- Email: [email protected]
Wound Healing: The Role of Gap Junctional Communication in Rat Granulation Tissue Maturation
Experimental and Molecular Pathology 72, 10–16 (2002) doi:10.1006/exmp.2001.2406, available online at http://www.idealibrary.com on
Wound Healing: The Role of Gap Junctional Communication in Rat Granulation Tissue Maturation
K. E. Moyer, A. Davis, G. C. Saggers, D. R. Mackay, and H. P. Ehrlich1 Division of Plastic Surgery, Milton S. Hershey Medical Center, Hershey, Pennsylvania 17033
Received July 17, 2001
Granulation tissue maturation is dependent upon the orientation of collagen fibers and cell differentiation. Gap junctions are intercellular membrane gated channels that facilitate direct communication between cells known as gap junctional intercellular communication (GJIC). The hypothesis is that GJIC modulates the maturation of granulation tissue during wound repair. In vitro, GJIC optimizes fibroblast-populated collagen lattice contraction and influences cell morphology. It is reported that LiCl increases GJIC in cultured cardiac myocytes. Polyvinyl alcohol (PVA) sponge implants with central reservoirs were placed within separate subcutaneous pockets on the backs of adult male Sprague–Dawley rats. Each PVA implant received either 20 mM LiCl or saline injections on days 5, 7, and 10 after implantation. On day 11 implants were harvested and processed for light microscopy. By H&E staining LiCl-treated implants showed increased vascularization and decreased cell density compared to saline controls. Polarized light microscopy of Sirius red-stained specimens revealed more intense collagen fiber birefringence secondary to dense, parallel-organized collagen fiber bundles after LiCl treatment. This suggests that LiCl enhancement of GJIC between fibroblasts advances the maturation of granulation tissue. It is proposed that the degree of GJIC between granulation tissue fibroblasts influences both the quantity and the quality of granulation tissue deposited during the wound healing process. 䉷 2002 Elsevier Science Key Words: granulation tissue; wound healing; lithium chloride; PVA sponge implant; gap junctions; rat.
INTRODUCTION
The process of wound healing is dependent upon the coordination of numerous cellular processes. The initial phase of the wound repair process, the inflammatory phase, is heralded by the infiltration of neutrophils, followed closely by macrophages and finally by fibroblasts. The second phase of the wound healing process is the proliferative phase, which involves the migration and expansion of fibroblast cell numbers along with the deposition of a new connective tissue matrix. The connective tissue matrix, granulation tissue, is made up of collagen fiber bundles, capillary beds, and glycosaminoglycans. The final phase of repair, known as the remodeling phase, includes the reorientation and reorganization of granulation tissue into a scar. A high fibroblast density, thin randomly oriented collagen fiber bundles, and a rich capillary bed all characterize early granulation tissue development (Diegelmann et al., 1979). As granulation tissue matures, the collagen fiber bundles become denser and more tightly packed (Gabbiani et al., 1978). The capillary bed also matures, developing into distinct blood vessels that percolate throughout the developing scar tissue (Ehrlich, 1988; Ehrlich and Krummel, 1996; Berry et al., 1998). Considering the multiple changes in cellular and connective tissue composition during the maturation of granulation tissue, there needs to be an efficient, coordinated method of intercellular communication. One possible avenue for that
1
To whom correspondence should be addressed at the Division of Plastic Surgery H071, Hershey Medical Center, 500 University Drive, Hershey, PA 17033-0850. Fax: (717) 531-4339. E-mail: pehrlich@ psu.edu.
10
0014-4800/02 $35.00 䉷 2002 Elsevier Science All rights reserved.
GAP JUNCTIONS AND SCAR MATURATION
type of coordinated communication is through structures known as gap junctions. Gap junctions are anatomical intercellular channel structures that allow the direct passage of molecules 1000 MW or less between cells (Kumar and Gilula, 1996). Each gap junction channel is composed of two intramembranous hexameric subunits known as connexons. Two adjacent cells contribute one connexon each. Each connexon is composed of six connexins (Cx), which are members of a family of structural proteins (Musil and Goodenough, 1991; Kumar and Gilula, 1996). Currently there are about 13 different connexin gene products that are capable of creating a gap junction. Cx43 is the most abundant connexin protein identified within dermal fibroblasts. It is well accepted that gap junctions are responsible for cell–cell communications that play a role in embryological development as well as some disease processes. However, a role for gap junctional intercellular communication (GJIC) in the coordination of wound healing is not reported. Previous studies show gap junctions are present in fibroblasts derived from healing wounds (Abdullah et al., 1999; Ko et al., 2000). GJIC optimizes the contraction of fibroblast-populated collagen lattices, representing an in vitro model of wound contraction (Ehrlich et al., 2000). However, no experiments, in vitro or in vivo, identify a role for GJIC in granulation tissue maturation. The hypothesis is that GJIC plays a role in the maturation of granulation tissue. GJIC will be demonstrated in vivo by the parachute technique, employing granulation tissue derived from subcutaneously implanted sponges in rats (Goldberg et al., 1995). Further, the enhancement of GJIC is proposed to modulate the rate and quality of granulation tissue deposition. LiCl is known to upregulate the expression of Cx43 transcription and increase GJIC (Hedgepeth et al., 1997; Ai et al., 2000). We will investigate LiCl modulation of GJIC in cultured fibroblasts, utilizing the scrape loading technique as well as its modulation in vivo of the maturation of granulation tissue within sponge implants in rats (Meda et al., 1987; Ehrlich et al., 2000).
