Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Interstitial pulmonary fibrosis (IPF) continues to remain a challenging and perplexing problem for both researchers and clinicians. IPF is a crippling disease with a survival rate of less than 5 years. There is no known effective therapy for treating this disease. The current and most common therapy involves treatment with corticosteroids. However, fewer than 20% of the patients respond to this line of therapy. Moreover, corticosteroid therapy is associated with severe side effects (Hunninghake and Kalica, 1996).

IPF is the result of a wide variety of injuries to the lung (DePaso and Winterbauer, 1991). The development of fibrosis initially involves all aspects of inflammation, including microcirculatory changes, production of lipid-derived different mediators and cytokines, infiltration of immune cells, and cell-cell interaction and fibroblast proliferation followed by scarring (Sheppard and Harrison, 1992). The hallmark of lung fibrosis is an increase in extracellular matrix production in response to proliferating lung fibroblasts, and this results in an excess deposition of collagen in the lung interstitium (Crouch, 1990).

In the bleomycin (BL) rodent model of lung fibrosis, the increase in collagen production has been attributed to both the increase in the number of fibroblasts and the increase in the amount of collagen synthesized by each fibroblast (Goldstein and Fine, 1986). The amount of collagen deposited in the lung is controlled by a balance between synthesis, regulated transcriptionally, translationally, and post-translationally, and degradation, regulated by a whole host of enzymes and their inhibitors (McAnulty and Laurent, 1995). The balance between collagen synthesis and degradation can be tipped off in favor of synthesis by numerous factors like increases in fibrogenic cytokines and growth factor productions. This would then result in an excess buildup of collagen in the lung interstitum. In fact, accumulation of collagen in response to increased synthesis has been demonstrated in experimental models of lung fibrosis in hamsters and rabbits and a decreased degradation in rabbits as well (Seyer et al., 1976; Clark et al., 1983; Laurent et al., 1983). An excess accumulation of collagen has also been reported in humans with idiopathic pulmonary fibrosis (Seyer et al., 1976). The molecular basis of increased collagen synthesis during the course of the development of lung fibrosis appears to be an overexpression of procollagen (PC) I and III and transforming growth factor- (TGF-) and fibronectin genes (Raghow et al., 1985), and this involves both transcriptional and post-transcriptional mechanisms (Raghow et al., 1985, 1987; Breen et al., 1992).

Several animal models are used to investigate the molecular mechanisms of pulmonary fibrosis. The most widely used model is the intratracheal (i.t.) instillation of BL in the rodents. This model produces histological lesions and biochemical changes that resemble IPF seen in humans (Giri et al., 1980). The BL rodent model is used for studying the pathogenesis of lung fibrosis (Thrall et al., 1979) and for rapid screening of drugs for their potential antifibrotic effects (Zuckerman et al., 1980; Iyer et al., 1995).

Our laboratory has demonstrated the antifibrotic effects of pirfenidone (PD) in the BL hamster model of lung fibrosis (Iyer et al., 1995). PD was also found to retard the progression of an on-going fibrotic process in the same model (Iyer et al., 1998a). However, the molecular mechanisms for the antifibrotic effect of this novel compound are not known. This study was designed as part of a long-term goal of exploring the possible mechanisms for the antifibrotic effect of PD in the BL hamster model of lung fibrosis.

