THE JOURNAL OF COMPARATIVE NEUROLOGY 309~218-230(1991)
Patterning in the Regeneration of Electroreceptors in the Fin o Kryp top terus MICHELE MILLER BEVER AND RICHARD B. BORGENS Department of Anatomy, Center for Paralysis Research, School of Veterinary Medicine, Purdue University, West Lafayette, Indiana 47907
ABSTRACT The influence of the target tissue on afferent nerve regeneration was studied in the adult glass catfish, Kryptopterus. In this fish, electroreceptors in the anal fin are distributed in a characteristic pattern in the proximal part of the fin and are absent in the distal portion of the fin. We tested whether axons were more likely to induce electroreceptors in certain regions of fin epidermis than in others. We rotated fin transplants so that the location of the degenerating electroreceptors was altered with respect to the regenerating axons in the host tissue dorsal to the fin. The effects of these rotations were observed in the living animal with differential interference contrast optics over a period of 10 weeks. When transplants were reversed rostrocaudally, new electroreceptors formed in the caudal half of the interradial zone, where degenerating electroreceptors were at the time of transplantation. When transplants were rotated so that the dorsoventral and rostrocaudal axes were reversed, some new receptors formed in the old target site regions that were located in the caudal interradial zones (in the distal half of the graft with respect to the host). Regenerating axons reached these regions of the transplant by taking unusual routes around the electroreceptor-free regions of fin. Very few electroreceptors formed in the distalicaudal or proximacaudal interradial quadrants of grafts where the original orientation of the tissue was maintained. We suggest that old target sites have a neurotropic influence on the regenerating afferent axons and discuss the possibility that the distal fin epidermis is not as permissive to electroreceptor formation as proximal fin epidermis. Key words: ampullary organs, catfish, lateral line nerve
Over 60 years ago, Ramon y Cajal ('28) postulated that regenerating nerve fibers were guided to their proper targets by "alluring or attracting substances." He suggested that growing nerve fibers were grossly attracted to the periphery where a more local and specific attracting mechanism was present to guide the fibers further to their terminal structures. Investigations into the nature of these guiding forces have continued to present times. In the chick embryo, developing motor neurons appear to be able to seek out their correct targets and make appropriate connections even when they or their targets in the limb have been shifted in the anterior-posterior axis of the embryo (Lance-Jones and Landmesser, '81).These authors suggested that a diffusable substance originating from the target was responsible for this directed growth. However, later studies provided evidence that motoneurons were capable of correct pathway selection even in the absence of their target (Tosney and Landmesser, '86; Phelan and Hollyday, '90). Neurotropic substances have also been implicated in the guidance of regenerating type I sensory axons to old mechanoreceptor sites (touch domes) after nerve transection in cats (Horch, '79, '82) and in the O 1991 WILEY-LISS, INC.
guidance of sprouting or regenerating mechanosensory axons to Merkel cells in the skin of salamanders (Diamond, '82). Here, we explore the possibility of axon guidance to targets of another sensory modality: electroreception. The electroresensory system of fish and amphibia plays an integral role in the animal's behavior, assisting in, among other things, feeding, orientation, and reproduction (Bullock, '74). Tuberous andor ampullary electroreceptor organs are distributed over the head and body in characteristic patterns that differ from species to species (Szabo, '74). In the nonelectric transparent catfish, Kryptopterus, the distribution of ampullary organs includes five longitudinal rows on each side of the trunk, a row along each of the pectoral, anal, and tail fin rays, and a scattering over the head and gills (Wachtel and Szamier, '69). These sensory receptors will atrophy in response to nerve transection (Bailey, '37; Roth and Szabo, '69; Szamier and Bennett, '73; Roth, '85), but will reform with the reinstatement of their innervation (Bailey, '37; Roth and Szabo, '69; Yialamas and Zakon, '84; Zakon, '86; Bever and Borgens, '91). Accepted March 29,1991.
219
PATTERNING IN RECEPTOR REGENERATION In our companion paper, we showed that after nerve transection, electroreceptorsregenerate in appropriate locations of an anal fin transplant that originally contained electroreceptors (Bever and Borgens, '91). Comparatively, when regenerating axons were faced with a fin transplant from a receptor-free region of fin, very few penetrated the graft and, in fact, most new electroreceptors formed in the proximal or peripheral zones (Bever and Borgens, '91). The latter results hint that the distal portion of the anal fin, which usually does not contain electroreceptors, is not permissive to their formation or does not possess the "cues" necessary to lead to their patterned development. In the present study, we used transplantation experiments to test the possibility that certain regions of the anal fin are attractive or inhibitory to the regenerating axons. Pieces of fin containing electroreceptors in characteristic locations were rotated and substituted for the normal fields of regenerating nerves. We show that although some nerve fibers induce receptors at new sites, other fibers will traverse substantial distances to induce electroreceptors in regions of old receptor sites.
distal portion of the graft (Fig. 1)and in the caudal half of the interray, relative to the regenerating axons dorsal to the graft. Therefore, in these grafts, an expanse of electroreceptor-free fin was interposed between regenerating axons projecting from the adjacent and dorsal body segment, and a region of fin graft originally containing electroreceptors. These electroreceptors would eventually degenerate and disappear since the nerves innervating them were severed during the transplantation procedure (see Bever and Borgens, '91). Moreover, the remnants of these axons were located at the distal margin of the graft, far from the regenerating nerves of the host.
Data analysis
Composite drawings. The grafts were observed weekly with DIC optics and were photographed with Kodak Technical Pan film. Tracings of the fin rays and electroreceptors in the grafts were made from photomicrographs and the axon pathways were superimposed after visually following axons through multiple planes of focus in the living animal. Analgsis. Newly formed electroreceptors in and around the grafts were classified according to their proximodistal position in the fin and placed in one of the following MATERIALS AND METHODS categories: peripheral zone (grafthost marginal region), Animals and observation proximal zone (portion of the graft nearest to the base of Adult transparent catfish, Kryptopterus (4-6 cm) were the fin; within 400 p m from the peripheral zone), middle maintained in fresh water aquaria at 25°C and fed brine zone (400 pm band in the central region of the graft), or shrimp. Observation techniques have been described previ- distal zone (portion of the graft furthest from the base of ously in our companion paper (Bever and Borgens, '91). the fin; greater than 800 p m from the peripheral zone) (see Briefly, fish were anesthetized in 0.05% tricaine methane Bever and Borgens, '91). Electroreceptors within a graft (all sulfonate (MS 222) and placed in a special viewing chamber zones except peripheral) were also classified according to (see Bever and Borgens, '91) that was suited for attachment their location within the interradial zone and placed in one to the rotating stage of a Nikon Diaphot inverted micro- of the following quadrants: proximal-caudal (quadrant 11, scope fitted with differential interference contrast (DIC) proximal-rostra1 (quadrant 21, distal-caudal (quadrant 31, optics. Wet gauze wrapped loosely around the head o f the or distal-rostral (quadrant 4).Electroreceptors that formed directly over a fin ray or in places where fin rays were fish kept the gills from drying during the viewing period. closely surrounding them, or exactly halfway between fin Surgical procedures rays, making rostral or caudal classification impossible, The surgical methods used have been described in detail were classified as undetermined (see Bever and Borgens, in the companion paper (Bever and Borgens, '91). In the '91). Regenerating axons induced electroreceptorsin the transpresent study, rectangular pieces (approximately2 mm by 4 mm) of caudal anal fin were removed by dissection and plants by projecting toward the target sites in one of two transplanted to a prepared site in the rostral part of the ways: 1) they projected into the interradial zones of the anal fin in two different orientations: 1) transplants in transplant in a proximal-distal direction; or 2) they trawhich the rostral-caudal axis was reversed (RC grafts); and versed a large expanse of fin either in the interradial zones 2) transplants in which the dorsal-ventral and rostral- or within the fin rays in a proximal-distal direction, and then reversed their direction of growth to project over 100 caudal axes were reversed (DV grafts). RC graft sites were prepared by removing a small rectan- pm in the opposite, distal-proximal direction. All axons gular piece of anal fin that contained electroreceptors and were classified by grouping them into one of these two portions of three fin rays (Fig. 1). Another piece of fin that categories of projection. The results of these grafting experiments were also contained electroreceptors was obtained from the proximal half of a more caudal region of anal fin in the same animal compared to prior results obtained from electroreceptorand was transplanted into this graft site. Before suturing, containing (EC) grafts, in which the position of receptors the donor tissue was rotated 180"around the dorsal-ventral within the fin was not disturbed after grafting, and electroreaxis. The resulting orientation of the graft tissue placed ceptor-free (EF) grafts (Bever and Borgens, '91). The degenerating electroreceptors in the caudal half of the Mann-Whitney,Student's t, and Chi-square tests were used interray with respect to the regenerating axons in the host for statistical comparisons. tissue dorsal to the graft (Fig. 11, instead of in the rostral interray (the normal position as observed in the remainder RESULTS of the fin). The transplanted tissue was sutured to intact fin RC grafts rays on either side of the graft with 10-0 Ethilon suture. DV grafts were prepared as above with the exception that In the rostral-caudal reversed grafts, the donor tissue the donor tissue was rotated 180"around a transverse axis (Figs. 1, 2A) was rotated so that electroreceptors and their (Fig. 1). In these double reversed cases, the resulting graft axons were relocated to the caudal half of the interray with orientation placed degenerating electroreceptors in the respect to the normal fin architecture (Figs. 1,2B).Because
220
Fig. 1. Diagram demonstrating rostral-caudal (RC)and dorsalventral (DW graft procedures. A rectangular piece of rostral anal fin containing electroreceptors was excised to create a graft site (G) with dimensions of approximately 2 mm by 1 mm. This process removed portions of three fin rays and also transected the nerves that normally project into the rostral half of the adjacent interrays. A piece of caudal anal fin (A) containing electroreceptors was transplanted into this graft site. Before suturing, the donor tissue was rotated one of two ways. The rotation placed electroreceptors in abnormal locations with respect to the host tissue and its regenerating nerve fibers. In the first case (RC),
M.M. BEVER AND R.B. BORGENS
donor tissue was rotated 180” around the dorsal-ventral axis so that electroreceptors in the graft appeared to be in the caudal half of the interrays with respect to the regenerating nerve fibers in the host. In the second case (DV), the donor tissue was rotated 180” around a transverse axis so that electroreceptors appeared to be in the caudal half of the interrays and in the distal half of the graft in relation to the regenerating nerve fibers in the host. Dorsal (D), ventral (V), rostral (R), and caudal ( C ) denotes orientation of host. d, v, r, and c denote original orientation of the transplant tissue.
