Factors Influencing - Fusion of R a t Palates Grown in vitro GORDON S. MYERS? NICHOLAS L. PETRAKIS A N D MELVIN LEES Graduate Group in Nutrition, Department of Epidemiology and International Health, University o f California Medical Center, Sun Francisco, California ABSTRACT A technique is described for growth in vitro of embryonic rat palatal tissue on defined and semi-defined media. Palatal fusion closely resembling that occurring i n vivo is found in these preparations. Fusion occurs in a large number of cases even in the absence of an animal extract. However, development is improved if fetal calf serum is added to the synthetic medium. Using Leibovitz Medium G 1 5 with 20% fetal calf serum, complete fusion was obtained in 87% of the palates, partial fusion in 9% and no fusion was found in only 4% of the palates cultured. Incorporation of galactoflavin, a riboflavin antagonist in a riboflavin-deficient diet of the mother or i n the culture medium results in a decreased number of complete fusions and corresponding increase in partial fusions. A suggestion is made as to the manner in which galactoflavin acts i n vivo to produce palatal cleft, based on the incidence and type of clefts and the histological appearance of the palatal tissues. A method is described in which the tissue sections are sliced prior to culture so that the behavior of individual palatal shelves can be studied during the fusion process. Fusion occurs between shelves if the medial margins are apposed, regardless of whether the two shelves come from the same embryo, from littermate embryos, or from non-littermate embryos. Some possibilities for application of this technique are suggested. The part that nutritional factors play in normal development of the mammalian embryo is both basic and complex. Studies of the effects of nutrients on development have often made use of agents which are metabolic inhibitors or nutrient antagonists and which thereby alter the nutritional state of the embryo, (Nelson et al., '55, '56; Chamberlain et al., '63). Maternal dietary deficiencies of riboflavin (Deuschle et al., 'Sl), folic acid (Asling et al., 5 5 ) , niacin (Chamberlain and Nelson, '63), or vitamin A (Kalter and Warkany, '59), as well as hypervitaminosis A (Mauer, '64), have been shown to interfere with normal palatal fusion as demonstrated by the occurrence of palatal clefts in the young of rats subjected to these treatments. However, many other factors may produce the same anomaly. Because the cleft is only the end-result of an altered morphogenetic process, little knowledge has been gained concerning the cell and tissue changes responsible for the occurrence of the defect. The use of in vitro techniques employing synthetic or semi-synthetic media may permit study of the action of specific nutrients on the palatal tissues and show the ANAT.REC., 162: 71-82. tissue changes which result in palatal cleft. Moriarty, Weinstein and Gibson ('63) have reported that the palatal structures of the embryonic rat can be grown in culture, during the period of normal palatal fusion, on a plasma clot containing chick embryo extract. Pourtois ('66) has also reported a procedure for culturing isolated palatal shelves in vitro. Palatal fusion closely resembling that seen in viuo was demonstrated in both instances. This investigation was undertaken to devise a modification of these methods which would circumvent the use of the plasma clot and permit a more precise control of the nutrients available to the tissues as well as to facilitate manipulation of the tissue fragments during the culture period. At the same time, a high percentage of palatal fusions under normal conditions of culture was desirable. We shall describe some of the conditions found necessary for suc1This research was supported in part by the National Institute of Health Grant DE-20 and by a grant from the Research Cdmmittee of the. Academic Senate. Universitv of California. San Francisco. 2 Present address: Faculty of Dentistry, University of British Columbia Vancouver, British Columbia. 3 Present address: School of Home Economics, University of British Columbia, Vancouver, British Columbia. 