Molecular Vision 1999; 5:17 <http://www.molvis.org/molvis/v5/p17/> Received 26 March 1999 | Accepted 3 August 1999 | Published 19 August 1999

Download Reprint

Hsp47-dependent and -independent intracellular trafficking of type I collagen in corneal endothelial cells

MinHee K. Ko,¹ EunDuck P. Kay^1,2
 
 

¹Doheny Eye Institute and ²Department of Ophthalmology, University of Southern California School of Medicine, Los Angeles, CA

Correspondence to: Correspondence to: EunDuck P. Kay, D.D.S., Ph.D., Doheny Eye Institute, 1450 San Pablo Street DVRC 203, Los Angeles, CA, 90033; Phone: (323) 442-6625; FAX: (323) 442-6688; email: ekay@hsc.usc.edu

Abstract

Purpose: Type I collagen is post-translationally regulated in corneal endothelial cells (CEC): CEC synthesize procollagen I and degrade it intracellularly. We investigated whether there is a Hsp47-independent pathway during intracellular trafficking of procollagen I.

Methods: Specific inhibitors were used to block intracellular transport of procollagen I and Hsp47. Immunocytochemical analysis was performed to determine the intracellular localization of the proteins of interest.

Results: When cells were treated with [alpha],[alpha]'-dipyridyl, this specific inhibitor for collagen promoted retention in the endoplasmic reticulum (ER) of some of the underhydroxylated procollagen I, which was colocalized with Hsp47 in CEC. At the same time, another fraction of the [alpha],[alpha]'-dipyridyl-induced underhydroxylated procollagen I was not located in the ER. When CEC were treated with brefeldin A, procollagen I and Hsp47 demonstrated a high degree of colocalization at the ER, whereas the inhibitor had less of an effect on the compartmentalization of procollagen I and prolyl 4-hydroxylase. When CEC were treated with either monensin or bafilomycin A1, procollagen I and Hsp47 were not colocalized: procollagen I was mostly localized at the Golgi area, while Hsp47 predominantly showed ER distribution. When colocalization of procollagen I and prolyl 4-hydroxylase was examined, a major population of procollagen I was not colocalized with prolyl 4-hydroxylase in the ER.

Conclusions: These results indicate that some procollagen I and Hsp47 travel together from the ER to the cis-Golgi compartment and that a major population of procollagen I that may not be properly hydroxylated may be destroyed intracellularly via the Hsp47-independent pathway in CEC.

Introduction

Corneal endothelium is a monolayer of endothelial cells covering the posterior surface of the cornea. The integrity of this layer is vital for maintenance of corneal transparency. Corneal endothelium in vivo responds to diverse types of pathology by converting to fibroblast-like cells [1,2]. These morphologically modulated cells, in turn, produce fibrillar collagens, the predominant species of which is type I collagen, and deposit an abnormal fibrillar extracellular matrix (retrocorneal fibrous membrane) between Descemet's membrane and the corneal endothelium [3]. The characteristic collagen phenotype in the retrocorneal fibrous membrane is type I collagen, whereas in normal corneal endothelium, both in vivo and in vitro, the predominant forms are types IV and VIII collagen [3-5]. Our previous study demonstrates that the steady-state level of [alpha]2(I) collagen RNA in normal corneal endothelial cells (CEC) is much higher than that in the modulated CEC that produce and secrete type I collagen. Further analysis of the 5'- and 3'-untranslated regions of [alpha]2(I) collagen RNA of normal CEC shows that the sequences in these regions are identical to those of modulated fibroblast-like CEC [6]. Therefore, the possibility of translational regulation based on mRNA sequences, as well as the possibility of transcriptional regulation of type I collagen expression in CEC, has been eliminated. Post-translational regulation is an obvious mechanism to explain the unique features of type I collagen expression in CEC [6,7]. Immunofluorescent staining of corneal tissue in vivo demonstrates that the corneal endothelium stains with anti-type I collagen antibodies; but there is no staining in the underlying Descemet's membrane. This observation suggests that type I collagen is synthesized in CEC in vivo, but there is no evidence of its secretion into the extracellular matrix. Pulse-chase experiments have further confirmed the synthesis and degradation of procollagen I in normal CEC [6]. These findings suggest that the newly synthesized procollagen I must undergo intracellular degradation in order not to be secreted under physiologic conditions.

