Effect of polylysine on the early stages of infection of wild type pseudorabies virus and of mutants defective in gIII

VIROLOGY

179,330-338

(1990)

Effect of Polylysine Pseudorabies LASZLO ZSAK,* THOMAS *Department

of Microbiology,

on the Early Stages of Infection of Wild Type Virus and of Mutants Defective in glll

C. MElTENLEITER,t

NANCY

SUGG,*

AND

Vanderbilt University School of Medicine, Nashville. Tennessee for Virus Diseases of Animals, D-7400 Tubingen, Federal Republic Received

April 30, 1990; accepted

TAMAR

BEN-PORAT*,’

37232; and tFederal of Germany

Research

Centre

July 3, 1990

The main pathway of adsorption of pseudorabies virus (PrV) to its host cells is via interactions between viral glycoprotein glll and a cellular heparin-like receptor. Mutants of PrV deficient in glycoprotein glll adsorb by an alternative, slower pathway. Penetration into the cells of glll- mutants is also delayed compared to penetration of wild type virus. We show here that polylysine enhances the adsorption of glll- mutants. Furthermore, in the presence of polylysine the adsorption of wild type virus involving the interactions of viral glycoprotein glll and the heparin-like cellular receptor is efficiently bypassed. Polylysine appears to promote virus adsorption by bridging the cellular and viral membranes. Polylysine not only stimulates adsorption of glll- mutants but also promotes their internalization; the delay in the initiation of viral protein synthesis that is observed in cells infected with gIlI- mutants compared to wild type infected cells is abrogated. Because it is unlikely that polylysine can substitute for two different functions of gIlI, adsorption and penetration, the delay in the initiation of the infectious cycle in glll--infected cells is probably related to the defect in adsorption. Furthermore, polylysine can completely overcome the inhibitory effects of antisera against gIlI, but not the inhibitory effects of antisera that affect a later stage of infection. It is unlikely therefore that polylysine can promote penetration directly and that glll is involved directly in penetration. These results, as well as those obtained previously, show that o 1990 Academic while glll is essential for the efficient adsorption of PrV, it affects virus penetration only indirectly. Press, Inc.

INTRODUCTION

tors which are responsible for the initial adsorption of most wild type virions to the cells. Not only do mutants defective in glll adsorb poorly but the internalization of glll- virions that have adsorbed to the cells is also delayed compared to that of wild type virus (Zuckermann et a/., 1989; Mettenleiter, 1989). Thus, either glll plays a direct role in both the adsorption and the penetration of the virus or the delay in penetration of glll- virus is linked to the alternative mode of adsorption the glll- mutants use. During the course of studies designed to elucidate the factors affecting adsorption of wild type PrV and of mutants of this virus that lack gIlI, we have tested the effects on virus adsorption of several polycationic and polyanionic substances that have been shown to inhibit the growth of some herpesviruses (Vaheri, 1964; Takemoto and Fabish, 1964; Nahmias and Kibrick, 1964; Langeland et al., 1987, 1988; WuDunn and Spear, 1989). We show here that polylysine enhances the adsorption of PrV by a pathway that is independent of the viral gill/cellular heparin-like moiety interaction. Furthermore, polylysine promotes the rapid penetration of glll- virions and overcomes the lag in the initiation of the infectious process that is observed in gIlI-infected cells. Polylysine also reverses the inhibitory effects of antibodies on virus adsorption but not their inhibitory effect on later stages of virus infection. The

We have been interested in ascertaining the functions of the nonessential glycoproteins of pseudorabies virus (PrV) and have focused on a major nonessential glycoprotein, glycoprotein gIlI. This glycoprotein is a homolog of glycoprotein gC of herpes simplex virus (Robbins et a/., 1986). Cells infected with mutants of PrV deficient in glll produce virus populations that have a lower titer of infectious virus than do cells infected with wild type virus (Keeler et al., 1986; Schreurs et a/., 1988; Whealy et al,, 1988). The lower titer of infectious virus produced by glll- mutant-infected cells is attributable, at least in part, to the reduced infectivity of the glll- mutants and reflects the important role that glll plays in the stable adsorption of the virus to its host cell (Schreurs et al., 1988). Glycoprotein glll binds to a heparin-like receptor on the cell surface (Mettenleiter et al., 1989). In the absence of gIlI, the virus adsorbs to the cells by a pathway that is not mediated by this cellular receptor (Zuckermann et a/., 1989). Thus, PrV can adsorb to the cells by a mode that is independent of the presence of both the glll viral attachment protein and the cellular recep-

’ To whom 0042-6822190

requests

for reprints

$3.00

CopyrIght 0 1990 by Academic Press. Inc All rights of reproduction in any form reserved

should

be addressed 330

POLYLYSINE

EFFECT

results of the experiments presented in this paper reinforce our previous (tentative) conclusion that the lag in the initiation of the infectious process observed when cells are infected with glll- virus is related to the alternative mode of adsorption the glll- mutants use and that glll is not directly involved in promoting penetration of the virus. MATERIALS Virus strains

