A blocking factor in amniotic fluid causing leukocyte migration enhancement N.
GLEICHER,
M.D.
C. J. COHEN, T.
M.D.
KERENYI,
M.D.
S. B. GUSBERG,
M.D.,
Neul
D.
York,
New
D.Sc.
York
Amniotic fluid was found to cause significant leukocyte migration enhancement during the second and third trimester of pregnancy and in the early postpartum period when compared to the migration area obtained with an ovarian tumor homogenate antigen (p < 0.01) choriocarcinoma spent medium (p < O.Ol), and placental pool homogenate (p < 0.01). Only borderline significance (p < 0.1) was obtained when migration enhancement with AF was compared between pregnant and nonpregnant female control patients, indicating minimal unspecific activity of AF. Migration enhancement with autologous amniotic fluid was slightly larger than with homologous amniotic fluid, but the difference did not reach significance (p < 0.4). None of the control antigens caused migration enhancement; placental pool homogenate in concentrations above 4 mg. per cent caused migration inhibition but did not in lower concentrations. The enhancing effect of AF could be abolished by dilution but not by addition of excessive antibody to estrogen or HCG. It is suggested that a blocking factor is present in AF preventing recognition of fetoplacental antigen by the maternal immune system. Thus in vitro leukocyte migration enhancement may correlate to in vivo graft enhancement. (AM. J. OBSTET. GYNECOL. 133:388, 1979.)
POSSESSING paternally derived antigens of the ABO, Rh, histocompatibility (HLA), and organ-specific types, the products of conception represent an allograft which survives within a potentially hostile immune system for the period of gestation. The antigenicity of placental and fetal tissue has been established by various methods.‘+7 During early gestation trophoblast is functionally hypoantigenic. Antigenicity increases significantly during the second From the Department oj Obstetrzcs and Gynecology, The Mount Sinai School of Medicine qf the City University ojf‘ New York. RPceivrd
for publiratzon
Revised April
27, 197X
.4ccepted Mar
24, 1978.
October
31, 1977.
Reprint requests: Dr. N. G&her, Department of Obstetrics and G.wecology, Mount Sinai School of Medicine, 100th St. and Fiph Ave., New York, New York 10029. Abbreviatrons u.\ed: AF, amnioticjuid; a-AF, autologous amnioticjluid; h-,4F, hnmologous amniotic&id; OCA, ovarian carcinoma antigen; CCSM. choriocarcinoma .spent medium: PPA, placental pool antigen; HCG, human chorionic gonadotropin; Mz4, area of migration; MI, migration index.
and third trimester”’ and continues to exert its effect into the postpartum period.‘, ’ AF, being mainly fetal in its origin, should therefore represent the fetoplacental unit antigenically. The leukocyte migration assay, as described by Soborg and Bendixen was chosen for the detection of cellular hypersensitivity against “foreign” fetoplacental antigen. Presensitized maternal leukocytes are thus expected to exhibit leukocyte migration inhibition not later than in the second trimester of gestation when challenged with AF (representing free fetoplacental antigen). The nonoccurrence of such migration inhibition in the above-described system could therefore represent one of the following conditions: (1) Absence of antigen, or (2) the nonrecognition of fetoplacental antigen by maternal effector cells, either due to lack of previous sensitization or due to blockage of receptor sites by a blocking factor as described in maternal serum.” It is the purpose of this paper to investigate the above-mentioned mechanism. Materials and methods This study included 20 pregnant trimesters of gestation, 5 women 0002-9378/79/040386+05$00.50/0
patients in all three on postpartum day
0 1979 The C. V. Mosby Co.
Volume Number
IX4 4
Leukocyte
four (PPD-4) and 11 nonpregnant female control subjects. Relevant information on all patients is presented in Table I. Serial measurements of the fetal biparietal diameter by ultrasonogram were utilized for dating of gestation. Preparation
Table
I.
migration
Clinical
enhancement
Control Parity
and PPD-4
Range Mean Range Mean Range Mean Range Mean
subjects
Pregnant
and PPD-4
of antigens.
.JmGotic fi&. AF was obtained from 16 second and third trimester patients by either sterile transabdominal amniocentesis or transvaginal amniotomy. Clinical indications for these procedures are reported in Table I. The AF was immediately iced and centrifuged at 1,000 g for 30 minutes at 5” C. The clear supernate was used as antigen after dilution 1: 1 with complete medium (Roswell Park Memorial Institute 1640* supplemented with 20 per cent heat-inactivated fetal calf serum.* 100 units per milliliter penicillin, and 1 mg. per milliliter of streptomycin). AF tested against the same patient’s lymphocytes from whom it was obtained was termed autologous (a-AF). Amniotic fluid 1~001 consisted of fluid from five to ten patients in the second and third trimester of gestation thoroughly mixed in equal amounts and diluted 1: 1 with comJ)lete medium. a-AF was never included in the pool against which a patient’s leukocytes were tested. Thus amniotic fluid pool represented homologous amniotic fluid (h-AF). In a few instances dilutional studies were performed at dilutions of 1: 10, 1:50, and 1: 100 with complete medium. a-AF only was used for these experiments. To exclude the possibility of a hormone effect as the cause of migration enhancement, controls were set up in single cases utilizing (1) sheep antiserum to estrogen,+ provided by Dr. P. G. Satyaswaroop from the laboratory of Dr. E. Gurpide, and (2) anti-human chorionic gonadotropin., f Both antibodies were added in excess to the necessary amount for neutralization of estrogen and HCG in AF. 0~1arircr~ carcinomu antigen. OCA-homogenate had been prepared previously and stored at -20” C. Its antigenicity at a protein concentration of 0.3 mg. per milliliter had been established.g Dilution was obtained with complete medium. Placental pool rLntigr,l. Three term placentas were obtained immediately after delivery, iced, and homogenized.g Equal amounts of homogenate from all three placentas were mixed thoroughly. Protein concentra*Grand Island Biological Co., Grand Island, N. Y. +Lot No. 22: Titers: 1:3,500 against E, with 50 per cent binding; 1: 7,000 against E2 with 50 per cent binding; 1: 10,000 against E3 with 50 per cent binding. *Lot No. 224: Index Biomedical Division Wilson Pharmaceutical and Chemical Corporation. Glenwood. Ill.
