HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Langbein, L. Articles by Schweizer, J. Search for Related Content PubMed PubMed Citation Articles by Langbein, L. Articles by Schweizer, J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 28, 19874-19884, July 9, 1999

The Catalog of Human Hair Keratins
I. EXPRESSION OF THE NINE TYPE I MEMBERS IN THE HAIR FOLLICLE^*

Lutz Langbein^§, Michael A. Rogers^¶, Hermelita Winter^¶, Silke Praetzel, Ulrike Beckhaus^¶, Hans-Richard Rackwitz, and Jürgen Schweizer^¶

From the  Division of Cell Biology and the ^¶ Division of Tumor Cell Regulation, German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The human type I hair keratin subfamily comprises nine individual members, which can be subdivided into three groups. Group A (hHa1, hHa3-I, hHa3-II, hHa4) and B (hHa7, hHa8) each contains structurally related hair keratins, whereas group C members hHa2, hHa5, and hHa6 represent structurally rather unrelated hair keratins. Antibodies produced against these individual hair keratins, first analyzed for specificity by one- dimensional Western blots of total hair keratins, were used to establish the two-dimensional catalog of the human type I hair keratin subfamily. The catalog comprises two different series of type I hair keratins: a strongly expressed, Coomassie-stainable series containing hair keratins hHa1, hHa3-I/II, hHa4, and hHa5, and a weakly expressed, immunodetectable series harboring hHa2, hHa6 hHa7, and hHa8. In situ hybridization and immunohistochemical expression studies on scalp follicles show that two hair keratins, hHa2 and hHa5, define the early stage of hair differentiation, i.e. hHa5 expression in hair matrix and hHa5/hHa2 coexpression in the early hair cuticle cells. Whereas cuticular differentiation proceeds without the expression of further type I hair keratins, matrix cells embark on the cortical pathway by sequentially expressing hHa1, hHa3-I/II, and hHa4, which are supplemented by hHa6 at an advanced stage of cortical differentiation, and hHa8, which is expressed heterogeneously in cortex cells. Thus, six type I hair keratins are involved in the terminal differentiation of anagen hairs. The expression of hHa7 is conspicuously different from that of the other hair keratins in that it does not occur in the large anagen follicles of terminal scalp hairs but only in central cortex cells of the rare and small follicle type that gives rise to vellus hairs.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The keratin multigene family consists of the epithelial cytokeratins or "soft" -keratins, which are differentially expressed in the various single- and multilayered epithelia, and the hair keratins or "hard" -keratins, which essentially contribute to the formation of hair, nails, claws, and other hard keratinized structures. Both keratins are divided into type I (acidic) and type II (basic to neutral) members, which form the 10-nm intermediate filament cytoskeletal network of epithelial cells through obligatory association of equimolar amounts of type I and type II keratins (for review, see Refs. 1-3). Disturbances of intermediate filament formation, caused by distinct mutations in epithelial or hair keratin molecules, can weaken the structural integrity of the respective epithelial cells and result in hereditary diseases of the skin, mucosa, nails, and hair (4-8).

Previous studies on the composition of the hair keratin family by one- and two-dimensional gel electrophoresis of the chemically unmodified keratins indicated the presence of four type I (44-48 kDa) and four type II members (55-60 kDa), which were invariably found in hairs of different species (9, 10). According to a proposal by Heid et al. (9), the hair keratin proteins were collectively designated H for hair, b for the basic hair keratins (Hb),¹ and a for the acidic hair keratins (Ha), respectively, with the two dimensionally resolved and Coomassie-stained protein spots of each subfamily being numbered from 1 to 4 in an counterclockwise manner. In addition to the major hair keratins Ha1-4 and Hb1-4, a weakly expressed additional keratin pair was designated Hax/Hbx (9, 11). Independent of the species, the hair keratin family was therefore assumed to comprise 10 individual members and thus to be numerically smaller than that of the epithelial keratins. This notion seemed to be confirmed by the subsequent molecular cloning of four type I mouse hair keratins, mHa1-mHa4, and four type II sheep wool keratins, for which the designation K2.9-K2.12 has been proposed (12-18). Expression studies in mouse skin revealed a differential synthesis of hair keratins in the central hair-forming compartment of the follicle. Apparently, only one type I hair keratin, mHa2, was expressed in the peripheral hair cuticle. In contrast, the inner hair cortex exhibited the sequential expression of three type I keratins, mHa1, mHa3, and mHa4, whereas keratins of the hair matrix were unknown (15).

Recent investigations in our laboratory aimed at characterizing human hair keratin genes have shown that the hair keratin family is distinctly more complex than previously assumed on the basis of protein studies. Screening of a human scalp cDNA library with probes derived from mouse and sheep hair keratin clones yielded seven type I hair keratin clones (hHa1, hHa2, hHa3-I, hHa3-II, hHa4, hHa5, and hHa6) and four type II hair keratin clones (hHb1, hHb3, hHb5, and hHb6) (Refs. 19-23 and unpublished data). Fluorescence in situ hybridization analyses to human metaphase chromosomes using preliminarily characterized genomic clones for the type I hair keratin hHa2 and the type II hair keratin hHb1 showed that, just like the human epithelial keratin genes, the hair keratin genes were located in a type-specific manner on chromosomes 17q12-q21 (type I) and 12q13 (type II), respectively (21). Considering further evidences for keratin gene clustering (Ref. 24 and references therein), we recently subjected a human P1 artificial chromosome (PAC) library to a PCR-based screening with specific primers for the arbitrarily selected type I hair keratins hHa6 and hHa3-II. We were able to characterize 190 kilobase pairs of genomic DNA, which contained 10 type I hair keratin genes. They were clustered within a 140-kilobase pair domain and displayed the same direction of transcription (24). Besides the genes for the seven type I hair keratins previously characterized as cDNA clones, the DNA contig contained functional genes for two novel type I hair keratins, hHa7 and hHa8, and one transcribed hair keratin pseudogene, hHaA. Because there was no evidence for further hair or epithelial keratin genes on the 190-kilobase pair PAC contig, the 140-kilobase pair gene domain was assumed to harbor the entire complement of human type I hair keratin genes (24).

