Approximately 80% of the world population are infected with a form of herpes viruses [1
]. Herpes viruses remain for the life of the host and can establish latency, a non-productive state that allows the virus to go undetected and unaffected by antiviral drugs [2
]. There are nine herpes viruses that infect humans and their effects can range from easily manageable to life threatening [4
]. Herpes simplex viruses 1 and 2 (HHV-1 and HHV-2) are associated with blisters on the lips and genitals, respectively, whereas HHV-6 and HHV-7 cause Roseola in infants [5
]. Human cytomegalovirus (CMV or HHV-5) is linked to lethal diseases involving the lungs, gastrointestinal tract, liver, retina, and central nervous system in immune compromised individuals, and is the leading cause of post-transplant infection [6
]. Congenital CMV results in an estimated 8000 cases of permanent neurological disabilities a year, affecting over 90% of infants who survive the initial infection [8
]. Chicken pox and shingles, caused by varicella-zoster virus (VZV), are manageable in immunocompetent people, but can become deadly in immunocompromised persons [4
]. Herpes viruses also are associated with several cancers. Epstein–Barr virus (EBV) causes Burkett’s lymphoma, nasopharyngeal carcinoma, and Hodgkin’s disease, and has been detected in approximately 70% of advance breast cancer tumors [12
]. Kaposi’s sarcoma-associated herpesvirus (KSHV) can result in body cavity-based lymphoma, and Kaposi’s sarcoma, the most common malignancy present in HIV-1 patients [14
]. The pervasiveness and deleterious effects of herpes viruses provides incentive for new approaches to antiviral drugs.
Widely used as a model to study herpesvirus replication, gene expression, and pathogenesis, the prototype virus of the family herpesviridae, HHV-1, has a linear double-stranded DNA genome of approximately 152 kbp. Several HHV-1 genomes have been sequenced to date, including strain 17 [16
], strain KOS [17
], and strain McKrae [18
]. Similar to other members of the herpesvirus family, the HHV-1 genome consists of two unique regions, unique long (UL) and unique short (US), flanked by terminal inverted repeats [19
]. The UL sequence contains 107,943 residues and has a base composition of 66.9% GC [16
]. Among 56 genes identified in the UL region, accounting for most of the DNA sequence, the UL1, UL43, US7, UL23 and UL49 genes accumulated the majority of non-synonymous mutations across the genome [20
]. The UL23 gene, for instance, has one of the highest genetic variabilities at 35.2%, confirming its role in the development of drug resistance [21
]. The assembly of accurate, full-length HHV-1 genomes is critical to identify the genetic determinants of drug resistance, virulence, and pathogenesis.
The most common antiviral strategy against herpes viruses is based on nucleoside analogs targeting the viral DNA polymerase and thymidine kinase. While these drugs are effective, many of them are associated with significant toxicities, poor bioavailability, and resistance in immunocompromised persons [22
]. The nucleoside analogues acyclovir and ganciclovir are the standard therapy for HHV-1 and CMV, respectively [22
]. These compounds target HHV-1 pUL30, the viral DNA polymerase, and HHV-1 pUL23, the viral thymidine kinase. Regarding CMV, mutations in the viral kinase pUL97 and polymerase pUL54 mediate resistance to ganciclovir and valganciclovir [25
]. The prevalence of resistance against acyclovir is 5% in immunocompromised persons and as high as 30% in allogeneic bone marrow transplant patients [26
], whereas the incidence of resistance to ganciclovir is 5–10% in organ transplant recipients [27
] and 40–50% in patients receiving repeated treatments or prolonged prophylaxis [25
]. HHV-1 strains that are resistant to acyclovir are typically cross-resistant to thymidine kinase-dependent drugs such as penciclovir and famciclovir. They also may be cross-resistant to polymerase dependent drugs, foscarnet or cidofovir [26
]. Ganciclovir-resistant strains also have shown cross resistance to second-line treatments foscarnet and cidofovir [28
]. Targeting the DNA polymerase and thymidine kinase, while widely used, is not the only option for herpesvirus antivirals.