71 72 GORDON S . MYERS, NICHOLAS L. PETRAKIS A N D MELVIN LEE cessful fusion of the palatal shelves and tioned on the canvas with the mandible present the results of studies on the effects upward. Iridectomy scissors were used to of galactoflavin ( a riboflavin antagonist) cut through the ramus of the mandible on on fusion. We shall also report on experi- either side, allowing the mandible and ments which indicate that fusion can occur tongue to be removed without disturbing the between individual palatal shelves from palatal shelves. Further trimming was done both littermate and non-littermate rat with cataract knives until a section approxiembryos, which may have a bearing on the mately 4 mm wide, 5 mm long and 1 mm role of genetic constitution in palatal thick was obtained, which consisted of fusion. maxillary tissue, palatine shelves and portions of the nasal septum and capsule. The MATERIALS AND METHODS two shelves at this time were usually separated by a space of 0.25-0.50 mm although Experimental animals Female Sprague-Dawley rats, 100-150 connected anteriorly and posteriorly by the days of age, were housed two or three to a tissues of the lip and pharynx. The secage and given stock rat pellets and water quence of steps and final preparation are ad libitum. A timed-pregnancy method was shown in figure 1. The procedure involves minimal maused with the day of finding sperm in the nipulation and disturbance of the palatine vaginal smear designated as day zero of pregnancy. With this timing procedure, processes and permits optimum alignment palatal fusion normally takes place dur- of the shelves during culture. ing the sixteenth gestational day in this In vitro culture strain of rat. Culture chambers were prepared by placThe pregnant rats were caged individually or in pairs, given water and either ing stainless steel gauze in the form of a stock diet or an experimental diet ad low platform, in each section of a compartmented disposable Petri dish. A square ( 2 libitum, and weighed twice weekly. On day 14 8 hours or day 15 8 hours cm x 2 cm) of ethanol-washed rayon acethe female was lightly anesthetized with tate was placed on the wire platform. Culether and killed by cervical dislocation. The ture medium was introduced beneath the abdomen was swabbed with an alcohol platform only until the rayon acetate square sponge and the peritoneal cavity opened was wetted. The dissected tissue section with sterile instruments. The uterus was was lifted from the microscope stage with removed and rinsed in cold sterile Tyrode’s a small spatula, rinsed in Tyrode’s solution, solution. The embryos were removed indi- and placed on the rayon acetate square vidually from the uterus and embryonic with the oral surface upward. In some exmembranes and also rinsed in Tyrode’s so- periments, the rayon acetate was replaced lution. Strict sterility was maintained in with a Millipore filter6 with no adverse effect. these and all subsequent procedures. The cultures were incubated at 37°C for 72 hours, with partial or complete removal Palate dissections and replacement of the medium after 24 The stage of a dissecting microscope was hours. At the end of the incubation period, covered with a piece of sterile canvas kept tissue was carefully detached and lifted moist with Tyrode’s solution. The embryo the from the rayon acetate square and placed was placed on the canvas and an initial in Bouin’s fixative for 24 hours. Following slice made from the tip of the snout the section was embedded in parthrough the region of the eye, to the occiput, fixation, affin, cut serially in a transverse plane at using a pair of mounted cataract knives (Beaver-small).6 A parallel slice was made 10 p and mounted sections stained with through the neck, severing the remainder hematoxylin and eosin. of the head from the body. The resulting 4 Foodstuff Processing Co., San Francisco, Calitissue section, consisting of the maxillary fornia. 5 Rudolph Beaver, Belmont Massachusetts. vault and associated structures plus the at6 0.45 B Type SM, 25 mm.,’Millipore Filter Co., Bedtached mandible and tongue, was posi- ford, Massachusetts. + + 73 FACTORS INFLUENCING PALATAL FUSION Fig. 1 Diagrammatic representation of technique used in obtaining embryonic palatal tissue. ( a ) Head of embryo illustrating first and second cuts, as described in text. ( b ) Oral view of section as cultured, showing morphology of palatal shelves and surrounding tissues. Culture media Two synthetic media were tested: Waymouths Medium MB 752/17 (Waymouth, '59) which contains a bicarbonate buffer and was used with a 5% C02-95% compressed air