Timing the commitment to a wound-healing responseof denervated limb stumps in the adult newt,Notophthalmus viridescens

Judith D. Salley-Guydon, PhDn; Roy A. Tassava, PhD

Department of Molecular Genetics, The Ohio State University, Columbus, Ohio

Reprint requests:Roy A. Tassava, PhD, Department of

Molecular Genetics, The Ohio State

University, 484 W. 12th Avenue, Columbus,

OH 43210.

Fax: 1 (614) 292-4466;

Email: tassava.1@osu.edu

nPresent address: Judith Salley-Guydon,

PhD, Department of Biological and Physical

Sciences, South Carolina State University,

Orangeburg, SC.

Manuscript received: January 19, 2006

Accepted in final form: February 15, 2006

DOI:10.1111/j.1743-6109.2006.00154.x

ABSTRACT

Adult newt limbs that are denervated 1 day after amputation undergo a wound-healing response and, although they become reinnervated, will not regenerateunless reinjured. Experiments were designed to determine when denervated limbstumps of adult newts become committed to a wound-healing response. In Ex-periment I, limbs were amputated and denervated 1 day after amputation. Ondays 7, 14, 21, 28, and 35 days after the initial amputation and denervation,stumps were reamputated to remove the distal tip. This design varied the time thedistal stump was devoid of nerves before reamputation. None of the limbs ream-putated at 7 days regenerated. About half of the limbs reamputated at 14 daysregenerated and almost all of those reamputated at days 21, 28, and 35 regener-ated. In Experiment II, limbs were denervated and then amputated on day 7 or14, at two different levels. Limbs with a short stump became innervated earlierand regeneration occurred more frequently at both levels of amputation. The re-sults of these experiments show that denervated limb stumps become committedto a wound-healing response between days 7 and 14 after amputation/denerva-tion. If sufficient nerves arrive before day 7, a regeneration response is initiated. Ifthe stump is denervated for 14 days or longer, commitment to wound healingoccurs and ingrowing nerves cannot initiate a regeneration response.

Urodele limb regeneration is characterized by the forma-tion of the blastema, a bud-like outgrowth of undifferen-tiated cells at the amputated end of the limb.1,2 Formationof the blastema depends on a threshold supply of nerves3

and a functional wound epithelial covering.4,5 If a limb isdenervated on the day of amputation or any time before ablastema has formed, regeneration does not occur.6–8

Nerves will ultimately grow back into a denervated limbstump and reach the original level of amputation. In Am-bystoma larvae, as ingrowing nerves reach the level of am-putation, cell division resumes, blastema formationoccurs, and the limb regenerates.7,9 In contrast, in theadult newt, regeneration is not initiated by ingrowingnerves.3

In an attempt to define the factors that prevent limbstumps of adult newts from regenerating upon reinnerva-tion, Salley and Tassava10 reinjured the ends of previouslyamputated, denervated, now reinnervated limb stumps onday 35 after amputation/denervation in various ways. Re-generation capability was restored by reamputation of thedistal 1mm of the healed stump (100% of the cases), a sin-gle razor incision (60% of the cases), and removal by dis-section of the healed tissues from the end of the stump(70% of the cases). It was concluded that these experi-mental manipulations reestablished the opportunity forinteractions between the important prerequisites for re-generation: injury, wound epithelium, and nerves.10

The denervated limb stump of the adult newt thus be-comes committed to a wound healing response at some

time before reinnervation so that suitable conditions forregeneration no longer exist. Here, we investigated thetiming of that commitment to wound healing. Experi-ments were designed to answer the following questions:Can regeneration of a completely denervated adult newtlimb be initiated without reinjury? When does a denervat-ed limb stump in the adult newt become committed to awound-healing response?

