Metamorphic shortening of the alimentary tract in anuran larvae (Rana catesbeiana).код для вставкиСкачать
THE ANATOMICAL RECORD 242:417-423 (1995) Metar iorphic Shortening of the Alimentary Tract in An1 Larvae ( R a m catesbeiana) ROBERT PRETTY, TOM10 NAITOH, AND RICHARD J. WASSERSUG Department of Anatomy and Neurobiology, Dalhousie University, Halifcu, Nova Scotia, Canada ABSTRACT Background: The premetamorphic alimentary tract in anurans can be more than 10 times a tadpole's body length but then dramatically shortens to a third or less of that length by the end of metamorphosis. Although there have been many studies on histological changes in the anuran gut with metamorphosis, the broader question of where the major shortening occurs has not been previously addressed. This topic is investigated here. Methods: We began our study by labeling intestinal coils in situ in preserved Rana catesbeiana tadpoles and then uncoiling their intestines, locating, and measuring the labeled points. This allowed us to map the coiled gut of the tadpole, such that the distance along the oral-anal axis could be determined by simply counting coils. We next implanted markers into the intestinal coils of live R. catesbeiana tadpoles at five known locations along the oral-anal axis, established from the prior mapping. The tadpoles were then induced to metamorphose by immersion in thyroid hormone. After the gut had shortened to a third of its premetamorphic length, the positions of the implanted markers were determined through dissection. Results: Relative distances between the marked points did not change when the gut shortened. Conclusions: The results indicate that during metamorphosis the intestine shortens uniformly along its length and not preferentially from one region or another. Although metamorphosis was artificially induced, the shortening that we observed matches that occurring during natural metamorphosis. o 1995 Wiley-fiss, Inc. Key words: Metamorphosis, Alimentary tract, Intestine, Anura, Rana catesbeiana The metamorphosis of a tadpole into a frog is one of nature's most spectacular events. Part of this spectacle includes profound changes in the alimentary tract. During metamorphosis there are both microscopic changes in the intestines at the histological level and macroscopic changes in its overall length. These histological changes are well documented. They occur in both the epithelial and extraepithelial layers (Dauca and Hourdry, 1985; Hourdry and Beaumont, 1985).During metamorphosis, the larval gut lining is degraded (Fox, 1984; Yoshizato, 1989). There is increased production of acid phosphatase by lysosomes along the alimentary tract (Kaltenbach et al., 1981)' which helps break down the premetamorphic epithelium. Epithelial cells are autolysed and engulfed by phagocytes. Finally, much of the degenerated tissue is sloughed off into the intestinal lumen. Simultaneously, the adult intestinal epithelium differentiates. Nests of undifferentiated precursor cells divide and produce a new epithelium (Marshall and Dixon, 1978).The connective tissue and musculature of 0 1995 WILEY-LISS. INC the gut wall also proliferate greatly (Ishizuya-Oka and Shimozawa, 1987). While all of these histological changes are taking place, the intestine shortens extensively. Before metamorphosis the intestine in common frogs, such as those of the genus Rana, can be 10 or more times the body length of the tadpole (Rugh, 1951; this study). Intestinal shortening begins before the tail is absorbed (Dauca and Hourdry, 1985), and by completion of metamorphosis the absolute length of the gut is usually less than a third its premetamorphic length. Some representative values for the percentage that the gut shortens at metamorphosis are: 58.15%in Rana temporaria, 82.2% in Rana pipiens (Janes, 1934); 90% in Alytes Received November 7, 1994; accepted January 31, 1995. Address reprint requests to Richard Wassersug, Department of Anatomy and Neurobiology, Sir Charles "upper Medical Building, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada. Tomio Naitoh is now at the Department of Biology, Shimane University, Matsue, 690, Japan. 