JC Polyomavirus Entry by Clathrin-Mediated Endocytosis Is Driven by β-Arrestin

Viruses usurp cellular factors to invade host cells. Activation and utilization of these proteins upon initiation of viral infection are therefore required for productive infection and resultant viral disease. The majority of healthy individuals are asymptomatically infected by JC polyomavirus (JCPyV), but if the host immune system is compromised, JCPyV can cause progressive multifocal leukoencephalopathy (PML), a rare, fatal, demyelinating disease. Individuals infected with HIV or taking prolonged immunomodulatory therapies have a heightened risk for developing PML. The cellular proteins and pathways utilized by JCPyV to mediate viral entry are poorly understood. Our findings further characterize how JCPyV utilizes the clathrin-mediated endocytosis pathway to invade host cells. We have identified specific components of this pathway that are necessary for the viral entry process and infection. Collectively, the conclusions increase our understanding of JCPyV infection and pathogenesis and may contribute to the future development of novel therapeutic strategies for PML.

ABSTRACT

JC polyomavirus (JCPyV) establishes a persistent, lifelong, asymptomatic infection within the kidney of the majority of the human population. Under conditions of severe immunosuppression or immune modulation, JCPyV can reactivate in the central nervous system (CNS) and cause progressive multifocal leukoencephalopathy (PML), a fatal demyelinating disease. Initiation of infection is mediated through viral attachment to α2,6-sialic acid-containing lactoseries tetrasaccharide c (LSTc) on the surface of host cells. JCPyV internalization is dependent on serotonin 5-hydroxytryptamine subfamily 2 receptors (5-HT₂Rs), and entry is thought to occur by clathrin-mediated endocytosis (CME). However, the JCPyV entry process and the cellular factors involved in viral internalization remain poorly understood. Treatment of cells with small-molecule chemical inhibitors and RNA interference of 5-HT₂R endocytic machinery, including β-arrestin, clathrin, AP2, and dynamin, significantly reduced JCPyV infection. However, infectivity of the polyomavirus simian virus 40 (SV40) was not affected by CME-specific treatments. Inhibition of clathrin or β-arrestin specifically reduced JCPyV internalization but did not affect viral attachment. Furthermore, mutagenesis of a β-arrestin binding domain (Ala-Ser-Lys) within the intracellular C terminus of 5-HT_2AR severely diminished internalization and infection, suggesting that β-arrestin interactions with 5-HT_2AR are critical for JCPyV infection and entry. These conclusions illuminate key host factors that regulate clathrin-mediated endocytosis of JCPyV, which is necessary for viral internalization and productive infection.

IMPORTANCE Viruses usurp cellular factors to invade host cells. Activation and utilization of these proteins upon initiation of viral infection are therefore required for productive infection and resultant viral disease. The majority of healthy individuals are asymptomatically infected by JC polyomavirus (JCPyV), but if the host immune system is compromised, JCPyV can cause progressive multifocal leukoencephalopathy (PML), a rare, fatal, demyelinating disease. Individuals infected with HIV or taking prolonged immunomodulatory therapies have a heightened risk for developing PML. The cellular proteins and pathways utilized by JCPyV to mediate viral entry are poorly understood. Our findings further characterize how JCPyV utilizes the clathrin-mediated endocytosis pathway to invade host cells. We have identified specific components of this pathway that are necessary for the viral entry process and infection. Collectively, the conclusions increase our understanding of JCPyV infection and pathogenesis and may contribute to the future development of novel therapeutic strategies for PML.

INTRODUCTION

JC polyomavirus (JCPyV) is the etiological agent of the fatal demyelinating disease progressive multifocal leukoencephalopathy (PML) (1, 2). Although PML is rare, ∼40% to 75% of the human population is thought to be seropositive for JCPyV (3, 4), which establishes a lifelong, asymptomatic infection in the kidney and B lymphocytes of healthy individuals (5–8). Under conditions of severe immunosuppression or immune modulation, JCPyV can become reactivated, leading to infection in the central nervous system (CNS) (7, 9, 10). In the CNS, JCPyV infects glial cells, astrocytes, and oligodendrocytes, resulting in severe demyelination and PML (11–15). PML occurs in HIV-1-positive individuals (2, 16) and those undergoing prolonged immunomodulatory therapies for immune-mediated diseases (17–19), such as those with multiple sclerosis (MS) taking natalizumab (18, 20). Progression of PML is rapid and can result in fatality within months if left untreated (21). Due to a lack of effective therapies, PML treatment typically involves addressing the underlying immunosuppression, which can improve life expectancy but does not cure PML (16, 22–24). Thus, understanding the mechanisms underlying JCPyV infection is critical to the development of therapies for PML.

JCPyV is a member of the Polyomaviridae family (25), which also includes simian virus 40 (SV40) and BK polyomavirus (BKPyV). Polyomaviruses are nonenveloped double-stranded DNA (dsDNA) viruses comprised of three viral structural proteins, viral protein 1 (VP1), VP2, and VP3 (26, 27). VP1 is expressed on the exterior of the capsid and serves as the major viral attachment protein (28). JCPyV attachment to host cells is mediated by a direct interaction between VP1 and α2,6-glycan lactoseries tetrasaccharide c (LSTc) (28–31). The serotonin 5-hydroxytryptamine (5-HT) receptors of subfamily 2 (5-HT_2AR, 5-HT_2BR, and 5-HT_2CR) are required for JCPyV entry and infection (32, 33), yet the mechanism by which they mediate entry is not understood. Serotonin receptors are G-protein-coupled receptors (GPCRs) that comprise a family of 15 members (34) and are widely expressed within the CNS (35, 36), associated with regulation of mood and psychiatric function (34, 37). Interestingly, the expression of 5-HT_2A, 5-HT_2B, and 5-HT_2C receptors is consistent with sites of JCPyV infection (38–40). While 5-HT₂R expression enhances JCPyV entry, the mechanism of viral internalization remains poorly understood (33). JCPyV internalization into cells has been reported to be mediated by clathrin-mediated endocytosis (CME), as viral infection is sensitive to chlorpromazine, a chemical inhibitor of CME (33, 41). Additionally, expression of dominant negative mutants of epidermal growth factor receptor (EGFR) substrate kinase subclone 15 (eps15), a clathrin adaptor protein (42), impairs JCPyV infection (43). However, the clathrin inhibitor chlorpromazine is also a 5-HT₂R antagonist (44), and thus, the role of CME in JCPyV internalization remains poorly characterized.

Viruses use multiple strategies to internalize into host cells, including clathrin- and caveolin-mediated endocytosis, macropinocytosis, and nonclathrin and noncaveolin mechanisms (45, 46). Of these, CME is the most common viral entry mechanism, serving as an enticing pathway for both enveloped and nonenveloped viruses to gain entry into host cells (47, 48). While JCPyV has been suggested to utilize CME, other polyomaviruses, including SV40 and BKPyV, those most closely related to JCPyV, enter cells via caveola-mediated or non-clathrin-, non-caveola-mediated uptake mechanisms (49–53). SV40 enters cells via a non-clathrin-dependent mechanism (52), including caveola-dependent and -independent lipid-mediated endocytic mechanisms (49, 50). BKPyV entry also requires caveolae, and entry is independent of clathrin (51, 54, 55). Regardless of the mechanism utilized for internalization, following entry events, all polyomaviruses ultimately traffic to the endoplasmic reticulum (ER) for partial uncoating prior to retrotranslocation to the nucleus (56–61).

