Insulin improves survival but does not maintain function of cultured chick wing bud apical ectodermal ridge.код для вставкиСкачать
THE ANATOMICAL RECORD 233~467-477 (1992) Insulin Improves Survival but Does Not Maintain Function of Cultured Chick Wing Bud Apical Ectodermal Ridge EUGENIE L. BOUTIN AND JOHN F. FALLON Department of Anatomy, University of California-SF, San Francisco, California 94143-0452 (E.L.B.); Department of Anatomy, University of Wisconsin, Madison, Wisconsin 53706 (J.F.F.) ABSTRACT Previously we demonstrated that high levels of insulin (5 kg/ml) permit the survival of isolated chick apical ectodermal ridge in culture (Boutin and Fallon, Dev. Biol., 104:lll-116, 1984). Here we address whether lower levels of insulin or insulin-like growth factors (IGFs) can also improve the survival of cultured apical ectodermal ridge and whether ridge function is maintained along with ridge survival. Neither IGF I nor IGF I1 (100 ng/ml) decreased ridge cell death; however, cell death was significantly decreased with 50 ng/ml insulin. No further improvement was obtained in the presence of both IGF I and insulin. These data suggest that insulin improved the survival of the isolated apical ectodermal ridge by binding its own receptor. To test for the maintenance of function, stage 20 ridges were cultured for 0, 6, 12, 18, or 24 hr with or without insulin (5 cg/ml or 5 ng/ml) and used to make recombinant limbs. Isolated ridges cultured for 12 hr or more produced fewer outgrowths and these were rarely distally complete. The medium in which the ridges had been cultured did not influence ridge activity, despite the major differences in cell survival. Recombinants made with ridges cultured with limb mesoderm for 18 hr did not yield outgrowths as often as those with freshly isolated ridges, but most of the limbs that did form were distally complete. These results suggest that the decline in function of cultured, isolated apical ectodermal ridge was not due merely to ridge cell death but rather, at least in part, to its separation from limb mesoderm. o 1992 WiIey-Liss, Inc. The apical ectodermal ridge is a morphologically distinct region of the amniote limb ectoderm located a t the distal tip of the developing limb bud. Several experiments have demonstrated that reciprocal interactions between this specialized epithelium and the underlying limb mesoderm are essential for the proper development of the limb (Saunders, 1948; Zwilling, 1961; Fallon et al., 1986). Thus, in the developing chick, the apical ectodermal ridge is required for the survival of the distal wing mesoderm through stage 21 (Rowe et al., 1982) and for the specification of the limb elements through stage 29 (Saunders, 1948; Summerbell, 1974; Rowe and Fallon, 1982).Both the induction of the ridge in competent ectoderm (Kieny, 1960; Carrington and Fallon, 1984a) and the maintenance of its pseudostratified (Saunders, 1948) morphology (Saunders, 1949; Zwilling, 1956) are dependent on limb mesoderm. Furthermore, while normally there is a low level of cell necrosis in the ridge (Boutin and Fallon, 1984; Todt and Fallon, 1986),this becomes extensive if the ridge is cultured in the absence of limb mesoderm (Searls and Zwilling, 1964; Cairns, 1975; Boutin and Fallon, 1984). At least some of the interactions occurring between limb mesoderm and limb ectoderm appear to involve diffusible factors (Cairns, 1975; Jorquera et al., 19791, and the presence of a continuous basement membrane between these two tissues (Kaprio, 1977) suggests that most, if not all, exchanges may be mediated by diffusible or extracellular substances. 0 1992 WILEY-LISS, INC. The ability of the limb mesoderm to maintain the morphology and the activity of the apical ectodermal ridge has been attributed to limb maintenance factor (Zwilling and Hansborough, 1956). Its presence was postulated based on the phenotypes of limb mutants (Zwilling and Hansborough, 1956); however, certain in vitro data also support its existence. For example, Jorquera and coworkers (1979) reported that ridge ectoderm cultured for 24 hr in limb mesodermal extract remains viable and retains the ability to induce outgrowths, while Feinberg and Saunders (1979) found that isolated ridges cultured for more than 12 hr in serum-containing medium lacking mesodermal factors could not promote limb development. The latter authors, however, did not examine the fate of the ridge cells in culture, and since it has been reported that isolated limb ectoderm dies when cultured in serumsupplemented medium (Searls and Zwilling 1964; Boutin and Fallon 1984), the loss of ridge function in Feinberg and Saunders’ cultures may have been caused by cell death rather than the absence of mesodermal maintenance activity. We have demonstrated previously that the addition of high levels of insulin (5 pg/ml) to serum-containing medium will permit the survival of ridge ectoderm for Received October 18, 1991; accepted December 24, 1991. 468 E.L. BOUTIN AND J.F. FALLON a t least 24 h r in vitro (Boutin and Fallon, 1984). This observation raises the questions of whether insulin improves the survival of ridge cells by binding its own receptor or that of an insulin-like growth factor (IGF) and whether ridge function is maintained along with cell survival in cultured ridges. At supraphysiological levels (5 pg/ml), insulin can bind both insulin and insulin-like growth factor (IGF) receptors (for reviews, see Czech, 1982; Jacobs and Cuatrecasas, 1983), but at lower levels (50 ng/ml), insulin binds primarily to its own receptor (Massague and Czech, 1982). Although a functioning pancreas is not present at the early limb bud stages we examined (Benzo and Green, 1974), insulin is present in early chick embryos (De Pablo et al., 1982), and IGFs are synthesized by limb mesoderm (D’Ercole et al., 1980; Engstrom et al., 1987; Stylianopoulou et al., 1988; Romanus et al., 1988; Bondy et al., 1990) and are localized in the peripheral limb mesenchyme (Ralphs et al., 1990). Furthermore, receptors for insulin and IGF I have been detected in the limb buds of 4-day chick embryos (Bassas et al., 1985)and receptors for IGF I and I1 in the limb buds of embryonic mice (Bhaumick and Bala, 1987; Bondy et al., 1990). Hence, the apical ectodermal ridge may be exposed to these molecules during early limb development; any of which might influence ridge cell survival or function in ovo as well a s in vitro. Therefore, we have examined the ability of IGFs and lower levels of insulin (50 ng/ml) to permit ridge cell survival. Furthermore, we have tested the ability of isolated apical ectodermal ridge to induce limb outgrowths following culture in medium with or without added insulin to see if ridge cell survival is accompanied by ridge cell function. MATERIALS AND METHODS Ridge Isolation and Culture Conditions Fertile White Leghorn eggs were incubated a t 37”C, windowed, and staged according to the Hamburger and Hamilton (1951) series. Apical ectodermal ridges were