JOURNAL OF CELLULAR PHYSIOLOGY 149:444--150 (19911

Sheep Amniotic Fluid Has a Protein Factor Which Stimulates Human Fibroblast Populated Collagen Lattice Contraction 1. RITTENBERG, M.T. LONGAKER, N.S. ADZICK, AND H.P. EHRLICH* Wound Healing Laboratory, Shriners Burns Institute, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts (J.R., H.P.E.) and The Fetal Treatment Program, Department of Surgery, University of California at San Francisco, Sari Francisco, California (M.J.L., N.S.A.) Sutured incisional wounds made in fetal sheep and rabbits heal without scarring. Fetal sheep excisional wounds can close by contraction, but those in fetal rabbits do not. In vivo and in vitro evidence suggests that rabbit amniotic fluid inhibits wound contraction. The question arises: does sheep amniotic fluid promote wound contraction because their fetal wounds close by contraction? Sheep amniotic fluid (SAF) from 100 and 125 days gestation was tested in fibroblast populated collagen lattice (FPCL) system, an in vitro model of wound contraction. SAF stimulated FPCL contraction in a dose responsive manner. SAF from a 100 day fetus was more stimulating than a 125 day SAF. SAF enhanced FPCL contraction in the presence or absence of serum. SAF was fractionated by size, using column chromatography. It yielded a fraction with an estimated molecular weigh near 40,000 daltons, which stimulated FPCL contraction. The factor was inactivated by proteolytic digestion and heat denaturation. This protein fraction which stimulates FPCL contraction i s not related to 1) actin-myosin filaments enhanced contraction by ATP-induced cell contraction, 2) promotion of fibroblast elongation on glass surface or in collagen, or 3) increased cell number by enhanced fibroblast duplication in a collagen matrix. A mechanism for SAF promotion of FPCL contraction was investigated but not identified.

Soft tissue repair in post-natal mammals differs from that of the fetus. The initial inflammatory response in the adult is intense, and the deposition of a new connective tissue matrix abundant, usually resulting in a scar. However, the inflammatory cellular response in the fetus is sparse and the deposition of new connective tissue minimal, leading to the absence of scarring. The reason for these differences is unclear. Suture-closed incisional wounds mend without developing a scar in the rabbit fetus. However, unsutured incisions or excisional wounds do not close and wound contraction is absent in the rabbit fetus (Somasundaram and Prathap, 1970). Sutured incisional wounds in the sheep fetus heal without scarring, identical to the rabbit fetus, but an open wound in a sheep fetus will close by wound contraction (Burrington, 1971). A possible cause for the absence of open wound closure in the rabbit fetus was described by Somasundaram and Prathap (1972). They showed that covering open rabbit fetal wounds with a silastic sheet, to exclude amniotic fluid, initiated wound contraction. They concluded that amniotic fluid inhibited wound contraction. In contrast, Burrington (1971)showed that amniotic fluid had no inhibitory effect on wound contraction in the sheep fetus. Open wounds in fetal sheep undergo wound contraction in the presence of amniotic Q

1991 WILEY-LISS, INC.

fluid (Longaker et al., 1990a, b). Sheep amniotic fluid (SAF) was examined in an in vitro wound contraction model that utilizes human dermal fibroblasts suspended in a collagen matrix. The rapid mixing of fibroblasts, native soluble collagen, and serum-supplemented culture medium at 37°C produces a polymerized collagen matrix with entrapped fibroblasts throughout. In time, this fibroblast-populated collagen lattice (FPCL) undergoes a reduction in size, called lattice contraction (Bell et al., 1979). The fibroblasts in contact with surrounding collagen fibrils produce forces, which cause lattice shrinkage. Lattice contraction requires serum, viable cells, functioning microtubules and microfilaments, as well as a collagen matrix, which can be readily organized (Bell et al., 1979; Ehrlich et al., 1989). DNA synthesis and cell duplication are not necessary, but protein synthesis is essential (Bell et al., 1979). A FPCL made with type I11 collagen contracts greater and faster than one composed of type I collagen (Ehrlich, 1988). In this study the effect of added SAF on lattice contraction with and without the addition of serum was examined. Received February 20, 1991; accepted June 23, 1991.

