HSV-1 significantly alters the phosphorylation modification of Pol II after infecting host cells, and ICP22 is a key protein involved in this effect (29, 54, 55). Pol II is composed of 12 subunits, and the amino acid sequence of the CTD of its LS consists of multiple repeating heptapeptide Tyr1-Ser2-Pro3-Thr4-Ser5-Pro6-Ser7 sequences. The CTD acts as a scaffold during transcription to recruit factors required for mRNA processing and chromatin modification of the Pol II complex, but this role depends on the phosphorylation of CTD (56), where the phosphorylation of CTD Ser-2 and Ser-5 are the most critical ( Figure 3A ). Phosphorylation of Ser-5 triggers recruitment of the capping enzyme for formation of the 5´ cap structure. Ser-2 phosphorylation is closely related to Pol II’s ability to overcome transtranscriptional pauses and can serve as a key point for transcriptional regulation (57). In addition to its role in transcription elongation, Ser-2 phosphorylation also recruits cleavage and polyadenylation factors for the generation of the mRNA 3´ end (58, 59), and it has been implicated in recruiting splicing and mRNA export factors (60, 61).

Pol II CTD Ser2 phosphorylation is mediated by P-TEFb. The CDK9 subunit of P-TEFb provides enzymatic activity, while cyclin has a regulatory role (62, 63). Many elongation regulators play a role in regulating transcription elongation by interacting with P-TEFb (64). Multiple studies have shown that HSV-1 ICP22 directly interacts with CDK9 and inhibits CDK9 enzyme activity, thereby inhibiting Ser-2 phosphorylation and inhibiting the transcription of cellular genes (3, 29, 65) ( Figure 3B ).

ICP22 mediates the loss of Ser-2 phosphorylation on the Pol II CTD, which not only inhibits host cell gene transcription but also affects the expression of viral genes (66). The HSV-1 virus can overcome this effect through other mechanisms. Studies have shown that ectopic expression of ICP22 can inhibit the transcription of HSV-1 IE (ICP4), E (TK), and L (gC, VHS) promoter reporter genes. Studies have shown that VP16 activates the transcription of IE genes by recognizing the core motif TAATGARAT in the promoter sequence of IE genes and binding to the promoter of IE genes (67, 68). The coexpression of VP16 and ICP22 protein indicates that VP16 can eliminate the transcriptional inhibition of ICP22 on the IE genes (65, 69). Immunoprecipitation assays showed that VP16 can interact with P-TEFb, but there was no direct interaction with ICP22 (29). ICP22, P-TEFb and VP16 form the ICP22-P-TEFb-VP16 multiprotein complex in the cell. This suggests that the cooperative regulation of viral genes by ICP22 and VP16 is related to P-TEFb. We speculate that VP16 interacts with P-TEFb to relieve the transcriptional inhibitory effect of ICP22 on IE genes ( Figure 3C ). However, the specific mechanism of cooperative regulation of VP16 and ICP22 is still unclear. The latest research shows that VP16 competes with ICP22 to interact with P-TEFB, thereby eliminating the inhibition of P-TEFb by ICP22 (29). In addition, studies have shown that ICP22 is necessary for the efficient transcription of viral L genes in some cell lines (54), but it is unclear how HSV-1 eliminates the inhibitory effects of ICP22 on E and L genes ( Figure 3D ). It is possible that after VP16 releases the transcriptional suppression of the IE gene by ICP22, the activated IE proteins ICP0, ICP4 and ICP27 are beneficial to the transcriptional expression of the E and L genes, but this requires further research and verification. However, it has been suggested that HSV-1 transcription may not require the Ser-2 phosphorylated form of Pol II CTD, although this form is necessary for cellular transcriptional extension and RNA processing (70). However, experiments have verified that HSV-1 transcription requires Ser-2 phosphorylation of the Pol II CTD (70). In addition, inhibition of Pol II CTD Ser2 phosphorylation suggests that NELF and DSIF phosphorylation may also be inhibited, but this remains to be determined. CDK9 also phosphorylates p53, a tumor suppressor that plays a central role in the response of cells to a range of stress factors. The interaction of ICP22 with the CDK9 target p53 is also important for efficient HSV-1 replication (71).