gas phase for maintenance of pH, and Leibovitz Medium L-15 (Leibovitz, '63), buffered with phosphate and used in free gas exchange with the atmosphere. Both media contain inorganic salts, amino acids and water soluble vitamins, including riboflavin, but neither contain fatty acids, fat soluble vitamins, or purine or pyrimidine components. The vitamin antagonist, galactoflavin? was used only in conjunction with L-15 medium containing 20% fetal calf serum. Leibovitz medium L-15 normally contains 0.1 pg/ml riboflavin. Galactoflavin was added to L-15 to produce concentrations of 1, 2 or 3 pg/ ml, an amount believed sufficient to antagonize the riboflavin present in the defined medium as well as the unknown concentration in the fetal calf serum. galactoflavin was included in the culture medium as previously noted. Control diets contained equivalent quantities of riboflavin in place of galactoflavin. The basic diet was composed of (in gm per 100 g m ) : vitamin-free casein, 20; powdered sucrose, 43; corn starch, 23; peanut oil, 8; OsborneMendel salt mix, 3; cod liver oil, 2; vitamin mix (without riboflavin), 1. The vitamin mix contained (Gms./Kg.) : choline dihydrogen citrate, 150; inositol, 10; thiamine mononitrate, 0.5; pyridoxine hydrochloride, 0.6; calcium pantothenate, 0.6; niacinamide, 2; folic acid, 0.2; biotin ( 1% triturate), 1; para-aminobenzoic acid, 0.6; vitamin E (33% triturate), 0.5; menadione, 0.5; vitamin BIZ (0.1% triturate), 20; Mannitol, 813.6. Experimental diets For studies on the effect of galactoflavin, either a powdered riboflavin-deficient diet containing 55, 100 or 250 mg galactoflavin per kilogram diet was offered to the pregnant rat from the first day of gestation or 7 Waymouth's Medium was obtained from Hyland Laboratories Los Angeles California. 8 Leibovitz' L-15 was obtained from Microbiological Associates. Washington. D. C. L-15 contains 0.1 fig Shelf repositioning experiments In order to examine the competency of the palatal shelves to fuse independently of.other aspects of palatal development, a procedure was devised whereby the individual palatal shelves could be repositioned in- riboflavinjml. ' 9 Galactoflavin was obtained from Microbiological Associates Inc Bethesda, Maryland, through the courtesy & the %ancer Chemotherapy National Service Center, N.I.H., Department of Health, Education and Welfare, Bethesda, Maryland. 74 GORDON S. MYERS, NICHOLAS L. PETRAKIS AND MELVIN LEE dependently of each other or recombined with shelves from other embryos. Preliminary dissection of the palatal section was carried out as previously described. The section was then placed, oral surface upward, on a Millipore filter. With paired cataract knives, an initial cut was made through the lip, continued posteriorly between the separated palatal shelves and carried through the pharyngeal tissues to the posterior margin of the section (fig. 2). D i r e c t i o n af Slice ! i I f I I I I I Fig. 2 Diagrammatic representation of slicing technique used to separate palatal shelves in shelf-repositioningexperiments. Extreme care was taken to avoid touching the shelves at any time. In the process, the filter was also cut in two parts, each with one-half of the original section. The halves were moved apart slightly to assure complete separation, then placed on the stainless steel platform and positioned so that the medial borders of the two palatal shelves appeared to contact. This procedure permitted culturing together shelves from the same embryo or shelves from littermate or non-Iittermate embryos. Subsequent handling and examination of these sections was by the routine used for control sections. RESULTS Figure 3 represents a transverse section of the palatal region from a 14 + day old embryo, immediately following dissection. At this stage the palatine processes were well formed and consisted of a shelf of mesenchymal cells and ground substance, covered by a thin layer of columnar epithelium. Because the preparations are cultured with the oral surface upward, the shelves tend to assume a horizontal position soon after dissection regardless of the age of the embryo, Figures 4a and 4b show a corresponding preparation after growth for 72 hours on L-15 medium with 20% fetal calf serum. The shelves had contacted and fused in the midline, the epithelial covering in the area of contact had disappeared and there was confluence of the