Lung fibrosis involves the overexpression of ECM proteins, primarily collagen, by α-smooth muscle actin (ASMA)-positive cells. Caveolin-1 is a master regulator of collagen expression by cultured lung fibroblasts and of lung fibrosis in vivo. A peptide equivalent to the caveolin-1 scaffolding domain (CSD peptide) inhibits collagen and tenascin-C expression by normal lung fibroblasts (NLF) and fibroblasts from the fibrotic lungs of scleroderma patients (SLF). CSD peptide inhibits ASMA expression in SLF but not NLF. Similar inhibition of collagen, tenascin-C, and ASMA expression was also observed when caveolin-1 expression was upregulated using adenovirus. These observations suggest that the low caveolin-1 levels in SLF cause their overexpression of collagen, tenascin-C, and ASMA. In mechanistic studies, MEK, ERK, JNK, and Akt were hyperactivated in SLF, and CSD peptide inhibited their activation and altered their subcellular localization. These studies and experiments using kinase inhibitors suggest many differences between NLF and SLF in signaling cascades. To validate these data, we determined that the alterations in signaling molecule activation observed in SLF also occur in fibrotic lung tissue from scleroderma patients and in mice with bleomycin-induced lung fibrosis. Finally, we demonstrated that systemic administration of CSD peptide to bleomycin-treated mice blocks epithelial cell apoptosis, inflammatory cell infiltration, and changes in tissue morphology as well as signaling molecule activation and collagen, tenascin-C, and ASMA expression associated with lung fibrosis. CSD peptide may be a prototype for novel treatments for human lung fibrosis that act, in part, by inhibiting the expression of ASMA and ECM proteins.
- α-smooth muscle actin
interstitial lung disease (ILD) frequently complicates scleroderma (systemic sclerosis, SSc) and can be a debilitating disorder with a poor prognosis. SSc-ILD is now the leading cause of death in SSc (46), and there are few, if any, safe and effective treatment options (47). Lung tissue from SSc-ILD patients is fibrotic, containing abnormally high levels of collagen I (hereafter referred to as collagen) and other ECM proteins such as tenascin-C. These ECM proteins are synthesized primarily by an activated population of α-smooth muscle actin (ASMA)-positive fibroblasts known as myofibroblasts. Normal human lung fibroblasts (NLF) and fibroblasts from the lung tissue of scleroderma patients (SLF) are useful for studying lung fibrosis because they retain their in vivo phenotypes, i.e., SLF overexpress collagen, tenascin-C, and ASMA (5, 50), for at least 4 passages in vitro.
The progression of lung fibrosis in vivo can be studied in an animal model in which rodents receive intratracheal bleomycin. Although intratracheal bleomycin is not a perfect model for human disease, it is a very convenient and reliable model and is the best available model for scleroderma lung disease and for idiopathic pulmonary fibrosis (IPF) (16, 56). Bleomycin-induced fibrosis is similar to scleroderma and to IPF in that the disease can progress rapidly, the distribution of fibrosis within the tissue is patchy rather than diffuse, and proliferating fibroblasts overexpress collagen and tenascin-C and are ASMA-positive (62, 63).
Caveolin-1, the principal coat protein of caveolae, is a promising therapeutic target for treating lung fibrosis (52, 59). Caveolae were originally observed in electron microscopic images as flask-shaped invaginations in the plasma membrane. These cholesterol- and sphingolipid-rich organelles function in endocytosis, vesicular trafficking, and in the compartmentalization of specific signaling cascades (1). The caveolin family of caveolae coat proteins contains three members. Caveolin-1 and -2 are abundantly expressed in adipocytes, endothelial cells, and fibroblasts; caveolin-3 is muscle specific (39, 44). Caveolins serve as scaffolds for signaling molecules including members of the MAP kinase family, isoforms of PKC, Akt, G proteins, Src-family kinases, and growth factor receptors (12, 25, 37, 38, 40, 41). The interaction of kinases with caveolins frequently inhibits their activity (30). Conversely, caveolin-1 depletion hyperactivates signaling molecules both in vitro (17, 52) and in vivo (10).
There are extensive observations linking caveolin-1 to the regulation of collagen expression in vitro and the progression of lung fibrosis in vivo. High levels of caveolin-1 are found in NLF, whereas much lower levels are found in SLF and in fibroblasts from the fibrotic lung tissue of IPF patients (52, 59). This difference in caveolin-1 levels appears to be responsible for the overexpression of collagen in cells from fibrotic tissue because when caveolin-1 expression is knocked down in NLF using short interfering RNA (siRNA), collagen expression increases dramatically (52). Conversely, adenovirus-mediated caveolin-1 overexpression inhibits TGF-β-induced collagen and fibronectin expression (59). In vivo, two groups observed that, in caveolin-1 null mice, lung tissue shows significant pathology: the diameter of alveolar spaces is reduced, alveolar walls are thickened and hypercellular, and ECM deposition is significantly increased (14, 36). Similarly, caveolin-1 levels are strikingly decreased in lung tissue induced to become fibrotic by irradiation or by bleomycin treatment (22, 52) and in fibrotic lung tissue from human IPF patients (59). Finally, intratracheal administration of adenovirus mediating the overexpression of caveolin-1 (59) blocks the progression of bleomycin-induced lung fibrosis in mice. Thus there is a clear cut causal relationship between low caveolin-1 levels and lung fibrosis.
The ability of caveolin-1 to bind to a variety of kinases and thereby inhibit their activity has been mapped to a sequence known as the caveolin-1 scaffolding domain (CSD, amino acids 82–101 of caveolin-1; Ref. 31). A peptide equivalent to the CSD can cross the plasma membrane when synthesized as a fusion peptide on the COOH terminus of the antennapedia internalization sequence (7) or when myristoylated (20). Like the intact molecule, this CSD peptide binds to PKC and ERK and inhibits their activity (17, 20, 30). The CSD peptide is particularly useful because it is functional when delivered in vivo (7, 35).
Our goal in the current studies was to determine how the interplay between caveolin-1 and several other signaling molecules [MEK, ERK, JNK, phosphatidylinositol 3-kinase (PI3K), and Akt] regulates the expression of collagen, tenascin-C, and ASMA in NLF and SLF in vitro and the progression of lung fibrosis in vivo. We chose to use the CSD peptide rather than adenovirus as our primary method of upregulating caveolin-1 function because the results of such studies may lead to novel treatments that can be used in human patients and because the practical and regulatory hurdles to using virus to upregulate caveolin-1 in humans would appear to be significant. Major results of our studies are as follows. First, although caveolin-1 inhibits the activation of MEK, ERK, JNK, and Akt in both NLF and SLF and inhibits the expression of collagen, tenascin-C, and ASMA, details of the signaling cascades involving these molecules differ between NLF and SLF. Second, studies on the expression of signaling molecules in fibrotic lung tissue from human patients and bleomycin-treated mice validate the relevance of the bleomycin model to human disease. Finally, treatment with the CSD peptide blocks both changes in tissue morphology and changes in signaling molecule expression associated with lung fibrosis. In summary, our results strongly suggest that the CSD peptide inhibits collagen, tenascin-C, and ASMA expression in vitro and the progression of lung fibrosis in vivo through similar molecular mechanisms.
