Potential Viral Mechanisms Contributing to the Development and Persistence of Long COVID: The Possible Role of Viral Coinfections

2.1. Reactivation of Сhronic DNA Virus Infections in Patients with Long COVID

2.1.1. Reactivation of Latent Herpesviridae Infections in Patients with Long COVID

Herpesviruses are of particular interest because, following primary infection, they establish lifelong latency in specific host-cell reservoirs and can reactivate when antiviral immune surveillance is altered or impaired [44,50,114]. SARS-CoV-2 infection can induce sustained perturbations of innate and adaptive immunity, including alterations in T-cell responses and inflammatory signalling [23,24,27], thereby providing biologically plausible conditions that may facilitate herpesvirus reactivation [44,48,50]. Importantly, such reactivation should not necessarily be regarded solely as a consequence of immune dysregulation: it may, in turn, contribute to persistent immune activation and potentially modify the development or clinical course of Long COVID [27,38,42,46].

Long COVID is increasingly recognized as a heterogeneous condition involving multiple biological mechanisms that may differ between patients. Therefore, additional factors, including viral reactivation and possible interactions between concurrent viral infections, may contribute to this heterogeneity [27,45]. Recent reviews indicate that herpesvirus reactivation is among the virological mechanisms discussed in association with SARS-CoV-2 infection and post-acute sequelae [42,44]. However, the strength of evidence differs considerably among individual herpesviruses [42,44]. To date, the most substantial body of evidence concerns EBV, whereas evidence supporting a role for CMV and HHV-6 in Long COVID is more limited [38,42,44,46]. For HSV, VZV and other herpesviruses, the available evidence is substantially more limited and is derived predominantly from reports of acute COVID-19 and relatively small observational datasets [44,50].

Reactivation of Epstein–Barr virus (EBV) chronic infectionsin patients with Long COVID

Epstein–Barr virus (EBV; Human gammaherpesvirus 4) is highly prevalent in the adult population and establishes lifelong latency following primary infection [42,44]. Accumulating evidence suggests that SARS-CoV-2 infection may trigger EBV reactivation in a subset of patients [36,37,38,42,44]. Serological and molecular indicators of EBV reactivation, including EBV DNAemia and antibodies to viral antigens, have been investigated in relation to Long COVID and have been associated in some studies with fatigue, neurocognitive dysfunction, and other persistent symptoms [36,37,38,40,41]. However, these findings are not uniform across studies, and a causal role of EBV in Long COVID remains unproven [38,39,42].

Evidence linking EBV reactivation to Long COVID is suggestive but heterogeneous. In a cohort of 280 individuals recovering from SARS-CoV-2 infection, serological evidence of recent EBV reactivation was independently associated with Long COVID, particularly fatigue and neurocognitive impairment, although active EBV viremia was generally absent in the post-acute phase, suggesting that reactivation may have occurred during acute or early post-acute infection [37,38]. Other studies found no convincing evidence that EBV reactivation causes post-COVID syndrome, particularly after predominantly mild or asymptomatic COVID-19, indicating that associations may depend on acute-disease severity, timing, detection methods, and Long COVID phenotype [24,39]. In 2025, EBV-reactivation markers were reported more frequently in patients with Long COVID and persistent fatigue than in controls (28.6% vs 11.3%) [40]. Studies combining assessment of blood microaggregates and EBV-specific immune responses also suggest a possible association with part of the clinical phenotype, although retrospective therapeutic observations require cautious interpretation [41].

Reactivation of Cytomegalovirus (CMV) chronic infectionsin patients with Long COVID

Compared with EBV, considerably fewer studies have investigated CMV reactivation in patients with Long COVID. The available evidence suggests that CMV reactivation may occur in some patients following SARS-CoV-2 infection; however, its clinical significance and contribution to the development and persistence of Long COVID remain to be established. For example, Peluso et al. (2023) compared several viruses and the reactivation of latent infections caused by EBV, CMV, and human immunodeficiency virus (HIV) [38]. The authors made an interesting observation: while EBV has been shown to be associated with Long COVID, no such association was found for CMV. Furthermore, CMV seropositivity was associated with a lower likelihood of certain neurocognitive manifestations of Long COVID [38]. Similarly, a study by other authors examining the reactivation of EBV, CMV, and HHV-6 as possible contributors to Long COVID supports the idea that the evidence for CMV reactivation is significantly weaker than that for EBV [42].

Unlike EBV, most available evidence regarding CMV reactivation concerns the acute phase of COVID-19 rather than Long COVID. A systematic review by Chorya and Naik (2025), based on 15 publications (10 case reports and 5 case series) including a total of 34 hospitalized patients, demonstrated that CMV reactivation occurred predominantly in critically ill patients with severe acute COVID-19. Most patients required intensive care and invasive mechanical ventilation, with ICU stays ranging from 15 to 40.5 days. Overall, 19 of the 34 patients (55.9%) died, indicating that CMV reactivation was associated with poor clinical outcomes. The authors concluded that severe COVID-19–associated immune dysregulation, together with underlying comorbidities and the use of immunosuppressive therapy (particularly corticosteroids and tocilizumab), were major factors contributing to CMV reactivation. However, whether CMV reactivation contributes to the development or persistence of Long COVID remains uncertain because direct evidence is still limited [43].

A comprehensive multi-omics study of 1,154 hospitalized patients with acute COVID-19 demonstrated that chronic viral reactivation, including CMV, occurred primarily during severe acute SARS-CoV-2 infection and was associated with distinct host immune responses and adverse clinical outcomes [115]. Collectively, these studies demonstrate that CMV reactivation is well documented in severe acute COVID-19, whereas direct evidence linking CMV reactivation to the development or persistence of Long COVID remains limited and inconsistent [38,42,43,44,115].

A systematic review by Naderi et al. (2025) [44] identified CMV among the viruses most frequently reported to reactivate following SARS-CoV-2 infection. Reported CMV reactivation occurred predominantly in patients with severe acute COVID-19 and was frequently associated with immune dysregulation, corticosteroid or other immunosuppressive treatment, and unfavorable clinical outcomes [44].

Thus, as conclusion, reactivation of latent herpesviruses, including EBV and CMV, has been associated with more severe clinical outcomes and increased mortality in patients with acute COVID-19 and the biological mechanisms underlying this association remain incompletely understood.

Reactivation of Human herpesvirus 6 (HHV-6)and other herpesviruschronic infectionsin patients with Long COVID

Most published studies have reported for human herpesvirus 6 (HHV-6) without consistently distinguishing between HHV-6A and HHV-6B, which limits assessment of their potentially distinct contributions to COVID-19 and Long COVID. In a study of 61 hospitalized patients with moderate, severe, or critical acute COVID-19 in which HHV-6A and HHV-6B were analyzed separately by qPCR, HHV-6A DNA was not detected in any patient, whereas HHV-6B DNA was detected in approximately 10.7–15.2% of patients [47]. Thus, the available evidence specifically concerning HHV-6A remains particularly limited, and current data are insufficient to determine the respective contributions of HHV-6A and HHV-6B to either acute COVID-19 or Long COVID.

In addition to EBV and CMV, reactivation of HHV-6 has also been reported in Long COVID, although currently available evidence remains limited [42]. Compared with EBV, substantially fewer studies have investigated HHV-6 reactivation in Long COVID. HHV-6 reactivation has been reported in association with persistent fatigue and neuropsychiatric manifestations in some Long COVID cohorts; however, whether this reactivation contributes causally to these manifestations remains unknown [27,42,46]. Based on data from 90 patients with Long COVID, Maes et al. (2024) concluded that HHV-6 reactivation represents one of the key biological processes potentially contributing to Long COVID. Their findings suggest that persistent SARS-CoV-2 infection and HHV-6 reactivation, acting together with immune-inflammatory and autoimmune mechanisms, may contribute to the development and maintenance of chronic fatigue syndrome and neuropsychiatric manifestations in Long COVID [46]. In addition to EBV and CMV, this article also examines HHV-6—a member of the latent herpesvirus family and a likely reactivating virus—which may have been reactivated during SARS-CoV-2-induced immune response [27]. This article discusses the reactivation of EBV, HHV-6, and other herpesviruses as one of the possible mechanisms underlying Long COVID. However, it was noted that it has not yet been established whether HHV-6 reactivation is a cause of Long COVID, a consequence of immune dysregulation, or merely a concomitant phenomenon [27]. Thus, recent comprehensive reviews identify HHV-6 reactivation as one of the potential biological mechanisms associated with Long COVID. However, current evidence remains limited and inconsistent, and it is still unclear whether HHV-6 reactivation plays a causal role in disease pathogenesis, represents a consequence of SARS-CoV-2-induced immune dysregulation, or merely reflects an epiphenomenon. Further prospective studies using standardized diagnostic approaches are required to clarify its clinical significance. The authors link HHV-6 to chronic fatigue, affective symptoms, and a possible role in the pathogenesis of Long COVID [27,42,46]. An interesting comparative study using qPCR to detect the herpesviruses EBV, CMV, HHV-6A, and HHV-6B revealed that EBV was reactivated very frequently: 72.7% in cases of severe/critical COVID and 67.9% in cases of moderate COVID. HHV-6A was not detected in any patient, HHV-6B was detected, but significantly less frequently—in approximately 10.7–15.2% of patients—and CMV was also detected rarely—in 7.1–12.1% [47]. However, Rizzo S, Ferraresi M, Strazzabosco G, et al., 2026 [48], reached an interesting conclusion. The authors demonstrated that SARS-CoV-2 infection was associated with HHV-6A reactivation in a subset of patients, suggesting that SARS-CoV-2-induced immune dysregulation may facilitate reactivation of this latent herpesvirus. In other words, HHV-6A can be reactivated in the context of COVID-19. HHV-6A reactivation was significantly associated with the inhibitory KIR2DL2/HLA-C1 immunogenetic profile, suggesting that host genetic background may influence susceptibility to herpesvirus reactivation during SARS-CoV-2 infection [48]. These findings suggest that host genotype may influence susceptibility to the reactivation of specific latent viruses and may thereby contribute to interindividual differences in the clinical course of COVID-19.

