Aberrant Signal Peptide Cleavage of Collagen X in Schmid Metaphyseal Chondrodysplasia

Schmid metaphyseal chondrodysplasia results from mutations in the collagen X (COL10A1) gene. With the exception of two cases, the known mutations are clustered in the C-terminal nonhelical (NC1) domain of the collagen X.In vitro and cell culture studies have shown that the NC1 mutations result in impaired collagen X trimer assembly and secretion. In the two other cases, missense mutations that alter Gly18at the −1 position of the putative signal peptide cleavage site were identified (Ikegawa, S., Nakamura, K., Nagano, A., Haga, N., and Nakamura, Y. (1997) Hum. Mutat. 9, 131–135). To study their impact on collagen X biosynthesis using in vitrocell-free translation in the presence of microsomes, and cell transfection assays, these two mutations were created inCOL10A1 by site-directed mutagenesis. The data suggest that translocation of the mutant pre-α1(X) chains into the microsomes is not affected, but cleavage of the signal peptide is inhibited, and the mutant chains remain anchored to the membrane of microsomes. Cell-free translation and transfection studies in cells showed that the mutant chains associate into trimers but cannot form a triple helix. The combined effect of both the lack of signal peptide cleavage and helical configuration is impaired secretion. Thus, despite the different nature of the NC1 and signal peptide mutations in collagen X, both result in impaired collagen X secretion, probably followed by intracellular retention and degradation of mutant chains, and causing the Schmid metaphyseal chandrodysplasia phenotype.

Collagen X is the most abundant extracellular matrix component synthesized by hypertrophic chondrocytes during the transition from cartilage to bone in endochondral ossification. It is classified as a short-chain nonfibrillar collagen and consists of three distinct protein domains; a central, short triple helical COL1 domain (463 amino acids) flanked by a small N-terminal nonhelical NC2 domain (38 amino acids) and a larger, more conserved, nonhelical C-terminal NC1 domain (161 amino acids). The chains are synthesized with an Nterminal signal peptide, which is proteolytically removed from the pre-␣1(X) chains during biosynthesis (1,2). In collagen biosynthesis, once the polypeptide chains are translocated into the lumen of the ER, the chains of the trimer associate via their C-terminal globular domains. This allows nucleation and folding of the triple helix to occur sequentially from this end of the molecule.
Mutations in the collagen X gene (COL10A1) result in Schmid metaphyseal chondrodysplasia (SMCD), 1 an autosomal dominant skeletal disorder characterized by short to normal stature, bowed legs, coxa vara, and flaring of the metaphyses of long bones. To date, all reported SMCD mutations in COL10A1, except two, are localized to the C-terminal NC1 domain of the protein (see review by Chan and Jacenko,Ref. 3). These NC1 domain mutations (COL10-NC1 m ) were proposed to affect the initial stages of the folding and chain assembly of collagen X (4,5), hindering nucleation of the triple helix and, therefore, impairing secretion of collagen X trimers.
The disruption of collagen X assembly as a mechanism underlying SMCD is supported by studies using in vitro transcription and translation of mutant and normal COL10A1 cDNAs in a cell-free system (1). In this system, the inability of COL10-NC1 m chains to trimerize was clearly demonstrated. Expression of COL10-NC1 m cDNAs in cells resulted in poor expression levels, with little or no secretion of the mutant chains (6). Together, these results indicate that the inability of mutant collagen ␣1(X) chains to assemble can lead to their intracellular retention and subsequent degradation. As a consequence, in SMCD, collagen X in the matrix could be reduced, potentially to 50% of normal levels, because of haploinsufficiency. The absence of detectable mutant COL10A1 mRNA in extracts of growth plate cartilage from a SMCD patient with a nonsense mutation in the NC1 domain, probably caused by nonsensemediated mRNA degradation, is consistent with haploinsufficiency as a mechanism in SMCD (7).