Several studies have provided maps of Pol II localization on the virus and host genome after HSV-1 infection (27, 28). HSV-1 infection leads to extensive transcription termination defects in host cell genes (72). Studies have shown that Pol II was almost completely removed from two-thirds of the host genes in the early stage of HSV-1 infection. The function of a few genes that increase the occupancy rate of Pol II is related to the upregulation of exosomal secretion and the downregulation of apoptosis, which may be beneficial to virus production. The HSV-1 genome contains a large amount of Pol II, which reflects the high-level transcription of viral genes. However, the Pol II occupancy level in the genome of the ICP22 mutant virus is reduced, and the accumulation of mRNA for almost all viral genes is reduced (73). Among them, ICP22 may downregulate cellular gene expression and promote viral gene expression by regulating the CDK9-mediated Pol II phosphorylation event in the P-TEFb complex, thereby promoting viral survival. ICP22 is not the only viral factor that affects Pol II during HSV-1 infection, which is the synergistic effect of HSV-1 gene products (74). This highlights the transcriptional regulation of the HSV-1 protein to maximize the expression of the viral genome while downregulating the expression of the host genome. As a result, the expression of host cell genes that are not conducive to virus replication is turned off, and the antiviral immune response of the host is destroyed. Therefore, this promotes the production of viral infectious virus particles.

Nucleosomes uniquely positioned on high-affinity DNA sequences present a polar barrier to transcription by Pol II (75). When the Pol II transcription elongation complex extends to the gene body, it needs to cross the nucleosome barrier (76, 77). There are a variety of regulatory mechanisms in cells that can remove or weaken the nucleosome barrier to assist Pol II extension, such as nucleosome remodeling and histone modification. Currently, a number of regulatory factors involved in regulating Pol II transcription elongation complex crossing nucleosome barriers have been identified in vitro using biochemical experiments, such as Facilitates chromatin transcription (FACT), Spt6, PAF complex and PARP (76). FACT was originally isolated and identified from an extract of the human HeLa cell nucleus (78), and it is a complex composed of SSRP1 and Spt16. FACT functions to disassemble an H2A-H2B dimer from nucleosomes, and Pol II can be transcribed through the remaining histone hexamer without being displaced (62, 79). When the Pol II transcription complex passes through the nucleosome, FACT can also promote the reassembly of the H2A-H2B dimer and the remaining hexamer into the histone octamer and maintain a highly activated state (80). Studies have found that after HSV-1 infects host cells, FACT relocates to the virus replication compartment in the nucleus, and FACT is abundant in the HSV-1 genome during the replication of HSV-1 (81, 82). This change is related to ICP22 protein (29, 73). Studies have shown that FACT interacts with ICP22 throughout infection. Transcriptome sequencing (RNA-seq) and chromatin immunoprecipitation-sequencing (ChIP-seq) experiments showed that compared with wild-type, regardless of the type of kinetics, the accumulation of almost all viral mRNAs of ICP22 mutant virus late infection was reduced. The Pol II occupancy level on the mutant virus genome is reduced (73). In contrast, the association of Pol II with the transcription initiation site in the mutant was not decreased. This shows that in the absence of ICP22, the viral gene transcription elongation rate is reduced. This indicates that ICP22 can recruit elongation factors (such as the FACT complex) into the HSV-1 genome to play a role and achieve effective viral transcription elongation. This promotes the production of infectious virus particles in late viral infection. Except for the FACT complex, in the absence of ICP22, the number of transcription elongation factors (Spt6 and Spt5) recruited by the viral genome was significantly reduced. These proteins all promote the elongation of cellular transcription and are known to interact with Pol II (83, 84). Based on the above research, we developed a model ( Figure 4 ): when HSV-1 infects cells, ICP22 can recruit FACT complexes to the viral genome by interacting with the FACT complex, thereby promoting Pol II to cross the nucleosome barrier on the viral genome and achieve efficient viral transcription elongation late in viral infection and ultimately infectious virion production.