isolated from stage 20 wing buds by cutting just proximal to the ridge and then separating ectoderm from mesoderm by a 3.5 min, 37°C incubation in 0.15% trypsin (Gibco; 1:300)/Simm’s balanced salt solution (BSS). The ectoderms were rinsed three times in the appropriate fetal bovine serum (Gibco) containing medium or in Tyrode’s solution containing 1% glucose (Sigma) and 1%bovine serum albumin (Sigma) and then either used immediately to make recombinant limbs or cultured. The ectoderms to be cultured were transferred to a tissue culture dish in drops of medium. The epithelia were straightened, oriented with their basal surfaces down, and gently pressed against the surface of the dish at their anterior and posterior ends. The anteroposterior axis of each ridge was marked on the bottom of the dish. The tissues were incubated for 2 h r at 37°C and 5% CO, in the drops of medium to allow the tissues time to attach to the surface of the dish before 2 ml medium was added. The ectoderms were cultured for a total of 6,12,18, or 24 hr. Normally, the cultures were checked 1 hr after the start of incubation, and any tissues that had detached from the dish were restraightened and pressed to the dish. Only those tissues that were still attached to the dish a t the end of the culture period were analyzed further. Normally tissues were cultured on uncoated tissue culture dishes. In other experiments, ridges were incubated in control or insulin-supplemented medium on plates coated with rat tail collagen, which had been extracted according to the procedure of Michalopoulos and Pitot (1975) and dried on the surface of the plate. These ectoderms were analyzed only for cell survival. Some tissues to be tested for ridge cell function were cultured on a layer of Matrigel (Collaborative Research), 0.1 ml per 35 mm tissue culture dish. Since fluids do not bead on this surface, tissues were transferred to it and excess medium withdrawn. The tissues were straightened but could not be pressed to the surface due to the flexibility of the gel. Nevertheless, the tissues remained flat for they were unable to curl up off the surface in the thin layer of fluid. After 2 hr, 2 ml medium was added. Control medium consisted of Ham’s F12X (Marzullo and Lash, 1970; prepared a s in Boutin and Fallon, 1984) plus 10% fetal bovine serum (Gibco). The amount of insulin in serum varies but normally does not exceed 0.6 ng/ml (data supplied by Gibco). Therefore, the control serum contained at most 0.06 ng/ml insulin. Supplements added to the control medium included insulin [5 Fg/ml (high-insulin cultures) or 50 ng/ml (lowinsulin cultures); Collaborative Research], multiplication stimulating factor (rat IGF 11, 100 ng/ml; Collaborative Research), and met-IGF I (100 ng/ml; Kabi Vitrum AB, Stockholm, Sweden). In addition, initially bacitracin (800 pglml; Sigma) was added to these media to inhibit the extracellular degradation of insulin (Roth, 1981), but because bacitracin can also modify the intracellular processing of insulin (Peavy et al., 1985), it was omitted from the later cultures. Most of the cultures analyzed for cell survival had bacitracin in the medium; however, bacitracin did not modify the amount of cell death, so these results are not reported separately. Fresh medium was added a t 6-9 hr intervals in the control and low-insulin- or IGF-supplemented cultures. Medium was not changed in the highinsulin cultures. To test the ability of mesodermal factors to maintain ridge survival and function, mesodermal extract was prepared according to the procedure of Jorquera et al. (1979). Stage 21 and 22 wing and leg buds were homogenized in Tyrode’s plus 1% glucose (one part limb buds to two parts Tyrode’s) by brief sonication. The homogenate was centrifuged a t 10,OOOg for 20 min. Enough x 100 antibiotic-antimycotic solution (Gibco) was added to the supernatant to make the extract x 1.5. Due to the small amount of extract obtained, ridges were cultured either individually in microtiter plates or en masse in a drop of extract in the center of a 35 mm plate surrounded by drops of Tyrode’s solution to prevent desiccation. As a n alternate way to expose the cultured ridges to diffusible mesodermal factors, ridges were cultured in high-insulin medium in the center of a 35 mm tissue culture dish. Distal mesoderm that had been separated from ectoderm by trypsinization was placed around the outer portion of the dish. Two rings were scratched on the surface of the plate to keep the cells of the two tissues separated. The tissues were incubated for 2 hr in separate, small amounts of medium to promote adhesion, prior to the addition of 2 ml medium, which was SURVIVAL AND FUNCTION OF CULTURED RIDGES STAGE 18-19 WING MESODERM FOLD ECTODERM t TO FORM RECOMBINANT LIMB GRAFT TO CHECK DEVELOPMENT SGMITES AT DAY II STRGE 22-23 BACK ECTODERM Fig. 1 . Diagram of the procedure used to make recombinant limbs. allowed to cross the rings. The tissues were incubated for a total of 18 hr, and the ridges were used to make recombinant limbs. In other experiments, ridges were not separated from distal limb mesoderm prior to culture. The ridge and distal mesoderm were pressed to the culture dish with the cut surface down and cultured for 18 hr in control or high-insulin medium. Following culture, the ectoderm was isolated from the mesoderm by an 8 min, 37°C incubation in 1%ethylenediaminetetraaceticacid (disodium salt, EDTA; Sigma) (Errick and Saunders, 1976) in double-strength calcium-magnesium free Tyrode’s solution (Kato, 1969). The isolated ectoderms were tested for functional activity. 469 was discarded. The remaining grafts were allowed to develop for a total of 8 days. At this time, the embryos were fixed in 10% formalin and stained for cartilage with Victoria blue. Some limb recombinants were prepared with quail (Coturnix coturnix japonica) back ectoderm rather than chick. These recombinants were fixed after 2 or 3 days in 70% ethanol, 40% formaldehyde, and glacial acetic acid in a ratio of 17:2:1. The buds were stained en bloc with the Feulgen reagent (Carrington and Fallon, 1984a) to detect the heterochromatin clumps characteristic of quail nuclei (LeDouarin and Barq, 1969), embedded in paraffin, and sectioned. In those recombinants allowed to develop for 8 days, ridge activity was rated according to the most distal limb element formed as 0 (nodules or small rods of cartilage), 1 (humerus or long bone), 2 (radius and/or ulna), or 3 (digital elements). Distal elements were never observed without proximal elements. Grafts that yielded any digital elements were considered distally complete, even if the elements themselves were incomplete. The average quality of the outgrowths was calculated by summing the ratings for each recombinant that produced an outgrowth and dividing by the number of outgrowths obtained (recombinants rated as 0 were not included in this calculation); an average of 3 would be obtained if all outgrowths were distally complete. Statistical