The IgG Receptor Induced by Herpes Simplex Virus: studies using Radioiodinated IgG

Transcription

1 J. gen. Virol. (1974), 24, 167 I78 Printed in Great Britain I67 The IgG Receptor Induced by Herpes Simplex Virus: studies using Radioiodinated IgG By DIANA WESTMORELAND AND J. F. WATKINS Sir William Dunn School of Pathology, Oxford University, Oxford, England (Accepted I March I974) SUMMARY Further evidence is presented that the receptor induced on the surface of cells infected with herpes simplex virus is specific for the Fc fragment of the IgG molecule of several species. A method has been developed for quantitative studies of this receptor using rabbit IgG labelled with [125I]. The effects of actinomycin D, puromycin, and cytosine arabinoside on the development of the receptor during the virus growth cycle are described. Evidence was obtained of a turnover of membrane receptors after infection. INTRODUCTION The plasma membranes of cells infected with herpes simples virus develop the ability to bind sheep erythrocytes sensitized with rabbit anti-sheep erythrocyte serum (Watkins, I964, I965). The failure of the (Fab')~ fragment of immune gamma-globulin against sheep erythrocytes to mediate the reaction led to the suggestion that the phenomenon was caused by the appearance of receptors for the Fc fragment of immune gamma-globulin in the cell membrane (Yasuda & Milgrom, I968). Evidence has also been presented that the receptor has properties in common with the cytophilic antibody receptor of macrophages (Shimizu, I97I). The adsorption of erythrocytes to cells is not accurately quantifiable. The main purpose of this paper is to describe a simple quantitative method of studying the receptor using purified immune gamma-globulin labelled with [1~5I], and to report some of the results obtained. METHODS Virus. Herpes simplex virus type I (HFEM strain) was grown in BHK cells cultured in minimal Eagle's medium with 2 ~o foetal calf serum and maintained at 32 C to maximize virus yield. After 24 h incubation the infected cells were subjected to ultrasonic disintegration in Eagle's medium. The suspension was centrifuged at 5oo g for 5 min and the supernatant fluid stored at -7o C. The stock virus was titrated on monolayers of CVI cells in Microtest trays (Falcon Plastics Ltd). The titration end-point was determined by the method of Reed & Muench (1938). One TCDso thus obtained has been shown in this laboratory to be approx. I p.f.u. Experiments were arranged such that IOO ~ of the cells were infected; this was achieved by using an input multiplicity of 5 to 2o TCDs0cell. Replicate assays of a given virus stock did not differ by more than o'5 log TCDs0. Confluent cultures of cells on coverslips were placed in a 60 mm plastic Petri dish (Falcon plastics) and overlayed with 0"5 ml virus suspension containing IO 7 to Ios TCDs0. After incubation at 37 C for I h, 4 ml of medium was added and the incubation continued

2 I68 D. WESTMORELAND AND J. F. WATKINS at 37 C in an atmosphere of 5 % CO2 in air. Infection of Ioo % of cells was confirmed by examination of Giemsa stained coverslips for herpes induced nuclear changes. Cells. All cells were routinely cultured in Eagle's minimal essential medium (MEM, Bio-Cult Laboratories Ltd, Scotland), supplemented with ~o % foetal calf serum and Ioo units per ml Crystapen Benzylpenicillin and zoozgml Streptomycin Sulphate. Cells were grown in plastic tissue culture flasks (Falcon Plastics Ltd) and harvested using o-i25 % trypsin + o-zo % versene solution. The line of HeLa cells used was obtained from Porton Microbiological Research Establishment, England. The line of baby hamster kidney cells (BHK 20 was obtained from Flow Laboratories, Irvine, Scotland (Macpherson & Stoker, I962). The