sign in sign up
<<
Home , Keloid, Necrobiosis, Necrobiosis lipoidica, Scleroderma

Lipoidica: Ultrastructural and Biochemical Demonstration of a Defect

Aarne Oikarinen, M .D ., Ph. D., Minna Mo rtenhul11 cr, M .D ., M atti Kallioincn, M .D., Ph.D., and Eeva-Riitta Savolainen, M. D., Ph.D. Coll agen Resea rch Unit, Unive rsity of O Ulll , Departmcnts of Dc n1l3tology (AO, MM ), Anatomy (AO), (MK), and Medi ca l Biochcmistry (AO, E-RS), Uni ve rsity of O ulu , O Ulll , Finl and

T en pati ents with lesions were stud­ lagen was unchanged in the affected skin . Fibrobl asts ied. Five pati ents had mellitus. The age of the es tablished fr o m affected skin synthesized less coll agen than patients vari ed from 15 to 73 yea rs and the durati on o f the ce ll s deri ved fr om hea lth y-looking skin. T he decreased col­ skin lesions w as fro m 2 to 20 yea rs. Histologica ll y, the lagen synthesis was due to a decreased amount of m essen­ lesions w ere characterized by degeneration of coll agen and ger RN A fo r type I procoll agen, m easured by hybridization elas tin. In som e lesions el as tin fibers could be seen in areas w ith a specific human cDN A cl one. T he producti on of devoid of normal-looking coll agen. Electron microscopy colla genase by these fib robl as ts was not in creased. O ur revea led loss of cross-striation of coHa gen fibrils and a marked res ults thus indica te that in necro bi osis lipoidica lesions, vari atio n in the diameter o f individual coll agen fibrils. T he coll agen fi brils are defecti ve and the amount of coll agen is concentrati on o f co ll agen, m easured by assay of hydroxy­ reduced, probabl y due to decreased synthesis of coll agen proline, a coll agen-specifi c amino acid, was markedl y de­ by affected fibrobl as ts. J f' lllesl Del'lll(1/o/ 88:227-232, 1987 creased in the lesional skin , but the ratio of type lIIll co1-