Due to the resistance, toxicities, and other adverse side effects of nucleoside analogs, new targets to treat herpes viruses are being perused. The viral DNA packaging motor, namely its large terminase subunit, has become a promising target for potential herpes antivirals [29
]. The HHV-1 terminase is composed of three subunits which play several important roles in the packaging process, including identifying the viral concatemeric DNA, as well as endonuclease and ATPase activity. This provides several critical mechanisms that could be interrupted, provided a feasible drug binding site. Inhibiting the viral terminase or other components of the viral DNA packaging motor is expected to be more effective and has less target-related toxicity than nucleoside analogs. This is due to the fact that the molecular functions of the DNA packaging motor, capsid formation, DNA cleavage, and packaging of DNA into capsids, are virus-specific and not found in mammalian cells [30
]. Current CMV terminase inhibitors include benzimidazoles and the 3,4-dihydro-quinazoline-4-yl-acetic acid derivative, letermovir [32
The capsid needs to be filled with the appropriate DNA for a viable viral particle to be formed. Replication of the HHV genome first creates a concatemer and, subsequently, the packaging motor identifies the genetic sequence and begins packaging. At a certain point in the packaging process it must cleave the DNA in the proper location so only a monomeric strand of DNA remains in the capsid [33
]. The first antiviral agents to inhibit the cleavage of concatenated DNA into monomeric genomes, through inhibiting the terminase, were 2,5,6-trichloro-1-β-D-ribofuranosyl benzimidazole (TCRB) and 2-bromo-5,6-dichloro-1-(β-D-ribofuranosyl) benzimidazole (BDCRB). These compounds are suspected to affect CMV proteins encoded by genes UL56 coding the small terminase subunit and UL89 coding the large terminase subunit, the key components of the packaging motor. Mutations in UL56 result in resistance to TCRB, whereas mutations in UL89 cause resistance to both TCRB and BDCRB [34
]. The high resistance caused by mutations in UL89 were mapped to amino acid substitutions D344E and A355T [35
]. Since there are no acidic residues corresponding to D344 of CMV in the HHV and VZM UL89 homologs, BDCRB and TCRB are unlikely to interact with HHV and VZV. This accords with in vitro results showing that benzimidazole ribonucleosides have little to no effect against HHV, VZV, HHV-6, and HHV-8 [30
]. Along with the specific nature of these compounds, they also are metabolized too rapidly in vivo, despite their effectiveness in cell culture [36
]. To develop more biologically stable compounds, analogs have been derived from BDCRB, including acetylated, tetrahalogenated benzimidazole D-ribonucleosides, 2-bromo-4,5,6-trichloro-1-(2,3,5-tri-O
-acetyl-β-d-ribofuranosyl) benzimidazole (BTCRB) and 2,4,5,6-tetrachloro-1-(2,3,5-tri-O
-acetyl-β-d-ribofuranosyl benzimidazole (Cl4RB), which inhibit DNA cleavage and packaging [38
]. BTCRB is suspected to inhibit the ATPase activity of pUL56 [38
], while Cl4RB is believed to interfere with the interaction between pUL56 and the portal protein pUL104 [40
]. These two compounds are shown to be active against VZV, rat cytomegalovirus, and human cytomegalovirus, however Cl4RB has no effect on HHV-1 and the effects of BTCRB are minimal [39
Raltegravir, an HIV integrase inhibitor, exhibits efficiency against herpes viruses. The effect of raltegravir on herpes viruses is attributed to an inhibition of the large subunit of the viral terminase because it has the same RNase H-like fold as the HIV integrase [31
]. Recently however, drug resistance to raltegravir has been traced to HHV-1 UL42 coding for the DNA polymerase accessory factor [41
]. Due to the fact that pUL42 is not a part of the terminase, it was determined that raltegravir likely inhibits DNA replication through the polymerase accessory factor rather than DNA cleavage and packaging through the terminase [41
]. Letermovir (AIC246 or MK-8228), is another promising new antiviral drug for the treatment of CMV. Letermovir is believed to inhibit the viral terminase complex by targeting pUL56 because L241P, R369S, and C325Y mutations in pUL56 correlate with resistance to letermovir [31
]. The inhibitory effect of letermovir is believed to be distinct due to the lack of cross-resistance of letermovir-resistant CMV strains to benzimidazoles [31
]. While letermovir has no target-related toxicity, and has a good safety profile, it is specific for human CMV and is ineffective against other viruses, including the remaining herpesviruses [44
Although numerous antivirals have been developed against HHV, many of these compounds are prone to resistance, have poor bioavailability and high toxicities, and are too specific to be utilized against other HHVs [45
]. We focus on the most conserved protein in HHVs. We propose DNA packaging terminase pUL15 as a drug target that also could be utilized in other herpesviridae and, possibly, a larger set of viruses utilizing a DNA packaging motor. We also suggest that the hinged region between the two domains of pUL15 is a promising drug target to inhibit its overall function.
