Herpes simplex virus type 1 infection of polarized ... - CiteSeerX

Journal of General Virology (2004), 85, 775–786

DOI 10.1099/vir.0.19530-0

Herpes simplex virus type 1 infection of polarized epithelial cells requires microtubules and access to receptors present at cell–cell contact sites Sabrina Marozin,3 Ute Prank and Beate Sodeik Correspondence Beate Sodeik [email protected]

Received 25 July 2003 Accepted 2 December 2003

Department of Virology, Hannover Medical School, Carl-Neuberg-Str. 1, D-30623 Hannover, Germany

Mucosal epithelia are invaded from the apical surface during a primary infection by herpes simplex virus type 1 (HSV-1). HSV-1 progeny virus, synthesized from latently infected peripheral neurons that innervate such epithelia, reinfects the epithelia most likely from the basolateral surface. The epithelial cell lines MDCK and Caco-2 can be induced in vitro to differentiate into polarized cells with distinct apical and plasma membrane domains separated by tight junctions if they are cultured on porous membrane filters. Our data using these culture systems showed that highly polarized epithelial cells were not susceptible to apical HSV-1 infection. However, HSV-1 infected these cells if added from the basolateral surface or if a depletion of extracellular Ca2+ had weakened the strength of the cell–cell contacts. Basolateral infection and apical infection after the Ca2+ switch required an intact microtubule network for genome targeting to the nucleus. This system can be used to identify the microtubule motors that HSV-1 uses during virus entry in polarized epithelial cells.

INTRODUCTION Herpes simplex virus type 1 (HSV-1) is a human doublestranded DNA virus, which initially infects epithelial cells in the mucous membranes of the oral cavity (Enquist et al., 1998; Roizman & Knipe, 2001). After replication at this site, progeny virus invades the peripheral nervous system and is transported to the trigeminal ganglion where it establishes a lifelong latent infection. Upon stress, the virus is reactivated and the neurons synthesize progeny virus, which is transported back to the epithelial cells in and around the mouth where it causes recurrent cold sores and blisters (Wagner & Bloom, 1997). Like many DNA viruses, HSV-1 has to deliver its genome of 152 kbp from the plasma membrane to the cell nucleus, where viral transcription and DNA replication take place (Roizman & Knipe, 2001; Whittaker et al., 2000). The genome is packaged into a pre-assembled nuclear capsid with a diameter of 125 nm, which is then further coated with an amorphous protein layer called the tegument. To complete virus assembly, a lipid membrane, the virus envelope, is wrapped around the tegument (reviewed by Enquist et al., 1998; Mettenleiter, 2002). The entry of HSV-1 into cells involves a series of interactions between several viral envelope proteins and molecules of the host plasma membrane. Infection is initiated 3Present address: Max-von-Pettenkofer Institut, Ludwig-MaximiliansUniversita¨t, Mu¨nchen, Germany.

0001-9530 G 2004 SGM

Printed in Great Britain

by the attachment of HSV-1 glycoprotein C (gC) and gB to heparan or chondroitin sulfate proteoglycans (reviewed by (Spear et al., 2000). Nectin-1 (HveC) is considered to be the most important receptor for HSV-1 gD and entry into epithelial cells (Geraghty et al., 1998). Additional HSV-1 gD receptors are HveA (Montgomery et al., 1996) and a heparan sulfate modified by a 3-O-sulfotransferase (Shukla et al., 1999). Nectin-1, like nectin-2, -3, -4 and the homologous poliovirus receptor, belong to the Ca2+-independent cell adhesion proteins of the immunoglobulin superfamily. Nectins mainly co-localize with the Ca2+-dependent cell adhesion protein E-cadherin and catenin in the adherens junctions of epithelial cells. Most forms of nectin can bind with their cytosolic tail to the PDZ domain of l-afadin, an actin-binding protein (Takai & Nakanishi, 2003). HSV-1 gD in combination with gB and the gH/gL complex trigger the fusion of the viral envelope with the cell membrane, which leads to the release of the tegument proteins and the viral capsid into the cytosol (Spear et al., 2000). While alphaherpesvirus infection is often mediated by pH-neutral fusion with the plasma membrane, certain cell types support an endocytic, pH-dependent route for HSV-1 entry (Nicola et al., 2003). After penetration of the actin cortex underneath the plasma membrane, the capsids are transported in many cell types along microtubules to the cell nucleus (Mabit et al., 2002; Sodeik et al., 1997). Microtubules are hollow polar protein cylinders polymerized from a/b-tubulin heterodimers. In 775

S. Marozin, U. Prank and B. Sodeik

cultured cells, their so-called minus ends are usually fixed at a microtubule-organizing centre, which is often localized in close proximity to the cell nucleus. The plus ends of microtubules extend into the cell periphery (Do¨hner & Sodeik, 2004). Like many host organelles, the capsids are transported in unpolarized epithelial cells by the minusend directed microtubule-activated ATPase dynein and its cofactor dynactin to the microtubule-organizing centre in the cell centre (Do¨hner et al., 2002; Sodeik et al., 1997). The capsid binds via importin-b to the nuclear pore and this interaction triggers the release of the HSV-1 genome into the nucleoplasm and allows the onset of virus replication (Ojala et al., 2000). The cell lines Madin–Darby canine kidney (MDCK) and Caco-2, a human Caucasian colon adenocarcinoma cell line, can be induced to differentiate and form an epitheliallike monolayer when grown on porous filters for several days (Rothen-Rutishauser et al., 1998; Rousset, 1986; Simons & Virta, 1998). For this reason, they have been extensively used as models to study epithelial polarization and virus infection of epithelial cells (Compans, 1995). Previous studies have suggested that polarized MDCK and Caco-2 cells can be infected from both the apical and basolateral surface with HSV-1 or a replication-incompetent HSV-1 helper virus (Griffiths et al., 1998; Hayashi, 1995; Murphy et al., 1997; Sears et al., 1991; Topp et al., 1997; Tran et al., 2000). In highly polarized epithelial cells, microtubules are usually arranged in an apical–basal direction with their minus ends oriented towards the apical surface while the plus ends point towards the basal region of the cell (Grindstaff et al., 1998; Meads & Schroer, 1995; Rothen-Rutishauser et al., 1998). Microtubule minus-end-directed motors like cytoplasmic dynein (Fath et al., 1997; Lafont et al., 1994; Wang et al., 2003) or KIFC3 (Noda et al., 2001) transport cargo such as membrane vesicles towards the apical surface, whereas conventional kinesin catalyses transport to the basolateral surface (Lafont et al., 1994). Since HSV-1 virions and capsids are too large to be transported efficiently in the cytoplasm by diffusion (Sodeik, 2000), this raises the question of how incoming HSV-1 capsids reach the nuclear pores after inoculation of highly polarized epithelial cells from either the apical or basolateral plasma membrane. If HSV-1 fused with the apical plasma membrane and the cytosolic capsids bound to dynein and dynactin, they would be transported back to the apical surface rather than to the nucleus. Therefore, entry of HSV-1 by fusion with the apical plasma membrane would require a plus-end-directed microtubule motor for capsid transport to the nucleus. Alternatively, HSV-1 could enter by endocytosis and intact virions could use endocytic membrane traffic for transport from the apical surface to the nucleus. This scenario requires that viral fusion does not occur until the endosomes have moved close to the nuclear pores. Moreover, on entry from the basolateral surface, the minus-end-directed microtubule motor dynein 776