However, there is increasing evidence suggesting that other mechanisms could also underlie SMCD. For example, trace amounts of heterotrimers can be detected during cell-free coexpression of normal and mutant chains with amino acid substitutions in the NC1 domain (6). In addition, the clustering of COL10A1 mutations in the NC1 domain, the absence of null mutations and the autosomal dominant inheritance of SMCD in patients are more consistent with a dominant-negative mechanism. That is, the expression of mutant collagen X chains could impact on the assembly and secretion of normal chains as well. We (8) and others (9,10) have recently demonstrated through in vitro approaches that a dominant-negative mechanism could also underlie some SMCD mutations. Molecular modeling of the NC1 domain based on the crystal structure of ACRP30 (9) showed that NC1 amino acid substitutions in SMCD are localized to two regions of the folded domain and that these mutations may not totally abolish the ability of the mutant chains to form trimers.
Recently, two missense mutations in SMCD patients were identified in the putative signal peptide of the molecule at nucleotide positions 148 and 149 (G148A and G149A) 2 (11), altering Gly 18 at the Ϫ1 position of the signal peptide cleavage site (2). The molecular consequences of these mutations on collagen X biosynthesis are unknown but the predicted consequence is that secretion of the molecules will be impaired as the signal sequence plays a key role in this process (12). We have aimed to obtain a better understanding of the molecular mechanisms underlying SMCD in these patients by studying the impact of these two signal peptide mutations (SP m ) on collagen X biosynthesis and assembly. Our data show that the G148A and G149A mutations do result in impaired secretion of collagen X.

EXPERIMENTAL PROCEDURES
Construction of ␣1(X) Signal Peptide Mutations-Overlap extension PCR (13) was used to reproduce the two human missense mutations in collagen X identified by Ikegawa et al. (11). Both mutations altered Gly 18 at the Ϫ1 position of the putative signal peptide cleavage site. These were the G148A and G149A nucleotide substitutions which change the codon for Gly 18 to codons for Arg (G18R) and Asp (G18D), respectively. To create these changes, 5 ng of a plasmid pTM1-h10wt (6) containing a full-length human collagen X cDNA was used as a template for the primary rounds of PCR with primer pairs: sense, pTM1-1; antisense, G18R-2 or G18D-2; and sense, G18R-1 or G18D-1; antisense, HX10 (Table I) to generate independent fragments with overlapping sequences. Second round PCR reactions were carried out with primers pTM1-1 and HX10 and 5 ng of the two corresponding overlapping fragments from the primary PCR reactions.
PCR reactions were performed in 50 l of 10 mM Tris/HCl, pH 8.0, containing 1.5 mM MgCl 2 , 0.2 mM dNTPs, and 0.75 M of each of the primers. The reactions were carried out using the GeneAmp PCR system 2400 (Perkin-Elmer). Cycle one was performed at 96°C for 2 min, 60°C for 1 min, and 72°C for 1 min and then followed by 25 cycles at 96°C for 20 s, 60°C for 20 s, and 72°C for 30 s. The reaction was terminated at 72°C for 1 min. The recombinant fragment was purified and digested with KpnI and XhoI. A 337-base pair fragment containing the mutation was cloned into appropriate sites of pTM1 (14). Two positive clones, pTM1-G18R and pTM1-G18D were selected for sequencing to ensure that the correct mutation was introduced and that there was no PCR error within the 337-base pair KpnI/XhoI fragment. A full-length cDNA was constructed by cloning a 3-kilobase fragment generated by XhoI and SalI digestion of pTM1-h10wt (6) into the XhoI site of pTM1-G18R or pTM1-G18D.
