JC Virus Minor Capsid Proteins Vp2 and Vp3 Are Essential for Virus Propagation

Abstract

Virus-encoded capsid proteins play a major role in the life cycles of all viruses. The JC virus capsid is composed of 72 pentamers of the major capsid protein Vp1, with one of the minor coat proteins Vp2 or Vp3 in the center of each pentamer. Vp3 is identical to two-thirds of Vp2, and these proteins share a DNA binding domain, a nuclear localization signal, and a Vp1-interacting domain. We demonstrate here that both the minor proteins and the myristylation site on Vp2 are essential for the viral life cycle, including the proper packaging of its genome.

JC virus (JCV) is the causative agent of the fatal demyelinating disease progressive multifocal leukoencephalopathy (PML) (22). PML develops from a lytic infection of the myelin-producing oligodendrocytes in the central nervous system. JCV is widespread in the human population, where it is estimated that 70% of people are seropositive for the virus (21). JCV is generally latent but can traffic to the central nervous system upon immunosupression, lytically infecting oligodendrocytes and causing PML (1, 15, 24).

JCV contains a small, nonenveloped, double-stranded DNA genome that is organized into early and late coding regions. These regions are divided by the regulatory region containing a bidirectional promoter and the viral origin of DNA replication. The late region contains agno, Vp1, Vp2, and Vp3. The V antigens make up the virus capsid, consisting of 360 molecules of the major coat protein Vp1. These are arranged in 72 pentamers, creating the icosahedral shape. One of the minor coat proteins, Vp2 or Vp3, lies in the center of each pentamer. Vp3 is identical to the C′-terminal two-thirds of Vp2; this shared domain is comprised of the nuclear localization signal (NLS), the DNA binding domain, and the Vp1-interacting domain (2, 4, 5, 8) (Fig. 1B). Vp2 is also modified by a myristylation moiety on its N′ terminus. Myristylation is the cotranslational addition of a fatty acid. Myristyl proteins may be either cytoplasmic or membrane associated. The myristyl group is transferred by the enzyme N-myristyl transferase after the methionine is removed and an N′-terminal glycine residue is recognized (23).

FIG. 1.

(A) Diagram of capsid proteins that will remain after site-directed mutagenesis. (B) Diagram of Vp2 and Vp3 and their functional domains.

The viral DNA is packaged with histones H2A, H2B, H3, and H4 and creates a minichromosome structure that is almost indistinguishable from the host's chromatin (20). The host cell type specificity can be controlled through transcriptional blocks to infection as well as virus-receptor interactions (3, 7, 19).

Previous work evaluating the role of minor coat proteins in the related polyomavirus simian virus 40 (SV40) found that Vp2 was dispensable but that Vp3 was necessary for infection. It also suggested that the importance of Vp3 could lie in its activation of poly(ADP-ribose) polymerase (9). This overactivation of poly(ADP-ribose) polymerase is thought to deplete intracellular ATP and cause necrosis, thereby releasing the virus (9, 10). The SV40 minor proteins have also been shown to contain lytic properties in bacteria that could aid in virus release from the cell (6).

Similar work has also been performed for mouse polyomavirus where Vp2, Vp3, and Vp2 myristylation mutants have been produced by two different labs. Both labs determined that Vp2 and Vp3 are essential for virus production and infection. The myristylation site is also necessary and has been postulated to be important for both exit from the cell and reinfection (16, 25).

To determine the role of the minor proteins in the human polyomavirus JCV, we used site-directed mutagenesis to eliminate Vp2 and Vp3 and the myristylation site on Vp2. The Mad1-SVEΔ strain of JCV was linearized at the BamHI sites and subcloned into pUC19 for viral DNA propagation in bacteria. The start sites of Vp2 and Vp3 were replaced by alanine residues by site-directed mutagenesis using GeneEditor (Promega). The results of the mutagenesis were confirmed by sequencing (the primers used for mutagenesis were 5′-gtgttttcaggttcGCgggtgccgcacttg-3′ [Vp2] and 5′-cagcagccagctGCggctttacaatta-3′ [Vp3], where mismatched nucleotides are shown in uppercase). The double mutant was also created, using both primers to eliminate both Vp2 and Vp3. Additional mutants were created by the removal of the myristylation site at position 2 of Vp2 by changing the glycine to either an alanine, glutamate, glutamine, or histidine (the primers used for mutagenesis were 5′-ggttcatcgCtgccgcac-3′ [G2A], 5′-gttttcaggttcatgGAAgccgcacttgcac-3′ [G2E], 5′-gttttcaggttcatgCATgccgcacttgcac-3′ [G2H], and 5′-gttttcaggttcatgCAAgccgcacttgcac-3′ [G2Q], where mismatched nucleotides are shown in uppercase) (Fig. 1A).