line of African green monkey kidney cells (CVI) was obtained from Bio-Cult Laboratories. Rabbit fibroblast cells were obtained from explants of a biopsy taken from the ear of an adult rabbit. Human fibroblast cells were derived from a sample of human embryo lung tissue. Primary mouse kidney cells were derived by trypsinization of kidneys of baby mice. Primary rabbit kidney cells were obtained by trypsinization of kidneys of a baby rabbit. Erythrocytes. Sheep erythrocytes preserved in Alsever's solution were obtained from Wellcome Research Laboratories, r Beckenham,' England. Cells used in experiments were not more than 2 weeks old. Chick erythrocytes were obtained from the wing vein of Rhode Island Red hens and preserved in Alsever's solution. Fresh cells were used for each experiment. Erythrocytes of other species were taken either from the heart or from the abdominal aorta of the appropriate animal. All erythrocytes were washed three times in phosphatebuffered saline (PBSA, Dulbecco & Vogt, 1954) before use in experiments. Antisera. All antisera were heated at 56 C for 30 min to inactivate complement, and stored at -20 C, except those from Wellcome Research Laboratories which were preserved in 20 % glycerol and stored at 4 C. Rabbit anti-sheep erythrocyte serum[was purchased from Wellcome Research Laboratories, Beckenham, England (haemagglutination titre ~512). Horse anti-sheep erythrocyte serum was also obtained from Wellcome Research Laboratories (haemagglutination titre I7oo). Rabbit anti-chick erythrocyte serum was obtained from a rabbit which had received repeated injections of Io 7 chick erythrocytes intravenously (haemagglutination titre I4OO). Rabbit anti-herpes type I antiserum was produced in a rabbit which had received repeated intravenous injections of Io 6 infected primary rabbit kidney cells (24 h post-infection) in I ml of PBSA. The rabbit was subsequently bled and the neutralizing titre of the serum for herpes simplex virus determined. Foetal calf serum was purchased from Flow Laboratories Ltd, Irvine, Scotland. Normal rabbit serum was obtained by bleeding from the ear vein of an adult rabbit. Normal rabbit immune gamma globulin (IgG) was purchased from Nordic Pharmaceuticals Sera Service Ltd, Tilburg, The Netherlands, as a lyophilized powder. Fc fragment of Rabbit IgG was a gift from Dr E. M. Press, M.R.C. Immunochemistry Unit, Oxford University. Haemagglutination titration. Twofold dilutions of serum in PBSA were prepared in a Salk pattern haemagglutination tray in 0"5 ml vol. To each well was added 0"5 ml of' a 2 % suspension of erythrocytes in PBSA. The tray was gently agitated, and incubated at 4 C for I to 2 h. The end-point of the titration was estimated as the serum dilution at which 5o % agglutination had occurred. Virus neutralization assay. Sera from immunized rabbits were serially diluted in MEM and incubated at 37 C for I h with an equal vol. of a virus suspension of known titre. A sample of virus was incubated in the absence of serum as a control. The incubation mixtures

3 HSV-induced IgG receptor 169 were then diluted and re-titrated in CVI cells. It was found that a 114oo dilution of antiserum reduced the titre of stock virus by 99 ~. Haemadsorption assay. Haemadsorption was routinely carried out by a slight modification of 'Procedure I' of Yasuda & Milgrom (I968). Infected or uninfected cells on glass coverslips (I I mm diam.) were incubated at 37 C for I h in rabbit anti-sheep erythrocyte serum diluted I in 5o in PBSA, washed thoroughly in PBSA and then incubated at 37 C for 30 min in a I ~o (vv) suspension of erythrocytes in PBSA. At the end of this incubation period the coverslips were again washed thoroughly in PBSA. Adhering erythrocytes could be seen on examination