ecrobiosis lipo idica is a chronic skin that is lesio ns fro m 2 to 20 yeJrs. Nine of 10 were fe m ales and 5 had o ften associated w ith diabetes mellitus. Typica l le­ d iabetes mellitus. sio ns are irregul arl y delnarca ted , yell owish lesions on the shins. O n histologic examination there is Skin Biopsies Contro l sa m plcs (s ite-matchcd) were taken dur­ degenerati o n o r necro bi osis o f coll agen and poly­ ing therapeutic operati ons on 3ge-matched patients in the Der­ Nm o rp hi c cellular infiltrates composed o f ly mphoid ce ll s, fibro­ n13to logiea l C linic, O ulu University Central Hospi ta l, Finland . bl as ts, and histi ocytes [11 . Sometimes, the contains gran­ All the sa mples were taken in accordance w ith the Decl aration of ul o m atous foci composed of epithclo id cell s and giant cell s. T he Helsinki. degeneratio n and hyaliniza ti o n o f coll agen bundles adj acent to the Skin samples fro m necrobio ti c lesions of 10 patien ts wcre cx­ g ranulo m as are va ri able. The bas ic eti o logy of necrobi osis li ­ cised fo r li ght microscopy, and, in 4 patients, fo r electro n mi­ po idica is unknown. H owever, focal degeneratio n o f coll agen has croscopy. been suggested to have a central role in its pathogenesis II]. The Sa mples fo r li g ht microscopy w ere fi xed in 10% phosphate­ purpose o f the present study was to examine coll agen by ultra­ buffered fo rmalin , processed ro uti nely, and embedded in paraffi n. structural and biochemica l m eans in necro biosis Iipo idica les io ns. Secti o ns were cu t at 5 j.Lm and stai ned w ith H & E and Verhoeff­ van Gieson stains. Specimens fo r electron microscopy were fixed PATIE NTS AND M ETH O D S in 4% g lutarald ehyde, postfixed in 1 % osm ium tetroxide, and embedded in E pon. Ultrathin secti ons were stained w ith uran yl The bi o psy sa mples w ere taken fro m affected and nonaffected acetate and lea d citrate and examined in a Philips 410 LS trans­ (s ite-matched) skin of l O patients. T he clinica l chara cteri zation of mission electro n microscopc. the patients g iven in T able I dem onstrates heterogeneity in the Primary cell cultures were established by routine m ethods, and ages o f t he patients and the durati on of disease. T he age of the subcul tivated on plasti c cul ture dishes in D ulbccco's m odified pati ents va ri ed fro m 15 to 73 yea rs, and durati on of necrobio ti c Eagle 's medium (DMEM ) supplemented with 10% fetal ca lf serum, 50 j.L g/ml of ascorbate, 290 j.L g/ml L-glu ta m ine, penicillin (100 Manu sc ript recc iv cd Ma y 28, 1986; acce pted for publication August 22, U / ml), and strepto m ycin (100 j.L g/ml). Analyses of fib ro blast 1986. cul tures were ca rried out at 4-8 passages of subculti va ti o n. Supported in part by a grant from the Medica l Resca rch Coun cil of th e Acade my of Finland. Collagen Biosynthesis Studies Fibro blasts at confluence were Reprint req ues ts to: Aarn c O ik arin cn, M . D., Dc partmcnt of De rm a­ in cubated fo r 24 h in D ME M supplcmented as above, except that to logy, O ulu Unive rsity Ccntral Hos pital, K'\iaanif1(ie 50, SF-90220 O ulll , the serum w as repla ced w ith 2% dialyzed feta l calf serum, and Finl and . [I 4C]proline (2 j.L C i/ ml) w as added . After the labeling peri od the Abbrcv iati ons: mcdium was coll ected and pro teinase inhibitors were added to DMEM: Dul becco's modificd Eagle's medium give fin al con centrati ons of 25 111 M N a2EDT A, 10 m M N-cthyl­ GGT: gal actosy lh yd roxylysy l glu cosyltransferase PH: prolyl 4-hydroxyla sc m aleil11i dc, I m M phen ylmerhylsul fo nyl fl uoride, and I mM SDS-PA GE: sodium dodecy l sul fa te-polyac rylamide gel paraa mino benzam id ine. T he m ediul1l p roteins wen: then precip­ el ec trophores is itated by add ing ammoniul11 sul fa te to a fi nal concentratioll of

0022-202X/87/S03.50 Copyri ght © 1987 by T hc Socicty fo r In ves ti gati ve DC r11l3 tology, In c.

227 228 O IK A IUNEN lOT AL THE JOURNAL O F INV ESTIGATIV E

Table I. C linica l Data on Patients With N ecrobiosis Lipoidica

Age Diabetes, D uration Code Sex (years) (yea rs) C linica l Findings

1 F 69 Yes, 17 O n the anteri o r site of legs, lesio ns fo r 10 yea rs 2 F 73 N o O n the left leg, Ilumero us les ions fo r 8 years 3 F 56 No O n th e left leg, some les io ns for 3 yea rs 4 F 18 Yes, 6 O n the lower legs, lesions for 5 yea rs 5 F 65 N o Above the ankles, numerous les ions for 20 years 6 F 50 Yes, 10 O n th e legs, les ions fo r 2 yea rs 7 F 15 Yes , 14 O n the lower legs, lesions for 8 yea rs 8 F 49 N o O n the legs, lesions for 7 yea rs 9 F 53 N o O n ri ght legs, 2 les ions fo r 5 yea rs 10 M 20 Yes, 13 O n legs, some lesio ns

