INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Diseases associated with pathological fibroproliferation represent a major cause of morbidity and mortality (1). As a group, they are characterized by an accumulation of fibroblasts and deposition of extracellular matrix components, resulting in organ dysfunction (2). Fibroproliferation occurs in blood vessels leading to cardiac disease, cerebral vascular disease, and peripheral vascular disease as well as in all major tissues and organs including liver, kidney, and lung. Despite representing a major cause of death in industrialized countries, no effective therapy targeting the effector fibroblasts exists for this class of disorders. Two new therapeutic directions have been reported. One promising direction is to inhibit fibroblast deposition of durable matrix components, such as collagens type I and III (3). A second area involves therapies designed to block expansion of the fibroblast population (4). An attractive adjunct to these approaches would be therapy targeted to reduce the fibroblast population size within existing fibrotic lesions by the induction of fibroblast apoptosis.

It has been well documented in several cell systems that constitutive expression of growth-promoting genes can sensitize cells to undergo apoptosis (7). Studies in our laboratory have demonstrated that lovastatin potently induces apoptosis in fibroblasts constitutively expressing Myc (10). Lovastatin is a pharmacologic agent widely used in the treatment of hypercholesteremia. It acts by irreversibly inhibiting 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase, the rate-limiting enzymes in the pathway leading to the production of cholesterol. Lovastatin depletes cells of mevalonic acid, the direct product of the enzyme reaction, and thereby prevents the synthesis of downstream products such as cholesterol, heme A, and dolichol (11). In addition, mevalonic acid serves as a precursor to lipid moieties that are covalently attached to isoprenylated proteins (12). Post-translational modification of these proteins is unable to occur, diminishing the levels of mature protein products. As a result, lovastatin has an impact on many cellular functions essential for normal cell homeostasis including proliferation and viability (11).

One key cell signaling molecule affected by lovastatin is Ras, a small guanosine triphosphate (GTP) binding protein. Lovastatin can inhibit the activation of Ras as a result of interfering with its prenylation, a modification necessary for Ras localization to the inner plasma membrane. Ras is a key signaling molecule acting downstream of growth factors. Activation of Ras requires its localization to the inner plasma membrane, allowing it to interact with factors facilitating the binding of Ras to GTP. Ras-GTP, the active conformation of Ras, is capable of stimulating downstream signaling molecules including phosphoinositide-3 (PI3) kinase which has been directly implicated in cell survival (13) and mitogen-activated protein (MAP) kinase which is involved in proliferation (14).

Pertinent to the therapeutic potential of lovastatin for nonmalignant fibroproliferative diseases, prior studies in our laboratory indicate that fibroblasts isolated from fibrotic lungs constitutively express growth-promoting genes (15). In this study, we sought to determine if nontransformed fibroblasts would manifest susceptibility to lovastatin-induced apoptosis similar to that observed in fibroblasts ectopically expressing Myc. Here we show that clinically achievable concentrations of lovastatin induced apoptosis in normal and fibrotic fibroblasts in vitro and in vivo. In vitro, apoptosis of human lung fibroblasts was dose- and time-dependent, and blocked by exogenous mevalonic acid. In vivo, lovastatin dramatically reduced granulation tissue formation in a guinea pig wound chamber model with ultrastructural evidence of fibroblast apoptosis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell Lines

Normal lung fibroblasts were previously described and characterized (15). Fibrotic fibroblasts were prepared identically from the lungs of two patients with idiopathic pulmonary fibrosis (IPF) undergoing lung transplantation. Each IPF patient manifested progressive breathlessness, crackles, and clubbing on physical examination, a high-resolution computed tomography (CT) scan showing honeycombing without ground glass opacities, and a diagnostic histological pattern on lung biopsy (16). Neither IPF patient had an associated collagen vascular disease, and neither patient was receiving anti-inflammatory therapy for the 3 mo prior to transplantation. Cells designated here as fibroblasts had the following properties: ultrastructural characteristics of myofibroblasts; immunologically they were vimentin and smooth muscle actin positive; and by phase contrast microscopy they had a typical appearance in tissue culture (15).

