Neonatal rat cartilage has the capacityfor tissue regeneration

WOLFGANG WAGNER, MDa; JOACHIM REICHL, PhDa; MANFRED WEHRMANN, MDb; HANS-PETER ZENNER, MDa

One of the most relevant issues in future medicine is tissue regeneration. Transplantation medicine alone cannot solvethe problem of incurable conditions of vital organs. One approach to this might be the replication of the spontaneousregeneration that is found in embryonic/neonatal tissue. In this study, a tissue model for basic investigation of regenera-tion mechanisms in vivo was established. We demonstrated by histology and immunohistochemical staining for typesI and II collagen that neonatal rat cartilage unlike adult cartilage has the capacity for rapid scarfree regenerationafter full-thickness incision. The underlying mechanism was identified in the preserved proliferative capacity of neonatalchondrocytes. This in vivo model should prove useful in further studies of the role of cellular (e.g., GA cell cycleregulators) and extracellular (e.g., cytokines) factors in tissue regeneration and wound healing. (WOUND REP REG2001;531–536)

Many tissues have the capacity for regeneration in the PCNA Proliferating cell nuclear antigenfetal stage, while this ability is lost in the adult organism TGF-� Transforming growth factor-�(overview in 1–3). Both cellular properties (e.g. prolifera-tive capacity) and extracellular factors (e.g. cytokines)are relevant for these differences. A clinical example is

ity of cartilage, which could be an even better tissuethe healing of fetal skin, which results in a scar-free tissue

model than skin, because it is avascular, noninnervated,restitution without inflammatory reaction unlike tissue

and comprised of only one cell type. These propertiesrepair in adult tissue. Instead of tissue replacement with

would allow for more isolated investigation of cellularconnective tissue, a regenerative healing process leading

regeneration and wound healing mechanisms in vivo. Into a restitutio ad integrum is observed (Table 1). Thus

order to establish cartilage as a tissue model for regenera-fetal skin provides us with a physiological model for

tion processes, we investigated whether fetal/neonatalspontaneous tissue regeneration.

cartilage has the same healing and regenerative capacityFetal healing properties have been found to be tissue-

as fetal skin.specific. For example, sheep of middle gestational age

During gestation we find a transition phase betweenshow scar-free healing in skin and bone, a combination

fetal and adult healing patterns. In most mammals, in-of scar-free and scar-forming healing in diaphragm mus-

cluding humans, this stage is between the second andcle, and at the same time distinct scar formation after

third trimester.8 However small rodents such as mice andwounding in the gastrointestinal tract.4–7 However, to

rats show fetal healing patterns up to the neonatal phase,date limited information is available on the healing capac-

which is due to their relative tissue immaturity at thetime of birth.9,10 Therefore, we used the rat as our animal

From the Departments of Otorhinolaryngologya and Pa- model, enabling research of fetal healing mechanisms inthologyb, University of Tubingen, Tubingen, Ger- the neonatal stage (days p0 and p1).many

Reprint requests: Wolfgang Wagner, MD, Department ofOtorhinolaryngology, University of Tubingen, MATERIALS AND METHODSSilcher Str. 5, 72076 Tubingen, Germany. Fax:

Neonatal (immediately postpartum, 2 hours of age) and�49–7071–293311; Email: wolfgwagner@yahoo.com. adult (9- to 10-month old) Wistar rats were used as our

531

WOUND REPAIR AND REGENERATIONNOVEMBER–DECEMBER 2001532 WAGNER ET AL.

Table 1. Fetal versus adult wound healing: important performed using antibodies against proliferating cell nu-differences on the cellular and molecular level as found in skin

clear antigen (PCNA). Wound tissue was embedded in Tis-Type of Healing sue Tek (Sakura Co., Tokyo, Japan) after fixation with 2%

Process or activity adult fetal formalin. Ten-micron sections were stained with a mono-collagen deposition disorganised (scar) organised (no scar) clonal antibody against mouse PCNA (#M0879-Dako Co.,wound contraction � �

Hamburg, Germany; 1 : 50 dilution) and development withInflammation � �recruitment of � � diaminobenzidine.fibroblastsrecruitment of � �myofibroblastsdeposition of � ��hyaluronic acid RESULTSactivity of � �

Standardized full-thickness incision wounds in the auri-hyaluronidaseactivity of fibronectin � �� cle of neonatal (n � 8) and adult (n � 8) rats wereTGF-� presence � �

examined after 4 weeks. After this relatively long healingperiod, the acute inflammatory processes are completedand one can observe a preliminary final stage of healing.In all of the neonatal ears the cartilage healed without any

animal model (n � 8 ears of each age, four ears without histologic evidence of scar formation. Normal cartilagetreatment as control). The auricles were incised with a tissue architecture was entirely restored, making the inci-no. 15 scalpel. The blade of the scalpel was held at a sion area indistinguishable from the surrounding unin-constant angle and pushed through the auricle (full-thick- jured tissue. There were no signs of inflammatoryness incision), producing an approximately constant inci- reactions (immigration of granulocytes, macrophags, his-sion with a length of 2 mm. The incisions were placed tiocytes) (see Figure 1). In contrast all the adult earscentrally, leaving the outer perimeter of the auricle intact displayed a distinct cartilage scar with two separatedto ensure adaptation of the wound margins. The incision cartilage stumps being reconnected by a fibrotic bridgearea was marked with special tissue ink (WAK-Chemie but not by chondrocytes (see Figure 2). A semiquantitaveMedical GmbH, Bad Soden, Germany) and the markings analysis of the healing outcome was neither possiblewere refreshed weekly. After 4 weeks the animals were nor useful because the healing pattern was distributedeuthanized. The animal experiments were performed ac- strictly bimodally between the neonatal and the adultcording to the German animal welfare legislation (ap- group; i.e., we did not observe any transitional findingsproval form HN 2/99, University of Tubingen. such as discrete scarring in the neonatal animals or par-

tial regeneration without scarring in the adult group.Histological analysis of wound tissueThe auricular tissue was fixed in 2% formalin, embedded

Immunostaining for type I and II collagen fourweeksin paraffin and 5 �m sections taken. The sections wereafter incisionstained with hematoxylin-eosin and immunostained withThe histological findings could be affirmed at the protein-antibodies against type I collagen (fibroblast specific)expression level by immunostaining for different types ofand type II collagen (chondrocyte specific). A mono-collagen. The healed area in the adult sections displayedclonal antibody against type I rabbit collagen (#PS065significant expression of type I collagen, which is pro-Cell Systems Co,-dilution 1 : 20), and a polyclonal anti-duced by fibroblasts but not by chondrocytes (scar-likebody against type II rabbit collagen (#MS235, Dunn La-healing, Figure 3). In contrast there was no type I collagenbortechnik Co. Asbach, Germany, dilution 1 : 100), wereexpression in the neonatal cartilage incision area, show-used as primary antibodies. A Vectastain ABC Kiting that fibroblasts played no part in the healing process.(#PK6101 Vector Laboratories, Burlingame, CA) wasHere a moderate staining in the subcutis, which physio-used to detect the primary antibodies with diaminobenzi-logically contains connective tissue, served as positivedine as the chromogen.control (Figure 4). Type II Collagen, which is specificfor chondrocytes, was strongly expressed in both neona-Incision and four-day healing time

In a second trial, neonatal and adult auricles (n � 8 ears tal and adult cartilage (positive control), as expected(Figures 5 and 6). However, in the adult scar expressionof each age, four ears without treatment as control) were

examined after a 4-day period following the same incision was comparatively weak, therefore affirming the histo-logical diagnosis of weak or absent chondrocyte-depen-procedure.

In this case the immunohistochemical staining was dent regeneration (Figure 5).

WOUND REPAIR AND REGENERATIONVOL. 9, NO. 6 WAGNER ET AL. 533

Figure 1. Neonatal auricles (four repre-sentative examples) area of incision,after 4 weeks of healing showing regen-erative noninflammatory healing (fetalhealing type). Arrow indicates cartilagestrip; M, muscle; SC, subcutis; E, epider-mis. Hematoxylin-eosin, original magnifi-cation �200 (bar � 25 �m).

Figure 2. Adult auricles (four representa-tive examples) area of incision, after 4weeks of healing showing scar formation(adult healing type). Top arrow indicatesthe cartilage strip and the bottom arrowscar tissue. Hematoxylin-eosin, originalmagnification �100 (bar � 50 �m)

WOUND REPAIR AND REGENERATIONNOVEMBER–DECEMBER 2001534 WAGNER ET AL.