In Vitro Cell-free Transcription and Translation-Cell-free transcription/translation was performed as described previously (1,6), in the presence or absence of canine pancreatic microsomal membranes (Promega), using the TNT T7 polymerase-coupled transcription and translation reticulocyte lysate system (Promega). The reactions were carried out in a total volume of 12.5 l and labeled with 10 Ci of translation grade L-[ 35 S]methionine (1000 Ci/mmol, PerkinElmer Life Sciences). To determine whether the mutant chains can form heterotrimers with normal chains, cotranslation experiments using an assembly-competent helix deletion ␣1(X) reporter construct, helix⌬ (6), were performed with either the G18R or G18D mutant constructs.
Coupled transcription and translation experiments were carried out at 30°C for 90 min using a total of 100 ng of purified plasmids. When required, the microsomes were separated from the reticulocyte lysate by centrifugation and analyzed separately. To assay for the location of the translation products, proteins external to the isolated microsome vesicles were digested with trypsin and chymotrypsin at a final concentration of 50 g/ml each (15).
Samples for SDS-PAGE analysis were dissolved or mixed with 40 l of sample loading buffer (10 mM Tris/HCl, pH 6.8, containing 2% SDS (w/v), 2 M urea, 10 mM dithiothreitol, and 20% sucrose (w/v)), and incubated at room temperature for 10 min prior to electrophoresis on a 7.5% SDS-polyacrylamide gel. Electrophoretic conditions and fluorography of radioactive gels have been described previously (16). Radioactive bands were imaged and quantified using a PhosphorImager (Molecular Dynamics).
Extraction of Proteins from Microsomal Vesicles with Sodium Carbonate-The sodium carbonate extraction procedure of Fujiki et al. (17) was used, with minor modifications, to determine whether the mutant chains remained as an integral component of the lipid bilayer. Following cell-free synthesis, microsome vesicles were separated from the reticulocyte lysate by centrifugation at 14,000 rpm, 4°C for 15 min, and washed with 0.5 ml of KHM buffer (110 mM KOAc, 20 mM HEPES, pH 7.2, 2 mM Mg(OAc) 2 ). Unbound proteins were extracted from the microsome vesicles with 0.1 M Na 2 CO 3 , pH 11.5, on ice for 30 min and then centrifuged as described above. The lipid bilayer was washed with 0.5 ml of KHM buffer. The supernatants were neutralized to pH 7.5 with 1 M HCl and the extracted proteins precipitated with 75% (v/v) ethanol. Proteins in both the Na 2 CO 3 -soluble, and the membrane fractions were dissolved in 40 l of sample loading buffer, as described above, prior to analysis on a 7.5% SDS-polyacrylamide gel.
Expression of Normal and Mutant Collagen X in Mammalian Cells-COL10A1 cDNA constructs were expressed in a rat osteogenic sarcoma cell line, UMR 106 -01 (American Type Culture Collection, ATCC-CRL 1661), using the vaccinia-driven T7 bacteriophage expression system previously described (6). Pulse-chase analysis was used to study collagen X secretion and degradation. Transfected cells in 6-well plates were preincubated in 1 ml of Dulbecco's modified Eagle's medium without L-methionine (Life Technologies, Inc.) for 1 h, then pulse-labeled for 2 h with 150 Ci of L-[ 35 S]methionine (1110 Ci/mmol, PerkinElmer Life Sciences). For secretion studies, labeled cells were chased with fresh medium containing excess unlabeled methionine over a period of 2 h. Collagen X from the cell and medium fractions was recovered by immunoprecipitation using a specific antibody (gift from Dr. Olena Jacenko) (7) and protein G-Sepharose (Roche Molecular Biochemicals) and analyzed by 7.5% (w/v) SDS-PAGE. To study the formation of stable triple helical molecules, aliquots of the cell fractions were subjected to limited pepsin digestion (16).
For analysis of intracellular collagen X degradation, cells were treated with either a proteasome inhibitor (5 M clasto-lactacystin ␤-lactone, CalBiochem) or an ER-Golgi transport inhibitor for the endosome/lysosome pathway (1 g/ml brefeldin A, Roche Molecular Biochemicals) throughout the preincubation, pulse labeling and chase periods. Cells were harvested after a 1-h chase, and collagen X was recovered by immunoprecipitation.