418 R. PRETTY ET AL. Switchback Point ‘ G Fig. 1. Left: Schematic ventral view of the Rana catesbeiana intestine inside the body cavity. SB = switchback point; i.e., the midpoint of the intestinal mass, where the direction of coiling reverses itself. The portion of the gut oral to the switchback point is shaded and individual coils numbered, with respect to distance from the switchback point, using Roman numerals. The caudal ( = anal) portion of the gut, with respect to the switchback point, is unshaded and coils there are numbered with Arabic numerals. In this view, proximal coils i, iv, and v are hidden by the overlying caudal coils. Typically in R . catesbeiana, as one moves laterally from the switchback point, the first 2-4 coils observed in ventral view are from the caudal portion of the alimentary tract. L = liver, P = pancreas. Right: Schematic representation of the coiled gut showing the location of oral (light dashed line) and anal (dark continuous line) portions of the gut. Here the coils have been spread out and tipped en masse rostrally to help expose both superficial and deep coils. obstetricans (Dauca and Hourdry, 1985); and 84% in R . catesbeiana (Carver and Frieden, 1977). As these figures indicate, the shortening is both massive and well documented. Yet, surprisingly, it is not known where along the oral-anal axis most shortening actually occurs. That topic is addressed here. By way of background, the intestines in tadpoles form a tight double spiral, which fills most of the body cavity. Nodzenski et al. (1989) report that in one anuran species, the coiled intestines occupied 80% of the body cavity, leaving little space for other viscera. The center of the double coil is the switchback point (SB)-the point at which the spiral reverses itself. Hence, the anterior and posterior portions of the small intestine concurrently spiral outward from this midpoint (Delsol and Zervudacki, 1962). The typical arrangement of the coils in premetamorphic Rana catesbezana is shown in Figure 1. We reasoned t h a t we could determine where the greatest amount of gut shortening took place if we could mark sites along the oral to anal axis of the in- testine before metamorphosis and then locate these same sites after metamorphosis. To do this we first needed a map of the larval gut, which would allow us to determine the distance along the intestine of specific coils. We constructed the map in the first part of our study using R . catesbeiana larvae. In the second part, the same points were surgically labeled in live R. catesbeiana and their positions relative to each other determined once the gut had shortened. Hypothetically there are four basic topological ways in which the intestine could shorten at metamorphosis: (1)the gut could preferentially shorten from the oral end, (2) the gut could preferentially shorten from the anal end, (3) the gut could preferentially shorten from a more central region (i.e., the switchback point), or (4) the gut could shorten uniformly along its oral-anal axis. These possibilities, which are not mutually exclusive, are schematically outlined in Figure 2. If the gut shortens preferentially from one end, then, after metamorphosis the marked points should be found nearer to each other at that end. If the gut shortens predomi- 419 Oral End Switchback Point Anal End PREMETAMORPHIC POSITION OF GUT MARKERS 1 . Preferential shortening from the oral end 3. Preferential shortening from switchback point 2. Preferential shortening from the anal end 4. Uniform shortening along intestinal length POSSIBLE POSTMETAMORPHIC POSITIONS OF GUT MARKERS Fig. 2. Schematic drawing of the uncoiled intestine of an anuran tadpole showing five marked positions in the larval intestine (above) and four hypothetical ways in which the intestine could shorten with metamorphosis and displace the points (below).The switchback point is a permanent feature represented here by a node. The possible ways the gut could preferentially shorten are: (1)from the oral end, (2) from the anal end, (3) from the middle, or (4)uniformly along its total length. nately from the center, then, after metamorphosis the marked points should be found closer to each other in that region. Alternatively, if the gut shortens uniformly along its oral-anal axis, then, there will be no change apparent in the relative distribution of the marked points along the gut. The alimentary tract was removed en masse from the body cavity of each specimen by cutting the esophagus, terminal rectum, post caval, and dorsal mesenteries. The gut was then uncoiled by cutting away remaining mesenteries and major vessels. The total length of the alimentary tract and the distances from the oral end to each marked point were recorded. The percentage distance along the oral-anal axis for each of the five mapped points was calculated. MATERIALS AND METHODS Intestinal Mapping in Tadpoles Tadpoles for this