In the present study we have explored the expression pattern of the nine functional type I hair keratin genes in the human hair follicle. To this purpose, antibodies were raised against the individual keratins, which were first analyzed for specificity by one-dimensional Western blots of total hair keratins and then used to establish a catalog of human type I hair keratins on the basis of two-dimensional Western blots. Subsequent in situ hybridization experiments and indirect immunofluorescence studies not only revealed a complex and differential keratin expression in all areas of the hair-forming compartment, which was indicative of an early and late phase of hair differentiation, but also yielded unexpected findings of hair phenotype specific keratin expression.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Extraction of Hair Keratins, One- and Two-dimensional Gel Electrophoresis, and Western Blot Procedure-- Plucked beard hairs from two volunteers were freed from adhering outer and inner root sheaths as described (9). The lower part of the follicles was cut off and extracted for hair keratins as described (25). One-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; 10% polyacrylamide) and two-dimensional resolution of hair keratins by isoelectric focusing (IEF) were carried out as described (25, 26). For IEF, the following mixture of ampholines was used: pH 5-7 (0.8% w/v), pH 4-6 (0.8% w/v), and pH 3-10 (0.4% w/v) (Bio-Rad, München, Germany). One- and two-dimensional gels were stained with Coomassie Blue. For Western blot procedures, gels were transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore, Eschborn, Germany) by semi-dry blotting. After staining (0.1% Coomassie Blue R250, 40% methanol, 1% acetic acid), destaining (50% methanol), and blocking with 5% nonfat milk powder in Tris-buffered saline, membranes were incubated with primary antibodies (see "Antibodies"). Peroxidase-coupled secondary antibodies (see "Antibodies") were diluted 1:10,000 in 2.5% nonfat milk powder, 0.1% Triton X-100 in Tris-buffered saline. Secondary antibodies were detected by chemiluminescence (ECL, Amersham Pharmacia Biotech).

In Situ Hybridization (ISH) and Indirect Immunofluorescence (IIF)-- ISH on cryostat sections of human scalp or plucked hair follicles was carried out as described previously in detail (23, 25, 27). Human scalp samples were kindly provided by Drs. Bernard Cribier and J. H. Asch (Dermatological Hospital, Strasbourg, France). For the detection of the various hair keratin mRNAs, the following probes were used: hHa3-I, a 250-bp 3' PCR fragment; hHa4, a 150-bp 3' PCR fragment; hHa6, a 170-bp PCR fragment from exon 1; hHa7, a 300-bp PCR fragment from exon 1; and hHa8, a 250-bp PCR fragment from exon 1 (all of the preceding were cloned into pMosBlue); hHa3-II, a 450-bp BamHI/XhoI 3' fragment cloned into Bluscript II(KS) hHa5 (22); hHa1 (25); and hHa2 (26). Probes were radiolabeled by in vitro transcription using ³⁵S rCTP.

For the recording of the ISH signals by reflection microscopy, a confocal laser scanning microscope (LSM 410 UV, Carl Zeiss, Oberkochen, Germany) was used. The instrument allows simultaneous visualization of ISH in epi-illumination for the detection of reflection signals and transmitted light in bright field for hematoxylin staining. The two signal channels were combined by an overlay in pseudocolor (transmission image in green, electronically changed into black/white, reflection image, i.e. IHS signals, in red).

IIF, on cryostat sections of human scalp or plucked hair follicles, was carried out as described previously in detail by Winter et al. (25, 26). Results were documented with a photomicroscope (Axiophot 2, Carl Zeiss).

Antibodies-- The following primary monoclonal antibodies used were: hHa1 antibody LHTric-1 (25, 28) (ECL dilution 1:200, IIF undiluted); hHa2 antibody (26) (ECL dilution 1:50, IIF dilution 1:10). Antisera against human hair keratins were produced in guinea pigs by the injection of specific synthetic oligopeptides (100 µg/injection; see Table II), coupled to keyhole limpet protein. After the third booster injection, the following antisera were obtained: hHa3-II (ECL dilution 1:5000, IIF dilution 1:500); hHa4 (ECL dilution 1:1,000, IIF dilution 1:300); hHa5 (ECL dilution 1:30,000, IIF dilution 1:2,500); hHa6 (ECL dilution 1:40,000, IIF dilution 1:1000); hHa7 (ECL dilution 1:2,000; IIF dilution 1:1,000); and hHa8 (ECL, dilution 1:80,000; IIF, dilution 1:5,000). For ECL, peroxidase-coupled rabbit anti mouse IgG + IgM or peroxidase-coupled goat anti-guinea pig IgG (H + L) were used as secondary antibodies. For immunofluorescence, Cy2-, Cy3-, or Texas Red-coupled goat anti mouse IgG + IgM or anti-guinea pig IgG were used at dilutions of 1:50-1:200. All secondary antibodies were from Dianova (Hamburg, Germany).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Western Blot Analyses of Human Type I Hair Keratins-- Multiple sequence comparisons of the hair keratins encoded by the nine functional genes contained in the human type I hair keratin gene domain led to a striking sorting into two groups, A and B, each containing highly related hair keratins, and a third group, C, of structurally unrelated hair keratins (Table I). Group A comprised hair keratins hHa1, hHa3-I, hHa3-II, and hHa4, and group B contained hair keratins hHa7 and hHa8. In addition to highly homologous rod domains, the members of groups A and B each possessed head domains of identical length (24), for which the sequence homology was nearly as high as that of the rod domains (Table I). The third group, C, comprised hair keratins hHa2, hHa5, and hHa6. Their structural relationship was restricted solely to the rod domains, for which homology values were distinctly lower than those found for the members of groups A and B (Ref. 24 and Table I).

