| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, Florida, USA
1Correspondence: Department of Cell Biology and Anatomy, RMSB 2030 A, University of Miami School of Medicine, 1600 NW 10th Ave., Miami, FL 33136, USA. E-mail: ggrotend{at}miami.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-SMA or elevated levels of collagen synthesis. Cells expressing -SMA do not exhibit DNA synthesis but do coexpress higher levels of types I and III collagen mRNA. Thus, fibroblast proliferation and differentiation are controlled by combinatorial signaling pathways involving not only components of the TGF-ß/CTGF pathway, but also signaling events induced by EGF and IGF-2-activated receptors. Collectively, our studies indicate TGF-ß functions as a classic embryonic inducer, initiating a cascade that is controlled by other factors in the cellular environment. We propose a model for this process with regard to wound repair and fibrotic lesion formation that is likely applicable to other instances of CTGF action during embryogenesis.—Grotendorst, G. R., Rahmanie, H., Duncan, M. R. Combinatorial signaling pathways determine fibroblast proliferation and myofibroblast differentiation.
Key Words: myofibroblasts • TGF-ß • CTGF • fibrotic tissue
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. They are characterized by expression of the smooth muscle isoform of -actin (-SMA) and, when activated, synthesize high levels of extracellular matrix proteins, particularly collagen (2
3
4
5)
. During normal wound repair, myofibroblasts are transiently present in the wound bed, where they are believed to play essential roles in wound contraction and connective tissue restoration. In contrast, these cells persist in fibrotic lesions and are believed to be responsible for the excessive collagen production that leads to alteration of tissue architecture and ultimately to organ failure. Collectively, fibrotic disorders are the leading cause of morbidity and mortality in North America, Europe, and Japan (6)
. Although myofibroblasts represent a key target cell for anti-fibrotic therapeutics, little is known about the regulatory mechanisms controlling the proliferation, differentiation, and apoptosis of these cells.
TGF-ß has been implicated as an initiating cytokine in numerous fibrotic disorders of the skin and internal organs and tissues (7
8
9
10
11)
. A diverse range of biological activities can be induced by TGF-ß in a wide range of target cells, making it an ideal candidate as an inducer of tissue formation. Initially identified by its unique ability to stimulate anchorage-independent growth of normal fibroblastic cells (12
, 13)
, sarcoma growth factor (SGF) was separated into its two component parts, TGF- and TGF-ß (14)
, demonstrating that the presence of an EGF-related peptide is essential for TGF-ß mitogenic activity. TGF-ß has since been shown to function as a potent mitogen on fibroblasts in monolayer culture (15
16
17)
, to inhibit the growth of epithelial cells (18)
and vascular endothelial cells (19
, 20)
, and to limit the activation and proliferation of leukocytic cells (21
, 22)
. Other studies have revealed that TGF-ß can stimulate extracellular matrix protein synthesis and deposition in vivo (23
24
25)
and induce a differentiated phenotype in certain mesodermal cell types (26
27
28)
. The molecular mechanism whereby TGF-ß induces these diverse and paradoxical (proliferation vs. differentiation) biological effects must occur postreceptor, as the same TGF-ß receptors appear to be responsible for initiating the various signaling events (29
, 30)
. We have determined that the growth stimulatory action of TGF-ß (mitogenic activity) and the induction of collagen synthesis and accumulation (matrigenic activity) occur via a post-TGF-ß receptor mechanism mediated by connective tissue growth factor (CTGF) (31
32
33)
.
