Ann Otol Rhlnol Laryngoll00: 1991

KERATINS AS MOLECULAR MARKERS OF EPITHELIAL DIFFERENTIATION: DIFFERENTIAL EXPRESSION IN CRYPT EPITHELIUM OF HUMAN PALATINE TONSILS DIETMAR MISCHKE, PHD BERLIN, GERMANY

TOMOHIRO GENKA, MD

GABRIELE WILLE

OKINAWA, JAPAN

BERLIN, GERMANY

HARTMUT LOBECK, MD

ARTUR G. WILD, MD

BERLIN, GERMANY

MUNICH, GERMANY

The expressionof keratins in the stratified squamous nonkeratinizing epithelium lining the surface and the crypts of human palatine tonsils was analyzed by high-resolution gel electrophoresis, immunoblotting, and immunohistochemical techniques. In contrast to the superficial epithelium, which showed a fairly constant keratin composition consisting of the neutral-to-basic keratins K4, K5, K6, and K8 and the acidic keratins K13, K14, K16, and K19, the keratin profiles of tonsillar crypt epithelial cells were found to be more variable, particularly with respect to the expression levels of K4 and K13. These were identical to those of surface epithelium, reduced, or abolished. Since K4 and KI3 characterize the mature stage of differentiation in squamous nonkeratinizing epithelia, their decrease is indicative of an incomplete epithelial differentiation. Immunohistochemical analyses confirmed this hypothesis and allowed us, furthermore, to correlate the expression of KI3 with the morphologic alterations of tonsillar crypt epithelium in the course of reticulation. KEY WORDS - biochemical and immunohistochemical analysis, epithelial differentiation, keratin expression, palatine tonsil, tonsillar crypt epithelium.

INTRODUCTION

weight from 40,000 to 70,000 d. Of the more than 20 human keratins described and catalogued so far,":" different members are detected in the diverse epithelial tissues, and each epithelial cell expresses only a few, frequently as coordinated pairs consisting of one neutral-to-basic and one acidic keratin. 11,12

Human palatine tonsils are lined, on their surface, by a nonkeratinizing stratified squamous epithelium that extends continuously into the branching system of tonsillar crypts. The particular function of the palatine tonsil with respect to the immunoprotection of an individual, however, brings about profound morphologic changes in the tonsillar crypt epithelium, termed reticulation. The regular stratified arrangement of epithelial cells is loosened and immunoreactive cells such as lymphocytes, plasmocytes, and macrophages invade the epithelium to directly contact the outside and handle antigens. While these fine structural changes in the course of reticulation have been studied,I-5 the underlying concomitant molecular events and alterations still remain to be investigated.

In this paper we report on the patterns of keratin expression in epithelial cells lining the crypts of human palatine tonsils. In comparison to the superficial cell layers of tonsils, the cryptal epithelium showed qualitative and quantitative alterations that concerned particularly keratins K4 (a v b, where v = and/or) and K13. The differential expression of these keratins could be correlated in immunohistochemical analyses to the morphologic alterations that tonsillar crypt epithelium undergoes in the course of reticulation and thus should reflect the different capabilities of tonsillar crypt epithelial cells to reach an advanced stage of differentiation.

Recent studies have indicated that the keratin proteins can serve as excellent markers for different types and pathways of epithelial differentiation and have also established their usefulness in tumor typing. 6 - 8 As specific constituents of the lO-nm (intermediate-size) filaments in almost all epithelial cells, keratins (or cytokeratins) belong to a heterogeneous family of acidic (pI < 5.5) and neutral-to-basic (pI >5.5) proteins ranging in apparent molecular

MATERIALS AND METHODS

Tissues. Palatine tonsils were obtained from 79 patients, 40 of whom were female and 39 of whom were male, who underwent elective tonsillectomy. Ages ranged from 3 to 45 years (mean, 15.3 years).