fluid
387
summary
Pregnant
Age
by amniotic
Control
subjects
14-41 27.8 22-41 29.5 O-6 1.7 o-4 0.6
Indications for amniocentesis Maternal age Midtrimester abortion Repeat cesarean section Postmaturity Previous anencephalus Severe hypertensive disease Amniotomy in labor Total *All of our control labor floor employing
patients were birth control.
7 3 2
16 hospital
personnel
on the
tion of this pool was determined by the method of Lowry and associateslO and adjusted to a concentration of 0.3 mg. per milliliter by dilution with complete medium. In special instances higher concentrations up to 4.0 mg. per milliliter were used. Choriocarcinoma spent medium. Spent medium from a choriocarcinoma cell line established by Patti110 and Gey” and grown by Dr. P. G. Satyaswaroop in our department was diluted with complete medium 1: 1 and frozen at -20” C. until use. This cell line is known to secrete 1 x lo-” 1.U. of HCG per cell per 24 hours.” Leukocyte migration. Leukocyte migration tests were performed according to the method of Sbborg and Bendixen with minor modifications.g Each experiment consists of four duplicates, which are averaged to obtain the mean area of migration (MA). The migration index (MI) was calculated as follows: MI =
area area
of migration of migration
with antigen without
antigen
x 100.
Thus a MI above 100 indicates migration enhancement and a MI below 100, migration inhibition. Percentage migration enhancement or inhibition equals MI minus 100 or 100 minus MI, respectively, and is displayed in Table II as (+) for enhancement and (-) for inhibition. Statistical analysis. Significance between two sample means was established by Student t test using Fisher’s distribution.
Results Fig. 1 represents the MI of all 20 patients and 11 control subjects when challenged with different antigens. Table II represents the mean percentage migration
enhancement
for
each
trimester
of pregnancy
and
388
Gleicher
February Am. J. Obstet.
et al.
Table II. Mean percentage migration enhancement on PPD4, and in nonpregnant control patients
t standard
deviation
with different
antigens
15. 1979 Gynecol.
in pregnancy,
Trim&u Control
Total Jar First
Third
Second
PPD-4
pregnancy
subjects
a-AF LE
+53.a
11 k 32.Q
+43.4
5 2 40.5
+50.5
16 k 34.4
+40.6
6 + 37.1
+34.2
4 4 12.9
+31.1
14 2 27.1
+4Q.l
17 2 33.Q
+39.3
Y 2 30.1
+41.5
30 + 32.2
8 ‘+- 11.6
+14.1
17 k 9.3
+8.8
3 + 13.0 2 3.9
h-AF 4
ZME AFt
+13.7ze
11.9
ZME OCA iME CCSM
+lO.l
4 k 7.2
+15.1
EME PPA
+3.3
2 T 4.7
9 + 15.9 + 12.0
4 + 16.8 % 7.6
+20.6
10 ‘- 10.4
5 + 15.7 -+ 7.5
+8.3
4 ” 8.7
+12.2
15 k 11.2
+15.0
.5 t- 21.4
+18.3
IY f
+26.7’?
18.1
11 + 16.4 2 14.6
5 + 5.3
13.1
*N = number of cases. tTotal AF (h-AF + a-AF). PPD-4 separately. In addition results for a-AF and h-AF are listed separately facilitating direct comparison of their activity wherever both were obtainable. The MI with AF during the second and third trimester of pregnancy was significantly larger when compared to OCA (p < 0.01). CCSM (p < O.Ol), and PPA (p < 0.0 1). The enhancing activity of a-AF exceeds that of h-AF only slightly (p < 0.04), which allows direct comparison of results obtained with either. This is of importance because obviously h-AF only was available during the first trimester of gestation and in the postpartum period. .1s can be seen from Table II a trend toward enhancement was observed when control patients were challenged with h-AF resulting in decreased significance (p < 0.1) when compared to the amount of enhancement with h-AF in pregnant women of second and third trimester. OCA, CCSM, and PPH showed equal migration areas in pregnant women of all trimesters and nonpregnant control subjects (p < 0.5), not causing enhancement. A trend toward enhancement with AF on PPD-4 was present but did not reach statistical significance, probably due to the small number of cases. None of the antigens caused significant migration enhancement during the first trimester of pregnancy but enhancement with h-AF reached borderline significance when compared in the third trimester with the migration area obtained in the first trimester (p < 0.1). None of the antigens showed migration inhibition in
routine concentrations. When the concentration of PPA was increased to 4.0 mg. per milliliter migration inhibition was encountered. Fig. 2 shows the decrease of MA with increased concentrations of PPA in one patient. The enhancing activity of AF could be abolished by increasing the grade of dilution. Significant enhancement with AF was usually abolished at a dilution of 1 : 50 (Fig. 3). A hormone effect (estrogen, HCG) as the cause of migration enhancement could be excluded by proving that abolition of hormone activity by specific antibodies did not inhibit migration enhancement (Fig. 4).
