Type XIV collagen, a new homotrimeric molecule extracted from fetal bovine skin and tendon, with a triple helical disulfide-bonded domain homologous to type IX and type XII collagens

Previously undescribed disulfide-bonded collagenous pepsin-derived fragments have been isolated from fetal calf tendon and skin. One fragment, 10.5 kDa after reduction, was shown to be similar but distinct to the COL1 domain of the recently characterized type XII collagen (64% primary structure identity). The similarity includes important features such as size, location of the cysteine residues, and nature and position of an imperfection of the triple helix. From fetal calf skin, two approximately 34-kDa disulfide-bonded trimeric fragments were isolated in the unreduced form. Amino acid sequencing showed that one fragment contained solely the COL1 domain of type XII collagen while the other one only contained the COL1 domain of the new chain. Like type XII collagen, the new chain is therefore part of a homotrimeric molecule and should thus be considered as a distinct collagen type. We propose to call the molecule from which this fragment is derived, type XIV collagen, with a chain composition (alpha 1 (XIV]3. The presence of a domain similar to the COL1 domain of collagens types IX and XII suggests that type XIV collagen belongs to the group of fibril-associated collagens with interrupted triple helices (FACIT). Two other fragments, 13.5 and 17 kDa after reduction, were also purified. They were shown to contain the same triple helical domain with different pepsin cleavage sites at the amino terminus. Several tryptic peptides were sequenced, and the derived sequences could be aligned with the COL2 domain of type XII collagen or with flanking sequences in the NC2 and NC3 domains (61% sequence identity). These fragments are very likely to be also derived from type XIV collagen.     

Large and small splice variants of collagen XII: differential expression and ligand binding

Collagen XII has a short collagenous tail and a very large, three-armed NC3 domains consisting primarily of fibronectin type III repeats. Differential splicing within this domain gives rise to a large (320 kD) and a small (220 kD) subunit; the large but not the small can carry glycosaminoglycan. To investigate whether collagen XII variants have distinct expression patterns and functions, we generated antibody and cDNA probes specific for the alternatively spliced domain. We report here that the large variant has a more restricted expression in embryonic tissue than the small. For example, whereas the small variant is widespread in the dermis, the large is limited to the base of feather buds. Distinct proportions of mRNA for the two variants were detected depending on the tissue. Monoclonal antibodies allowed us to separate collagen XII variants, and to show that homo- and heterotrimers exist. Collagen XII variants differ in ligand binding. Small subunits interact weakly with heparin via their COOH-terminal domain. Large subunits have additional, stronger heparin-binding site(s) in their NH2-terminal extra domain. In vivo, both large and small collagen XII are associated with interstitial collagen. Here we show biochemically and ultrastructurally that collagen XII can be incorporated into collagen I fibrils when it is present during, but not after, fibril formation. Removal of the collagenous domain of collagen XII reduces its coprecipitation with collagen I. Our results indicate that collagen XII is specifically associated with fibrillar collagen, and that the large variant has binding sites for extracellular ligands not present in the small variant.     

Mapping of a putative surface-binding site of human coagulation factor XII

We have localized the binding epitope(s) of two murine monoclonal antibodies (B7C9 and P5-2-1) that were shown previously to inhibit the activation of human coagulation factor XII by negatively charged surfaces. A factor XII cDNA expression library in lambda gt11 was screened with antibody B7C9, and 16 immunoreactive bacteriophage were isolated. Fusion proteins from each of the recombinant phage were reactive with both monoclonal antibodies. Two of the phage cDNA inserts were found to code for amino acid residues -6-+31 and +1-+47 of factor XII, respectively, thereby defining the limits of the antigenic peptide to amino acids +1-+31. Each of the remaining 14 recombinant phage contained longer factor XII cDNA inserts that included sequences coding for the amino-terminal 31 amino acid residues. These results were confirmed by direct binding of antibody B7C9 to synthetic peptides containing amino acids 1-14 and 1-28 of factor XII. Further experiments with a set of nested peptides also indicated that amino acid residues 1-4 were essential but not sufficient for binding of B7C9 to the peptides. Hydrophobicity analysis of the amino-terminal region of plasma factor XII revealed a highly hydrophilic region between amino acid residues 5 and 15 that contained positively charged lysine residues at positions 8, 11, and 13. We conclude that a major epitope(s) recognized by monoclonal antibodies B7C9 and P5-2-1 is present in the amino-terminal 28 amino acids of factor XII. It is proposed that binding of these antibodies to factor XII blocks interaction of the positively charged region between residues 5 and 15 with negatively charged surfaces, thereby inhibiting activation.     