Reactivation of other human herpesviruses, including HHV-7, HSV-1/2 and VZV, has also been reported during acute COVID-19 and, less frequently, in Long COVID [44,49]. However, compared with EBV and CMV, the available evidence remains limited and is based mainly on case reports or small observational studies. Therefore, their contribution to the pathogenesis of Long COVID requires further investigation. In addition, some authors, after analyzing 32 studies, concluded that herpesvirus reactivation is a common occurrence in patients with COVID-19, especially in severe cases [50], which indicates significant immune dysregulation during SARS-CoV-2 infection. They also emphasize that detecting herpesvirus reactivation may have clinical significance for patient management and the prevention of complications. The authors quantitatively assessed, for the first time, the frequency of reactivation of various herpesviruses in hospitalized patients with severe COVID-19: EBV(45%), HHV-7 (44%),HSV (38%), CMV (19%), HHV-8(19%); HHV-6 (18%) [50], The authors concluded that herpesvirus reactivation may have important clinical implications in severe COVID-19 and emphasized the need for further studies to clarify the interactions between herpesviruses and SARS-CoV-2. Further research is required to elucidate the interaction between HHVs and COVID-19 [50].

Thus, when considering herpesvirus reactivation in the context of COVID-19 and post-COVID-19, the information presented in Table 3 can be summarized as follows.

The evidence levels presented in Table 3 were assigned using a semi-quantitative assessment of the available literature. The grading considered (i) the number of independent studies reporting an association, (ii) study design and sample size, (iii) consistency of findings across studies, and (iv) whether the evidence was derived from case reports or small observational studies, larger cohort studies, prospective investigations, or systematic reviews. Evidence was classified as + (very limited evidence, predominantly individual case reports or small case series), ++ (limited evidence supported by several observational studies), +++ (moderate evidence supported by multiple independent studies, including larger cohorts or prospective investigations), or ++++ (relatively strong and consistent evidence supported by multiple independent studies and/or systematic reviews). The grading reflects the relative strength, consistency, and quality of the currently available published evidence and should not be interpreted as a quantitative meta-analytic assessment or as evidence of causality. The key studies supporting the evidence level assigned to each virus are cited in Table 3 and discussed in the corresponding sections of the text. For most herpesviruses, particularly CMV, HHV-6, HHV-7, HSV, and VZV, further prospective studies are required to determine whether viral reactivation contributes directly to the development of Long COVID or primarily reflects SARS-CoV-2-induced immune dysregulation. Moreover, most studies have evaluated individual herpesviruses separately, whereas potential interactions among multiple concurrently reactivated herpesviruses remain insufficiently investigated.

Collectively, these findings suggest that reactivation of latent herpesviruses is not restricted to a single viral species. Instead, SARS-CoV-2–associated immune dysregulation appears capable of disturbing the latency of several herpesviruses, although the magnitude of evidence differs substantially among individual viruses.

Based on these observations, we propose the hypothesis that concurrent reactivation of multiple latent viruses may exert additive or potentially synergistic effects on cellular stress responses, including DNA damage response (DDR) pathways. This possibility should not be interpreted as an established mechanism and requires direct experimental validation. It could be tested in controlled in vitro models by comparing the effects of single-virus and multiple-virus infections or reactivations on markers of DNA damage, DDR activation, chromosome instability, and cellular repair responses. Such an experimental approach would also extend earlier observations that viral infections can induce chromosome alterations and perturb cellular genome stability [90,91,92,93,94].

Because SARS-CoV-2 and several herpesviruses, particularly human cytomegalovirus (HCMV), have independently been reported to modulate the cellular DNA damage response (DDR) and DNA repair pathways [53,54,55], similar mechanisms raise the possibility that concurrent viral infections may exert additive or even synergistic effects on genome maintenance. This possibility has received little experimental attention and warrants further investigation.

2.2. Reactivation and Coinfections with Non-Herpesviruses During COVID-19 and Long COVID

In addition to herpesviruses, coinfections or reactivation involving several non-herpesviruses have been reported during SARS-CoV-2 infection. Coinfections with SARS-CoV-2 have been reported for influenza viruses, respiratory syncytial virus (RSV), adenoviruses and rhinoviruses, as well as for enteroviruses and several other viral pathogens [56,57,58,59,60,61,62,63,64]. However, their reported frequency and clinical significance of these coinfections vary considerably among studies. In the post-COVID era, the epidemiology and seasonality of several traditional respiratory viruses have changed, partly because of altered population immunity, changes in public-health measures and their renewed co-circulation with SARS-CoV-2 [63].

A systematic review and meta-analysis including 59 studies and 16,643 SARS-CoV-2-positive patients reported a pooled prevalence of respiratory viral coinfections of 5.01% (95% CI: 3.34–7.27%). Influenza viruses were identified in 1.54% of patients, whereas enterovirus detections were reported in 1.32%; however, differentiation between rhinoviruses and non-rhinovirus enteroviruses was not uniform across all diagnostic studies [59,63]. Coinfections were more frequent in children than in adults (9.39% vs. 3.51%, P=0.02) and were associated with a higher risk of fatal outcome (OR 1.66) [59]. In a pooled analysis of 11 prospective cohorts, de Hoog et al. identified another respiratory virus in 7.7% of 1,606 SARS-CoV-2-positive episodes, with rhinovirus being the most frequently detected copathogen [58].

RSV is commonly evaluated together with influenza viruses and other respiratory pathogens [62]. Most available evidence concerns co-circulation, seasonality, disease severity, hospitalization, and other clinical outcomes during acute SARS-CoV-2 infection rather than mechanisms contributing to Long COVID [56,58,63]. SARS-CoV-2–adenovirus and enterovirus coinfections have also been documented [63], although much of the available evidence is derived from individual cases, small observational series, or studies involving patients with underlying disorders [56,58]. Importantly, longitudinal studies specifically assessing whether acute respiratory viral coinfections modify the subsequent risk, phenotype, or severity of Long COVID remain scarce. Thus, although respiratory viral coinfections may influence the severity and immune response during acute COVID-19, whether these effects translate into an altered risk of persistent post-acute sequelae remains largely unknown and represents an important gap in the current literature.

Clinical evidence for coinfection with non-rhinovirus enteroviruses remains particularly limited. Many diagnostic studies report combined rhinovirus/enterovirus results, and only small numbers of strictly identified enterovirus–SARS-CoV-2 coinfections have been described [63,64]. Available studies indicate that enteroviruses may occur as coinfecting pathogens, supporting the use of multiplex diagnostic approaches when clinically appropriate [64]. Consequently, the clinical consequences of these combinations remain uncertain.

Several studies have associated SARS-CoV-2–influenza coinfection, particularly influenza A coinfection, with an increased risk of severe clinical outcomes, including ICU admission, mechanical ventilation, and mortality [108,109]. However, other studies have reported no significant increase in mortality or severe outcomes, suggesting that the clinical consequences of viral coinfection may depend on the coinfecting virus, patient population, and epidemiological context [110,111]. Importantly, viral interactions during coinfection are not necessarily synergistic. They may also be antagonistic because of viral interference.

Svyatchenko et al. investigated simultaneous and sequential co-infection of SARS-CoV-2 with either the live attenuated Coxsackievirus A7 vaccine strain LEV-8 or the pathogenic enterovirus A71 (EV-A71) [65]. In Vero E6 cells, both enteroviruses markedly interfered with SARS-CoV-2 replication, with the competitive inhibitory effect being more pronounced against SARS-CoV-2 than against the enteroviruses themselves. In Syrian hamsters, enterovirus pre-infection resulted in a more than 100-fold reduction in SARS-CoV-2 titres in the respiratory tract, accelerated clearance of infectious SARS-CoV-2 from the lower respiratory tract, and attenuated clinical manifestations and SARS-CoV-2-induced pulmonary pathology. The authors suggested that this viral interference may involve competition between viruses and activation of innate antiviral responses, including interferon- and cytokine-mediated mechanisms. These findings provide experimental evidence of viral interference in cellular and animal models [65], but should not be interpreted as evidence that naturally occurring enterovirus co-infection exerts a protective effect against SARS-CoV-2 infection or its post-acute consequences in humans.