CME involves direct interactions between clathrin and other scaffolding proteins, like adaptor protein 2 (AP2), to coordinate internalization (62, 63). AP2 is comprised of four subunits (α2, β2, σ2, and μ2) and serves multiple roles in CME, including assisting in the assembly of clathrin-coated pits, providing a proteinaceous link between internalizing cargo and clathrin, and coordinating interactions between clathrin and other CME proteins (42, 64–67). Following the activation of and interaction between these proteins, a clathrin-coated pit is formed, which becomes pinched off through GTPase activity of dynamin into an endocytic vesicle (68, 69). The endocytic vesicle is then free to traffic cargo, including viruses, to the appropriate intracellular compartment (70).

GPCRs, including 5-HT₂Rs, can be internalized through clathrin-mediated endocytosis in a process involving specific scaffolding proteins, leading to receptor recycling, receptor degradation, or activation of downstream signaling cascades (71). 5-HT₂R-mediated CME requires β-arrestins, which bind to cytoplasmic domains within the receptor (72–75) and drive receptor internalization (76, 77). β-Arrestin binding to intracellular domains prevents G-protein coupling and directs the receptor to the clathrin-coated pit with the assistance of endocytic proteins, including clathrin and AP2 (72, 78, 79). 5-HT₂Rs have a conserved tripeptide motif, Ala-Ser-Lys (ASK), that serves as a β-arrestin binding domain in the C terminus, which has been demonstrated to modulate internalization of 5-HT_2AR (80). In addition to receptor interactions, β-arrestins can bind to endocytic scaffolding proteins, including clathrin (81, 82) and the β2 subunit of AP2 (83), and modulate their activity. Therefore, β-arrestins are critical for mediating GPCR signaling events and determine the outcome of agonist-mediated receptor endocytosis (71).

The focus of this study was to characterize the host cell proteins required for JCPyV internalization and to further define the role of 5-HT₂Rs in JCPyV internalization and infection. We have further defined the cell entry pathway by identifying specific endocytic proteins required for JCPyV entry and infection. Collectively, these results elucidate the internalization strategy usurped by JCPyV to infect host cells and highlight the role of 5-HT₂R scaffolding proteins in viral internalization strategies.

RESULTS

Clathrin is required for JCPyV infection.

To define whether clathrin is required for JCPyV infection, SVG-A cells, a model human glial cell line, were treated with the clathrin inhibitor Pitstop 2, which specifically inhibits the clathrin terminal domain (84). Cells were pretreated and infected with JCPyV in the presence of Pitstop 2 and fed with medium containing JCPyV-specific antiserum at 1 h postinfection (hpi) to neutralize noninternalized virus. Infectivity was scored by a fluorescence focus assay (FFA) at 72 hpi to measure newly synthesized VP1 following a single replication cycle. Pitstop 2 treatment resulted in a dose-dependent decrease in JCPyV infection (Fig. 1A). Additionally, to determine whether Pitstop 2 was impacting JCPyV infection at a time consistent with JCPyV entry, Pitstop 2 was added at specific times based on the reported kinetics of JCPyV internalization (33, 61). SVG-A cells were either pretreated for 15 min with Pitstop 2, treated at the time of infection, or incubated with Pitstop 2 at 4 hpi, when the majority of virus should be internalized (33, 61) (Fig. 1B). Additionally, JCPyV-specific antiserum was added to the medium at 4 hpi. Pitstop 2 reduced infection when cells were pretreated or when the inhibitor was included in the viral inoculum (Fig. 1B). There was no impact on JCPyV infection when Pitstop 2 was added at 4 hpi, suggesting that clathrin inhibition specifically reduces JCPyV infection at times consistent with viral attachment and entry. To further confirm the role of clathrin in JCPyV infection, SVG-A cells were treated with a small interfering RNA (siRNA) targeting clathrin heavy chain (CHC), and knockdown was measured by Western blot (WB) analysis (Fig. 1D). Additionally, CHC siRNA-treated cells were infected with either JCPyV or SV40, at multiplicities of infection (MOIs; indicated in the figure legends) resulting in comparable levels of infectivity (41, 85, 86). Viral infectivity was scored by indirect immunofluorescence at 72 hpi, following a single round of infection, for newly synthesized VP1. CHC knockdown resulted in a significant reduction in JCPyV infection in comparison to EGFR and scrambled control siRNAs (Fig. 1C). Conversely, infectivity of SV40, which utilizes a clathrin-independent internalization pathway (49, 50, 52), was not impacted by CHC knockdown (Fig. 1C). To determine if JCPyV colocalizes with clathrin at times consistent with viral internalization, SVG-A cells were transfected with plasmids to express either emerald-clathrin (em-clath) or emerald-caveolin (em-cav) (both derivatives of green fluorescent protein [GFP]) and infected with JCPyV labeled with Alexa Fluor 647 (JCPyV-647) on ice for 1 h to allow for synchronized viral attachment. Cells were then fixed or shifted to 37°C for 5 min for viral internalization prior to fixation (Fig. 1E). Samples were imaged by confocal microscopy (Fig. 1E), and fields of view were analyzed for colocalization between em-clath or em-cav and JCPyV-647. Virus particles colocalized with em-clath at 5 min postinternalization yet did not demonstrate significant colocalization with em-cav (Fig. 1F). These data suggest that clathrin is critical for JCPyV infection and that JCPyV colocalizes with clathrin at times consistent with viral internalization.

FIG 1.

Clathrin inhibition decreases JCPyV infection. (A) SVG-A cells were pretreated with the clathrin inhibitor Pitstop 2 or a DMSO control at the indicated concentrations at 37°C for 15 min prior to infection with JCPyV, and at 1 hpi, cells were fed with medium containing JCPyV-specific neutralizing antibody (JC-specific Ab). (B) A total of 20 μM Pitstop 2 or the DMSO control was added at the times indicated prior to or following infection with JCPyV, and at 4 hpi cells were fed with medium containing JC-specific Ab. (C) SVG-A cells were transfected with siRNA targeting clathrin heavy chain (CHC), the EGFR control, or a nontargeting scrambled (Scr) control. Cells were infected with JCPyV (MOI = 0.5 FFU/cell) (A to C) or SV40 (MOI = 0.001 FFU/cell) (C). (D) Cellular lysates from panel C were processed for Western blotting to confirm protein knockdown of CHC. Infected cells were fixed at 72 hpi and analyzed by indirect immunofluorescence utilizing an Ab that detects both JCPyV and SV40 VP1. Data represent the percentages of JCPyV- or SV40-infected VP1⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples (values for all samples were normalized to that for EGFR control siRNA-treated cells [100%]) from at least three independent experiments. Error bars indicate standard deviations (SD). (E and F) SVG-A cells were transfected with either emerald-clathrin (em-clath) or emerald-caveolin (em-cav) (green), prechilled, and then incubated with JCPyV-647 (MOI = 5 FFU/cell) (pseudocolored magenta) on ice for 1 h (viral attachment), followed by incubation at 37°C for viral entry. (E) Cells were fixed at the times indicated, nuclei were stained with DAPI (blue), and 10 fields of view per sample were measured by confocal microscopy at a ×60 magnification. (F) Colocalization was analyzed using ImageJ software by measuring the percent correlation of em-clath or em-cav and JCPyV-647 using Mander’s coefficient. Data represent the percentages of em-clath and em-cav colocalized with JCPyV/visual field for at least 10 fields of view for triplicate samples. Data are depicted in a box-and-whisker plot denoting the median and the distribution of percent colocalization for both samples and are representative of results from at least 3 independent experiments. Upper and lower whiskers are 1.5 times the interquartile range. Data points in gray (+) are outliers. Bars = 5 μm. Arrows indicate sites of colocalization. *, P < 0.05.

JCPyV infection requires adaptor protein AP2 subunits β2 and μ2.

