Much of what is known about the structure of polyomavirus T antigens comes from studies of SV40. All polyomavirus T antigens share an N-terminal region of approximately 80 residues that shows structural and sequence homology to the DnaJ or J domain found in host cell HSP40 homologs (Fig. 53.4). Full-length large T antigen is a nuclear phosphoprotein of approximately 700 residues. The molecule’s atomic coordinates have been assembled from crystallography of isolated domains, including the DnaJ and the retinoblastoma protein–binding domains; the DNA-binding domain, also known as the origin-binding domain; and a central domain consisting of residues 251 to 627 that forms a hexamer and contains the adenosine triphosphatase (ATPase) and helicase activities required for viral replication.¹⁶³ Studies using scanning transmission electron microscopy, negative staining with atomic force microscopy, and single-particle reconstruction of cryoelectron microscopy (cryo-EM) images revealed that large T antigen forms a double hexamer in a head-to-head arrangement when bound to the origin of DNA.⁵⁰

A number of functional domains contained within large T antigen are required for viral replication. Functions intrinsic to SV40 large T antigen include the ATPase/helicase domain and
the DNA-binding domain that mediates direct interactions with specific DNA sequences at the origin of replication. In addition to these intrinsic functions, the SV40 large T antigen domains serve to recruit host factors important for viral replication (Fig. 53.5). For example, the N-terminal J domain, containing the canonical residues HPDK, binds and activates the ATPase activity of host cell HSC70.³¹ The DNA-binding domain binds to replication protein A, while the helicase domain binds to the DNA polymerase α/primase complex.¹¹⁸ In addition, the helicase domains of many but not all polyoma large T antigens bind to p53. The outer surface of each SV40 large T antigen hexamer subunit can bind directly to the DNA-binding domain of p53.¹⁶⁷ The large T antigen from MPyV is a notable exception among polyomaviruses because it does not bind to p53, although the ability of all the other polyomavirus large T antigens to bind to p53 has not yet been reported.

Smaller functional motifs within large T antigen include the nuclear localization signal (NLS).¹³²,¹⁵⁴ Mutations that disrupt the NLS result in the cytoplasmic localization of SV40 large T antigen and inability to support the viral lytic life cycle. All mammalian polyomavirus large T antigens contain the conserved residues LXCXE (where X is any residue) that bind directly to the retinoblastoma family of tumor suppressor proteins including pRb (RB1), p107 (RBL1), and p130 (RBL2).⁶⁰,⁶⁸ The large and small T antigens from GHPyV, CPyV, and FPyV each contain the LXCXE motif, while those from APyV contain a related sequence, LXAXE. It is not known if any of the bird polyomavirus T antigens can bind to pRb or p53.

SV40 large T antigen contains a series of posttranslational modifications, including phosphorylation, O-glycosylation, acylation, poly(ADP)-ribosylation, and acetylation. In addition, phosphorylation of C-terminally located threonine residues in SV40 large T antigen creates a phospho-degron motif that binds directly to FBXW7, an F-box substrate adapter. Large T antigen binding to FBXW7 blocks binding to cyclin E and prevents its degradation by the CUL1^FBXW7 RING ubiquitin ligase (Fig. 53.5).²⁷⁷ Phosphorylation regulates some of the functions of the molecule, including its subcellular localization and its ability to participate in the initiation of viral DNA synthesis. These regulatory events will be discussed later in the context of the life cycle.

Polyomavirus small T antigen is found in both the nucleus and the cytoplasm. Small T antigen is a cysteine-rich protein ranging in size from 124 to 198 residues and shares its N-terminus with large T antigen (i.e., those residues encoded up to the 5′ large T antigen splice site) but contains a unique C-terminal region. Small T antigen contains the same N-terminal J domain as large T antigen and a unique C-terminal domain (Fig. 53.4). The unique domain of the mammalian polyomavirus small T antigens contains a highly conserved set of cysteine and histidine residues that bind to two zinc molecules.⁴³,⁴⁴ These zinc-binding domains serve an important role in binding to the cellular protein phosphatase 2A (PP2A).²⁰⁰ PP2A is a trimeric complex consisting of two regulatory subunits A and B that bind to the catalytic C subunit. Small T antigen binds directly to the PP2A A subunit where the B subunit normally binds and thereby displaces or replaces the B subunit. SV40 small T antigen binds specifically to the Aα (PPP2R1A) subunit, while MPyV small T antigen binds to both the Aα and Aβ (PPP2R1B) subunits.⁴ The small T antigen–PP2A complex contains also the Cα (PPP2CA) or Cβ (PPP2CB) catalytic subunit. There are at least 18 different B PP2A subunits identified in mammalian cells. At the very least, SV40 small T antigen can displace the B56α (PPP2R5A), B56γ (PPP2R5C), and PR72/PR130 (PPP2R3A).²¹⁹ It is likely that the polyomavirus small T antigen–PP2A complex not only serves to disrupt the cellular PP2A complexes but also is likely to retain specific phosphatase activity directed toward substrates. Notably, the bird polyomavirus small T antigens do not contain the conserved cysteine/histidine residues that serve to bind zinc, and it is not known if they are capable of PP2A binding.