differences were determined using a logistic regression analysis. A P value of 0.01 was used to determine significant differences. Recombinant Limbs Cell Death Analysis Following culture, the apical ridges were rinsed in Simm’s BSS and used to make recombinant limbs using a modification of the procedure of Errick and Saunders (1976). The procedure is shown in Figure 1. Mesodermal cores were isolated from right and left wing buds of stage 18,19, or 20 embryos (normally stage 18) by an 8 min, 37°C incubation in 1%EDTMCMF Tyrode’s. Stage 22-24 back ectoderm (normally stage 22), which overlay the neural tube and somites, was isolated from dissected embryos using the same procedure. [At these stages, body ectoderm is not competent to respond to the inductive signals of early limb mesoderm and will not form a ridge (Carrington and Fallon, 1984a)l.Both ectoderm and mesoderm were held on ice in 10% horse serum (Gibco) in Simm’s BSS until recombined. A piece of back ectoderm that had been trimmed to an appropriate size was placed flat on a tissue culture dish. A cultured or freshly isolated ridge was applied to approximately the center of the back ectoderm so that its basal surface faced up. A wing bud mesodermal core was placed on top of the back ectoderm such that the distal edge abutted the ridge along its anteroposterior axis. The back ectoderm was then folded over and its cut edges pinched together. The recombinant limb was allowed to heal a t room temperature (RT) for at least 30 min, excess ectoderm was trimmed away, and the bud was grafted to the somites of a stage 21 or 22 host. Two drops of x 5 penicillin-streptomycin solution (Gibco) were added at the time of grafting and 24 and 48 hr later. After healing for 30 min a t RT, the embryos were returned to the incubator. Any recombinant that had evidence of pooled blood or necrosis the next day Tissues to be analyzed for cell survival were fixed and embedded in Embed 812. The entire tissue was sectioned in 1 pm increments and the number of live and dead cells counted in every tenth section as previously described (Boutin and Fallon, 1984). These data were analyzed using a logistic regression analysis and the method of Williams (1982) to adjust for betweentissue variability. The mean percentage necrosis and the standard deviation for each group are reported in the tables. A P value of 0.01 was used to determine significant differences. RESULTS Cell Survival in Cultured Ridges General observations The isolated ridges were thin and elongate. An opaque region was visible along the anteroposterior axis. This area was assumed to be the apical ectoderma1 ridge whose greater cell density and height would reflect more light than the thinner adjacent dorsal and ventral limb ectoderms. No major change in the gross morphology of the ectoderms was detected after 6 hr in culture. At 12 hr, the tissues were still elongate and thin, but many of the tissues had begun to spread slightly. Spreading was more extensive after 18hr. The well-spread tissues remained elongate but were much wider than a t the earlier times. In addition, the opaque region was less well defined. In cross sections, spread tissues exhibited a thin layer of cells, two to four cells deep, flanking a thicker stratified epithelium. A peridermal layer was still evident over both regions. Spreading was even more extensive in the 24 hr cul- E.L. BOUTIN AND J.F. FALLON 470 TABLE 1. Influence of 50 ng/ml insulin and of a collagen-coated substratum on cell death in cultured apical ectodermal ridges Hours in 6 12 18 TC dish3 Collagen3 TC dish Collagen TC dish Collagen Control medium Insulin medium 5 2 2 (10) 6 2 3 (7) 35 f. 16 (19) 35 11(6) 42 2 8 (10) 28 2 12 (7) 5 t 2 (9) 5 t l(10) 13 2 9 (18) 9 2 6 (7) 17 t 6 (10) 15 t 4 (6) * ‘Data are expressed as the average percentage necrosis 2 the standard deviation. The numbers in parentheses are the number of ridges examined. 2The data for cell death at 0 hr were taken from Boutin and Fallon (1984). ‘TC dish, tissue culture plate; collagen, collagen-coated plate. tures that were not examined for cell survival but were tested for ridge cell function. Although no effort was made to quantitate the degree of spreading, in general, tissues grown on collagen-coated plates spread more than those grown on uncoated plates and ridges grown in insulin-supplemented media spread more than those in control cultures. Two of the six tissues grown on collagen in the presence of insulin showed extreme spreading. The tissues were very broad and were only a few cells thick. They showed little evidence of a thick, stratified ridge. The edges of these tissues were two to three cells thick, while the central region was only three to four cells deep. The cells in these tissues were squamous, in contrast to the cuboidal to columnar morphology normally seen in vivo. Control Medium The pattern of cell death obtained in the control cultures supplemented with 10% fetal bovine serum was similar to that reported in our earlier experiments (Boutin and Fallon, 1984). By 12 hr, the amount of cell death had increased significantly over that present in freshly isolated tissues (Table 1; P < 0.001) with a similar amount of necrosis evident at 18 h r (Fig. 2). Although the percentages of necrotic cells were somewhat lower than previously reported (6% +- 2% a t 6 hr, 47% 2 12% a t 12 hr, and 52% -t 11% a t 18 hr; Boutin and Fallon, 1984), the data were not significantly different (0.1 < P < 0.2). In addition, while the presence of a collagen substratum has been demonstrated to affect the function of other cell types (Hauschka and Konigsberg, 1966; Gospodarowicz et al., 19801, there was no statistically significant decrease in the amount of cell death in the presence of a collagen substratum (P > 0.51, despite the lower average detected a t 18 h r (Table 1).This average was skewed by the inclusion of a few tissues with minimal necrosis and very low total cell counts (less than one-half the average total cell count). It seems likely that in these cases ridge cells had died but had been eliminated from the tissue prior to harvesting, resulting in data that artificially lowered the average necrosis. Fig. 2. Section through a stage 20 ridge cultured for 18 hr in Ham’s F12X plus 10% fetal bovine serum. Numerous necrotic cells are present (arrowheads). x 600. Fig. 3. Section through a stage 20 ridge cultured for 18 hr in Ham’s F12X plus 10% fetal bovine serum and 50 ngiml insulin. A few dead cells are present (arrowheads), but most cells appear healthy. x 538. Insulin (50 ng/ml) and IGF I or I1 (100 ng/ml) After 12 h r of culture in insulin-containing media (50 ng/ml), the amount of cell death was increased over that of the 0 h r level (Table 1; P < 0.001; Fig. 3); however, the amount of cell death was also significantly less than that obtained in the control cultures (P < 0.001), indicating that insulin did enhance the survival of the apical ectodermal ridge a t this concentration. The survival