Effect of injected human immunoglobulins on fetal rat development. Spinal neural and osseous changesкод для вставкиСкачать
Effect of Injected Human Immunoglobulins on Fetal Rat Development. Spinal, Neural and Osseous Changes BENJAMIN B. KAMRIN 1 Institute of Rehabilitation Medicine, New York University Medical Center, New York ABSTRACT Rabbit anti-rat-brain serum immunoglobulins injected into pregnant rats on the ninth, tenth, eleventh or twelfth day of gestation resulted in a fetal resorption rate 15 to 30 times higher than that found in normal untreated pregnant rats. Human serum immunoglobulins obtained from normal postpartum mothers produced a similar percentage of fetal resorption when injected by the same routes into pregnant rats of the same gestational age. In neither of the above experiments were malformations detected among the delivered 20-day fetuses. Injection of human serum immunoglobulins obtained from mothers of children with spina bifida manifesta into pregnant rats along similar routes and periods gave comparable fetal resorption rates. Injection of the above immunoglobulin into the lumen of the uterus adjacent to implantation sites gave a low fetal resorption rate and produced varying degrees of skeletal and soft tissue malformations among the viable survivors. The only difference which could be discerned between normal postpartum immunoglobulins and those obtained from mothers of spina bifida manifesta children was characterized in the latter by a two-fold increase in the IgG levels and the immunoelectrophoretic reactivity of its immunoglobulins with human spinal cord antigens, The developmental defects observed were: 1. Cranialthinning and bleb formation of skull bones; widening of the foramen magrium; descent of the obex closer to the foramen magnum. 2. Skeletal-delayed or inhibited calcification of the bodies and spinous processes of the thoracic and lumbar vertebrae; widening of the vertebral canal and central canal of the spinal cord. Alterations in the growth patterns of developing embryos and fetuses have been induced by exposing them to tissue antiserum. Brent et al. (’61) demonstrated that rabbit anti-rat kidney serum injected intravenously into pregnant rats on the eighth day of gestation resulted in severe skeletal congenital malformations in 100% of the fetuses. McCallion (’68) reported that rabbit anti-chicken brain serum placed on 33-hour incubated chick embryos and incubated for a further five to seven days resulteld in the formation of localized or generalized spina bifida in 10% of the surviving embryos. Kamrin (’70) reported that immunoglobulins from mothers of children with spina bifida manifesta produced lesions suggestive of spina bifida and the Arnold-Chiari syndrome when injected into rats on the tenth day of gesANAT. REC., 173: 173-180. tation. It is evident in all of these experiments that the exogenous immunoglobulins arrested or slowed the differentiation of somite cells to cartilage and their subsequent calcification. These experiments also appear to confirm the findings of Holtzer et al. (’55, ’61, ’68) that ( a ) using somite tissue taken from chick embryos and grown in the absence of spinal cord or notochord, the production of cartilage will be prevented; (b) cartilage could be induced to appear by the addition of embryonic intermediate somitic mesoderm (from which the kidney forms) and ( c ) “there is no such thing as an undifferentiated, uninstructed cell” and the inducers or inhibitors are therefore acting on cells Received Nov. 15, ’71. Accepted Feb. 4, ‘72. 1 Present address : Biophysics Research Laboratory. Department of Physics, N.Y.U., New York, N.Y. 10003. 173 174 BENJAMIN B. KAMRIN which have received some genetic information. The following experiments were devised to test whether the immunoglobulins (predominantly IgG) obtained from mothers of children with spina bifida manifesta would produce similar malformations in the rat fetus after exposure to the material during gestation. Prior to performing such an analysis, a series of experiments involving the injection of rabbit anti-rat brain serum and normal postpartum serum via various routes into the pregnant rat was attempted. METHODS Animal eqerimentation. Adult and 20-day fetal Wistar rats were anesthetized and the cerebrum, cerebellum and medulla oblongata dissected out under aseptic conditions and pooled according to age and anatomical entity. Each pooled specimen was homogenized in small volumes of physiological saline (pH 7.2), placed in dialyzing sacs and dialyzed against repeated changes of cold distilled water to remove contained sodium chloride. After lyophilization, each specimen was stored in a separate vial in the freezer (-5°C) until used. Rabbit antiserum was prepared from each specimen by injection of the reconstituted material ( 1 ml