Crystallization and preliminary X-ray analysis of human transglutaminase 3 from zymogen to active form.

Transglutaminases(TGases; protein-glutamine-glutamyl-transferases) are a large family of calcium-dependent acyl-transfer enzymes that catalyze the formation of covalent cross links in proteins. Of these, the "epidermal" or "hair follicle" TGase 3 isoform is critically involved in barrier formation in epithelia. It is a zymogen, requiring proteolytic activation to achieve maximal specific activity. In order to understand its structure and function, we have devised methods for the rapid large-scale expression of the TGase 3 zymogen in the baculovirus system, and here we describe the purification of the zymogen and activated forms. We describe methods for the formation of high-quality, well-diffracting crystals within 3-5 days, using both dioxane and beta-octylglucoside to overcome severe twinning problems. The crystal of the zymogen belongs to the triclinic space group P1 and diffracts to 2.2-A resolution, and the crystal of the active form belongs to the P2(1) space group at 2.7-A resolution.

In vitro assembly and structure of trichocyte keratin intermediate filaments: a novel role for stabilization by disulfide bonding.

Intermediate filaments (IF) have been recognized as ubiquitous components of the cytoskeletons of eukaryotic cells for 25 yr. Historically, the first IF proteins to be characterized were those from wool in the 1960s, when they were defined as low sulfur keratins derived from "microfibrils." These proteins are now known as the type Ia/type IIa trichocyte keratins that constitute keratin IF of several hardened epithelial cell types. However, to date, of the entire class of >40 IF proteins, the trichocyte keratins remain the only ones for which efficient in vitro assembly remains unavailable. In this paper, we describe the assembly of expressed mouse type Ia and type IIa trichocyte keratins into IF in high yield. In cross-linking experiments, we document that the alignments of molecules within reduced trichocyte IF are the same as in type Ib/IIb cytokeratins. However, when oxidized in vitro, several intermolecular disulfide bonds form and the molecular alignments rearrange into the pattern shown earlier by x-ray diffraction analyses of intact wool. We suggest the realignments occur because the disulfide bonds confer substantially increased stability to trichocyte keratin IF. Our data suggest a novel role for disulfide bond cross linking in stabilization of these IF and the tissues containing them.

Transglutaminase cross-linking properties of the small proline-rich 1 family of cornified cell envelope proteins. Integration with loricrin.

Small proline-rich 1 (SPR1) proteins are important for barrier function in stratified squamous epithelia. To explore their properties, we expressed in bacteria a recombinant human SPR1 protein and isolated native SPR1 proteins from cultured mouse keratinocytes. By circular dichroism, they possess no alpha or beta structure but have some organized structure associated with their central peptide repeat domain. The transglutaminase (TGase) 1 and 3 enzymes use the SPR1 proteins as complete substrates in vitro but in different ways: head domain A sequences at the amino terminus were used preferentially for cross-linking by TGase 3, whereas those in head domain B sequences were used for cross-linking by TGase 1. The TGase 2 enzyme cross-linked SPR1 proteins poorly. Together with our data base of 141 examples of in vivo cross-links between SPRs and loricrin, this means that both TGase 1 and 3 are required for cross-linking SPR1 proteins in epithelia in vivo. Double in vitro cross-linking experiments suggest that oligomerization of SPR1 into large polymers can occur only by further TGase 1 cross-linking of an initial TGase 3 reaction. Accordingly, we propose that TGase 3 first cross-links loricrin and SPRs together to form small interchain oligomers, which are then permanently affixed to the developing CE by further cross-linking by the TGase 1 enzyme. This is consistent with the known consequences of diminished barrier function in TGase 1 deficiency models.

Structural and transglutaminase substrate properties of the small proline-rich 2 family of cornified cell envelope proteins.