of the ridge in the presence of 50 nglml of insulin was not, however, a s good as it had been in the presence of 5 pglml insulin (P < 0.001; average necrosis 2 standard deviation was 4% 2% a t 12 h r and 5% 2 2% at 18 hr; Boutin and Fallon, 1984). There was no further improvement in the survival of the ridge in the presence of a collagen substratum (P > 0.5). The addition of rat IGF I1 (100 ng/ml) or IGF I (100 ng/ml) to the control medium did not change the percentage of necrotic cells in cultured ridges (P > 0.5; Table 2). Furthermore, the presence of both 50 ng/ml insulin and 100 ng/ml IGF I did not improve the rate of cell survival over that found in the presence of insulin (50 ng/ml) alone (0.2 < P < 0.5). Additional information Although detailed cell counts were not tallied, representative examples of ridges cultured for 18 h r in * SURVIVAL AND FUNCTION O F CULTURED RIDGES TABLE 2. Influence of insulin-like growth factors on cell death in cultured apical ectodermal ridges Hours in culture 12 18 Supplements (%) IGF 111 IGF I1 IGF I + insulin' 36 t 10 (11)3 38 t 8 (19) 18 t 9 (11) nd 35 t 10 (10) 23 7 ( 7) * 'IGF I and I1 100 ng/ml. 'IGF I 100 ngiml, insulin 50 ngiml. 3Data are expressed as the average percentage necrosis 2 the standard deviation. The numbers in parentheses are the number of ridges examined. high-insulin medium on Matrigel or in mesodermal extracts were examined histologically. Ridges cultured in insulin medium on Matrigel remained healthy but lost their elongated morphology and did not spread. This appeared to arise from a contraction of the gel by the tissue and resulted in a more compact piece of ectoderm. Ridges cultured for 18 hr in limb mesodermal extract prepared according to Jorquera and coworkers (1979) tended to lose their attachment to the dish and round up. Cross sections of representative ridges exhibited massive necrosis (not shown). Ridge Function in Cultured Ridges Control experiments: Noncultured ridges Data from two groups of control experiments were collected. One group consisted of grafting isolated limb mesoderm in back ectoderm without an apical ectoderma1 ridge. The majority were made with stage 18 mesoderms, but four were constructed with stage 19 mesoderms and two with stage 20 mesoderms. None gave a positive outgrowth (Table 3). The most that was obtained was a small rod or nodule of cartilage covered by a shoulder girdle pattern of feathers (Fig. 4). The other control group included grafts of limb mesoderms and freshly isolated stage 20 ridges in back ectoderm. Fourteen operations survived the additional 8 days to day 11, and all produced outgrowths. The average level of outgrowth was 2.5. Eight outgrowths were distally complete, three of which contained at least portions of all limb elements, the humerus, radius, ulna, and digits 2, 3, and 4 (Fig. 5). To examine the fate of the grafted ridges, three recombinants were made with a freshly isolated, chick apical ridge and quail back ectoderm and fixed 3 days after grafting. Two limb buds (67%)had chick ectoderm distally and both of these demonstrated ridge morphology. Chick cells were always located at the center of the ridge, but quail cells often populated the edges of the ridge and could be found basally. In some sections, it appeared as if the quail cells were displacing the remaining chick cells. If no chick cells were present in a particular section, ridge morphology was not evident; the quail cells at the tip of the limb bud formed a simple, columnar epithelium capped by a squamous periderm. The third recombinant limb contained no chick ectodermal cells distally and lacked a ridge. A small amount of cell death was present in its distal mesoderm. Mesodermal cell death is consistent with a lack of ridge function (Rowe et al., 1982). 471 Ridges cultured without mesoderm General observations. Compared with the activity of freshly isolated ridges, cultured ridges demonstrated a significant decrease in ridge function with increased time in culture. This was reflected both in the number (P < 0.001) and the quality (P < 0.01) of outgrowths obtained. The medium in which the ridges had been cultured did not make a significant difference in the quality or quantity of the outgrowths obtained (P > 0.5) with two exceptions. Therefore, except where differences were found, the results for all media are combined in the text but are presented individually in Table 3. Six hour cultures. Seventeen recombinants were analyzed for limb morphogenesis; 13, or 76%,produced outgrowths, including seven distally complete limbs (Fig. 6). The average level of distal completeness was 2.2. While both the percentage of outgrowths and the quality of the outgrowths were lower than those obtained with freshly isolated ridges, neither was significantly different (0.05 < P < 0.1 and 0.1 < P < 0.25, respectively). Eight grafts were made with quail back ectoderm and chick ridges that had been cultured for 6 hr in control medium; five were fixed 2 days after grafting and three after 3 days. Five recombinants (62%) showed evidence of ridge morphology a t the apex of the sectioned wing bud. All five had chick cells in the ridge (Fig. 7). In fact, thick ridge was always associated with most or all of the ridge cells being of chick origin, suggesting, as was reported by Saunders et al. (1976), that ridge function was attributable to the grafted ridge and was not assumed by the flank ectoderm. When the distal ectoderm was populated with mainly quail cells, the ridge became thinner. The two limb buds fixed after 2 days that lacked ridge morphology had evidence of cell death in the distal mesoderm. This is consistent with a lack of ridge function (Rowe et al., 1982). The third bud lacking a ridge was fixed 3 days after grafting and did not have cell death in its distal mesoderm. Twelve hour cultures. The 25 surviving recombinants gave outgrowths in 56% of the cases, with an average quality of outgrowth of 1.6. These were significantly different from the 0 hr results (P < 0.005 and P < 0.005, respectively). Only one recombinant from the low-insulin group yielded a distally complete limb (Fig. 8). Many of the outgrowths that were rated as 2 consisted of two long bones fused in tandem, but this was not seen in grafts constructed with ridges that had been cultured for 12 hr in the high-insulin medium. Eighteen hour cultures. As was mentioned above, after 18 hr in culture, the epithelial sheets had spread considerably and the thickening of the ridge was less well defined. Ridges cultured in control, low-insulin, or high-insulin medium without bacitracin gave outgrowths at a lower rate than the 12 hr cultures (7121 or 33%,P < 0.001), but the average level of completeness of those that did develop, 1.7, was essentially the same (P > 0.5). The three outgrowths rated as 2 contained two long bones fused in tandem. Only one graft that had received a ridge cultured in control medium produced digital elements (Fig. 9). Twenty-one recombinants received ridges which had been cultured in medium containing bacitracin; none of E.L. BOUTIN AND J.F. FALLON 472 TABLE 3. Outgrowths induced by ridges cultured in the absence of mesoderm Hours in culture 01 6 12 18 24 Medium With ridge No ridge Control Control Low insulin High insulin Control Control Low insulin High insulin Control Control Low insulin High insulin High insulin Insulin + matrigel Bacitracin 0 + 31 1 Control Low insulin High insulin High insulin - __ 2 1 1 2 5 3 8 6 - 4 + + - + + + - + - Level of outgrowth 1 2 1 5 2 1 1 1 5 2 1 2F2 2F 2 1 1 2F 1 1F 3 8 2 1 2 2 1 1 13 4 7 22 7 8 3 2F 'Two control groups a t 0 hr consisted of recombinants made with freshly isolated ridges or recombinants made without ridges. 