Hsp47, a stress protein in the ER, is assumed to be a molecular chaperone specific to collagen [8-11]. It has been demonstrated that Hsp47 interacts with polysome-associated [alpha]1(I) procollagen chains; this interaction suggests that Hsp47 has a specific function relative to chain selection or completion of stable folding in procollagen I [10]. A recent report demonstrated the coexpression of the Hsp47 gene and both [alpha]1(I) and [alpha]1(III) collagen genes in carbon tetrachloride-induced rat liver fibrosis [12]. We have recently reported that the level of Hsp47 is greater in the modulated CEC than in normal CEC and that the amount of type I collagen associated with Hsp47 is higher in modulated CEC than in normal cells [13]. However, there is differential intracellular localization of Hsp47 and procollagen I in normal CEC: some fraction of the two proteins is colocalized, but their localization is not largely coincidental. On the other hand, modulated CEC demonstrate colocalization of these two proteins at the cis-Golgi area [13]. Therefore, we proposed that there may be differential pathways for the procollagen I that is fated for intracellular degradation. In the present study, we analyzed the intracellular trafficking of procollagen I and Hsp47 using specific inhibitors to block transport of newly synthesized proteins between organelles. We showed that there are both a Hsp47-dependent intracellular transport pathway and a Hsp47-independent pathway for the procollagen I synthesized by CEC. Interestingly, there is a large population of procollagen I in CEC that is not colocalized with prolyl 4-hydroxylase within the ER. This population of procollagen I may not be properly hydroxylated and, therefore, may ultimately be subjected to intracellular degradation. This subset of procollagen I may utilize the Hsp47-independent pathway before it is targeted to the degradation site.

Methods

Antibodies and probes

The following primary antibodies and immunofluorescent probes were used: goat anti-human type I collagen antibody (Chemicon International, Inc., Temecula, CA); mouse anti-rat Hsp47 (colligin) antibody (StressGen Biotechnologies Corp., Victoria, BC, Canada); mouse anti-Golgi 58K protein (Sigma, St. Louis, MO); rat anti-lysosomal associated membrane protein 2 (LAMP-2) antibody (StressGen Biotechnologies Corp.); sheep anti-chick prolyl 4-hydroxylase antibody for ER marker (a gift from Dr. Winston Kao, University of Cincinnati, Cincinnati, OH). Texas red-conjugated secondary antibodies were purchased from Vector Laboratories, Burlingame, CA, and FITC-conjugated secondary antibodies were purchased from Sigma.

Cell cultures and inhibitor treatment

Isolation and establishment of the rabbit CEC culture were performed as previously described [4]. Briefly, the Descemet's membrane-corneal endothelium complex was treated with 0.2% Type 2 collagenase (Worthington Biochemical Co., Lakewood, NJ) and 0.05% hyaluronidase (HSE; Worthington Biochemical Co.) for 60 min at 37 °C. Cultures were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco-BRL, Irvine, CA), supplemented with 10% fetal bovine serum and 50 µg/ml of gentamycin, in a 5% CO₂ incubator. This procedure has been shown to promote cell proliferation during the early phase of culture and to maintain the culture as a contact-inhibited monolayer when the cells reach confluence. First-passaged CEC were used for all experiments.