The small proline-rich (SPR) proteins are components of the cornified cell envelope of stratified squamous epithelia and become cross-linked to other proteins by transglutaminases (TGases). The SPR2 family is the most complex, as it consists of several differentially expressed members of the same size. To explore their physical and cross-linking properties, we have expressed in bacteria a human SPR2 family member, and purified it to homogeneity. By circular dichroism, it possesses no alpha or beta structure but has some organized structure associated with the central peptide repeat domain. The TGase 1, 2, and 3 enzymes expressed in epithelia use the recombinant SPR2 protein as a complete substrate in vitro, but with widely differing kinetic efficiencies, and in different ways. With TGase 1, only one glutamine on the head domain and one lysine on the tail domain were used for limited interchain cross-linking. With TGase 3, multiple head and tail domain residues were used for extensive interchain cross-linking. The total usage of glutamine and lysine residues in vitro by TGase 3 was similar to that seen in earlier in vivo studies. We conclude that SPR2 proteins are cross-linked in epithelia primarily by the TGase 3 enzyme, a minor extent by TGase 1, and probably not by TGase 2.

The fate of trichohyalin. Sequential post-translational modifications by peptidyl-arginine deiminase and transglutaminases.

Trichohyalin (THH) is a major structural protein of the inner root sheath cells and medulla layer of the hair follicle and, to a lesser extent, of other specialized epithelia. THH is a high molecular weight insoluble alpha-helix-rich protein that forms rigid structures as a result of postsynthetic modifications by two Ca2+-dependent enzymes, transglutaminases (TGases) (protein cross-linking) and peptidyl-arginine deiminase (conversion of arginines to citrullines with loss of organized structure). The modified THH is thought to serve as a keratin intermediate filament matrix protein and/or as a constituent of the cell envelope. In this paper, we have explored in vitro the order of processing of THH to fulfill these functions, using an expressed truncated, more soluble form THH-8. THH-8 is a complete substrate for three known TGases expressed in epithelia, but the kinetic efficiency with TGase 3 is by far the greatest. Following maximal conversion of its arginines to citrullines, THH-8 is cross-linked even more efficiently by TGase 3, using most glutamines partially and all lysines. In addition, we show that insoluble aggregates of THH-8 or native pig tongue THH can be solubilized following peptidyl-arginine deiminase modification. Together, these data suggest an in vivo model in which THH located in insoluble cytoplasmic droplets is first modified by peptidyl-arginine deiminase which denatures it and makes it more soluble. This renders it available for efficient cross-linking by TGase 3 to form highly cross-linked rigid structures in the cells. This temporal order of reaction is supported by the observation that THH is expressed in hair follicle cells before the TGase 3 enzyme.

Profilaggrin is a major epidermal calcium-binding protein.

Profilaggrin is a major highly phosphorylated protein component of the keratohyalin granules of mammalian epidermis. It contains 10 to 12 tandemly repeated filaggrin units and is processed into the intermediate filament-associated protein filaggrin by specific dephosphorylation and proteolysis during terminal differentiation of the epidermal cells. Later, filaggrin itself is degraded to free amino acids that participate in maintenance of epidermal flexibility. The present paper describes the structural organization of the 5' region of the human profilaggrin gene as well as the amino terminus of the profilaggrin protein. The primary profilaggrin transcript consists of three exons and two introns. The first exon (exon I) is only 54 bp and is untranslated. The coding sequences are distributed between exon II (159 bp) and exon III, which contains the information for 10 to 12 filaggrin repeats (972 bp each) and the 3' noncoding sequences. A very large intron separates exons I and II. The combination of a very short exon I with an unusually long intron 1 makes the structure of the profilaggrin gene unique among the epidermally expressed genes investigated so far. Comparison of the expression patterns revealed by primer extension and RNase protection analysis of foreskin epidermal and cultured keratinocyte RNAs suggests that alternately spliced messages, which are different from profilaggrin mRNA, are transcribed from the profilaggrin gene system at earlier stages of epidermal differentiation. The amino terminus of profilaggrin exhibits a significant homology to the small calcium-binding S100-like proteins. It contains two alpha-helical regions, termed EF-hands, that bind calcium in vitro. This is the first example of functional calcium-binding domains fused to a structural protein. We suggest that in addition to its role in filament aggregation and the maintenance of epidermal flexibility, profilaggrin may play an important role in the differentiation of the epidermis by autoregulating its own processing in a calcium-dependent manner or by participating in the transduction of calcium signal in epidermal cells.