HSV-1 ocular infections are the most common cause of corneal blindness in developed countries (102). Recurrent infections and the prolonged inflammatory response after viral clearance both contribute to corneal scarring (103). HSV-1 infection is recognized by antigen-presenting cells such as dendritic cells (DCs), natural killer cells, and macrophages (104, 105), which induce the secretion of IFNs and cytokines and the activation of CD4+ and CD8+ T cells. This T cell activation is strictly controlled and requires at least two signals, which involve the binding of CD28, CTLA-4 or PD-1 on the surface of T cells corresponding to the APC costimulatory molecule CD80 (B7-1) or CD86 (B7-2) (106). The costimulatory molecule CD80/86 drives T-cell activation and proliferation by binding to CD28 (107). CD80 plays a critical role in increased inflammatory responses in HSV-1-infected mouse corneas (108). Studies have shown that increasing CD80 levels promote increased CD8+ T cells, leading to exacerbated eye disease in HSV-1-infected mice (109). A study described a new mechanism of HSV-1 immune escape through ICP22-dependent downregulation of the host T cell costimulatory molecule CD80 in DCs (110, 111). According to reports, ocular infection of mice with HSV-1 suppressed the expression of the costimulatory molecule CD80 but not CD86 in the cornea (110). This effect was specifically mediated by the binding of HSV-1 ICP22 to the CD80 promoter, which is located between positions 151 and 462 of the CD80 promoter, and this interaction was required for HSV-1-mediated inhibition of CD80 expression. The latest research shows that the binding site of ICP22 and the CD80 promoter is located at aa 305–345 of ICP22. The use of HSV-1 recombinant virus expressing truncated ICP22 lacking CD80 promoter binding increased the expression of CD80 in DCs and the expression of IFN-γ in CD8+ T cells but did not increase CD4+ T cells in mouse corneas (111). Other viral proteins, such as ICP0, ICP27 or ICP47, can significantly increase CD80 promoter activity, and these increases did not offset the inhibition of CD80 by ICP22. The large increase in CD80 promoter activity observed in ICP22-deficient HSV-1 virus-infected DCs indicates that only ICP22 can counteract the stimulatory effect of other IE genes on CD80 expression, and no other HSV-1 genes have the effect of inhibiting CD80 expression. In contrast, overexpression of CD80 by infecting mouse eyes with recombinant HSV-1 lacking ICP22 will exacerbate corneal scarring in the infected mice, leading to greater eye disease. Interestingly, studies have shown that although D22 (recombinant virus lacking ICP22) replicates poorly in the eyes of infected mice, the level of T-cell infiltration caused by D22 infection is similar to that of its parental strain WT KOS (108). HSV-1 uses the ICP22-CD80 promoter interaction to specifically downregulate CD80 as a key immune escape mechanism and weakens the host’s immune response (109) ( Figure 6 ). ICP22 plays a critical role in reducing HSV-1-mediated immunopathology.