From the Biochemistry Laboratory of the Department of Otorhinolaryngology-Head and Neck Surgery (Mischke, Cenka, Wille) and the Department of Pathology (Lobeck), Rudolf Virchow Medical Center, Free University of Berlin, Berlin, Germany; and the Department of Otorhinolaryngology-Head and Neck Surgery, Grosshadern Medical Center, University of Munich, Munich, Germany (Wild), Presented in part at the 58th Annual Meeting of the German Society of Otorhinolaryngology-Head and Neck Surgery, Bad Neuenahr, Germany, May 31June 4, 1987, REPRINTS - Dietmar Mischke, PhD, Laboratorium fiir Biochemie der Hals-Nasen-Ohren-Klinik, Unlversitatsklinikum Rudolf Virchow (Charlottenburg), Spandauer Damm 130, D-I000 Berlin 19, Germany,

372

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015

373

Mischke et al, Keratins as Markers of Epithelial Differentiation

13~

14~ 16~

19~

-

4a c>

4ac>

5b c>

6 c>

6 c>

8 c>

2

3

4

5

6

2

3

4

5

6

Fig 1. Sodium dodecyl sulfate polyacrylamide gel electrophoresis and immunoblot analysis of cytoskeletal proteins present in superficial and cryptal epithelium of human palatine tonsils. Cytoskeletal residues represent surface epithelium (lane 1) and five different preparations of cryptal epithelium (lanes 2 through 6). Keratins are numbered according to catalogue of human cytokeratins' with consideration of polymorphic keratins. 1' . lS A) Coomassie blue-stained gel. A - actin. B-D) Corresponding immuno recognition patterns on nitrocellulose paper. B) Polyclonal rabbit antikeratin antiserum 10-2/2 exclusively reactive to neutral-to-basic keratins. C) Reprobing of Busing polyclonal antiserum 8-2/4 to display acidic keratins. D) Monoclonal antikeratin antibody KL-1. Arrow - ubiquitous keratinous contaminant" detected by monoclonal KL-I antibody in addition to its ability to react with keratins K4, K5, K6, and K8 as well as keratin KID (not shown; see Viac et al").

For routine histopathologic and immunohistochemical examinations, tissues were fixed in formalin and embedded in paraffin, while tissue samples for biochemical analyses were processed immediately or after storage at -70° C.

amide gel electrophoresis, two-dimensional separations employing the nonequilibrium pH gradient technique, and immunoblotting of separated polypeptides were performed as described. 14.15 Nitrocellulose filters were incubated with a set of polyclonal antikeratin antisera raised in rabbits against chromatographically purified keratins,":" the monoclonal antikeratin antibody KL-l, 17 or the monoclonal antivimentin antibody V9. 18 Secondary antibodies were horseradish peroxidase-conjugated goat anti rabbit immunoglobulins being developed with 4-chloro-l-naphthol and hydrogen peroxide as described. IS The mouse monoclonal primary antibodies were bridged via rabbit anti mouse immunoglobulins to the peroxidase-coupled antibodies.

Cytoskeletal Preparations. Tonsillar crypts were irrigated several times with ice-cold Ca 2+- and Mg2 +_ containing Hepes-buffered Earle's balanced salt solution with use of a blunt-ended needle in order to remove the cryptal contents (mainly detritus). Specimens were then heated in the solution at 60°C for 5 minutes to allow for easy dissection of the epithelium. I 3,1 4 After removal of all of the superficial epithelium, the irrigated crypts were incised and the cryptal epithelium was obtained under microscopic control with a small sharp spoon from an area as deep as possible to avoid contamination with surface epithelium. Epithelial tissue samples were homogenized and extracted in the presence of protease inhibitors as described. 14 The resulting Triton-high salt resistant pellet, highly enriched in keratins, was solubilized in sample buffer for gel electrophoresis.