portion of the study were all colSurgical Procedures lected in the vicinity of Halifax County, Nova Scotia. Live R . catesbeiana tadpoles were purchased from They were euthanized in MS-222 (Sigma, St. Louis, MO) and preserved in 10% neutral-buffered formalin. NASCO (Fort Atkinson, WI) and kept at 15°C in a reTen specimens with conspicuous hindlimbs, but no ex- frigerated aquarium to slow natural metamorphosis posed forelimbs (stages 37-41; Gosner, 1960) were dis- and maintained on a diet of parboiled lettuce. Ten of sected. Prior to dissection, each specimen’s develop- the tadpoles were implanted with intestinal markers; two were kept as controls. The control animals were mental stage and snout-vent length were recorded. The skin and musculature of the anterior abdominal not subjected to surgery but were otherwise treated the wall were cut away to reveal the coiled gut. Five points same as the specimens with the intestinal markers. All along the alimentary tract were marked with a loop of these specimens were closely matched to those used in 6 - 0 monofilament polypropylene suture (SurgileneB, the premetamorphic mapping study (above) in size and Davis & Geck, Baie d’Urf6, PQ) tightly knotted around metamorphic stage. The animals selected for surgery were anesthetized the gut. In each specimen the marked points were: the third and fifth anterior intestinal coils, the third and by immersion in a 0.15% solution of Benzocaine (Sigfifth posterior intestinal coils, plus the “central switch- ma) for -6.5 min (following Vanable, 1985). Surgery was performed under a dissecting microscope with fiber back” point (Fig. 1). 420 R. PRETTY ET AL. optic illumination. The tadpoles were placed on a bed of ice to slow metabolism during surgery. Throughout surgery the tadpoles were bathed in a cold tadpole Ringer's solution (NASCO), and a n antibiotic antimycotic solution (Sigma) of penicillin, streptomycin, and amphotericin was applied to minimize the possibility of infection. A longitudinal ventral incision was made slightly to the righi of the midline, to avoid cutting sagittal blood vessels. The incision measured no more than 25 mm in any specimen. Using a wet spatula, the intestinal coils were gently lifted out of the surgical wound and placed directly on the abdominal wall. Care was taken not to tear the mesentery. In preliminary experiments a variety of in vivo marking techniques was tried, such as histological dyeing and branding of points along the gut with a hot needle. None of these labels proved to be permanent enough. Ultimately we settled on inserting short lengths of nonabsorbable 4-0 monofilament polypropylene suture (SurgileneB)directly through the gut wall. Alternating black and blue sutures were used to distinguish between neighboring points. At each of the five previously mapped points, a length of suture was passed completely through the gut wall into the lumen of the alimentary tract and back out again. The ends of the sutures were flared with a hot spatula to hold them in place. After all five sutures were in place, the intestines were tucked back into the body cavity and the incision closed using 4-6 absorbable 6-0 sutures (Ethicone, Ethicon Suture, Peterborough, ON). The operation typically lasted between 35-60 min, depending on the size of the tadpole and the accessibility of the individual gut coils. After surgery, the tadpoles were kept individually in aerated aquarium water, with a small amount of tetracycline to avoid wound infection. Total behavioral recovery from anesthesia was observed within 2 hr. Metamorphic Induction Postoperatively the animals were kept at room temperature (c. 20-22°C). Initially we planned to let animals metamorphose spontaneously, but our R. catesbeiana tadpoles did not spontaneously metamorphose after such major intestinal surgery. Therefore metamorphosis was induced with the hormone thyroxine. Thyroxine is the endogenous biochemical promoter of metamorphosis, and thus it emulates natural metamorphosis of the gut at physiologically realistic dosages (Etkin, 1981; Shi and Brown, 1993). Thyroid induction was performed with two groups of six tadpoles, each consisting of five tadpoles with the implanted markers and one control tadpole. Generic SynthroidO (L-thyroxine sodium, Boots Pharmaceuticals, Etobicoke, ON) was dissolved in water to give concentrations of 7.5 x lo-' M and 1.5 x l o p 7 M. Postsurgical tadpoles were kept individually in 1.0 1 of the thyroid solution. The molarities were maintained by replacing 62.5 ml of the thyroid