of the coverslips under a low-power microscope. The preparations were fixed in 1"25 ~ (wv) glutaraldehyde in PBSA and stained in IIO Giemsa stain for examination under higher power (Fig. I). Labelling antiserum with [1~5I]. Purified rabbit IgG (Nordic Pharmaceuticals) was labelled with [125I] by the chloramine T method (Byrt & Ada, I969) modified as described by Jensenius & Williams (I973). Ten #l of sodium iodide (sp, act. 14 mci#g) in NaOH ph 8 to II was mixed with IO #g of IgG and IO #1 chloramine T solution at ph 7"3. The reaction was stopped by the addition of excess tyrosine. The reaction mixture was fractionated on a Sephadex G5o column, which gave good separation of labelled IgG from small mol. wt. contaminants. Rabbit Fc and (Fab')~ fragments were treated similarly. Labelled antibody preparations were diluted in 5 ~ FCS in PBSA. The sp. act. of the [125I]-IgG solution was estimated to be IO 7 ctmin~g, on the assumption that all the IgG loaded on to the Sephadex G5 o column was recovered. On this basis the sp. act. was approx. 2 IO -6 ctminmol. [1251]-IGG binding assay. All binding assays were carried out in the following way: The uptake of [125I]-IgG by cells was determined by incubating washed unfixed coverslip cultures of cells (about lo s cells per coverslip) in IO #1 (approx. lo 5 ctmin) of diluted [125I]- IgG at 37 C for 30 min except when otherwise stated. After incubation in labelled IgG, the coverslips were washed thoroughly in PBSA, allowed to dry and immersed in 5 ml of scintillation fluid, IO g 2,5-diphenyloxazole (Fisons, Loughborough, England) and 0.2 g p-bis-( -methylstyryl)-benzene (Packard Instrument Co., Illinois, U.S.A.) in 2"5 1 toluene). The radioactivity was measured in a Nuclear Chicago Scintillation Counter Model 724 using integral counting procedure. Inhibitors. Cytosine arabinoside HC1 was purchased from Koch-Light Laboratories, Colnbrook, England. IO #gml reduced the incorporation of [3H]-thymidine into 5 ~ trichloracetic acid (TCA) insoluble material in herpes simplex infected HeLa ceils by 95 ~- Actinomycin D was obtained from. BDH Biochemicals, London, England. One #gml reduced uptake of [~H]-uridine into TCA-insoluble material by 95 ~o within a few minutes. Puromycin was supplied by Serva Feinbiochemica, Heidelberg, West Germany. At concentrations of inhibitor in excess of IOO #gml incorporation of [14C]-amino acids into TCA-insoluble material was reduced by at least 9o ~. RESULTS Evidence against a common antigen in herpes-infected cells and sheep erythrocytes Further evidence against the idea that haemadsorption could be due to the appearance on herpes-infected cells of an antigen in common with sheep erythrocytes was obtained by studying the adsorption of several species of erythrocytes to infected cells treated with varying dilutions of rabbit antisera against sheep erythrocytes and chicken erythrocytes. The results in Table I show that haemadsorption was produced only by the homo-specific antiserum. If common antigens are responsible at least two antigens must therefore be

4 D. W E S T M O R E L A N D I7O A N D J. F. W A T K I N S 5, te~ p i e* N e ID O i i ~i~ ~ : i,~:~5~&... +,,2 Fig. I. Sensitized sheep erythrocyte adsorbed to HeLa cells 16 h after infection with herpes simplex virus. Photographed at 54o x magnification. T a b l e I. Haemadsorption to herpes-infected HeLa cells of erythrocytes sensitized with rabbit antiserum against sheep or chick erythroeytes* Antiserum used and the highest dilution at which it produced haemadsorption f Source of erythrocytes Sheep Chick Rat Rabbit Mouse Anti-chick erythrocyte Anti-sheep erythrocyte Normal serum i8 I4oo < I8 116oo < 14 < I8 < I8 < 18 < I8 I8 18 < 18 < 18 < I8 < 18 * Haemadsorption as described in Methods.