“To whom correspondence should be sent a t Shriners Burns Institute, 51 Blossom Street, Boston, MA 02114.

SHEEP AMNIOTIC FLUID STIMULATES CONTRACTION

445

MATERIALS AND METHODS Eight time-dated pregnant ewes from Torrel Farms (Utah, CA) underwent general halothaneioxygen anesthesia using techniques for fetal lamb surgery described previously (Harrison e t al., 1980). Four lambs were at 100 days gestation and four were a t 125 days gestation (145 days equals term). In short, a n ewe was prepared and dropped under sterile conditions. A midline laparotomy was made to expose the uterus. Amniotic fluid was aspirated through the uterine wall using a 16-gauge needle and a 20 cc syringe. Collected amniotic fluid was frozen and stored at -70°C until studied.

To manufacture FPCL a suspension of fibroblasts was added to culture medium and additives, totalling 1.5 ml. The mixture was then blended rapidly with 0.5 ml of collagen solution and transferred to a 35 mm Petri dish. It was placed in a 37°C incubator with a 5%’ CO,, 95% air, and water-saturated atmosphere, where the collagen polymerized within 90 seconds, entrapping the fibroblasts within it. The area of the FPCL was determined by measuring the diameter of the lattice with a ruler to the nearest 0.5 mm. FPCL made under identical conditions in 4 days varied less than 1.5 mm in lattice diameter. All additions and modifications of FPCL were made a t the time of manufacture.

Fibroblasts populated collagen lattice The FPCLs were composed of culture medium supplemented with serum, cultured human dermal derived fibroblasts, and soluble native collagen. The culture media were the same for both FPCL and the maintaining fibroblasts in monolayer culture. They contained Dulbecco’s Modification of Eagle’s Medium (DMEM) supplemented with 4 mM glutamine and gentamicin a t 15 pg/ml. Serum was present in all experiments, except where noted. Fetal bovine serum (FBS) (Gibco, BRL, Grand Island, NY) was at 10% viv, when present but not specified. Human dermal fibroblasts derived from human foreskin were a gift from Dr. David Wyler, Department of Medicine, Tufts New England Medical Center, Boston, MA. All cells used in these studies were between passages 12 and 18. Fibroblasts were released from monolayer culture by trypsinization, counted by a Coulter Counter ZM (Luton, England) and diluted with culture medium to the appropriate concentrations. Native collagen solutions were derived from two sources. The initial studies used pepsin solubilized collagen extracted from a benign uterine tumor (leiomyoma) which had been discarded after examination by the MGH Pathology Department. Briefly, 1 gm of tissue was homogenized in 100 ml of acetic acid (0.5 M) on ice. Pepsin a t 10 mg/100 ml was added and the homogenate stirred for 48 hours a t 4°C. The homogenate was centrifuged at 10,OOOg for 10 minutes and the supernatant saved. Sodium chloride (10% w/v) was added to the supernatant and the mixture was stirred overnight a t 4°C. The salt precipitated collagen was collected by centrifugation (10,OOOg for 10 minutes) and the pellet suspended in 145 mM potassium phosphate buffer pH 7.6. Any insoluble material was removed by centrifugation (10,OOOg for 20 minutes), and the supernatant dialyzed exhaustively against 1 mM HC1. The salt-free collagen solution was frozen, lyophilized, and taken up a t 5 mgiml in sterile 1mM HC1. Later experiments used rat tail tendon collagen. Rat tail tendons isolated from 300 to 400 gm Sprague Dawley male rats were stirred in 0.5 M acetic acid at 4°C for 48 hours. Undissolved material was removed by centrifugation (10,OOOg for 10 minutes). Collagen was precipitated from solution with sodium chloride and then purified and diluted as described with the leiomyoma collagen preparation. SDS polyacrylamide gel electrophoresis of both preparations showed no noncollagen protein staining bands. Both collagen solutions were stored a t 4°C until needed.