Fibroblasts were derived from lung tissue obtained at autopsy from scleroderma patients (SLF) and from age-, race-, and sex-matched normal subjects (NLF) and cultured as previously described (52). Cells were used in passages 2-4. The group of scleroderma patients fulfilled the criteria of American College of Rheumatology for the diagnosis of scleroderma with lung involvement. Normal human lung tissue was obtained from the Brain and Tissue Bank for Developmental Disorders (Baltimore, MD) or from the National Disease Research Interchange (Philadelphia, PA). Scleroderma lung tissue was obtained from the Division of Pathology and Laboratory Medicine at the Medical University of South Carolina (MUSC). The study was approved by the Institutional Review Board for Human Subject Research at the MUSC.
Perturbation of Caveolin-1 Expression
The CSD peptide (amino acids 82–101 of caveolin-1; DGIWKASFTTFTVTKYWFYR) and a scrambled control peptide (WGIDKAFFTTSTVTYKWFRY) were synthesized as fusion peptides to the COOH terminus of the antennapedia internalization sequence (RQIKIWFQNRRMKWKK). Before each experiment, desiccated peptides were dissolved at a 1 mM final concentration in 10% DMSO as described by Bernatchez et al. (4). Cells in six-well plates at 70–80% confluence were then incubated with 1 ml of serum-free DMEM containing 5 μM CSD or scrambled peptide (Scr). After 6 h, the culture medium and cell layer were harvested.
Recombinant adenovirus containing myc-tagged caveolin-1 and control, empty adenovirus (45) were used. Amplification, titering, and infection were performed as described previously (32). After 48 h, the medium was replaced with fresh serum-free medium. After an additional 6 h, cells and medium were harvested.
siRNA was used to knock down caveolin-1 expression as we have previously described (52).
U0126 (MEK inhibitor), SP-600125 (JNK inhibitor), LY-294002 (PI3K inhibitor), and Akt inhibitor VIII were purchased from Calbiochem (La Jolla, CA) and dissolved in DMSO. Cells in six-well plates were treated with 10 μM inhibitors in serum-free medium for 6 h (U0126) or 24 h for the other inhibitors before the harvesting of cells and medium.
Western Blot Analyses
Cell layers were harvested in boiling SDS-PAGE sample buffer; medium was dialyzed, lyophilized, and resuspended in sample buffer. Aliquots of culture medium or of cell layer representing material derived from the same number of cells were probed using the following primary antibodies and appropriate secondary antibodies. Culture medium: goat anti-human collagen I (AB758P) and rabbit anti-human tenascin-C (AB19011) from Millipore (Temecula, CA). Cell layer: rabbit antibodies against PKCα (sc-208), PKCε (sc-214), activated JNK (sc-12882), activated Akt Thr308 (sc-166646R), and caveolin-1 (sc-894) from Santa Cruz Biotechnology (Santa Cruz, CA); rabbit antibodies against ERK 1/2 (9102), activated ERK 1/2 (9106), JNK (9252), Akt (9272), MEK 1/2 (9122), and activated MEK 1/2 (9121) from Cell Signaling (Beverly, MA); mouse monoclonal anti-ASMA (clone 1A4) from Sigma (St. Louis, MO); and mouse monoclonal anti-actin (MAB1501) from Millipore.
Immunocytochemistry and Laser Confocal Microscopy
NLF and SLF were cultured in four-well glass chamber slides (Nalge Nunc International, Naperville, IL) and stained as previously described (51) using the indicated primary antibodies and appropriate secondary antibodies tagged with Alexa Fluor 488 (Molecular Probes, Eugene, OR). Nuclei were stained with 7-aminoactinomycin D (Molecular Probes). Images were acquired using a Zeiss LSM510 laser confocal microscope (excitation S490/20, emission D528/38) fitted with an oil-immersion objective (×40/1.4).
Bleomycin-Induced Lung Fibrosis and CSD Treatment
This procedure was approved by the MUSC Institutional Animal Care and Use Committee. Ten-week-old, male CD-1 mice (Charles River, Boston, MA) were anesthetized and received bleomycin (Calbiochem) or saline solution by intratracheal instillation as previously described (52). Two approaches were used for CSD peptide treatment. In both approaches, the peptide treatment was initiated 1 day before bleomycin or saline treatment. Mice were killed 14 days after bleomycin treatment. In the first approach, ALZET Osmotic Pumps delivering 0.5 μl/h of a 1 mM solution of CSD peptide were implanted into 12 mice; 12 additional mice received no pump. The next day, one-half of each group of mice received bleomycin treatment, and one-half received saline treatment. In the second approach, 12 mice received daily intraperitoneal injections of 100 μl of a 0.15 mM solution of the CSD peptide; 12 additional mice received the Scr. On the 2nd day, one-half of each group received intratracheal bleomycin, and one-half received saline. After 14 days, lungs were harvested. Half of each set of lungs was fixed, sectioned, and stained with Masson's trichrome stain as previously described (51). Briefly, lungs were removed, inflated, and fixed overnight in 2% paraformaldehyde in PBS, and the tissue was dehydrated through an alcohol series and embedded in paraffin. Sections were cut, deparaffinized, and stained. The other half of each set of lungs was dissolved in SDS-PAGE sample buffer for Western blotting experiments. Because very similar results were obtained using the two approaches to CSD peptide treatment, results were pooled.
Immunohistochemistry of Human Lung Tissue Sections
Lung tissue samples from scleroderma patients and healthy individuals (see above) were fixed without inflation overnight in 2% paraformaldehyde in PBS, and sections were cut, deparaffinized, blocked with 3% bovine serum albumin-1% goat serum-0.1% Triton X-100-PBS, and incubated overnight with blocking buffer containing appropriate primary antibodies and for 1 h with blocking buffer containing horseradish peroxidase-conjugated goat anti-rabbit IgG. Primary antibodies were rabbit anti-human tenascin-C (AB19011) from Millipore and rabbit polyclonal anti-caveolin-1 (sc-894) from Santa Cruz Biotechnology. Mouse monoclonal anti-ASMA (A-5691, Sigma) directly conjugated with alkaline phosphatase was developed using alkaline phosphatase substrate kit SK-5100.