MATERIALS AND METHODS
Scrape loading. Human dermal fibroblasts (HF) derived from discarded foreskin were grown in monolayer and used between their 10th and 12th passage. Cells were plated in 35-mm tissue culture dishes and maintained in Delbecco’s modification of Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, referred to as complete
11 DMEM. Medium was changed twice a week until cells neared confluence. When the cells reached confluence, the plates of cells were divided into two experimental groups. The first group received 2 ml of complete DMEM containing 100 l of phosphate-buffered saline (PBS) and was designated as the control group. The second group received 2 ml of complete DMEM containing 20 mM LiCl and was the experimental group. The cells from each group were incubated at 37⬚C for 24 h, after which the cells were rinsed with PBS. A dye solution consisting of Lucifer Yellow (LY) 20 mg/ml in PBS and rhodamine (Rh)–dextran 5 mg/ml (Molecular Probes, Eugene, OR) was added to cover the fibroblast monolayer. A glass cutter purchased from a hardware store was used to scratch a fine linear wound in the cell monolayer. The injured fibroblasts were permeable to both dyes, which readily entered the damaged cells. Undamaged cells were unable to take up either dye through membrane diffusion. Instead, the uptake of LY dye by uninjured cells was limited to those cells with patent gap junction channels linking them to the scrape-injured fibroblasts. Within injured cells, only the LY dye could pass through patent gap junction channels established with the uninjured neighbors. The large-size Rh– dextran particles remained trapped within the scrape-injured cells. After a 2-min incubation period at room temperature, the dye solution was removed and the cells were rinsed with PBS (Boitano et al., 1992; Parker et al., 1994). The scratched monolayer was fixed in buffered 4% paraformaldehyde, pH 7.2. Only cells injured by the glass cutter wheel were labeled with both Rh–dextran and LY dyes. Uninjured cells linked to injured cells by gap junction channels were labeled with LY dye only. A coupling index was determined and reported by viewing the fibroblasts at the scratch mark with a Leica MZ12 fluorescence steromicroscope (Leica Microsystems Inc., Buffalo, NY) equipped with appropriate green and red fluorescent filters. The ratio of LY-labeled cells, which appeared as yellow-green fluorescent cells, to Rh–dextran-labeled cells, which appeared as red fluorescent cells, produced the coupling index. The greater the ratio of yellow-green cells to red cells, the higher the coupling index and the greater the number of fibroblasts associated by GJIC. Rat studies. Polyvinyl alcohol (PVA) sponge disc donuts 1.2 cm in diameter and 0.3 cm thick with a central hole 0.5 cm in diameter were manufactured with a thin polycarbonate plastic 1.2-cm-diameter disc attached to the bottom of each PVA donut with silastic glue. The PVA implants had a central reservoir for repeated injections after surgical implantation. Three Sprague–Dawley male rats (each 350 g) were anesthetized by halothane inhalation. Each rat’s dorsum was shaved
12 and cleaned with 70% alcohol. A 2-cm incision was made with a scalpel over the dorsal midline, and a subcutaneous pocket was created at four specific locations in each rat. The first two pockets were located 2 cm caudal to the forelimbs in a transverse plane. The remaining two pockets were created 2 cm cranial to the hindlimbs, again in a transverse plane. A PVA donut sponge was inserted into each subcutaneous pocket with the plastic base resting on the muscle bed and the PVA sponge located just below the dermis. After the implantation of all four sponge donuts, the incision was closed with 4-0 nylon suture. The rats were then returned to their cages upon recovery from anesthesia. Two of the rats on postimplantation days 5, 7, and 10 were again anesthetized with halothane, the skin over the PVA implant was cleaned with 70% alcohol, and 0.1 ml of wound fluid was removed from each PVA donut implant reservoir. Following the removal of wound fluid, each PVA sponge received an injection of 0.1 ml of either PBS or 20 mM LiCl in the reservoir. PBS was injected into the front left and rear right PVA donuts and 0.1 ml of 20 mM LiCl in the remaining implants in a similar fashion. On postimplantation day 11, the rats were sacrificed and the PVA sponge donuts removed along with their subcutaneous capsules. The PVA implants were fixed in 10% buffered formaldehyde, paraffin embedded, sectioned, and examined by light microscopy. The sections were stained with either hematoxylin and eosin (H&E) or Sirius red. The stained slides were viewed with an Olympus BH-2 microscope (Olympus America Inc., Melville, NY) equipped with bright and polarized light capabilities. Photographs were taken with Ektacrome 400 ASA slide film (Eastman Kodak Co., Rochester, NY). Parachute technique. The remaining rat with implants was used to identify GJIC in vivo between fibroblasts within granulation tissue by the parachute technique on postimplantation day 7. Using the protocol devised by Goldberg et al. (1995), HF fibroblasts were grown to confluence in complete DMEM in a 60-mm tissue culture dish. The medium was removed and the cells were washed four times in 2 ml of 300 mM glucose in 30 mM Hepes buffer. The cells were labeled by adding 20 l of Calcein AM at 1 mg/ml in DMSO and 16 l of 1,1⬘-dihexadecyl-3,3,3⬘,3⬘-tetramethylindocarbocyanine perchlorate (DiI) Molecular Probes) at 2.5 mg/ ml in 70% ethanol with 2 ml of 300 mM glucose in 30 mM Hepes. DiI is incapable of passing through patent gap junction channels, since once taken up by the cell it is permanently incorporated into the cell membrane. However, Calcein AM remains free in the cytoplasm and can easily pass through patent gap junction channels. The cells were
MOYER ET AL.
incubated for 10 min at 37⬚C. The dye solution was removed, and the cells were rinsed with PBS and released into suspension by trypsinization. The dye-loaded cells, or paratroopers, were suspended in 2 ml of complete DMEM and taken up in a 3-ml sterile syringe with an attached 16-gauge needle. Each reservoir had 0.1 ml of wound fluid removed, which was replaced with 0.1 ml of the paratrooper cell suspension. After 24 h, the PVA reservoir was removed with its subcutaneous capsule intact. The PVA reservoir was then divided into vertical sections with a scalpel and fixed in buffered 4% paraformaldehyde. Utilizing the fluorescence stereomicroscope, the DiI- and Calcein-positive cells were visualized by using their appropriate filters. Dye-loaded fibroblasts, or paratrooper cells, had both green fluorescence, indicating Calcein AM dye, and red fluorescence, indicating DiI dye presence. Only the paratroopers had both red and green fluorescence. Any cell with only green fluorescence indicated a cell coupled to a paratrooper by patent gap junction channels.
RESULTS
Scrape loading. Under the fluorescence stereomicroscope, scrape-injured cells accumulated the Rh–dextran-labeled particles, which gave the cells a red appearance when viewed with the red fluorescent filter. The LY dye accumulated in both scrape-injured cells and in cells that were coupled to them through patent gap junctions. With a green fluorescent filter, cells containing LY appeared yellow-green. The coupling of injured and uninjured cells to one another through GJIC was directly proportional to the coupling index ratio of cells showing yellow-green fluorescence to those showing only red fluorescence. Both the control and LiCltreated fibroblasts demonstrated some degree of coupling because cells with only yellow-green fluorescence were seen in both groups. LY-containing cells, identified by their yellow-green flouresence, were counted and compared to fluorescent-red Rh–dextran-labeled cells in each group. The ratios were reported as the coupling index (Table 1). The coupling index for control PBS-treated fibroblasts was 3.1 ⫾ 0.5 (sampling areas counted ⫽ 8). A coupling index of 3.1 indicated that for every red fluorescent cell there were 3.1 yellow-green fluorescent cells. The coupling index for LiCl-treated fibroblasts was 6.9 ⫾ 0.5 (sampling areas counted ⫽ 8). The higher coupling index of 6.9 for LiCltreated cells indicated that for every red fluorescent cell there
13
GAP JUNCTIONS AND SCAR MATURATION TABLE 1 Coupling Index of PBS Control Cultured Fibroblasts and LiCl-Treated Cultured Fibroblasts after Scape Loading with LY and Rh–Dextran Treatment group PBS Control LiCl Treated
Coupling index mean
SD
3.1 6.9 a
0.5 0.5
Note. The coupling index mean represents eight sampling areas on each scrape-loaded tissue culture plate. The coupling index represents a quantification of GJIC, where the higher the coupling index, the greater the cell–cell communication through gap junction channels. a Based upon the Student t test, the difference is significant at P ⱕ 0.01.