'F indicates that the long bones were fused in tandem. these gave outgrowths (P < 0.001). This group included eight ridges that had been cultured in control medium and 13 that had been cultured in high-insulin medium. Prior to this time, bacitracin did not have any inhibitory effects on either the number or the quality of the outgrowths obtained (P > 0.5). Of seven ridges cultured for 18 hr in high-insulin medium on the basement membrane substratum Matrigel, none produced any evidence of limb growth. This was in contrast t o other systems in which Matrigel enhances function (Hadley et al., 1985; Kleinman et al., 1986).The decrease in inductive activity may have been due to a direct inhibitory effect of Matrigel or to the shorter area that the limb mesoderm was in contact with the limb ectoderm since the ridges contracted on the gel. Alternately, since the ectoderm and gel were separated only by mechanical means, an intervening layer of Matrigel may have remained and inhibited the transfer of inductive influences between the ridge and the limb mesoderm. A layer of nonridge ectoderm delays the response of the mesoderm to the ridge (Murphy et al., 1983); a nonliving barrier might also be inhibitory. In that we did not investigate the action of Matrigel further, we could not distinguish between these possibilities. Twenty-four hour cultures. As in the 18 hr cultures, eight ridges cultured in the presence of bacitracin and high insulin were unable to induce outgrowths. No outgrowths were obtained from 29 recombinants made with ridges incubated in the control or low-insulin medium lacking bacitracin, while ridges cultured in highinsulin medium gave evidence of ridge function in two of five cases (P < 0.001). Both positive high-insulin recombinants were rated as two fused long bones (Fig. 10). Ridges cultured with limb extract or limb mesoderm Of 11 grafts made with ridges cultured in limb mesodermal extracts prepared according to Jorquera and coworkers (19791, only one produced an outgrowth (Table 4).This graft yielded two long bones, the others small rods or nodules of cartilage. As in the cultures without mesoderm, isolated ridges cultured for 18hr in high-insulin medium with isolated pieces of limb mesoderm spread on the surface of the dish. When used in recombinants, they did not exhibit increased function over ridges cultured without mesoderm (11%vs. 33%). Only one outgrowth was obtained from nine grafts. A long bone developed in this graft. When intact distal tips were cultured, both the mesoderm and the ectoderm exhibited some spreading; however, this was minimal compared with that observed with isolated ridge ectoderm. In addition, a distinct ridge capped the mesodermal mass. Eighteen recombinants were made with ridges that had been cultured while in contact with limb mesoderm. Outgrowths were obtained in 56% of the grafts. While the percentage of outgrowths was lower (P < 0.005), the average level of completeness, 2.1, was comparable to that obtained with freshly isolated ridges (P > 0.5). Four of the limbs were distally complete (Fig. 11). Ridges in control and high-insulin medium yielded comparable results (P > 0.5). DISCUSSION These studies addressed two questions regarding the fate of the apical ectodermal ridge in culture. The first was since high, nonphysiological levels of insulin (5 pg/ml) increased the survival of cultured apical ectodermal ridge (Boutin and Fallon, 1984), could lower levels of insulin or IGFs produce the same effect. We SURVIVAL AND FUNCTION OF CULTURED RIDGES 473 Fig. 4. Outgrowth obtained from a control recombinant limb lacking an apical ridge. A nodule of cartilage (arrow) covered by a shoulder pattern of feathers has developed from the stage 18 wing mesoderm. Fig. 6. Outgrowth obtained from a stage 19 wing mesoderm and a stage 20 ridge cultured for 6 hr in high-insulin medium. A humerus, radius, ulna, and digits 2 (shortened), 3, and 4 have developed. Fig. 5. Outgrowth obtained from a stage 18 (33 somite) wing mesoderm and a freshly isolated stage 20 ridge wrapped in stage 22 back ectoderm. A humerus, radius, ulna, and parts of digits 2,3, and 4 have developed. Fig. 7. Section through the tip of a 3-day-old recombinant limb made from chick ridge cultured for 6 hr in control medium and quail back ectoderm. The heterochromatin clumps characteristic of quail nuclei (arrows) are evident in the dorsal and ventral limb ectoderm; however, the apical ectodermal ridge is primarily of chick origin. x 550. have found that 50 ng/ml of insulin significantly decreased the amount of cell death in cultured, isolated apical ectodermal ridges but that 100 ng/ml of IGF I or IGF I1 was ineffective. These data suggest that insulin improves ridge cell survival by acting via its own receptor. The second set of experiments addressed whether conditions that maintain ridge cell survival also maintain the ability of the ridge to induce limb outgrowth. While apical ectodermal ridges cultured in association with limb mesoderm for 18 hr retained the ability to induce distally complete limbs, isolated ridges rapidly lost this ability irrespective of whether ridge viability was maintained. This suggests that more than ridge cell survival is required to maintain full ridge function. The fact that 50 nglml insulin improves ridge cell survival suggests that the apical ectodermal ridge contains insulin receptors. Although the embryonic chick does not appear to possess distinct receptors for IGF I1 (Bassas et al., 1987, 19881, receptors for insulin and IGF I are present in the early chick limb bud (Bassas et al., 1987). These have not been localized to ectodermal or mesodermal limb components, but interestingly, in prelimb bud stage embryos, insulin and IGF receptors are richest in ectodermal specializations (Girbau et al., 1989). Since insulin improves the survival of cultured apical ectodermal ridge, this raises the possibility that insulin plays a vital role in limb development in ovo. Insulin is present in the chick embryo from day 2 of development (De Pablo et al., 1982), prior to the development of the pancreas (Benzo and Green, 1976; Serran0 et al., 1989), when it may be obtained from the yolk (De Pablo et al., 1982) or synthesized at extrapancreatic sites (Serrano et al., 1989). Thus insulin is accessible to ridge cells in ovo. Limb mesoderm produces IGF I1 (Engstrom et al., 1987; Stylianopoulou et al., 1988; Romanus et al., 1988; Bondy et al., 1990) and IGF I (D’Ercoleet al., 1980), and these are localized to the peripheral limb bud mesenchyme (Ralphs et al., 1990); however, neither IGF alone nor IGF I in concert with insulin improved ridge cell survival in culture. We do not know whether insulin increased ridge cell survival through a specific influence on the differentiated state of the ridge cells or through a general metabolic effect such as an increase in glucose or amino acid uptake (Strauss, 1984). Antibodies to insulin (De Pablo et al., 1985) o r the insulin receptor (Girbau et al., 474 E.L. BOUTIN AND J.F. FALLON Fig. 8. Outgrowth obtained from a stage 18 (34 somite) wing mesoderm and a stage 20 ridge cultured for 12 hr in low-insulin medium. A humerus, radius, ulna, and parts of digits 2 and 3 have developed. Fig. 9. Outgrowth obtained from a stage 18 (33 somite) wing mesoderm and a stage 20 ridge cultured for 18 hr in control medium. A humerus, radius, ulna, and part of an unidentifiable digit have developed. Fig. 10. Outgrowth obtained from a stage 18 (33 somite) wing mesoderm and a stage 20 ridge cultured for 24 hr in high-insulin medium. Two long bones have developed. Two small fleshly outgrowths (arrowheads) also are evident. Fig. 11. Outgrowth obtained from a stage 18 (33 somite) wing mesoderm and a stage 20 ridge cultured for 18 hr in contact with limb mesoderm in high-insulin medium. A humerus, ulna, and digits 2 , 3 , and 4 have developed. 1984), a process believed to mimic that occurring during limb development. While insulin has profound effects on the metabolic activity of the regenerating Level of outmowth blastema (Vethamany-Globus et al., 19841,it is unclear Type of mesoderm 0 1 2 3 whether any of this is due to effects mediated via the wound epidermis, a structure analogous to the apical 10 1 Mesodermal extract Mesodermal pieces 8 1 ectodermal ridge. Tips in control medium 4 2 2 2 Despite the fact that insulin improves the survival of 4 1 1 2 cultured apical ectodermal ridge, insulin did not sigTips in high-insulin medium nificantly improve the ability of cultured ridges to induce distally complete limbs. The function of stage 20 1988) significantly retard the development of 2 day apical ectodermal ridges declined with increasing time chick embryos, indicating that insulin may have a gen- in culture irrespective of whether the medium coneral growth promoting activity very early in develop- tained insulin. This indicates that more than the mainment. However, in other systems, insulin appears to tenance of viability was necessary for ridge function have specific effects on differentiation (Chapman et al., and that neither insulin nor the IGFs was sufficient to 1984;Wolinsky et al., 19851, at least some of which are maintain full ridge function in vitro. independent of its growth promoting activity (Nicholas It is unlikely that the spreading of the ridge ectoet al., 1983).It is interesting to note that serum insulin derms in culture contributed to the decline in ridge levels rise during the initial stage of limb regeneration function and that limb mesoderm was able to prevent in Xenopus laevis (Liversage et al., 1987) and that in- this decrease when intact distal tips were cultured sulin appears to be an essential component for limb purely by acting in a structural manner to maintain regeneration (Vethamany-Globus and Liversage, 1973; compact ridge morphology. Decreased ridge function Vethamany-Globus, 1987; Vethamany-Globus et al., was evident prior to any major spreading of the culTABLE 4. Outgrowths induced by ridges cultured in the presence of mesodermal factors for 18 hr SURVIVAL AND FUNCTION OF CULTURED RIDGES tured ridges, and there is evidence that neither the flattening (Rubin and Saunders, 1972) nor the gross rearrangement (Errick and Saunders, 1976) of ridge cells correlates with its potential to induce limb outgrowths. Furthermore, the decline in ridge function probably was not related merely to normal aging of the ridge during culture. A stage 20 ridge in situ continues to induce outgrowth for an additional 3 days (Summerbell, 1974) and has the capacity to maintain this function even longer when combined with competent mesoderm (Rubin and Saunders, 1972).Thus, if the cultured ridges merely had aged on schedule, ridges cultured for 24 hr still should have been able to induce digital elements (Summerbell, 1974; Rubin and Saunders, 1972). However, digital elements rarely developed in grafts made with ridges cultured for more than 6 hr, indicating that the cultured ridges had not maintained full functional capacity. Ridges cultured in association with distal limb mesoderm were able to induce the same level of limb outgrowth as freshly isolated ridges, although the incidence of outgrowths was lower. Hence, limb mesoderm appeared to prevent or a t least slow the loss of ridge function, supporting the concept of a limb mesoderm maintenance factor. While experiments performed with wingless and polydactylous mutants support the existence of mesodermal factors which maintain ridge function (Zwilling, 1956; Zwilling and Hansborough, 1956; Carrington and Fallon, 1984b), the in vitro evidence for mesodermal maintenance activity is more ambiguous. Jorquera and coworkers (1979) reported that apical ridges cultured for 24 hr in limb mesodermal extracts remained viable and capable of inducing distally complete limbs. In contrast, Feinberg and Saunders (1979) found that isolated apical ectodermal ridges cultured in serum-containing medium for more than 12 hr lost ridge function. However, since under these culture conditions isolated apical ectodermal ridges die (Searls and Zwilling, 1965; Boutin and Fallon, 1984; this report), it is unclear if the loss of ridge function in Feinberg and Saunders’ experiments (1979) was due t o the absence of limb factors or to ridge cell death. Furthermore, we were unable to repeat the results of Jorquera et al. (1979); limb extracts did not maintain the viability, morphology, or function of cultured ridges. We do not know why we obtained such disparate results. Nevertheless, our data demonstrate that ridge function can be diminished in the absence of ridge cell death and that limb mesoderm supports the maintenance of ridge function. The fact that bacitracin led to a precocious loss of ridge function is intriguing. Ridges cultured in highinsulin medium containing bacitracin are viable at 18 and 24 hr (data not shown), so the loss of function was not due to accelerated cell death. Bacitracin inhibits some degradative enzymes (Roth, 1981; Peavy et al., 1985),perhaps some of these are essential for the maintenance of ridge function. If it could be determined how bacitracin accelerates the loss of ridge function, valuable insight might be gained on how function is normally maintained. While it