Immunofluorescent staining

CEC (3 x 10⁴/chamber) were seeded on the 4-well chamber slide and maintained in culture until they reached 80% confluence. Cells were treated with inhibitors under the conditions described below: 0.3 mM [alpha],[alpha]'-dipyridyl for 2 h; 2 µg/ml brefeldin A for 30 min; 5 µM monensin for 60 min; or 5 µg/ml bafilomycin A1 for 60 min. Brefeldin A and monensin were dissolved in 100% methanol, while [alpha],[alpha]'-dipyridyl and bafilomycin A1 were dissolved in 100% ethanol. Cells were washed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde in PBS for 15 min at room temperature. Cells were simultaneously permeabilized and blocked with 0.1% Triton-X-100 + 1% BSA in PBS for 15 min at 4 °C. Cells were incubated with primary antibodies prepared in 0.1% Triton-X-100 + 1% BSA in PBS for 1 h at 37 °C, and then washed with PBS. Cells were next incubated with the biotinylated secondary antibody (Vector Laboratories; 1:200 dilution) for 1 h followed by an extensive rinse in PBS and incubation with fluorescein conjugated to avidin (Vector Laboratories; 1:100 dilution) for 30 min. For double-staining studies, following the incubation of cells with two primary antibodies, cells were simultaneously incubated with Texas red-conjugated secondary antibody (1:50 dilution) and FITC-conjugated secondary antibody (1:50 dilution) for 30 min at 37 °C in the dark. After an extensive washing with PBS, the slides were mounted in a drop of Vectashield mounting medium (Vector Laboratories) to reduce photobleaching. For double-staining experiments in which both primary antibodies were produced in the mouse, the experimental procedures were modified as follows. Fixation, blocking, permeabilization and rinsing procedures were the same as described above. Cells were incubated with the first monoclonal antibody at 37 °C for 1 h, rinsed in PBS and incubated with the FITC-conjugated secondary antibody (anti-mouse IgG) at 37 °C for 30 min in the dark. Following the first monoclonal reactions, remaining mouse Fc sites were blocked by sequential incubation of cells with normal mouse serum and 0.01 mg/ml of goat anti-mouse IgG Fab fragment (ICN, Aurora, Ohio) at 37 °C for 1 h in order to saturate any open antigen binding sites on the first secondary antibody. Following extensive washing in PBS, the second monoclonal antibody was incubated at 37 °C for 1 h and cells were rinsed in PBS. The corresponding secondary Texas red-conjugated antibody (anti-mouse IgG) was incubated for 30 min at room temperature in the dark. After extensive washing with PBS, the coverslips were mounted in a drop of Vectashield mounting medium. The dilution of the primary antibodies was 1:25 for anti-type I collagen antibody and 1:100 for the other antibodies.

Confocal microscopy and image analysis

Antibody labeling was examined using a Zeiss LSM-20 laser scanning confocal microscope equipped with barrier filters for fluorescein (DTAF filter and Argon 488 nm as light source) and Cy3 epi-fluorescence (Helium Neon 543 nm as light source). A plan-neofluar x40 (N.A. 1.3) oil immersion objective was used for imaging of fluorescently labeled samples. Separate optical images of FITC or Texas-red were captured from the same optical section. The captured images were then pseudocolored green or red and digitally overlaid to visualize any colocalization. Regions of codistribution appear in yellow, reflecting the additive effect of superimposing green and red pixels. Image analysis was performed using the standard system operating software provided with the Zeiss LSM microscope (Version 2.08). Color photomicrographs were taken using a Sony printer connected to the video output of the microscope.

Results

Probes for ER, Golgi apparatus, and lysosome in CEC

The identification of cellular organelles was first determined by morphologic observations using antibodies for organelle-specific proteins. When intracellular localization of the ER was determined using anti-prolyl 4-hydroxylase antibody made to both [alpha]- and ß-subunits of the enzyme, the ER demonstrated a perinuclear distribution in CEC (Figure 1A). Distribution of Golgi apparatus was examined using anti-Golgi 58K protein antibody: the Golgi 58K protein was localized to the cytoplasmic face of the Golgi apparatus and bound to microtubules, which are the peripheral membrane components of the Golgi apparatus. Golgi 58K protein showed a compact, perinuclear distribution in CEC (Figure 1B). Distribution of lysosomes was determined using anti-LAMP-2 antibody in order to distinguish the staining profiles among the cellular organelles. LAMP-2 was distributed in a punctate globular fashion in CEC (Figure 1C). These organelle staining patterns were very similar to those reported previously [14].