AND

METHODS

and cell culture

PrV(Ka) is a strain that has been carried in our laboratory for more than 25 years. The isolation and characterization of the gIlI- mutants of PrV(Ka) used have been described previously (Schreurs et al., 1988). The PrV(Be) strain and the gill- mutant of this strain (BlO) were made available to us by Lynn Enquist (DuPont). Madin-Darby bovine kidney (MDBK) or rabbit kidney (RK) cells were cultivated in EDS. Virus was titrated by plaque assay on MDBK monolayer cultures. Media and solutions The media and solutions used were EDS, Eagle’s synthetic medium plus 5% dialyzed bovine serum; EDS-FU, EDS plus 5-fluorouracil(20 pg/ml) and thymidine (5 pg/ml); PBS, phosphate-buffered saline containing 0.136 Iw NaCI, 2.6 mN1 KCI; 8 mM Na,HPO,, 1 mll/l KH,PO,, 20 mM MgCI,, 1.8 mM CaCI,, (pH 7.0); TBS; Tris-buffered saline (the same as PBS, but with phosphate replaced by 0.01 M Tris hydrochloride, pH 7.5); TBSA, TBS containing 1% crystalline bovine albumin; and TBS-SDS, TBS containing 2% sodium dodecyl sulfate. Radiochemicals

and chemicals

[3H]Thymidine (sp act, 45 Ci/mmol) and [3H]leucine (sp act, 68 Ci/mmol) were purchased from New England Nuclear Corp. Polyaminoacids (approximate MW 50,000) were purchased from Sigma Chemical Co. Labeling

and purification

ON

PrV

INFECTION

331

was resuspended gently in 1 ml TBSA and centrifuged on a sucrose gradient, as described previously (BenPorat et al., 1974; Ben-Porat and Kaplan, 1976). Fractions were collected and the virus peak was localized. The virus was diluted with TBSA and sedimented through a TBSA-30% sucrose cushion by centrifugation at 15,000 rpm for 1 hr in an SW27 rotor. Adsorption

of labeled virus

Purified [3H]thymidine-labeled virus in TBSA, prepared as described above, was incubated with the monolayers (which had been preincubated in TBSA for 15 min to minimize unspecific binding). The monolayers were further incubated for 1 hr at 37” and then washed extensively to remove unadsorbed virus. The cells were scraped into TBS-SDS, and the amount of radioactivity that was associated with the cell monolayers was determined. Entry kinetics This was done as described previously (Mettenleiter, 1989). Briefly, MDBK cells in 5-cm plastic dishes were preincubated for 20 min at 4” either in TBSA or in TBSA containing polylysine (50 pg/ml). The monolayers were then infected with 1 ml (approximately 200 PFU) of virus and virus adsorption was allowed to proceed at 4” for 30 min (at that temperature, virus adsorption but not penetration occurs). The unadsorbed virus was removed by washing and the monolayers were incubated with prewarmed medium at 37”. At various times thereafter, the medium was removed and the plates were washed once with PBS, incubated for 2 min either with PBS (control plates) or with citric acid buffer [40 mM citric acid, 10 mM KCI, and 135 rnM NaCl (pH 3.0)], washed again, and overlaid with agarose. Polyacrylamide

gel electrophoresis

This was performed

as described

previously

(Hampl

et al., 1984).

of virus

To obtain [3H]thymidine-labeled virus, cell monolayers were incubated for 24 or 48 hr in EDS-FU [a procedure that inhibits cellular DNA synthesis without affecting virus growth (Kaplan and Ben-Porat, 1961)]. The cells were then infected (m.o.i., 5 PFU/cell) and incubated in EDS containing [3H]thymidine (20 &i/ml) for 24 hr. To purify the virions, the medium was collected and clarified by centrifugation at 4000 g for 10 min, and the supernatant containing the extracellularvirus was centrifuged on a TBSA-30% sucrose cushion at 15,000 rpm for 1 hr in a Beckman SW27 rotor. The virus pellet

Antisera Polyvalent anti-glll (286) and anti-gll (284) are goat sera prepared against denatured cloned cro-gll and cro-glll fusion proteins (generous gifts from L. Enquist). Anti-PrV serum was obtained from pigs recovering from infection with PrV. RESULTS Effect of polyamino

acids on adsorption

of PrV

Langeland et al. (1988) and WuDunn and Spear (1989) have previously reported that polylysine inhibits

332

ZSAK

ET AL.

TABLE EFFECT OF POLYAMINO

ACIDS ON PLAQUE FORMATION

Polylysine Concentration (ha/ml) 0 5 15 50 a MDBK either wild adsorption * PFU/ml c ND, not

OF WILD TYPE AND gIlI-

Polyarginine

MUTANTS OF PrV” Polyhistidine

L-Lysine

wt

gIlI-

wt

gIlI-

wt

gIlI-

wt

gIlI-

90.96 83.0 46.3 19.3

0.96 0.7 2.3 11.3

72.3’ 44.6 19.3 12.6

0.8’ 0.6 2.3 7.8

35.56 ND ND 30.3

0.96 ND ND 0.8

35.0b ND 28.9 30.3

ND” ND ND ND

monolayers were pretreated type PrV(Ka) or a PrV(Ka)glllperiod of 1 hr, the monolayers X 1 07. done.