AP2 is a key regulator in clathrin-coated pit assembly and clathrin-mediated endocytosis (65–67) and is required for internalization of 5-HT₂Rs (72). AP2 subunit μ2 interacts with the internalizing receptor and cargo (87, 88), while subunit β2 interacts with other endocytic proteins, such as clathrin and β-arrestin (89–92). To determine if AP2 is required for JCPyV infection, SVG-A cells were treated with siRNA targeting subunits AP2 μ2 and β2, and knockdown was measured by WB analysis (Fig. 2B). Following siRNA treatment, cells were infected with JCPyV (Fig. 2A). Knockdown of μ2 resulted in a slight though significant reduction in JCPyV infection, while knockdown of β2 significantly reduced infection in comparison to EGFR and scrambled control siRNAs (Fig. 2A). Collectively, these data demonstrate a requirement for AP2 subunits μ2 and, predominantly, β2 for JCPyV infection.

FIG 2.

JCPyV infection requires adaptor protein AP2 subunits μ2 and β2. SVG-A cells transfected with siRNA targeting the EGFR control, the nontargeting scrambled (Scr) control, or AP2 subunit μ2 or β2 were infected with JCPyV (MOI = 0.5 FFU/cell) (A), or cellular lysates were analyzed by Western blotting to confirm protein knockdown of AP2 μ2 and AP2 β2 (B). Infected cells were fixed at 72 hpi, stained using a VP1-specific antibody, and analyzed by indirect immunofluorescence. Data represent the percentages of VP1⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples (values for all samples were normalized to that for EGFR control siRNA-treated cells [100%]) from at least three independent experiments. Error bars indicate SD. *, P < 0.05.

Dynamin is required for JCPyV infection.

Clathrin-coated pits are pinched off into vesicles by the molecular GTPase dynamin (68, 69). To test the role of dynamins in JCPyV infection, SVG-A cells were pretreated for 30 min with increasing concentrations of Dynole 34-2, an inhibitor of dynamin isoforms 1 and 2 (93), which are expressed within the brain and function in CME (94). Cells were subsequently infected with JCPyV in the presence of Dynole, resulting in a dose-dependent decrease in JCPyV infection (Fig. 3A). To further explore the role of dynamin in infection, dynamin isoforms 1 and 2 were individually knocked down in SVG-A cells by siRNA interference (Fig. 3C). Following siRNA treatment, cells were infected with JCPyV, and infectivity was measured (Fig. 3B). Knockdown of dynamin 1 resulted in a significant decrease in infection, while knockdown of dynamin 2 led to a lesser although significant reduction in infectivity (Fig. 3B). Collectively, these data demonstrate the requirement of dynamins, mainly dynamin 1, for JCPyV infection.

FIG 3.

Dynamin is required for JCPyV infection. (A) SVG-A cells were pretreated with the dynamin 1/2 inhibitor Dynole at the indicated concentrations or with the Dynole-specific negative control at 37°C for 30 min before infection with JCPyV (MOI = 0.5 FFU/cell) in the presence of Dynole. (B) SVG-A cells transfected with siRNA targeting the EGFR control, the nontargeting scrambled (Scr) control, dynamin 1 (Dyn1), or dynamin 2 (Dyn2) were infected with JCPyV (MOI = 0.5 FFU/cell). Infected cells were fixed at 72 hpi, stained using a VP1-specific antibody, and analyzed by indirect immunofluorescence. (C) Protein knockdown of Dyn1 and Dyn2 was confirmed by Western blotting. Data represent the percentages of VP1⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples (values for all samples were normalized to that for EGFR control siRNA-treated cells [100%]) from at least three independent experiments. Error bars indicate SD. *, P < 0.05.

Knockdown of β-arrestin significantly reduces JCPyV infection in glial cells.

β-Arrestin acts as an adaptor protein and assists in directing internalizing GPCRs into clathrin-coated pits through direct interactions with clathrin and AP2 (78, 79). While there are reported species-specific, cell type-specific, and ligand-specific differences in β-arrestin-mediated internalization (95–97), human 5-HT_2AR has been reported to be β-arrestin dependent (80), suggesting that β-arrestin may play a role in JCPyV infection. SVG-A cells were transfected with siRNA targeting β-arrestin (Fig. 4B) and infected with JCPyV or SV40, and infectivity was measured. β-Arrestin knockdown resulted in a significant reduction in JCPyV infection while not impacting SV40 infection (Fig. 4A). These data suggest that JCPyV infection requires the activity of the 5-HT₂R scaffolding protein β-arrestin.

FIG 4.

JCPyV infection requires β-arrestin. SVG-A cells transfected with siRNA targeting β-arrestin (β-arr), EGFR control, or nontargeting scrambled (Scr) control were infected with JCPyV (MOI = 0.5 FFU/cell) or SV40 (MOI = 0.001 FFU/cell) (A) or processed for Western blotting to confirm β-arrestin knockdown (B). Infected cells were fixed at 72 hpi, stained using a VP1-specific antibody, and analyzed by indirect immunofluorescence. Data represent the percentages of VP1⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples (values for all samples were normalized to that for EGFR control siRNA-treated cells [100%]) from at least three independent experiments. Error bars indicate SD. *, P < 0.05.

Reduction of clathrin heavy chain or β-arrestin does not affect JCPyV attachment but specifically reduces internalization.

Reduction of CHC and β-arrestin, key proteins in the CME pathway, significantly reduced JCPyV infection (Fig. 1 and 4), and JCPyV was found to colocalize with clathrin during internalization (Fig. 1). To determine whether knockdown of CHC and β-arrestin specifically reduces JCPyV internalization, viral attachment and entry were measured following siRNA treatment (Fig. 5). SVG-A cells were treated with siRNAs targeting CHC, β-arrestin, or a nonrelevant control, EGFR, and subsequently incubated with Alexa Fluor 488-labeled JCPyV (JCPyV-488) virions on ice to allow for viral attachment. Attachment of JCPyV-488 to siRNA-treated cells was measured by flow cytometry (Fig. 5A and B). Analysis of mean fluorescence intensities indicated that binding of JCPyV to SVG-A cells is not reduced by siRNA silencing of CHC (Fig. 5A) or β-arrestin (Fig. 5B). To measure JCPyV entry, SVG-A cells were treated with CHC, β-arrestin, or EGFR siRNAs and incubated on ice at 4°C with JCPyV-488 for synchronized viral attachment. Cells were then incubated at 37°C for 1.5 h to allow for viral entry (61). Following fixation, JCPyV internalization was measured by confocal microscopy (Fig. 5C and D, right panels) by quantifying the relative fluorescence intensity for internalized JCPyV-488 within individual cells, excluding the plasma membrane. Treatment of cells with either CHC (Fig. 5C) or β-arrestin (Fig. 5D) siRNAs resulted in a significant reduction in JCPyV internalization in comparison to EGFR control siRNA-treated samples. Taken together, these results suggest that reduction of β-arrestin or CHC does not impact viral attachment to cells but specifically reduces JCPyV internalization.

FIG 5.

