Chemotaxis of polymorphonuclear leukocytes to varicella-zoster virus antigens

Microbial Pathogenesis 1991 ; 10 : 451-458

Chemotaxis of polymorphonuclear leukocytes to varicella-zoster virus antigens Toshiaki Ihara,' Naoki Yasuda,' Hitoshi Kamiya,' Sadayoshi Torigoe 2 and Minoru Sakurai 2

'Department of Pediatrics, Mie National Hospital, 357, Ohsato, Kubota, Tsu, Mie, 514-01, Japan and 2Department of Pediatrics, Mie University, 2-174, Edobashi, Tsu, Mie, 514, Japan (Received April 3, 1990 ; accepted in revised form February 5, 1991)

Ihara, T . (Dept of Pediatrics, Mie National Hospital, 357, Ohsato, Kubota, Tsu, Mie, 514-01, Japan), N . Yasuda, H . Kamiya, S . Torigoe and M . Sakurai . Chemotaxis of polymorphonuclear leukocytes to varicella-zoster virus antigens . Microbial Pathogenesis 1991 ; 10 : 451-458 . Chemotaxis of polymorphonuclear leukocytes (PMNs) to various varicella-zoster virus (VZV) antigens was studied using a membrane filter method . Chemotactic activity of PMNs was detected in the presence of sonicated VZV antigen and soluble VZV skin test antigen . This activity was reduced when sonicated VZV antigen was treated with human seropositive serum or murine monoclonal antibodies which reacted with glycoprotein (GP) I or GP II of VZV . However, chemotaxis of PMNs was not reduced when sonicated VZV antigen was treated with human seronegative serum or a murine monoclonal antibody which reacted with GP IV . These results suggest that GP I and GP II act as chemoattractants to PMNs, and this mechanism might contribute to the resolution of the skin lesions of varicella and herpes zoster .

Key words : varicella-zoster virus ; polymorphonuclear leukocytes ; chemotaxis ; glycoproteins .

Introduction Polymorphonuclear leukocytes (PMNs) are the predominant cells in vesicular fluids of patients with herpes zoster . The appearance of increased numbers of PMNs usually precedes the cessation of new lesion formation .' PMNs in vesicular fluids may be capable of killing varicella-zoster virus (VZV)-infected cells via antibody-dependent and independent cytotoxic mechanisms . 2 Exact mechanisms of migration of PMNs into vesicular fluid have not yet been defined . Recently, several biological roles have been proposed for the glycoproteins (GPs) of VZV . G P I and GP II participate in antibody-dependent cellular cytotoxicity against VZV-infected fibroblasts, and GP I participates antibody- plus-complement mediated lysis against VZV-infected targets .' In the present study, we investigated the chemotaxis of PMNs to various VZV antigens . Pretreatment of antigens with murine monoclonal antibodies, which reacted with specific GPs of VZV, identified that those GPs contributed to chemotaxis .

Results

Chemotaxis of PMNs to VZV antigens Small numbers of PMNs migrated in the presence of RPMI 1640 in the lower chamber (see Materials and methods) . The chemotactic index (C .I .) of PMNs in the presence 0882-4010/91 /060451 +08 $03 .00/0

© 1991 Academic Press Limited



452

T . Ihara et al.

Table 1 Chemotaxis of PMNs to VZV antigen, control antigen, and medium Chemoattractant

No . of expt .

Chemotactic index

8 5 4

26 .8+2 .9* 8 .3+3 .7 9 .2+1 .7

Sonicated VZV Control RPM11640

Each value represents mean + SD . * P < 0 .001, Student's t-test compared with either control .

of control antigen was similar to that in the presence of RPMI 1640 . In contrast, PMN migration was significantly increased in the presence of sonicated VZV antigen compared with control antigen and RPMI 1640 (P <0 .01 ; Table 1) . The magnitude of chemotaxis of PMNs decreased with increasing dilutions of VZV antigen (Fig . 1) . Undiluted antigen stimulated the highest chemotaxis of PMNs . Therefore, undiluted antigen was used in subsequent experiments . In other experiments, soluble skin test antigen to VZV and supernatant antigens of VZV-infected and uninfected cells were used as chemoattractants . The mean C .l .s (±SD) of PMNs to soluble skin test antigen to VZV and supernatant antigen of VZV infected cells were 22 .0±4 .5 and 25 .0±3 .4 respectively . These values were significantly higher than those observed in the presence of supernatant antigen of uninfected cells (Table 2) . 40