Mutations at Gly 18 Prevent Cleavage of the Signal Peptide-
Sequencing confirmed that the G148A and G149A mutations had been introduced in the cDNA constructs ( Fig. 1). In vitro transcription and cell-free translation demonstrated that the mutant cDNAs were translated into pre-␣1(X) chains with the same molecular size as the wild type (Fig. 2a, lanes 1-3). The lower molecular weight band in lanes 1-3 represents translation products initiated from the second methionine residue. In the presence of microsomes (Fig. 2a, lanes 4 -6), the initiation of translation was more accurate, with only a single translation product for each of the constructs. A reduction in the molecular size can be clearly demonstrated when wt pre-␣1(X) chains (Fig. 2a, lane 1) were processed to ␣1(X) chains (Fig. 2a, lane 4). In contrast, the mutant translation products remained as pre- ␣1(X) chains with no apparent reduction in the molecular size (Fig. 2a, lanes 5 and 6). Translation in the presence of microsomes promoted trimer association for wt as well as G18R and G18D translation products (Fig. 2a, lanes 4 -6). The homotrimers for G18R and G18D chains migrated with a slightly increased molecular size, suggesting that the mutant trimers contain chains with an unprocessed signal peptide. When microsomes were separated from the reticulocyte lysate prior to analysis, both the mutant and wild-type translation products were found predominately in the microsome fraction (Fig. 2b). The radioactive trimer and monomer components of the translation products were quantified using phosphor imaging. Trimer quantity expressed as percentage of the total translation product (trimers ϩ monomers) was used as a measure of the efficiency of trimer assembly. The percentages of mutant pre-␣1(X) chains that assembled into homotrimers were determined to be 20 Ϯ 4.5% (n ϭ 3) for G18R, 27 Ϯ 5.5% (n ϭ 3) for G18D, and 40 Ϯ 6.5% (n ϭ 3) for wild type.

Mutant Chains Can Associate in Vitro via the NC1 Domain to Form Heterotrimers with Signal Peptide-cleaved Chains-Be-
cause SMCD is an autosomal dominant disorder, an important consideration is whether the mutant pre-␣1(X) chains are able to associate into heterotrimers with normal ␣1(X) chains. To address this issue, reporter (helix⌬) and SP m plasmids were cotranscribed and translated in the presence of microsomes. The helix⌬ plasmid is an ␣1(X) construct, which contains a shortened triple helical domain but has normal NC1 and NC2 domains, and a normal signal peptide sequence. This construct has been used previously to generate a protein reporter, which allows the ability of mutant chains to assemble into heterotrimers in cotranslation experiments to be assessed (6). If the SP m chains can assemble with the helix⌬ chains, the stoichiometry of these complexes can readily be determined by assessing the electrophoretic migration of the multimers.
All translation products were localized to the microsomal membrane fraction (Fig. 3). Cotranslation of the SP m or wt transcripts with the helix⌬ transcripts showed that in addition to the trimeric components of each product, two additional intermediate multimeric bands, labeled as a and b, were also observed (Fig. 3, lanes 8 -10). This is consistent with band a containing heterotrimers of two pre-␣1(X) or wt chains and one helix⌬ ␣1(X) chain, and band b containing heterotrimers of one pre-␣1(X) or wt chain and two helix⌬ ␣1(X) chains. When compared with the cotranslation products of wt and helix⌬ transcripts (Fig. 3, lane 8), bands a and b in lanes 9 and 10 (Fig.  3) migrated with a slight increase in their apparent molecular sizes.