290 mg/ml, and th e precipitates were coll ected by centrifuga ti on by addin g soya-bea n trypsin inhibito r (50 /-L g/ml) , and coll agen ase for 30 min at 1 0,000,~ after stirring overnight at 4°C. This material activity was assayed by incubatin g samples with radioactive type was used fo r the assays of 1"'C lh yd roxypro lin e and total in cor­ I coll agen, as described previo usly [1 2], The coll agenase activity po ration of I'IC radioactivity by a specifi c radi ochemi ca l method was expressed as degradation of 31-:1-labeled coll agen, dpm X 12], h - '/ mg D NA, The ce ll layer was rinsed with phosphate-buffered saline and Analyses of Total Collagen and Collagen Types The amount the ce ll s scraped w ith a rubber poli ce lllan into 2 ml of 0.4 M of hydroxyprolin e, a measure of colla gen, was determined by a N aCI, 0, 1 M Tris-HC I, pH 7,5, contain ing th e proteinase inhib­ specific colorimetric assay [1 3], itors described above, T he ce ll s were soni cated at 60 Hz for 30 For determination of genetica ll y distinct coll agen types, ti ssu e s, T hese sa mples were used for assays o f total ce ll layer protein specimens were ho mogenized in 0,5 M acetic acid and submitted [31 and DNA 141, to limited proteolysis by pepsin (Worthington, 2 X crystalli zed), Part of the am 111 0 niull1 sulfate precipitate of the culture medium at a fin al concentration of 300 /-L g pepsin per mL The sa mples was used for 6% sodium dodecyl sulfate-polyacrylamide gel elec­ were in cubated for 3 h at 24°C, fo ll owe'd by 16 h at 4°C The trophoresis (SDS-PAGE) after reducti on 151 and radioactive pep­ pepsin -solu bilized m aterial was recovered by centrifugation for tides were visualized by flu orography 16J, 60 min at 37,000g at 4°C, and the insoluble material was subjected Assay of Type I Procol\agen mRNA Steady-State Level to further pepsinization as above, The supernatants ' containing Total RNA was isolated as described previo usly 17] and used for th e pepsin -solubilized materi al were combined, and protease in­ dot-blot hybridization assay [8J, The RNA sa mples were dotted hibitors at the concentrations indicated above were added, The (0,3-0,9 /-L g of total RNA) onto nitrocellulose paper, and the pH o f the sa mples was adjusted to 8,5 by adding 1 M Tris, and fi lters were hybridized w ith recombinant plas mid Hf677 con­ th e sa mples were in cubated for 60 min at 4°C to in activate pepsin, taining cDNA for human proal (I ) coll agen mRNA [9], The re­ T he sa mples were then dia lyzed ,against 0.4 M NaC l, 10 mM Tris­ combinant pl asmid was labeled w ith 132 P lnucleotides to a specifi c HC I, pH 7,5, containing the proteinase inhibitors, Coll agen was acti vity of5-8 x 10M cpm/ /-Lg by ni ck-translati on [1 0], T he amount precipitated by adding N aCI to a fin al concentration of 4.4 M of recom binan t 32 p_la beled plasmid hybrid ized to III RNA was and the precipitate was coll ected by centrifugation, visuali zed by autoradiograph y usin g Kodak X-Omat film and The SDS-PAGE was performed Ll sin g 8% polyacrylamide gels, cassettes w ith intensifyin g screens, T he autoradiograms were w ith and without delayed reduction w ith 2-mercaptoethanol [14) , quantitated by sca nning w ith a Kontes