Immunohistochemistry. Immunohistochemical analyses were performed according to the alkaline phosphatase-antialkaline phosphatase technique of Cordell et al. 19 The mouse monoclonal antibodies Ks13.1 (Dianova, Hamburg, Germany) and V9 (Dakopatts, Hamburg) were used to detect keratin K13 and vimentin, respectively.

Gel Electrophoresis and Immunoblot Analysis. One-dimensional sodium dodecyl sulfate polyacryl-

Keratin Expression Patterns in Tonsillar Epitheli-

RESULTS

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015

374

Mischke et al, Keratins as Markers of Epithelial Differentiation

J-NEPHG

S

D

S

A

Fig 2. Two-dimensional separation of cytoskeletal proteins obtained from A-D) surface epithelium and E-H,I-L) two different crypts of human palatine tonsils. A,E,I) Coomassie blue-stained gels. B,F) Corresponding immunoblots with polyclonal antiserum 10-212 to detect basic keratins. C,G) Reprobing of Band F with polyclonal antiserum 8-214 against acidic keratins to establish complete keratin pattern. D,H,L) KL-I monoclonal antibody. J) Antivimentin monoclonal antibody V9. K) Reprobing of J with antiserum 10-212. NEpHG - direction of first dimension using nonequilibrium pH gradient electrophoresis (basic polypeptides are to left), SDS - direction of second dimension in presence of sodium dodecyl sulfate. Keratins are numbered as described in Fig 1. Brackets enclose isoelectric variants of same protein. Coelectrophoresed marker polypeptides are bovine serum albumin (B; apparent molecular weight 68 kd; pI 6.35) and yeast phosphoglycerate kinase (P; apparent molecular weight 43 kd; pI 7.4). A - residual actin, v - vimentin, asterisks - "staircase" of spots typical of proteolytically affected vimentin molecules," arrows - ubiquitous keratin contaminants detected by some antikeratin antibodies.

urn. One-dimensional gel electrophoretic separa-

tions of cytoskeletal proteins prepared from the superficial epithelium of palatine tonsils (lane 1) and five different tonsillar crypts (lanes 2-6) are shown in Fig 1.9 • 14 • 15.1 7 . 20 Typically, the surface epithelium displayed only a few abundant proteins on Coomassie blue-stained gels, while the profiles of cryptal samples appeared to be more heterogeneous, occasionally lacking some of the proteins prominent in superficial epithelium and containing additional, although mainly minor, polypeptides (Fig lA). Identification of keratins was accomplished in the corresponding immunoblots: keratins of the neutral-to-basic subfamily, namely the members K4a,

K5b, and K6, were elicited by the polyclonal antikeratin antiserum 10-2/2 (Fig IB) and the members of the acidic subfamily, ie, K13, K14, K16, and K19, were detected on the same blot following reaction with antiserum 8-2/4 (Fig lC). In one, this immunoblot presented also the entire spectrum of keratins expressed in these specimens except for keratin K8, which was detected by the monoclonal antikeratin antibody KL-l (Fig ID). The keratin pattern shown for the surface epithelium of palatine tonsils (lane 1) is typical of the stratified squamous, primarily nonkeratinizing epithelium lining the major part of the upper digestive

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015

Mischke et al, Keratins as Markers of Epithelial Differentiation

~ kerali

4(avb) 5(a"b) 6

13 8

14 16

19

II

III

e:::::::3

c::::::=J

~

c:::::J

I

-- - --- - ---

--

- -

n

54

143

86

'10

19.1

50.5

30.4

Fig 3. Classification of keratin patterns in tonsillar crypt epithelium and their frequency. Schematic presentation according to one-dimensional separations. P - and/or, filled bar - amount as in surface epithelium of palatine tonsils, split bar - variably reduced amounts, empty bar - low to undetectable levels.