The human loricrin gene.

Loricrin is the major protein component of the cornified cell envelope of terminally differentiated mammalian epidermal (stratum corneum) cells. Using a specific human cDNA clone, we have isolated and characterized the human loricrin gene. We show that it has a very simple structure of a single intron of 1188 base pairs (bp) in the 5'-untranslated region; there are no introns in coding sequences. By use of rodent-human somatic cell hybrids, followed by in situ hybridization with a biotin-labeled genomic DNA clone, the single-copy gene maps to chromosome location 1q21. Polymerase chain reaction analyses of genomic DNAs from different individuals show that human loricrin consists of two allelic size variants, due to sequence variations in its second glycine loop domain, and these variants segregate in the human population by normal Mendelian mechanisms. Furthermore, there are multiple sequence variants within these two size class alleles due to various deletions of 12 bp (4 amino acids) in the major loop of this glycine loop domain. By use of a specific loricrin antibody, we show by immunogold electron microscopy that loricrin initially appears in the granular layer of human epidermis and forms composite keratohyalin granules with profilaggrin, but localizes to the cell periphery (cell envelope) of fully differentiated stratum corneum cells.

Structure and organization of the human transglutaminase 1 gene.

Membrane-associated transglutaminases (TGase1) have recently been found to be common in mammalian cells, but it is not clear whether these derive from the same or different genes. In order to determine the complexity of this system, we have isolated and characterized the human gene (TGM1). The gene of 14,133 base pairs was found to contain 15 exons spliced by 14 introns. Interestingly, the positions of these introns have been conserved in comparison with the genes of two other transglutaminase-like activities described in the literature, but the TGM1 gene is by far the smallest characterized to date because its introns are relatively smaller. On the other hand, the TGase1 enzyme is the largest known transglutaminase (about 90 kDa), apparently because its gene acquired tracts that encode additional sequences on its amino and carboxyl termini that confer its unique properties. Southern blot analyses of total human genomic DNA cut with several restriction enzymes reveal only one band. Use of human-rodent cell hybrid panels and chromosomal in situ hybridization with biotin-labeled probes revealed that the human TGM1 gene maps to chromosome position 14q11.2-13. Such data suggest there is a single gene copy per haploid human genome. Comparisons of sequence identities and homologies indicate that the transglutaminase family of genes arose by duplications and subsequent divergent evolution from a common ancestor but later became scattered in the human genome. Although our present Southern blot and chromosomal localization studies revealed no restriction fragment length polymorphisms, comparisons of published sequences and our genomic clone indicate there are two sequence variants for TGase1 within the human population. The rare smaller variant contains a two-nucleotide deletion near the 5'-end, uses an alternate initiation codon, and differs from the common larger variant only in the first 15 amino acids. Furthermore, the DNA sequences of intron 14 possess several tracts of dinucleotide repeats that by polymerase chain reaction analysis show wide size polymorphism within the human population. Accordingly, this gene system constitutes a useful polymorphic marker for genetic linkage analyses.

The complete amino acid sequence of the human transglutaminase K enzyme deduced from the nucleic acid sequences of cDNA clones.

In order to study the expression and role of transglutaminases in the formation of the cross-linked cell envelope of human epidermis, we have used a synthetic oligonucleotide encoding the consensual active site sequence of known transglutaminase sequences. By Northern blot analysis, newborn foreskin epidermis expresses three different mRNA species of about 3.7, 3.3, and 2.9 kilobases while normal cultured epidermal keratinocytes express only the 3.7- and 2.9-kilobase species. The largest species corresponds to a known ubiquitous tissue type II or transglutaminase C activity, the smallest corresponds to a known type I or transglutaminase K activity, and the mid-sized component apparently encodes a transglutaminase E activity that has recently been shown to be expressed in terminally differentiating epidermis (Kim, H. C., Lewis, M. S., Gorman, J. L., Park, S. C., Girard, J. E., Folk, J. E. & Chung, S. I. (1990) J. Biol. Chem., in press). Using the active site oligonucleotide as a probe, we have isolated and sequenced cDNA clones encoding the transglutaminase K enzyme. The deduced complete protein sequence has 813-amino acid residues of 89.3 kDa, has a pl of 5.7, and is likely to be an essentially globular protein, which are properties expected from the partially purified enzyme. It shares 49-53% sequence homology with the other transglutaminases of known sequence, especially in regions carboxyl-terminal to the active site, and possesses sequences likely to confer its Ca2+ dependence. Interestingly, its larger size is due to extended sequences on its amino and carboxyl termini, absent on the other transglutaminases, that may define its unique properties.

Organization, structure, and polymorphisms of the human profilaggrin gene.

Profilaggrin is a major protein component of the keratohyalin granules of mammalian epidermis. It is initially expressed as a large polyprotein precursor and is subsequently proteolytically processed into individual functional filaggrin molecules. We have isolated genomic DNA and cDNA clones encoding the 5'- and 3'-ends of the human gene and mRNA. The data reveal the presence of likely "CAT" and "TATA" sequences, an intron in the 5'-untranslated region, and several potential regulatory sequences. While all repeats are of the same length (972 bp, 324 amino acids), sequences display considerable variation (10-15%) between repeats on the same clone and between different clones. Most variations are attributable to single-base changes, but many also involve changes in charge. Thus, human filaggrin consists of a heterogeneous population of molecules of different sizes, charges, and sequences. However, amino acid sequences encoding the amino and carboxyl termini are more conserved, as are the 5' and 3' DNA sequences flanking the coding portions of the gene. The presence of unique restriction enzyme sites in these conserved flanking sequences has enabled calculations on the size of the full-length gene and the numbers of repeats in it: depending on the source of genomic DNA, the gene contains 10, 11, or 12 filaggrin repeats that segregate in kindred families by normal Mendelian genetic mechanisms. This means that the human profilaggrin gene system is also polymorphic with respect to size due to simple allelic differences between different individuals. The amino- and carboxyl-terminal sequences of profilaggrin contain partial or truncated repeats with unusual un-filaggrin-like sequences on the termini.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of a cDNA clone encoding human filaggrin and localization of the gene to chromosome region 1q21.

Filaggrins are an important class of intermediate filament-associated proteins that interact with keratin intermediate filaments of terminally differentiating mammalian epidermis. They show wide species variations and their aberrant expression has been implicated in a number of keratinizing disorders. We have isolated a cDNA clone encoding human filaggrin and used this to demonstrate that the human gene encodes a polyprotein precursor containing numerous tandem filaggrin repeats. This structure is similar to that of mouse; however, the human filaggrin repeat is much longer (972 base pairs; 324 amino acids) and shows little sequence homology to the mouse protein. Also, data presented here reveal that the human filaggrin repeats show considerable sequence variations; such polymorphism is not found in the mouse. Furthermore, chromosomal mapping data revealed that the human gene is located at 1q21, indicating that the polymorphism is confined to a single locus. By peptide mapping, we define a short linker sequence within the human filaggrin repeat that is excised by proteolysis to yield functional molecules. Finally, we show by in situ hybridization that human filaggrin precursor gene expression is tightly regulated at the transcriptional level in terminally differentiating epidermis and that this represents a useful system in which to study intermediate filament-intermediate filament-associated protein interactions as well as disorders of keratinization.

The complete sequence of the human intermediate filament chain keratin 10. Subdomainal divisions and model for folding of end domain sequences.

We present the complete amino acid sequence of the human keratin 10 (type I) intermediate filament chain expressed in terminally differentiated epidermal cells. Comparisons of this sequence with its mouse and bovine counterparts allow us to describe structural features of the functional end domains. First, sections of their respective end domains are highly conserved and permit a redefinition of earlier models for their subdomainal organization. The amino-terminal end domain consists of El, the first 57-58 residues that are basic, glycine-rich, and have been highly conserved among the three species; V1, a region of well-defined quasi repeats of the motif aliphatic-serine/glycinen; and H1, a newly recognized short acidic sequence that has been conserved among the type I keratin family. The carboxyl-terminal end consists of V2 and E2 whose properties but not sequence resemble V1 and E1, respectively. Second, since the E1, H1, and E2 sequences have been highly conserved between the three species, we suggest they are critical elements in defining intermediate filament function. Third, we note that the E and V sequences of the keratin 10 (and other keratin) chains share many properties in common with protein chain turns found in globular proteins. We therefore propose a model in which these sequences form omega loop-like structures (Leszczynski, J. N. & Rose, G. D. (1986) Science 234, 849-855) on the surface of keratin intermediate filaments. This represents the first specific proposal for the end domain structure of any intermediate filament chain.

The gene for mouse epidermal filaggrin precursor. Its partial characterization, expression, and sequence of a repeating filaggrin unit.

Filaggrin is an important keratin intermediate filament-associated protein of terminally differentiated mammalian epidermis. Its aberrant expression has been implicated in a number of keratinizing disorders. We have isolated and sequenced a cDNA clone to mouse filaggrin, of 1.479 kilobase pairs, which represents less than 10% of the full-length mRNA estimated by Northern blot analysis to be 17 kilobases long. The cDNA clone delineates a 744-base pair repeat. This encodes a protein of 248 amino acids or 26,330 Da, which is almost identical to the known properties of mouse filaggrin in size, amino acid composition, and charge. Total mouse genomic DNA and the filaggrin gene isolated from a cosmid library were found to contain a super-stoichiometric repeat of the same size. These data support the hypothesis (Haydock, P.V., and Dale, B.A. (1986) J. Biol. Chem. 261, 12520-12525) that filaggrin is initially synthesized as a polyprotein precursor containing many tandem copies. However, our data suggest that the repeating filaggrin units of the precursor are not separated by "large linker" peptides as suggested by these authors. In situ hybridization was used to show that the filaggrin precursor mRNA is located precisely over the granular layer of the epidermis, indicating that expression of this gene is regulated at the transcriptional level as for the differentiation-specific keratin 1 protein. These probes will now permit detailed studies on the regulation of expression of the filaggrin gene.

Amino acid sequences of mouse and human epidermal type II keratins of Mr 67,000 provide a systematic basis for the structural and functional diversity of the end domains of keratin intermediate filament subunits.

From the nucleotide sequences of specific cDNA clones, we present partial amino acid sequences (75-90% of the total) of 67-kDa type II keratin subunits expressed in terminally differentiating mouse and human epidermis. Analysis of the sequence information reveals that their secondary structures conform to the pattern common for all intermediate filament (IF) subunits. Together with the previously published sequence of the mouse 59-kDa type I keratin (Steinert, P. M., Rice, R. H., Roop, D. R., Trus, B. L., and Steven, A. C. (1983) Nature 302, 794-800) these data allow us to make comparisons between two keratins which are coexpressed in an epithelial cell type and which coassemble into the same IF. Moreover, these comparisons suggest a systematic plan for the general organization of the end domains of other keratin subunits. We postulate that each end domain consists of a set of subdomains which are distributed with bilateral symmetry with respect to the central alpha-helical domain. Type II (but not type I) keratins contain short globular sequences, H1 and H2, immediately adjacent to the central domain, that have been conserved in size and sequence and which account for most of the difference in mass between coexpressed type II and type I keratins. These are flanked by subdomains V1 and V2 that are highly variable in both length and sequence, often contain tandem peptide repeats, and are conspicuously rich in glycines and/or serines. At the termini are strongly basic subdomains (N and C, respectively) that are variable in sequence. Among keratins of a given type, their variability in mass appears to reside in the size of their V1 and V2 subdomains. However, coexpressed type I and type II keratins have generally similar V1 and/or V2 sequences. By virtue of the ease with which large portions of these subdomain sequences can be removed from intact keratin IF by limited proteolysis, we hypothesize that they lie on the periphery of the IF where they participate in interactions with other constituents of epithelial cells.

Structure of a gene for the human epidermal 67-kDa keratin.