Amino acid sequence of the heavy chain of human alpha-factor XIIa (activated Hageman factor)

The amino acid sequence of the heavy chain of human alpha-factor XIIa (activated Hageman factor) was determined by automated Edman degradation using the peptides produced by chemical and enzymatic cleavages of intact factor XII and alpha-factor XIIa. Combining this sequence with the previously determined sequence of beta-factor XIIa (Fujikawa, K., and McMullen, B. A. (1983) J. Biol. Chem. 258, 10924-10933), the complete amino acid sequence of human factor XII has been established. The heavy chain of alpha-factor XIIa is composed of 353 amino acid residues containing one Asn-linked and six probable O-linked carbohydrate chains. The heavy chain of alpha-factor XIIa appears to contain four different domains including a "kringle," a "growth factor" domain, and the "type I" and "type II" domains of fibronectin. The domain organization of factor XII is analogous to those of several fibrinolytic proteins, including tissue plasminogen activator and urokinase, suggesting that factor XII belongs to the same protease subfamily as these two proteins.     

Identification and partial purification of a large, variant form of type XII collagen.

A large, alternate form of type XII collagen has been identified in cultures of the human epidermoid cell line WISH. This form, designated XIIA, is comprised of alpha chains that are approximately 90 kDa larger than the 220-kDa alpha chain previously characterized in extracts of fetal chicken and bovine tissues. Results from both collagenase digestion and rotary shadow analysis of partially purified material show that the increase is due to a larger NC3 domain. While both the large (XIIA) and the small (XIIB) forms of type XII collagen are identified in pulse-chase radiolabeling of fetal bovine skin explant culture, they are not related in a precursor-product fashion. Inhibition studies with alpha, alpha'-dipyridyl indicate that proper folding of the collagen helix is required for complete assembly and secretion of type XIIA in WISH cell culture. The 310-kDa alpha 1A chain is likely to represent the bovine equivalent of a second translation product, estimated to be 340 kDa, predicted from analysis of one complete chick cDNA sequence. Additionally, the amino-terminal amino acid sequence of the 220-kDa bovine alpha 1B chain was determined. This sequence is very near a potential alternate splice site predicted from analysis of chicken type XII cDNA.     

Cloning of the full-length coding sequence of rat liver-specific organic anion transporter-1 (rlst-1) and a splice variant and partial characterization of the rat lst-1 gene

The full-length coding sequence of rat liver-specific organic anion transporter-1 (lst-1) and its splice variant have been cloned. The full-length rat lst-1 (designated rlst-1a) encodes a protein containing 687 amino acids and has 12-putative transmembrane domains, multiple potential N-glycosylation and phosphorylation sites. Therefore, rat lst-1a has 35 additional amino acid residues compared to the previously reported rat lst-1. A splice variant (designated rlst-1c) reported in this communication encodes a protein containing 654 amino acids and has 10-putative transmembrane domains. PCR analysis suggests that rlst-1a is the most abundant form in liver. Phylogenetic analysis reveals that rat lst-1a is an ortholog of human LST-1 (hLST-1) and mouse lst-1 (mlst-1). The rlst-1 gene is composed of 15 exons and 14 introns. Analysis of exon-intron boundary reveals that the splice variant rlst-1c lacks the entire exon 7, while the previously reported rat lst-1 (designated herein as rlst-1b) lacks approximately half of exon 10, and the splicing has occurred through alternative usage of a splice donor site within exon 10.     