Mutant and wild-type genomes were linearized, and equal amounts of viral DNA were transfected into permissive SVG-A cells by using Lipofectamine (Invitrogen) and Plus (Invitrogen) reagents. They were scored at 84 h posttransfection for viral expression by indirect immunofluorescence assay of Vp1. At 84 h posttransfection, Vp1 was expressed in approximately 5% of the cells, which is consistent with the transfection efficiency of these cells. Vp1 has a weak monopartite NLS, in contrast to the strong bipartite NLSs of its family members SV40 and BK virus (11, 26). JCV therefore relies on the presence of the minor proteins for efficient nuclear localization. All samples containing one minor protein correctly localized Vp1 to the nucleus, indicating the mutant DNA is able to be expressed and that only one minor protein is needed for Vp1 import (Fig. 2A). The double mutant (Vp2⁻/Vp3⁻) did not localize Vp1 to the nucleus as efficiently as either of the single mutants, which is consistent with the weak NLS in JCV Vp1 (Fig. 2A). The infection was followed for 19 days, and the mutants were unable to efficiently propagate the infection. In contrast, wild-type virus spread to 85% of the cells (Fig. 2B and C). This suggests that JCV requires Vp2, the myristylation site on Vp2, and Vp3 for completion of its life cycle. This also shows that Vp2 and Vp3 have distinct functions in the viral life cycle, as neither can be complemented by the other protein.

FIG. 2.

Viral propagation requires the minor proteins. (A) Nuclear localization was monitored by indirect immunofluorescence of Vp1. All mutants except the double mutant properly localize Vp1. (B) Viral growth was monitored by indirect immunofluorescence of Vp1. All mutants are expressed at 4 days posttransfection but fail to grow over time. (C) Growth curves were created for wild-type (WT) and mutant viruses by determining the average number of Vp1-positive cells/250 cells counted.

The lack of the minor proteins could affect many aspects of the viral life cycle. These proteins have been shown to localize to ND10 domains in the nucleus, so their absence could cause mislocalization and therefore assembly defects (27). To address an assembly defect in the life cycle, a DNase protection assay was performed. Virus was harvested from transfected cells at 4, 13, and 22 days posttransfection, and genome protection by the capsid was tested with DNase treatment. The virus was harvested by freeze-thawing, sonication, membrane lysis with 2.5% deoxycholate, and centrifugation to remove cell debris. Equal amounts of virus harvested from the cells were treated with a 20-fold excess of DNase I or were mock treated with water. The capsids were removed using proteinase K, and the genomes were purified and amplified by PCR (primers JCV_for401 [5′-GTG AAG ACA GTG TAG ACG G-3′] and JCV_rev1070 [5′-GAA TTT CCT GAG AGG TTA AGC-3′]). Cells that received no viral DNA were used as a control to show that cellular DNA was not being amplified. At all of the time points, wild-type virus showed protection of its DNA, whereas the mutants were unable to protect their DNA (Fig. 3, compare lanes 3, 5, 8, and 10). This lack of amplification in the mutant samples is not due to a lack of viral DNA as shown by the mock-treated lanes. (Fig. 3, compare lanes 4, 6, 9, and 11). At the 4-day time point the wild type and mutants were expressed equally, suggesting that the result is not due to excess wild-type virus.

FIG. 3.

DNase sensitivity assay. Equal amounts of virus was obtained for each time point. The virus samples were subjected to treatment with a 20-fold excess of DNase I or were mock treated with water. The capsids were then removed by proteinase K treatment and the genomes amplified by PCR. Control, uninfected cells; N, PCR control that received no input DNA.

From these results we concluded that both of the minor proteins and the myristylation of JCV Vp2 are necessary for efficient propagation of the virus. Neither minor protein is able to functionally substitute for the other. Both minor proteins and the myristylation site are needed for the correct packaging of the virus. Studies to further understand the assembly defects are under way in our lab. These studies will continue to evaluate whether there is a difference in the localization of the capsid proteins or the genome in the mutant viruses.

Our results are also consistent with recent findings using mouse polyomavirus. When the myristylation moiety of polyomavirus is mutated to histidine, glutamine, and glutamate, the viruses have a lower viral burst and therefore have fewer infected cells over time (16). Also, when the myristylation site is mutated to a glutamate, electron microscopy studies show an altered morphology. These mutants were able to create capsid-like structures but were less regular and less compact (14).

Myristylated proteins have been known to play a key role in assembly for other nonenveloped viruses, such as poliovirus (17, 18). It is possible that the myristyl moiety makes key contacts within the virion that are necessary for the structural integrity of the virus. Additionally, the assembly defects could be because of unique protein interactions between the minor proteins and cellular chaperones.

The inability of the virus to protect DNA when it is lacking the minor proteins is also consistent with studies using virus-like particles (VLPs). One study clearly showed that JCV VLPs, consisting of Vp1 alone, were unable to protect genome-size DNA from DNase degradation (28). It has also been shown that under physiological conditions, Vp2 and VP3 are both required for SV40 Vp1 to form VLPs (12, 13).