Comment The assumption that AF represents the fetoplacental unit antigenically seems reasonable in view of out knowledge (although incomplete) about its formation and rapid turnover. Because fetal tissue achieves increased antigenicity during the second and third trimester of gestation it can be concluded that AF should represent this antigenicity at approximately the same time.+’ Using in this study the leukocyte migration assay, as developed by Soborg and Bendixen we anticipated migration inhibition to occur as evidence for the presence of hypersensitivity toward the AF-antigen pool. A possible lack of inhibitory activity by AF was theoretically thought to represent either the absence of fetoplacental antigen in AF or the blockage of antigen by a blocking factor. Such a blockage could render the antigen nonantigenic or, more probable, as in tumor
Volume Number
133 4
Leukocyte migration enhancement
oa-AF
240-I 220 200 -
2
180 160 140 120
-
100
__-
g -
80-
E
60-
Oh-AF
’
6 OCA
’ ’0 ’ . .
’
~02%
’
x PPA
’ ’
1
’
0 l
---t-E.
_-_____
. .
z
i -)r.-
--.1--
0
l
-
.
_
CONC. OF PPA DILUTED WITH COMPLETE MEDIUM
5 ” r
40 200
ttIIIIII;I
----
60 4020 o! ’ ! ! ’ I I ’ ’ ’ ’ 0 4 8 12 16 20 24 28 32 36 4044pP,0,f;MCoNTRoLS WEEKS
389
0
0 0.
-,..-
by amniotic fluid
Fig. 2. Mean migration area (MA) and migration index (MI) in one of the patients (H.D.) when tested with PPA in increasing concentrations and with full medium (ME) as control. Note the decrease in MA and MI with increasing PPA concen-
trations until migration inhibition tion of 4.0 mg. per cent.
is obtained at a concentra-
DAY4
OF PREGNANCY
Fig. 1. Migration index (MI) in pregnant and postpartum patients and nonpregnant control subjects when tested with AF, OCA, CCSM, and PPH.
systems prevent recognition of antigen by maternal effector cells. The additional explanation of nonpresensitized effector cells was excluded by the demonstration of migration inhibition with PPA in high concentrations, as had been used by other authors.2-“, g Thus antigenicity of fetal tissue as well as immunologic competence of maternal effector cells were confirmed by the present study. Surprisingly a new immunologic phenomenon-migration enhancement-was encountered at exactly that period of gestation when migration inhibition was expected to occur with AF as antigen. None of the other antigens exhibited migration enhancement of significant proportions. As a consequence we conclude that AF has to be the causative factor for migration enhancement and exclude the possibility that maternal lymphocytes acquire special properties per se during pregnancy. While the mechanism of migration inhibition is caused by the secretion of migration inhibition factor (MIF) by presensitized T-lymphocytes on contact with the “recognized” antigen, nothing is known about the causative factors involved in migration enhancement. It could be speculated that inhibition of MIF secretion by any mechanism might result in migration enhancement. However, inhibition of migration inhibition factor only
by but
protein does
synthesis
inhibitors
not
in migration
result
prevents
inhibition
enhancement.‘*
”
I:I
I:10 I:50
DILUTION COMPLETE
WITH MEDIUM
Fig. 3. Mean migration area (MA) and migration index (MI) in one of the patients (H.D.) when tested with increasing dilutions of AF and with complete medium (ME) as control. Note the decreasing MA and MI with increasing dilutions of AF
until inhibition medium.
is reached at a dilution of 1: 50 with complete
400 $300 a p 200 2 ” 100 E 0
ESTROGEN ANTIBODY
IiCGANTIBODY
Fig. 4. Mean migration area (MA) in one of the patients (B.K.) when antibodies to estrogen and HCG were added to AF and control chambers with complete medium (ME). Note the significant migration enhancement with AF with or without antibody in comparison to the MA obtained with ME only, with or without antibody.
390
Gleicher et al.