Figure 5.Distinct expression of chondrocyte-specific type II colla-Figure 3. Adult auricle after 4 weeks of healing stained for typegen in cartilage but relatively weaker expression in scar (com-I collagen. Distinct expression of fibroblast-specific type I colla-pare expression of type I collagen in scar in Figure 3), confirminggen in the healed area (scar), indicating healing through fibro-that adult healing was accomplished primarily through fibro-blasts. Weaker expression is observed in subcutis, which containsblasts. Adult auricle, area of incision, after 4 weeks of healing. Topfibroblast derived connective tissue and no staining of cartilagearrow: scar; bottom arrow: cartilage strip. Original magnificationand epidermis indicating specificity of the method. Top arrow:�100 (bar � 50 �m).cartilage strip; bottom arrow: scar. Original magnification �100

(bar � 50 �m)

Figure 6.Distinct expression of chondrocyte-specific type II colla-Figure 4. Neonatal auricle after 4 weeks of healing stained forgen in cartilage, indicating scarless healing accomplishedtype I collagen. No dehiscence in the cartilage strip as notedthrough chondrocytes. Neonatal auricle, area of incision, afterwith no expression of type I collagen in the healed area. Top4 weeks of healing. Original magnification �200 (bar � 25 �m).arrow: cartilage strip; bottom arrow: subcutis, area of incision.

Note the expression in subcutis of type I collagen. Original magni-fication �200 (bar � 25 �m).

Immunostaining for PCNA four days after incisionIn a second study, neonatal (n � 8) and adult (n � 8)ears were examined four days after incision. Already theneonatal cartilage had healed free of scarring. The onlyevidence of the incision was a thickening in the incisedarea. Immunohistological staining for the proliferationmarker PCNA showed strong expression in the nucleiof the neonatal chondrocytes but no expression in theperichondral area. Epidermis and cutaneous glands—other areas of tissue proliferation—were clearly stained

Figure 7. Distinct expression of PCNA in chondrocytes but no(positive control) (figure 7). In the adult auricle, on theexpression in perichondral area, indicating that cartilage healingother hand, the lesion was still clearly dehiscent. Thewas accomplished not through appositional growth but throughPCNA stain was positive only in the connective tissuechondrocyte proliferation. Neonatal auricle, area of incision,

around the incision, indicating a beginning fibroblast- after 4 days of healing. Top arrow: healed cartilage strip; bottomdependent reparation, but not in the cartilage itself (Fig- arrow: perichondral area (area of perichondrium in adult carti-

lage). Original magnification �400 (bar � 12 �m).ure 8).

WOUND REPAIR AND REGENERATIONVOL. 9, NO. 6 WAGNER ET AL. 535

able to demonstrate that the observed healing was ac-complished by the latter mechanism. There was prolifer-ation of neonatal chondrocytes in the lesion area,whereas the adult chondrocytes failed to show this re-sponse. Furthermore, our histological samples showedno evidence of cell migration from the outer cartilageperimeter (as witnessed in adult perichondrial growth).Namba et al.12 made similar observations. The regenera-tion of articular cartilage in their trial was also achievedby chondrocyte proliferation in the vicinity of the lesionin fetal sheep. The mechanism was not cell migration

Figure 8. No expression of PCNA in cartilage but distinct expres- from the synovial membrane as has been described insion in surrounding tissue, indicatingmissingproliferativecapacity adult rabbits.13

of adult chondrocytes and beginning repair of incision by con-Which factors are responsible for the differencesnective tissue. Adult auricle, area of incision, after 4 days of

between the fetal and adult systems? Fetal as well ashealing. Left arrow: cartilage strip; right arrows: connective tissuein area of incision. Original magnification �200 (bar � 25 �m). adult skin retain their healing properties when trans-

ferred into a different milieu.14,15 It was concluded thatintrinsic, tissue-specific factors cause fetal healing pat-

DISCUSSION terns and not extrinsic conditions such as intrauterinesterility or amniotic fluid.14It could be shown in the animal model that incised neona-

tal rat cartilage is capable of scar-free regeneration, The embryonic ability for the proliferation of differ-entiated cells is one of the primary cellular prerequisiteswhereas adult tissue shows a reparative healing pattern.

To our knowledge only two studies have been published for regenerative fetal wound healing. The adult systemhas lost this ability primarily because of terminal cellconcerning fetal/neonatal cartilage healing. Goss11 stud-

ied the healing of full-thickness oro-facial incisions (skin, differentiation and inhibition of mitotic cell division (se-nescence). This means that the cell is arrested in the G1mucosa, cartilage) placed 1 day prior to the end of gesta-

tion in rats. The dermis and mucosa healed entirely with- phase, unable to enter the S phase with DNA synthesisand proliferation. The G1/S phase transition in mammalsout visible scarring. The cartilage (septum) showed quick

wound healing free of inflammation, however the carti- is regulated by cellular regulators (cyclin-dependant ki-nases, cyclins and their inhibitors).16 Consequently, thelage stumps remained separated. Namba et al.12 placed

incisions 100 �m deep (partial-thickness) into articular differential expression of these cellular regulators, whichare again regulated by transcriptional control throughcartilage of fetal sheep in mid-gestation in an attempt to