would transport capsids towards the apical compartment if the capsids did not detach from the microtubules in time to switch over to the nucleus, which is located approximately in the basolateral half of the cells. To address these questions, we set up polarized epithelial cell culture systems using MDCK and Caco-2 cells. In contrast to previous reports, our data showed that highly polarized epithelial cells were not susceptible to apical infection. However, HSV-1 infected these cells if added from the basolateral surface or if a depletion of extracellular Ca2+ had weakened the strength of cell–cell contacts. Basolateral infection and apical infection after a Ca2+ switch required an intact microtubule network. This system can now be used to identify the microtubule motors that HSV-1 uses in polarized epithelial cells.

METHODS Cells. All cell lines were maintained in plastic dishes or flasks as adherent cultures at 37 uC and 5 % CO2. MDCK-II (from Kai Simons), Vero (ATCC CCL-81) and baby hamster kidney (BHK-21) cells (ATCC CCL-10) were passaged twice a week and Caco-2 cells (from Ju¨rgen Wehland) once a week. To induce differentiation and full polarization, MDCK and Caco-2 cells were plated on polycarbonate permeable membrane filters with a diameter of 12 mm and a mean pore size of 0?4 or 3 mm (Transwell; Costar) at a density of 0?8–16105 cells cm22. The 3 mm filters were used to analyse HSV-1 infection from the basolateral surface. The medium of cells cultured on filters was changed daily. To open tight and adherens junctions, the polarized cells were washed with Ca2+- and Mg2+-free PBS (pH 7?4) and incubated at 37 uC with PBS containing 0?2 % (w/v) tissue-culture grade BSA and 10 mM EGTA to deplete extracellular Ca2+. The transepithelial electrical resistance (TER) was measured at various times after seeding with an electrical voltohmmeter with ‘chopstick’ electrodes (EVOM; World Precision Instruments; Simons & Virta, 1998). To study HSV-1 infection of unpolarized MDCK cells, the cells were seeded in 24-well culture plates with or without cover slips at a density of 4–56104 cells cm22. Antibodies. To monitor age-related changes of the epithelial cells,

we used an affinity purified rabbit antiserum to occludin (Zymed Laboratories), the mouse monoclonal antibody (mAb) 34 to E-cadherin (BD Biosciences Clontech), the mouse mAb DM1A (Sigma-Aldrich) to label tubulin, and mouse mAbs to gp114 as marker for the apical and to gp 58 for the basolateral surface (Balcarova-Stander et al., 1984). The major receptor for HSV-1, nectin-1, was detected with the mouse mAb CK6 (Krummenacher et al., 2000), the immediate-early HSV-1 protein ICP4 with the mouse monoclonal 58S (Showalter et al., 1981) and the late HSV-1 protein gB with a rabbit polyclonal serum R68 (Eisenberg et al., 1987). Filamentous actin was visualized using FITC–phalloidin (Sigma-Aldrich). Virological techniques. Wild-type HSV-1 strain F (ATCC VR-

733), wild-type HSV-1 strain 17+ (from John Subaq-Sharpe) and the b-galactosidase-expressing strain [KOS]tk12 (Warner et al., 1998), which expresses the bacterial lacZ gene encoding the enzyme b-galactosidase under the control of the immediate-early ICP4 promoter of HSV-1, were amplified in BHK cells, purified and titrated in Vero cells as described (Do¨hner et al., 2002; Sodeik et al., 1997). To depolymerize microtubules, the cells were treated with nocodazole (Sigma-Aldrich,) for 1 h prior to virus infection, then kept in Journal of General Virology 85

HSV infection of polarized epithelial cells nocodazole for the duration of the experiment. To deplete extracellular Ca2+, cells were incubated with 10 mM EGTA at 37 uC for 30 or 45 min prior to the addition of virus. Cells were infected with virus diluted in CO2-independent culture medium (Gibco Life Technologies) supplemented with 0?2 % (w/v) cell-culture grade BSA (Sigma-Aldrich). The virus suspension was added at 0?5 ml per well or filter to the cells for 1 h on ice for cells grown on plastic and at room temperature for cells grown on filters, since the latter could not be maintained well at 4 uC. After inoculation, the virus was removed and the cell dishes were incubated in a water bath at 37 uC in CO2-independent culture medium supplemented with 10 % foetal calf serum for another 4?5 h. Galactosidase assay. Immediate-early viral gene expression was

quantified using the mutant HSV1[KOS]tk12 as described (Mabit et al., 2002). The amount of b-galactosidase was determined after lysis in 0?5 % (v/v) TX-100/PBS with 1 mg BSA ml21 and protease inhibitors using O-nitrophenyl b-D-galactosylpyranoside as a substrate. The lysate was incubated with the substrate for about 2 h at room temperature and the enzymic activity (A420) was measured using a plate reader (Spectra Count Microplate Photometer; Packard Instruments Company). The cell density was estimated with a parallel set of plates or filters by staining fixed cells with 0?25 mg crystal violet ml21 in 5 % (v/v) ethanol for 5 min. After drying, bound crystal violet was dissolved in 100 % ethanol and the A590 was read. Microscopy. Cells grown on cover slips were fixed and labelled