The structure of avian type XII collagen. Alpha 1 (XII) chains contain 190-kDa non-triple helical amino-terminal domains and form homotrimeric molecules

The monoclonal antibody 75d7, specific for type XII collagen (Sugrue, S.P., Gordon, M.K., Seyer, J., Dublet, B., van der Rest, M., and Olsen, B. R. (1989) J. Cell Biol., in press), was used to characterize the intact form of type XII collagen from chick embryo leg tendons. On an immunoblot of a 6% polyacrylamide gel of tendon extracts, one sharp band is recognized by the antibody at Mr = 220,000, while two fuzzy and poorly resolved bands are seen at Mr = 270,000 and Mr = 290,000. By immunoprecipitation of radiolabeled tendon culture media and electrophoresis of the precipitated material, bands with the same mobilities are observed, indicating that type XII collagen is not proteolytically processed in the extracellular space. Type XII collagen was extracted from tendons with 1 M NaCl in a Tris-HCl buffer and partially purified by concanavalin A-Sepharose and gel permeation chromatographies, using dot immunoblots to monitor the purification. Fractions highly enriched in bacterial collagenase-sensitive proteins with the same electrophoretic properties as type XII collagen were obtained. These fractions did not stain with Alcian blue and neither they nor the immunostained type XII collagen were affected by chondroitinase ABC digestion, indicating that type XII collagen is not a proteoglycan. A disulfide-bonded trimeric CNBr peptide was isolated by affinity chromatography on an antibody column and further purified by gel electrophoresis. Its NH2-terminal amino acid sequence was shown to be unique, demonstrating that type XII collagen is a homotrimer [alpha 1 (XII)]3. After bacterial collagenase digestion, both the immunopurified radiolabeled preparation and the purified tendon extract fraction showed by gel electrophoresis the presence of a large disulfide-bonded, 3 x 190-kDa, collagenase-resistant domain. Rotary shadowing and electron microscopy of the purified type XII fraction demonstrated that the molecule has the structure of a cross consisting of a 75 nm collagenase-sensitive tail, a central globule, and three 60 nm arms each ending in a small globule. After heat denaturation and renaturation, only a very large globule can be seen, attached to the triple helical tail. These results show that type XII collagen has a unique structure and is different from the other matrix constituents described so far.     

Isolation and differential tissue distribution of two human cDNAs encoding PDE1 splice variants.

A cDNA selection technique has been used to isolate full-length human cDNAs of the phosphodiesterase 1 (PDE1) calcium calmodulin (CaM)-regulated phosphodiesterase gene family. We isolated cDNAs representing multiple splice variants of PDE1A, 1B and 1C from a variety of tissues. Included among these were two novel splice variants for PDE1A and 1B. The first, PDE1A5, encodes a 519-residue protein, which is different from PDE1A1 by the insertion of 14 residues, a divergent carboxy terminus and also differs from PDE1A3 through a divergent amino terminus. Our second novel splice variant represents the first occurrence of a splice variant of the PDE1B gene. PDE1B2 encodes a 516-residue protein and diverges from PDE1B1 by the replacement of the first 38 residues by an alternative 18, which is predicted to be functionally significant. Using the splice variant sequence differences to perform comparative Northern analysis, we have demonstrated that each variant has a differential tissue distribution.     

Characterisation of proteoglycans and their catabolic products in tendon and explant cultures of tendon.

Tendons are collagenous tissues made of mainly Type I collagen and it has been shown that the major proteoglycans of tendons are decorin and versican. Little is still known about the catabolism of these proteoglycans in tendon. Therefore, the aim of the study was to characterise the proteoglycans including their catabolic products present in uncultured bovine tendon and in the explant cultures of tendon. In this study, the proteoglycans were extracted from the tensile region of deep flexor tendon and isolated by ion-exchange chromatography and after deglycosylation analysed by SDS-polyacrylamide electrophoresis, Western blotting and amino-terminal amino acid sequence analysis. Based on amino acid sequence analysis, approximately 80% of the total proteoglycan core proteins in fresh tendon was decorin. Other species that were detected were biglycan and the large proteoglycans versican (splice variants V(0) and/or V(1)) and aggrecan. Approximately 35% of decorin present in the matrix showed carboxyl-terminal proteolytic processing at a number of specific sites. The analysis of small proteoglycans lost to the medium of tendon explants showed the presence of biglycan and decorin with the intact core protein as well as decorin fragments that contained the amino terminus of the core protein. In addition, two core protein peptides of decorin starting at residues K(171) and D(180) were observed in the matrix and one core protein with an amino-terminal sequence commencing at G(189) was isolated from the culture medium. The majority of the large proteoglycans present in the matrix of tendon were degraded and did not contain the G1 globular domain. Furthermore the aggrecan catabolites present in fresh tendon and lost to the medium of explants were derived from aggrecanase cleavage of the core protein at residues E(373)-A(374), E(1480)-G(1481) and E(1771)-A(1772). The analysis of versican catabolites (splice variants V(0) and/or V(1)) also showed evidence of degradation of the core protein by aggrecanase within the GAG-beta subdomain, as well as cleavage by other proteinase(s) within the GAG-alpha and GAG-beta subdomains of versican (variants V(0) and/or V(2)). Degradation products from the amino terminal region of type XII collagen were also detected in the matrix and medium of tendon explants. This work suggests a prominent role for aggrecanase enzymes in the degradation of aggrecan and to a lesser extent versican. Other unidentified proteinases are also involved in the degradation of versican and small leucine-rich proteoglycans.     