were 6.9 yellow-green fluorescent cells. By the Student t test this difference was significant at P ⱕ 0.01. The coupling index for fibroblasts treated with LiCl was significantly increased compared to that of the PBS controls, indicating greater GJIC. Parachute technique. Utilizing the fluorescence stereomicroscope, the PVA sponges treated with the preloaded HF fibroblasts as described above were visualized. DiI was permanently incorporated into the cell membrane and identified the preloaded HF fibroblasts, or paratroopers. Hence the red fluorescent-stained cells identified the preloaded human fibroblast paratroopers. Calcein AM dye taken up by the HF fibroblast paratroopers accumulated in the cytoplasmic pool of the cell and gave the paratroopers a green fluorescence. Calcein dye, unlike DiI, is capable of passing from a paratrooper cell to a cell in developing granulation tissue through patent gap junction channels. Figure 1A shows a red fluorescence DiI-labeled HF fibroblast paratrooper (arrow). Figure 1B shows the same HF fibroblast paratrooper (arrow), now under a green fluorescence filter. Another adjacent cell also stained with green fluorescence but was free of any red fluorescence due to the transfer of the Calcein AM dye through a patent gap junction channel into a coupled rat granulation tissue fibroblast. The green fluorescent-onlystaining cell was free of any red fluorescence. The abovementioned green fluorescent-stained cell could only be a coupled rat granulation tissue fibroblast and not a paratrooper. Histology. H&E-stained sections from the PVA sponge implants revealed differences between control and LiCl treatment groups. A higher density of fibroblasts was noted in the control PVA sponge implants than in the LiCl-treated implants. It appears that there was more connective tissue separating the cell types in the sponge implant treated with LiCl (Figs. 2A and 2B). In addition, the density of blood
vessels within the LiCl-treated PVA implants was increased compared to that in the saline-treated controls. LiCl appeared to increase the number of blood vessels that accumulated within the developing granulation tissue, indicating that the granulation tissue was more mature at 11 days postimplantation. Sirius red staining coupled with polarized light viewing revealed the organization of collagen fiber bundles within the capsule of the PVA implants. LiCl-treated PVA implants demonstrated increased collagen fiber birefringence intensity compared to the saline-treated control implants (Figs. 3A and 3B). The increased birefringence is attributed to the thicker, more organized collagen fiber bundles deposited within the LiCl-treated implants, indicating that the granulation tissue in the LiCl-treated sponge implant was more
FIG. 1. (A) Prelabeled HF fibroblast, or paratrooper (arrow), in a PVA sponge implant viewed under a fluorescence filter specific for DiI (magnification, 20⫻). (B) Prelabeled HF fibroblast, or paratrooper (arrow), and its coupled granulation tissue fibroblasts in a PVA sponge implant viewed under a fluorescence filter specific for Calcein AM (magnification, 20⫻). The image area is the same as presented in (A).
14
MOYER ET AL.
inositol triphosphate have all been shown to pass between cells through gap junction channels (Charles et al., 1991; Boitano et al., 1992; Kumar and Gilula, 1996). Gap junctions are gated pores that are in either an opened or a closed configuration. Alterations in tyrosine phosphorylation have been shown to affect channel conformation and gap junction channel patency (Musil and Goodenough, 1991). GJIC and tyrosine phosphorylation are associated with changes in intracellular calcium concentration (Lee et al., 1992). Gap junctions allow cells to directly communicate with surrounding cells based on intracellular and extracellular signals or events. In response to trauma, gap junctions have been shown to close, preventing the transfer of molecules between cells through GJIC (Parker et al., 1994). By the termination of
FIG. 2. (A) Histology of PVA sponge implant stained with H&E. The implant was treated with saline and represents the control. Note the increased cell density and minimal vascularization (magnification, 160⫻). (B) Histology of PVA sponge implant stained with H&E. The implant was treated with 20 mM LiCl. Note the decreased cell density and increase in vascularization compared to (A) (magnification, 160⫻).
mature at 11 days postimplantation than that in the salinetreated control sponge.
DISCUSSION The process of granulation tissue maturation is a complex one involving the coordination of numerous cellular activities. A possible mechanism for coordination is a communication link between fibroblasts within granulation tissue by GJIC. Gap junctions facilitate a number of processes, such as the transfer of second messengers between cells without the second messengers entering the extracellular space. Second messengers such as cAMP, cGMP, calcium waves, and
FIG. 3. (A) Histology of PVA sponge implant stained with Sirius red and viewed under polarized light. The figure represents the salinetreated control implant (magnification, 90⫻). (B) Histology of PVA sponge implant stained with Sirius red and viewed under polarized light. The image represents the LiCl-treated implant. Note the increased collagen birefringence and increased collagen fiber density compared to that in (A), indicating the presence of more mature collagen fiber bundles (magnification, 90⫻).
15
GAP JUNCTIONS AND SCAR MATURATION
GJIC, in which the channel pore between cells is closed after cellular trauma, the integrity of uninjured cells will be maintained by minimizing their coupling to injured cells. Termination of that communication allows uninvolved cells to be independent from injured, dying cells. Effective transfer of nutrients, such as oxygen between fibroblasts, is required for optimizing the wound healing process (Knighton et al., 1981). Does the patency of gap junction channels play a role in the metabolism of fibroblasts within granulation tissue? The speculation is that more efficient transfer of nutrients (glucose, amino acids, and oxygen) from the capillaries to the fibroblasts, as well as the transfer of CO2 from the fibroblasts to the capillaries, will benefit from GJIC. In addition to diffusion through the extracellular compartment, there will be direct transfer between cells through gap junction channels. The more efficient the fibroblast metabolism is, the more efficient the synthesis and deposition of a new connective tissue matrix will be and the less time will be required for granulation tissue maturation. The passage of Calcein dye through gap junctions within fibroblasts residing in granulation tissue is demonstrated by the parachute technique. Calcein dye, which is smaller than 1000 MW, is able to freely pass through patent gap junction channels between the paratrooper fibroblasts and the granulation tissue fibroblasts (Salomon et al., 1988; White et al., 1995). The cell membrane covalently linked DiI dye cannot pass through gap junction channels; thus, the paratrooper and granulation tissue fibroblasts can be readily identified (Goldberg et al., 1995). Through this technique we demonstrate in vivo that the granulation tissue of a PVA sponge implant contains fibroblasts that exhibit GJIC. Granulation tissue accumulates within PVA sponge implants. The degree of cellularity, the organization of the collagen fibers, and the presence of distinct blood vessels are all markers of granulation tissue maturation. Granulation tissue maturation involves a reduction in fibroblast density and an increase in connective tissue deposition between cells. As granulation tissue matures, capillaries coalesce into blood vessels (Knighton et al., 1981; Ehrlich, 1988; Ehrlich and Krummel, 1996; Berry et al., 1998). These morphological characteristics occur earlier in LiCl-treated PVA sponge implants. Collagen fibers are known to become thicker and more organized as granulation tissue matures (Li et al., 1980). Utilizing Sirius red staining, the birefringence of collagen fibers becomes more intense as they become thicker and more organized (Constantine and Mowry, 1968; Li et al., 1980). When compared to saline-treated controls, LiCltreated PVA implants showed an increase in collagen fiber birefringence intensity and organization; again, evidence that granulation tissue from LiCl-treated PVA implants matures
more rapidly. It is proposed that the enhancement of gap junctions and GJIC advances the maturation of granulation tissue in vivo. How does LiCl alter GJIC? One effect of LiCl is to mimic the wingless/int (wnt-1) signaling pathway, leading to the accumulation of the effector protein, -catenin (Hedgepeth et al., 1997; Ai et al., 2000). The accumulation of -catenin in the cell nucleus increases the expression of connexins, specifically Cx43, and GJIC in cardiac myocytes (Ai et al., 2000). The increase in the coupling index observed with scrape-loaded HF fibroblasts that received LiCl confirms those reported findings. GJIC influences the rate and final degree of rat osteoblastand human fibroblast-populated collagen lattice contraction (Bowman et al., 1998; Ehrlich et al., 2000). By polarized light microscopy, the increase in the rate of collagen lattice contraction is consistent with an improved organization and orientation of collagen fibers. Those studies support a role for GJIC in the organization of collagen fiber bundles in vitro similar to that for those deposited in granulation tissue in vivo. Here an increase in GJIC through the addition of LiCl leads to the earlier maturation of granulation tissue. It is proposed that increasing GJIC between fibroblasts within granulation tissue enhances the organization of the collagen fiber bundles and accelerates the onset of the remodeling phase of repair. Further studies examining the role of GJIC in the repair process may increase our understanding of how to better control the fibrotic process.
REFERENCES
Abdullah, K. M., Luthra, G., et al. (1999). Cell-to-cell communication and expression of gap junctional proteins in human diabetic and nondiabetic skin fibroblasts: Effects of basic fibroblast growth factor. Endocrine 10, 35–41. Ai, Z., Fischer, A., et al. (2000). Wnt-1 regulation of connexin43 in cardiac myocytes. J. Clin. Invest. 105, 161–171. Berry, D. P., Harding, K. G., et al. (1998). Human wound contraction: Collagen organization, fibroblasts, and myofibroblasts. Plast. Reconstr. Surg. 102, 124–134. Boitano, S., Dirksen, E. R., et al. (1992). Intercellular propagation of calcium waves mediated by inositol trisphosphate. Science 258, 292–295. Bowman, N. N., Donahue, H. J., et al. (1998). Gap junctional intercellular communication contributes to the contraction of rat osteoblast populated collagen lattices. J. Bone Miner. Res. 13, 1700–1706. Charles, A. C., Merrill, J. E., et al. (1991). Intercellular signaling in glial cells: Calcium waves and oscillations in response to mechanical stimulation and glutamate. Neuron 6, 983–992.
16 Constantine, V. S., and Mowry, R. W. (1968). Selective staining of human dermal collagen. II. The use of picrosirius red F3BA with polarization microscopy. J. Invest. Dermatol. 50, 419–423. Diegelmann, R. F., Cohen, I. K., et al. (1979). Growth kinetics and collagen synthesis of normal skin, normal scar and keloid fibroblasts in vitro. J. Cell. Physiol. 98, 341–346. Ehrlich, H. P. (1988). The role of connective tissue matrix in wound healing. Prog. Clin. Biol. Res. 266, 243–258. Ehrlich, H. P., Gabbiani, G., et al. (2000). Cell coupling modulates the contraction of fibroblast-populated collagen lattices. J. Cell. Physiol. 184, 86–92. Ehrlich, H. P., and Krummel, T. M. (1996). The regulation of wound healing from a connective tissue aspect. Wound Repair Regener. 4, 203–210. Gabbiani, G., Chaponnier, C., et al. (1978). Cytoplasmic filaments and gap junctions in epithelial cells and myofibroblasts during wound healing. J. Cell. Biol. 76, 561–568. Goldberg, G. S., Bechberger, J. F., et al. (1995). A pre-loading method of evaluating gap junctional communication by fluorescent dye transfer. Biotechniques 18, 490–497. Hedgepeth, C. M., Conrad, L. J., et al. (1997). Activation of the Wnt signaling pathway: A molecular mechanism for lithium action. Dev. Biol. 185, 82–91.
MOYER ET AL.
Knighton, D. R., Silver, I. A., et al. (1981). Regulation of woundhealing angiogenesis-effect of oxygen gradients and inspired oxygen concentration. Surgery 90, 262–270. Ko, K., Arora, P., et al. (2000). Biochemical and functional characterization of intercellular adhesion and gap junctions in fibroblasts. Am. J. Physiol. Cell. Physiol. 279, C147–C157. Kumar, N., and Gilula, N. (1996). The gap junction communication channel. Cell. 84, 381–388. Lee, S. W., Tomasetto, C., et al. (1992). Transcriptional downregulation of gap-junction proteins blocks junctional communication in human mammary tumor cell lines. J. Cell. Biol. 118, 1213–1221. Li, A. K., Ehrlich, H. P., et al. (1980). Differences in healing of skin wounds caused by burn and freeze injuries. Ann. Surg. 191, 244–248. Meda, P., Bruzzone, R., et al. (1987). Gap junctional coupling modulates secretion of exocrine pancreas. Proc. Natl. Acad. Sci. USA 84, 4901–4904. Musil, L. S., and Goodenough, D. A. (1991). Biochemical analysis of connexin43 intracellular transport, phosphorylation, and assembly into gap junctional plaques. J. Cell. Biol. 115, 1357–1374. Parker, S. B., Hertzberg, E. L., et al. (1994). Modulation of gap junctionmediated intercellular communication in embryonic chick mesenchyme during tissue remodeling in vitro. Cell. Tissue Res. 275, 215–224. Salomon, D., Saurat, J. H., et al. (1988). Cell-to-cell communication within intact human skin. J. Clin. Invest. 82, 248–254. White, T. W., Bruzzone, R., et al. (1995). The connexin family of intercellular channel forming proteins. Kidney Int. 48, 1148–1157.
Wound Healing: The Role of Gap Junctional Communication in Rat Granulation Tissue Maturation
K. E. Moyer, A. Davis, G. C. Saggers, D. R. Mackay, and H. P. Ehrlich1 Division of Plastic Surgery, Milton S. Hershey Medical Center, Hershey, Pennsylvania 17033
Received July 17, 2001
Granulation tissue maturation is dependent upon the orientation of collagen fibers and cell differentiation. Gap junctions are intercellular membrane gated channels that facilitate direct communication between cells known as gap junctional intercellular communication (GJIC). The hypothesis is that GJIC modulates the maturation of granulation tissue during wound repair. In vitro, GJIC optimizes fibroblast-populated collagen lattice contraction and influences cell morphology. It is reported that LiCl increases GJIC in cultured cardiac myocytes. Polyvinyl alcohol (PVA) sponge implants with central reservoirs were placed within separate subcutaneous pockets on the backs of adult male Sprague–Dawley rats. Each PVA implant received either 20 mM LiCl or saline injections on days 5, 7, and 10 after implantation. On day 11 implants were harvested and processed for light microscopy. By H&E staining LiCl-treated implants showed increased vascularization and decreased cell density compared to saline controls. Polarized light microscopy of Sirius red-stained specimens revealed more intense collagen fiber birefringence secondary to dense, parallel-organized collagen fiber bundles after LiCl treatment. This suggests that LiCl enhancement of GJIC between fibroblasts advances the maturation of granulation tissue. It is proposed that the degree of GJIC between granulation tissue fibroblasts influences both the quantity and the quality of granulation tissue deposited during the wound healing process. 䉷 2002 Elsevier Science Key Words: granulation tissue; wound healing; lithium chloride; PVA sponge implant; gap junctions; rat.
INTRODUCTION
The process of wound healing is dependent upon the coordination of numerous cellular processes. The initial phase of the wound repair process, the inflammatory phase, is heralded by the infiltration of neutrophils, followed closely by macrophages and finally by fibroblasts. The second phase of the wound healing process is the proliferative phase, which involves the migration and expansion of fibroblast cell numbers along with the deposition of a new connective tissue matrix. The connective tissue matrix, granulation tissue, is made up of collagen fiber bundles, capillary beds, and glycosaminoglycans. The final phase of repair, known as the remodeling phase, includes the reorientation and reorganization of granulation tissue into a scar. A high fibroblast density, thin randomly oriented collagen fiber bundles, and a rich capillary bed all characterize early granulation tissue development (Diegelmann et al., 1979). As granulation tissue matures, the collagen fiber bundles become denser and more tightly packed (Gabbiani et al., 1978). The capillary bed also matures, developing into distinct blood vessels that percolate throughout the developing scar tissue (Ehrlich, 1988; Ehrlich and Krummel, 1996; Berry et al., 1998). Considering the multiple changes in cellular and connective tissue composition during the maturation of granulation tissue, there needs to be an efficient, coordinated method of intercellular communication. One possible avenue for that
1
To whom correspondence should be addressed at the Division of Plastic Surgery H071, Hershey Medical Center, 500 University Drive, Hershey, PA 17033-0850. Fax: (717) 531-4339. E-mail: pehrlich@ psu.edu.
10
0014-4800/02 $35.00 䉷 2002 Elsevier Science All rights reserved.
GAP JUNCTIONS AND SCAR MATURATION
type of coordinated communication is through structures known as gap junctions. Gap junctions are anatomical intercellular channel structures that allow the direct passage of molecules 1000 MW or less between cells (Kumar and Gilula, 1996). Each gap junction channel is composed of two intramembranous hexameric subunits known as connexons. Two adjacent cells contribute one connexon each. Each connexon is composed of six connexins (Cx), which are members of a family of structural proteins (Musil and Goodenough, 1991; Kumar and Gilula, 1996). Currently there are about 13 different connexin gene products that are capable of creating a gap junction. Cx43 is the most abundant connexin protein identified within dermal fibroblasts. It is well accepted that gap junctions are responsible for cell–cell communications that play a role in embryological development as well as some disease processes. However, a role for gap junctional intercellular communication (GJIC) in the coordination of wound healing is not reported. Previous studies show gap junctions are present in fibroblasts derived from healing wounds (Abdullah et al., 1999; Ko et al., 2000). GJIC optimizes the contraction of fibroblast-populated collagen lattices, representing an in vitro model of wound contraction (Ehrlich et al., 2000). However, no experiments, in vitro or in vivo, identify a role for GJIC in granulation tissue maturation. The hypothesis is that GJIC plays a role in the maturation of granulation tissue. GJIC will be demonstrated in vivo by the parachute technique, employing granulation tissue derived from subcutaneously implanted sponges in rats (Goldberg et al., 1995). Further, the enhancement of GJIC is proposed to modulate the rate and quality of granulation tissue deposition. LiCl is known to upregulate the expression of Cx43 transcription and increase GJIC (Hedgepeth et al., 1997; Ai et al., 2000). We will investigate LiCl modulation of GJIC in cultured fibroblasts, utilizing the scrape loading technique as well as its modulation in vivo of the maturation of granulation tissue within sponge implants in rats (Meda et al., 1987; Ehrlich et al., 2000).
MATERIALS AND METHODS
Scrape loading. Human dermal fibroblasts (HF) derived from discarded foreskin were grown in monolayer and used between their 10th and 12th passage. Cells were plated in 35-mm tissue culture dishes and maintained in Delbecco’s modification of Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, referred to as complete
11 DMEM. Medium was changed twice a week until cells neared confluence. When the cells reached confluence, the plates of cells were divided into two experimental groups. The first group received 2 ml of complete DMEM containing 100 l of phosphate-buffered saline (PBS) and was designated as the control group. The second group received 2 ml of complete DMEM containing 20 mM LiCl and was the experimental group. The cells from each group were incubated at 37⬚C for 24 h, after which the cells were rinsed with PBS. A dye solution consisting of Lucifer Yellow (LY) 20 mg/ml in PBS and rhodamine (Rh)–dextran 5 mg/ml (Molecular Probes, Eugene, OR) was added to cover the fibroblast monolayer. A glass cutter purchased from a hardware store was used to scratch a fine linear wound in the cell monolayer. The injured fibroblasts were permeable to both dyes, which readily entered the damaged cells. Undamaged cells were unable to take up either dye through membrane diffusion. Instead, the uptake of LY dye by uninjured cells was limited to those cells with patent gap junction channels linking them to the scrape-injured fibroblasts. Within injured cells, only the LY dye could pass through patent gap junction channels established with the uninjured neighbors. The large-size Rh– dextran particles remained trapped within the scrape-injured cells. After a 2-min incubation period at room temperature, the dye solution was removed and the cells were rinsed with PBS (Boitano et al., 1992; Parker et al., 1994). The scratched monolayer was fixed in buffered 4% paraformaldehyde, pH 7.2. Only cells injured by the glass cutter wheel were labeled with both Rh–dextran and LY dyes. Uninjured cells linked to injured cells by gap junction channels were labeled with LY dye only. A coupling index was determined and reported by viewing the fibroblasts at the scratch mark with a Leica MZ12 fluorescence steromicroscope (Leica Microsystems Inc., Buffalo, NY) equipped with appropriate green and red fluorescent filters. The ratio of LY-labeled cells, which appeared as yellow-green fluorescent cells, to Rh–dextran-labeled cells, which appeared as red fluorescent cells, produced the coupling index. The greater the ratio of yellow-green cells to red cells, the higher the coupling index and the greater the number of fibroblasts associated by GJIC. Rat studies. Polyvinyl alcohol (PVA) sponge disc donuts 1.2 cm in diameter and 0.3 cm thick with a central hole 0.5 cm in diameter were manufactured with a thin polycarbonate plastic 1.2-cm-diameter disc attached to the bottom of each PVA donut with silastic glue. The PVA implants had a central reservoir for repeated injections after surgical implantation. Three Sprague–Dawley male rats (each 350 g) were anesthetized by halothane inhalation. Each rat’s dorsum was shaved
12 and cleaned with 70% alcohol. A 2-cm incision was made with a scalpel over the dorsal midline, and a subcutaneous pocket was created at four specific locations in each rat. The first two pockets were located 2 cm caudal to the forelimbs in a transverse plane. The remaining two pockets were created 2 cm cranial to the hindlimbs, again in a transverse plane. A PVA donut sponge was inserted into each subcutaneous pocket with the plastic base resting on the muscle bed and the PVA sponge located just below the dermis. After the implantation of all four sponge donuts, the incision was closed with 4-0 nylon suture. The rats were then returned to their cages upon recovery from anesthesia. Two of the rats on postimplantation days 5, 7, and 10 were again anesthetized with halothane, the skin over the PVA implant was cleaned with 70% alcohol, and 0.1 ml of wound fluid was removed from each PVA donut implant reservoir. Following the removal of wound fluid, each PVA sponge received an injection of 0.1 ml of either PBS or 20 mM LiCl in the reservoir. PBS was injected into the front left and rear right PVA donuts and 0.1 ml of 20 mM LiCl in the remaining implants in a similar fashion. On postimplantation day 11, the rats were sacrificed and the PVA sponge donuts removed along with their subcutaneous capsules. The PVA implants were fixed in 10% buffered formaldehyde, paraffin embedded, sectioned, and examined by light microscopy. The sections were stained with either hematoxylin and eosin (H&E) or Sirius red. The stained slides were viewed with an Olympus BH-2 microscope (Olympus America Inc., Melville, NY) equipped with bright and polarized light capabilities. Photographs were taken with Ektacrome 400 ASA slide film (Eastman Kodak Co., Rochester, NY). Parachute technique. The remaining rat with implants was used to identify GJIC in vivo between fibroblasts within granulation tissue by the parachute technique on postimplantation day 7. Using the protocol devised by Goldberg et al. (1995), HF fibroblasts were grown to confluence in complete DMEM in a 60-mm tissue culture dish. The medium was removed and the cells were washed four times in 2 ml of 300 mM glucose in 30 mM Hepes buffer. The cells were labeled by adding 20 l of Calcein AM at 1 mg/ml in DMSO and 16 l of 1,1⬘-dihexadecyl-3,3,3⬘,3⬘-tetramethylindocarbocyanine perchlorate (DiI) Molecular Probes) at 2.5 mg/ ml in 70% ethanol with 2 ml of 300 mM glucose in 30 mM Hepes. DiI is incapable of passing through patent gap junction channels, since once taken up by the cell it is permanently incorporated into the cell membrane. However, Calcein AM remains free in the cytoplasm and can easily pass through patent gap junction channels. The cells were
MOYER ET AL.
incubated for 10 min at 37⬚C. The dye solution was removed, and the cells were rinsed with PBS and released into suspension by trypsinization. The dye-loaded cells, or paratroopers, were suspended in 2 ml of complete DMEM and taken up in a 3-ml sterile syringe with an attached 16-gauge needle. Each reservoir had 0.1 ml of wound fluid removed, which was replaced with 0.1 ml of the paratrooper cell suspension. After 24 h, the PVA reservoir was removed with its subcutaneous capsule intact. The PVA reservoir was then divided into vertical sections with a scalpel and fixed in buffered 4% paraformaldehyde. Utilizing the fluorescence stereomicroscope, the DiI- and Calcein-positive cells were visualized by using their appropriate filters. Dye-loaded fibroblasts, or paratrooper cells, had both green fluorescence, indicating Calcein AM dye, and red fluorescence, indicating DiI dye presence. Only the paratroopers had both red and green fluorescence. Any cell with only green fluorescence indicated a cell coupled to a paratrooper by patent gap junction channels.
RESULTS
Scrape loading. Under the fluorescence stereomicroscope, scrape-injured cells accumulated the Rh–dextran-labeled particles, which gave the cells a red appearance when viewed with the red fluorescent filter. The LY dye accumulated in both scrape-injured cells and in cells that were coupled to them through patent gap junctions. With a green fluorescent filter, cells containing LY appeared yellow-green. The coupling of injured and uninjured cells to one another through GJIC was directly proportional to the coupling index ratio of cells showing yellow-green fluorescence to those showing only red fluorescence. Both the control and LiCltreated fibroblasts demonstrated some degree of coupling because cells with only yellow-green fluorescence were seen in both groups. LY-containing cells, identified by their yellow-green flouresence, were counted and compared to fluorescent-red Rh–dextran-labeled cells in each group. The ratios were reported as the coupling index (Table 1). The coupling index for control PBS-treated fibroblasts was 3.1 ⫾ 0.5 (sampling areas counted ⫽ 8). A coupling index of 3.1 indicated that for every red fluorescent cell there were 3.1 yellow-green fluorescent cells. The coupling index for LiCl-treated fibroblasts was 6.9 ⫾ 0.5 (sampling areas counted ⫽ 8). The higher coupling index of 6.9 for LiCltreated cells indicated that for every red fluorescent cell there
13
GAP JUNCTIONS AND SCAR MATURATION TABLE 1 Coupling Index of PBS Control Cultured Fibroblasts and LiCl-Treated Cultured Fibroblasts after Scape Loading with LY and Rh–Dextran Treatment group PBS Control LiCl Treated
Coupling index mean
SD
3.1 6.9 a
0.5 0.5
Note. The coupling index mean represents eight sampling areas on each scrape-loaded tissue culture plate. The coupling index represents a quantification of GJIC, where the higher the coupling index, the greater the cell–cell communication through gap junction channels. a Based upon the Student t test, the difference is significant at P ⱕ 0.01.
were 6.9 yellow-green fluorescent cells. By the Student t test this difference was significant at P ⱕ 0.01. The coupling index for fibroblasts treated with LiCl was significantly increased compared to that of the PBS controls, indicating greater GJIC. Parachute technique. Utilizing the fluorescence stereomicroscope, the PVA sponges treated with the preloaded HF fibroblasts as described above were visualized. DiI was permanently incorporated into the cell membrane and identified the preloaded HF fibroblasts, or paratroopers. Hence the red fluorescent-stained cells identified the preloaded human fibroblast paratroopers. Calcein AM dye taken up by the HF fibroblast paratroopers accumulated in the cytoplasmic pool of the cell and gave the paratroopers a green fluorescence. Calcein dye, unlike DiI, is capable of passing from a paratrooper cell to a cell in developing granulation tissue through patent gap junction channels. Figure 1A shows a red fluorescence DiI-labeled HF fibroblast paratrooper (arrow). Figure 1B shows the same HF fibroblast paratrooper (arrow), now under a green fluorescence filter. Another adjacent cell also stained with green fluorescence but was free of any red fluorescence due to the transfer of the Calcein AM dye through a patent gap junction channel into a coupled rat granulation tissue fibroblast. The green fluorescent-onlystaining cell was free of any red fluorescence. The abovementioned green fluorescent-stained cell could only be a coupled rat granulation tissue fibroblast and not a paratrooper. Histology. H&E-stained sections from the PVA sponge implants revealed differences between control and LiCl treatment groups. A higher density of fibroblasts was noted in the control PVA sponge implants than in the LiCl-treated implants. It appears that there was more connective tissue separating the cell types in the sponge implant treated with LiCl (Figs. 2A and 2B). In addition, the density of blood
vessels within the LiCl-treated PVA implants was increased compared to that in the saline-treated controls. LiCl appeared to increase the number of blood vessels that accumulated within the developing granulation tissue, indicating that the granulation tissue was more mature at 11 days postimplantation. Sirius red staining coupled with polarized light viewing revealed the organization of collagen fiber bundles within the capsule of the PVA implants. LiCl-treated PVA implants demonstrated increased collagen fiber birefringence intensity compared to the saline-treated control implants (Figs. 3A and 3B). The increased birefringence is attributed to the thicker, more organized collagen fiber bundles deposited within the LiCl-treated implants, indicating that the granulation tissue in the LiCl-treated sponge implant was more
FIG. 1. (A) Prelabeled HF fibroblast, or paratrooper (arrow), in a PVA sponge implant viewed under a fluorescence filter specific for DiI (magnification, 20⫻). (B) Prelabeled HF fibroblast, or paratrooper (arrow), and its coupled granulation tissue fibroblasts in a PVA sponge implant viewed under a fluorescence filter specific for Calcein AM (magnification, 20⫻). The image area is the same as presented in (A).
14
MOYER ET AL.
inositol triphosphate have all been shown to pass between cells through gap junction channels (Charles et al., 1991; Boitano et al., 1992; Kumar and Gilula, 1996). Gap junctions are gated pores that are in either an opened or a closed configuration. Alterations in tyrosine phosphorylation have been shown to affect channel conformation and gap junction channel patency (Musil and Goodenough, 1991). GJIC and tyrosine phosphorylation are associated with changes in intracellular calcium concentration (Lee et al., 1992). Gap junctions allow cells to directly communicate with surrounding cells based on intracellular and extracellular signals or events. In response to trauma, gap junctions have been shown to close, preventing the transfer of molecules between cells through GJIC (Parker et al., 1994). By the termination of
FIG. 2. (A) Histology of PVA sponge implant stained with H&E. The implant was treated with saline and represents the control. Note the increased cell density and minimal vascularization (magnification, 160⫻). (B) Histology of PVA sponge implant stained with H&E. The implant was treated with 20 mM LiCl. Note the decreased cell density and increase in vascularization compared to (A) (magnification, 160⫻).
mature at 11 days postimplantation than that in the salinetreated control sponge.
DISCUSSION The process of granulation tissue maturation is a complex one involving the coordination of numerous cellular activities. A possible mechanism for coordination is a communication link between fibroblasts within granulation tissue by GJIC. Gap junctions facilitate a number of processes, such as the transfer of second messengers between cells without the second messengers entering the extracellular space. Second messengers such as cAMP, cGMP, calcium waves, and
FIG. 3. (A) Histology of PVA sponge implant stained with Sirius red and viewed under polarized light. The figure represents the salinetreated control implant (magnification, 90⫻). (B) Histology of PVA sponge implant stained with Sirius red and viewed under polarized light. The image represents the LiCl-treated implant. Note the increased collagen birefringence and increased collagen fiber density compared to that in (A), indicating the presence of more mature collagen fiber bundles (magnification, 90⫻).
15
GAP JUNCTIONS AND SCAR MATURATION
GJIC, in which the channel pore between cells is closed after cellular trauma, the integrity of uninjured cells will be maintained by minimizing their coupling to injured cells. Termination of that communication allows uninvolved cells to be independent from injured, dying cells. Effective transfer of nutrients, such as oxygen between fibroblasts, is required for optimizing the wound healing process (Knighton et al., 1981). Does the patency of gap junction channels play a role in the metabolism of fibroblasts within granulation tissue? The speculation is that more efficient transfer of nutrients (glucose, amino acids, and oxygen) from the capillaries to the fibroblasts, as well as the transfer of CO2 from the fibroblasts to the capillaries, will benefit from GJIC. In addition to diffusion through the extracellular compartment, there will be direct transfer between cells through gap junction channels. The more efficient the fibroblast metabolism is, the more efficient the synthesis and deposition of a new connective tissue matrix will be and the less time will be required for granulation tissue maturation. The passage of Calcein dye through gap junctions within fibroblasts residing in granulation tissue is demonstrated by the parachute technique. Calcein dye, which is smaller than 1000 MW, is able to freely pass through patent gap junction channels between the paratrooper fibroblasts and the granulation tissue fibroblasts (Salomon et al., 1988; White et al., 1995). The cell membrane covalently linked DiI dye cannot pass through gap junction channels; thus, the paratrooper and granulation tissue fibroblasts can be readily identified (Goldberg et al., 1995). Through this technique we demonstrate in vivo that the granulation tissue of a PVA sponge implant contains fibroblasts that exhibit GJIC. Granulation tissue accumulates within PVA sponge implants. The degree of cellularity, the organization of the collagen fibers, and the presence of distinct blood vessels are all markers of granulation tissue maturation. Granulation tissue maturation involves a reduction in fibroblast density and an increase in connective tissue deposition between cells. As granulation tissue matures, capillaries coalesce into blood vessels (Knighton et al., 1981; Ehrlich, 1988; Ehrlich and Krummel, 1996; Berry et al., 1998). These morphological characteristics occur earlier in LiCl-treated PVA sponge implants. Collagen fibers are known to become thicker and more organized as granulation tissue matures (Li et al., 1980). Utilizing Sirius red staining, the birefringence of collagen fibers becomes more intense as they become thicker and more organized (Constantine and Mowry, 1968; Li et al., 1980). When compared to saline-treated controls, LiCltreated PVA implants showed an increase in collagen fiber birefringence intensity and organization; again, evidence that granulation tissue from LiCl-treated PVA implants matures
more rapidly. It is proposed that the enhancement of gap junctions and GJIC advances the maturation of granulation tissue in vivo. How does LiCl alter GJIC? One effect of LiCl is to mimic the wingless/int (wnt-1) signaling pathway, leading to the accumulation of the effector protein, -catenin (Hedgepeth et al., 1997; Ai et al., 2000). The accumulation of -catenin in the cell nucleus increases the expression of connexins, specifically Cx43, and GJIC in cardiac myocytes (Ai et al., 2000). The increase in the coupling index observed with scrape-loaded HF fibroblasts that received LiCl confirms those reported findings. GJIC influences the rate and final degree of rat osteoblastand human fibroblast-populated collagen lattice contraction (Bowman et al., 1998; Ehrlich et al., 2000). By polarized light microscopy, the increase in the rate of collagen lattice contraction is consistent with an improved organization and orientation of collagen fibers. Those studies support a role for GJIC in the organization of collagen fiber bundles in vitro similar to that for those deposited in granulation tissue in vivo. Here an increase in GJIC through the addition of LiCl leads to the earlier maturation of granulation tissue. It is proposed that increasing GJIC between fibroblasts within granulation tissue enhances the organization of the collagen fiber bundles and accelerates the onset of the remodeling phase of repair. Further studies examining the role of GJIC in the repair process may increase our understanding of how to better control the fibrotic process.
REFERENCES
Abdullah, K. M., Luthra, G., et al. (1999). Cell-to-cell communication and expression of gap junctional proteins in human diabetic and nondiabetic skin fibroblasts: Effects of basic fibroblast growth factor. Endocrine 10, 35–41. Ai, Z., Fischer, A., et al. (2000). Wnt-1 regulation of connexin43 in cardiac myocytes. J. Clin. Invest. 105, 161–171. Berry, D. P., Harding, K. G., et al. (1998). Human wound contraction: Collagen organization, fibroblasts, and myofibroblasts. Plast. Reconstr. Surg. 102, 124–134. Boitano, S., Dirksen, E. R., et al. (1992). Intercellular propagation of calcium waves mediated by inositol trisphosphate. Science 258, 292–295. Bowman, N. N., Donahue, H. J., et al. (1998). Gap junctional intercellular communication contributes to the contraction of rat osteoblast populated collagen lattices. J. Bone Miner. Res. 13, 1700–1706. Charles, A. C., Merrill, J. E., et al. (1991). Intercellular signaling in glial cells: Calcium waves and oscillations in response to mechanical stimulation and glutamate. Neuron 6, 983–992.
16 Constantine, V. S., and Mowry, R. W. (1968). Selective staining of human dermal collagen. II. The use of picrosirius red F3BA with polarization microscopy. J. Invest. Dermatol. 50, 419–423. Diegelmann, R. F., Cohen, I. K., et al. (1979). Growth kinetics and collagen synthesis of normal skin, normal scar and keloid fibroblasts in vitro. J. Cell. Physiol. 98, 341–346. Ehrlich, H. P. (1988). The role of connective tissue matrix in wound healing. Prog. Clin. Biol. Res. 266, 243–258. Ehrlich, H. P., Gabbiani, G., et al. (2000). Cell coupling modulates the contraction of fibroblast-populated collagen lattices. J. Cell. Physiol. 184, 86–92. Ehrlich, H. P., and Krummel, T. M. (1996). The regulation of wound healing from a connective tissue aspect. Wound Repair Regener. 4, 203–210. Gabbiani, G., Chaponnier, C., et al. (1978). Cytoplasmic filaments and gap junctions in epithelial cells and myofibroblasts during wound healing. J. Cell. Biol. 76, 561–568. Goldberg, G. S., Bechberger, J. F., et al. (1995). A pre-loading method of evaluating gap junctional communication by fluorescent dye transfer. Biotechniques 18, 490–497. Hedgepeth, C. M., Conrad, L. J., et al. (1997). Activation of the Wnt signaling pathway: A molecular mechanism for lithium action. Dev. Biol. 185, 82–91.
MOYER ET AL.
Knighton, D. R., Silver, I. A., et al. (1981). Regulation of woundhealing angiogenesis-effect of oxygen gradients and inspired oxygen concentration. Surgery 90, 262–270. Ko, K., Arora, P., et al. (2000). Biochemical and functional characterization of intercellular adhesion and gap junctions in fibroblasts. Am. J. Physiol. Cell. Physiol. 279, C147–C157. Kumar, N., and Gilula, N. (1996). The gap junction communication channel. Cell. 84, 381–388. Lee, S. W., Tomasetto, C., et al. (1992). Transcriptional downregulation of gap-junction proteins blocks junctional communication in human mammary tumor cell lines. J. Cell. Biol. 118, 1213–1221. Li, A. K., Ehrlich, H. P., et al. (1980). Differences in healing of skin wounds caused by burn and freeze injuries. Ann. Surg. 191, 244–248. Meda, P., Bruzzone, R., et al. (1987). Gap junctional coupling modulates secretion of exocrine pancreas. Proc. Natl. Acad. Sci. USA 84, 4901–4904. Musil, L. S., and Goodenough, D. A. (1991). Biochemical analysis of connexin43 intracellular transport, phosphorylation, and assembly into gap junctional plaques. J. Cell. Biol. 115, 1357–1374. Parker, S. B., Hertzberg, E. L., et al. (1994). Modulation of gap junctionmediated intercellular communication in embryonic chick mesenchyme during tissue remodeling in vitro. Cell. Tissue Res. 275, 215–224. Salomon, D., Saurat, J. H., et al. (1988). Cell-to-cell communication within intact human skin. J. Clin. Invest. 82, 248–254. White, T. W., Bruzzone, R., et al. (1995). The connexin family of intercellular channel forming proteins. Kidney Int. 48, 1148–1157.