Inhibition of transport of procollagen I from the ER

To determine whether hydroxylation of procollagen I plays a role in intracellular trafficking of the molecule and whether Hsp47 is involved in the transport of underhydroxylated procollagen I, CEC were treated with [alpha],[alpha]'-dipyridyl, an iron chelator that inhibits hydroxylation of prolyl residues in procollagen chains and subsequently inhibits triple-helix formation of procollagen molecules. Intracellular localization of procollagen I and Hsp47 were examined using doubling-staining with antibodies against type I collagen and Hsp47. In the absence of [alpha],[alpha]'-dipyridyl, the staining pattern in CEC showed that some of the two proteins colocalized (Figure 2A, yellow staining), but most of the two proteins did not colocalize (Figure 2A, green and red staining). However, underhydroxylated procollagen I, induced by [alpha],[alpha]'-dipyridyl, was mostly found to colocalize with Hsp47 (Figure 2B). When the ER-localization of the underhydroxylated procollagen I was further determined using anti-prolyl 4-hydroxylase antibody, there was as much noncolocalization of the underhydroxylated procollagen I and prolyl 4-hydroxylase (Figure 2C, red and green staining) as there was colocalization of the two proteins (Figure 2C, yellow staining).

In addition to [alpha],[alpha]'-dipyridyl, brefeldin A, which inhibits transport of the newly synthesized proteins from the ER to the cis-Golgi compartment, was employed in order to understand the role of Hsp47 in the export of procollagen I from the ER [15,16]. When CEC were treated with brefeldin A, procollagen I and Hsp47 demonstrated colocalization at the ER (Figure 3A, yellow staining); however, a minor population of Hsp47 was simultaneously present in structures similar to the small vesicles observed in chick embryo fibroblasts (Figure 3A, red staining). These structures were identified as an intermediate compartment between the ER and the cis-Golgi [17]. Although a small amount of procollagen I was colocalized with prolyl 4-hydroxylase at the ER (Figure 3B, yellow staining) in the presence of brefeldin A, a large population of procollagen I was not colocalized with the enzyme, suggesting that the two proteins may not be in the same compartment of the ER (Figure 3B, green and red staining). On the other hand, most of the Hsp47 was located at the ER (Figure 3C), as determined by double-staining with the ER marker, prolyl 4-hydroxylase. A fraction of Hsp47 was also present in the small vesicles (Figure 3C, red staining). Phase-contrast microscopic analysis further demonstrated that there are small vesicular structures in addition to the condensed ER structure in the brefeldin A-treated CEC (Figure 3D).