tract. 6 •7 •9,14,21 In contrast, the keratin patterns of tonsillar crypts were found to be much more variable, both in composition and in quantity of particular members expressed. Thus, we have found keratin patterns of crypt epithelium that were almost identical to those of the surface epithelium (lane 2), while in others the amount of K4 and K13 was reduced to a different (lanes 3 and 4) or even an undetectable (lanes 5 and 6) level. In addition, K6 was reduced in some samples, while K8 and K19, being hardly detected in superficial epithelium, were present in increased amounts but did not vary considerably among the different cryptal samples. Two-dimensional separations of cytoskeletal proteins of surface epithelium (Fig 2A_022) and of two different samples of crypt epithelium (Fig 2E-H and Fig 2I-L, respectively) allowed us to distinguish the keratins according to charge and size and provided the basis for the nomenclature used. 9 •14 Again, the major difference concerning the expression of keratins resided in the considerable reduction of K4 and K13 seen in crypt epithelium. Vimentin, the intermediate filament protein specific to mesenchymal cells, gave evidence of the contribution of lymphatic tissue to the samples taken from tonsillar crypts and was identified by the monoclonal antibody V9 (Fig 2E,I,],K). Its presence in cryptal samples does not, however, interfere with our conclusions outlined below.

Classification oj Keratin Patterns Expressed in Tonsillar Crypt Epithelium. On the basis of the analyses of 238 tissue samples of tonsillar crypt epi-

375

thelium obtained from 79 patients, three types of keratin patterns could be discerned (Fig 3). Type I patterns (eg, Fig 1, lane 2) were almost identical to the pattern of superficial epithelium and were found in about 20 % of the specimens. Type II patterns occurred most frequently (50 %) and displayed a variable reduction of K4 and K13 as exemplified in lanes 3 and 4 of Fig 1. Type III patterns were observed with a frequency of about 30 % and were characterized by very low to almost undetectable levels of these keratins (compare Fig 1, lanes 5 and 6). This variation apparently did not depend on the clinical state of the tonsils or the age of the patient; rather, we have found patterns belonging to different types within one tonsil and even along one particular crypt (see also below and Fig 4).

Expression oj Keratin K13 in Tonsillar Crypt Epithelium. Immunohistochemical analyses were performed to investigate how the alterations observed in the keratin patterns correlated to the morphologic facets of this epithelium. Figure 4 shows formalin-fixed and paraffin-embedded tissue sections of tonsillar crypts and the surface epithelium stained either with a monoclonal antibody that allowed us to monitor the specific distribution of K13 or with an antivimentin antibody. In surface epithelium the K13 antibody characteristically stained the suprabasal cell layers but not the basal cells (Fig 4B). This is quite similar to the staining pattern found in tonsillar crypts showing a regularly stratified epithelium (Fig 4F). In areas showing reticulation the epithelial organization is loosened and the epithelial cells are typically arranged in a netlike formation (Fig 4G). Among these, several cells still expressed keratin K13, but many others, usually located more basally, were negative (compare Fig 4C and 40). In reticulated epithelium the number of keratin K13-positive cells as well as the staining intensity generally decreased and thus led us to conclude that the reduction of keratin K13, and consequently its basic partner K4, is correlated with the stage of reticulation in tonsillar crypt epithelium. DISCUSSION

In the context of the now widely accepted notion that regards keratins as sensitive molecular indicators for the different routes and stages of epithelial differentiation.v" the present paper is concerned with the analysis of keratins in epithelial cells of human palatine tonsils, in particular those lining the tonsillar crypts. The crypt epithelium is, like the surface epithelium, a squamous nonkeratinizing tissue. The surface epithelium of palatine tonsils can be easily dissected in sheets from the underlying connective tissue upon heating." In contrast, the close association of epithelial and lymphatic tissue, characteristic of the lymphoepithelial organs belonging to the ring of Waldeyer and particularly re-

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015

376

Mischke et aI, Keratins as Markers of Epithelial Differentiation

...... ....

. G.