Thus, cellular processes during co-infections may vary; one possible scenario involves viral interference that suppresses SARS-CoV-2 replication. These findings demonstrate that attenuation does not eliminate the capacity of a live virus to replicate, activate innate immune pathways and modify the replication of another virus. However, these results were obtained in cell culture and animal models and cannot yet be directly extrapolated to naturally occurring coinfections in humans or to Long COVID. Whether active viral replication is also required for virus-induced alterations in DNA damage-response pathways or genome stability remains unknown and requires direct experimental investigation. So, let’s take a moment to summarize. Clinical evidence for enterovirus coinfection with SARS-CoV-2 remains limited and is based primarily on small observational studies. Moreover, many multiplex diagnostic assays report combined rhinovirus/enterovirus results, making the contribution of individual enteroviruses difficult to assess. Experimental evidence suggests that certain enteroviruses may antagonize SARS-CoV-2 replication through viral interference; however, these findings have not yet been confirmed in naturally occurring human coinfections.

Reactivation or altered persistence of chronic viruses, including HBV, HPV, and polyomaviruses, has also been described, although the available evidence differs substantially among individual viruses and clinical settings [66,68,69,70].

With regard to human hepatitis B virus (HBV), the available evidence primarily concerns the potential reactivation of chronic HBV infection, particularly in the context of immunosuppressive treatment for COVID-19, including corticosteroid therapy [66]. Although current data do not indicate a substantial overall increase in the risk of HBV reactivation associated with COVID-19, this remains clinically relevant in HBsAg-positive and immunocompromised patients [66].

A large multicenter retrospective cohort study from the United States included 4,206,774 patients with COVID-19, of whom 8,293 had chronic HBV infection [67]. Chronic HBV infection was associated with worse clinical outcomes following SARS-CoV-2 infection, including increased odds of ICU admission and mortality, particularly among patients with cirrhosis. An increased risk of all-cause mortality was also observed in the overall chronic HBV cohort (OR 1.18; 95% CI 1.06–1.33) [67]. Importantly, these findings concern outcomes following acute SARS-CoV-2 infection and do not establish a role for chronic HBV infection or HBV reactivation in the pathogenesis of Long COVID.

Şahin et al. (2026) reported that a history of COVID-19 was independently associated with persistence of high-risk human papillomavirus (HPV), particularly HPV16 [68]. The authors discussed prolonged SARS-CoV-2-associated immune dysregulation, including alterations in T-cell and interferon responses, as a potential explanation for impaired viral clearance. These findings suggest that the immunological consequences of SARS-CoV-2 infection may influence the persistence of another chronic viral infection. However, the study did not establish that HPV persistence contributes to the pathogenesis or clinical manifestations of Long COVID.

Meshram et al. (2021) described the development or intensification of BK polyomavirus (BKPyV) replication following COVID-19 in kidney transplant recipients [69]. The authors suggested that even mild SARS-CoV-2 infection may disrupt immune control over latent viruses, thereby facilitating their reactivation [69]. However, because these patients were transplant recipients receiving immunosuppressive therapy, the contribution of SARS-CoV-2 infection cannot be clearly separated from the effects of underlying immunosuppression. Thus, these observations support the biological plausibility that SARS-CoV-2-associated immune perturbation may facilitate loss of control over latent viruses, but they do not establish BKPyV reactivation as a mechanism of Long COVID [69]. Evidence regarding JC polyomavirus (JCPyV) in this context is currently even more restricted, and no firm conclusions regarding its contribution to Long COVID can be drawn.

Taken together, these observations suggest that the immunological consequences of SARS-CoV-2 infection may influence the persistence or reactivation of other chronic or latent viruses. On this basis, we propose the hypothesis that SARS-CoV-2-induced immune dysregulation may create conditions favoring the persistence or reactivation of otherwise controlled viruses, which could, in turn, contribute to the heterogeneity or persistence of Long COVID in susceptible individuals. This proposed relationship remains to be tested directly.

Pessoa-Gonçalves et al. (2023) described a case of progressive multifocal leukoencephalopathy (PML) in which SARS-CoV-2 and JC polyomavirus (JCPyV; HPyV-2) were simultaneously detected in the cerebrospinal fluid [70]. The authors proposed that COVID-19-associated inflammation and immune dysregulation might have contributed to JCPyV reactivation and the development or progression of PML. However, as this observation was based on a single case, it provides evidence of a possible temporal and biological association rather than a causal relationship between SARS-CoV-2 infection and JCPyV reactivation [70]. Moreover, it does not provide direct evidence for a role of JCPyV in Long COVID.

Human immunodeficiency virus (HIV) is a globally prevalent human retrovirus and the causative agent of acquired immunodeficiency syndrome (AIDS). People living with HIV (PLWH) represent a heterogeneous population with respect to COVID-19 outcomes [116]. Emerging evidence suggests that PLWH may also experience an increased burden of post-acute sequelae following SARS-CoV-2 infection. In a longitudinal cohort, Peluso et al. (2022) reported a higher prevalence of PASC among PLWH and identified differences in SARS-CoV-2-specific immune responses associated with post-acute symptoms [116]. A subsequent systematic review and meta-analysis including 17 studies and 39,405 PLWH with COVID-19 estimated that approximately 52% experienced at least one Long COVID symptom and found increased odds of Long COVID compared with people without HIV (OR 2.20; 95% CI 1.25–3.86) [117]. However, estimates vary substantially among studies, and a more recent systematic review and meta-analysis of eight studies (4,489 participants) found that the overall association between HIV status and Long COVID remained inconclusive (OR 1.16; 95% CI 0.58–2.29), although an increased risk was observed when the analysis was restricted to covariate-adjusted estimates (OR 2.21; 95% CI 1.12–4.36) [118].

HIV disease status may also be important when interpreting these associations. Well-controlled HIV infection under effective ART, with suppressed HIV viral load and preserved CD4+ T-cell counts, has generally been associated with more favorable COVID-19 outcomes than advanced or uncontrolled HIV infection. Low CD4+ T-cell counts, persistent HIV viremia, absence of effective ART, and comorbidities have been associated with increased risks of severe acute COVID-19 and mortality [71,72,73,74,75]. Of particular relevance to the present review, severe HIV-associated immunodeficiency may permit prolonged SARS-CoV-2 replication, providing a setting in which impaired antiviral immune control can influence the persistence and evolution of another virus [71,72,73,74,75]. Such observations provide a biological rationale for considering HIV-associated immune dysfunction as a potential modifier of post-acute SARS-CoV-2 outcomes, although prolonged SARS-CoV-2 replication in advanced HIV infection should not itself be regarded as evidence of a mechanism underlying Long COVID.

Collectively, current evidence indicates that SARS-CoV-2-associated immune dysregulation may influence a broad spectrum of latent, persistent, and acute viral infections. However, the strength of evidence varies markedly among individual viruses. While herpesvirus reactivation is supported by the largest body of evidence, observations involving HPV, polyomaviruses, HBV, and other chronic viral infections remain limited. Importantly, most available studies evaluate individual viruses separately. To our knowledge, the potential biological consequences of concurrent or sequential infections involving SARS-CoV-2 and multiple other viruses, particularly their combined effects on genomic stability, DNA damage, and chromosomal alterations, remain largely unexplored. This knowledge gap forms the basis of our hypothesis that concurrent or sequential viral infections may produce additive or synergistic cellular stress, resulting in greater perturbation of DNA damage response (DDR) pathways and genomic stability than that induced by SARS-CoV-2 infection or individual viral infections alone. This hypothesis is testable by directly comparing single-virus, concurrent, and sequential infection models using quantitative markers of DNA damage, DDR activation, chromosome instability, and cellular repair responses.

3.1. Genomic Alterations in Human Cells Induced by SARS-CoV-2 During Acute COVID-19 and Long COVID

Genomic instability and dysregulation of DNA damage response (DDR) pathways may be relevant to Long COVID because persistent DNA damage, oxidative stress, and impaired cellular repair mechanisms can contribute to prolonged cellular dysfunction, inflammatory signalling, and altered immune responses. In the context of viral persistence, reactivation, or co-infection, these effects may not necessarily arise from SARS-CoV-2 alone. We hypothesize that concurrent or sequential exposure to multiple viruses may impose additional genotoxic and cellular stress, potentially producing additive or synergistic perturbations of DDR pathways and genomic stability. Thus, examining virus-induced genomic alterations provides a mechanistic framework for assessing whether viral interactions could contribute to the persistence and heterogeneity of Long COVID.

