THF .luliR.'A'- or hn:sn<:ATI\ £ 0£RMAT0Lot;) , 6:!: 361 366. llli-1 • Cop~·ru:ht@ l9i·l hy The William' & \\ ilk1n, Cu.

\ul 62. '1/o ~ Pr1nred 111 l.'.S.A

IMMUNOCYTOCHEMICAL LOCALIZATTOK OF COLLAGE1 ASE IN HUMAN SKIN AND FIBROBLASTS r~ MOI\OLAYER CULTURE'"

:VIAX E REDDICK :\J.D .. EL'GE'-E A. BAl'ER. M.D .. A'IIJ ARTHl'R Z. EISEN. M .D.

ABSTRACT The development ot a lunctionall~ mono~pecifie ant i~erum Lo human skin collagenase has

permined the precise localization or this enzyme using lluorescence microscop~· in tissue sections of human skin and in fibroblasts from primary cell culture. Ln human skin. collagenase is primarily localized to the upper or pupillary dermis. Cnllagen fibers show pronounced staining sugge~ting that much of the enzymt> present 111 skin is t>xtracellular and perhap:- bound to its collagen ~ubstrate. Cytopla~mic staining was also seen in fibroblast-like rells scattered in the upper dermis. Specifit' fluorescent ~tainmg was \'irtually absent in the mid- and lower dermis and in the epidermis.

Fibroblasts obtained from human skin explants displa~·ed a granular and frequently reticulated !';taining pattern primarily in the perinuclear region when examined etther hy immunof1unrescent or tmmunoperoxidase techntques. By radiCllmmunoassay the culture medium contains material immunologically idem ical to skin collagenase. The~e findings suggest thai the fibroblast is the major cell respon:-ible for C!•llagenase production.

'Specific neut rat collagenase;. have been b.olat ed from organ cultures of a number of human and ammal tissue~ [11 and recent studies mdicate that this j.!J'oup of enz~·mes is acti\·ely invoh·ed in the

~ remodehng ol collagen in vivo. For example. colla­gena;.e ha~ been detected in extracts of' normal

~ human "kin [2 ). 111 rheumatoid syno,•ial 11 uid J :~ J. and in involuting rat uteru~ [4j. De::.pite the apparent importance of rollagenases m collagen metahultsm. the cell type responsibl£> for the pro­duction of the enzyme has not been established. In order tn obtain an accurate as:-es!;ment ol the role of collagena~c in the in vi\'0 degradat inn ot collagen

.. in normal and pathologic state;;. it i;, essential to

identity the cell of ongin of the enzyme as well as ~ the localization of eollagenase in the extracellular

elements of' the t tssue. A major deterrent to the cellular localization of

collagenase production has been the inahtlity to maintain cell cultures in serum-free medium. a

~ condttion necessary for the enzymatk detel·tion of collagenase, s inct' these enzymes are in hibited by

; whule serum [5]. The availability of a functionall) monospecifH' ant 1serum to hum on skin collagena,.,e !61 suggested that an immunologit' approach might he useful in est ablishing the cytologic ori~-rn1 oft hts enzyme in the intact organism and in celleultures of human skin fibroblasts. Immunologic detection

• ot collagenase is more sensitive than the enzymatic assay [7] and the enzyme can be measured in ttssue extra('ts in the prelience ot .:;erum and perhap!< other tissue inhibitors ot tollagena:-;e act i\­lly [2, .'i j.

This mve~llgatiom 11a~ suppnrted b~ l l.:-i. Puhlil' ~ Health Service Research Grant~ Al\1 12129 and Ai\1 O.'ifH 1

from the National Institute <Jf Arthrius and :V1etnhultc ~ Dtsea~e,.

· From the Divi·,iun nl Dermawlu~ . Depart men! ut Medtctne. Washington Univer::.ity School or :OVledicme.

~ ~t Lnui!<, Missouri G:HlO.

This report describes the cellular localization of human skin cr>llagenase by immunohistochemical techniques in normal human skin under ccmditions where enzymutic activity cannot be detected. In addition. immunologic localization ol collagena~e has been accomplished in fibroblast cultures at a time when immunoreacti1·e t•ollagenase can he demonstrated in the culture medium in which the fibroblasts ha\'e been grown.

