Pair formation and promiscuity of cytokeratins: formation in vitro of heterotypic complexes and intermediate-sized filaments by homologous and heterologous recombinations of purified polypeptides

Altered keratin expression in buccal mucosal squamous cell carcinoma

Cytokeratin pattern was analyzed in 14 moderately differentiated and 12 well-differentiated squamous cell carcinomas of buccal mucosa by SDS-PAGE, immunoblotting and two dimensional electrophoresis. These were compared with patterns of normal buccal mucosa and surrounding areas whenever possible. Normal buccal mucosa expresses keratin No. 4 (59Kd), 5 (58Kd), 13 (54Kd) and 14 (50Kd). Keratin No. 4 (59Kd) and 14 (50Kd) were expressed by 20 of 26 tumors studied, while many of the tumors did not express keratins No. 5 (58Kd) and 13 (54Kd). Keratin No. 1 (67Kd) and 16 (48Kd) were aberrantly expressed by 9 well-differentiated tumors. Keratin No. 17 (46Kd) and 18 (45Kd) were expressed by 10 and 8 tumors of 14 moderately differentiated tumors. Six tumors which showed involvement of alveolar mucosa, expressed some keratins expressed by its normal counterpart. Their altered expression was consistent with the differentiation pattern as stated earlier. Non-expression of keratins 5 and 13 seems to be the result of malignant transformation and is seen in the majority of tumors, while appearance of aberrant keratins seems to be related more to the degree of differentiation of the tumor.

Posttranslational regulation of keratins: degradation of mouse and human keratins 18 and 8

Human keratin 18 (K18) and keratin 8 (K8) and their mouse homologs, Endo B and Endo A, respectively, are expressed in adult mice primarily in a variety of simple epithelial cell types in which they are normally found in equal amounts within the intermediate filament cytoskeleton. Expression of K18 alone in mouse L cells or NIH 3T3 fibroblasts from either the gene or a cDNA expression vector results in K18 protein which is degraded relatively rapidly without the formation of filaments. A K8 cDNA containing all coding sequences was isolated and expressed in mouse fibroblasts either singly or in combination with K18. Immunoprecipitation of stably transfected L cells revealed that when K8 was expressed alone, it was degraded in a fashion similar to that seen previously for K18. However, expression of K8 in fibroblasts that also expressed K18 resulted in stabilization of both K18 and K8. Immunofluorescent staining revealed typical keratin filament organization in such cells. Thus, expression of a type I and a type II keratin was found to be both necessary and sufficient for formation of keratin filaments within fibroblasts. To determine whether a similar proteolytic system responsible for the degradation of K18 in fibroblasts also exists in simple epithelial cells which normally express a type I and a type II keratin, a mutant, truncated K18 protein missing the carboxy-terminal tail domain and a conserved region of the central, alpha-helical rod domain was expressed in mouse parietal endodermal cells. This resulted in destabilization of endogenous Endo A and Endo B and inhibition of the formation of typical keratin filament structures. Therefore, cells that normally express keratins contain a proteolytic system similar to that found in experimentally manipulated fibroblasts which degrades keratin proteins not found in their normal polymerized state.

Complete structure of the gene for human keratin 18

The complete sequence of the human keratin 18 (K18) gene was determined. The K18 gene is 3791 bp in length and the K18 protein is coded for by seven exons. The exon structure of K18 has been conserved compared to that of other keratin genes, with the exception of a single 3' terminal exon that codes for the tail domain of the protein that is represented by two exons in epidermal keratins. The K18 gene contains an unusual AG/GC donor splice site of intron 3 instead of the consensus AG/GT sequence. This variation is not seen in any other intermediate filament genes. The promoter region of the gene contains a TATA box, six potential SP1 binding sites, and 10 copies of CACCC boxes but lacks any CCAAT boxes and is surprisingly different from the immediately 5' flanking region of the homologous mouse Endo B gene. However, both genes contain small CpG islands surrounding the 5' end of exon 1 and, in addition, conserve repetitive Alu potential transcription units approximately 300 nt upstream of the transcriptional start site.

Expression of mutant keratin cDNAs in epithelial cells reveals possible mechanisms for initiation and assembly of intermediate filaments

We have deleted cDNA sequences encoding portions of the amino- and carboxy-terminal end of a human type I epidermal keratin K14, and examined the molecular consequences of forcing the expression of these mutants in simple epithelial and squamous cell carcinoma lines. To follow the expression of our mutant products in transfected cells, we have tagged the 3' end of the K14 coding sequence with a sequence encoding an antigenic domain of the neuropeptide substance P. Using DNA transfection and immunohistochemistry (with an antibody against substance P), we have defined the limits of K14 sequence necessary to incorporate into a keratin filament network in vivo without disrupting its architecture. We have also uncovered major differences in the behavior of carboxy- and amino-terminal alpha-helical mutants which do perturb the cytoskeletal network of IFs: whereas carboxy terminal mutants give rise to aggregates of keratin in the cytoplasm, amino-terminal mutants tend to produce aggregates of keratins which seem to localize at the nuclear surface. An examination of the phenotypes generated by the carboxy and amino-terminal mutants and the behavior of cells at late times after transfection suggests a model whereby initiation of filament assembly occurs at discrete sites on the nuclear envelope and filaments grow from the nucleus toward the cytoplasm.

Expression of hair-related keratins in a soft epithelium: subpopulations of human and mouse dorsal tongue keratinocytes express keratin markers for hair-, skin- and esophageal-types of differentiation

The dorsal surfaces of mammalian tongues are covered with numerous projections known as filiform papillae whose morphology varies in different species. Using a panel of monoclonal antibodies to keratins as probes, we have established that, in both human and mouse, the interpapillary epithelia express mainly the "esophageal-type" keratins, while the papillary epithelia express "skin-type" keratins as well as some keratins reacting with a monoclonal antibody (AE13) to hair keratins. The AE13-reactive proteins of the mouse were found to be very similar to those of authentic mouse hair keratins. However, the corresponding protein of human tongue appears to be different from all known human keratins. This protein has a MW of 51K; it is relatively acidic; it is sulfhydryl-rich, as revealed by iodoacetic acid-induced charge and apparent size shift; it shares an epitope with all the known acidic human hair keratins; and it is associated with keratin fibrils in vivo. This protein may therefore be regarded as a novel type I "hard" keratin. These data establish that mammalian dorsal tongue epithelia can be divided into at least three compartments that undergo mainly "esophageal-", "skin-" and "hair"-types of differentiation. Different keratin filaments, e.g., those of the esophageal- and hair-types, exhibit strikingly different degrees of lateral aggregation, which can potentially account for the different physical strength and rigidity of various cellular compartments. Our data also suggest the possibility that variations in papillary structure in human and mouse may arise from different spatial arrangements of specific keratinocytes, and/or from the expression of specialized hair-related keratins.