30

0 E m

U

10

0

Undiluted

1 :5

Dilution of sonicated

1 :25 VZV

Control

antigen

Fig . 1 . Effect of various concentrations of sonicated VZV antigen on chemotaxis of PMNs . VZV antigen was diluted with RPMI 1640 . Each bar represents mean±SD for three different experiments . *P<0 .05; **P < 0 .01 as compared with control antigen .

Table 2 Chemotaxis of PMNs to VZV skin antigen and supernatants of cultured cells Chemoattractant Varicella skin antigen Supernatant of cultured mediums VZV-infected Uninfected

No . of expt .

Chemotactic index

3

22 .0±4 .5*

3 3

25 .0±3 .4** 11 .0±3 .0

Each value represents mean±SD for three different experiments . 'Cultured medium was centrifuged at 750 xg for 10 min and supernatant was used as a chemoattractant . * P <0 .05; ** P <0 .01 as compared with supernatant of uninfected cells .



Chemotaxis of PMNs to VZV antigens

4 53

Table 3 Migration of PMNs for 2 h to VZV antigen and control antigen using the agarose gel method Migration (µm) No . of expt .

No . of wells

VZV antigen

Control antigen

P value

1 2

5 5

85 .50 + 7 .63 72 .35+10 .31

57 .50+ _ 9 .10 51 .20+9 .16

P <0 .001 P <0 .01

In the agarose gel method PMN migration toward sonicated VZV antigen was significantly increased compared with control antigen (Table 3) . Chemotaxis of PMNs from seronegative donors Chemotaxis of PMNs from four seronegative normal children was performed . The C .I . in the presence of sonicated VZV antigen ranged from 19 .1 to 30 .8 with a mean (±SD) of 23 .4±5 .1, which was significantly higher than the C .I . (10 .8±2 .2) in the presence of sonicated control antigen (P < 0 .01) (Table 4) . Effect of antibodies on chemotaxis of PMNs Human sera and murine monoclonal antibodies were inactivated at 56°C for 30 min . Human sera were used undiluted and murine monoclonal antibodies were diluted 1 :20 with phosphate buffered saline (PBS) . Sonicated VZV antigen (150 ul) was treated with 50 .tl of human sera, diluted murine monoclonal antibodies, or PBS at 4°C for 30 min . After treatment, VZV antigen was dispensed into the lower chamber and chemotaxis was performed as described in the Materials and methods section . After treatment with seropositive serum, the C .I . was significantly depressed (P < 0 .001, Table 5) . However, the C .I . was not depressed when VZV antigen was pretreated with seronegative serum . Chemotaxis of PMNs was also reduced after pretreatment of sonicated VZV antigen with mAb 8 which reacted with GP II, and mAb 9 which reacted with G P I (P <0 .01 and P <0 .02, respectively, as compared with pretreatment with seronegative serum) . In contrast, chemotaxis of PMNs was not inhibited by pretreatment of VZV antigen with mAb 12 which reacted with GP IV . These inhibitory effects of mAb 8 and mAb 9 decreased with increasing dilution of these monoclonal antibodies, and when VZV antigen was pretreated with combination of mAb 8 and mAb 9, an inhibitory effect did not increase (Fig . 2) . However, the C .I . of PMNs to zymosan activated serum (ZAS) was not depressed, when 150 MI of ZAS Table 4 Chemotaxis of PMNs obtained from seronegative donors Chemotactic index obtained with different sonicated antigens Donor

VZV

1 2 3 4

19 .1 21 .1 30 .8 22 .7

Mean+SD

23 .4±5 .1 °

" P < 0 .01, Student's t-test .