The middle T antigen of MPyV shares its N-terminal J domain and PP2A-binding domain with small T antigen (Figs. 53.4 and 53.6). The MPyV middle and small T antigens are identical for the N-terminal 191 residues until a splice junction that removes four nucleotides resulting in an additional 230 residues at the C-terminus of middle T antigen. In contrast, small T antigen contains only four unique amino acids after this intron. The unique middle T antigen C-terminus is encoded by an alternate reading frame used for coding the second exon of large T antigen. Notably, middle T antigen contains an N-terminal J domain in common with both large and small T antigen and, like small T antigen, binds to the A and C subunits of PP2A. The additional MPyV middle T residues mediate binding to several proteins involved in signal transduction, including the SRC tyrosine kinase, the SHC1 phosphotyrosine docking protein, phospholipase C (PLCG1), and phosphatidylinositol 3-kinase (PIK3CA and PIK3R1).²²³ MPyV middle T antigen can also bind to the SRC-related tyrosine kinases YES1 and FYN. The C-terminus of middle T antigen contains
a 22-residue hydrophobic domain that moves the newly translated middle T antigen from the cytoplasm through the endoplasmic reticulum (ER) to the inner plasma membrane.²⁸⁵ The combination of membrane localization with recruitment and activation of several enzymes enables middle T antigen to function as a constitutively activated tyrosine kinase that triggers downstream signaling in the RAS and MEK pathways.

Prior to the 21st century, the identity of the cell surface receptors for polyomaviruses was poorly understood. Early studies on SV40 indicated that it used the major histocompatibility complex (MHC) class I antigens to bind to cells.²⁹ Supporting evidence for this model included observations that antibodies against MHC class I blocked virus binding to rhesus monkey kidney cells and the inability of virus to bind well to human cells that do not express MHC class I antigens. Experimentally induced expression of MHC class I in nonexpressing cells restored binding. Although the MHC class I antigens were implicated as the SV40 receptor, they were not sufficient to account for all binding. For example, virus binding occurred only on the apical surface of polarized monkey epithelial cells, while MHC class I antigens were expressed on both the apical and basolateral surfaces. In addition, expression of MHC class I on human kidney epithelial cells was not sufficient for SV40 infection.

Subsequent results challenged the notion that SV40 uses a protein molecule as its receptor, indicating instead that it uses the branched ganglioside, GM1 (Fig. 53.7).²⁶⁵ This finding is more consistent with the route of entry of the virus through endosomes. In these studies, a rat cell line that did not express gangliosides and was unable to be infected by SV40 was rendered susceptible by preincubation with GM1. Gangliosides are glycosphingolipids that combine a sialylated oligosaccharide with ceramide consisting of sphingosine and a fatty acid. The sialic acid is critical for viral binding to the cell. In addition to providing a binding site for the virus, the gangliosides direct the virus to the correct endocytic pathway.²⁰⁹ The efficiency of SV40 infection is dependent on the relative concentration of GM1 on the cell surface as well as its ability to activate focal adhesion kinase (PTK2).²⁴⁴

Other polyomaviruses use different forms of gangliosides as receptors (Fig. 53.7). For example, MPyV uses GT1b or GD1b.⁹²,²⁶⁵ After initial binding to the ganglioside, MPyV interacts with α4β1 integrin that may serve as a secondary or cell type–specific receptor for viral entry.³⁵ BKPyV also uses gangliosides to enter the cell, as judged by restoration of infectivity to otherwise resistant cells upon preincubation with GT1a and GD1a.¹⁷¹ These branched gangliosides are found on the renal tubular epithelial cells that BKPyV normally infects. Another report indicated that an N-linked glycoprotein containing sialylated oligosaccharides can also mediate BKPyV binding.⁶⁶ Both the glycolipid and glycoprotein contain sialic acid, consistent with early reports that BKPyV can hemagglutinate human red blood cells and that this activity was neuraminidase sensitive. There are conflicting reports regarding the receptor for MCPyV. Although it has been shown that MCPyV VP1 capsomeres can bind the ganglioside GT1b,⁷² pseudovirions can bind heparin moieties.²²⁵