Reduction of clathrin heavy chain or β-arrestin prevents JCPyV entry. (A and B) SVG-A cells were transfected with EGFR control siRNA or siRNA targeting CHC (A) or β-arrestin (β-arr) (B). Cells were stripped from plates, prechilled at 4°C for 45 min, incubated with JCPyV-488 (MOI = 5 FFU/cell) on ice for 1 h, and analyzed by flow cytometry. Histograms represent the mean fluorescence intensities for cells transfected with CHC siRNA, β-arrestin siRNA, or EGFR siRNA or cells alone (gray). Data are representative of 10,000 events for triplicate samples from at least 3 independent experiments. (C and D) To measure viral entry, SVG-A cells were transfected with siRNAs for CHC (C) or β-arrestin (D) or EGFR control siRNA, incubated at 37°C for 72 h, prechilled, and then incubated with JCPyV-488 (MOI = 5 FFU/cell) (green) on ice for 1 h (viral attachment), followed by incubation at 37°C for 1.5 h (viral entry). Cells were fixed, nuclei were stained with DAPI (blue), and viral entry was analyzed by confocal microscopy at a ×60 magnification (right panels). Utilizing DIC overlay, JCPyV internalization was quantified as the relative fluorescence per cell for at least 30 cells for CHC, β-arrestin, and EGFR siRNA-treated samples using Olympus FluoView 10-ASW software. Data are depicted in box-and-whisker plots denoting medians and the distributions of percent internalization for both samples and are representative of results from at least three independent experiments. Values for all samples were normalized to that for EGFR-siRNA-treated cells (100%). Upper and lower whiskers are 1.5 times the interquartile range. Bars = 5 μm. *, P < 0.0001.

Clathrin, AP2, and β-arrestin are required for JCPyV infection of kidney cells.

JCPyV infects both kidney and glial cells yet demonstrates poor permissivity in established kidney cell lines (33). Human embryonic kidney (HEK293A) cells express the JCPyV attachment factor LSTc but lack sufficient expression of 5-HT₂Rs to support JCPyV infection and are poorly permissive for infection unless 5-HT₂Rs are expressed (33). HEK293A cells stably expressing 5-HT_2A, 5-HT_2B, and 5-HT_2C receptors support JCPyV infection and demonstrate a specific increase in viral entry in comparison to control HEK293A cells (33). To determine whether JCPyV infection of kidney cells is dependent upon clathrin-mediated endocytosis, 5-HT_2AR-, 5-HT_2BR-, and 5-HT_2CR-expressing HEK293A cells (33) were treated with siRNA targeting CHC. Protein knockdown was determined by WB analysis (Fig. 6A). Following siRNA treatment, cells were infected with JCPyV, and infectivity was measured at 48 hpi for newly synthesized T antigen (T-Ag). Reduction of CHC in 5-HT_2AR-, 5-HT_2BR-, and 5-HT_2CR-expressing HEK293A cells resulted in significantly reduced JCPyV infection (Fig. 6A). To further determine the CME components necessary for JCPyV infectivity of kidney cells, 5-HT_2AR-, 5-HT_2BR-, and 5-HT_2CR-expressing HEK293A cells were treated with siRNAs targeting AP2 subunit β2 or β-arrestin. Protein reduction was determined by WB analysis (Fig. 6B and C). Following siRNA treatments, cells were infected with JCPyV, and infectivity was measured at 48 hpi. Reduction of AP2 β2 in 5-HT_2AR-, 5-HT_2BR-, and 5-HT_2CR-expressing HEK293A cells significantly reduced JCPyV infectivity (Fig. 6B). Moreover, treatment with a β-arrestin siRNA resulted in significant reductions in JCPyV infection (Fig. 6C) while not impacting infection by SV40 (Fig. 6C). These data suggest that clathrin, AP2, and β-arrestin are required for JCPyV infection of 5-HT₂R-expressing HEK293A cells (Fig. 6).

FIG 6.

Reduction of clathrin, AP2 β2, and β-arrestin in HEK-5-HT_2AR-YFP, HEK-5HT_2BR-YFP, and HEK-5HT_2CR-YFP cells significantly reduces infection. HEK293A cells stably expressing 5-HT_2AR–YFP, 5-HT_2BR–YFP, and 5-HT_2CR–YFP were transfected with siRNA targeting CHC (A), AP2 β2 (B), or β-arrestin (β-arr) (C); the EGFR control; and the nontargeting scrambled (Scr) control. Cells were infected with JCPyV (MOI = 0.5 FFU/cell) (A to C) or SV40 (MOI = 0.001 FFU/cell) (C) or processed for Western blotting to confirm CHC, AP2 β2, and β-arrestin knockdown. Infected cells were fixed at 48 hpi, stained using a JCPyV or SV40 T-antigen (T-Ag)-specific antibody, and analyzed by indirect immunofluorescence. Data represent the percentages of JCPyV- or SV40-infected T-Ag⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples (values for all samples were normalized to that for EGFR control siRNA-treated cells for each receptor subtype cell line [100%]) from at least three independent experiments. Error bars indicate SD. *, P < 0.05.

Site-directed mutagenesis of a conserved β-arrestin binding domain within 5-HT_2AR reduces JCPyV infection.

The human 5-HT₂R subtypes contain conserved domains within the intracellular loops that serve as docking sites for endocytic scaffolding machinery that can promote receptor internalization (80, 98). The ASK motif, conserved in human 5-HT₂Rs, is a reported direct interaction site for β-arrestin (80). The role of this protein binding domain has been extensively characterized for internalization and trafficking of 5-HT_2AR; expression of the ASK motif confers β-arrestin-mediated 5-HT_2AR internalization (80) (depicted in Fig. 7A). Furthermore, β-arrestin scaffolds 5-HT_2AR to clathrin and AP2 to mediate CME (99). To determine the role of the ASK motif in JCPyV infection, single- and double-amino-acid point mutations were engineered in the 5-HT_2A–yellow fluorescent protein (YFP) receptor, in which the serine and lysine at amino acid positions 457 and 458 were replaced with alanine residues (Fig. 7A). To ensure accessibility of the mutated 5-HT_2ARs for viral utilization, cell surface expression of wild-type (WT) and mutated receptors was measured by confocal microscopy. The cell surface was defined by staining the plasma membrane with an antibody (Ab) specific for pan-cadherin, and colocalization of wild-type and mutated 5-HT_2ARs with pan-cadherin was measured. Mutagenesis of 5-HT_2A–YFP receptors did not reduce the surface expression of the receptor in comparison to wild-type cell surface expression (Fig. 7B). Wild-type and mutated 5-HT_2A–YFP receptors were expressed in HEK293A cells and infected with JCPyV or SV40, and viral infectivity was measured by expression of the newly synthesized viral protein T-Ag by indirect immunofluorescence at 48 hpi (Fig. 7C). Cells expressing 5-HT_2AR-S457A (AAK) and 5-HT_2AR-SK457-58AA (AAA) demonstrated significant reductions in JCPyV infection in comparison to wild-type 5-HT_2AR–YFP (Fig. 7C). However, 5-HT_2AR-K458A (ASA) supported JCPyV infection to the level of wild-type 5-HT_2AR–YFP (Fig. 7C). Furthermore, mutation of residues in the 5-HT_2AR ASK motif had no significant effect on SV40 infection of HEK293A cells (Fig. 7D). Collectively, these data suggest that the β-arrestin binding domain in 5-HT_2AR is essential for JCPyV infection and specifically that the serine residue within the ASK motif is critical for JCPyV infection in HEK293A cells.

FIG 7.

The β-arrestin binding motif in 5-HT_2AR is required for JCPyV infection of HEK293A cells. (A) Schematic demonstrating the 5-HT_2AR ASK motif. HEK293A cells were transfected with wild-type unmutated 5-HT_2AR–YFP (WT), 5-HT_2AR–YFP receptors containing point mutations in the ASK motif, or the empty YFP vector. (B) Cell surface expression of WT and mutated receptors was measured by confocal microscopy at a ×60 magnification (right). Cell surface expression was analyzed using ImageJ software by measuring the percent correlation of receptor expression (green) and a pan-cadherin cell surface marker (red) using Mander’s coefficient. Data represent the percentages of the receptor that are cell surface expressed/visual field normalized to the value for the WT for at least 10 fields of view from triplicate samples. (C and D) Following transfection, cells were infected with JCPyV (MOI = 0.5 FFU/cell) (C) or SV40 (MOI = 0.001 FFU/cell) (D), fixed at 48 hpi, stained with a JCPyV or SV40 T-Ag-specific antibody, and analyzed by indirect immunofluorescence. Data represent the percentages of JCPyV-infected T-Ag⁺ cells/visual field normalized to the number of DAPI⁺ cells/field for five ×10 fields of view for triplicate samples from at least three independent experiments. Bars = 5 μm. Error bars indicate SD. *, P < 0.05.