5 HSV-induced IgG receptor I 7~ Table 2. Inhibition of haemadsorption by normal rabbit serum ~* of infected HeLa cells showing haemadsorption when the rabbit anti-erytbrocyte serum was diluted in: Dilution of c ' anti-erythrocyte Normal rabbit Foetal calf PBSA serum t serum serum I5o I3'6 7~'7 I 1 oo "4 l500 1 "9 59"7 1]ooo o.3 57"6 I I "2 * Mean of three coverslip cultures. t Diluted I in 20 in PBSA. 69"I 63q 63'I 49" Table 3. Haemadsorption by herpes-infected cells treated with rabbit or horse anti-sheep erythrocyte serum and sheep erythrocytes Cell Rabbit serum* Horse serum* HeLa Human epithelial line HF Human fibroblast strain CVI African green monkey kidney epithelial line BSC African green monkey kidney epithelial line + + Not done BHK Hamster fibroblast line RF Rabbit fibroblast line MK Mouse primary kidney + + Not done * + +, All infected cells covered in erythrocytes. +, Most cells with many erythrocytes adhering. +, Occasional cells (i ~) with a few erythrocytes adhering. -, No haemadsorption. present on the surface of a herpes infected cell, one in common with sheep erythrocytes, and the other in common with chick erythrocytes, which seems improbable. Yasuda & Milgrom 0968) found that the (Fab')~ fragment produced by pepsin digestion of rabbit anti-sheep erythrocyte IgG failed to mediate haemadsorption, and suggested that the phenomenon was due to attachment of antibody by its Fc end. The failure reported by Watkins (1965) to inhibit the reaction by pre-treating infected cells with normal rabbit serum before addition of sensitized erythrocytes was at variance with this interpretation. The effect of normal rabbit serum on the reaction was therefore re-examined. It was possible that the attachment of sensitized erythrocytes could have occurred to new receptors for immunoglobulin which appeared after the removal of normal serum. Serial dilutions of rabbit anti-sheep erythrocyte antibody were therefore made in either normal rabbit serum or foetal calf serum before their addition to herpes infected cells. A clear inhibition of haemadsorption was seen in the dilutions made in normal serum. Foetal calf serum, which contains little immunoglobulin, did not inhibit haemadsorption (Table 2). Rabbit antiserum against herpes simplex virus was made as described, by intravenous injections of infected rabbit kidney cells. The antiserum produced after three such injections neutralized over 99 ~ of herpes simplex virus at a dilution of ~4oo but did not agglutinate sheep erythrocytes at a dilution greater than I8. Normal rabbit serum agglutinated sheep erythrocytes at this dilution. Samples of the infected rabbit cells which had been used to induce anti-herpes antibody showed strong haemadsorption with sensitized erythrocytes. If the receptor had been identical with an antigen on sheep erythrocytes, antiherpes antiserum would be expected to agglutinate these cells. Finally, when rabbit anti-sheep erythrocyte serum and horse anti-sheep erythrocyte

6 I72 D. WESTMORELAND AND J. F. WATKINS I I I I I [ < = 10, O l ~ I I! I Virus Time after infeciion (h) Fig. 2. A0pearance of receptor for [12sI]-lgG on HeLa cells infected with herpes simplex virus. Ordinate: mean ctmin of [12sI]-IgG adsorbed to three coverslip cultures., infected cells; O O, uninfected cells. serum were compared for their ability to produce haemadsorption to cells of different species which had been infected 24 h previously with herpes simplex virus it was found that horse serum would not produce haemadsorption to human or monkey cells (Table 3). This result not only rules out the possibility of a common antigen, but also shows that some of the properties of the receptor must be determined by the host cell. Uptake of [125I]-labelled normal rabbit IgG by herpes-infected cells The strong evidence that the haemadsorption phenomenon was not mediated by a common antigen, but by some portion of the Fc end of the antibody molecule, suggested that purified IgG from a rabbit which had not been immunized against erythrocytes would bind as well as anti-erythrocyte serum to herpes-infected cells. This supposition proved to be correct. No agglutinating activity against sheep erythrocytes could be demonstrated :in the preparation of IgG used in these experiments; nevertheless the labelled IgG bound strongly. In Fig. 2 is shown the result of an experiment in which coverslip cultures of HeLa cells were treated for 3o min periods at 37 C with labelled IgG at intervals after infection. The IgG-binding capacity of the infected cells increased exponentially from the time of infection. The shape of the curve is broadly similar to that reported by Watkins (I965) fi~r the rate of appearance of the ability of infected cells to bind sensitized erythrocytes. From the estimated sp. act. of the IgG it was calculated that about 5 x Io 4 molecules of IgG were bound per infected cell in 3o min at 37 C at ~ I h after infection.