To assess trimer assembly efficiency, the trimer bands, ␣1(X) 3 /pre-␣1(X) 3 , band a, band b, and helix⌬␣1(X) 3 in lanes 8 -10, were quantified using phosphor imaging to estimate trimer assembly preference (Table II). The data showed that relative to the formation of helix⌬␣1(X) 3 , the formation of trimers containing one or more SP m chains in cotranslations with helix⌬ is less efficient compared with cotranslations of wt with helix⌬. In cotranslational experiments of SP m and helix⌬ chains, the assembly of heterotrimers containing one or two SP m chains appears to be more favored then the assembly of SP m homotrimers.
Mutant Chains with Uncleaved Signal Peptide Behave as Integral Membrane Proteins-Treatment of membrane vesicles at high pH solubilizes nonmembrane proteins and leaves only integral membrane proteins associated with the sedimentable lipid bilayers (18). Following cell-free synthesis in the presence of microsomes, the vesicles were treated with 0.1 M Na 2 CO 3 , pH 11.5 and centrifuged to determine whether the mutant chains remain integrated with the lipid bilayer. This approach revealed that the majority of the monomers and trimers containing uncleaved SP m chains associate primarily with the membrane pellet (Fig. 4a). Thus it appears that, in the absence of cleavage by signal peptidase, mutant pre-␣1(X) chains remain anchored to the translocons and behave as integral membrane components. We observed a similar pattern for cotranslation products of trimers containing SP m chains with an uncleaved signal peptide, which were preferentially retained in the membrane-bound fraction following treatment with Na 2 CO 3 (Fig.  4b). These included mutant pre-␣1(X) 3 homotrimers and the heterotrimers labeled a and b described above (Fig. 4b, lanes 2  and 3). In contrast, the homotrimers of the reporter (helix⌬), wt chains, and heterotrimers of helix⌬ and wt chains, were found predominantly in the soluble fraction (Fig. 4b, lanes 4 -6).
Mutations in the Signal Peptide Cleavage Site Impair Stable Triple Helix Formation and Secretion from Cells-The in vitro cell-free system does not promote stable triple helix formation because of the lack of appropriate post-translational modifying enzymes (6). To address the issue of helix formation, secretion, and the fate of the mutant chains, normal and mutant plasmids were transiently transfected into cells. Expression studies in UMR-106-01 cells, a collagen-producing osteoblast-like cell line, indicated that the wt chains assembled as trimers (Fig. 5), forming stable triple helical molecules (Fig. 6), which were secreted efficiently by the cells (Fig. 5). In contrast, whereas the SP m chains (G18R and G18D) associated into homotrimers intracellularly (Fig. 5), they were sensitive to pepsin digestion (Fig. 6), demonstrating a lack of helical configuration. In Fig. 6, the pepsin-resistant ␣1(I) and ␣2(I) chains are endogenous collagen I expressed by the UMR106-01 cells and were present in control untransfected cells that were only infected with the vTF7-3 virus. Secretion of the mutant chains was also severely impaired as no immunoprecipitable products could be detected in the medium fractions over the 2-h chase period (Fig. 5).
Unsecreted Mutants Chains Are Degraded Intracellularly in UMR106-01 Cells-Even though SP m trimers or chains were not secreted, the intracellular concentration decreases over the chase period, suggesting that the mutant chains are being degraded intracellularly (Fig. 5). To determine the likely degradative pathway involved, cells were treated from the preincubation period with either a proteasome inhibitor (clasto-lac-tacystin ␤-lactone) or a vesicular transport inhibitor for the endosome/lysosome pathway (brefeldin A). The concentration of mutant collagen X was compared after 1 h of the chase period. Cells transfected with the G18R or G18D constructs showed a similar response (Fig. 7). Incubation with brefeldin A resulted in a 4-fold increase in intracellular mutant chains compared with untreated cells and an approximately 8-fold increase was observed in the presence of clasto-lactacystin ␤-lactone.