for 15 min with 1 ml of TBSA containing the indicated concentrations of polyamino acids. Stocks mutant were then titrated on these cells by adding 0.1 ml of lo-fold dilutions of the virus. After were extensively washed and overlaid with agarose. Plaques were counted 4 days later.

plaque formation of HSV-1 by inhibiting adsorption of the virus. Furthermore, WuDunn and Spear (1989) have concluded that, in addition to its effect on adsorption of HSV-1, polylysine also inhibits some later stage of infection. Table 1 shows the effect of the presence of different polyamino acids during the adsorption period on plaque formation of wild type PrV(Ka) and of a mutant of this virus lacking glycoprotein gIlI. (Each set of data presented in this paper was obtained from a single experiment. The experiments were repeated at least three times with essentially the same results.) Even at high concentrations (50 pg/mI) both polylysine and polyarginine inhibited plaque formation of wild type virus by not more than 80%. Thus, the inhibitory effect of polylysine on plaque formation of PrV is considerably less pronounced than it is on HSV-1 (Langeland et al., 1988; WuDunn and Spear, 1989). While plaque formation of wild type PrV was somewhat decreased, plaque formation of glll- mutants of PrV was increased more than lo-fold in the presence of polylysine and polyarginine. Indeed, the titer of the glll- stocks (which is considerably lower than that of wild type when the virus is allowed to adsorb to the cells in the absence of the drugs) was almost the same as that of wild type virus when the adsorption was carried out in the presence of polylysine. The effects of the polyamino acids on the adsorption of radiolabeled wild type and glll- PrV(Ka) was also ascertained (Table 2). Both polylysine and polyarginine had little effect on the adsorption of radiolabeled wild type virus, but considerably enhanced the adsorption of the glll-virus. Similar results were also obtained with two other independently isolated glll- mutants of PrV(Ka) as well as one glll- mutant of PrV(Be), PrVl 0. Effect of the The plaque

1

of pretreatment of the cells or virus with polylysine stimulation by polylysine of adsorption and of formation of glll- mutants of PrV appeared to

of an

be of sufficient interest to warrant further investigation. To ascertain whether it was the result of the interaction of polylysine with the virus or with the cells, each one was pretreated with polylysine and the effect of this pretreatment on subsequent adsorption and plaque formation was ascertained. Pretreatment of the cells with polylysine did not greatly affect the titer or the amount of radiolabeled wild type virus that adsorbed to the cells. However, pretreatment of the cells with polylysine increased the titer and the number of glll-virions that adsorbed to the cells by approximately 1O-fold (Table 3). Pretreatment of wild type virus with polylysine (polylysine was diluted out prior to assay) did not greatly affect its titer; pretreatment of glll- mutants with polylysine enhanced it considerably (Table 4). Similar results were also obtained with another, independently isolated glll- mutant of PrV(Ka) (data not shown). Since polylysine appears to interact with both the virus and the cells it seemed likely that polylysine may enhance the adsorption and the infectivity of the glllmutants by bridging the virus to the cells. Polylysine

TABLE EFFECTOF

TBSA TBSA TBSA

2

POLYLYSINE AND POLYARGININE ON ADSORPTION RADIO~ABELED WILD TYPE AND glll- PrV’

plus polylysine plus polyarginine

OF

wt

gIlI-

11 .8b 10.9 8.6

0.8b 9.8 11.8

a RK cells were incubated in TBSA with 1 O5 cpm of [3H]thymidinelabeled purified wild type PrV(Ka) or a glll- mutant of this virus in the presence or absence of polylysine or polyarginine (50 rglml). After a 60.min adsorption period, the cells were washed and the number of counts associated with the cell monolayers was determined. bcpm X 103.

POLYLYSINE TABLE

EFFECT

ON

PrV INFECTION

333

3

TABLE POLYLYSINE-INDUCED

EFFECT OF PRETREATMENT OF CELLS WITH POLYLYSINE ON VIRUS ADSORPTIONAND PLAQUEFORMATION~

5

AGGLUTINATION

OF PrVa

Polylysrne Adsorptron (cpm x 103)

Pretreatment with polylysine

wt

-

24.5 24.9

+

Plaque formation (PFWml)

glll 1.5 17.3

Dilution facto?

wt

glll

8x10’ 3x lo*

8X lo6 1X108

a Monolayers of MDBK cells were incubated for 60 min with TBSA or with TBSA containing polylysrne (50 ~g/ml). The monolayers were then washed extensively and used either to plaque assay stocks of wild type PrV(Ka) or a gIlI- mutant of this virus or to ascertain the ability of these radiolabeled purified vinons to adsorb to the cells. Plaque formation was tested as described in footnote a of Table 1; adsorption of radiolabeled virus was ascertained as described In footnote a of Table 2.