7 HS V-induced lgg receptor I73 I I I I 37 C ? 7 x.~ 6 ~- 15 :C C 1 I I I I Incubation time ill [lzsl]-igg (min) Fig. 3- Binding of [125I]-IgG to HeLa cells infected with herpes simplex virus and incubated at 4, I5 or 37 C in p~5i]-igg between I6 and 18 h after infection. The results shown in Fig. 3 to 7 are corrected for [l~ii-igg binding to uninfected cells. Table 4. Adsorption of p~5i]-igg or [125I]-IgG fragments to infected and uninfected HeLa cells Ratio of ctmin bound to infected Ctmin Infected cells* Uninfected cells and Material added per bound per coverslip bound per coverslip uninfected added coverslip (ctmin) (ctmin) cells Rabbit IgG 3'2 I0 ~ (s.e. = 512) 3668 (s.e. = +235 ) 6'19 Human IgG 085 IO ~ (s.e. = + 302) 2242 (s.e. = + 22) 4'68 Horse IgG 3"6x IO 5 3 IO8 (s.e. = + I48) 28oo (s.e. = ) i-ii Rabbit Fc 3-I x io (s.e. = + 867) 3536 (s.e. = +34o) 3"oi Rabbit Fab'~ 6.2 x lo 5 55o8 (s.e. = + I311) 3578 (s.e. = +203) 1'53 * Cells infected I8 h previously. When coverslip cultures of cells which had been infected 16 h previously were incubated at 37 C in labelled IgG, and the amount of binding determined on groups of three coverslips removed from the IgG at intervals, it was found that, unexpectedly, the amount of igg bound did not reach saturation during the 2 h of observation; at 4 C and ]5 C, however, a plateau was reached (Fig. 3). These results suggested that synthesis of the 12 VIR 24

8 174 D. WESTMORELAND AND J. F. WATKINS I I I I I I O P loc I I I I 0" y D f = 50 u I I I I I I I I I Time after infection (h] Time (rain) Fig. 4 Fig. 5 Fig. 4- Effect of cytosine arabinoside 0o #gml) added at intervals after infection on the amount of [125I]-IgG bound at Io h after infection. Ordinate: mean ctmin of three coverslip cultures expressed as a percentage of the ctmin of untreated, infected coverslip cultures at Io h after infection. ---0, infected cultures not treated with cytosine arabinoside;, infected cultures treated with ara C from a time indicated by the start of the solid line. Fig. 5. Effect of continued incubation in inhibitors from I6 to I8 h after infection on the binding of p25i]-igg (also present from 16 to 18 h after infection). Ordinate: mean ctmin of three coverslip cultures expressed as percentage of ctmin of infected cultures incubated from 16 to 18 h after infection in [12~I]-IgG alone, x x, no inhibitor;, cytosine arabinoside io #gml; - - -, actinomycin D (I #gml);, puromycin (2oo #gml). receptor and its incorporation into the membrane were still occurring as late as I6 h after infection. Purified human IgG and purified Fc fragment of rabbit IgG labelled with [125I] also adsorbed to infected cells. Radioiodinated horse IgG and Rabbit (Fab')~ fragment bound very little more to infected HeLa cells than to uninfected cells (Table 4). Effect of cytosine arabinoside (ara C) on IgG binding Ara C was added to coverslip cultures of HeLa cells at a concentration of Io~gml at intervals after infection (when the amount of igg binding to control infected cells was determined), and incubation was continued until IO h after infection, when IgG binding was measured (Fig. 4)- Incubation in ara C from I, 2, 3 or 4 h after infection to Io h after infection did not prevent some increase in the amount of IgG which could be bound, After 5 h, however, incubation in ara C led to a decrease in the amount of IgG which couht