We hypothesized that the reduction in collagen content due to PD treatment could be due to initially by suppression of BL-induced lung inflammation followed by a down-regulation of BL-induced overexpression of PC I and III genes. The present study was designed to test this hypothesis by studying the effects of dietary intake of PD on BL-induced increases in lung lipid peroxidation (index of inflammation), prolyl hydroylase (PH) activity (enzyme responsible for post-translational modification of collagen), hydroxyproline content (index of collagen) and PC I and III mRNA accumulation. To determine whether transcriptional and/or posttranscriptional processes are involved, we also investigated whether the alteration in PC I and III gene expression by PD occurred at the level of gene transcription after BL instillation at 14 days and whether PD had any direct effect on PH activity in an in vitro study.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Treatment of Animals. Male golden Syrian hamsters weighing 90 to 110 g were purchased from Simonsens, Inc. (Gilroy, CA). Hamsters were housed in groups of four in facilities with filtered air and constant temperature and humidity. All care was in accordance with the National Institutes of Health guidelines for animal welfare. The hamsters were acclimatized for 1 week before any treatment. A 12/12-h light/dark cycle was maintained, and the animals had access to water and either pulverized Rodent Laboratory Chow 5001 (Purina Mills, Inc., St. Louis, MO) or the same pulverized chow containing PD (0.5% w/w). PD was generously supplied by Dr. Margolin (Marnac, Inc., Dallas, TX). The hamsters were fed these diets 2 days before i.t. instillation and throughout the study period. Hamsters were instilled i.t. with saline (SA) or BL (7.5 units/kg/5 ml). The animals were randomly divided into four experimental groups: 1) SA-instilled animals fed the control diet (CD) (SA + CD), 2) SA-instilled animals fed PD in the same diet (SA + PD), 3) BL-instilled animals fed the CD (BL + CD), and 4) BL-instilled animals fed PD in the same diet (BL + PD). The animals were sacrificed at 3, 7, 10, 14, and 21 days after the BL or SA instillation by decapitation, and their lungs were removed and freeze-clamped, dropped in liquid nitrogen, and stored at 80°C until used for mRNA analysis. The major portion of the sample was used for direct total RNA isolation, and the remainder was used for other biochemical studies.

Tissue Processing for Biochemical Study. The frozen lungs were thawed and homogenized in 0.1 M KCl/0.02 M Tris·HCl buffer, pH 7.6, with a Polytron homogenizer (Brinkmann Instruments Inc., Weastbury, NY). After recording the total homogenate volume (5-6 ml), it was mixed, divided into aliquots, and stored at 80°C, except for the aliquots for lipid peroxidation and hydroxyproline assays, which were processed and assayed the same day on which the lungs were homogenized.

Determination of Lipid Peroxidation. The lung malondialdehyde equivalent (MDAE) level, an index of lipid peroxidation, was determined in the whole homogenate according to the method of Ohkawa et al. (1979).

Determination of PH Activity. The method for PH assay was the same as reported previously and is based on the release of tritiated water from 3,4-[³H]proline-labeled unhydroxylated PC substrate prepared in vitro using 10-day-old embryonic chick tibiae (Giri et al., 1983). During the reaction, tritium is released in stoichiometric proportion to prolyl hydroxylation as tritiated water, which is collected and counted as a measure of the PH activity (Hutton et al., 1966). The activity was expressed as the total dpm released/lung/30 min.

Determination of PH Activity In Vitro. Five control hamsters not subjected to any treatment were first anesthetized with sodium pentobarbital (80-100 mg/kg). Their lungs were perfused with ice-cold isotonic SA; then, all lung lobes were dissected out and rinsed in SA. They were immediately homogenized in buffer containing 0.1 M Tris and 0.05% Triton X. The homogenate was spun down at 6000g for 20 min at 4°C. The supernatant was gently aspirated and used to determine PH activity and its protein content. The procedure to measure the PH activity was essentially the same as described above. Briefly, the reaction mixture in a total volume of 2.2 ml consisted of 200 µl of -ketoglutarate (0.001 M), 200 µl of ferrous ammonium sulfate (0.005 M), 250 µl of supernatant as enzyme source, 200 µl of PD to produce the desired final concentration, 200 µl of Tris·HCl (1 M), and 20 µl of ³H-unhydroxylated PC substrate (400,000 cpm). The reaction mixture was first preincubated with PD at different concentrations for 30 min at 37°C in a shaking water bath. The reaction was started by adding 200 µl of ascorbic acid (0.005 M); 30 min later, the reaction was terminated by adding 200 µl of 50% trichloroacetic acid. The tritiated water released was collected by vacuum distillation. Then 1 ml of the tritiated water was mixed with 10 ml of Ready Safe liquid scintillation cocktail (Beckman), and the radioactivity was determined at 45% counting efficiency in a scintillation counter. Protein content of the supernatant was determined according to the method of Lowry et al. (1951). The PH activity was expressed as total dpm associated with tritiated water released in the reaction mixture/mg protein/30 min.