the end of the fin rays normally slant caudally from in Table 1. Seventy-eight percent (mean of quadrants proximal to distal, a 180”rotation around the dorsal ventral 1 + 3, Table 2) of these newly formed receptors were axis resulted in fin rays within the graft slanting in the located in the caudal half of an interradial zone. In the absence of interference with the sensory innervaopposite direction relative to the intact fin rays surrounding them. Therefore, when the transplant was sutured into the tion, electroreceptors in a given interradial zone are supgraft site, it was difficult to align the proximal ends of plied by axons that enter the interray from the adjacent grafted fin rays with the distal ends of the host fin rays body segment dorsal to it (Bever and Borgens, ’91).Thus, in dorsal to the graft (Fig. 2B,C). In all of these animals the order to determine the effect of graft reorientation on axon host fin ray stumps and the fin rays within the transplant regeneration, the projection of the axon supplying each new itself alternated to some degree rather than aligning. electroreceptor within and around each RC graft was Sometimes host site fin ray stumps that were not aligned followed. An example of the axon paths for one RC graft is with graft rays regenerated distally a short distance to form shown in Figure 3. In this example, the axons regenerated an incomplete fin ray or “spike” (Fig. 2D) or grew even distally into the interradial zones by fairly direct pathways, further distally to produce an additional or supernumerary with only three axom growing into an adjacent interray. It new fin ray within the graft. Of the eight animals surgically is interesting to note that even though the fin rays of the manipulated in this way, seven completed 10 weeks of graft were not aligned with the fin ray stumps at the host observation and one died of unknown causes 1week after site, the two parts often appeared to fuse along their overlapping edges (Fig. 2D). Fin rays are comprised of two surgery. One week after transplantation, all RC grafts were viable opposing pieces, one beneath the epidermis of each side of and well vascularized. However, the electroreceptors within the fin. Some of the axons shown in Figure 3 grew between the grafts had degenerated since being denervated at the the paired components of the fin rays in this fused region to original surgery (Fig. 2C). By 2 weeks posttransplantation, enter the interray while other axons grew around the new receptors had appeared in all seven grafts, and had proximal end of a graft fin ray to enter an adjacent interray. greatly increased in number by 10 weeks after surgery (Fig. All 114 axons traced in RC grafts had induced electrorecep2D). A summary of the number of electroreceptors and tors after projecting to the sites in a proximal-to-distal their proximodistal positions within each graft is presented direction.
221
PATTERNING IN RECEPTOR REGENERATION
DV grafts
tors appeared near the proximal grafthost tissue boundary. During the fourth to tenth weeks after surgery, additional In the dorsal-ventral reversed grafts, the transplant receptors appeared at the proximal interface and in more tissue (Fig. 4A) was rotated so that degenerating electrore- distal locations in nine of the transplants (Fig. 4D). A ceptors were in the distal half of the graft and in the caudal summary of the number and proximodistal distribution of half of the interradial zone with respect to regenerating new receptors within each DV graft is shown in Table 3. axons penetrating the graft from the more dorsal body The rostral or caudal interray location of electroreceptors segment (Fig. 4B). After reorientation, the distal segments was determined for all DV grafts. As mentioned above, of axons innervating these electroreceptors extended away electroreceptors in the normal animal are typically found in from them toward the distal grafthost margin. Of the 17 the rostral half of each interradial zone. Thirty-one percent animals receiving this treatment, four lost their grafts (mean of quadrants 2 4, Table 2) of electroreceptors within the first week and three died of unknown causes. Of within all ten DV grafts were in such a “normal” locathe remaining ten animals, seven survived an observation tion-in the rostral half of an interradial zone. Fifty-nine period of greater than 9 weeks, while the other three fish percent (mean of quadrants 1 + 3) of all electroreceptors lived 4 , 5 , and 7 weeks. were located in the caudal interray (Table 2). However, The DV grafts were viable and appeared to be well whether a new electroreceptor appeared in the rostral or vascularized at 1 week after transplantation. As in RC caudal half of an interradial zone depended upon whether it grafts, endogenous electroreceptors had degenerated and was in the proximal or distal half of the transplanted fin were no longer visible within the transplant (Fig. 4 0 . New (Table 2). In the proximal half of the transplants, the mean receptors began appearing within the DV grafts at 2 to 3 percent of new receptors that formed in the rostral weeks after surgery. In one graft, the first electroreceptors (“normal”) portion of an interradial zone was 28.0%(quadappeared in the distal end of the graft, while in two rant 2). A similar mean of 28.8% formed in the caudal additional grafts the first electroreceptors appeared simul- interradial zone (quadrant 1).In the distal half of the DV taneously at the proximal boundaries of the grafts and in grafts, a mean of 29.8%of the receptors was in the caudal distal regions of the transplants. In a fourth graft, electrore- half of the interradial zone (quadrant 3), while a mean of ceptors never appeared in the 10 weeks of observation. In only 3.4% formed in the rostral half (quadrant 4). the remainder of the grafts (6/10),the first new electrorecepA qualitative analysis of the pattern of innervation of new electroreceptors in the DV grafts revealed that it was not uncommon for the axons to grow between the paired TABLE 1. Number and Distribution of Newly Formed components of the fin rays (Fig. 5 ) . In several cases (11 Electroreceptors in RC Grafts axons in five different grafts), these axons would grow Proximodistal position in fin distally inside the fin ray for the entire width of the graft Middle Distal and then exit the fin ray and turn through 180“ to grow Peripheral Proximal Total Animal proximally for some distance exceeding 100 km (at times zone zone zone zone no. receptors 400-500 km) before inducing the formation of an electrore7 4 15 RC-1 26 ceptor (Fig. 5). Thirty-three percent of the total number of 7 5 2 RC-2 14 4 11 3 2 RC-3 20 axons traced in DV grafts approached their target sites RC-4 24 3 17 5 from a distal-to-proximal direction; all 14 of these axons 6 12 4 RC-5 29 7 24 1 4 RC-6 29 supplied electroreceptors that were located in the distal 11 3 RC-8 14 half of the transplant.
+
TABLE 2. Summary of Electroreceptor Location in Grafts Quadrant’ Graft
Animal
EC’
EF’ RC
DV
1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10
3
4
N3
Undetermined’
Peripheral4
76.6 51.2
3.1 4.8
9.2
T = 85 T = 30
0 0
23.1
0.6 5.2 26.9 7.1 5.0 12.5 0 3.4 28.6 11.9 0 28.6 0 0 5.3 17.9 0 0 5.9 0 5.8
14.3 69.6 15.4 0 20.0 8.3 24.1 82.8
1
2
Mean 0.8 Mean 29.8 100.0 76.9 73.3 78.9 59.1 0
90.0 Mean 68.3 50.0 0 0 0
100.0 18.2 100.0
20.0 0 0 Mean 28.8
13.3 15.8 9.1 75.0 10.0 17.6 50.0 75.0
0 0
5.3 13.6 25.0 0
9.6 0
25.0
0
0
20.0 0 54.5 0
80.0 0 18.2
60.0 0
20.0 75.0 80.0 29.8
20.0 28.0
0
‘Numberofelectroreceptors In each quadrant #ven as a prrwntage of N ’Only the summnry data has been reproduced from Table 4 , Uever and Rorpn3 ‘91 ’N.tuwl receptors in graft thar could beraregorired ‘Suntber uf electroreceptor:. g i v m sm a prrcentege of t hr total wcepmrs.
0 0 0
13.3 0.0 18.2 0 0 4.5 0 0 0 0 0
9.1 0 0 25.0 0 3.4
15 13 15 19 22 4 10 T = 98 2 4 0
5 1 11 1 5
4 10 T = 43
0
21.5 0
14.3
Total receptors T = 103 T = 106 26 14 20 24 29 29 14 T = 156 2 7
0
0
68.8 89.5 42.9 92.3 72.2 70.6 41.2 49.2
16 19 28 13 18 17 17 T = 137
M.M. BEVER AND R.B. BORGENS
222
Figure 2
PATTERNING IN RECEPTOR REGENERATION
Figure 2 continued
223
224
M.M. BEVER AND R.B. BORGENS
Fig. 3. A composite drawing of newly formed electroreceptors and their respective innervation in an RC graft. Same graft as presented in Figure 2. All of the new electroreceptors in this graft were located along the rostral aspect of the fin rays (with respect to host tissue) except for
one (arrowhead), which could not be classified as rostral or caudal. Arrows indicate axons that grew into neighbor interrays. Broken lines signify portions of axons that grew between the paired components of the fin rays.
DISCUSSION
receptors with respect to the regenerating nerves. The results suggest that regions once containing electroreceptors may somehow influence the destination of the ingrowing axons.
In this study, we tested the ability of old receptor sites to attract regenerating nerve fibers by rotating the fin transplant, thus altering the relative location of the degenerating
Patterning in axonal regeneration Fig. 2. (pages222-223) Regeneration of electroreceptors in an RC graft. A Donor region of the caudal anal fin to be transplanted to a new site in the rostral anal fin. Severed fin rays at the distal margin of the presumptive graft are marked with arrows. The graft margins that remain to be cut are delimited by dashed lines. Note that electroreceptors are located only along the caudal aspect of each fin ray. B: Donor tissue (same tissue as in A) sutured (arrows) to intact fin rays on either side of the graft site. Transplant was rotated 180" around the dorsalventral axis before suturing so that electroreceptors (arrowheads) within the graft were relocated along the new rostral aspect of the fin rays with respect to the surrounding host tissue. Note that the proximal ends of the transplant fin rays are not aligned with the fin ray stumps of the host. C: The same graft as presented in B at 1week posttransplantation. The graft margins have healed and the graft is well vascularized, but the electroreceptorsthat were within thegraft have disappeared. D: Ten weeks after grafting, electroreceptors have reappeared within the graft tissue. In this animal, all of the new receptors appeared along the rostral aspect of graft fin rays (compare to normal electroreceptors outside of the graft (arrows). Two of the host fin ray stumps have regenerated bony spicules within the graft tissue (arrowheads). Scale bar for A, B, C, D = 0.5 mm.