MATERIALS AND METHODS

Adult newts, Notophthalmus viridescens, of unknown ages,were collected from ponds in southern Ohio and main-tained on raw beef liver in the laboratory in aerated waterat 22–23 1C. All surgeries were performed while newts wereanesthetized in MS222 (ethyl m-amino methane sulfonate;Sigma-Aldrich, St. Louis, MO). Amputations were per-formed with a sterile razor blade. Protruding bone and softtissues were immediately trimmed to provide a level am-putation surface. This animal research was carried out un-der The Ohio State University Animal Research Protocol# 99A0111.

H Humerus

R/U Radius/ulnar

Wound Rep Reg (2006) 14 479–483 c� 2006 by the Wound Healing Society 479

Wound Repair and Regeneration

Experimental protocol I

Previously, we showed that amputated, once-denervated,adult newt limbs will not regenerate but if these limbs arereamputated on day 35 after the original amputation, re-generation commences in 100% of the cases.10 Here, weinclude reamputations beginning 7 days after the initialamputation/denervation. Left and right limbs of 102 newtswere amputated through the mid-radius/ulna (mid-R/U).One day after amputation, the left limbs were completelydenervated by cutting the nerves at the brachial plexus.Care was taken to not cut the brachial artery. By delayingdenervation, 1 day soft-tissue resorption is prevented andwound epithelial covering is enhanced. The right limbswere sham operated without cutting the brachial nervesand served as innervated controls.

On each of days 7 (18 newts), 14 (28 newts), 21 (14newts), 28 (12 newts), and 35 (20 newts) after the originalamputation, the left and right limbs were reamputated byremoving 1mm of the distal stump. The reamputationprovided a new injury and a fresh wound epithelium. Therationale was that if nerves had returned to the distalstump in sufficient quantity, then the reamputated limbshould have regenerated. Limbs were observed for thepresence or absence of regeneration for a period of 10weeks.

Histological procedures

To assess the degree of nerve ingrowth, at each of theabove reamputation days, two of the denervated left limbsand two innervated right limbs were sampled for nervestaining (10 newts total). The limbs were fixed in Bouin’sfluid, embedded in paraffin, sectioned at 10mm, and ana-lyzed for nerves using Samuel’s nerve stain.11 The degreeof nerve ingrowth was also assessed by testing the sensitiv-ity of limb stumps to touch, by examining the angle of thelimb in relation to the body, and by observing movement.

Experimental protocol II

In Experiment II, left limbs of 24 newts were completelydenervated as above but limbs were not amputated. Theright limbs were sham operated and served as innervatedcontrols. On days 7 (14 newts) and 14 (14 newts) after den-

ervation, both right and left forelimbs were amputated. Sixof the 16 limbs were amputated through the mid-R/U,while the other six limbs were amputated through the mid-humerus (mid-H). Protruding bone and soft tissues weretrimmed level. Thus, nerves had an opportunity to rein-nervate the limb for either 7 or 14 days before amputation.Utilizing the two amputation levels provided either a shortstump or a long stump for reinnervation. The 24 limbswere observed through a period of 10 weeks for the pres-ence or absence of regeneration. Four limbs from Experi-ment II at the H-level were stained for nerves (two at day 7and two at day 14) and were also observed for sensitivity,position, and movement. Experiment II controls includedsix newts, whose left limbs were amputated through themid-R/U and denervated 1 day after amputation and sixnewts, whose left limbs were amputated through the mid-H and denervated 1 day after amputation. Right limbswere amputated throught the mid-R/U or mid-H, respec-tively, as normally regenerating controls.

RESULTS

None of the 18 amputated, denervated, left limb stumpsreamputated 7 days after the original amputation regener-ated (Table 1). All of the right, control limbs regenerated.Reamputation on day 14 resulted in 43% of the limbs re-generating but only after a delay of 1 week compared withthe control, right limbs, all of which regenerated. Whenreamputation was 21 days after the original amputation,86% of the limbs regenerated, with only a 2–3-days delaycompared with the right, control limbs, all of which regen-erated. Reamputation at 28 days resulted in 92% of thelimbs regenerating, and without delay compared with con-trol, right limbs. Reamputation on day 35 resulted in100% regeneration with no delay compared with the con-trol, right limbs, all of which regenerated (Table 1).

