All obtained genomes belonged to the HBoV1 species as determined by phylogeny.The mean nucleotide diversity between these genomes was 99.61 ± 0.18% and ranged between 98.70% and 100.00%, corresponding to a mean number of 13 ± 7 nucleotides (range, 0-49).Congruently with this overall low intergenome genetic diversity, bootstrap values in the phylogenetic tree were very low, most often <60%, which prevented a classification into lineages within the HBoV1 genotypic cluster.Besides, no genome clustering was observed according to the date of collection of the samples.
To get a better insight in the genetic evolution of these HBoV1 genomes over time, we performed a hierarchical clustering analysis based on a list of 27 variable positions exhibiting mutations present in at least 10% of the 213 genomes.This allowed assigning genomes in five genotypic clusters we named cluster α, β, γ, δ, and ε.Some genomes could not be unambiguously assigned and were assigned to an "unclassified" cluster (Figure 4).We looked at how the circulation of these clusters occurred over time and observed they were scattered over the study period, being each detected in different years although with different prevalence (Figure 5).As HBoV1 was diagnosed for some samples in coinfection with other respiratory viruses, we were interested in analyzing the distribution of these viral coinfections according to the HBoV1 cluster.For 52% of the respiratory samples from which the 213 genomes were obtained, between one or up to four additional respiratory viruses were simultaneously diagnosed, rhinoviruses and HAdVs being those the most frequent, in 32% and 46% of the cases, respectively.We observed that before Year 2020 (before the SARS-CoV-2 pandemic), epidemic peaks of HBoV1 were observed just after those of HAdVs and rhinoviruses.Then, during and after 2020, we observed that HBoV1, HAdVs, and rhinoviruses exhibited similar seasonal prevalence curves (Figure 6).
We then tested using PCA if the prevalence of HBoV1 genotypic clusters that we previously defined (α, β, γ, δ, ε, and "unclassified") was correlated with those of the co-diagnosed respiratory viruses.This allowed revealing a tendency toward a correlation between the prevalence of HAdVs and that of the HBoV1 cluster α (Figure 7).Of the 18 nasopharyngeal samples for which this association between HBoV1 cluster α and HAdV was observed, a sufficient volume of sample was available for 15 to perform the classification of HAdVs into genotypes A-G.These 15 samples were tested concurrently with 15 other nasopharyngeal samples collected during the same year and week and positive by PCR for HAdV but negative for HBoV1.Four HBoV1-positive samples and one HBoV1-negative sample were co-infected with HAdVs.Three HAdV genotypes including B, C, and F were detected.There was no significant association between the prevalence of HBoV1 and that of HAdV C and F, but we found a significant association with the prevalence of HAdV B (Table 1).Indeed, 6/15 HBoV1 genotypic cluster α-positive samples showed co-diagnosis of HAdV B infection, compared with only 1/15 HBoV1-negative samples (p = 0.08; Fisher exact test).
Finally, to investigate further the association HBoV1 cluster α with HAdV C, we determined the HAdV C cluster and observed that HAdV C2 predominated in both HBoV1 cluster α-positive and HBoV1-negative samples (Table 2), and that co-infection with both HAdV C1 and C2 serotypes was observed in both groups without significant differences in prevalence.published studies showing that HBoV1 predominates in human respiratory infections, while HBoV2, HBoV3, and HBoV4 predominate in gastroenteritis. 35,36The evolution of HBoV, which have relatively stable genomes, differs from that of other epidemic viruses such as SARS-CoV-2, whose temporal evolution is associated with the successive circulation of a few viral variants, which replace those that circulated during the previous months 37 (https://covariants.org/