CTGF is an original member of the CCN family of growth factors (34)
that have been reported to play regulatory roles as growth stimulators and growth inhibitors in a wide range of biological processes (35
36
37
38
39
40
41
42
43)
. Originally identified as a chemoattractant and mitogen for smooth muscle cells (44)
, CTGF is selectively induced by TGF-ß (35
, 45
, 46)
. TGF-ß and CTGF are coordinately expressed at sites of tissue repair and fibrosis (32
, 47)
. We have found that TGF-ß-stimulated fibroblast proliferation (31
, 48)
and collagen synthesis (33)
can be blocked with anti-CTGF antibodies or by inhibition of CTGF synthesis, demonstrating that CTGF acts as a secondary cytokine for TGF-ß biological effects on fibroblastic cells and indicating that CTGF plays a key role in TGF-ß-stimulated connective tissue formation in human fibrotic disease.
Several studies have indicated that TGF-ß is a potent inducer of the myofibroblast phenotype (49
50
51
52)
. We report here that, as with TGF-ß-induced proliferation or collagen synthesis, TGF-ß induction of the myofibroblast phenotype is mediated by CTGF. Expression of the myofibroblast phenotype correlates with induction of elevated collagen synthesis but is prevented by induction of cell proliferation. The results of our studies demonstrate that the presence of other growth factors (EGF and IGF) determines whether TGF-ß/CTGF induces DNA synthesis and cell proliferation or myofibroblast differentiation and concurrent -SMA expression and elevated collagen synthesis. Based on these studies, we present a model for regulation of fibroblast growth and myofibroblast differentiation where the determination of cell behavior is the result of combinatorial signaling by multiple growth factor families.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. RhTGF-ß1, acidic and basic FGF, and PDGF-AA and AB were obtained from Gibco-BRL (Gaithersburg, MD, USA); rhEGF and PDGF-BB were from Upstate Biotechnology (Lake Placid, NY, USA) and rhIGF-2 from Collaborative Biomedical (Bedford, MA, USA). Affinity-purified goat anti-rhCTGF was raised against rhCTGF and purified as described (33)
.
Cell cultures
NRK fibroblasts (clone NRK-49F, a continuous line of cultured normal rat kidney fibroblasts) (53)
were originally obtained from Dr. R. Assoian, University of Pennsylvania. NRK fibroblasts were maintained in Dulbecco’s modified Eagle media (DME) containing 2.5% fetal bovine serum (FBS) and 2.5% Nu-Serum I (NS) (Collaborative Biomedical) and passaged before confluence. We and others have shown NRK cells and other fibroblast cell types to be sensitive target cells for studying TGF-ß and CTGF regulation of fibroblast proliferation and collagen synthesis (23
, 25
, 32
, 33
, 45
, 54)
.
Mitogenic assay
NRK fibroblasts were seeded at 10,000 cells/well in 48-well plates and grown to a confluent monolayer over a 6 day period in DME + 2.5% FBS/NS. Media was then changed to DME containing 25 mM HEPES and insulin, transferrin, and selenium [ITS premix] (Collaborative Biomedical) and cells cultured for 8 or 9 days to permit NRK monolayers to become quiescent. The effect of CTGF and other growth factors on NRK fibroblast mitogenesis was then assessed by their direct addition, along with 50 µg/mL of ascorbic acid, to the starved culture media. DNA synthesis during the last 24 h of a 48 h treatment period was assessed by measuring incorporation of 3H-thymidine (2 µCi/mL) into TCA-precipitated DNA, as we have described previously (32
, 45)
. Data from each experimental condition were derived from duplicate or triplicate wells and averages are reported as cpm/well. SEs were generally <5%.
Fibroblast collagen synthesis assay
Quiescent fibroblast monolayers identical to those in the mitogenic assay were used to evaluate fibroblast collagen synthesis during the last 24 h of a 48 h treatment period by measuring the incorporation of 3H-proline (2 µCi/mL) into pepsin-resistant salt-precipitated extracellular collagen (33
, 55
, 56)
. Each experimental condition was done in duplicate or triplicate wells; averaged results are expressed as cpm/well, or where treatments resulted in cellular proliferation as cpm/103 cells, after counting trypsinized cell monolayers. SEs were generally <5%.