Inhibition of transport of procollagen I in the Golgi complex

Monensin is known to inhibit intracellular transport of proteins within the Golgi complex, preferentially between the cis- or medial-Golgi and trans-Golgi compartments [18,19]. When monensin-treated cells were double-stained with anti-type I collagen antibody and anti-Golgi 58K protein antibody, an accumulation of procollagen I within the compressed Golgi complex was clearly shown (Figure 4A, yellow staining). CEC treated with monensin exhibited flattened and enlarged morphology and demonstrated numerous vacuoles throughout the cytoplasm when examined by light microscopy (Figure 4B). This observation is consistent with our earlier report, in which electron microscopic analysis revealed numerous vacuoles in the absence of the characteristic Golgi structures in the monensin-treated CEC [20]. In contrast to the distribution of procollagen I in monensin-treated cells (Figure 4A), a large fraction of Hsp47 was localized at the ER as determined by colocalization with prolyl 4-hydroxylase (Figure 4C, yellow staining). A small amount of Hsp47 had a reticular staining pattern in small vesicular structures (Figure 4C, red staining). This staining pattern of Hsp47 differs from that of procollagen I. Double-staining of procollagen I and Hsp47 further confirmed their differential staining with virtually no coincidence between the two proteins (Figure 4D, green and red staining), suggesting that dissociation of procollagen I and Hsp47 takes place before these proteins reach the trans-Golgi network. This finding confirms the previous observation that procollagen I and Hsp47 dissociated between the post-ER and the cis-Golgi compartments [17]. Similar results were obtained when secretion was inhibited by treatment with bafilomycin A1, which is known to inhibit secretion at the trans-Golgi network [21]. When the staining profiles of Hsp47 and procollagen I were examined in the bafilomycin A1-treated CEC, the intracellular localization of the two proteins was completely different (Figure 5A, green and red staining). When the location of Hsp47 was determined by double-staining with anti-prolyl 4-hydroxylase, Hsp47 was colocalized at the ER with prolyl 4-hydroxylase (Figure 5B, yellow staining) and in the intermediate compartment (Figure 5B, red staining). On the other hand, procollagen I was not colocalized at the ER with prolyl 4-hydroxylase (Figure 5C, red and green staining) and most of the procollagen I was localized at the Golgi network (Figure 5D, yellow staining). Differential compartmentalization of procollagen I and Hsp47 in the cells treated with either monensin or bafilomycin A1 further confirms that procollagen I has already dissociated from Hsp47 in the cis- or medial-Golgi compartments. Like cells treated with monensin, CEC treated with bafilomycin A1 contained numerous vacuoles (data not shown).

Colocalization of procollagen I with prolyl 4-hydroxylase

Since [alpha],[alpha]'-dipyridyl promotes retention of the underhydroxylated procollagen I in the ER, albeit in part, we further explored whether procollagen I synthesized by CEC in the absence of specific inhibitor was colocalized with prolyl 4-hydroxylase in the same compartment of the ER. When CEC were double-stained with antibodies against type I collagen and prolyl 4-hydroxylase, anti-type I collagen antibody demonstrated a diffuse cellular staining along with the perinuclear ER staining (Figure 6A). Anti-prolyl 4-hydroxylase antibody showed a strictly perinuclear ER staining (Figure 6B). When the staining qualities of the two proteins were compared, it was clear that a large population of procollagen I was not located in the same compartment as prolyl 4-hydroxylase in the perinuclear ER (Figure 6C).

Discussion

The study reported here grew out of an attempt to resolve an apparently intriguing observation on the expression of type I collagen in CEC. Procollagen I with a correct composition of two pro [alpha]1(I) and one pro [alpha]2(I) chains is synthesized by CEC [6]. The intracellular presence of the molecule in vivo was demonstrated using immunocytochemical analysis. Pulse-chase experiments further demonstrate that procollagen I is intracellularly degraded; thus, secretion and subsequent deposition of the biologically undesirable type I collagen into the basement membrane environment is blocked [6]. In the subsequent study to elucidate the mechanism of the intracellular degradation of procollagen I in CEC, we assumed that Hsp47, an ER-resident protein known as a chaperone molecule specific for collagen, played a role in this pathway. We began with the assumption that CEC that must intracellularly degrade procollagen I may have a diminished ability to produce Hsp47. We demonstrate that the expression level of Hsp47 is lower in normal CEC than in the modulated CEC that secrete type I collagen into the matrix and that Hsp47 associates with procollagen I much less in the normal CEC than in the modulated cells [13]. We also showed that Hsp47 and procollagen I are colocalized in part of the ER, but a large population of the two proteins is not coincidental in normal CEC. On the other hand, the two proteins are colocalized at the Golgi apparatus in the modulated CEC [13]. Taken together, these findings suggest that procollagen I destined for intracellular degradation may have a different intracellular transport pathway than the molecules secreted into the matrix. Although CEC contain a low level of Hsp47, there are a few problems with assigning Hsp47 to the specificity on type I collagen: the association constants of Hsp47 with types I-V collagens are in the same range [8], and Hsp47 is also involved in the translation-translocation of type IV collagen [22]. In addition there is the question of whether procollagen I is selectively degraded in the presence of physiological basement membrane collagens (types IV and VIII) in CEC.