Tissue-specific and differentiation-specific expression of a human K14 keratin gene in transgenic mice

A construct containing approximately 2500 base pairs (bp) of 5' upstream and approximately 700 bp of 3' downstream sequence was used to drive the expression of an intronless human K14 gene in vitro and in vivo. To track the expression of the gene, a small sequence encoding the antigenic portion of neuropeptide substance P was inserted in frame 5' to the TGA translation stop codon of the gene. Surprisingly, this gene was expressed promiscuously in a wide variety of cultured cells transiently transfected with the construct. In contrast, when introduced into the germ line of transgenic mice, the construct was expressed in a fashion analogous to the endogenous K14 gene--namely, in the basal layer of stratified squamous epithelia. Our results suggest that some regulatory mechanism is overridden as a consequence of transient transfection but that sequences that can control proper K14 expression are present in the construct. The appropriate tissue-specific and differentiation-specific expression of K14.P in transgenic mice is an important first step in characterizing a promoter that could be employed to drive the foreign expression of drug-related genes in the epidermis of skin grafts.

Density-dependent modulation of synthesis of keratins 1 and 10 in the human keratinocyte line HACAT and in ras-transfected tumorigenic clones

The spontaneous human keratinocyte line HaCaT and c-Ha-ras oncogene-transfected cell clones are capable of expressing an unusually broad spectrum of keratins, not observed so far in epithelial cells. This expression is, however, strongly modulated by environmental conditions, including cell density. Both cells of the nontumorigenic HaCaT line and the tumorigenic HaCaT-ras clones, I-7 and II-3 (giving rise to benign and malignant tumors, respectively), constitutively expressed the keratins K5, K6, K14, K16 and K17, which are also common in cultures of normal keratinocytes. In addition keratins K7, K8, K18 and K19, generally associated with simple epithelia, were synthesized (to a most pronounced extent in sparse cultures), while keratins K4, K13 and K15 appeared at confluence, presumably with the onset of stratification. Moreover, in both HaCaT and HaCaT-ras clones the epidermal "suprabasal" keratins, K1 and K10, were expressed in conventional submerged cultures (at normal vitamin A levels), markedly rising with cell density, but not strictly correlated with the degree of stratification. This property was maintained in HaCaT cells up to the highest passages. According to immunofluorescence, this was due to increasing numbers of strongly stained cells, and not due to a gradual increase in all cells. Most strikingly, there was a significant delay in the appearance of K10 compared to K1, and this dissociation of expression was most evident in dispase-detached cell sheets (submerged cultures) and organotypic cultures of the ras clones (grown at the air-liquid interface). While on frozen sections bright staining for K1 was seen in some basal and virtually all suprabasal cell layers, K10 was largely restricted to the uppermost layers. Thus, obviously synthesis of K1 and K10 can be regulated independently, although generally in this given sequence. The apparent compatibility of K1 synthesis with proliferation and particularly the extended delay of K10 expression (as a postmitotic event) might be causally related to altered growth control and as such imply the significance of this disturbance. Finally, the highly preserved epidermal characteristics, in terms of expression of keratins (and other differentiation markers [5]) and their regulation, makes these cell lines excellent candidates for studying external modulators of differentiation and also underlying molecular mechanisms.

Identification of an orthologous mammalian cytokeratin gene. High degree of intron sequence conservation during evolution of human cytokeratin 10

Among the human acidic (type I) cytokeratins, components 10 and 11 are especially interesting, as they are under various kinds of expression control. They are synthesized in the suprabasal cell layers of certain stratified epithelia, notably epidermis, in an endogenous differentiation program; they are expressed in certain epithelial tumours but not in others; they can appear de novo in certain pathological situations such as in squamous metaplasias; and their expression in vivo and in vitro is under positive influence of extracellular calcium concentrations and is reduced in the presence of vitamin A or other retinoids. To provide a basis for studies of the various regulatory elements, we have isolated the human gene encoding cytokeratin 10, using a cDNA probe derived from the corresponding bovine gene, and have sequenced the mRNA coding region as well as adjacent regions approximately 1500 bases 5' upstream and 1000 bases 3' downstream. The eight exons encode a polypeptide 59,535 Mr, i.e. somewhat larger than the corresponding bovine and murine proteins. The deduced amino acid sequences display a high degree of homology, which is not restricted to the exons and the 5' and 3' adjacent regions but, surprisingly, is also evident in the seven introns, some of which contain extended sequence elements with 70% identical nucleotides and more, i.e. similar to the homology in the adjacent exons. This exceptionally high level of conservation of intron sequences is discussed in relation to the recently accumulating evidence of the occurrence of intron sequences important in the regulation of the expression of members of other multigene families during development.

Chromosomal mapping of human keratin genes: evidence of non-linkage

We have determined the chromosomal location of the genes for the human keratin intermediate filament proteins K1 (type II; 67 kDa) and K10 (type I; 57 kDa) by the use of specific cDNA clones in conjunction with somatic cell hybrid analysis and in situ hybridization. The K1 keratin gene maps to chromosome region 12q11----q13; the K10 keratin gene maps to chromosome region 17q12----q21. Each gene has been mapped relative to other genes known to be localized on chromosomes 12 and 17, respectively. In somatic cell hybrid analysis, the K1 gene segregates concordantly with the Hox-3 homeo box gene cluster at chromosome region 12p12----q13. The K10 gene localizes to a region proximal to a breakpoint at 17q21 which is involved in a t(17;21)(q21;q22) translocation associated with an acute leukemia. K10 appears to be distal (telomeric) to the gene loci for G-CSF, erb-A, and Her-2, which map to chromosome region 17q12----q21. The NGFR gene and Hox-2 homeo box locus are localized distal to the 17q21 break point and thus distal to the K10 gene. These data demonstrate that keratin genes K1 and K10, which are coexpressed in terminally differentiated epidermis, are not linked in the human genome, implying the existence of trans-acting factors involved in the regulation of expression of these genes.

Isolation, sequence, and differential expression of a human K7 gene in simple epithelial cells

Simple epithelial cells synthesize a different set of keratins than epidermal cells. In experiments reported in this manuscript, we show that the base level of keratin expression in simple epithelial cells is variable for different cell types, and that, in some simple epithelia, this level can be upregulated by increasing the exposure of cells to retinoids, but not glucocorticoids or estradiol. To elucidate the molecular mechanisms underlying simple epithelial keratin gene regulation, we have isolated and characterized a human gene encoding the simple epithelial keratin K7. By examining the possible regulatory elements of this gene and by investigating the behavior of this gene introduced transiently into simple epithelial cells, we have uncovered a possible basis for the differential expression of epidermal and simple epithelial keratin genes.