A curiosity in the regeneration of peripheral nerve fibers is the reestablishment ofpattern in the nerve field (Ramony Cajal, '28; Weiss, '39; Guth, '57; Sperry, '63; Diamond, '82). Sensory fibers supplying muscle spindles, taste buds, lateral line organs, Merkel cells, touch domes, pacinian corpuscles, or electroreceptors are somehow "directed" to the correct region of skin or muscle to renew contact with, or induce the formation of, their sensory end organs (Speidel, '64; Zalewski, '72; Burgess et al., '74; Brown and Butler, '76; Diamond, '82; Zelena, '84; Bever and Borgens, '91). One explanation for the return of regenerating nerve fibers to specific foci is that they have access t o empty Schwann tubes that could lead them directly to previously innervated sites (Young, '49; Burgess et al., '74). This notion is well supported in nerve crush cases (Horch, '79; Cheal and Oakley, '77; Zelena, '84; Diamond et al.,'87) and may also be relevant to the projection pattern after nerve trunk transection (Ramon y Cajal, '28; Speidel, '63, '64; Terzis and Dykes, '80). However, there are examples of
225
PATTERNING IN RECEPTOR REGENERATION
Regeneration of electroreceptors within transplants
TABLE 3. Number and Distribution of Newly Formed Electroreceutorsin DV Grafts
Proximodistal position in fin
The RC and DV transplantation experiments with catfish fin have provided results that are most readily interpreted in terms of attracting properties of the dermis or epidermis DV-12 2 2 in regions of old receptor sites. In RC grafts, reinnervation DV-13 7 1 5 1 of transplants occurred according to the original orientaDV-14 0 tion of the graft prior to transplantation. After the transDV-19 16 11 1 1 3 DV-23 19 17 1 1 planted tissue was rotated at the time of surgery, degeneratDV-24 28 12 5 8 3 ing electroreceptorswere in the caudal half of the interradial DV-25 13 12 1 DV-26 18 13 4 1 zones with respect to the host. The terminal segments of DV-27 17 12 1 2 2 their axons extended dorsally toward the proximal graft/ DV-28 16 7 2 5 2 host interface. Weeks later, the majority of the newly formed receptors were in that same region, that is, in the caudal half of the interradial zones with respect to the host. A comparison of the mean percent of receptors occurring in nerve transections in which regenerating nerve fibers show quadrant 1in RC and EC grafts (in which no manipulation a strong predilection to return to specific sites but do not of receptor location occurred at the time of graft transplanreach them via “preexisting pathways” (Speidel, ’64; Horch, tation) is shown in Figure 6. Receptors were more likely to form in quadrant 1 in RC grafts than in EC grafts ’79, ’82; also see below). Patterned regeneration can also be viewed in terms of (P < 0.005). These results could possibly be explained by positional information. Wolpert (’71) suggested that migrat- the capability of regenerating axons to find old Schwann ing “cells would ‘test’ the postional information of the cells tubes that then direct them to old target sites. One way to over which they were moving and would stop at that value eliminate this possibility is to create a situation in which for which they were programmed.” One proposed mecha- there are degenerating receptors, but no preexisting fascicnism €or conferring such a positional cue to growing nerves ulation pathways accessible to regenerating nerves. This involvesa differential adhesiveness of the substrate; growth situation was produced in the DV grafts. In DV grafts, transplanted tissue was rotated so that would continue in the direction of increasing adhesivity (Nardi and Kafatos, ’76). Nardi (’83)has demonstrated a degenerating electroreceptors were in the distal portion of proximodistal asymmetric distribution of matrix molecules the graft and in the caudal half of the interray (quadrant 3) in the moth wing that could be involved in the guidance of with respect to the regenerating axons dorsal to the graft. developing neuron processes proximally toward their tar- Figure 7 compares the mean percent of receptors that gets. In this example, axons actually avoided grafts with the formed in quadrant 3 of DV grafts with the mean percent that formed in the same quadrant in EC grafts. Receptors “incorrect” proximodistal orientation. were more likely to form in quadrant 3 in DV grafts than in A third explantion for reestablishment of pattern is that the target sites are neurotropic (Ram6n y Cajal, ”28;Horch, EC grafts (P < 0.05). Indications of target site attraction properties are re’79; Diamond, ’82; English et al., ’83). In this view, axons are directed to appropriate regions of skin or even specifi- vealed by comparing the results of dorsoventrally (DV) cally to old sites of innervation because of a specific rotated grafts with the results of the electroreceptor-free attraction emanating from those sites. Cells of the target (EF) grafts (Bever and Borgens, ’91). These two types of site could presumably release an attractant or “sprouting grafts are similar in that a portion of the transplant, a band factor,” the production of which could be inhibited by about 600 pm wide closest to the regenerating axons uptake or neutralization after the successful establishment emanating from the trunk of the fish, is initially free of of new nerve connections (Ram6n y Cajal’ ’28; Diamond, receptor organs. The two types of grafts differ, however, in ’82). Scott and colleagues (’81) have demonstrated that that electroreceptors and their terminal axon segments Merkel cells persist when deprived of their innervation. existed in the distal half of DV grafts after rotation and there were no such structures in the EF grafts. If we These cells subsequently become innervated by sprouting compare the innervation patterns and the locations of new or regenerating nerve fibers. Furthermore, new Merkel cells appear in regenerating skin in the absence of innervation. They become innervated when the nerves regenerate. This latter result supports the notion that the Merkel cells Fig. 4. (pages 226-227) Regeneration of electroreceptors in a DV are neurotropic, since there were no old nerve pathways for graft. A: Donor tissue in the caudal anal fin. The margins of the the regenerating axons to follow to these targets, but does presumptive transplant are delimited with dashed lines. Electrorecepnot exclude the possibility that innervation was random tors are found only along the caudal aspect of each fin ray. B: Donor and that axons unsuccessful in finding targets did not tissue transplanted into the graft site. Before suturing, the transplant was rotated 180” around a transverse axis so that electroreceptors survive. In another system, Horch (’82) demonstrated that (arrowheads) within the graft appear to be in the caudal half of the uncauterized touch domes in the cat were more likely to be interradial zone and in the distal half of the graft with respect to the reinnervated after nerve transection than ones that had orientation of the host. C: Same graft as in B, 1week after transplantabeen cauterized. Since regenerating axons in both cases tion. Graft is viable and well vascularized. Original electroreceptors are would have access to degenerate perineural tubes, and since absent. D: Same graft as B, 10 weeks after transplantation. Receptors have formed at the dorsal margin of the graft and in the distal half of there was not equal reinnervation of the two groups, he the graft. All of the new receptors (arrowheads) in the distal half of the concluded that the growing axons were influenced by some graft are located on the rostra1 aspect of the fin ray (in the caudal half of intrinsic property of the unlesioned target sites. an interray with respect to the host). Scale bar for A, B, C, D = 0.5 mm. ~
Animal no. ~
Total receutors
Peripheral zone
~
Proximal zone
Middle zone
Distal zone
M.M. BEVER AND R.B. BORGENS
226
Figure 4
PATTERNING IN RECEPTOR REGENERATION
Figure 4 continued
227
M.M. BEVER AND R.B. BORGENS
228
Figure 5
PATTERNING IN RECEPTOR REGENERATION receptors in these two types of grafts, we find three outstanding differences between them: 1. Very few axons (mean 4.8%,quadrants 3 + 4; Table 2) grew into the distal region of the EF grafts to induce electroreceptors, whereas one-third (mean 33.2%;Table 2) of the axons grew into distal regions of DV grafts and induced electroreceptors(P < 0.05, Mann-Whitneytest). 2. The pattern of newly formed receptors in the distal half of the DV transplants was reestablished according to the original position of these structures within the transplant tissue, that is, a mean 29.8%(quadrant 3, Table 2) of the new receptors in the graft were in the distalcaudal quadrant of an interradial zone with respect to the host (in the same regions where degenerating receptors were prior to rotation at the time of surgery). A mean of only 3.4% of the new receptors were in the distal-rostral quadrant (quadrant 4) of an interray (P < 0.05, Mann-Whitney test). On the other hand, in the proximal half of the DV grafts, the mean percent of new receptors was similar for the rostral (quadrant 2, 28%)and caudal (quadrant 1,28.8%)half of the interradial zones (P = 0.7207, Mann-Whitney test). EF grafts displayed a greater mean percent of new electroreceptors in the proximal-caudalquadrant (quadrant 2) of the interradial zones. A mean of 51%of new receptors were in quadrant 2 of an interradial zone, while a mean of 30%of new receptors formed in quadrant 1. 3. Although the number of axons that invaded each transplant was similar for EF and DV grafts (P > 0.20, Mann-Whitney test), a considerably larger proportion (14/43) of the axons in DV grafts made their final approach to new receptor sites in a distal-to-proximal direction, compared to only a few (2/36) in the EF grafts (P < 0.01, Chi-square test). Proximal-distal location of new electroreceptors. It is important to note that the proximal portions of the DV and EF grafts were nearly identical, except that the proximodistal orientation of DV grafts was altered by 180”. Keeping this in mind, regenerating axons in each case were faced with “foreign” electroreceptor-free fin in the proximal half of the grafts. It appears, though, that axons found the distal half of the EF grafts less congenial or less attractive than the distal half of DV grafts. Only 5% of new receptors formed in the distal half of the EF grafts compared with 33%in DV grafts (Table 2). To state this another way, the distal portion of the DV grafts was attractive enough that axons crossed substantial distances, penetrating a proximal region of electroreceptor-free fin, to reach this area. This distal region once contained electroreceptors; however,
229 there were no Schwann tubes extending proximally to guide the axons from the dorsal grafthost interface to the distal half of the transplant. Pattern of new electroreceptors. If we consider the influence of the proximal or distal half of the DV grafts on the pattern of the new electroreceptors, we find that in the distal half the receptors tended to form in a location reminiscent of their original orientation prior to transplantation, while in the proximal half, receptors showed no tendency to form in a particular region of interray. These results are consistent with the concept that old receptor sites have attracting properties. Axon pathways to target sites. A striking number of axons (33%)that induced electroreceptors in the DV grafts reached their targets via “unusual” routes. In addition to these axons growing first distally for long distances and then proximally for at least 100 pm, 11 of these axons also traversed the fin by growing between the paired components of fin rays. The central region of a fin ray contains nerves and blood vessels (Goss, ’69).Although one might
100
coal-
A A
80
8
OF3
60
A
!?