Both, mono- and polyclonal antibodies were raised against specific synthetic peptides that, in most cases, corresponded to the last 15-20 amino acid residues of the structurally variable tail domains (Table I) of the various type I hair keratins (Table II). One of the exceptions was the group A hair keratin isoforms hHa3-I and hHa3-II, in which the tail domains possessed close structural resemblance (24). We selected a hHa3-II carboxyl-terminal peptide region exhibiting the only perceptible sequence deviation from the hHa3-I tail for antibody production. Nevertheless, of its 17 amino acid residues, the hHa3-II peptide still contained nine amino acid residues in common with the hHa3-I isoform (Table II).

Total hair keratins extracts of human anagen hair bulbs from which the outer and inner root sheaths had been mechanically removed were separated by SDS-PAGE. Because the one-dimensional resolution of hair keratins was primarily aimed at verifying the specificity of the antibodies, no special efforts were made to optimally resolve the individual type I and type II hair keratins. Hence, the Coomassie-stained pattern shown in Fig. 1, lane a, consisted of a broad but distinct protein band ranging from 55 to 61 kDa, which contained the type II hair keratins, and a protein doublet between 44 kDa and 48 kDa, which harbored the smaller type I hair keratins.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Identification of human type I hair keratins by SDS-PAGE and Western blot analysis. Lane a, Coomassie staining of total hair keratins transferred to a nylon membrane. In addition to Coomassie staining, the position of the collective type II hair keratins (Type II HK) and type I hair keratins (Type I HK) was marked by perforations of the membrane (arrows in lanes a and c-j). Western blots were performed with antibodies against hHa1 (lane c), hHa2 (lane d), hHa3-II (lane e), hHa4 (lane f), hHa5 (lane g), hHa6 (lane h), hHa7 (lane i), and hHa8 (lane j). The open arrowhead in lane d marks a barely visible protein also detected by the hHa2 antibody. Lane b contains the following molecular mass markers (from top to bottom): 66, 43, 36, 29, and 24 kDa.

In accordance with previous investigations (25), hHa1 immunodetection revealed a strong protein band apparently covering the entire upper portion of the Coomassie-stained bipartite type I hair keratin pattern (Fig. 1, lane c). In contrast, the hHa2 antiserum detected a protein above hHa1 that was not stainable with Coomassie Blue (Fig. 1, lane d). Consistently, the hHa2 antiserum reacted with a second, slightly larger, but extremely weak band, which in most cases was hardly separated from the major hHa2-reactive band (arrowhead in lane d). The hHa3-II, hHa4, and hHa5 antisera each detected single proteins with similar mobilities in the lower portion of the Coomassie-stained type I hair keratin pattern (Fig. 1, lanes e-g). The hHa6 antiserum reacted with a protein, which similar to hHa2, was located outside of the Coomassie-stained type I hair keratin pattern (Fig. 1, lane h). The same held true for the hHa7 and hHa8 antisera; each of them detected a single protein band at ~54 kDa, immediately subjacent to the Coomassie-stained collective type II hair keratin band at 55-61 kDa (Fig. 1, lanes i and j). In most cases, the molecular mass values of the hair keratin proteins deduced from their migration in the gel were close to the values calculated from their sequences (Table II). This was, however, not the case for both hHa5, which despite a calculated molecular mass of 47,719 kDa migrated at the level of the small hHa3-II or hHa4 hair keratins, and hHa7/hHa8, for which the deduced molecular mass values deviated even more from the respective calculated values (Table II).

The two-dimensional pattern of human hair keratins obtained by IEF in an ampholine mixture that differed only slightly from that used by Heid et al. (11) is shown in Fig. 2a. It can be clearly seen that the high molecular mass portion (47-48 kDa) of the type I hair keratins was resolved into two protein spots of decreasing intensity, whereas the low molecular mass portion (44-46 kDa) consisted of three distinct protein spots, of which the most acidic one was slightly weaker in intensity than the others.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Identification of human type I hair keratins by two-dimensional gel electrophoresis and Western blot analysis. Panel a, Coomassie staining of total hair keratins transferred to a nylon membrane. Type II HK, type II hair keratins; Type I HK, type I hair keratins. Western blots were performed successively with antibodies against hHa1, hHa2, and hHa3-II (panel b); hHa4 and hHa5 (panel c); hHa5 and hHa6 (panel d); hHa3-II and hHa7 (panel e); and hHa8 (panel f).The proteins were separated by IEF in the first dimension and by SDS-PAGE in the second dimension (SDS, 8% polyacrylamide). The two bars at the right of the figure indicate the mean molecular mass values of the type II and type I hair keratins.

After electroblotting and Coomassie staining, membranes were first exposed separately to the various antibodies (results not shown). For a more rational approach toward the arrangement of the reacting protein spots relative to each other, the membranes were then reacted consecutively with several antibody combinations. Fig. 2b shows that the hHa1 antibody recognized the entire upper series of Coomassie-stainable protein spots, indicating that they represent isoelectric hHa1 variants. Re-exposure of the membrane to the hHa2 antibody confirmed the localization of this hair keratin above the Coomassie-stained pattern. hHa2 showed up as a single protein spot, which was located exactly above the main hHa1 variant (Fig. 2b). Moreover, the antibody reacted with a slightly larger and weakly expressed protein component, which was also seen in one-dimensional blots (arrowheads in Figs. 1d and 2b). Additional exposure of the membrane to the hHa3-II antiserum revealed a specific reaction with the least acidic spot of the low molecular mass protein series (Fig. 2b; see also Fig. 2e). The detection of only one protein by the hHa3-II antibody in both one- and two-dimensional Western blots strongly indicated that, despite a substantial sequence homology of the hHa3-II peptide selected for immunization with the corresponding region of hHa3-I, this antibody specifically reacted with the hHa3-II isoform. Use of a new blot and the hHa5 antiserum clearly showed that the main variant of this hair keratin colocalized with hHa3-II. However, the antibody also recognized the second, more acidic spot of the series. This reaction is not visible in Fig. 2c (see, however, Fig. 2d), because the subsequent probing of the membrane with the hHa4 antiserum revealed the colocalization of the hHa5 satellite spot with the main hHa4 variant. In addition, the hHa4 antiserum recognized the most acidic spot of the series as a minor isoelectric hHa4 variant. The hHa6 antiserum, which was probed together with the hHa5 antiserum on a new blot membrane, detected two isoelectric variants of this hair keratin at the height of hHa2 in the unstained region of the membrane (Fig. 2d). Simultaneous exposure of a separate blot membrane with the hHa6 and hHa2 antibodies revealed a colocalization of the main hHa6 variant with hHa2 (results not shown). Fig. 2e shows the reaction of another blot membrane with antisera against hHa3-II and hHa7. The latter appeared as a single protein spot located at ~54 kDa. Also the hHa8 antibody detected a single protein (Fig. 2f), which exhibited complete colocalization with hHa7 upon coexposure of the membrane with the respective antisera (results not shown).