is not known what molecules or factors are required in ovo for maintaining the function of limb 475 ectoderm, our data demonstrate that ridge cell survival did not guarantee the retention of full ridge function, S O , a t least under these conditions, ridge cell survival and ridge cell function could be uncoupled. Obviously, function can be lost in a nonviable tissue even if functional maintenance activity is present. Thus it will be important t o distinguish between factors that regulate function and those that influence cell survival, if we are to understand what is required for maintaining full ridge function. A number of growth factors are produced or at least present in limb mesoderm (Bell, 1986; Goedert, 1986; Engstrom et al., 1987; Heine et al., 1987; Seed et al., 1988; McLachlan et al., 1988; Munaim et al., 1988; Joseph-Silverstein et al., 1989). It remains to be addressed whether these growth factors are instrumental in maintaining the ability of the apical ectodermal ridge to induce distally complete limb outgrowth. Since we have established conditions that permit ridge cell survival, it is now possible to test whether the addition of other growth factors will permit the retention of ridge cell function in the absence of limb mesoderm. This would permit a more biochemical approach to be taken in investigating the complex interactions required for limb development. ACKNOWLEDGMENTS We thank Diane Widegran a t KabiVitrum AB for the generous gift of met-IGF I. We appreciate the helpful discussions with Drs. Robert DeHaan, Arnold Ruoho, and Michael Shannahan. We thank Kay Simandl and Tamara Bucci for technical assistance. This work was funded by NIH grant T32HD07118, NSF grant PCM8406338, and NIH grant POlHD20743. LITERATURE CITED Bassas, L., F. De Pablo, M.A. Lesniak, and J . Roth 1985 Ontogeny of receptors for insulin-like peptides in chick embryo tissues: Early dominance of insulin-like growth factor over insulin receptors in brain. Endocrinology, 117:2321-2329. Bassas, L., M.A. Lesniak, M. Girbau, and F. De Pablo 1987 Insulinrelated receptors in the early chick embryo: From tissue patterns to possible function. J. Exp. Zool., Suppl. 1 :299-307. Bassas, L., M.A. Lesniak, J . Serrano, J . Roth, and F. De Pablo 1988 Developmental regulation of insulin and type I insulin-like growth factor receptors and absence of type 11 receptors in chicken embryo tissues. Diabetes, 37:637-644. Bell, K.M. 1986 The preliminary characterization of mitogens secreted by embryonic chick wing bud tissues in vitro. J . Embryol. Exp. Morphol., 93:257-265. Benzo, C.A., and T.D. Green 1974 Functional differentiation of the chick endocrine pancreas: insulin storage and secretion. Anat. Rec., 180:491-496. Bhaumick, B., and R.M. Bala 1987 Receptors for insulin-like growth factors I and I1 in developing embryonic mouse limb bud. Biochim. Biophys. Acta 927:117-128. Bondy, C.A., H. Werner, C.T. Roberts Jr., and D. LeRoith 1990 Cellular pattern of insulin-like growth factor-I (IGF I) and type I IGF receptor gene expression in early organogenesis: Comparison with IGF-I1 gene expression. Mol. Endocrinology, 90t1386-1398. Boutin, E.L., and J.F. Fallon 1984 An analysis of the fate of the chick wing bud apical ectodermal ridge in culture. Dev. Biol., 104t111116. Cairns, J.M. 1975 The function of the apical ectodermal ridge and distinctive characteristics of adjacent distal mesoderm in the avian wing-bud. J . Embryol. Exp. Morphol., 34:155-169. Carrington, J.L., and J.F. Fallon 1984a The stages of flank ectoderm capable of responding to ridge induction in the chick embryo. J. Embryol. Exp. Morphol., 84t19-34. Carrington, J.L., and J.F. Fallon 1984b Evidence that the ectoderm is the affected germ layer in the wznggless mutant chick embryo. J. Exp. Zool., 232:297-308. 476 E.L. BOUTIN AND J.F. FALLON Chapman, A.B., D.M. Knight, B.S. Dieckmann, and G.M. Ringold 1984 Analysis of gene expression during differentiation of adipogenic cells in culture and hormonal control of the developmental program. J . Biol. Chem., 259:15548-15555. Czech, M.P. 1982 Structural and functional homologies in the receptors for insulin and the insulin-like growth factors. Cell, 31 53-10. De Pablo, F., M. Gibrau, J.A. Gomez, E. Hernandez, and J. Roth 1985 Insulin antibodies retard and insulin accelerates growth and differentiation in early embryos. Diabetes, 34:1063-1067. De Pablo, F., J . Roth, E. Hernandez, and R.M. Pruss 1982 Insulin is present in chicken eggs and early chick embryos. Endocrinology, I 1 1:1909-1916. DErcole, A.J., G.T. Applewhite, and L.E. Underwood 1980 Evidence that somatomedin is synthesized by multiple tissues in the fetus. Dev. Biol., 75:315-328. Engstrom, W., K.M. Bell, and P.N. Schofield 1987 Expression of the insulin like growth factor I1 gene in the developing chick limb. Cell Biol. Int. Rep., 11t415-421. Errick, J.E., and J.W. Saunders Jr. 1976 Limb outgrowth in the chick embryo induced by dissociated and reaggregated cells of the apical ectodermal ridge. Dev. Biol., 5Ot26-34. Fallon, J.F., E.L. Boutin, and J.L. Carrington 1986 Apical ectodermal ridge maintenance. In: Cellular Endocrinology: Hormonal Control of Embryonic and Cellular Differentiation. G. Serrero and J . Hayashi, eds. Alan R. Liss, New York, pp. 103-114. Feinberg, R.N., and J.W. Saunders J r . 1979 Morphogenetic competence of chick apical ectodermal ridge after isolation in culture. J . Cell Biol., 83t396a. Girbau. M.. L. Bassas. J. Alemany, and F. De Pablo 1989 In situ autoradiography and ligand-dependent tyrosine kinase activity reveal insulin receptors and insulin-like growth factor I receptors in prepancreatic chicken embryos. Proc. Natl. Acad. Sci. USA, 86.5868-5872. Girbau, M., M.A. Lesniak, J . Serrano, J.A. Gomez, and F. De Pablo 1988 Insulin action in early embryonic life: Anti-insulin receptor antibodies retard chicken embryo growth but not muscle differentiation in vivo. Biochem. Biophys. Res. Commun., 153t142148. Goedert, M. 1986 Molecular cloning of the chicken nerve growth factor gene: mRNA distribution in developing and adult tissues. Biochem. Biophys. Res. Commun., 141:1116-1122. Gospodarowicz, D., D. Delgado, and I. Vlodavsky 1980 Permissive effect of the extracellular matrix on cell proliferation in vitro. Proc. Natl. Acad. Sci. U.S.A., 77:4094-4098. Hadley, M.A., S.W. Byers, C.A. Suarez-Quian, H.K. Kleinman, and M. Dym 1985 Extracellular matrix regulates Sertoli cell differentiation, testicular cord formation, and germ cell development in vitro. J . Cell Biol., 101:1511-1522. Hamburger, V., and H.L. Hamilton 1951 A series of normal stages in the development of the chick embryo. J . Morphol., 88t49-92. Hauschka, S.D., and I.R. Konigsberg 1966 The influence of collagen on the development of muscle clones. Proc. Natl. Acad. Sci. USA, 55t119-126. Heine, U.I., E.F. Munoz, K.C. Flanders, L.R. Ellingsworth, H.