Fibroblast ATP-induced cell contraction For ATP-induced cell contraction studies, 20,000 fibroblasts were transferred to a 12 mm circular glass cover slip, resting on the bottom of a well in a 24-wellcluster-tissue-culturedish. The total volume of culture medium in each well was 0.4 ml. Fibroblasts were incubated with additives for 24 hours. The medium was removed and the cells were prepared for ATP-induced cell contraction as described previously (Ehrlich e t al., 1986). Briefly, 50% glycerol in 10 mM Tris HCl, pH 7.5, 50 mM KC1, and 5 mM MgC1, were added and incubated a t room temperature for 30 minutes. The mixture was replaced every 30 minutes in the sequence of 25%, 12%, and 5% glycerol in the same buffer. Fibroblasts, which have their plasma membranes glycerol permeabilized, were no longer viable. To test cell contraction, the 5% glycerol solution was replaced by 10 mM ATP in 30 mM KC1, 5 mM MgCl,, 0.1 mM CaCl,, and 10 mM Tris HC1, pH 7.0. The ATP solution was immediately replaced in half the wells by buffered 4% paraformaldhyde, which fixed the cells. These cell preparations were referred to as time 0. The ATP solution was replaced with paraformaldhyde in the remaining wells after 10 minutes of incubation at room temperature. These cell preparations had completed their microfilament contraction capacity by this time. The fixative was removed after 5 minutes and the coverslips washed three times with phosphate-buffered saline, (PBS) (10 mM sodium phosphate buffer with 140 mM sodium chloride) and left in it until microscopic examination. To measure cell length, coverslips were viewed with a n inverted microscope equipped with phase-contrast optics and a video camera. Cell length, reported in arbitrary units, was measured with a metric ruler on the screen of the video monitor. All experiments were done in triplicate with 50 cells measured from each coverslip. Owing to a n abundant supply, 125-day SAF was examined further. SAF was concentrated 33-fold by vacuum dialysis versus PBS. The concentrated solution was passed through a Fast Protein Liquid Chromatography system Superose-12 Chromatography Column (Pharmacia, Inc., Piscataway, NJ). The column was equilibrated in PBS a t a flow rate of 1mliminute. Two minute fractions were collected, filter sterilized (0.2 p pore membrane), and stored sterile a t 4°C until testing. A 30 pl aliquot from each of three fractions were pooled and tested in a n FPCL. The two pooled fractions that

446

RITTENBERG ET AL.

700

B

A

E 500 E

125 D A Y SAF

700

100 D A Y S A F

E 500 E

Q

a

U

W

[I

a:

a

a 300

300

1 oc

100

(

1

2

- 3. -

-,

4

C 1

DAYS

2

3

4

DAYS

Fig. 1. SAF stimulation of FPCL contraction. FPCL composed of 70,000 fibroblasts, with 2.5 mg of leiomyoma collagen in a total volume of 2 ml and in presence or absence of 10%FBS, were measured over a 4 day period. FPCLs were supplemented with either 100, 200, or 400 111of SAF. A: SAF harvested from 100-day gestation period was added at 5%,viv squares, 10%viv triangles, and 20% viv diamonds. The

open symbols represent lattice size when 10% FBS was present and the solid symbols represent the omission of serum The open circles are 10%FBS alone B: SAF harvested from a 125-daygestation period was substituted for 100-day SAF and the symbols mean the same as described above in A

showed the greatest stimulating activity were tested at 0.1 ml from each 2 ml fraction. The three 2 ml fractions with the greatest promoting activity were pooled and used in further studies. Student's "t" test was utilized for the statistical analysis.

TABLE 1. Amniotic fluid harvested from 100- and 125-day-oldsheep

RESULTS Lattice contraction Lattice contraction is the size reduction of a FPCL with human fibroblasts incorporated in a collagen matrix. As a n example, a n initial 960 mm2 FPCL in a 35 mm Petri dish was reduced to 79 2 1 5 mm2 by 48 hours, when 70,000 fibroblasts were combined with 2.5 mg of leiomyoma collagen and 10% FBS in a 2 ml volume. When PBS was substituted for FBS, the FPCL a t 48 hours showed no contraction and the area remained a t 960 mm'. When SAF from 125-day-old fetus was substituted for FBS, at concentrations of 5%, lo%, and 20% vlv, then lattice sizes were reduced to 544 2 126 mm2, 288 t 100 mm', and 148 2 38 mm', respectively a t 48 hours. As demonstrated in Figure 1, the addition of SAF from 100- or 125-day-old fetus stimulated lattice contraction in a dose-dependent manner, increasing both its rate of contraction and final size. The importance of fetal age and the presence of this lattice contraction was investigated. SAF from a 100day fetus was compared to that collected from a 125-day fetus. As presented in Table 1and Figure 1, the 100-day fetus amniotic fluid stimulated lattice contraction more than the 125-day amniotic fluid. This difference was significant at 1day (22%),2 days (29%), and 4 days (33%).It appears that under the conditions of this study, the SAF a t 100 days was a better stimulator of lattice contraction than the SAF a t 125 days. In the presence of FBS, the inclusion of SAF further

