Type I Keratinocyte Transglutaminase: Expression in Human Skin and Psoriasis

Wanda T. Schroeder, Scott M. Thacher, Shelley Stewart-Galetka, Mary Annarella, Deidra Chema, Michael J. Siciliano, Peter J.A. Davies, Hsiao-Yuan Tang, Blair A. Sowa, and Madeleine Duvic Departments of Dermatology and Internal Medicine (WTS, SS-G, MA, DC, MD) Pharmacology (PJAD), University of Texas Medical School , Houston, Texas; Department of Medical Biochemistry and Genetics (SMT, H-YT), College of Medicine, and Department of Veterinary Pathobiology (BAS) , College of Veterinary Medicine, Texas A&M University, College Station, Texas; Department of Molecular Genetics (MJS), and Section of Dermatology (MD), M.D. Anderson Cancer Center, Houston, Texas, U.S.A.

A 92-kD trans glutaminase (TGase K), expressed in human cultured keratinocytes and stratum corneum, catalyzes a criti­cal step in the formation of the cornified envelope of terminal differentiation. A rabbit polyclonal antibody to TGase K was us.ed to isolate overlapping eDNA clones from a human ker­atl110cyte eDNA expression library. The eDNA clones were sequenced and unequivocally identified as TGase K by com­parison to the N -terminal amino acid sequences of two cyano­gen bromide fragments from the purified enzyme. The mRN A for T gase K is expressed in cuI tured keratinocytes but not in A431 squamous carcinoma cells, in fibroblasts, or in other non-epithelial tissues and cells. Although TGase K pr?tein expression is limited to the upper layers of normal epidermis, the mRNA is generally present throughout the

T erminal differentiation of epidermal keratinoeytes in­volves formation of a covalently crosslinked protein layer beneath the cell membrane. This step is cata­lyzed by the enzyme transglutaminase, which links adjacent proteins by formation of the N-(y-gluta­

myl)-lysine isopeptide bond [1]. A keratinocyte transglutaminase , also referred to as type I transglutaminase, or TGase K, has been Identified in cultured epidermal keratinocytes and related cell types as a 92~kD protein [2 - 4]. It is largely membrane bound and is found rrllnanly in stratified squamous epithelia, where its expression is Imlted to the more differentiated cell layers [2,4,5]. A mRNA from

rabbit tracheal cells has been described that contains sequence ho-mology to other more extensively described transglutaminases [6] .

epidermis, suggesting the possibility of post-transcriptional regulation. Precocious expression ofTGase K protein occurs in psoriasis, and quantitative Northern blot analysis ofTGase K mRNA from normal and involved epidermal biopsies from psoriasis patients suggests that TGase K mRNA levels are incre~sed in psoriat~c lesions. By using quantitative laser scannmg confocallnlcroscopy (LSCM) and in situ hybridiza­tion, the increase of the TGase K mRNA was in the range of 3 - 7 times in the psoriatic epidermis and was significantly higher compared with normal skin and with paired adjacent skin. Quantitative LSCM provides a powerful and direct method for analysis of gene expression in skin. ] Invest Der­matoI99:27-34,1992

Its transcription is regulated negatively by retinoids and positively by Ca++ concentration as described for TGase K activity [6].

At least one other trans glutaminase activity has been reported in epidermis. This enzyme, TGase E, is soluble, requires proteolytic activation, and ap£ears to be distinct from TGase K in physical characteristics [7,8 j. Tissue transglutaminases have recently been cloned and are expressed in macrophages and endothelial cells [9,10]. In this study we describe a partial human cDNA for TGase K, which we have established unequivocally as the membrane­bound enzyme expressed in cultured cells by partial amino acid sequencing of the purified protein. W e show that although the mRNA is expressed in a tissue-specific pattern, it is unexpectedly found throughout normal epidermis rather than at the upper layers

---------------------------------------------------------------------------------------------------------------Manuscript received December 31, 1991 ; accepted for publication Febru­

ary 27, 1992. AI[his work was supported in part by NIH grants RD36546, AR39915 , and 35 40520 (MD), TAES H-6194 and AG1001 (BS), and ACS grants CD-

2, and CD-352A (ST) . Dr. Schroeder was the recipient offe llowship support from the Dermatol­

ogy Foundation sponsored by Roche Pharmaceuticals and Squibb Laborato­ries.

This work was presented in part at the Society for Investigative Dermatol­igYdrneetll1gs, in Seattle, Washington, May 1991 and in Baltimore, Mary­an ,April 1992.

:eprint requests to: Dr. Madeleine Duvic, Departments of Dermatology ~1\8Internal Medicine, Univers ity of Texas Medical School , 6431 Fannin, . 4, Houston, Texas 77030. Abbreviations:

Ca++: calcium

eDNA: complementary deoxyribonucleic acid CNBr: cyanogen bromide cRNA: complementary ribonucleic acid IF: immunofluorescence LSCM: laser scanning confocal microscope mRNA: messenger ribonucleic acid Nl Ps: normal skin from psoriasis Oligo: oligonucleotide PEG: polyethylene glycol Ps: psoriasis lesion SDS: sodium dodecyl sulfate SSC: sodium citrate buffer TGase E: epidermal transglutaminase TGase K: keratinocyte transgl utaminase UV: ultraviolet

0022-202X/92/S05.00 Copyright © 1992 by The Society for Investigative Dermatology, Inc.

27

28 SCHROEDER ET AL

of the stratum granulosum as expected by the distribution of the protein. Compared with adjacent normal skin and control skin, the mRN A is increased in psoriasis lesions.