simulate microlesions that mark the pathophysiological cytokines, is likely to be relevant in healing and regenera-tion in fetal and adult tissues.begining of articular arthrosis. Four weeks later the de-

fect was entirely filled with chondrocytes. The healed As far as extracellular influences are concerned,there is extensive evidence that the three isoforms ofinjured region could not be distinguished from sur-

rounding normal tissue. This is quite similar to the com- the cytokine transforming growth factor-� (TGF-�) playa crucial role (reviews in 2,17). TGF-� in its active formplete tissue regeneration observed in our study, although

the fact that the healing was only seen in superficial is a 25 kD dimeric protein released by thrombocytes aftertissue damage. In addition to a chemotactic effect onarticular cartilage defects limits comparability of the re-

sults with our study. Furthermore, the cartilage regenera- neutrophils, T-lymphocytes and monocytes, it also regu-lates proliferation and the cytokine secretion rate oftion model introduced by Namba et al. requires prenatal

intrauterine manipulation. these inflammatory cells. TGF-� also up-regulates theproduction of extracellular matrix components (colla-Generally, there are two possible mechanisms that

could explain the neonatal regenerative healing observed gen, fibronectin, proteoglycans) by fibroblasts, down-regulates immunologic activity of B and T-lymphocytes,in this study. First, immigration of premature mesenchy-

mal cells from the perichondrium could occur, which on and influences cellular toxicity of macrophages (over-views in 18,19). The overall effect of these multiple actionstheir way differentiate into chondrocytes (appositional

growth). This is the regular growth mechanism in postna- on wound healing can basically be characterized as astimulation of the inflammatory reaction and subsequenttal organisms. Secondly proliferation (mitotic growth) of

the resident already-differentiated chondrocytes could tissue remodeling by fibroblasts. Comparative analysisof fetal and adult wound healing has shown a significantlyoccur in the defect area (interstitial growth). We werelower level of TGF-� in fetal tissue.20 This suggested that

WOUND REPAIR AND REGENERATIONNOVEMBER–DECEMBER 2001536 WAGNER ET AL.

Adzick NS. Scar formation in the fetal alimentary tract. J Pediatrmanipulation of this factor might be a key to improvedSurg 1995;30:392–5.

healing. 6. Longaker MT, Whitby DJ, Jennings RW, Duncan BW, FergusonThere have been successful attempts in several ani- MW, Harrison MR, Adzick NS. Fetal diaphragmatic wounds heal

with scar formation. J Surg Res 1991;50:375–85.mal tissues to influence healing processes by aligning7. Longaker MT, Moelleken BR, Cheng JC, Jennings RW, Adzick NS,the adult cytokine distribution to a fetal pattern by means

Mintorovich J, Levinsohn DG, Harrison MR, Simmons DJ. Fetalof addition and neutralization of TGF-� (in skin,21,22

fracture healing in a lamb model. Plast Reconstr Surg 1992;90:161–71.bone,23–25 and corneal epithelium26). The effects of TGF-

8. Cass DL, Bullard KM, Sylvester KG, Yang EY, Longaker MT, Adzick� in humans are being tested in a phase I clinical studyNS. Wound size and gestational age modulate scar formation in

whose aim is to reduce excessive scarring in skin (Fergu- fetal wound repair. J Pediatr Surg 1997;32:411–5.9. Boon L, Manicourt D, Marbaix E, Vandenabeele M, Vanwijck R.son MWJ, 2000, personal communication).

A comparative analysis of healing of surgical cleft lip correctedWe could demonstrate histologically and immunohis-in utero and in neonates. Plast Reconstr Surg 1991;89:11–7.

tochemically that neonatal rat cartilage has the capacity 10. Oberg KC, Kirsch WM, Hardesty RA. Prospectives in cleft lip andfor rapid, scar-free healing. Now an appropriate in vivo palate repair. Clin Plast Surg 1993;20:815–21.

11. Goss AN. Intra-uterine healing of fetal rat oral mucosal, skin andmodel is at hand for examination of both spontaneouscartilage wounds. J Oral Pathol 1977;6:35–43.regeneration of differentiated cells in neonatal organisms

12. Namba RS, Meuli M, Sullivan KM, Le AX, Adzick NS. Spontaneousand the missing regenerative capacity in adults. The next repair of superficial defects in articular cartilage in a fetal lamb

model. J Bone Joint Surg 1998;80:4–10.step will be the analysis of relevant G1 cell cycle regula-13. Hunziker EB, Rosenberg LC. Repair of partial-thickness defects intors and cytokines using different healing times. Knowl-

articular cartilage: cell recruitment from the synovial membrane. Jedge of differential expression of the responsible factors Bone Joint Surg Am 1996;78:721–33.