solutions daily. Once a week the containers were cleaned and new solutions were started. For the first group of six tadpoles we used a thyroxine concentration of 1.5x lop7 M and thyroxine treatment continued until metamorphic climax began (defined as the appearance of front limb bulges and/or shortening of the tail). However, because of high mortality among the metamorphosing froglets, the dosage was reduced in the second group of tadpoles. In that group we used a thyroxine concentration of 7.5 x lop8 M and stopped treatment when we saw the first signs of transformation (e.g., forelimbs about to emerge). Treatment periods were 15-18 days (1721 day; TikSE) and 11-16 days of treatment (13+2 days), respectively. When animals died, they were promptly preserved in neutral-buffered formalin. Postmetamorphic Observation of Implanted Markers The procedures for determining the position of the gut markers in the hormonally treated specimens were identical to those used to establish the premetamorphic intestinal map. Prior to dissection, the size and developmental stage of each specimen were recorded. Gut lengths and distances from the oral end to each marker were taken. These measurements were then converted to percent distance of each marker along the oral-anal axis. RESULTS Intestinal Mapping Figure 3 (top) shows the relative distribution of the five mapped points along the oral-anal gut axis in 10 premetamorphic but mature (i.e., Gosner stage 2-37)R. catesbeiana tadpoles. The switchback point was a n easily recognized permanent feature in both the coiled and uncoiled gut (Figs. l A , 2); in our R. catesbeiana tadpoles, it was 6210.9% (Z+SE) along the gut axis. Because of its location toward the caudal end of the gut, the remaining mapped points are unevenly distributed. The data reveal that the mapped points represent completely independent regions; there was no overlap in the ranges except for two marked points anal to the switchback point, and the mean location for those two points differed significantly (Mann-Whitney U, P<.0002). Metamorphic Induction All tadpoles submerged in thyroid hormone solution began to transform within 3 weeks of surgery, Unfortunately, all died during the 10 days following forelimbs emerging (Gosner stage 42). Control tadpoles, however, did not differ in metamorphosis from experimental animals. The experimental group that experienced the lower dose of exogenous thyroxine lived for 22-27 days (2322 days) after thyroid treatment started, whereas survival times for the group receiving the higher thyroxine dosage were 16-30 days (23k5 days). Survival times did not differ between the two groups (Mann-Whitney U, P = 0.52). Gut Shortening Gut lengths of the transformed froglets were compared with those of the mature tadpoles measured during the mapping experiment. The mean gut length in mature tadpoles (i.e., Gosner stages 37-41) was 1522.9 times their mean body length. In contrast, the mean gut length of the partially transformed froglets (of stages 42-44) was only 5.141.6 times their mean body length. Since the tadpoles in our premetamorphic sample were at metamorphic size, i.e., essentially the same 421 GUT METAMORPHOSIS I S X S U R A K S PREMETAMORPHIC Oral End 25% 47% t ? Coil v 62% T Coll Ili 52% 26% f Switchback point Coil 3 Coil 5 74% 62% 86% POSTMETAMOR PHlC I 0 I 10 I I I 20 30 I Anal End 84% 73% m k l S D I I I I I I 40 50 60 70 80 90 I I 100 PERCENTAGE DISTANCE ALONG ALIMENTARY TRACT Fig. 3. Above: Distribution of mapped points along the alimentary tract in R . catesbeiana tadpoles. Below: The same marked points in metamorphic R . catesbeiana after the gut has shortened to a third of its length. The intestines are shown here scaled to the same length. The ruler below shows the 410 distance of the marked points from the oral end. Arrows indicate the mean values obtained from 10 specimens, whereas the banded regions indicate i one standard deviation. The switchback point is a permanent feature represented here by a node. Roman numerals are assigned to points anterior to the switchback point, whereas Arabic numerals refer to points posterior. There is no overlap in the mapped regions in either the tadpole or metamorphic sample. The similar distribution above and below confirms that shortening is uniform along the length of the intestine. body length a s the froglets, this means that the intestinal tract in our R. catesbeiana had shortened both relatively and absolutely to a third of its premetamorphic length (i.e,, shortened by 66%). As