-Y. Lam, N.L. Thompson, A.B. Roberts, and M.B. Sporn 1987 Role of transforming growth factor+ in the development of the mouse embryo. J . Cell Biol., 105t2861-2876. Jacobs, S., and P. Cuatrecasas 1983 Insulin receptors. Annu. Rev. Pharmacol. Toxicol., 23,461-479. Jorquera, B., E. Molinari, 0. Goicoechea, and 0. Garrido 1979 Involucion in vitro, regeneracion y actividad funcional de la zona de la cresta apical ectodermica del esbozo del miembro de pollo. Z. Vetinarmed. C, 8.262-276. Joseph-Silverstein, J., S.A. Consigli, K.M. Lyser, and C. Ver Pault 1989 Basic fibroblast growth factor in the chick embryo: Immunolocalization to striated muscle cells and their precursors. J. Cell Biol., 108:2459-66. Kaprio, E.A. 1977 Ectodermal-mesenchymal interspace during the formation of the chick leg bud. Rouxs Arch. Dev. Biol., 182t213225. Kato, Y. 1969 Epithelial metaplasia induced on extraembryonic membranes. I. Induction of epidermis from chick chorionic epithelium. J . Exp. Zool., 17Ot229-252. Kieny, M. 1960 Role inducteur du mesoderme dans la differenciation precoce du bourgeon de membre chez l’embryon de Poulet. J . Embryol. Exp. Morphol., 8t457-467. Kleinman, H.K., M.L. McGarvey, J.R. Hassell, V.L. Star, F.B. Cannon, G.W. Laurie, and G.R. Martin 1986 Reconstituted basement membrane: A defined supramolecular complex with biological activity. Biochemistry, 25:312-318. LeDouarin, N., and G. Barq 1969 Sur l’utilisation des cellules de la caille japonaise comme “marqueurs biologiques” en embryologie experimentale. C.R. Acad. Sci. Paris [DI, 269:1543-1546. Liversage, R.A., R.A. Foty, and P.D.G. Miles 1987 Serum immunoreactive insulin levels in intact and regenerating postmetamorphic Xenopus laeuis. J . Exp. Zool., 243:453-460. Marzullo, G.,a nd J.W. Lash 1970 Control of phenotypic expression in cultured chondrocytes: Investigations on the mechanisms. Dev. Biol., 22:638-654. Massague, J., and M.P. Czech 1982 The subunit structures of two distinct receptors for insulin-like growth factors I and I1 and their relationship t o the insulin receptor. J . Biol. Chem., 257:50385045. McLachlan, J.C., J . Macintyre, D.D. Hume, and J . Smith 1988 Direct demonstration of production of transforming growth factor activity by embryonic chick tissue. Experientia, 44:351-352. Michalopoulos, G., and H.C Pitot 1975 Primary culture of parenchymal liver cells on collagen membranes. Morphological and biochemical observations. Exp. Cell Res., 94:70-78. Munaim, S.I., M. Klagsbrum, and B.P. Toole 1988 Developmental changes in fibroblast growth factor in the chicken embryo limb bud. Proc. Natl. Acad. Sci. U.S.A., 85:8091-8093. Murphy, M.J., M.T. Gasseling, and J.W. Saunders Jr. 1983 Formation of a functional apical ectodermal ridge from nonridge ectoderm in limb buds of chick and quail embryos. J . Exp. Zool., 226t221-230. Nicholas, K.R., L. Sankaran, and Y.J. Topper 1983 A unique and essential role for insulin in the phenotypic expression of rat mammary epithelial cells unrelated to its function in cell maintenance. Biochim. Biophys. Acta 763t309-314. Peavy, D.E., F.G. Hamel, V.L. Kincke, and W.C. Duckworth 1985 Evidence that bacitracin alters intracellular insulin metabolism in isolated rat hepatocytes. Diabetes, 34217-221. Ralphs, J.R., L. Wylie, and D.J. Hill 1990 Distribution of insulin-like growth factor peptides in the developing chick embryo. Development, 109:51-58. Romanus, J.A., Y.W.-H. Yang, S.O. Adams, A.N. Sofair, L.Y.-H. Tseng, S.P. Nissley, and M.M. Rechler 1988 Synthesis of insulinlike growth factor I1 (IGF-11) in fetal rat tissues: Translation of IGF-I1 ribonucleic acid and processing of pre-pro-IGF-11. Endocrinology, 122:709-716. Roth, R.A. 1981 Bacitracin: An inhibitor of the insulin degrading activity of glutathione insulin transhydrogenase. Biochem. Biophys. Res. Commun., 98:431-438. Rowe, D.A., J.M. Cairns, and J.F. Fallon 1982 Spatial and temporal patterns of cell death in limb bud mesoderm after apical ectoderma1 ridge removal. Dev. Biol., 93:83-91. Rowe, D.A., and J.F. Fallon 1982 The proximodistal determination of skeletal parts in the developing chick leg. J . Embryol. Exp. Morphol., 68:l-7. Rubin, L., and J.W. Saunders Jr. 1972 Ectodermal-mesodermal interactions in the growth of limb buds in the chick embryo: Constancy and temporal limits of the ectodermal induction. Dev. Biol., 28: 94-112. Saunders. J.W.. J r . 1948 The aroximo-distal seauences of oriein of the parts of the chick wing and the role of the e‘ctoderm. J . E>p. Zool., 108t363-404. Saunders, J.W.,Jr. 1949 An analysis of the role of the apical ridge of ectoderm in the development of the limb bud in the chick. Anat. Rec., 105t567-568. Saunders, J.W., Jr., M.T. Gasseling, and J.E. Errick 1976 Inductive activity and enduring cellular constitution of a supernumerary apical ectodermal ridge grafted to the limb bud of the chick embryo. Dev. Biol., 50:16-25. Searls, R.L., and E. Zwilling 1964 Regeneration of the apical ectoderma1 ridge of the chick limb bud. Dev. Biol., 9t38-55. Seed, J., B.B. Olwin and S.D. Hauschka 1988 Fibroblast growth factor levels in the whole embryo and limb bud during chick development. Dev. Biol., 128:50-57. Serrano, J., C.L. Bevins, S.W. Young, and F. De Pablo 1989 Insulin gene expression in chicken ontogeny: Pancreatic, extrapancreatic, and prepancreatic. Dev. Biol., 132r410-418. Straws, D.S. 1984 Growth-stimulating actions of insulin in vitro and in vivo. Endocrine Rev., 5:356-369. Stylianopoulou, F., A. Efstratiadis, J. Herbert, and J . Pintar 1988 Pattern of the insulin-like growth factor I1 gene expression during rat embryogenesis. Development, 103t497-506. Summerbell, D. 1974 A quantitative analysis of the excision of the AER from the chick limb bud. J. Embryol. Exp. Morphol., 32: 651-660. Todt, W.L., and J.F. Fallon 1986 Develoament of the auical ectoderma1 ridge in the chick leg bud and comparison Gith the wing bud. Anat. Rec., 215:288-304. ~ ~ ~~ a SURVIVAL AND FUNCTION OF CULTURED RIDGES Vethamany-Globus, S. 1987 Hormone action in newt limb regeneration: Insulin and endorphins. Biochem. Cell Biol., 65t730-738. Vethamany-Globus, S.,M. Globus, A. Darch, G. Milton, and B.L. Tomlinson 1987 In vitro effects of insulin on macromolecular events in newt limb regeneration blastemata. J . Exp. Zool., 231t63-67. Vethamany-Globus, S., and R.A. Liversage 1973 Effects of insulin insufficiency on forelimb and tail regeneration in adult Diemictylus viridesens. J. Embryol. Exp. Morphol., 30~427-447. Williams, D.A. 1982 Extra-binomial variation in logistic linear models. Appl. Statist., 31t144-148. 477 Wolinsky, E.J., P.H. Patterson, and A.L. Willard 1985 Insulin promotes electrical coupling between cultured sympathetic neurons. J . Neurosci., 5:1675-1679. Zwilling, E. 1956 Interaction between limb bud ectoderm and mesoderm in the chick embryo. IV. Experiments with a wingless mutant. J . Exp. Zool., 132:241-254. Zwilling, E. 1961 Limb morphogenesis. Adv. Morphogen., lr301-330. Zwilling, E., and L.A. Hansborough 1956 Interaction between limb bud ectoderm and mesoderm in the chick embryo. 111. Experiments with polydactylous limbs. J. Exp. Zool., 132t219-240.