August t992
Volume 309, number 1, 82-84 FEB$11494 O 1992 F=derntion of European Biochemical Societies 0014S793/9~JSS,00
Inhibitory effect of human chorionic gonadotropin (hCG) on HIV-1 transmission from lymphocytes to trophoblasts Aldar S. BourinbaiaP and Raisa Nagomy" "118 East l Oth Street, New 7ork, N Y 10003. USd and bThe Population Council Ce, ter fiJe Biomedical Research. 1230 York Avenue. Nee' Yor~'. N}' 10021, USA
Re¢iv©d 7 July 1992 It has ~ dcmonstrnted that human chorioni¢ ilonadotropin (hCG) inhibits HIV production in vitro, subletting that this ~lubl¢ placental illycoprotein can control viral replication and s ~ d in rive. hCG - the major product of fetal tr~phoblasts - was teated on an in vitro model ¢onsi=Itinllorchoriocarclnoma..derivcd ENAM! trophoblasts exposedto HlV.in reeled MOL'F-4 lymphocytes, The results show a U.sbaped antiviral dose- efr¢¢t and ,,ullllestthat hOG may contribute to prol~tion airiest intrauterine tran,=miudon of HIV-I, hOG: HIV: Lymphocyte: Trophoblast; Placenta
1. INTRODUCTION
2, MATERIALS AND METHODS 2,1. Anti¢lral as~l),
Recent observations seem to indicate that CD4-negative placental trophoblasts - the only fetal cells ;n direct contact with maternal blood - can be susceptible to HIV infection depending on the mode of virus delivery [1-3]. Trophoblast choriocarcinomic tines were resistant to infection by c¢ll.fr~ virus [3]. Furthermore, there were no signs of infection when trophoblasts were exposed to HIV-¢arrying ells with impaired adhesion capacity [2]. However, the exposure to lymphocytes that adhere to substrate e~lls invariably resulted in infection of placental cells[1,2]. Neverthele.~, the low transmission rate of HIV across the placenta cannot be accounted for solely by such an explanation [4]. The possibility that pregnancy proteins and hormones may regulate retroviral infection has been recently ~plored [2] and it was found that human chorionic gonadotropin (hOG) can inhibit viral production in HIV-inf¢cted ceils [5].h C O is the most abundant glycoprot¢in hormone produ~d by fetal trophoblasts and is known as a potent suppressor of ~ll-mediated allogeneic reactions involving the interaction of maternal lymphocytes with fetal implant [6,7]. In this study we evaluated the effect of hCG on viral transmission using an in vitro model of pia~ntal infection occurring upon lymphocyte-to-trophoblast contact.
Correspondence address: A.S, Bourinbaiar, 118 E, 10th Street, New York, NY 10003, USA, Fax: (1) (212) 50S-8081,
82
The seedin8 concentration of ¢horiocarcinoma.dcrived ENAM[ trophoblasts (kindly from Dr, N, Matiuzaki, Osaka University Medical School. Japan) was lxl0 ~cells i:atr wall (l ca") and plates wtrc cultured in a humidified incubator with 5% CO~ at ~,?°C for 24 h prior to co-incubation wit h lymphoc>,tes.Serial ten-fold dilutions of purified preparation o1' hCG (ape=trio activity 4.000 IUImB; Stilton, St. Louis, MO) were added to washed inflict ENAMI trophoblast ¢¢11st'or 0,5 h at 4"C, followed by the addition ot" 10S/ml of MOLT.4/YH$ cells (HIV-i stain YH5 inl'¢¢tcd MOLT-4 Tedl line we,=provided by Dr, .tun Minowada, Fujhtaki Coil Center, Okayama, Japan); the co..culture was continued in the presence of hCG for I h at 37'C. By the end of incubation suspensio.-ilrown donor lymphocytic cells were thor. Oalihly waslted from the plastic-adherent trophoblasts by repeated rinsing with ice.cold Ca:'. Mg~" t'r¢¢ PUS. Following the removal ot" non.absorbed HIV and adherent HIV.infected lymphocytes, the virus. exposed tropltoblasts were trypsinlzed and replated into new culture dishes, Residual lymphocytcs whiclt could ~rsist after wathing were eliminated by further culture in a medium (Serumlcss Medium, Ncu. man & Tytell, Gib¢o) that has no effect on trophoblast viability. One month post.infection mock.inl'cctcd and HIV inre¢ted trophoblasts remained morphologically identical and no cytopathic effect could bc observed in trophoblast ¢~lls exposed to HIV.I. The infection of trophoblnsts was p~rdstent, with very low spontaneous viral produc. tion, as them was no detectable p24 release or RT activity. 2.2. Reco~'er),o/virus from late, tI.Pinfected rrophob/asts by ¢acuhure
with cord.bloud.derired MT.4 lymphoeytes The cnineubation of Iow.producin8 trophoblaits with non.infected indicator MT-4 lymphocytcs (fron~ Dr, J, Minowada) always rcsult~ in a heavy viral production accompani~ by syncytium formation and lymphocyte death by day 7. Therefore, the recovery and amplification of the virus from low-producing placent='l cells was achieved by coin. cubation with HlV-susccptibl= MT-4 lymphoc),tes carried out two w=ks alter the initial inoculation, MT-4 ells (10' ~llgmL) were washed twice e.nd added to virus.~rryinB ENAMI cells l'or 1 h. In. ."eared-by.contact MT-4 cells were gently removed, washed and cultut'¢d at 37"C in S% CO~ for 7 days. Samples of the culture m~ium were tested on day 4 with p24 AB ELISA (Cellular Products, Buffalo, NY).
Pubt;dtcd by Eh'evicr 8¢1¢n¢¢ Publishers B. V.
Volume 309, number l
FEBS L E T T E R S 1000
3, RESULTS The antiviral effect of hCG a$ainst c¢ll,.contact-mediated transmission was evaluated by measuring the levels of p24 from MT.4 lymphocytes after coincubation with ENAMI trophoblasts exposed beforehand to MOLT-4/ Y H S in the prccnc¢ of the hCG, This indirect procc. dure is time consuming. However, considering that HIV-I replication in trophoblasts results in very low spontaneous viral production that is at the sensitivity limit of ELISA. the simplest mean of characterizing the degree of infection is the amplification of the virus by co-culture with MT-4, Productive HIV infection in MT4 lymphocytcs ischaracterized by a fommtion ofsyncytial cells and extensive cell death. Inst~d of relying on visual evaluation based on the manual counting of multinucleated ceils, we quantitatcd th, dos~-response by standardized ELISA for p24 viral antigen, The resuits shown in Fig, l seem to indicate that hCO can prevent cell-mediated HIV- 1 infection in a dose-dependcat fashion, A O-sha~d dose-response curve was observed, implying that the intermediate doses of hCG appear to prevent celL-mediated HIV.I infection more efficiently than the lower or higher do~es at the edges of the curve, Cell-to-cell HIV-I transmission was blocked most efficiently within a dose range of 0.1-100 IU/ml. Lower and higher concentrations of hCG. to O.01 and 1,000 IUlml, resulted in approximately 10% and 50% inhibition, respectively.Despite such n peculiar dose-response all tested concentrations of hCG were at least partially protective against HIV infectivity. 4. D I S C U S S I O N In previous studies we have obtained two sets of data. First, trophoblasts were permissive to HIV delivered by physical contact with infected lymphocyte-, but not when exposed to cell-free HIV or infected cells with impaired adhesion capacity [2], Second, low doses of hCO can inhibit reverse transcriptase activity and release of p24 HIV gag antigen from infected lymphocytcs; the dose-rc,,ponsc was U-shaped, hCG s~mcd to have a specific antiviral effect since there was no disccrnibl¢ cytotoxic effect [5], Based on these observations we decided to evaluate the effect of hCG in cell-cell infection using an in vitro system representing the model of placental infection. The results presented here indicate that hCG inhibits the onset of viral spr~td via cell- cell contact and the dose-response curve appeared to be of the same U-shape observed previously [5]. h C G is the earliestand most abundant polypcptida hormone produced during gestation and is mostly known for its involvement in hormon~l interplay at the establishment and maintenance of pregnancy. Th© function of h C G as an immunorcgulator is not well understood itispossible el'a:locallys~re~d hCG interferes with the cell-cellinteraction required for lymphocyte-
August 1992
IOO, |00,
l
700, 600, 500, 400,
$00, 200, tOO, 0
•
t~..z t~-t ta' tat
t~a t~tl
(IUtml) FiM. l, Effect of hCG (Sigma) on cell.mediated HLV trafllmimion occurring us a result o1"[ymphosyt~'.to-trophobiast contact in the pros¢11¢eor absent,; ,.-fconcentrutions of hCG ranllin$ from 10" lUIml to tO) [Ulml, The experiments were repeated three tim~ and means from duplicate wells of a reprc'sentativ¢ experiment an: shown. Not¢ that levels of p24 below l0 piVmi arc beyond Ih~ reliable sensitivity limit of the ELISA kit,
mediated r,.'jection of the scmi-allogeneic fetal implant [6,7], Although there is still a con:siderable controversy regarding this issue [8,9]. purified hCO preparations have been demonstrated to inhibit the reactivity of lymphocytes [11+14] in an alloreactive environment, We have suggested ~rlier that hCO may have an inhibitory effect on release of cytoplasmic content, i,e. viral pardties from iymphocytes triggered by contact with the substrat¢ cell [5]. The normal levels of hCG in plasma fluctuate from 160 IU/,.:I in the first trimester to under 80 IU/ml during the rest of the pregnancy [6]. It was recently reported that HIV.I infection of placental explains results in a ten-fold decline of hCG production and it has been proposed that this may decimate normal fetal develop., meat [15]. Effective antiviral concentrations of hCG in our study were within the range that may correspond to the decreased levels of hCG in plasma of HIV-Ipositive pregnant women, However, there are no clinical data that correlate hCG concentrations in viruscarrying pregnant women and the incidence of fetal infection. It is difficult to predict whether clinical applications of this glycoprotein hormone can be projected. The possibility that exogenous pregnancy products may prevent fetal infection andJor control viral infection in rive cannot be excluded. Except for interferon alpha [16]. hCO is one of the few proteins secreted by the human body that has been claimed to have antiviral activity [5], Although hCG was shown to inhibit reverse transcriptase activity and p24 production in infected cells, the mechanism of its action is not yet known, Further studies are need'~l to establish the mechanism of hCG and other pregnancy factors such as steroid hormones [2] and placental interferon against HIV replication and spread in rive,
83
Volume 309. number I
FIZBS LETTERS
.4ck.mt'ledtlement::: Carol Colem,~nprovided editorial .ulst.nce. ~ . Jun Minowada and Noboru Mntsuzaki provided cell lin~s and viral strains, Dr. Phillipos prnvidr.d lahora|oW Ipac¢. We thank Dr, W~i Lu for |ethnical auistance, This work was supported in p.r~ by The RockeNller Foundation. REFERENCES [l] Douslas, G,C,, Fry, G.N,, Thirkill, T,, Holmes, ~ . Hakim, H., Jstnninlp, M. and Kinll, B,F. 0991) AIDS Res, Human. Rctrovir. 7. ?35-740. [2] Bourir.baiar, A.S,, Nasorn~,, R, and Tan, X. (1992) FEltS Lctt. 302. 20~..20B, [3] Zachar, V,, Spire, B,, Hirfch, L, Chermann, J,C, ;,nd Eb~_.en. P. (199t) J, Virol, 65, 2102-2|07, [4] Profit. C.G. and Gershon A,A, (1991) Pediatr, infect. Dist. J. 10, 6~14-.695, [5] Dourinbaiar, A,S, and NaBorny, R. (1992) FEMS Microbiol, Lctt, (in prcu), [6] $tim,,on, W,H, (19:~3) in: [mmumolosy of Reproduction (Well. mann, T,G, and Gill ill, T,J., eds,) pp. 2~1-301, Oxl'ord University Prstis. New York,
84
August 1992
['/] Menu. E. and Cl'mouui. G, (1990) in: The lmmunoloBy of the Fetus (G. Chnoual, ed,) pp. 153-160, CRC Pm~, Boca Raton. [8) D'.ck, D,, Ginsburll, H, and Nao|, Y, (1977) Am, J. C)b~tet, Gyn¢col, 129, 14-20. [9] Kimniiler, J.W. ,rid Rocklin, R,E. (lg~10)J, Reprod. [mrnunoh 1. 29?-306, [i0] B,rlocci, A,, Pai~d©metriou. V,, ~hlick, E,, Nilula, B.C. and CldriBos, M.A, (1982) Ccll. [mmunol. 71,326-3•3. [11] Fuchs, T., Hammararom, L, Smith, C,LE. and Brundin, J. (1982) J. h¢prod. Xmmunol. 4, lt|5-1g0. [12] Rickc|ts, R.M. nnd Joncs, D,i~, (191f5) J, R¢i~ro¢l, [mmun=l, 7, ~5-232, [13] Scholar, R. and Bureau, J.P. (1979) J. Exp. Pnihol. 60, 483-48~t. (14] Harbour-McMstnamin. D.. Smith, E,M. and Bialock. J.E. (L9116) Pro¢, Nazi./v.ad. Sci. USA ~3, ~;-0838. [15] Amhire~,mmi-Allhili, N, and $1xs:ter S,A, (1991) J, Vixol. 65. 2231-2236, [16] Bourinbaiar, A,S,. N-Borny, R. and Tun, X, (199Z) in: Vir~! Qunntitation in HW [nl'ection (Andrieu0 J.M.e.d,) pp. 41-52, John Libbcy Eurotcxt, Paris.
Volume 309, number 1, 82-84 FEB$11494 O 1992 F=derntion of European Biochemical Societies 0014S793/9~JSS,00
Inhibitory effect of human chorionic gonadotropin (hCG) on HIV-1 transmission from lymphocytes to trophoblasts Aldar S. BourinbaiaP and Raisa Nagomy" "118 East l Oth Street, New 7ork, N Y 10003. USd and bThe Population Council Ce, ter fiJe Biomedical Research. 1230 York Avenue. Nee' Yor~'. N}' 10021, USA
Re¢iv©d 7 July 1992 It has ~ dcmonstrnted that human chorioni¢ ilonadotropin (hCG) inhibits HIV production in vitro, subletting that this ~lubl¢ placental illycoprotein can control viral replication and s ~ d in rive. hCG - the major product of fetal tr~phoblasts - was teated on an in vitro model ¢onsi=Itinllorchoriocarclnoma..derivcd ENAM! trophoblasts exposedto HlV.in reeled MOL'F-4 lymphocytes, The results show a U.sbaped antiviral dose- efr¢¢t and ,,ullllestthat hOG may contribute to prol~tion airiest intrauterine tran,=miudon of HIV-I, hOG: HIV: Lymphocyte: Trophoblast; Placenta
1. INTRODUCTION
2, MATERIALS AND METHODS 2,1. Anti¢lral as~l),
Recent observations seem to indicate that CD4-negative placental trophoblasts - the only fetal cells ;n direct contact with maternal blood - can be susceptible to HIV infection depending on the mode of virus delivery [1-3]. Trophoblast choriocarcinomic tines were resistant to infection by c¢ll.fr~ virus [3]. Furthermore, there were no signs of infection when trophoblasts were exposed to HIV-¢arrying ells with impaired adhesion capacity [2]. However, the exposure to lymphocytes that adhere to substrate e~lls invariably resulted in infection of placental cells[1,2]. Neverthele.~, the low transmission rate of HIV across the placenta cannot be accounted for solely by such an explanation [4]. The possibility that pregnancy proteins and hormones may regulate retroviral infection has been recently ~plored [2] and it was found that human chorionic gonadotropin (hOG) can inhibit viral production in HIV-inf¢cted ceils [5].h C O is the most abundant glycoprot¢in hormone produ~d by fetal trophoblasts and is known as a potent suppressor of ~ll-mediated allogeneic reactions involving the interaction of maternal lymphocytes with fetal implant [6,7]. In this study we evaluated the effect of hCG on viral transmission using an in vitro model of pia~ntal infection occurring upon lymphocyte-to-trophoblast contact.
Correspondence address: A.S, Bourinbaiar, 118 E, 10th Street, New York, NY 10003, USA, Fax: (1) (212) 50S-8081,
82
The seedin8 concentration of ¢horiocarcinoma.dcrived ENAM[ trophoblasts (kindly from Dr, N, Matiuzaki, Osaka University Medical School. Japan) was lxl0 ~cells i:atr wall (l ca") and plates wtrc cultured in a humidified incubator with 5% CO~ at ~,?°C for 24 h prior to co-incubation wit h lymphoc>,tes.Serial ten-fold dilutions of purified preparation o1' hCG (ape=trio activity 4.000 IUImB; Stilton, St. Louis, MO) were added to washed inflict ENAMI trophoblast ¢¢11st'or 0,5 h at 4"C, followed by the addition ot" 10S/ml of MOLT.4/YH$ cells (HIV-i stain YH5 inl'¢¢tcd MOLT-4 Tedl line we,=provided by Dr, .tun Minowada, Fujhtaki Coil Center, Okayama, Japan); the co..culture was continued in the presence of hCG for I h at 37'C. By the end of incubation suspensio.-ilrown donor lymphocytic cells were thor. Oalihly waslted from the plastic-adherent trophoblasts by repeated rinsing with ice.cold Ca:'. Mg~" t'r¢¢ PUS. Following the removal ot" non.absorbed HIV and adherent HIV.infected lymphocytes, the virus. exposed tropltoblasts were trypsinlzed and replated into new culture dishes, Residual lymphocytcs whiclt could ~rsist after wathing were eliminated by further culture in a medium (Serumlcss Medium, Ncu. man & Tytell, Gib¢o) that has no effect on trophoblast viability. One month post.infection mock.inl'cctcd and HIV inre¢ted trophoblasts remained morphologically identical and no cytopathic effect could bc observed in trophoblast ¢~lls exposed to HIV.I. The infection of trophoblnsts was p~rdstent, with very low spontaneous viral produc. tion, as them was no detectable p24 release or RT activity. 2.2. Reco~'er),o/virus from late, tI.Pinfected rrophob/asts by ¢acuhure
with cord.bloud.derired MT.4 lymphoeytes The cnineubation of Iow.producin8 trophoblaits with non.infected indicator MT-4 lymphocytcs (fron~ Dr, J, Minowada) always rcsult~ in a heavy viral production accompani~ by syncytium formation and lymphocyte death by day 7. Therefore, the recovery and amplification of the virus from low-producing placent='l cells was achieved by coin. cubation with HlV-susccptibl= MT-4 lymphoc),tes carried out two w=ks alter the initial inoculation, MT-4 ells (10' ~llgmL) were washed twice e.nd added to virus.~rryinB ENAMI cells l'or 1 h. In. ."eared-by.contact MT-4 cells were gently removed, washed and cultut'¢d at 37"C in S% CO~ for 7 days. Samples of the culture m~ium were tested on day 4 with p24 AB ELISA (Cellular Products, Buffalo, NY).
Pubt;dtcd by Eh'evicr 8¢1¢n¢¢ Publishers B. V.
Volume 309, number l
FEBS L E T T E R S 1000
3, RESULTS The antiviral effect of hCG a$ainst c¢ll,.contact-mediated transmission was evaluated by measuring the levels of p24 from MT.4 lymphocytes after coincubation with ENAMI trophoblasts exposed beforehand to MOLT-4/ Y H S in the prccnc¢ of the hCG, This indirect procc. dure is time consuming. However, considering that HIV-I replication in trophoblasts results in very low spontaneous viral production that is at the sensitivity limit of ELISA. the simplest mean of characterizing the degree of infection is the amplification of the virus by co-culture with MT-4, Productive HIV infection in MT4 lymphocytcs ischaracterized by a fommtion ofsyncytial cells and extensive cell death. Inst~d of relying on visual evaluation based on the manual counting of multinucleated ceils, we quantitatcd th, dos~-response by standardized ELISA for p24 viral antigen, The resuits shown in Fig, l seem to indicate that hCO can prevent cell-mediated HIV- 1 infection in a dose-dependcat fashion, A O-sha~d dose-response curve was observed, implying that the intermediate doses of hCG appear to prevent celL-mediated HIV.I infection more efficiently than the lower or higher do~es at the edges of the curve, Cell-to-cell HIV-I transmission was blocked most efficiently within a dose range of 0.1-100 IU/ml. Lower and higher concentrations of hCG. to O.01 and 1,000 IUlml, resulted in approximately 10% and 50% inhibition, respectively.Despite such n peculiar dose-response all tested concentrations of hCG were at least partially protective against HIV infectivity. 4. D I S C U S S I O N In previous studies we have obtained two sets of data. First, trophoblasts were permissive to HIV delivered by physical contact with infected lymphocyte-, but not when exposed to cell-free HIV or infected cells with impaired adhesion capacity [2], Second, low doses of hCO can inhibit reverse transcriptase activity and release of p24 HIV gag antigen from infected lymphocytcs; the dose-rc,,ponsc was U-shaped, hCG s~mcd to have a specific antiviral effect since there was no disccrnibl¢ cytotoxic effect [5], Based on these observations we decided to evaluate the effect of hCG in cell-cell infection using an in vitro system representing the model of placental infection. The results presented here indicate that hCG inhibits the onset of viral spr~td via cell- cell contact and the dose-response curve appeared to be of the same U-shape observed previously [5]. h C G is the earliestand most abundant polypcptida hormone produced during gestation and is mostly known for its involvement in hormon~l interplay at the establishment and maintenance of pregnancy. Th© function of h C G as an immunorcgulator is not well understood itispossible el'a:locallys~re~d hCG interferes with the cell-cellinteraction required for lymphocyte-
August 1992
IOO, |00,
l
700, 600, 500, 400,
$00, 200, tOO, 0
•
t~..z t~-t ta' tat
t~a t~tl
(IUtml) FiM. l, Effect of hCG (Sigma) on cell.mediated HLV trafllmimion occurring us a result o1"[ymphosyt~'.to-trophobiast contact in the pros¢11¢eor absent,; ,.-fconcentrutions of hCG ranllin$ from 10" lUIml to tO) [Ulml, The experiments were repeated three tim~ and means from duplicate wells of a reprc'sentativ¢ experiment an: shown. Not¢ that levels of p24 below l0 piVmi arc beyond Ih~ reliable sensitivity limit of the ELISA kit,