Thus, for migration enhancement there has to be either an additional mechanism to what is obtained with protein synthesis inhibitors or total blockage of MIF secretion, not present in the usual resting state. Although this report represents the first evidence for the presence of a blocking factor in 4F, similar factors have been described previously in serum. Thus Hellstriim and co-workers” described a serum factor abrogating cellular immunity to antigenically foreign mouse embryogenic cells, thereby causing enhanccment of’ growth. In human beings it was shown that maternal serum prevents the cytotoxic effects of maternal lymphocytes 011 trophoblasr.” Abrogation of cellular immunity to placental antigen by maternal serum has in addition been reported by many authors, using different in vitro techniques, including the migration assay. Although migration enhancement has in those cases not been observed, it could be speculated that the difference between no inhibition and enhancement is a quantitative difference in the amount of blocking factor present in serum and AF. This would indicate a very high concentration of blocking factor in AF, making it the perfect vehicle for future purification. Clinical application of such a purified blocking factor does not seem to lie in the too distant future, taking into account the recent report by Rocklin and coworkers,‘” who reported the absence of such a blocking factor in women with repeated idiopathic abortions. IgG was implicated as the decisive factor.‘” The recent description of antigen-antibody complexes in sera of
normal
pregnant
patients
allows
suggestions
for
a
mechanism of nonrejection of pregnancy, very similar to the one described in the nonrejection of malignant tumors.‘” In view of these findings studies are underway in OUI laboratory to correlate migration enhancement with IgG concentration in pregnancy sera and AF and the possible presence of immune complexes. Preliminary data suggest an inverse relationship between migration enhancement and IgG concentration in AF. We are presently unable to explain the large fiuctuations
in
seems
challenging
the
enhancement
pattern
to assume
that
caused they
bp
might
AF.
It
represent
different concentrations of blocking factor in AF. If this assumption proves to be correct a wide variety of clinical applications seems feasible. Thus if hypertensive disease of pregnancy in fact represents an immunologically induced disease process, caused bp partial rejection of the fetal alograft, as believed by many. a lower degree of migration enhancement (because of less blocking factor) could be expected in patients with hypertensive the
early
disease diagnosis
of’ pregnancy. of
this
condition
This
could through
lead
to
simple
amniocentesis in the midtrimester. Changes in lymphocyte function in mild pre-eclampsia have recentl) been reported.” In this connection it seems of interest that the only patient in our study with severe hypertensive disease of pregnancy did not show significant migration enhancement.
REFERENCES
1. Hulka. 2. 3.
4.
5.
6.
7.
8.
J. F., Hsu, K. C., and Beiser, S. M.: Antibodies to trophoblast during the post-partum period. Nature 91: 519, 1961. Koren, Z., Behrman, S. J., and Paine, P. J.: Antigenicity of trophoblastic cells indicated by fluorescein techniques, AM. J. OBSTET. GYNECOL. 104: 50, 1969, Koren, Z., Abrams, G., and Behrman, S. J.: Antigenicity of mouse placenta1 tissue. AM. J. OBSTET. GYNECOL. 102: 340, 1968. Yoitannukorn, V., and Matangkasombut. P.: Human maternal cell-mediated immune reaction to placental antigens, Clin. Exp. Immunol. 11: 549, 1972. Taylor, P. V.. and Hancock, K. W.: Antigenicity of trophoblast and possible antigen-masking effects during pregnancy, Immunology 28: 973. 1975. Yoitananukorn, V., Matangkosombut, P., and Osathanondh, V.: Onset of human maternal cell-mediated immune reaction to placental antigens during the first pregnancy, Clin. Ex’p. Immunol. r5: 593, 19%. Tavlor. P. V.. Gowland. G.. Hancock. K. W.. et al.: Effect of length of gestation on maternal cellular immunity to human trophoblast antigens, AM. J. OBSTET. GYNECOL. 125: 528, 1976. SBborg, M., and Bendixen, G.: Human lymphocyte migration as a paramenter of hypersensitivity, Acta Med. Sand. 181: 247, 1967.
9. Faiferman.
10. Il.
12.
13.
14.
15.
16.
I., Gleicher. N., Cohen, C. J., et al.: Leukocyte migration in ovarian carcinoma: Comparison of inhibitory activity of tumor extracts, J. Natl. Cancer Inst. 59: 1593, 1977. Lowry, 0. H.. Rosenbrough, N. J., Lewis Farr, A., et al.: Protein measurement with the Folin phenol reagent, J. Biol. Chem. 193: 265, 1951. Patillo, R. A.. and Gey, G. 0.: The establishment of a cell line of human hormone-synthesizing trophoblastic cells in vitro, Cancer Res. 28: 1231, 1968. Henney, C. S., Gaffney. J., and Bloom, B. R.: On the relation of products of activated lymphocytes to cellmediated cytolysis, J. Exp. Med. 140: 837. 1974. Hellstriim, E., and Brawn, J.: Abrogation of cellular immunity to antigenically foreign mouse embryogenic cells by a serum factor, Nature 224: 914, 1969. Rocklin. R. E., Kitzmiller, J. L., Carpenter, C. B., et al: Absence of immunologic blocking factor in women who chronically abort, N. Engl. J. Med. 295: 1209, 1976. Petrucco, 0. M., Seamark, R. F.. Holmes, K., et al.: Changes in lymphocyte function during pregnancy, Br. J. Obstet. Gynecol. 83: 245, 1976. Masson. P. L., Delire, M., and Cambiasso, C. L.: Circulating immune complexes in normal human pregnancy, Nature 266: 542, 1977.