Mutagenesis of the ASK motif in 5-HT_2AR reduces internalization of JCPyV.

To determine whether the ASK motif is required for JCPyV internalization, viral attachment and entry were measured. HEK293A cells transiently expressing wild-type and mutated 5-HT_2A–YFP receptors were removed from plates and incubated with JCPyV-647 virions on ice. Attachment of JCPyV-647 to cells expressing wild-type and mutated 5-HT_2ARs was measured by flow cytometry (Fig. 8A). Mean fluorescence intensities of JCPyV-647-bound virus were equivalent in cells expressing wild-type and mutated 5-HT_2A–YFP receptors, suggesting that mutagenesis of the ASK motif does not reduce JCPyV binding to cells (Fig. 8A). Furthermore, expression of wild-type 5-HT_2AR did not enhance JCPyV binding in comparison to untransfected cells, in correlation with previously reported results (33). To measure viral entry, HEK293A cells expressing wild-type and mutated 5-HT_2A–YFP receptors were incubated on ice with JCPyV-647 for synchronized viral attachment, followed by incubation at 37°C for 1.5 h for viral entry. Following fixation, viral internalization was measured by confocal microscopy by quantifying the relative fluorescence intensity for internalized JCPyV-647 within individual cells, excluding the plasma membrane (Fig. 8C). Cells expressing 5-HT_2AR-S457A (AAK) and 5-HT_2AR-SK457-58AA (AAA) demonstrated a significant reduction in JCPyV internalization, while 5-HT_2AR-K458A (ASA) supported levels of JCPyV internalization comparable to the wild-type 5-HT_2A–YFP receptor (Fig. 8B). Collectively, these data demonstrate that the serine residue within the ASK motif is critical for JCPyV entry into host cells.

FIG 8.

Mutagenesis of the β-arrestin binding motif in 5-HT_2AR decreases JCPyV internalization in HEK293A cells. HEK293A cells were transfected with wild-type unmutated 5-HT_2AR–YFP (WT); the mutated receptor 5-HT_2A-S457A-R–YFP (AAK), 5-HT_2A-SK457-58AA-R–YFP (AAA), or 5-HT_2A-K458A-R–YFP (ASA); or the empty YFP vector. (A) Cells were stripped from plates; prechilled at 4°C, followed by incubation with JCPyV-647 (MOI = 5 FFU/cell) on ice for 1 h for viral attachment; and then analyzed by flow cytometry. Histograms represent the mean fluorescence intensities for cells (no virus) (gray), YFP (no virus), or samples treated with JCPyV-647. Data are representative of 10,000 events for triplicate samples from at least 3 independent experiments. (B and C) For viral entry, HEK293A cells were transfected with the WT, AAK, AAA, or ASA. Cells were chilled to 4°C and incubated with JCPyV-647 (MOI = 3 FFU/cell) (pseudocolored magenta) on ice for 1 h (viral attachment), followed by incubation at 37°C for 1.5 h (viral entry). Cells were fixed, nuclei were stained with DAPI (blue), and entry was analyzed by confocal microscopy at a ×60 magnification. Utilizing DIC overlay, JCPyV internalization was quantified as the relative fluorescence within individual cells for at least 30 cells for 5-HT_2AR WT, AAK-, AAA-, and ASA-transfected samples using Olympus FluoView 10-ASW software. Data are depicted in a box-and-whisker plot denoting the medians and the distributions of percent internalization across samples measured and are representative of results from at least three independent experiments. Values for all samples were normalized to the value for WT-transfected cells (100%). Upper and lower whiskers are 1.5 times the interquartile range. Data points in gray (+) indicate outliers. Bars = 5 μm. *, P < 0.0001; NS, not significant.

DISCUSSION

Viral entry is a complex process, requiring numerous host cell factors, including cell surface receptors and intracellular proteins, to facilitate internalization (46). In this study, we have further characterized the endocytic proteins in the clathrin-mediated pathway that facilitate JCPyV internalization. Disruption of clathrin (Fig. 1) and the CME accessory proteins β-arrestin (Fig. 4), AP2 (Fig. 2), and dynamin (Fig. 3) significantly reduced JCPyV infection. Furthermore, disruption of CHC and β-arrestin specifically reduced JCPyV entry into host cells while not impacting viral attachment (Fig. 5). Interestingly, the chemical inhibitors and siRNAs utilized did not impact infection by SV40, which is known to enter by clathrin-independent pathways (49, 50, 52). Additionally, mutagenesis of the ASK motif, a conserved β-arrestin interaction motif within the C terminus of human 5-HT_2AR, reduced JCPyV infection (Fig. 7) and entry (Fig. 8), indicating that the β-arrestin binding domain of 5-HT_2AR is critical for JCPyV infection and internalization. These results demonstrate that JCPyV utilizes CME to infect multiple host cell types and characterize the key host cell proteins involved, further defining the role of 5-HT₂Rs in JCPyV internalization strategies (Fig. 9).

FIG 9.

Model of JCPyV entry. JCPyV initially interacts with LSTc via direct interactions with VP1. Internalization of JCPyV is facilitated by 5-HT₂Rs. 5-HT₂R-mediated entry likely first activates β-arrestin at the site of infection, facilitating an interaction with the receptor through the conserved 5-HT₂R ASK motif. Clathrin and AP2 subunits β2 and μ2 are recruited, likely forming a clathrin-coated pit around the viral cargo. Internalization is completed by the GTPase scission mechanism of dynamins 1 and 2, allowing for JCPyV to traffic through the endocytic compartment in a clathrin-coated vesicle.

Viruses usurp many endocytic pathways to enter host cells (46), including clathrin-mediated endocytosis, macropinocytosis, caveolae, clathrin- and caveola-independent mechanisms, and cholesterol-mediated or -independent pathways (45). Interestingly, polyomaviruses have been reported to enter cells by a number of the above-mentioned routes (51, 52, 100), yet one distinct feature is clear: following internalization, all polyomaviruses reach the endoplasmic reticulum (ER) for uncoating (56–61). Internalization strategies and targeting of cargo to the proper cellular compartment for viral uncoating events are largely reliant on receptors utilized for viral attachment and entry (46). Following internalization, SV40 is deposited into a caveolin-positive vesicle and then traffics to the ER for partial uncoating (101). Like other polyomaviruses, localization of SV40 to the ER is a direct result of binding to and internalization by specific ganglioside receptors (57, 102–105). Although JCPyV binds to gangliosides with low affinity, it does not utilize gangliosides as functional viral receptors (31). Instead, JCPyV infection is dependent on attachment to the sialic acid-containing LSTc receptor motif (28) and requires 5-HT₂Rs (32) for internalization by CME (33). This dichotomy in polyomavirus entry pathways suggests that JCPyV may utilize proteinaceous receptors, 5-HT₂Rs, in order to facilitate clathrin-mediated endocytosis, rather than the ganglioside-dependent, non-clathrin-dependent pathways utilized by other polyomaviruses (49, 50), for arrival to the ER.