physiological saline = 50 mg of antigen) with complete Freund’s adjuvant into the rabbit’s footpads over a 30-day period. The animals were bled at one and two weeks after the last injection. Each of the antiserum was titrated against the original antigen and only antibody titers of 128 or more used in subsequent experiments. The antiserum was also examined for its IgG, IgA and IgM content by the use of Hyland Immunoplates (Hyland Laboratories, California) against standards of pooled non-pregnant rat serum and the enclosed standards. The immunoglobulins in the serum were precipitated from the antiserum by the addition of an equal volume of saturated ammonium sulfate, mixed thoroughly and centrifuged at 2000 rpm for 20 minutes at 5°C. The precipitate was dialyzed against frequent changes of distilled water in a refrigerator for approximately 72 hours. The cleared immunoglobulins were lyophilized and stored in individual vials in the freezer (-5°C) until reconstituted for use by the addition of 1 ml of physiological saline (pH 7.2) for each 25 mg of material. Treatment schedule. On the ninth, tenth, eleventh or twelfth day of gestation, pairs of pregnant rats were laporotomized under ether anesthesia and the exact site of each implantation charted for each horn. One of each control/experimental pair received needle pricks into the uterus between implantation sites or into the site itself; or an injection of saline into the same areas. Or one of the pair would receive an injection of the specific reconstituted immunoglobulins (0.015 to 0.04 ml) intranvenously, into the uterine lumen or intrafetally so that all the fetuses were involved on one side. Each experimental animal received only one type of injection on a specific day with a specific antiserum. Recovery and examination of animals. On the twentieth day of gestation (term = 21 days), each animal was again anesthetized and a laporotomy performed. The two horns were exteriorized and the number and position of each fetus compared with the original findings on the animal. Each fetus was removed and inspected for gross malformations. Abnormalities such as alterations in the eye bulge, cranial defects, cleft lip or cleft palate, position of limbs and number of toes, defects along the dorsal midline such as depressions or bubble-like bulges, or constricted vascularized areas of the limbs and tail were noted if present. Each animal was weighed, then pinned fully extended to a small corkboard which carried the notations of animal number, uterine position and malformation data. Each animal was fixed in 10% formalin. The position of each resorption site was noted and related to the particular treatment; nodular masses when found, were fixed for histological examination or clearing. The protocol for selection and ultimate type of examination was as follows: The implantation site nearest the bifurcation (right or left) was designated as A, the next B, C , and so forth to the tip of the horn. If unresorbed, the fetuses lettered “B’ on the experimental side and in the control animal were cleared and stained with alizarin. The “C” animals from the experi- I I - ~ -. HUMAN IMMUNOGLOBULINS ON FETAL RAT DEVELOPMENT mental and control mothers were sectioned sagitally and stained alternately in sequence with hematoxylin and eosin, Masson’s trichrome, alizarin and Bodian’s silver stain. Sectioned material was examined at X 25 magnification and measurements made with a calibrated ocular reticle. The cleared animals were examined for calcification defects in the skull and vertebrae and for malformations of ribs and limbs under the stereomicroscope at 6 X, 12 X, and 25 X magnifications. Human material. Blood serum from mothers giving birth to normal off-spring (NS-M) was obtained from 24 hours to five days postpartum. Blood serum was obtained from mothers of children with spina bifida manifesta and meningomyelocele (SpBi-M) at times ranging from one week to seven months postpartum. Blood serum was obtained from three spina bifida children ( SpBi-Inf) 72 hours after birth. Fresh infant spinal cord and associated meningomyelocele were obtained at the death of a seven-month-old child. The spinal cord was separated from the meningomyelocele and associated tissues, then homogenized by sonic vibration and grinding with physiological saline and glass beads in a high speed blender. The resulting mixture was filtered under suction in the cold, dialyzed to remove the saline and finally lyophilized. The meningomyelocele was also homogenized, dialyzed and lyophilized. Both of these materials, when reconstituted, served individually as antigen in double gel diffusion and immunoelectrophoresis studies against the human serum. Prior to the precipitation of the immunoglobulins from the human serum, each individual serum was tested for IgG, IgA and IgM