Determination of Hydroxyproline. For assay of lung hydroxyproline as a measure of collagen content, 1 ml of whole homogenate was precipitated with 0.25 ml of ice-cold 50% (w/v) trichloroacetic acid and centrifuged, and the precipitate was hydrolyzed in 2 ml of 6 N HCl for 18 h at 110°C. The hydroxyproline content was measured according to the method of Woessner (1961).

Total RNA Isolation and Hybridization Analysis. Total RNA isolation from hamster whole lung samples were carried out according to the method of Chomczynski and Sacchi (1987). The Northern blot experiments were performed as described previously by the method of Gurujeyalakshmi et al. (1996). Total RNA (10 µg/lane) was electrophoresed through 1% agarose/2.2 M formaldehyde gels and transferred to a nylon membrane. The membranes were UV cross-linked (UV Stratalinker; Stratagene, La Jolla, CA). One set of the membranes were stained with methylene blue to confirm uniformity of RNA loading, transfer, and integrity. Prehybridization was carried out on another set of membranes at 42°C for 2 to 3 h in a solution containing 50% formamide, 5× sodium chloride sodium phosphate EDTA, 0.5% SDS, and 200 µg/ml sheared salmon sperm DNA. For hybridization, the radiolabeled probes were prepared by the random primer method (Bio-Rad, Richmond, CA). The denatured, ³²P-labeled cDNA probes of either PC I, PC III, or 18S rRNA were added to the prehybridization buffer, and hybridization was carried out at 42°C for 16 for 20 h. After hybridization, the membranes were washed as described previously (Gurujeyalakshmi et al., 1996) and exposed to Fuji X-ray film at 80°C with intensifying screens. The intensity of each band was determined by using the dual-wavelength flying spot scanning densitometer (model CS-9301PC; Shimadzu, Columbia, MD). The band intensity for PC I or PC III was divided by the band intensity of 18S rRNA to correct for variations in the quantity of RNA loaded. The ratio of the signal intensities of PC I mRNA/18S mRNA and PC III mRNA/18S mRNA are expressed as arbitrary units.

Nuclear Run-Off Transcription Analysis. The isolation of nuclei from the whole hamster lungs and the in vitro transcriptional reactions involving the labeling of the nascent RNA were performed according to the method of Gurujeyalakshmi et al. (1996). A slot-blot apparatus was used to prepare nytran membranes containing 20 µg/slot of plasmid DNA with cDNA inserts encoding PC I, PC III, and 18S rRNA. Then 20 µg of insert-free pBR322 vector was also included as a control for nonspecific binding. The plasmids were linearized, denatured, and neutralized before being blotted onto the membranes. The membranes were air dried for 20 min and UV cross-linked (UV Stratalinker; Stratagene). Then, they were prehybridized in 6× sodium chloride sodium phosphate EDTA, 0.5% SDS, and 200 µg/ml sheared salmon sperm DNA for 2 to 3 h at 65°C and hybridized for 36 to 40 h at 65°C with ³²P-labeled nascent RNAs (2-4 × 10⁶ cpm) per assay and washed in solutions of increasing stringency. The membrane strips were exposed to Fuji X-ray film for 48 to 72 h at 80°C with intensifying screens. The intensity of the DNA-RNA binding was determined by using the dual-wavelength flying spot scanning densitometer (model CS-9301PC; Shimadzu, Columbia, MD).

Molecular Probes. The collagen and 18S rRNA probes were obtained from American Type Culture Collection (Rockville, MD). They included the clone HF677, containing the cDNA insert coding for PC I (₁), a 1.8-kb EcoRI fragment; clone HF934 containing the cDNA insert coding for PC III (₁), a 1.3-kb EcoRI/HindIII fragment and 18S ribosomal RNA clone PN29III with a 0.752-kb BamHI and SphI cDNA insert. The plasmid pBR322 was purchased from Pharmacia (Piscataway, NJ). The plasmids were isolated using standard procedures; after endonuclease restriction digestion, the inserts were purified with a Qiagen gel extraction kit (Qiagen, Chatsworth, CA).