.... . (i~""""'la-r:'·",

Fig 4. Immunohistochemical detection of keratin K13 in human tonsillar crypt epithelium. Formalin-fixed, paraffinembedded sections reacted with K13 monoclonal antibody Ks13.1 (B,D-G) or vimentin antibody V9 (A,e) and detected by alkaline phosphatase--antialkaline phosphatase technique. A,B) Surface epithelium of palatine tonsil (original x500) and C,D) tonsillar crypt descending from surface as found in same specimen (original x330). E) Bottom of crypt showing high degree of reticulation and very few K13-positivecells as well as luminal detritus (original x350). F ,G) Cryptal areas from one specimen with F) almost regularly stratified epithelium (original xBOO) and G) typical netlike formation of epithelial cells (original x500).

fleeted in the reticulation of the epithellum.v':" did not permit us to retrieve pure epithelial tissue samples from the tonsillar crypts. However, the use of antibodies of known specificity in combination with high-resolution gel electrophoresis and immunoblotting allowed us to identify unequivocally the constituent water-insoluble (ie, intermediate) filament proteins of their cytoskeleton and revealed changes in the keratin pattern that reflected the various degrees of reticulation of tonsillar crypt epithelium. Since all of the patterns analyzed contained keratin K5 (a " b) (in accordance with the notion that the keratin pair K5/K14 may serve as a molecular marker for keratinocytes, the major cell type of stratified epithelia" 7.9,11), these proteins were used as a reference to adjust the sample loadings for equal staining on gels and immunoblots. In comparison to the superficial cell layers of tonsils, the cryptal epithelium showed qualitative and quantitative alterations that concerned particularly keratins K4 and K13. These two keratins were 1) present

in the same amounts as in surface epithelium, 2) reduced, or 3) undetectable. As the keratin pair K4/ K13 is considered to characterize biochemically the most advanced stage of differentiation in squamous nonkeratinizing epithelia lining the upper digestive tract,'·9.14 reduced levels may be indicative of an incomplete and/or disrupted epithelial differentiation in about 80 % of the tonsillar crypts sampled. An increased contribution of immature keratinocytes (ie, basal and parabasal cells) in the samples is also manifested by the moderate increase of keratins K8 and K19. Differentiation of tonsillar crypt epithelium may then be classified as ranging from the "surface type" to the "basal cell type." Moreover, since the differential expression of these keratins and, in particular, of keratin K13 could be correlated to the morphologic alterations that tonsillar crypt epithelium undergoes in the course of reticulation, 1.2.4.5 the different capabilities of tonsillar crypt epithelial cells to reach an advanced stage of differentiation can be determined by use of keratins as molecular markers.

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015

Mischke et al, Keratins as Markers of Epithelial Differentiation

That the observed spectrum of keratin expression patterns in tonsillar crypt epithelium resembles that of the epithelial lining of lateral cervical cysts" confirms our previous conclusion that lateral cervical cyst epithelia faithfully perform the differentiation programs typical of upper digestive tract squamous epithelia. With respect to the question of the origin of lateral cervical cysts, ie, whether they are con-

377

genital or acquired (see Wild et al " and references therein), our results then lend further support to our hypothesis that the inner lining of lateral cervical cysts is derived from Waldeyer's ring epithelial cells that settle and "transform" a neck lymph node to form an epithelium that is competent to imitate the crypts of the palatine tonsil and to cover part of their function, for example in antigen handling.

ACKNOWLEDGMENTS - WethankProfessor Jobst vonScheel forhiscontinuing Interest and encouragement. Dr Genka particularly acknowledges the supportprovided by Professor Yutaka NOOa (Head of the Department of Otolaryngology, University of Ryukyus, Okinawa, Japan). Part of thiswork wasdoneby Dr Genka In partial fulfillment of the requirements for the MDdegree of the Free University of Berlin.