K 495000 densito meter The colla gen polypeptid es were visuali zed by stai'ling with connected to a Spectra-Physics SP4 \ 00 computing integrator. Coom3ssie Brilli ant Blue and quantitated with an automatic C0111- puting densitometer. Assays of Enzyme Activities T he ce ll s were gro wn to ea rl y conAu ence, harvested by trypsini za tion, and stored in the form Assay of DNA Synthesis The DNA synthes is activity was of a pell ct at -70°C for up to 3 weeks, A fter thawin g, they were estimated by detennining the in corporation of[3H]thymidine into ho mogenized w ith a tight Teflon-glass ho mogenizer (1200 rpm , the trichl oroacetic ac id-precipitable material. The cells were trans­ 50 ) in a cold solu tion contai ning 0,2 M NaC l, 0,1 M fer red to a 24- well p13te, approximately 2 X 10" cell s per well. glycine, 0, 1% (wt/vol) Triton X-IOO, 0,01 % (wt/vol) soya-bean After 48-h preincubation the cells were labeled with [3H)thymidine trypsin inhibitor, and 0,02 M T ris-HC I buffer, pH 7,5 (3-4 x (1 /-L C i/ well ) for 4 h, T he cell s were then washed thoroughly with 106 cell s/ml), The ho mogenates were centrifuged at 15,000 g for phosphate-buffered sa line and di srupted by incubating for 1 h at 30 min at 4°C and aliquots of the supernatants were taken for the 40°C in 1 % (wt/ vol) SDS, 1 mM Na2EDTA, and 50 mM Tris­ enzy me assays, HC I, pH 7,5, supplemented with 65 /-L g/m l of proteinase K , The Skin sa mples were homogeni zed w ith a Po lytron tissue ho­ lysates were then precipitated with 1 vol of 20% (wt/vol) tri­ mogeni zer in the solution described above, T he homogenates chl oroaceti c acid containing 100 /-L g of ca lf thymus DNA as a were centrifugcd at 15,000 g fo r 30 min, and ali quots o f the carrieI'. The precipitates were co ll ected by centrifugation at 15,000 supernatant were used for protein and enzy me assays, ,~ for 20 min, washed twice with 10% (wt/vol) trichloroacetic Pro lyl 4-hydroxylase (PH) activity was assayed by measuring acid , dissolved in Lumagel, and the radioactivity counted, the formati on of radioactive 4-hydroxyproline using a l'4C)proline­ For statistical analyses Student's I.-test was used , label ed type I protocoll agen substrate 1111 , RESULTS Galactosylh ydroxylysyl glu cosylt ransferase (GGT) activity was assayed by determining the radioactive glu cosylga lactosylh y­ Light Microscopy Findings The samples from the clinically droxylysine formed in a gelatinized ca lf sk in coll agen substrate necro biotic skin showed coll agen degeneration and at r111 , va riolls levels of the dermis, The necrotic fo ci were surrounded For assay o f coll agenase activity, fibroblas ts were cultured in by mononuclear cell infiltrates, hi stiocytic cel ls, lymphocytes, and se rum-free DMEM for 6 h , The m ediulll was coll ected and ali­ plas ma cell s, Histi ocyti c cell granulo l1la s were seen in 6 of 10 qu ots were subj ected to brief trypsin proteolys is usin g 0,1-10 /-L g cases (F ig 1), The elastica stain , Verhoeff-van Gieson, revealed trypsi n per ml for 10 min at 25°C, Trypsin was th en in activated the clasti c fibers of degenerated and necrotic areas to be short and VOL. IlH . NO. 2 FEU I ~UA I (Y I 'JH7 COLLAGEN DEFECT IN NECI( Ul OS IS LlPO IDIC A 229