MATERIALS AND METHODS

Isolation and Description ofTGase K cDNA Clones TGase K clones were obtained by screening a lambda gt11 expression li­brary prepared from poly A+ mRNA from keratinocytes cultured in 1 mM calcium (Clonetech) [11] using a rabbit polyclonal antibody raised against purified human TGase K [1 2]. Out of2 X 106 plaques screened, seven positives were chosen for further characterization by epitope selection [13]. The fusion protein expressed by a cDNA clone containing a 0.9-kb imert (pKTG900) bound antibody that specifically recognized purified human TGase K by Western blot­ting [14]. Rescreening with pkTG900 identified cDN A clones with Eco Rl inserts of 1.5 kb (pKTG1500) and 0.66 kb (pKTG660), which were subcloned into the pGEM-3Z vector (Strategene). DNA sequencing [15] on both strands was performed and homol­ogy with other transglutaminases was found using W ordsearch [16]. pKTG 1500 began at the second amino acid G of the ac tive site (YGQCWVFA) and contained the pKTG660 clone within the 3' sequences. The sequence was deposited in EMBL (accession number X57974) and is identical to other published TGase K cDNA se­quences [17 -20]. pKTG1500 mapped to Chromosome 15 by hy­bridization to hamster-human hybrid panels, by methods described previously [21], confirming the map assignment reported by Pola­kowlska et al [22].

Protein Purification and Sequencing The identity of the TGase K clones was established by protein sequencing. TGase K was purified by a modification of a monoclonal antibody affinity­chromatography [23], in which the purified monoclonal antibody B.Cl [2] was coupled to N-hydroxysuccinimide biotin (Pierce) and attached to avidin-agarose [24]. A non-ionic detergent extract of confluent human foreskin' keratinocytes [2] was passed over the column and TGase K eluted as described [17]. TGase K was concen­trated by precipitation in 10% polyethylene glycol (PEG) 6000 (Sigma), purified on an SDS gel, and cleaved by cyanogen bromide. Individual fragments were submitted to Edman degradation on an Applied Biosystems 470A gas-phase sequencer following electro­blotting of SDS-gel-separated bands to an Immobilon PVDF membrane [25].

Northern Analysis Human keratinocytes, fibroblasts, mela­noma cells, and A431 squamous carcinoma lines were cultured as reported previously [26]. Total mRNA was extracted with guani­dine isothiocyanate [27] . Shave biopsies taken from established pso­riasis vulgaris plaques and normal skin within several centimeters from the lesions were processed immediately in RNAsol (Biotecx) for total RNA extraction [28]. Total RNA was quantitated by UV spectroscopy. RNA quality and size was analyzed by ultraviolet shadowing [28]. RNA was electrophoresed through agarose for­maldehyde gels, transferred to Zetabind membranes (Cuno), and hybridized to random primer oligo labeled pKTG 1500 insert [26,30]. The Northern blots were rehybridized to an oligolabeled GAPDH cDNA probe [31] or to an end-labeled 28S RNA oligo (Oncogene Sciences) as a standard. The filters were assessed by direct counting for 60 min on a Betagen scanner [3~] and by autora­diography against Kodak XAR 5 film for 1 - 10 d. An RNA ladder (Bethesda Research Laboratories) was used as a molecular size stan­dard.

Immunofluorescence and Immunoperoxidase Staining Fresh frozen tissue sections of normal cheek, forearm, and foreskin were incubated with B.Cl monoclonal antibody specific for human TGase K and detected using fluorescein-labeled goat anti-mouse IgG Qackson Immunolaboratories) as described [2]. Psoriasis lesions were stained usin g B.Cl monoclonal detected with biotinylated goat-anti mouse and the avidin biotinylated peroxidase complex method (Vectastain ABC kit) as described [33J. N egative controls were performed with mouse serum and goat anti-mouse IgG 1 (kindly provided by William Geoghegan) .

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

ATGCCAGTGGCCTACAAGGAATACCGGCCCCATCTTGTGGACCAGGGGGCC 274 M P V A Y KEY R P H L V D Q G A KTG3 P V A Y KEY P L V Q

ATGCTGCTCAATGTCTCAGGCCACGTCAAGGAGAGCGGGCAGGTGCTGGCC 292 M L L N V S G H V K E S G Q V L A KTG5 L L N V S V K V L A

Figure 1. Comparison of deduced amino acids with protein sequences. Amino acid numbering begins at the first codon of the active site (YGQCWVFA). The amino acids (lille 2) deduced from nucleotide se­quences (lille 1) show good correlation with the sequenced CNBr peptide fragments (lille 3) . The N-terminal amino acid sequence ofKTG3 (M Wt. 21.3 kD) is PVAYKEY-P-LV-Q--M. KTG5 and KTG6 (M Wt 19.6 and 19.1 kD) gave identical N-terminal sequences: LLNVS--VK----VLA. At positions indicated by a dash, the amino acid residue was not clearly identifi­able.

