Relating the structure of the HSV-1 UL25 DNA packaging protein to its function

The herpes simplex virus type 1 (HSV-1) UL25 protein (pUL25) is a minor capsid protein that is essential for packaging the full-length viral genome into preformed precursor capsid. It is also important in virus entry and recently has been implicated in the egress of the virus from the cell (Coller et al., 2007, Preston et al., 2008). The crystallographic structure of an N-terminally truncated form of pUL25 (residues 134-580) has been determined to 2.1 Å, revealing a protein with a novel fold that consists mostly of a-helices and a few minor b-sheets (Bowman et al., 2006). An unusual feature of the protein is the presence of numerous flexible loops extending out from the stable core and its distinctive electrostatic distribution. Five of the extended loops contain unstructured regions, L1-L5, with three additional unstructured amino acids, L6, located at the carboxyl terminus of the protein (Bowman et al., 2006). Four potentially functional clusters of residues, C1-C4, were identified on the surface of the protein using evolutionary trace analysis (Lichtartge and Sowa, 2002). To examine the function of the protein in relation to its structure, site-directed mutations were engineered into the UL25 gene in a protein expression plasmid. A series of mutant proteins was generated, each protein containing a deletion of the unstructured residues in one of the six regions, L1-L6. Another set of mutant proteins were constructed with each member containing substitutions of selected amino acids within one of the four potentially functional clusters, C1-C4, or substitutions of the three disordered amino acids in L6. The amino acid substitutions were generally to alanine, but in one case where the SIFT program predicted alanine would not affect the function of the protein an alternative residue was substituted. To determine the functional significance of the uncrystallised part of pUL25, residues 1-133, three deletion mutant proteins that spanned this region (pUL25D1-45, pUL25D1-59 and pUL25D1-133) were included in the study. Although an existing UL25 null mutant, KUL25NS, was available at the beginning of the project for analysis of the mutant proteins, it had been made by the insertion of multiple stop codons in the UL25 ORF and as a result some UL25 sequences were still present within the virus genome. Consequently, during complementation assays recombination between the UL25 sequences in the KUL25NS genome and the transfected expression plasmid generated low levels of wild-type (wt) progeny virus. To improve the sensitivity of the assay, a new deletion mutant, ΔUL25MO, was created that lacked the entire UL25 gene. This mutant failed to form plaques in non-permissive Vero cells and grew well in the complementing cell line, 8-1. However, contrary to previously published work, electron microscopic (EM) analysis revealed that DNA-containing capsids as well as A- and B-capsids were present in the nuclei of both ΔUL25MO- and KUL25NS-infected cells. As expected, none of the progeny from ΔUL25MO-infected Vero cells expressing the wt pUL25 formed plaques on non-permissive cells. Of the 17 mutant UL25 proteins screened in the complementation assay, nine failed to complement the growth of ΔUL25MO in Vero cells. Three of the non-complementing mutant proteins examined altered the phenotype of ΔUL25MO in a transient DNA packaging assay, allowing the mutant virus to package full-length genomes in U2OS cells co-infected with ΔUL25MO and a mammalian baculovirus vector containing the mutant UL25 gene. These results indicate that viral assembly was disrupted in these cells following DNA packaging. However, five of the mutant proteins did not change the pattern of DNA encapsidation of ΔUL25MO in this system, suggesting that the wild-type residues mutated in these proteins are critical for packaging virus DNA. To determine at which point in the virus growth cycle the post-packaging blocks occurred, EM was used to investigate the pattern of virus assembly in ΔUL25MO-infected cells expressing either of the three packaging-competent mutant proteins. In addition, fluorescent in-situ hybridisation (FISH) analysis was performed to establish the distribution of virus DNA in these cells. The results showed that in ΔUL25MO-infected cells expressing two of the mutant pUL25s the C-capsids failed to exit the nucleus, whereas in cells expressing the third post-packaging mutant protein C-capsids were seen in both the nucleus and the cytoplasm. The FISH data confirmed the EM observations. These studies show that two regions of pUL25 are important for egress of the C-capsids from the nuclei. Since these two regions lie in close proximity to each other on the surface of the molecule they may represent a single functional interface of the protein. In addition, another region of pUL25 was identified that was essential for the interactions required for virus assembly after the C-capsids are released into the cytoplasm. The 62 carboxyl-terminal region of the UL36 gene product (pUL36) has previously been shown to contain a capsid-binding domain (CBD) that interacts with pUL25 (Coller et al., 2007). A GST-pull down assay was used to determine whether the mutations in the post-packaging mutant proteins disrupted the interaction of pUL25 with the CBD of pUL36. However, all of these mutant proteins and the wt pUL25 bound to the pUL36 CBD GST fusion protein. In summary, three different classes of pUL25 mutants, each of which affect a different essential function of pUL25, have been identified, revealing that pUL25 is indeed a versatile viral protein. These mutants provide the first evidence that this DNA packaging protein is crucial for virus assembly at two different stages after DNA encapsidation, one in nuclear egress of C-capsids and the other in the assembly of the virus in the cytoplasm