Figure 2. E lectron microscopic pi cture fro lll necrobiotic skin . The de­ generated coll agen fibers show granularity, indisti nct bo rders, and loss of periodic banding. Bar = 0.25 J.L1ll .

T he so lubility o f coll agen in 0 .5 N acetic acid was also m easured. It is known that the solu bi lity of collagen is increased in situations w ith increased synthesis of co ll agen [1 5]. In affe cted skin, 2. 1 ± 1.3% (SD) and in contro l skin, 2.6 ± 0.3% of the coll agen was Figure 1. Lig ht microscopic picture from ncer'obiosis lipoidica skin. The acid-solu ble. collJ gcn and bundles are regular in no ninvolvcd "reas but arc lost in arcas of degencrati o n and of g iant and epithcloid ce ll g ranulo mas. Some granulo m as show traces of el as tin (a rrow). Vcrhoeff-van G ieson. X 140.

uncven in thickness, and disori entated, few in Ilumbcr, o r totall y lost. As a rule, the g ranulo m as werc totall y devoid of coll agen and elasti c fibers, but some g ranulo mas contained traces of elasti c fibers (Fig 1). Electron Microscopy Findings Coll agcn dcgencrati on and necrosis appeared as swoll en or frazzled fibers w ith disa ppearing periodic banding (Fi g 2), and as loose filamcntous and amorpho ll s, sli g htly g ranular material in mo re advanced cases. The clastic fibers were secn to be thick masses of ho m ogcno us elastica w ith central skeleton fibril s and peripheral elastic- fiber micro fibrils (not shown). Thc g ranulo m as were composed of histiocyti c and fi­ broblastic cells and som etimes of multinuclca ted histiocytic cell s. Som e g ranulomas were in the vicinity of the ca pillari es. Both histiocytcs and showed degenerati vc changes, cyto­ pl asmic peripheral villi , swellings, dividing lines in the cytopl as m , and pinching off of parts of the cyto plas m (Fi g 3). Lysosomal dense bo dies and other phagocytized m ateri al w ere also no ti ced, as well as a prominent roug h endopla smic reti culum in som e fib roblas ts. Studies The conccntration of h ydroxypro li nc, a measure of coll agen, w as m arkedly decreased in the skin bio psy specimens taken fro lll the affectcd skin o f thc paticnts (Fig 4). The m ean value was 9.4 ± 3 .7 (SD) J1.g/m g wet weight in affected skin, and 15.9 ± 11. 9 J1. g/m g in the nonaffected skin (p < 0.05). The concentratio n o f D NA, reRccting ti ssue cellularity, and the acti vities of PH (1' < 0.01) and GGT (p < 0 .01) wcre increascd in the affected skin compared to thc samples taken from nonaf­ fected skin or from hea lthy controls (Table II ). The ratios of types I and III coll agen were estimated fro m skin Figure 3. Electron micrograph o f epithelo id cell g ra nulo ma. T he hi sti o­ biopsies after limitcd pepsin proteolysis and 4.4 M N aCI prccip­ cyti c and fibroblastic ce ll s arc degenerating and show peripheral cyto­ itation. T he m ean proportion of type II [ co ll agen in the biopsy plasmi c vill i, i.e. , pinching off o f the periphera l cytoplasm (1 1, ;/1 nrnJ/I's). specimens from the affected skin of 3 patients (nos. 5, 6, and 9) A histi ocyti c ce ll w ith fa r-advanced degeneratio n and lys is contains dense was 23.1 ± 5.6% (SD), and in control sa mples 18.4 ± 8.1 % . lysosom;t1 bodies (Ih;ck (11"'0111). Bar = 1.0 J.L1ll . 230 O IKARINEN ET A L T H E JOURNAL O F INVESTIGAT IVE D ERMAT OLOGy