DISCUSSION
Signal peptide sequences of secreted proteins share common features that include a net positive charge at the N terminus, a central hydrophobic region, and a C-terminal region with small nonpolar amino acids at positions Ϫ1 and Ϫ3 from the cleavage site (12). The Ϫ1 position of the putative signal peptide in collagen X is Gly 18 (2). In view of the common requirement for signal peptide sequences, the G18R and G18D mutations are significant, changing the nonpolar glycine residue at the Ϫ1 position to the highly charged arginine or aspartate residues. Based on reports of signal sequence mutations in other secretory proteins (15,19), the most likely consequence of these mutations is impaired cleavage of the signal peptide and inhibition of collagen X secretion.
Cell-free translations in the presence of microsomal membranes will allow the interaction of the signal recognition particle with the newly synthesized polypeptides, followed by docking and translocation into the microsome vesicles, where cleavage of the signal peptide will occur (20). This system was utilized to test whether the G18R and G18D mutations in COL10A1 alter these events. When microsomes were separated from the reticulocyte lysate following translation, the translation products of wt, G18R, and G18D were found associated with the microsome vesicles. Only small amounts of translation product remained in the reticulocyte lysate thus showing no indication of a preferential retention of mutant chains in the reticulocyte lysate when compared with wt translation. This suggests that interaction with the signal recognition particle and targeting to the translocons are not affected, as the mutant chains are delivered to the microsome vesicles during cell-free synthesis. Translocation into the lumen of the microsome vesicles was also not affected because we were able to demonstrate that the translation products are inside the microsomes with a protease protection assay using exogenous trypsin and chymotrypsin (data not shown). As the exogenous proteases cannot enter the vesicle, the presence of intact pre-␣1(X) chains implies that they are inside the microsomes (15).
Our in vitro data suggest that signal peptidase failed to recognize or cannot cleave the mutant sites. With no cleavage, the mutant pre-␣1(X) chains would not be released from the translocons and remain anchored to the microsomal membranes. The molecular consequence of mutations involving the signal peptide has not been characterized in the collagen family of proteins or other matrix components. However, there are a number of reports characterizing signal peptide mutations in secreted proteins that are associated with several diseases. For example, in human coagulation factor X deficiency (15) and diabetes insipidus (19,21), signal peptide mutations have been identified in factor X and prepro-vasopressin, respectively. For these nonmatrix-secreted proteins, failure to cleave the signal peptide resulted in intracellular retention of the mutant products, with a severe reduction in the amount of circulating protein and protein activity.
As shown by pulse-chase analysis in transiently transfected cells, impaired secretion of the mutant collagen X chains is clearly evident in the two cases reported here. That the nonsecreted SP m chains are targeted for intracellular degradation is TABLE II Relative trimer concentrations formed in cell-free co-translation experiments To assess trimer assembly efficiency, the trimer bands, ␣1(X) 3 /pre-␣1(X) 3 , band a, band b, and helix ⌬ ␣1(X) 3 , in lanes 8 -10 of Figure 3 were quantified using phosphor imaging to estimate trimer assembly preference. For each co-translation experiment, the values were calculated relative to helix ⌬ ␣1(X) 3 , which was set to unity. The values represent the average of three experiments.
pre-␣1(X) 3 Heterotrimer band a supported by the continued decline of intracellular levels of SP m chains over the chase period in the absence of secretion. The precise mechanism(s) for this intracellular degradation is  7. Degradation of SP m collagen X chains in UMR106-01 cells is prevented by proteasome and vesicular transport inhibitors. Transfections and biosynthetic pulse labeling with L-[ 35 S]methionine were performed as described under "Experimental Procedures." Cells were treated throughout the preincubation, pulse, and chase periods with (ϩ) or without (Ϫ) inhibitors for intracellular protein degradation; 5 M clasto-lactacystin ␤-lactone (Lac) for the proteasome, or 1 g/ml brefeldin A (BFA) for the endosome/lysosomal pathways. Cells were harvested after 1 h of chase and collagen X recovered by immunoprecipitation and analyzed on a 7.5% SDS-polyacrylamide gel. A standard (std) of pepsin-treated collagens from human skin fibroblasts was included as a molecular size reference. The position of the ␣1(I) and ␣2(I) chains of collagen I and the trimer of the unreduced collagen III are shown. Radioactive bands were imaged and quantified using a phosphor imager. The quantified value for the bands indicated as collagen X trimers and monomers were combined as a measure of total collagen X present in the cell fractions. The degree of protection against intracellular degradation by the addition of inhibitors was estimated relative to the corresponding untreated cells. not clear at this stage. Our preliminary study using inhibitors of lysosomal and proteasome pathways suggests that both degradative pathways may be involved. However, the more significant level of protection from degradation shown by the proteasome inhibitor, clasto-lactacystin ␤-lactone, suggests this could be the major degradative pathway.