mesenchymal cells across the midline. The fusion line was, in some cases, marked by a slight depression of the epithelium or a midline groove on the oral surface. Persistence of a laminated epithelial layer in the midline was rarely found. The number of fused, partially fused, or non-fused palates after culture for 72 hours on Waymouth's Medium or on Leibovitz Medium L-15 is shown (table 1 ) . Complete fusion was defined as fusion over more than two-thirds of the length of the palatai suture and partial fusion as fusion of onethird to two-thirds of the length of the suture. Fusion of one or the other type was obtained in a large number (88% ) of the palate sections grown on MB752/1 containing no fetal calf serum. In preliminary studies, M e r e n t concentrations of fetal calf serum were used with L-15 to determine an optimum concentration for palatal fusion. The percentage of complete fusions increased as the concentration of fetal calf serum was raised from 5% to 20%. Of a large series of cultures (89) grown on L-15 with 20% fetal calf serum, 96% showed partial or compIete fusion of the palatal shelves and this concentration was used for all subsequent cultures. Although fusion was obtained with Waymouths Medium without calf serum and with Leibovitz Medium with 5 and 10% calf serum, the tissues appeared less vital and were smaller in size after a 72-hour culture period when compared with sections grown on L-15 with 20% fetal calf serum. In experiments involving palatal slices and repositioning of the shelves, fusion was found in nine of a total of 1 4 palates (64% ) 75 FACTORS INFLUENCING PALATAL FUSION Fig. 3 Transverse section through embryonic rat palate upon removal from 14 + day embryo. Palatal shelves (PS) and nasal septum (NS) Hematoxylin and eosin. x 20. TABLE I Fusion of rut embryo palates cultured on synthetic media with and without added fetal calf serum 1 Medium Waymouth’s MB 75211 2 Leibovitz L-15 1 2 3 % Fetal Number of calf serum palates Partially fused Not fused 0 26 19 (73% ) 4 (15% ) 3 (12%) 5 10 20 6 5 89 4 (66% ) 4 (80%) 77 (87% ) 2 (34% ) 1 (20% ) 8( 9%) 4 ( 4%) + 14 + and 15 day palates cultured for 72 hours at 37°C. Cultured in 5% C02-95% Air. 1% glutamine added. grown on L-15 with 20% fetal calf serum. The cutting process usually resulted in a loss of the nasal septum, making proper alignment of the shelves for the full length of the fusion line difficult. In those cultures in which the medial margin of one shelf was in contact with an area other than the medial margin of the other shelf, fusion did not occur. When a section containing a palatal shelf from one embryo was combined with a similar section from a littermate embryo, the medial margins fused in a manner grossly and histologically indistinguishable from fusion of shelves from the same embryo. Fusion was also found between sections of this kind taken from non-littermate embryos. A combination of non-littermate sections is shown in figures 5a and 5b as i t appeared following a culture period of 72 hours on L-15 medium with 20% fetal calf serum. 76 GORDON S. MYERS, NICHOLAS L. PETRAKIS AND MELVIN LEE + Fig. 4a Transverse section through 14 day embryonic rat palate following growth in vitro for 72 hours. The fusion line ( F ) is denoted. Hematoxylin and eosin. X 15. Fig. 4b Higher magnification of the fusion line shown in figure 4a. X 100. 77 FACTORS INFLUENCING PALATAL FUSION + Fig. 5a Transverse section of 14 day embryonic rat palate consisting of palatal shelves from two separate non-littermate embryos grown in vitro for 72 hours. Hematoxylin and eosin. X 10. Fig. 5b Higher magnification of the fusion line shown in figure 5a. X 40. 78 GORDON S. MYERS, NICHOLAS L. PETRAKIS AND MELVIN LEE amounts of galactoflavin in the maternal diet resulted in a corresponding increase in the percentage of non-fusions, this relationship did not hold true when galactoflavin was added to the culture medium. If palates from mothers fed galactoflavin (100 mg/Kg of diet) were grown on L-15 medium containing 20% fetal calf serum Table 2 presents the results obtained when galactoflavin was included in the maternal diet or when it was incorporated in the culture medium. In both cases, there was a decrease in the percentage of complete fusions and a n increase in percentage of partial fusions or non-fusions when compared with controls. Although increasing TABLE 2 E f f