mRNA in the fibro blasts derived from the affected skin of the ~ pati ents (nos. 3 and 5) studied (Fi g 6). The amo unt o f typ e t 10\61 coll agen mRNA was 50% in cell line no. 3 and 57 % in cell lin!;) no. 5 compared w ith the corresponding va lue in ce ll s deri vecl. from the healthy skin of the sa me pati ents. :;:30 mi g h ~ .s:: Since in creased degradation of collagen by coll agenase .2- contribute to the reduced amo un t of collagen in necrobiosis Ii" :J'" ~ poidica lesions, coll agenase acti vity was determined from the m e" :J'" di a of ce ll cultures after brief trypsin acti va ti on. The res ults in" E'" 4 di cated that in the 3 cell lin es analyzed, the mea n acti vity Of' 'd,20 6 2 coll agenase was 14.3 X 10 dpm / h/ mg DNA in cell cultu res w 6 z I established fr o m hea lthy skin and 12.0 X 10 dpm/ h/ mg DNA. :; • 0 in fibro blas ts derived from affected skin. 2: • x>- 2 • ~ 1 0 6 DISC USSION >- :I: 5 In th e present study we demonstrated that large areas of skil1 8 • lesions in pati ents with necrobiosis lipoidica were devoid of nor, mal-lookin g collagen. In electron microscopy studies , the dis, appea rance of regular cross-striati ons of co ll agen fibrils was t he NA A most prominent finding . Variation in the di ameter of coll agen fibrils was also noted in the lesional skin. The reason for this is Figure 4. Conccntrati on of hydro xy prol in c in th e nonaffec ted (NA) and unknown. It is possible that at earl y stages of the degenerative affected (A) skin of pati ents with nec robi os is lipoidica and in th e co ntrols process, the regular structures of collagen fibrils are des troyed, (C). Hyd roxyprolin e was assayed as described in Palietl ls alld Merf lO ris. lea ding to abno rmal-looking fibrils. Another possibility could be T he ,,,,,,,bers bes ide the sy mbols re fer to the subj ects in Table I. that during regeneration , newly synthes ized coll agen m olecules are unable to form structurall y normal co ll agen fibri ls. Elastin, another maj or component o f dermal connecti ve ti ssue, was also Culture Studies T o examine th e mechanisms be­ decreased in the affected skin. Ho wever, there were cl ea rl y d e­ hind the decreased concentrati ons of coll agen in the skin o f nec­ tectable areas w here coll agen had totall y disappeared, and some robiosis lipoidica patients , fibro bl as t cultures were established elas tin fibers were still to be seen. This may refl ect the fac t that fr o m affected and nonaffected skin. Radioacti ve thymidine in­ elas tin is highl y resistant to the attack of proteolyti c enzymes [1 6] . corporati on, refl ectin g cell proliferation, was 3.04 X 103 dpm l lLg Sin ce va ri ous infl ammatory ce ll s produce elas ta se-like enzym es DNA in cell s from affected skin and 2.43 X 103 dpml lL g DNA [1 7], the decrease of elas tin in most of the lesions is plausi ble. in healthy skin fIbroblas ts. T otal protein synthes is, measured as In the bi ochemica l studies, th e concentration of hydro xyproli nc the in corpo rati on of rl4C ]prolin e in to the nondialyza ble fracti on, was markedl y reduced in skin specimens obtain ed from thc le­ was m arkedly decreased in 3 necro bi osis Jipoidica ce ll lines studied sional skin. T he concentrati on of hydroxyproline in the nonaf­ (patients 1, 3, and 5). C ollagen synthesis, assayed by the fo r­ fected skin was w ithin the range of hea lthy controls, which in­ mati on of [14 C]hydroxyproline, was also significantly reduced in di ca tes th at the defi ciency of coll agen was li mited to the lesion a] these sa me ce ll cultures (T able Ill) . The rel ative rati o of type I to skin of the pati ents. The concentration of DN A and the ac ti vities III coll agens was un changed (n ot shown). of PH and GGT were in creased in the affected skin. The in creased Reduced coll agen synthesis was also dem onstrated by SDS-gel DNA concentrati on could have been due to the accumulation of ele ctro phores is. T here was a decrease in the relati ve am ount of cellu lar infiltrate, w hi ch was o bvio us on hi stologic examination. proteins corresponding to the positions of proal (1) and proa 2(1) The elevated number of inflammato ry cell s could also have COn­ coll agen chains, especiall y in tb e cell s of patient no. 3 (Fi g 5) . tributed to the in creased activi ties o f PH and GGT , sin ce it is T he molecul ar sizes of the procoll agen molecules w ere un­ known that PH and GGT ca n be fo und in inflammatory cells changed. . [1 8], and PH and GGT activities have been fo und to be in creased, T he PH acti vity was sli ghtly decreased and the GGT activity e.g., in sa rcoidosis and lesions [1 9,20]. Constant was unaltered in fibroblas ts fr om affected skin (T able Ill). degenerati on and regenerati on of coll agen m ay also induce the Decreased coll agen synthesis by fibro blas ts established from syntheti c acti vity of fib roblasts in vivo, leading to increased levels the affected skin of necrobiosis lipoidica pati ents could be due of PH and GGT . either to a decreased amount o f procollagen mRNA or to de­ In order to study the mechanisms behind the reduced concen­ creased translati onal effi ciency. To study the mechani sm behind trati on of coll agen in ,necro biosis lipoidica, fibro bl as t cultures decreased coll agen synthesis, cellular RNA was dot-blotted onto w ere establi shed and studied in detail. The fibroblasts derived nitrocellulos<; fil ters and hybridized w ith a cDNA cl one comple­ from the les ional skin cl earl y show ed a decreased ca pacity to mentary to human proal (1) coll agen mRNA sequences. The hy­ synthesize coll agen in vitro. The reason for this is currently un­ bridiza ti on studies revealed a marked decrease in type I coll agen known. H o wever, it is possible that cell s were affected in vivo,