JCPyV binds to lactoseries tetrasaccharide c(LTSc), a linear sialylated oligosaccharide that differs from the branched forms reported for other polyomaviruses (Fig. 53.7). LTSc is a pentasaccharide with the terminal sialic acid linked by an α2,6 bond to the penultimate galactose.¹⁹⁴,²⁶⁶ In addition, JCPyV uses the 5HT_2A serotonin receptor, perhaps as a cell type– specific receptor, for viral entry.⁶⁹ This receptor is expressed on
glial cells, the major target cell for JCPyV. Antibodies to the 5HT_2A receptor and receptor antagonists block JCPyV infection, while expressing the receptor in otherwise noninfectible cells renders them susceptible to infection.

The polyomaviruses use several pathways to enter into the cell and pass through the endosomes to the ER (Fig. 53.8). The pathway from the endosomal compartment to the ER and from the ER to the nucleus is not well understood. After binding to gangliosides on the cell surface, SV40 and BKPyV enter the cell using a pathway involving caveolin. The virus is delivered to a neutral pH organelle called the caveosome by endocytosis and then to the ER, where it could be detected by electron microscopy. Recent work has indicated that SV40 may also traffic through a more traditional endocytic pathway to the ER that does not involve a caveosome.⁷⁰ Evidence also indicates that entry of SV40 into the cell requires engagement of a signal transduction cascade through interactions at the cell surface.⁷⁰

Studies with various inhibitors of intracellular structures and processes have shed some light on how the virus travels within the cytosol. Molecules that interfere with microtubules and prevent movement of vesicles from the endosome to the ER interfere with SV40 infection.²³⁵ Infection can be blocked by the drug, brefeldin A, which inhibits trafficking between the ER and the Golgi apparatus,¹⁹⁷ and a number of other drugs that interfere with endosome maturation and trafficking to the ER.⁷⁰

The polyomavirus capsid begins to disassemble in the ER. Evidence for this includes the appearance of epitopes on VP2 and VP3 that become accessible for immunostaining.¹⁹⁷ In addition, structural changes to the capsid, including disulfide bond reduction and isomerization mediated by ER-resident protein disulfide isomerases, can be detected biochemically at the point when the virus enters the ER.⁹³,¹⁷²,²²⁴,²⁷⁴ Furthermore, transmission EM experiments have detected changes in the morphology of virions that were isolated from the ER.¹²¹ It is thought that this change leads to exposure of hydrophobic surfaces of the VP1 molecule on the capsid that facilitates transport across the ER membrane. This step in trafficking also involves ER chaperone proteins. Similar to SV40, it has been shown that the conformation of the MPyV capsid begins to change in the ER due to the action of ERp29, a member of
the protein disulfide isomerase family.¹⁷²,²⁶⁵ Capsid disassembly in the ER or the cytosol may be required because the viral particle is bigger than the functional capacity of the nuclear pore. Evidence for a multistep disassembly process for SV40 includes the exposure of certain VP2 and VP3 epitopes in the ER, while further disassembly occurs in the cytoplasm at a later time point as evidenced by immunoassay detection of the viral genome.¹⁵⁰ The ER-associated degradation pathway, which functions to target misfolded cellular ER proteins for proteasome-mediated degradation, has also been implicated in polyomavirus disassembly.⁹¹,⁹⁶,¹²⁵

The question of how the genome gets carried to the nucleus remains to be answered. For SV40, it has been speculated that release from the vesicular compartment might involve VP1, as noted earlier, or VP2, by virtue of its myristoylated N-terminus inserting into the lipid bilayer.¹⁹⁷ An NLS in VP3 may mediate entry through the nuclear pore complex, because mutations in this NLS inhibit entry of the virion into the nucleus but do not affect capsid assembly or the production of new virions.¹⁹¹