essentially as described previously (Do¨hner et al., 2002; Sodeik et al., 1997). Cells grown on membrane filters were fixed with 3 % (w/v) paraformaldehyde in PBS for 30 min followed by treatment with 50 mM NH4Cl in PBS for 20 min and 0?1 % Triton X-100 in PBS for 10 min. Non-specific antibody binding was quenched using 0?2 % (v/v) cold-water-fish skin gelatin (Sigma-Aldrich) and 0?5 % (w/v) BSA in PBS for 30–60 min prior to labelling with the antibodies, diluted in the same buffer, for 30–60 min at room temperature. Where indicated, nuclei were labelled with 20 mg propidium iodide ml21 after treating the permeabilized cells with 1 mg RNase A ml21 (Sigma-Aldrich) in PBS for 10 min at room temperature (Reinsch et al., 1998). Affinity-purified secondary antibodies were purchased from Dianova. The cells were mounted in Moviol containing 50–100 mg 1,4-diazabicyclo(2,2,2)octane (DABCO) ml21 or 25 mg N-propylgallate ml21 and examined with a fluorescence microscope equipped for laser-scanning confocal light microscopy (DM IRB/E; Leica). Optical sections were recorded using a 1006 oil immersion objective with a numerical aperture of 1?4 at a resolution of 5126512 pixels. Digitalized images were further processed using Adobe Photoshop version 4.0.

RESULTS Differentiation of epithelial cells in culture Fully differentiated epithelial cells in situ differ greatly from the non-polarized cultured cells that are commonly used to study HSV-1 infection. However, certain cell lines mature into a highly polarized epithelial monolayer if grown on porous filters that allow access to the culture medium from both the apical and basolateral plasma membrane. Such cells form a sealed epithelium with closed tight junctions and develop a considerable electrical resistance (Rousset, 1986; Simons & Virta, 1998). MDCK cells were seeded on permeable polycarbonate membrane filters and the transepithelial electrical resistance (TER) was monitored continuously to assess the degree of differentiation during http://vir.sgmjournals.org

culture (Fig. 1A, filled symbols). The TER was lowest at day 1 and peaked after 2 days at up to 400 Ohm?cm2. Around day 4, the cultures reached a stable TER above 160 Ohm?cm2, which was maintained until day 9, the longest time point analysed. Confocal laser scanning microscopy demonstrated that, after 2 days on filters, the tight junction protein occludin was already localized in many cells in a ring along the borders of the MDCK cells (data not shown). Nectin-1, the major receptor for HSV-1 gD (Fig. 1B) and E-cadherin (data not shown), both markers for adherence junctions, appeared in a wrinkled solid line at the sites of cell–cell contact. As polarity markers, we used specific antibodies to the glycoproteins gp114 and gp58 (Balcarova-Stander et al., 1984). After 2 days of filter culture, few cells expressed gp114 (Fig. 1C) and pg58 was localized to the basolateral as well as to the apical plasma membrane (Fig. 1D). After 5 days on filters, gp114 was expressed in all cells and had been correctly targeted to the apical (Fig. 1E) and gp58 to the lateral plasma membrane (Fig. 1F), suggesting that cell differentiation and polarization in culture had been completed. These experiments showed that the peak in TER resistance at day 2 corresponded with the formation of a confluent monolayer of epithelial cells of which some already showed a polarized distribution of integral plasma membrane proteins. Around day 4–5, the TER was stabilized and the separation, as well as polarization, of apical and basolateral plasma membrane domains had been completed. These observed changes of the TER over the culture period agreed well with results from Rothen-Rutishauser et al. (1998). HSV-1 infection of unpolarized MDCK cells Several laboratories have reported the infection of polarized epithelial cells with HSV-1 after virus challenge from both the basolateral and the apical plasma membrane (Griffiths et al., 1998; Sears et al., 1991). When we used a viral mutant expressing the lacZ gene under the control of an immediate-early HSV-1 promoter (Warner et al., 1998) to infect our clone of MDCK cells cultured on plastic dishes, there was a dose-dependent increase in the synthesis of galactosidase (Fig. 2A). If, however, the cells were infected in the presence of nocodazole, which reversibly depolymerizes microtubules, the synthesis of b-galactosidase decreased in a concentration-dependent manner (Fig. 2B). Infecting cells grown on cover slips with wild-type HSV-1 for 4 h resulted in a prominent nuclear labelling for the immediate-early protein ICP4. Interestingly, the cells were not homogeneously infected; HSV-1 showed a clear preference for isolated cells. As also reported by Schelhaas et al. (2003), the peripheral cells of an MDCK islet contained more ICP4 when compared with cells completely surrounded by other cells in a confluent region (Fig. 2C). When MDCK cells grown on porous filters for 2 days were infected with virus added to the apical culture chamber, the 777


400

TER (Ohm.cm2)

320

240

160

80 0 A

2 4 6 Time post-seeding (days)

8 B

C

D

E

F

778

Journal of General Virology 85

HSV infection of polarized epithelial cells

Fig. 1. Differentiation of MDCK epithelial cells in culture. (A) MDCK cells were seeded on to 0?4 mm filters and supplemented with medium on the apical and basolateral surfaces. The transepithelial electrical resistance (TER) of the cell monolayer (filled symbols) was measured in triplicate every day in three independent experiments. As a control, we also determined the TER of filters without cells (open symbols). The TER peaked 2 days after seeding and reached a stable TER of about 160–180 Ohm?cm2 around day 4. (B–E) Laser-scanning confocal microscopy of MDCK cells cultured on filters for 2 (B, C, D) or 5 (E, F) days, fixed and labelled with antibodies (green) directed against the HSV-1 receptor nectin-1 (B), the apical marker protein gp114 (C, E) or the basolateral marker protein gp58 (D, F). Propidium iodide (red) was used to label the nuclei. The right panels in (D)–(F) show xz scans perpendicular to the filters, whereas (C) and the left panels of (D) and (E) show xy scans more or less parallel to the filters. After 2 days on filters, nectin-1 was localized at the plasma membrane in one or two optical sections just above the nuclei representing the area of the adherens junctions. Moreover, there was a prominent labelling of internal vesicles. The nectin-1 antibody showed some cross-reactivity with the nuclei, as exemplified in a mitotic cell in the top row of cells (yellow) in (B). In younger cultures (C, D), gp114 (C) was expressed in few cells at the apical surface, whereas gp58 (D) was mainly localized in the basal and lateral plasma membrane, but many cells also showed an apical localization of gp58. In addition, many cytoplasmic vesicles contained gp58. In older cultures (E, F), all cells expressed pg114 on the apical plasma membrane domain and pg114 was almost exclusively expressed on the lateral and, albeit more weakly, on the basal but not on the apical domain.