Control 9 .1 10 .4 13 .9 9 .6 10.8±2 .2°

T. Iha ra et al

454

Table 5 Chemotaxis of PMNs to VZV antigen : effect of treatment with human and murine antibodies' Antibody

No . of expt.

Chemotactic index

Human sera Seropositive Seronegative

4 4

9 .5±2 .9° 21 .0±2 .7

Murine monoclonal mAb 8 mAb 9 mAb 12

5 5 4

11 .1 +2 .9° 14 .2±0 .8t 23 .0+3 .0

None

3

23,1+3 .5

Each value represents mean + SD . a Heat-inactivated human sera and murine ascites fluids were used at final dilutions of 1 :4 and 1 :80 respectively. VZV antigen was incubated with each antibody at 4°C for 30 min prior to testing . P < 0 .01 ; t P < 0 .02 as compared to treatment with seronegative sera, Student's t-test .

was pretreated with 50 yl of 1 :20 diluted mAb 8 and mAb 9 for 30 min at 4°C (Fig . 3) . Discussion Polymorphonuclear leukocytes predominate in the vesicular fluids of patients with herpes zoster, and the number of PMNs increases before cessation of cutaneous and

20 T

V

1

a T T **

T

t U 9FiflF

0

vzv antigen alone

vzv

vzv

vzv

antigen

antigen

mAb8 (1 :80)

mAb8 (1 :400)

antigen + mAb9 (1 :80)

vzv

vzv

antigen antigen + + mAb9 mAb8 (1 :400) (1 :80) mAb9 (1 :80)

VZV Control antigen antigen + mAb8 (1 :400) mAb9 (1 :400)

Fig . 2 . Effect of various dilutions of mAb 8 and mAb 9 to VZV on chemotaxis of PMNs to VZV antigen . Monoclonal antibodies were diluted with PBS and final dilutions were 1 :80 and 1 :400 . After sonicated VZV antigen was pretreated with mAb 8, mAb 9, and combination of mAb 8 and mAb 9 for 30 min at 4°C, chemotaxis of PMNs to VZV antigen was tested by the filter-membrane method . Each bar represents mean±SD for three different experiments . ***P<0 .001 ; **P<0 .01 ; *P<0 .05 as compared with VZV antigen alone .



Chemotaxis of PMNs to VZV antigens

455

40

00

30

x

v c U

u 20 O 0

E

d s

v 10

0 Zymosan

Zymosan

Zymosan

mAbS (1 :80)

mAb9 (1 :80)

PBS

Fig . 3. Effect of monoclonal antibodies to VZV on chemotaxis of PMNs to ZAS . After ZAS was pretreated with mAb 8 and mAb 9 for 30 min at 4°C, chemotaxis of PMNs to ZAS was tested by the filter-membrane method . Each bar represents mean +SD for three different experiments .

visceral dissemination .' The detailed mechanisms involved in PMN migration into vesicular fluid have not been defined . We have speculated that glycoproteins of VZV, which are the major components of the envelope of the virus, might induce chemotaxis of PMNs . We have shown that PMNs migrated significantly in the presence of VZV antigen as compared with control antigen using the filter-membrane method and the agarose gel method . These results suggested that VZV antigen acted as a chemoattractant . We have also shown that PMNs obtained from the peripheral blood of seropositive and seronegative donors are capable of migrating in the presence of various VZV antigens . The magnitude of chemotaxis of PMNs from seropositive donors is similar to that of PMNs from seronegative donors . These results support the in vivo observation that PMNs migrated to the VZV-infected sites of infected individuals regardless of their serological status . A similar phenomenon was observed in studies of chemotaxis to herpes simplex virus antigen ."' The formation of antigen, antibody and complement complexes results in the production of C5a which can induce chemotaxis of PMNs . However, antigen and antibody complexes do not induce chemotaxis . In our study, antibodies to VZV antigens, which had been heat-inactivated at 56°C for 30 min to inactivate complement, reduced chemotaxis of PMNs . We speculated that these antibodies might cover the sites on VZV antigen which induced chemotaxis of PMNs . Several biological properties of GPs of VZV have been reported? In this study, chemotaxis of PMNs was reduced when VZV antigen was pretreated with seropositive human serum and each of two different murine monoclonal antibodies, which reacted with GP I and GP II of VZV . However, chemotaxis of PMNs to ZAS was not reduced when ZAS was pretreated with these monoclonal antibodies . These results suggested that GP I and GP II of VZV might have the ability of inducing PMN migration . In contrast, GP IV did not have such an ability, since mAb 12 did not inhibit chemotaxis of PMNs .



4 56

T . thara et al.

Monoclonal antibodies 8 and mAb 9 inhibited chemotaxis of PMNs with nearly equal magnitude and an inhibitory effect did not increase when VZV antigen was pretreated with combination of mAb 8 and mAb 9 . These results suggested that GP I and GP II might cross-react with a PMN surface to induce chemotaxis . There are two possibilities that GP I and GP II of VZV induce PMN migration . One is GP I and GP II attach to PMN surface and induce PMN migration, and another is GP I and GP II induce intermediary cytokines that mediate the chemotaxis . Interleukin 1 (IL-1) and tumor necrosis factor (TNF), which are secreted from activated monocytes/macrophages, enhance PMN migration .' Monocytes were not contaminated in PMN preparation, and IL-1 and TNF were not detected in the solution of the lower and upper chamber after 4 h incubation in two experiments of the filtermembrane method (data not shown) . These results might suggest that GP I and GP II induce PMN migration without intermediary cytokines . We have not tested murine monoclonal antibody against GP III of VZV . However, G P III and G P IV are secreted from VZV-infected cells, and the soluble skin test antigen to VZV is chiefly composed of GP III and GP IV .' - ' In our studies soluble skin test antigen and cultured media obtained from VZV-infected fibroblasts had significant chemoattractant activity . We speculate that GP III might act as a chemoattractant, since GP IV did not have this property . PMNs are capable of using reactive oxygen species-dependent mechanisms to lyse VZV-infected cells in vitro .' Under appropriate conditions chemoattractants induce granule enzyme secretion and production of reactive oxygen species ." Thus, PMNs, might make some contribution to vesicular formation and have capacity to lyse VZVinfected cells with reactive oxygen species-dependent mechanisms . GP I, GP II and GP III of VZV appear to act as chemoattractants for PMNs . Since GP III are secreted from VZV-infected cells and GP I and GP II are not,' GP III might be the earliest acting attractant in vivo . G P I and GP II are released from lysed VZVinfected cells, which might be lysed by migrated PMNs and natural killer cells ." These GPs might induce more migration of PMNs to the VZV-infected lesions and also induce specific humoral and cellular immunity to VZV . 12, ' 3 In summary, GP I, GP II, and GP III of VZV appear to act as chemoattractants for PMNs, and chemoattracted PMNs might have the ability to lyse VZV-infected cells with reactive oxygen species-dependent mechanism . We speculate that these mechanisms may contribute to the cessation of cutaneous lesions in VZV infections .

Materials and methods

Preparation of PMNs. PMNs were obtained as described by Ferrante and Thong ." Briefly, heparinized peripheral blood from seropositive normal adults and seronegative normal children was layered on Ficoll-Hypaque of 1 .114 density . After centrifugation at 300xg for 30 min, two distinguishable bands were present . The upper band consisted of mononuclear cells and lower band consisted of PMNs . The lower band was harvested with a Pasteur pipette, washed twice with RPMI 1640, and adjusted to a concentration of 2x10 6 /ml in RPMI 1640 with 10% fetal bovine serum (FBS) . Mean purity (±SD) of PMNs was 98 .3±0 .5% in 10 experiments in Wright-stained preparation . Less than 3% of lymphocytes were contaminated in the PMN preparation . Except where specifically noted, PMNs from the seropositive adult donors were used in all studies . Preparation of VZV antigens . For preparation of whole VZV antigen, monolayers of human embryonic fibroblasts (HEF) in 75 cm 2 flasks were infected with the Kawaguchi strain . When the cultures showed 3 to 4+ cytopathic effects, cells were suspended with trypsin, centrifuged once at 750xg (1500 rpm) for 30 min, resuspended with 2 ml PBS per one 75 cm 2 flask, and sonicated for 5 s . Sonicates were clarified by centrifugation at 1200xg (2000 rpm) for 20 min .



Chemotaxis of PMNs to VZV antigens

457

Supernatants were harvested and stored at -70°C prior to use . Control antigen was prepared from uninfected HEF in the same fashion . Both antigens were heat-inactivated at 56°C for 30 min prior to use . For preparation of supernatant antigen, when VZV-infected cells showed 3 to 4+ cytopathic effects, supernatant was harvested and centrifuged at 750xg (1500 rpm) for 10 min . Supernatants were then stored at -70°C prior to use . Supernatant control antigen was prepared from uninfected HEF in the same fashion . Skin test antigen of varicella-zoster virus . A soluble varicells-zoster virus antigen for skin testing was kindly provided by Dr Takahashi, Research Institute for Microbial Diseases, Osaka University, Suita, Japan .' The soluble skin test antigen to VZV was purified from cultured media of VZV-infected cells and did not contain any preservatives . Preparation of ZAS. Zymosan activated serum was prepared as described previously." Briefly, 25 mg zymosan was put into 1 ml fresh human serum and the mixture was treated at 37°C for 30 min . After 10 min centrifugation at 500xg, the supernatant was harvested . The supernatant was treated at 56°C for 30 min and diluted to 10% with PBS . Zymosan activated serum was stored at -20°C prior to use . Chemotaxis of PMNs. Assays for chemotaxis of PMNs were performed by the filter-membrane method described by Boyden ." Briefly, 0.2 ml of VZV antigens or control antigen was dispensed into the lower chamber, and 0 .2 ml of PMN preparations were dispensed into the upper chamber . The two chambers were separated by a 3 .0 pm filter (Millipore, Bedford, Massachusetts) . The chambers were incubated for 4 h at 37°C in 5% CO 2 . After incubation, filters were fixed with ethanol, stained with hematoxylene-eosin and made transparent with xylene . The numbers of PMNs on the upper surface and those on the lower surface in the same field were counted with a microscope . More than 10 different fields were counted on each filter . Total numbers of PMNs on different fields of the upper surface were defined as A and those of the lower surface were defined as B . The C .I . was calculated using the following formula : C .I . = B/A+B x 100,

Another assay for chemotaxis of PMNs was performed by the agarose gel method described by Nelson et al." Briefly, the center well received a 5 ul volume of the PMN suspension containing 5x10 5 purified PMNs . The outer well received 5 p1 of sonicated VZV antigen and inner well received 5 yl of control antigen . The completed dishes were incubated for 2 h at 37°C in a humidified atmosphere containing of 5% of CO 2 . Quantification of migration was done by measurement of linear distance the cells had moved from margin of the well toward the VZV antigen and the linear distance the cells had moved from the margin of the well toward the control antigen with an inverted microscope . Preparation of seropositive serum, seronegative serum and murine monoclonal antibodies to GPs of VZV. A seropositive serum [titer of 64 by fluorescent antibody to membrane antigen (FAMA)] 18 was obtained from a normal adult and a seronegative serum (titer less than 2 by FAMA) was obtained from a normal child . Three monoclonal antibodies to VZV were kindly provided as ascites fluid by Dr Yamanishi, the Research Institute for Microbial Diseases, Osaka University, Suita, Japan .' Characteristics of these monoclonal antibodies are shown in Table 6 .

Table 6 Characteristics of murine monoclonal antibodies to VZV Antibody

GP'

NT'

IgG type

FAMA`

mAb 8 mAb 9 mAb 12

11 1 IV

Cd

IgG 2 b IgG 2 a IgG 2 a

25600 25600 25600

-

Glycoprotein of VZV . ' Neutralization . Fluorescent antibody titers to membrane antigen . "Complement dependent neutralization . 'Negative . a



458

T . Ihara et al

Statistics. Differences between chemotactic indexes and linear distances were analysed using Student's t-test .