levels on Hyland Immunoplates against the contained controls. Yn double gel diffusion studies, the antigens derived from the spinal cord and from the meningomyelocele were placed in central wells in agar and the antibodies in the various immunoglobulins placed in the surrounding peripheral wells. They were allowed to diffuse toward each other in the agar. Where interaction occurs between the antigen and antibody, whitish bands are formed indicating that lines of pre- 175 cipitation are formed where antigen and antibody meet. When the precipitation lines are indistinct or multiple, as often occurs when one is dealing with a complex antigen mixture, immunoelectrophoresis is used. The mixture of antigen is placed in a small well in the center of a microscope slide on which has been placed a 3 mm thick layer of agar. The agar has been mixed with a buffer solution of about pH 8.2 and is connected to reservoirs at either end containing a similar buffer and electrodes. Under the influence of the current, the various proteins in the antigen mixture spread out toward the anode and cathode because of varying electrophoretic mobilities. After several hours, the current is stopped and the slides removed from the machine. A long horizontal trough is cut at one side of the slide. Immunoglobulins derived from specific individuals are placed in the trough and double diffusion takes place. Each protein antigen forms an arc of precipitate as it interacts with its corresponding antibody. In this study, human immunoglobulins from NS-M, SpBi-M and SpBi-Inf sera were processed against the antigens contained in the spinal cord and meningomyelocele. RESULTS In the control group (1) of 256 fetuses, subjected to no treatment and laporotomy, the spontaneous fetal resorption rate was 0.7%. Group 2, subjected to pricking of a needle into the various sites where injections were to be made plus laporotomy gave the following fetal resorption rate: intraperitoneal - 0.8% ; intravenous 1.8; intrauterine - 2.0% ;intrafetal - 2.5%. Group 3, subjected to injection of experimental amounts of physiological saline and laparotomy showed that the vehicle for the lyophilized rabbit anti-rat brain antibodies and human immunoglobulins had a deleterious effect on the fetal resorption rate: intraperitoneal - 1.0; intravenous - 5.0% ; intrauterine-6.0% and intrafetal-12.0%. The injection of equivalent amounts of rabbit anti-brain serum immunoglobulins, as well as NS-M, SpBi-M, and SpBi-Inf immunoglobulins showed a marked increase in the resorption rate of fetuses. Comparison of the lethal effects produced by the 176 BENJAMIN B. KAMRIN injection of different immunoglobulins into specific sites adjacent to the developing fetus demonstrated that the closer the material is deposited to the fetus, the greater the risk of resorption: ~ Antirat-brain Intrauterine (lumen) Intr af e t a1 NS-M SpBi-M SpBi-Inf % % % % 15.0 10.2 3.0 12.5 30.0 33.0 29.0 26.5 testing and confidence limits were made by the use of the “t-test” suggested by Mainland (’52). Using these criteria, it must be noted that normal human serum also caused increases in width except for the first thoracic level where it was similar to the controls. Measurement of sagittal sections confirmed the data obtained from the cleared, alizarin-stained animals. In addition, sagittal sections showed widening at the levels of the foramen magnum, tenth thoracic and first lumbar vertebrae (table 2). The finding of a significantly enlarged foramen magnum suggested the presence of an Arnold-Chiari-likesyndrome and directed attention to an examination of the position of the cerebellum in the sagittal sections. Comparison between the controls and SpBi-M injected experimental animals showed a marked downward movement of the cerebellum and a thickening of the cervical spinal cord (in a greatly enlarged third cervical level of the vertebral canal). In control fetuses, the level of the obex in sagittal sections is approximately 3 mm (range 2.5 to 4.0 mm) above the upper limits of the mineralized occipital condyle. In SpBi-M treated fetuses, the obex descends to the upper limits of the occipital condyle. Such caudal movement signifies a corresponding downward movement of the cerebellum. Measurement of the spinal cord central canal failed to show significant (P = 0.1) enlargement except in the region of the fourth lumbar vertebra (table 3 ) . Precipitin reaction tests in agar of the various immunoglobulins against antigens derived from human spinal cord and the meningomyelocele demonstrated interaction only between SpBi-M immunoglobulins and spinal cord antigen. Since the precipitate obtained was diffuse and not clear-cut, immunoelectrophoretic studies were made of the immunoglobulins obtained from each of the mothers of children with spinal bifida manifesta. The dispersed spinal