All human fibroblasts were cultured (10% CO₂, 90% air at 37° C) in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% calf serum, 100 U/L penicillin, 100 µg/L streptomycin, and 250 ng/L amphotericin B, and used prior to the eighth subcultivation. For routine maintenance, medium was replaced twice weekly and cells were subcultivated at a one to three split ratio weekly. All fibroblasts were found to have a doubling time of 21 to 24 h. The basal rate of apoptosis for normal lung fibroblasts was less than 1%, whereas in the fibrotic strains it ranged from 4 to 18%.

Analysis of Apoptosis

Cells were examined for apoptosis morphologically using three methods: phase contrast microscopy, fluorescence microscopy after acridine orange staining, and transmission electron microscopy (17, 18). DNA integrity was assessed by flow cytometry (FACSCaliber; Becton Dickinson, Lincoln Park, NJ) after propidium iodide staining, agarose gel electrophoresis, and cytochemically by TUNEL, as previously described (19).

Induction of Apoptosis by HMG-CoA Reductase Inhibitors

Cell strains were seeded into 6-well clusters at 1.5 × 10⁴ cells/cm² and cultured in DMEM containing 10% calf serum for 24 h. HMG-CoA reductase inhibitors or vehicle were added to cells for the times and concentrations indicated prior to preparation for analysis. Because serum survival factors are known to modulate apoptosis, fibrotic fibroblasts were treated with 5 µM of lovastatin for 5 d in DMEM containing calf serum ranging from 0.1% to 10%. The greatest difference between lovastatin-treated and control samples occurred at a serum concentration of 0.1%. All experiments were therefore performed using 0.1% calf serum. Lovastatin and simvastatin were obtained from Merck, Sharpe, and Dohme (Rahway, NJ), and 5 mM of acid-activated solutions were prepared from 40-mg tablets as described by Jakobisiak and coworkers (20). In some experiments, 250 µM mevalonic acid (Sigma, St. Louis, MO) was added in conjunction with lovastatin.

Ras Immunoblots

Fibroblasts seeded at 1.5 × 10⁴ cells/cm² were treated with lovastatin at 5 µM or 10 µM for 48 h. To some cultures, 250 µM mevalonic acid was also added. Adherent and suspended cells were combined and lysed in reducing gel sample buffer (20% glycerol, 5% mercaptoethanol, 2.3% sodium dodecyl sulfate (SDS), 0.001% bromphenol blue, and 0.0625 M TRIS). For each sample, 40 µg of total protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (12%) and blotted onto nitrocellulose. Ras was identified using anti-Ras antibody (1:100; Santa Cruz Biotechnologies, Inc., Santa Cruz, CA), sheep anti-mouse IgG antibody (1:5,000; Sigma), and enhanced chemiluminescence (ECL; Amersham, Arlington, IL).

In Vivo Assessment of Apoptosis

A guinea pig wound chamber model of granulation tissue formation was adapted from Dvorak and colleagues (21) with minor modifications. Individual wound chambers were constructed by gluing a 13-mm round thermamox coverslip (Nunc Inc., Naperville, IL) to one surface of a Plexiglas ring (2 mm height; outer diameter 14 mm; internal diameter 10 mm) with MF cement formulation 1 (Millipore Corporation, Bedford, MA). Seven regularly spaced pores, 0.6 mm diameter were drilled in a second coverslip which was glued to the opposite surface of the Plexiglas ring to form the chamber roof. Chambers were sterilized and filled to capacity with DMEM containing thrombin (0.2 U/ml) and guinea pig fibrinogen (3 mg/ml), and in some of the chambers, 5 µM lovastatin. Gels were allowed to polymerize at 37° C for 2 h in a humidified incubator before implantation.