Statistical Analysis. All data are expressed as mean ± S.E.M. BL treatment increases the amount of proteins of extrapulmonary origin that can result in the artificial lowering of all values (Karlinski and Goldstein, 1980). Thus, the in vivo data are expressed on the basis of per lung. The data were compared within the four groups at the corresponding times using the two-way ANOVA where four groups were involved and the t test between the two groups. A value of P  .05 was considered to be the minimum level of statistical significance.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Lung Malondialdehyde Level. The levels of lung MDAE for various treatment groups are shown in Fig. 1. The MDAE values for both SA-treated control groups SA + CD and SA + PD averaged 60 ± 5 and 65 ± 4 nmol/lung, respectively. The i.t. instillation of BL significantly increased the MDAE values to 226%, 376%, and 202% of the corresponding SA + CD control groups at 10, 14, and 21 days, respectively. Treatment with PD significantly reduced these values to 81%, 123%, and 111% of their respective BL + CD groups at the corresponding times.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Effect of i.t. instillation of SA (control) or BL with and without dietary intake of PD on lipid peroxidation measured as MDAE levels of hamster lungs at different times. MDAE levels were measured as described in Materials and Methods. Each value represents mean ± S.E.M. of four animals. Treatment groups are indicated along the x-axis. *Significantly higher (P  .05) than other groups at corresponding times. +Significantly lower (P  .05) than BL + CD groups at corresponding times.

PH Activity. The PH activity in various groups is shown in Fig. 2. There was no significant difference in the PH activity between the two SA-treated SA + CD and SA + PD control groups at 3, 7, 10, 14, and 21 days. Treatment with BL increased the PH activity from 7 through 21 days in the BL + CD groups. Significant increases in PH activity occurred in these groups at 7, 10, 14, and 21 days by 208%, 359%, 240%, and 148% of their corresponding SA + CD control groups, respectively. Treatment with PD significantly reduced the PH activity in the BL + PD groups by 110%, 170%, and 85% at 10, 14, and 21 days compared with their corresponding BL + CD groups, respectively.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Effect of i.t. instillation of SA (control) or BL with and without dietary intake of PD on the PH activity of hamster lungs at different times. Lung PH activity was measured as described in Materials and Methods. Each value represents mean ± S.E.M. of four animals. *Significantly higher (P  .05) than other groups at corresponding times. +Significantly lower (P  .05) than BL + CD groups at corresponding times.

Lung PH Activity In Vitro. The direct effect of different concentrations of PD on the PH activity in vitro is shown in Table 1. The PH activity remained unchanged at any concentrations of PD because the activity was more or less the same in all cases with or without the PD preincubation of the enzyme . 

Lung Hydroxyproline Level. The effects of i.t. instillation of BL with and without PD treatment on the lung hydroxyproline levels at various times are summarized in Fig. 3. There were no significant differences in the lung hydroxyproline levels among the four groups at 3, 7, and 10 days after i.t. instillation of BL. However, the lung hydroxyproline levels in the BL-treated hamsters in the BL + CD groups were significantly increased to 1405 ± 37 and 1553 ± 14 µg/lung at 14 and 21 days compared with their corresponding SA + CD control groups, respectively. Treatment with PD caused significant reductions in the hydroxyproline levels to 1001 ± 114 and 941 ± 42 µg/lung in the BL + PD groups at 14 and 21 days compared with the corresponding BL + CD groups, respectively.

View larger version (45K): [in this window] [in a new window]   Fig. 3.   Effect of i.t. instillation of SA (control) or BL with and without dietary intake of PD on hydroxyproline content of hamster lungs at different times. Lung hydroxyproline content was measured as described in Materials and Methods. Each value represents mean ± S.E.M. of four animals. *Significantly higher (P  .05) than other groups at corresponding times. +Significantly lower (P  .05) than BL + CD groups at corresponding times.