JCPyV enters the cell through clathrin-coated pits, as indicated by use of inhibitors of this pathway as well as the demonstration that JCPyV co-localizes with transferrin, which is known to use clathrin-mediated entry into the cell.²⁰⁶ Both microtubules and microfilaments play a role in trafficking of JCPyV to the nucleus.⁷ As with SV40, binding of JCPyV to its receptor induces a signal transduction cascade required for efficient entry that can be inhibited by genistein, a tyrosine kinase inhibitor.²¹¹ JCPyV appears to signal through the extracellular signal-regulated kinase (ERK) or mitogen-activated protein kinase (MAPK) pathway. Using fluorescently labeled virus-like particles (VLPs), it has been shown that JCPyV particles do not disassemble before they reach the nuclear pore and that the nuclear localization signal of VP1 is required for entry into the nucleus.²¹⁰

After the polyomavirus genome enters the nucleus, it serves as a template for transcription by the cellular RNA polymerase II. Once inside the nucleus, the viral genome becomes wrapped in nucleosomes containing histone H1 in addition to the four core histones that are present in the virus particle.²⁷⁰ Within the cell, the SV40 viral genome contains 24 nucleosomes with a nucleosome-free region of 400 base pairs encompassing the NCCR of early and late promoters and origin of replication. This is in contrast to the virion, where all regions of the SV40 genome are covered with nucleosomes devoid of H1. Chromatin immunoprecipitation can detect transcriptionally active chromatin, defined by the presence of RNA polymerase II, as early as 30 minutes postinfection with SV40.¹¹ The chromatin at this time also contains hyperacetylated histone H3 and H4,¹⁰,¹⁸² which have been associated with chromatin undergoing transcription initiation and elongation.

Transcription of polyomavirus genes is governed by cis-acting sequences in the regulatory region. The SV40 regulatory region has been the most intensively studied and serves as a paradigm for the other polyomaviruses. Seminal studies involving SV40 include the demonstration that AT-rich sequences designated TATA boxes act to direct RNA polymerase II to the proper initiation site for transcription.¹⁵ The SV40 early promoter also contains a series of GC-rich sequences within the 21–base-pair repeat region.⁷³,⁸⁴ The protein SP1, one of the first eukaryotic transcription factors to be cloned, binds to these sequences.⁶⁷ The SV40 early promoter also contains a duplicated element called the 72–base-pair repeat, which was the first cis-acting DNA sequence to be deemed a transcriptional enhancer because it could activate transcription when placed several thousand base pairs distant from the transcription start site.¹⁰⁴,¹⁸⁷ Interestingly, most clinical isolates of SV40 carry only a single copy of the 72–base-pair repeat, and it appears that duplication can be selected during passage in culture.¹⁵⁷

In MPyV, the enhancer element consists of two neighboring enhancers called A and B, alternatively α and β, that can act independently. The activity of the MPyV enhancers is dependent on the cell environment, as viral variants that are selected for growth on differentiated or undifferentiated embryonal carcinoma cells have mutations that map to the enhancers.⁵⁴,¹⁸⁰ The MPyV early promoter is regulated by the cellular factors characterized as polyomavirus enhancer A binding proteins (PEA) or RUNX1, CBFB, and ETV4 that are expressed when cells are growth stimulated with serum.¹²²,¹⁷⁶

The JCPyV promoter shows a distinct tissue-specific activity that correlates with the sites of acute infection. While normal JCPyV virions can attach and enter many types of cells, its host range is restricted to those cells that have the appropriate DNA-binding proteins.²⁶⁷ The JCPyV early promoter
contains sequences that act as transcriptional promoters. These sequences include the TATA box as well as binding sites for the transcription factors SP1, YB1 (YBX1), Pur α (PURA), AP-1, a heterodimer of JUN and FOS, nuclear factor 1 (NFAT), and NF1X. A nuclear factor-κB (NF-κB) binding site that includes the NFAT site has been shown to be active for late transcription and possible positioning of other DNA-binding proteins to initiate efficient transcription.⁸⁶ NFAT consists of a family of proteins that are expressed in many cells but have multiple classes with tissue-specific expression. The NF-1 class X (NFIX) is highly expressed in human glial cells, stromal cells, B lymphocytes, and CD34+ stem cells that have all been reported to support JCPyV transcription and replication.⁷⁹ Interactions between YB1 and PURA may also provide cell-specific regulation.⁸⁶

The transcription factors that govern kidney tissue-specific activity of the BKPyV early promoter are not known. While numerous candidate transcription factor binding sites have been identified,¹⁸⁵ it remains unclear which of those factors bind the BKPyV regulatory region in an infected kidney cell.