14. Longaker MT, Whitby DJ, Ferguson MW, Lorenz HP, Harrison MR,in fetal and adult tissues is a prerequisite for a targetedAdzick NS. Adult skin wounds in the fetal environment heal withintervention of healing and regeneration processes. Asscar formation. Ann Surg 1994;219:65–72.

far as cytokines are concerned, this could be accom- 15. Lorenz HP, Lin RY, Longaker MT, Whitby DJ, Adzick NS. The fetalplished by direct addition or neutralization (through neu- fibroblast: the effector cell of scarless fetal skin repair. Plast

Reconstr Surg 1995;96:1251–61.tralizing antibodies) of the appropriate growth factors,16. Morgan DO. Cyclin-dependent kinases: engines, clocks and micro-similar to the previously described method in skin22, or

processors. Ann Rev Cell Dev Biol 1997;13:261–91.genetic cell transfection leading to local expression and 17. Rosier RN, O’Keefe RJ, Hicks DG. The potential role of trans-

forming growth factor beta in fracture healing. Clin Orthop 1998;secretion of the cytokine.27 For cell cycle regulators, ma-355 (Suppl.):S294–300.nipulation of the protein synthesis of identified target

18. Alevizopoulous A, Mermod N. Transforming growth factor-�: themolecules is possible by means of antisense-oligonucleo- breaking open of a black box. Bioessays 1997;19:581–91.

19. Border WA, Noble NA. Transforming growth factor beta in tissuetides.28

fibrosis. N Engl J Med 1994;331:1286–92.20. Whitby DJ, Ferguson MWJ. Immunohistochemical localization of

growth factors in fetal wound healing. Dev Biol 1991;147:207–15.21. Mustoe TA, Glenn FP, Thomason A, Gramates P, Sporn MB, DeuelACKNOWLEDGMENTS

TF. Accelerated healing of incisional wounds in rats induced byWe thank Marlies Knipper, PhD (Clinical Research transforming-growth-factor beta. Science 1987;237:1333–6.

22. Shah M, Foreman DM, Ferguson MWJ. Neutralisation of TGF-Group, Department of Otorhinolaryngology, Universitybeta1 and TGF-beta2 or exogenous addition of TGF-beta3 to cuta-of Tubingen) for scientific discussions and Karl Sotlar,neous rat wounds reduces scarring. J Cell Science 1995;108:

MD (Institute of Pathology, University of Tubingen), for 985–1002.23. Beck LS, DeGuzman L, Lee WP, Xu Y, McFatridge LA, Gillett NA,help in conducting the experiments.

Amento EP. TGF-beta induces bone closure of skull defects. JBone Mineral Res 1991;6:1257–65.

24. Bostrom MP, Asnis P. Transforming-growth-factor beta in fracturerepair. Clin Orthop 1998;355 (Suppl.):S124–31.REFERENCES

25. Mont MA, Jones LC, Einhorn TA, Hungerford DS, Reddi AH. Os-teonecrosis of the femoral head. Potential treatment with growth1. Dostal GH, Gamelli RL. Fetal wound healing. Surg Gyn Obstet

1993;176:299–306. and differentiation factors. Clin Orthop 1998;355 (Suppl.):314–35.26. Er H, Uzmez E. Effects of transforming-growth-factor beta 2, in-2. Ferguson MWJ, Whitby DJ, Shah M, Armstrong J, Siebert JW,

Longaker MT. Scar formation: the spectral nature of fetal and terleukin 6 and fibronectin on corneal epithelial wound healing.Eur J Ophthalmol 1998;8:224–9.adult wound repair. Plast Reconstr Surg 1996;97:854–60.

3. Wagner W, Plinkert PK. Fetale Chirurgie: aktueller stand und neue 27. Davidson JM, Whitsitt JS, Pennington B, Ballas CB, Eming S, BennSI. Gene therapy of wounds with growth factors. Curr Top Pathmoglichkeiten durch das endoskop. HNO 1998;46:903–13.

4. Stern M, Dodson TB, Longaker MT, Lorenz HP, Harrison MR, 1999;93:111–21.28. Coats S, Flanagan M, Nourse J, Roberts JM. Requirement of p27Kip1Kaban LB. Fetal cleft lip repair in lambs: histologic characteristics

of the healing wound. Int J Maxillofac Surg 1993;22:371–4. for restriction point control of the fibroblast cell cycle. Science1996;272:877–80.5. Meuli M, Lorenz HP, Hedrick MH, Sullivan KM, Harrison MR,