PC I Gene Expression. The kinetics of PC I gene expression are shown in Fig. 4. The Northern blot analysis was carried out to determine the effect of PD in down-regulating the BL-induced overexpression of PC I mRNA. Treatment with BL gradually increased the PC I gene expression in the BL + CD groups from 7 to 21 days, peaking at 10 days. There was a significant increase in the PC I gene expression at 10, 14, and 21 days by 777%, 475%, and 462% of the corresponding SA + CD control groups, respectively. Treatment with PD gradually reduced the PC I gene expression in the BL + PD groups and showed significant reductions at 10, 14, and 21 days by 71%, 62%, and 77% compared with the corresponding BL + CD groups, respectively. Figure 5 represents the Northern blot and densitometric reading showing the overexpression of PC I gene in the BL + CD group and its down-regulation in the BL + PD group.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Effect of dietary intake of PD on BL-induced increases on steady-state level of PC I mRNA in hamster lungs at different time points after i.t. instillation of SA or BL. PC I mRNA was quantified as described in Materials and Methods. Each value represents mean ± S.E.M. of four animals. *Significantly higher (P  .05) than other groups at corresponding time points. +Significantly lower (P  .05) than BL + CD groups at corresponding times.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Demonstration of effect of PD on PC I mRNA accumulation in BL-instilled hamster lungs. Total RNA was extracted, and (A) PC I mRNA was quantified as described in Materials and Methods. Membrane shown in A was hybridized with PC I cDNA and then rehybridized with a cDNA probe for 18S rRNA as shown in B. C, densitometric analysis of autoradiogram shown in A.

PC III Gene Expression. The kinetics of PC III gene expression with and without PD treatment are shown in Fig. 6. Treatment with BL in BL + CD groups significantly increased the PC III gene expression at 14 and 21 days to 278% and 428% of their corresponding SA + CD groups. Dietary intake of PD significantly reduced the PC III gene expression in BL + PD groups by 76% and 85% of the corresponding BL + CD group at 14 and 21 days, respectively. Figure 7 represents the Northern blot and densitometric reading showing the overexpression of PC III gene in BL + CD groups and its down-regulation in BL + PD group.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Effect of dietary intake of PD on BL-induced increases on steady-state level of PC III mRNA in hamster lungs at different time points after i.t. instillation of SA or BL. PC III mRNA was quantified as described in Materials and Methods. Each value represents mean ± S.E.M. of four animals. *Significantly higher (P  .05) than other groups at corresponding time points. +Significantly lower (P  .05) than BL + CD groups at corresponding times.

View larger version (31K): [in this window] [in a new window]   Fig. 7.   Demonstration of effect of PD on PC III mRNA accumulation in BL-instilled hamster lungs. Total RNA was extracted, and (A) PC III mRNA was quantified as described in Materials and Methods. Membrane shown in A was hybridized with PC III cDNA and then rehybridized with a cDNA probe for 18S rRNA, as shown in B. C, densitometric analysis of autoradiogram shown in A.

Regulation of PC Genes at Transcriptional Level. We carried out a nuclear run-off transcription assay to determine whether the inhibition of PC I and III gene expression by PD is mediated by an alteration in gene transcription. Newly synthesized RNA transcripts from lung nuclei of hamsters in SA + CD, SA + PD, BL + CD, and BL + PD groups were isolated at 14 days after i.t. instillation of SA or BL as described in Materials and Methods. The transcription of PC I and III mRNAs was easily detectable in the nuclei prepared from lungs of hamsters in BL + CD and BL + PD groups. In contrast, we were not able to detect any transcripts of PC I or PC III in nuclei prepared from the lungs of hamsters in SA + CD and SA + PD groups. As shown in Fig. 8 and Table 2, the labeled transcripts that hybridized to cDNA of PC I were reduced significantly by 50% in the BL + PD group compared with the BL + CD group at 14 days. Although the transcription of PC III gene in the BL + PD group was less than that of BL + CD group, it was not significantly different. There was no difference at 14 days in the transcription of the gene encoding for 18S rRNA in the lung nuclei of hamster between the BL + CD and BL + PD groups. These results demonstrate that PD treatment down-regulates the BL-induced overexpression of PC I mRNA specifically at the transcriptional level.