cord antigen usually showed one definite arc and occasionally one or two more fuzzy arcs. No explanation can be offered why the SpBi-M immunoglobulins showed a low resorption rate to intrauterine injection, except that they proved effective in obtaining viable malformed animals. Intravenous , intrauterine and intr af etal injections of SpBi-M immunoglobulins into 9, 10, or 11 day pregnant rats produced little overt external malformation in the delivered fetuses. Several of the fetuses exposed to intrauterine injection of SpBi-M had irregularly shaped heads and some encircling constrictions (vascularized depressions) on their limbs associated with the umbilical cord. There were no signs of cleft palate or deformed limbs. Clearing of these specimens in alizarin disclosed that the frontal and parietal skull bones were thinned and had transparent blisterlike swellings. Sagittal sections through these areas demonstrated that there were small (approximately 1.0 mm) outpocketings of the skull in which the membranous bone was not mineralized (not stained with alizarin). Further evidence of malformation or delayed development was seen in those cleared animals in which the vertebral bodies and spinous processes appeared to be missing (fig. 1). Examination of the sagittal sections showed that these structures were present but not mineralized. Some animals showed varying degrees of vertebral body mineralization with kidneyshaped or paired small centers of ossification at the lateral borders. Measurement of the interpedicular widths of the cleared, alizarin-stained specimens at the levels of the third cervical, first thoracic, and fourth DISCUSSION lumbar vertebrae demonstrated a signifiAside from the increased resorption rate cant enlargement of the width in the SpBi-M Immunoglobulin-injected animals of fetuses, no specific malformations could over the controls (table 1 ) . Significance be induced by the injection of rabbit anti- HUMAN IMMUNOGLOBULINS ON FETAL RAT DEVELOPMENT 177 Fig. 1 Cleared, alizarin-stained fetal rats delivered by laparotomy on twentieth day of gestation. Left fetus, 3 in left horn, received a total intrauterine injection of 0.03 ml (0.75 m g ) SpBi-M immunoglobulins on tenth day of gestation. Right fetus 3 in right horn, received a total intrauterine injection of 0.03 m l physiological saline. Rib and skull distortions caused by manipulation during photography. rat cerebrum, cerebellum or medulla oblongata antisera. The same results were obtained with human NS-M and SpBi-Inf sera. Only with serum obtained from mothers of children with spina bifida manifesta (SpBi-M) could malformations be induced in injected fetuses. The SpBi-M serum differed from all of the other sera in its high IgG level and the ability to react (precipitate) with spinal cord antigen. The average IgG level for NS-M was 1200 350 mg/100 ml with a range of 1200 to 1750 mg/100 ml. The SpBi-M range was 2200 to 3100 mg/ml or a two-fold increase. The IgA and IgM levels were in the same range for NS-M and SpBi-M sera. If confirmed by a study of a much greater number of mothers of children with spina bifida manifesta, these high IgG levels may provide a clue in determining the etiology of this affliction. It is well known that IgG readily crosses the +_ 178 BENJAMIN B. KAMRIN TABLE 1 Measurements o f vertebral interpedicular width derived f r o m the examination o f cleared alizarinstained rat fetuses delivered by laparotomy o n the twentieth day o f gestation Measurements ( m m ) Treatment 3rd Cervical vertebra 12 1.09%0.04 1,4 1.09 f0.04 1.01 f0.04 rare 2 none 1.08 f0.05 1.06 f0.04 rare none 1.3620.02 4,5 1.18f0.03 frequent frequent 1. None 2. Normal serum 3 8 1.24 % 0.02 3. SpBi-M serum 9 1.42f0.10 i34 1st Thoracic vertebra Non-c alcific ation Total subiects 4th Lumbar vertebra Thorac. bodies Lumbar bodies 1 Mean and standard error. 2 Infrequent kidney-shaped partially calcified bodies in untreated and i n normal serum injected animals; particularly i n 10, 11, and 12 thoracic vertebrae. 3 Saline controls ( 5 ) gave measurements intermediate between untreated and normal serum. 4 Significant difference - P = 0.05 to 0.1. 5 Significant difference - P = 0.1. TABLE 2 Measurements o f histological sagittal sections o f whole 20 day rat fetuses delivered by laparotomy on twentieth day o f gestation Dimension of bony canal ( m m ) Treatment 1. None Number exam. 8 2. Normal serum 10 3. SpBi-M serum 7 1 2 8 Foramen magnum 3rd Cerv. region 1st Thorac. region 10th Thorac. region 1st Lumb. region 4th Lumb. region 1.03f0.04 0.96f0.04 0.96 f0.06 s 1.14 f0.05 1.05 2 0.05 1.41-10.042 1.41f0.103 1.42e0.09 1.26k0.07 1 . 0 1 ~ 0 . 0 2 1 1.06%0.062 ~2 1.01-10.042 1.00%0.02 1.25" 0.03 3 1.23 2 0.02 3 1 . 7 1 f 0 . 