We present the structure and nucleotide sequence of a gene encoding the human epidermal 67-kDa keratin. Three genomic clones were isolated from a lambda Charon 4A human genomic library by hybridization to a specific cDNA probe. One clone of 12.3 kilobase pairs was shown by R-loop, DNA sequence, and primer-extension analyses to encode an entire gene of about 6.25 kilobase pairs. Of eight identified introns, seven are located within the region that encodes the central coiled-coil alpha-helical domain of the protein. Except for one intron located at the end of the region encoding this domain, these do not delineate apparent structural subdomains. The positions of five of the introns exactly coincide with the positions of introns previously reported in the hamster gene for the intermediate filament protein vimentin [Quax, W., Egberts, W.V., Hendricks, W., Quax-Jeuken, Y. & Bloemandal, H. (1983) Cell 35, 215-233]. These findings suggest that the human 67-kDa keratin and vimentin genes arose from a common ancestral gene.

The complete cDNA and deduced amino acid sequence of a type II mouse epidermal keratin of 60,000 Da: analysis of sequence differences between type I and type II keratins.

We present the complete nucleotide and deduced amino acid sequences of a mouse epidermal keratin subunit of 60,000 Da. The keratin possesses a central alpha-helical domain of four tracts (termed 1A, 1B, 2A, and 2B) that can form coiled-coils, interspersed by short linker sequences, and has non-alpha-helical terminal domains. This pattern of secondary structure is emerging as common to all intermediate filament subunits. The alpha-helical sequences conform to the type II class of keratins. Accordingly, this is the first type II keratin for which complete sequence information is available, and thus it facilitates elucidation of the fundamental distinctions between type I and type II keratins. It has been observed that type I keratins are acidic and type II keratins are neutral--basic in charge. We suggest that the basis for this empirical correlation between type and charge resides in the respective net charges of the 1A and 2B tracts. Calculations on interchain interactions between charged residues in the alpha-helical domains indicate that this keratin prefers to participate in dimers according to an in-register parallel arrangement. The terminal domains of this keratin possess characteristic glycine-rich sequences, and the carboxyl-terminal domain is highly homologous to that of a human epidermal keratin of 56,000 Da. According to the hypothesis that end-domains are located on the periphery of keratin filaments, we conclude that the corresponding mouse and human keratins are closely related, both structurally and functionally.

In vitro reconstitution of intermediate filaments form mammalian neurofilament triplet polypeptides.

Intermediate filaments (IF) were reconstituted in vitro from bovine neurofilament triplet polypeptides. Neural IF, solubilized in either low salt or 8 M urea solution, assembled into IF when returned to near-physiological solution conditions. The 68,000-dalton component of the triplet, purified to homogeneity by preparative NaDodSO4 electrophoresis, was renatured and reassembled into short (approximatley 0.05-micrometer) approximatley 10 nm-diameter filaments. These results demonstrate that the triplet polypeptides are components of neural IF and that the 68,000-dalton polypeptide is an IF structural protein.

O-phosphoserine content of intermediate filament subunits.

Purified subunits of intermediate filaments obtained from a variety of tissues and cell types contain O-phosphoserine and, in some cases, smaller amounts of O-phosphothreonine. The O-phosphoserine content was estimated by reaction of performic acid oxidized subunits with methylamine in NaOH. Decamin of BHK-21 and CHO fibroblasts contained about 1 mol/mol. Avian and mammalian desmin consists of two subunits, an acidic (alpha) subunit which contained 2 mol/mol and a more basic (beta) nonphosphorylated subunit. The principal (Mr approximately 60 000) subunit of squid brain neurofilaments contained 5 mol/mol. Most mouse and bovine keratin subunits contained 3--6 mol/mol, although certain bovine subunits of higher molecular weight contained none. The O-phosphoserine contents of keratin subunits purified from the viable and stratum corneum layers were the same. The O-phosphoserine was located in non-alpha-helical regions of the subunits which presumably project out from the alpha-helical wall of the intermediate filaments. Most subunits could be partially dephosphorylated in vitro with alkaline phosphatase. It was found that the capacity of such partially dephosphorylated subunits for assembly into native-type filaments in vitro was independent of their phosphate content.