Formation of the Nucleocapsid

Capsid assembly is coordinated by the assembly protein precursor (pAP, pUL80.5, 38 kDa) and the genetically related protease precursor (pPR, pUL80a, 74 kDa), both of which are ultimately eliminated from the maturing particle. These proteins are encoded by 3'-coterminal in-frame genes, with the consequence that the car-boxyl approximately 60% of pPR is identical to pAP (Fig. 2). Two smaller proteins encoded by the same set of genes (pUL80.4 and pUL80.3) have unknown functions that are dispensable for growth in cell culture (N. Nguyen and W. Gibson, unpublished data) from mutant viruses having one or both translational start methionines replaced with isoleucines. Key amino acid sequences within these proteins that enable their function are illustrated in Fig. 3.

A working model of CMV capsid formation is illustrated in Fig. 4. The earliest steps in the assembly process begin in the cytoplasm. One pathway leads to proto-capsomers and is initiated when the amino conserved domain (ACD) promotes pAP

l-Site l-Site

Fig. 3 Landmarks on the assembly protein and protease precursors. The assembly protein precursor (pAP) has the same amino acid sequence as the carboxyl half of the protease precursor (pPR) and includes the following sequences of interest: the amino conserved domain (ACD), which promotes self-interaction of pAP and pPR; the carboxyl conserved domain (CCD), which promotes

Fig. 3 Landmarks on the assembly protein and protease precursors. The assembly protein precursor (pAP) has the same amino acid sequence as the carboxyl half of the protease precursor (pPR) and includes the following sequences of interest: the amino conserved domain (ACD), which promotes self-interaction of pAP and pPR; the carboxyl conserved domain (CCD), which promotes

Fig. 4 Capsid assembly model. Shown here are interactions between the major capsid protein (MCP, pUL85, narrow trapezoids), assembly protein precursor (pAP, pUL80.5 lines with empty circles), and protease precursor (pPR, pUL80a, lines with filled circles), and some of the putative complexes they form (1). The largest represents a complete capsomer precursor (protocapsomer) (2), but there is no direct evidence that its cytoplasmic assembly reaches completion. The minor capsid protein (mCP, pUL85, 35 kDa, light ovals) and mCP-binding protein (mCBP, pUL46, 33 kDa, darker oval) also interact with each other in the cytoplasm to form heterotrimers, called triplexes (3). The two types of oligomers are translocated into the nucleus (4) and coalesce to form the procapsid (Pro), incorporating the portal-protein complex (pUL104, 78 kDa, broken trapezoid at bottom of capsid) (5). The terminase subunits are indicated by shaded ellipses below the portal complex. Activation of pPR (6) results in cleavage and elimination of the internal scaffolding proteins (pPR and pAP) from the capsid, before or during the process of DNA packaging (7). Brackets around the B-capsid (B) indicate uncertainty about the nature of putative intermediate(s) between procapsids and DNA-containing nucleocapsids. (Modified from Gibson 2006)

Fig. 4 Capsid assembly model. Shown here are interactions between the major capsid protein (MCP, pUL85, narrow trapezoids), assembly protein precursor (pAP, pUL80.5 lines with empty circles), and protease precursor (pPR, pUL80a, lines with filled circles), and some of the putative complexes they form (1). The largest represents a complete capsomer precursor (protocapsomer) (2), but there is no direct evidence that its cytoplasmic assembly reaches completion. The minor capsid protein (mCP, pUL85, 35 kDa, light ovals) and mCP-binding protein (mCBP, pUL46, 33 kDa, darker oval) also interact with each other in the cytoplasm to form heterotrimers, called triplexes (3). The two types of oligomers are translocated into the nucleus (4) and coalesce to form the procapsid (Pro), incorporating the portal-protein complex (pUL104, 78 kDa, broken trapezoid at bottom of capsid) (5). The terminase subunits are indicated by shaded ellipses below the portal complex. Activation of pPR (6) results in cleavage and elimination of the internal scaffolding proteins (pPR and pAP) from the capsid, before or during the process of DNA packaging (7). Brackets around the B-capsid (B) indicate uncertainty about the nature of putative intermediate(s) between procapsids and DNA-containing nucleocapsids. (Modified from Gibson 2006)