Processing in the C-terminal domain of minicollagen XII removes a heparin-binding site

A minicollagen comprising the two C-terminal domains of collagen XII (COL1 and NC1) has been expressed in insect cells and characterized biochemically. An interaction with heparin is demonstrated, which depends on the correct folding of the molecule. After secretion, minicollagen XII is immediately processed to a form lacking heparin binding ability. Processed and unprocessed trimers differ only at the level of the eight or nine C-terminal residues but they reveal different structures as judged from rotary shadowing images. Similar processing is also observed in the medium of transfected human HeLa cells. These data show that a heparin-binding site is present in the C-terminal end of the chicken collagen XII sequence and strongly suggest that proteolytic processing in the NC1 domain can occur in vivo and regulate the interactive properties of collagen XII.     

Sites of the NUDT9-H domain critical for ADP-ribose activation of the cation channel TRPM2

TRPM2 is a cation channel unique within the transient receptor potential family because of its gating by ADP-ribose (ADPR). ADPR gating is enabled by a cytosolic C-terminal Nudix box sequence motif embedded into a region homologous to the NUDT9 ADPR pyrophosphatase. A recently discovered splice variant of TRPM2 (TRPM2-DeltaC) lacks 34 amino acid residues in the NUDT9 domain and is insensitive to ADPR. To analyze in detail which parts of the deleted sequence (DeltaC-stretch) are critical for ADPR gating, we tested mutants that lacked 19, 25, and 29 amino acid residues in the N-terminal part or had amino acid residues substituted in the remaining C-terminal part of the DeltaC-stretch. All of these mutants displayed typical ADPR-induced currents. However, the deletion or substitution of the amino acid residue Asn-1326 immediately downstream of the DeltaC-stretch abrogated ADPR gating. We furthermore analyzed the mutation I1405E/L1406F in the Nudix box of TRPM2, because a considerably decreased AD-PRase activity of the TRPM2 NUDT9-H protein in comparison to the NUDT9 pyrophosphatase has been attributed to the reverse exchange EF --> IL. The I1405E/L1406F variant of TRPM2 failed to respond to ADPR even at concentrations up to 10 mM. We concluded that the DeltaC-stretch contains no individual amino acid residues essential for ADPR gating but may act as a spacer segment stabilizing a conformation necessary for the essential residue Asn-1326 to interact with other channel regions. Enhancing the enzymatic activity of the Nudix box abolishes the ADPR gating of TRPM2, pointing to the requirement of prolonged binding rather than degradation.     

The roles of types XII and XIV collagen in fibrillogenesis and matrix assembly in the developing cornea

Corneal transparency depends on the architecture of the stromal extracellular matrix, including fibril diameter, packing, and lamellar organization. The roles of collagen types XII and XIV in regulation of corneal fibrillogenesis and development were examined. The temporal and spatial expression patterns were analyzed using semi-quantitative RT-PCR, in situ hybridization, Western analysis, and immunohistochemistry. Expression of types XII and XIV collagens in cornea development demonstrated that type XII collagen mRNA levels are constant throughout development (10D-adult) while type XIV mRNA is highest in early embryonic stages (10D-14D), decreasing significantly by hatching. The spatial expression patterns of types XII and XIV collagens demonstrated a homogeneous signal in the stroma for type XIV collagen, while type XII collagen shows segregation to the sub-epithelial and sub-endothelial stroma during embryonic stages. The type XII collagen in the anterior stroma was an epithelial product during development while fibroblasts contributed in the adult. Type XIV collagen expression was highest early in development and was absent by hatching. Both types XII and type XIV collagen have different isoforms generated by alternative splicing that may alter specific interactions important in fibrillogenesis, fibril-fibril interactions, and higher order matrix assembly. Analysis of these splice variants demonstrated that the long XII mRNA levels were constant throughout development, while the short XII NC3 mRNA levels peaked early (12D) followed by a decrease. Both type XIV collagen NC1 splice variants are highest during early stages (12D-14D) decreasing by 17D of development. These data suggest type XII collagen may have a role in development of stromal architecture and maintenance of fibril organization, while type XIV collagen may have a role in regulation of fibrillogenesis.     