It was previously reported that HSV-1 ICP22 prevents the aggregation of nonnative proteins and recruits cell heat shock protein 70 (Hsc70) into the nuclear domain (112, 113). Recent studies have shown that ICP22 resembles a cellular J-protein/HSP40 family cochaperone, interacting specifically with Hsc70 (114). These findings showed that ICP22 reduced cytoplasmic protein aggregation, and ICP22 provided protection against the heat inactivation of firefly luciferase. Sequence homology analysis indicated that ICP22 contains an N-terminal J-domain and a C-terminal substrate binding domain, similar to type II cellular J-proteins. ICP22 may thus be functionally similar to J-protein/Hsp40 cochaperones that function together with their HSP70 partners to prevent aggregation of nonnative proteins. HSP40/HSP70 complexes have been shown to play critical roles in a myriad of cellular processes, including the regulation of gene expression and cell cycle control (115–117). This also suggests that ICP22 may be involved in chaperone-mediated autophagy (CAM) as a cochaperone protein similar to HSP40 and Hsc70. The CMA process involves three stages: recognition of molecular chaperones and substrates, unfolding of substrates and translocation across lysosomal membranes (118). This process requires the participation of two protein complexes: one is the substrate recognition complex, and the other is the lysosomal transmembrane translocation complex (118). The substrate recognition complex is composed of Hsc70 and other cochaperone molecules, including HSP90, HSP40, Hip, Hop and Bag-1, and recognizes protein substrate molecules with KFERQ sequences (118–120). Among them, Hsp40 stimulates the ATPase activity of Hsc70, leading to increased rates of binding and release of substrate proteins (118). In addition, Hsc70 is also involved in other antiviral defense mechanisms, such as aggregate formation and proteolytic degradation of viral proteins (121, 122). During HSV-1 infection, the ability of ICP22 to relocate Hsc70 to the VICE domain in the nucleus may prevent Hsc70 from exerting antiviral activity in other parts of the cell (113). Therefore, we speculate that ICP22, as a common partner similar to the cell J protein/HSP40 family, may participate in some functions of Hsc70 in this manner, but this needs to be verified by further experiments. This is the first known example of a virus obtaining a complete J-like protein, indicating that HSV has used the adaptability of the J protein to evolve a multifunctional helper chaperone that works with Hsc70 to promote lytic infection.

It is generally believed that ICP22 not only promotes apoptosis but can also inhibit apoptosis. HSV-1 in the absence of ICP22 induced more apoptotic cells, indicating that ICP22 can block apoptosis (123, 124). However, the antiapoptotic activity of ICP22 was not strong. Therefore, ICP22-mediated antiapoptosis is unlikely to directly act on apoptosis signal transduction but precisely regulate the expression of the E and L genes (including antiapoptotic genes) of the virus. For example, in the absence of ICP22 in HSV-1, the expression of US5, which inhibits apoptosis, is delayed (125–127).

In addition, ICP22 was cleaved by caspases in cells infected with HSV-1 d120 (apoptosis inducer), revealing another product called Mr 37500 (128). This process can be inhibited by overexpression of Bcl-2, transfection with US3 or addition of caspase-3 inhibitors. In the process of viral infection, the cleavage of viral proteins by caspases can lead to a variety of consequences, such as the adverse effect of apoptosis, the enhancement or reduction of replication, and the spread of the virus (129, 130).

In addition, HSV-1 ICP22 can interact with p53 and antagonize p53 (71). p53 is a key cellular transcription factor that plays a central role in cellular responses to a broad range of stress factors through its regulation of a variety of cellular pathways, such as apoptosis, the cell cycle, cellular senescence, DNA repair, autophagy, and innate immune control (131, 132). Studies have shown that p53 plays a positive role in HSV-1 replication. However, p53 has no effect on the replication of the ICP22-deficient strain (71). It is speculated that the effect of p53 on HSV-1 replication depends on ICP22, but this requires experimental verification. In addition, p53 promotes the expression of ICP27 in early infection without relying on ICP22, but the negative effect of p53 on reducing the expression of ICPO in late infection can be offset by ICP22. However, it is not clear whether ICP22 can directly regulate cell apoptosis by interacting with p53.

On the other hand, the antiapoptotic effects of HSV-1 US1.5 and ICP22 during infection were the opposite. Overexpression of US1.5 with a baculovirus vector has been shown to trigger the activation of caspase 3 in rabbit skin cells (133). This result indicates that US1.5 may have proapoptotic activity (130).