Fig. 3 (continued) interaction of pAP and pPR with MCP; nuclear localization signals 1 and 2 (NLS1, NLS2); casein kinase II phosphorylation site (black dots between NLS1 and NLS2) of undetermined significance; mitogen-activated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK-3) sites whose phosphorylation antagonizes pAP self-interaction (Casaday et al. 2004); and the five pPR self-cleavage sites: the maturational site (M site, VNA^S), which severs the linkage of pPR and pAP to MCP; the release site (R site, YVKA^S), which separates the proteolytic domain (assemblin) of pPR from the scaffolding portion (carboxyl end); the internal site (I site, VEAJ- A), which converts assemblin from an active single-chain form to a two-chain form that retains activity, the cryptic site (C site, VDA^S) that interrupts the assemblin dimer interface, and the tail site (T site, VLA^- A) detected upon refolding denatured pPR (Brignole and Gibson 2007). Also shown is the amino acid sequence of the ACD and location of the critical Leu47 (red; Leu382 in context of pPR sequence) within it. (Modified from Gibson 2006)

self-interaction, which in turn potentiates or stabilizes pAP binding through its carboxyl conserved domain (CCD) to the major capsid protein (MCP, pUL86) (Wood et al. 1997; see Figs. 3 and 4, step 1). This interaction enables MCP, which lacks its own nuclear localization sequence (NLS), to be translocated into the nucleus (Fig. 4, step 4) as part of the pAP-MCP complex via two NLSs present in pAP (Wood et al. 1997; Plafker and Gibson 1998). Because its carboxyl end is identical to pAP (Figs. 2, 3), the proteinase precursor can interact with itself and pAP through its ACD, and with MCP through its CCD. Mimicking these pAP interactions ensures incorporation of pPR into the capsid cavity, where its enzymatic function is required. The composition and variety of the complexes formed (e.g., pAP-MCP; pAP-pPR-MCP; pPR-MCP) and the extent to which they progress in the cytoplasm toward completely preassembled protocapsomers (Fig. 4, step 2) is unknown.

The biological relevance of ACD-mediated pAP self-interaction was established by using mutant viruses, which were found to replicate slowly, assemble nucleo-capsids inefficiently, and yield roughly 20-fold less virus than wild type (Loveland et al. 2007). Mutant viruses have also been used to verify the biological requirement for pAP NLS, by showing that loss of both is lethal, loss of either one alone slows nuclear translocation of MCP, and loss of NLS2 inhibits virus replication more profoundly than loss of NLS1 (Nguyen et al. 2007). Restriction of NLS2 to the P-herpesvirus pAP homologs and its comparatively greater impact on virus replication suggests it may have a different or additional group-specific function.

Through a similar subunit assembly process, the triplex proteins associate in the cytoplasm before translocation into the nucleus (Fig. 4, step 3) (Baxter and Gibson 1997; Spencer et al. 1998). Like MCP, the minor capsid protein (mCP, pUL85) does not enter the nucleus when expressed alone in transfected cells, even though its size is small enough to be allowed entry by diffusion. The similar-size mCBP, in contrast, does enter the nucleus on its own and when mCP and the mCP-binding protein (mCBP, pUL46) are expressed together, they co-localize to the nucleus (Baxter and Gibson 1997). This is attributed to the rapid formation of approximately 70-kDa mCP homodimers that require interaction with an NLS-bearing mCBP for nuclear translocation. These sequential cytoplasmic interactions of pAP and pPR with each other and with MCP, and of mCP with itself and with mCBP, initiate and direct the assembly process and consequently have the potential to help modulate it. Cytoplasmic preassembly may also increase the fidelity and efficiency of procapsid formation by ensuring delivery into the nucleus of correctly organized capsid substructures.