The capacity of viral infections to induce genotoxic stress has been recognized for several decades, beginning with early observations of virus-induced chromosome damage and subsequently extending to direct DNA damage, DDR dysregulation, genomic instability, and oxidative stress [22,25,35,52,53,54,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94].

A related but distinct level of alterations involves genomic instability and structural chromosomal damage, including chromosome aberrations and micronucleus formation [54,75,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96]. Micronuclei are particularly relevant because they represent not only markers of chromosomal instability but also potential mediators of inflammatory signalling through cytosolic DNA-sensing pathways, including cGAS–STING [92,94,95,96]. Thus, persistent DNA damage or impaired DDR may contribute to genomic instability, but these phenomena should not be considered synonymous.

The genotoxic and DDR-related effects of SARS-CoV-2 identified during the early years of the pandemic have been reviewed previously [54,55]. Therefore, rather than reiterating these observations, the present section focuses primarily on more recent evidence addressing the persistence of DNA damage, genomic instability and DDR alterations beyond acute infection and their potential relevance to Long COVID.

More recent studies suggest that alterations associated with genome integrity may persist beyond the acute phase of COVID-19. In patients examined more than four weeks after infection, persistent DNA damage detected by comet assay and evidence of genomic instability were reported, with the extent of these alterations associated with the severity of the preceding acute disease [97]. These observations have raised questions regarding possible links with cellular senescence, chronic inflammation, autoimmunity, and other long-term consequences of SARS-CoV-2 infection [97]. A further, mechanistically distinct level involves persistent epigenetic and transcriptional alterations. Changes in DNA methylation, miRNA profiles, and long-term gene regulation have been reported in patients with Long COVID [98,112]. Such epigenetic alterations do not themselves demonstrate persistent DNA damage or defective DNA repair; rather, they indicate prolonged changes in genome regulation that may coexist with, or potentially arise downstream of, cellular stress and altered DDR pathways. The extent to which these processes are causally connected remains to be established.

Recent mechanistic studies have further expanded our understanding of the interactions between SARS-CoV-2 and host DNA damage response pathways. Gioia et al. (2023) demonstrated that SARS-CoV-2 induces DNA damage and interferes with DDR (direct DDR disruption) through at least two distinct mechanisms. The viral proteins ORF6 and NSP13 promoted CHK1 degradation, leading to reduced RRM2 levels and depletion of the cellular dNTP pool, thereby inducing replication stress and DNA damage. In parallel, the N protein impaired recruitment of 53BP1 to DNA damage sites, resulting in defective non-homologous end joining (NHEJ) and reduced DNA repair efficiency [113]. These alterations were associated with genomic instability, inflammatory signalling, and cellular senescence and were demonstrated in vitro, in infected mice, and in samples from patients with COVID-19 [113].

A complementary mechanism was identified by Tapryal et al. (2023), who demonstrated that SARS-CoV-2 infection affects the DNA glycosylase NEIL2, a component of the base-excision DNA repair machinery [119]. Reduced NEIL2 levels were observed in the lungs of patients with severe COVID-19 and were further confirmed in SARS-CoV-2-infected patients, hamsters, and human A549-ACE2 cells. Importantly, NEIL2 appeared to have a dual function linking maintenance of host genome integrity with antiviral defence: in addition to its role in DNA repair, NEIL2 interacted with the 5′ untranslated region of the SARS-CoV-2 genomic RNA and inhibited viral protein synthesis [119]. These findings therefore suggest a direct mechanistic intersection between host DNA repair pathways and cellular control of SARS-CoV-2 replication.

More recently, Zhao et al. (2025) demonstrated another dimension of the SARS-CoV-2–DDR interaction by showing that Spike-induced cell fusion activates the ATR–CHK1 and ATM–CHK2 signalling axes and promotes accumulation of the DNA damage-associated proteins γH2AX, 53BP1, and RAD51 in syncytia [120]. Pharmacological inhibition or siRNA-mediated depletion of ATM or ATR reduced syncytium formation through decreased Spike protein levels, indicating that DDR activation is not merely a consequence of SARS-CoV-2-induced cellular stress but may also contribute to stabilization of the Spike protein and promotion of cell fusion [120]. These observations raise the possibility that SARS-CoV-2 can exploit components of the host DDR machinery to facilitate virus-induced cellular alterations.

Wang et al. (2023) extended these observations from individual DNA lesions and repair pathways to higher-order genome organization. Using three-dimensional genomic and epigenomic profiling, the authors demonstrated extensive restructuring of host chromatin after SARS-CoV-2 infection, including weakening of the active A compartment, increased A–B compartment mixing, reduced intra-TAD contacts, depletion of cohesin from intra-TAD regions, and decreased H3K27ac. These changes were accompanied by suppression of interferon-response genes and increased H3K4me3 at promoters of strongly induced pro-inflammatory genes [121]. Thus, SARS-CoV-2 may affect not only DNA damage and repair but also the higher-order spatial organization and regulation of the host genome.

Nong et al. (2024) identified an additional mechanism linking the SARS-CoV-2 Spike protein to cellular senescence. Spike increased Cdc42 activity and activated Wnt/β-catenin signalling, promoting a senescent phenotype [122]. Inhibition of Cdc42 reduced cellular senescence, lung injury, and inflammation, suggesting that the Cdc42–Wnt/β-catenin axis contributes functionally to Spike-induced cellular stress and tissue damage [122].

Hornung et al. (2023) further showed that virus-induced cellular responses may differ between SARS-CoV-2 variants. Using primary human alveolar epithelial cells in vitro and human lung slices ex vivo, the authors compared Delta (B.1.617.2) and Omicron (B.1.1.529) and observed a distinct Omicron-associated senescence programme [123]. Omicron infection preferentially altered cell-cycle, inflammatory, and integrin-related pathways and was associated with increased expression of p21, p16, and p38 together with an enhanced senescence-associated secretory phenotype (SASP) [123]. These findings indicate that the magnitude and qualitative pattern of virus-induced cellular senescence may depend on the viral variant.

A further mechanistic link between DNA damage, inflammation, and tissue dysfunction was demonstrated in glial models. SARS-CoV-2 infection of human iPSC-derived astrocytes and HMC3 microglial cells induced cellular senescence and activation of the pro-inflammatory cGAS–STING pathway. In primary rat cortical cultures, infected glial cells were associated with increased DNA-damage foci, enhanced inflammatory mediator production, loss of synaptic connections, and impaired neuronal-network activity. Pharmacological inhibition of cGAS–STING partially ameliorated the early decline in neuronal electrical activity, supporting a functional connection between SARS-CoV-2-induced glial damage, cGAS–STING-dependent inflammation, and neuronal dysfunction [124].

Recent evidence also indicates that coronavirus-induced stress can reshape host RNA modification pathways. SARS-CoV-2 and HCoV-OC43 reprogrammed specific host tRNA modifications by altering the expression of tRNA-modifying enzymes, thereby improving translation of viral proteins [125]. Because both viruses induced DNA damage and oxidative stress—conditions known to alter the tRNA epitranscriptome—these findings suggest that coronaviruses can exploit stress-associated host responses to optimize their own protein synthesis [125].

Although mechanistic studies provide substantial evidence that SARS-CoV-2 can interfere with DNA integrity and DNA damage-response pathways, considerably fewer studies have directly examined whether genomic damage persists in humans after the acute phase of infection. In a longitudinal pilot study of individuals with post-COVID fatigue, Hofmann et al.(2023) detected oxidative stress and DNA strand breaks by comet assay approximately three months after recovery. During follow-up eight weeks later, improvement in fatigue was accompanied by a significant reduction in reactive oxygen species and a decrease in DNA damage that narrowly missed statistical significance (p=0.053), suggesting a possible relationship between persistent oxidative stress, genomic damage and post-COVID symptoms [126].

However, persistent DNA damage has not been demonstrated consistently across post-COVID cohorts. Martins et al. examined 231 participants, including controls and individuals with acute or chronic post-COVID conditions, and found no significant differences in overall DNA damage measured by alkaline comet assay among the three groups [127]. Nevertheless, sex-specific differences were observed, with higher DNA damage in men with acute post-COVID condition than in women, while age, sex and waist circumference were identified as significant predictors of DNA damage [127]. A subsequent study by the same group expanded the analysis to alkaline and enzyme-modified comet assays together with inflammatory, hematological and metabolic parameters. Oxidative DNA lesions were detected, but again no significant overall differences in DNA damage were observed among control, acute post-COVID and chronic post-COVID groups [129]. These findings indicate that persistent genomic damage may not be a uniform feature of post-COVID conditions and may depend on individual biological and clinical factors.

In contrast, Abiri et al. (2025) also reported DNA damage during the post-COVID period, with higher levels associated with greater severity of the preceding disease [128]. In a subsequent study of 29 post-COVID patients examined four weeks after a positive RT-PCR test, the same group found significantly greater DNA damage than in healthy controls, with a clear severity-dependent gradient: the highest levels occurred in previously intubated ICU patients, followed by non-intubated ICU and non-ICU patients [130]. Thus, these observations support an association between the severity of acute COVID-19 and the magnitude of persistent post-acute DNA damage, although the relatively small sample size limits the strength of this conclusion.