Expression of Type I Hair Keratins in the Human Hair Follicle-- The expression patterns of the members of the three type I hair keratin groups, A, B, and C, in human scalp hair follicles were investigated by both ISH, with specific cRNA probes of the respective hair keratin genes, and IIF, with the specific antibodies described above.

Group A Hair Keratins hHa1, hHa3-I, hHa3-II, and hHa4 Are Sequentially Expressed in the Hair Cortex-- ISH with the hHa1 cRNA probe on human hair follicles revealed a strong expression of the hHa1 mRNA, which began at the transition of the matrix and the cortex and remained prominent throughout the lower and mid-cortex region before it gradually vanished in the uppermost cortex cells. Cells lining the apex of the dermal papilla were free of hHa1 transcripts (Fig. 3a). The overall expression of the hHa3-I mRNA was considerably weaker than that of hHa1 and clearly limited to a much smaller area of the cortex (Fig. 3b). First hHa3-I transcripts were observed only in cells of the mid-cortex region and obviously disappeared somewhat earlier than the hHa1 mRNA. The clearly delayed onset of hHa3-I mRNA expression, when compared with hHa1, could best be demonstrated in the three follicles sectioned at different levels of the hair shaft (Fig. 3b).The investigation of hHa3-II mRNA expression in serial sections of the same scalp sample showed, that in terms of location and intensity, the resulting transcript pattern represented virtually an exact replica of the pattern found in hHa3-I mRNA (compare Fig. 3, b and d). In contrast, ISH with the hHa4 cRNA probe showed that this hair keratin, in which the mRNA was almost as strongly expressed as in hHa1, constituted the latest expressed group A member. First cortical hHa4 transcripts clearly occurred subsequent to the onset of hHa3-I/II mRNA expression and were maintained almost up to the uppermost cortex cells (Fig. 3c).

View larger version (190K): [in this window] [in a new window]   Fig. 3.   Demonstration of group A hair keratin mRNA synthesis. ISH is shown on human scalp sections with specific cRNA probes for mRNAs of hHa1 (a, note mRNA-free cells lining the dermal papilla), hHa3-I (b), hHa3-II (d), and hHa4 (c). Red arrows in each panel demarcate the site of mRNA expression in the cortex. ORS, outer root sheath; IRS, inner root sheath; co, cortex; cu, cuticle; ma, matrix; dp, dermal papilla. Bars = 200 µm.

Indirect immunofluorescence studies with the antisera against hHa1, hHa3-II, and hHa4 faithfully reflected the sequential onset of mRNA expression of these hair keratins in the cortex but showed that all of these hair keratin proteins could be demonstrated up to the zone of hair fiber hardening (Fig. 4 a-c).

View larger version (67K): [in this window] [in a new window]   Fig. 4.   Immunohistochemical labeling of group A hair keratins. IIF is shown with antibodies against hHa1 (a), hHa3-II (b), and hHa4 (c) on human scalp sections. Red arrows demarcate the site of mRNA synthesis; green arrows indicate the site of hair keratin proteins in the cortex. All sections were counterstained with DAPI. For abbreviations, see the legend to Fig. 3. Bars = 200 µm.

The Expression Sites of the Group B Hair Keratins hHa7 and hHa8 in the Hair Follicle Are Unrelated-- ISH with specific hHa7 and hHa8 cRNA probes led to extremely weak signals, which were difficult to interpret with regard to their exact location in the hair-forming compartment of the follicle (results not shown). The expression of these hair keratins was therefore investigated by IIF. The hHa8 antiserum revealed the presence of stained cells from the lower to mid-cortex up to the zone of hair hardening (Fig. 5a).The onset of hHa8 synthesis was thus comparable with that of hHa1. However, unlike the group A hair keratins, hHa8 synthesis clearly occurred in distinct cells, which were randomly scattered throughout the entire cortex. Although this labeling was best visible in the lower cortex region in which some of the earliest hHa8-positive cells had not yet acquired a vertical orientation and a spindle-like shape (see magnification in Fig. 5a'), cross-sections through different levels of hair follicles unambiguously showed that notwithstanding the gradually narrowing of the hair shaft and the resulting compression of the cortex, isolated hHa8-positive cells could also be identified in the very late cortex region (Fig. 5b). Double-label IIF studies with antibodies against hHa1 and hHa8 demonstrated that, with the exception of some early hHa8-positive cortex cells, the two hair keratins were coexpressed (Fig. 5c).

View larger version (58K): [in this window] [in a new window]   Fig. 5.   Immunolabeling of group B hair keratin hHa8. IIF is shown with the hHa8 antibody on longitudinal sections of hair follicles (a) and at a higher magnification of the lower cortex region (a'). Cross-section showing three follicles cut at different heights of the hair shaft (b). Double-label IIF with antibodies against hHa8 (red) and hHa1 (green; yellow color indicates coexpression) on human scalp sections (c). The arrows in panels a and a' point to very early hHa8-positive cells. The arrows in panel c indicate hHa8-positive cells that have not yet started hHa1 synthesis. Panels a and c were counterstained with DAPI. For abbreviations, see the legend to Fig. 3. Bars = 200 µm.