The HCMV portal protein (pUL104, 78 kDa) may also interact with pAP, as it does in HSV (Singer et al. 2005). However, little is known about when and where its self-interaction and interaction with pAP occur. Unlike MCP, the portal protein contains its own NLS (Patel and MacLean 1995; Patel et al. 1996) and would not seem to require interaction with pAP as an NLS-bearing nuclear-translocation escort.

Following translocation into the nucleus, the pAP-MCP, pPR-MCP, and pAP-pPR-MCP complexes or protocapsomers interact more extensively with one another and associate with the triplexes and portal protein complex to form procapsids (Fig. 4, step 5). These unstable particles (Newcomb et al. 1999; Rixon and McNab 1999), first evidenced in HSV (Newcomb et al. 1999; Rixon and

McNab 1999; Newcomb et al. 2000) and constituted or closely approximated in vitro (Newcomb et al. 1999, 2000), are less angular than late-stage capsids and contain scaffolding proteins but no viral DNA. An involvement of the pAP amino-conserved domain during this nuclear stage of capsid formation was discovered with the L47A mutant virus, which showed a dramatically altered distribution of pAP within the nucleus (Fig. 5) and gave rise to overall fewer and aberrant capsids (Loveland et al. 2007).

Absence of detected procapsid formation in the cytoplasm, where all of the necessary proteins are present, may be explained by the comparatively higher protein concentrations in the nucleus or by the presence of specific nuclear initiating or enhancing factors. Alternatively or additionally, there may be assembly-enhancing changes in the complexes that signal or result from nuclear translocation. In HSV, where it has been possible to examine capsid assembly in isolation from other viral proteins, the homologs of MCP, pAP, mCP, and mCBP (i.e., HSV VP5, preVP22a, VP23, VP19c) are all necessary and collectively sufficient to assemble the capsid shell (Tatman et al. 1994; Thomsen et al. 1994). Similar attempts to make CMV capsids from their recombinantly cloned and expressed counterpart genes have not yet succeeded and it is unresolved whether this is due to technical factors (e.g., CMV protein expression levels too low) or perhaps to differences in the minimal compliment of proteins required.

Fig. 5 Distribution of UL80 proteins in nuclei of HCMV-infected cells. An HCMV-AD169 bac-mid was mutated to block self-interaction of the UL80 proteins (i.e., point mutation L47A in pAP sequence) (Loveland et al. 2007). Both the HCMV-AD169 bacmid and the parental wild-type bacmid were also modified to incorporate a tetra-cysteine tag (CCPGCC) into the UL80 proteins. Virus derived from each bacmid was used to infect human foreskin fibroblasts, which were stained with the biarsenical dye FIAsH when strong cytopathic effects were observed. Shown here are images of the stained, living cells taken by confocal fluorescence microscopy. The FIAsH-stained UL80 proteins are organized in tubular and rod-shaped structures in nuclei of cells infected with the L47A mutant (first three panels from the left), but in an entirely different pattern reminiscent of the intranuclear inclusions typically observed in nuclei of cells infected with wild type virus (right-hand panel). (Modified from Loveland et al. 2007)

Fig. 5 Distribution of UL80 proteins in nuclei of HCMV-infected cells. An HCMV-AD169 bac-mid was mutated to block self-interaction of the UL80 proteins (i.e., point mutation L47A in pAP sequence) (Loveland et al. 2007). Both the HCMV-AD169 bacmid and the parental wild-type bacmid were also modified to incorporate a tetra-cysteine tag (CCPGCC) into the UL80 proteins. Virus derived from each bacmid was used to infect human foreskin fibroblasts, which were stained with the biarsenical dye FIAsH when strong cytopathic effects were observed. Shown here are images of the stained, living cells taken by confocal fluorescence microscopy. The FIAsH-stained UL80 proteins are organized in tubular and rod-shaped structures in nuclei of cells infected with the L47A mutant (first three panels from the left), but in an entirely different pattern reminiscent of the intranuclear inclusions typically observed in nuclei of cells infected with wild type virus (right-hand panel). (Modified from Loveland et al. 2007)