METIIOl)S

PreparatiOn of antisera. Functionally monospecific an­ti::.erum to human skin cnllagenru;e tami- HSC' J was prepared in raboits as previou;.ly described [61 and was taken tn 40" sarur1Uinn \\Jlh ammOnium sullate at 0°(' and pH 7.0 to obtain the gamma globulin fraction. Th1s preparation "a~ dissoh-ed in O.O:i M Tr1:;- HCI tpH 'i.:il with n. 15 M NaCJ tTri:.- :"JaCJ buffer!, dialvzed agaim;t the same buller and adjusted to a protem concentraunn of 20 mltfml.

Nonimmune rabbit gamma globulin was obtained I rom rabbits prinr tu Immunization. Rabbit ani ihuvine serum albumin and goat antirabbit lgG were purchased l'Ommercially tGatewa\ lmmunochemirahd. These ~era were ~ubjerted to an identical 40':; ammonium sulJate fractwnalton prtor lU use and theu protem concentra­tions were adju~ted to 10 mg/ml in Tris NaCI buffer.

Preparatwn •>f {luorescem-conju~?ated }(Oat anllrabbll [f~G. Gnat antirahhit l!!G was tonjugated to lluorescein •snthwcyanate (FTTCI a~ de~cribed hy :'\airn [8]. The react ion mixture contained~ ml of0.2 M :\a,HPO, buller I pH 9.51, !(){) n g ol the gamma globulin fraction oi goat antirabhit lgG. and I :.lfo m~: of FITC !Isomer. I. S1gma Chemical Co.l. l'he reaction mixture was stirred for :lO min at :.l5°C. alter which the FITC-Iabeled goat anllrab· bit lgG was separated frr>m the rree FITC by gel filtration on a ('Uiumn (:2 . .1 15 ern) of ::iephadex G-:2;} equilibrated with O.l)J M pho~phate buffer I pH 'i.:l l. The conJugated antiserum was applied to a column 11.1 . :30 cml of DEAE ce llulose 1\\'hatman DE-321 equilibrated in the same buller. Eluttnn wru; ac,·omplibhed with a linear gradwnt In I \ol NaC'l [9, Ill J The FlTC-labeled an·

:36!

:362 THE JOUR!\AL OF 1110\'ESTIGATIVE DERMATOLOGY

tberum obtained by this procedure had an F 1P molar ra£io nl 3.8 and a protem concentration of 4.9 m!dml. Follow­in){ dialysis into Tris :-\aC'l buffer. the samples were di\'idcd into I 0 ml aliquot" and l~·ophilized. They were suh;;equently ~tored At 20°(' and prott:n"d from light until used The conjugated antiserum retamed precipitin activity against rabbit lgG by Ouchterlony double diffu­sum 1111.

Prl'puration of peroxrdo-'e- antrbnd\ con)U/!atP,, Goat aniirabbii IJ!;G was conjugat~d to horl'eradi:;h peroxidase !Tvpe I\'. Sigma Chemica I Cn. l using glutaraldehvde th

the hitunctinnal cross.Jinkina: reagent 1121. A typical reaction mixture mntaint>d .'10 mg of horseradi,.h peroxi­dal't' and 20 mg of goat ant irabhit lgG in phosphate buller !pH 6.91 to which was added 0.2 ml ut \ 'JC glutaraldehyde in dropwise ra~hion with con,.tanl stirring. After allowin~r the reartion to proeeed fur 6 hr at room temperature, the mixture wa~ placed on o ('olumn ( 1.6 70 ern) of Sephudex G-200 equilibrated with Tris '\;aCI buffer. Eluent tractions were monitored at an absorbance ot ·103 nm and peroxidase acti\'ity wa::. asse:-sed with guaiacol rea~rent and H,O,. C'nnjuj!:ated active lractionb werE> pooled, dial~·zed. and lyophilized fur storage at ~n·c.

Ti.•~ru:source.,, For localization ol (·ollagena~e in intact human ~kin. 3-mm punch biopsie~ lrom healthy adult volunteer;; were obtained under l 'lt lidocaine local ane;;­thesiil. SpecimenF were quick-trozen and cut into P.·JJ thi<-k sertrons on a ci')'O::.lat at 2s•c.