0
$3 -u 4-3 CCS
8 .E
z
40
a 20
Fig. 6. Percent of electroreceptors in quadrant 1 (shaded area) in EC and RC grafts. Mean f S.E.M. was 0.8 f 0.8 and 68.3 f 12.4, respectively. Differences between EC and RC grafted animals: P = 0.0026 with the Mann-Whitney test and P = 0.0001 with the t test.
90
AA ?!!
0
c
a m
a*
8:
60
z’op - 3
Fig. 5. Composite drawings o f newly formed electroreceptors and their respective innervation in DV grafts. A Same graft as shown in Figure 4. Four of the new eledroreceptors (large arrowheads) in the distal half of the graft were induced by axons that grew between (broken lines) the paired components of a fin ray to the distal margin of the transplant before exiting and growing back proximally for several hundred micrometers (small arrowheads). B : Composite drawing of a second DV graft. Three of the axons (arrows) that innervated new electroreceptors grew between (broken lines) the paired components of the fin rays to the distal margin of the graft, where they then exited and grew proximally along the rostral edge of the fin ray. Of these three axons, two induced electroreceptors on the rostral aspect of a fin ray (large arrowheads) while the third axon crossed the interradial zone and induced an electroreceptor on the caudal aspect of a fin ray (small arrowhead).
C
0,
0
z
U
.r
30
a
EC
DV
Fig. 7. Percent of receptors in quadrant 3 (shaded area) in EC and DV grafts. Mean -+ S.E.M. was 3.1 f 2.0 and 29.8 ? 11.0, respectively. Differences between EC and DV grafted animals: P = 0.0361 with the Mann-Whitney test andP = 0.0323 with the t test.
M.M. BEVER AND R.B. BORGENS
230 argue that the intrinsic nerves of the fin ray could serve to direct regenerating axons to the distal end of the DV grafts, there are two lines of evidence that do not support this view. First, in RC grafts, fin rays of the transplant were frequently misaligned with the fin ray stumps of the host site (see Results). This misalignment placed fin rays of the transplant in an excellent position to be penetrated by regenerating axons from the adjacent body segment. However, we did not observe one example of an axon growing through the center of a fin ray in these grafts. Second, even though the proximal portions of the EF and DV grafts were similar in that neither contained any remnant of previous afferent electroreceptor innervation, in E F grafts there were only seven axons that grew between fin rays, and only two of those seven made it as far as the distal half of the transplant (neither of these two traversed the entire width of the transplant). This can be compared to DV grafts where 11 axons grew inside fin rays for the entire width of the graft and induced electroreceptors in the distal portion of those grafts. In summary, we have shown that although the preexistence of receptors is not required for their induction, axons have a propensity for inducing receptors in regions of old receptor sites. This may, in part, be due to the use of degenerate perineural tubes that lead the regenerating axons directly to the previous target sites. However, we also show evidence that, in the absence of such “pathways,” axom will traverse great expanses of fin to induce electroreceptors in regions of the fin that previously contained receptors. We suggest from these results that an intrinsic property of the regions of fin that previously contained receptors is responsible for the guidance of regenerating axons to those regions.
ACKNOWLEDGMENTS This work was supported by NIH grant R 0 1 HD 20664 (R.B.B.), and the David Ross Fellowship Program of Purdue University (M.M.B.). We wish to thank Mary Jo Maslin for manuscript preparation and Barbara Sturonus for the artwork. We thank J.W. Vanable, Jr., A.R. Blight, and M.E. McGinnis for their helpful critique of this manuscript.
LITERATURE CITED Bailey, S.W. (1937) An experimental study of the origin of lateral-line structures in embryonic and adult teleosts. J. Exp. Zool. 76:187-233. Bever, M.M., and R.B. Borgens (1991) The regeneration of electroreceptors in Kryptopterus. J. Comp. Neurol. 309:ZOO-217. Brown, M.C., and R.G. Butler (1976) Regeneration of afferent and efferent fibers to muscle spindles after nerve injury in adult cats. J. Pbysiol. (Lond.) 260:253-266. Bullock, T.H. (1974) General introduction. In A. Fessard (ed): Electroreceptors and Other Specialized Receptors in Lower Vertebrates. New York: Springer-Verlag, pp. 1-12. Burgess, P.R., K.B. English, K.W. Horch, and L.T. Stensaas (1974) Patterning in the regeneration of the type I cutaneous receptors. J. Physiol. 236.57-82. Cheal, M., and B. Oakley (1977) Regeneration of fungiform taste buds: Temporal and spatial characteristics. J. Comp. Neurol. 172r609-626. Diamond, J. (1982) Modeling and competition in the nervous system: Clues from the sensory innervation of the skin. In R.K. Hunt (ed): Current Topics in Developmental Biology, Vol. 17. New York: Academic Press, pp. 147-205. Diamond, J., M. Coughlin, L. Macintyre, M. Holmes, and B. Visheau (1987) Evidence that endogenous nerve growth factor is responsible for the
collateral sprouting, but not the regeneration, of nociceptive axons in adult rats. Proc. Natl. Acad. Sci. U.S.A. 84:65966600. English, K.B., D. Kavka-Van Norman, and K. Horch (1983) Effects of chronic denervation i n type I cutaneous mechanoreceptors (Haarscheiben). Anat. Rec. 207:79-88. Goss, R.J. (1969) Principles of Regeneration. New York: Academic Press. Guth, L. (1957) The effects of glossopharyngeal nerve transection on the circumvallate papilla of the rat. Anat. Rec. 128t715-731. Horch, K. (1979) Guidance of regrowingsensory axons after cutaneous nerve lesions in the cat. J. Neurophysiol. 421437-1449, Horch, K. (1982) The influence of mechanoreceptor structures on regenerating sensory axons after cutaneous nerve transection in the cat. Neurosci. Lett. 32:281-284. Lance-Jones, C., and L. Landmesser (1981) Pathway selection by embryonic chick motoneurons in a n experimentally altered environment. Proc. R. SOC. Lond. [Biol.]214:19-52. Nardi, J.B. (1983) Neuronal pathfinding in developing wings of the moth Manduca sexta. Dev. Biol. 95t163-174. Nardi, J.B., and F.C. Kafatos (1976) Polarity and gradients in lepidopteran wing epidermis 11. The differential adhesiveness model: Gradient of a non-diffusible eel1 surface parameters. J. Emhryol. Exp. Morpbol. 36r489512. Phelan, K.A., and M. Hollyday (1990) Axon guidance in muscleless chick wings: The role of muscle cells in motoneuronal pathway selection and muscle nerve formation. J. Neurosci. lOr2699-2716. Ram6n y Cajal, S. (1928) Degeneration and Regeneration of the Nervous System. (R.M. May, trans.). New York: Hafner Publishing Co., (reissued 1959). Roth, A. (1985) Axonal flow in the afferent fiber maintains the electroreceptor in the skin of fish. Naturwissenschaften 72380-381. Roth, A,, and T. Szabo (1969) The effect of sensory nerve transsection on the sensory cells and on the receptor potential of the tuberous (Knollen) organ in mormyrid fish (Gnathonernus sp.) 2. Vergl. Physiol. 62r395410. Scott, S.A., E. Cooper, and J. Diamond (1981) Merkel cells as targets of the mechanosensory nerves in salamander skin. Proc. R. SOC.Lond. [Biol.] 211:455-470. Speidel, C.C. (1963) In vivo studies of myelinated nerve fibers. Int. Rev. Cytol. 16:173-231. Speidel, C.C. (1964) Correlated studies of sense organs and nerves of the lateral-line in living frog tadpoles. IV. Patterns of vagus nerve regeneration after single and multiple operations. Am. J. Anat. 114:133-160. Sperry, R.W. (1963) Chemoaffinity in the orderly growth of nerve fiber patterns and connections. Proc. Natl. Acad. Sci. U.S.A. 50:703-710. Szaho, T. (1974) Anatomy of the specialized lateral line organs of electroreception. In A. Fessard (ed): Electroreceptors and Other Specialized Receptors in Lower Vertebrates. New York: Springer-Verlag, 13-58. Szamier, R.B., and M.V.L. Bennett (1973) Rapid degeneration of ampulliuy electroreceptor organs after denervation. J. Cell Biol. 56:466477. Terzis, J.K., and R.W. Dykes (1980) Reinnervation of glabrous skin in baboons: Properties of cutaneous mechanoreceptors subsequent to nerve transection. J. Neurophysiol. 44: 1214-1225. Tosney, K.W., and L.T. Landmesser (1984) Pattern and specificity of axonal outgrowth following varying degrees of chick limb bud ablation. J. Neurosci. 4:25 18-2527. Wachtel, A.W., and R.B. Szamier (1969) Special cutaneous receptor organs of fish IV.Ampullary organs of the non-electric catfish, Kryptopterus. J. Morphol. 128t291-308. Weiss, P. (1939) Principles of Neurogenesis. New York: Henry Holt and Company. Wolpert, L. (1971) Positional information and pattern formation. In A.A. Moscona (ed): Current Topics in Developmental Biology, Vol. 6. New York: Academic Press. Yialamas, D., and H.H. Zakon (1984) Tuning of newly generated electroreceptors. SOC.Neurosci. Abstr. 10:193. Young, J.Z. (1949) Factors influencing the regeneration of nerves. Adv. Surg. 1:165-220. Zakon, H.H. (1986) The emergence of tuning in newly generated tuberous electroreceptors. J. Neurosci. 6t3297-3308. Zalewski, A. (1972) Regeneration of taste buds after transplantation of tongue and ganglia grafts to the anterior chamber of the eye. Exp. Neurol. 35.519-528. Zelena, J. (1984) Multiple axon terminals in reinnervated pacinian corpuscles of adult rat. J. Neurocytol. 133365-684.