Transient coexpression of desmin and cytokeratins 8 and 18 in developing myocardial cells of some vertebrate species

During myogenesis the intermediate-sized filament (IF) cytoskeleton is characterized by increasing proportions of desmin. While skeletal and smooth muscle formation occurs in free mesenchymal cells containing vimentin-type IFs, myocardial development starts from a polar epithelium containing cytokeratin IFs and desmosomes. Therefore, we have studied the formation of the epicardium and the myocardium in different vertebrate species, combining light and electron microscopic immunolocalization techniques with gel-electrophoretic analyses of cytoskeletal proteins of microdissected myocardial tissue at differing developmental stages. In this report, we describe results obtained from advanced stages of myocardial differentiation. In all species studied the myocardial cell possess IFs abundant in desmin, often together with smaller amounts of vimentin, and the mesothelial layer of the epicardium contains cytokeratin IFs. However, we have observed remarkable interspecies differences with respect to the occurrence of cytokeratins in embryonic myocardial cells. In fetal human myocardium, from week 10 of pregnancy on, but not in juvenile and adult myocardium, and in chicken myocardium of all stages examined (until several days after hatching) specific immunostaining was seen with certain broad-range cytokeratin antibodies as well as with antibodies specific for cytokeratins 18 (in both species) and 8 (showing significant reaction only in human). This cytokeratin immunoreaction, however, did not appear in IFs extending throughout the cytoplasm or at Z-lines, but was localized in punctate arrays representing aggregates of dense material. The aggregates were often enriched at, but not restricted to, the desmosomal plaques of the intercalated discs. These observations were supported by gel-electrophoretic demonstration of small but significant amounts of cytokeratins 18 (in both species) and 8 (detected only in human) in microdissected myocardial tissue. We also observed cytokeratins in smooth muscle cells of some cardiac blood vessels. In contrast, bovine myocardium of advanced fetal age as well as rat and mouse myocardium (from fetal day 12 on) were negative for cytokeratins with all methods, although epicardial cytokeratin IFs were demonstrable. These observations are discussed in relation to myocardial histogenesis and to general problems of cytokeratin gene expression control in epithelial and nonepithelial cells.

Nucleotide sequence of mouse EndoA cytokeratin cDNA reveals polypeptide characteristics of the type-II keratin subfamily

EndoA cytokeratin (EndoA) belongs to a family of intermediate filaments (IFs) and is coordinately expressed with EndoB cytokeratin during early mouse embryogenesis. We have isolated and sequenced a cDNA from a library constructed from mRNA of parietal yolk sac-like cells, PYS-2, which are derived from mouse teratocarcinoma. Sequence analysis reveals that EndoA is composed of 490 amino acids, its Mr is 54,362, and it contains a central alpha-helical coiled-coil structure flanked by non-alpha-helical domains. The amino acid sequence of EndoA is highly homologous with human cytokeratin No. 8 (93%) and with bovine cytokeratin No. 8 (91%) not only in the central domain, but also in its tail portion, which is less conserved among other intermediate filaments. A comparison with the other cytokeratin proteins characterizes this polypeptide as a non-epidermal type of cytokeratin of the basic (type-II) subfamily. The C-terminal sequence of EndoA is identical to that of human and bovine cytokeratin No. 8 and also highly conserved in other intermediate filaments (desmin, vimentin, glial fibrillary acidic protein and EndoB). It suggests that these may be involved in interaction with some cell component(s), or in more general roles to form IFs. The N-terminal head region is rich in Ser residues including possible phosphorylation sites.

Cloning of the human keratin 18 gene and its expression in nonepithelial mouse cells

Human keratin 18 (K18) and the homologous mouse protein, Endo B, are intermediate filament subunits of the type I keratin class. Both are expressed in many simple epithelial cell types including trophoblasts, the first differentiated cell type to appear during mouse embryogenesis. The K18 gene was identified and cloned from among the 15 to 20 similar sequences identified within the human genome. The identity of the cloned gene was confirmed by comparing the sequence of the first two exons to the K18 cDNA sequence and transfecting the gene into various murine cell lines and verifying the encoded protein as K18 by immunoprecipitation and partial peptide mapping. The transfected K18 gene was expressed in mouse HR9 parietal endodermal cells and mouse fibroblasts even though the fibroblasts fail to express endogenous Endo B. S1 nuclease protection analysis indicated that mRNA synthesized from the transfected K18 gene is initiated at the same position as authentic K18 mRNA found in both BeWo trophoblastoma cells and HeLa cells. Pulse-chase experiments indicated that the human K18 protein is stable in murine parietal endodermal cells (HR9) which express EndoA, a complementary mouse type II keratin. Surprisingly, however, K18 was degraded when synthesized in cells which lack a type II keratin. This turnover of K18 may be an important mechanism by which epithelial cells maintain equal molar amounts of both type I and II keratins. In addition, the levels of the endogenous type I Endo B in parietal endodermal cells were compensatingly down regulated in the presence of the K18 protein, while the levels of the endogenous type II Endo A were not affected in any of the transfected cell lines.

Isolation and characterization of a cDNA clone coding for human epidermal keratin K5. Sequence of the carboxyterminal half of this keratin

A cDNA clone encoding human epidermal keratin No. 5 (K5) has been isolated from a lambda gt10 cDNA library on the basis of crosshybridization with cDNAs coding for other basic keratins and differential hybridization with total cDNA probes. This clone contains a 1100-bp insert able to inhibit specifically translation of K5 in a hybrid-arrested translation test. Northern blots show that this insert hybridizes with a 2.4-kb message present in epidermal mRNAs. The 1100-bp sequence reported here corresponds to the 3' half of the K5 message. Comparison of the deduced aminoacid sequence shows that the protein exhibits characteristic features of a basic keratin. The 242 aminoacid sequence reported here extends from the carboxyterminal end up to the last helical portion of the central rod domain.

Cytokeratins in normal and malignant transitional epithelium. Maintenance of expression of urothelial differentiation features in transitional cell carcinomas and bladder carcinoma cell culture lines

The pattern of cytokeratins expressed in normal urothelium has been compared with that of various forms of transitional cell carcinomas (TCCs; 21 cases) and cultured bladder carcinoma cell lines, using immunolocalization and gel electrophoretic techniques. In normal urothelium, all simple-epithelium-type cytokeratins (polypeptides 7, 8, 18, 19) were detected in all cell layers, whereas antibodies to cytokeratins typical for stratified epithelia reacted with certain basal cells only or, in the case of cytokeratin 13, with cells of the basal and intermediate layers. This pattern was essentially maintained in low-grade (G1, G1/2) TCCs but was remarkably modified in G2 TCCs. In G3 TCCs simple-epithelial cytokeratins were predominant whereas the amounts of component 13 were greatly reduced. Squamous metaplasia was accompanied generally by increased or new expression of some stratified-epithelial cytokeratins. The cytokeratin patterns of cell culture lines RT-112 and RT-4 resembled those of G1 and G2 TCCs, whereas cell line T-24 was comparable to G3 carcinomas. The cell line EJ showed a markedly different pattern. The results indicate that, in the cell layers of the urothelium, the synthesis of stratification-related cytokeratins such as component 13 is inversely oriented compared with that in other stratified epithelia where these proteins are suprabasally expressed, that TCCs retain certain intrinsic cytoskeletal features of urothelium, and that different TCCs can be distinguished by their cytokeratin patterns. The potential value of these observations in histopathologic and cytologic diagnoses is discussed.