fetus stimulation of FPCL contraction' Addition v/v

15%PBS 15%FBS 15% 100-dav SAF 15%~ 125-da; SAF

Area of FPCL measured in mm2 N 1 day 2 day 3 3 3 3

960 6 2 3 i 158 462 f 97* 597 f 103*

960 2 3 4 3 ~17 245 48* 346 S 87*

+

4 day 842 f 22 10458 181 f 28** 273 f 58**

'FPCLS were manufactured with 80,000 human fibroblasts with 2.5 mg of leiomyoma soluble collagen, having a total volume of 2 ml in a 35 mm Petri dish. *P 5 0.05 comparing 100-dayto 125-day SAF. '*P C 0.01 comparing 100-day to 125-day SAF.

enhanced lattice contraction (Figure 2). FBS a t 6% vlv concentration promotes lattice contraction to 24% of its initial size within 48 hours. SAF (125-day1, a t 15% vlv concentration, enhanced lattice contraction to 36% of its initial size. When 15% SAF was combined with 6% FBS, the lattice area was further reduced to 9% of its initial size. SAF is additive to FBS promotion of lattice contraction. SAF collected from a 125-day-old fetus was fractionated by size using column chromatography. As shown in the insert in Figure 3, the fractions with SAF lattice contraction stimulating activity appeared in the 62 and 64 ml volume fractions, with a n estimated molecular weight between 40,000 and 47,000. This 40,000 plus molecular weight lattice contraction stimulator (40KLCS) was studied further. An inhibitory-like activity was detected in fraction volumes 96 and 102 ml (Fig. 2). However, this inhibitory activity could not be confirmed when studied further.

Mechanism of action The mechanism for 40K-LCS activity was investigated for its possible potential to promote human

447

SHEEP AMNIOTIC FLUID STIMULATES CONTRACTION

-

I

I

=

z

I

I

i*

I

I I I.

I I I I I

z

I I I

r

I E

z z

I 5

I

I I

~

1

3

2

4

DAYS

Fig. 2. SAF stimulates lattice contraction in the absence of added serum. FPCL composed of 80,000 fibroblasts, with 2.5 mg rat tail tendon collagen in a total volume of 2 ml, were measured daily for 4 days. The PBS lattice had 450 ~1 of PBS added. FBS 6%had 120 1 of FBS added. The SAF-l5%-FPCL had 300 p1 of 125-day SAF acbed. FBS and SAF had 120 ~1 of FBS and 300 p1 of 125-day SAF present. The lines represent the standard deviation of 3 FPCL from each treatment eroua.

42

54

66

78

90

102

114

FRACTION VOLUME (ml)

