Experimental study on pathogenesis of polydactyly of the thumb The pathogenesis of polydactyly of the thumb was investigated by administration of single dose of cytosine arabinoside (ara-C, 100 mg/kg) to pregnant SD rats on day 11. Marked growth discrepancy between ectoderm and mesoderm of the limb bud was observed both morphologically and biochemically 24 hours after treatment. The growth discrepancy was induced by interference with proliferation process of mesodermal cells. The different susceptibility of ectodermal cells and mesodermal cells to the teratogenic agent was found to depend on the stage difference of cellular differentiation of each tissue. Cephalocaudal asymmetry of the limb bud, as well as relative overgrowth of ectodermal tissue, was postulated to be the cause of protrusion of the preaxial border of the limb bud, which was an initial sign of extra first digit j()rmation.
Hiroshi Nogami, M.D., and Atsuhiko Oohira, Ph.D., Aichi, Japan
Polydactyly of the thumb is the most common congenital abnormality of the hand. It is frequently unilateral and sporadic, suggesting the environmental factor rather than the genetic one might playa role in production of polydactyly of the thumb. However, teratogenesis of polydactyly of the thumb has not yet been elucidated. Administration of cytosine arabinoside (ara-C) , an antitumor drug, to pregnant rats or mice is known to produce polydactyly of the first digit in fetuses at a very high rate.I-:l Increase of digital number in spite of reduced cell number in the limb bud, and different susceptibility of ectoderm and mesoderm of the limb bud to the drug were reported in those experiments,2' 3 but the mechanism to produce polydactyly has not been explained. On the other hand, we have found that 2,2' -dipyridyl (DIP), a chelator for ferrous iron, produced only oligodactyly in rat fetuses. 4 Therefore, comparative studies on productions of polydactyly of the first digit by ara-C and oligodactyly by DIP in relation to the susceptibility of ectoderm and mesoderm of the limb bud to the teratogenic agent may bring us closer to understanding of pathogenesis of polydactyly of the thumb in the human being.
From the Department of Orthopaedic Surgery, Central Hospital, and the Department of Embryology, Institute for Developmental Research, Kasugai, Aichi, Japan. Received for publication Aug. 30, 1979. Reprint requests: Hiroshi Nogami, M.D., Department of Orthopaedic Surgery, Central Hospital, Kamiya-Cho, Kasugai, Aichi 480-03, Japan.
0363-5023/801050443+08$00.8010
Material and methods Sprague-Dawley rats (Japan Clea Co., Osaka, Japan), weighing about 280 gm, were caged overnight with males in rat breeding rooms with a 12-hour light cycle and constant temperature (24° C) and humidity (55%). Nine 0 'clock of the following morning was considered to be the beginning of day 0 of gestation if vaginal plugs were noted. Ara-C (l00 mg/kg body weight) was administered intraperitoneally at 7 P.M. on day II to each pregnant rat. Skeletal malformations. Ten rats were killed on day 21 of pregnancy, and 110 fetuses were fixed in 10% formalin at 4° C for 4 days, stained with 0.25% methylene blue in 70% ethanol containing 1% (v/v) HC!" and cleared in tricresyl phosphate- tributyl phosphate (83: 17)6 for examination of cartilaginous skeleton of the digit. Five litters (50 newborns) were left to grow for roentgenographic observations until 5 weeks after birth. Electron microscopy. Ten embryos from three litters were removed 24 hours after administration of ara-C and the hindlimb buds excised and placed in 3% glutaraldehyde buffered to pH 7.4 with O.IM sodium cacodylate at 4° C for 4 hours. Postfixation was carried out in cacodylate-buffered 2% osmium tetroxide at 4° C for 3 hours. Some limb buds were fixed in a solution of two parts of cacodylate-buffered 3% glutaraldehyde and one part of ruthenium red (TAAB Lab.) at 4° C for 2 hours, and postfixed in equal parts of 2% osmium tetroxide in cacodylate buffer and ruthenium red for 3 hours. 7 The samples were dehydrated in a series of solution of ethanol and acetone and embedded in Epon
© 1980 American Society for Surgery of the Hand
THE JOURNAL OF HAND SURGERY
443
444
Nogami and Oohira
The Journal of HAND SURGERY
Fig. 1. A through C are views of the cartilaginous skeleton of the rat hind foot on day 21 of fetal age, showing polydactyly of the first digit caused by ara-C administration. D through F are roentgenograms 5 weeks after birth.