Bleomycin-Induced Apoptosis and CSD Treatment
Mice were treated with bleomycin and received daily intraperitoneal injections of either the scrambled or the CSD peptide as described above. After 7 days, lungs were harvested, and apoptosis was detected by terminal deoxynucleotidyl transferase dUTP-mediated nick-end labeling (TUNEL) labeling of DNA strand breaks. Briefly, lungs were removed, inflated, and fixed with 2% paraformaldehyde in PBS. Tissue sections were permeabilized and labeled using ApopTag Plus Peroxidase In Situ Apoptosis Kit (S7101) as recommended by the manufacturer (Millipore). Counterstaining was done with 4,6-diamidino-2-phenylindole (DAPI; Molecular Probes). Using an Olympus IX71 fluorescence microscope equipped with a ×40 objective, DAPI images were acquired by fluorescence microscopy. TUNEL images were acquired using transmitted light and a monochromatic camera (causing the nuclei of apoptotic cells to appear black). DAPI and TUNEL images were superimposed.
Immunoreactive bands were quantified by densitometry using the NIH ImageJ (ImageJ 1.32j; National Institutes of Health) software. For statistical analysis, the raw densitometric data were processed and analyzed using the Prism 3.0 (GraphPad Software) statistical analysis software.
Student's t-test was used to analyze protein expression levels in lung tissue samples obtained from normal individuals and from scleroderma patients. Protein expression levels for samples obtained from in vitro experiments and from mice were analyzed by two-way ANOVA followed by Bonferroni post-test. For experiments in which the degree of tissue damage in mice treated with bleomycin was scored on an arbitrary scale, data was analyzed using the Mann-Whitney rank sum test. In all tests, results were regarded as statistically significant if P < 0.05.
We (52) previously demonstrated that caveolin-1 depletion upregulates collagen expression in lung fibroblasts via a mechanism involving the activation of MEK and ERK. In the current study, we build on these observations by: 1) demonstrating that collagen expression is inhibited when cells are treated with a membrane-permeable peptide from the scaffolding domain of caveolin-1 (CSD peptide) that is known to mimic the kinase-inhibiting activity of full-length caveolin-1 (17, 20, 30); 2) demonstrating that caveolin-1 also regulates the expression of the myofibroblast differentiation marker ASMA and of the ECM protein tenascin-C; 3) extending the signaling cascade that regulates the expression of collagen, tenascin-C, and ASMA to include JNK, Akt, and PI3K; 4) validating the relevance of this signaling cascade to the progression of lung fibrosis in vivo; and 5) demonstrating that CSD peptide treatment blocks the progression of bleomycin-induced lung fibrosis in vivo.
CSD Peptide Inhibits Collagen Expression by Lung Fibroblasts
Throughout these studies, collagen expression in cultured lung fibroblasts was quantified in terms of its accumulation in the culture medium. When collagen expression by NLF or SLF treated with CSD peptide or with the control Scr was compared, the CSD peptide was found to inhibit collagen expression by >95% (Fig. 1). This effect was rapid, occurring in less than 5 h, and powerful, occurring at 5 μM CSD. In further agreement with our previous studies demonstrating that the ability of caveolin-1 to regulate collagen expression is downstream from its regulation of MEK/ERK activation (52), the CSD peptide decreased MEK/ERK activation by ∼50%. In contrast, the CSD peptide did not affect the expression of caveolin-1 itself, PKCα, or PKCε.
To provide additional confirmation for the link between caveolin-1, ERK activation, and collagen expression in lung fibroblasts, NLF and SLF were infected with adenovirus encoding caveolin-1 or with control virus lacking a cDNA insert. Figure 2, top, demonstrates that infecting cells with adenovirus encoding caveolin-1 results in partial inhibition of ERK activation and almost complete inhibition of collagen expression. As with CSD peptide treatment, this treatment did not affect the expression of PKCα or PKCε. Figure 2, bottom, demonstrates the dose-dependence of treatment with adenovirus encoding caveolin-1 on collagen expression. Thus each of two very different treatments that increase caveolin-1 expression in cells results in the partial inhibition of ERK activation and in the almost complete inhibition of collagen expression.
Caveolin-1 Regulates the Expression of ASMA and Tenascin-C
Because myofibroblasts overexpress collagen (52), we evaluated the possibility that caveolin-1 regulates myofibroblast differentiation in addition to regulating collagen expression. Therefore, we determined the effects of up- and downregulating caveolin-1 expression on the expression of the myofibroblast differentiation marker ASMA in NLF (i.e., fibroblasts) and SLF (a myofibroblast-rich population expressing ASMA at high levels). We observed striking differences between NLF and SLF. siRNA-mediated inhibition of caveolin-1 expression increased ASMA expression in NLF but not in SLF (Fig. 3, A and B). Conversely, overexpression of caveolin-1 mediated by the CSD peptide (Fig. 3, C and D) or by adenovirus (Fig. 2) inhibited ASMA expression in SLF but not in NLF. In summary, decreasing caveolin-1 expression increases ASMA expression in cells expressing relatively low levels of ASMA but not in cells already expressing high levels of ASMA; increasing caveolin-1 expression decreases ASMA expression in cells expressing high levels of ASMA but not in cells expressing relatively low levels of ASMA.
In addition to overexpressing collagen, myofibroblasts also overexpress tenascin-C (50). Therefore, to determine whether caveolin-1 regulates both the differentiation and function of myofibroblasts, we evaluated the effect on tenascin-C expression of perturbing caveolin-1 expression. siRNA-mediated inhibition of caveolin-1 expression increased tenascin-C expression in NLF but not in SLF, just as it increased ASMA expression (Fig. 3) and collagen expression in NLF but not SLF (52). Caveolin-1 overexpression mediated by the CSD peptide (Fig. 3) or by adenovirus (Fig. 2) inhibited tenascin-C expression in both SLF and NLF, just as it inhibited collagen expression (Fig. 1). In contrast, the CSD peptide inhibited ASMA expression in SLF but not NLF (Fig. 3, C and D). The combined observations support the idea that the decreased level of caveolin-1 present in SLF regulates both their differentiation into myofibroblasts and their overexpression of ECM proteins.
Tenascin-C Expression is MEK/ERK Dependent, and ASMA Expression is MEK/ERK Independent
We previously demonstrated that collagen expression is regulated by MEK/ERK signaling in both NLF and SLF (52). To determine whether the expression of tenascin-C and ASMA are also regulated by MEK/ERK, cells were treated with the MEK inhibitor U0126. Whereas U0126 inhibited tenascin-C expression in both NLF and SLF, ASMA expression was unaffected by U0126 in both cell types (Fig. 4). Thus the signaling cascade through which caveolin-1 regulates collagen and tenascin-C expression appears to include MEK/ERK, whereas the cascade through which caveolin-1 regulates ASMA expression must include different intermediates.