Six-week-old male Hartley guinea pigs weighing 400 to 500 g were obtained from Harlan Sprague Dawley (Indianapolis, IN). All experiments were carried out in accordance with the guidelines of the University of Minnesota Animal Care Committee. Guinea pigs were anesthetized for all surgical procedures with intraperitoneal sodium pentothal (Nembutal; Abott Laboratories, North Chicago, IL) (40 mg/kg). On each flank, three skin incisions were made (i.e., six total), and one chamber was implanted in each individual subcutaneous pocket. Each animal had six chambers implanted: three containing lovastatin and three containing vehicle only. In each position (shoulder, back, and flank), a control chamber was implanted on one side and a lovastatin chamber was implanted on the opposite side. Assignment of position was random. After 8 d, some chambers were prepared for ultrastructural analysis. Chambers were opened and the gels were fixed in 2% glutaraldehyde and 2% formalin overnight at 4° C, rinsed in sodium cacodylate buffer (0.1 M, pH 7.4) and postfixed in 1% osmium tetroxide. After washing with the same buffer, gels were dehydrated in an ethanol series and embedded in epon. Thin sections were viewed in a Philips 301 electron microscope. Other gels were harvested and fixed overnight in 10% buffered formalin for morphometric image analysis or additionally processed and stained with hematoxylin-eosin for light microscopy.

Morphometric Image Analysis

Wound chambers were photographed and the images were scanned and acquired into Adobe Photoshop as a digitized picture. Angiogenesis, as evidenced visually by a red coloration in the gels, was used to estimate granulation tissue formation. The model of Dvorak and colleagues is a granulation tissue model. The wound chambers are known to contain fibroblasts, macrophages, and interstitial collagen in the same areas where clusters of microvessels are localized. Studies by other laboratories similarly observe that wound chambers accumulate fibroblasts and collagen along with microvessels in reproducible proportions (22, 23). Therefore, wound chambers were used to study the genesis and regression of fibroproliferative tissue.

Image analysis was performed taking advantage of the difference in the red color intensities obtained on each wound chamber photograph (24). The image analysis software was able to delineate the area of a specific range of red density. Results were expressed as a percentage of the total wound chamber surface (S) occupied with granulation tissue [(GT/S) × 100%]. Statistical analysis was performed using the Mann-Whitney U test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lovastatin-induced Fibroblast Apoptosis In Vitro

When fibroblasts (fibrotic strain FF1) were incubated with lovastatin (5 µM) for 72 h, they underwent apoptosis. By phase contrast microscopy, membrane ruffling and blebbing were apparent in some cells within 24 h and in about 40% of the remaining cell population after 72 h (data not shown). After fixation and acridine orange staining, nuclei were found to be condensed and fragmented (Figure 1A). Fibroblasts incubated in lovastatin demonstrated a loss of DNA integrity assessed by agarose gel electrophoresis which exhibited the intranucleosomal cleavage pattern typical of myofibroblasts (Figure 1B). Corroboration of this result was obtained by TUNEL analysis (Figure 1C), which indicated that 6% of cells had positive nuclei after 24 h, increasing to between 20% and 30% of the adherent cell population by 72 h. Further evaluation of the kinetics of lovastatin-induced apoptosis by flow cytometry verified it was time-dependent, with a monotonically increasing apoptotic response ranging from 18 ± 3.1% after 24 h up to 63 ± 3% after 5 d in lovastatin (Figure 1D).

View larger version (78K): [in this window] [in a new window]   View larger version (120K): [in this window] [in a new window]   View larger version (53K): [in this window] [in a new window]   View larger version (97K): [in this window] [in a new window]   View larger version (73K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 1.   Lovastatin induces apoptosis in fibroblasts in vitro. Fibrotic fibroblasts (FF1) were incubated with 5 µM lovastatin for 72 h. Apoptosis was demonstrated by (A) acridine orange staining (top, control; bottom, 5 µM lovastatin); original magnification ×200, inset ×400; (B) agarose gel electrophoresis (lane 1 = 1 kb ladder, lane 2 = control, lane 3 = 5 µM lovastatin), and (C ) TUNEL (top, control; bottom, 5 µM lovastatin); original magnification ×200, inset ×400. (D) Kinetic profile of lovastatin-induced apoptosis was determined by incubation of fibroblasts with 5 µM lovastatin for 5 d with apoptosis quantified daily. Cells were fixed with 70% ethanol and stained with propidium iodide for analysis by flow cytometry. Hypodiploid DNA content was used as a measure of apoptosis.