Immunohistological assay of myofibroblast formation
The myofibroblasts present in NRK fibroblast monolayers after 48 h treatments with growth factors were detected immunohistologically using a standard avidin-biotin amplification method (57)
. After fixation in methanol at –20°C and blocking with 2% milk/10% horse serum, red-stained, SMA-positive myofibroblasts were visualized by sequentially incubating fixed monolayers with 1) anti--SMA monoclonal IgG (clone 1A4: Sigma Chemical, St. Louis, MO, USA), 2) biotinylated horse anti-mouse IgG secondary antibody (Vector Labs, Burlingame, CA, USA), 3) alkaline phosphatase conjugated streptavidin-biotin complex (Dako, Carpenteria, CA, USA), and 4) Vector Red alkaline phosphatase visualization substrate (Vector Labs). The percentage of -SMA-positive cells was determined by counting five high-power fields per well and calculated as the number of -SMA-positive cells per total cells. If fibroblast monolayers were to be further processed for collagen mRNA in situ hybridization studies, SMA was visualized with a peroxidase conjugated streptavidin-biotin complex (Dako) and Nova Red peroxidase substrate (Vector Labs). During subsequent in situ hybridization studies, the Nova-Red stain was extracted by formamide-containing prehybridization solutions, permitting unobstructed detection of colorimetric hybridization signals.
Immunohistological assay of BrdU incorporation
Proliferating cells in NRK fibroblast monolayers were tagged with BrdU by incubation with 10 µM BrdU during the last 24 h of 48 h treatments with CTGF or other growth factors. After fixation with 70% ethanol at –20°C, BrdU-labeled nuclei were detected with a mouse monoclonal anti-BrdU antibody (clone BU-33: Sigma Chemical) (58)
using the standard avidin-biotin amplification method and a NBT/BCIP alkaline phosphatase substrate (55
, 57)
. Mitotic index was calculated as the percentage of BrdU-positive nuclei vs. total in five high-power fields per well.
Dual detection of BrdU incorporation and myofibroblast formation
Dual immunohistological detection on the same fibroblast monolayer was done sequentially with anti-BrdU and anti-SMA as described above. After 70% ethanol fixation, BrdU-labeled nuclei were detected with NBT/BCIP substrate; after a second fixation with methanol, -SMA-positive myofibroblasts were visualized with Vector Red substrate.
In situ hybridization assay for collagen mRNA synthesis
In situ hybridization for collagen mRNAs was performed on fibroblast monolayers using described methods modified to accommodate the use of digoxigenin-labeled riboprobes (59)
. Briefly, fibroblast monolayers that either had or had not been processed for anti--SMA immunohistology were fixed or refixed with 4% paraformaldehyde, treated with proteinase K, acetylated, and dehydrated. After prehybridization in a 50% formamide/Denhardt’s-based solution (Sigma), hybridizations were performed at 50°C for 16 h with 500 ng/mL of digoxigenin-labeled riboprobes for type I or III collagen. After stringency washes and digestion of unhybridized RNA with RNase A, hybridized digoxigenin-labeled riboprobe was detected with an anti-digoxigenin antibody coupled to alkaline phosphatase using NBT/BCIP visualization (Roche Applied Science, Indianapolis, IN, USA). Sense and antisense digoxigenin-labeled RNA probes for type I and III collagen mRNAs were prepared by recloning EcoRI-cleaved open reading frame fragments of 1.5 and 0.6 kb, respectively, from plasmids Hf677 (COL1A1) and Hf934 (COL3A1) (ATCC, Manassas, VA, USA) into vector pBluescript II KS+ (Stratagene, La Jolla, CA, USA). After linearization with appropriate restriction enzymes, single-strand anti-sense and sense digoxigenin-labeled riboprobes were transcribed from pBluescript SP6/T7 promoters using a DIG-RNA labeling kit (Roche Applied Science). Both antisense probes gave positive hybridization signals, as shown in Fig. 3
. The sense probes served as negative controls and resulted in no hybridization signals (data not shown).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 33
, 48)
. To determine whether TGF-ß induced the myofibroblast phenotype via a CTGF-dependent pathway, we treated NRK cells with TGF-ß or CTGF and measured myofibroblast formation, determining the expression of -SMA in individual cells by immunostaining. Counting the number of -SMA-positive cells revealed that both factors induced >60% of the total cells per high-power field to express -SMA and to adapt a myofibroblast morphology, with large numbers of actin stress fibers (Fig. 1
A–C). Complete inhibition of TGF-ß-induced myofibroblast formation by anti-CTGF antibody provides proof that this activity is via a CTGF-dependent pathway (Fig. 1D
). Nonimmune goat IgG had no inhibitory effect on the TGF-ß induction of myofibroblasts (not shown).