Upon activation, GPCRs, including 5-HT₂Rs, are regulated by two distinct mechanisms, G-protein- and β-arrestin-dependent signaling (78, 106). The activation of these pathways is based on the ligand that induces either G-protein and/or β-arrestin coupling to the receptor (72, 95, 107–110). Upon recruitment, β-arrestins interact with the intracellular loops of 5-HT₂Rs, leading to further recruitment of endocytic proteins (39, 95, 97, 111–114) and activation of signaling events (97, 107, 110, 115). In addition to regulation of 5-HT₂R internalization and recruitment of β-arrestin, internalization of GPCRs into clathrin-coated vesicles also implicates the potentiation of downstream signaling cascades, including the mitogen-activated protein kinase (MAPK) pathway (97, 116, 117). Interestingly, the MAPK extracellular signal-regulated kinase (ERK) is activated at time points coinciding with JCPyV entry (43, 85) and is required for viral infection (85). We have demonstrated that β-arrestin, a known activator of MAPK and ERK1/2 signaling, is essential for JCPyV infection and internalization. However, the influence of β-arrestin-mediated internalization of 5-HT₂Rs on ERK signaling in the context of JCPyV pathogenesis remains to be explored.

Conserved intracellular motifs of 5-HT₂Rs have been implicated in serving as points of interaction between endocytic scaffolding machinery and the receptors (80, 98). Among these motifs are domains associated with interactions for β-arrestins, including the Asp-Arg-Tyr (DRY), Asn-Pro-X-X-Tyr (NPXXY), and ASK motifs in human 5-HT₂Rs (118–120). The ASK motif is required for β-arrestin-induced internalization, and disruption of this domain has been demonstrated to impede normal trafficking of 5-HT_2AR (80). Interestingly, expression patterns of these motifs are not conserved across all 5-HT receptor subtypes, and only 5-HT₂Rs of human origin express the ASK motif (determined by MacVector amino acid alignment). Our findings demonstrate that the ASK motif within 5-HT_2AR is essential for JCPyV entry (Fig. 8) and infection (Fig. 7), suggesting that the presence of the ASK motif within 5-HT₂Rs may confer specificity for JCPyV utilization of this receptor family. However, the presence of multiple β-arrestin-interacting domains suggests that more than one interaction may be necessary to facilitate ligand-mediated internalization of these receptors and warrants further exploration.

In this study, we have characterized the key proteins required for JCPyV internalization, including β-arrestin, clathrin, AP2, and dynamin. Furthermore, we have identified a β-arrestin interaction motif within 5-HT_2AR that is required for JCPyV infection, suggesting that it may contribute to JCPyV internalization strategies utilizing 5-HT₂Rs. Collectively, these findings contribute to our understanding of JCPyV internalization and infection, identifying new targets for the development of antiviral therapies to treat JCPyV infection and prevent PML pathogenesis.

MATERIALS AND METHODS

Cells, viruses, antibodies, plasmids, and reagents.

SVG-A cells (13) were maintained in complete minimum essential medium (MEM) (Corning) containing 10% fetal bovine serum (FBS), 1% penicillin-streptomycin (P/S) (Mediatech, Inc.), and 0.2% Plasmocin prophylactic (InvivoGen). HEK293A cells were cultured in complete Dulbecco’s modified Eagle medium (DMEM) (Corning) supplemented with 10% FBS, 1% P/S, and 0.2% Plasmocin prophylactic. HEK293A cells stably expressing YFP fusions of 5-HT₂R subtypes 2A, 2B, and 2C (33) were maintained in complete DMEM additionally supplemented with 1% G418 (Corning). All cell types were stored in a humidified incubator at 37°C with 5% CO₂. Cell lines were generously provided by the Atwood Laboratory (Brown University) and were verified by the ATCC.

The generation and proliferation of JCPyV strain Mad-1/SVEΔ and SV40 strain 777 (provided by the Atwood Laboratory [Brown University]) were described previously (121, 122). Alexa Fluor 488- and 647-labeled JCPyVs were prepared as previously described (31). Determination of the titers of JCPyV and SV40 was performed in SVG-A cells by a fluorescent focus unit (FFU) infectivity assay. The clathrin inhibitor Pitstop 2 (catalog number ab120687; Abcam) and the dynamin inhibitor Dynole series kit (catalog number ab120474; Abcam) were reconstituted, according to the manufacturer’s instructions, in dimethyl sulfoxide (DMSO) and used at the indicated concentrations at the time points specified. JCPyV-neutralizing antiserum was used to neutralize noninternalized virus (1:10,000; generously provided by Walter Atwood). Antibodies used for FFU infectivity assays include PAB597 (1:10), a supernatant containing a monoclonal antibody (mAb) against JCPyV VP1 derived from a hybridoma (generously provided by Ed Harlow); PAB962 (neat or 1:2), a mAb for large T antigen specific to JCPyV which does not react to SV40 large T antigen (generously provided by the Tevethia Laboratory [Penn State University]) (123); Ab-2 (1:50), a polyclonal antibody (pAb) that detects large T antigens of SV40 and JCPyV (catalog number DP02; Calbiochem); and secondary polyclonal goat anti-mouse Alexa Fluor 488 or 594 antibodies (Thermo Fisher). Antibodies used for Western blot analysis include anti-β-arrestin 1/2 monoclonal Ab (generously provided by Cheryl Craft, Institute for Genetic Medicine [described in reference124]; used for detection in SVG-A cells) or β-arrestin 1/2 monoclonal Ab (1:500) (catalog number 4674S; Cell Signaling Technology) (used for detection in HEK293A cells), polyclonal anti-clathrin heavy chain (1:250) (catalog number ab21679; Abcam), anti-dynamin 1 polyclonal (1:250) and anti-dynamin 2 polyclonal (1:250) antibodies (catalog numbers PA1-660 and PA1-661; Thermo Fisher), monoclonal anti-AP2 μ2 (1:50) (catalog number 611350; BD Biosciences), monoclonal anti-AP50 β-adaptin (1:250) (catalog number 610381; BD Biosciences), monoclonal anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mouse-specific (1:2,000) (catalog number ab8245; Abcam) or anti-GAPDH rabbit-specific (1:2,000) (catalog number ab8245; Abcam) antibodies, and the corresponding infrared 680 goat anti-mouse and infrared 800 goat anti-rabbit secondary antibodies (1:15,000) (Li-Cor).

siRNA treatment.

SVG-A or HEK293A cells were seeded in 12-well plates to ∼50% confluence. siRNA transfections were performed using RNAiMax (Thermo Fisher) according to the manufacturer’s instructions or modified as indicated in Table 1. Following the addition of siRNA, cells were incubated at 37°C in a humidified incubator with 5% CO₂ for 72 h. Transfection efficiency was monitored through the use of Block-iT Alexa Fluor red control oligonucleotide (Life Technologies) at 48 h posttransfection. Following siRNA treatments, cells were processed for Western blot analysis or infected with JCPyV or SV40 (at the MOIs indicated in the figure legends). siRNA-induced toxicity was measured by a propidium iodide flow cytometry-based viability assay (Thermo Fisher).

TABLE 1.

Conditions for siRNAs used in this study

Target or control	Concn of siRNA added (pmol/well)	Time(s) of siRNA addition (h)	Supplier, sequence targeted, and/or catalog no.
Clathrin heavy chain	37.5	0 and 24	Ambion, CLTC, s475
Dynamin 1	10	0	Ambion, DNM1L, s19559
Dynamin 2	10	0	Ambion, DNM2, s4212
β-Arrestin	7.5	0	Cell Signaling, SignalSilence beta-arrestin 1 siRNA I, 6218s
AP2 μ2	7.5	0	Dharmacon, AP2M2, custom/AAGUGGAUGCCUUUCGGGCAUU
AP2 β2	7.5	0	Ambion, AP2B2, s36
EGFR	Control used at same concn as target	Control added at same time as target	Cell Signaling, SignalSilence EGF receptor siRNA II, 6482s
Scrambled nontargeting	Control used at same concn as target	Control added at same time as target	Cell Signaling, SignalSilence control siRNA (unconjugated), 6568s
Block-iT Alexa Fluor red control oligonucleotide	Control used at same concn as target	Control added at same time as target	Life Technologies, 465318

siRNA Western blot analysis.