fibroblast proliferation in a collagen matrix. Human fibroblasts do not divide when suspended in collagen for the first 4 days (Sarber e t al., 1981; Schor, 1980). If 40K-LCS stimulates human fibroblasts to divide, the greater number of cells within the collagen matrix would be expected to enhance lattice contraction. To test this theory, lattice contraction was measured in the presence and absence of hydroxyurea. Hydroxyurea at 1mM concentration inhibits DNA synthesis and blocks cell division. FPCLs were cast with 2.5 mg of rat tail tendon collagen. Fibroblasts were released from collagen by collagenase digestion for 30 minutes a t 37°C (Buttle and Ehrlich, 1983). The cells were harvested by centrifugation and counted with a hemocytometer. The addition of 1 mM hydroxyurea to FPCL was over a 2-day period. FPCLs with 10% FBS and 120,000 cells contracted to 42% of their initial size and contained 88,700 ? 52,000 cells. FPCLs with 10% FBS and 1mM hydroxyurea contracted to 41% of their initial size and contained 117,000 c 56,500 cells. FPCL made with 125-day SAF contracted to 52% of their initial size and contained 130,000 64,700 cells. FPCL made with 15%SAF and 1mM hydroxyurea had contracted to 50% of their initial size and had 91,300 2 51,500 cells. No significant enhancement of cell duplication occurred from added SAF. In another series of experiments, the isolated 40KLCS a t 5% vlv inclusion over a 3-day period was examined. Little or no lattice contraction occurred without added 40K-LCS during the 3-day observation period (Table 2). Including 40K-LCS with and without 1 mM hydroxyurea promoted a equivalent degree of lattice contraction during that same time interval. Thus, SAF and 40K-LCS were shown not to promote human fibroblast proliferation in collagen. Another possible cause for the promotion of lattice

*

Fig. 3. The isolation of the SAF stimulator activity of FPCL contraction by column chromatography. FPCL composed of 119,000 human fibroblasts, with 2.5 mg of rat tail tendon collagen in a total volume of 2 ml, were supplemented with three pooled column fractionations. Fractionation of SAF was described in the text. FPCL size was measured a t day 2. The fractions with stars were further tested as 2 minute fractions and the results from that study are presented in the insert.

contraction is enhanced cytoplasmic microfilament contractile activity. For example, enhanced myosin ATPase activity is expected to enhance lattice contraction (Ehrlich et al., 1986). A direct examination of this theory utilized a model of ATP-induced cell contraction. Will ATP-induced cell contraction be increased by incubating fibroblasts, in the presence of 40K-LCS? The results are presented in Figure 4. The 40K-LCS a t 10% vlv with and without serum was included in the monolayer culture medium for 24 hours before testing. No enhancement of ATP-induced cell contraction oc-

TABLE 2. FPCL contraction and DNA synthesis modulation by amniotic fluid 4OK-LCS fraction'

Addition v/v -

1

5%PBS % Change 5%40K-LCS

960 0

-HU Change +HU Change

*

734 97 24% 738 5 32 23%

FPCL area in mm2 Day 2 921 4% 548 f 76 43% 572 f 135 40%

3 903

* 57

6%

471 f 56 51% 491 f 111 49%

'FPCLS were manufactured with 120,000 human dermal fibroblasts with 2.5 mg of rat tail tendon collagen having a 2 rnl volume in a 35 mm Petri dish. Three FPCLs were used for each addition. Hydroxyurea, HU, was added at 1 mM concentration to inhibit DNA synthesis.

448

RITTENBERG ET AL.

TABLE 3. The protein nature of lattice contraction stimulating factor from fetal sheea amniotic fluid' Addition v/v

N

Lattice area at 48 hours

Area change

10% PBS 10% FBS 10%40K-LCS2

3 3

948 mm2 494 79 mm2

48%

3 3 3

501 f 19 mm2 757 f 13 mm2* 781 f 87 mm2*

alone trypsin3 9ooc3

*

1%

48% 21%

19%

'FPCLS were manufactured with 120,000 human fibroblasts and 2.5 mg rat tail tendon collagen in a volume of 2 ml in a 35 mm Petri dish. 'Fraction 4OK-LCS is a 40,000 molecular weight fraction isolated from 125-day-old sheep amniotic fluid by molecular sieve chromatography. 3Limited trypsin digestion of 4OK-LCSinvolved mixing 10 wg of trypsin with 0.5 ml of 4OK-LCS and incubating at room temperature for 30 minutes. The trypsin was inhibited by adding 1 wg of phenylmethylsulfonyl fluoride. *P 5 0.01 compared to 10%4OK-LCS.