The ability of SP m chains to associate into homo-and heterotrimers contrasts with our previous reports on SMCD mutations in the NC1 domain, where in vitro association of NC1 mutant collagen X chains was impaired (1,6). Although SP m chains do associate into homo-and heterotrimers, there appears to be some constraint on the nature of the trimers that are formed. In cell-free translation experiments expressing SP m or wt chains, we noted that the assembly of SP m pre-␣1(X) chains into homotrimers was less efficient relative to the assembly of wt homotrimers. When cotranslated with helix⌬ ␣1(X) chains, SP m chains showed a preference for assembly into heterotrimers over SP m homotrimers, with heterotrimers containing one SP m chain most favored.
Collagen X triple helix formation is initiated via nucleation and extension following the assembly of the C-terminal NC1 domains, so a possible reason why heterotrimers are preferred is the physical constraint imposed by the anchoring of SP m chains to the membrane. For the formation of homotrimers, only chains that are close enough will be able to associate; however, this constraint will have a lesser effect on the formation of heterotrimers containing two SP m chains, and be smallest for heterotrimers containing one SP m chain. We reason that whereas SP m chains can associate into homo-and heterotrimers via the NC1 domain in vivo, folding of SP m chain-containing molecules into a triple helix is likely to be inefficient because the N termini will be anchored to the ER membrane. A lack of stable triple helical structure was demonstrated for SP m homotrimers by the absence of intracellular pepsin resistance SP m collagen X molecules in transfected cells. The precise impact on the level of available normal molecules in vivo is not clear. In transient transfected cells, it is difficult to make reliable assessments of the yield of trimers because of the variable transfection efficiency and the added complication of intracellular degradation. However, we noted a reduction of ϳ60% in the amount of collagen X molecules secreted in a cotransfection experiment (data not shown), supporting a dominant-negative effect of SP m chains that could affect the level of normal collagen X molecules. We propose that SP m chains anchored to the translocons impair triple helix formation, and the unfolded chains are removed from the ER by retrograde translocation and targeted for degradation via the proteasome pathway (22,23).
It would appear that although the signal peptide mutations and the NC1 SMCD mutations reported previously (6) are at opposite ends of the molecule and have different molecular effects on collagen X biosynthesis, the net phenotype is similar, with intracellular retention and degradation of mutant collagen X. The fact that both types of mutations should give rise to SMCD is consistent with a similar molecular consequence for the mutations. Whereas an apparent 50% reduction in collagen X level has been shown in one patient to be a mechanism for SMCD (7), this may not be the only mechanism. The current finding together with other in vitro studies (8 -10) suggest that a dominant-negative effect may be a more common mechanism. Intracellular retention and active degradation processes of collagen molecules containing mutant chains could lead to deregulated cellular metabolism, altering cell differentiation, proliferation, and apoptosis. These are the critical cellular events in endochondral ossification that induce precise alterations in the ECM. An imbalanced ECM not only compromises matrix integrity but also alters cell-matrix interactions, eliciting aberrant cellular responses with unknown consequences. Further understanding of the impact of SP m and COL10-NC1 m SMCD mutations on the metabolism and differentiation of chondrocytes in the growth plate will need to await investigation of these mutations in animal models.