When limbs were amputated at the proximal-H level 7days after denervation, regeneration occurred in five of sixlimbs (Table 2). A blastema outgrowth was first observedat 4 weeks after the amputation. When limbs were ampu-tated at the mid-R/U level 7 days after denervation, zero ofsix limbs regenerated. When limbs were amputated at theproximal-H level 14 days after denervation, regenerationoccurred in five of six limbs. When limbs were amputated

Table 1. Regeneration responses after reinjury/reamputation of completely denervated newt limbs on days 7, 14, 21, 28, and 35

after the initial amputation

Reinjury: days

after amputation/

denervation

Total number

of limbs

Regeneration response? Number of days

after reamputation

that regeneration was

first observed1Yes % No %

7 18 0 0 18 100

14 28 12 43 16 57 28

21 14 12 86 2 14 23

28 12 11 92 1 8 21

35 20 20 100 0 0 21

1Regeneration was based on the presence of an early-bud stage blastema. Control reamputated limbs consistently showed early-

bud stage blastemas at 21 days after amputation.

Wound Rep Reg (2006) 14 479–483 c� 2006 by the Wound Healing Society480

Wound-healing response in the adult newt Salley-Guydon and Tassava

at the mid-R/U level 14 days after denervation, six of sixlimbs regenerated (Table 2). In both of the 14-day groups,a blastema outgrowth was observed at 3 weeks after am-putation. Regardless of amputation level, no regenerationwas seen in any of the 12 amputated limbs that were den-ervated 1 day after amputation (and not reamputated).Control, innervated, right limbs all regenerated typically.

The results of nerve staining of limbs of Experiment Ishowed that control, sham-operated limb stumps typicallyhad many nicely staining nerve fibers in the various nervetracts at the mid-R/U level (Figure 1A). However, at 7 and14 days after denervation, nerve tracts in the limb had de-generated and no new ingrowing fibers could be detectedat the R/U level (Figure 1B, C). By day 21, regrowing fib-ers had reached the R/U level in relative abundance, esti-mated to be 1/3 or more of the normal supply (Figure 1D).By day 28, almost full reinnervation had occurred (Figure1E) and on day 35 reinnervation was complete (Figure1F). It should be emphasized that the extent of reinnerva-tion was a gross estimation based on examination of lon-gitudinal sections as actual nerve fiber counts were notperformed. Note that most stumps at the R/U level weresensitive to touch at the H-level at day 14 and all were sen-sitive at day 21 at the R/U level. Stumps were beginning toshow some movement at day 14 and more so at day 21.Results from Samuel’s nerve stain analysis of ExperimentII limb stumps indicated that nerves reached the mid-H

Table 2. Regeneration responses after delayed amputation of denervated newt limbs 7 and 14 days after denervation

Day of amputation

after denervation Level of amputation

Regeneration Number of days after

amputation regeneration

first observedYes No

7 Proximal humerus 5 1 28 (35)1

7 Radius/ulna 0 6

14 Proximal humerus 5 1 19 (33)

14 Radius/ulna 6 0 22 (36)

1Number in parentheses is the age of the limb in days after denervation.

Figure 1. Photomicrographs showing the progress of ingrow-

ing nerves into denervated limb stumps of the adult newt. (A)

Normal nerve fibers in a major nerve at the R/U level of an in-

nervated, control limb. (B) Denervated, nerve at the R/U level 7

days after denervation. Nerve fiber degeneration has occurred.

(C) Degenerated nerve at the R/U level 14 days after denerva-

tion. Regenerating fibers have not yet reached this level. (D) In-

growing nerve fibers at the R/U level at 21 days after

denervation. Approximately 1/3 of the normal nerve supply has

returned. (E) Ingrowing nerve fibers fill up the nerve tract at the

R/U level by 28 days after denervation. Almost complete rein-

nervation has occurred. The denervated limb stump at the mid-

H level has become innervated to this degree by day 21. (F)

Complete filling of nerve tracts has occurred at the R/U level by

day 35 after denervation. The H-level shows full innervation, as

seen here, at day 28 after denervation (scale bar5100mm

(A–E); scale bar530mm (F)).

