The advent of chromosome conformation capture (3C) technologies has provided unprecedented insights into the mechanisms by which cellular DNA is spooled and packaged into the nuclear microenvironment, and how this packaging impacts biological processes ( Figure 1 ). The original 3C methodology, described by Dekker and colleagues in 2002, enabled the detection of the frequency of contacts between pairs of genomic loci (1). Various iterations of this technology, reviewed elsewhere, were subsequently developed, providing essential tools to interrogate the principles of spatial organization of the genome (2, 3). Though these techniques have greatly improved our understanding of cis-folding processes of cellular DNA, these assays have also revealed that there exists little trans-interaction between host chromosomes (4). In spite of these features of mammalian genome looping, viral infection of host cells represents a common biological process where two genomes, host and viral pathogen, can interact with one another in-trans. Emerging studies have thus focused on unraveling how viral pathogens utilize the principles of genome folding to navigate the nuclear environment and establish infection. Here we briefly review some of the recent discoveries that have advanced our understanding of how DNA viruses co-opt host architectural proteins to organize their genomes, and in doing so, facilitate essential processes such as infection, lytic replication, and latency. In particular, we highlight the ability of a diverse array of DNA viruses to utilize genome-organizing protein CCCTC-Binding Factor (CTCF) in one or both of two ways: 1. to impact the chromatin topology of the host genome or 2. to organize its own genome in ways that enable sophisticated epigenetic control ( Figure 1 ). Indeed, a wealth of evidence has established that human tumor viruses Kaposi’s sarcoma-associated herpesvirus (KSHV) and Epstein-Barr virus (EBV) persist as latent episomes in B lymphocytes, and that CTCF and/or Cohesin play important roles in the maintenance of these episomes (reviewed in 5). Episome maintenance is correlated with various virus-induced cancers, in part because latent episomes and latency gene products can impact host genome topology and, by extension, gene expression. Thus, a deeper understanding of the interactions among viral episomes, viral latency products, and host genomes will prove critical in our understanding of virus-induced oncogenesis as well as the design of targeted cancer therapeutics by engineered oncolytic viruses.

At the simplest level, the mammalian genome is folded according to a distinct hierarchical order made up of promoter-enhancer loops [reviewed in (6)]. Many such functional loops, combined with structural loops, all generated within a distinct region of a chromosome spanning hundreds of kilobases, creates a topologically associating domain [TAD; Figure 1 ; (7)] Structurally defined by the high frequency of intradomain contacts over a large region of the chromosome, TADs are packaged in chromatin that has a similar transcriptional status. Functionally, this suggests that genes in the same TAD are coregulated (8, 9). This structural organization segregates the genome into chromatin types, broadly defined as Type A, which is largely associated with permissive chromatin, and Type B, which is composed of repressive chromatin [ Figure 1 ; (4, 10)]. In addition to expression, the boundaries of TADs coincide with replication boundaries, suggesting that TADs can potentially replicate as stable units (8). Indeed, these units are established early in G1 phase of cell cycle and are dissolved pre-mitotically at G2 phase (11, 12). As the fidelity of the genetic code must be maintained during cell cycle, and since TAD structures oscillate between formation and dissolution throughout cell division, we can infer that TADs may play a role in maintaining the stability of the cellular genome (discussed below).

One of the most visible features of TAD boundaries is the binding of the architectural protein CTCF, which, when bound in a convergent orientation, enables the formation of a loop of the intervening DNA (4). By serving as a border between the high frequency intra-TAD interactions and low frequency inter-TAD interactions with neighbors, CTCF sites at TAD boundaries influence the maintenance of the local chromatin environment (13–15). The loop extrusion model, proposed to explain the mechanics of loop formation, posits that chromatin is extruded through the Cohesin complex until it encounters CTCF-bound distal elements, which form TAD boundaries (16). Convergence of the CTCF binding sites is critical for this process. The formation of a CTCF/Cohesin-dependent loop anchor sets up the milieu for the generation of subsequent smaller DNA loops in conjunction with Mediator, cohesin and Ying-Yang (YY1) proteins that constitute the entire TAD (6, 17, 18). These loop anchors at TAD boundaries, bound by CTCF and Cohesin, are vulnerable to DNA damage in a transcription-, replication-, and cell-type-independent, but Cohesin-dependent manner until they encounter convergent CTCF-bound distal elements (19, 20). Additionally, the topoisomerase subunit protein Top2b regulates the torsional stress at these loop anchors to modulate genome stability (20). Taken together, replication stress initiated at these sites during genome replication predisposes them to becoming fragile genomic regions, which can serve as the initiation sites for chromosomal rearrangements and genome instability (21).