Primary human libroblnst culture" were initiated from two :l-mm skin punch biopsies which were mmced finely with 11 razor blade and placed in ~0 ml of Dulbeccu' !< modiJied Eagle's medium containing 2 mg/ml of bacterial collagenase ( Clu.,tridium histoiJtirum, Wort hin,non Bio­rhemu·aiJ. ;; mg,ml of tryp,in. :!llO units fml of penicillin and :WO ,..gtml of streptomycin. The suspension \\Rl'

incubated at 37 °C for ~i) min wnh ~:entle stirrin)!. Celb and tissue were sedimented by low-speed centrifugation at room temperature. the ~upematant fractions were di~carded. and the procedure repeated it the tissue was nnt sutficiently lragmented. Alter incuhation the tissue w11s suspendl'd in Dulhecro's rnlldified 8agle·~ medium­HC: ~:lutrunine contorntng 1.'11'1 fe-tal calf serum, 200 units/ ml uf penicillin and 200 ,..gtml of ~treptomyrin and placed in disposable sterile pla~tic culture llask.o; !Falcon Plastir,l or glass slidr-bottom 11a.~kettes (Lah Te\'hl. Cultures were maintained at :l'i°C in an air/CO, 19;;/!l \' ' \'I incubator and the medium changed e\'ery 3 dan. Subcultures were handled in A ~imilar fa,.hion .

To ensure that the majority ol rell~ prest>nt in thP mono laver culture~ were pr1marilv fibroblasts and nut macrophages. which have ret'ently heen ~hown to contnm collagenase [13], tbe phagocytic index was determined IL4 ). Conlluent cultures were incubated for :!·1 br in frehh medium containing poly,;ivrene latex particules 11.811J in diameter t Oifco) The celt,; w·ere washed and the number ol part1cles rngested per cell were counted usinl!: pha~e contrast micr<>!'copy. In no instance was the a\·erage numht>r nf particles ingested greater than one, ~uggesting that there was little contamination ol the fibrobla~t cultures with macrupha~:es.

Starnrn11 procedure~. Afu•r sectioning. tissue sped­mens were htained by the indirect immunofluorescent tl'chmque. The >.ecllons wen: incuhuted with a I: 30 dilutwn nt anri-HSC gamma globulm (20 mg/mll for 30 mJn at rOtJm temperature. The slide~ were l(ently waghed :-1 -6 limes in Tri~- '\;aCt buUer, alter which they wen:• incubated with 11 1.20 dilutmn of FITC-labeled goat andrabbit fgG tor :lll mm at roum temperature and again washed in Tris :\aCt buJfer. Tissue sections were mounted in buffered glycerol !pH 8 .11) and stored. pro-

tected from light , at t•c until \'iewed . Fibroblasts adhering to the glass ~tide bot toms oi cell culture lla,kettes were ;;tained in an identical manner except that following removal ol the cultun· medium the slides wwt· rinsed with bufier and placed in 95% ethanol fixative tnr 30 ~ec. Specimens were exammed with a Le1tz Ort hoplan l1uurescence micro~cupc equipped with a Ploem vertic<~l illuminator racti\'ntinn filter BG-12. bar­rier filters K-510 and K-530). The light source Wlb a h11rh-prl'"sure mercur)' vapor :!00-lllltt lamp.

Pt-rm.idase stainin~: wa!< a(·l·nmplished using rabhit anti·HSC gamma globulin in the ~nme concentration as w1th the 11uorescence technique and goat antirabbit lgG conjugated to horseradish peroxidase as the second layer. The peroxidase-goat antirabhit lgG was incubated on the tissue ,E•ctimh for 30 min at room temperature and tht-n rim;ed with Tris· :\aCI hufter. The ~eC'tionh were subse­quently incubated for :10 rnlll in O.ll.'l :VI Tris HCI Cpll 7.Cil r.ontainin~t 0.3 ml{/ml diaminnbl'nzidine and 0.001 % H,O, and examined by !ltandltTd li~ht mkrosrop~'.