To determine the efficacy of siRNA knockdown, following siRNA treatment, SVG-A or HEK293A cells were washed in phosphate-buffered saline (PBS), removed from wells with a cell scraper, washed in PBS, and then pelleted by centrifugation at 414 × g at 4°C for 5 min. Cell pellets were resuspended in 50 μl of Tris-HCl lysis buffer (1 mM EDTA, 50 mM Tris-HCl, 120 mM NaCl) or radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors (1:10; Sigma-Aldrich) and phosphatase inhibitors (1:100; Sigma-Aldrich) and incubated on ice for 10 min or 20 min, respectively. Cellular insoluble material was pelleted at 21,130 × g at 4°C for 10 min. For CHC siRNA-treated samples, cellular pellets were sonicated on ice at 35% amplitude for 10 s, three times, prior to centrifugation. Lysates containing protein were mixed 1:1 with Laemmli sample buffer (Bio-Rad), boiled at 95°C for 5 min, and resolved by SDS-PAGE through a 10% TGX minigel (Bio-Rad). Proteins were transferred to nitrocellulose membranes (Bio-Rad) using a semidry transblot system (Bio-Rad) at 2.5 Å (25 V) for 3 min. Membranes were equilibrated in 1× Tris-buffered saline (TBS) for 10 min before blocking in 5% nonfat milk–TBS-T (1× TBS-Tween 20 [0.1%]) or 5% bovine serum albumin (BSA)–TBS-T at room temperature (RT) for 1.5 h before 3 washes with TBS-T at RT. Membranes were then incubated with the primary antibody indicated and a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (housekeeping) antibody in 5% BSA–TBS-T or 5% nonfat milk–TBS-T at 4°C overnight, with rocking. Membranes were washed extensively in TBS-T at RT before incubation with secondary antibodies (Li-Cor) in 5% nonfat milk–TBS-T at RT for 1 h, with rocking. Membranes were washed before imaging on a Li-Cor Odyssey CLx apparatus. Reduction in protein expression was calculated by defining the relative fluorescence of each target protein band and the loading control in each lane utilizing Li-Cor ImageStudio Western blot analysis software (version 5.2). Target protein bands were normalized to the loading control, and protein reduction was calculated in comparison to the irrelevant EGFR siRNA control. Knockdown by siRNA was optimized for reductions of at least 70%.

JCPyV and SV40 FFU infectivity assays.

Following chemical treatments, siRNA treatments, or transfections, cells were infected with JCPyV or SV40 at the MOIs indicated in the figure legends. Virus was diluted in the infection inoculum, either MEM or DMEM, containing 2% FBS and 1% P/S, at a total volume of 200 μl per well (24-well plates) or 300 μl (12-well plates). Cells were incubated in the viral inoculum at 37°C for 1 h, cell-type-specific complete medium was added, and cells were incubated at 37°C for 48 h (T antigen) or 72 h (VP1). Following infection, cells were washed in 1× PBS and then fixed in ice-cold methanol (MeOH) at −20°C for 20 min or in 4% paraformaldehyde (PFA) at RT for 10 min before staining by indirect immunofluorescence. Following fixation, cells were washed extensively with 1× PBS and permeabilized with 1% Triton X-100–PBS (Thermo Fisher) at RT for 15 min. Cells were incubated with mAb PAB962 for T antigen or VP1 antibody PAB597 to detect newly synthesized T antigen or VP1 protein in SVG-A cells at 48 and 72 hpi, respectively (123), or with pAb AB-2 to detect T antigen for SV40 in HEK293A cells in 1× PBS at 37°C for 1 h; washed 3 times; and incubated at 37°C for 1 h with secondary polyclonal goat anti-mouse Alexa Fluor 488 or 594 antibodies. Infection was quantified by determining the number of VP1-positive (VP1⁺) or T-antigen-positive (T-Ag⁺) nuclei divided by the total number of DAPI-positive (DAPI⁺) cells per 10× visual field (percent infection). The number of DAPI (4′,6-diamidino-2-phenylindole)-positive cells per visual field was determined through the use of a binary screen described previously (85), utilizing Nikon NIS-Elements Basic Research software (version 4.50.00, 64 bit). Average percent infection was then normalized to the control indicated.

JCPyV and SV40 infection in the presence of inhibitors.

SVG-A cells were seeded to ∼70% confluence in a 24-well plate before pretreatment with the clathrin inhibitor Pitstop 2 at the indicated concentrations in 2% FBS-containing MEM at 37°C for 15 min. Cells were infected with JCPyV in 200 μl of MEM (2% FBS) at the MOIs indicated in the figure legends, in the presence of chemical inhibitors, at 37°C for 1 h. Cells were fed with 1 ml complete medium containing JCPyV antiserum (1:10,000) and incubated at 37°C for 72 h. For Pitstop 2 addback experiments, cells were treated with Pitstop 2 at the time points indicated, either prior to, at the time of, or 4 h following infection by JCPyV. After infection had progressed for 4 h, cells were fed with 1 ml of complete medium containing JCPyV antiserum (1:10,000). For infectivity in the presence of Dynole, SVG-A cells were seeded to ∼70% confluence in a 24-well plate before pretreatment with the dynamin 1/2 inhibitor Dynole 34-2 or the negative control Dynole 31-2 at the indicated concentrations in 2% FBS-containing MEM at 37°C for 30 min. Cells were infected with JCPyV in 200 μl of MEM (2% FBS) at the MOIs indicated in the figure legends, in the presence of chemical inhibitors, at 37°C for 1 h. Cells were fed with 1 ml complete medium and incubated at 37°C for 4 h. At 4 hpi, the viral inoculum was removed and replaced with 1 ml complete medium. Following infection, cells were fixed and stained by indirect immunofluorescence for nuclear expression of VP1. Inhibitor-induced toxicity was measured by a 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) viability assay (Promega) to define optimal working concentrations of inhibitors.

Site-directed mutagenesis of 5-HT₂ receptors.

5-HT_2AR–YFP fusion construct plasmid generation was described previously (123). Potential β-arrestin binding sites were altered by site-directed mutagenesis with the QuikChange II site-directed mutagenesis kit (Agilent). Primers for each mutation were designed by the use of Agilent QuikChange Primer Design software and were manufactured and high-performance liquid chromatography (HPLC) purified by Integrated DNA Technologies (IDT) (Table 2). Following mutagenesis, plasmids were transformed into DH5α-competent cells (Invitrogen). Mutated sequences were verified by DNA sequencing (Genewiz [Plainfield, NJ] or The University of Maine DNA Sequencing Facility) and analysis via MacVector and subsequently transiently expressed in HEK293A cells. GenBank accession numbers are provided in Table 2.

TABLE 2.