Signaling Cascades Regulating the Expression of Collagen, Tenascin-C, and ASMA
To identify the distinct signaling cascades that regulate the expression of collagen, tenascin-C, and ASMA in normal and fibrotic lung tissue, we evaluated the levels of activated MEK, ERK, Akt, and JNK in NLF and SLF treated with the CSD peptide or the Scr peptide. We (52) previously observed that the activation of MEK and ERK is upregulated in SLF compared with NLF. We now report that the activation of Akt and JNK, in accord with Shi-Wen et al. (43), is also upregulated in SLF (Fig. 5). Consistent with the ability of caveolin-1 to inhibit the activation of a variety of kinases, in both NLF and SLF, treatment with the CSD peptide inhibited the activation of Akt and JNK (Fig. 5) in addition to MEK and ERK (Fig. 1). In contrast, the CSD peptide did not affect the expression of total MEK, total ERK, total Akt, or total JNK.
To further elucidate cross talk between MEK/ERK, Akt, PI3K, and JNK and the ability of these kinases to regulate the expression of collagen, tenascin-C, and ASMA, NLF and SLF were treated with the MEK inhibitor U0126, the JNK inhibitor SP-600125, Akt inhibitor VIII, and the PI3K inhibitor LY-294002. Whereas U0126 blocked ERK activation and collagen and tenascin-C expression as described above (Fig. 4), it had little or no effect on the activation of JNK and Akt (Fig. 6A) or on the expression of ASMA (Fig. 4).
SP-600125 inhibited the activation of JNK, ERK, and Akt in both NLF and SLF (Fig. 6B). The inhibition of JNK activation by SP-600125 raises the possibility that the form of JNK present in these cells is JNK2, which is known to undergo autophosphorylation (13). SP-600125 also had a consistent inhibitory effect on collagen and tenascin-C expression in both NLF and SLF; however, it inhibited ASMA expression in SLF yet had no effect on ASMA expression in NLF (Fig. 6B).
Akt inhibitor VIII inhibited the activation of Akt and enhanced the activation of ERK in both NLF and SLF (Fig. 6C). This inhibition of Akt activation is consistent with its known autophosphorylation (41). Whereas Akt inhibitor VIII had no effect on JNK activation in NLF, it enhanced JNK activation in SLF. Akt inhibitor VIII inhibited collagen and tenascin-C expression and had no effect on ASMA expression in NLF (Fig. 6C). Conversely, Akt inhibitor VIII inhibited ASMA expression and had no effect on collagen and tenascin-C expression in SLF (Fig. 6C).
LY-294002 inhibited Akt activation, enhanced ERK activation, and had no effect on JNK activation in both NLF and SLF (Fig. 6C). These observations are consistent with the fact that PI3K is frequently placed immediately upstream from Akt in signaling cascades. Like Akt inhibitor VIII, LY-294002 inhibited collagen and tenascin-C expression in NLF, had no effect on ASMA expression in NLF, and inhibited ASMA expression in SLF (Fig. 6C). However, LY-294002 inhibited collagen and tenascin-C expression in SLF, even though Akt inhibitor VIII had no effect on these cells (Fig. 6C). These observations suggest that in SLF, PI3K may regulate collagen and tenascin-C expression via an Akt-independent mechanism. A model summarizing all our data on the regulation of collagen, ASMA, and tenascin-C expression by signaling cascades involving caveolin-1, MEK/ERK, JNK, PI3K, and Akt is presented in discussion.
Immunohistochemical Detection of Signaling Molecules
To confirm and extend the results of the Western blotting experiments described above, we examined the effects of the CSD peptide on the levels of expression and distribution of caveolin-1, activated ERK, activated JNK, and activated Akt. The low level of caveolin-1 in SLF compared with NLF (see Fig. 1 and Ref. 52) was confirmed by immunohistochemistry (IHC; Fig. 7). In both cell types treated with the control Scr peptide, caveolin-1 was detected primarily in a punctate pattern in the cytoplasm. In accord with Fig. 1, the CSD peptide did not affect the level of caveolin-1 expression detected by IHC (Fig. 7). However, the distribution of caveolin-1 staining was altered in both cell types, with staining associated with the plasma membrane becoming more prominent, particularly in SLF.
The high level of activated ERK present in SLF compared with NLF (Fig. 1) was also confirmed by IHC (Fig. 7). Whereas activated ERK in NLF treated with the control peptide is primarily localized in the nucleus and the perinuclear region, activated ERK in SLF treated with the control peptide is present in the perinuclear region and in an intensely labeled punctate pattern in the cytoplasm. As observed in Fig. 1, the CSD peptide decreases the level of activated ERK in both NLF and SLF (Fig. 7). This residual staining is primarily perinuclear and cytoplasmic in both NLF and SLF.
The high level of activated JNK present in SLF compared with NLF (Fig. 5) was also confirmed by IHC (Fig. 8). Although activated JNK in NLF treated with the control peptide is primarily perinuclear, intense punctate cytoplasmic staining is observed in addition to perinuclear staining in SLF. As observed in Fig. 5, the CSD peptide decreases the level of activated JNK in both NLF and SLF (Fig. 8). In both cell types, this treatment preferentially decreased perinuclear staining and increased staining associated with the plasma membrane. In SLF, CSD peptide treatment also eliminated the intense punctate cytoplasmic staining observed in cells treated with the control peptide.
Finally, the high level of activated JNK present in SLF compared with NLF (Fig. 5) was also confirmed by IHC (Fig. 8). Both cell types exhibited nuclear and perinuclear staining; punctate cytoplasmic staining was also apparent in both cell types but was particularly prominent in SLF. As observed in Fig. 5, the CSD peptide decreases the level of activated Akt in both NLF and SLF (Fig. 8). In both cell types, this treatment preferentially decreased perinuclear, nuclear, and cytoplasmic staining while increasing staining associated with the plasma membrane.
In summary, the IHC studies in Figs. 7 and 8 confirmed the results of the Western blotting experiments in Figs. 1 and 5 regarding the relative levels of caveolin-1, activated ERK, activated JNK, and activated Akt in NLF and SLF, as well as the effects of the CSD peptide on these levels. In addition, Figs. 7 and 8 revealed differences in the distribution patterns of these signaling molecules in NLF and SLF and cell type-specific differences in the effect of the CSD peptide on these distribution patterns. These observations raise the possibility that the CSD peptide alters the trafficking of these signaling molecules between organelles and that the trafficking of these molecules is different in NLF and SLF.