Controls . Control specimens were stained in an identi­<'JII fa,.hinn except that gamma globulin fractions of nnnimmun., rabbit serum or nt rabbit antibo\'ine semm albumin "·en• used as the tll'!\l layer in place of anti -HSC gamma ~lobulin. To funher deline the specificity of the staining, anti-HSC ~ramma globulin wa.~ absorbed with an exeess nt enzymaticall~· active human ~kin colla~:en ­nse. l'rcclplllltcs were removed by centrifugation prior to inl·ubutwn ot' the absnrhed antiserum with the tiJ:tsue ::;pccimens. Collagenase tor thesC' experiments was pre­pun•cl a~ previously de~cribed llo] and it;. activity a;;­SIH'.~ed on nati\'l' rct'on,;tituted "C'-laheled collagen gel~ 1!6]. Protein was determined according to Lowry et al IIi].

Raduummunoa.,,,ay {11r human .•km rolla11tmase. C ut. turl· medium was har\·e,ted at the termination of cell cultures and examined lnr immunoreactive collagenase hy radioimmunoasl>ay 171. Prior to use m the assay, medium was concentrated bv aU 'iO'ii> ammomum sullilte fractionation and re<'onstit'utf.'d in o small volume ot Tri~o -NuCl huffer. After dialy~i~. ,erial doubling dilution~ oft hi~ material wew examined for extent or no:;s-reartiv­ity with purified human skin cullu!!enase in standard inhihllinn curves.

RESl'l.l'S

Who{P ti:;sue sections. ;:\ormal human ~kin. stained by indirect immunnlluorescence using an­ti-HSC gamma globulin ns the first layer and FITC-Iabeled goat antirubbil lgG as the second layer is shown in Figure lA. Collagenase is localized primarily to the upper or papillar~ portion of the dermis. There is specific green fluorescence not only of cellular elements but also diffuse staining of the collagen fibers. suggesting that much of the enzyme present in vivo is extracellular and perhaps bound to its collagen substrate. Epidermal stain­ing is absent. and. although brightly refractil<' nonl1uorescent tihers are present in the mid- and lower dermis, specific fluorescence appears to be limited primarily to the papillary dermis with minimal enzyme lt>calized Ill the deeper dermal layers. Controls (Fig. lBJ in which nonimmune rabbit gamma globulin or rabbit antibo\'ine serum albumin was substituted for anti-HSC gamma globulin sbow no staining eit ber of the cellular elements or of the collagen. Only refractile tiber.:;

LOCALIZATION OF HUMAN SKIN COLLAGENASE 363

FIG. 1: Fluorescence photomicrograph!; of whole human skm (A) Section stained with rabbit anti-HSC gamma globulin followed by F'ITC-labeled goat anti rabbit lgG. Note the inten~e specific fluorescence of the papillary dermi!;. x 100. (8) Control section st.ained with non immune rabbit gamma globulin and FITC'-labeled goat antirabbit lgG. Only nonfluorescent, bri,;ht, refractilE' fibers can be seen in the mid- and lower dermis. 100.

are present 111 th«' mid- and lower dermis. Direct stainmg of the sect ions with FITC -labeled goat

~ antirahbit lgG was al~;;o nonreactive Fibroblast culture:.. Fibroblasts obtained from

human skin explants and allowed w reach near confluency were examined in a fashion identical to sections of human skin. As shown in Figure 2A and B. the fibroblasts display a granular and frequently reticulated staimng pattern in the perinuclear

t- region consistent with the organization of the endo­plasmic reticulum. At this level of resolution it is not possible to determine whether the granular stairung pattern indicates that collagenase is pres­ent within cytoplasmic vacuoles. Diffuse staining is occasionally present in some of the long cell

~ processes. ln cells undergoing mitosis, fluorescence l staining is homol{eneouH and of greater intensity

(Fig. 2C). Whether this merely reflects greater t cytoplasmic dens t ,. or represents greater synt he!iis

ot the enzyme b~ cells during mitosis is unclear at present. Staining of the fibroblasts is not evident when either nonimmune rabbit gamma globulin or rabbit antibovim· serum albumin is used in the

~ place of specific anti-H 'C gamma globulin as the fJTSt anti bod~.