|
Relationship of myofibroblast phenotype to cell proliferation and collagen synthesis
To determine whether induction of the myofibroblast phenotype occurred in both nonproliferating and proliferating cells, we investigated the relationship between cell proliferation and myofibroblast phenotype, as both responses can be induced by TGF-ß and CTGF. NRK fibroblasts were grown to confluence in complete medium and then shifted to serum-free media containing DMEM and ITS for 8 days. Cells were treated with various growth factors separately or in combination. Individual cell proliferation or differentiation was monitored by double immunostaining for incorporated BrdU or -SMA, respectively. The results of these studies are summarized in Fig. 2
A. In control cultures, little DNA synthesis occurred (mitotic index of 4%) and no myofibroblasts were detected (Fig. 2a
). Treatment of the cells with either IGF-2 or EGF alone (Fig. 2b, c
, respectively) or in combination (Fig. 2d
) did not induce any significant increase in DNA synthesis or myofibroblast formation. Cultures treated with TGF-ß exhibited an increase in myofibroblast and a small increase in the number of cells positive for DNA synthesis (Fig. 2e
). However, large increases in myofibroblasts were found when cells were activated by TGF-ß in the presence of IGF-2 (Fig. 2f
). In marked contrast to these cultures, when cells were activated by TGF-ß in the presence of EGF, no myofibroblasts were detected and a majority of the cells were positive for DNA synthesis (Fig. 2g
). A maximal proliferative response was detected in cultures activated with TGF-ß in the presence of EGF and IGF-2 (Fig. 2h
). Under no circumstances did we detect cells positive for both BrdU incorporation and -SMA. These data demonstrate TGF-ß-mediated myofibroblast differentiation and proliferation of the NRK fibroblasts are mutually exclusive responses to TGF-ß. Differentiation into myofibroblasts appears to be the default pathway whereas proliferation becomes the dominant and sole response when both TGF-ß and EGF are present.
|
We next evaluated the effect of cell proliferation on collagen synthesis in NRK cells. Cultures parallel to those used in the above experiment were used to determine collagen synthesis as described in Materials and Methods. Confluent growth-arrested cultures were treated with the indicated growth factors individually or in combination. Cell numbers were determined by direct cell counting to normalize for cell proliferation and the data are presented as collagen synthesis per cell. Only when cells were activated with either TGF-ß alone or in the presence of IGF-2 was there a detectable and significant increase in collagen synthesis per cell (2.5- and 4.2-fold, respectively) (Fig. 2B
). In the presence of EGF, TGF-ß did not stimulate any increase in collagen synthesis per cell with or without IGF-2 present (Fig. 2B
). These findings parallel those in the myofibroblast formation studies and link the expression of the myofibroblast phenotype with elevation of collagen synthesis. We also evaluated whether other growth factors (PDGF AA, AB, BB, acidic and basic FGF and serum) could interfere with TGF-ß-induced myofibroblast differentiation and collagen synthesis. Results of these experiments are similar to the studies with EGF and indicate that any growth factor that stimulated cells to proliferate suppressed expression of -SMA and elevation of collagen synthesis in response to TGF-ß (data not shown). These data support the hypothesis that cell division and myofibroblast differentiation are mutually exclusive cellular responses to TGF-ß.