9 HSV-induced IgG receptor I75 I I I I 100 P ~ -,c c\ "'---o 7 o o - v o l0 i I I i i Time alicr infection (h) Fig. 6. Effect of actinomycin D (Itgml) added at intervals after infection on the amount of [:~q]- IgG bound at IO h after infection. Ordinate: mean ctmin of three coverslip cultures expressed as a percentage of the mean ctmin of untreated, infected cultures at so h after infection., infected cultures not treated with actinomycin D; O, infected cultures treated with actinomycin D from a time indicated by the start of the solid line. be bound compared with the amount which could be bound at the time ara C was applied. This result is consistent with loss of receptors from the membrane during the period, possibly because of turnover. When infected cells were incubated from I6 h after infection in a mixture of ara C (Io~gmg) and labelled IgG, the binding of IgG did not differ significantly from the control curve (Fig. 5)- Watkins 0965) reported that 5-iodo-deoxyuridine had no effect on haemadsorption. In parallel with the IgG binding experiments, therefore, the effect of ara C on haemadsorption was examined. It was found that after incubation in the presence of io #gml of ara C for I8 h from the time of infection, infected cells bound erythrocytes in a similar pattern to that reported for cells in 5-iododeoxyuridine. That is, at multiplicities of infection of less than one infectious unit per cell, the initially infected cell showed strong haemadsorption, but there was no spread of infection to surrounding cells, or syncytium formation. 12-2

10 176 D. WESTMORELAND AND J. F. WATKINS I I I I I I I00 c.i )v'-2-" - * I I I I I I Incubation time in puromycin (min) Fig. 7. Effect of continued incubation in puromycin (2oo#gml) from 16 to 18 h after infection on the binding of [12~I]-IgG. Groups of three coverslips were removed from inhibitor at the times shown, and the amount of [125I]-IgG binding in a 30 min application was determined. O, continued incubation in puromycin;, puromycin removed after 6o min and incubation continued in medium without inhibitor;, puromycin removed after 9o rain and incubation continued in medium without inhibitor. Ordinate: mean ctmin of three coverslip cultures expressed as percentage of mean ctmin of three infected coverslip cultures treated with [a~5i]-igg at I6 h after infection. Effect of actinomycin D on IgG binding Actinomycin D was added to coverslip cultures of HeLa cells at a concentration of I #gml at intervals after infection, and incubation was continued until IO h after infection, when IgG binding was measured. In Fig. 6 the amount of IgG bound to actinomycin-treated cells at IO h after infection is compared with the amount of IgG binding to control infected cells at the times that actinomycin was added. For most of the intervals there was little or no difference between the amounts of IgG which could be bound at the beginning and the end of incubation in actinomycin D. From 6 h after infection incubation in actinomycin D led to a decrease in the amount of IgG which could be bound at the end of incubation in the inhibitor compared with what could be bound at the time it was applied. This could be explained by an increase in turnover of receptors from about 6 h after infection, at which time progeny virus DNA would be available for transcription. The constancy of the amounts of IgG bound in the earlier part of the figure can be attributed either to stability of messenger RNA combined with turnover, or instability of messenger RNA with no turnover of membrane receptor. When infected cells were incubated continuously from 16 to 18 h after infection in the presence of [xzsi]-igg and actinomycin D, the amount of IgG binding steadily increased as it did in the control infected cells (Fig. 5). The result shows that the steady increase in the amount of IgG bound to the control infected cells at this late stage did not depend upon continuing synthesis of messenger RNA, and is consistent with stability of messenger RNA. This, in turn, suggests that the hypothesis of stable messenger combined with some membrane turnover in the first 6 h after infection is the correct one.