Patterning in the Regeneration of Electroreceptors in the Fin o Kryp top terus MICHELE MILLER BEVER AND RICHARD B. BORGENS Department of Anatomy, Center for Paralysis Research, School of Veterinary Medicine, Purdue University, West Lafayette, Indiana 47907
ABSTRACT The influence of the target tissue on afferent nerve regeneration was studied in the adult glass catfish, Kryptopterus. In this fish, electroreceptors in the anal fin are distributed in a characteristic pattern in the proximal part of the fin and are absent in the distal portion of the fin. We tested whether axons were more likely to induce electroreceptors in certain regions of fin epidermis than in others. We rotated fin transplants so that the location of the degenerating electroreceptors was altered with respect to the regenerating axons in the host tissue dorsal to the fin. The effects of these rotations were observed in the living animal with differential interference contrast optics over a period of 10 weeks. When transplants were reversed rostrocaudally, new electroreceptors formed in the caudal half of the interradial zone, where degenerating electroreceptors were at the time of transplantation. When transplants were rotated so that the dorsoventral and rostrocaudal axes were reversed, some new receptors formed in the old target site regions that were located in the caudal interradial zones (in the distal half of the graft with respect to the host). Regenerating axons reached these regions of the transplant by taking unusual routes around the electroreceptor-free regions of fin. Very few electroreceptors formed in the distalicaudal or proximacaudal interradial quadrants of grafts where the original orientation of the tissue was maintained. We suggest that old target sites have a neurotropic influence on the regenerating afferent axons and discuss the possibility that the distal fin epidermis is not as permissive to electroreceptor formation as proximal fin epidermis. Key words: ampullary organs, catfish, lateral line nerve
Over 60 years ago, Ramon y Cajal ('28) postulated that regenerating nerve fibers were guided to their proper targets by "alluring or attracting substances." He suggested that growing nerve fibers were grossly attracted to the periphery where a more local and specific attracting mechanism was present to guide the fibers further to their terminal structures. Investigations into the nature of these guiding forces have continued to present times. In the chick embryo, developing motor neurons appear to be able to seek out their correct targets and make appropriate connections even when they or their targets in the limb have been shifted in the anterior-posterior axis of the embryo (Lance-Jones and Landmesser, '81).These authors suggested that a diffusable substance originating from the target was responsible for this directed growth. However, later studies provided evidence that motoneurons were capable of correct pathway selection even in the absence of their target (Tosney and Landmesser, '86; Phelan and Hollyday, '90). Neurotropic substances have also been implicated in the guidance of regenerating type I sensory axons to old mechanoreceptor sites (touch domes) after nerve transection in cats (Horch, '79, '82) and in the O 1991 WILEY-LISS, INC.
guidance of sprouting or regenerating mechanosensory axons to Merkel cells in the skin of salamanders (Diamond, '82). Here, we explore the possibility of axon guidance to targets of another sensory modality: electroreception. The electroresensory system of fish and amphibia plays an integral role in the animal's behavior, assisting in, among other things, feeding, orientation, and reproduction (Bullock, '74). Tuberous andor ampullary electroreceptor organs are distributed over the head and body in characteristic patterns that differ from species to species (Szabo, '74). In the nonelectric transparent catfish, Kryptopterus, the distribution of ampullary organs includes five longitudinal rows on each side of the trunk, a row along each of the pectoral, anal, and tail fin rays, and a scattering over the head and gills (Wachtel and Szamier, '69). These sensory receptors will atrophy in response to nerve transection (Bailey, '37; Roth and Szabo, '69; Szamier and Bennett, '73; Roth, '85), but will reform with the reinstatement of their innervation (Bailey, '37; Roth and Szabo, '69; Yialamas and Zakon, '84; Zakon, '86; Bever and Borgens, '91). Accepted March 29,1991.
219
PATTERNING IN RECEPTOR REGENERATION In our companion paper, we showed that after nerve transection, electroreceptorsregenerate in appropriate locations of an anal fin transplant that originally contained electroreceptors (Bever and Borgens, '91). Comparatively, when regenerating axons were faced with a fin transplant from a receptor-free region of fin, very few penetrated the graft and, in fact, most new electroreceptors formed in the proximal or peripheral zones (Bever and Borgens, '91). The latter results hint that the distal portion of the anal fin, which usually does not contain electroreceptors, is not permissive to their formation or does not possess the "cues" necessary to lead to their patterned development. In the present study, we used transplantation experiments to test the possibility that certain regions of the anal fin are attractive or inhibitory to the regenerating axons. Pieces of fin containing electroreceptors in characteristic locations were rotated and substituted for the normal fields of regenerating nerves. We show that although some nerve fibers induce receptors at new sites, other fibers will traverse substantial distances to induce electroreceptors in regions of old receptor sites.
distal portion of the graft (Fig. 1)and in the caudal half of the interray, relative to the regenerating axons dorsal to the graft. Therefore, in these grafts, an expanse of electroreceptor-free fin was interposed between regenerating axons projecting from the adjacent and dorsal body segment, and a region of fin graft originally containing electroreceptors. These electroreceptors would eventually degenerate and disappear since the nerves innervating them were severed during the transplantation procedure (see Bever and Borgens, '91). Moreover, the remnants of these axons were located at the distal margin of the graft, far from the regenerating nerves of the host.
Data analysis
Composite drawings. The grafts were observed weekly with DIC optics and were photographed with Kodak Technical Pan film. Tracings of the fin rays and electroreceptors in the grafts were made from photomicrographs and the axon pathways were superimposed after visually following axons through multiple planes of focus in the living animal. Analgsis. Newly formed electroreceptors in and around the grafts were classified according to their proximodistal position in the fin and placed in one of the following MATERIALS AND METHODS categories: peripheral zone (grafthost marginal region), Animals and observation proximal zone (portion of the graft nearest to the base of Adult transparent catfish, Kryptopterus (4-6 cm) were the fin; within 400 p m from the peripheral zone), middle maintained in fresh water aquaria at 25°C and fed brine zone (400 pm band in the central region of the graft), or shrimp. Observation techniques have been described previ- distal zone (portion of the graft furthest from the base of ously in our companion paper (Bever and Borgens, '91). the fin; greater than 800 p m from the peripheral zone) (see Briefly, fish were anesthetized in 0.05% tricaine methane Bever and Borgens, '91). Electroreceptors within a graft (all sulfonate (MS 222) and placed in a special viewing chamber zones except peripheral) were also classified according to (see Bever and Borgens, '91) that was suited for attachment their location within the interradial zone and placed in one to the rotating stage of a Nikon Diaphot inverted micro- of the following quadrants: proximal-caudal (quadrant 11, scope fitted with differential interference contrast (DIC) proximal-rostra1 (quadrant 21, distal-caudal (quadrant 31, optics. Wet gauze wrapped loosely around the head o f the or distal-rostral (quadrant 4).Electroreceptors that formed directly over a fin ray or in places where fin rays were fish kept the gills from drying during the viewing period. closely surrounding them, or exactly halfway between fin Surgical procedures rays, making rostral or caudal classification impossible, The surgical methods used have been described in detail were classified as undetermined (see Bever and Borgens, in the companion paper (Bever and Borgens, '91). In the '91). Regenerating axons induced electroreceptorsin the transpresent study, rectangular pieces (approximately2 mm by 4 mm) of caudal anal fin were removed by dissection and plants by projecting toward the target sites in one of two transplanted to a prepared site in the rostral part of the ways: 1) they projected into the interradial zones of the anal fin in two different orientations: 1) transplants in transplant in a proximal-distal direction; or 2) they trawhich the rostral-caudal axis was reversed (RC grafts); and versed a large expanse of fin either in the interradial zones 2) transplants in which the dorsal-ventral and rostral- or within the fin rays in a proximal-distal direction, and then reversed their direction of growth to project over 100 caudal axes were reversed (DV grafts). RC graft sites were prepared by removing a small rectan- pm in the opposite, distal-proximal direction. All axons gular piece of anal fin that contained electroreceptors and were classified by grouping them into one of these two portions of three fin rays (Fig. 1). Another piece of fin that categories of projection. The results of these grafting experiments were also contained electroreceptors was obtained from the proximal half of a more caudal region of anal fin in the same animal compared to prior results obtained from electroreceptorand was transplanted into this graft site. Before suturing, containing (EC) grafts, in which the position of receptors the donor tissue was rotated 180"around the dorsal-ventral within the fin was not disturbed after grafting, and electroreaxis. The resulting orientation of the graft tissue placed ceptor-free (EF) grafts (Bever and Borgens, '91). The degenerating electroreceptors in the caudal half of the Mann-Whitney,Student's t, and Chi-square tests were used interray with respect to the regenerating axons in the host for statistical comparisons. tissue dorsal to the graft (Fig. 11, instead of in the rostral interray (the normal position as observed in the remainder RESULTS of the fin). The transplanted tissue was sutured to intact fin RC grafts rays on either side of the graft with 10-0 Ethilon suture. DV grafts were prepared as above with the exception that In the rostral-caudal reversed grafts, the donor tissue the donor tissue was rotated 180"around a transverse axis (Figs. 1, 2A) was rotated so that electroreceptors and their (Fig. 1). In these double reversed cases, the resulting graft axons were relocated to the caudal half of the interray with orientation placed degenerating electroreceptors in the respect to the normal fin architecture (Figs. 1,2B).Because
220
Fig. 1. Diagram demonstrating rostral-caudal (RC)and dorsalventral (DW graft procedures. A rectangular piece of rostral anal fin containing electroreceptors was excised to create a graft site (G) with dimensions of approximately 2 mm by 1 mm. This process removed portions of three fin rays and also transected the nerves that normally project into the rostral half of the adjacent interrays. A piece of caudal anal fin (A) containing electroreceptors was transplanted into this graft site. Before suturing, the donor tissue was rotated one of two ways. The rotation placed electroreceptors in abnormal locations with respect to the host tissue and its regenerating nerve fibers. In the first case (RC),
M.M. BEVER AND R.B. BORGENS
donor tissue was rotated 180” around the dorsal-ventral axis so that electroreceptors in the graft appeared to be in the caudal half of the interrays with respect to the regenerating nerve fibers in the host. In the second case (DV), the donor tissue was rotated 180” around a transverse axis so that electroreceptors appeared to be in the caudal half of the interrays and in the distal half of the graft in relation to the regenerating nerve fibers in the host. Dorsal (D), ventral (V), rostral (R), and caudal ( C ) denotes orientation of host. d, v, r, and c denote original orientation of the transplant tissue.