View larger version (63K): [in this window] [in a new window]   Fig. 8.   Effect of PD on PC I and III gene transcription in BL-instilled hamster lungs at 14 days. Each value represents mean ± S.E.M. of five animals. See Materials and Methods for details. 32P-labeled nascent RNAs (2-4 × 106) from each assay were hybridized to 20 µg of plasmid PC I, PC III, and 18S rRNA containing cDNA inserts, which had been linearized, denatured, and immobilized on Nytran membranes. Inset, free plasmid pBR322 was included as control for nonspecific binding.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

BL-induced lung injury involves damage to the epithelial and endothelial cells of the lung, followed by infiltration of the inflammatory cells. These cells include predominantly neutrophils and macrophages, with some lymphocytes and eosinophils (Giri et al., 1986). The inflammatory process is highly amplified by the production of reactive oxygen species (ROS). In the BL model of lung injury, neutrophils are known to be sequestered very early into the lung, generating various proteolytic enzymes and free radicals. The neutrophils contain myeloperoxidase which catalyzes a reaction of H₂O₂ with Cl to form HOCl, which is reactive, cytotoxic, and injurious to the parenchymal cells (Klebenoff, 1988). The macrophages also generate ROS during phagocytosis, but more importantly, they release various proinflammatory cytokines and growth factors that cause fibroblast proliferation at the sites of the injury (Rennard et al., 1981; Kelly, 1990). In addition, BL binds to DNA and Fe²⁺, undergoes redox cycling, and generates ROS like superoxide and hydroxy radicals (Suguira et al., 1978; Caspary et al., 1982). These radicals are known to cause DNA strand scission and lipid peroxidation, resulting in damage to the lung (Giri et al., 1983). We found significant increases in the lung lipid peroxidation as reflected by the MDAE content in the BL + CD groups from day 10 through day 21, and treatment with PD significantly reduced the lung lipid peroxidation content at these time points. We previously demonstrated that dietary intake of PD decreased the BL-induced increases in the lung superoxide dismutase activity (Iyer et al., 1995, 1998a), indicating a lower level of ROS formation in the BL + PD than in the BL + CD groups. Furthermore, PD was found to be an effective scavenger of O₂, OH, and H₂O₂ generated in vitro (Valeyathan et al., 1996). Thus, it is possible that the antifibrotic action of PD may partly reside in its ability to scavenge the ROS generated by inflammatory cells and/or by the redox cycling of BL/DNA/Fe²⁺ complex. Thus it is highly likely that a reduction in the extent of ROS-induced lung damage by PD treatment in BL + PD group may subsequently attenuate the ensuing cascade of events which generally lead to the development of lung fibrosis.

Fibrosis is generally a final outcome of the inflammatory process. One of the strategies against the development of lung fibrosis is to down-regulate the excess collagen production. The synthesis of collagen involves transcription of the PC mRNA in the nucleus, which is translocated to the endoplasmic reticulum to translate PC polypeptide chains (McAnulty and Laurent, 1995). We observed a marked up-regulation in the PC I and PC III mRNA levels in the lungs of hamsters in BL + CD groups beginning from day 10 through day 21. This overproduction in the PC mRNAs in response to BL treatment is attributed to an increase in the rate of mRNA transcription and a prolonged half-life of the mRNAs (Raghow et al., 1985, 1987). The observed down-regulation of the mRNAs levels by PD treatment in the BL + PD group from day 10 through day 21 could be due to a reduction in the transcription rate and the message stability resulting from increased degradation or other factors that could affect the transcription or mRNA accumulation. Our study demonstrates that treatment with PD down-regulated the PC I gene transcription by 50% in the BL + PD group compared with the BL + CD group at 14 days but had little effect on PC III gene transcription. Our results also indicate that treatment with PD did not alter the rate of PC III gene transcription. The reasons for this differential effects of PD on gene transcription of PCs I and III are not clear; however, it is possible that the transcriptional down-regulation of PC III by PD might have occurred earlier than 14 days. In any event, the present study suggests that the observed reduction in PC I gene expression by PD treatment is regulated at least in part at the transcriptional level.