11 HS V-induced IgG receptor 177 Effect of puromycin on IgG binding When infected cells were incubated between 16 and 18 h after infection in the continued presence of puromycin (2oo #gml) and IgG, the cumulative amount of IgG bound was less than to control cells; this difference was apparent after lo rain incubation (Fig. 5). This result was in contrast to the results obtained with actinomycin D and cytosine arabinoside under the same conditions (Fig. 5), and showed that the increase in binding observed when infected cells were incubated with either of these two inhibitors required continuing protein synthesis. Furthermore, the absence of a decrease in the amount of IgG remaining bound in the absence of protein synthesis showed that there was no loss of receptors and their associated IgG into the medium. In another series of experiments infected cells were incubated between 16 and 18 h after infection in the continued presence of ioo #gml puromycin, and the amount of IgG binding in a 30 min application was determined at intervals after the addition of puromycin. At two points, 6o min and 9o min after the addition of puromycin some coverslips were removed from the inhibitor and washed, and their incubation was continued in inhibitorfree medium. The amount of IgG which could be bound decreased steadily during incubation in puromycin, but rose again in the cells which were removed from puromycin (Fig. 7). Since the earlier experiments had shown that IgG receptors were not being shed into the medium when infected cells were incubated in puromycin, the decline in IgG binding seen in Fig. 7 must have been due either to destruction of receptors which were not 'stabilized' in some way by binding with IgG, or else to their removal from the cell surface by pinocytosis or a similar mechanism. DISCUSSION These results establish the fact that a receptor for the Fc fragment of rabbit IgG appears at the surface of cells infected with herpes simplex virus. This receptor is not present on uninfected cells, even after brief trypsin treatment (unpublished observation). As judged by the results of haemadsorption tests (Watkins, I965; Yasuda & Milgrom, 1968) production of the receptor is probably associated with infection by all herpes simplex type strains; haemadsorption has also been observed to cells infected with type 2 strains (unpublished observation). So far, in this laboratory, no cell species has been found not to show haemadsorption after infection with herpes simplex virus. Since receptors for the Fc fragment of IgG are present on normal macrophages (Berken & Benacerraf, 1966) and B lymphocytes (Basten, Warner & Mandel, 1972) the potentiality for production of the receptor must exist in all cells. The receptor could therefore either be the product of a cell gene switched on by infection, or of a virus gene. The results obtained with actinomycin D are consistent with the view that messenger RNA synthesis is required for receptor formation during the first lo h after infection; cellular RNA synthesis, on the other hand, is greatly reduced after infection with herpes simplex virus (Flanagan, i967). These facts favour the view that a virus gene is involved. It is also possible that the insertion of virusdetermined glycoproteins into the cell membrane (Roizman, I972) unmasks Fc receptors present in the membrane in a cryptic form. The turnover of receptor demonstrated in the puromycin experiments implies continuing synthesis of receptor in infected cells rather than unmasking and masking of molecules synthesized and inserted in the membrane before infection. A solution of the problem must await further experiments, but at present the explanation which makes the results easiest to understand is that herpes simplex virus codes for a protein which is either the receptor itself, or is glycosylated by cellular enzymes