The nuclear factor kappa-B (NF-κB) transcription factor family can be found in almost all animal cells, and they are involved in the response of cells to external stimuli. NF-κB transcription factors are activated in response to a variety of signals, including cytokines, pathogens, injury, and other stress conditions. The activation of this factor leads to the expression of several immune response genes, such as proinflammatory cytokines (IFN-β, TNF-α, IL-6, IL-8), chemokines and adhesion molecules. Therefore, NF-κB plays a key role in the inflammatory response and immune response of cells (134, 135). In unstimulated cells, NF-κB binds to the inhibitory protein IκB and isolates the NF-κB·IκB complex in the cytoplasm, thereby preventing NF-κB from binding to DNA. The activation of NF-κB signaling is caused by extracellular stimulation. These stimuli are recognized by the receptor and transmitted to the cell, which ultimately leads to the activation of IκB kinase (IKK). IKK phosphorylates the inhibitory IκB subunit of the NF-κB·IκB complex in the cytoplasm. This phosphorylation causes IκB to be degraded by the proteasome and releases NF-κB from the inhibitory complex (136, 137). The released NF-κB protein is then transported to the nucleus, where it binds to the target sequence DNA and activates gene transcription ( Figure 7 ).

Studies have reported that transiently expressed VZV IE63 can downregulate the activity of several cellular NF-κB-responsive gene promoters, such as IL-8 and IL-6 (49). Subsequently, Habran et al. (48) showed that IE63 does not change the basic expression of these genes, which may be due to the issue of promoter accessibility related to chromatin. After treating HeLa cells with TNFα, a cytokine known to increase the utilization of several chromatins, the researchers measured the transcription levels of these genes (IL-8, IL-6, ICAM-1 and IκBα). The results showed that the presence of IE63 can reduce the expression of IL-8, IL-6 and ICAM-1 mediated by TNFα and increase the expression of the IκBα gene. In addition, after the degradation of IκBα, IE63 can promote the resynthesis of IκBα. This result indicates that the presence of IE63 promotes complementation of the IκBα library. The phosphorylation status of IE63 is crucial for affecting TNFα-mediated gene transcription. When alanine replaces the phosphorylation site, IE63-S224/T222A does not significantly regulate TNFα-induced gene transcription. Chip assays showed that IE63 reduced the accessibility of IL-8 and ICAM-1 promoters and increased the accessibility of TNFα to IκBα promoters. IE63 expressed in TNFα-treated HeLa cells reduced the binding of p65 to the NF-κB proximal sites on the IL-8 and ICAM-1 promoters and enhanced the recruitment of p65 to the IκBα promoter. It is speculated that the inhibition of IL-8 and ICAM-1 gene expression may be the result of IκBα induction because IκBα is an inhibitor of NF-κB (136). IE63 interferes with the TNF-inducing ability of several NF-κB-dependent genes through the accelerated resynthesis of IκBα. Although the molecular mechanism of these effects has not yet been determined, the treatment of IE63-expressing cells with TNFα inhibits the NF-κB pathway, suggesting that VZV may use such a strategy to resist VZV-infected cells from exogenous proinflammatory cytokine-induced antiviral reactions.