To visualize which cells had elevated collagen production and the myofibroblast phenotype, we compared the expression of -I type I and -I type III collagen mRNA transcripts with immunostaining for -SMA. In situ hybridization of mRNA for collagen had to be used, as there are no pro-collagen antibodies available for immunohistological detection of nonprimate pro-collagens types I or III. During the processing for hybridization, the stain to detect -SMA is removed so that the in situ signal can be clearly visualized. The data are presented as identical regions of the culture, first stained for -SMA using immunohistochemical methods, then for collagen transcripts using in situ hybridization with digoxigenin-labeled RNA probes. Cells were cultured using standard conditions as described above, and control untreated cultures (IGF-2, 10 ng/mL) were compared with those activated with TGF-ß (5 ng/mL) in the presence of IGF-2 (10 ng/mL) after 48 h. As seen in Fig. 3
, control cultures (undifferentiated cells) expressed very low levels of I type I (Fig. 3A
) or I type III (Fig. 3D
) collagen transcripts. Activation of the cultures with TGF-ß (5 ng/mL) for 48 h under conditions that induce differentiation significantly increased levels of both -I type I collagen mRNA (Fig. 3B
) and -I type III collagen mRNA (Fig. 3E
). It appeared that whereas all cells exhibited some increase in collagen transcripts, a subset of cells presented very high levels of collagen mRNAs (red arrows). Comparison of these photomicrographs to those of the identical region of the culture previously stained for -SMA (Fig. 3C, F
, black arrows) demonstrated a concordance of expression of high levels of collagen transcripts with high levels of -SMA. These results support the concept that a characteristic of the myofibroblast phenotype is elevated production of collagen. The elevation of transcripts for types I and III collagen is consistent with a smooth muscle phenotype instead of a fibroblast phenotype characterized by the predominant expression of type I collagen.
CTGF mediates EGF-stimulated proliferation and IGF-2-induced differentiation
We next compared the effectiveness of CTGF to synergize with EGF to stimulate cell proliferation and IGF-2 to stimulate myofibroblast differentiation and collagen synthesis in the absence of TGF-ß. We used conditions identical to those used in previous experiments, but now performed a dose response curve of EGF or IGF-2 in the presence or absence of CTGF (5 ng/mL). As seen in Fig. 4
A, EGF alone at concentrations of up to 2 ng/mL was not effective as a mitogen for the NRK cells in monolayer culture under these conditions (open circles). However, in the presence of CTGF (5ng/mL), a potent DNA synthesis response was induced at EGF concentrations as low as 0.1 ng/mL (filled circles). The induction of myofibroblast formation and stimulation of collagen synthesis were followed in identical cultures except that IGF-2 was used in place of EGF (Fig. 4B
). As with EGF, in these experiments IGF-2 alone did not induce myofibroblast formation (data not shown) or elevation of collagen synthesis at concentrations of up to 50 ng/mL (open circles). However, in the presence of CTGF (5 ng/mL), IGF-2 at 2 ng/mL induced myofibroblast formation (filled squares) and collagen synthesis (filled circles) with identical dose response curves. These data demonstrate that at the concentrations tested, none of these growth factors exhibit activity alone. However, in combination, CTGF and EGF or IGF-2 potently stimulate proliferation or differentiation and collagen synthesis, respectively, in the target cells.