The increase in collagen synthesis is also regulated by various cytokines and growth factors (Kelly et al., 1990). In this regard, TGF- is of particular interest because it is well known to be involved in the pathogenesis of lung fibrosis (Khalil et al., 1991), and it probably functions as a master switch in triggering the sequentially interconnected events that lead to an excess deposition of collagen for the following reason. It has been demonstrated that increased TGF- mRNA transcription is followed by TGF- mRNA accumulation and TGF- protein, which precedes PC gene transcription followed by increased collagen synthesis and deposition (Breen et al., 1992; King et al., 1994). The data presented in the present report suggest that PD exerts its effect in part proximally at the transcriptional level and down-regulates the BL-induced overexpression of PC mRNAs by down-regulating the BL-induced overexpression of TGF- mRNA and TGF- protein, as shown in our preliminary study (Iyer et al., 1997, 1998b).

After the translocation of the PC mRNA to the endoplasmic reticulum to form polypeptide chain, it undergoes a number of cotranslational and post-translational modifications: initially, the hydroxylation of the proline residues catalyzed by PH, followed by glycosylation and secretion into extracellular space and cross-linking. The prolyl-4-hydroxylase catalyzed hydroxylation of proline is required for both stability and secretion of tropocollagen (Kivirikko et al., 1990). If this enzyme is inhibited, triple helix tropocollagen formation will not occur, and this will promote a rapid degradation of unhydroxylated pro- chains as demonstrated by various investigators (Philajameni et al., 1991). The data presented here are consistent with those of the latter study; BL treatment in the BL + CD group had significantly higher lung PH activity, which preceded the accumulation of collagen in the lung, and PD treatment, which inhibited the PH activity in the BL + PD group, also reduced the lung collagen accumulation in this group. Interestingly, PD structurally resembles certain PH inhibitors like pyridine 2,4-dicarboxylate (Dowell and Hadley, 1992). It is possible that PD could be acting as an inhibitor of prolyl-4-hydroxylase, thus reducing the hydroxylation of proline residues and allowing the deposition of a more pliable and soluble form of collagen. However, our in vitro experiment completely ruled out that possibility and demonstrated unequivocally that PD had no direct inhibitory effect on PH activity. Nevertheless, it is possible that PD may be indirectly inhibiting the activity of this enzyme in vivo by suppressing the BL-induced increased production of various cytokines and chemokines and lipid-derived mediators of inflammation.

From our study, it appears that the antifibrotic effect of PD could be attributed to attenuation of the inflammatory events by both the reduced influx of inflammatory cells into the lung and the ability of PD to scavenge free radicals. The observed down-regulation of PC gene expression by PD treatment in the BL + PD group can be attributed to its direct effect at the transcription of PC I gene and/or via fibrogenic cytokines like TGF-. Our studies also indicate that PD is not a direct inhibitor of prolyl-4-hydroxylase. The results of this study strongly suggest the involvement of a transcriptional mechanism at least in part for the decreased level of PC mRNA and collagen content.

This is the first mechanistic study that demonstrates multiple sites of action of PD in minimizing the BL-induced lung collagen accumulation. The antifibrotic effect of PD has been demonstrated in other models, including cyclophosphamide mouse model (Keher and Margolin, 1997). Although the latter study was not designed to define the mechanism, it is possible that PD might be exerting its antifibrotic effect even in that model via one of the mechanisms reported here. The data presented in this investigation once again demonstrate that PD is a novel antifibrotic agent, and if pretreatment or treatment with this compound is instituted at early phases of the development of lung fibrosis, it may offer beneficial effects against this crippling disease, which has so far defied all known therapeutic modalities.