Rearrangements in the NCCR have also been found in BKPyV and JCPyV.³⁰,¹²⁴,²¹⁴ Viral regulatory regions referred to as archetype are thought to be associated with naturally circulating polyomaviruses, while rearranged regulatory regions arise when high levels of viral replication occur in culture or during disease. For example, JCPyV early promoter undergoes duplication in virus isolated from patients with PML. These rearrangements may serve to allow higher levels of large T antigen expression and viral replication that are tolerated by the patient’s immunocompromised state.⁹⁸ Notably, the δSV40/JCV hybrid virus contains JCPyV T antigen and VP1, VP2, and VP3 coding sequences but a hybrid regulatory region with elements of the SV40 enhancer. This laboratory-generated hybrid virus grows with faster kinetics and to higher titers than wild-type JCPyV and has an expanded host range to human and monkey kidney cells as well as monkey glial cells.²⁶⁷

The SV40 early promoter undergoes negative feedback regulation by large T antigen.²⁵⁹ When cells are infected with a temperature-sensitive large T antigen mutant of SV40 at the nonpermissive or restrictive temperature, early mRNAs are overproduced. The ability to repress early transcription is dependent on the binding of large T antigen to the promoter. Mutations that disrupt the DNA-binding domain of large T antigen also typically lead to higher levels of large T antigen.¹³³ Three binding sites, referred to as sites I, II, and III, in the regulatory region are all involved in repression by large T antigen. Because site III overlaps the early promoter elements, it is possible that large T antigen binding acts to prevent binding or displace other transcription factors to that region. Early transcription of MPyV is also regulated by large T antigen, but to a lesser degree and in a DNA binding site–independent manner.

Early gene expression in SV40 is down-regulated by miRNAs encoded by the late transcript.²⁵³ The other primate polyomaviruses have similar sequences that are predicted to be able to form the characteristic hairpin structure of miRNA.²²⁸ Somewhat surprisingly, an SV40 mutant virus lacking the miRNA does not produce more virus than wild type in cultured monkey kidney cells. However, cells infected with the mutant virus are more sensitive to killing by cytotoxic T lymphocytes in vitro. This observation led to the proposal that the miRNA serves to limit production of antigens recognized by the immune system and thereby protecting the infected cell.²⁵³ A mutant MPyV that cannot express its miRNA shows no difference in pathogenesis in animals.²⁵⁴ Therefore, the role of the miRNAs during infection remains to be fully elucidated.

Similar to the early promoter, the late promoter elements span the regulatory region and have been defined in a variety of in vivo and in vitro systems. In SV40, maximal late transcription requires sequences in the 21–base-pair GC repeats and the 72–base-pair enhancers. Late gene transcription occurs concomitantly with the onset of DNA replication, although replication is not required for activation of late transcription. Large T antigen can promote viral replication as well as late gene expression. Large T antigen can stimulate late transcription, although large T antigen binding the origin of replication is not required for this activity. Rather, transcription activation is accomplished through large T antigen interactions with components of the basal transcription machinery such as TATA-binding protein, a component of TFIID and TBP-associated factor 1 (TAF1), as well as transcription factors late SV40 factor (LSF, TFCP2), TEF-1 (TEAD1), and SP1. These interactions increase the binding of TBP and another basal transcription factor, TFIIA, to the TATA element.⁵³

The control of late gene expression in MPyV is more complicated. Host range or hr-t mutants that do not express functional small or middle T antigen produce equivalent amounts of late proteins, as does wild-type virus.⁸⁷ However, infection with a virus containing similar mutations in cells with a different genetic background demonstrated a stimulatory role of the two T antigens during the late phase. This does not appear to be solely due to an indirect effect because of amplification of the genome, but also involves a direct stimulation of transcription because the RNA/DNA ratio increases.⁴⁰ These same studies also indicated a role for middle and small T antigens in the induction of early gene expression. As these two T antigens induce signaling pathways that lead to activation of transcription factors known to bind the polyomavirus enhancer, this effect is not surprising. The murine virus also differs from its primate counterparts in how its late primary RNA transcript is processed. The MPyV late poly(A) site is a relatively weak site, resulting in the RNA polymerase circling the viral genome multiple times.¹,¹⁶¹ The first exon of the late transcript can therefore be spliced to itself multiple times, although the protein-coding sequence is only present once on each mature mRNA.