In Situ Hybridization The method of in situ hybridization de­scribed by Stoler et al [34] was used on paraformaldehyde-fixed, paraffin-embedded sections from normal human cheek, forearm, foreskin, and from psoriasis lesions paired with adjacent skin. In the first experiment, normal skin specimens were compared to psoriasis lesions from different patients. In the second experiment, psoriasis lesions were compared with adjacent normal skin from the same

1 2 3 4 5

2.9 kb

1.3 kb

Figure 2. Northern blotting shows tissue-specific expression in vitro. 30 fJ.g per lane of total RNA extracted from cultured A431 sqnamous carci­noma cells (lalle 1) , human foreskin keratinocytes in 1 mM Ca++ (lalle 2), human diaphragm tissue (lalle 3), WM9 melanoma cell line (lalle 4) , and cultured human fibroblasts lalle 5) were electrophoresed on formaldehyde agarose gels, blotted onto Zetabind membranes, and hybridized with o ligo­labeled cDNA insert. GAPDH control cDNA probe is shown in the bottom pallel. The 2.9-kb mRNA hybridizing to the TGase K cDNA probe was seen only in cultured keratinocytes (lalle 2) and was not seen in the other ce lls that did express the 1.3-kb GAPDH mRNA band. Markers used were the 1-kb ladder (BRL).

VOL. 99, NO. 1 JULY 1992

patient. Sense and antisense riboprobes were used on each biopsy specimen.

Riboprobes were synthesized from linearized pKTG660 using th.e Riboprobe Gemini system (Promega) and 35S-labeled uridine tnphosphate (Amersham). The pKTG660 3' clone was used be­cause it lacks the active site that is highly conserved among all transglutaminases and gave only one mRNA band on Northern blotting. cRNA were then hydrolyzed and hybridized to sections at 55 a C for 3 h. Following hybridization, sections were washed in 50% formamide, 2 X SSC for 10 min at 55 a C and again for 20 min at 55 a C. Slides were dipped in Kodak NB2 emulsion, developed for 6 d (experiment 1) or 21 d (experiment 2). Slides were counter­stamed with Wright's Giemsa.

I":Iage Analysis Analysis of in situ hybridization was performed usmg a Biorad 600 Laser Scanning Confocal Microscope (LSCM) as described [35,36]. Briefly, the images were viewed by transmitted Itght (tissue section) in parallel with the confocal reflected mode (in Situ hybridization signal). The reflected confocal image was visual­Ized in a 0.5-micron plane within the upper emulsion, representing the specific in situ hybridization signal. The transmitted and confo­cal Images were then merged to localize the signal within areas of the skin.

The digital analysis system integrated with the Biorad LSCM was used to quantify intensity of the in situ hybridization signals in normal and psoriatic skin biopsies. Hybridization signals repre­sented by reflected intensity (pixels) within areas ranging from 15,000 to 25,000 square microns were determined from the confo­cal on the line image (100 X magnification) from four to six separate areas per field in two or three sections from each lesion. The mean Intensities in pixels per square micron and standard deviations were falcu lated for each area. Intensity readings taken from the upper and OWer epidermis of psoriatic lesions were generally similar. Specific

mean intensities for each lesion were obtained by subtracting the sense from the antisense signals. Analysis of variance was used to analyze the sampling of individual lesions and the Student-New­man-Keuls (two-tailed) t test was used to identify statistically signifi­c.ant differences between individual sense and antisense hybridiza­tion signals and between psoriasis and paired normal skin [37].

TGase K INCREASED IN PSORIASIS 29

RESULTS

Nucleotide and Protein Sequences We confirmed our clones' identities as TGase K by protein sequence determination. The N­terminal sequences of three CNBr fragments of TGase K were obtained by protein microsequencing at the lowest limit of sensitiv­ity (10 pmoles/residue or less) and showed excellent correspon­dence to the predicted amino acid sequence (Fig 1). The CNBr fragment KTG3 includes the two smaller fragments sequenced, KTG5 and KTG6. The sizes of these fragments as estimated by SDS-PAGE [19] were 21.3,19.6, and 19.1 kD. Their relative sizes are in agreement with the sizes predicted from the eDNA (17.9, 16.0, and 15.4 kD). The difference between KTG5 and KTG6 is due to partial CNBr cleavage at a methionine close to the C-ter­minal. This is the first TGase K protein sequence reported.

Expression of mRNA is Specific for Epidermal Cell In Vitro In order to determine the TGase K mRNA size and tissue distribution, all three eDNA inserts were hybridized to Northern blots containing total mRNA extracted from cultured skin cells (Fig 2) . All three hybridized to a 2.9-kb mRNA from foreskin keratino­cytes cultured in 1 mM Ca++ but not in human vulvar squamous carcinoma A431 cells, WM9 melanoma cells, cultured human fibro­blasts, or human diaphragm tissue.

Expression of Protein in Normal Human Epidermis and Psoriasis Keratinocyte transglutaminase protein has previously been shown to be present in a narrow distribution in the stratum granulosum just under the stratum corneum [23]. In Fig 3A the restricted pattern of protein expression as determined by immuno­fluorescence using the B.Cl mouse monoclonal antibody [17] is shown m normal cheek skin, and is representative. No epidermal staining was seen using the secondary antibody alone (Fig 3B). It has previously been reported that theenzyme becomes active in living skm, as demonstrated by Its ability to use dansyl-cadaverine as a substrate, in an even more restricted region of the stratum granulo­sum [38].

