Poliovirus Replication and Spread in Primary Neuron Cultures

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Virology 340 (2005) 10 – 20 www.elsevier.com/locate/yviro

Poliovirus replication and spread in primary neuron cultures

John K. Daley, Lisa A. Gechman, Jason Skipworth, Glenn F. Rall*

Fox Chase Cancer Center, Institute for Cancer Research, 333 Cottman Avenue, Philadelphia, PA 19111, USA

Received 24 November 2004; returned to author for revision 20 December 2004; accepted 23 May 2005 Available online 11 July 2005

Abstract

While some neurotropic viruses cause rapid central nervous system (CNS) disease upon entry into the brain parenchyma, other viruses that are cytolytic in the periphery either result in little neuropathology or are associated with a protracted course of CNS disease consistent with persistent infection. One such virus, poliovirus (PV), is an extremely lytic RNA virus that requires the expression of CD155, the poliovirus receptor (PVR), for infection. To compare the kinetics of PV infection in neuronal and non-neuronal cell types, primary hippocampal neurons and fibroblasts were isolated from CD155+ transgenic embryos and infected with the Mahoney and Sabin strains of PV. Despite similar levels of infection in these ex vivo cultures, PV-infected neurons produced 100-fold fewer infectious particles as compared to fibroblasts throughout infection, and death of PV-infected neurons was delayed approximately 48 h. Spread in neurons occurred primarily by trans-synaptic transmission and was CD155-dependent. Together, these results demonstrate that the magnitude and speed with which PV replication, spread, and subsequent cell death occur in neurons is decreased as compared to non-neuronal cells, implicating cell-specific effects on replication that may then influence viral pathogenesis. D 2005 Elsevier Inc. All rights reserved.

Keywords: Poliovirus; CD155; Primary neurons; Mouse embryonic fibroblasts; CNS; Viral spread; Trans-synaptic transmission

Introduction terminus of the internal VP1 protein is extruded (Everaert et al., 1989; Fricks and Hogle, 1990; Gomez et al., 1993; Poliovirus (PV) is an enterovirus belonging to the family Lonberg-Holm et al., 1975; Wien et al., 1996). It is believed Picornaviridae and is the etiological agent of poliomyelitis that the exposure of the hydrophobic N-terminus of the VP1 in humans. PV possesses a positive strand RNA genome of protein forms a pore in the host cell membrane, allowing the approximately 7.5 kb, packaged in a non-enveloped viral particle to enter the cytoplasm. Additionally, these icosahedral protein capsid (Hogle et al., 1985). The capsid conformational changes in the virion result in a reduced consists of 60 copies of each of four virally encoded sedimentation coefficient in the viral particle from the native structural proteins VP1, VP2, VP3, and VP4. The icosahe- 160S to an infectious intermediate or 135S sub-particle. It is dral structure formed by these capsid proteins produces deep thought that viral engagement with CD155 triggers this surface depressions, or Fcanyons_, that have been identified conformational change (Tsang et al., 2001); however, as receptor binding sites (Colston and Racaniello, 1994, whether the loss of VP4 is the basis of the altered 1995; Hogle et al., 1985; Racaniello and Ren, 1996). PV sedimentation coefficient remains debated (Basavappa et infection initiates when the viral particle binds to its cellular al., 1998). receptor, CD155 (Belnap et al., 2000; He et al., 2000). Upon The normal route of PV infection in naturally permissive binding, the PV particle undergoes a conformational change hosts begins with infection of the enteric system through in which the VP4 capsid protein is lost and the amino- oral ingestion of the virus. Viral particles initially replicate in the gastrointestinal system, but replication at this site does not result in any detectable pathology (Bodian, 1955; Sabin, * Corresponding author. Fax: +1 215 728 2412. 1956). In approximately 0.1–1% of acute PV infections, the E-mail address: [email protected] (G.F. Rall). virus invades the CNS, producing the paralytic disease

0042-6822/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2005.05.032 J.K. Daley et al. / Virology 340 (2005) 10–20 11 symptoms associated with poliomyelitis. This disease 160S form of the viral particle, not the 135S sub-particles results from PV infection, translational inhibition, and thought to be required for infection (Arita et al., 1999). One destruction of motor neurons in the central nervous system possible interpretation of these results is that interaction (CNS) and is clinically manifested as limb paralysis. with the viral receptor may not be required for PV Interestingly, despite a rapid course of viremia that can be transmission across the synapse, a theory for which detected soon after viral exposure, PV appears to adopt a evidence exists in a model of neuronal infection by another less productive course in the CNS (Girard et al., 2002; RNA virus, measles (Lawrence et al., 2000). Minor, 1996). The ability of polio to spread in a somewhat Although PV has been studied extensively in vivo, its restricted manner within the CNS may foster the establish- replication and spread in purified primary neuron cultures ment of a slow or persistent infection. While controversial, have not been extensively explored. Based on observations this observation has led to suggestions by others that chronic that PV infection of the CNS is distinct from infection of the infection may be linked to diseases such as post-polio periphery with respect to viral yield and cytopathicity, we syndrome, in which patients develop signs of poliomyelitis undertook studies to address whether PV adopts an altered months to years after primary exposure (Destombes et al., replication profile in purified primary neurons and to further 1997; Pavio et al., 1996). explore the basis of PV spread in such cultures. Using The natural expression of CD155 in humans and some highly enriched primary neuron cultures isolated from primates, but not in experimental animals such as mice, CD155+ transgenic embryos, as well as control mouse limited early studies of PV spread and pathogenesis. In embryo fibroblasts (MEFs) obtained from the same animals, 1990–91, transgenic mice expressing CD155 were estab- we compared the kinetics of infection of neurons with lished; these mice were susceptible to PV infection by all MEFs, established the mode of interneuronal viral spread, three viral serotypes (Koike et al., 1991; Ren et al., 1990). and clarified the role played by CD155 in neuronal PV The development of these valuable models was responsible transport. for substantial advances in our understanding of PV replication and pathogenesis; however, important aspects of PV interaction with its natural host remain obscure. For Results example, while expression of CD155 is required for poliovirus entry (Koike et al., 1990; Mendelsohn et al., Poliovirus growth kinetics in permissive neurons and mouse 1989), it is not sufficient to confer permissivity, as polio- embryo fibroblasts virus cannot establish an infection in CD155-expressing mice by the intra-oral route, despite expression of CD155 To assess PV replication kinetics in different primary cell within the gut (Ren et al., 1990). cultures, type 1 PV infection by the Sabin strain and the The pathway of PV spread from the periphery to the CNS more virulent Mahoney strain of cultured CD155+ neurons in susceptible hosts is not fully understood. As viremia is a was compared to infection of mouse embryo fibroblasts prerequisite for the development of poliomyelitis and (MEFs) isolated from the same embryos. dendritic cells and macrophages can be infected with To confirm that CD155+ explant cultures indeed express poliovirus (Wahid et al., 2005; Buisman et al., 2003; this receptor, immunofluorescence using a monoclonal Freistadt et al., 1993), polio may gain entry to the brain antibody to CD155 was performed on CD155+ and via transport of infected blood cells across the blood–brain CD155À cells. As shown in Fig. 1, neurons (A–C) and barrier (BBB) (Yang et al., 1997). A second proposed MEFs (G–I) isolated from CD155+ transgenic mice were mechanism of PV neurotropism is transport along neural immunopositive, whereas neurons isolated from CD155À pathways, perhaps provoked by muscle injury at the site of (c57Bl/6) embryos (D–F) were not. The slight background infection, as in the case of provocation poliomyelitis signal detected in nontransgenic neurons is likely due to (Gromeier and Wimmer, 1998). In immunoelectron micro- nonspecific binding of the secondary antibody. scopy studies of spinal cord sections of mice infected with PV replication and transmission in CD155+ primary the type 2 Lansing strain, PV antigens were primarily neurons and MEFs were then examined. Cells were infected detected in neuronal axons and synapses, suggesting that with PV-Sabin or PV-Mahoney (MOI = 0.1); at 8 h spread may be trans-synaptic (DalCanto et al., 1986). intervals, cells and supernatants were collected and analyzed Subsequently, PV was shown to travel to the CNS from for infection by flow cytometry and immunocytochemistry infected muscle tissue in transgenic mice via major nerves, (Figs. 2A and B), extent of cell death by cellular such as the sciatic nerve (Ponnuraj et al., 1998; Ren and morphology and counting of residual live cells (Fig. 2C), Racaniello, 1992), providing additional support for trans- and virus production by plaque assay and infectious center synaptic transmission. The speed of PV spread from the assay (Fig. 3). As shown by flow cytometry (Fig. 2A) at 24 periphery to the CNS via neuron–neuron interactions was and 40 h post-infection (hpi), the extent of infection in later shown to be consistent with fast retrograde axonal MEFs and neurons was comparable. Although more MEFs transport (Ohka et al., 1998). Interestingly, however, PV were infected at each timepoint evaluated, both neurons and particles isolated from the sciatic nerve were of the intact MEFs became heavily infected over time, and the rate of 12 J.K. Daley et al. / Virology 340 (2005) 10–20

Fig. 1. CD155 expression in primary hippocampal neurons and mouse embryonic fibroblasts (MEFs). CD155+ primary hippocampal neurons (A–C) and mouse embryonic fibroblasts (G–I) were prepared from PVR transgenic mice. CD155À control neurons (D–F) were prepared from c57Bl/6 mice. All cells were immunostained for CD155 expression. Phase (A, D, G), fluorescence (B, E, H), and merged (C, F, I) are shown. Representative images were taken from one of three identical experiments. Magnification: 100Â. spread within the cultures was similar. For the Mahoney 2B, panel f; Fig. 2C), and even at 40 hpi, neuronal death was strain, 72% and 84% of MEFs were immunopositive at 24 still less than 50% (Fig. 2B, panel j). Although MEF cell and 40 hpi respectively, compared to 36% and 58% death was less pronounced with the Sabin strain, most infection of neurons (Fig. 2A; Table 1). Although the infected MEFs showed signs of nuclear condensation and number of antigen-positive cells was lower following rounding (Fig. 2B, panels g and k), whereas neurons at the infection by the less virulent Sabin strain, a similar profile same timepoints appeared healthy, with intact processes of infection was observed. Interestingly, two distinct peaks (panels h and l). Thus, while neurons eventually do die as a were observed in the FACScan histograms of Sabin-infected result of PV infection with both strains, virus-induced MEFs at 40 hpi (Fig. 2A); this was also observed as two cytopathicity was noted significantly earlier in MEFs than immunopositive populations that differed in staining inten- in neurons: dying fibroblasts were seen as early as 8–16 hpi sity (Fig. 2B, panels k and m). for both strains of PV; in contrast, neuronal cytopathicity was Despite efficient PV infection and spread in both cell not observed until 36–48 hpi, and even at this late timepoint, types, virus-induced cytopathicity in neurons was reduced as viable neurons were abundant. compared with matched infected MEFs (Fig. 2C). For these The delay in PV-induced cytopathic effect (CPE) in studies, the number of live and dead cells was determined at neurons was consistent with a 100- to 500-fold decrease in each timepoint by morphology, TUNEL staining, or trypan extracellular infectious viral titers, relative to matched MEFs blue exclusion. Mahoney infection of MEFs was highly (Fig. 3A). For example, in the representative experiment lytic, killing virtually all infected cells by 24 hpi (Fig. 2B, shown (of five such experiments), at 24 hpi, MEF titers panel e; Fig. 2C). In contrast, neuronal infection at this same following PV-Mahoney challenge were 9 Â 107 PFU/ml as timepoint was not associated with extensive cell loss (Fig. compared to neuronal titers of 2 Â 105 PFU/ml. Though

Fig. 2. PV spread and cytotoxicity in primary neurons and fibroblasts. (A) Flow cytometry was performed on single cell suspensions of PV-infected MEFs and neurons at 24 and 40 hpi. As described in Materials and methods, cells were labeled with anti-PV antibodies and visualized with an FITC-conjugated secondary antibody. PV-infected: blue lines; uninfected: red lines. (B) CD155+ MEFs (a, c, e, g, i, k, m) and neurons (b, d, f, h, j, l) were infected with either the Mahoney strain (a–b, e–f, i–j) or Sabin strain (c–d, g–h, k–m) of PV (MOI = 0.1) and collected at 16, 24, and 40 h post infection. PV antigen was detected by immunostaining using a monoclonal antibody to PV and visualized by DAB precipitation. For panel (m), arrows denote low antigen-expressing cells, whereas arrowheads denote high-expressing cells. Representative images from one of three identical experiments are shown. Magnification: (a–l): 100Â; (m): 400Â. (C) The total number of live and dead cells was ascertained, as determined by cellular morphology, TUNEL staining, and/or trypan blue exclusion. Viability was independently confirmed by trypan blue exclusion and absence of TUNEL staining (data not shown). Results are shown for both Mahoney (left) and Sabin (right) at 0, 16, 24, and 40 hpi. Black bars: MEFs; stippled bars: neurons. J.K. Daley et al. / Virology 340 (2005) 10–20 13 14 J.K. Daley et al. / Virology 340 (2005) 10–20

extent of infection and the magnitude of infectious particles produced were substantially reduced in neurons as compared to control fibroblasts.

Interneuronal poliovirus spread is primarily trans-synaptic

In a mouse model of neuronal infection by measles virus, reduced cytopathicity was associated with selective transmission at the synapse (Lawrence et al., 2000). To determine if a similar process was occurring for PV, we turned our attention to the basis of spread in these primary cultures. Immunocytochemical data indicated that clusters of infected neurons were connected by antigen-positive processes (e.g., Fig. 2B, panels b, f, j, and l), supporting studies that demonstrated trans-synaptic spread in vivo (Ohka et al., 1998; Ren and Racaniello, 1992). These immunopositive processes were not observed when poliovirus-challenged control PVRÀ neurons were stained for PV antigen (data not shown). To ascertain the relative contributions of extracellular virus and trans-synaptic transfer to PV spread in infected neurons, neutralizing antibodies were employed. Serum from previously infected PVR+ transgenic mice (greater than 25 days post-infection) was used; similar experiments were repeated with a commercially available type 1 neutralizing antibody (BioSource) with identical results. The neutralizing antiserum was added to the cultures 1 h post-infection with PV-Sabin (MOI = 0.1); cells and supernatants from antibody-treated and non-treated cultures Fig. 3. Poliovirus-infected primary neuron cultures produce less extra- were collected at 24 and 32 h post-infection and were cellular infectious virus than infected MEFs. (A) Equivalent numbers of MEFs and primary neurons were plated and infected with PV-Mahoney or immunostained for PV antigens as described previously. As Sabin, as described. At the indicated times post-infection, supernatants were shown in Fig. 4A, addition of neutralizing antibodies collected, and plaque assays were performed; data are presented as log reduced viral titers by greater than 2 orders of magnitude PFU/ml. A representative experiment of five is shown. Black circles: for both cell types. For neurons, this represented a decline in Mahoney titers in MEFs; white circles: Mahoney titers in neurons; black titers from 105 to 103, and for fibroblasts, a decline from 106 triangles: Sabin titers in MEFs; white triangles: Sabin titers in neurons. (B) 3.5 The average number of viral particles produced per infected cell was to 10 . Different sources of antibodies, combinations of calculated for each timepoint post-infection with PV-Mahoney. These antibodies, or addition of higher doses did not appreciably values were obtained by dividing the amount of total extracellular virus by decrease viral titers further (data not shown). the total number of immunopositive cells in the cultures. Black bars: MEFs; As expected, the presence of PV neutralizing antibody in stippled bars: neurons. ND = none detected. infected MEF cultures almost entirely inhibited viral spread, reducing the percent of immunopositive cells at 24 hpi from infection with the Mahoney strain resulted in higher titers in >50% to <10% (Figs. 4B and C, panels a and b). Note that, both cell types, overall differences in titer between the two since the neutralizing antibody was added to the cultures cell populations remained consistent. This difference was after viral challenge, cells infected by the original inoculum also apparent on a per-infected cell basis; as indicated in the representative Mahoney experiment shown in Fig. 3B (one of four such experiments), approximately 20-fold more infec- Table 1 Infectious center analysis of PV-infected neurons and fibroblasts tious PV particles were produced in MEFs than in neurons. Similar results were obtained when identical comparisons HPI Neurons Fibroblasts were done with the Sabin strain of PV (data not shown). % Infected PFU/infected cell % Infected PFU/infected cell To confirm that antigen-positive cells were productively 16 22 0.13 64.9 1.13 infected with PV, infectious center assays were performed. 24 32.6 0.33 77.6 0.92 As shown in Table 1, at 32 h post-infection with PV- 32 46.3 1.1 85 4.95 Mahoney, fibroblasts produced approximately 5 PFU/ Primary neurons and fibroblasts were cultured and infected with PV- infected cell, whereas neurons produced only 1 particle/ Mahoney as described in Materials and methods. Infectious center assays were performed on the infected cells, and the percent of infected cells infected cell, mirroring the data obtained by histochemistry. determined. The total PFU in the supernatants (determined by plaque assay) Thus, by both immunochemical and functional assays, the was then divided by this value to obtain PFU/infected cell. J.K. Daley et al. / Virology 340 (2005) 10–20 15

neurons (Figs. 4B and C, panels c and d). Similar observations were obtained following treatment of Mahoney virus-infected cultures (data not shown). Thus, the absence of an effect on viral spread in neurons, despite a significant impact on extracellular viral levels following neutralizing antibody treatment, supports the hypothesis that the predominant mode of spread in neurons is by trans-synaptic transmission. Because the neutralizing antibody did not completely eliminate extracellular virus, we cannot formally exclude the possibility that some spread in neuron cultures was due to residual infectious particles; however, this seems unlikely as spread in MEFs in the presence of the antibody was completely ablated and neuronal spread still occurred efficiently in antibody-treated neuron and MEF co-cultures (data not shown).

Poliovirus spread between infected CNS neurons requires the expression of CD155

CD155 is required for initiation of PV infection in susceptible mice (Koike et al., 1991; Ren et al., 1990). However, it is not clear if subsequent PV spread between neurons requires CD155, as a recent study indicated that only the intact 160S, and not the ‘‘infectious’’ 135S sub- viral particle, could be found in the sciatic nerve of infected mice (Arita et al., 1999). To address this question, primary neurons from c57Bl/6 mice were loaded with an orange vital dye and were mixed in varying ratios with unlabeled CD155+ neurons. These ‘‘mixed’’ neuron cultures were then infected with either PV-Sabin or PV-Mahoney at an MOI = 0.1 and were collected and fixed 24 hpi. As shown in Figs. 5a, d, and g, neurons infected with either Sabin or Mahoney strains were connected by immunopositive processes. Nevertheless, no spread to nontransgenic neurons (Panels 5b and e) was observed when the images were merged (Figs. 5c and f). This was also true in cultures collected at 40 hpi, a relatively late timepoint post-infection (data not shown). As a control, CD155+ neurons were infected in suspension (Mahoney; MOI = 0.1) and plated + Fig. 4. Poliovirus spread in infected CD155+ neuron cultures occurs in the with uninfected, vital-dye-labeled CD155 neurons in the presence of type 1 PV neutralizing antibody. (A) Supernatants from PV presence of neutralizing antibody. As shown in Figs. 5g–i, Sabin-infected MEFs and neurons, in the presence or absence of type 1 PV spread to vital-dye-labeled neurons occurred readily, as neutralizing antibody, were collected and assayed for infectious particles expected. Thus, we conclude from these experiments that by plaque assay through 32 hpi. Black bars: no neutralizing antibody; expression of CD155 is required for both viral entry and stippled bars: antibody added. The experiment shown is representative of three experiments. (B) CD155+ MEFs and neurons infected with PV-Sabin subsequent trans-synaptic spread. (MOI = 0.1) were incubated for 16 and 24 hpi in the presence or absence of type 1 PV neutralizing antibody, and the percent of infected cells was then determined by immunocytochemistry. (C) CD155+ MEFs (a, b) and Discussion neurons (c, d) were infected with PV-Sabin (MOI = 0.1) and were incubated for 24 h with (+) or without (À) PV type 1 neutralizing antibody. PV antigen was visualized as described in Materials and While many viruses can infect the human CNS, they do methods. Magnification = 100Â. so rarely. For many of these viruses, once access to the CNS is achieved, the profile of replication is distinct from that in the periphery: often, a more chronic or persistent infection would be expected. In contrast, the presence of PV occurs, usually associated with a progressive course, as with neutralizing antibody in neuron cultures infected with PV- measles and the rare but fatal disease, subacute sclerosing Sabin had virtually no impact on viral spread in infected panencephalitis. Poliovirus is the causative agent of polio- 16 J.K. Daley et al. / Virology 340 (2005) 10–20

Fig. 5. Interneuronal spread of PV requires the expression of the poliovirus receptor, CD155. c57Bl/6 neurons (b and e) or CD155+ neurons (h) were loaded with an orange vital dye and plated with unlabeled, PV-infected CD155+ neurons (Sabin: a; Mahoney: d, g). After 24 h, the neurons were fixed and immunostained for PV antigen using an FITC-conjugated secondary antibody. (a, d and g) PV-infected neurons; (b, e, and h) vital dye-labeled neurons; (c, f, and i) merge. Magnification = 100Â. myelitis, a rare and progressive virus-induced degeneration modulate viral replication. While the reduced productivity of motor neurons that results in limb paralysis. In an effort in neurons may be attributable to viral mutations, as to determine whether PV replication and spread were previously shown for some strains of PV (LaMonica and impacted upon neuronal infection, primary cultures from Racaniello, 1989), other reasonable explanations exist. For susceptible PV receptor-expressing transgenic mice were example, it is possible that the cellular machinery available established, and the infection profiles compared. Two to viruses upon neuronal entry may be sufficiently unique to conclusions can be drawn from the experiments described support a different mode of replication. Key differences in in this work. First, the extent of viral cytopathicity and the availability of translation-associated proteins (Guest et extracellular virus production is significantly reduced in al., 2004), the cellular proteins that are required for viral neurons compared to MEFs, despite similar levels of budding (Lawrence et al., 2000), the reduced mitotic infection. These data support previous observations in capacity of parenchymal cells, and the level of cyclic transgenic mice and in established neuroblastoma cells that nucleotides (Miller and Carrigan, 1982) may impact on viral PV replication is restricted upon neuronal infection (Cou- replication and resultant pathogenesis. As these studies derc et al., 2002; Girard et al., 2002). Secondly, despite the continue, it will be of interest to further identify the cellular production of some extracellular viral progeny, the predom- factors that influence the poliovirus life cycle in neurons. inant mechanism of PV spread is via trans-synaptic transfer, Our studies support and extend those of other laborato- a process that is CD155-dependent. The ability of virus to ries who have assessed the spread and pathogenesis of PV in spread via the synapse (rather than through cell lysis, as neurons. While care must be taken not to infer too broadly occurs in non-neuronal cells) may contribute to the reduced from purified primary neurons to the more complex cytopathicity of PV within neurons. interaction of poliovirus with other cell types in vivo, The reduced productivity and cytopathicity of both Sabin studies from our laboratory and others indicate that such and Mahoney strains of PV in neurons are consistent with cultures can reproducibly parallel observations made in observations that many DNA and RNA viruses undergo a permissive mice. For example, work from Blondel and change in the viral replication cycle upon neuronal colleagues using permissive transgenic mice and mixed infection. For example, upon entry into spinal cord neurons, brain cell cultures showed that PV infection of neurons herpes simplex virus-1 undergoes a switch to a latent state in resulted in apoptosis, as caspase inhibitors prevented cell which viral gene expression is virtually ablated (reviewed in death in these cultures (Couderc et al., 2002; Girard et al., Johnson, 1998). Moreover, MV adopts a noncytolytic, non- 1999). Consistent with our observations of delayed neuronal fusogenic, trans-synaptic mode of spread in neurons that is death, apoptosis in mixed cultures was apparent only after not associated with the emergence of viral variants 28 hpi. We have since confirmed in our purified cultures that (Lawrence et al., 2000) (Gechman et al., in preparation). neuronal loss due to PV infection is attributable to Thus, our observations using PV-infected neurons may apoptosis, whereas MEF death is more likely due to classic reflect a common mechanism by which neurons restrict or virus-induced CPE (data not shown). J.K. Daley et al. / Virology 340 (2005) 10–20 17

Work from other laboratories has begun to define the (Banker and Goslin, 1991; Pasick et al., 1994; Rall et al., trafficking of PV within neurons. Wimmer and colleagues 1995), with the exception that cells were maintained in have reported that the cytoplasmic domain of the CD155 neurobasal medium (Life Technologies, Grand Island, NY) receptor binds to Tctex-1 (Mueller et al., 2002), a light chain containing 4 Ag/ml glutamate, in the absence of an astrocyte of the dynein motor complex that governs retrograde feeder layer. These mixed sensory and motor neuron transport in polarized cells. The apparent ‘‘hand-off’’ of cultures are routinely >95% pure, as determined by staining the incoming viral particle to molecular motor proteins with microtubule-associated protein-2 (MAP-2) (data not likely facilitates the fast retrograde axonal transport used by shown). For all studies in this report, plated neurons were PV in vivo to spread from muscle to the CNS through the allowed to differentiate for 72–96 h before infection, sciatic nerve (Ohka et al., 1998, 2004). Given that the coinciding with the formation of functional synapses and mature 160S form of PV was isolated from infected nerves expression of neuron-specific markers (Rall et al., 1995). (Arita et al., 1999; Ohka and Nomoto, 2001; Ohka et al., Mouse embryonic fibroblasts (MEFs) were prepared 1998), it was possible that the interaction of PV with its from the same day 15.5 CD155+ transgenic embryos. receptor may be either different at the synapse than at the Briefly, livers were removed, and the remaining tissues plasma membranes of non-neuronal cells or that CD155 were then finely minced in 10 ml of 0.4% trypsin followed may be dispensable for viral spread altogether. As support by trituration with a 10 ml pipette. The single cell for the latter hypothesis, our laboratory previously demon- suspension was incubated at 37 -C for 10 min, washed strated that measles virus, while requiring a receptor to and re-trypsinized. After large tissues pieces were allowed initially infect neurons, could spread from infected neurons to settle and were discarded, the resulting cell suspension to uninfected neurons in a receptor-independent manner was pelleted by centrifugation at 1100 rpm for 5 min. The (Lawrence et al., 2000). Our studies here indicate that resulting pellet was resuspended in complete DMEM (Life CD155 is required for neuron-to-neuron spread; therefore, if Technologies, supplemented with 5% FCS, 2 mM l- the interneuronal spread of poliovirus is distinct from glutamine, 100 U/ml penicillin, and 100 ng/ml streptomy- infection by an extracellular particle, it is nevertheless cin) and plated into culture flasks. Vero fibroblasts were also CD155-dependent. cultured in complete DMEM. All cells were maintained at The delayed course to PV-induced cytolysis and reduced 37 -C in a humidified incubator with 5% CO2 and were viral productivity that we observed in cultured neurons may routinely tested for mycoplasma contamination. have direct bearing on the development of persistent The Sabin and Mahoney strains of type 1 poliovirus (PV) infection within the CNS. Several studies have demonstra- were a gift from Luis Sigal and were passaged and titered in ted that PV RNA can persist in susceptible mouse neurons, Vero fibroblasts. both in vivo (Destombes et al., 1997) and in vitro (Pavio et al., 1996). Moreover, PV RNA has been detected in human Infections cerebrospinal fluid from individuals who previously had an acute case of poliomyelitis (Leparc-Goffart et al., 1996). Primary neurons and MEFs were plated onto poly-l- PV remains high on the World Health Organization’s list lysine-treated 15-mm-diameter glass coverslips at a density of viruses to be eradicated in the near future, and some have of 500 cells/mm2. Neurons were differentiated for 72–96 h argued that the risks of a sustained vaccination program are prior to infection, whereas MEFs were plated 24 h before greater than the risks of viral recrudescence. Nevertheless, infection. Cells were infected with type 1 PV, Sabin or as recently seen with outbreaks of MV in England, viruses Mahoney strain, at a multiplicity of infection (MOI) = 0.1. can make a rapid and serious recurrence in populations with The coverslips were covered with the viral inoculum and declining vaccine coverage. The diseases associated with incubated for 1 h at 37 -C, after which the inoculum was PV infection, including both the risk of poliomyelitis in the replaced with either conditioned neurobasal medium (pri- short term and post-polio syndrome in the long term, argue mary neurons) or fresh complete DMEM (MEFs). Slips and in favor of further studies to define the interaction of this supernatants were collected every 8 h after infection. Slips virus with its cellular host. were washed briefly in fresh PBS, and the cells were fixed for 5 min in a 1:1 solution of ice-cold methanol: acetone. After fixation and drying, the slips were stored at 4 -Cin Materials and methods PBS.

Cells and viruses Immunofluorescence

Poliovirus receptor (PVR; CD155+) transgenic mice, To detect CD155 expression, a mouse antibody to originally developed by Racaniello and colleagues (Ren et CD155 (BioSource International, Camarillo, CA) was used al., 1990), were obtained from Luis Sigal at Fox Chase. at a 1:100 dilution followed by an Alexa-Fluor 488 Primary hippocampal neurons were prepared from day 15.5 conjugated anti-mouse IgG (Molecular Probes, Eugene, CD155+ transgenic embryos, as previously described OR; 1:500). 18 J.K. Daley et al. / Virology 340 (2005) 10–20

For viral spread studies, CD155-negative neurons were at 4 -C. Cells were centrifuged and resuspended in 400 Alof isolated from c57Bl/6 day 15.5 embryos as described above, 0.5% formaldehyde in PBS and analyzed on a BD FACScan counted, and loaded with Vybrant CMDil cell labeling fluorescence-activated cell scanner. solution (Molecular Probes). The vital dye was added to neurons in suspension at a concentration of 5 Al/2.0 Â 106 Plaque assay cells/ml in serum-free neurobasal media and incubated at 37 -C for 30 min. Labeled neurons were then washed twice Supernatants from infected cells were serially diluted and in fresh serum-free neurobasal medium and centrifuged added to Vero fibroblasts plated in 6-well cell culture dishes through a 2 ml FCS cushion. Immunofluorescence detection in complete DMEM at a cell density of 106 cells/well. After of PV antigen utilized a type 1 PV-specific mouse incubation at 37 -C for 1 h, the viral inoculum was removed, monoclonal (Accurate, Westbury NY; 1:200 dilution in and the cells were covered with a 1:1 mixture of 2Â PBS) followed by incubation with a biotinylated anti-mouse Medium 199 (Life Technologies) and 1% molten agarose in IgG2a (Serotec, Raleigh NC; 1:300 dilution in PBS). Finally, water. The agarose was allowed to solidify, and the plates an FITC-conjugated avidin (1:500 dilution; Vector, Burlin- were then returned to the 37 -C incubator. Two days later, game, CA) was incubated with the slips for 30 min followed the cells were fixed with 10% formaldehyde in PBS. by three 5 min washes. Agarose overlays were removed, and plaques were visual- All slides were visualized using a Nikon fluorescence ized by staining the remaining fibroblasts with 0.1% crystal microscope model TE300, equipped with appropriate violet. filters for the fluorescence studies. For the viral trafficking studies, fluorescent orange and green images from Infectious center assay identical fields were captured and merged using Adobe Photoshop. Infectious center assays (ICA) were performed using PVR+ neurons and MEFs that were infected with the Immunohistochemistry Mahoney strain of poliovirus at an MOI = 0.1. At 16, 24, and 32 hpi, supernatants were removed and assayed for Fixed cells were blocked with 1% (vol/vol) goat serum in infectious PV particles by plaque assay (see above). The PBS, further blocked with avidin and biotin (Vector), and attached cells from each timepoint were subsequently stained for PV antigens as described above. However, after washed with PBS and removed with 0.4% trypsin (2 ml incubation with the secondary antibody, a streptavidin– final volume). The cells were counted (to obtain the total peroxidase conjugate (Vector) was added, and PV was number of cells), serially diluted, and added to Vero visualized by diaminobenzidine precipitation (0.7 mg/ml fibroblasts plated on 6-well cell culture dishes. Two days DAB with 1.6 mg/ml H2O2 in 60 mM Tris; Sigma post-plating, the cells were fixed and subsequently visual- Chemical). In some cases, cells were counterstained with ized with crystal violet. The percentage of infected cells was hematoxylin to visualize nuclei. To quantify cell death in calculated using the following equation: # of plaques (from infected cultures, live and dead cells (as determined by ICA) Â dilution factor / total number of cells. Infectious morphology) were counted from 5 representative fields at virus per-infected cell was calculated using the following 400Â for each timepoint post-infection. Samples were equation: total amount of infectious virus released (deter- coded until after counting. mined by plaque assay) / # of infected cells (determined by ICA). Flow cytometry Neutralizing antibody assay Primary PVR+ hippocampal neurons and MEFs were plated on poly-l-lysine-treated 6-well culture plates at a cell CD155+ MEFs and neurons, plated on coverslips, were density of 500 cells/mm2. Cells were infected with the infected with the Sabin strain of PV at an MOI = 0.1 as Mahoney or Sabin strain of PV at an MOI = 0.1 or were described above. For some slips, 30 Al of type 1 PV mock-infected. Cells were removed from the plates at 24 neutralizing antibody (1:320; BioSource) was added 1 h and 40 hpi using Cell Stripper (Cellgro Mediatech, Inc., post-infection. Infected cells and corresponding superna- Herndon, VA). Cells were pelleted at 1200 rpm for 5 min tants were collected at various times post-infection. Cover- and resuspended in 2 ml cytofix/cytoperm buffer (BD slips were fixed and immunostained for PV antigen as Biosciences/Pharmingen, San Diego, CA) and incubated for described above; supernatants were subjected to plaque 20 min at 4 -C. Cells were then washed twice with perm/ assay. The percent of infected cells at each timepoint was wash buffer and transferred to wells of a 96-well microtiter calculated by counting 5 representative fields/slip at 400Â plate. Poliovirus type 1 antibody at a 1:100 dilution or PBS magnification; identification of the slips was coded until (for no primary control) was added to each well for 30 min after counting. Viral titers from each timepoint were at 4 -C. An anti-mouse IgG FITC-conjugated secondary determined by plaque assay of supernatants from infected antibody (Vector; 1:500) was added to all wells for 30 min cells as described above. J.K. Daley et al. / Virology 340 (2005) 10–20 19

Acknowledgments Guest, S., Pilipenko, E., Sharma, K., Chumakov, K., Roos, R.P., 2004. Molecular mechanisms of attenuation of the Sabin strain of poliovirus type 3. J. Virol. 78, 11097–11107. The authors acknowledge support from the following He, Y., Bowman, V.D., Mueller, S., Bator, C.M., Bella, J., Peng, X., Baker, sources: NIH grants MH 56951 (G.R.) and T32-AI07492 T.S., Wimmer, E., Kuhn, R.J., Rossmann, M.G., 2000. Interaction of the (J.D.) and the F.M. Kirby Foundation. We would also like to poliovirus receptor with poliovirus. Proc. Natl. Acad. Sci. 97, 79–84. thank Isabelle Cruz for secretarial support and Drs. Kerry Hogle, J.M., Chow, M., Filman, D.J., 1985. Three-dimensional structure of Campbell, Wesley Rose, Christine Matullo, and Luis Sigal poliovirus at 2.9 A resolution. Science 229, 1358–1365. Johnson, R.T., 1998. Viral Infections of the Nervous System. Lippincott- for helpful suggestions. Raven, Philadelphia, PA. Koike, S., Horie, H., Ise, I., Okitsu, A., Yoshida, M., Iizuka, N., Takeuchi, K., Takegami, T., Nomoto, A., 1990. The poliovirus receptor protein is References produced both as membrane-bound and secreted forms. EMBO J. 9, 3217–3224. Arita, M., Ohka, S., Sasaki, Y., Nomoto, A., 1999. Multiple pathways for Koike, S., Taya, C., Kurata, T., Abe, S., Ise, I., Yonekawa, H., Nomoto, S., establishment of poliovirus infection. Virus Res. 62, 97–105. 1991. Transgenic mice susceptible to poliovirus. Proc. Natl. Acad. Sci. Banker, G., Goslin, K., 1991. Culturing Nerve Cells. MIT Press, Cam- 88, 951–955. bridge. LaMonica, N., Racaniello, V.R., 1989. Differences in replication of Basavappa, R., Gomez-Yafal, A., Hogle, J.M., 1998. The poliovirus empty attenuated and neurovirulent poliovirus in human neuroblastoma cell capsid specifically recognizes the poliovirus receptor and undergoes line SH-SY5Y. J. Virol. 63, 2357–2360. some, but not all, of the transitions associated with cell entry. J. Virol. Lawrence, D.M.P., Patterson, C.E., Gales, T.L., D’Orazio, J.L., Vaughn, 72, 7551–7556. M.M., Rall, G.F., 2000. Measles virus spread between neurons requires Belnap, D.M., McDermott, B.M., Filman, D.J., Cheng, N., Trus, B.L., cell contact but not CD46 expression, syncytium formation or Zuccola, H.J., Racaniello, V.R., Hogle, J.M., Steven, A.C., 2000. Three extracellular virus production. J. Virol. 74, 1908–1918. dimensional structure of poliovirus receptor bound to poliovirus. Proc. Leparc-Goffart, I., Julien, J., Fuchs, F., Janatova, I., Aymard, M., Kopecka, Natl. Acad. Sci. 97, 73–78. H., 1996. Evidence of presence of poliovirus genomic sequences in Bodian, D., 1955. Emerging concept of poliomyelitis infection. Science cerebrospinal fluid from patients with post-polio syndrome. J. Clin. 122, 105–108. Microbiol. 34, 2023–2026. Buisman, A.M., Sonsma, J.A., Wijk, M.G., Vermeulen, J.P., Roholl, P.J., Lonberg-Holm, K., Gosser, L.B., Kauer, J.C., 1975. Early alteration of Kimman, T.G., 2003. Pathogenesis of poliovirus infection in PVRTg poliovirus in infected cells and its specific inhibition. J. Gen. Virol. 27, mice: poliovirus replicates in peritoneal macrophages. J. Gen. Virol. 84, 329–342. 2819–2828. Mendelsohn, C.L., Wimmer, E., Racaniello, V.R., 1989. Cellular receptor for Colston, E.M., Racaniello, V.R., 1994. Soluble receptor-resistant poliovirus poliovirus: molecular cloning, nucleotide sequence, and expression of a mutants identify surface and internal capsid residues that control new member of the immunoglobulin superfamily. Cell 56, 855–865. interaction with the cell receptor. EMBO J. 13, 5855–5862. Miller, C.A., Carrigan, D.R., 1982. Reversible repression and activation Colston, E.M., Racaniello, V.R., 1995. Poliovirus variants selected on of measles virus infection in neural cells. Proc. Natl. Acad. Sci. 79, mutant receptor-expressing cells identify capsid residues that expand 1629–1633. receptor recognition. J. Virol. 69, 4823–4829. Minor, P.D., 1996. Poliovirus. In: Nathanson, N. (Ed.), Viral Pathogenesis. Couderc, T., Guivel-Benhassine, F., Calaora, V., Gosselin, A., Blondel, B., Lippincott-Raven, Philadelphia, pp. 555–574. 2002. An ex vivo murine model to study poliovirus-induced apoptosis Mueller, S., Cao, X., Welker, R., Wimmer, E., 2002. Interaction of the in nerve cells. J. Gen. Virol. 83, 1925–1930. poliovirus receptor CD155 with the dynein light chain Tctex-1 and DalCanto, M.C., Barbano, R.L., Jubelt, B., 1986. Ultrastructural immuno- its implication for poliovirus pathogenesis. J. Biol. Chem. 277, histochemical localization of poliovirus during virulent infection of 7897–7904. mice. J. Neuropathol. Exp. Neurol. 45, 613–618. Ohka, S., Nomoto, A., 2001. Recent insights into poliovirus pathogenesis. Destombes, J., Couderc, T., Thiesson, D., Girard, S., Wilt, S.G., Blondel, Trends Microbiol. 9, 501–506. B., 1997. Persistent poliovirus infection in mouse motoneurons. J. Virol. Ohka, S., Yang, W.X., Terada, E., Iwasaki, K., Nomoto, A., 1998. 71, 1621–1628. Retrograde transport of intact poliovirus through the axon via fast Everaert, L., Vrijsen, R., Boeye, A., 1989. Eclipse products of axonal transport system. Virology 250, 67–75. poliovirus after cold-synchronized infection of HeLa cells. Virology Ohka, S., Matsuda, N., Tohyama, K., Oda, T., Morikawa, M., Kuge, S., 171, 76–82. Nomoto, A., 2004. Receptor (CD155)-dependent endocytosis of polio- Freistadt, M.S., Fleit, H.B., Wimmer, E., 1993. Poliovirus receptor on virus and retrograde axonal transport of the endosome. J. Virol. 78, human blood cells: a possible extraneural site of poliovirus replication. 7186–7198. Virology 195, 798–803. Pasick, J.M., Kalicharran, K., Dales, S., 1994. Distribution and trafficking Fricks, C.E., Hogle, J.M., 1990. Cell-induced conformational change in of JHM coronavirus structural proteins and virions in primary neurons poliovirus: externalization of the amino terminus of VP1 responsible for and the OBL-21 neuronal cell line. J. Virol. 68, 2915–2928. liposome binding. J. Virol. 64, 1934–1945. Pavio, N., Buc-Caron, M.H., Colbere-Garapin, F., 1996. Persistent polio- Girard, S., Couderc, T., Destombes, J., Thiesson, D., Delpeyroux, F., virus infection of human fetal brain cells. J. Virol. 70, 6395–6401. Blondel, B., 1999. Poliovirus induces apoptosis in the mouse central Ponnuraj, E.M., John, T.J., Levin, M.J., Simoes, E.A., 1998. Cell-to-cell nervous system. J. Virol. 73, 6066–6072. spread of poliovirus in the spinal cord of bonnet monkeys (Macaca Girard, S., Gosselin, A.S., Pelletier, I., Colbere-Garapin, F., Couderc, T., radiata). J. Gen. Virol. 79, 2393–2403. Blondel, B., 2002. Restriction of poliovirus RNA replication in Racaniello, V.R., Ren, R., 1996. Poliovirus biology and pathogenesis. Curr. persistently infected nerve cells. J. Gen. Virol. 83, 1087–1093. Top. Microbiol. Immunol. 206, 305–325. Gomez, Y.A., Kaplan, G., Racaniello, V.R., Hogle, J.M., 1993. Character- Rall, G.F., Mucke, L., Oldstone, M.B.A., 1995. Consequences of cytotoxic ization of poliovirus conformational alteration mediated by soluble cell T lymphocyte interaction with major histocompatibility complex- receptors. Virology 197, 501–505. expressing neurons in vivo. J. Exp. Med. 182, 1201–1212. Gromeier, M., Wimmer, E., 1998. Mechanism of injury-provoked polio- Ren, R., Racaniello, V.R., 1992. Poliovirus spreads from muscle to the CNS myelitis. J. Virol. 72, 5056–5060. by neural pathways. J. Infect. Dis. 166, 747–752. 20 J.K. Daley et al. / Virology 340 (2005) 10–20

Ren, R., Costantini, F., Gorgacz, E.J., Lee, J.J., Racaniello, V.R., 1990. Wahid, R., Cannon, M.J., Chow, M., 2005. Dendritic cells and macrophages Transgenic mice expressing a human poliovirus receptor: a new model are productively infected by poliovirus. J. Virol. 79, 401–409. for poliomyelitis. Cell 63, 353–362. Wien, M.W., Chow, M., Hogle, J.M., 1996. Poliovirus: new insights from Sabin, A.B., 1956. Pathogenesis of poliomyelitis: reappraisal in the light of an old paradigm. Structure 4, 763–767. new data. Science 123, 1151–1157. Yang, W.X., Terasaki, T., Shiroki, K., Ohka, S., Aoki, J., Tanabe, S., Tsang, S.K., McDermott, B.M., Racaniello, V.R., Hogle, J.M., 2001. Nomura, T., Terada, E., Sugiyama, Y., Nomoto, A., 1997. Efficient Kinetic analysis of the effect of poliovirus receptor on viral uncoating: delivery of circulating poliovirus to the central nervous system the receptor as catalyst. J. Virol. 75, 4984–4989. independently of poliovirus receptor. Virology 229, 421–428.