has been reported to integrate into cellular membranes (Bashford et al., 1986; Shen and Ryser, 1979). To ascertain whether polylysine inserts into the viral membranes and can bridge the membranes of different virions, we determined whether it would agglutinate the virions and whether it would do so in a virion concentration-dependent manner (Table 5). The presence of polylysine (but not of lysine; data not shown), caused the virus to sediment under conditions which normally do not affect it. The magnitude of this effect was dependent on the concentration of the virus; the fraction of the virions that sedimented after lowspeed centrifugation decreased considerably as the concentration of the virus decreased. We conclude from these results that polylysine probably bridges the virus particles by inserting itself into their membranes, thereby agglutinating them. It is likely that polylysine promotes the adsorption of the glll- virus to the cells by a similar process, i.e., by inserting itself into both the

TABLE

Titer (PFWml)

+

wt 2x lo* 1X108

235,830’ 46,035 9,334 1,989 410 95

+ 11.660C 10,156 5,108 1,381 330 110

Polylysine Polylysine

+ -

0.05 0.22 0.55 0.69 0.81 1.16

a A purified preparation of [3H]thymrdine-labeled PrV(Ka) virions (approximately 5 X 10” particles/ml) in TBS was clarified by centrifugation at 5000 g for 5 min and diluted serially (5X) In TBS. The samples were counted. To part of the preparatron polylysine (45 pg/ml) was added. The samples were Incubated at 37” for 15 mm and then centrifuged at 5000 g for 3 min. Part of the supernatants were counted. In all cases less than 3% of the counts sedimented in the sample devoid of polylysrne. ’ Dilution of labeled virus. ’ cpm in 0.1 ml of supernatant.

cellular and viral membranes, thereby binding the virus to the cells. Since polylysine causes the agglutination of the virus, it should, in principle, consistently reduce its titer. However, it increases the titer of glll- virus (see Table 4). The reason no decrease in the titer of glll- virus is observed in the experiment summarized in Table 4 is twofold. (1) In this experiment, the concentration of the virus particles that were treated with polylysine was at least one thousand times lower than the initial virus concentration in the experiment summarized in Table 5. At a similarly low concentration, only a small proportion of the virus was agglutinated by polylysine (see Table 5). (2) The small inhibitory effect of polylysine on the virus titer resulting from agglutination would not be detectable because of the much larger enhancing effect the drug has on adsorption. Effect of polylysine on the ability of PrV to adsorb to cells treated with heparinase

4

EFFECT OF PRETREATMENT OF glll- MUTANTS WITH POLYLYSINE ON ITS INFECTIOUS TITERS

Pretreatment with polylysine

5 25 125 625 3,125 15,625

-

gIlI4.6 X 1 O6 4.7 x lo7

a Stocks of wild type virus or gIlI- mutants of PrV(Ka) were diluted 1: 100 in TBSA containing polylysine (50 pg/ml) and were Incubated for 60 min at 37”. Serial lo-fold dilutions of the virus were then plaque-assayed on MDBK cells. Plaques were counted 4 days later.

We have shown previously that the majority of wild type PrV adsorbs to a cell surface heparin-like receptor and that the viral glycoprotein glll is essential to this process; the absence of glll on the virion surface or the treatment of the cells with heparinase greatly reduces virus adsorption (Mettenleiter et a/., 1989). In addition to the viral gllI/cellular heparin-like receptor-mediated adsorption, PrV can also adsorb by another, less efficient, pathway used primarily by glll- virus (Zuckermann et a/., 1989). Since polylysine promotes the adsorption of glll- mutants, it was of interest to determine

334

ZSAK TABLE

EFFECTOF

6

POLYLYSINE ON PLAQUE FORMATION OF WILD TYPE VIRUS ON HEPARINASE-TREATED CELLS Heparinase

Adsorption medium TBSA TBSA

+ polylysine

treatment monolayers

of cell

6.7 X 1 O7 2.0 x 1 o7

+ 3.2 X 10’ 5.9 x 1 o7

a MDBK monolayers were treated for 60 min at 37” with either PBS or PBS containing heparinase (5 units/ml), as described previously (Mettenleiter eta/., 1989). They were then washed and incubated for 10 min with 1 ml TBSA or TBSA containing polylysine (50 rg/ml). Subsequently, 0.1 ml of 1 O-fold dilutions of a PrV(Ka) wild type virus stock was added to the cultures, which were further incubated for 1 hr at 37”, washed to remove unadsorbed virus, and overlaid with agarose. Plaques were counted 4 days later.

ET AL

treatment (Mettenleiter, 1989). Unadsorbed virus was then removed by washing and the cultures were shifted up to 37”. At various times thereafter, the proportion of adsorbed virus that remained sensitive to low pH treatment was ascertained (Fig. 1). Wild type PrV(Ka) quickly became resistant to low pH treatment; 20 min after shift-up, approximately 75% of the wild type virus had become insensitive to this treatment. On the other hand, only 5% of a glll- mutant of this virus had become insensitive to low pH treatment by 20 min after shift-up. When the glll- virus was adsorbed to the cells in the presence of polylysine, it became resistant to low pH treatment much more rapidly and by 20 min after shift-up, almost 40% was resistant.

Time course of viral protein synthesis after adsorption of glll- mutants in the presence or absence of polylysine In the hope of gaining an understanding of the early events that initiate the infectious cycle and to ascertain

whether in its presence wild type virus could also bypass the viral gill/cellular heparin-receptor mode of adsorption. Plaque formation of wild type virus on monolayers that had been treated with heparinase was considerably reduced compared to plaque formation on untreated monolayers (Table 6). The presence of polylysine, however, restored to heparinase-treated monolayers the ability to support plaque formation. Polylysine also restored the ability of heparinasetreated cells to adsorb radiolabeled virus (data not shown). Thus, in the presence of polylysine, the cellular heparin-like receptor is not required for the adsorption of eitherwild type or glll-virus, and theviral glIl/cellular heparin-like moiety interaction which normally is the pathway by which most of the wild type virus adsorbs to the cells is successfully bypassed.

Resistance to low pH treatment after adsorption gill- mutants in the presence or absence of polylysine

00

60

E B i!

40

s

of

Not only do mutants of PrV lacking glll adsorb less efficiently than does wild type virus but once they have stably adsorbed to the cells (i.e., cannot be removed by extensive washing) they also become resistant more slowly than does wild type virus to treatment with low pH and to neutralization by antisera (Zuckermann eta/., 1989; Mettenleiter, 1989). The experiments described below show that polylysine promotes the rapid acquisition by adsorbed glll- virus of resistance to treatment with low pH. Virus was allowed to adsorb at 4’ in the presence or absence of polylysine. At this temperature the virus adsorbs to the cells but remains susceptible to low pH

0

10

20

JO

Time (minutes) FIG. 1. Acquisition of low pH resistance by adsorbed wild type virus and gIli- mutants. MDBK cells in 5-cm plastic dishes were preincubated for 20 min at 4” either in TBSA or in TBSA containing polylysine (50 pg/ml). The monolayers were then infected with approximately 200 PFU of virus and adsorption was allowed to proceed at 4” for 30 min (at that temperature virus adsorption but not penetration occurs). The unadsorbed virus was removed by washing and the monolayers were incubated with prewarmed medium at 37”. At various times after the temperature shift the medium was removed and the plates were washed once with PBS, incubated for 2 min either with PBS (control plates) or with citric acid buffer [40 mAY citric acid, 10 mM KCI, and 135 mM NaCl (pH 3.0)], washed again, and overlaid with agarose. Closed circles, wild type virus; closed squares, gill- virus; open circles, glll- virus adsorbed in the presence of polylysine.

POLYLYSINE

WT

gIlI-

”

EFFECT

‘UT

MCI MDBF

FIG. 2. Autoradiogram of proteins synthesized by cells at various times after infection with wild type and gIlI- virus. (A) MDBK cells (1 06/cells grown in 50.mm petri plates) were Infected with 3 X 1 O6 PFU of wild type PrV(Ka) or with 2 X 10’ PFU of a gIlI- mutant of thus virus and adsorption was allowed to proceed for 2 hr. At the indicated times (2,4, or 6 hr), they were incubated in EDS without amino acrds containing argrnine and [3H]leucine (25 &i/ml) for 1 hr. The cells were harvested and the proteins separated on acrylamide gels. (6) MDBK cells were infected with 2 X 1 O6 PFWml(2a. 4a, 6a) or with 5 x 1 O5 PFU/ml(2b, 4b, 6b) and labeled as described for A.

the role of glll in virus penetration, we ascertained how the presence of polylysine during adsorption affects the time course of Synthesis of viral proteins in wild type- and gIlI--infected cells (Figs. 2 and 3). Since virus penetration and initiation of DNA synthesis is delayed in cells infected with glll-virus compared to cells infected with wild type virus (Zuckermann eta/., 1989; Mettenleiter, 1989), it was not surprising to find that the initiation of viral protein synthesis in glll--infected cells was also delayed (Fig. 2). At 2 hr postinfection in wild type virus-infected cells (Fig. 2A, first lane), the major DNA binding protein (MDBP) (an early protein) is abundantly represented relative to the major capsid protein (MCP) (an early-late protein). In glll--infected cells, at 2 hr postinfection (Fig. 2A, fourth lane), neither of these viral proteins is detectable. By 4 hr postinfection the MCP is much more prominent than the DBP in wild type virus-infected cells (Fig. 2A, second lane), but in glll- mutant-infected cells both are present in approximately equal amounts (Fig. 2A, fifth lane). We estimate that in this experiment, the adsorbed multiplicity of the glll- mutant was approximately twice that of wild type virus. A decrease in the multiplicity of infection of wild type virus also did not significantly affect the pattern of protein synthesis (Fig. 2B). The delay in the initiation of protein synthesis observed in gIlI--infected cells is therefore not related to a lower multiplicity of virus infection in these cells but is due to a slower initiation of infection of the adsorbed gIlI- virus.

ON

PrV INFECTION

335

The presence of polylysine during the adsorption period did not significantly affect the pattern of viral protein synthesis in wild type virus-infected cells; it did, however, promote a more rapid initiation of viral protein synthesis in the gIlI- mutant-infected cells (Fig. 3). The pattern of protein synthesis in cells infected in the presence of polylysine with glll- mutants (Fig. 3, lanes 7-9) or with wild type virus (Fig. 3, lanes l-3) was approximately the same during the first few hours of infection. (In this experiment, the cells were labeled at earlier times postinfection than in the experiment shown in Fig. 2.) Thus, polylysine not only promotes the adsorption and the acquisition of low pH resistance but also stimulates the penetration and rapid initiation of the infectious process of glll- mutants. Polylysine antibodies

reverses against

the neutralizing effects glll but not against gll

of

The experiments described in this section were designed to ascertain whether polylysine can restore infectivity to virus that has been neutralized by specific antisera. The effects of polylysine on the ability of virus that had been neutralized by sera against glll (that inhibit adsorption and plaque formation), against gll (that do not inhibit adsorption but inhibit plaque formation), and against polyvalent anti-PrV serum (that inhibits both) to adsorb and to form plaques was ascertained. Wild type PrV was incubated with sufficient antiserum to reduce its infectivity by approximately 99%, as determined by assay on MDBK cells. The ability of the neutralized virus to adsorb and form plaques on MDBK cells treated or untreated with polylysine was ascer-

12345

6769

FIG. 3. Autoradiogram of proteins synthesized by cells infected with wild type and glll- mutants in the presence or absence of polylysine. MDBK cells were infected with wild type PrV(Ka) (2 x 10” PFU) (lanes l-3), wrth a gIlI-- mutant of this virus (2 x 10’ PFU) (lanes 46) or with the gIlI- vrrus (5 X 1 O6 PFU) In the presence of polylysine (50 @g/ml) (lanes 7-9). At hourly intervals (1, 2, and 3 hr postinfection), the cells were incubated for 1 hr In EDS without amino acids contarning arginrne and [3H]leucrne (25 &i/ml). The cells were harvested and the proteins separated on polyacrylamide gels.

336

ZSAK TABLE

7

ADSORPTION AND PLAQUE FORMATION OF WILD TYPE PrV AFTER NEUTRALIZATION WITH VARIOUS ANTIBODIES IN THE PRESENCE OR ABSENCE OF POLYLYSINE Adsorption b of 3H-labeled virus (cpm)

Virus titera PFU) Polylysine -

Antiserum Preimmune Anti-glll (286) Anti-gll (284) Anti-PrV

6.0 2.1 1.5 6.2

X x x X

10’ 106 lo6 lo6

Polylysine +

-

+

2.2 x 108 1.9x lo8 1.4x lo6 7.8 X lo6

59.3 6.2 52.0 2.0

61.6 64.3 56.0 59.0

‘Wild type PrV(Ka) was incubated for 1 hr with the indicated antisera at a concentration that had previously been determined to be sufficient to reduce the infectious titer by approximately 99%. The virus was serially diluted (1 O-fold) in TBS or TBS containing polylysine (50 pg/ml) and 1 ml of the appropriate virus dilutions was added (in duplicates) to monolayers of MDBK cells. After a GO-min adsorption period, the monolayers were washed and overlaid with agarose. Plaques were counted 4 days later. The results are expressed as PFU/ml. b Radiolabeled, purified wild type PrV(Ka) was incubated with the indicated antiserum (at a concentration sufficient to neutralize approximately 99% of the virus) for 1 hr at 37”. The virus was then diluted fivefold into either TBS or TBS containing polylysine (50 pglml). One milliliter of the virus suspension (approximately 3 X 1 O5 cpm) was then added to monolayers of MDBK cells which were incubated for 1 hr at 37”. The monolayers were washed extensively and the number of counts that remained associated with the cell monolayers was ascertained. The results are expressed as cpm X 1 03.

tained. Since glycoprotein glll is the viral component that interacts with the heparin-like cellular receptor (Schreurs et a/., 1988) adsorption of wild type virus should not be affected by neutralizing antibodies against glll if it is carried out in the presence of polylysine, which allows the virus to bypass these interactions. Indeed, when polylysine was present during the adsorption period, the adsorption and the titer of the virus that had been neutralized by anti-9111 serum was approximately the same as that of the nonneutralized virus (Table 7). Thus, the neutralizing effect of antibodies against glll can be reversed when the virus is adsorbed to or assayed on polylysine-treated cells. Polylysine also reversed the effects of anti-PrV sera on adsorption of radiolabeled virus but not on the ability of the virus to form plaques (Table 7). Since the anti-PrV serum probably contains antibodies against most viral glycoproteins and since polylysine can overcome the inhibitory effect of this serum on adsorption, it appears that the effect of antibodies against any virus proteins that may affect adsorption can be overcome by polylysine.

ET AL

The inhibitory effects of anti-911 serum and anti-PrV serum on plaque formation were not overcome by the presence of polylysine. Thus, while polylysine can overcome the inhibition of virus adsorption by antibodies, it cannot overcome the effects of antisera that affect later events such as penetration (i.e., fusion of the cellular membrane and the viral membranes). DISCUSSION During the course of studies designed to elucidate the processes which control adsorption and penetration of PrV, we have tested the effect on wild type PrV and on mutants of this virus deficient in glll of several polyanionic and polycationic substances that had been previously shown to affect adsorption or growth of herpesviruses. In this paper we focus on the effects of polylysine on adsorption and on penetration into the cells of wild type PrV and of mutants of this virus deficient in gIlI. Polylysine has been shown to integrate into the plasma membrane in which it induces “pores” and to increase the permeability of the cell membranes (Bashford et a/., 1986). It has also been reported to induce membrane fusion (Walter et al., 1986; Uster and Deamer, 1985; Gad et a/., 1985) and to be endocytosed (Shen et al., 1985). Furthermore, polylysine efficiently blocks the binding of HSV-1 to susceptible cells (Langeland et a/., 1988; WuDunn and Spear, 1989). Polylysine does not significantly inhibit adsorption of wild type PrV but somewhat affects plaque formation of this virus. Thus, polylysine appears to interfere with the infectious cycle of PrV at some step subsequent to binding to the cell. HSV-2 is similarly affected by polylysine (WuDunn and Spear, 1989). While polylysine does not affect the adsorption of wild type PrV, it considerably enhances the adsorption and the infectivity of glllmutants of this virus. Our results indicate that this enhancement is probably due to the fact that in the presence of polylysine the usual pathway of virus adsorption that involves the interaction of the viral glycoprotein glll with the cellular heparin-like receptor is successfully bypassed. That this is the case is indicated by the finding that wild type PrV adsorbs well and forms plaques on heparinase-treated cells provided the cells are also treated with polylysine during the adsorption period. Furthermore, polylysine overcomes the inhibitory effects of antibodies on virus adsorption. The increase in the ability of glll- mutants to adsorb to cells in the presence of polylysine is probably due to the ability of the drug to insert itself into the viral and cellular membranes, thereby bridging them. This conclusion is based on the following observations: (1) Pretreatment of either the cells or the virus with polylysine

POLYLYSINE

EFFECT

promotes the adsorption of gIlI- virus. Therefore, it is clear that polylysine can interact with both. (2) The addition of polylysine to highly concentrated virus suspensions causes the virus to aggregate. Since this effect of polylysine is virus concentration-dependent, it is likely that polylysine agglutinates the virions, probably by inserting into membranes of neighboring virions. glll- mutants use an alternative inefficient mode of adsorption and are also delayed in penetration into the cells (Zuckermann et al., 1989; Mettenleiter, 1989). We show here that not only do glll- virions adsorb more efficiently to the cells in the presence of polylysine, but they become resistant more rapidly to low pH treatment as well. The patterns of viral protein synthesis in cells infected with wild type and glll-virus also indicate that while initiation of infection is delayed in glll- infected cells compared to wild type infected cells, this delay is abrogated in the presence of polylysine. Thus, polylysine promotes adsorption as well as internalization of glll- virus, and also overcomes the delay in the initiation of the infectious process that occurs in glllmutant-infected cells. The finding that adsorbed gIlI- virions initiate infection more slowly than does wild type virus raised the possibility that glll may play a direct role in virus penetration. Alternatively, the delay in the initiation of infection could be related to the alternative mode of adsorption the glll- virus uses. Three arguments favor the latter possibility: (1) Polylysine promotes both adsorption and penetration of glll- virus. It is unlikely that polylysine can substitute for two distinct functions of gIlI. Rather, the two functional defects observed in glll- mutants may be linked, i.e., efficient penetration of the virus may be coupled to its correct adsorption. Thus, while the adsorption of glll- virus by the alternative mode would result in a delay in the penetration of the virus, in the presence of polylysine the mode of adsorption would be such that rapid penetration can occur. (2) Although polylysine overcomes the inhibitory effects on adsorption of antibodies against glll and restores infectivity to virus neutralized by anti-glll sera, it cannot overcome the inhibitory effects of antisera against gll. Thus, polylysine does not promote virus internalization, i.e., fusion of the cellular and viral membrane (at least not after the virus has been neutralized). It is unlikely, therefore, that polylysine directly promotes internalization. (3) Because polylysine can completely overcome the neutralizing effects of antibodies against glll but cannot overcome the neutralizing effects of antibodies that inhibit events that occur after the virus has adsorbed, glll does not appear to be directly involved in these later events. On the basis of these findings, we conclude that glll is directly involved in virus adsorption but that it plays

ON

PrV

INFECTION

337

no direct role in virus penetration. The reason penetration of glll- virus is delayed is linked to the alternative mode of adsorption these mutants use. Adsorption via the viral glll/heparin-like cellular receptor probably promotes the interactions of some viral membrane proteins with the appropriate cellular proteins, leading to the rapid penetration of the virions into the cells. Adsorption of glll-virus by the alternative mode of adsorption probably results in a “loose” contact between the cellular and viral membranes, and the interactions between the appropriate viral and cellular proteins leading to virus penetration consequently cannot occur efficiently. However, after adsorption of glll- virus to polylysine-treated cells, the juxtaposition of the viral and cellular membranes is such that their proteins can interact rapidly and penetration can occur efficiently.

ACKNOWLEDGMENT This investigation was supported by Public Health Service Grant Al-10947 from the National Institutes of Health and by a research grant (Me 854/2-l) from the Deutsche Forschungsgemeinschaft.

REFERENCES BASHFORD, C. L., ALDER, G. M., MENESTRINA, G., MICKLEN, K. J., MURPHY, 1. J., and PASTERNAK, C. A. (1986). Membrane damage by hemolytlc viruses, toxins, complement and other cytotoxic agents. /. Biol. Chem. 261,9300-9308. BEN-P• RAT, T., DEMARCHI, J. M., and KAPLAN, A. S. (1974). Characterization of defective interfering viral particles present in a population of pseudorabies vinons. J. Virol. 61, 29-37. BEN-P• RAT, T., and KAPLAN, A. S. (1976). A comparison of two populations of defective Interfering pseudorabies virus particles. Virology72,471-479. GAD, A. E., BENTAL, M., ELYASHIV, G., and WEINBERG, H. (1985). Promotion and inhibition of vesicle fusion by polylyslne. Biochemistry 2496277-6282. HAMPL, H., BEN-P• RAT, T., EHRLICHER, L., HABERMEHL, K.-O., and KAPLAN, A. S. (1984). Characterization of the envelope proteins of pseudorabies virus. /. Viral. 52, 583-590. KAPLAN, A. S., and BEN-P• RAT, T. (1961). The action of 5-fluorouracil on the nucleic acid metabolism of pseudorabies virus-infected and noninfected rabbit kidney cells. Virology 13, 78 -92. KEELER, C. K., WHEALY, M. E., and ENQUIST. L. W. (1986). Construction of an infectious PrV recombinant expressing a glycoprotein glll beta-galactosidase fusion protean. Gene 50, 215-224. LANGELAND, N. H., HOLMSEN, H., LILLEHAUG. J. R., and HAARS, L. (1987). Evidence that neomycin inhibits binding of herpes simplex virus type 1 to the cellular receptor. /. Viral. 61, 3388-3393. LANGELAND, N., MOORE, L. J., HOLMSEN, H., and HAARS, L. (1988). Interaction of polylysine with the cellular receptor for herpes simplex virus type 1. J. Gem Viral. 69, 1 137- 1 145. METTENLEITER, T. C. (1989). Glycoproteln glll deletion mutants of pseudorabies virus are impalred in virus entry. Virology 171, 623625. ME~TENLEITER, T. C.. ZSAK. S.. ZUCKERMANN. F., SUGG, N., KERN, H., and BEN-P• RAT, T. (1989). Interaction of glycoprotein glll with a cellular heparin-like substance mediates adsorption of pseudorabies virus. /. Viral. 64, 278-286.

338

ZSAK ET AL.

NAHMIAS, A. J., and KIBRICK,S. (1964). Inhibitory effect of heparin on herpes simplex virus. I. Entry. 1. Bacferiol. 87, 1060-l 066. ROBBINS,A. K., WATSON, R. J., WHEALY, M. E., HAYS,W. W., and ENQUIST,L. W. (1986). Characterization of a pseudorabies virus glycoprotein gene with homology to herpes simplex virus type 1 and type 2 glycoprotein C. J. Viral. 58, 339-347. SCHREURS,C., MET~ENLEITER,T. C., ZUCKERMANN, F., SUGG, N., and BEN-P• RAT. T. (1988). Glycoprotein glii of pseudorabies virus is multifunctional. i. i/iiol. 62,225 7-2257. SHEN, W.-C., and RYSER.H. J.-P. (1979). Poiy(L-lysine} and poly(~-lysine} conjugates of methotrexate: Different inhibitory effects on drug-resistant cells. Mol. Pharmacot. 16,614-622. SHEN, W.-C., RYSER,H. J.-P., and LAMANNA, L. (1985). Disulfide spacer between methotrexate and poly(o-lysine). J. Bid. Chem. 260, 10,905-l 0,908. TAKEMOTO, K. K., and FABISCH,P. (1964). Inhibition of herpes virus by natural and synthetic acid polysaccharides. Proc. Sot. fxp. Biol. Med. 116140-144.

USTER,P. S., and DEAMER, D. W. (1985). pH-dependent fusion of liposomes using titratable polycations. Biochemistry 24, l-8. VAHERI,A. (1964). Heparin and related polyionic substances as virus inhibitors. Acta Pathol. Microbial. Stand. Suppl. 171, 7-97. WALTER, A., STEER. C. J., and BLUMENTHAL, R. (1986). Polylysine induces pH-dependent fusion of acidic phospholipid vesicles: a model for polycation-induced fusion. Biochimica Biophys. Acta 861.319-330. WHEAL~, M. E.. RO~BINS,A. K., and ENQUIST,L. W. (1988). Pseudorabies virus glycoprotein gill is required for efficient growth in tissue culture. 1. Viral. 62, 25 12-25 15. WUDUNN, D., and SPEAR, P. G. (1989). Initial interaction of herpes simplex virus with cells is binding to heparan sulfate. J. Viral. 63, 52-58. ZUCKERMANN,F., ZSAK, L., REILLY, L., SUGG, N., and BEN-P• RAT, T. (1989). Early interactions of pseudorabies virus with host cells: Functions of glycoprotein glll. J. Viral. 63, 3323-3329.