Signaling Molecule Expression in Normal and Fibrotic Human Lung Tissue In Vivo
We (52) previously observed that caveolin-1, PKCε, phosphorylated-MEK (ph-MEK), and ph-ERK were involved in the regulation of collagen expression in NLF and that their expression was altered in SLF, leading to the overexpression of collagen by these cells. In the current study, we have demonstrated that ph-JNK and ph-Akt are also involved in the regulation of collagen expression in NLF and that their expression is altered in SLF. To evaluate whether similar alterations could be detected in the expression of these signaling molecules in vivo, the levels of these proteins in autopsy specimens of normal lung tissue and of fibrotic lung tissue from scleroderma patients were examined. Extremely consistent results were obtained in a Western blotting experiment using three samples from each source (Fig. 9, A and B). In particular, in accord with results obtained with NLF and SLF, caveolin-1 and PKCε levels were very low in lung tissue from scleroderma patients whereas ph-MEK, ph-ERK, ph-JNK, and ph-Akt were present at very high levels. The levels of actin (loading control) and of total ERK, JNK, and Akt were similar in normal and scleroderma lung tissue (Fig. 9, A and B). Whereas total MEK was present at elevated levels in scleroderma lung tissue (2.5-fold increase), this elevation was still much less than for ph-MEK (6.5-fold increase). The observed decrease in caveolin-1 expression in scleroderma lung tissue does not appear to be due simply to a decrease in the number of caveolin-1-rich endothelial cells, since we (3) showed previously that there are more endothelial cells present in scleroderma lung tissue than in normal tissue at this stage of the disease. (A comparison of Fig. 10 with Ref. 3 demonstrates that the current samples are from the stage designated as SSc I.) In general, the changes in these signaling molecules associated with fibrosis in vivo were more extreme than the differences that we observed between NLF and SLF (52), suggesting that whereas NLF and SLF retain the signaling properties of fibroblasts and myofibroblasts in vivo, these differences are attenuated during culture.
In the case of PKCα, much lower levels were observed in scleroderma lung tissue than in normal lung tissue (Fig. 9, A and B), even though no difference had been observed between NLF and SLF (52). This observation would be consistent with our observation that PKCα regulates caveolin-1 expression (52). It is also possible that cell types other than fibroblasts present in vivo contribute to the observed decrease in PKCα expression in fibrotic lung tissue from scleroderma patients.
In addition, we also compared the levels of caveolin-1, of the ECM proteins collagen and tenascin-C, and of the myofibroblast marker ASMA in normal and scleroderma lung tissue by histochemistry and IHC (Fig. 10). In accord with the Western blotting data, there was much less caveolin-1 in scleroderma lung tissue than in normal lung tissue. As expected, collagen (detected using Masson's trichrome stain) was present at much higher levels in scleroderma lung tissue than in normal lung tissue (Fig. 10). Although tenascin-C and ASMA were present at higher levels in scleroderma lung tissue than in normal lung tissue by Western blot (Fig. 9, A and B), the differences were not statistically significant. Nevertheless, IHC data support the idea that tenascin-C and ASMA are present at higher levels in scleroderma lung tissue than in normal lung tissue. Indeed, examination of ASMA staining (Fig. 10) demonstrates that there is a striking increase in the number of ASMA-positive cells present in fibrotic scleroderma lung tissue. It may be that a loss of vascular smooth muscle and bronchiolar smooth muscle associated with fibrosis limits the level of the increase in ASMA that can be detected by Western blotting.
Validation of the Bleomycin Model
When bleomycin is administered to mice intratracheally, within 2 wk, 30–50% of the mice die. Those surviving show massive fibrosis with distortion of the alveoli and with filling of the tissue and the alveolar air space with various types of cells and with collagen and other ECM proteins. To verify the validity of bleomycin treatment of mice as a model system for human lung fibrosis, we performed Western blotting for a variety of signaling molecules in control (saline-treated) and fibrotic (bleomycin-treated) mouse lung tissue. The results obtained with bleomycin-induced lung fibrosis (Fig. 11) were strikingly similar to the results obtained with fibrotic human (SSc) lung tissue (Fig. 9). In particular, in both cases, PKCα, PKCε, and caveolin-1 were downregulated in fibrotic tissue whereas ph-ERK, ph-JNK, and ph-Akt were all upregulated. In addition, the ECM protein tenascin-C and the myofibroblast marker ASMA were upregulated to a greater extent during bleomycin-induced lung fibrosis (Fig. 11) than in fibrotic human lung tissue (Fig. 9). All of these changes in protein expression between saline- and bleomycin-treated mouse lung tissues were statistically significant (P < 0.01). These results strongly suggest that the same molecular mechanisms that regulate lung fibrosis in human patients also regulate lung fibrosis in the bleomycin model.
CSD Provides Protection against Bleomycin-Induced Lung Fibrosis
Having verified the importance of caveolin-1 and other associated signaling molecules in the progression of lung fibrosis in vivo, we then tested the possibility that administration of the CSD peptide might provide protection against bleomycin-induced fibrosis. The results of these experiments were strikingly positive. Whereas bleomycin-treated mice that did not receive peptide routinely showed severe tissue damage and collagen deposition or died, bleomycin-treated mice that received peptide routinely showed only slight to moderate tissue damage and collagen deposition and improved survival (Fig. 12). The improvement in tissue morphology caused by peptide treatment was statistically significant (P < 0.02) as determined using the Mann-Whitney rank sum test; the improvement in survival, however, did not achieve statistical significance, perhaps due to the relatively small number of animals studied. Interestingly, when deaths were plotted on a Kaplan-Meier curve (Fig. 12), it was observed that there was a similar level of mortality at 6 days after bleomycin treatment in both groups (i.e., the time when inflammation has peaked, but fibrosis is only beginning). However, mice that did not receive the CSD peptide continued to die during the period of fibrosis, whereas no mice that received the CSD peptide died during the period of fibrosis. Finally, the CSD peptide inhibited weight loss in mice killed 14 days after bleomycin treatment. On average, mice that did not receive the peptide lost 30% of their original body weight whereas mice that received the CSD peptide lost only 14%.
Changes in the expression of signaling molecules, ASMA, and tenascin-C usually associated with bleomycin treatment were also inhibited by the peptide (Fig. 11). In particular, the increases in ph-ERK, ph-Akt, ph-JNK, ASMA, and tenascin-C and the decrease in PKCε were all inhibited by at least 75%, and the decrease in PKCα was inhibited by ∼50%. Each of these changes was statistically significant (P < 0.01). The CSD peptide did not affect the expression of caveolin-1 itself, strongly suggesting that the CSD peptide alters the expression of proteins that are downstream from caveolin-1 in signaling cascades. These observations on the lack of effect of CSD peptide treatment on caveolin-1 expression and on the effect of CSD peptide treatment on ASMA and tenascin-C expression were confirmed by IHC (Fig. 13). In summary, treatment with the CSD peptide provides a remarkable degree of protection against bleomycin-induced lung fibrosis.
CSD Provides Protection against Bleomycin-Induced Apoptosis and Inflammation
To determine whether the beneficial effects of the CSD peptide on the progression of lung fibrosis might involve additional mechanisms, we evaluated the effect of CSD peptide treatment on apoptosis and on the accumulation of inflammatory cells in lung tissue. These experiments were performed in mice 7 days after bleomycin treatment when part of the tissue shows normal morphology and part is infiltrated with inflammatory cells. TUNEL labeling showed that the CSD peptide has a remarkable ability to inhibit bleomycin-induced epithelial cell apoptosis (Fig. 14). In a typical field from a region of normal morphology from a bleomycin-treated mouse receiving control peptide, ∼60% of the cells are apoptotic (Fig. 14C). These apoptotic cells must be primarily epithelial cells because of their location lining the alveolar surface and because epithelial cells are the predominant cell type in alveoli. Treatment with the CSD peptide almost completely eliminates apoptosis in these cells (Fig. 14D); almost no cells are apoptotic in sections from mice that did not receive bleomycin (Fig. 14, A and B). When apoptosis was quantified by counting the number of TUNEL-positive cells per field (average of 5 fields with normal morphology from each of 6 mice), the data obtained were: control mice, control peptide 0.8 ± 0.2 labeled cells per field; control mice, CSD peptide 1.0 ± 0.3 labeled cells per field; bleomycin-treated mice, control peptide 38.6 ± 5.7 labeled cells per field; bleomycin-treated mice, CSD peptide 5.2 ± 1.2 labeled cells per field. Little apoptosis of infiltrating inflammatory cells was observed in bleomycin-treated mice whether or not they received the CSD peptide (data not shown).
To evaluate the effect of the CSD peptide on inflammatory cell infiltration, the portion of hematoxylin and eosin (H&E)-stained tissue containing densely packed cells with prominently stained nuclei and little cytoplasm was quantified. CSD peptide treatment strongly inhibited inflammatory cell infiltration in bleomycin-treated mice (Fig. 14). Whereas in a typical section from a bleomycin-treated mouse receiving control peptide, ∼50% of the tissue was infiltrated with inflammatory cells (Fig. 14E), only ∼15% of the tissue was infiltrated in bleomycin-treated mice receiving the CSD peptide (Fig. 14F). Little, if any, dense inflammatory cell staining was ever observed in mice that did not receive bleomycin (data not shown). When this parameter was quantified in six mice, we found that, on average, 52.8% ± 3.1% of the tissue was infiltrated with inflammatory cells in bleomycin-treated mice receiving control peptide whereas only 14.8% ± 1.4% of the tissue was infiltrated with inflammatory cells in bleomycin-treated mice receiving the CSD peptide. In summary, it appears that the CSD peptide provides protection against the progression of bleomycin-induced lung fibrosis by inhibiting epithelial cell apoptosis and inflammatory cell infiltration as well as by directly inhibiting the overproduction of collagen.
The results of our in vitro experiments support the model shown in Fig. 15 depicting the signaling cascades that regulate the expression of collagen, tenascin-C, and ASMA in NLF and SLF. In our previous work (52), we demonstrated that the activation of MEK/ERK increases collagen expression and that knocking down caveolin-1 expression using siRNA increases MEK/ERK activation, again leading to increased collagen expression. Conversely, we now observe that upregulating caveolin-1 function using either the CSD peptide (Fig. 1) or adenovirus (Fig. 2) inhibits MEK/ERK activation and collagen expression in both NLF and SLF. Although upregulating caveolin-1 function also inhibits tenascin-C expression in NLF and SLF, upregulating caveolin-1 function inhibits ASMA expression only in SLF (Figs. 2 and 3). Given that the MEK inhibitor U0126 inhibits ERK activation and the expression of collagen and tenascin-C but has no effect on ASMA expression (Fig. 4), these results demonstrate that caveolin-1 regulates ASMA expression in SLF via a mechanism not involving MEK/ERK.
In the current report, we have extended our studies to include JNK, PI3K, and Akt. Like activated MEK and ERK (Fig. 1), activated JNK and Akt are present at higher levels in SLF than in NLF, and their activation is inhibited by the CSD peptide (Fig. 5). Our results are in accord with published observations that JNK and Akt activation promote a fibrotic phenotype in cultured lung fibroblasts (42, 43, 59) and that both PI3K and Akt are activated in fibrotic lung tissue from mice in which TGF-β is overexpressed only in the lung (21). To the best of our knowledge, this is the first demonstration that Akt is activated in fibroblasts derived from any human fibrotic lung disease and the first demonstration that JNK is activated in scleroderma lung fibroblasts.
In addition to serving as an inhibitor of signaling proteins, CSD peptide altered the subcellular localization of caveolin-1 and of activated ERK, JNK, and Akt (Figs. 7 and 8). These observations support the idea that the CSD peptide affects caveolin-1-mediated vesicular trafficking. Although the mechanism through which the CSD peptide affects trafficking is currently unknown, it is reasonable to believe that it acts by inhibiting the interaction of caveolin-1 with other proteins.
The use of specific inhibitors was extremely helpful in developing the hierarchy of signaling molecule functions shown in Fig. 15. Although the MEK inhibitor U0126 did not affect ASMA expression in either NLF or SLF, the JNK inhibitor SP-600125, Akt inhibitor VIII, and the PI3K inhibitor LY-294002 all inhibited ASMA expression in SLF but not NLF (Fig. 6). Because the JNK inhibitor and the PI3K inhibitor each block the activation of Akt (Fig. 6), we show Akt as being downstream from JNK and PI3K (Fig. 15). In the simplest scenario combining these observations (Fig. 15), ASMA must be downstream from Akt. In addition, the JNK inhibitor inhibited the activation of ERK and Akt in both NLF and SLF. Therefore, we have placed JNK upstream of ERK and Akt.