Immunoidentification of type XII collagen in embryonic tissues

We have generated a monoclonal antibody against a synthetic peptide whose sequence was derived from the nucleotide sequence of a cDNA encoding alpha 1(XII) collagen. The antibody, 75d7, has been used to identify the alpha 1(XII) chain on immunoblots of SDS-PAGE tendon extracts as a 220-kD polypeptide, under reducing conditions. Amino-terminal amino acid sequence analysis of an immunopurified cyanogen bromide fragment of type XII collagen from embryonic chick tendons gave a single sequence identical to that predicted from the cDNA, thus confirming that the antibody recognizes the type XII protein. Immunofluorescence studies with the antibody demonstrate that type XII collagen is localized in type I-containing dense connective tissue structures such as tendons, ligaments, perichondrium, and periosteum. With these data, taken together with previous results showing that a portion of the sequence domains of type XII collagen is similar to domains of type IX, a nonfibrillar collagen associated with cross-striated fibrils in cartilage, we suggest that types IX and XII collagens are members of a distinct class of extracellular matrix proteins found in association with quarter-staggered collagen fibrils.     

The residues determining differences in ion affinities among the alternative splice variants F, A, and B of the mammalian renal Na-K-Cl cotransporter (NKCC2)

Three alternatively spliced variants of the renal Na-K-Cl cotransporter (NKCC2) are found in distinct regions of the thick ascending limb of the mammalian kidney; these variants mediate Na(+)K(+)2Cl(-) transport with different ion affinities. Here, we examine the specific residues involved in the variant-specific affinity differences, utilizing a mutagenic approach to change the NKCC2B variant into the A or F variant, with functional expression in Xenopus oocytes. The splice region contains the second transmembrane domain (TM2) and the putative intracellular loop (ICL1) connecting TM2 and TM3. It is found that the B variant is functionally changed to the F variant by replacement of six residues, half of the effect brought about by three TM2 residues and half by three ICL1 residues. The involvement of the ICL1 residues strongly suggests that this region of ICL1 may actually be part of a membrane-embedded domain. Changing six residues is also sufficient to bring about the smaller functional change from the B to the A variant; three residues in TM2 appear to be primarily responsible, two of which correspond to residues involved in the B-to-F changes. A B-variant mutation reported in a mild case of Bartter disease was found to render the cotransporter inactive. These results identify the combination of amino acid variations responsible for the differences among the three splice variants of NKCC2, and they support a model in which a reentrant loop following TM2 contributes to the chloride binding and translocation domains.     

Evidence that the final turn of the last transmembrane helix in the lactose permease is required for folding.

Although truncation of the hydrophilic C-terminal tail of the lactose (lac) permease of Escherichia coli (residues 401-417) has no significant effect on membrane insertion, stability, or transport activity, sequential substitution of stop codons for amino acid codons 398-401 leads to a progressive increase in transport activity and in the lifetime of the permease in the membrane (McKenna, E., Hardy, D., Pastore, J. C., and Kaback, H. R. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 2969-2973). Thus, either the last turn of putative helix XII or the region immediately distal to helix XII is important for proper folding, and hence, activity and resistance to proteolysis. In an effort to determine whether this 3-4-amino acid sequence comprises the final turn of the last transmembrane helix of the permease or the beginning of the hydrophilic C-terminal tail, we deleted residues 401-417 and replaced amino acid residues 397-400 with either 4 Leu residues ("helix making") or Gly-Pro-Gly-Pro ("helix breaking"). Permease with 4 Leu residues at positions 397-400 is fully functional with respect to transport and completely stable, as judged by [35S]methionine labeling experiments. In marked contrast, permease with Gly-Pro-Gly-Pro at the same positions exhibits minimal activity and is unstable. The results imply that the amino acid sequence ... Val397Phe398Thr399 Leu400 ... in lac permease may comprise the last turn of transmembrane helix XII, rather than the beginning of the C-terminal tail.     