0 7 2 ~ 31.55f0.063 1.18 f0.05 Mean and standard error. Significant difference P = 0.05 to 0.1. Significant difference - P = 0.1. - placental barrier by active transport; if it carried an antibody against neural tissues, the IgG molecule could alter the development of these tissues. It must be noted, however, that examination of the IgG levels of three babies with spina bifida manifesta showed readings of 150,575 and 1750 mg/ 100 ml, comparable to the findings of McKay et al. ('67). Therefore, assuming that the IgG can transport anti-tissue molecules through the placenta, it is possible that the SpBi-M immunoglobulins interact directly with the antigens on the predetermined mesenchymal somite cell membrane, thereby altering its metabolic activity (Kamrin, '69). Or, in view of the evidence of Holtzer ('61) and Lash et al. ( ' 6 2 ) evidence that spinal cord or notochord cells or their products are necessary for the differentiation of somite cells to cartilage, the IgG carrying anti-neural antibodies may in some manner inhibit or neutralize the factors elaborated by the developing neural cells. CONCLUSIONS Pregnant rats injected with rabbit antirat cerebrum, cerebellum or medulla oblongata serum immunoglobulins via various routes on the ninth, tenth, eleventh or twelfth day of gestation did not produce offspring with gross or microscopic (as studied at 25 X ) deformities. The injected immunoglobulins did cause an increase in the resorption rates of the implanted fetuses; the major increase occurred from intrafetal injections. Injections of anti-rat brain, NS-M, and SpBi-Inf sera into the lumen of the uterus between the implantation sites resulted in a reduction to onehalf of the fetal resorption rate. Similar injections of SpBi-M serum into the same site resulted in a very low fetal resorption rate (3.0% versus 10 to 1 5 % ) . It was in HUMAN IMMUNOGLOBULINS ON FETAL RAT DEVELOPMENT N a I: 0 2t I 0 0, U 9 +I m 2 8 0 m M 0 0 8 8tI $1 2 0, m 2 ? u, m 8t i 0 0 $1 a ? -! 0 r E -E 0 0 aF r c( 0 9 0 c aF n 0 0 (r c v: C F m u) z z 0 E a I wish to thank Doctors C. A. Swinyard and C. Sansaricq at the Institute of Rehabilitation Medicine, N. Y. U. Medical Center for their aid and support in making this study. I appreciate Miss Patricia Muck's fine technical assistance. 0 9 n c? (D 0 3 0 ? 3 ti 3 2$1 3 0 a 7 5 0 9 0, 0 ? 9 0 tl $1 m 7 $ 0 1 0 0 I: H 8 2 L ' 5 c ACKNOWLEDGMENTS tI M 0, Y 9 0 01 * z this latter group that the spinal, neural and osseous changes were found. The major differences found in the SpBi-M serum compared with the serum of NS-M were: a two-fold increase in the IgG levels of the SpBi-M serum and the ability of the SpBi-M serum to form reactivity bands and arcs in double gel diffusion tests and immunoelectrophoresis preparations. The teratogenic tendency of the SpBi-M immunoglobulins was seen in the widening of the foramen magnum, the vertebral canal and the central canal of the spinal cord. These findings are accompanied by apparent delay in calcification of the vertebral bodies and spinous processes and downward movement of the cerebellum and obex toward the foramen magnum. Whether these findings apply to the formation of human spina bifida manifesta and the Arnold-Chian syndrome is unclear. 0 $1 Y yi 179 Q Z S d LITERATURE CITED Brent, R. L., E. Averich and V. A. Drapiewski 1961 Production of congenital malformations using tissue antibodies. I. Kidney antisera. Proc. SOC.Exptl. Biol. Med., 106: 523-526. Grobstein, C., and H. Holtzer 1955 In vitro studies of cartilage induction i n mouse somite mesoderm. J. Exp. Zool., 128: 333-358. Holtzer, H. 1961 Aspects of chondrogenesis and myogenesis. In: Synthesis of Molecular and Cellular Structure. D. Rudnick, ed. Ronald Press, New York, pp. 35-87. 1968 Induction of chondrogenesis: A concept in quest of mechanisms. I n : EpithelialMesenchymal Interactions. R. Fleischmajer and R. E. Billingham, eds. Williams and Wilkins, Baltimore, pp. 15fG164. Kamrin, B. B. 1969 Role of alpha globulins in immuno-suppression: Reactive site occlusion hypothesis. Transpl. Proc., 1: 506-510. 1970 Possible influence of serum immunoglobulins on embryological malformations. Abstract p. 192 in IXth International Congress of Anatomists. Leningrad, Russia. Lash, J. W., F. A. Hommes and F. Zilliken 1962 Induction of cell differentiation. I. The in vitro induction of vertebral cartilage with a low- 180 BENJAMIN B. KAMRIN molecular-weight tissue component. Biochem. Biophys. Acta, 56: 313-319. Mainland, D. 1952 Elementary medical statistics. W. B. Saunders Co., Philadelphia, pp. 147-208. McCallion, D. J. 1968 The production of spina bifida in the chick embryo by brain antibodies. Abstract 10 in The Teratology Society. 8th Annual Meeting. Buck Hills, Pa. McKay, E., H. Thom and P. Gray 1967 Immunoglobulins in umbilical cord plasma. 11. Congenital deformities, other abnormalities and multiple pregnancies. Arch. Dis. Child., 42: 264-274.