e c t of galactoflavin on in vitro palatal fusion in Sprague-Dawley rat embryos Galactoflavin 1 No. palates cultured Fused Partially fused Not fused 42 31 14 32 (76%) 19 (61% ) 7 (50% ) 9 (21% ) 9 (29% ) 3 (22% ) 1( 3%) 3 (10%) 4 (28% ) 11 13 6 6 (55% ) 5 (45%) 3 (23% ) 2 (33%) 0 9 (69% ) 4 (67% ) 1( 8%) 0 9 7 (78%) 1 (11% ) l(ll%) 14 13 (93% ) 1 ( 7%) 55 mg/kg of diet 100 mg/kg of diet 250 mg/kg of diet 1 pg/ml medium 2 pg/ml medium 3 p g / m l medium 100 mg/kg of diet + riboflavin 4 mg/ml in medium None (control) 1 Cultured o n Leibovitz Medium L-15 with 20% fetal calf serum for 72 hours at 37°C. + day embryonic rat palate grown i n vitro for Fig. 6 Transverse section through 14 72 hours on medium containing 1 p g/ml galactoflavin. Note persistence of epithelial lamina in the fusion line. Hematoxylin and eosin. X 100. FACTORS INFLUENCING PALATAL FUSION and supplemented with riboflavin, the percentage of complete fusions increased slightly. Figure 6 shows a transverse section through a palate cultured on standard medium with 1 g/ml galactoflavin added. No generalized histological changes were evident. The shelves and surrounding tissues appeared well developed, with normal tissue organization. The principal difference between galactoflavin-treated and non-treated palates was a persistence of the epithelial layers in the fusion line of the treated palates in many instances. Single and double strands of epithelial cells were found as well as epithelial remnants. DISCUSSION In vitro procedures offer several advantages for study of the effect of environmental factors on developmental processes. Direct observation is possible during the process of shelf fusion. Nutrients are directly available to the tissues and are not subject to maternal alteration or rejection. The quantity and quality of the available nutrient factors can be controlled to some extent. Most important, developmental changes in the whole embryo do not influence the course of development of particular tissues or organs in vitro. In the case of palatal fusion, neither the position of the tongue at the time of closure and fusion of the shelves nor continued growth in width of the head need be considered to influence the fusion process during culture. Therefore one may distinguish between physical factors which prevent or disrupt seIf apposition and factors inherent in the tissues of the shelves which inhibit fusion. Finally, manipulation of the tissue sections is possible in vitro which cannot be carried out in the whole organism. On the other hand, no in vitro procedure can completely reproduce conditions occurring in viva Tissue metabolism is altered by the new environment as well as by the trauma of dissection. The anatomical relationships of organs are disrupted so the influence of morphogenetic movements and inductive processes may be greatly decreased. In the culture of palatal tissues, disruption of this kind is produced by shelf displacement, movement or loss of the nasal septum, and growth in an inverted 79 position. Swelling of tissues in culture is a constant finding, and its effects are not subject to interpretation. Increased hydrostatic pressure in the shelves may affect shelf apposition and must be considered as a factor in palatal fusion in vitro. The experiments reported here indicate that the dissection techniques employed and the use of a liquid defined or semidefined medium will produce tissues which are viable and capable of a pattern of development similar to that found in vivo. Palatal fusion in vitro was obtained in many cases which was similar, histologically, to that seen in sections from embryos of the same developmental stage fixed immediately after removal from the uterus. Cell size and morphology appeared normal although intercellular spaces were more evident due to edema of the tissues. Mitotic figures were commonly seen in sections from palates after they had been grown in culture for 72 hours. Grooving or lack of fusion was often found in the anterior third of the fusion line in control palates. This resulted in short partial clefts in some preparations. The cause is not known but may be due to the fact that this area is in the center of the tissue section and therefore the least well nourished. Unphysiological pH might also occur in this region. These conditions may not permit completion of the fusion process during the period of