Table II. C oncentra ti on o f DN A a ~d the Acti vities of Prolyl 4-Hydroxyla se (PH) an d Gala ctosylh ydroxylysyl Glu cosyltransferase (GGT ) in the Skin o f Patients With N ecro bi osis Lipoidica" No. of DNA PH GGT Site of Biopsy Patients (ILg/mg wet weight) (dpm x 10- 3/mg protein ) (dpm x 1O - 3/ mg protein) P3ticnts· Nonaffcc tcd skin 10 2.41 ± 0.78 7.49 ± 3.48 6.06 ± 4.07 Les iona l skin 10 4.06 ± 2.8 1 15.59 ± 6. 14b 1"1. 54 ± 4.75" Control subj ects 6 2.04 ± 0.56 6.53 ± 2.20 6.26 ± 1. 53

"The biochemica l analyses were perfo rmed as descri bed in Paril'lIfs (l ml Me th ods and th e va lu es arc th e mea n :t SO. "S ignificantly different fro m nonafTcctcd skin; I' < 0.01. VOL. 88, N O.2 FEBRUARY 1987 CO LLAGEN DEFECT I N NEC ROBIOS IS L1PO ID IC A 231

Table III. Coll agen Productio n and the Activities of Pro lyl 4-H ydroxylase (PH) and Ga l ac t ~sy lh y dro xy l ysy l G lucosyltransferase (GGT) in N ecro biosis Lipoidica Fibroblas ts in C ulture Total In co rporation" 1"'C[Hydroxyproline" PH I, GGT" J Cell Lin e (code) (dplll x IO - / mg DNA) (dplll X IO - -'/ mg DNA) (dp ml /-L g protein ) (dpm l /-L g protein) Necrobiosis lipoidi ca Nonaffected (I ) 9.42 2.01 105.7 39. 1 Nonaffected (3) 8.89 2.00 170 .1 45. 1 Nonaffected (5) 7.88 1.85 179.5 40.0 Mea n ± SO 8.73 ± 0.78 1. 95 ± 0.09 15 1. 6 ± 40.5 41. 4 ± 3.2 Lesion (1) 3. 19 0.70 88. 1 57. 8 Lesion (3) 1. 10 0. 16 98. 1 43.6 Les ion (5) 1.87 0.43 131.3 49.7 Mea n ± SO 2.05 ± 1. 06' 0.43 ± 0.27' 105.8 ± 22.6 50.4 ± 7. 1 Controls (n = 5) 12.9 ± 3.8 3.00 ± 0.46 185.7 ± 42.6 49.5 ± 8.0