812. Preliminary sections were cut at 0.5 p,m and stained with toluidine blue for the light microscopic observation and for the purpose of orientation. Thin sections were stained with uranyl acetate and lead citrate. Sections from samples stained with ruthenium red were stained with uranyl acetate. Observations were carried out under a lEaL JEM-IOOB electron microscope. Epon-embedded sections of the hindlimb buds of additional 10 embryos from three litters were examined under light microscope 48 hours after injection of ara-C. Biochemical analysis. Forty-two embryos (three litters) were removed 24 hours after administration of ara-C and 21 of them were incubated in 0.25% trypsin at 37° C for 20 minutes to loosen the ectodermal envelope of the limb buds. Embryos were then transfered
into ice-cold physiological saline solution, and the ectoderm was separated from the hindlimb buds with a pair of sharpened microsurgery forceps under a dissecting microscope. Following separation of ectoderm, the hindlimb buds were excised from the embryos and pooled for DNA measurement. The hindlimb buds with ectoderm were removed from other 21 ara-C treated embryos in physiological saline solution at 4° C. The hindlimb buds from 40 control embryos (three litters) were also divided into the above-described two groups-the limb buds without ectoderm and with ectoderm. The whole limb buds and dissociated mesoderm were homogenized in ice-cold 4M guanidine hydrochloride containing 0.05M Tris-HCl, pH 7.5 (50 p,l/limb), with a Potter homogenizer. The homogenates were diluted with 2 volumes of distilled water, fol-
Vol. 5, NO.5 September 1980
Pathogenesis of polydactyly of thumb
445
Fig. 2. Photomicrographs of epon-embedded sections of the hindlimb buds 24 hours after administration of ara-C show a marked decrease of the mesodermal tissue (A, relatively early stage; and B, relatively late stage) and deformity of the preaxial area (B, arrow) and control (C). (Toluidine blue stain.)
lowed by the addition of 9 volumes of 95% ethanol containing 1.3% potassium acetate. After standing at 0° C for 1 hour, the mixtures were centrifuged at 13,000 x g for 20 minutes. The precipitation with ethanol in the presence of potassium acetate was carried out three
additional times. DNA in the final pellet were determined by the method of Burton. 8 The hindlimb buds of the same number as the ara-C- treated and the control groups were similarly processed, and amounts of DNA were determined 24
446
Nogami and Oohira
The lournal of HAND SURGERY
Fig. 3. Electron micrographs of mesodermal cells in the preaxial area of the hindlimb bud 24 hours after administration of ara-C show sparse distribution of cells and cells with autophagic vacuoles (A) and control (B). (Uranyl acetate and lead citrate stain .)
hours after administration of DIP (75 mg/kg of body weight) on day 13.5 of gestation. This day was found to be the critical stage for producing oligodactyly in the hindlimb of rat fetuses at the highest rate (86%) by DIP.4 Administration of ara-C could also produce oligodactyly in the hindlimb , but the highest rate was approximately 40%. Therefore, ara-C was not employed to produce oligodactyly in this experiment. Results Skeletal malformations. Polydactyly of the first digit of the hindlimb was observed in 92% of fetuses on 21 st day of fetal age. Seventy-six percent of them were bilateral. They took various forms, from a vague suggestion of bifurcation of the distal phalanx to triplication of the first digit. Other digital malformations were not detected in the present experiment. Representative forms of polydactyly of the first digit are shown in Fig.!.
Light and electron microscopy. Thick epon sections examined by light microscope revealed marked decrease of mesodermal cells 24 hours after administration of ara-C (Fig. 2). Some hindlimb buds already showed slight deformity at the preaxial border (Fig. 2, B). Compared with mesodermal cells, ectodermal cells appeared to be intact, and the apical ectodermal ridge was well-maintained. Intercellular space was widened due to decreased number of mesenchymal cells in mesoderm and some mesenchymal cells showed degenerative change (Fig. 3, A) . The basal membrane between ectoderm and mesoderm was slightly wrinkled at the preaxial area where the first digit might be formed (Fig . 4, A). There were many ruthenium red staining granules palisading along the surface of the basal membrane and the plasma membrane in both treated (Fig . 5, A) and control (Fig . 5, B) limb buds. Growth retardation of the mesodermal tissue was quickly recovered and almost no difference was ob-
Vol. 5, No.5 September 1980
Pathogenesis of polydactyly of thumb
447
Fig. 4. Electron micrographs of the ectoderm -mesoderm interface in the preaxial area of the limb bud 24 hours after administration of ara-C show wrinkling of the basal membrane and a decrease in mesodermal cells (A) and control (8). (Uranyl acetate and lead citrate stain .)