Taking procapsids as the end point of capsid assembly, two major changes occur during its maturation: elimination of the scaffolding proteins and incorporation of viral DNA. The two processes appear to be coupled. Elimination of the internal scaffolding proteins is effected by proteolysis catalyzed by pPR (Fig. 4, step 6). Although able to cleave itself at five sites, pPR activity appears to be modulated during infection such that its autoproteolysis is initiated following procapsid formation. It has recently been determined using purified HCMV pPR that the active enzyme is a trimer or tetramer whose primary subunit interaction is through the amino-conserved domain of its scaffolding region (Brignole and Gibson 2007). This quaternary structure is very different from that of purified assemblin, whose monomer activates by dimerizing through sequences located in its carboxyl end (Chen et al. 1996; Cole 1996; Darke et al. 1996; Margosiak et al. 1996; Qiu et al. 1996; Shieh et al. 1996; Tong et al. 1996). Active pPR cleaves the R site to release the well-characterized proteolytic domain, assemblin, and cleaves the M site to sever the tail sequence linking itself and pAP to MCP in the capsid shell (Figs. 2, 3).

Although having comparable overall enzymatic rates (e.g., kcat/KM) , pPR and assemblin are distinguished in ways thought to reflect mechanistic differences (Brignole and Gibson 2007). First, whereas imidazole can chemically rescue the enzymatic activity of assemblin substituted with Gly at its catalytically critical His63, it fails to restore activity to the same mutation in pPR (McCartney et al. 2005), indicating a difference in the catalytic sites of the two forms of the enzyme. Second, the sequences driving assemblin dimerization have comparatively little effect on pPR oligomerization, yet this interaction of assemblin is thought to induce catalytic-site changes required for its activity (Tong et al. 1996; Buisson et al. 2002). And third, there is evidence that the enzymatic rates of pPR and assemblin may be comparable because of offsetting differences in their kc at (higher for pPR) and KM, again suggesting catalytic-site differences between pPR and assemblin that may be important to regulating activity.

Proteolysis results in essentially all M and R sites being cleaved. HCMV pPR makes three additional cleavages. Two are at the internal (I) and cryptic (C) sites within assemblin and reduce production of infectious virus when blocked (Chan et al. 2002; Loveland et al. 2005). These cleavages reduce the size and interactions of the scaffolding proteins, facilitating their elimination from the capsid in preparation for DNA packaging (Fig. 4, step 7). Unlike HSV, which retains assemblin in its mature virion, all remnants of HCMV pAP and pPR, including assemblin, are eliminated from the capsid. Absence of counterpart I and C sites in the HSV assemblin homolog, and persistence of HSV assemblin in the mature virion, may reflect a requirement for additional space within the CMV capsid to accommodate its 51% longer DNA genome (Chan et al. 2002; Loveland et al. 2005). The significance of a fifth cleavage site just discovered in the carboxyl tail (T site) of purified pPR is undetermined. Electrostatic repulsion by the incoming viral DNA has been suggested to play a role in displacing the internal scaffolding proteins (McClelland et al. 2002), and specific pAP phosphorylations that weaken scaffolding protein self-interactions may by important to this process (Casaday et al. 2004; Gibson 2006).

Maturational proteolysis converts at least some procapsids to B-capsids (Fig. 6b), which differ by having an angular appearance (instead of round), and containing cleaved forms of the internal scaffolding proteins (instead of precursors). B-capsids are depicted in Fig. 4 as intermediates in the assembly pathway, but it has been difficult to demonstrate their maturation to DNA-containing nucleocapsids (O'Callaghan and Randall 1976; Ladin et al. 1982; Lee et al. 1988; Sherman and Bachenheimer 1988; Church and Wilson 1997). One explanation is that only a small percentage of the relatively large B-capsid pool incorporates DNA, making their loss from the pool difficult to detect. Moreover, once DNA packaging begins, the particles involved may become compositionally heterogeneous (e.g., a decreasing amount of scaffolding proteins and an increasing amount of DNA) and, consequently, escape detection by methods routinely used to recover and characterize capsids (e.g., sedimentation and equilibrium centrifugation). An alternate and plausible interpretation is that B-capsids are formed when early steps in the procapsid maturation process fail, such as timely cleavage and elimination of the scaffolding proteins or successful initiation of DNA packaging.