the end of the fin rays normally slant caudally from in Table 1. Seventy-eight percent (mean of quadrants proximal to distal, a 180”rotation around the dorsal ventral 1 + 3, Table 2) of these newly formed receptors were axis resulted in fin rays within the graft slanting in the located in the caudal half of an interradial zone. In the absence of interference with the sensory innervaopposite direction relative to the intact fin rays surrounding them. Therefore, when the transplant was sutured into the tion, electroreceptors in a given interradial zone are supgraft site, it was difficult to align the proximal ends of plied by axons that enter the interray from the adjacent grafted fin rays with the distal ends of the host fin rays body segment dorsal to it (Bever and Borgens, ’91).Thus, in dorsal to the graft (Fig. 2B,C). In all of these animals the order to determine the effect of graft reorientation on axon host fin ray stumps and the fin rays within the transplant regeneration, the projection of the axon supplying each new itself alternated to some degree rather than aligning. electroreceptor within and around each RC graft was Sometimes host site fin ray stumps that were not aligned followed. An example of the axon paths for one RC graft is with graft rays regenerated distally a short distance to form shown in Figure 3. In this example, the axons regenerated an incomplete fin ray or “spike” (Fig. 2D) or grew even distally into the interradial zones by fairly direct pathways, further distally to produce an additional or supernumerary with only three axom growing into an adjacent interray. It new fin ray within the graft. Of the eight animals surgically is interesting to note that even though the fin rays of the manipulated in this way, seven completed 10 weeks of graft were not aligned with the fin ray stumps at the host observation and one died of unknown causes 1week after site, the two parts often appeared to fuse along their overlapping edges (Fig. 2D). Fin rays are comprised of two surgery. One week after transplantation, all RC grafts were viable opposing pieces, one beneath the epidermis of each side of and well vascularized. However, the electroreceptors within the fin. Some of the axons shown in Figure 3 grew between the grafts had degenerated since being denervated at the the paired components of the fin rays in this fused region to original surgery (Fig. 2C). By 2 weeks posttransplantation, enter the interray while other axons grew around the new receptors had appeared in all seven grafts, and had proximal end of a graft fin ray to enter an adjacent interray. greatly increased in number by 10 weeks after surgery (Fig. All 114 axons traced in RC grafts had induced electrorecep2D). A summary of the number of electroreceptors and tors after projecting to the sites in a proximal-to-distal their proximodistal positions within each graft is presented direction.
221
PATTERNING IN RECEPTOR REGENERATION
DV grafts
tors appeared near the proximal grafthost tissue boundary. During the fourth to tenth weeks after surgery, additional In the dorsal-ventral reversed grafts, the transplant receptors appeared at the proximal interface and in more tissue (Fig. 4A) was rotated so that degenerating electrore- distal locations in nine of the transplants (Fig. 4D). A ceptors were in the distal half of the graft and in the caudal summary of the number and proximodistal distribution of half of the interradial zone with respect to regenerating new receptors within each DV graft is shown in Table 3. axons penetrating the graft from the more dorsal body The rostral or caudal interray location of electroreceptors segment (Fig. 4B). After reorientation, the distal segments was determined for all DV grafts. As mentioned above, of axons innervating these electroreceptors extended away electroreceptors in the normal animal are typically found in from them toward the distal grafthost margin. Of the 17 the rostral half of each interradial zone. Thirty-one percent animals receiving this treatment, four lost their grafts (mean of quadrants 2 4, Table 2) of electroreceptors within the first week and three died of unknown causes. Of within all ten DV grafts were in such a “normal” locathe remaining ten animals, seven survived an observation tion-in the rostral half of an interradial zone. Fifty-nine period of greater than 9 weeks, while the other three fish percent (mean of quadrants 1 + 3) of all electroreceptors lived 4 , 5 , and 7 weeks. were located in the caudal interray (Table 2). However, The DV grafts were viable and appeared to be well whether a new electroreceptor appeared in the rostral or vascularized at 1 week after transplantation. As in RC caudal half of an interradial zone depended upon whether it grafts, endogenous electroreceptors had degenerated and was in the proximal or distal half of the transplanted fin were no longer visible within the transplant (Fig. 4 0 . New (Table 2). In the proximal half of the transplants, the mean receptors began appearing within the DV grafts at 2 to 3 percent of new receptors that formed in the rostral weeks after surgery. In one graft, the first electroreceptors (“normal”) portion of an interradial zone was 28.0%(quadappeared in the distal end of the graft, while in two rant 2). A similar mean of 28.8% formed in the caudal additional grafts the first electroreceptors appeared simul- interradial zone (quadrant 1).In the distal half of the DV taneously at the proximal boundaries of the grafts and in grafts, a mean of 29.8%of the receptors was in the caudal distal regions of the transplants. In a fourth graft, electrore- half of the interradial zone (quadrant 3), while a mean of ceptors never appeared in the 10 weeks of observation. In only 3.4% formed in the rostral half (quadrant 4). the remainder of the grafts (6/10),the first new electrorecepA qualitative analysis of the pattern of innervation of new electroreceptors in the DV grafts revealed that it was not uncommon for the axons to grow between the paired TABLE 1. Number and Distribution of Newly Formed components of the fin rays (Fig. 5 ) . In several cases (11 Electroreceptors in RC Grafts axons in five different grafts), these axons would grow Proximodistal position in fin distally inside the fin ray for the entire width of the graft Middle Distal and then exit the fin ray and turn through 180“ to grow Peripheral Proximal Total Animal proximally for some distance exceeding 100 km (at times zone zone zone zone no. receptors 400-500 km) before inducing the formation of an electrore7 4 15 RC-1 26 ceptor (Fig. 5). Thirty-three percent of the total number of 7 5 2 RC-2 14 4 11 3 2 RC-3 20 axons traced in DV grafts approached their target sites RC-4 24 3 17 5 from a distal-to-proximal direction; all 14 of these axons 6 12 4 RC-5 29 7 24 1 4 RC-6 29 supplied electroreceptors that were located in the distal 11 3 RC-8 14 half of the transplant.
+
TABLE 2. Summary of Electroreceptor Location in Grafts Quadrant’ Graft
Animal
EC’
EF’ RC
DV
1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10
3
4
N3
Undetermined’
Peripheral4
76.6 51.2
3.1 4.8
9.2
T = 85 T = 30
0 0
23.1
0.6 5.2 26.9 7.1 5.0 12.5 0 3.4 28.6 11.9 0 28.6 0 0 5.3 17.9 0 0 5.9 0 5.8
14.3 69.6 15.4 0 20.0 8.3 24.1 82.8
1
2
Mean 0.8 Mean 29.8 100.0 76.9 73.3 78.9 59.1 0
90.0 Mean 68.3 50.0 0 0 0
100.0 18.2 100.0
20.0 0 0 Mean 28.8
13.3 15.8 9.1 75.0 10.0 17.6 50.0 75.0
0 0
5.3 13.6 25.0 0
9.6 0
25.0
0
0
20.0 0 54.5 0
80.0 0 18.2
60.0 0
20.0 75.0 80.0 29.8
20.0 28.0
0
‘Numberofelectroreceptors In each quadrant #ven as a prrwntage of N ’Only the summnry data has been reproduced from Table 4 , Uever and Rorpn3 ‘91 ’N.tuwl receptors in graft thar could beraregorired ‘Suntber uf electroreceptor:. g i v m sm a prrcentege of t hr total wcepmrs.
0 0 0
13.3 0.0 18.2 0 0 4.5 0 0 0 0 0
9.1 0 0 25.0 0 3.4
15 13 15 19 22 4 10 T = 98 2 4 0
5 1 11 1 5
4 10 T = 43
0
21.5 0
14.3
Total receptors T = 103 T = 106 26 14 20 24 29 29 14 T = 156 2 7
0
0
68.8 89.5 42.9 92.3 72.2 70.6 41.2 49.2
16 19 28 13 18 17 17 T = 137
M.M. BEVER AND R.B. BORGENS
222
Figure 2
PATTERNING IN RECEPTOR REGENERATION
Figure 2 continued
223
224
M.M. BEVER AND R.B. BORGENS
Fig. 3. A composite drawing of newly formed electroreceptors and their respective innervation in an RC graft. Same graft as presented in Figure 2. All of the new electroreceptors in this graft were located along the rostral aspect of the fin rays (with respect to host tissue) except for
one (arrowhead), which could not be classified as rostral or caudal. Arrows indicate axons that grew into neighbor interrays. Broken lines signify portions of axons that grew between the paired components of the fin rays.
DISCUSSION
receptors with respect to the regenerating nerves. The results suggest that regions once containing electroreceptors may somehow influence the destination of the ingrowing axons.
In this study, we tested the ability of old receptor sites to attract regenerating nerve fibers by rotating the fin transplant, thus altering the relative location of the degenerating
Patterning in axonal regeneration Fig. 2. (pages222-223) Regeneration of electroreceptors in an RC graft. A Donor region of the caudal anal fin to be transplanted to a new site in the rostral anal fin. Severed fin rays at the distal margin of the presumptive graft are marked with arrows. The graft margins that remain to be cut are delimited by dashed lines. Note that electroreceptors are located only along the caudal aspect of each fin ray. B: Donor tissue (same tissue as in A) sutured (arrows) to intact fin rays on either side of the graft site. Transplant was rotated 180" around the dorsalventral axis before suturing so that electroreceptors (arrowheads) within the graft were relocated along the new rostral aspect of the fin rays with respect to the surrounding host tissue. Note that the proximal ends of the transplant fin rays are not aligned with the fin ray stumps of the host. C: The same graft as presented in B at 1week posttransplantation. The graft margins have healed and the graft is well vascularized, but the electroreceptorsthat were within thegraft have disappeared. D: Ten weeks after grafting, electroreceptors have reappeared within the graft tissue. In this animal, all of the new receptors appeared along the rostral aspect of graft fin rays (compare to normal electroreceptors outside of the graft (arrows). Two of the host fin ray stumps have regenerated bony spicules within the graft tissue (arrowheads). Scale bar for A, B, C, D = 0.5 mm.