GLEICHER,
M.D.
C. J. COHEN, T.
M.D.
KERENYI,
M.D.
S. B. GUSBERG,
M.D.,
Neul
D.
York,
New
D.Sc.
York
Amniotic fluid was found to cause significant leukocyte migration enhancement during the second and third trimester of pregnancy and in the early postpartum period when compared to the migration area obtained with an ovarian tumor homogenate antigen (p < 0.01) choriocarcinoma spent medium (p < O.Ol), and placental pool homogenate (p < 0.01). Only borderline significance (p < 0.1) was obtained when migration enhancement with AF was compared between pregnant and nonpregnant female control patients, indicating minimal unspecific activity of AF. Migration enhancement with autologous amniotic fluid was slightly larger than with homologous amniotic fluid, but the difference did not reach significance (p < 0.4). None of the control antigens caused migration enhancement; placental pool homogenate in concentrations above 4 mg. per cent caused migration inhibition but did not in lower concentrations. The enhancing effect of AF could be abolished by dilution but not by addition of excessive antibody to estrogen or HCG. It is suggested that a blocking factor is present in AF preventing recognition of fetoplacental antigen by the maternal immune system. Thus in vitro leukocyte migration enhancement may correlate to in vivo graft enhancement. (AM. J. OBSTET. GYNECOL. 133:388, 1979.)
POSSESSING paternally derived antigens of the ABO, Rh, histocompatibility (HLA), and organ-specific types, the products of conception represent an allograft which survives within a potentially hostile immune system for the period of gestation. The antigenicity of placental and fetal tissue has been established by various methods.‘+7 During early gestation trophoblast is functionally hypoantigenic. Antigenicity increases significantly during the second From the Department oj Obstetrzcs and Gynecology, The Mount Sinai School of Medicine qf the City University ojf‘ New York. RPceivrd
for publiratzon
Revised April
27, 197X
.4ccepted Mar
24, 1978.
October
31, 1977.
Reprint requests: Dr. N. G&her, Department of Obstetrics and G.wecology, Mount Sinai School of Medicine, 100th St. and Fiph Ave., New York, New York 10029. Abbreviatrons u.\ed: AF, amnioticjuid; a-AF, autologous amnioticjluid; h-,4F, hnmologous amniotic&id; OCA, ovarian carcinoma antigen; CCSM. choriocarcinoma .spent medium: PPA, placental pool antigen; HCG, human chorionic gonadotropin; Mz4, area of migration; MI, migration index.
and third trimester”’ and continues to exert its effect into the postpartum period.‘, ’ AF, being mainly fetal in its origin, should therefore represent the fetoplacental unit antigenically. The leukocyte migration assay, as described by Soborg and Bendixen was chosen for the detection of cellular hypersensitivity against “foreign” fetoplacental antigen. Presensitized maternal leukocytes are thus expected to exhibit leukocyte migration inhibition not later than in the second trimester of gestation when challenged with AF (representing free fetoplacental antigen). The nonoccurrence of such migration inhibition in the above-described system could therefore represent one of the following conditions: (1) Absence of antigen, or (2) the nonrecognition of fetoplacental antigen by maternal effector cells, either due to lack of previous sensitization or due to blockage of receptor sites by a blocking factor as described in maternal serum.” It is the purpose of this paper to investigate the above-mentioned mechanism. Materials and methods This study included 20 pregnant trimesters of gestation, 5 women 0002-9378/79/040386+05$00.50/0
patients in all three on postpartum day
0 1979 The C. V. Mosby Co.
Volume Number
IX4 4
Leukocyte
four (PPD-4) and 11 nonpregnant female control subjects. Relevant information on all patients is presented in Table I. Serial measurements of the fetal biparietal diameter by ultrasonogram were utilized for dating of gestation. Preparation
Table
I.
migration
Clinical
enhancement
Control Parity
and PPD-4
Range Mean Range Mean Range Mean Range Mean
subjects
Pregnant
and PPD-4
of antigens.
.JmGotic fi&. AF was obtained from 16 second and third trimester patients by either sterile transabdominal amniocentesis or transvaginal amniotomy. Clinical indications for these procedures are reported in Table I. The AF was immediately iced and centrifuged at 1,000 g for 30 minutes at 5” C. The clear supernate was used as antigen after dilution 1: 1 with complete medium (Roswell Park Memorial Institute 1640* supplemented with 20 per cent heat-inactivated fetal calf serum.* 100 units per milliliter penicillin, and 1 mg. per milliliter of streptomycin). AF tested against the same patient’s lymphocytes from whom it was obtained was termed autologous (a-AF). Amniotic fluid 1~001 consisted of fluid from five to ten patients in the second and third trimester of gestation thoroughly mixed in equal amounts and diluted 1: 1 with comJ)lete medium. a-AF was never included in the pool against which a patient’s leukocytes were tested. Thus amniotic fluid pool represented homologous amniotic fluid (h-AF). In a few instances dilutional studies were performed at dilutions of 1: 10, 1:50, and 1: 100 with complete medium. a-AF only was used for these experiments. To exclude the possibility of a hormone effect as the cause of migration enhancement, controls were set up in single cases utilizing (1) sheep antiserum to estrogen,+ provided by Dr. P. G. Satyaswaroop from the laboratory of Dr. E. Gurpide, and (2) anti-human chorionic gonadotropin., f Both antibodies were added in excess to the necessary amount for neutralization of estrogen and HCG in AF. 0~1arircr~ carcinomu antigen. OCA-homogenate had been prepared previously and stored at -20” C. Its antigenicity at a protein concentration of 0.3 mg. per milliliter had been established.g Dilution was obtained with complete medium. Placental pool rLntigr,l. Three term placentas were obtained immediately after delivery, iced, and homogenized.g Equal amounts of homogenate from all three placentas were mixed thoroughly. Protein concentra*Grand Island Biological Co., Grand Island, N. Y. +Lot No. 22: Titers: 1:3,500 against E, with 50 per cent binding; 1: 7,000 against E2 with 50 per cent binding; 1: 10,000 against E3 with 50 per cent binding. *Lot No. 224: Index Biomedical Division Wilson Pharmaceutical and Chemical Corporation. Glenwood. Ill.