REFERENCES 1. Fioretti A. Die Gaumenmandel. Stuttgart, Federal Republic of Germany: G Thieme, 1961. 2. Falk P, Mootz W. Morphologische Untersuchungen zur Retikulierung des Tonsillenepithels. Acta Otolaryngol (Stockh) 1973;75:85-103. 3. Brandtzaeg P. Immune functions of human nasal mucosa and tonsils in health and disease. In: Bienenstock J, ed. Immunology of the lungs and upper respiratory tract. New York, NY: McGraw-Hill, 1984:28-96. 4. Kawabata I. The surface structure of human tonsillar crypt. Scanning electron microscopic observation. Jpn J Tonsil 1978;17:204-10. 5. Howie AJ. Scanning and transmission electron microscopy on the epithelium of human palatine tonsils. J PathoI1980;130: 91-8. 6. Quinlan RA, Schiller DL, Hatzfeld M, et al. Patterns of expression and organization of cytokeratin intermediate filaments. Ann NY Acad Sci 1985;455:126-43. 7. Cooper D, Schermer A, Sun TT. Classification of human epithelia and their neoplasms using monoclonal antibodies to keratins: strategies, applications, and limitations. Lab Invest 1985; 52:243-56. 8. Osborn M. Summary: intermediate filaments 1984. Ann NY Acad Sci 1985;455:669-81. 9. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The catalogue of human cytokeratin polypeptides: patterns of expression of specific cytokeratins in normal epithelia, tumors, and cultured cells. Cell 1982;31:11-24. 10. Heid HW, Werner E, Franke WW. The complement of native a-keratin polypeptides of hair-forming cells: a subset of eight polypeptides that differ from epithelial cytokeratins. Differentiation 1986;32:101-19. 11. Eichner R, Sun TT, Aebi U. The role ofkeratin subfamilies and keratin pairs in the formation of human epidermal intermediate filaments. J Cell Bioi 1986;102:1767-77.

12. Hatzfeld M, Franke WW. Pair formation and promiscuity of cytokeratins: formation in vitro of heterotypic complexes and intermediate-sized filaments by homologous and heterologous recombinations of purified polypeptides. J Cell Bioi 1985;101:182641. 13. Milstone LM, McGuire J. Different polypeptides form the intermediate filaments in bovine hoof and esophageal epithelium and in aortic endothelium. J Cell Bioi 1981;88:312-6. 14. Wild GA, Mischke D. Variation and frequency of cytokeratin polypeptide patterns in human squamous non-keratinizing epithelium. Exp Cell Res 1986;162:114-26. 15. Mischke D, Wild G. Polymorphic keratins in human epidermis. J Invest DermatoI1987;88:191-7. 16. Wild GA, Wille G, Mischke D. Lateral cervical (branchial) cyst epithelia express upper digestive tract-type cytokeratins. Polyclonal antibody studies. Ann Otol Rhinol Laryngo11988; 97:365-72. 17. Viac J, Staquet MJ, Thivolet J, Goujon C. Experimental production of antibodies against stratum corneum keratin polypeptides. Arch DermatoI1980;267:179-88. 18. Osborn M, Debus E, Weber K. Monoclonal antibodies specific for vimentin. Eur J Cell Bioi 1984;34:137-43. 19. Cordell JL, Falini B, Erber WN, et al. Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP complexes). J Histochem Cytochem 1984;32:219-29. 20. Ochs D. Protein contaminants of sodium dodecyl sulfatepolyacrylamide gels. Anal Biochem 1983;135:470-4. 21. Ouhayoun JP, Gosselin F, Forest N, Winter S, Franke WW. Cytokeratin patterns of human oral epithelia: differences in cytokeratin synthesis in gingival epithelium and the adjacent alveolar mucosa. Differentiation 1985;30:123-9. 22. Geisler N, Weber K. Amino acid sequence data on glial fibrillary acidic protein (GFAP): implications for the subdivision of intermediate filaments into epithelial and nonepithelial members. EMBO J 1983;2:2059-63.

Downloaded from aor.sagepub.com at Purdue University Libraries on July 4, 2015