A curiosity in the regeneration of peripheral nerve fibers is the reestablishment ofpattern in the nerve field (Ramony Cajal, '28; Weiss, '39; Guth, '57; Sperry, '63; Diamond, '82). Sensory fibers supplying muscle spindles, taste buds, lateral line organs, Merkel cells, touch domes, pacinian corpuscles, or electroreceptors are somehow "directed" to the correct region of skin or muscle to renew contact with, or induce the formation of, their sensory end organs (Speidel, '64; Zalewski, '72; Burgess et al., '74; Brown and Butler, '76; Diamond, '82; Zelena, '84; Bever and Borgens, '91). One explanation for the return of regenerating nerve fibers to specific foci is that they have access t o empty Schwann tubes that could lead them directly to previously innervated sites (Young, '49; Burgess et al., '74). This notion is well supported in nerve crush cases (Horch, '79; Cheal and Oakley, '77; Zelena, '84; Diamond et al.,'87) and may also be relevant to the projection pattern after nerve trunk transection (Ramon y Cajal, '28; Speidel, '63, '64; Terzis and Dykes, '80). However, there are examples of
225
PATTERNING IN RECEPTOR REGENERATION
Regeneration of electroreceptors within transplants
TABLE 3. Number and Distribution of Newly Formed Electroreceutorsin DV Grafts
Proximodistal position in fin
The RC and DV transplantation experiments with catfish fin have provided results that are most readily interpreted in terms of attracting properties of the dermis or epidermis DV-12 2 2 in regions of old receptor sites. In RC grafts, reinnervation DV-13 7 1 5 1 of transplants occurred according to the original orientaDV-14 0 tion of the graft prior to transplantation. After the transDV-19 16 11 1 1 3 DV-23 19 17 1 1 planted tissue was rotated at the time of surgery, degeneratDV-24 28 12 5 8 3 ing electroreceptorswere in the caudal half of the interradial DV-25 13 12 1 DV-26 18 13 4 1 zones with respect to the host. The terminal segments of DV-27 17 12 1 2 2 their axons extended dorsally toward the proximal graft/ DV-28 16 7 2 5 2 host interface. Weeks later, the majority of the newly formed receptors were in that same region, that is, in the caudal half of the interradial zones with respect to the host. A comparison of the mean percent of receptors occurring in nerve transections in which regenerating nerve fibers show quadrant 1in RC and EC grafts (in which no manipulation a strong predilection to return to specific sites but do not of receptor location occurred at the time of graft transplanreach them via “preexisting pathways” (Speidel, ’64; Horch, tation) is shown in Figure 6. Receptors were more likely to form in quadrant 1 in RC grafts than in EC grafts ’79, ’82; also see below). Patterned regeneration can also be viewed in terms of (P < 0.005). These results could possibly be explained by positional information. Wolpert (’71) suggested that migrat- the capability of regenerating axons to find old Schwann ing “cells would ‘test’ the postional information of the cells tubes that then direct them to old target sites. One way to over which they were moving and would stop at that value eliminate this possibility is to create a situation in which for which they were programmed.” One proposed mecha- there are degenerating receptors, but no preexisting fascicnism €or conferring such a positional cue to growing nerves ulation pathways accessible to regenerating nerves. This involvesa differential adhesiveness of the substrate; growth situation was produced in the DV grafts. In DV grafts, transplanted tissue was rotated so that would continue in the direction of increasing adhesivity (Nardi and Kafatos, ’76). Nardi (’83)has demonstrated a degenerating electroreceptors were in the distal portion of proximodistal asymmetric distribution of matrix molecules the graft and in the caudal half of the interray (quadrant 3) in the moth wing that could be involved in the guidance of with respect to the regenerating axons dorsal to the graft. developing neuron processes proximally toward their tar- Figure 7 compares the mean percent of receptors that gets. In this example, axons actually avoided grafts with the formed in quadrant 3 of DV grafts with the mean percent that formed in the same quadrant in EC grafts. Receptors “incorrect” proximodistal orientation. were more likely to form in quadrant 3 in DV grafts than in A third explantion for reestablishment of pattern is that the target sites are neurotropic (Ram6n y Cajal, ”28;Horch, EC grafts (P < 0.05). Indications of target site attraction properties are re’79; Diamond, ’82; English et al., ’83). In this view, axons are directed to appropriate regions of skin or even specifi- vealed by comparing the results of dorsoventrally (DV) cally to old sites of innervation because of a specific rotated grafts with the results of the electroreceptor-free attraction emanating from those sites. Cells of the target (EF) grafts (Bever and Borgens, ’91). These two types of site could presumably release an attractant or “sprouting grafts are similar in that a portion of the transplant, a band factor,” the production of which could be inhibited by about 600 pm wide closest to the regenerating axons uptake or neutralization after the successful establishment emanating from the trunk of the fish, is initially free of of new nerve connections (Ram6n y Cajal’ ’28; Diamond, receptor organs. The two types of grafts differ, however, in ’82). Scott and colleagues (’81) have demonstrated that that electroreceptors and their terminal axon segments Merkel cells persist when deprived of their innervation. existed in the distal half of DV grafts after rotation and there were no such structures in the EF grafts. If we These cells subsequently become innervated by sprouting compare the innervation patterns and the locations of new or regenerating nerve fibers. Furthermore, new Merkel cells appear in regenerating skin in the absence of innervation. They become innervated when the nerves regenerate. This latter result supports the notion that the Merkel cells Fig. 4. (pages 226-227) Regeneration of electroreceptors in a DV are neurotropic, since there were no old nerve pathways for graft. A: Donor tissue in the caudal anal fin. The margins of the the regenerating axons to follow to these targets, but does presumptive transplant are delimited with dashed lines. Electrorecepnot exclude the possibility that innervation was random tors are found only along the caudal aspect of each fin ray. B: Donor and that axons unsuccessful in finding targets did not tissue transplanted into the graft site. Before suturing, the transplant was rotated 180” around a transverse axis so that electroreceptors survive. In another system, Horch (’82) demonstrated that (arrowheads) within the graft appear to be in the caudal half of the uncauterized touch domes in the cat were more likely to be interradial zone and in the distal half of the graft with respect to the reinnervated after nerve transection than ones that had orientation of the host. C: Same graft as in B, 1week after transplantabeen cauterized. Since regenerating axons in both cases tion. Graft is viable and well vascularized. Original electroreceptors are would have access to degenerate perineural tubes, and since absent. D: Same graft as B, 10 weeks after transplantation. Receptors have formed at the dorsal margin of the graft and in the distal half of there was not equal reinnervation of the two groups, he the graft. All of the new receptors (arrowheads) in the distal half of the concluded that the growing axons were influenced by some graft are located on the rostra1 aspect of the fin ray (in the caudal half of intrinsic property of the unlesioned target sites. an interray with respect to the host). Scale bar for A, B, C, D = 0.5 mm. ~
Animal no. ~
Total receutors
Peripheral zone
~
Proximal zone
Middle zone
Distal zone
M.M. BEVER AND R.B. BORGENS
226
Figure 4
PATTERNING IN RECEPTOR REGENERATION
Figure 4 continued
227
M.M. BEVER AND R.B. BORGENS
228
Figure 5
PATTERNING IN RECEPTOR REGENERATION receptors in these two types of grafts, we find three outstanding differences between them: 1. Very few axons (mean 4.8%,quadrants 3 + 4; Table 2) grew into the distal region of the EF grafts to induce electroreceptors, whereas one-third (mean 33.2%;Table 2) of the axons grew into distal regions of DV grafts and induced electroreceptors(P < 0.05, Mann-Whitneytest). 2. The pattern of newly formed receptors in the distal half of the DV transplants was reestablished according to the original position of these structures within the transplant tissue, that is, a mean 29.8%(quadrant 3, Table 2) of the new receptors in the graft were in the distalcaudal quadrant of an interradial zone with respect to the host (in the same regions where degenerating receptors were prior to rotation at the time of surgery). A mean of only 3.4% of the new receptors were in the distal-rostral quadrant (quadrant 4) of an interray (P < 0.05, Mann-Whitney test). On the other hand, in the proximal half of the DV grafts, the mean percent of new receptors was similar for the rostral (quadrant 2, 28%)and caudal (quadrant 1,28.8%)half of the interradial zones (P = 0.7207, Mann-Whitney test). EF grafts displayed a greater mean percent of new electroreceptors in the proximal-caudalquadrant (quadrant 2) of the interradial zones. A mean of 51%of new receptors were in quadrant 2 of an interradial zone, while a mean of 30%of new receptors formed in quadrant 1. 3. Although the number of axons that invaded each transplant was similar for EF and DV grafts (P > 0.20, Mann-Whitney test), a considerably larger proportion (14/43) of the axons in DV grafts made their final approach to new receptor sites in a distal-to-proximal direction, compared to only a few (2/36) in the EF grafts (P < 0.01, Chi-square test). Proximal-distal location of new electroreceptors. It is important to note that the proximal portions of the DV and EF grafts were nearly identical, except that the proximodistal orientation of DV grafts was altered by 180”. Keeping this in mind, regenerating axons in each case were faced with “foreign” electroreceptor-free fin in the proximal half of the grafts. It appears, though, that axons found the distal half of the EF grafts less congenial or less attractive than the distal half of DV grafts. Only 5% of new receptors formed in the distal half of the EF grafts compared with 33%in DV grafts (Table 2). To state this another way, the distal portion of the DV grafts was attractive enough that axons crossed substantial distances, penetrating a proximal region of electroreceptor-free fin, to reach this area. This distal region once contained electroreceptors; however,
229 there were no Schwann tubes extending proximally to guide the axons from the dorsal grafthost interface to the distal half of the transplant. Pattern of new electroreceptors. If we consider the influence of the proximal or distal half of the DV grafts on the pattern of the new electroreceptors, we find that in the distal half the receptors tended to form in a location reminiscent of their original orientation prior to transplantation, while in the proximal half, receptors showed no tendency to form in a particular region of interray. These results are consistent with the concept that old receptor sites have attracting properties. Axon pathways to target sites. A striking number of axons (33%)that induced electroreceptors in the DV grafts reached their targets via “unusual” routes. In addition to these axons growing first distally for long distances and then proximally for at least 100 pm, 11 of these axons also traversed the fin by growing between the paired components of fin rays. The central region of a fin ray contains nerves and blood vessels (Goss, ’69).Although one might
100
coal-
A A
80
8
OF3
60
A
!?