ATP

ADDED FOR

10 M I N U T E S

Fig. 4. ATP-induced cell contraction. Six monolayer fibroblast preparations from each experimental group were tr;a?d with ATP as described in text. Three of the preparations - were fixed in paraformaldhyde immediately following the addition of ATP. The three other preparations "+" were incubated with ATP for 10 minutes before fixation. The length of fibroblasts was measured with a metric ruler on a video screen and cell lengths were recorded in arbitrary units. Fibroblasts in monolayer were incubated for 24 hours in 0.4ml of medium with 40 pl of FBS, or 40 p1 of 40K-LCS or both 40 pl of FBS and 40 p1 of 40K-LCS.

curred in the 40K-LCS with and without FBS. Also, 40K-LCS on a glass surface did not promote increased cell elongation or spreading. Thus, the amniotic fluid stimulator did not promote enhanced cytoplasmic muscular contraction and/or myosin ATPase activity. When the 40K-LCS fraction was treated with trypsin or heated to 90°C for 30 minutes, stimulation of lattice contraction was reduced by 50% and 55% (Table 3). FPCL size reduction of 52% to 501 t 19 mm2 a t 48 hours with FBS in rat tail tendon collagen occurred with 120,000 human fibroblasts. FPCLs made with r a t tail tendon collagen contract slower than those composed of leiomyoma collagen. With the omission of either 40K-LCS or FBS, lattice size was 948 mm2, 99% of its initial size. With trypsin pretreatment of 40KLCS, lattice size was 757 2 13 mm2, a 21% reduction of the initial size. Preheated 40K-LCS produced lattice size of 781 ? 87 mm2 at 48 hours, representing a 19% reduction or 81% of its initial size. This suggests that $OK-LCS is a protein which requires secondary protein structure for biological activity. Both reducing the size of the protein by trypsin digestion or heat denaturing the protein, which destroys secondary protein structure, decreased its biological activity. Within collagen, SAF and FBS promoted cell elongation and spreading, as shown by phase contrast microscopy at 48 hours (Fig. 5). In the absence of either SAF or FBS, fibroblasts showed reduced cell spreading and elongation; instead they retained a spherical shape (Fig. 5A). The morphological shift from spherical to elongated cell morphology is complete within 2 days and is required for lattice contraction. Human fibroblasts lengthened and spread in a rat tail tendon collagen lattice with COK-LCS (Fig. 5B). Fibroblasts in

a collagen lattice with FBS appear identical to cells with added 40K-LCS (Fig. 5C).

DISCUSSION A factor has been isolated from 125-day fetal sheep amniotic fluid (SAF) that stimulates fibroblast populated collagen lattice contraction. This same factor appears to be present in 100-day SAF in a more effective concentration. This SAF factor has a molecular weight between 40,000 and 47,000, is titled 40K-LCS, and will advance lattice contraction in the absence of serum. Based upon molecular size, this factor is too large to be Transforming Growth Factor beta (TGF,) 25,000 MW, which enhances lattice contraction in the presence of serum (Montesano and Orci, 1988). In the absence of serum TGF, does not support lattice contraction (unpublished data). The SAF 40K-LCS factor stimulates FPCL contraction without serum. The 40K-LCS promotes cell elongation in collagen. This suggests that its action may be related to the cytoskeleton of the cell (Adelstein, 1982; Ehrlich et al., 1986). Microfilament sliding may be enhanced by added 40K-LCS. The addition of ATP to glyceroltreated fibroblast preparations induces cell contraction, which measures the dynamic changes in cell size produced by actin myosin contraction. The inclusion of 40K-LCS produced no enhancement of ATP induced cell contraction. This finding does not support the concept that myosin ATPase or microfilament contraction was enhanced by 40K-LCS. If i t did, a greater degree of cell contraction would have been expected. The factor's stimulation of lattice contraction was additive with serum. It is possible that the factor is also found in serum and t h a t the combination of 40K-LCS and FBS increased its concentration. Also, the cause of action for the SAF promotor may be different from that of serum. One possibility is that it may promote a n increase in cell number. Human fibroblasts do not divide when suspended in collagen (Bell et al., 1979). Serum does not promote human fibroblast proliferation when suspended in collagen (Sarber et al., 1981; Schor, 1980); the possibility that 40K-LCS promotes human cell proliferation was examined. We found human cell division was not stimulated by SAF. Other possible mechanisms may be related to cell attachments to collagen or a direct action on collagen fibrils, where

SHEEP AMNIOTIC FLUID STIMULATES CONTRACTION

Fig. 5. Phase-contrast view of fibroblasts in the central region of 2-day-old FPCL. FPCL composed of 120,000 fibroblasts, with 2.5 mg rat tail tendon collagen in a final volume of 2 ml, were supplemented with looP1 Of PBS looJL1 40K-LCS (B), Or looP1 Of FBS ()‘. Microscopic magnification is x 100.

449

they can be more efficiently packed together. Studies of these theories are planned for the future. Differences in collagen deposition in fetal rabbit wounds have been reported (Adzick et al., 1985; DePalma et al., 1987; Krummel et al., 1989; Siebert et al., 1990). In rabbit fetus healing, hyaluronic acid is the most abundant connective tissue component) while the deposition of collagen appears to be minimal (DePalma et al., 1987; Siebert e t al., 1990). In contrast, collagen deposition in a the fetal sheep wound is greater than in a n adult wound (Longaker et al., 1990b). A recent reported finding showed that the elimination of hyaluronic acid by hyaluronidase in rabbit fetal wounds promoted both collagen deposition and wound contraction (Mast et al., 1990). This shows that hyaluronic acid has a n influential role in fetal healing. Rabbit fetal wound contraction in vivo and FPCL lattice contraction in vitro have been shown to be inhibited by rabbit amniotic fluid (Krummel et al., 1986; Somasundaram and Prathap, 1972). The prevention of direct contact of amniotic fluid with a n open rabbit fetal wound advanced wound contraction (Somasundaram and Prathap, 1972). It was concluded that rabbit amniotic fluid inhibits wound contraction. In collaboration with Krummel’s group, we showed that rabbit amniotic fluid inhibited FPCL contraction (Krummel et al., 1989). FPCL, composed of rabbit fibroblasts and rabbit amniotic fluid, at concentrations greater than 20% viv inhibited lattice contraction. That finding is in direct conflict with findings presented here, It appears that differences in lattice contraction caused by added amniotic fluid are related to species differences. The rabbit amniotic fluid inhibitory factor is concentration dependent but has not been characterized by size. In another study examining added biological fluid‘s modulation of lattice contraction, 3- and 7-day-old adult r a t wound fluid was studied and found to be inhibitory (Rittenberg e t al., 1990). A small molecular weight factor between 10,000 and 20,000 was isolated from rat wound fluid, which inhibited lattice contraction. The early nature of r a t wound fluid harvesting suggests that the factor may be a product of wound macrophages. Inflammatory cells within fetal wounds are sparse in number and are expected to play a minor role in fetal wound repair compared to adults (Harrison et al., 1980; Rowsell, 1984). The absence of these soluble products in SAF may contribute to its promotion of lattice contraction. The theory that a species specific factor could have a major effect on wound closure may have clinical importance. Its inclusion in a healing defect may promote wound closure. In a healed scar, its exclusion or a n inhibitor of its action may prove beneficial in reducing the severity of scar contractures. In either case, work is continuing to purify and identify the mechanism of action Of the factor. LITERATURE CITED Adelstein, R.S. (1982) Calmodulin and the regulation of the actinmyosin interaction in smooth muscle and non-muscle cells. Cell, 30,349-350. Adzick, N.J., Harrison, M.R., Glick, P.L., Beckstead, J.H., Villa, R.L., Scheuenstuhl, H., and Goodson, W.H. (1985) Comparison of fetal, newborn and adult wound healing by histologic, enzyme-histochem-