At first glance, the pattern of hair keratin hHa7 synthesis shown in Fig. 6a deviated from that of hHa8 (Fig. 5a) only by the striking restriction of hHa7-positive cells to the innermost region of the mid- to upper cortex of the hair follicle. At a higher magnification (Fig. 6b), the hHa7-positive cell column were seen to be composed of only a few cells, which in cross-sections seemed to be more or less concentrically arranged around a hHa7-negative core column (Fig. 6d). The most striking feature of hHa7-expressing follicles was, however, that they clearly represented a folliclular type that, in several aspects, was fundamentally different from the group A hair keratin and hHa8-expressing follicles. This difference is demonstrated in Fig. 6, which shows a scalp section stained with the hHa7 antibody (Fig. 6c) and the corresponding phase contrast visualization of the same section (Fig. 6c'). Although the latter shows that, from the morphological point of view, the hHa7-positive hair follicle at the right fulfills all of the criteria of an anagen follicle (see also the phase contrast visualization in Fig. 6a'), it is clearly located much higher in the scalp dermis than the anagen hair follicle at the left, which represents the overwhelming type of scalp anagen follicles. Moreover, in addition to their higher location in the scalp dermis, hHa7-positive follicles were also distinctly smaller and more slender in size than hHa7-negative follicles (Fig. 6, a' and c').

View larger version (85K): [in this window] [in a new window]   Fig. 6.   Demonstration of group B hair keratin hHa7 synthesis. IIF is shown with the hHa7 antibody. a and a', longitudinal scalp section that contains a vellus hair follicle (V). b, higher magnification of the cortex region. c and c', longitudinal sections of terminal anagen (A) and vellus hair follicle; d and d' show respective cross-sections. Green arrows indicate the site of hHa7 synthesis. Open yellow arrows in b and d indicate heterogeneous onset of hHa7 synthesis. The asterisks in panels c and c' denote an air bubble. For abbreviations, see the legend to Fig. 3. Bars = 200 µm.

Group C Hair Keratins hHa2, hHa5, and hHa6 Are Differentially Expressed in the Hair-forming Compartment of the Follicle-- ISH with a hHa2 specific cRNA probe revealed transcripts exclusively in cells of the hair cuticle. First signs of hHa2 mRNA expression were seen in cuticular precursor cells of the lower hair bulb immediately above the germinative cell pool. The hHa2 mRNA-positive single-cell layer ascended vertically and could be followed up to the transition of the lower to mid-cortex region where the cuticular hHa2 mRNA expression gradually ceased (Fig. 7a). The corresponding IIF studies with the hHa2 antibody confirmed the synthesis of the hHa2 keratin in the entire hair cuticle (Fig. 8a). Cuticular precursor cells in the lower hair bulb were cuboidal and contained large nuclei (Fig. 8a and lower inset). During their upward journey, they expanded horizontally, then flattened, and eventually acquired a distally oriented slant that was best seen by the likewise obliquely oriented nuclei of the cells (Fig. 8a). Later, cuticular cells became anucleated and showed up as a continuous band that terminated rather abruptly at the level of hair fiber hardening (Fig. 8a). However, deviating from the in situ hybridization data (Fig. 7a), the hHa2 antibody also decorated a 10-12 cell layers-thick central area that extended from the uppermost matrix into the lower cortex. Consistently, the labeling intensity of these cells was distinctly weaker than that seen in cuticular cells (asterisks in Fig. 8a and lower inset).

View larger version (188K): [in this window] [in a new window]   Fig. 7.   Demonstration of group C hair keratin mRNA synthesis. IHS is shown on longitudinal scalp sections with specific cRNA probes for hHa2 mRNA (a, the inset shows a cross-section through the widest point of the hair bulb), hHa5 mRNA (b), and higher magnification (b'). The open white arrows in b' indicate hHa5 mRNA-positive cells bordering the dermal papilla and hHa6 mRNA (c). The red arrows in panels a, b, and b' indicate the site of expression of the respective mRNAs. gp, germinative pool (for further abbreviations, see the legend to Fig. 3). Bars = 150 µm.

View larger version (91K): [in this window] [in a new window]   Fig. 8.   Immunolabeling of group C hair keratins. IIF is shown with antibodies against hHa2 (a, upper inset, cross-section through the upper hair shaft; lower inset, higher magnification of the suprabulbar region), hHa5 (b), and hHa6 (c) on anagen hair follicles (DAPI counterstaining). Red arrows demarcate the site of mRNA expression, and green arrows indicate the labeling sites of hair keratin proteins. For the asterisks in a, see text ("Results"). Double-label IIF is shown with antibodies against hHa2 (green) and hHa5 (red; yellow indicates coexpression) on longitudinal scalp sections (d and d') and a cross-section thereof (e). Double-label IIF is shown with antibodies against hHa5 (red) and hHa1 (green; yellow indicates coexpression) on a longitudinal section (f) and a cross-section (g). For abbreviations, see the legend to Fig. 3. Bars = 200 µm.