mediated r,.'jection of the scmi-allogeneic fetal implant [6,7], Although there is still a con:siderable controversy regarding this issue [8,9]. purified hCO preparations have been demonstrated to inhibit the reactivity of lymphocytes [11+14] in an alloreactive environment, We have suggested ~rlier that hCO may have an inhibitory effect on release of cytoplasmic content, i,e. viral pardties from iymphocytes triggered by contact with the substrat¢ cell [5]. The normal levels of hCG in plasma fluctuate from 160 IU/,.:I in the first trimester to under 80 IU/ml during the rest of the pregnancy [6]. It was recently reported that HIV.I infection of placental explains results in a ten-fold decline of hCG production and it has been proposed that this may decimate normal fetal develop., meat [15]. Effective antiviral concentrations of hCG in our study were within the range that may correspond to the decreased levels of hCG in plasma of HIV-Ipositive pregnant women, However, there are no clinical data that correlate hCG concentrations in viruscarrying pregnant women and the incidence of fetal infection. It is difficult to predict whether clinical applications of this glycoprotein hormone can be projected. The possibility that exogenous pregnancy products may prevent fetal infection andJor control viral infection in rive cannot be excluded. Except for interferon alpha [16]. hCO is one of the few proteins secreted by the human body that has been claimed to have antiviral activity [5], Although hCG was shown to inhibit reverse transcriptase activity and p24 production in infected cells, the mechanism of its action is not yet known, Further studies are need'~l to establish the mechanism of hCG and other pregnancy factors such as steroid hormones [2] and placental interferon against HIV replication and spread in rive,
83
Volume 309. number I
FIZBS LETTERS
.4ck.mt'ledtlement::: Carol Colem,~nprovided editorial .ulst.nce. ~ . Jun Minowada and Noboru Mntsuzaki provided cell lin~s and viral strains, Dr. Phillipos prnvidr.d lahora|oW Ipac¢. We thank Dr, W~i Lu for |ethnical auistance, This work was supported in p.r~ by The RockeNller Foundation. REFERENCES [l] Douslas, G,C,, Fry, G.N,, Thirkill, T,, Holmes, ~ . Hakim, H., Jstnninlp, M. and Kinll, B,F. 0991) AIDS Res, Human. Rctrovir. 7. ?35-740. [2] Bourir.baiar, A.S,, Nasorn~,, R, and Tan, X. (1992) FEltS Lctt. 302. 20~..20B, [3] Zachar, V,, Spire, B,, Hirfch, L, Chermann, J,C, ;,nd Eb~_.en. P. (199t) J, Virol, 65, 2102-2|07, [4] Profit. C.G. and Gershon A,A, (1991) Pediatr, infect. Dist. J. 10, 6~14-.695, [5] Dourinbaiar, A,S, and NaBorny, R. (1992) FEMS Microbiol, Lctt, (in prcu), [6] $tim,,on, W,H, (19:~3) in: [mmumolosy of Reproduction (Well. mann, T,G, and Gill ill, T,J., eds,) pp. 2~1-301, Oxl'ord University Prstis. New York,
84
August 1992
['/] Menu. E. and Cl'mouui. G, (1990) in: The lmmunoloBy of the Fetus (G. Chnoual, ed,) pp. 153-160, CRC Pm~, Boca Raton. [8) D'.ck, D,, Ginsburll, H, and Nao|, Y, (1977) Am, J. C)b~tet, Gyn¢col, 129, 14-20. [9] Kimniiler, J.W. ,rid Rocklin, R,E. (lg~10)J, Reprod. [mrnunoh 1. 29?-306, [i0] B,rlocci, A,, Pai~d©metriou. V,, ~hlick, E,, Nilula, B.C. and CldriBos, M.A, (1982) Ccll. [mmunol. 71,326-3•3. [11] Fuchs, T., Hammararom, L, Smith, C,LE. and Brundin, J. (1982) J. h¢prod. Xmmunol. 4, lt|5-1g0. [12] Rickc|ts, R.M. nnd Joncs, D,i~, (191f5) J, R¢i~ro¢l, [mmun=l, 7, ~5-232, [13] Scholar, R. and Bureau, J.P. (1979) J. Exp. Pnihol. 60, 483-48~t. (14] Harbour-McMstnamin. D.. Smith, E,M. and Bialock. J.E. (L9116) Pro¢, Nazi./v.ad. Sci. USA ~3, ~;-0838. [15] Amhire~,mmi-Allhili, N, and $1xs:ter S,A, (1991) J, Vixol. 65. 2231-2236, [16] Bourinbaiar, A,S,. N-Borny, R. and Tun, X, (199Z) in: Vir~! Qunntitation in HW [nl'ection (Andrieu0 J.M.e.d,) pp. 41-52, John Libbcy Eurotcxt, Paris.