~ To confirm these observations, fibrobla::.ts were ~tained with horseradish peroxidase coupled to

goat antirabbit IgG. Indirect immunoperoxida;;e stainmg abo reveal:. a granular perinuclear stain­ing pattern in 11hrohlasts virtually identical to that obtained with Jluore»cent techniques !Fig. :J1 . The nonimmune rabbit gamma ~-tlobulin control was again negati\1!. The u;;e of the immunoperoxidal-e technique should permit more preci:;e ~;>nzyme localization at the resolution ol the electron micro­scope.

Further ;;pec!licit) of the fibroblast staining was ascertained h~· reatting the celb with anti -HSC gamma globulin which had previous!) been ah­!iOrbed with an excess of human skin collagenast>. Pretreatment of the unti:,erum in this f~:~shion completely abolished all specd'ic stainmg of the cell;;.

Radioimmunoassay of wlture medium . In addi­tion tu demons! rating collagenase tmmunocvto­chemically. fibroblasts are capable of elahorattng collagenase into the culture medium as a!>sessed bv radioimmunoassa~. F'igure 4 shows that the stan­dard curve obtained with serial doubling dilutions of culture medium ts identicalm its cross-reacth­tty with the standard cune of purified human skin collagenase. ldenttty of the immunoreactiw mate­rial with pure human skin collagenase b\ this !'ensitive technique lend" further support to the

364 THE JOURNAL OF 11\'VESTIGATIVE DERMATOLOGY

Frc. 2: FluoreM·ence photomicrowapho of human skin fibroblasts. (A and Bl Fibroblasts at near confluency stained with rabbit anti-HSC' and FlTC-labeled goal antirabbit lgG. The cells show a granular and frequently reticulated perinuclear staining pattern. (Cl Fibroblast undergoing mitosi~ demonstrating an intense homogene-ous 11uorescence. Staining as in A. 1350.

concept that the fihroblast is the cellular source of collagenase.

OISCt.:SSION

lmmun(lcytochemicnl techniques used in this study have permitted the pred.se localization of human skin colla~ennse in whole sect ions of human skin. Pre' ioul:i in vitro studies utilizing explants of human skin mainrained on a short­term basi'-' in serum-free medium IISI!:iuggested that rhe papillary dermis wa..., responsible for collagenase })roduction. In addition, under certain circum~tances !18. 19), the epidermis also ap­peared to be capable of elaburatrng the enzyme. The present study shows that under normal in vi,·o eondirions. the upper. or papillary. dermili is the major site of collagenase production. U!';inl{ in­direct immunofluorel'cent ::;taining. ::;pecific 11uoresccnce was ohsen·ed in tibrohlast-like cells and on collagen fibers principally located in £he upper dermis Whether the epidermis is indeed capable of elaborating the enzyme under condi­tion,; such as wound healing ll8] is currently under inve~ligation.

LOCALIZATION OF HUMAN SKIN COLLAGENASE 365

' ·~ . Ftc. . :l: Light micrograph ot connuent human skin fibroblasts stained "'ith rabbit anti-HSC followed by

p!!ruxidase-labeled goat antirabbit l~G. 1000

N

·o >< 18 c: .E ' !!! c: 6 ~

] >- 14 '0 0 .0

c 0 12 $2 '0 c: ~ 0 .0

l) CJ')

I I

" ~

I 5 10 50 I()()

Human skm collagenase (ng/0 I ml)

ftG . 4: Cross-relctivitv of fibroblast culture medium with human skin t·ollagenase in tht> radioimmunoassay. Aliqunts of 100 1-l of either electrophoretically pure human skm collag£nase !Ot containing 0-50 ng of protein

• or -erial doubling dilutions ol fibroblast culture medium re l wert> reacted with ant1-HSC gamma globulm (I : :.!!)00) and '"1-HSC 19600 cpm). In controls, using an equivalent con cent ration of nonimmune rabbit gamma globulin. 8.3 percent of the radioactivity was precipi ­tatt>d

Our previous studies have indicated that colla­genase can be detected immunologically in tissue extracts of human skin [2, 19] and rheumatoid synovium [20] and that following chromatographic separation from serum inhibitors of collagenase, at least part of this immunoreactive material is also enzymatically acti\•e. Immunofluorescent staining patterns indicate that much of the enzyme in vivo is bound to its collagen substrate extracellularly. This suggests that the enzyme may be synthesized continuously and that at least partial control of collagenase activity occurs at an extracellular le,·el. Whether the extracellular control of enzyme activity is accomplished by serum or perhaps other tissue collagenase inhibitors and/or by a zymogen form of the enzyme is unknown. Thus far, our studies have failed to disclose a procollagenase for human skin [20 ], although the existence of a zymogen for tadpole [22, 23] and mouse bone collagenases [24, 2fl] ha!> been suggested.

Further localization of the cellular source of collagenase has been accomplished with cell cul­tures of human skin fibroblasts. The reticular staining seen in the cytoplasm of these cell~:> is consistent with the organization of the endoplas­mic reticulum. However, the granular appearance of both the immunofluorescent and peroxidase staining material in the fibroblasts may indicate that the enzyme is present, at least in part, in stor­age granules. Studies on the subcellular localiza­tion of the enzyme at the electron microscopic level will be needed before it can be determined whether human collagenase in skin fibroblasts can be demonstrated in lysosomes.

366 TilE .JOl H~AL Of (:-;\'t:STH:ATI\ E llF.R~1ATOLOC:\'

Although immunnrt'at'ti\'(~ <·ollogenase i~ dt'tt'Cl· abll· in fihrohlast cullurl' medium, it ha ... not ,.ct been possible to dett>ct enzym<· activity dire<·tl): an the medium. This is unden;tanduhiE' :-ince thr culture medium is rich in !ierum. a potent inhibitor of human skin collagenase [fi ]. The pm;sihility that the enzyme may be complexed to other as yet unidentifit>d inhibitors nr is prest•nt as an inactin• precur:-or ha~ not heen ruled out by these ;;tudies. Nevertheles .... the fact thut lhe immunoreactive material is identical with pure. lll'ti\e human skin collagenase in the radioimmunoassay pro\'ides strong evidence that fihrohlnst:- nn· capable ot elaboratinl! collagenase into the culture medium.

It is of interest that in cell cultures most. iJ not all, of the fibroblasts sh11w positive staining f(Jr colla~enase <Figs. 2. :H. Sin<'(• numerous studie!i [26) have shown that fibroblast ... in culture produre collagen, and hydroxyprolint>·rnntainin~ peptide:­are pre~ent in the cellrulture,.. used in this "tudy (unpublished observations), it b pos ... ihle that the same cell elaborates both ('()llal!rn and collagenase. The ust' ol double labeling flunresn•nt antihody techniques should provide a basis for future experi­ments un thi:-; aspect of coiiiiJ!E'Il mPLaholism.

Our appreciation goet> to Or. Hovel K. Hartman for h1s interest and man~· helpful ~UI(I!l·stlons .

REFERE;o.;CES

1. Eispn AZ, Bauer EA .. Jeffrey .JJ: Animal and human collagenoses. ,J Invest Dermatol 5li:359-373. 1970

2. Risen AZ. Bauer ~A .. Jeflrt>V .I.J: Human skin colla· gc.>nnse. The role nl serum ulphu·glnhulins in tht­control of act ivitv in t•it·o und in t'itro. Proc ~at I Acad Sci USA 68:248 251. 1971

3. Harris ED, DiBona DR. KranE' SM: Collagena~es in human synovial fluid. ,J Clin lnvel-ot 48:2104· '2113, 1969

4. Ryan J:O.:. Woessner ,Jf': Mammalinn collagena~t·: direct demonstration in honwgtnate~ ot involuting rat uterus. Biocbem Biophys Res Commun 44: 144-1·\9, 1971

i'> , Eisen AZ. Bloch K.J, Sakat 1': Inhibition ot human skin collagenase by humnn sPrum .• J Lab Clin 1\1ed 75:2.58-26.1. 1970

G. Bauer EA. Eisen AZ •. Jeltrcy .J.J: Immunologic rela ­tionship of a purified human -kin collagenase In uther human and animnl collagenases. Biochim Biophys Acta 206:152-160. 1970

7. Bauer EA. Eisen AZ .• JetTrev .J.J: Hadinimmunoassav of human collagenase. Specilicit\' of the assay and

quunutati\'e dctermmauon 111 m L'lt o and tn vitro human skin l"OIIagcnasc • • J Hiol Chern 2-17:1i679 f>68.'), ] 972

8. :-\nirn RC: Stundardtzlltton tn tmmunulluorc,ct'ncc. Clin Exp lmmunol a:41l'i -Ui7, 1968

9. Tlw TH. Feltkamp Tf<:W: l'nnjugllt ion of tluor~sn•in isu1 hioryanate tn antihnrlit·~ . I. l!:xpl'riml•nt" ont hi' conditions of runjugutwn. lmmunoln~ty 18:Hil5 1{7:1, 1970

10. The TH, Feltkamp TE\\: Conjugal um nf tluore.•cr m t.SOthiocyannte w nntibt)(lil' . II. :\ rl'produrible mrthod. Immunology 18:875-&11. Hl7<)

11. Ouchterlon} 0 : DiffuJon·m·gel methods for immu· nological analy:;is. l'rog Allergy 5:1-78, 1958

12 AHnmeas S. Rnutl'ille M: llhr<~structural I<X·uli~;a. tu•n of antihudy hy antiJ:l'll laheled with peroxt· dase. Exp Cell Re, 1i3:Hifl 167, 1961'1

IJ. ~~nior R.\1. Rielefeld OB, .Jetlrry .J,J : ('ollugf'nulyti!' 111·1 i\' ity in alvcnlnr mnnophn~es. Clin Rt>S 20:&'i, 1972

1-1 l't-arsall :0.::-\, \\ t'bl'r HS: The :\1onophagc. l'hilndel · phtn, Lea & Frhiger. 1!1711

15. Ei!ien AZ. Jeffre,· .J.J , nro,s J : !Iuman -kin collttgen­n"e Isolation and mechanism of attack on the collagen molecuh~. liiochim Biophys Acta 151: 637~4:.. 1968

lil. ='agai Y, Lapiere C:\1 , r.n~s .J : Tnrlpnlt• collagPnn"t·: preparation and punlirntinn. Biochemistry i1: :112:t-!H30, 1966

17 l~>wn· 0. Ro,;ebrouJ(h N, Furr A. Randall H: l'rutt•ln measurement wtth tht• Fnlin phenol reut.:cnl. ,J Btnl C'hem 193:265-271>, 19r,1

Ill. Eiwn AZ: Human skin n•Ilag~nn-.e: localization nnd distrthution in nnrmul hum11n ~kin. ,J lnvc~t D..r· matol 52:442-44H, 191).'1

19. Eisen AZ. Humnn skin c·ollugl!nase. Relut ionshp tn the p!!tho~tencsb of cpidermolysi' bullosa dystro· phica .• J Invest Derrnatnl .'\2:449--153. 1969

20. Bauer EA. E~>;cn AZ • • Jcfft'r\ .J.J: He!rulatinn ot \'Crtl'bratl' cnllagena'e Ill I iviiy 1n t TIL'O anclm VItro. ,J Invest Oermatol 59::10 f1f>, 1972

21. Rauer EA. Ei~en AZ, .Jl'ftre~ ,J,J: Studies on 1>uril it>d rheumatoid -.ynnvial rnllagcnnsl' in L'itro and in vil•o. ,J Clin Invest 50:211.'>6-206-1. 1971

22. llorpcr E. Bloch K.J, C. ross ,J: Tht· lynwgen nl tadpole collagena,c. Bwthl'mistry to::;o:\f>..:!IH 1, 1971

2.1 Harper E. Gru,, .J : Collnt.:enase, prul'ollagcnn~e 11nd acti\ a tor relationship~ in tadpole ti~sue culturl!ll Biochem Binph\s Hell C'ommun -18:1117 - llii2, 1972

24. Vacs r.: The releuse uf c·nllugenast: as nn inal't iw proenzyme b~ hone l'Xplnnt,.. in rulturc. Biorhem ,J 126:275-289, 1972

25 \'aes <: : :\lulttpl~ -11:p~ 111 the activation ol thto 1

inaetl\·e precursc:•r nf hone cullagena~;e b.Y trvpsm . FEBS Letterll 28: 19~ :!Oo. 1972

26 Gre£'n H. Gnldh!'rg B; Kim·tics ut collngL·n syntlws is II\' established mnmmr•li11n n•ll line> • .:-.:nture (f..ondl 200:1097- 10911. 1963