Lateral cervical (branchial) cyst epithelia express upper digestive tract-type cytokeratins. Polyclonal antibody studies

The epithelial lining of lateral cervical cysts (LCCs) was analyzed for keratin polypeptide composition by means of high resolution gel electrophoresis and immunoblotting using polyclonal rabbit antikeratin antisera of defined specificity. The keratin phenotype expressed in branchial mass epithelia was found to be homologous to the profiles obtained for the squamous epithelium of corresponding palatine tonsils, but was clearly different from related polypeptide complements of both epidermis and simple (columnar) epithelium. The presence of particular keratin members (pairs 5/14 and 4/13) strongly indicates that branchial mass inner lining derives from keratinocytes that are programmed to form a stratified squamous epithelium and reveal, at least biochemically, an upper digestive tract or esophageal type of differentiation. On the basis of these data and the recent finding that a neck lymph node is involved as a target tissue in LCC formation, hypotheses concerning branchial mass histogenesis in general appear to be highly unsettled. We propose an alternative model that may explain the conflicting clinical, anatomic, and morphologic findings associated with LCC disease.

The sequence of the human epidermal 58-kD (#5) type II keratin reveals an absence of 5' upstream sequence conservation between coexpressed epidermal keratins

We report the isolation and sequencing of cDNA and genomic clones encoding the complete sequence of the human 58-kD epidermal keratin (#5). The sequence specifies a protein of 62,471 daltons that contains a central alpha-helical segment capable of forming a coiled-coil structure flanked by regions that are not alpha-helical. A comparison of the primary sequence with the known sequences of other intermediate filament proteins reveals many common motifs. The 58-kD keratin is highly similar to other type II keratins and less similar to type I keratins and other intermediate filament proteins. The 58-kD keratin is regulated by retinoids in several tissues and is one of four keratins abundantly expressed in epidermal keratinocytes, where it may be important in maintaining structural integrity of the integument. A comparison of the keratin 5 sequence with coexpressed keratin 14 reveals an absence of sequence conservation in regulatory regions and suggests that common sequence elements may not be necessary for coordinate expression of type I and type II keratin partners. Interestingly, keratin 5 contains only one region weakly resembling the SV40 enhancer-like sequence found in some other keratins indicating that this sequence motif may not be necessary for regulation or abundant expression of all epidermal keratins.

A group of type I keratin genes on human chromosome 17: characterization and expression

The human type I keratins K16 and K14 are coexpressed in a number of epithelial tissues, including esophagus, tongue, and hair follicles. We determined that two genes encoding K16 and three genes encoding K14 were clustered in two distinct segments of chromosome 17. The genes within each cluster were tightly linked, and large parts of the genome containing these genes have been recently duplicated. The sequences of the two K16 genes showed striking homology not only within the coding sequences, but also within the intron positions and sequences and extending at least 400 base pairs 5' upstream and 850 base pairs 3' downstream from these genes. Despite the strong homologies between these two genes, only one of the genes encoded a protein which assembled into keratin filaments when introduced into simple epithelial cells. While there were no obvious abnormalities in the sequence of the other gene, its promoter seemed to be significantly weaker, and even a hybrid gene with the other gene's promoter gave rise to a much reduced mRNA level after gene transfection. To demonstrate that the functional K16 gene that we identified was in fact responsible for the K16 expressed in human tissues, we made a polyclonal antiserum which recognized our functional K16 gene product in both denatured and filamentous form and which was specific for bona fide human K16.

Concerted gene duplications in the two keratin gene families

Evolutionary trees were derived from the keratin protein sequences using the Phylogeny Analysis Using Parsimony (PAUP) set of programs. Three major unexpected conclusions were derived from the analysis: The smallest keratin protein subunit, K#19 (Moll et al. 1982), is not the most primitive one, but has evolved to fulfill a highly specialized function, presumably to redress the unbalanced synthesis of keratin subunits. Second, the ancestors of keratins expressed in the early embryonic stages, K#8 and K#18, were the first to diverge from the ancestors of all the other keratins. The branches leading to these two keratins are relatively short, indicating a comparatively strong selection against changes in the sequences of these two proteins. Third, the two keratin families show extraordinary parallelism in their patterns of gene duplications. In both families the genes expressed in embryos diverged first, later bursts of gene duplications created the subfamilies expressed in various differentiated cells, and relatively recent gene duplications gave rise to the hair keratin genes and separated the basal cell-specific keratin from those expressed under hyperproliferative conditions. The parallelism of gene duplications in the two keratin gene families implies a mechanism in which duplications in one family influence duplication events in the other family.

Expression of simple epithelial type cytokeratins in stratified epithelia as detected by immunolocalization and hybridization in situ

Multi-layered ("stratified") epithelia differ from one-layered ("simple") polar epithelia by various architectural and functional properties as well as by their cytoskeletal complements, notably a set of cytokeratins characteristic of stratified tissue. The simple epithelial cytokeratins 8 and 18 have so far not been detected in any stratified epithelium. Using specific monoclonal antibodies we have noted, in several but not all samples of stratified epithelia, including esophagus, tongue, exocervix, and vagina, positive immunocytochemical reactions for cytokeratins 8, 18, and 19 which in some regions were selective for the basal cell layer(s) but extended into suprabasal layers in others. In situ hybridization with different probes (riboprobes, synthetic oligonucleotides) for mRNAs of cytokeratin 8 on esophageal epithelium has shown, in extended regions, relatively strong reactivity for cytokeratin 8 mRNA in the basal cell layer. In contrast, probes to cytokeratin 18 have shown much weaker hybridization which, however, was rather evenly spread over basal and suprabasal strata. These results, which emphasize the importance of in situ hybridization in studies of gene expression in complex tissues, show that the genes encoding simple epithelial cytokeratins can be expressed in stratified epithelia. This suggests that continual expression of genes coding for simple epithelial cytokeratins is compatible with the formation of squamous stratified tissues and can occur, at least in basal cell layers, simultaneously with the synthesis of certain stratification-related cytokeratins. We also emphasize differences of expression and immunoreactivity of these cytokeratins between different samples and in different regions of the same stratified epithelium and discuss the results in relation to changes of cytokeratin expression during fetal development of stratified epithelia, in response to environmental factors and during the formation of squamous cell carcinomas.

Patterns of expression of trichocytic and epithelial cytokeratins in mammalian tissues. II. Concomitant and mutually exclusive synthesis of trichocytic and epithelial cytokeratins in diverse human and bovine tissues (hair follicle, nail bed and matrix, lingual papilla, thymic reticulum)

The hair-forming cells (trichocytes) and the mature hair contain four major trichocytic cytokeratins from each of the subfamilies, basic (Hb1-4) and acidic (Ha1-4); these are related - but not identical - to the epithelial cytokeratins. Here we show, by biochemical methods and immunofluorescence microscopy using antibodies specific for either epithelial or trichocyte cytokeratins, that the same set of hair-type cytokeratins, including two newly identified minor components, designated Hax (type I) and Hbx (type II), are also expressed in cells forming nails, in the filiform papillae of the dorsal surface of human and bovine tongue, and, most surprisingly, in some cells of the epithelial reticulum of bovine and human thymus. By double-label immunofluorescence microscopy, we also show that the expression of the two subsets of cytokeratins, i.e., the epithelial and the trichocytic ones, is not necessarily mutually exclusive, but that certain cells of hair follicles, nail matrix and bed, lingual papillae, and the nonlymphoid cell system of the thymus contain both trichocytic and certain epithelial cytokeratins. This indicates that these cells coexpress representatives of both kinds of cytokeratin. Implications of these findings with respect to problems of regulatory control of cytokeratin synthesis in tissue development and differentiation, and the possible functional meaning of the occurrence of trichocytic cytokeratins in such histologically diverse tissues, are discussed.