Dose dependency was examined using two fibrotic strains and one normal fibroblast strain. Each was subjected to increasing concentrations of lovastatin for 4 d before analysis of apoptosis by flow cytometry (Figure 2A). FF1 was most sensitive to lovastatin, exhibiting an apoptotic frequency of 47 ± 4.6% at 5 µM, with a plateau value of 56 ± 3.5% beginning at 10 µM to 50 µM. The second fibrotic fibroblast strain (FF2) was less susceptible to apoptosis with an apoptotic frequency of 36 ± 2.9% at 50 µM. Normal human lung fibroblasts (HLF) were also susceptible to lovastatin-induced apoptosis with an apoptotic frequency of 12 ± 0.6% observed at 5 µM, and a plateau value of 64 ± 0.2% at 50 µM.

View larger version (16K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Figure 2.   Lovastatin-induced apoptosis is dose-dependent. (A) Two fibrotic fibroblast strains, FF1 (open squares) and FF2 (closed squares), and one normal human lung fibroblast strain, HLF (closed circles) were incubated with lovastatin for 4 d. (B) FF1 cells were treated with lovastatin (open squares), lovastatin + 250 µM mevalonic acid (closed squares), or 5 µM simvastatin (open circles) for 4 d. Apoptosis was quantified by flow cytometry.

To determine whether the apoptotic response observed with lovastatin was related to its ability to inhibit HMG-CoA reductase, simvastatin, another member of this class of drugs was examined. Simvastatin is a semisynthetic 2,2-dimethylbutyrate analogue of lovastatin that can potently inhibit cell growth (25, 26). FF1 treated with 5 µM simvastatin demonstrated a pattern of susceptibility similar to that observed with lovastatin (Figure 2B). To further examine whether inhibition of HMG-CoA reductase inhibition was the mechanism, FF1 were incubated with increasing concentrations of lovastatin in the presence of mevalonic acid for 4 d. Exogenous mevalonic acid, the metabolic intermediate whose production is blocked by HMG-CoA reductase inhibitors, at a concentration of 250 µM was able to completely eliminate apoptosis at lovastatin concentrations up to 50 µM (Figure 2B). These data indicated that apoptosis was caused by the deficiency of key metabolic intermediates downstream of HMG-CoA reductase.

Lovastatin-induced Apoptosis Was Associated with Inhibition of Ras Processing

One potential target of the mevalonic acid pathway known to regulate cell viability is Ras, which requires the addition of an isoprenoid group resulting in its insertion into the plasma membrane (27). The isoprenoid moiety is a downstream metabolite of the mevalonic acid pathway, and its production is inhibited by lovastatin (11). Immunoblot analysis was performed to determine if apoptotic concentrations of lovastatin blocked Ras activation in part through alterations in Ras processing. Mature, active Ras is known to migrate faster than unmodified Ras (27, 28). We found that fibroblasts incubated with lovastatin exhibited a slower migrating species of Ras, whereas untreated fibroblasts contained predominantly the faster migrating active form of Ras (Figure 3). The addition of exogenous mevalonic acid prevented lovastatin-induced alterations in Ras processing. These data indicated that doses of lovastatin that induce apoptosis also increased the proportions of unprocessed Ras.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Lovastatin altered Ras processing. Crude cell lysates were prepared from FF1 cells treated with lovastatin (5 µM, 10 µM, or 10 µM + 250 µM mevalonic acid) for 48 h. 40 µg/lane of total protein was separated by SDS-PAGE (12%). Detection of Ras was performed by immunoblot analysis and ECL (lane 1 = control, lane 2 = 5 µM lovastatin, lane 3 = 10 µM lovastatin, lane 4 = 10 µM lovastatin + 250 µM mevalonic acid).

Lovastatin-induced Fibroblast Apoptosis In Vivo

These in vitro results prompted an examination of lovastatin using an in vivo model of fibroproliferation. For this purpose we chose the well-characterized guinea pig wound chamber model of Dvorak and coworkers (21). This model afforded us the experimental advantages of direct administration of drug, obviating the need for daily oral dosing, and allowed for the precise control of the lovastatin concentration at the target site. Studies by other investigators have demonstrated that implanted wound chambers accumulate granulation tissue composed of myofibroblasts and new blood vessels after 10 d, with an associated increase in collagen content within the gels (22, 23). Consistent with these published observations, chambers filled with polymerized fibrin and placed under the skin of guinea pigs for 8 d formed granulation tissue (Figure 4A) containing the characteristic infiltrate of fibroblasts, microvessels, and inflammatory cells (Figure 4B). Digital image analysis of red coloration was used to estimate fibroproliferation because prior studies have documented that microvessels, fibroblasts, and collagen deposits are found within the chambers in constant proportions (21).