| DISCUSSION
).This low level of genetic variability for HBoV can be at least partly explained by the fact that they are DNA viruses, which are known to be more genetically stable than RNA viruses. 38Compared with a reference HBoV genome obtained from a patient sampled between December 2003 and March 2004, some of the amino acid substitutions identified in all the genomes in our study were also hosted by viral genomes available in GenBank.Four of these mutations were present in all genomes deposited in GenBank since 2005.The diversity observed in this study may have an impact on the contagiousness and severity of HBoV1 infections.This could be linked to mutations in the capsid proteins that are involved in attachment of the virus to its host cell, as has been observed for other viruses such as SARS-CoV-2 for which mutation in the spike glycoprotein was associated with a high contagiousness 39 or the respiratory syncytial virus for which capacities of immune response were reported to vary according to the viral genotype. 40Further studies are needed to gain a better knowledge about these points and the impact of given mutations.Intriguingly, stop codons were observed in three genomes from the present study and also in genomes from GenBank at different genome positions.The significance of these stop codons is unclear.It was for instance observed in SARS-CoV-2 variants of concern, indicating that some accessory viral genes are dispensable in some settings. 41Nonetheless, here stop codons are in overlapping genes that encode the minor capsid proteins which are essential for viral replication.Therefore, there could be mechanisms such as translational readthrough reported in some viruses. 42her respiratory viruses were concurrently detected with HBoV1 in more than half of the nasopharyngeal samples studied here, which is congruent with other studies conducted on HBoV that showed a relatively high proportion of coinfections ranging from 45% to 72% of cases. 43,44Here, the most predominant coinfecting viruses were rhinoviruses followed by HAdVs, which is also congruent with previous studies. 21We observed that prevalence rates of these different viruses varied between before, during, and after Year 2020.As a matter of fact, the epidemiological features of co-circulation of respiratory viruses F I G U R E 4 Hierarchical clustering of HBoV1.and the processes of possible interference between these viruses appear to be complex issues.[47][48] Hierarchical cluster analysis carried out on 27   the study period, which is in agreement with previous studies. 49,50 further found that among HAdV-positive and HBoV1-negative samples, HAdVC and HAdVB genotypes were those the most frequently detected, which has been also reported previously in respiratory infections. 51Furthermore, we found that HAdVC2 predominated in both HBoV1 cluster α-positive and HBoV1negative control samples.We also found that five nasopharyngeal samples were coinfected with two different HAdV genotypes.Such coinfections between HAdVs were already reported. 52,53Most interestingly, we observed a significant association of HBoV1 cluster α with HAdVB.This questions if some HBoV1 might need an auxiliary virus to replicate, as for the case of Adeno-Associated Virus 54 that belongs to the same viral family.This observation requires further study.It should be considered that the different HBoV1 genotypic clusters identified here were detected during different years and at different prevalence.6][57][58] Molecular monitoring of these viruses in mono-and co-detection by complete sequencing would help to clarify this question in future work.Taken together, previous findings suggest that in respiratory infections HAdVB may be linked specifically to HBoV1 genotypic cluster α, but further analyses with larger cohorts are required to investigate this point.
The present study has some limitations.Thus, there is a bias linked to the sampling as it involved samples collected in the setting of university public hospitals, and from a single geographical area, South of France; in addition, our sampling is limited in time sequencing and analyses.However, the greater severity is a muchdiscussed factor which has been the subject of contradictory results, which may suggest a lack of severity 56 but in fact very probably depends on the viruses considered and on the age and clinical background of the patients. 57,59Another issue is whether these coinfections are concurrent acute infections with the two viruses or alternatively an acute, superinfection with one virus in a patient protracted carrier of a second virus.This latter case is suspected when one of the two viruses is detected at a very low viral load, which is particularly common in respiratory virus co-infections involving HBoV. 60Beyond, here, the association of Adenovirus C and HBoV1 caught our attention because HBoVs (genus Bocaparvovirus) are members of the subfamily Parvovirinae, as the Adenoassociated viruses (genus Dependoparvovirus) that require coinfection with a helper virus to replicate efficiently. 61 as of 05/06/2023; this is mighty limited for example in comparison with over 8 million SARS-CoV-2 genomes available in GenBank (and over 15 million available in the GISAID database (https://gisaid.org/).Here we aimed to perform retrospectively the nextgeneration sequencing (NGS) and characterization of HBoV1 genomes from respiratory samples of patients diagnosed between 2017 and 2022 at university and public hospitals of Marseille, Southern France.

2 |
MATERIALS AND METHODS 2.1 | Clinical samples Clinical samples analyzed were nasopharyngeal samples sent to the clinical microbiology and virology laboratory of university and public hospitals of Marseille, southeastern France, for the purpose of routine clinical diagnosis of respiratory viruses.These samples were collected between February 2017 and August 2022 (67 months).DNA was extracted from nasopharyngeal samples using the KingFisher Flex system (Thermo Fisher Scientific), following the manufacturer's instructions.HBoV diagnosis was performed by qPCR with the FTD Respiratory pathogens 21 (FTD) assay (Fast Track Diagnosis) that targets HBoV1 DNA.
Within the 213 genomes, the total number of nucleotide substitutions compared with reference GenBank Accession no.NC_075120.1 ranged from 7 to 44, with an average of 23 ± 5.These mutations tended to be scattered over the whole genome length.Only 7 of the 213 genomes exhibited nucleotide deletions; 4 exhibited one deletion and 3 exhibited 2. These nucleotide mutations were mostly synonymous mutations.Regarding amino acid mutations, the total number of substitutions ranged from 5 to 18, with an average of 9 ± 3. One genome harbored an amino acid deletion while 2 harbored two or one amino acid insertions.A total of 167 different amino acid substitutions were recorded in the set of 213 genomes, and only 33 of these were present in at least three genomes.The most predominant amino acid mutations were NS1:K649E, which also corresponds to UP1:K28E and was present in all of our genomes, followed by VP1:K613E corresponding to VP2:K524E and VP3:K484E, both present in 211 of the 213 genomes (Figure1).Stop codons were present in four genomes.Two of these genomes harbored the same stop codons, namely VP1:Q418*, corresponding to VP2:Q329*, and VP3:Q289*, therefore involved the VP1, VP2 and VP3 overlapping genes that encode the minor capsid proteins.These stop codons are located approximately in the middle of these genes.Another genome harbored ORFx:L93* and another VP2:G81* corresponding to VP3:G41*.ORFx:L93* is located near the end of the protein, while VP2:G81* corresponding to VP3:G41* is found at the beginning of the gene.