More recent evidence further implicates persistent redox imbalance as a potential contributor to post-COVID DNA damage. Dikshit et al. examined symptomatic post-COVID patients and age- and sex-matched healthy controls and demonstrated substantial depletion of antioxidant defences together with significantly increased DNA damage across all comet-assay parameters; % tail DNA was approximately 24% higher in the post-COVID group [131]. The accompanying disruption of normal relationships among antioxidant and oxidative-stress markers suggests that persistent DNA damage may occur within a broader alteration of redox homeostasis rather than as an isolated genomic phenomenon [131].

Collectively, these human studies provide evidence that DNA damage may persist beyond acute SARS-CoV-2 infection in at least a subset of individuals, particularly following more severe disease and in association with persistent oxidative stress [126,128,130,131]. However, the absence of significant differences in genomic damage in other comparatively large post-COVID cohorts [127,129] indicates that this phenomenon is neither universal nor yet sufficiently characterized. Differences in study populations, timing after infection, severity of the preceding disease, clinical phenotype and methods used to assess DNA damage may contribute to these apparently divergent findings. Thus, current human evidence supports persistent genomic damage as a potential component of post-COVID pathophysiology, but not as an established universal feature of Long COVID.

3.2. Chromosomal and Genomic Instability Associated with SARS-CoV-2 Infection

A total of 100 patients with COVID-19 were examined using Papanicolaou staining to assess the presence of micronuclei (MN) and the frequencies of pyknotic, karyolytic, and karyorrhectic cells. SARS-CoV-2 infection was confirmed by reverse transcription polymerase chain reaction (RT-PCR). The study demonstrated a significantly increased frequency of micronucleus formation, indicating a genotoxic effect on buccal mucosal cells. Specifically, the numbers of MN, micronucleated cells, pyknotic cells, karyolytic cells, and karyorrhectic cells in patients with COVID-19 were 24.12, 15.24, 3.08, 2.88, and 4.40, respectively, compared with 5.69, 8.17, 1.08, 1.00, and 2.43, respectively, in the control group (COVID-19-negative individuals). The genotoxicity ratio for SARS-CoV-2 was therefore calculated as 2.68. Thus, the authors demonstrated that SARS-CoV-2 infection was associated with increased frequencies of micronucleated, pyknotic, karyolytic, and karyorrhectic cells, providing evidence of a cytogenotoxic effect of SARS-CoV-2. Importantly, these cytogenetic alterations may serve as useful markers for the early detection of potential carcinogenic effects associated with SARS-CoV-2 infection [132]. Micronuclei are well-established cytogenetic biomarkers of genomic damage and chromosomal instability; their formation may result from the loss of whole chromosomes and/or the generation of acentric chromosome fragments following chromosomal damage [133,134,135]. Micronuclei may arise from lagging acentric chromosome or chromatid fragments generated as a consequence of misrepair or failure to repair DNA strand breaks [136]. Consistent with this mechanism, increased MN frequencies have been reported in cells with defective DNA damage repair systems and impaired checkpoint machinery [137]. A second study of SARS-CoV-2-infected patients, also using Papanicolaou staining together with immunohistochemical assessment of oxidative DNA damage, confirmed these observations. The numbers of micronucleated cells, total micronuclei, neutrophils, lymphocytes, and inflammatory cells in RT-PCR-positive COVID-19 patients were 11.32 ± 7.00, 17.32 ± 12.53, 1.93 ± 1.17, 3.98 ± 2.55, and 5.90 ± 3.15, respectively, and were increased compared with the control cohort (RT-PCR-negative for COVID-19), in which the corresponding values were 6.09 ± 3.83, 9.04 ± 5.82, 0.18 ± 0.49, 2.13 ± 0.84, and 2.31 ± 1.01. In addition, increased oxidative DNA damage was observed [138].

Taken together, these studies provide convergent human cytogenetic evidence that acute SARS-CoV-2 infection is associated with increased micronucleus formation, linking virus-induced DNA damage and impaired repair processes to chromosome-level genomic instability. However, micronucleus assays do not establish whether such damage progresses to persistent structural or numerical chromosome alterations. Direct assessment of classical chromosomal aberrations or aneuploidy by metaphase chromosome analysis/karyotyping remains a major unresolved gap, particularly in post-COVID and Long COVID.

Importantly, these mechanisms may not be unique to SARS-CoV-2. Different viruses can perturb overlapping cellular pathways involved in DNA repair, genome maintenance, and inflammatory signalling, raising the possibility that concurrent or sequential viral infections could converge on common mechanisms of genomic stress.

3.2.1. Host Genomic Background, Chromatin Reorganization, and Broader Genomic Consequences of SARS-CoV-2 Infection

Notably, a study of 9,578 individuals diagnosed with COVID-19 identified 133 cases (1.42%) of mosaic chromosomal alterations (mCA) and 226 men (5.08%) with acquired loss of the Y chromosome (LOY). In individuals with clonal mosaic alterations (mCA and/or LOY), the risk of death from COVID-19 was increased by 54%. LOY was associated with transcriptomic biomarkers of immune dysfunction, procoagulant activity, and cardiovascular risk. Interferon-induced genes involved in the initial immune response to SARS-CoV-2 were also suppressed in the presence of LOY. Thus, mCA and LOY underlie, at least in part, sex-related differences in COVID-19 severity and mortality among older patients (more frequently men), which should be taken into account [139]. It may be assumed that in individuals with pre-existing chromosomal alterations, the additional impact of SARS-CoV-2 may aggravate these abnormalities, leading to more severe consequences.

In addition, it has been discussed that specific genetic and structural features of the X chromosome may contribute to interindividual and sex-related differences in susceptibility to SARS-CoV-2 infection and in the clinical severity of COVID-19. In particular, it has been proposed that single-nucleotide polymorphisms (SNPs), rearrangements, and epigenetic alterations affecting X-linked genes involved in viral entry and immune-response regulation, including ACE2, TLR7/8, FOXP3, and RGN, influence SARS-CoV-2-associated lung pathology, disease severity, and mortality [140]. Therefore, it can be concluded that the genetic background at the chromosomal level may not only modulate the host response to SARS-CoV-2 but also contribute to the marked heterogeneity of clinical outcomes observed in COVID-19.

The functional significance of this host genetic and chromosomal background extends beyond susceptibility to infection and disease severity, as demonstrated by its measurable influence on the magnitude of the adaptive immune response to SARS-CoV-2 antigens. Accordingly, vaccine immunogenicity is also influenced by the genetic background of the vaccinated individual [141]. Post-vaccination titres of SARS-CoV-2-specific antibodies were associated with the immunoglobulin heavy-chain (IGH) locus and the major histocompatibility complex (MHC) locus, whereas T-cell responses were associated with the MHC. Variants associated with antibody titres modulate circulating immunoregulatory proteins (e.g., LILRB4 and FCRL6). Age-related expanded hematopoietic mosaic chromosomal alterations (mCA) involving the MHC and IGH also impair antibody production. mCA affecting the MHC and IGH increase the risk of infectious and immune-mediated diseases, including sepsis and Graves’ disease. The effects of expanded mosaic loss of the X/Y chromosomes on these phenotypes were also examined. It was concluded that both germline and somatic mutations influence adaptive immune functions [141]. SARS-CoV-2 genetic variability further modifies this host–virus interaction: distinct variants differ in transmissibility, pathogenesis, immune escape, and susceptibility to vaccine-induced, monoclonal, polyclonal, and convalescent antibodies [142,143,144,146]. These variant-specific effects may interact with host genetic background, contributing to heterogeneous consequences of infection and vaccination. Comparative analysis of 44 Sarbecovirus genomes also showed that the functional protein-coding capacity of SARS-CoV-2 remains incompletely resolved, complicating attribution of potential genotoxic effects to individual viral genes or proteins [146].

Tan et al. (2023) identified SARS-CoV-2-induced alterations at the level of spatial chromatin organization, including changes in the position, morphology, and intermingling of chromosome territories; in particular, chromosome 1 shifted toward the nuclear periphery in infected patients. At the same time, disruptions of interchromosomal interactions and reorganization of local chromatin structure at the level of (sub)TADs were detected. Although these are not classical chromosome aberrations in the cytogenetic sense, they bring the observations closer to alterations of the hereditary apparatus of the cell at the chromosomal level rather than merely molecular DNA damage [145].