In wound healing and in psoriasis, we and others have reported precocious expression ofTGase K protein throughout the epidermis [23,39,40]. As shown in Fig 3C, immunoperoxidase staining of

c

p' d~~ure 3. Expression ofTGase K protein in normal skin and psoriasis. A) Immunofluorescence detection ofTGase K with the E.C! monoclonal antibody stra:

cted With fluorescein-conjugated goat anti-mouse IgG. There is a narrow band of fluorescence (left) in the upper stratum granulosum just under the

spec~rn ~orneum in normal skin specimens studied. (Magnification X 400.) B) Same as specimen A with fluorescein-conjugated goat anti-mouse IgG shows no broa~ c uor.escence. There is co llagen autofluorescence in the dermis (X 400). C) Immunoperoxidase detection of TGase K in psoriatic lesions shows a layers~r distribution than normal skin as represented by the black staining around keratinocytes in the upper epidermis. Protein staining is not seen in the basal

30 SCHROEDER ET AL

;- J •

"', I

. '

Figure 4. Expression ofTGase mRNA by in situ hybridization in normal skin. A) Normal human cheek skin with epidermis (at left sjde) and hair follicles (righI, arrows) in transmitted mode hybridized with antisense TGase K cRNA probe pKTG660. (Magnification X 100.) B) Reflected confocal image of section A. The specific hybridization overlying the epidermis is more clearly revealed in the reflected image as a band of tiny white dots and in the dermal hair follicles (X 100). C) Normal human check section (A) with sense negative control probe, transmit­ted image. D) Reflected image of C shows no specific hybridiza­tion (X 100). E) Transmitted image of human neonatal foreskin hybridized with antisense TGase K cRNA probe (X 100). F) Reflected image of E shows narrower distribution of the TGase K hybridization signal than in section of cheek skin seen in A,B.

,. -.~~ (

..

,~ ... t

"

..

.~ , " ..

,. f ' ,

psoriatic epidermis has a broader distribution than in normal skin, but the enzyme does not appear in the basal layer. Secondary anti­body alone gave no staining (data not shown). It was therefore of interest to determine the spatial expression of TGase K mRNA in normal human skin and in psoriasis.

Expression ofTGase K mRNA and Protein is Discordant in Normal Skin and Psoriasis Using in situ hybridization we first studied the expression ofTGase K mRNA in normal cheek skin (Fig 4A - D), in neonatal foreskin (Fig 4E,F) , in normal forearm skin (not shown), and in psoriasis lesions (Fig 5) and paired normal skin samples (not shown). T he sense hybridization signals as negative controls are shown in Fig 4C,D and SC,D . The transmitted tissue images (4A,C,E) that show the morphology best are compared to the reflected signals (4B,D,F) that show the in situ hybridization signals as white dots. Specific hybridization due to the antisense TGase K probe is shown in Fig 4A,B,D,E.

I,

, It · ,-

" • . 1

~ ~ ' " .. \- ' ., .,.. ~ . \ . ~ ... &~ / • II

~/ ' .

I ,

I ~, U.

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

In contrast to the more restricted pattern of immunoflu orescence staining for TGase K protein (Fig 3A) [2,5,23,38]' mRNA (as judged by in situ hybridization) is observed throughout the epider­mis in normal skin specimens. Because the normal cheek skin was removed adjacent to a basal cel l carcinoma, we were concerned that the expression of TGase K might be abnormal, as reported for an­other differentiation marker, involucrin [41]. However, in other specimens of normal sk in examined including two forearms and three normal skin biopsies from trunk and extremities within 3 cm from psoriatic plagues, a similar distribution was found. The one exception noted was the neonatal foreskin (shown in Fig 4E,F), where the mRNA appeared to be concentrated more strongly in a band- li ke distribution at the upper epidermis as expected by the protein distribution.

Interestingly, given the recent report of keratin 14 expression in hair follicles [42], we noted that the expression ofTGase K mRNA was most prominent in the outer root sheath areas of follicles

VOL. 99, NO. 1 JULY 1992 TGase K INCREASED IN PSORIASIS 31

Figur~ 5. Expression ofTGase K mRNA in psoriasis lesions by in situ hybridization. A) Psoriasis lesion (lesion 1, Table J) hybridized with antisense cRNA, t7nsmItted image (magnification Xl 00) . B) Reflected image oflesion in F shows very intense signal throughout the epidermis (X 100). C) Transmitted image o( another section of same psoriasis lesion hybridized with sense TGase cRNA probe (X 100). D) Reflected image of H shows lack of specific hybridization >(100).

Imaged in the upper dermis (Fig 4A,B), but appeared equal in inten­Sity to the epidermal signal. The other areas of the dermis, exclud-1i1g hair follicles, were not above background in the hybridization signal.

' . The LSCM digital imaging system allows the confocal reflected 1i1 Situ colorized image to be superimposed on the transmitted Image. This showed that in normal skin sections (forearm and

cheek), the cells in the basal layer, as well as that in the upper layers, contained TGase K mRNA. Therefore, constitutive expression of the TGase K message may occur throughout the epidermis without translation, and a signal for translation may occur at the level of the stratum granulosum. Alternatively, there may be very minute amounts of protein made in the basal byers that are not detectable using IF or immunoperoxidase methods.

Table I. Quantitative In Situ Hybridization from Nonnal and Psoriatic Skin Lesions· ------------________________ 2-______________ ~ __________________________________________________________ _

Anti-Sense Sense Anti-Sense - Sense Hybridization Signalb

Experiment 1 --------~------------------------------------------------------------------------~----~----------

Mean SD Mean SD Specific Hybrid

Normal skin 1 Ps lesion 2 Ps lesion

Experiment 2 3 Ps normal skin 3 Ps lesion 4 Ps normal skin 4 Ps lesion 5 Ps normal skin 5 Ps lesion 3,4,5 normal skin

1.38 0.62 7.73 2.42 3.88 2.12

7.28 0.85 10.48 1.89 7.88 0.78

17.83 4.76 7.04 0.28 9.52 2.53 7.40 0.43

12.61 4.55

0.23 0.23 1.15 0.15 0.13 7.58' 0.03 0.05 3.85'

5.93 0.24 1.35 5.91 0.39 4.57J

7.93 1.27 -0.05 7.49 0.89 10.34J

5.89 0.25 1.15 5.43 0.27 4.09J

6.58 1.17 0.82 6.28 1.08 6.33J _________ 3,4,5 psoriasis

'p ~----~-------------------------------------------------------------------------------• 5, pSoriasis.