F I G U R E 1
Representation of amino acid substitutions in the 213 genomes obtained in the present study.F I G U R E 2 Representation of amino acid substitutions in NCBI genomes.
Here, we were able to obtain and characterize 213 HBoV1 genomes, which span the most recent 6 years period from 2017 to 2022.This represents approximately two-thirds of the 309 HBoV1 genomes available worldwide in GenBank at the beginning of our study, and therefore expands considerably the global set of genomes available.For these 309 genomes deposited in GenBank, no information regarding the date of sample collection was available in 32 cases.For the remaining 277 genomes, only 55 sequences were obtained from samples collected between 2017 and 2021.In addition, no HBoV1 genome from samples collected in 2022 was deposited in GenBank.The 213 genomes obtained in the present study were obtained from nasopharyngeal samples collected between 2017 and 2022, 64 being from samples collected in 2022.Moreover, none of the HBoV1 genomes available in GenBank as of June 05, 2023 originated from France.The present work therefore provides and reports on the first genomes from this country.All genomes obtained in the present study belong to the Bocaparvovirus primate 1 species, which is in line with previously F I G U R E 3 HBoV1 distribution by year.
variable positions present in at least 10% of the HBoV1 genomes delineated five genotypic clusters, and this enabled us identifying an association between HAdV and HBoV1 genotypic cluster α in coinfection.HAdV genotyping showed that HAdVC was the predominant HAdV during F I G U R E 5 Distribution of HBoV1 genotypes over time.

F I G U R E 6
Distribution per year and season of the numbers of diagnoses of HBoV1, HAdV and rhinovirus infections.

ap
(2017-2022).Therefore, we acknowledge that results could have been different if the study would have involved respiratory samples from patients who received cares outside hospitals or in other European or world areas, notably due to different indications or policies for the collection of respiratory samples, or due to differences in the panel of respiratory viruses tested.Nevertheless, the results found in this study can be expected to be similar to those from other geographical areas since we analyzed the diversity of all HBoV1 genomes available in GenBank regardless of the countries they originated and the time periods during which the samples were collected, and we found genotypic patterns comparable to those based on the sequences we obtained in the present study (Figure 2).However, more in-depth analyses of genetic changes over time and in different countries would be necessary to verify these results, and our findings would hopefully contribute generating interest from other laboratories in several countries to start implementing global F I G U R E 7 Principal component analysis of respiratory virus coinfections according to the HBoV1 genotypes.T A B L E 1 Prevalence of HAdV genotypes according to HBoV1 genotypic cluster α positivity or HBoV1 negativity.Five samples were coinfected by two HAdVs.b One sample was coinfected with two HAdVs.c p value = 0.08 (Fisher exact test).d p Value = 1 (Fisher exact test).T A B L E 2 Genotyping of HAdV C for HBoV1 genotypic cluster α-positive and HBoV1-negative samples.value = 1 (Fisher exact test).b p value = 0.37 (Fisher exact test).HBoV genomic surveillance.Also, we acknowledge that data availability for a longer study period could have provided here a more comprehensive picture of the genetic evolution and dynamics of HBoVs.Overall, the present study expands considerably the set of HBoV1 genomes available worldwide and provides the first genomes for France.It highlights limited genetic variability at the genomic level over a 5-year recent period of time, and unexpectedly, reveals an association between an HBoV1 genotypic cluster and HAdV in coinfection, which prompts conducting further studies, particularly considering the knowledge of auxiliary viruses in the same viral family.The clinical significance of these coinfections were not investigated in the present work, which only consisted in genome Therefore, this questions if HBoVs may have a similar role as helper.In fact, the present work opens the question of the interaction of certain HBoV genotypes with certain adenovirus genotypes, even if many other factors could account for this nonfortuitous association.Beyond, present data warrant performing genomic surveillance for HBoVs as already performed for influenza viruses, respiratory syncytial virus, and SARS-CoV-2. 62AUTHOR CONTRIBUTIONS Conceived and designed the experiments: Bernard La Scola, Philippe Colson.Contributed materials/analysis tools: Houmadi Hikmat, Lorlane Le Targa, Celine Boschi, Justine Py, Marielle Bedotto, Aurélie Morand, Nadim Cassir, Sarah Aherfi, Bernard La Scola, Philippe Colson.Analyzed the data: Houmadi Hikmat, Celine Boschi, Bernard La Scola, Philippe Colson.Wrote the paper: Houmadi Hikmat, Bernard La Scola, and Philippe Colson.