expected, we found that the length of the thyroid treatment (in days) and the relative amount of gut shortening were correlated (Spearman Rank, 0.1 > P>.05). Not all gut implants were recovered from the partially transformed froglets; recovery ranged from a single implant in one specimen to all five implants in another (Table 2). However, all implants that were recovered properly fit the color scheme used a t the time of implantation and fell within the region (as % of gut length) described by the larval intestinal map (Table 1, Fig. 3). Most significantly, the relative distances of the markers along the oral-anal axis did not differ from the those in the premetamorphic specimens (Mann-Whitney U, P>O.lO). just the spiral portion of the small bowel. If the foregut (stomach and duodenum) were excluded from our measurement of gut length, the switchback point would be located much closer to the midpoint of the intestine. A concern to us was the fact that our surgically treated animals did not metamorphose spontaneously and that all of the hormonally treated specimens died before completing metamorphosis. The high mortality in surgical specimens is one reason the experiment was terminated and our sample sizes are small. An overdose of thyroxine may have been a factor, but our dosages were not exceptionally high compared to previous studies on metamorphosis where this hormone has been used. The possibility that the animals died because of either blockage of the digestive tract or starvation can be ruled out because we used large, latestage tadpoles about to metamorphosis and tadpoles normally stop feeding during metamorphosis (Duellman and Trueb, 1986). Despite these technical problems there is strong evidence that the intestines of the marked animals shortened in a natural fashion. First, the shortening occurred a t a constant rate. This conclusion is based on data of gut length versus length of time that the tadpoles were exposed to thyroid hormone. Also, the mean amount of shortening observed was consistent with the amount that has been reported in the literature for R . catesbeiana. We found a mean shortening of 66% between stages 41 and 43. Carver and Frieden (1977) DISCUSSION The coiling pattern of the R. catesbeiana larval gut varies little among individuals. Thus one can identify individual gut coils and determine how far along the oral-anal axis particular coils lie (Fig. 4).The most obvious feature in both the coiled and uncoiled gut is the switchback point, located 62?.91% from the oral end of the gut. This point is not halfway along the total length of the alimentary tract because we included in our measurements the whole alimentary tract and not 422 R. PRETTY ET AL. TABLE 1. Position of marker along alimentary tract as % distance from oral end (%+SE) Implant location Coil v At implantation in larvae (n = 10) 25 0.7 Coil iii 47 t 1.5 Switchback point 62 2 0.9 Coil 3 73 2 1.0 Coil 5 84 * -t 1.2 After metamorphosis 26 t 2.3 (n = 3) 52 -t 0.5 (n = 2) 62 2 0.5 (n = 8) 74 -t 1.1 (n = 5) 86 -t 0.9 (n = 8 ) TABLE 2. Implants recovered from surgery specimens after thyroxine-inducedmetamorphosis Specimen number 1 2 3 4 5 6 7 8 9 Oral end Coil v Coil iii (blue) (black) SB Point (blue) \’ \ \ \ \ \ \’ \ \ \ \ \ \ \ \ Anal end Coil 3 Coil 5 (black) (blue) \. \ \ \ \ ---Anal End -c-- reported a total shortening of 83% by stage 46 in this species. Our data indicate that intestinal shortening in R. catesbeiana occurs uniformly along the length of the gut. This conclusion is supported indirectly by Kaltenbach et al. (1981),who found that in natural metamorphosis, thyroid hormone was located in the gut epithelium along its entire length. Curiously, the surgical implants were not recovered uniformly along the intestine. Sutures placed at the switchback point and farther down the alimentary canal were recovered most often (Table 2). In contrast, those in the anterior 60% of the gut were more often lost. Although only 26/50 of the implants were recovered, there was little positional variation among the recovered implants (Fig. 3), consistent with our primary conclusion that gut shortening occurs uniformly along the intestine. The low recovery rate of implants from the surgical specimens can be explained by one or more of the following: (1)because of the delicate nature of the surgery, some implants may not have been properly secured in the first place; loose implants could have become dislodged and fallen either into the body cavity or into the gut lumen, then been passed by the