Wound Rep Reg (2006) 14 479–483 c� 2006 by the Wound Healing Society 481

Wound-healing response in the adult newtSalley-Guydon and Tassava

amputation level about 1 week earlier, already at 14 daysafter denervation, than the mid-R/U level, and had abovethreshold innervation by day 21 (Figure 1E).

Denervated limb stumps typically point caudally, areheld at approximately a 451 angle with the body, are notsensitive to touch, and do not move. This denervated con-dition was maintained until day 14, at which time somesensitivity, movement, and change of angle toward 901were seen. By day 21, all limbs showed increased sensitivityand swinging movement rostrally and caudally. Limbswere no longer held back at a 451 angle to the body. Bydays 28 and 35, all limbs showed movement similar to thecontrol, innervated, right limbs.

We performed a chi-square test of homogeneity on thedata in Table 1 testing the null hypothesis: the proportionof limbs that regenerated was the same in all categories re-gardless of the days after initial amputation when limbswere reamputated; differences were due to chance alone.The null hypothesis was rejected (p < 0.001). To deter-mine which of the categories of reamputation were signif-icantly different from one another, a Tukey multiplecomparisons test was performed and showed that the day7 reamputation category was significantly different fromall other reamputation categories (p < 0.05).

The data in Table 2 were analyzed by the nominal logis-tic fit for response method. The null hypothesis was: thedata differences due to level of initial amputation and dayof reamputation are due to chance alone. The test showedthat both level of initial amputation and day of reamputa-tion had significant effects (p < 0.002), thus rejecting thenull hypothesis.

DISCUSSION

The present experiments were designed to vary the time ofinnervation of denervated limb stumps in an effort to timethe critical healing events that block regeneration. Thus,limbs were reamputated at different times after the originalamputation/denervation to determine when in-growingnerves could still initiate regeneration. Also, limb stumpsof short or long lengths were utilized so that ingrowingnerves had different distances of the limb to traverse.

Limbs in Experiment I were reamputated and subse-quently observed for regeneration as an indication of thetiming of both commitments to wound healing and thresh-old innervation. The results showed that sufficient nerveshad returned to the limb stump by day 21 to initiate re-generation. Of the limbs reamputated on day 21, 86% re-generated without delay compared with the control, rightlimbs. Nerve staining confirmed threshold innervation onday 21. As could be predicted, limbs reamputated 28 and35 days after the original amputation/denervation regen-erated in 92 and 100% of the cases, respectively. However,when reamputation was on day 14, only 43% of the limbsregenerated and blastema formation was delayed by 1week. These 14-day reamputated limb stumps would nothave acquired threshold reinnervation for another 7 days.Thus, while nerves were returning to the limb stump, cov-ert wound-healing events presumably were taking place inover half of the limbs. It was of interest that regenerationof the 14-day limbs was delayed. It may be that some de-differentiated cells had committed to wound healing andothers had not. The latter could have been present in rel-

atively small numbers and a week’s time was therefore re-quired for them to mount a regeneration response. Noneof the limbs reamputated at 7 days regenerated. These limbstumps would not have acquired threshold innervation foranother 2 weeks, during which time all of these limbs com-mitted to a wound-healing response. We conclude fromExperiment I results that commitment to wound healingoccurs between 1 and 2 weeks after amputation/denerva-tion.