amount of b-galactosidase synthesized increased with increasing m.o.i. (Fig. 3A). When cells were infected in the presence of nocodazole, virus infection was again suppressed in a dose-dependent manner (Fig. 3B). However, the effect of nocodazole was not as strong as in cells grown in plastic dishes. The reasons for this are unclear and were not further characterized. It is possible that the microtubules were more stable after culture for 2 days on filters relative to 1 day in plastic dishes. Moreover, the cells grown on plastic were kept on ice during virus inoculation and microtubules depolymerize more readily in the cold. Single-cell analysis using confocal laser-scanning microscopy demonstrated that most cells showed a strong nuclear labelling for ICP4 (Fig. 3C), unless they had been treated with nocodazole (data not shown). Thus, our clone of MDCK cells was permissive for HSV-1 infection. Moreover, efficient viral gene expression was shown to require an intact microtubule network, as has been shown for the epithelial cell lines Vero and PtK2 (Mabit et al., 2002; Sodeik et al., 1997). Basolateral infection of polarized epithelial cells requires microtubules We next wanted to analyse the infection of fully polarized MDCK cells via the apical plasma membrane. However, we could detect only very low levels of b-galactosidase synthesis when cells cultured on filters for 7 days were infected with HSV-1 over a wide range of multiplicities (not shown). Confocal laser-scanning microscopy analysis revealed a few isolated cells labelled for ICP4, but overall there were few signs of an HSV-1 infection (Fig. 4A). As previously reported, we were also unable to infect MDCK cells grown on 0?4 mm pore filters from the basal chamber with herpesvirus (Hemmings & Guilbert, 2002; Topp et al., 1997). If, however, MDCK cells were grown for 8 days on filters with a larger pore diameter of 3 mm, which allowed passage of HSV-1 virions, and infected from the basolateral chamber, the cells were clearly permissive for HSV-1 http://vir.sgmjournals.org

infection and many were labelled by antibodies against ICP4, an immediate-early HSV-1 protein, and gB, a late structural protein (Fig. 4B). Infection via the basolateral plasma membrane was also inhibited by nocodazole in a dose-dependent manner (Fig. 4C). Thus, MDCK cells differentiated on filters were susceptible to infection, but only if HSV-1 had access to the basolateral surface of the cells and not if the virus was added to the apical chamber. Apical infection of fully polarized MDCK cells requires access to microtubules and to receptors present at cell–cell contact sites Previous experiments on unpolarized MDCK cells suggested that susceptibility to HSV-1 could be increased if cell–cell contacts were opened by the depletion of extracellular Ca2+ (Hayashi, 1995; Topp et al., 1997; Yoon & Spear, 2002). We therefore incubated fully polarized MDCK cells grown on filters for 8 days with EGTA to sequester extracellular Ca2+ and challenged them from the apical chamber with HSV-1. After 30 min of Ca2+ depletion, and more so after 45 min, the cells had become susceptible to HSV-1 and showed robust b-galactosidase expression (Fig. 5A). Confocal laser-scanning microscopy using wild-type HSV-1 showed many cells labelled for the immediate-early protein ICP4 and the late structural protein gB (Fig. 5B). This apical infection also required an intact microtubule network, as it was inhibited by nocodazole in a dose-dependent manner (Fig. 5C). The depletion of extracellular Ca2+ leads to the disruption of both types of cell– cell contacts, the tight junctions as well as the adherens junctions (Rothen-Rutishauser et al., 2002). Confocal laserscanning microscopy showed that the EGTA treatment also led to an overall increase in the signal for nectin-1, the major HSV-1 receptor, but not to a dramatic relocalization of nectin-1 to intracellular membranes (Fig. 5D). Since the cells were treated for only a very short time, it seems likely that the incubation with EGTA improved the accessibility of the antibody to nectin-1 present at the plasma membrane but did not cause an increased synthesis of nectin-1. 779


0.1 0.5

Galactosidase Crystal violet 0 A

0 4

0

8

HSV-1 ( 105 p.f.u. per well)

Crystal violet (A)

Galactosidase (A)

1

We also measured the TER during the course of the experiment. The initial TER of approximately 160 Ohm?cm2 dropped to 90 after the EGTA treatment, but then returned to approximately 120 after HSV-1 inoculation for 1 h at room temperature in the presence of Ca2+ and remained there until the cells were harvested after further incubation for another 4?5 h at 37 uC. Cells treated in the same way but omitting the virus showed an identical progression of TER. Labelling the actin cytoskeleton with phalloidin revealed that after EGTA treatment there were holes in the cell monolayer, demonstrating that the cells could no longer maintain their lateral cell–cell contacts (data not shown; but see also Fig. 5D). Confocal laser-scanning light microscopy also showed that the depletion of extracellular Ca2+ did not lead to a major reorganization of the microtubule network (data not shown). There was still a basolateral microtubule network, peripheral microtubule bundles in the focal plane of the nuclei and an apical microtubule network, as described for fully polarized MDCK cells (Grindstaff et al., 1998; Meads & Schroer, 1995; Rothen-Rutishauser et al., 1998).

100

Apical infection of fully polarized human Caco-2 cells is facilitated by depleting extracellular Ca2+

% of control

75

The lack of susceptibility of the canine MDCK cells once they are fully polarized to apical infection with the human pathogen HSV-1 might be due to a host species barrier. Therefore, we performed additional experiments using the human cell line Caco-2, which also differentiates and polarizes if grown on porous filters (Rousset, 1986). Initial experiments using cells grown on plastic showed that our

50 Crystal violet Galactosidase 25

0 B

0

10

20

30

[Nocodazole] (mM)

C

780

40

50

Fig. 2. Subconfluent unpolarized MDCK cells are more susceptible to HSV-1 than confluent monolayers and the infection is less efficient without microtubules. (A) MDCK cells grown in 24-well dishes were infected with HSV1[KOS]tk12 at increasing m.o.i.s and immediate-early viral gene expression was measured at 4 h post-infection by quantifying the amount of b-galactosidase (&; in triplicate) synthesized from the lacZ gene under the control of an immediate-early HSV-1 promoter. The cell density was estimated from parallel plates stained with crystal violet (%; in triplicate). Maximal virus infection was achieved at 1?56105 p.f.u. cm”2 (36105 p.f.u. per well) without any significant cell loss. (B) Infection of unpolarized MDCK cells was inhibited in the presence of increasing concentrations of nocodazole, which depolymerizes microtubules. The assay was performed in triplicate. (C) Single-cell analysis using laserscanning immunofluorescence microscopy. Viral gene expression of wild-type HSV-1 (1?56107 p.f.u. per well), detected with an antibody to the immediate-early HSV-1 protein ICP4 (green), started earlier in cells with exposed free cell margins than in cells in a confluent cell islet that were completely surrounded by other cells. The inoculum was labelled with an antibody directed against HSV-1 gB (red). To visualize all cells, nuclei were labelled with propidium iodide (red). Journal of General Virology 85