|
Kinetics of myofibroblast differentiation and collagen synthesis
To determine the temporal relationship of myofibroblast phenotype expression relative to the elevation of collagen synthesis, we measured the expression of -SMA and collagen synthesis at different times after activation with TGF-ß. Standard culture conditions were used with growth-arrested and serum-starved NRK fibroblasts. Cells were activated with TGF-ß (5 ng/mL) and IGF-2 (10 ng/mL); collagen synthesis was measured by incorporation of 3H-proline into collagen and myofibroblast formation determined by immunostaining of -SMA. Myofibroblast formation took longer than 8 h to occur, but by 16 h 60–75% of the cells in the culture were -SMA positive. The percentage of -SMA-positive cells remained high but gradually decreased over the next 56 h, with 65% positive 48 h after addition of TGF-ß and 38% positive 72 h after addition of TGF-ß. The percentage of -SMA-positive cells continued to decrease such that by 6 days after addition of TGF-ß, only 5% of the cells exhibited a myofibroblast phenotype. This decrease was not due to a decrease in TGF-ß levels in the culture media, as addition of supplemental TGF-ß (5 ng/mL) on day 3 resulted in no significant difference in the percentage of -SMA-positive cells present in the culture on day 6 compared with the culture that received only a single addition of TGF-ß on day 0 (Fig. 5
A).
|
In contrast to the kinetics of myofibroblast formation, collagen synthesis was much slower. Slightly elevated levels of collagen synthesis are seen at 8, 16, and 24 h (Fig. 5B
). Dramatic increases in collagen synthesis are detected 48 and 72 h after addition of TGF-ß, reaching levels in excess of sevenfold over non-TGF-ß-treated control cultures (Fig. 5B
). Furthermore, unlike the more rapidly reversible expression of the -SMA-positive phenotype, elevated collagen synthesis persisted at high levels (4- to 5-fold) 9 days after addition of TGF-ß. It was not until 16 days after addition of TGF-ß that the level of collagen synthesis returned to basal levels (data not shown). These studies indicate the myofibroblast phenotype is induced before up-regulation of collagen synthesis and does not persist for the same duration as collagen synthesis in response to TGF-ß.
Transient exposure to TGF-ß but not CTGF is sufficient to stimulate cell proliferation, differentiation, and increased collagen synthesis
We earlier demonstrated that brief exposure of cells to TGF-ß is sufficient to induce long-term expression of CTGF gene expression and protein production (60)
. This occurs even when cellular protein synthesis was completely inhibited during exposure to TGF-ß, demonstrating the CTGF gene is directly controlled by TGF-ß. We wanted to determine whether transient exposure of cells to TGF-ß or CTGF would be sufficient to stimulate cell proliferation. Cells were exposed to either TGF-ß or CTGF in the presence of EGF. TGF-ß or CTGF was removed and the medium was replaced with EGF only; 3H-thymidine was added during the final 24 h of incubation, with the cells harvested 48 h after the initial addition of growth factor. The data presented in Fig. 6
A demonstrate that a 30 min exposure of the cells to TGF-ß was sufficient to induce a 50% maximal DNA synthesis response compared with cells that were exposed to the growth factor continuously for the 48 h period. This percentage increased to 75% for the culture exposed to TGF-ß for 6 h. In contrast, even a 6 h exposure to CTGF stimulated a response of only 8% of maximal, indicating CTGF must be present for substantially longer periods. We tested the media removed from the treated cultures on parallel naive cultures for mitogenic activity. These studies revealed that all media samples stimulated a 100% mitogenic response in the target cells compared with positive controls, indicating that that no significant depletion of growth factor from the media had occurred up to the 6 h incubation time and that the CTGF added to the media had full biological activity provided cells were exposed for a sufficient period. Similar observations were made with both myofibroblast induction and up-regulation of collagen synthesis (Fig. 6B
). These studies indicate that TGF-ß functions similar to an embryonic inducer and that its effects persist long after its removal from the culture medium. One of the mediators of these effects is CTGF, as neutralization of CTGF with anti-CTGF antibodies completely blocks all of these TGF-ß-induced biological responses (31
, 33
, Fig. 1
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-SMA and do not exhibit elevated levels of collagen synthesis. These findings strongly indicate there are two distinct biological response pathways that are mutually exclusive for cells that have been activated by TGF-ß. For cells to proliferate in response to TGF-ß/CTGF, EGF or another suitable mitogenic factor must be present. For cells to differentiate into myofibroblasts and to increase collagen synthesis, IGF-2 is required. We propose that these responses represent a developmental program initiated by TGF-ß and maintained by and controlled by CTGF in concert with other co-growth factors during tissue repair and wound healing. We think similar principles may be applicable to the many instances in embryogenesis, where CTGF is expressed to control normal tissue formation and development (61
, 62)
, and in the pathological fibrotic disorders.
While both proliferation and differentiation are mediated via a CTGF pathway, the factor controlling whether TGF-ß stimulates either cell proliferation or myofibroblast formation is the presence of other growth factors such as EGF. Both of these responses occur over the same period (within 24 h of TGF-ß addition to the media) and are mutually exclusive, as we never detected any cell that was both BrdU and SMA positive, even under conditions where both myofibroblast formation and cell proliferation were occurring in the same culture. Furthermore, the results of these studies argue strongly that differentiation is the default pathway unless sufficient mitogenic growth factors are present to support cell proliferation in response to TGF-ß. The suppression of differentiation is not limited to EGF, as other factors such as PDGF can function in a similar fashion. This supports our hypothesis of TGF-ß functioning as an inducer of differentiation, which can be suppressed by an environment supplying appropriate co-mitogens for the target cells.
The molecular basis for enhanced EGF signaling in TGF-ß-activated cells appears to be via an increased number of high-affinity EGF receptors (19
, 63)
. The mechanism for increased receptor affinity is not fully understood, but recent studies have provided a firm linkage between cell adhesion molecules and growth factor receptor signaling. Our laboratory and others have reported that TGF-ß (32
, 64)
and CTGF (32
, 65)
can up-regulate 5ß1 integrin, which enhances fibronectin-mediated adhesion and is an essential component for anchorage-independent growth stimulated by TGF-ß (64)
. More important, recent studies have indicated that increased expression of 5ß1 integrin selectively enhances EGF mitogenic signaling (66)
, which provides a likely mechanism for CTGF enhancement of EGF activity.
Regarding CTGF collaboration with IGF-2 in signaling differentiation and up-regulation of collagen synthesis, a model involving direct interaction of CTGF and IGF-2 is possible. The first motif in the CTGF amino-terminal domain has a strong sequence homology to the IGF binding domain of IGF binding proteins (34)
. Biochemical studies indicate that at high concentrations, CTGF binds to IGF in solution (67)
. Although these concentrations are orders of magnitude in excess of those that occur physiologically in the extracellular space (or those used in our assays), this raises the intriguing possibility that a CTGF-IGF complex is the active cytokine that may form only in association with the receptor and would be much weaker in the absence of the third partner of this trimer. Such an active complex could signal via one of the known IGF receptors or through some yet unidentified receptor. Even though our studies were conducted in the presence of high concentrations of insulin, which can fully activate the type I IGF-R, it was essential to add physiological concentrations of IGF-2 for biological activity of the purified recombinant CTGF. This would argue for involvement of the type II IGF-R, also known as the cation-independent mannose 6 phosphate receptor, involved in intracellular trafficking (68)
as opposed to the type I IGF-R, which can be activated by insulin at the concentrations used in our assays (69)
. Previous studies have reported that TGF-ß can induce IGF synthesis (70
, 71)
, which we believe explains the lack of an absolute requirement for IGF in TGF-ß-activated cells compared with CTGF-treated cells. In this manner, TGF-ß induces the expression of multiple growth factors that are required for the completion of the developmental program.
Tissue formation during embryogenesis and regeneration after injury demands highly coordinated cell migration and proliferation, followed by cellular differentiation and matrix synthesis and assembly. The regulatory mechanisms that control these events appear to be linked in a cascade fashion, so that early factors function as initiators of these complex biological processes (72
73
74)
. Dysfunctional cascades lead to tissue malformations in the embryo and are likely the underlying basis for wound healing disorders and many fibrotic disorders. The results of our studies support a model where the production of active TGF-ß at sites of injury or inflammation functions in a fashion analogous to an inducer during embryonic development. That is, given a permissive environment, a complex series of biological responses proceeds in a programmed fashion no longer dependent on the presence or action of the inducer. Our finding that a brief exposure (30 min) of the NRK cells to TGF-ß is sufficient to induce a 50% maximal response of either proliferation or myofibroblast formation and collagen synthesis compared with a continuous exposure supports this hypothesis. In contrast, even a 6 h exposure of the same cells to CTGF induced a <10% maximal response compared with continuous 48 h exposure to CTGF. This pattern of biological response correlates with the kinetics and pattern of expression of the CTGF gene, directly regulated by TGF-ß, and remains expressed at high levels for at least 30 h even after a 1 h exposure of the cells to TGF-ß (60)
. Thus, TGF-ß exhibits all of the characteristics of a true embryonic inducer activating a long-term response without a requirement for its continuous presence through a secondary mediator CTGF.
Our findings indicate that CTGF serves to control a pivotal switch point in the cascade for connective tissue formation. CTGF alone is not responsible for determining whether cells respond to TGF-ß by proliferation or differentiation. Rather, CTGF acts in conjunction with other growth factors such as EGF and IGF-2 to control the events. Our model for the roles of TGF-ß, CTGF, EGF, and IGF is diagramed in Fig. 7
. In this model, TGF-ß serves to initiate the cascade by activation of a responsive cell (potentially a stem cell) to produce CTGF and to prime the cell to become receptive to CTGF. Initially, target cells are stimulated to amplify their number by proliferation if sufficient co-mitogens (i.e., EGF or PDGF in our studies) are present in the environment. When co-mitogen levels fall below concentrations that support cell proliferation, these amplified cells are then stimulated to differentiate by signals induced by CTGF as long as sufficient levels of IGFs are present. In the case of NRK cells or other fibroblasts, these cells express the myofibroblast phenotype and synthesize increased amounts of collagen. Collectively, these studies demonstrate that a complex combinatorial signaling mechanism involving multiple growth factors from different growth factor families functions together to regulate the commitment of TGF-ß-activated fibroblasts to either proliferate or differentiate and produce collagen. We suggest that other CCN family members may also function downstream of a priming growth factor or with cytokine partners, including other members of the TGF-ß superfamily, during embryogenesis and in tissue regeneration and repair processes. Future studies directed at elucidating the actions of the CCN gene family should provide interesting and important insight into the molecular control mechanisms of a wide array of developmental systems and pathological disorders involving connective tissues.
|
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 4, 2003. Accepted for publication November 21, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Haydont, B. L. Riser, J. Aigueperse, and M.-C. Vozenin-Brotons Specific signals involved in the long-term maintenance of radiation-induced fibrogenic differentiation: a role for CCN2 and low concentration of TGF-{beta}1 Am J Physiol Cell Physiol, June 1, 2008; 294(6): C1332 - C1341. [Abstract] [Full Text] [PDF]
|
|
|
0.02 ng/mL. B) Collagen synthesis and myofibroblast formation. Confluent growth-arrested and serum-starved cultures of NRK cells were stimulated with increasing amounts of IGF-2 in the absence (open circles) or presence (filled circles) of CTGF (5 ng/mL). Collagen synthesis was determined using incorporation of 3H-proline. Vertical bars represent SE. In the absence of CTGF, IGF-2 did not stimulate an increase in collagen synthesis. In the presence of CTGF, IGF-2 stimulated a 5- to 6-fold increase in collagen synthesis. As in the mitogenic assay, CTGF was not active alone or at IGF-2 concentrations of 