In the present study, an attempt was made to investigate whether there is a Hsp47-independent intracellular transport pathway for procollagen I, which ultimately leads to the intracellular degradation in CEC. First, specific inhibitors were employed to block specific routes of the intracellular trafficking of procollagen I and Hsp47. Treatment of cells with [alpha],[alpha]'-dipyridyl promotes retention of the under- (or non-) hydroxylated procollagen I in the ER, and the underhydroxylated procollagen I is colocalized with Hsp47 at the ER. However, an equal amount of underhydroxylated procollagen I induced by [alpha],[alpha]'-dipyridyl is not colocalized with prolyl 4-hydroxylase in the ER from its modifying enzyme, suggesting that this population of the underhydroxylated procollagen I may be located either in a different compartment of the ER or in the cytoplasm, perhaps at proteasome for degradation. This is highly speculative; and the involvement of proteasome in the intracellular degradation of procollagen I is under investigation. Unlike [alpha],[alpha]'-dipyridyl, which is a specific inhibitor for collagen, brefeldin A inhibits transport of the newly synthesized proteins from the ER to the cis-Golgi compartment. Brefeldin A affects the intracellular location of Hsp47: it retains most of the Hsp47 in the ER; and it induces some population of Hsp47 located in the intermediate compartment between the ER and cis-Golgi, as shown in chick embryo fibroblasts [17]. On the other hand, the inhibitor has no effect on the compartmentalization of procollagen I with its modifying enzyme: a large population of procollagen I is not colocalized with prolyl 4-hydroxylase. When inhibitors that affect transport between the Golgi apparatus were used, monensin and bafilomycin A1 demonstrated a differential staining pattern between procollagen I and Hsp47. Hsp47 is retained in the ER and in the small vesicles; procollagen I is retained in the Golgi apparatus. The observation clearly shows that dissociation of procollagen I and Hsp47 takes place before these proteins reach the trans-Golgi network. This finding confirms the previous observation that procollagen I and Hsp47 dissociate between the post-ER and the cis-Golgi compartment [17]. Of interest, monensin is able to retain procollagen I in the compressed Golgi area. This population of procollagen I may be properly folded and therefore able to be transported from the ER to Golgi. At present, the fate of the procollagen I retained in the Golgi area mediated by monensin or bafilomycin A1 is not known.

We further explored whether prolyl 4-hydroxylase has any role in the intracellular degradation of procollagen I under physiologic conditions. Prolyl 4-hydroxylase plays a central role in the synthesis of all collagens [23,24], since the 4-hydroxyproline residues formed in its reaction are essential for folding of the newly synthesized collagen polypeptide chains into triple-helical molecules. Therefore, we explored whether procollagen I synthesized in CEC was colocalized with prolyl 4-hydroxylase. Our data demonstrate that a large population of procollagen I is not located in the same compartment with prolyl 4-hydroxylase at the ER. This observation suggests that a large amount of procollagen I is not associated with prolyl 4-hydroxylase and that this population of the molecule may remain nonhydroxylated. These nonhydroxylated procollagen chains do not participate in a proper triple helical formation and are susceptible to intracellular degradation. This assumption is, in part, supported by the findings on the intracellular localization of the underhydroxylated procollagen I induced by [alpha],[alpha]'-dipyridyl: the intracellular localization of the molecule is not completely restricted to the ER and some fraction of the molecule appears to be distributed in the cytoplasm, perhaps at the site of proteasome, in which the improperly folded procollagen I may be degraded due to lack of hydroxylation. Taken together, these data indicate a likelihood that procollagen I synthesized by CEC under physiologic conditions is bound to specific ER chaperones and that assessment of the molecule to the enzyme is therefore impeded. It has been reported that immunoglobulin heavy chain-binding protein (Bip or Grp78), a Hsp70-related, ER-resident protein, binds pro[alpha]chains of procollagen I with mutations in the carboxyl-terminal peptide synthesized by cells from patients with osteogenesis imperfecta [25]. Bip is also involved in the export of proteins destined for degradation to the cytosol for ER degradation [26]. This possibility is currently being examined.