DNA viruses enter cells through receptor-mediated endocytosis, traverse the cytoplasm, and enter the nucleus through nucleopore dependent or independent mechanisms, where they usurp cellular factors to replicate their genomes and generate progeny virions (22, 23). Upon infection of host cells, viruses can induce or inhibit a cellular DNA damage response (DDR), potentially downstream of innate immune signals, caused by the presence of foreign DNA or viral proteins (24). Modulating the host DDR has the potential to impede [adenovirus and herpesviruses (25–30)] or facilitate [DNA tumor viruses and parvoviruses (31–35)] virus infection. As viral pathogens continue to replicate in the host nucleus, the viral nucleic acids and proteins or the DDR they modulate, have the capacity to modulate the topology of host chromatin to benefit viral infection. This has recently been observed in for Influenza A Virus (IAV), which in spite of being an RNA virus, replicates in the nucleus using the IAV Non-Structural Protein (IAV-NS1), which inhibits transcription termination at the ends of highly transcribed genes. This leads to the displacement of Cohesin from CTCF sites (36), converting repressive chromatin compartments (Type B; (4) into permissive ones [mostly Type A; [ Figure 1 ; (4)]. These changes in the epigenetic landscape of infected cells by readthrough-transcription modify the host’s TADs, especially at the borders maintained by CTCF and Cohesin. Interestingly, changes in TAD structure by readthrough-transcription of the host genome have also been observed during infection of the unrelated DNA virus Herpes Simplex Virus [HSV (37)], indicating that targeting the host genome’s topology may be an efficient mechanism for viral pathogens to gain a foothold in the nuclear environment. By modulating cellular TADs, viral pathogens may increase the chromatin environment that is available to them. They can then use this environment to establish and expand their replication centers while simultaneously evading host antiviral defense factors.

In addition to its impact on the host genome’s topology, viruses must also utilize cellular factors to appropriately express proteins essential for their life cycle. In this regard, the large 150 kb genome of HSV forms multiple DNA loops, not unlike TADs in the host cells, which are maintained by CTCF-bound elements and are essential for viral infection, latency and reactivation (38, 39). Additionally, a wealth of evidence now indicates that human gammaherpesviruses, known for their biphasic life-cycles involving latency and lytic replication, utilize CTCF and Cohesin to establish and maintain latency, and thus, life-long infection. Recent work has established that CTCF and/or Cohesin contribute to viral latency by acting as boundary factors and, in some cases, coordinating 3D genome looping; these events enable viral genomes to adopt a heterochromatin structure that limits viral gene expression to a few latent genes (40). This manipulation of viral 3D genomic architecture is linked to the ability of herpesviruses to elicit cellular transformation and oncogenesis (41, 42).

The human gammaherpesviruses, Kaposi’s-sarcoma-associated herpesvirus (KSHV) and Epstein-Barr Virus (EBV), can establish latency in B lymphocytes, allowing both viruses to simultaneously avoid immune surveillance while creating a long-term infective niche. Indeed, both KSHV and EBV have evolved sophisticated mechanisms to preserve latent genome maintenance in proliferating cells; this is thought to be a major contributing factor in the onset of oncogenesis (42, 43). Both KSHV and EBV establish latency through a wide array of epigenetic mechanisms that have been reviewed elsewhere (40, 44). Briefly, however, these mechanisms include regulation by non-coding RNAs (ncRNAs), post-translational modification (PTM) of histones, and DNA methylation. As each of these events has the potential to disrupt TAD boundaries, all are potential drivers of tumorigenesis (40, 45). Recent work, however, has focused on the mechanisms by which these viruses manipulate higher-order chromatin architecture in order to establish and preserve latency, as well as the impact of latent virus-induced epigenetic modifications on the host genome (40, 42, 44).

With regard to small DNA viruses, it is tempting to speculate that they also adopt a looped conformation akin to cellular promoter-enhancer loops. Interestingly, the genome of the Dependoparvovirus Adeno-Associated Virus 2 (AAV2), which is 5 kilobases long, has been observed to adopt a looped structure in electron microscopy studies [ Figure 1 ; (80)]. Additionally, Cohesin subunits SMC5/6 interact with the HBx protein of Hepatitis B Virus (HBV) during infection to block extrachromosomal DNA transcription (81). If the degradation of SMC5/6 complex is required for HBV expression in host cells, then this would also impact the host genome’s topology. Alternatively, HBx-mediated degradation of SMC5/6 may be targeted to the subnuclear locations in proximity to the HBV genome. However, further studies on the cis-topology of small DNA viruses are yet to be investigated using chromosome conformation capture techniques.

The genome of the carcinogenic Human Papillomavirus 18 (HPV18) adopts a looped configuration maintained by interaction between distally bound CTCF and YY1, which inhibits the expression of the HPV oncogenes E6 and E7 in undifferentiated cells (82). The HPV oncogene E2 interacts with SMC5/6 to facilitate genome maintenance, which may involve maintaining viral genome looping (83). Strikingly, mutation of the CTCF binding site on the HPV genome leads to alterations in processing of viral RNA transcripts (84). These findings are consistent with newly identified roles of CTCF in RNA processing for both the host and small DNA viruses such as Parvoviruses (85–87). In order to investigate where the genome of the non-integrating Parvovirus Minute Virus of Mice (MVM) localizes, Majumder et al., 2018 utilized a modified form of 4C-seq assay to investigate the localization of small DNA viruses to cellular sites, a technique that has been dubbed V3C-seq [Viral Chromosome Conformation Capture Assays; (35)]. These studies revealed that MVM, which requires cellular DDR to replicate, also localizes to cellular sites of DNA damage, many of which are previously identified Early Replicating Fragile sites [ERFs (21)]. Strikingly, many of these sites also coincide with TADs and contact domains that are packaged in Type A chromatin (35). In subsequent studies, Majumder and colleagues discovered that ectopically expressed viral non-structural phosphoprotein NS1 binds to cellular sites of DNA damage, and, when bound cognate sequences on the viral genome, can transport the viral genome to these cellular DDR sites [ Figure 1 ; (88)]. These findings suggest that the NS1 protein of MVM, a small DNA virus, can help a viral genome navigate the nuclear milieu to essential TAD/DDR regions that promote viral infection.

In the years since the first report of chromosome conformation capture technology (1) we have gained unprecedented insight into the mechanisms of genome folding and how this regulates cellular physiology. A wealth of recent work has focused on unraveling how viral pathogens utilize the principles of genome folding to navigate the nuclear environment and establish infection. Future studies are likely to shed light on additional mechanisms at work between viral and host genomes. In this regard, Chromatin Interaction Analysis with Paired-End Tag Sequencing [ChIA-PET; (89)] assays of viral proteins, which combine Chromatin Immunoprecipitation with high-throughput Chromosome Conformation Capture (3C) techniques, will yield critical insights into how both viral and host proteins modulate nuclear topology, establish viral replication centers, affect latency and impact reactivation as a part of viral pathogenesis. Thus far, these studies have only been performed for RNA Polymerase II in EBV-infected B cells to characterize the EBV regulome in lymphoblastoid cells, identifying critical insights into how spatial organization underlies EBV-dependent cellular transformation (79). Additionally, as viral infection can activate or inhibit DNA damage signaling [reviewed in (22)], and the cellular DDR induces alterations in cellular chromosome conformation (90), the cause-effect relationship between viral infection, DNA damage and chromatin conformation are yet to be unraveled. These mechanisms may be unraveled by ChIA-PET studies of bridging molecules such as CTCF, Cohesin and gamma H2AX between the virus and host genomes, which will yield critical insights into the mechanisms of viral latency, reactivation, virus-induced oncogenesis and how viral oncolytic agents may function.

KM and AM both made substantial and direct contributions to the writing and editing of the manuscript and have approved it for publication. All authors contributed to the article and approved the submitted version.

KM is supported by a NIH K99/R00 Pathway to Independence Award (AI148511), University of Wisconsin Office of Vice Chancellor for Research and Graduate Education and the University of Wisconsin Carbonne Cancer Center’s Human Virology Program. Additional support for this work was provided by a PSC-CUNY Award, jointly funded by The Professional Staff Congress and the City University of New York (TRADA-51-497; AM).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.