served between ara-C treated and control limb buds after 48 hours except for a small bump or protrusion of the preaxial border of the treated hindlimb bud (Fig . 6) . A polydactylous feature was manifested externally on day 14 of fetal age (Fig. 7, A) . Quick recovery of mesodermal cells seems to be due to the fact that ara-C is cytotoxic only during the S phase of the cell cycle .9 Biochemical analysis. The amount of DNA in ectoderm and mesoderm of hindlimb buds 24 hours after administration of ara-C is shown (Table I) . There was no difference in DNA levels in ectoderm between the ara-C-treated and the control limb buds, whereas the DNA level decreased to 58.8% in mesoderm of the ara-C-treated limb buds as compared with the control. The amount of DNA in the ectoderm of the hindlimb buds was determined by subtracting the amount of mesodermal DNA from that of the whole limb bud DNA . The mesoderm and ectoderm of the hindlimb buds were similarly affected by administration of DIP. DNA
Table I. DNA content of the rat hind limb bud 24 hours after administration of cystosine arabinoside* DNA (mnole liimb bud) Ectoderm Dose (mg/kg of body weight) 100
0
Amount
0.04 0 .04
I
Mesoderm
%
Amount
100 100
1.90 3.25
I
% 58.8 100
'The data represent the average o f three determinations, with a standard deviation of less than 10%.
levels were reduced to 40,5% in ectoderm and 60 .8% in mesoderm, respectively (Table II) . Discussion It has been reported that the normal development of a limb bud depends on reciprocal influences between ectoderm and mesoderm in the embryonic stage. This
448
The Journal of HAND SURGERY
Nogami and Oohira
Fig. S. Electron micrographs show the basal membrane between ectoderm and mesoderm in the preaxial area of the hindlimb bud 24 hours after administration of ara-C (A) and control (B). There are many ruthenium red staining granules palisading along the basal membrane and the plasma membrane of both the treated and the control hindlimb bud, indicating normal function of cells which survived ara-C treatment (A). (Ruthenium red and uranyl acetate stain.)
Fore limb bud
Hind limb bud
Hind 1 imb bud (ara-C treated)
Fig. 6. Schema of external features of developing rat limb buds on days 11, 12, 13, and 14 of fetal age, showing the forelimb buds, the hindlimb buds, and the ara-C treated hindlimb buds. The preaxial border of the hindlimb bud is more flattened than that of the forelimb bud on days 11 and 12. Arrow shows a protrusion of the preaxial border which is an initial sign of polydactyly formation. (Magnifications of the limb buds on each day are not the same.)
Vol. 5, NO.5 September 1980
Pathogenesis of polydactyly of thumb
449
Fig. 7. External features of the hind foot plate on day 14 of fetal age show excessive growth of the first digit area (A) and control (8).
Table II. DNA content of the rat hind limb bud 24 hours after administration of 2,2' -dipyridyl* DNA (nmole/limb bud) Ectoderm
Dose (mg /kg of body weight)
Amount
75 0
0.85 2.10
I
Mesoderm
%
Amount
40.5 100
9.40 15.45
j
%
60.8 100
Legend: DIP, 2,2' = dipyridyl. 'The data represent the average of three determinations, with a standard deviation of less than 10%.
mechanism has been studied extensively in normal development, and there are many descriptive informations,10-12 but the process of organogenesis and teratogenesis has not been fully established. Continuous growth of ectoderm and retarded growth of mesoderm of the hindlimb bud resulted in marked growth discrepancy between these two tissues 24 hours after administration of ara-C in the present experiment. On the other hand, DNA synthesis was similarly inhibited in both ectoderm and mesoderm of the hindlimb buds 24 hours after administration of DIP. Therefore, it was postulated that relative overgrowth of ectoderm might be related to the production of polydactyly of the first digit and a lesion of both ectoderm and mesoderm of the limb bud might lead to formation of oligodactyly and/ or syndactyly. Administration of ara-C with increased dose decreased the rate of polydactyly forma-
Fig. 8. Diagrammatic representation of the hypothetical protrusion force, which induces polydactyly of the first digit. The limb bud in the early stage is represented as triangle and the protrusion forces in the preaxial border (8) and in the postaxial border (C) are shown as component forces of the proximodistal growing force (A) of the limb bud. The smaller the angle of the preaxial border (D) is, the stronger the protrusion force (8) becomes. Loosened ectoderm at the preaxial border due to growth discrepancy between ectoderm and mesoderm may help to make a protrusion.