Our data suggest that PI3K and Akt differ in their ability to regulate collagen and tenascin-C expression in NLF and SLF. PI3K and Akt are circled together in Fig. 15 because Akt is frequently, but not always, immediately downstream from PI3K in signaling cascades. Given that the Akt inhibitor inhibited collagen and tenascin-C expression in NLF but not SLF and that the PI3K inhibitor inhibited collagen and tenascin-C expression in both NLF and SLF, our working hypothesis is that PI3K regulates collagen and tenascin-C expression via an Akt-independent mechanism in SLF (and possibly NLF), whereas PI3K also regulates collagen and tenascin-C expression via an Akt-dependent mechanism in NLF but not SLF. Intermediates involved in this Akt-independent mechanism are currently unidentified; one possibility would be a member of the Tec family of kinases known to be involved in PI3K-dependent, Akt-independent signaling (8). NLF and SLF also differed in Akt signaling in that the Akt inhibitor promoted JNK activation only in SLF even though it promoted ERK activation in both cell types. Therefore, we have placed Akt as a negative upstream regulator of ERK in both NLF and SLF and of JNK only in SLF. The fact that the Akt inhibitor promotes JNK activation only in SLF raises the possibility of an alternative interpretation of our data in which PI3K signals primarily through Akt in both NLF and SLF. In this scenario, our failure to observe inhibition of collagen and tenascin-C expression by the Akt inhibitor in SLF may occur because the reversal of the negative regulation of JNK activation by Akt that occurs only in SLF may compensate for the reversal of the positive regulation of collagen and tenascin-C expression by Akt.
Because of the complexity of the signaling loops in Fig. 15 and the fact that a particular signaling molecule can be found both upstream and downstream from another signaling molecule, it would be risky to use Fig. 15 to predict that altering the activation of Akt or PI3K would be a beneficial treatment for scleroderma patients. For example, although it is possible that upregulating Akt activity would inhibit ERK and JNK and thereby inhibit collagen expression and fibrosis, it is also possible that upregulating Akt activity would promote ASMA expression and thereby promote the fibrotic phenotype. One source of this complexity may be the fact that there are a large number of isoforms of both PI3K and Akt (Table 1). The available inhibitors block the kinase activity of all forms. If one studied each form of PI3K and Akt independently, then it might become possible to identify a therapeutic target of value in treating scleroderma. Although it is likely that such work would be of great importance, it is far beyond the scope of the current study.
In vivo experiments demonstrate that the same signaling cascades that regulate the expression of collagen, tenascin-C, and ASMA in NLF and SLF also regulate their expression during the progression of a murine model of lung fibrosis. Collagen, tenascin-C, and ASMA expression are upregulated in SLF and in the fibrotic lung tissue of SSc-ILD patients and bleomycin-treated mice (Figs. 9–12). Similarly, caveolin-1 and PKCε are downregulated, and activated MEK, ERK, JNK, and Akt are upregulated in SLF, the fibrotic lung tissue of SSc-ILD patients, and bleomycin-treated mice (Figs. 1, 9, and 11). In addition, PKCα is downregulated in the fibrotic lung tissue of SSc-ILD patients and bleomycin-treated mice (Figs. 9 and 11), although its expression is similar in NLF and SLF (Fig. 1). These observations validate the relevance of both SLF as an in vitro model and bleomycin-treated mice as an in vivo model for the fibrosis observed in SSc-ILD.
Because we observed that the CSD peptide inhibits collagen, tenascin-C, and ASMA expression in vitro, we reasoned that this peptide should also inhibit the progression of lung fibrosis in vivo. Indeed, systemic treatment with the peptide had a striking positive effect on the survival and lung tissue morphology of bleomycin-treated mice (Figs. 12 and 13). These results are totally consistent with those of Wang et al. (59), who used adenovirus rather than the CSD peptide to upregulate caveolin-1 expression, thereby inhibiting the progression of bleomycin-induced lung fibrosis in mice. In addition to its effects on survival and tissue morphology, the CSD peptide blocked the changes in collagen, tenascin-C, ASMA, PKCε, and PKCα expression and the changes in the activation of ERK, JNK, and Akt that are normally associated with bleomycin-induced lung fibrosis (Figs. 11 and 13). Interestingly, CSD peptide treatment did not block the bleomycin-induced decrease in caveolin-1 expression (Fig. 11), strongly suggesting that all the changes that were blocked are downstream from caveolin-1 in signaling cascades. In summary, our results strongly suggest that the low levels of caveolin-1 present in SLF and in the fibrotic lung tissue of SSc-ILD patients and bleomycin-treated mice lead to the activation of MEK, ERK, JNK, and Akt, which, in turn, lead to the overexpression of collagen, tenascin-C, and ASMA.
To determine whether the CSD peptide might provide protection against bleomycin-induced lung fibrosis via additional mechanisms, we also examined its effects on apoptosis and inflammation (Fig. 14). Indeed, the CSD peptide strikingly inhibited epithelial cell apoptosis and inflammatory cell infiltration. Whether the CSD peptide blocks apoptosis by inhibiting the killing of epithelial cells by bleomycin or by blocking the secretion of toxic agents by inflammatory cells or myofibroblasts (24) is an open question of great interest to us. The fact that the CSD peptide inhibits inflammatory cell infiltration is consistent with the known anti-inflammatory effects of caveolin-1 (58).
Myofibroblasts are contractile, ASMA-positive fibroblasts that secrete high levels of ECM proteins and thus are key participants in the tissue remodeling that occurs during wound healing and in various fibrotic disorders (55, 57). Originally, myofibroblasts were viewed as resident fibroblasts that became activated and proliferated due to their interaction with effector molecules present in fibrotic lung tissue such as thrombin and TGF-β. In accord with this idea, thrombin and TGF-β induce the expression of collagen, tenascin-C, and ASMA by NLF (5, 6, 18). More recently, it has been proposed that myofibroblasts are generated by epithelial-mesenchymal transformation (23, 61) and by the differentiation of bone marrow-derived stem cells into circulating, collagen-positive fibrocytes that traffic into injured lung tissue (29, 33, 34). Again, TGF-β and other profibrotic cytokines promote epithelial-mesenchymal transformation (23, 61), the differentiation of fibrocytes into myofibroblasts, and their expression of ECM proteins (2, 19, 34). Currently, it remains an open and controversial question as to whether all three potential sources contribute to the population of myofibroblasts present in fibrotic human lung tissue or whether one source is predominant. In any case, we propose that low levels of caveolin-1 expression will be a general feature of myofibroblasts whether it turns out that myofibroblasts in human disease are derived from one source or from multiple sources. Therefore, treatments that increase the expression/function of caveolin-1 (e.g., CSD peptide) should be beneficial in either scenario.
There have been relatively few studies on the signaling mechanisms controlling the myofibroblast phenotype (overexpression of ASMA and ECM proteins) in fibroblasts from normal and fibrotic lung tissue. Our results are consistent with those in the most closely related previous study (59) in which it was found that using adenovirus to upregulate caveolin-1 expression in cultured fibroblasts inhibited TGF-β-induced ECM protein expression by inhibiting the activation of JNK and ERK and that adenovirus-mediated upregulation of caveolin-1 expression in vivo ameliorated bleomycin-induced lung fibrosis (59). Our results are also consistent with the findings that JNK and PI3K/Akt signaling are involved in the regulation of ASMA expression (19, 42, 43). Other signaling molecules that are involved in the regulation of the myofibroblast phenotype and thus are likely to fit within the signaling pathways that we have proposed (Fig. 15) include focal adhesion kinase, PTEN, and endothelin (28, 42, 43, 48, 60).
