Monoclonal antibody studies of mammalian epithelial keratins: a review

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.

Schistosomiasis-associated squamous lesions of the bladder. Expression of low molecular weight cytokeratin proteins

The monoclonal antibody CAM 5.2 was used to investigate the expression of lower molecular weight cytokeratin proteins immunocytochemically in conventionally processed samples of schistosoma-associated squamous metaplasias and carcinomas of the urinary bladder. The antibody did not differentiate between squamous metaplasias and well differentiated squamous carcinomas, though it did focally label moderate to poorly differentiated tumours. The results suggest that squamous carcinomas of the bladder in this setting differ fundamentally from such tumours at other sites in their patterns of cytokeratin protein expression.

Structural distinctions among human breast epithelial cells revealed by the monclonal antikeratin antibodies AE1 and AE3

Two monoclonal antikeratin antibodies, AE1 and AE3, were used in indirect immunocytochemistry to examine keratin expression in normal, benign proliferative, and malignant human breast epithelium. Both antibodies reacted strongly with most luminal cells in ducts and acini of normal gland. While AE1 did not stain myoepithelium, AE3 recognized myoepithelial cells of ducts but not acini, implying a cytoskeletal difference between the myoepithelium of these two components. Moreover, the antibodies reacted differently with the myoepithelium of intracanalicular as compared with pericanalicular types of fibroadenomas. Tumour cells of infiltrating ductal carcinomas with a prominent intraductal component stained more homogeneously with AE1 and AE3 than those without intraductal growth. The results provide evidence for two phenotypes of myoepithelial cells and for the presence of cryptic keratin epitopes in human breast epithelial cells. The finding that neither AE1 nor AE3 is a universal detector of these cells has important clinical and experimental implications.

Cytokeratin polypeptides in normal and metaplastic human salivary gland epithelia

Immunofluorescent and immunoperoxidase labelling of normal and metaplastic human submandibular salivary glands with a battery of cytokeratin-specific monoclonal antibodies was carried out. Labelling with a broad spectrum cytokeratin antibody (KG 8.13), as well as with antibody to cytokeratin polypeptide No. 18 (Ks 18.18) was observed in all the epithelial elements of the gland. Polypeptide No. 19, however, was present in ductal cells only, sparing the acini and the associated myoepithelium. Antibody KS 8.58, specific for cytokeratins Nos. 13 and 16, selectively labelled basal cells along the large ducts. Examination of squamous metaplasia associated with chronic suppurative sialadenitis indicated that the metaplastic cells display the same cytokeratin profile as the normal ductal cells and are not labelled with antibodies KS 8.58 and KK 8.60 which usually stain normal and pathological stratified epithelia. The significance of these observations for the histogenesis of normal salivary glands, as well as for the development of the metaplastic process, is discussed.

Differentiation-related changes of cytokeratin expression in cultured keratinocytes and in fetal, newborn, and adult epidermis

Cytokeratin expression in differentiating cultured foreskin keratinocytes was studied using chain-specific anti-cytokeratin monoclonal antibodies directed against cytokeratins 4, 8, 10, 13, 18, and 19, respectively. Keratinocytes were cultured at low Ca2+ concentration (0.06 mM) to repress differentiation. At confluency, the cells were switched to high Ca2+ concentration (1.6 mM) to induce differentiation. Cells were harvested 0, 3, 8, 16, 24, 48, and 72 h after the switch. Keratinocytes cultured throughout at high Ca2+ concentration were also harvested. Immunoblots of cytokeratin preparations isolated from these cultures showed that cytokeratins 4, 13, and 19 were not present in nondifferentiating keratinocytes but could be detected from about 16 h after the Ca2+ switch. Immunohistochemical studies were performed on frozen sections of cell sheets incubated with anti-cytokeratin and anti-vimentin. Expression of cytokeratins 4, 13, and 19 was seen in superficial cells. Cytokeratin 10 was locally present in suprabasal and superficial cells. Vimentin was present in 40-70% of the basal cells and in only a few differentiating keratinocytes. Expression of cytokeratins 8 and 18 could not be detected. The same antibodies were also used to stain sections from fetal (15, 20, and 29 weeks), newborn (40 weeks), and mature (5 and 75 years) epidermis. In the 15-week-old epidermis, basal cells were positive for cytokeratins 8 and 19 and locally for cytokeratin 4; intermediate cells expressed cytokeratins 4, 10, 13, and 19; and the periderm contained cytokeratins 4, 8, 13, 18, and 19. In the 20-week-old epidermis, cytokeratin 4 had disappeared from the basal cell layer and cytokeratin 19 was present only locally; in the intermediate cell layer, cytokeratins 4 and 19 had disappeared; and in the periderm, the expression of the cytokeratins studied was the same as that in the 15-week-old epidermis. The basal cells of the 29-week-old fetal epidermis, the newborn epidermis, and the mature epidermis are negative with all antibodies tested, except for some scattered cells in the fetal and newborn skin, presumably Merkel cells, that were positive for cytokeratins 8, 18, and 19. Suprabasal cells in all specimens were positive only for cytokeratin 10. With respect to the cytokeratins studied, our results show that cultured differentiating keratinocytes resemble the suprabasal cells of early fetal epidermis. Basal cells of cultured keratinocytes resemble the basal cells of late fetal, newborn, and adult epidermis and therefore support previous observations.

Potential applications of anti-keratin antibodies in oral diagnosis

Recent progress in understanding the biology of keratins together with the development of monoclonal antibodies to individual keratin proteins provide the foundation for studying keratin expression in normal and pathological oral epithelia. Areas of oral pathology in which the examination of epithelial keratin profiles may yield information of potential diagnostic value are discussed. Examples of altered keratin expression in epithelial dysplasia, oral cancer and odontogenic cysts and tumours are presented. Immunocytochemical demonstration of individual keratins can clarify the composition of complex epithelia and may help to establish epithelial lineage or indicate progression changes in tumours. Some problems in the application and interpretation of keratin immunocytochemistry are also considered.

Cytokeratin polypeptides expression in different epithelial elements of human salivary glands

Immunofluorescent labeling of human salivary glands was carried out with a battery of monoclonal antibodies reactive with specific cytokeratin polypeptides. All the epithelial elements of the glands were positively labelled by a broad-spectrum cytokeratin antibody (KG 8.13) and by antibody Ks 18.18, which reacts with cytokeratin No. 18 exclusively. Labelling of frozen sections with antibody KM 4.62, which is reactive with the 40 Kd (No. 19) cytokeratin, was confined to the ductal system and apparently absent from the acini. Antibody KA-1, reactive with polypeptides 4, 5 and 6 stained both the myoepithelial cells and the basal cells of the large ducts. Antibody KS 8.58, however, reacted with the basal cells exclusively. It is thus proposed that the combined use of the various monoclonal antibodies may provide a most useful probe in studies on epithelial cell diversity in normal salivary glands as well as in pathological disorders of that gland.

Expression of monoclonal antibody-defined epitopes of keratin 19 in human tumours and cultured cells

The monoclonal antibodies BA16 and BA17, reacting specifically with human keratin 19 (40 kD) have been tested by immunohistochemical staining methods for their reaction with a wide range of human tumours and cultured cells. Primary adenocarcinomas and their metastases showed a homogeneously positive reaction with greater than 95% of the tumour cells staining. Non-epithelial tumours, basaliomas and squamous cell carcinomas were unstained, while benign breast lesions and a thyroid adenoma show a mosaic pattern of stained and unstained (5-40%) cells. These three staining patterns were also seen in cultured cells. Positive homogeneous staining was seen in all breast cancer cell lines examined with the exception of PMC42, which exhibits stem cell characteristics, and which showed the heterogeneous pattern of staining seen in milk cell cultures. Non-epithelial lines and strains, two cell lines from cervical carcinomas and three SV40 transformed breast epithelial lines were unstained. The antibodies BA16 and 17 are potentially useful reagents for distinguishing adenocarcinomas (and their metastases) from non-epithelial tumours and from squamous carcinomas. They may also discriminate between benign and malignant breast lesions, and identify a specific differentiation phenotype in the secretory cell lineage.

Acidic and basic hair/nail ("hard") keratins: their colocalization in upper cortical and cuticle cells of the human hair follicle and their relationship to "soft" keratins

Although numerous hair proteins have been studied biochemically and many have been sequenced, relatively little is known about their in situ distribution and differential expression in the hair follicle. To study this problem, we have prepared several mouse monoclonal antibodies that recognize different classes of human hair proteins. Our AE14 antibody recognizes a group of 10-25K hair proteins which most likely corresponds to the high sulfur proteins, our AE12 and AE13 antibodies define a doublet of 44K/46K proteins which are relatively acidic and correspond to the type I low sulfur keratins, and our previously described AE3 antibody recognizes a triplet of 56K/59K/60K proteins which are relatively basic and correspond to the type II low sulfur keratins. Using these and other immunological probes, we demonstrate the following. The acidic 44K/46K and basic 56-60K hair keratins appear coordinately in upper corticle and cuticle cells. The 10-25K, AE14-reactive antigens are expressed only later in more matured corticle cells that are in the upper elongation zone, but these antigens are absent from cuticle cells. The 10-nm filaments of the inner root sheath cells fail to react with any of our monoclonal antibodies and are therefore immunologically distinguishable from the cortex and cuticle filaments. Nail plate contains 10-20% soft keratins in addition to large amounts of hair keratins; these soft keratins have been identified as the 50K/58K and 48K/56K keratin pairs. Taken together, these results suggest that the precursor cells of hair cortex and nail plate share a major pathway of epithelial differentiation, and that the acidic 44K/46K and basic 56-60K hard keratins represent a co-expressed keratin pair which can serve as a marker for hair/nail-type epithelial differentiation.

Differentiation-related expression of a major 64K corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells

In this paper we present keratin expression data that lend strong support to a model of corneal epithelial maturation in which the stem cells are located in the limbus, the transitional zone between cornea and conjunctiva. Using a new monoclonal antibody, AE5, which is highly specific for a 64,000-mol-wt corneal keratin, designated RK3, we demonstrate that this keratin is localized in all cell layers of rabbit corneal epithelium, but only in the suprabasal layers of the limbal epithelium. Analysis of cultured corneal keratinocytes showed that they express sequentially three major keratin pairs. Early cultures consisting of a monolayer of "basal" cells express mainly the 50/58K keratins, exponentially growing cells synthesize additional 48/56K keratins, and postconfluent, heavily stratified cultures begin to express the 55/64K corneal keratins. Cell separation experiments showed that basal cells isolated from postconfluent cultures contain predominantly the 50/58K pair, whereas suprabasal cells contain additional 55/64K and 48/56K pairs. Basal cells of the older, postconfluent cultures, however, can become AE5 positive, indicating that suprabasal location is not a prerequisite for the expression of the 64K keratin. Taken together, these results suggest that the acidic 55K and basic 64K keratins represent markers for an advanced stage of corneal epithelial differentiation. The fact that epithelial basal cells of central cornea but not those of the limbus possess the 64K keratin therefore indicates that corneal basal cells are in a more differentiated state than limbal basal cells. These findings, coupled with the known centripetal migration of corneal epithelial cells, strongly suggest that corneal epithelial stem cells are located in the limbus, and that corneal basal cells correspond to "transient amplifying cells" in the scheme of "stem cells----transient amplifying cells----terminally differentiated cells."