View larger version (125K): [in this window] [in a new window]   View larger version (110K): [in this window] [in a new window]   Figure 4.   Implanted fibrin gels formed granulation tissue with an associated fibroblast influx. Wound chambers were implanted in the backs of guinea pigs for 8 d before gels were harvested, processed, and stained with hematoxylin-eosin. (A) Cross section of fibrin gel demonstrating the formation of granulation tissue; original magnification ×100. (B) Longitudinal section of fibrin gel showing an infiltrate of mononuclear cells and fibroblasts; original magnification ×200.

Quantitative image analysis of the wound chambers revealed that lovastatin significantly decreased the total area occupied by granulation tissue (Figure 5). The mean values for granulation tissue formation in control and lovastatin chambers were 17.43% and 6.16%, respectively. This represents a statistically significant absolute reduction of 11.27% and a relative reduction of 64.7% in granulation tissue formation (p = 0.008, Mann-Whitney U test). Histological evaluation revealed a marked diminution in new blood vessels and fibroblasts in lovastatin-filled gels (data not shown), with total cellularity within these gels dramatically decreased from control. Ultrastructural examination performed on the cells remaining in the lovastatin-filled gels indicated that most were apoptotic. Furthermore, the apoptic cells had features of mesenchymal cells, excluding the possibility that the apoptotic cells were neutrophils or other inflammatory cells (Figures 5C and 5D). These studies indicated a powerful in vivo effect of lovastatin on fibroproliferation which in part resulted from the induction of fibroblast apoptosis.

View larger version (59K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   View larger version (97K): [in this window] [in a new window]   View larger version (190K): [in this window] [in a new window]   Figure 5.   Lovastatin induces fibroblast apoptosis in vivo. Wound chambers containing fibrin gels ± 5 µM lovastatin were implanted in the backs of guinea pigs for 8 d. Analysis of the wound chambers was performed by (A) comparison of photographs of wound chambers (left, control; right, 5 µM lovastatin). (B) Summary of wound chamber data. Photographs of gels were scanned and densitometric analysis was performed to determine the area of the gel covered by a specific range of red color, which was taken to represent new granulation tissue formation. Percent granulation tissue formation was calculated as the percent of granulation tissue (GT) over the total surface area (S), [(GT/S) × 100%]. Each symbol represents chambers from a specific guinea pig (n = 3). Statistical analysis was performed by the Mann-Whitney U test. (C ) Ultrastructural analysis of wound chambers was performed using transmission electron microscopy (left, control; right, 5 µM lovastatin). (D) Transmission electron microscopy showing evidence of many apoptotic cells in wound chambers containing lovastatin.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Therapy of fibroproliferative diseases has largely been relegated to blocking inflammation. However, clinical benefit is limited despite excellent anti-inflammatory efficacy of the agents used, suggesting therapy targeting fibroblasts is essential. At least three approaches can be considered. One approach is to block expansion of the fibroblast population by interfering with growth factors, their cognate receptors, or downstream signaling (4). A second approach is to block synthesis of collagen and other extracellular matrix components (3). A third approach which we propose, is to induce fibroblast apoptosis. We favor this strategy because it seeks to restore the physiological process for elimination of fibroproliferation during integumentary and visceral wound healing (29, 30). In this report, we show that lovastatin, a widely used HMG-CoA reductase inhibitor, is a potent inducer of fibroblast apoptosis. Apoptosis was found to be dose- and time-dependent and occurred at clinically achievable concentrations (31). Studies in a guinea pig wound chamber model indicated a potent inhibitory effect on fibroproliferation in vivo.