References 1 . Stevens DA, Ferrington RA, Jordan GW, Merigan TC . Cellular events in zoster vesicles : relation to clinical course and immune parameters . J Infect Dis 1975 ; 131 : 509-15 . 2 . Ihara T, Starr SE, Ito M, Douglas SD, Arbeter AM . Human polymorphonuclear leukocyte-mediated cytotoxicity against varicella-zoster virus infected fibroblasts . J Virol 1984; 51 : 110-16 . 3 . Ito T, Ihara T, Grose C, Starr SE . Human leukocytes kill varicella-zoster virus-infected fibroblasts in the presence of murine monoclonal antibodies to virus-specific glycoproteins . J Virol 1985; 54 : 98-103 . 4 . Snyderman R, Wohlenberg C, Notkins AL . Inflammation and infection : chemotactic activity resulting from the interaction of antiviral antibody and complement with cells with herpes simplex virus . J Infect Dis 1972; 126 : 207-9 . 5 . Ward P, Cohen S, Flanagan TD . Leukotactic factors elaborated by virus-infected tissue . J Exp Med 1972; 135 : 1095-104 . 6 . Steinbeck MJ, Roth JA . Neutrophil activation by recombinant cytokines . Rev Infect Dis 1989 ; 11 : 549-68 . 7 . Asano Y, Shiraki K, Takahashi M, Nagai H, Ozaki T, Yazaki Y . Soluble skin test antigen of varicellazoster virus prepared from the fluid of infected cultures . J Infect Dis 1981 ; 143 : 684-92 . 8 . Okuno T, Yamanashi K, Shiraki K, Takahashi M . Synthesis and processing of glycoproteins of varicellazoster virus (VZV) as studied with monoclonal antibodies to VZV-antigen . Virol 1983 ; 129 : 357-68 . 9 . Namazue J, Campo-Vera H, Kitamura K, Okuno T, Yamanishi K . Processing of virus-specific glycoproteins of varicella-zoster virus . Virology 1985 ; 143 : 252-9 . 10 . Lehmyer JE, Snyderman R, Johnston RB, Jr. Stimulation of neutrophil oxidative metabolism by chemotactic peptides: influence of calcium on concentration and cytochalasin B and comparison with stimulation by phorbol myristate acetate . Blood 1979 ; 54 : 35-45 . 11 . Ihara T, Starr SE, Arbeter AM, Plotkin SA . Effects of interferon on natural killing and antibodydependent cellular cytotoxicity against varicella-zoster virus-infected and uninfected target cells . J Interferon Res 1983 ; 3 : 263-9 . 12 . LaRussa PS, Gershon AA, Steinberg SP, Chartrand SA . Antibodies to varicella-zoster virus glycoproteins I, II and III in leukemic and healthy children . J Infect Dis 1990; 162 : 627-33 . 13 . Watson B, Keller PM, Ellis RW, Starr SE . Cell-mediated immune responses after immunization of healthy seronegative children with varicella vaccine : kinetics and specificity . J Infect Dis 1990 ; 162 .794-99 . 14 . Ferrante A, Thong YH . Optimal conditions for simultaneous purification of mononuclear and polymorphonuclear leukocytefrom human peripheral blood by Hypaque-Ficoll method . J Immunol Methods 1980 ; 36 : 109-17 . 15 . Ward PA, Cochrane CG, Muller-Eberhard HJ . The role of serum complement in chemotaxis of leukocytes in vitro . J Exp Med 1965 ; 125 : 327-46 . 16 . Boyden S . The chemotactic effect of mixtures of antibody and antigen on polymorphonuclear leukocytes . J Exp Med 1962 ; 115: 453-66 . 17 . Nelson RD, Quie PG, Simmons RL. Chemotaxis under agarose : a new and simple method for measuring chemotaxis and spontaneous migration of human polymorphonuclear leukocytes and monocytes . J Immunol 1975; 115 :1650-6 . 18 . Wiliams A, Gershon A, Brunell P . Serological response to varicella-zoster membrane antigens measured by indirect immunofluorescence. J Infect Dis 1974 ; 130 : 669-72 .