HSV infection of polarized epithelial cells 0.4

A

0.2 1 Galactosidase Crystal violet

0.0

Crystal violet (A)

Galactosidase (A)

2

0 0

5

10

15

Caco-2 cell clone was fully permissive to HSV-1 infection and synthesized large amounts of galactosidase (Fig. 6A). We next monitored cell differentiation during culture on porous filters. Caco-2 cells developed differently compared with the MDCK cells (cf. Fig. 1A). The TER of the Caco-2 cells was very low at day 1, but steadily rose to reach 300 Ohm?cm2 by day 13 (Fig. 6B). The development of TER was similar to the results reported by Griffiths et al. (1998) for Caco-2 cells, but the absolute overall resistance was lower in our experiments. Infection of this Caco-2 clone with HSV-1 from the apical chamber yielded similar results to those obtained with the MDCK cells. The Caco-2 cells were susceptible to HSV-1 infection after culture for 6 days but not for 12 days. If, however, the cells were treated with EGTA to deplete Ca2+, they became susceptible to HSV-1 infection from the apical surface (Fig. 6C).

HSV-1 ( 106 p.f.u. per well)

DISCUSSION

% of control

100

B

Sears et al. (1991) were the first to analyse HSV-1 infection of MDCK cells grown on porous filters. They reported that different receptors were used for infection via the apical and the basolateral plasma membrane and, moreover, that apical infection involved viral gC. However, Griffiths et al. (1998) could not detect any difference in the apical adsorption when they compared gC-deficient mutants with wildtype strains. These studies, and also Hayashi (1995), Topp et al. (1997) and Tran et al. (2000), suggested that polarized MDCK cells are infected with HSV-1 via the apical plasma membrane. We therefore started this study to analyse the role of microtubules in cytoplasmic transport from the apical plasma membrane to the nucleus, with the overall aim of identifying the microtubule motors responsible for nuclear targeting.

Crystal violet Galactosidase

50

0 0

10

20 [Nocodazole] (mM)

C

http://vir.sgmjournals.org

30

40

Fig. 3. Infection of MDCK cells cultured on porous membrane filters for 2 days. (A) MDCK cells grown on 0?4 mm pore filters for 2 days were infected from the apical chamber with HSV1[KOS]tk12 at increasing m.o.i.s and immediate-early viral gene expression was measured by quantifying the amount of bgalactosidase (&; in triplicate) synthesized from the lacZ gene under the control of an immediate-early HSV-1 promoter. The cell density was estimated from parallel plates stained with crystal violet (%; in triplicate). Maximal virus infection was achieved at 86106 p.f.u. cm”2 without any significant cell loss. (B) The infection of partially polarized MDCK cells with 26107 p.f.u. cm”2 was inhibited in the presence of increasing concentrations of nocodazole, which depolymerizes microtubules. The assay was performed in triplicate. (C) Single-cell analysis using laser-scanning immunofluorescence microscopy showing strong viral gene expression detected with antibodies to the immediate-early HSV-1 protein ICP4 (green) and to the late HSV-1 protein gB (red) in many cells. MDCK cells grown on 0?4 mm pore filters for 2 days were infected from the apical chamber with wild-type HSV-1 at a concentration of 86107 p.f.u. cm”2 for 1 h at room temperature and incubated for a further 4?5 h at 37 6C. 781


However, our data showed that fully polarized MDCK cells could not be infected with HSV-1 added to the apical culture chamber. We first determined that our clone of MDCK cells was fully susceptible to HSV-1 by showing that unpolarized cells were easily infected by HSV-1, irrespective of whether they were grown on a plastic surface, on glass cover slips or on filters for 2 days. We then used two assays to assess the differentiation state of the filter cultures: the development of the TER and the subcellular localization of integral plasma membrane proteins known to develop a polarized distribution. According to these criteria, we focused our experiments on MDCK cells that had either been cultured on filters for 2 days and had not yet developed two separated plasma membrane domains, or had been cultured for longer than 5 days, which we classified as maximally polarized. The latter is most likely the cause of major differences compared with previously published reports, which used younger epithelial cultures on filters for their experiments.

A

As reported previously for the epithelial Vero and PtK2 cell lines (Do¨hner et al., 2002; Mabit et al., 2002; Sodeik et al., 1997), efficient infection of unpolarized as well as highly polarized MDCK cells from the basolateral or apical plasma membrane after Ca2+ depletion required an intact microtubule network. Nocodazole, a reversible inhibitor of microtubule assembly (Jordan & Wilson, 1999), at a concentration of 33 mM was sufficient to completely depolymerize the MDCK microtubule network under all culture conditions used (data not shown; but see also Grindstaff et al., 1998; Lafont et al., 1994). There are numerous reports that nocodazole does not have non-specific effects on protein synthesis in polarized epithelial cells (see, for example, Eilers et al., 1989; Grindstaff et al., 1998). Moreover, nocodazole does not affect either HSV-1 binding to B

100

% of control

75

Galactosidase

50

Crystal violet 25

0 C

0

10


782

30

40

Fig. 4. Basolateral infection of fully polarized MDCK cells. (A) MDCK cells grown on 0?4 mm pore filters for 7 days were infected from the apical chamber with wild-type HSV-1 at a concentration of 86107 p.f.u. cm”2 for 1 h at room temperature and incubated for a further 4?5 h at 37 6C. Very few cells were infected, as shown by single-cell analysis using laser-scanning immunofluorescence microscopy with antibodies to the immediate-early HSV-1 protein ICP4 (green). (B) MDCK cells were grown on 3 mm pore filters for 8 days and infected from the basal chamber with wild-type HSV-1 at a concentration of 86107 p.f.u. cm”2 for 1 h at room temperature and incubated for a further 4?5 h at 37 6C. Single-cell analysis using laserscanning immunofluorescence microscopy showed a strong viral gene expression detected with antibodies to ICP4 (green) and gB (red) in many cells. (C) Infection of MDCK cells grown on 3 mm pore filters for 7 days with HSV1[KOS]tk12 at a concentration of 2?46107 p.f.u. cm”2 was inhibited in the presence of increasing concentrations of nocodazole, which depolymerizes microtubules. b-Galactosidase activity (&) and cell density (%) were measured in triplicate. Journal of General Virology 85