12 I78 D. WESTMORELAND AND J. F. WATKINS to form the receptor, possibly because of similarities between the virus protein and the protein portion of the normal cellular Fc receptor of macrophages and B lymphocytes. The latter possibility would explain the apparent role of the host cell in the specificity of the receptor observed in the experiments with horse anti-sheep erythrocyte serum. The protein is synthesized throughout virus development, as a result of transcription from both infecting and progeny DNA. The biological significance of the receptor is at present a matter for speculation only. The fact that human IgG binds to herpes-infected cells suggests a possible role for the receptor in maintaining latent herpes infection, since the coating of infected cells with IgG molecules attached by their Fc end may conceivably protect the cell from destruction by cytotoxic antibodies and lymphocytes. This possibility is under investigation. Whether or not the phenomenon plays any role in oncogenesis by herpes viruses must await further investigation. There is one report in the literature of adsorption of sensitized sheep erythrocytes to Burkitt's lymphoma cells (Kumagai & Minowada, i97o), but this is so far unconfirmed. Duff & Rapp 0970 reported the transformation of Syrian hamster embryo cells by herpes simplex virus inactivated by u.v. light. We have examined several of their transformed lines and have found that many of them showed adsorption of sensitized erythrocytes and radio-iodinated IgG (D. Westmoreland, J. F. Watkins & F. Rapp, unpublished observations). We are grateful to Miss Sue Freebury for technical help, to Miss Christine Court for drawing the graphs, and to Mr S. Buckingham for the photography. Diana Westmoreland is in receipt of a Scholarship for Training in Research Methods from the Medical Research Council of Great Britain. REFERENCES BASTEN, A., WARNER, N. L. & MANDEL, X. (I972). A receptor for antibody on B lymphocytes. II. Immunochemical and E.M. characteristics. Journal of Experimental Medicine x35, BERKEN, A. & BENACERRAr, B. (X 966). Properties of antibodies cytophilic for macrophages. Journal of Experimental Medicine xz3, I ~9-t44. BYRT, V. & ADA, G. L. 0969). An in vitro reaction between labelled flagellin or haemocyamin and lymphocytelike cells from normal animals. Immunology XT, 5o DUFF, R. & RAPP, r. (I97I). Oncogenic transformation of hamster cells after exposure to herpes simplex virus type 2. Nature New Biology 233, DULBECCO, R. & VO6T, M. (I954). Plaque formation and isolation of pure lines with poliomyelitis viruses. Journal of Experimental Medicine 99, I67-I82. rlanagan, J. e. (I967). Virus-specific ribonucleic acid synthesis in KB cells infected with herpes simplex virus. Journal of Virology x, o. JENSENIUS, J. C. & WILLIAMS, A. V. (I973)- The IgG receptor induced by herpes simplex virus: studies using radioiodinated IgG. European Journal oflmmunology (in the press). KUMAGAI, K. & MINO~VADA, J. 0970). Immunoglobulin G receptors on Burkitt cells. BacteriologicalProceedings, p. I8I. MACPHERSON, I. A. & STOKER M. O. P. 0962). Polyoma transformation of hamster cell clones - an investigation of genetic factors affecting cell competance. Virology x6, I47-I51. REED, L. J. & MUENCH, H. (1938). A simple method of estimating 5o per cent endpoints. American Journal of Hygiene 27, ROIZMAN, B. (1972). The biochemical features of herpesvirus-infected cells, particularly as they relate to their potential oncogenicity. In Oneogenesis and Herpesviruses, pp. I-I7. Edited by P. M. Biggs, G. de The & L. N. Payne. Lyon, France: International Agency for Research on Cancer. SnIMIZU, Y Modification of host cell membrane after herpes simplex virus infection. Arehivfiir die gesamte Virusforschung 33, WATKINS, S. r. 0964). Adsorption of sensitized sheep erythrocytes to HeLa cells infected with herpes simplex virus. Nature, London 202, 1364-I365. WATKINS, J. r. (I965). The relationship of the herpes simplex haemadsorption phenomenon to the virus growth cycle. Virology 26, YASUDA, J. & MILGROM, F. 0968). Haemadsorption by herpes simplex infected cell cultures. International Archives of Allergy and Applied Immunology 33, o. (Received 30 January I974)