tadpole; (2) during metamorphosis, the epithelium of the gut degenerates and sloughs off into the lumen of the gut (Marshall and Dixon, 1978); this could have loosened some implants and caused them to be carried into the lumen of the intestine; (3) the circumference of the intestine changes during metamorphosis, and as this happens, implants could have been excluded from Fig. 4. Intestine of R . catesbeiana larva shown coiled and uncoiled. Shaded areas refer to the anterior portion of the gut while white areas refer to the posterior portion of the gut. See Figures 1 and 3 for an explanation of other labels. Coil v is shown in this view but would more commonly be hidden by overlying coils in ventral view. the gut wall. The possibility, however, that some implants were digested by the tadpoles can be ruled out because, if this were the case, the implants that were recovered would have borne some sign of degradation, and that was never observed. In summary, this experiment shows that as the intestine of R. catesbeiana shortens to a third of its premetamorphic length, it does so uniformly along its oral-anal axis. Although metamorphosis was artificially induced, the shortening we observed closely matched the regression in intestinal length that occurs during natural metamorphosis. ACKNOWLEDGMENTS This study was part of the joint JapaneseICanadian program on “Visceral Function in Amphibians.” We thank Scott Pronych, Edward Hitchcock, and Ruth Waldick for help with animal collection and general laboratory assistance. Brian Hall, Jane Kaltenbach, and Monika Fejtek critically reviewed manuscript drafts. We are particularly grateful to Monika Fejtek who helped with computer graphics, statistical analysis, and manuscript formatting. The research was sup- GUT METAMORPHOSIS IN ANURANS ported by the Natural Sciences and Engineering Research Council of Canada, Pacific 2000 Program of the Canadian Department of External Affairs, and the International Scientific Research Program of Monbusho, Japan. LITERATURE CITED Carver, V.H., and E. Frieden 1977 Gut regression during spontaneous and triiodothyronine induced metamorphosis in Rana catesbeiana tadpoles. Gen. Comp. Endocrin., 31:202-207. Dauca, M., and J. Hourdry 1985 Transformations in the intestinal epithelium during anuran metamorphosis. In: Metamorphosis: The Eighth Symposium of the British Society for Developmental Biology. Oxford University Press, New York, pp. 37-58. Delsol, M., and M. Zervudacki 1962 Metamorphose de l'intestin entier du t6tard d'Alytes obstetricans isole en survie: Modalites et conditions de ce phenomene. C. R. Acad. Sci. Fr., 254:759-761. Duellman, W.E., and L. Trueb 1986 Biology of Amphibians. McGrawHill, New York. Etkin. W. 1981 Hormonal control of amDhibian metamoruhosis. In: Metamorphosis: A Problem in DevAopmental Biology. Plenum Press, New York, pp. 313-347. Fox, H. 1984 Amphibian Morphogenesis. Humana Press, Clifton, New Jersey. Gosner, K.L. 1960 A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica, 16:183-190. Hourdry, J., and A. Beaumont 1985 Les Mbtamorphoses des Amphibiens. Masson, Paris. ~~ 423 Ishizuya-Oka, A., and A. Shimozawa 1987 Development of the connective tissue in the digestive tract of the larval and metamorphosing Xenopus laeuis. Anat. Anz. Jena, 164:81-93. Janes, R.G. 1934 Studies on the amphibian digestive system. I. Histological changes in the alimentary tract of anuran larvae during involution. J . Exp. Zool., 67.73-91. Kaltenbach, J.C., C.H.K. Wang, and M.J. Lipson 1981 Histochemical study of the amphibian digestive tract during normal and thyroxine induced metamorphosis. 11. Acid phosphatase. J . Exp. Zool., 216:247-259. Marshall, J.A., and K.E. Dixon 1978 Cell specialization in the epithelium of the small intestine of feeding Xenopus laeuis tadpoles. J . Anat., 126:133-144. Nodzenski, E., R.J. Wassersug, and R.F. Inger 1989 Developmental differences in visceral morphology of megophryine pelobatid tadpoles in relation to their body form and mode of life. Biol. J . Linn. SOC., 38:369-388. Rugh, R. 1951 The Frog, Its Reproduction and Development. McGrawHill, New York. Shi Y-B., and D.D. Brown 1993 The earliest changes in gene expression in tadpole intestine induced by thyroid hormone. J . Biol. Chem., 268t20312-20317. Vanable, J.W. 1985 Benzocaine: a n excellent amphibian anesthetic. Axolotl Newsletter, 14:19-21. Yoshizato, K. 1989 Biochemistry and cell biology of amphibian metamorphosis with a special emphasis on the mechanism of removal of larval organs. Int. Rev. Cytol., 119:97-149.