We also investigated possible genomic mechanisms underlying the rapid development of and recovery from anosmia in patients, a potential diagnostic indicator of the early stage of COVID-19 infection. Based on previous observations of how olfactory receptor (OR) gene expression is regulated through chromatin structure in mice, we hypothesized that disruption of OR gene expression and, consequently, impairment of OR function could be caused by chromatin reorganization occurring during SARS-CoV-2 infection. We obtained chromatin ensemble reconstructions from patients with COVID-19 and control samples using our original computational platform for reconstructing three-dimensional whole-genome chromatin ensembles. Specifically, we used megabase-scale structural units and effective interactions between them, derived from Markov-state modelling of the Hi-C contact network, as input for a stochastic embedding procedure to reconstruct three-dimensional whole-genome chromatin ensembles. In addition, we developed a new method for analysing fine structural hierarchy using (sub)TAD-sized units in local chromatin regions, which we applied to chromosomal regions containing OR genes and the corresponding regulatory elements. We observed structural changes in patients with COVID-19 at different levels of chromatin organization, ranging from alterations in whole-genome structure and chromosome intermingling to reorganization of contacts between chromatin loops at the level of topologically associating domains [145]. While additional data on known regulatory elements are required for a more complete understanding of the mechanisms regulating olfactory receptor genes, the findings indicate a possible link between SARS-CoV-2-induced reorganization of chromatin architecture, disruption of OR-gene expression, and the development of anosmia [145].

At the same time, the effects of individual SARS-CoV-2 proteins on the cellular genetic apparatus may exhibit pronounced substrate specificity. Li et al. showed that the viral helicase nsp13 directly binds to episomal DNA and suppresses its transcription through NTPase- and helicase-dependent mechanisms, whereas no comparable suppression of the expression of chromosomally integrated genes was detected. Thus, these data limit the assumption of a direct effect of nsp13 on chromosomal DNA transcription and simultaneously emphasize the need to distinguish the effects of SARS-CoV-2 on different forms and levels of organization of cellular genetic material [147].

Although these are epidemiological observations rather than evidence of a genotoxic mechanism of SARS-CoV-2 effects on pregnancy, it should be noted that among 1,127 pregnancies, the authors found a higher frequency of congenital defects following SARS-CoV-2 infection in the first trimester; chromosomal abnormalities were among the types of defects observed [148]. Evidence regarding the association between pregestational or early-pregnancy SARS-CoV-2 infection and congenital anomalies remains inconsistent. Although large studies have not demonstrated a consistent association, epidemiological observations during the pandemic reported increased rates of specific congenital defects, including neural tube defects, situs inversus, and chromosomal anomalies; however, a causal role of SARS-CoV-2 has not been established. Importantly, potential effects of maternal infection on fetal development may not require transplacental viral transmission and could be mediated by maternal immune responses, pro-inflammatory cytokines, hypoxia, and fever [149].

Caution is warranted when interpreting reports suggesting that the cytogenetic profile of hematologic malignancies remained largely stable throughout the COVID-19 pandemic. In particular, the striking decrease in abnormal karyotypes observed in CLL during the post-COVID period (5.7% vs. 40.0% pre-COVID; p < 0.001) was not robust in sensitivity analysis and, in the absence of individual SARS-CoV-2 infection or Long COVID data, cannot be unequivocally attributed to changes in healthcare access or referral patterns rather than to potential biological effects of infection; therefore, its biological significance remains uncertain [150].

In a multicenter cohort study of 585 infertile couples, Tian et al. found that SARS-CoV-2 infection during controlled ovarian stimulation was associated with significantly impaired embryo development, including reduced rates of top-quality embryos (47.3% vs. 52.1%; p = 0.01), blastocyst formation (33.3% vs. 45.3%; p < 0.001), available blastocysts (28.2% vs. 35.8%; p < 0.001), and top-quality blastocysts (7.1% vs. 11.5%; p < 0.001) compared with SARS-CoV-2-negative couples [151]. Importantly, the absence of a demonstrated association in individual cytogenetic studies should not be interpreted as evidence for the absence of biological effects of SARS-CoV-2, particularly when infection status, post-COVID conditions, and potentially relevant cellular or molecular endpoints were not directly assessed [150,151].

SARS-CoV-2 RNA has been detected in semen or seminal plasma in a small proportion of men during acute infection and after recovery, with a pooled detection rate of approximately 7%; moreover, post-COVID alterations in spermatogenesis may persist for at least one complete spermatogenic cycle after recovery, although some sperm parameters tend to normalize over time [156]. A systematic review and meta-analysis of seven studies demonstrated a significant increase in sperm DNA fragmentation during the post-COVID period, temporally corresponding to the Long COVID timeframe, persisting for 2–3 months following SARS-CoV-2 infection (WMD = 12.558; 95% CI: 4.482–20.635; p < 0.0001), indicating persistent impairment of sperm DNA integrity; however, substantial between-study heterogeneity (I² = 99%) warrants caution in estimating the magnitude of this effect [152]. This finding is supported by a larger meta-analysis of 39 studies from 15 countries, including 1,887 men with a history of SARS-CoV-2 infection and 2,097 controls, which evaluated multiple semen parameters and identified the sperm DNA fragmentation index (DFI) as a separate endpoint. Based on four studies reporting DFI, sperm DNA fragmentation was significantly increased following SARS-CoV-2 infection (MD = 4.01; 95% CI: 3.02–5.00; p < 0.001) [153]. The meta-analysis additionally categorized follow-up as short-term (<90 days) and long-term (≥90 days), allowing post-infectious changes in male reproductive parameters to be considered in relation to time since infection [153].

Importantly, evidence of persistent DNA damage was demonstrated in the ovarian microenvironment 2–9 months (mean, 4.5 months) after SARS-CoV-2 infection. Follicular fluid (FF) obtained from post-COVID-19 patients induced significantly increased expression of γH2AX, a molecular marker of DNA damage and chromatin alterations associated with double-strand breaks, in both granulosa COV434 cells (p < 0.05) and endothelial EA.hy926 cells (p < 0.01), compared with cells exposed to control FF. These effects were observed without significant changes in cell proliferation. The authors proposed that SARS-CoV-2-associated systemic and/or local inflammation may alter ovarian cytokine signaling and promote ROS generation, potentially disrupting DNA integrity in follicular and endothelial cells [154].

Notably, these alterations in DNA integrity were detected during the 2–9-month post-COVID period, whereas a subsequent study by the same group showed normalization of DNA integrity between 9 and 18 months after SARS-CoV-2 infection, indicating a prolonged but potentially reversible post-infectious effect [154,155]. Taken together, these findings indicate that post-COVID alterations are dynamic and temporally heterogeneous: while some disturbances in DNA integrity may persist for several months and subsequently recover, other cellular and tissue-level abnormalities can remain detectable for considerably longer periods [152,153,154,155].

The possibility of SARS-CoV-2 genomic integration into the human genome remains a subject of scientific debate. A recent review specifically addressing this controversy discusses evidence both supporting and questioning virus–host genomic integration, including the possibility of LINE-1-mediated reverse transcription and retrotransposition of viral RNA, the occurrence and interpretation of viral–host chimeric sequences, and methodological approaches and limitations in detecting potential integration events. Thus, the biological occurrence and significance of SARS-CoV-2 genomic integration remain under discussion [157].

However, there are also reports indicating no significant association between prior SARS-CoV-2 infection and embryo chromosomal abnormalities: in a retrospective study of 498 PGT-A cycles, the rates of euploidy (38.5% vs. 40.3%; p = 0.451), aneuploidy (45.0% vs. 43.8%; p = 0.616), and mosaicism (16.5% vs. 15.9%; p = 0.746) were comparable between SARS-CoV-2-positive and control groups [158].

Importantly, the long-term persistence of post-COVID-19 symptoms also appears to differ across SARS-CoV-2 epidemic periods. In a cohort of 2,689 individuals, persistent symptoms at 6 months were reported in 47% of adults infected during the Delta period, compared with 23% and 21% during the Omicron-2022 and Omicron-2024 periods, respectively. Even two years after infection, approximately 20% of adults infected during the pre-Omicron periods and approximately 10% of those infected during the Omicron periods continued to report symptoms; moreover, symptoms persisting beyond two years showed little further resolution [159].

The marked duration of these post-infectious manifestations, together with their dependence on the epidemic period and therefore, at least indirectly, on the predominant viral variant, raises an important question in the context of the genomic and chromosomal alterations discussed above. Given accumulating evidence that SARS-CoV-2 can affect DNA integrity, DNA-damage responses, chromatin organization, epigenetic regulation, and chromosome stability, could persistent or incompletely resolved alterations of the cellular hereditary apparatus contribute to the long-term phenotype observed in at least a subset of individuals after COVID-19? Moreover, could differences in the biological properties of SARS-CoV-2 variants result not only in different clinical courses of infection but also in quantitatively or qualitatively different effects on host genome stability? These possibilities remain hypothetical and require direct experimental and longitudinal investigation.