Fig. 6 Capsids in nucleus of CMV-infected cells. These selected particles from electron micrographs of CMV-infected human foreskin fibroblasts show features consistent with the DNA packaging scheme illustrated in Fig. 4. a Capsid at top appears to be at comparatively early stage of DNA packaging. The elongated putative nascent core is smaller and less electron dense than those at the top of panel b and bottom of panel c, and is off center to the extent of appearing to touch the inner wall of the capsid and to be continuous with more filamentous material extending through the capsid (presumably via portal) into the nucleoplasm (arrow). b Capsid at lower left appears to contain scaffolding proteins absent in the particle just above it. Capsid at top contains DNA core with filamentous extension through the capsid shell and into the nucleoplasm (arrow). c Lower capsid appears to contain scaffolding proteins and upper capsid is interpreted to be nearly finished packaging its DNA, the end of which may extend out to the left of the capsid (arrow)

Fig. 6 Capsids in nucleus of CMV-infected cells. These selected particles from electron micrographs of CMV-infected human foreskin fibroblasts show features consistent with the DNA packaging scheme illustrated in Fig. 4. a Capsid at top appears to be at comparatively early stage of DNA packaging. The elongated putative nascent core is smaller and less electron dense than those at the top of panel b and bottom of panel c, and is off center to the extent of appearing to touch the inner wall of the capsid and to be continuous with more filamentous material extending through the capsid (presumably via portal) into the nucleoplasm (arrow). b Capsid at lower left appears to contain scaffolding proteins absent in the particle just above it. Capsid at top contains DNA core with filamentous extension through the capsid shell and into the nucleoplasm (arrow). c Lower capsid appears to contain scaffolding proteins and upper capsid is interpreted to be nearly finished packaging its DNA, the end of which may extend out to the left of the capsid (arrow)

By analogy with the bacteriophage DNA packaging system, once pAP and pPR are eliminated, DNA is incorporated with the help of the terminase-portal complex (Fig. 4, step 7). Examples of intranuclear capsids apparently involved in this process are shown in Fig. 6. The portal protein forms a 12-subunit homo-oligomeric ring at a single vertex of the capsid through which the viral DNA can enter and leave (Newcomb et al. 2001; Trus et al. 2004). At 29-Ä resolution the HSV portal complex resembles that of bacteriophage (Trus et al. 2004). Its CMV homolog, pUL104, also appears restricted to a single capsid vertex (Dittmer and Bogner 2005). The herpesvirus DNA cleavage/packaging enzyme (terminase) is composed of two subunits (Poon and Roizman 1993; Baines et al. 1994). The larger (HCMV pUL56, 96 kDa) has properties consistent with it being a counterpart of the large subunit of bacteriophage terminase (Bogner et al. 1993, 1995, 1998; Holzenburg and Bogner 2002; Scheffczik et al. 2002): (a) it associates with a smaller subunit (pUL89, 77 kDa, (b) it binds double-stranded viral DNA, and (c) it binds to the capsid (White et al. 2003). Retention of DNA in the capsid is believed to be stabilized by a protein (HCMV pUL7, 71 kDa) whose HSV homolog (pUL25, 60 kDa) has a mutant phenotype that fails to stably package DNA (McNab et al. 1998; Ogasawara et al. 2001; Sheaffer et al. 2001). This protein is considered a possible counterpart of the bacteriophage cap protein, but may exert its effect by binding at multiple sites on the capsid surface (Newcomb et al. 2006).

Was this article helpful?

0 0

Post a comment