**4. Conclusion**

The genomic and structural diversity of parvovirus is today classified by phylogeny analysis showing an expected separation between parvoviruses and densoviruses, but its robustness is relative, suggesting that the introduction of new sequences could change our perception of their evolutionary history [32]. The diversity of sequences, structures, and genomic organizations of parvoviruses suggest evolutionary histories that are probably more complex than those illustrated by current phylogenies. These observations led us to analyze and characterize the intriguing terminal sequences present in all parvoviruses, namely the telomeres.

This chapter highlights the diversity of *Parvoviridae* telomeres through a complete analysis of the terminal ends folding and secondary structures. Their length, GC content, and global shape vary even within a genus between phylogenetically closely related viruses. Evolution also led to heterotelomeric viruses with completely different left and right extremities. The diversity suggests high importance of these particular structures. Yet, factors involved in the TR selective pressure are unknown. Cotmore and Tattersall suggested a link between the resolution mechanism, the strand polarity, and the TR conformation [4, 33], while Tijssen et al. suggested that the significant differences in size and secondary structure of genome end between genera might reflect a dependence on specific cellular factors necessary for replication and encapsidation [34]. Consistently, we hypothesized that TR may have evolved according to the interactions with their replicase, helper virus co-factors, and/or cell

host proteins. Based on data integration of predictive DNA secondary structures in a PCA, new groups were made that were distinct from the ICTV phylogenetic classification conducted from the NS replicase sequences.

Additionally, the significance of specific secondary structures in the parvovirus life cycle and the relation with strand polarity of the packaged linear genome are interesting topics deserving further investigations. The MVM, canine parvovirus (CPV), BPV1 (*Parvovirinae*), and the AalDV2 (*Densovirinae*) encapsidate only or predominantly negative-strand polarity genomes and possess heterotelomeric TR [35, 36]. On the contrary, the homotelomeric AAV2 encapsidates both strands polarities at the same level. By having different shapes and different secondary structure elements, the TR directly impacts the polarity of the encapsidated strand.

Finally, a special emphasis was put on the ITRs of the adeno-associated virus serotype 2, taking into consideration its importance in the world of gene transfer using viral vectors. Particular motifs and secondary structures within AAV ITR may have a significant impact on gene transfer efficiency. Indeed, it has already been demonstrated that AAV2 ITRs are detected by cellular factors belonging to the NHEJ and HR-DNA damage pathways [37]. The viral telomeres may also be recognized by DNA sensors which subsequently could restrict AAV vectors transduction or activate innate immune responses [21]. Consistent with this hypothesis, a variety of cellular proteins have been shown to interact with AAV2 ITR, such as nucleophosmin (NPM1), a protein involved in ribosome biogenesis and nucleolus transport of basic proteins. Notably, NPM1 binds preferentially G4. The restriction factor FKBP52 in its phosphorylated form also binds to the ITR in the D region, inhibits the second strand synthesis, and consequently decreases transgene expression [38]. Thus, the involvement of ITR recognition by cellular factors is central to understand the extent of subsequent responses to the rAAV DNA that can negatively impact the therapeutic gene expression and cause potential safety concerns for the patients. Using drastic parameters, no putative G4 or triplex were found in AAV2 ITR contrary to a previous study [14]. The formation of these non-conventional DNA motifs highly depends on the adjacent sequences as well as pH and ion concentration conditions and thus requires to be confirmed experimentally.