In summary, our study is a particularly comprehensive analysis of the signaling mechanisms underlying the regulation of the myofibroblast phenotype because we have examined fibroblasts from both normal lung tissue and fibrotic lung tissue from SSc-ILD patients; we have examined the expression of ASMA, collagen, and tenascin-C; we have examined signaling both in vitro and in vivo; and we have examined several signaling molecules (caveolin-1, MEK, ERK, JNK, PI3K, Akt, PKCε, and PKCα). Of particular importance are the several pathways involving PI3K and Akt that we have observed to function differently in NLF and SLF (Fig. 15).
The etiology of scleroderma is believed to start with the influence of the environment on genetically susceptible individuals leading to autoimmunity. There are no doubt a variety of environmental influences and biochemical alterations (e.g., faulty DNA repair mechanisms; Ref. 27) that can promote autoimmunity. In turn, autoimmunity leads to tissue damage, inflammation, and fibrosis. These intertwined processes result in the generation of a plethora of mediators (cytokines, growth factors, bioactive lipids, ECM molecules, and proteolytic enzymes) that activate all of the cell types present in the tissue (epithelial cells, endothelial cells, inflammatory cells, and fibroblasts) through a variety of molecular mechanisms. We propose that several of these mechanisms converge on a final common pathway regulating fibrosis in which the expression of caveolin-1 is significantly decreased in myofibroblasts and other cell types. This idea is strongly supported by the observations that caveolin-1 expression is inhibited in fibroblasts treated with the profibrotic cytokine TGF-β and in the tissue of patients with IPF (a fibrotic lung disease not involving autoimmunity; Ref. 59).
The identification of direct targets for the CSD peptide (i.e., the kinases to which it binds directly) is a difficult and interesting question. Although the list of signaling proteins that coimmunoprecipitate with the CSD peptide and/or for which activity is inhibited by the CSD peptide is long (26), many of these proteins may be indirect targets that are complexed with direct targets. To determine whether a protein is a direct target for the CSD peptide, it is necessary to study the interaction between the purified protein and the CSD peptide. The list of proteins shown in this manner to be directly affected by the CSD peptide includes eNOS, PKC, MEK, ERK, Ras, G proteins, and the EGF receptor (4, 11, 12, 15, 22, 36). The sequences ΦXΦXXXXΦ and ΦXXXXΦXXΦ (and the combined sequence ΦXΦXXXXΦXXΦ), where Φ stands for any of the aromatic amino acids (F, W, or Y) and X stands for any amino acid, were identified as consensus ligands for the CSD peptide in a phage display experiment (11). At least one of these caveolin-1-binding domain (CBD) sequences is present in almost every protein known to interact with caveolin-1 (11, 23). This definition of the CBD domain may be overly stringent. G proteins and G protein-coupled receptor kinases bind to the CSD peptide via sequences in which a hydrophobic amino acid (I, V, or L) is substituted for an aromatic amino acid at certain positions in consensus CBD sequences (9, 11). These more inclusive consensus CBD sequences are ΦXZXXXXΦ and ZXXXXΦXXZ where Z stands for F, W, Y, I, V, or L.
To address the question of which of the signaling proteins studied in this paper may be direct targets for the CSD peptide, we compiled a table of proteins involved in MAP kinase and PI3K signaling that contain “classic” and inclusive consensus CBD domains (Table 1). The MAP kinase family has four levels, MAP kinases, MAP kinase kinases, MAP kinase kinase kinases, and MAP kinase kinase kinase kinases. To avoid confusion between the name of the family and the level of the kinase, we will refer to these as M1K, M2K, M3K, and M4K. Although none of the M1Ks (including ERK and JNK) contain classic CBD sequences, several M2Ks (including the ERK kinases MEK1 and MEK2 and the JNK kinase MP2K7), several M3Ks, and several M4Ks contain classic CBD sequences (Table 1). These data suggest that M2Ks, M3Ks, and M4Ks may be direct ligands for caveolin-1, whereas M1Ks such as ERK and JNK may be indirect ligands that interact with caveolin-1 via the higher kinases. However, if inclusive consensus CBD sequences are considered, then every M1K contains multiple CBD sequences (Table 1), and every member of the MAP kinase family may be a direct target for the CSD peptide.
Whether we examined the classic or the inclusive set of consensus CBD sequences similarly affected our analysis of which proteins involved in PI3K signaling might be direct targets for the CSD peptide. In this paper, we observed both Akt-dependent and Akt-independent PI3K signaling. In particular, the regulation of collagen and tenascin-C expression by PI3K appeared to be Akt-dependent in NLF but Akt-independent in SLF. Although several members of the PI3K family and several members of the Tec family of kinases (which are involved in Akt-independent PI3K signaling; Ref. 8) contain classic CBD sequences, Akt does not contain these sequences (Table 1). However, all three Akt polypeptides contain several of the inclusive consensus CBD sequences (Table 1). Thus, if only the classic sequences are considered, then PI3K and Tec kinase family members may be direct targets for the CSD peptide whereas Akt may be an indirect target. However, if inclusive consensus CBD sequences are considered, then PI3 kinases, Tec kinases, and Akt may all be direct targets for the CSD peptide.
The current studies have demonstrated that the CSD peptide can be used to upregulate the function of caveolin-1 both in vitro and in vivo. These in vivo experiments provide proof of principle that the CSD peptide or a related agent can be used to treat lung fibrosis in human patients. Although experiments using adenovirus are an elegant way to demonstrate the ability of caveolin-1 to ameliorate the progression of lung fibrosis in an animal model (59), it would seem that a pharmacological agent such as the CSD peptide that mimics the function of caveolin-1 is more likely than adenovirus infection to gain approval for use in human patients. Of course, there is still considerable room for improvement in the CSD peptide. The version used in the current study contains 16 amino acids from the antennapedia internalization sequence to mediate its entry into cells and 20 amino acids from the caveolin-1 scaffolding domain. In future studies, it will be exciting to determine the activity of CSD peptide variants constructed using alternative means of promoting internalization and using known functional subdomains of the caveolin-1 scaffolding domain (4). Such studies will be critical in identifying the optimal version of the CSD peptide to be used as a treatment for lung fibrosis in human patients.
This work was supported by a grant from the Scleroderma Foundation (to E. Tourkina), National Heart, Lung, and Blood Institute Grant R01-HL-73718 (to S. Hoffman), National Center for Research Resources Facilities Construction Grant C06-RR-015455 and National Institute of Arthritis and Skin Diseases Grant K01-AR-054143 (to E. Tourkina).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society