culture. The studies involving various concentrations of fetal calf serum in the medium indicate that at least 10% serum is required for optimum growth and development. Mesenchymal fusion was obtained with as little as 5% calf serum but palatal development was for the most part unsatisfactory. It should be noted, however, that a high percentage of fusions was obtained with the completely defined medium, Waymouths MB 752/1, although the extent and type of fusion was not as useful for this study as that found with L-15 plus 20% fetal calf serum. The findings are in agreement with other studies indicating that large tissue sections cannot be successfully cultured unless an animal component such as plasma, embryo extract, or serum is added to the defined medium. It has been recognized for many years that congenital defects of the skeletal sys- 80 GORDON S. MYERS, NICHOLAS L. PETRAKIS AND MELVIN LEE tem can be produced experimentally by dietary riboflavin deficiency (Warkany and Nelson, '40). Soft tissue anomalies in addition to skeletal defects have been produced by riboff avin-deficient diets to which has been added the riboflavin antagonist, galactoflavin (Nelson et al., '56). The tissue sections used in the present experiments were at a developmental stage in which the only cartilage present occurred in the nasal capsule and septum and therefore the defects in palatal development represent anomalies of soft tissue development. The levels of galactoflavin used by other workers (Nelson et al., '56) to produce clefts in the Long-Evans rat did not cause clefts in 18-day-old Sprague-Dawley embryos and higher levels were consequently used. A strain difference in degree of susceptibility to the antagonist is indicated by this finding. Similar strain differences in response to galactoflavin have been reported for mice (Kalter and Warkany, '57). The increase in the incidence of partial fusions may be significant in view of the increase also noted in the number of fusions of the epithelial type. The epithelial fusions were often incomplete and resulted in deep grooves which occasionally opened into the nasopharynx to produce partial clefts. The incidence of partial clefts and the observed epithelial fusions suggested that galactoflavin may act in vitro by interfering with the final stage of palatal fusion; i.e., merging of the mesenchyme. This could involve a delayed breakdown of the epithelial lamina which prevents merging of the mesenchymal celIs during the culture period. If this were the case, the primary effect of galactoflavin might be on the lysosomal or macrophagic processes responsible for epithelial degeneration and removal. On the other hand, the effect may be an inhibition of the mesenchymal cells themselves so that the influence of these cells on the epithelial lamina is lost. If persistence of the epithelial lamina occurred in uivo, it seems possible that continued growth in width of the head could exert enough tension on the weakened fusion line that it would rupture and result in palatal cleft. It is widely recognized that shelf position at the normal time of palatal fusion is of great importance in the production of clefts of the secondary palate. If the shelves remain in a vertical or semi-vertical position on either side of the tongue past a certain developmental stage, growth of the other cephalic structures wiU make contact (and therefore fusion) of the shelf borders in the midline impossible and result in palatal cleft. At the same time it is suggested that factors which interfere with the fusion process itself may produce cleft palate even though the shelves are in a horizontal position at the proper time and perhaps with medial borders which contact each other. The cleft thus produced may be indistinguishable in the newborn animal from one produced by improper shelf position. The experiments involving slicing and repositioning of the individual palatal shelves suggest that the medial borders of the shelves may have a competency to fuse which is not present in other surfaces of the shelves. Other epithelial surfaces of the palate sections which were forced into close contact due to conditions of the in vitro technique (e.g., embryonic lip and future alveolar ridge) did not fuse even though the conditions of culture resulted in palatal fusion. If the medial border of one shelf was grown in contact with the oral or nasal