Cytokeratin intermediate filaments of rat hepatocytes: different cytoskeletal domains and their three-dimensional structure

A new method of visualizing the three-dimensional architecture of the cytokeratin filaments of the intact rat hepatocyte in situ has been achieved. Frozen sections of liver cut 10 micron thick were serially extracted to remove all elements of the cells except the intermediate filaments. Parallel sections were stained with monoclonal antibodies to the two main cytokeratins found in bile duct and liver cells. Immunofluorescent antibody and immunogold electron microscopy techniques were used to identify the proteins morphologically. Several new observations resulted from these studies. The pericanalicular sheath of intermediate filaments was visualized using steropairs as an uninterrupted branching tubular structure composed of cytokeratins located in the cell cortex of adjacent hepatocytes. Intermediate filaments in the cell cortex formed a distinct sheet of matted filaments which enveloped the entire hepatocyte. The cortical intermediate filaments were in continuity with the pericanalicular sheath and the filaments located within the cytoplasm. The intermediate filaments are attached to the centrioles and appeared to tent the nuclear lamina-pore complex at points of contact. Monoclonal antibodies to rat liver intermediate filament cytokeratins (CK49 and CK55) each stained intermediate filaments located in the cell cortex, within the cytoplasm and at the nucleus. By immunogold staining, some of the intermediate filament filaments were shown to contain both cytokeratins. Filaments which did not stain were thought to be either actin at the cell periphery or nuclear lamins around the nucleus. It is concluded that the cytokeratins form a specialized framework for the cell cortex, canaliculus, centrioles and the nucleus of hepatocytes. The filaments run continuously throughout the cytoplasm without terminating.

Molecular characterization and expression of the stratification-related cytokeratins 4 and 15

A number of human cytokeratins are expressed during the development of stratified epithelia from one-layered polar epithelia and continue to be expressed in several adult epithelial tissues. For studies of the regulation of the synthesis of stratification-related cytokeratins in internal tissues, we have prepared cDNA and genomic clones encoding cytokeratin 4, as a representative of the basic (type II) cytokeratin subfamily and cytokeratin 15, as representative of the acidic (type I) subfamily, and determined their nucleotide sequences. The specific expression of mRNAs encoding these two polypeptides in certain stratified tissues and cultured cell lines is demonstrated by Northern blot hybridization. Hybridization in situ with antisense riboprobes and/or synthetic oligonucleotides shows the presence of cytokeratin 15 mRNA in all layers of esophagus, whereas cytokeratin 4 mRNA tends to be suprabasally enriched, although to degrees varying in different regions. We conclude that the expression of the genes encoding these stratification-related cytokeratins starts already in the basal cell layer and does not depend on vertical differentiation and detachment from the basal lamina. Our results also show that simple epithelial and stratification-related cytokeratins can be coexpressed in basal cell layers of certain stratified epithelia such as esophagus. Implications of these findings for epithelial differentiation and the formation of squamous cell carcinomas are discussed.

Cloning of the human keratin 18 gene and its expression in nonepithelial mouse cells

Human keratin 18 (K18) and the homologous mouse protein, Endo B, are intermediate filament subunits of the type I keratin class. Both are expressed in many simple epithelial cell types including trophoblasts, the first differentiated cell type to appear during mouse embryogenesis. The K18 gene was identified and cloned from among the 15 to 20 similar sequences identified within the human genome. The identity of the cloned gene was confirmed by comparing the sequence of the first two exons to the K18 cDNA sequence and transfecting the gene into various murine cell lines and verifying the encoded protein as K18 by immunoprecipitation and partial peptide mapping. The transfected K18 gene was expressed in mouse HR9 parietal endodermal cells and mouse fibroblasts even though the fibroblasts fail to express endogenous Endo B. S1 nuclease protection analysis indicated that mRNA synthesized from the transfected K18 gene is initiated at the same position as authentic K18 mRNA found in both BeWo trophoblastoma cells and HeLa cells. Pulse-chase experiments indicated that the human K18 protein is stable in murine parietal endodermal cells (HR9) which express EndoA, a complementary mouse type II keratin. Surprisingly, however, K18 was degraded when synthesized in cells which lack a type II keratin. This turnover of K18 may be an important mechanism by which epithelial cells maintain equal molar amounts of both type I and II keratins. In addition, the levels of the endogenous type I Endo B in parietal endodermal cells were compensatingly down regulated in the presence of the K18 protein, while the levels of the endogenous type II Endo A were not affected in any of the transfected cell lines.

A group of type I keratin genes on human chromosome 17: characterization and expression

The human type I keratins K16 and K14 are coexpressed in a number of epithelial tissues, including esophagus, tongue, and hair follicles. We determined that two genes encoding K16 and three genes encoding K14 were clustered in two distinct segments of chromosome 17. The genes within each cluster were tightly linked, and large parts of the genome containing these genes have been recently duplicated. The sequences of the two K16 genes showed striking homology not only within the coding sequences, but also within the intron positions and sequences and extending at least 400 base pairs 5' upstream and 850 base pairs 3' downstream from these genes. Despite the strong homologies between these two genes, only one of the genes encoded a protein which assembled into keratin filaments when introduced into simple epithelial cells. While there were no obvious abnormalities in the sequence of the other gene, its promoter seemed to be significantly weaker, and even a hybrid gene with the other gene's promoter gave rise to a much reduced mRNA level after gene transfection. To demonstrate that the functional K16 gene that we identified was in fact responsible for the K16 expressed in human tissues, we made a polyclonal antiserum which recognized our functional K16 gene product in both denatured and filamentous form and which was specific for bona fide human K16.

Sequence analysis of murine cytokeratin endo A (no. 8) cDNA. Evidence for mRNA species initiated upstream of the normal 5' end in PCC4 cells

Keratin 8 (Endo A) is expressed in simple epithelia, together with keratin 18 (Endo B). Filaments formed by this keratin pair are the first cytoskeletal elements induced during mouse embryogenesis. We have isolated Endo A cDNA clones from lambda gt11 libraries prepared with mRNA isolated from PCC4 embryonal carcinoma (EC) cells. Sequencing of three overlapping cDNAs and of a genomic clone allowed us to determine the complete sequence of the Endo A message. Analysis of the protein sequence deduced showed that the Endo A protein presented all the characteristics of intermediate filaments, including an alpha-helical central rod domain and nonhelical N- and C-termini. In the rod domain, the degree of similarity to the other members of the basic keratin family was high. A high degree of homology to keratin 8 of other species was observed, even in the non-helical domains. During these analyses, we found clones extending upstream of the normal 5' end of the mRNA. Sequence comparison between these cDNAs and the 5' upstream region of the Endo A gene suggested that they corresponded to transcripts initiated at an upstream alternative promoter. These observations supported previous results showing the presence of Endo A transcripts initiated upstream of the normal 5' end in mouse morulae and blastocysts.

Sequence of the human 40-kDa keratin reveals an unusual structure with very high sequence identity to the corresponding bovine keratin

The complete amino acid and DNA sequences of the human 40-kDa keratin are reported. The DNA sequence encodes a protein of 44,098 Da, which is unique in that it lacks the terminal non-alpha-helical tail segment found in all other keratins. When the human 40-kDa keratin amino acid sequence is compared to the corresponding bovine keratin, the overall identity is 89%. The coil-forming regions are 89% identical and the head regions are 88% identical. This similarity is also evident in the DNA sequence of the coding region, the 5' upstream sequences, and the 3' noncoding sequences. The high degree of cross-species identity between bovine and human 40-kDa keratins suggests that there is strong evolutionary pressure to conserve the structure of this keratin. This in turn suggests an important and universal role for this intermediate filament subunit in all species.

Maturation of hagfish gland thread cells: composition and characterization of intermediate filament polypeptides

Previous studies with the hagfish, a primitive vertebrate, have shown that the gland thread cells (GTCs) each contain a single thread (approximately 60 cm long in average-sized cells) in the form of a concisely coiled cytoskeletal entity destined for export by holocrine secretion. The thread in relatively immature GTCs consists almost entirely of intermediate filaments (IFs) bundled in parallel alignment with far fewer microtubules (MTs). The three thread polypeptides described earlier (alpha, basic; beta, acidic; gamma, most acidic; each with a Mr of 63-64 kD) are now further evaluated with respect to in vitro assembly, cross-reactivity with IF polypeptides from higher vertebrates, and peptide sequence homology with known IF polypeptides. The overall results mainly suggest that the hagfish polypeptides are keratinlike substances but lamins or a new type of IF is not ruled out. However, cross-reactivity is weak with mammalian keratins; the 8-11-nm filaments formed from mixtures of alpha and gamma in vitro are generally linear rather than the curvilinear structures usually formed by keratin and nonkeratin IFs; and mixtures of alpha and beta tend to yield 9-12-nm granules or granular strings. Polypeptide analyses on GTCs segregated on the basis of maturational stage show a progressive increase in beta/gamma values which correlates with cell maturation, but the alpha/(beta + gamma) ratios remain near 1. Inasmuch as beta and gamma have many similar properties, the documented increase in the amount of the beta component in aging GTCs might in part be the result of a failure in a posttranslational modification system and may contribute to the ultrastructural changes that accompany thread maturation in preparation for holocrine secretion and subsequent modulation of the viscoelastic properties of mucus.

Turnover of cytokeratin polypeptides in mouse hepatocytes

The turnover of cytokeratin polypeptides A (equivalent to No. 8 of the human cytokeratin catalog) and D (equivalent to human cytokeratin No. 18) of mouse hepatocytes was studied by pulse-labeling of mouse liver proteins after intraperitoneal injection of L-[guanido-14C]arginine and [14C]sodium bicarbonate. At various times after injection cytoskeletal proteins were prepared and separated by SDS-polyacrylamide gel electrophoresis, and the specific radioactivities of polypeptides recovered from excised gel slices were determined. With L-[guanido-14C]arginine a rapid increase in the specific radioactivity of both cytokeratins was observed which reached a plateau between 12 and 24 h. With [14C]sodium bicarbonate maximal specific radioactivity was obtained at 6 h followed by a rapid decrease to half maximum values within the subsequent 6 h and then a slower decrease. Half-lives were determined from the decrease of specific radioactivities after pulse-labeling by least-squares plots and found to be 84 h (for cytokeratin component A) and 104 h (component D) for arginine labeling. Values obtained after bicarbonate labeling were similar (95 h for A and 98 h for D). These results show that liver cytokeratins are relatively stable proteins and suggest that components A and D are synthesized and degraded at similar rates, probably in a coordinate way.

Monoclonal cytokeratin antibody recognizing a heterotypic complex: immunological probing of conformational states of cytoskeletal proteins in filaments and in solution

A novel type of monoclonal murine antibody (Ks18.18) directed against an epitope depending on human cytokeratin (CK) 18, a member of the acidic (type I) CK subfamily, is described. We show by SDS-PAGE immunoblots and dot-blot assays that this antibody is unreactive with both the denatured and the renatured individual polypeptides but binds strongly to heterotypic coiled-coil complexes of CK 18 with several members of the complementary basic (type II) CK subfamily, notably with CK 8; i.e., its most frequent natural partner. We also show that specific interactions between complementary CK polypeptides take place during the incubation steps of immunoblotting procedures as polypeptides, or fragments thereof, that detach from the substrate can bind to complementary polypeptides attached to the substratum, which may result in false assignments of antibody reactivities. The conformation-specific, CK 18-dependent epitope of Ks18.18 was detected in intermediate filaments (IFs) of cultured cells, simple epithelia, and many carcinomas and, surprisingly, also in the basal cells of some stratified epithelia. Ks18.18 also reacts with altered CK configurations as present in the spheroidal bodies of mitotic cells and in the Mallory bodies of hepatocytes intoxicated with certain drugs, thus indicating that the heterotypic CK complexes are maintained in these structures. We have also used antibody Ks18.18 to demonstrate the existence of heterotypic CK 8 and 18 complexes in a distinct soluble form among supernatant proteins from cell homogenates which is indistinguishable from the heterotypic tetramer obtained after experimental disintegration of IFs. The potential value of such IF conformation-specific antibodies in cell biological research and pathology is discussed.

Cytokeratin domains involved in heterotypic complex formation determined by in-vitro binding assays

Cytokeratins are constituent proteins of intermediate filaments (IFs) that form heterotypic tetrameric IF subunits containing two polypeptide chains of each of the two cytokeratin subfamilies, i.e. the acidic (type I) and the basic (type II). To locate the molecular domains involved in the formation of these heterotypic complexes, we have developed a binding assay in which total cellular or cytoskeletal polypeptides, or proteolytically prepared cytokeratin fragments, are separated by one-, or two-dimensional gel electrophoresis, blot-transferred on to nitrocellulose paper and probed with radio-iodinated purified cytokeratin polypeptides or fragments thereof, using buffers of various ionic strengths with or without 4 M-urea. Using these polypeptides in the binding assay, specific heterotypic binding was observed between complementary cytokeratin polypeptides of the two subfamilies (but not with other IF proteins) and between the corresponding alpha-helical rod domain fragments. Both rod coils 1 and 2 of the type II cytokeratin 8 bound to the rod (coils 1 and 2) fragment of type I cytokeratins, and this binding occurred at both low and high ionic strengths. The results obtained indicate that: (1) the binding between cytokeratin polypeptides of the complementary type is stronger and more selective than interactions of cytokeratins with other IF and non-IF proteins; (2) both the head and the tail portions of the proteins are not required for heterotypic complex formation; (3) the complementarity information located in the alpha-helical portions of the rod domain, and in short sequences immediately flanking them, is sufficient to discriminate between the two types of cytokeratins and to secure the formation of heterotypic cytokeratin complexes; (4) both coils 1 and 2 of the rod can contribute to this association; and (5) the formation of the heterotypic cytokeratin complex is not critically dependent upon ionic interactions. Our results are further compatible with the concept that the heterotypic binding takes place between cytokeratin homodimer coiled-coils.