It has been known that human fibroblasts, when they are induced to make nonhelical, defective collagen, have mechanisms for degrading up to 30% of their newly synthesized collagen intracellularly prior to secretion [27]. Protein degradation is the final stage in cellular regulation of gene expression and represents the last chance for a cell to suppress the potentially pathologic effects of expression of aberrant proteins. The present study demonstrates that Hsp47 may not be the sole choreographer involved in the intracellular transport of the type I collagen that is intracellularly degraded. The intracellular degradation of procollagen I in CEC strongly suggests that the molecule is structurally favorable for intracellular degradation; it is likely to be underhydroxylated and improperly folded. These findings, therefore, bring us back to the fundamental questions regarding collagen biosynthesis, such as hydroxylation (a major post-translational modification) and proper folding into the triple-helix necessary for secretion. This reopened question requires a conceptual frame shift from addressing the question of why such an unorthodox process (synthesis and subsequent degradation) has taken place to determining what really happens to the molecule within the cells before they allow a mistake to occur. CEC appear to use the final stage of the regulatory mechanism for the gene expression of type I collagen, a pathologic phenotype in basement membrane. The study on the hydroxylation and triple-helical formation of procollagen I synthesized in CEC is conducted to test the hypothesis.

Acknowledgements

We are grateful to Dr. Bjorn R. Olsen (Harvard Medical School, Boston) for his insight on collagen gene expression. Support for this work was provided by NIH grants EY06431 and EY03040 and by an unrestricted grant from Research to Prevent Blindness.

References

1. Michels RG, Kenyon KR, Maumence AE. Retrocorneal fibrous membrane. Invest Opthalmol 1972; 11:822-31.

2. Brown SI, Kitano S. Pathogenesis of the retrocorneal membrane. Arch Ophthalmol 1966; 75:518-25.

3. Kay EP, Cheung CC, Jester JV, Nimni ME, Smith RE. Type I collagen and fibronectin synthesis by retrocorneal fibrous membrane. Invest Ophthalmol Vis Sci 1982; 22:200-12.

4. Kay EP, Smith RE, Nimni ME. Basement membrane collagen synthesis by rabbit corneal endothelial cells in culture. Evidence for an alpha chain derived from a larger biosynthetic precursor. J Biol Chem 1982; 257:7116-21.

5. Sawada H, Konomi H, Hirosawa K. Characterization of the collagen in the hexagonal lattice of Descemet's membrane: its relation to type VIII collagen. J Cell Biol 1990; 110:219-27.

6. Kay EP, Gu X, Choi SH, Ninomiya Y. Posttranslational regulation of type I collagen in corneal endothelial cells. Invest Opthalmol Vis Sci 1996; 37:11-9.

7. Kay EP, He YG. Post-transcriptional and transcriptional control of collagen gene expression in normal and modulated rabbit corneal endothelial cells. Invest Ophthalmol Vis Sci 1991; 32:1821-7.

8. Nagata K. Hsp47: a collagen-specific molecular chaperone. Trends Biochem Sci 1996; 21:22-6.

9. Natsume T, Koide T, Yokota S, Hirayoshi K, Nagata K. Interactions between collagen-binding stress protein HSP47 and collagen. Analysis of kinetic parameters by surface plasmon resonance biosensor. J Biol Chem 1994; 269:31224-8.

10. Sauk JJ, Smith T, Norris K, Ferreira L. Hsp47 and the translation-translocation machinery cooperate in the production of alpha 1(I) chains of type I procollagen. J Biol Chem 1994; 269:3941-6.

11. Nakai A, Satoh M, Hirayoshi K, Nagata K. Involvement of the stress protein HSP47 in procollagen processing in the endoplasmic reticulum. J Cell Biol 1992; 117:903-14.

12. Masuda H, Fukumoto M, Hirayoshi K, Nagata K. Coexpression of the collagen-binding stress protein HSP47 gene and the alpha 1(I) and alpha 1(III) collagen genes in carbon tetrachloride-induced rat liver fibrosis. J Clin Invest 1994; 94:2481-8.

13. Gu X, Ko MK, Kay EP. Intracellular interaction of Hsp47 and type I collagen in corneal endothelial cells. Invest Ophthalmol Vis Sci 1999; 40:289-95.

14. Ripley CR, Bienkowski RS. Localization of procollagen I in the lysosome/endosome system of human fibroblasts. Exp Cell Res 1997; 236:147-54.

15. Fujiwara T, Oda K, Yokota S, Takatsuki A, Ikehara Y. Brefeldin A causes disassembly of the Golgi complex and accumulation of secretory proteins in the endoplasmic reticulum. J Biol Chem 1988; 263:18545-52.

16. Lippincott-Schwartz J, Yuan LC, Bonifacino JS, Klausner RD. Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 1989; 56:801-13.

17. Satoh M, Hirayoshi K, Yokota S, Hosokawa N, Nagata K. Intracellular interaction of collagen-specific stress protein HSP47 with newly synthesized procollagen. J Cell Biol 1996; 133:469-83.

18. Tartakoff AM. Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 1983; 32:1026-8.

19. Rosa P, Mantovani S, Rosboch R, Huttner WB. Monensin and brefeldin A differentially affect the phosphorylation and sulfation of secretory proteins. J Biol Chem 1992; 267:12227-32.

20. Kay EP, Jester JV, Smith RE, Nimni ME. Effect of monensin on secretion of basement membrane collagen by cultured rabbit corneal endothelial cells in comparison with rabbit tendon fibroblasts. Connect Tissue Res 1984; 12:179-90.

21. Henomatsu N, Yoshimori T, Yamamoto A, Moriyama Y, Tashiro Y. Inhibition of intracellular transport of newly synthesized prolactin by bafilomycin A1 in a pituitary tumor cell line, GH3 cells. Eur J Cell Biol 1993; 62:127-39.

22. Ferreira LR, Norris K, Smith T, Hebert C, Sauk JJ. Hsp47 and other ER-resident molecular chaperones form heterocomplexes with each other and with collagen type IV chains. Connect Tissue Res 1996; 33:265-73.

23. Kivirikko KI, Myllyharju J. Prolyl 4-hydroxylases and their protein disulfide isomerase subunit. Matrix Biol 1998; 16:357-68.

24. Kivirikko KI, Myllyla R, Pihlajaniemi T. Protein hydroxylation: prolyl 4-hydroxylase, an enzyme with four cosubstrates and a multifunctional subunit. FASEB J 1989; 3:1609-17.

25. Chessler SD, Byers PH. BiP binds type I procollagen pro alpha chains with mutations in the carboxyl-terminal propeptide synthesized by cells from patients with osteogenesis imperfecta. J Biol Chem 1993; 268:18226-33.

26. Sommer T, Wolf DH. Endoplasmic reticulum degradation: reverse protein flow of no return. FASEB J 1997; 11:1227-33.

27. Berg RA, Schwartz ML, Rome LH, Crystal RG. Lysosomal function in the degradation of defective collagen in cultured lung fibroblasts. Biochemistry 1984; 23:2134-8.