The results of Experiment II support the conclusionfrom Experiment I. Both the level of amputation and thetime allowed for nerves to begin reinnervation before am-putation were instrumental in determining the number oflimbs that regenerated in Experiment II. Denervated limbsamputated through the proximal-H always regenerated,regardless of whether amputation was 7 or 14 days afterdenervation. Proximal limb stumps are short, and ingrow-ing nerves reached the amputation surface quickly, beforecommitment to stump wound healing. However, dener-vated limbs amputated 7 days after denervation were de-layed in initiating regeneration by approximately 7 dayscompared with denervated limbs amputated 14 days afterdenervation. It seems that sufficient nerves were alreadypresent at 14 days after denervation (7 days after amputa-tion) of these short limb stumps to prevent further cellularcommitment to wound healing, but a week’s time was re-quired for the noncommitted cells to initiate regeneration.Denervated R/U level limb stumps failed to regeneratewhen amputated 7 days after denervation, as in Experi-ment I. However, if R/U-level limb stumps were amputat-ed 14 days after denervation, all of the limbs regenerated.It should be noted that the R/U level stumps in Experi-ment I that were reamputated on day 14 regenerated in just43% of the cases, and were delayed by 1 week. The lowerpercentage regeneration and the delay may be due to thefact that the limb stumps of Experiment I were reamputat-ed 1mm into the stump. Some of the cells at that reampu-tation level were likely in the dedifferentiation area andwere thus possibly already committed to wound healing.

Results from Experiments I and II lead to the conclu-sion that events inhibitory to regeneration occur in dener-vated limb stumps between days 7 and 14 afteramputation. During this interval of time, the limb stumpbecomes committed to a wound-healing response. Previ-ous histological examination of denervated adult newtlimb stumps between days 7 and 14 after amputation re-vealed no signs of overt tissue healing (cartilaginous callus,dermal pad, cicatrix)12 at the distal tip. Instead, dediffer-entiated cells and a wound epithelium were present.13,14

This commitment to wound healing must therefore occurcovertly. It seems likely that the cells are being develop-mentally programed to undergo tissue healing rather thanepimorphic regeneration. Thus, even though nerves ulti-mately are present in sufficient supply for regeneration,wound healing occurs instead.

Schotte15 denervated limbs of Triton (European newt)and amputated at different levels. Some of the shortestlimb stumps, those amputated nearest the body, showeddelayed regeneration. Schotte15 reasoned that nervesreached the distal tip of the shortest limb stumps beforeinhibitory healing events occurred. In the present study,none of the control denervated limb stumps in which caseamputation was through the mid-H regenerated. It may be

Wound Rep Reg (2006) 14 479–483 c� 2006 by the Wound Healing Society482

Wound-healing response in the adult newt Salley-Guydon and Tassava

that if the amputation had been nearer the body wall, somelimbs may have regenerated. A ‘‘resistant cicatrix’’ wasobserved in the nonregenerating Triton stumps thatSchotte15 believed was the mechanical obstacle thatblocked regeneration. Note that Schotte15 carried out his-tological examination on limbs several weeks after ampu-tation and thus overt tissue healing had already occurred.Salley and Tassava10 showed that when the dermal padwas removed by amputation or dissection or the stumpwas sufficiently injured by a razor incision, long-term den-ervated newt limbs regenerated. Thus, with sufficientnerves already having grown in, two additional conditionsnecessary for regeneration, an injury and a wound epithe-lium, were established.

We show here that even in the absence of sufficientnerves, the amputated limb is still capable of regeneratingfor 1 week. Then, between 1 and 2 weeks after amputation,if nerves continue to be absent, covert events occur thatpreclude regeneration and favor wound healing. These ob-servations are consistent with those of previous studieswith adult newts, showing that many events occur similar-ly in innervated and denervated limb stumps. Cells dedif-ferentiate and enter the cell cycle,13 and the ST1 matrixmolecule undergoes breakdown in both cases.16 It may bethat during the first week after amputation/denervation, asubstantial fraction of dedifferentiated cells are removedby apoptosis leading to premature differentiation of theremaining cells.17,18 While mitotic activity, and DNA,RNA, and protein synthesis are reduced in denervatedlimb stumps, no specific RNA or protein has been associ-ated with the denervated condition.1 Subtle differences inone or more of these events/molecules may be instrumen-tal in regeneration on the one hand or the commitment towound healing on the other. Nerves may favorably mod-ulate the response of the immune system during the firstweek after amputation.19 Alternatively, outgrowth and re-generation vs. commitment to wound healing may involveone or more growth factors.20 The system described heremay provide an opportunity to identify the cellular andmolecular events associated with regeneration vs. woundhealing by comparing innervated and denervated limbstumps between the first and second weeks after amputa-tion.