Hybridization' 1 d . . . . 1 . f C • fi ld . . b' t P < 0 05 . s l~na cxprcssc as mean mtenslty In plxe s per square mJCron rom rour to SIX e S III two or nlore sections per tOpsy. J p < O· psor~as~s versus control normal skin .

. 05 pSonasls versus paired normal skin.

32 SCHROEDER ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table II. Expression ofTGase K mRNA by Northern Blotting"

Lane 2 3 4 5 6 7 8

Source NL PS NL PS NL PS Fibroblasts Keratinocytes TGase K 1.6 12.9 3.4 19.0 1.8 25.7 1.7 7.5 Normalized 4.7 97 37 79 5.9 79 3.6 19 Increase 20 X 2X 13X 5X

, The signals of the bands shown in Fig 5 were counted (cpm) and the background signal from the filter was subtracted. The TGase K signal was normalized by dividing by the 28S signal and multiplying by 1,000. NL, normal; PS, psoriasis.

Expression ofTGase K mRNA is Increased in Psoriasis Les­ions Psoriatic epidermis (Fig SA,B) showed more intense hybrid­ization compared with normal skin (4A,B) when hybridized in par­allel under identical conditions. The expression of the mRNA was also evident throughout the epidermis, including the basal layer. The in situ hybridization signal appears visually more intense per unit area in psoriasis than in normal skin.

To quantitate the intensity of hybridization within a small epi­dermal unit area, we used the digital image analysis system inte­grated with the Biorad LSCM. The results of two experiments are shown in Table 1. Psoriasis lesions (lesion 1 is shown in Fig 5) were compared with normal skin (Fig 4A,B) in experiment 1. In experi­ment 2, three paired lesions and normal adjacent skin biopsies from three additional patients were compared (in situ images not shown) . The two experiments cannot be compared with one another because different labeling reactions and exposure times were used. How­ever, in both experiments psoriatic epidermis had 3 - 7 -times greater hybridization intensity compared to normal or paired normal skin. The differences were statistically significant (Table I). Further­more, the numbers were consistent over several different fields and sections within biopsies, as determined by analysis of variance [30].

In order to confirm these findings by a more traditional method, quantitative Northern blotting was used to compare paired normal and involved skin biopsies from three additional patients with psori­asis. As shown in Table II, there was a 3 - 20-times greater hybridiza­tion signal to TGase K in psoriasis skin versus adjacent normal skin. By traditional autoradiography, shown in Fig 6, the signal on the filter from normal skin is almost undetectable at 20 h exposure (row B) but can be visualized if the blot is exposed for 10 d (row A) . The psoriasis signals are very intense, even at 20 h, and are overexposed at 10 d. Overexposure would prevent accurate quantitation by densi­tometry but is avoided by direct scanning of radioactive emission

1 2 345 6 7 8 9

A

B

c Figure 6. Quantitative Northern blotting shows increased TGase K mRNA in psoriasis lesions. Total RNA (5 Jig/lane) from uninvolved adja­cent normal skin (fanes 1, 3, and 5), psoriatic plaques (folies 2, 4, and 6), cultured fibroblasts lane 7) and keratinocytes (lane 8) was analyzed by Northern hybridization to oligo-labeled pKTG660 insert (1 X 109 cpm/ Jig). The filter was exposed to Kodak XAR film for 40 h (row B) and for 10 days (row A), stripped and rehybridized to an end-labeled 40-mer specific for 28S ribosomal RNA (Oncogene Sciences) (row C). Lalle 9 contains th e RNA ladder (BRL) as a marker. TGase K mRNA is 2.9 kb in size between the 4.4-and 2.3-kb marker bands. Quantitation of the signals shown in this figure arc shown in Table II.

from the filter. The 20-times difference seen between lanes 1 and 2 (Fig 6) representing psoriasis lesion 4 (Table II) may result from an underestimation of the control hybridization signal (row C) used to

normalize the values. The combination of quantitative in situ hy­bridization using the LSCM and quantitative Northern blotting using a betagen scanner may be useful for studying the regulation of other epidermal genes.

DISCUSSION

We identified a partial eDNA clone for keratinocyte transgluta­minase K (TGase K) in order to better understand the regulation of this enzyme critical for epidermal differentiation. Because of the reported presence of two transglutaminases in epidermis in addition to tissue transglutaminase [7 - 10,18]' we confirmed our identifica­tion of the eDNA clone by protein sequencing of cyanogen bromide fragments of membrane-bound TGase K affinity purified by the antibody B.C1. The identities of previous clones reported forTGase K [6,16 - 20] have been inferred by their homology to other TGase and by a pattern of mRNA expression that correlated with that reported for membrane-bound TGase K protein. For the first time, we provide unequivocal evidence that the reported clones corre­spond to TGase K.

An unexpected finding of this study was the distribution of the in situ mRNA signal in a pattern that is spatially broader than the restricted pattern of protein expression in the stratum granulosum previously reported by independent investigators [2,5,23,38]. The expression of the TGase K mRNA in the lower levels of the epider­mis, including basal layers, suggests that the expression of the en­zyme may be regulated post-transcriptionally. Regional differences in the expression of the mRNA may exist, as the pattern in foreskin was more restricted than in cheek, forearm, or trunk normal skin. It is also possible, however, that the restricted localization of the pro­tein itself is due to occlusion of the epitope recognized by RCI antibody in the lower layers of the epidermis. It is of note that B.Cl antibody does react to TGase K even in the absence of calcium. Dermal hair fo llicles were also similar in their in situ hybridization signal to epidermis but the other parts of the dermis showed hybrid­ization levels that were not above background. Within hair follicles seen in the upper dermis, the TGase K signal was somewhat more prominent in the outer root sheath area (Fig 4A,B).

We have shown by two independent methods that the epidermal expression ofTGase K mRNA is increased in psoriasis and demon­strated that laser scanning confocal microscopy may be a useful tool for quantitation of mRNA in skin biopsies. A major disadvantage of studying psoriasis biopsies by quantitative Northern blotting is con­trolling for the amounts of RN A loaded per lane, as well as for the proportion of mRNA obtained from epidermis versus dermis from shave or punch biopsies [43]. Therefore, one could interpret the increased TGase K mRNA found by Northern blots to be due solely to the increased proportion of the epidermal component. However, when similar unit areas of epidermis were compared for the inten­sity of in situ hybridization signal, the psoriasis biopsies consistently were several times stronger than normal skin, even in paired sam­ples, and the differences were statistically significant. The use of a Betagen scanner also improves the quantitation of Northern blot­ting, as it provides a direct analysis of the filter without film and densitometry [32] .

TGase K, as a catalyt ic enzyme for cornified envelope formation, participates in terminal differentiation of the squamous epithelium

VOL. 99. NO. 1 JULY 1992

[2.6). Calcium (> 1.0 mM) and phorbol esters are agents know n to Induce the expression of the enzyme and to induce terminal differ­entiation [44 ,45) . Retinoic acid, although upregulating tissue transglutaminase [46]' inhibits keratinocyte trans glutaminase ex­pression [6,44,45,47,48J. Transforming growth factor beta 1 upre­gulates tissue transglutaminase, but has no effect on T Gase K [49]. The mechanism of retinoic acid inhibition of the enzyme in culture has been suggested to be pre-translational [50] . The discordant ex­pressIOn of mRNA and protein that we observed has not been de­scribed in an in vitro system and requires further investigation.

It IS not yet known w here the primary defect occurs in psorias is: whether keratinocytes are intrinsically hyperproliferative and have Incr~ased sensitivity to growth fac tors or whether growth factors, possIbly T GF-alpha [51 J and interleukin 6 [52]' can induce the phenotype. There is growing recognition that the immune system and 1l1.f1ammation can play a major role through the generation of cytokines [53J.

Inc.reased expression of TGase K protein and mRNA in psoriasis may In part explain the hyperkeratosis characteristic of psoriatic lesIons. The formation of a more highly cross-linked stratum cor­neum may lead to greater retention of the squames observed in leslonal skin. It is also possible that retention of the stratum cor­neum could change the normal regulation of other epidermal genes. Alternatively and more likely, increased TGase K expression in psonasls may resu lt secondarily from the dysregulation of the nor­mal epidermal differentiation pathway [39J. Further study of the upregulation of T Gase K may contribute to understanding the pathogenesis of psoriasis. It is tempting to speculate that one benefi­cIal effect of retinoids for psorias is may result from their inhibition ofTGase K. This hypothesis can be tested experimentally with the technIques we have described here.

We t!' all k Dorothy Lewis, LOllis Smitlr, alld RobertJordoll fo r helpfll l disCIIssioll, Pat ShqJield, Willa Collill s, alld Kat lrerine Kelly fo r tecllllica l assistallce, alld Jea t/eue ~y fo r preparatiot/ of tire mat/ uscript.

REFERENCES

1. Eckert RL: Structure, function, and differentiation of the keratinocyte. Phys iol Rev 69 :131 6- 1346, 1989

2. Thacher SM , Rice RH: Keratinocyte-specific transglutaminase of cul­tured epidermal cells: relation to cross-linked envelope fo rmation and terminal differentiation. Cell 40:685 - 695, 1985

3. Shirley JE, J etten AM: C haracterization of transglutaminase activity in rabbit tracheal epithel ial cells. Regulation by retinoids. ] BioI Chem. 261:15097 -1 5101 , 1986

4. Schmidt R, Michel S, Shroot B, Reichert U : Transglutaminases in normal and transformed human keratinocytes in culture. ] Invest Dermatol 90:475 - 479, 1988

5. Parenteau NL, Pilato A, Rice RH: Induction of keratinocyte type-I transglutaminase in epithelial cells of the rat. D ifferentiation 33:130 - 141,1 986

6. Floyd EE,] etten AM: Regulation of type I (Epidermal) transglutamin­ase mRNA levels during squamous differentiation: down regulation by retinoids. Mol Cell Biochem 9:4846-4851,1989

7. Park SC, Kim SY, Kim HC , Thacher S, C hung SI: Differential expres­Sion of transglutaminases in human foreskin and cultured keratino­cytes (abstr) . ] Cell BioI 107:139A, 1988

8. Kim HC, Lewis MS, Gorman JJ , Park SC, Girard ]E, Folk JE, C hung SI: Protransglutaminase E from guinea pig skin: isolation and partial characterization. ] BioI C hem 265:21 971 -21978, 1990

9. Grundmann U , Amann E, Zettlemeissl G, Kupper HA: C haracteriza­tion of eDNA coding for human factor XlIIa. Proc Nat! Acad Sci

10. USA 83:8024 - 8028, 1986 Gentile V, Saydak M, C hiocca EA. Akande 0, Birckbichler P] , Lee

KN , Stein ]P, Davies PJA: Isolation and characterization of cDNA clones to mouse macrophage and human endothel ial cell t issue

11. transglutaminases. ] BioI C hem 266:478-483, 1991 YOung RA, Davis RW : Yeast RNA polymerase II gene: isolation with

antIbody probe. Science 222:778 - 782, 1983

TGase K INCREASED IN PSORIASIS 33

12. Podda r S, Hong WK, Thacher SM, Lotan R: Beta trans retinoic acid suppression of envelope competence, epidermal transglutaminase, involucrin and keratin Kl in human head and neck squamous cell carcinoma cells. Int] Cancer 48:239 - 247, 1991

13. Smith DE, Fisher PA: Identification, developmental regulation, and response to heat shock of rwo antigenically related forms of a major nuclear envelope protein in drosophia embryos: application of an improved method for affini ty purifica tion of antibodies using poly­peptide immobilized on nitrocellu lose blots. ] Cell BioI 99:20 - 28, 1984

14. T owbin HT, Staehlin T , Gordon]: Electrophoretic transfer of pro­teins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc N atl Acad Sci USA 76:4350 -4354, 1979

15. Sanger F, Nicklen S, Coulson AR: D N A sequencing with chain-ter­minating inh ibitors. Proc N at! Acad Sci USA 74: 5463 - 5467, 1977

16. Devereux], H aeberli P, Smithies 0 : A comprehensive set of sequence analys is programs for the VAX. Nucleic Acids Res 12:387-395, 1984

17. Phillips MA, Stewart BE, Qin Q , C hakravarty R, Floyd EE, ] etten AM, Rice RH : Primary structure of keratinocyte transglutaminase. Proc Natl Acad Sci USA 87:9333-9337, 1990

18. Kim H C, Idler WW, Kim IG, Han]H, C hung SI, Steinert PM: The complete amino acid sequence of the human transglutaminase K enzyme deduced from the nucleic acid sequences of cDNA clones.] BioI C hem 266:536- 539, 199 1

19. Yamin is hi K, Liew FM, Konishi K, Yasumo H , Doi H , H irano ] , and Fukushima S: Molecular cloning of human epidermal transgluta­minase cDNA from keratinocytes in culture. Biochem Biophys Res Commun 175:906-913, 1991

20. Polakowska R, Herting E, Goldsmi th LA: Epidermal type I transgluta­minase. ] Invest Demu tol 96:285 - 288, 1991

21. Thompson LH , Carrano AV, Sato K, Salazar EP, W hite BF, Stewart SA, Minkler SA, Minkler JL, Siciliano M]: Identification of nucleo­tide-excision repair genes on human chromosome 2 and 13 by func­tional complementation in hamster-human hybrids. SOmat Cell Mol Genet 13:539 - 551 , 1987

22. Polakowska RR, Eddy RL, Shows TB, Goldsmith LA: Epidermal type I transglutaminase (TGM1) is assigned to human chromosome 14. Cytogenet Cell Genet 56:105 -107, 199 1

23. T hacher SM: Purifica tion of keratinocyte transglutaminase and its express ion during squamous differentiation. ] Invest Dermatol 92:578-584, 1989

24. Levitt ML, Gazdar AF, Oie HK, Schuller H , Thacher SM: C ross­I inked envelope related markers for squamous diffe rentiation in human lung cancer cell lines. Cancer Res 50: 120-1 28, 1990

25. Sowa BA, Kelly KA, Muehleisen DP: Precycling polybrene-coated disks used to support electroblotted proteins improves sequencing yields. D NA Prot Engineering T ech 76:4350 -4354, 1988

26. Konohana A, Konohana I, Schroeder"WT , O'Brien WR, Amagai M, Greer] , Shimizu N , Gammon WR. Siciliano M], Duvic M: Placen­tal glutathione-S-transferase-Pi mRN A is abundantly expressed in human skin. ] Invest Dermatol 95: 119- 126, 1990

27. C hi rgwin ]M, Przybyla AE, MacDonald R] , Rutter WJ: Isolation of biologica ll y active ribonucleic acid from sources enriched in ribonu­clease. Biochemistry 18:5294 - 5299, 1979

28. Kedzierski W, Porter ] C: A novel non-enzymatic procedure for re­moving DNA template from RNA transcription mixture. Biotech­niques 10:210 - 21 4, 199 1

29. Thurston S], Saffer ]D: Ultraviolet shadowing of nucleic acids on nylon membranes. Analytical Biochemistry 178:41- 42, 1989

30. Fienberg AP, Vogelstein B: Addendum: A technique fo r radiolabeling DNA restriction endonuclease fragments to high specific activity. Analyt Biochem 137:266-267 , 1984

31. T okunaga K, Nakamura Y. Sakata K, Fujimori K, O hkubo M, Sawada K, Sakiyama S: Enhanced expression of a glyceraldehyde-3-phos­phate dehydrogenase gene in human lung cancers. Cancer Res 47:5616-5619, 1987

32. Sullivan DE, Auron PE, Quigley G], W atk ins PC, Stanchfield ] E, Bolon C:The nucleic acid blot analyzer. I: H igh speed imaging and quantltatlO n of 32P-Iabcled blots. BlOtechniques 5:672 - 678, 1987

33. Hsu SM , Riane L, Fanger H : A comparative study of the peroxidase­antiperox idase method and an avidin-biotin complex method fo r

34 SCHROEDER ET AL

studying polypeptide hormones with radioimmunoassay antibodies. Am J Clin Pathol 75:734- 738, 1981

34. Stoler A, Kopan R, Duvic M, Fuchs E: Use of monospecific antisera and cRNA probes to localize the major changes in keratin expression during normal and abnormal epidermal differentiation. J Cell BioI 107:427 - 446, 1988

35. Mahoney SE, Duvic M, Nickoloff BJ, Minshall M, Smith LC, Grif­fiths CEM, Paddock SW, Lewis DE: HIV transcripts identified in HIV -related psoriasis and Kaposi's sarcoma. J Clin Invest 88: 174-185,1991

36. Paddock S, Mahoney S, Minshal M, Smith L, Duvic M, Lewis D: Improved detection of in situ hybridization by laser scanning confo­cal microscopy. BioTechniques 11 :486 - 493, 1991

37. Glantz SA: Primer of biostatistics. 2nd Ed. McGraw-Hill, New York, 1987, pp 1-100

38. Michel S, Demarchez M: Localization and in vivo activity of epidermal transglutaminase.J Invest DermatoI90:472-474, 1988

39. Mansbridge IN, Knapp AM: Changes in keratinocyte maturation dur­ing wound healing. J Invest Dermatol 89:253 - 263, 1987

40. Bernard BA, Reano A, Darmon YM, Thivolet J: Precocious appear­ance of involucrin and epidermal transglutaminase during differen­tiation of psoriatic skin. Br J Dermatol 114:279 - 283, 1986

41. Said JW, Sasson AF, Shintaku IP, Banks-Schlegel S: Increased involu­crin expression near basal and squamous cell carcinomas. J Invest Dermatol 82:449-452, 1984

42. Coulombe PA, Koplan R, Fuchs E: Expression of keratin K14 in the epidermis and hair follicle: insights into complex programs of dif­ferentiation. J Cel l BioI 109:2295-2312, 1989

43. Elder JT, Tavakkol A, Klein SB, Zeigler ME, Wicha M, Voorhees JJ: Protoonocogene expression in normal and psoriatic skin. J Invest DermatoI94:19-25,1990

44. Yuspa SH, Ben T, Hennings H: The induction of epidermal transglu­taminase and terminal differentiation by tumor promoters in cul­tured epidermal cell s. Carcinogenesis 4: 1413 - 1418, 1983

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

45. Lichti U, Yuspa SH: Modulation of tissue and epidermal transgluta­minases in mouse epidermal cells after treatment with 12-0-tetrade­canoylphorbol-13-acetate and/or retinoic acid in vivo and in cul­ture. Cancer Res 48:74-81,1988

46. Murtaugh MP, Arend WP, Davies PJA: Induction of tissue transgluta­minase in human peripheral blood monocytes. J Exp Med 159:114-125,1984

47. Rubin AL, Rice RH: Differential regulation by retinoic acid and cal­cium of transglutaminases in cultured neoplastic and normal human keratinocytes. Cancer Res 46:2356 - 2361, 1986

48. Nagae S, Lichti U, De Luca SM, Yuspa SH: Effect of retinoic acid on cornified envelope formation: difference between spontaneous en­velope formation in vivo or in vitro and expression of envelope competence. J Invest Dermatol 89:51 - 58, 1987

49. George MD, Vollberg TM, Floyd EE, SteinJP,Jetten AM: Regulation of transglutaminase type II by transforming growth factor-beta 1 in normal and transformed human epidermal keratinocytes. J BioI Chern 265:11098-11104,1991

50. Michel S, Reichert U, Isnard JL, Shroot B, Schmidt R: Rctinoic acid controls expression of epidermal transglutaminase at the pre-trans­lational level. FEBS Letter 258:35 - 38, 1989

51. Elder JT, Fisher GJ, Lindquiest PB, Bennet GL, Pittelkow MR, Coffey R], Ellingsworth L, Derynck R, Voorhees J]: Overexpression of transforming growth factor alpha in psoriatic epidermis. Scicnce 243:811-814,1989

52. Sehgal PB: Interleukin-6: molecular pathophysiology. J Invest Der­matol94 (6 suppl):2S-6S, 1990

53. Baadsgaard 0, Tong P, Elder ]T, Hansen Ho, V, Hammerberg C, Lange-Vajlsgaard G, Fox DA, Fisher G, Chan LS, Voorhees JJ, Cooper K: UM4D4+ (CDw60) T cells are compartmentalizcd pso­riatic skin and release lymphokines that induce a keratinocyte pheno­type expressed in psoriatic skin lesions. J Invest Dermatol 95:275-282,1990 .

ANNOUNCEMENT

The 9th Annual Meeting of the Skin Pharmacology Society will be held October 14-16,1992 at the Stouffer Tower City Plaza Hotel in Cleveland, Ohio. There will be a special symposium on neuroendocrine immunologic interactions in skin, on novel phototherapies, and on clinical research, and the Albert Kligman Lecture will be presented October 16. There will be an opportunity for free oral and poster communications. The closing date for submission of abstracts is May 15, 1992. Abstract forms and meeting registration information can be obtained by writing to Pauline M Dowd, MD, FRCP, Secretary jTreasurer-S.P.S., Department of Der­matology, Middlesex Hospital, Mortimer Street, London W1N 8AA.