The mesonephric (wolffian) and paramesonephric (müllerian) ducts of golden hamsters express different intermediate-filament proteins during development

We analysed the expression of intermediate-filament proteins in the developing mesonephric duct (the precursor of the male genital ducts) and the paramesonephric duct (the precursor of the female genital ducts) of golden-hamster embryos using immunohistochemical methods. Embryos were investigated from the early stages of duct development, i.e. at 9.5 days post conceptionem (dpc), through sexual differentiation, until birth (15.5 dpc). Monospecific antibodies to vimentin or keratins 7, 8, 18 or 19 as well as two keratin antibodies that are pan-epithelial in human tissues were tested. Both ducts expressed vimentin to some degree from their early stages (mesonephric duct from 9.5 dpc onwards; paramesonephric duct from 10.5 dpc onwards) until birth. No keratins were detectable at these earliest stages. In the mesonephric duct, keratins 7, 18 and 19 appeared simultaneously at 10.5 dpc and persisted until birth. In the paramesonephric duct, only keratin 18 was detectable at first (at 12.0 dpc), with the expression of keratins 7 and 19 being delayed until 14.5 dpc. This feature was irrespective of sexual differentiation, which begins at 11.0 dpc, so that, in males, these keratins appeared on cue, even though the paramesonephric duct was regressing at this time. The expression of keratin 8 could not be demonstrated in either duct using the antibodies tested in our study. By 14.5 dpc, the differentiated male mesonephric duct and the differentiated female paramesonephric duct exhibited the same intermediate-filament protein pattern (weak vimentin expression and strong expression of keratins 7, 18 and 19), in spite of differences in the intermediate-filament protein patterns exhibited by the two ducts during early development. These different programmes of intermediate-filament protein regulation do not support the concept that the mesonephric duct makes a cellular contribution to the paramesonephric duct during the development of the latter.

An unusual type-II 70-kilodalton keratin protein of mouse epidermis exhibiting postnatal body-site specificity and sensitivity to hyperproliferation

Keratin extracts from the epidermis of adult mouse ears, footpads, and tail contain large amounts of a 70-kilodalton (kDa) protein which has not been detected in any other body site of the adult mouse or in the epidermis of neonatal mice. Two-dimensional immunoblotting using an antiserum which recognizes both type-I and type-II murine keratins revealed that the 70-kDa protein is indeed a keratin belonging to the type-II subfamily. Its postnatal induction occurs during the first 2 weeks after birth, being first observed in tail epidermis, then in footpad epidermis, and only rather late in ear epidermis. Although in vitro translation experiments with polyA+-RNA from adult tail and footpad epidermis consistently failed to reveal the 70-kDa protein among the translation products, we obtained evidence using a specific cDNA clone that, in vivo, the protein is encoded by a discrete mRNA. This clone, termed pke70, was isolated from a cDNA library of footpad epidermal mRNA. Homology comparisons with a variety of known keratin cDNAs indicated that pke70 contains sequence information for a type-II keratin that is substantially larger than the mouse 67-kDa keratin protein. Northern-blot analysis with a specific 3'-fragment of pke70 demonstrated a single 2.8 +/- 0.1 kb mRNA species exclusively in adult ear, footpad, and tail epidermis. In situ hybridization with the same fragment revealed the presence of the pke70-hybridizing mRNA in both basal and suprabasal cells of ear and footpad epidermis as well as in the orthokeratinizing parts of the tail epidermis; however in the epidermis covering the balls of the feet, labeling was restricted to suprabasal cells at the base of these nodular elevations. Continuous treatment of adult tail or ear epidermis with hyperplasiogenic agents, e.g., vitamin A acid and the tumor promoter, 12-O-tetradecanoylphorbol-13-acetate (TPA), leads to a gradual disappearance of the 70-kDa protein. We obtained evidence using in situ hybridization that the loss of the 70-kDa keratin is preceded by a specific suppression of the transcription of its putative mRNA in basal cells, whereas initially suprabasal cells are apparently still able to complete their original commitment. The particular properties of the 70-kDa keratin protein, i.e., its topological restriction, its postnatal and time-dependent acquisition, and its pronounced sensitivity to hyperplasiogenic stimuli, make this keratin subunit an especially suitable candidate for studies concerning the regulation of keratin expression and morphogenesis in general, as well as for studies of the factors that control its expression so specifically.

The expression of mutant epidermal keratin cDNAs transfected in simple epithelial and squamous cell carcinoma lines

We have deleted cDNA sequences encoding portions of the carboxy-terminal end of a human type I epidermal keratin K14, and examined the molecular consequences of forcing the expression of these mutants in simple epithelial and squamous cell carcinoma lines. To follow the expression of our mutant products in transfected cells, we have tagged the 3' end of the K14 coding sequence with a sequence encoding an antigenic domain of the neuropeptide substance P. Using DNA transfection and immunohistochemistry (with an antibody against substance P), we have identified a collection of mutants that have a wide range of morphological effects on the endogenous keratin filament networks of transfected cells. Mutants that are missing most of the nonhelical carboxy-terminal domain of K14 incorporate into the endogenous keratin filaments without any visible perturbations on the network. In contrast, mutants that are missing as few as 10 of the 310 amino acids of the central alpha-helical domain of the polypeptide cause gross alterations in the keratin network. In some cases, the entire cytoskeletal network of keratins was disrupted, leaving no evidence of 8-nm filaments. These results reveal the existence of a dynamic exchange between newly synthesized subunits and preexisting keratin filaments.

Retinoids as important regulators of terminal differentiation: examining keratin expression in individual epidermal cells at various stages of keratinization

When human epidermal cells were seeded on floating rafts of collagen and fibroblasts, they stratified at the air-liquid interface. The suprabasal cells synthesized the large type II (K1) and type I (K10/K11) keratins characteristic of terminal differentiation in skin. At earlier times in culture, expression of the large type II keratins appeared to precede the expression of their type I partners. At later times, all suprabasal cells expressed both types, suggesting that the accumulation of a critical level of K1 keratin may be a necessary stimulus for K10 and K11 expression. Expression of the terminal differentiation-specific keratins was completely suppressed by adding retinoic acid to the culture medium, or by submerging the cultures in normal medium. In submerged cultures, removal of vitamin A by delipidization of the serum restored the keratinization process. In contrast, calcium and transforming growth factor-beta did not influence the expression of the large keratins in keratinocytes grown in the presence of retinoids, even though they are known to induce certain morphological features of terminal differentiation. Retinoic acid in the raft medium not only suppressed the expression of the large keratins, but, in addition, induced the synthesis of two new keratins not normally expressed in epidermis in vivo. Immunofluorescence localized one of these keratins, K19, to a few isolated cells of the stratifying culture. In contrast, the other keratin, K13, appeared uniformly in a few outer layers of the culture. Interestingly, K13 expression correlated well with the gradient of retinoid-mediated disruptions of intercellular interactions in the culture. These data suggest that K13 induction may in some way relate to the reduction in either the number or the strength of desmosomal contacts between suprabasal cells of stratified squamous epithelial tissues.

Suprabasal expression of a 64-kilodalton keratin (no. 3) in developing human corneal epithelium

We have previously shown that a basic 64-kilodalton (no. 3 in the catalog of Moll et al.) and an acidic 55-kilodalton (no. 12) keratin are characteristic of suprabasal cell layers in cultured rabbit corneal epithelial colonies, and therefore may be regarded as markers for an advanced stage of corneal epithelial differentiation. Moreover, using an AE5 mouse monoclonal antibody, we showed that the 64-kilodalton keratin marker is expressed suprabasally in limbal epithelium but uniformly (basal layer included) in central corneal epithelium, suggesting that corneal basal cells are in a more differentiated state than limbal basal cells. In conjunction with previous data implicating the centripetal migration of corneal epithelial cells, our data support a model of corneal epithelial maturation in which corneal epithelial stem cells are located in the limbus, the transitional zone between the cornea and conjunctiva. In the present study, we analyzed the expression of the 64-kilodalton keratin in developing human corneal epithelium by immunohistochemical staining. At 8 weeks of gestation, the presumptive corneal epithelium is composed of a single layer of cuboidal cells with an overlying periderm; neither of these cell layers is AE5 positive. At 12-13 weeks of gestation, some superficial cells of the three- to four-layered epithelium become AE5 positive, providing the earliest sign of overt corneal epithelial differentiation. At 36 weeks, although the epithelium is morphologically mature (four to six layers), AE5 produces a suprabasal staining pattern, this being in contrast to the adult epithelium which exhibits uniform staining.(ABSTRACT TRUNCATED AT 250 WORDS)

The transfection of epidermal keratin genes into fibroblasts and simple epithelial cells: evidence for inducing a type I keratin by a type II gene

Through gene transfection studies, we have discovered that the forced expression of a foreign type II epidermal keratin in fibroblasts can trigger the expression of an endogenous type I epidermal keratin. Both the transfected and the induced proteins participate in the formation of filamentous structures. Interestingly, this regulation appears to be unidirectional: the expression of a transfected type I keratin does not induce type II expression. Rather, nonfilamentous aggregates of type I protein accumulate in the cytoplasm. In contrast, simple epithelial cells transfected with either a type I or a type II epidermal keratin gene do not respond by inducing the expression of a host epidermal keratin. In this case, the foreign protein is incorporated into the endogenous keratin network. These results suggest the possibility that type I keratin expression may be dependent on the accumulation of unpolymerized type II keratin.

Polymorphic keratins in human epidermis

Human epidermal keratins from many different individuals were identified and compared by both high-resolution 1- and 2-dimensional gel electrophoresis and immunoblotting. While the polypeptide patterns obtained for keratin-enriched cytoskeletal preparations could be considered typical of normal interfollicular epidermis, they also disclosed variations, among the individuals, concerning some of the constituent protein subunits. Three sets of interindividually varying keratins could be distinguished owing to their distinct, though small, differences in electrophoretic mobility on sodium dodecyl sulfate-polyacrylamide gels and their similar or identical charge characteristics upon nonequilibrium pH gradient electrophoresis: the basic keratins 1a and 1b as well as 5a and 5b and the acidic keratins 10a and 10b. Of each set either a doublet, showing a marked 1:1 ratio of polypeptides, or the one or the other variant protein was detected together with keratin 14, which did not display any variation in a series of 148 individual tissue samples tested. Thus, the keratin composition of human epidermis could be summarized in the formula: (1a v 1b) + (5a v 5b) + (10a v 10b) + 14. The systematic appearance of the variants suggested that each protein within a set is the product of an independent allele. In support of this hypothesis we have found that the same variant is expressed in other epithelia of a given individual. Moreover, the frequency of any of the keratins in our sampling concurred with the frequency predicted by the Hardy-Weinberg relation for the distribution of alleles in a population, as did the frequency distribution of particular keratin patterns.

Sequence of a cDNA encoding human keratin No 10 selected according to structural homologies of keratins and their tissue-specific expression

We present here the nucleotide sequence of a 1700 bp-long cDNA encoding human epidermal keratin No. 10 (56.5 kDa). cDNA clones of the acidic keratin family were first isolated from a pBR322 human epidermal cDNA library by hybridization with a probe coding for keratin No. 14. Differential hybridization using total cDNA probes prepared from poly(A)+ RNA extracted either from epidermis (which contains keratin No. 10) and from squamous carcinoma or hepatoma cell lines (which do not express keratin No. 10) made possible the selection of clones potentially coding for keratin No. 10. The 1.7 kb sequence exhibits the characteristic features of an acidic keratin with a constant central rod domain and C-terminal variable structures. Moreover, the sequence shows extensive homologies with the cDNA of murine keratin No. 10.

Identification of the conserved, conformation-dependent cytokeratin epitope recognized by monoclonal antibody (lu-5)

The epitope recognized by the murine monoclonal antibody (mAB lu-5) recently described as a formaldehyde-resistant, "pan-epithelial marker" of great value in tumour diagnosis is located on the surface of cytokeratin filaments. It has been preserved during vertebrate evolution from amphibia to man. As this epitope is not reactive after SDS-polyacrylamide gel electrophoresis (SDS-PAGE), the epitope-bearing protein has been identified by a dot-blot antibody binding assay, using purified proteins in which the epitope is reconstituted. We show that the epitope is present in most cytokeratin polypeptides of both the acidic (type I) and basic (type II) subfamily but does not occur in other cytoskeletal proteins. The location of this widespread epitope is discussed with respect to homologies of amino acid sequences of cytokeratins and their conformations.

Cytokeratin expression in simple epithelia. III. Detection of mRNAs encoding human cytokeratins nos. 8 and 18 in normal and tumor cells by hybridization with cDNA sequences in vitro and in situ

We describe cDNA clones of mRNAs encoding human cytokeratins nos. 8 and 18, and the amino acid sequences deduced from their nucleotide sequences. Human cytokeratin no. 8 is a typical cytokeratin of the basic (type II) subfamily, which is highly homologous to the corresponding bovine and amphibian (Xenopus laevis) proteins; however, unlike the amphibian protein, it does not contain glycine-rich oligopeptide repeats in its carboxyterminal 'tail' domain. Comparison with the reported amino acid sequences of two fragments of human 'tissue polypeptide antigen' (TPA), a widely used serodiagnostic carcinoma marker, revealed sequence identity, indicating that this serum component is derived from the intracellular cytokeratin no. 8 present in diverse kinds of epithelia and epithelium-derived tumors. Human cytokeratin no. 18 is very similar to the corresponding murine protein but contains two additional blocks of 4 and 5 amino acids in the 'head' portion. These cDNA clones and the RNA probes derived therefrom were used to detect specifically mRNAs by Northern-blot assays of RNAs from various carcinomas and cultured carcinoma cells. Using in situ hybridization on frozen sections of tumor-containing tissues, notably lymph nodes containing metastatic breast carcinoma, we were able to demonstrate the specificity and sensitivity of this procedure. The potential value for cell-biological research and pathology of being able to detect a mRNA encoding a given cytokeratin polypeptide in situ is discussed.