Identification and characterization of a novel alternative splice variant of mouse GMx33alpha/GPP34

We have cloned and characterized a novel splice variant of mouse GMx33alpha/Golgi-associated protein of 34 kDa (GPP34), hereby designated GMx33alphaV/GPP34V. This splice variant skips the second and third exons, and the resulting frame shift generates a stop codon in the fourth exon. GMx33alphaV/GPP34V is comprised of 81 amino acid residues derived from the N-terminal end of the full length protein and corresponds to approximately one-third of the full length GMx33alpha/GPP34 sequence with a calculated molecular mass of 8900. In contrast to GMx33alpha/GPP34 mRNA which is expressed at similar levels in various tissues, GMx33alphaV/GPP34V mRNA was differentially expressed when examined by RT-PCR. Compared to other tissues, skeletal muscle showed relatively strong expression of GMx33alphaV/GPP34V mRNA. This splice variant cDNA was also detected in a human cell line.     

The complete primary structure of type XII collagen shows a chimeric molecule with reiterated fibronectin type III motifs, von Willebrand factor A motifs, a domain homologous to a noncollagenous region of type IX collagen, and short collagenous domains with an Arg-Gly-Asp site

Extracellular matrix molecules are generally categorized as collagens, elastin, proteoglycans, or other noncollagenous structural/cell interaction proteins. Many of these extracellular proteins contain distinctive repetitive modules, which can sometimes be found in other proteins. We describe the complete primary structure of an alpha 1 chain of type XII collagen from chick embryonic fibroblasts. This large, structurally chimeric molecule identified by cDNA analysis combines previously unrelated molecular domains into a single large protein 3,124 residues long (approximately 340 kD). The deduced chicken type XII collagen sequence starts at the amino terminus with one unit of the type III motif of fibronectin, which is followed by one unit homologous to the von Willebrand factor A domain, then one more fibronectin type III module, a second A domain from von Willebrand factor, 6 units of type III motif and a third A domain, 10 consecutive units of type III motif and a fourth A domain, a domain homologous to the NC4 domain peptide of type IX collagen, and finally two short collagenous regions previously described as part of the partially sequenced collagen type XII molecule; an Arg-Gly-Asp potential cell adhesive recognition sequence is present in a hydrophilic region at the terminus of one collagenous domain. Antibodies raised to type XII collagen synthesized in a bacterial expression system recognized not only previously reported bands (220 kD et cetera) in tendons, but also bands with apparently different molecular sizes in fibroblasts and 4-d embryos. The antibodies stained a wide variety of extracellular matrices in embryos in patterns distinct from those of fibronectin or interstitial collagens. They prominently stained extracellular matrix associated with certain neuronal tissues, such as axons from dorsal root ganglia and neural tube. These studies identify a novel chimeric type of molecule that contains both adhesion molecule and collagen motifs in one protein. Its structure blurs current classification schemes for extracellular proteins and underscores the potentially large diversity possible in these molecules.     

Type XII collagen. A large multidomain molecule with partial homology to type IX collagen

Three overlapping cDNAs encoding alpha 1 (XII) collagen have been isolated and sequenced. The DNAs define five sequence domains within the chain. Three domains are nontriple-helical; two are relatively short triple-helical regions. The amino acid sequences of tryptic peptides derived from 16- and 10-kDa pepsin-resistant fragments isolated from tendon extracts are in full agreement with the deduced sequences of the triple-helical regions. Two of the five sequence domains in alpha 1 (XII), one triple-helical and one nontriple-helical, show a high degree of similarity to regions in type IX collagen chains. In addition, examination of seven exons in the alpha 1 (XII) gene shows that the gene is, in part, similar to the structure of type IX collagen genes. Therefore, collagen types IX and XII are partially homologous. The alpha 1 (XII) sequence data predict an asymmetric structure for type XII collagen molecules, fully consistent with the rotary shadowing images. These images show a triple-helical 75-nm tail attached through a central globule to three finger-like structures, each 60 nm long (Dublet, B., Oh, S., Sugrue, S. P., Gordon, M. K., Gerecke, D. R., Olsen, B. R., and van der Rest, M. (1989) J. Biol. Chem. 264, 13150-13156).