Fibroblast apoptosis was found to occur in both normal and fibrotic fibroblasts, and a difference in apoptotic response was not discernible between the two cell types. The fibroblast strain, FF1, exhibited the most apoptosis to lovastatin, but also had a higher basal rate of apoptosis. However, even after the absolute percentages are corrected to take into account spontaneous apoptosis, it is clear that lovastatin can induce significant apoptosis. The mechanism for the differences in apoptotic frequency among fibroblast strains is unknown, but may relate to cytokinetic properties which are known to correlate with apoptotic susceptibility (32). In vitro, both normal and fibrotic fibroblasts exhibit a similar proliferative capacity and apoptotic frequency. However, in vivo, normal fibroblasts are mainly quiescent whereas activated fibroblasts found in granulation tissue are highly proliferative. Therefore, it is possible that fibrotic fibroblasts will be more sensitive to apoptosis in response to inhibitors of HMG-CoA reductase than normal fibroblasts in vivo.

Prior studies have documented that lovastatin induces apoptosis in a number of cancer and transformed cell lines (33) including malignant glioma (37) and malignant mesothelioma (38). In most of these studies, the lovastatin concentrations used were within clinically obtainable limits, and in the human malignant glioma cells, doses as low as 100 nM over a 72-h period were effective. In nontransformed cells, only prostate stromal cells obtained from patients with benign prostatic hypertrophy (BPH) have manifested susceptibility to lovastatin-induced apoptosis (36). In that study, in vivo efficacy was not examined. Stromal cells from BPH bear similarity to fibroblasts isolated from fibrotic lung lesions in that they survive when deprived of growth factors, and have an enhanced proliferative capacity (15, 36).

The mechanisms of lovastatin-induced apoptosis are undefined. Our data allow us to speculate that the mechanism may involve Ras activation. We show a preponderance of slower migrating, inactive Ras in lovastatin-treated cells condemned to apoptosis compared with their untreated counterparts. A large body of evidence supports the role of Ras in promoting cell viability (13, 39, 40). Studies using dominant negative Ras demonstrated that Ras has a crucial role in the regulation of viability. More recent studies have demonstrated that survival signaling operates through activation of PI3 kinase and the Akt/protein kinase B (PKB) pathway. Together with studies in transformed cells (34, 38), our present findings implicate repression of Ras-mediated survival signaling as a possible mechanism of lovastatin-induced fibroblast apoptosis. In these reports, the production of mature Ras was found to be diminished by lovastatin, suggesting that viability signaling may have been blocked. Further support is provided by the ability of farnesyltransferase inhibitors to induce apoptosis of tumor cells by activating a latent apoptotic pathway (41). Definitive assignment of the mechanism by which HMG-CoA reductase inhibitors induce apoptosis awaits further studies.

Lovastatin potently inhibited granulation tissue formation in vivo. In addition to its proapoptotic effect on fibroblasts, lovastatin most likely blocked formation of granulation tissue by targeting the action of multiple cellular functions. Several steps in the growth factor signaling cascade are modulated by lovastatin, including growth factor receptor activity and signal transduction events (42). Mevalonic acid and its downstream intermediates are essential for cell cycle progression (11). In addition, lovastatin induction of the cyclin-dependent kinase inhibitor p27 leads to binding and inactivation of cell cycle progression factors, effectively arresting cells in G₁ (45). Other potential effects include inhibition of cell adhesion and migration (46, 47) which is supported by the observation that fewer cells entered the lovastatin-filled chambers when compared with controls. The inhibition of isoprenoid synthesis adversely affects other small GTP binding proteins, such as Rho, which are known to affect cytoskeletal function and cell migration. In Swiss 3T3 fibroblasts, lovastatin causes loss of focal adhesions and stress fibers (46), while in monocytes, it decreases expression of the adhesion molecule CD36 (47). Thus, the ability of lovastatin to inhibit multiple pathways in fibroproliferation makes assignment of mechanism speculative. Its robust functional impact, however, together with its safety as a drug, makes lovastatin an attractive candidate as a new therapy for fibrosis.

The results of our study should be interpreted cautiously. First, we have only limited data with other HMG-CoA reductase inhibitors. Second, we have explored only one model to show in vivo efficacy, and that model has only been tested using direct administration to establish proof of principle, not oral dosing as is necessary for a valid preclinical trial. However, the work reported here represents the first steps in demonstrating that lovastatin and other HMG-CoA reductase inhibitors can be effective in modulating fibroproliferation. We view these studies as promising because they support a therapeutic option that is safe and readily available for further preclinical and clinical testing in a class of diseases for which few if any effective alternatives exist.