ISH with the hHa5-specific cRNA probe led to a prominent, well circumscribed label apparently comprising the entire matrix compartment as well the lowermost cortex region about 7-10 cell layers above the vertex of the dome-shaped dermal papilla (Fig. 7b). Similar to the hHa2 transcripts, hHa5 transcripts could first be seen immediately above the germinative cell pool. Cells bordering the dermal papilla were devoid of hHa5 transcripts, although occasionally some of them, apparently in the process of leaving this borderline layer, were labeled (small arrows in Fig. 7b'). As expected, the hHa5 antibody decorated the entire region from the lowermost matrix up to the zone of hair fiber hardening (Fig. 8b). Again, cells lining the dermal papilla were free of label. However, a closer examination of the outermost hHa5-positive cells led us to doubt whether these cells represented matrix or cortex cells, considering that their morphology, in particular their cuboidal shape in the lower and midportion of the follicle, and the outward slope of the nuclei of cells in the upper cortex region resembled that of hHa2-positive cells of the hair cuticle (see Fig. 8a).To verify this result, we performed double-label IIF studies with both the hHa5 antiserum (visualized in red) and the hHa2 antibody (visualized in green). As shown in a longitudinal section (Fig. 8d) and a cross-section through different levels of the hair shaft (Fig. 8e), the entire hair cuticle appeared in yellow, thus clearly demonstrating hHa5/hHa2 coexpression in this compartment of the follicle. (Because of the strong hHa5 expression in the entire matrix/cortex region (Fig. 8d), the weakly hHa2-positive cell population in the lowermost cortex (asterisk in Fig. 8d', specifically visualized for hHa2) was largely obscured in Fig. 8d.) Subsequent double-label IIF with the hHa5 antiserum (visualized in red) and the antibody against the most prominent cortex keratin, hHa1 (visualized in green), not only confirmed the synthesis of hHa5 in the hair cuticle but also reinforced the restriction of hHa1 expression to the cortex (Fig. 8, f (longitudinal section) and g (cross-section)).

ISH with a hHa6 specific cRNA probe clearly identified this keratin as another cortex keratin in which mRNA expression extended over the mid-cortex (Fig. 7c), i.e. the area in which also hHa3-I/II and hHa8 mRNA expression was found. In terms of intensity, the hHa6 mRNA expression resembled that of the two weakly expressed hHa3 isoforms (see Fig. 3, b and d). This resemblance was confirmed by indirect immunofluorescence with the hHa6 antiserum, which revealed well discernible spindle-shaped cells from the mid-cortex region up to zone of hair fiber hardening (Fig. 8c).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Contrary to the previous view of the existence of four major and one minor hair keratin pair, we recently showed that the human type I hair keratin subfamily alone comprises as many as nine functional members, which, on the basis of sequence characteristics, can be divided into three groups, A, B, and C. Moreover, it was found that the genes of the members of each group form distinct subclusters within the 140-kilobase type I hair keratin gene domain (24). In the light of these findings, it became mandatory to not only reinvestigate the previous gel electrophoretic pattern of this hair keratin subfamily and to appropriately adapt the designation of its members (9) but also to determine the expression sites of these multiple keratins in the hair-forming compartment of the follicle.

For these purposes, we first generated antibodies against the individual type I hair keratin proteins. In several cases, the selection of immunogenic keratin-specific oligopeptide sequences, usually derived from the very end of the keratin tail domain, required some precautions, which are indicated in Table II. Noteworthy is the unexpected and striking sequence homology between the last 15 carboxyl-terminal amino acid residues of hair keratin hHa6 (GKVISSREHVQSRPL (24)) and the type I epithelial keratin K17 (GKVISSREQVHQTTR (29)); nine common residues are in italics). To circumvent potential cross-reactions, the very end of the hHa6 tail domain was therefore not used for antibody production. It is needless to emphasize that similar preventive measures should also be taken, in particular when an antibody against K17 is aimed at being generated by immunization with K17-derived oligopeptides.

The various type I hair keratin antibodies proved to be highly specific in that they recognized only single protein bands when analyzed in Western blots of one dimensionally resolved human hair keratins. One exception concerned the hHa2 antibody, which besides the prominent hHa2 protein band also reacted with a slightly larger minor band. As mentioned in Table II, the carboxyl-terminal hHa2 oligopeptide used for immunization shares some sequence homology with the corresponding hHa5 region. The occurrence of the minor band, however, can not be explained by a possible cross-reaction of the hHa2 antibody with hHa5, because the latter, revealed by an antibody against a highly specific hHa5 amino-terminal peptide, migrates distinctly faster in the gel. At present, the nature of the minor hHa2 reactive protein remains unknown.

Coomassie staining of the two dimensionally separated hair keratins revealed a pattern for the type I subfamily consisting of four major protein spots that were located essentially the same as in the pattern previously published by Heid et al. (9, 11). Successive Western blot analyses allowed the identification of these proteins as hair keratin hHa1, which represents the largest and most abundantly expressed member of the stainable type I hair keratins, the lower molecular mass keratins hHa3-II and hHa4, all being group A hair keratins, as well as the group C hair keratin hHa5. With the exception of hHa3-II, each of these hair keratins exhibits at least one satellite spot, so that the low molecular mass hair keratins hHa3-II, hHa5 and hHa4 form a complex pattern in which major and minor isoelectric variants of the various keratins are partly superposed. The remaining group C members hHa2 and hHa6, as well as group B hair keratins hHa7 and hHa8, can be detected only by Western blot analysis. The fact that these hair keratins are not revealed by Coomassie staining, although both hHa2/hHa6 and hHa7/hHa8 exhibit comigration, implies that these keratins represent only very weakly expressed members of the type I hair keratin subfamily. These predictable expression characteristics are highlighted schematically in the catalog of the human type I hair keratin subfamily shown in Fig. 9a, in which the strongly expressed hair keratins hHa1, hHa3-II, hHa4, and hHa5 are indicated by closed circles and the weakly expressed hair keratins hHa2, hHa6, hHa7, and hHa8 are represented by open circles. A comparison of this catalog with the respective pattern of human type I hair keratins published by Heid et al. (9, 11) shows that out of the previous nomenclature, only the designation for hHa1 can be maintained. The type I hair pattern derived from Heid et al. (9) also contains a minor hair keratin, Hax, which could only be revealed by means of Western blots with a pan-type I hair keratin antibody (11). It is clear that the position of this keratin corresponds to the protein spot in the catalog that is composed of both hHa2 and the main variant of hHa6. It will be shown below that the decision as to which of these two keratins represents Hax can be made by expression studies.