450

RITTENBERG ET AL

ical and hydroxyproline determinations. J. Pediatr. Surg., 20:315319. Bell, E., Ivarsson, B., and Merrill, C. (1979)Production of a tissue-like structure by contraction of collagen lattice by human fibroblasts of different proliferative potential in vitro Proc. Natl. Acad. Sci. U.S.A., 76:1274-1278. Burrington, J.D. (1971) Wound healing in the fetal lamb. J. Pediatr. Surg., 6.523-528. Buttle, D.J., and Ehrlich, H.P. (1983) Comparative studies of collagen lattice contraction utilizing a normal and transformed cell line. J. Cell. Physiol., 116:159-166. DePalma, R.L., Krummel, T.M., Nelson, J.M., Durham, L.A., Michna, B.A., Diegelmann, R.F., and Cohen, I.K. (1987)Fetal wound matrix is composed of proteoglycan rather than collagen. Surg. Form., 38,626-628. Ehrlich, H.P. (1988) The modulation of contraction of fibroblast populated collagen lattices by types I, XI, and 111 collagen. Tissue Cell 20t47-50. Ehrlich, H.P., Buttle, D.J., and Bernanke, D.H. (1989) Physiological variables affecting collagen lattice contraction by human dermal fibroblasts. Exp. Mol. Pathol., 50:220-229. Ehrlich, H.P., Rajaratnam, J.B.M., and Griswold, T.R. (1986) ATPinduced cell contraction in dermal fibroblasts: Effects of CAMPand myosin light-chain kinase. J. Cell Physiol., 128:223-230. Harrison, M.R., Jester, J.A., and Ross, N.A. (1980) Correction of congenital diaphragmatic in utero. I. The model: Intrathoracic balloon produces fetal pulmonary hypoplasia. Surgery, 88:174-182. Krummel, T.M., Ehrlich, H.P., Nelson, J.M., Michna, B.A., Thomas, B.L., Haynes, J.H., Cohen, I.K., and Diegelmann, R.F. (1989) Open fetal rabbit wounds do not contract. Surg. Forum, 40:613-615. Krummel, T.M., Nelson, J.M., Diegelmann, R.F., Lindblad, W.J., Salzberg, A.M., Greenfield, L.J., and Cohen, I.K. (1986) Wound healing in the fetal and neonatal rabbit. Surg. Forum, 38r595-598. Longaker, M.T., Burd, D.A.R., Yen, T.S.B., Gown, A.M., Jennings, R.W., Duncan, B.W., Siebert, J.W., Harrison, M.R., and Adzick, N.S.

(1990a) Midgestation fetal wounds contract in utero. Plastic Surg. Res. Council, 35:lOl-103. Longaker, M.T., Whitby, D.J., Adzick, N.S., Crombleholme, T.M., Langer, J.C., Ducan, B.W., Bradley, S.M., Stern, R., Ferguson, M.W.J., and Harrison, M.R. (1990b) Studies in fetal wound healing: VI. Second and early third trimester wounds demonstrate rapid collagen deposition without scar formation. J . Pediatr. Surg., 25: 63-69. Mast, B.A., Haynes, J.H., Krummel, T.M., Schatzhi, P.F., Flood, L.C., Diegelmann, R.F., and Cohen, I.K. (1990) The addition of hyaluronidase to a fetal rabbit wound site induced fibrogenesis. Plast. Surg. Res. Council, 35r103-105. Montesano, R., and Orci, L. (1988) Transforming growth factor B stimulates collagen-matrix contraction by fibroblasts: Implication for wound healing. Proc. Nat. Acad. Sci. U.S.A., 85:4894-4897. Rittenberg, T., Burd, D.A.R., and Ehrlich, H.P. (1990)Soluble factor(s1 in rat wound fluid inhibit fibroblast populated lattice contraction. Exp. Mol. Pathol., 52:132-140. Rowsell, A.R., (1984) The intra-uterine healing of foetal muscle wounds: Experimental study in the rat. Br. J. Plast. Surg., 37:635642. Sarber, R., Hull, B., Merrill, C., Soranno, T., and Bell, E. (1981) Regulation of proliferation of fibroblasts of low and high population doubling levels grown in collagen lattices. Mech. Age Dev., 17:107117. Schor, S.L. (1980) Cell proliferation and migration in collagen substrata in vitro. J. Cell Sci., 41:159-171. Siebert, J.W., Burd, D.A.R., McCarthy, J.G., Weinzwerg, J., and Ehrlich, H.P. (1990) Fetal wound healing: A biochemical study of scarless healing. Plast. Reconst. Surg., 85:495-502. Somasundaram, K., and Prathap, K. (1970) Intra-uterine healing of skin wounds in rabbit foetuses. J . Pathol., 100:81-86. Somasundaram, K., and Prathap, K. (1972) The effect of exclusion of amniotic fluid on intra-uterine healing of skin wounds in rabbit foetuses. J . Pathol., 107:127-130.