ACKNOWLEDGMENTS

We thank Eric V. Yang for help with Figure 1 and with theelectronic submission, and Greg Booton and Scott Kim-ball for the statistical analyses.

REFERENCES

1. Stocum DL. Wound repair, regeneration and artificial tissues.Austin: Landis RG, Company, 1995.

2. Tsonis P. Limb regeneration. New York: Cambridge Univer-sity Press, 1996.

3. Singer M. The influence of the nerve in regeneration of theamphibian extremity. Q Rev Biol 1952; 27: 169–200.

4. Singer M, Salpeter M. Regeneration in vertebrates: the roleof the wound epithelium. In: Zarrow MX, editor. Growth inliving Systems. New York: Basic Books, 1961.

5. Mescher AL. Effects on adult newt limb regeneration of par-tial and complete skin flaps over the amputation surface.J Exp Zool 1976; 195: 117–29.

6. Singer M, Craven L. The growth and morphogenesis of theregenerating forelimb of the adult newt, Triturus. J EmbryolExp Morph 1948; 108: 279–308.

7. Schotte OE, Butler EG. Morphological effects of denerva-tion and amputation of limbs in urodele larvae. J Exp Zool1941; 87: 279–322.

8. Schotte OE, Butler EG. Phases in regeneration of urodelelimbs and dependence on the nervous system. J Exp Zool1944; 97: 95–106.

9. Petrosky N, Tassava R, Olsen C. Cellular events in denervat-ed limb stumps of Ambystoma larvae during reinnervationand subsequent regeneration. Experientia 1980; 36: 601–3.

10. Salley J, Tassava RA. Responses of denervated adult newtlimb stumps to reinnervation and reinjury. J Exp Zool 1981;215: 183–9.

11. Samuel EP. Towards controllable silver staining. Anat Rec1953; 116: 511–7.

12. Schotte OE, Liversage RA. Effects of denervation and am-putation upon the initiation of regeneration of regenerates ofTriturus. J Morphol 1959; 105: 495–527.

13. Mescher AL, Tassava RA. Denervation effects on DNA rep-lication and mitosis during the initiation of limb regenerationin adult newts. Dev Biol 1975; 144: 187–97.

14. Salley JD. Responses of amputated adult newt limb stumpsto denervation, reinnervation, and reinjury: a morphologicaland histological study, 1981. PhD Dissertation, The OhioState University.

15. Schotte OE. Systeme nerveus et regeneration chez le Triton.Rev Suisse Zool 1926; 33: 1–211.

16. Yang EV, Shima DT, Tassava RA. Monoclonal antibodyST1 identifies an antigen that is abundant in the axolotl andnewt limb stump but is absent from the undifferentiatedregenerate. J Exp Zool 1992; 264: 337–50.

17. Globus M, Vethamany-Globus S, Lee YCI. Effect of apicalepidermal cap on mitotic cycle and cartilage differentiation inregeneration blastema in the newt, Notophthalmus viride-scens. Dev Biol 1980; 75: 358–72.

18. Mescher A, White GW, Brokaw JJ. Apoptosis in regenerat-ing and denervated, nonregenerating urodele forelimbs.Wound Rep Reg 2000; 8: 110–6.

19. Mescher A, Neff AW. Regenerative capacity and the devel-oping immune system. Adv Biochem Engin/Biotechnol 2005;93: 39–66.

20. Whitehead GG, Makino S, Lien C, Keating MT. Fgf20 is es-sential for initiating zebrafish fin regeneration. Science 2005;310: 1957–60.

Wound Rep Reg (2006) 14 479–483 c� 2006 by the Wound Healing Society 483

Wound-healing response in the adult newtSalley-Guydon and Tassava