View larger version (23K): [in this window] [in a new window]   Fig. 9.   The catalog of human type I hair keratins. a, schematic presentation of the two-dimensional pattern of human type I hair keratins identified in this study by means of specific antibodies. Hair keratins also detectable by Coomassie staining are indicated by black circles; hair keratins detectable only by the respective antibodies are given as white circles. b, schematic presentation of the two-dimensional pattern of Coomassie-stained human type I hair keratins (shaded circles) and their designation according to Heid et al. (9). (The pattern was derived from Fig. 4a of Heid et al. (9)). The scheme also contains the minor hair keratin Hax (white circle), detectable only by Western blots with a pan-type I hair keratin antibody (11). The proteins are resolved by IEF in the first dimension and by SDS-PAGE (SDS) in the second dimension. Abscissa, isoelectric pH; ordinate, molecular mass values.

Both ISH and IIF showed that the structurally highly related group A hair keratins represent cortex keratins that are sequentially expressed in the order hHa1, hHa3-I/II, and hHa4. Because hHa1 paves the way toward cortical differentiation, its preponderance at the protein level over the later cortex keratins becomes understandable. Although expression data for the murine ortholog of the late cortex keratin hHa4 (12) are not available, a similarly staggered cortical expression has at least been shown for the mRNAs of mHa1 and mHa3 (15). Moreover, the human type II hair keratins hHb1, hHb3, and hHb6 (23), as well as their sheep analogs K2.9, K2.10, and K2.11 (16), which also exhibit an extraordinarily high and species independent overall sequence homology among each other, have likewise been shown to be sequentially expressed in the given order in the cortex of the hair and wool follicles (16, 23). This sequential expression pattern strongly suggests that the differentiation of the human hair cortex involves the stepwise expression of at least the hair keratin pairs hHa1/hHb1, hHa3-I/II/hHb3, and hHa4/hHb6.

In contrast, each of the structurally unrelated group C hair keratins exhibits a conspicuously different expression pattern in the various parts of the hair-forming compartment. Similar to the mRNA of its murine ortholog (15), the expression of the hHa2 mRNA is virtually restricted to the vertically ascending cell row of early cuticle cells in the hair bulb, whereas the hHa2 protein can be followed up to terminally differentiated cuticle cells. Conversely, at first glance, the hHa5 mRNA seems to be strongly expressed in the entire matrix region above both the germinative cell pool and the cell layer apposed to the dermal papilla. However, the examination of hHa5 protein localization by means of a specific antiserum, particularly in combination with either the hHa2 (cuticle) or the hHa1(cortex)-specific antibody, clearly abrogated our previous view of a mutually exclusive hHa2/hHa5 expression in the hair-forming compartment (22). Instead, our data unambiguously revealed that hHa5 is coexpressed with hHa2 in the entire hair cuticle. The observed marginal reaction of the hHa2 antibody with some of the late matrix and early cortex cells may be due to the minor protein detected by the antibody in Western blots; or alternatively, the partial homology of the selected carboxyl-terminal hHa2 peptide with hHa5 (see Table II), obviously unremarkable on denatured keratins on blots, may lead to a subtle cross-reaction with native hHa5 proteins in cells exhibiting the highest degree of synthesis of this hair keratin. Generally, however, the differential expression of these hair keratins explains the gel electrophoretic identification of hHa5 as a major and hHa2 as a minor hair keratin.

In retrospect, it should be emphasized that a reliable demonstration of cuticular hHa5 expression solely on the basis of in situ hybridization is critical, because the expression of the hHa5 and hHa2 mRNAs ceases at a similar height of the hair shaft, and the power of resolution of hybridization signals does not allow a trustworthy distinction between matrix and early cuticle cells in the bulbar region. The cuticular hHa5 expression is, however, further confirmed by the findings that an antibody against the type II hair keratin hHb5, in which the mRNA exhibits the same onset of expression in the lowermost hair bulb as that of hHa5 (23), also reacts with both the matrix/cortex and the hair cuticle (unpublished preliminary data). These findings suggest that hHa5 and hHb5 form the earliest expressed hair keratin pair.

The timing and spacing of hHa2 and hHa5 expression in the lower hair-forming compartment reflects the stringency of the gene control mechanisms that govern the commitment of topologically adjacent pluripotential germinative cells into either the cuticular or the matricial cell lineage. To ensure the selective expression of hHa2, the underlying gene regulatory design must be particularly vigorous for the single-cell precursor of the cuticular pathway. Whereas matrix cells cease hHa5 mRNA expression when embarking cortical differentiation via the stepwise expression of the group A hair keratins, there is no evidence that the differentiation of cuticular cells above the bulbar region is associated with the expression of further hair keratins. Thus both hHa5/hHb5 and hHa2, together with its still unknown type II partner, form the wispy intermediate filament network in early cuticle cells (30, 31). Higher up, these intermediate filaments most probably aggregate with the members of two keratin-associated protein families, KAP5 and KAP10 (31), without, however, forming the highly organized keratin macrofibrils within an electron dense matrix that are typically seen in cortex cells (for review, see Ref. 31).

The third group C member, hHa6, was discovered to be another cortex keratin that is added to the group A hair keratins at a more advanced stage of cortex differentiation, but similar to hHa2, it does not figure among the major hair keratins. The expression pattern of these two minor hair keratins now enables the identification of the weak type I hair keratin previously designated Hax by Heid et al. (11), which occupies a position in two-dimensional gels similar to the comigrating main hHa2 and hHa6 variants. Besides a pan-type I hair keratin antibody, Hax was also selectively detected among human type I hair keratins by Western blots with monoclonal antibody K_s19.2, otherwise specific for the type I epithelial keratin K19 (11). IIF with K_s19.2 on human hair follicle sections not only revealed certain cells of the outer root sheath, which most probably represent K19-expressing Merkel cells (32, 33), but also led to a distinctly suprabulbar, albeit weak, staining of the cortex (11), i.e. the same region detected by the hHa6 specific antiserum. Thus, Hax is identical to the group C hair keratin hHa6. In this context it should be mentioned that despite their conspicuous structural deviation, the group C hair keratins fulfill the criteria of hair keratins in that, on an average, each member possesses ~20 positionally highly conserved cysteine residues, a value that is only marginally different from that encountered in the group A cortex keratins (i.e. ~23 residues (24)). Their true "hard" keratin nature is convincingly confirmed by the finding that, together with its type II partner, the murine ortholog of hHa6 is able to built up the hard keratinized parakeratotic mouse tail scale epidermis, which morphologically has many aspects in common with the differentiation of the hair cortex (34).