fluid
387
summary
Pregnant
Age
by amniotic
Control
subjects
14-41 27.8 22-41 29.5 O-6 1.7 o-4 0.6
Indications for amniocentesis Maternal age Midtrimester abortion Repeat cesarean section Postmaturity Previous anencephalus Severe hypertensive disease Amniotomy in labor Total *All of our control labor floor employing
patients were birth control.
7 3 2
16 hospital
personnel
on the
tion of this pool was determined by the method of Lowry and associateslO and adjusted to a concentration of 0.3 mg. per milliliter by dilution with complete medium. In special instances higher concentrations up to 4.0 mg. per milliliter were used. Choriocarcinoma spent medium. Spent medium from a choriocarcinoma cell line established by Patti110 and Gey” and grown by Dr. P. G. Satyaswaroop in our department was diluted with complete medium 1: 1 and frozen at -20” C. until use. This cell line is known to secrete 1 x lo-” 1.U. of HCG per cell per 24 hours.” Leukocyte migration. Leukocyte migration tests were performed according to the method of Sbborg and Bendixen with minor modifications.g Each experiment consists of four duplicates, which are averaged to obtain the mean area of migration (MA). The migration index (MI) was calculated as follows: MI =
area area
of migration of migration
with antigen without
antigen
x 100.
Thus a MI above 100 indicates migration enhancement and a MI below 100, migration inhibition. Percentage migration enhancement or inhibition equals MI minus 100 or 100 minus MI, respectively, and is displayed in Table II as (+) for enhancement and (-) for inhibition. Statistical analysis. Significance between two sample means was established by Student t test using Fisher’s distribution.
Results Fig. 1 represents the MI of all 20 patients and 11 control subjects when challenged with different antigens. Table II represents the mean percentage migration
enhancement
for
each
trimester
of pregnancy
and
388
Gleicher
February Am. J. Obstet.
et al.
Table II. Mean percentage migration enhancement on PPD4, and in nonpregnant control patients
t standard
deviation
with different
antigens
15. 1979 Gynecol.
in pregnancy,
Trim&u Control
Total Jar First
Third
Second
PPD-4
pregnancy
subjects
a-AF LE
+53.a
11 k 32.Q
+43.4
5 2 40.5
+50.5
16 k 34.4
+40.6
6 + 37.1
+34.2
4 4 12.9
+31.1
14 2 27.1
+4Q.l
17 2 33.Q
+39.3
Y 2 30.1
+41.5
30 + 32.2
8 ‘+- 11.6
+14.1
17 k 9.3
+8.8
3 + 13.0 2 3.9
h-AF 4
ZME AFt
+13.7ze
11.9
ZME OCA iME CCSM
+lO.l
4 k 7.2
+15.1
EME PPA
+3.3
2 T 4.7
9 + 15.9 + 12.0
4 + 16.8 % 7.6
+20.6
10 ‘- 10.4
5 + 15.7 -+ 7.5
+8.3
4 ” 8.7
+12.2
15 k 11.2
+15.0
.5 t- 21.4
+18.3
IY f
+26.7’?
18.1
11 + 16.4 2 14.6
5 + 5.3
13.1
*N = number of cases. tTotal AF (h-AF + a-AF). PPD-4 separately. In addition results for a-AF and h-AF are listed separately facilitating direct comparison of their activity wherever both were obtainable. The MI with AF during the second and third trimester of pregnancy was significantly larger when compared to OCA (p < 0.01). CCSM (p < O.Ol), and PPA (p < 0.0 1). The enhancing activity of a-AF exceeds that of h-AF only slightly (p < 0.04), which allows direct comparison of results obtained with either. This is of importance because obviously h-AF only was available during the first trimester of gestation and in the postpartum period. .1s can be seen from Table II a trend toward enhancement was observed when control patients were challenged with h-AF resulting in decreased significance (p < 0.1) when compared to the amount of enhancement with h-AF in pregnant women of second and third trimester. OCA, CCSM, and PPH showed equal migration areas in pregnant women of all trimesters and nonpregnant control subjects (p < 0.5), not causing enhancement. A trend toward enhancement with AF on PPD-4 was present but did not reach statistical significance, probably due to the small number of cases. None of the antigens caused significant migration enhancement during the first trimester of pregnancy but enhancement with h-AF reached borderline significance when compared in the third trimester with the migration area obtained in the first trimester (p < 0.1). None of the antigens showed migration inhibition in
routine concentrations. When the concentration of PPA was increased to 4.0 mg. per milliliter migration inhibition was encountered. Fig. 2 shows the decrease of MA with increased concentrations of PPA in one patient. The enhancing activity of AF could be abolished by increasing the grade of dilution. Significant enhancement with AF was usually abolished at a dilution of 1 : 50 (Fig. 3). A hormone effect (estrogen, HCG) as the cause of migration enhancement could be excluded by proving that abolition of hormone activity by specific antibodies did not inhibit migration enhancement (Fig. 4).
Comment The assumption that AF represents the fetoplacental unit antigenically seems reasonable in view of out knowledge (although incomplete) about its formation and rapid turnover. Because fetal tissue achieves increased antigenicity during the second and third trimester of gestation it can be concluded that AF should represent this antigenicity at approximately the same time.+’ Using in this study the leukocyte migration assay, as developed by Soborg and Bendixen we anticipated migration inhibition to occur as evidence for the presence of hypersensitivity toward the AF-antigen pool. A possible lack of inhibitory activity by AF was theoretically thought to represent either the absence of fetoplacental antigen in AF or the blockage of antigen by a blocking factor. Such a blockage could render the antigen nonantigenic or, more probable, as in tumor
Volume Number
133 4
Leukocyte migration enhancement
oa-AF
240-I 220 200 -
2
180 160 140 120
-
100
__-
g -
80-
E
60-
Oh-AF
’
6 OCA
’ ’0 ’ . .
’
~02%
’
x PPA
’ ’
1
’
0 l
---t-E.
_-_____
. .
z
i -)r.-
--.1--
0
l
-
.
_
CONC. OF PPA DILUTED WITH COMPLETE MEDIUM
5 ” r
40 200
ttIIIIII;I
----
60 4020 o! ’ ! ! ’ I I ’ ’ ’ ’ 0 4 8 12 16 20 24 28 32 36 4044pP,0,f;MCoNTRoLS WEEKS
389
0
0 0.
-,..-
by amniotic fluid
Fig. 2. Mean migration area (MA) and migration index (MI) in one of the patients (H.D.) when tested with PPA in increasing concentrations and with full medium (ME) as control. Note the decrease in MA and MI with increasing PPA concen-
trations until migration inhibition tion of 4.0 mg. per cent.
is obtained at a concentra-
DAY4
OF PREGNANCY
Fig. 1. Migration index (MI) in pregnant and postpartum patients and nonpregnant control subjects when tested with AF, OCA, CCSM, and PPH.
systems prevent recognition of antigen by maternal effector cells. The additional explanation of nonpresensitized effector cells was excluded by the demonstration of migration inhibition with PPA in high concentrations, as had been used by other authors.2-“, g Thus antigenicity of fetal tissue as well as immunologic competence of maternal effector cells were confirmed by the present study. Surprisingly a new immunologic phenomenon-migration enhancement-was encountered at exactly that period of gestation when migration inhibition was expected to occur with AF as antigen. None of the other antigens exhibited migration enhancement of significant proportions. As a consequence we conclude that AF has to be the causative factor for migration enhancement and exclude the possibility that maternal lymphocytes acquire special properties per se during pregnancy. While the mechanism of migration inhibition is caused by the secretion of migration inhibition factor (MIF) by presensitized T-lymphocytes on contact with the “recognized” antigen, nothing is known about the causative factors involved in migration enhancement. It could be speculated that inhibition of MIF secretion by any mechanism might result in migration enhancement. However, inhibition of migration inhibition factor only
by but
protein does
synthesis
inhibitors
not
in migration
result
prevents
inhibition
enhancement.‘*
”
I:I
I:10 I:50
DILUTION COMPLETE
WITH MEDIUM
Fig. 3. Mean migration area (MA) and migration index (MI) in one of the patients (H.D.) when tested with increasing dilutions of AF and with complete medium (ME) as control. Note the decreasing MA and MI with increasing dilutions of AF
until inhibition medium.
is reached at a dilution of 1: 50 with complete
400 $300 a p 200 2 ” 100 E 0
ESTROGEN ANTIBODY
IiCGANTIBODY
Fig. 4. Mean migration area (MA) in one of the patients (B.K.) when antibodies to estrogen and HCG were added to AF and control chambers with complete medium (ME). Note the significant migration enhancement with AF with or without antibody in comparison to the MA obtained with ME only, with or without antibody.
390
Gleicher et al.
Thus, for migration enhancement there has to be either an additional mechanism to what is obtained with protein synthesis inhibitors or total blockage of MIF secretion, not present in the usual resting state. Although this report represents the first evidence for the presence of a blocking factor in 4F, similar factors have been described previously in serum. Thus Hellstriim and co-workers” described a serum factor abrogating cellular immunity to antigenically foreign mouse embryogenic cells, thereby causing enhanccment of’ growth. In human beings it was shown that maternal serum prevents the cytotoxic effects of maternal lymphocytes 011 trophoblasr.” Abrogation of cellular immunity to placental antigen by maternal serum has in addition been reported by many authors, using different in vitro techniques, including the migration assay. Although migration enhancement has in those cases not been observed, it could be speculated that the difference between no inhibition and enhancement is a quantitative difference in the amount of blocking factor present in serum and AF. This would indicate a very high concentration of blocking factor in AF, making it the perfect vehicle for future purification. Clinical application of such a purified blocking factor does not seem to lie in the too distant future, taking into account the recent report by Rocklin and coworkers,‘” who reported the absence of such a blocking factor in women with repeated idiopathic abortions. IgG was implicated as the decisive factor.‘” The recent description of antigen-antibody complexes in sera of
normal
pregnant
patients
allows
suggestions
for
a
mechanism of nonrejection of pregnancy, very similar to the one described in the nonrejection of malignant tumors.‘” In view of these findings studies are underway in OUI laboratory to correlate migration enhancement with IgG concentration in pregnancy sera and AF and the possible presence of immune complexes. Preliminary data suggest an inverse relationship between migration enhancement and IgG concentration in AF. We are presently unable to explain the large fiuctuations
in
seems
challenging
the
enhancement
pattern
to assume
that
caused they
bp
might
AF.
It
represent
different concentrations of blocking factor in AF. If this assumption proves to be correct a wide variety of clinical applications seems feasible. Thus if hypertensive disease of pregnancy in fact represents an immunologically induced disease process, caused bp partial rejection of the fetal alograft, as believed by many. a lower degree of migration enhancement (because of less blocking factor) could be expected in patients with hypertensive the
early
disease diagnosis
of’ pregnancy. of
this
condition
This
could through
lead
to
simple
amniocentesis in the midtrimester. Changes in lymphocyte function in mild pre-eclampsia have recentl) been reported.” In this connection it seems of interest that the only patient in our study with severe hypertensive disease of pregnancy did not show significant migration enhancement.
REFERENCES
1. Hulka. 2. 3.
4.
5.
6.
7.
8.
J. F., Hsu, K. C., and Beiser, S. M.: Antibodies to trophoblast during the post-partum period. Nature 91: 519, 1961. Koren, Z., Behrman, S. J., and Paine, P. J.: Antigenicity of trophoblastic cells indicated by fluorescein techniques, AM. J. OBSTET. GYNECOL. 104: 50, 1969, Koren, Z., Abrams, G., and Behrman, S. J.: Antigenicity of mouse placenta1 tissue. AM. J. OBSTET. GYNECOL. 102: 340, 1968. Yoitannukorn, V., and Matangkasombut. P.: Human maternal cell-mediated immune reaction to placental antigens, Clin. Exp. Immunol. 11: 549, 1972. Taylor, P. V.. and Hancock, K. W.: Antigenicity of trophoblast and possible antigen-masking effects during pregnancy, Immunology 28: 973. 1975. Yoitananukorn, V., Matangkosombut, P., and Osathanondh, V.: Onset of human maternal cell-mediated immune reaction to placental antigens during the first pregnancy, Clin. Ex’p. Immunol. r5: 593, 19%. Tavlor. P. V.. Gowland. G.. Hancock. K. W.. et al.: Effect of length of gestation on maternal cellular immunity to human trophoblast antigens, AM. J. OBSTET. GYNECOL. 125: 528, 1976. SBborg, M., and Bendixen, G.: Human lymphocyte migration as a paramenter of hypersensitivity, Acta Med. Sand. 181: 247, 1967.
9. Faiferman.
10. Il.
12.
13.
14.
15.
16.
I., Gleicher. N., Cohen, C. J., et al.: Leukocyte migration in ovarian carcinoma: Comparison of inhibitory activity of tumor extracts, J. Natl. Cancer Inst. 59: 1593, 1977. Lowry, 0. H.. Rosenbrough, N. J., Lewis Farr, A., et al.: Protein measurement with the Folin phenol reagent, J. Biol. Chem. 193: 265, 1951. Patillo, R. A.. and Gey, G. 0.: The establishment of a cell line of human hormone-synthesizing trophoblastic cells in vitro, Cancer Res. 28: 1231, 1968. Henney, C. S., Gaffney. J., and Bloom, B. R.: On the relation of products of activated lymphocytes to cellmediated cytolysis, J. Exp. Med. 140: 837. 1974. Hellstriim, E., and Brawn, J.: Abrogation of cellular immunity to antigenically foreign mouse embryogenic cells by a serum factor, Nature 224: 914, 1969. Rocklin. R. E., Kitzmiller, J. L., Carpenter, C. B., et al: Absence of immunologic blocking factor in women who chronically abort, N. Engl. J. Med. 295: 1209, 1976. Petrucco, 0. M., Seamark, R. F.. Holmes, K., et al.: Changes in lymphocyte function during pregnancy, Br. J. Obstet. Gynecol. 83: 245, 1976. Masson. P. L., Delire, M., and Cambiasso, C. L.: Circulating immune complexes in normal human pregnancy, Nature 266: 542, 1977.