surface rather than the medial edge of the opposing shelf, fusion did not occur. Other in vitro investigations (Myers et al., '67) suggest that at least one teratological agent (6-aminonicotinamide,a niacin antagonist) can specifically affect the competency of the shelves to fuse since fusion fails to occur in the presence of this antimetabolite even though the medial borders are in close apposition during the culture period. The results with galactoflavin indicate that definitive fusion can be inhibited even though fusion of the surface layers does occur. Whether the competency to fuse resides in the surface epithelium or in the underlying mesenchyme or is the expression of an interaction between the two has yet to be determined. In any case, these studies suggest that palatal cleft may be the result of a localized alteration of the tissues involved in fusion as well as more generalized changes affecting the position or growth of the shelves as a whole. The slicing procedure offers a means of studying the effects of differential treatment of the palatal tissues in that one shelf FACTORS INFLUENCING PALATAL FUSION 81 may be treated with nutrient antagonists, Deuschle, F. M., E. Takacs and J. Warkany 1961 Postnatal dentofacial changes induced in rats vitamin excesses, chelating agents, or other by prenatal riboflavin deficiency. J. Dent. Res., substances while the other shelf remains 40: 366-377. untreated. Vital staining of one shelf would Kalter, H., and J. Warkany 1957 Congenital malformation in inbred s t r a i n s of mice induced offer one approach to the problem of morby riboflavin-deficient galactoflavin containing phogenetic movements of shelf elements diets. J. Fxp. Zool., 136: 531-565. during palatal fusion. - 1959 Experimental production of congenital malformations in mammals by metaThe ability to obtain fusion between bolic procedures. Physiol. Rev., 39: 69-115. shelves of W e r e n t embryos offers a tool for Leibovitz, A. 1963 The growth and maintethe study of genetic influences on palatal nance of tissue-cell cultures in free gas exfusion. The effect of embryonic age on the change with the atmosphere. Am. J. Hyg., 78: 173-180. competency of the shelves to fuse may also I. 1964 Vitamin A-induced congenital be investigated by combining shelves from Mauer, defects in hairless mice. Biol. Neonat., 6: 26. embryos of various ages. Moriarty, T. M., S. Weinstein and R. D. Gibson By utilizing the in vitro technique with 1963 The development in vitro and in vivo of fusion of the palatal processes of rat emdefined or semi-defined liquid media, varibryos. J. Embryol. Exp. Morph., 22: 605-619. ous factors of an environmental and heredi- Myers, G. S., N. L. Petrakis and M. Lee 1967 tary nature may be investigated and comEffect of 6-aminonicotinamide and of added pared with respect to their influence on vitamin A on fusion of embryonic rat palates in vitro. J. Nutr., 93: 252i-262. fusion of the palate. ACKNOWLEDGMENTS Grateful acknowledgment is made to Miss Hannah Oto for technical assistance and to Miss Charlott Havemann for preparation of the figures. LITERATURE CITED Asling, C. W., M. M. Nelson, H. V. Wright and H. M. Evans 1955 Congenital skeletal abnormalities in fetal rats resulting from maternal pteroylglutamic acid deficiency during gestation. Anat. Rec., 121: 775-789. Chamberlain, J. G., and M. M. Nelson 1963 Multiple congenital abnormalities in the rat resultine from acute maternal niacin deficiency during pregnancy. Proc. SOC.Exp. Biol. and Med., 112: 836-840. Nelson, M. M., H. V. Wright, C. W. Asling and H. M. Evans 1955 Multiple congenital abnormalities resultine from transitorv deficiencv of pteroylglutamic acid during gestation in the rat. J. Nutr., 56: 349-363. Nelson, M. M., C. D. C. Baird, H. V. Wright and H. M. Evans 1956 Multiple congenital abnormalities in the rat resulting from riboflavin deficiency induced by the antimetabolite galactoflavin. J. Nutr., 58: 125-134. Pourtois, M. 1966 Onset of the acquired potentiality for fusion in the palatal shelves of rats. J. Embryol. Exp. Morph., 16: 171-182. Warkany, J., and R. C. Nelson 1940 Appearance of skeletal abnormalities in the offspring of rats reared on a deficient diet. Science, 92: 383-384. Waymouth, C. 1959 Rapid proliferation of sublines of NTC 929 (Strain L) mouse cells in a simple chemically defined medium (MB 752/1). J. Nat. Cancer Inst., 22: 1003-1018.