tion in our preliminary experiment. It might be caused by cellular lesions in both the ectoderm and mesoderm. Krowke, Berg, and Merkefl reported that the affection in mesodermal cells was greater than in ectodermal ones when ara-C was administered and the different susceptibility of the cell types to teratogenic drugs might be an important factor for the induction of limb malformations. According to our experiments, different susceptibility of ectoderm and mesoderm may be due to the stage
450
The 10urnal of HAND SURGERY
Nogami and Oohira
difference of cellular differentiations of these two tissues, from ectodermal tissue to skin and from mesodermal tissue to cartilage. Mesodermal cells change their function of proteoglycan synthesis gradually from mesenchymal to cartilaginous cell type between days 12 and 14, while ectodermal cells begin to synthesize glycosaminoglycans of skin type on day 14.13 In general, susceptibility of cells is the highest at the stage just prior to onset of cellular differentiation. Therefore, a pronounced cellular lesion is caused in mesoderm of the hindlimb buds by treatment on day 11, whereas ectodermal cells are hardly affected, and in both ectoderm and mesoderm by treatment on day 13. Disappearance of the apical ectodermal ridge and formation of digital rays occur on days between 13 and 14. The reason why relative overgrowth of ectoderm induces polydactyly of the first digit in the hindlimb bud is not known, but we postulate the cephalocaudal asymmetry of the hindlimb bud in the early stage of morphogenesis as a cause of polydactyly. Morphologically, the preaxial border of the hindlimb bud (Fig. 6) is more flattened than that of the forelimb bud (Fig. 6) in rats and mice in the early stages of limb bud formation, and this asymmetry quickly diminishes between days 13 and 14. This indicates that the proximodistal morphogenetic activity becomes stronger preaxially than postaxially in the hindlimb bud during this period. Therefore, repairing activity of the mesodermal tissue 14 from the damage of ara-C administration may also be stronger in the preaxial area than in the postaxial area, and biomechanically it may playa role to produce a protrusion of the preaxial border (Figs. 6 and 7, A). This protrusion force may be regarded as a component force of the proximodistal growth activity of the underlying preaxial mesodermal tissue. The more striking the cephalocaudal asymmetry is, the stronger the protrusion force becomes as diagramatized in Fig. 8. It is reasonable to assume that the teratogen-induced or the genetically determined surplus ectodermal tissue helps to make a protrusion which provides additional direction of cellular proliferation to produce extra digit. It is interesting that the cephalocaudal asymmetry of the limb bud is more pronounced in the upper limb bud than in the lower limb bud in human embryos in which polydactyly of the thumb is predominant over that of the great toe. Pathogenesis of polydactyly produced in the axial or the postaxial region seems to be different from that of the thumb. Polydactyly of the axial area is often associated with reduction malformations, such as split hand deformity. Polydactyly of the ulnar side is often bilat-
eral, hereditary, 15 or associated with other abnormalities, such as chondroectodermal dysplasia 'H or short rib-polydactyly syndrome." The authors wish to express their appreciation to Miss Kuniko Ozeki for her assistance.
REFERENCES I. Chaube S, Kreis W, Uchida K, Murphy ML: The teratogenic effect of I-,B-D-arabinofuranosylcytosine. Biochern Pharmacol 17:1213-26, 1968 2. Scott W1, Ritter E1, Wilson 1G: Studies on induction of polydactyly in rats with cytosine arabinoside: Develop Bioi 45:103-11, 1975 3. Krowke R, Berg P, Merker H-1: Effects of cytosine arabinoside, 6-aminonicotinamide, and 6-mercaptopurine riboside on ectoderm and mesoderm of mouse limb buds. Teratology 15: 137-48, 1977 4. Oohira A, Nogami H: Limb anomalies produced by 2,2' -dipyridyl in rats. Teratology 18:63-70, 1978 5. Noback GL: The use of the Van Wijhe method for the staining of the cartilaginous skeleton: Anat Rec 11: 292-4, 1916 6. Groat RA: Clearing tissue with mixture of tributyl and tricresyl phosphates. Stain Tech 16: 111-7, 1941 7. Laros GS, Cooper RR: Electron microscopic visualization of protein polysaccharides. Clin Orthop 88: 179-92, 1972 8. Burton K: A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem 1 62:315-23, 1956 9. Cozzarelli NR: The mechanism of action of inhibitors of DNA synthesis. Ann Rev Biochem 46:641-68, 1977 10. Saunders 1W 1r: The interplay of morphogenetic factors, in Swinyard CA, editor: Limb development and deformity: problems of evaluation and rehabilitation. Springfield, 1969, Charles C Thomas, Publisher, pp 84-100 11. Zwilling E: Abnormal morphogenesis in limb development, in Swinyard CA, editor: Limb development and deformity: problems of evaluation and rehabilitation. Springfield, 1969, Charles C Thomas, Publisher pp 100-18 12. Stocum DL: Outgrowth and pattern formation during limb ontogeny and regeneration. Differentiation 3: 16782, 1975 13. Oohira A, Nogami H: In preparation 14. Nogami H: Digital malformations in the mouse foetus caused by X-radiation during pregnancy. 1 Embryol Exp Morphol 12:637-50, 1964 15. Bora FW: Congenital deformities, in Kilgore ES Jr, Graham WP, editors: The hand. Philadelphia, 1977, Lea & Febiger, Publishers, pp 379 16. Smith DW: Recognizable patterns of human malformation, ed 2. Philadelphia, 1976, WB Saunders Co, p 200 17. Spranger 1W, Langer LO, Wiedemann H-R: Bone dysplasia. Stuttgart, 1974, Gustav Fischer Verlag, p 53