Many dsDNA viruses utilize a motor to package their DNA into an empty capsid and, therefore, may be classified under type II DNA-packaging. However, the exact mechanism of the DNA packaging with detailed movements of the machinery are unknown. Most dsDNA viruses contain packaging proteins that are composed of two separate domains, an ATPase domain as well as a DNA-binding/nuclease domain. These two functions are linked together, which is necessary for proper DNA packaging to occur. Several hypotheses were formulated to try to answer this question. One popular theory involves the idea that proteins in the packaging motor push the DNA into the capsid in a linear manner. Considering the linear hypothesis, it is believed that the protein utilizes the hydrolysis of ATP to clamp onto the DNA and push it up into the empty capsid [91
]. This mechanism begins with the DNA-binding domain binding the viral DNA. Once the DNA is bound, the DNA-binding domain moves the viral DNA closer to the ATPase domain of the same protein. A conformational change takes place in the ATPase domain, prepping the ATPase active site for hydrolysis. The subsequent hydrolysis of ATP causes another conformational change that moves the DNA-binding domain in such a way that it pushes the viral DNA into the capsid. When ADP and Pi are released, the DNA-binding domain releases the DNA and returns to its original conformation. When the DNA is released from the DNA-binding domain, it ends up being bound to the DNA-binding domain of an adjacent subunit [92
A second hypothesis, the rotation theory, states that the DNA is packaged into the capsid with a screw-like motion as the packaging proteins rotate to facilitate this motion. More specifically, it refers to the possibility of a connecter protein in the packaging motor rotating around the viral DNA and, therefore, using this movement to screw the DNA into the capsid. However, more recent studies suggest that dsDNA viruses are unlikely to utilize the rotational mechanism. It was reported that rotational motors have relatively small channels that are typically smaller than the width of dsDNA, so when the DNA goes through that channel, it would require the dsDNA to split into ssDNA [91
]. A more plausible hypothesis has been suggested where the packaging motors utilize a revolving mechanism instead. This means that rather than the proteins rotating to push the DNA through, the proteins are relatively stationary and the DNA revolves around the channel due to the backbone interactions with the walls of the channel. Revolution motors have been found to have diameters larger than the width of dsDNA, allowing the DNA to go through the revolving mechanism [91
Finally, the “Scrunchworm Hypothesis” considers DNA to be what provides the energy for its own translocation rather than the protein doing all of the work [93
]. During the process of packaging, proteins bind to the DNA and remove it from the solvent, causing its dehydration and compression. The energy stored in the DNA at this compressed state then is responsible for the ultimate movement of the DNA into the empty capsid.
The GNM analysis of pUL15 indicates that it contains two domains connected by a hinge that move in opposite directions of one another. When dsDNA was docked onto our model, we found that the DNA was surrounded by three loop structures flanking it on both sides. Further, the ANM model suggests that the movement of the protein upon ATP hydrolysis most likely would be perpendicular to that of the DNA, contrary to the linear theory. Our data indicates that the packaging of DNA would follow more closely with either the rotation or revolution theory. Due to the limitations of the rotation theory, particularly the small size of the channel for the dsDNA to pass through, it is more likely that the DNA is packaged according to the revolution theory in HHV-1 due to its abundant space for the dsDNA to pass through the channel.
Proteins directly involved in the propagation of the concatemeric viral DNA into the capsid have a significant potential as drug targets due to their essential role in the formation of new virus particles. Out of all of the packaging proteins, terminase proteins maintain the crucial functions of utilizing ATP hydrolysis to push the DNA into the capsid and cleave it once packaging is complete. This makes them promising drug targets because there are multiple functions that may be disturbed to prevent the completion of virus particles. Using this information, the hinge region of pUL15, connecting the DNA and ATP binding domains, is an attractive allosteric target. While targeting DNA- and ATP- binding sites may inhibit the function of this protein, these binding pockets are not unique to the virus. Therefore, we hypothesize that the hinge region can provide an allosteric target to alter the conformational change in pUL15, preventing the movement of the dsDNA into the capsid. It is a favorable site for drug design as multiple HHVs share the same feature in their packaging motor, yet it is specific to the virus and absent in host proteins.