3.3. Virus-Induced Alterations in Nucleus–Cytoskeleton Interactions and Nuclear Mechanics

In normal cells, the microtubule cytoskeleton forms a dynamic three-dimensional mechanical network whose organization is regulated by the opposing activities of the molecular motors dynein and kinesin. Dynein generates predominantly inward-directed forces that promote microtubule network compaction and recruitment of microtubules to cell-adhesion sites, whereas kinesin generates opposing outward-directed forces that favor microtubule sliding and network expansion. The dynamic balance between these motor systems, acting in coordination with the actomyosin cytoskeleton, contributes to the maintenance of cell shape, three-dimensional architecture, mechanical properties, adhesion, migration, and mechanosensing. Thus, the dynein–kinesin balance represents an important component of the normal mechanical organization of the cell and its interactions with the surrounding microenvironment [160]. The microtubule network is also closely involved in the intracellular positioning and movement of the nucleus. Wang et al. demonstrated that microtubules, together with 10-nm intermediate filaments, form part of a force-generating cytoskeletal system responsible for nuclear movement and positioning within multinucleated cells [161]. The nuclear cytoskeleton, including nuclear actin and the nuclear lamina, plays an essential role in maintaining nuclear architecture and contributes to chromatin organization, transcriptional regulation, DNA replication and repair, and other nuclear processes. Viral infections can exploit and remodel these structural systems, altering nuclear actin and lamins to facilitate viral replication and induce changes in nuclear organization and function [162,163].

As early as 1968, evidence emerged that viral infection can profoundly affect nuclear movement and positioning through virus-induced alterations of the cytoskeletal system. Holmes and Choppin demonstrated that, in SV5-induced syncytia, nuclei actively migrated through the cytoplasm at rates of approximately 1–2 μm/min and became aligned in characteristic patterns, whereas disruption of microtubules by colchicine inhibited both nuclear movement and alignment, indicating a critical role of the microtubule network in these processes [164].

Microtubules are highly dynamic components of the cytoskeleton whose organization and stability are tightly regulated in normal cells. During infection, however, viruses can actively modify microtubule dynamics, stability, post-translational modifications, and spatial organization, thereby reorganizing the intracellular environment to support different stages of their life cycle. Such virus-induced remodeling affects not only intracellular trafficking of viral components but also fundamental microtubule-dependent cellular processes, including cell shape, polarity, signaling, and motility. These conclusions were based on evidence obtained across a broad range of viral systems, including herpes simplex virus type 1, HIV-1, poliovirus and other picornaviruses, influenza virus, vaccinia virus, African swine fever virus, as well as plant viruses, demonstrating that manipulation of microtubule organization represents a widely used strategy across diverse virus families [165].

In addition, viruses can exploit coordinated interactions between actin and microtubule networks during the early stages of infection. Following actin-dependent attachment, entry and short-range transport at the cell periphery, incoming viral particles can transition to microtubule-based transport through actin–microtubule cross-linking factors and +TIPs, while interactions with dynein and kinesin motors enable bidirectional intracellular trafficking, including transport toward the nucleus [166].

In an earlier study using primary normal human dermal fibroblasts, Procter et al. (2018) demonstrated that the HCMV cytoplasmic assembly compartment functions as a Golgi-derived microtubule-organizing centre: EB3-dependent acetylated microtubules generated by this structure drove persistent nuclear rotation, which occurred in approximately 80% (53/66) of infected cells and could reach rotations of approximately 720° [167]. A direct mechanistic link between virus-induced cytoskeletal remodeling, nuclear movement, and intranuclear organization was demonstrated by Procter et al. (2020) in primary normal human dermal fibroblasts (NHDFs) infected with HCMV. The authors showed that HCMV establishes a virus-associated microtubule-organizing centre from which acetylated microtubules exert dynein-dependent forces on the nuclear envelope through Nesprin-2G–SUN1 LINC complexes, resulting in nuclear rotation. Disruption of SUN1–Nesprin-2G interactions or inhibition of the dynein adaptor BICD2 suppressed this rotation. Importantly, these cytoplasmic mechanical forces also induced polarization of SUN1 and emerin and reorganized nuclear F-actin, leading to spatial segregation of viral DNA from host DNA and inactive histones and thereby creating a nuclear environment favorable for viral replication [168].

More recent studies extended these observations by demonstrating that HCMV can dynamically switch the cytoskeletal mechanisms controlling nuclear movement during the course of infection. In primary normal human dermal fibroblasts, Mahmud et al. showed that HCMV independently downregulates Lamin A/C and SUN2, accompanied by extensive loss of actin filaments and actin caps associated with the nuclear surface and, conversely, by the formation of acetylated microtubule arrays extending toward the nucleus and cell periphery. At later stages of infection, HCMV-infected fibroblasts exhibited a distinct migratory phenotype characterized by more exploratory movement with cellular projections and lamellipodia, in contrast to the predominantly linear movement of uninfected cells. This transition was accompanied by cessation of nuclear rotation and a switch to directional translocation and deformation of the nucleus within the migrating infected cell. Functional inhibition of the viral kinase pUL97, expression of phosphorylation-resistant Lamin A/C mutants, or depletion of the tubulin acetyltransferase ATAT1 impaired acetylated microtubule formation, nuclear movement, and cell migration. Acetylated microtubule arrays were observed in approximately 80% of control infected cells, whereas following pUL97 inhibition or expression of the Lamin A/C-S22A mutant, approximately 80% of cells failed to form prominent acetylated microtubule arrays. Importantly, Lamin A/C mutants markedly inhibited HCMV-induced cell migration without suppressing the normal movement of uninfected fibroblasts, supporting a virus-specific remodeling mechanism rather than a general impairment of cellular motility. Together, these findings demonstrate that HCMV actively rewires nucleus–cytoskeleton coupling, switching the mechanisms governing nuclear movement and cellular migration rather than merely exploiting pre-existing cytoskeletal structures [169].

Importantly, alterations in nuclear architecture are not restricted to the viral systems described above but have also been demonstrated during SARS-CoV-2 infection. In human lung epithelial Calu-3 and HeLa-ACE2 cells, SARS-CoV-2 induced cell–cell fusion and formation of multinucleated syncytia, accompanied by pronounced nuclear membrane blebbing, budding of chromatin from the nucleus, and accumulation of chromosomal DNA in the cytoplasm. Approximately 18% of cells underwent fusion under the experimental conditions, while cytoplasmic chromatin was detected predominantly in fused rather than non-fused cells. SARS-CoV-2-induced syncytia were also observed in THP-1 macrophages and in the lungs of infected K18-hACE2 mice, supporting the occurrence of this phenomenon across different cellular systems and in vivo. Importantly, expression of SARS-CoV-2 Spike together with cellular ACE2 was sufficient to induce cell fusion, and three-dimensional reconstruction and live-cell imaging demonstrated the formation of nuclear membrane blebs followed by budding of chromatin into the cytoplasm. The released chromatin was recognized by cGAS, activating the cGAS–STING pathway and downstream interferon and pro-inflammatory responses [170].

Further evidence of extensive SARS-CoV-2-induced cytoskeletal remodeling was obtained in human lung epithelial cells. Using A549-ACE2 and Calu-3 cells and a combination of three-dimensional electron microscopy, super-resolution microscopy, and live-cell imaging, Cortese et al.(2020) demonstrated pronounced reorganization of several cytoskeletal systems during infection. SARS-CoV-2-infected cells showed accumulation of cortical actin at the plasma membrane and formation of actin rings surrounding intracellular Spike-positive vesicles. Particularly striking changes involved intermediate filaments: perinuclear dsRNA-positive viral replication organelles, composed predominantly of clusters of double-membrane vesicles, became enclosed within a cage-like vimentin network, while microtubules surrounded this compartment and were largely excluded from its interior. Live-cell imaging showed that vimentin cage formation occurred early during infection, with most events detected at approximately 6.5 h post-infection. Functional experiments further demonstrated that disruption of intermediate filaments markedly reduced viral replication and infectious-virus production, whereas pharmacological alteration of microtubule dynamics with paclitaxel or vinblastine strongly impaired production of extracellular infectious virus. In contrast, disruption of actin microfilaments with latrunculin A had little effect on viral replication or infectious-particle production, despite the pronounced redistribution of cortical actin observed in infected cells. These findings demonstrate that SARS-CoV-2 induces extensive and functionally differentiated remodeling of actin, intermediate-filament, and microtubule networks in human respiratory epithelial cells [171]. Virus-induced structural alterations may arise within the first hours of infection and simultaneously involve different levels of cellular organization, ranging from cytoskeletal remodeling to alterations of the nuclear and chromosomal apparatus.