0
$3 -u 4-3 CCS
8 .E
z
40
a 20
Fig. 6. Percent of electroreceptors in quadrant 1 (shaded area) in EC and RC grafts. Mean f S.E.M. was 0.8 f 0.8 and 68.3 f 12.4, respectively. Differences between EC and RC grafted animals: P = 0.0026 with the Mann-Whitney test and P = 0.0001 with the t test.
90
AA ?!!
0
c
a m
a*
8:
60
z’op - 3
Fig. 5. Composite drawings o f newly formed electroreceptors and their respective innervation in DV grafts. A Same graft as shown in Figure 4. Four of the new eledroreceptors (large arrowheads) in the distal half of the graft were induced by axons that grew between (broken lines) the paired components of a fin ray to the distal margin of the transplant before exiting and growing back proximally for several hundred micrometers (small arrowheads). B : Composite drawing of a second DV graft. Three of the axons (arrows) that innervated new electroreceptors grew between (broken lines) the paired components of the fin rays to the distal margin of the graft, where they then exited and grew proximally along the rostral edge of the fin ray. Of these three axons, two induced electroreceptors on the rostral aspect of a fin ray (large arrowheads) while the third axon crossed the interradial zone and induced an electroreceptor on the caudal aspect of a fin ray (small arrowhead).
C
0,
0
z
U
.r
30
a
EC
DV
Fig. 7. Percent of receptors in quadrant 3 (shaded area) in EC and DV grafts. Mean -+ S.E.M. was 3.1 f 2.0 and 29.8 ? 11.0, respectively. Differences between EC and DV grafted animals: P = 0.0361 with the Mann-Whitney test andP = 0.0323 with the t test.
M.M. BEVER AND R.B. BORGENS
230 argue that the intrinsic nerves of the fin ray could serve to direct regenerating axons to the distal end of the DV grafts, there are two lines of evidence that do not support this view. First, in RC grafts, fin rays of the transplant were frequently misaligned with the fin ray stumps of the host site (see Results). This misalignment placed fin rays of the transplant in an excellent position to be penetrated by regenerating axons from the adjacent body segment. However, we did not observe one example of an axon growing through the center of a fin ray in these grafts. Second, even though the proximal portions of the EF and DV grafts were similar in that neither contained any remnant of previous afferent electroreceptor innervation, in E F grafts there were only seven axons that grew between fin rays, and only two of those seven made it as far as the distal half of the transplant (neither of these two traversed the entire width of the transplant). This can be compared to DV grafts where 11 axons grew inside fin rays for the entire width of the graft and induced electroreceptors in the distal portion of those grafts. In summary, we have shown that although the preexistence of receptors is not required for their induction, axons have a propensity for inducing receptors in regions of old receptor sites. This may, in part, be due to the use of degenerate perineural tubes that lead the regenerating axons directly to the previous target sites. However, we also show evidence that, in the absence of such “pathways,” axom will traverse great expanses of fin to induce electroreceptors in regions of the fin that previously contained receptors. We suggest from these results that an intrinsic property of the regions of fin that previously contained receptors is responsible for the guidance of regenerating axons to those regions.
ACKNOWLEDGMENTS This work was supported by NIH grant R 0 1 HD 20664 (R.B.B.), and the David Ross Fellowship Program of Purdue University (M.M.B.). We wish to thank Mary Jo Maslin for manuscript preparation and Barbara Sturonus for the artwork. We thank J.W. Vanable, Jr., A.R. Blight, and M.E. McGinnis for their helpful critique of this manuscript.
LITERATURE CITED Bailey, S.W. (1937) An experimental study of the origin of lateral-line structures in embryonic and adult teleosts. J. Exp. Zool. 76:187-233. Bever, M.M., and R.B. Borgens (1991) The regeneration of electroreceptors in Kryptopterus. J. Comp. Neurol. 309:ZOO-217. Brown, M.C., and R.G. Butler (1976) Regeneration of afferent and efferent fibers to muscle spindles after nerve injury in adult cats. J. Pbysiol. (Lond.) 260:253-266. Bullock, T.H. (1974) General introduction. In A. Fessard (ed): Electroreceptors and Other Specialized Receptors in Lower Vertebrates. New York: Springer-Verlag, pp. 1-12. Burgess, P.R., K.B. English, K.W. Horch, and L.T. Stensaas (1974) Patterning in the regeneration of the type I cutaneous receptors. J. Physiol. 236.57-82. Cheal, M., and B. Oakley (1977) Regeneration of fungiform taste buds: Temporal and spatial characteristics. J. Comp. Neurol. 172r609-626. Diamond, J. (1982) Modeling and competition in the nervous system: Clues from the sensory innervation of the skin. In R.K. Hunt (ed): Current Topics in Developmental Biology, Vol. 17. New York: Academic Press, pp. 147-205. Diamond, J., M. Coughlin, L. Macintyre, M. Holmes, and B. Visheau (1987) Evidence that endogenous nerve growth factor is responsible for the
collateral sprouting, but not the regeneration, of nociceptive axons in adult rats. Proc. Natl. Acad. Sci. U.S.A. 84:65966600. English, K.B., D. Kavka-Van Norman, and K. Horch (1983) Effects of chronic denervation i n type I cutaneous mechanoreceptors (Haarscheiben). Anat. Rec. 207:79-88. Goss, R.J. (1969) Principles of Regeneration. New York: Academic Press. Guth, L. (1957) The effects of glossopharyngeal nerve transection on the circumvallate papilla of the rat. Anat. Rec. 128t715-731. Horch, K. (1979) Guidance of regrowingsensory axons after cutaneous nerve lesions in the cat. J. Neurophysiol. 421437-1449, Horch, K. (1982) The influence of mechanoreceptor structures on regenerating sensory axons after cutaneous nerve transection in the cat. Neurosci. Lett. 32:281-284. Lance-Jones, C., and L. Landmesser (1981) Pathway selection by embryonic chick motoneurons in a n experimentally altered environment. Proc. R. SOC. Lond. [Biol.]214:19-52. Nardi, J.B. (1983) Neuronal pathfinding in developing wings of the moth Manduca sexta. Dev. Biol. 95t163-174. Nardi, J.B., and F.C. Kafatos (1976) Polarity and gradients in lepidopteran wing epidermis 11. The differential adhesiveness model: Gradient of a non-diffusible eel1 surface parameters. J. Emhryol. Exp. Morpbol. 36r489512. Phelan, K.A., and M. Hollyday (1990) Axon guidance in muscleless chick wings: The role of muscle cells in motoneuronal pathway selection and muscle nerve formation. J. Neurosci. lOr2699-2716. Ram6n y Cajal, S. (1928) Degeneration and Regeneration of the Nervous System. (R.M. May, trans.). New York: Hafner Publishing Co., (reissued 1959). Roth, A. (1985) Axonal flow in the afferent fiber maintains the electroreceptor in the skin of fish. Naturwissenschaften 72380-381. Roth, A,, and T. Szabo (1969) The effect of sensory nerve transsection on the sensory cells and on the receptor potential of the tuberous (Knollen) organ in mormyrid fish (Gnathonernus sp.) 2. Vergl. Physiol. 62r395410. Scott, S.A., E. Cooper, and J. Diamond (1981) Merkel cells as targets of the mechanosensory nerves in salamander skin. Proc. R. SOC.Lond. [Biol.] 211:455-470. Speidel, C.C. (1963) In vivo studies of myelinated nerve fibers. Int. Rev. Cytol. 16:173-231. Speidel, C.C. (1964) Correlated studies of sense organs and nerves of the lateral-line in living frog tadpoles. IV. Patterns of vagus nerve regeneration after single and multiple operations. Am. J. Anat. 114:133-160. Sperry, R.W. (1963) Chemoaffinity in the orderly growth of nerve fiber patterns and connections. Proc. Natl. Acad. Sci. U.S.A. 50:703-710. Szaho, T. (1974) Anatomy of the specialized lateral line organs of electroreception. In A. Fessard (ed): Electroreceptors and Other Specialized Receptors in Lower Vertebrates. New York: Springer-Verlag, 13-58. Szamier, R.B., and M.V.L. Bennett (1973) Rapid degeneration of ampulliuy electroreceptor organs after denervation. J. Cell Biol. 56:466477. Terzis, J.K., and R.W. Dykes (1980) Reinnervation of glabrous skin in baboons: Properties of cutaneous mechanoreceptors subsequent to nerve transection. J. Neurophysiol. 44: 1214-1225. Tosney, K.W., and L.T. Landmesser (1984) Pattern and specificity of axonal outgrowth following varying degrees of chick limb bud ablation. J. Neurosci. 4:25 18-2527. Wachtel, A.W., and R.B. Szamier (1969) Special cutaneous receptor organs of fish IV.Ampullary organs of the non-electric catfish, Kryptopterus. J. Morphol. 128t291-308. Weiss, P. (1939) Principles of Neurogenesis. New York: Henry Holt and Company. Wolpert, L. (1971) Positional information and pattern formation. In A.A. Moscona (ed): Current Topics in Developmental Biology, Vol. 6. New York: Academic Press. Yialamas, D., and H.H. Zakon (1984) Tuning of newly generated electroreceptors. SOC.Neurosci. Abstr. 10:193. Young, J.Z. (1949) Factors influencing the regeneration of nerves. Adv. Surg. 1:165-220. Zakon, H.H. (1986) The emergence of tuning in newly generated tuberous electroreceptors. J. Neurosci. 6t3297-3308. Zalewski, A. (1972) Regeneration of taste buds after transplantation of tongue and ganglia grafts to the anterior chamber of the eye. Exp. Neurol. 35.519-528. Zelena, J. (1984) Multiple axon terminals in reinnervated pacinian corpuscles of adult rat. J. Neurocytol. 133365-684.