Ambagala and Cohen reported that VZV IE63 is required to inhibit IFN-α-induced antiviral responses (138), and they used IE63 deletion virus to infect human melanoma cells and U2OS cells. The results showed that the virus is highly sensitive to the antiviral effect of human IFN-α but not to IFN-γ compared to the parental and other viral gene mutants. IFN-α inhibited the expression of viral genes in cells infected with IE63 deletion virus at the posttranscriptional level without affecting its mRNA level. An important component of the innate response enhanced by the activity of IFNs is the signaling of the double-stranded RNA sensor PKR (139). PKR is an IFN-induced, double-stranded RNA (dsRNA)-activated serine/threonine protein kinase (140). This latent enzyme needs to be activated by autophosphorylation. Unless blocked, activated PKR phosphorylates the alpha subunit of eukaryotic initiation factor 2 (eIF-2α), whose phosphorylation causes inhibition of translation and, therefore, inhibition of virus replication (141). Most viruses either synthesize double-stranded RNA or form a double-stranded RNA structure during infection, thereby activating PKR. Viruses have evolved several mechanisms to interfere with eIF-2α phosphorylation and prevent the inhibition of protein synthesis in infected cells (142). In HSV-1, PKR is blocked by several viral genes, which redirect protein phosphatase 2 (PP2A) to dephosphorylate eIF-2α (143, 144). It has been reported that cells infected with the IE63 mutant have increased eIF-2 phosphorylation compared to cells infected by the parental virus (138). In the same study, cells transiently expressing IE63 showed a decrease in the basal level of eIF-2α phosphorylation, indicating that IE63 is sufficient to inhibit this phosphorylation ( Figure 8 ). These results indicate that IE63 may inhibit eIF-2α phosphorylation through the PKR sensor pathway, resist IFN-α-induced inhibition of protein synthesis and promote viral protein synthesis and virus replication.

The deletion of IE63 makes the deletion virus sensitive to IFN-α(the replication of the deletion virus is significantly reduced), suggesting that IE63 may be involved in the regulation of JAK-STAT signaling (138). The latest research shows that VZV inhibits the signal transduction activated by type I IFN through the JAK-STAT pathway. VZV infection leads to a decrease in IRF9 and a decrease in STAT2 protein and its phosphorylation (145). IE63 is the key protein for this change. In the presence of IFN, the expression of IE63 alone led to a reduction in IRF9 protein, while STAT2 did not change. The expression of IE63 led to a decrease in the expression of luciferase induced by IFNs. This result indicated that VZV IE63 interferes with type I IFN-mediated activation of JAK-STAT signaling by reducing IRF9, thereby inhibiting the expression of interferon-stimulated genes ( Figure 5 ). Simian varicella virus (SVV) shares approximately 75% DNA homology with VZV and exhibits a highly similar genome organization (146). SVV IE63, which is homologous to VZV IE63, inhibits the JAK-STAT pathway in a similar manner and degrades IRF9 in a proteasome-dependent manner. HSV-2 ICP22, as an E3 protease, participates in the proteasome to degrade IRF9 and STATs, thereby interfering with the signal transduction of type I IFN. Unlike HSV-2, VZV IE63 does not affect STAT2 and its phosphorylation, and there may be other proteins encoded by VZV that are responsible for inhibiting STAT2 phosphorylation and reducing STAT2 expression. Whether the degradation mechanism of VZV IE63 IRF9 is the same as that of HSV-2 ICP22 needs to be verified through experiments. It has not been reported that VZV IE63 has the effect of the E3 protease, and its specific mechanism needs to be further studied. The above studies show that VZV IE63 plays a key role in regulating the innate immune response to VZV.

VZV IE63 can also block apoptosis (147, 148). rOkaORF63 (a recombinant virus unable to express one copy of the diploid IE gene) (ORF63)-infected neurons were more susceptible to apoptosis than parental rOka-infected neurons. Furthermore, the expression of IE63 protein in neurons alone can resist apoptosis induced by nerve growth factor (NGF) (149). These results showed that IE63 can suppress apoptosis of neurons. Compared with parental virus, IE63-deleted VZV has higher levels of phosphorylated eIF-2α to promote apoptosis. Meanwhile, the expression of IE63 alone can be sufficient to block eIF-2α phosphorylation (130, 138). This indicates that during active VZV infection, IE63 participates in a protective mechanism against apoptosis in neurons. Although this study was limited to productive infections, the rich expression of IE63 during latent infection suggests that this gene product may play an anti-apoptotic function during the incubation period or reactivation of neurons (150, 151). Similar to HSV-1 ICP22, VZV IE63 can inhibit cell apoptosis, but it functions through a completely different mechanism. The specific mechanism by which VZV IE63 inhibits cell apoptosis still needs further study.