The very early onset of SARS-CoV-2-induced cellular remodeling is further supported by temporal phosphoproteomic analyses. Bouhaddou et al. examined SARS-CoV-2-infected Vero E6 cells, an epithelial-like kidney cell line derived from the African green monkey, at 0, 2, 4, 8, 12, and 24 h post-infection and detected extensive changes in host phosphorylation signaling beginning as early as 2 h after infection. These alterations involved multiple regulatory pathways, including kinases associated with cytoskeletal organization, and were accompanied by marked activation of CK2 and p38 MAPK signaling and suppression of mitotic kinase activity. To examine cytoskeletal morphology in a human cellular model, the authors additionally studied human Caco-2 intestinal epithelial cells at 24 h post-infection and observed prominent, elongated F-actin-rich filopodial protrusions, with SARS-CoV-2 M protein localized along their shafts and tips and CK2 colocalizing with viral N protein in these structures. Electron microscopy of infected Vero E6 cells further demonstrated viral particles budding from filopodia. Thus, SARS-CoV-2-induced molecular signaling alterations can be detected within the first few hours of infection, preceding or accompanying pronounced remodeling of actin-based cellular structures observed at later stages [172].

The involvement of the actin cytoskeleton in SARS-CoV-2 infection has also been demonstrated in a three-dimensional human tissue model. Using 3D human gastric organoids, Wang et al. showed that primary SARS-CoV-2 infection through the apical surface depends on the apical actin-based cytoskeleton and its regulatory network. Pharmacological perturbation of actin dynamics significantly affected the response of gastric organoids to infection, while interference with kinesin-dependent processes altered components of the cytokine response. These findings extend evidence of SARS-CoV-2–cytoskeleton interactions beyond conventional two-dimensional cell cultures to a 3D human gastrointestinal epithelial model and implicate both actin-associated and motor-protein-dependent mechanisms in the cellular response to infection [173].

Dynamic live-cell imaging provided further evidence that SARS-CoV-2-induced cellular remodeling develops from the early stages of infection. Using human U2OS-ACE2 cells and label-free holotomographic microscopy combined with AI-based image analysis, Saunders et al. continuously monitored infected cells for up to 48 h, with quantitative analysis performed every 12 min beginning as early as 2 h post-infection. Although the authors did not identify a specific cytoskeletal alteration at this earliest time point, the experimental design enabled cellular and subcellular dynamics to be followed from the very early phase of infection. The first prominent morphological event was syncytium formation, visible from approximately 10 h post-infection, accompanied by rapid clustering of nuclei as soon as two or more cells began to fuse and by changes in nuclear/nucleolar organization. Importantly, comparison of SARS-CoV-2-induced syncytia with syncytia generated independently of infection by expression of the cellular fusogenic protein Syncytin-1 revealed a distinct organization of the microtubule network: a flat radial microtubule distribution extending from clustered nuclei toward the syncytium boundaries was detected in approximately 77% of SARS-CoV-2-induced syncytia compared with only 20% of Syncytin-1-induced syncytia. These observations indicate that SARS-CoV-2-associated cytoskeletal remodeling cannot be explained solely by cell fusion and support the concept that infection induces specific alterations in cellular structural organization [174].

The formation of multinucleated syncytia provides another potential link between SARS-CoV-2-induced alterations of cellular and nuclear organization and genomic instability. Histopathological examination of lung tissues from patients who died from COVID-19 revealed SARS-CoV-2-positive syncytial pneumocytes containing from 2 to more than 20 nuclei, demonstrating that extensive multinucleation occurs in vivo during human infection [175]. In an experimental human cell model, expression of SARS-CoV-2 Spike in ACE2-expressing cells induced rapid cell–cell fusion, generating giant syncytia containing more than 60 nuclei as early as 12 h after transfection [176]. Importantly, multinucleation was associated with micronucleus formation. Ren et al. detected micronuclei in more than 93% of SARS-CoV-2 Spike-induced syncytia, with a mean of 3.99 micronuclei per syncytium and a range of 0–21; moreover, the number of micronuclei positively correlated with the number of nuclei within individual syncytia. More than 59% of micronuclei generated following Spike expression and more than 72% of those observed after infection with authentic SARS-CoV-2 were γH2AX-positive, indicating DNA damage. Micronucleus formation was accompanied by activation of the DNA damage response and cGAS–STING signaling [177]. Together, these observations establish a quantitative association between the extent of SARS-CoV-2-induced multinucleation and micronucleus formation and suggest a potential mechanistic connection between virus-induced syncytial remodeling and genomic instability.

However, the chromosomal content of SARS-CoV-2-induced micronuclei has not yet been sufficiently characterized. In particular, centromere-specific analyses are needed to distinguish micronuclei arising from acentric chromosome fragments from those containing whole chromosomes and therefore reflecting chromosome missegregation or aneugenic events.

In mouse oocytes, 24-h exposure to SARS-CoV-2 N501Y mutant Spike protein at 20 or 50 μg/ml induced abnormal meiotic spindle morphology and significantly increased both spindle length and metaphase-II plate width, although chromosome misalignment and the proportion of oocytes reaching metaphase II did not differ significantly from controls; notably, the study demonstrated a treatment effect but did not establish a clear dose-dependent relationship [178]. Whether SARS-CoV-2 Spike-induced increases in spindle length and metaphase plate width are accompanied by prolonged metaphase and altered spindle-checkpoint dynamics remains unknown.

Although SARS-CoV-2 replicates predominantly in the cytoplasm and lacks an obligatory nuclear phase, accumulating evidence indicates that the nucleus is actively involved in the cellular response to infection. Chen et al. emphasized the functional continuum linking the cytoskeleton, LINC complexes, nuclear lamina, and chromatin, through which mechanical and structural alterations in the cytoplasm can be transmitted to the nucleus. In SARS-CoV-2-infected or Spike-induced fused cells, disruption of nuclear integrity, including reduced Lamin A/C, nuclear-envelope blebbing and rupture, cytoplasmic release of genomic DNA, and accumulation of γH2AX-positive DNA-damage foci, further links alterations of nuclear architecture to genome instability and innate immune activation. The authors concluded that nuclear biology remains an insufficiently explored aspect of SARS-CoV-2 infection and requires further mechanistic investigation. Importantly, these observations indicate that nuclear integrity itself may modulate host-cell immune responses during SARS-CoV-2 infection [179].

Taken together, these findings indicate that virus-induced cytoskeletal remodeling should be considered not merely as a structural consequence of infection but as a mechanism capable of altering the physical and functional state of the entire cell. The nucleus is mechanically integrated with actin filaments, microtubules, intermediate filaments, and their associated motor proteins through the LINC complex and nuclear lamina; therefore, changes in the organization or force balance of these cytoskeletal systems can be transmitted directly to the nuclear surface and interior. Altered activity of kinesin and dynein motors, microtubule reorganization, disruption of actin–nuclear connections, and changes in SUN proteins or lamins can modify the forces acting on the nucleus, resulting in abnormal nuclear rotation, displacement and positioning within the cytoplasm, directional translocation, deformation, or loss of nuclear-envelope integrity. Importantly, these phenomena are functionally relevant: nuclear movement and positioning normally contribute to cellular polarity, migration, intracellular organization, and the spatial coordination of cellular processes, whereas mechanical forces transmitted to the nucleus can influence nuclear shape and stiffness, chromatin organization, nuclear transport, and transcriptional regulation. Consequently, virus-induced changes in nuclear rotation or position may redistribute mechanical forces and intracellular spatial relationships and thereby contribute to a broader reprogramming of cellular behavior rather than representing an incidental morphological phenomenon.

The temporal characteristics of these alterations are particularly important. Available studies indicate that SARS-CoV-2-induced host-cell reprogramming can begin within the first hours after infection: phosphorylation-signaling changes have been detected as early as approximately 2 h post-infection, vimentin remodeling within the first several hours, and overt syncytial and cytoskeletal reorganization by approximately 10–24 h, depending on the experimental system and the endpoint examined. These observations suggest that the apparent timing of virus-induced structural damage is strongly influenced by what is measured and by the sensitivity of the method used; therefore, the absence of obvious morphological abnormalities at very early time points should not be interpreted as evidence that cellular reorganization has not yet begun. Rather, molecular and mechanical perturbations may precede later visible changes in nuclear and cytoskeletal architecture.

This relationship may become particularly important when mechanical remodeling progresses to nuclear damage. Disruption of the nuclear lamina and nucleus–cytoskeleton coupling can facilitate nuclear deformation and rupture, chromatin displacement and the appearance of micronuclei, thereby linking altered nuclear mechanics with genome instability. In SARS-CoV-2 infection, loss of nuclear integrity may additionally expose genomic DNA to the cytoplasm, where it can activate cGAS–STING-dependent innate immune and inflammatory signaling, indicating that nuclear integrity itself can modulate the host-cell immune response. Thus, the functional consequences of virus-induced remodeling may extend from very early molecular signaling disturbances and altered nuclear movement to changes in genome organization, gene regulation, inflammatory signaling, and ultimately the phenotype and function of the affected cell. Whether such mechanically and structurally altered states persist after the acute phase of SARS-CoV-2 infection and contribute to sustained cellular dysfunction in Long COVID remains unresolved, but the available evidence identifies nucleus–cytoskeleton coupling and nuclear mechanics as a biologically plausible level at which an initially rapid viral perturbation could generate broader and potentially longer-lasting alterations in cellular function.