Primers and GenBank accession numbers for mutated receptors used in this study

Mutation	Primer (5′–3′) for 5-HT2AR–YFP receptor mutation		GenBank accession no.
	Forward	Reverse
5-HT2A-S457A	CAGCATTCTGAAGAGGCTGCTAAAGACAATAGCGACG	CGTCGCTATTGTCTTTAGCAGCCTCTTCAGAATGCTG	MK086042
5-HT2A-SK457-58AA	AGCAGCATTCTGAAGAGGCTGCTGCAGACAATAGCGACGGAGTG	CACTCCGTCGCTATTGTCTGCAGCAGCCTCTTCAGAATGCTGCTT	MK086043
5-HT2A-K458A	CTCCGTCGCTATTGTCTGCAGAAGCCTCTTCAGAATGCTG	CAGCATTCTGAAGAGGCTTCTGCAGACAATAGCGACGGAG	MK086041

Transfection of mutated receptor plasmids in HEK293A cells.

HEK293A cells were seeded to ∼80% confluence into 24-well plates in 500 μl or into 96-well number 1.5 glass-bottomed plates (CellVis) in 100 μl of DMEM containing 10% FBS without antibiotics. Transfection complexes containing 1 μg of DNA–1 μl of Lipofectamine 2000 (Invitrogen) in incomplete medium (lacking FBS and antibiotics) at total concentrations of 1 μg/well (24-well plate) and 0.16 μg/well (96-well plate) were incubated at RT for 20 min. Following incubation, complexes were added to the cells, and the cells were incubated at 37°C for 4 h. Transfection medium was removed, complete medium containing antibiotics was added, and plates were then incubated at 37°C for an additional 20 h. Transfection efficiency was determined through visualization of YFP expression by fluorescence microscopy, and cells were fixed or infected as indicated.

Surface expression of mutant receptor constructs in HEK293A cells.

Following transfection of wild-type and mutated 5-HT_2AR–YFP plasmids in HEK293A cells, cells were washed with 1× PBS and then fixed in 4% PFA for 10 min, followed by 3 washes with 1× PBS. Cells were incubated in blocking reagent (1× PBS containing 2% goat serum, 0.2% Triton X-100, and 0.1% BSA) at RT for 1 h and then stained with a pan-cadherin-specific antibody (1:200) (catalog number ab6528; Abcam) in blocking reagent at 4°C overnight. Cells were extensively washed with 1× PBS and incubated with secondary goat anti-mouse Alexa Fluor 594 antibody (1:1,000) in blocking reagent at RT for 1 h. Cells were washed with 1× PBS, and DAPI was added to stain for the nucleus (1:1,000) for 10 min at RT prior to final 1× PBS washes and visualization by confocal microscopy. Fields of view were collected using an Olympus laser scanning confocal microscope (model IX81; Olympus America, Inc.) at a ×60 magnification (oil immersion) using FluoView software (version 04.01.01.05). Fields of view were defined using DAPI visualization, cells were then viewed via z sectioning, and fluorescence images were collected using 405/635- and 543/488-nm multiline argon lasers. At least 10 fields of view were captured for each sample measured. Cell surface expression of mutant receptors was determined by using ImageJ. Percent surface expression was determined by Mander’s overlap coefficient utilizing colocalization threshold analysis (ImageJ) (125). Each field of view in comparable z planes was analyzed and represents the percentage of overlap between the pan-cadherin cell surface marker and YFP expression for transfected receptors. Generation of box-and-whisker plots was performed using MATLAB and Statistics Toolbox (release R2018a; The MathWorks, Inc., Natick, MA).

Colocalization analysis of JCPyV in SVG-A cells.

SVG-A cells were seeded to ∼80% confluence in 96-well number 1.5 glass-bottomed plates in 100 μl of MEM containing 10% FBS without antibiotics. Transfection complexes of either emerald-clathrin (em-clath) or emerald-caveolin (em-cav) (gifts from Michael Davidson [described in references 126 and 127]; Addgene plasmids 54040 and 54026, respectively), containing 1 μg of DNA–1 μl of Lipofectamine 2000 in incomplete medium (lacking FBS and antibiotics) at a total concentration of 0.16 μg/well, were incubated at RT for 20 min. Following incubation, complexes were added to each well, and the mixtures were incubated at 37°C for 4 h. Transfection medium was removed, complete medium containing antibiotics was added, and plates were incubated an additional 20 h at 37°C. Cells were then prechilled on ice at 4°C for 45 min prior to the addition of 10% phenol-free medium containing JCPyV-647 (at the MOIs indicated in the figure legends). Cells were then incubated on ice at 4°C for 1 h, allowing for synchronized viral attachment, washed with 1× PBS, and fixed with 4% PFA or incubated at 37°C for 5 min, washed with 1× PBS, and fixed. Cells were incubated with DAPI nuclear stain (1:1,000), and imaging was done by confocal microscopy. Fields of view were defined using DAPI visualization, the cells were then viewed via z sectioning, and at least 10 fields of view were collected for each sample. Colocalization between JCPyV-647 and em-cav or em-clath was determined for each field of view in comparable z planes by Mander’s overlap coefficient using colocalization threshold software in ImageJ for each field of view (percent colocalization). Generation of box-and-whisker plots was performed using MATLAB and Statistics Toolbox.

Viral attachment by flow cytometry following siRNA treatment.

Following siRNA treatment or transfection of wild-type or mutated receptors, SVG-A or HEK293A cells were removed from plates using CellStripper (Corning) and pelleted by centrifugation at 414 × g at 4°C for 5 min. Cells were washed with 1× PBS, pelleted, resuspended in 10% complete phenol-free MEM (Corning), and incubated on ice at 4°C for 45 min. Cells were pelleted, resuspended in 200 μl of 10% complete phenol-free MEM containing Alexa Fluor 488-labeled JCPyV (siRNA-treated SVG-A samples) or Alexa Fluor 647-labeled JCPyV (transfected HEK293A samples), and incubated on ice at 4°C for 1 h to allow for viral attachment. Cells were then pelleted by centrifugation, washed with 1× PBS, fixed in 4% PFA for 10 min, and subsequently resuspended in 300 μl 1× PBS. Fixed cells were then analyzed for viral attachment by flow cytometry using a BD FACSCanto instrument (BD Biosciences) equipped with 488- and 633-nm AP-C laser lines (Becton, Dickinson and Company) for 10,000 events before analysis with BD FACSDiva (Becton, Dickinson and Company) and FlowJo (TreeStar, Inc.) software. Samples were gated to exclude complex and dead cells by utilizing FlowJo software.

Viral entry by confocal microscopy.

Cells were seeded in a 96-well number 1.5 glass-bottomed plate (CellVis) to ∼50% confluence before treatment with siRNA or to ∼80% confluence for transfection with wild-type or mutated receptors. Following siRNA treatment or transfection, cells were chilled on ice at 4°C for 45 min before the addition of Alexa Fluor 488-labeled JCPyV (SVG-A cells) or Alexa Fluor 647-labeled JCPyV (HEK293A cells) in MEM containing 2% FBS and 1% P/S and incubated on ice at 4°C for 1 h for viral attachment. Cells were then incubated at 37°C for 1.5 h, allowing for viral entry, washed in 1× PBS, and fixed in 4% PFA for 10 min. Cells were stained with DAPI and imaged by confocal microscopy. Fields of view were defined using DAPI visualization, cells were then viewed via z sectioning utilizing a 60× objective (oil immersion), and fluorescence and differential interference contrast (DIC) images for each field of view were collected. Viral entry was quantified using FluoView software single-measurement analysis by outlining individual regions of interest (ROIs) utilizing the corresponding DIC images to define the plasma membrane of the cell. ROIs were drawn to exclude the plasma membrane to measure internalized virus by relative fluorescence units per cell for background-corrected samples within an applied threshold intensity. Cross sections were quantified for >30 individual cells for control and target siRNA-treated samples (SVG-A cells) or cells expressing wild-type and mutated receptors (HEK293A cells) for at least three independent experiments.

Data availability.

The sequences were deposited in GenBank under accession numbers MK086042, MK086043, and MK086041.