Although the recently identified group B hair keratins hHa7 and hHa8 exhibit by far the highest sequence homology to each other, their expression patterns are fundamentally different. Thus, hHa8 clearly represents another cortex keratin, bringing the overall number of type I hair keratins involved in the cortical differentiation pathway to six. The most striking feature of hHa8, however, is the restriction of its expression to randomly scattered cortex cells, which implies even more stringent gene regulatory constraints than for the cuticular hHa2 gene. Neither by their position nor their shape can the hHa8-positive cells be distinguished from their hHa8-negative neighbor cells. There are other examples of locally restricted keratin expressions that add to a basic keratin expression scenario, i.e. those of the epidermal keratins K9 and K2 (27, 35-37). However, in these cases, the respective expression patterns comprise distinct and extended cell collectives, which may, for instance, confer a regionally varying tissue cohesiveness (38). Although, during their upward journey, cortex cells acquire an IF network that ultimately results from the assembly of as many as seven different pairs of hair keratins, it is thinkable that the growing hair fiber is further stabilized by the intercalation of randomly spaced hHa8-expressing cells.

Compared with all other type I hair keratins, the expression of hHa7 was literally out of place in that it did not occur in the large terminal anagen hairs of the human scalp but in a follicular type that, by its location and size, most probably represents a vellus hair. It has since long been recognized that vellus hairs, although at a very low frequency, are part of the normal scalp hair population of both male and female infants and adolescents, as well as adults. On the other hand, it is known that beginning in adolescence, vellus hairs also arise from the stepwise and long-lasting conversion of terminal hairs as a consequence of balding (for review, see Ref. 30). In the present case, the scalp sample was taken from the frontal and parietal scalp regions of a 57-year-old male exhibiting no obvious signs of balding. It has, however, frequently been found that whether the signs of baldness are visible or not, the hair line along the frontal region of the forehead in particular always contains a population of follicles in every conceivable transitional stage from perfect terminal hairs to fully converted vellus hairs (30). At present, we are not in a position to decide in which type of vellus hair, i.e. that possibly persisting from childhood or that resulting from terminal hair conversion, hHa7 is expressed. We are aware that this question can be answered only by extended hHa7 expression studies, not only in different areas of the normal and balding scalp but also in other body sites known to exhibit either uniquely vellus or composite hair populations.

In summary, the elucidation of the entire human type I hair keratin subfamily has enabled us, for the first time, to describe the maturation of the terminal anagen scalp hair on the basis of its most important structural proteins. The earliest step of hair differentiation, i.e. the commitment of pluripotential germinative cells at the base of the hair bulb for trichocytes, is associated with the differential expression of two structurally unrelated hair keratins, hHa2 and hHa5. Whereas hHa5 expression alone defines the matricial cell lineage, its coexpression with hHa2 determines the cuticular cell lineage. Later stages of cuticular differentiation do not involve the expression of further type I hair keratins. In contrast, matrix cells convert into terminally differentiating cortex cells along with the stepwise expression of hair keratins hHa1, hHa3-I, hHa3-II, and hHa4. Most probably, these abundantly expressed and structurally highly related hair keratins constitute the elemental keratin equipment of cortex cells that is, however, supplemented by two further hair keratins, one of which, hHa6, is added at an advanced stage of differentiation, whereas hHa8 is expressed in a subpopulation of cortical cells. It is evident that this complex scenario of early and late hair keratin expression requires highly intricate control mechanisms at the gene level. At least for the structurally related hair keratins hHa1-hHa4, the striking subclustering of their genes may indicate higher order control mechanisms, which govern their collective expression in the cortex; those regulatory mechanisms can most probably be excluded for the likewise subclustered, but differentially expressed, hHa2, hHa5, hHa6, and hHa7, hHa8 gene groups, respectively. In particular the totally divergent expression of hHa7 in vellus hairs should be reflected in its gene regulatory design. It has recently been found that lymphocyte enhancer factor-1 (LEF-1) plays an important role in the expression of hair keratin genes and probably in other hair-specific genes through the binding to a LEF-1 consensus sequence in the proximal promoter region of the corresponding genes (39). A LEF-1 consensus sequence is invariably located at a highly conserved position in the proximal promoter region of the members of the human type I hair keratin family but it is specifically lacking in the entire promoter of the hHa7 gene (24).

	ACKNOWLEDGEMENTS

We are grateful to Drs. Werner W. Franke (German Cancer Research Center) for helpful discussion and advice, to Bernard Cribier and Jean Henry Asch, Department of Dermatology, University of Strasbourg, France, for providing us with human scalp samples, and to Herbert Spring, German Cancer Research Center, for his help with confocal laser microscopy.

	FOOTNOTES

^* This work was supported in part by the Deutsche Forschungsgemeinschaft (Grant Schw 539/1-3).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed: German Cancer Research Center, Div. of Cell Biology, A0100, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. Tel.: 49-6221-42-3436; Fax: 49-6221-42-3404; E-mail: langbein@dkfz-heidelberg.de.

	ABBREVIATIONS

The abbreviations used are: Hb, basic hair keratin; Ha, acidic hair keratin; m, mouse (mHa); h, human (hHa); PCR, polymerase chain reaction; SDS-PAGE, SDS-polyacrylamide gel electrophoresis; IEF, isoelectric focusing; ISH, in situ hybridization; IIF, indirect immunofluorescence; bp, base pair(s); LEF-1, lymphocyte enhancer factor-1; DAPI, 4',6-diamidino-2-phenylindole.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike