Vertebral epidermal transamidases

Biochimica et Biophysica Aeta, 351 (1974) 113-125

© Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands BBA 36704 VERTEBRAL EPIDERMAL TRANSAMIDASES

LOWELL A. GOLDSMITH*, HOWARD P. BADEN, STANFORD I. ROTH, ROBERT COLMAN, LORETTA LEE and BARBARA FLEMING Departments of Dermatology, Pathology and Medicine, Harvard Medical School, Massachusetts General Hospital, Boston, Mass., 02114 (U.S.A.)

(Received October 22nd, 1973)

SUMMARY The a-helical fibrous proteins (keratin) of the stratum corneum undergo changes in their physicochemical properties which are not solely explained on the basis of disulfide bond formation. The precursor of the stratum corneum protein, prekeratin, and the final product keratin were studied in the hair-follicle-free skin of calf snout epidermis. All of the lysines of prekeratin can be cyanoethylated while keratin molecules has 7-8 nmoles lysine per mg protein not available to cyanoethylation. Keratin from the human stratum corneum has 8-9 nmoles/g, protein which cannot be cyanoethylated. Direct examination of these proteins failed to detect the presence of N-e-acetyllysine groups or e-(7-glutamyl)lysine groups as the explanation of the blocked lysines. The hair-follicle-free epidermis of several vertebrate species contains transamidases similar to the transglutaminases which form e-(7-glutamyl)lysine crosslinks, in fibrin, hair follicle, and seminal fluid. The epidermal transamidase has a molecular weight between 40 000 and 60 000, is stimulated by reducing agents and calcium, and has most of its activity in the highermost levels of the epidermis. A semipurified preparation of the epidermal transamidase was able to form ~ dimers and a polymers from Factor XIII-free fibrin. A transamidase substrate, [1,4-~H2]putrescine, when injected into mouse skin and studied autoradiographically, was localized in the stratum corneum. [1,4-a4c2]putrescine incorporation into epidermal proteins was inhibited by 50 m M iodoacetamide. The incorporation of [ 1,4-14C2]putrescine was highest in very insoluble protein fractions. The exact role of transamidases in epidermal metabolism is not yet identified.

INTRODUCTION Transglutaminases, specific transamidases, cross-link fibrin [1] medullary hair proteins [2, 3] and the basic proteins of semen [4]. Similar transglutaminases present * Reprint requests to Lowell A. Goldsmith, Department of Medicine, Division of Dermatology, Duke University Medical Center, Durham, N.C., 27710, U.S.A.

114 in other tissues do not have clearly defined physiological roles although it has been suggested that they may cross-link structural proteins of cell membranes [5]. Epidermal transamidases were studied to determine if changes in the physicochemical properties of epidermal structural proteins during keratinization could be explained by these enzymes forming e-(~-glutamyl)lysine cross-links. This paper reports the characteristics of epidermal transamidases in the hair-follicle-free skin of several vertebrate species and the results of the analysis of epidermal fibrous proteins for blocked eamino groups. MATERIALS AND METHODS Dansylcadaverine (N-(5-amino-pentyl)-5-dimethylamino- 1-naphthalene-sulfonamide) monohydrate, e-(~-glutamyl)lysine and N-e-acetyllysine were purchased from Cyclo; BioGel P-2 from BioRad; a-casein (Hammerstan), p-nitrophenyl acetate, DEAE- and CMC-cellulose, from Mann; Sephadex G-200 from Pharmacia, Penicillin from Pfizer; isoniazid from Squibb; TPCK-trypsin and chymotrypsin from Worthington; mercaptoethanol, iodoacetamide, pepsin, leucine amino-peptidase, carboxypeptidases A and B, acetylthiocholine chloride, ovalbumin and bovine serum albumin from Sigma; salts and other reagents were the highest grades available. [1,4-~H2]Putrescine (108 Ci/mole), and [1,4-1~C2]putrescine (20 Ci/mole) and aquasol were purchased from New England Nuclear. Hank's solution was purchased from Baltimore Biological Laboratories. Human fibrinogen (96 ~ clottable) was purchased from AB Kabi (Stockholm) and screened for the absence of Factor XIII by the monochloroacetic acid test [6]. Bovine thrombin, prepared from a single animal was purified by the method of Rosenberg and Waugh [7] and at the time of use had a specific activity of 2300 N.I.H. units/rag. Enzyme preparation: cow snout

Snouts from freshly slaughtered cows were obtained from a local abattoir. Epidermis visually and histologically free of hair follicles were carefully removed with a microtome knife and rinsed with cold 0.25 M sucrose. Gloves were worn during the preparation of the epidermis to prevent contamination with human enzymes and care was taken to prevent contamination with blood. This preparation and succeeding operations were conducted at 4 °C. A 20 ~ suspension of minced snout epidermis in 50 mM Tris-HCl, pH 7.5, containing, 1 mM EDTA, and 0.25 M sucrose was homogenized for 3 min at 45 000 rev./min in a Virtis 45 homogenizer. The homogenate was centrifuged at 20 000 x g for 15 min and the supernatant was concentrated to 11 ml with an Amicon ultrafilter XM-50. All of the enzyme activity was found in the concentrated extract. Gel filtration of the concentrated extract in 5 mM Tris-HCl, pH 7.5, 1 mM EDTA on a 2.5 c m x 83 cm column of Sephadex G-200 was performed. 4-ml fractions were collected. The fractions containing enzyme activity were pooled and applied to a DEAE-cellulose column (0.9 cm × 15 cm) equilibrated with 50 mM Tris-HC1, pH 7.5, 1 mM EDTA. The column was washed with 50 ml equilibrating buffer, and developed with a 200-ml linear gradient of NaC1 (0-0.6 M) in the same buffer. Enzyme activity coincided with the first peak of protein eluted at about 0.05 M NaC1, were

115 equilibrated with 5 m M Tris-acetate, p H 6.5, and applied to a 0.9 cm × 12 cm carboxmethylcellulose column equilibrated with the same buffer. After washing, a 200-ml linear gradient of NaC1 from 0-0.6 M was applied and enzyme activity coincided with the major peak of protein at approximately 0.2 M NaC1. The active peak was concentrated on an XM-50 membrane, diluted with 5 m M Tris-HC1, p H 7.5, 1 m M E D T A and applied to an equilibrated D E A E column and developed with a 200-ml linear gradient from 0 to 0.25 M NaC1. Two protein peaks eluted between 0 and 0.1 M NaC1 and the second peak which eluted at 0.05 M NaC1 contained enzyme activity. The material was used for cross-linking experiments, the measurements of esterase activity, thiolesterase activity, and acrylamide-gel electrophoresis [8]. In 4 separate experiments 18 m m diameter cores were taken from individual cow snouts with a specially constructed punch, and 0.5 m m thick wafers cut parallel to the skin surface with a Stadie-Riggs microtome, weighed, homogenized in sucrose, concentrated and assayed. The lowermost slice was still within the epidermis by inspection. One set of biopsies was dried at 110 °C to constant weight to determine the water content of individual epidermal layers.

Enzyme preparation: other animals Frogs (Rana pipiens and Rana catesbeiana), toads (Bufo marinus) and a soft shelled pond turtle (Chrysemys picta) were pithed and their skins removed by blunt dissection. The skins were then heated for 30 s at 56 °C and the epidermis removed with a scalpel. Gloves were worn during all these precodures and care was taken to prevent contamination from blood. Wax depilated rats were decapitated and their epidermis separated as above. Rat fetuses, surgically removed from the uterus, were weighed and measured, their fetal membranes removed and their epidermis scraped off. The various epidermal tissues were homogenized in 0.25 M sucrose and the supernatant after centrifugation at 100 000 × g for 45 min was concentrated on XM-50 filters and assayed. Freshly removed human stratum corneum was homogenized, concentrated and assayed in a similar fashion.

Enzyme assay Transamidase activity was assayed by measuring the incorporation of dansylcadaverine into a-casein at 37 °C with a modification of the method of Lorand et al. [9]. The 2.0-ml reaction mixture contained 1 m M dansylcadaverine, 0.1-0.15 ~ casein, 10 m M mercaptoethanol, 50 m M Tris-HC1, p H 7.5, and 5 m M CaCI~ and 0.3 ml of the enzyme mixture. Reactions were stopped by adding trichloroacetic acid to a final concentration of 5.0 ~ , and then washing the precipitate once with 5 ~ trichloroacetic acid and 5 times with a 50 ~ (v/v) mixture of ethanol and ethylether. The pellet was air dried, dissolved in 2 ml of a solution containing 0.05 M Tris-HC1, p H 8.5, 0.5 sodium dodecylsulfate and 8 M urea. The fluorescence of solution was read in a Bowman-Amicon spectrofluorimeter, excitation 355 nm, emission 525 nm. Monodansylcadaverine diluted in the same solvent served as a reference solution. Protein concentrations were measured by the Lowry method [10], and D N A by the method of Santoianai and Rothman [11 ] using calf thymus D N A as a standard.

Heat inactivation Aliquots of the cow snout extract were added to test tubes preheated to 56 °C

116 in a stirred water bath and at 2, 5, l0 and 15 min, rapidly cooled in an ice-water bath. The suspension was centrifuged in the cold at 20 000 × g for 15 min and the supernatant assayed.

Ca-EDTA effects An aliquot of the Sephadex G-200 gel-filtered cow snout enzyme was gelfiltered a second time in 0.05 M Tris-HC1, p H 7.5, on a 1.5 cm x 30 cm column of Sephadex G-25. The protein peak was concentrated on an XM-50 membrane and assayed in the standard assay mixture, the standard assay mixture prepared without calcium, and the standard assay mixture prepared without calcium but containing 2 mM EDTA.

Inhibitor experiments Penicillin at a final concentration of 20 units/ml, and isoniazid at 5 concentrations from 16 to 124 m M were tested in the standard assay mixture with cow snout enzyme. Iodoacetamide at 20 m M final concentration, was added before dansylcadaverine to the standard assay mixture. Hydroxylamine at 17 m M final concentration was tested with rat epidermal enzyme in the standard assay mixture.

Esterase activity Calcium-stimulated esterase activity [12] was measured in rat epidermis after 1130 COO × g centrifugation, in cow snout epidermis after 100 000 × g centrifugation and in the highly purified snout preparation used for cross-linking experiments. Thiolesterase activity [13] was also measured in unpurified and semipurified cow snout preparations.

Cross-linking activity 20 #g of the highly purified cow snout enzyme was incubated at 37 °C with 2.5 mg fibrinogen 4.5 N.I.H. units purified bovine thrombin in 0.1 M Tris-HC1, p H 7.5, 1 mM dithiothreitol, 5 m M CaC12, 80 m M NaC1. Control tubes included fibrinogen alone, fibrinogen and thrombin, fibrinogen and cow snout enzyme without thrombin, in the same buffer. After incubation for 1 h 1 ml of 0.05 M Tris-HC1, p H 8.5, with 8 M urea and 0.5 ~ sodium dodecylsulfate was added and the solution incubated at 50 °C overnight. Aliquots of the solubilized solutions were applied to sodium dodecylsulfate acrylamide gels [14].

Incorporation of putrescine into epidermal proteins For autoradiographic experiments 5 /zCi of [1,4-ZH]putrescine were injected intradermally into mouse back skin. Biopsies taken 1 and 4 h after injection were fixed in Bouin's solution, dehydrated, embedded in paraffin and 4-6-#m sections covered with K o d a k NBT2 emulsion and developed after 1, 2 and 3 weeks exposure

[15]. To isolate the proteins which incorporated putrescine, 1.7 g of cow snout epidermis were incubated with 16 ml Hank's solution, p H 7.0, and 8/zCi [1,4-14C] putrescine for 1 h at 37 °C. The mixture was washed twice with cold 0.25 M sucrose, then homogenized in a glass homogenizer with l ~ unlabelled putrescine in 0.25 M sucrose. After centrifugation at 20 000 × g the pellet was washed with 10 ml 0.05 M

117 Tris-HC1, p H 7.5, and sequentially extracted at 4 °C for 24-h periods in l0 ml of 0.05 M Tris, p H 8.5, with 6 M urea; 0.l M Tris-HCl, p H 8.4, with 6 M urea and 0.1 M mercaptoethanol; 0.1 Tris-HCl, p H 8.9, with 6 M urea and 0.1 M mercaptoethanol; and 0.1 M fi-alanine, p H 10.7, with 6 M urea and 0.1 M mercaptoethanol. The remaining pellets were extracted sequentially with 10 ml 0.1 M N a O H and 10 ml 0.2 M N a O H . The final pellet and the urea and urea-mercaptoethanol extracts were exhaustively dialyzed with water, lyophilized and weighed. 2-mg aliquots of the lyophilized material were wet with 0.4 ml water and dissolved in 5 ml NBS solubilizer, and 15 ml aquasol and counted in a Packard Tri-carb scintillation counter. One ml aliquots from the N a O H extracts were mixed with 4 ml NBS solubilizer and 15 ml aquasol and counted directly. In another set of incorporation experiments one incubation flask was incubated with 50 m M iodoacetamide for 10 min before the addition of 8 /~Ci of [1,4-14Cz]putrescine.

Cyanoethylation S-Carboxymethyl derivatives of prekeratin and the a-helical protein of cow snout stratum corneum were prepared [ 16, 17]. 10-rag samples of protein were suspended in 2.5 ml of 0.05 M N H 4 H C O 3 containing 0.01 M CaCl 2 and 1 mg TPCK-trypsin. After 48 h incubation at 37 °C the mixture was centrifuged at 2000 × g, the supernatant was evaporated at 60 °C and then cyanoethylated with 0.2 ml acrylonitrile with 20 #1 triethylamine at 37 °C for 120 h [18]. After incubation, the tubes were dried, 1.0 ml 6 M HC1 added and hydrolyzed in vacuo at l l0 °C for 24 h, and analyzed on a Beckman 116 automatic amino acid analyzer. A similar incubation was performed beginning with 60 mg of S-carboxymethylated human stratum corneum and 6 mg TPCK-trypsin, aliquots of the mixture were incubated at 37 °C for various times. The molar arginine/lysine ratios were calculated.

Attempts to demonstrate e-(v-glutamyl)lysine dipeptides To directly demonstrate the presence of e-(:y-glutamyl)lysine in stratum corneum a-helical fibrous protein the S-carboxymethylated urea-mercaptoethanol-extractable protein from h u m a n stratum corneum was treated with trypsin as above, and then sequentially with pepsin, chymotrypsin, pronase, leucine amino peptidase and carboxypeptidase A and B [19]. Part of the mixture was deproteinated with 3 ~o (w/v) picric acid, centrifuged, and the supernatant applied to a small Dowex 2-X8 colunm which was then eluted with 0.02 M HCI. The eluate was examined for s-(7-glutamyl )lysine on the Beckman 116 amino acid analyzer with resin and program. Another portion of the complete enzyme digestion was filtered through an UM-10 Amicon membrane and the filtrate analyzed for s-(~-glutamyl)lysine by chromatography on a Beckman 120 C amino acid analyzer with a custom AA-15 resin (0.9 cm × 55 cm) [20]. The zone where standard e-(y-glutamyl)lysine eluted was collected, acid hydrolysed and rechromatographed on a D u r r u m amino acid analyzer.

N-e-acetyllysine isolation Using a BioGel P-2 (200-4130 mesh) column (1 cm × 1130cm) in 0.25 acetic acid N-e-acetyllysine was consistently separated from the bulk of the amino acids. It eluted

118 at 105-115 ml between lysine and histidine standards. Portions of the complete enzyme digest were chromatographed on this column and the eluate between 100-120 ml pooled, lyphilized, dissolved in 0.2 M citrate buffer, p H 2.2, and then chromatographed on the long column of the Beckman 116 amino acid analyzer; authentic N-e-acetyllysin eluted as a separate peak just before glycine. A preparation of enzyme digest corresponding to 2.5 mg starting protein was analyzed using this method. RESULTS

Enzyme isolation and properties Transamidase activity was present in 12 different cow snout epidermis preparations. The transamidase activity of crude homogenates were stable at --40 °C up to four months, but isolated fractions of the enzyme after gel filtration or ion exchange chromatography rapidly lost activity and hindered enzyme purification; performing this procedure in the presence of dithiothreitol or sucrose did not prevent loss of enzyme activity. The transamidase activity eluted from the Sephadex G-200 column between bovine serum albumin and ovalbumin (Fig. 1). No activity was present in the first peak of the chromatogram where plasma Factor X I I I would be eluted. 0-200 2.5 x 8 3 c m

[

) 30

40

BSA

~)VALBUMIN

% 50 60

"~

70

~~ 3°I ~, 20

o

i

1o

100

80 90 k

_

200

VOLUME

300

400

J t00 500

(ml)

Fig. I. Transmission and transamidase activity of concentrated 100 000 x g supernatant of cow snout epidermis on fractions from Sephadex G-200 gel filtration. No transamidase activity was present in the first and last peaks of the chromatogram.

The most purified preparations after chromatography revealed 2 bands staining with Coomassie Blue one 20-50 times the intensity of the other. Attempts to elute enzyme activity after gel electrophoresis from unstained gels were unsuccessful. Transamidase activity was proportional to time and the amount of enzyme present in the assay mixture in cow snout and rat epidermis preparations. Km for dansylcadaverine with the cow snout enzyme was 8.8.10 -4 M. Thrombin activation was not necessary for enzyme activity and 2.4 N.I.H. units purified bovine thrombin added to an assay mixture of crude cow snout enzyme did not increase enzyme activity. 60 % of the

119 TABLEI HEAT STABILITY OF EPIDERMAL TRANSAMIDASE AT 56 °C Time (rain)

Relative fluorescence

0 2 5 10 15

5O 42 37 32 30

transamidase activity was stable to heating to 50 °C for 15 min (Table I). Although reducing agent was not necessary for enzyme activity 50 mM mercaptoethanol increased activity of freshly prepared cow snout 100 000 × g homogenates 75 ~ over the control activity (Table If). The gel filtered enzyme did not require exogenous TABLE II EFFECT OF MERCAPTOETHANOL ON COW SNOUT EPIDERMIS TRANSAMIDASE ACTIVITY Mercaptoethanol concn ( m M ) 0 1

2 5 10 50

Relative fluorescence 27 26 30 32 37 47

calcium for activity but was stimulated three and one-half times by 4 mM CaCl2 (Table III); there was no enzyme activity in the presence of 2 m M EDTA. No calcium stimulated esterase or thiolesterase activity was present in purified or crude cow snout extracts.

lnhibitors 20 mM iodoacetamide when added before substrate completely inhibited enzyme activity; isoniazid at a final concentration of 16 mM to 124 nM, and penicillin at a concentration of 20 units/ml reaction mixture were not inhibitors; 17 mM hydroxylamine inhibited enzyme activity by 96 ~ .

Cross-linking activity of purified cow transamidase 20 #g of chromatographically purified cow snout transamidase cross-linked plasma Factor X I I I free human fibrin. Sodium dodecylsulfate-acrylamide electrophoresis showed the a and ~ chains were decreased and ~ dimers and a polymers formed (Fig. 2). The reaction mixture in the absence of thrombin gelled slightly but cross-linking was not demonstrated electrophoretically.

120 T A B L E III INHIBITION OF EPIDERMAL TRANSAMIDASE

BY E D T A A N D A C T I V A T I O N BY Ca 2+

Semipurified cow s n o u t t r a n s g l u t a m i n a s e was gel filtered on a 1.5 cm × 30 cm c o l u m n of Sephadex G-25 in 0.05 M Tris-HC1, p H 7.5, the protein containing material was concentrated on a U M - 1 0 a m i c o n filter a n d assayed in the s t a n d a r d reaction mixture containing no a d d e d calcium or E D T A ; or with 2 m M E D T A , or with 4 m M C a C l v T h e reaction was stopped at 30 or 60 rain with trichloroacetic acid a n d quantitated as described in Materials a n d Methods. Incubation time (min)

30 60

E n z y m e activity (/tmoles dansylcadaverine incorporated) Gel-filtered enzyme, no added CaCI2 or E D T A

2 mM EDTA

4 mM CaC12

0.42 0.77

0.21 0.22

1.53 2.44

t--.- ¥.¥

•

¥

o

b

Fig. 2. S o d i u m dodecylsulfate-acrylamide gels of (a) Factor Xlll-free fibrinogen incubated with t h r o m b i n a n d 20 p g of semipurified cow s n o u t transamidase. Factor X l l l - f r e e fibrinogen incubated with t h r o m b i n (b). T h e / 7 b a n d s are of equivalent intensities in both preparations. In (a) 7 dimers a n d weak a polymers are seen.

121 TABLE IV LOCATION OF TRANSAMIDASE ACTIVITY IN THE COW SNOUT EPIDERMIS Layer of the cow snout *

¢tmoles dansylcadaverine incorporated/4 h/100 mg wet wt Expt : 1

l(stratumcorneum) 2 3 4(basallayer)

18.8 2.1 0.7 0.5

2

3

4

8.0 1.8 1.8 0.6

3.3 1.7 0.8 0.3

4.8 3.3 0.7 0.5

* Progressively deeper layers of cow snout epithelium removed as described in text.

Enzyme location in epidermis The location of enzymatic activity was determined by assaying t r a n s a m i d a s e activity in different layers of the cow s n o u t epidermis. E n z y m e activity was highest in the outermost layers of cow snoit epidermis (Table IV). These corresponded histologically to the s t r a t u m c o r n e u m and the g r a n u l a r layer. The water contents of the various layers of the cow s n o u t were very similar, 79 _-t~ 3 ~ . The variability between

Fig. 3. Light micrograph of radioautograph of mouse skin 4 h after the intradermal injection of [1,4-aH2]putrescine. Significant localization is limited to the stratum corneum (arrows). Malpighian cells, dermis and hair follicles show only background levels of grains. Hematoxylin and eosin. ×488

122 experiments is partially related to differences in the thickness of various epidermal layers and variation in the rete ridge pattern in different snouts. Autoradiographs 1 h after the intradermal injection of radioactive putrescine showed labelling of the viable epidermis and stratum corneum. 4 h after injection labelling was almost exclusively limited to the stratum corneum of the interfollicular epidermis with only minimal labelling of the viable epidermis and hair follicles (Fig. 3). In vitro incorporation and autoradiography of [ZH]putrescine into cow snout showed also early labelling of the stratum corneum. Enzyme activity could also be demonstrated and quantified in the scales of patients with various skin diseases such as psoriasis and ichthyosis confirming the retention of enzyme activity in the stratum corneum.

Nature of protein acceptor Jbr transamidase Although putrescine was incorporated into all fractions of the solubilized proteins the highest specific activity was observed in the sodium hydroxide extracts and in the final pellet (Table V). The specific activities of the urea of the urea and TABLE V INCORPORATION OF [I,4-14C2]PUTRESCINE INTO EPIDERMAL PROTEINS Extraction conditions

cpm/mg

6 M urea, pH 8.5 (prekeratin) 6 M urea, pH 8.4, 8.9, 10.7 0.1 M mercaptoethanol (keratin) 0.1 M NaOH 0.2 M NaOH Final pellet

mg protein

39

28

27 820 2000 135

70 6 3 11

urea-mercaptoethanol extracts, which contained prekeratin and keratin respectively, were low and showed no significant differences. The incorporation of putrescine into cow snout proteins was inhibited 84 ~ when the epidermis was preincubated with 50 m M iodoacetamide (11 600 cpm/g for control tissue and 1880 cpm/g for iodoacetamide-treated tissue). TABLE VI EPIDERMAL TRANSAMIDASE ACTIVITY IN DIFFERENT SPECIES /~moles dansylcadaverine incorporated

Rat skin, 5 days old Turtle leg skin Toad (Bufo marinus) Frog (Rana catesbeiana)

/~moles/h/g epidermal wet wt

/~moles/h/mg epidermal DNA

21.3 11.0 30.2 4.0

52.0 66.5 42.4 18.5

123 TABLE VII PRESENCE OF BLOCKED LYSINE IN EPIDERMAL FIBROUS PROTEINS AS DETERMINED BY CYANOETHYLATION Blocked lysine (nmoles/mg) Prekeratin Cow snout; urea-mercaptoethanolextractable protein Human stratum corneum ureamercaptoethanol-extractable fraction Chymotrypsin

0 to trace 8.8, 7.7, 7.1 8.7, 8.0 0

Comparative studies E p i d e r m a l t r a n s a m i d a s e activity was detected in a variety o f vertebrates tested (Table VI). It was present in n e w b o r n rats a n d also at 14 d a y s o f fetal life in rat epidermis. The e p i d e r m a l enzyme f r o m Rana pipiens h a d similar p r o p e r t i e s to cow snout transamidase. It was c o m p l e t e l y inhibited by 20 m M i o d o a c e t a m i d e a n d by 4.3 m M E D T A , and had a 3 0 ~ increase in its activity with 10 m M m e r c a p t o e t h a n o l .

Blocked lysines P r e k e r a t i n h a d no to trace a m o u n t s o f b l o c k e d lysine. C h y m o t r y p s i n , studied as a n o t h e r control, h a d no blocked lysines. The a-fibrous s t r a t u m c o r n e u m protein had 7-8 ~ b l o c k e d lysines; a similar level occurred in the a-fibrous p r o t e i n o f h u m a n s t r a t u m c o r n e u m (Table VII). A c o n s t a n t level o f b l o c k e d lysine r e m a i n e d after 14 days o f c y a n o e t h y l a t i o n (Table VIII). N-e-Acetyllysine or e-(7-glutamyl)lysine were n o t directly d e m o n s t r a t e d in c o m p l e t e enzyme digests o f h u m a n s t r a t u m corneum. TABLE VIII PERSISTANCE OF LYSINE IN HUMAN STRATUM CORNEUM a-FIBROUS PROTEIN AFTER PROLONGED CYANOETHYLATION Incubation time (days)

Arginine/lysine ratio*

0 2 5 10 14

0.8 13.7 18.5 17.4 19.0 * Mean of duplication samples for all times, except for the 14-day incubation which is one sample.

DISCUSSION Hair-follicle-free epidermis o f the cow and other vertebrates c o n t a i n e d a transa m i d a s e with m a n y o f the characteristics o f the transglutaminases including relative

124 heat stability, inhibition by E D T A and iodoacetamide, stimulation by reducing agents and the ability to cross-link fibrin. Of the several tissues transglutaminases, the epidermal enzyme most closely resembled the hair follicle transglutaminase in properties, molecular weight and lack of esterase activity. Due to the complex anatomy of hair follicles epidermal transglutaminase may have contaminated the hair follicle preparations of others. Cow snout epidermal transamidase activity was highest in the stratum corneum and granular layer. Arginase [21] and phospholipase [22] have similar distributions. This location of transamidase activity suggested the enzyme may function in the latter stages of keratinization. Post-translational modification of the epidermal fibrous proteins as a potential function of epidermal transamidase was investigated by quantitating the blocked eamino groups of prekeratin and the a fibrous protein of the stratum corneum in cow snout epidermis. None of the e-amino groups of prekeratin were blocked while 7-8 ~/o of those in the stratum corneum's alpha fibrous protein were blocked. After complete enzymatic digestion of the human stratum corneum fibrous protein, no e-(y-glutamyl)lysine dipeptides or N-e-acetyllysine were detected. Since only 60 ~ of the stratum corneum was digested to amino acids under the experimental conditions it was possible for such groups to have been undetected. The incorporation of radioactive putrescine was followed autoradiographically and chemically to study the distribution and nature of the acceptor protein. Putrescine labelling of the stratum corneum was much higher than that of the nucleated epidermis which was very different from the distribution of nucleic acid or amino acid precursors [23]. In vitro labelling of the epidermal proteins was decreased by iodoacetamide which indicated that most of the putrescine decreased by iodoacetamide which indicated that most of the putrescine incorporation was enzyme dependent. When putrescine-labelled cow snout was fractionated into prekeratin, stratum corneum a-helical fibrous protein and other less soluble fractions the highest specific activities were in those fractions requiring 0.1-0.2 M N a O H for solubilization and in the material insoluble under these conditions. These fractions probably contained highly cross-linked fibrous proteins, cells envelopes and cell membranes. This incorporation of putrescine into epidermal proteins suggested that post-translational cross-linking could occur within the epidermis but direct demonstration of e-(y-glutamyl)lysine cross-links has not yet been possible. ACKNOWLEDGMENTS We thank Dr J. Pisano (National Institutes of Health, Bethesda, Md.) for confirming some of our early cyanoethylation experiments, and Dr L. Lorand and Mr S. Pabalan (Northwestern University, Evenston, Ill.) for trying to identify e-(yglutamyl)lysine in stratum corneum fibrous protein. This research was supported by grants from the National Institutes of Health (AM-43414, AM-6838, CA-10744) and the National Foundation (CRBS-70). REFERENCES 1 Lorand, L. (1972) Ann. N.Y. Acad. Sci. 202, 6-30 2 Chung, S. I. and Folk, J. E. (1972) Proc. Natl. Acad. Sci. U.S. 69, 303-307

125 3 Harding, H. W. J. and Rogers, G. E. (1972) Biochemistry 11, 2858-2863 4 Williams-Ashman, H. G., Notides, A. C., Pabalan, S. S. and Lorand, L. (1972) Proc. Natl. Acad. Sci. U.S. 69, 2322-2325 5 Birckbichler, P. J., Dowben, R. M., Matacic, S. and Loewy, A. G. (1973) Biochim. Biophys. Acta 291, 149-155 6 Lorand, L. (1950) Nature 166, 694--695 7 Rosenberg, R. D. and Waugh, D. F. (1970) J. Biol. Chem. 245, 5049-5056 8 Davis, B. J. (1964) Ann. N.Y. Acad. Sci. U.S. 121, 404-418 9 Lorand, L., Urayama, T., DeKiewiet, J. W. C. and Nossel, H. L. (1969) J. Clin. Invest. 48, 10541064 10 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265275 11 Santoianai, P. and Rothman, S. (1963) J. Invest. Dermatol. 40, 317-323 12 Folk, J. E., Cole, P. W. and Mullooly, J. P. (1968) J. Biol. Chem. 243, 418-427 13 Lorand, L., Chou, C. H. J. and Simpson, I. (1972) Proc. Natl. Acad. Sci. U.S. 69, 2645-2648 14 Weber, K. and Osborn, M. (1969) J. Biol. Chem. 244, 4406 4411 15 Roth, S. I. and Baden, H. P. (1967) J. Exp. Zool. 165, 345-353 16 Baden, H. P. and Gifford, A. N. (1970) Biochim. Biophys. Acta (1970) 674-676 17 Baden, H. P. (1970) J. Invest. Dermatol. 55, 184-187 18 Pisano, J. J., Finlayson, J. S. and Peyton, M. P. (1969) Biochemistry 8, 871-876 19 Harding, H. W. J. and Rogers, G. E. (1971) Biochemistry 10, 624-630 20 Lorand, L., Downey, J., Gotoh, T., Jacobsen, A. and Tokara, S. (1968) Biochem. Biophys. Res. Commun. 31,222-230 21 Rossmiller, J. D. and Hoekstra, W. G. (1965) J. Invest. Dermatol. 45, 24-7 22 Long, V. J. W. and Yardley, H. J. (1972) J. Invest. Dermatol. 58, 148-154 23 Fukuyama, K., Nakamura, T. and Bernstin, I. A. (1965) Anat. Rec. 152, 525-536