Galactosidase Crystal violet

% of control

100

50

0 Control

A

30 min

B

45 min

125

% of control

100 Galactosidase Crystal violet

75

50

25

0 0 C

10


30

40 D

Fig. 5. Apical infection of fully polarized MDCK cells after Ca2+ depletion. (A) MDCK cells grown on 0?4 mm pore filters for 8 days were treated with EGTA for 30 or 45 min to deplete extracellular Ca2+, then infected from the apical chamber with HSV1[KOS]tk12 with a concentration of 3?26107 p.f.u. cm”2. Immediate-early viral gene expression was measured by quantifying the amount of b-galactosidase (grey bars; in triplicate) synthesized from the lacZ gene under the control of an immediate-early HSV-1 promoter. The cell density was estimated from parallel plates stained with crystal violet (white bars; in triplicate). If the cells were not treated with EGTA (control), viral gene expression was barely detectable, whereas prior treatment with EGTA made the cells susceptible to HSV-1 infection. (B) Single-cell analysis using laser-scanning immunofluorescence microscopy of MDCK cells grown on 0?4 mm pore filters for 7 days showing a strong viral gene expression detected with antibodies to the immediate-early HSV-1 protein ICP4 (green) and to the late HSV-1 protein gB (red) in many cells (compare with Fig. 4A). The cells were treated with EGTA for 30 min and then infected from the apical chamber with wild-type HSV-1 at a concentration of 86107 p.f.u. cm”2 for 1 h at room temperature and incubated for a further 4?5 h at 37 6C. (C) The apical infection of fully polarized MDCK cells that had been cultured on 0?4 mm pore filters for 7 days and treated with EGTA for 30 min, then infected with HSV1[KOS]tk12 at a concentration of 1?66107 p.f.u. cm”2 was inhibited in the presence of increasing concentrations of nocodazole, which depolymerizes microtubules. b-Galactosidase activity (&) and cell density (%) were measured in triplicate. (D) Laser-scanning immunofluorescence microscopy showing that the major HSV-1 receptor, endogenous nectin-1, is localized in MDCK cells cultured on 0?4 mm pore filters for 5 days at the plasma membrane in one or two optical sections just above the nuclei (top panels). The depletion of extracellular Ca2+ by incubating the cells with EGTA for 30 min led to an increase in the signal for nectin-1 but not to a relocalization of the HSV-1 receptor to internal membranes (bottom panels). The antibody to nectin-1 showed some cross-reactivity with the nucleus (see also Fig. 2B). http://vir.sgmjournals.org

783


cells or internalization, but reduces cytosolic capsid transport to the nucleus (Sodeik et al., 1997). In the MDCK cells analysed in this study, nocodazole most likely inhibited either the cytoplasmic transport of endocytosed HSV-1 or the cytosolic HSV-1 capsid transport. Both scenarios would reduce the number of viral genomes imported into the nucleus and, consequently, immediate-early gene transcription, as well as protein expression, which we quantified using a reporter virus expressing b-galactosidase. Interestingly, HSV-1 infected single MDCK cells or peripheral cells of a cell islet rather than confluent cells. In older cultures on filters, there were very few cells that showed any viral protein synthesis. However, this was not due to cell death since the cells were highly susceptible if the virus had access to the basolateral surface of cultures grown on large-pore filters or if the tight junctions and the adherens junctions had been opened experimentally by depleting extracellular Ca2+. One might argue that the canine MDCK cell line is not the best model to study human viruses. However, we obtained similar results with the human Caco-2 cell line. Again, less polarized cells were more readily infected from the apical chamber than fully polarized cells. Moreover, infection experiments after pre-incubation with EGTA demonstrated that the older cultures were also fully susceptible to HSV-1 infection if the cell–cell contacts had been opened. As the cultures differentiate, two barriers for apical HSV-1 infection develop. One is that the major HSV-1 gD receptor, Fig. 6. Apical infection of Caco-2 cells after Ca2+ depletion. (A) Caco-2 cells grown in 24-well plastic dishes were infected with HSV1[KOS]tk12 at increasing m.o.i.s for 4 h and immediateearly viral gene expression was measured by quantifying the amount of b-galactosidase (&; in triplicate) synthesized from the lacZ gene under the control of an immediate-early HSV-1 promoter. Cell density was estimated from parallel plates stained with crystal violet (open symbols; in triplicate). Maximal virus infection was achieved at 16106 p.f.u. cm”2 (26106 p.f.u. per well) without any significant cell loss. (B) Caco-2 cells were seeded on to 0?4 mm filters and supplemented with medium on the apical and basolateral surfaces. The transepithelial electrical resistance (TER) of the cell monolayer (&; in triplicate) was measured over a time period of 13 days. As a control, we also determined the TER of filters without cells (%; in triplicate). The TER reached an initial plateau 2 days after seeding and remained at 160 Ohm?cm2 until day 5. At 10– 13 days, there was a second plateau around 300 Ohm?cm2. (C) Caco-2 cells cultured on 0?4 mm filters were infected from the apical chamber with HSV1[KOS]tk12 at a concentration of 86106 p.f.u. cm”2 for 4 h. Cells seeded on to filters for 6 days showed a significant viral gene expression, while older cultures that had been cultivated on filters for 12 days were not infected (grey bars; in triplicate). However, at both time points, there was a dramatic increase in viral gene expression if the cells were pre-treated with EGTA to deplete extracellular Ca2+ and open the cell–cell contacts (white bars; in triplicate). 784



nectin-1, is increasingly sequestrated into adherens junctions. The second is the formation of tight junctions above the adherens junctions, which form such a tight seal between neighbouring cells that even small molecules such as sugars can hardly pass (Takai & Nakanishi, 2003). The second seems to be the major obstacle, since polarized cells were infected if the virus was able to access the cells from the basolateral chamber via large-pore filters. Moreover, heparan sulfate proteoglycans, important receptors for HSV-1 gC, are mainly localized in the basolateral plasma membrane (Caplan et al., 1987), while the apical plasma membrane contains chondroitin sulfate proteoglycans (Kolset et al., 1999), which cannot be used as efficiently by HSV-1 (Mardberg et al., 2002).

REFERENCES

Our experiments are in agreement with and extend the studies of Yoon & Spear (2002) and Schelhaas et al. (2003). Even MDCK cells grown to confluency on glass or plastic are rather resistant to apical HSV-1 infection until the monolayer has been wounded or the cell–cell contacts weakened by cytochalasin treatment (Schelhaas et al., 2003). Moreover confluent MDCK cells, stably transfected with human nectin-1 and grown on plastic or glass, bind more soluble gD and are more efficiently infected with HSV-1 when the adherens junctions have been disrupted by Ca2+ depletion (Yoon & Spear, 2002). These results and ours with highly polarized MDCK cells expressing endogenous nectin-1 suggest that, under steady-state conditions, most of the nectin-1 molecules are engaged in homotypic interactions. More nectin-1 receptors become available for virus binding if the depletion of extracellular Ca2+ releases these interactions. Thus, the entry of HSV-1 into polarized cells and infection require access to microtubules and to receptors present in the adherens junction. Our data suggest that an intact, fully polarized epithelium that has established tight junctions is not susceptible to apical HSV-1 infection. This supports the hypothesis proposed by Spear (2002) that a primary infection might only occur via small wounds or via undifferentiated and thus unpolarized cells present in the mucosa.

Eilers, U., Klumpermann, J. & Hauri, H. P. (1989). Nocodazole, a

ACKNOWLEDGEMENTS This project was funded by a program project grant on ‘Mucosal host pathogen interactions’ from the German Research Council (DFGGraduierten-Kolleg 745). We thank Doris Meder and Kai Simons (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany) for providing the MDCK cells as well as gp114 and gp58 antibodies, and Doris also for many helpful discussions on their use. Ju¨rgen Wehland (German Research Centre for Biotechnology, Braunschweig, Germany) kindly provided the Caco-2 cells. We are grateful to Pat Spear (Northwestern University, Chicago, IL, USA) for providing HSV1[KOS]tk12, to John Subaq-Sharpe (University of Glasgow, UK) for HSV-1 strain 17+ and to Roger Everett for antibodies to ICP4 (University of Glasgow, UK), as well as to Roslyn Eisenberg and Gary Cohen for antibodies to gB and nectin-1 (University of Pennsylvania, Philadelphia, PA, USA). Thomas Schulz (Virology, Hannover Medical School) and Rudi Bauerfeind (Cell Biology, Hannover Medical School) are acknowledged for helpful comments on the manuscript. http://vir.sgmjournals.org

Balcarova-Stander, J., Pfeiffer, S., Fuller, S. & Simons, K. (1984).

Development of cell surface polarity in the epithelial Madin–Darby canine kidney (MDCK) cell line. EMBO J 3, 2687–2694. Caplan, M., Stow, J., Newman, A., Madri, J., Anderson, H., Farquhar, M., Palade, G. & Jamieson, J. (1987). Dependence of

pH of polarized sorting of secreted proteins. Nature 329, 632–635. Compans, R. W. (1995). Virus entry and release in polarized

epithelial cells. Curr Top Microbiol Immunol 202, 209–219. Do¨hner, K. & Sodeik, B. (2004). The role of the cytoskeleton during

viral infection. Curr Top Microbiol Immunol 285 (in press). Do¨hner, K., Wolfstein, A., Prank, U., Echeverri, C., Dujardin, D., Vallee, R. & Sodeik, B. (2002). Function of dynein and dynactin in

herpes simplex virus capsid transport. Mol Biol Cell 13, 2795–2809. microtubule-active drug, interferes with apical protein delivery in cultured intestinal epithelial cells (Caco-2). J Cell Biol 108, 13–22. Eisenberg, R. J., Ponce de Leon, M., Friedman, H. M., Fries, L. F., Frank, M. M., Hastings, J. C. & Cohen, G. H. (1987). Complement

component C3b binds directly to purified glycoprotein C of herpes simplex virus types 1 and 2. Microb Pathog 3, 423–435. Enquist, L. W., Husak, P. J., Banfield, B. W. & Smith, G. A. (1998).

Infection and spread of alphaherpesviruses in the nervous system. Adv Virus Res 51, 237–347. Fath, K., Trimbur, G. & Burgess, D. (1997). Molecular motors and a

spectrin matrix associate with Golgi membranes in vitro. J Cell Biol 139, 1169–1181. Geraghty, R. J., Krummenacher, C., Cohen, G. H., Eisenberg, R. J. & Spear, P. G. (1998). Entry of alphaherpesviruses mediated by

poliovirus receptor-related protein 1 and poliovirus receptor. Science 280, 1618–1620. Griffiths, A., Renfrey, S. & Minson, T. (1998). Glycoprotein C-

deficient mutants of two strains of herpes simplex virus type 1 exhibit unaltered adsorption characteristics on polarized or nonpolarized cells. J Gen Virol 79, 807–812. Grindstaff, K. K., Bacallao, R. L. & Nelson, W. J. (1998). Apiconuclear

organization of microtubules does not specify protein delivery from the trans-Golgi network to different membrane domains in polarized epithelial cells. Mol Biol Cell 9, 685–699. Hayashi, K. (1995). Role of tight junctions or polarized epithelial

MDCK cells in the replication of herpes simplex virus type 1. J Med Virol 47, 323–329. Hemmings, D. & Guilbert, L. (2002). Polarized release of human cytomegaloviurs from placental trophoblasts. J Virol 76, 6710–6717. Jordan, M. A. & Wilson, L. (1999). The use and action of drugs in

analyzing mitosis. Methods Cell Biol 61, 267–295. Kolset, S., Vuong, T. & Prydz, K. (1999). Apical secretion of

chondroitin sulphate in polarized Madin–Darby canine kidney cells. J Cell Sci 112, 1797–1801. Krummenacher, C., Baribaud, I., de Leon, M. P., Whitbeck, J. C., Lou, H., Cohen, G. H. & Eisenberg, R. J. (2000). Localization of a

binding site for herpes simplex virus glycoprotein D on herpesvirus entry mediator C by using antireceptor monoclonal antibodies. J Virol 74, 10863–10872. Lafont, F., Burkhardt, J. K. & Simons, K. (1994). Involvement of

microtubule motors in basolateral and apical transport in kidney cells. Nature 372, 801–803. Mabit, H., Nakano, M. Y., Prank, U., Saam, B., Do¨hner, K., Sodeik, B. & Greber, U. F. (2002). Intact microtubules support adenovirus and

herpes simplex virus infections. J Virol 76, 9962–9971. 785


Mardberg, K., Trybala, E., Tufaro, F. & Bergstrom, T. (2002).

Herpes simplex virus type 1 glycoprotein C is necessary for efficient infection of chondroitin sulfate-expressing gro2C cells. J Gen Virol 83, 291–300.

cell receptors interact with different viral proteins. Proc Natl Acad Sci U S A 88, 5087–5091. Showalter, S. D., Zweig, M. & Hampar, B. (1981). Monoclonal

Meads, T. & Schroer, T. (1995). Polarity and nucleation of micro-

antibodies to herpes simplex virus type 1 proteins, including the immediate-early protein ICP 4. Infect Immun 34, 684–692.

tubules in polarized epithelial cells. Cell Motil Cytoskeleton 32, 272–288.

Shukla, D., Liu, J., Blaiklock, P. & 7 other authors (1999). A novel

Mettenleiter, T. C. (2002). Herpesvirus assembly and egress. J Virol

role for 3-O-sulfated heparan sulfate in herpes simplex virus 1 entry. Cell 99, 13–22.

76, 1537–1547. Montgomery, R. I., Warner, M. S., Lum, B. J. & Spear, P. G. (1996).

Herpes simplex virus-1 entry into cells mediated by a novel member of the TNF/NGF receptor family. Cell 87, 427–436. Murphy, C., Zacchi, P., Parton, R., Zerial, M. & Lim, P. (1997).

Simons, K. & Virta, H. (1998). Growing Madin–Darby canine kidney cells for studying epithelial cell biology. In Cell Biology: a Laboratory Handbook, pp. 101–106. Edited by J. E. Celis. San Diego: Academic Press.

HSV infection of polarized epithelial cells on filter supports: implications for transport assays and protein localization. Eur J Cell Biol 72, 278–281.

Sodeik, B. (2000). Mechanisms of viral transport in the cytoplasm.

Nicola, A. V., McEvoy, A. M. & Straus, S. E. (2003). Roles for

mediated transport of incoming herpes simplex virus 1 capsids to the nucleus. J Cell Biol 136, 1007–1021.

endocytosis and low pH in herpes simplex virus entry into HeLa and Chinese hamster ovary cells. J Virol 77, 5324–5332. Noda, Y., Okada, Y., Saito, N., Setou, M., Xu, Y., Zhang, Z. & Hirokawa, N. (2001). KIFC3, a microtubule minus end-directed

motor for the apical transport of annexin XIIIb-associated Tritoninsoluble membranes. J Cell Biol 155, 77–88. Ojala, P. M., Sodeik, B., Ebersold, M. W., Kutay, U. & Helenius, A. (2000). Herpes simplex virus type 1 entry into host cells: recon-

stitution of capsid binding and uncoating at the nuclear pore complex in vitro. Mol Cell Biol 20, 4922–4931. Reinsch, S., Eaton, S. & Stelzer, E. (1998). Confocal microscopy of

polarized epithelial cells. In Cell Biology: a Laboratory Handbook, pp. 170–178. Edited by J. E. Celis. San Diego: Academic Press. Roizman, B. & Knipe, D. M. (2001). Herpes simplex viruses and their

replication. In Fundamental Virology, 4th edn, pp. 1123–1183. Edited by D. M. Knipe & P. M. Howley. Philadelphia: Lippincott Williams & Wilkins.

Trends Microbiol 8, 465–472. Sodeik, B., Ebersold, M. W. & Helenius, A. (1997). Microtubule-

Spear, P. G. (2002). Viral interactions with receptors in cell

junctions and effects on junctional stability. Dev Cell 3, 462–464. Spear, P. G., Eisenberg, R. J. & Cohen, G. H. (2000). Three classes of

cell surface receptors for alphaherpesvirus entry. Virology 275, 1–8. Takai, Y. & Nakanishi, H. (2003). Nectin and afadin: novel organizers

of intercellular junctions. J Cell Sci 116, 17–27. Topp, K. S., Rothman, A. L. & Lavail, J. H. (1997). Herpes virus

infection of RPE and MDCK cells: polarity of infection. Exp Eye Res 64, 343–354. Tran, L., Kissner, J., Westerman, L. & Sears, A. E. (2000). A herpes

simplex virus 1 recombinant lacking the glycoprotein G coding sequences is defective in entry through apical surfaces of polarized epithelial cells in culture and in vivo. Proc Natl Acad Sci U S A 97, 1818–1822. Wagner, E. K. & Bloom, D. C. (1997). Experimental investigation of

herpes simplex virus latency. Clin Microbiol Rev 10, 419–443.

Rothen-Rutishauser, B., Kra¨mer, S., Braun, A., Gu¨nthert, M. & Wunderli-Allenspach, H. (1998). MDCK cell cultures as an epithelial

Wang, Y., Jerdeva, G., Yarber, F. & 8 other authors (2003).

in vitro model: cytoskeleton and tight junctions as indicators for the definition of age-related stages by confocal microscopy. Pharm Res 15, 964–971.

Cytoplasmic dynein participates in apically targeted stimulated secretory traffic in primary rabbit lacrimal acinar epithelial cells. J Cell Sci 116, 2051–2065.

Rothen-Rutishauser, B., Riesen, F., Gu¨nthert, M. & WunderliAllenspach, H. (2002). Dynamics of tight and adherens junctions

Warner, M. S., Geraghty, R. J., Martinez, W. M., Montgomery, R. I., Whitbeck, J. C., Xu, R., Eisenberg, R. J., Cohen, G. H. & Spear, P. G. (1998). A cell surface protein with herpesvirus entry activity (HveB)

under EGTA treatment. J Membr Biol 188, 151–162. and Caco-2: two in vitro models for the study of intestinal differentiation. Biochimie 68, 1035–1040.

confers susceptibility to infection by mutants of herpes simplex virus type 1, herpes simplex virus type 2, and pseudorabies virus. Virology 246, 179–189.

Schelhaas, M., Jansen, M., Haase, I. & Knebel-Mo¨rsdorf, D. (2003).

Whittaker, G. R., Kann, M. & Helenius, A. (2000). Viral entry into the

Herpes simplex virus type 1 exhibits a tropism for basal entry in polarized epithelial cells. J Gen Virol 84, 2473–2484.

Yoon, M. & Spear, P. G. (2002). Disruption of adherens junctions

Rousset, M. (1986). The human colon carcinoma cell lines HT-29

Sears, A. E., McGwire, B. S. & Roizman, B. (1991). Infection of

polarized MDCK cells with HSV1: two asymmetrically distributed

786

nucleus. Annu Rev Cell Dev Biol 16, 627–651. liberates nectin-1 to serve as receptor for herpes simplex virus and pseudorabies virus entry. J Virol 76, 7203–7208.