' Fibroblast cultures fro m necrobiosis li po idi ca patients and nOl'1nal human sllbj ects were bbclcd w ith / " C /prolinc for 24 h; / "C/hydroxyprolin c and to ta l radioactivit y were th en assayed from th e culrun.: medium :.15 described in Patii'lIl s ami N/I..' t" o(/s. I'Fibroblasts at ea rl y conAll cncy were harvested and ll scd for the assay of PH and GGT, as dcscribed in PllliclIIS mid Ml'IIi"ds. 'Signifi ca ntl y different fro lll no naffccted skin; p < 0.01.

i.e" by mediators of inflammato ry cell s [21-24], and were unable 4 to produce normal amounts of coll agen in vitro, T he reduced 1 2 3 synthesis of coll agen was found to be due to a decreased am ount of coll agen mRN A , indicating that coll agen synthesis was affected at the pretranslational level. T hi s is in agreement with previous 0 .9 t studies in kelo id cell s [25], fibroblasts [26]. and virus­ transformed cells [27], in which the rate of coll agen biosynthesis correlated well with the abundance of coll agen mRNA. T he reduced am o unt of coll agen could also be due to increased degradation of collagen in vivo. Vario us m etall o pro tein ases from 0 .6 • inflammato ry cell s and fibroblasts could participate in the deg­ radation o f coll agen in necrobiosis lipo idica (28). T he production of co ll agenase by fibroblasts was no r increased, h owever, in the present study, indicating that fibroblas t coll agenase m ay not have a central role in the degradative process. It thus seem s that in the 0.3 development of necrobiosis Iipoidica lesions, the accumulation of inflammatory cell s leads to degeneration of the connective ti ssue Figure 6, Determination of type I procoll agen mRNA ab undance in m atrix. In vivo there is still som e regenerative process, w hi ch is necrobiosis lipoidica fibroblasts. Total RN A was iso latcd and dotted on nitrocellulose filters (0.3, 0.6, and 0.9 /-L g/we ll). RNA was hybridi zed with .12 P-labelcd human proal (I) co ll agen cDNA probe. T he RNA­ [32p]DNA hybrids were visualized by autoradiograp hy and quantitated 2 3 4 5 6 by sca nning the band s with a densitometer. LIlII es 1 alld 2 represent assay of the cel l lin e fro m patient no. 3, and lalles 3 alld 4 that from patienr no...... 5. Lfllr es 1 alld 3 are fro m nonaffec ted skin and lall es 2 and 4 from affected . . . . skin of the patients. • • • I i I refle cted in increased level s of enzym es of collagcn biosynthesis. Future studies o n the interactio ns o f , inflam­ m atory cells, and vario us m ediato rs could further elucidate the m echanistic detail s behind the reduced connective tissue in nec­ robiosis lipo id ica [29).

The all,ilor5 ackll owledge 'h e expel" ,ec/lI1 ica l assislnll CC of Mrs. Eer)(l LciI'imiiki alld Mrs. Raija SO l'lllllllell , M.Sc. , alld 'h e helpjiil CO lllll re/ IlS by Professors Kari J. Kivirikko alld Ma rti Ha/ll/llksela. T il e proa1 (1) co llagell eDNA c/oll e was a giJi FO III Drs. M.- L. e hll , J. e. Myel'S, D. J. Pl'O ckop, "lid F. Ramirez.

Figure 5. The anal ysis of medium proteins of cell cultures fro m nec­ robiosis lipoidi ca patients by 6% SOS-PAGE. The cells were label ed with REFERENCES 14 [ C Jprolin e, th e medium proteins were precipitated with ammonium I. Lever WF, Schaumburg-Lever G: Nec robiosis Iipoidica, Hi stopa­ sulfate, and analyzed (10,000 dpm /well ) by el ectrophores is after reduc­ th ology of the Skin, 6th ed , Philadelphia, JB Lippinco tt, 1983, pp tion . The lhiek arrow indica tes the mi gration position of fibron ec tin , and 236-240 the Iilill fl rrorvs th e positi ons of proa I (I) and proa2(1) chains. Lalles 1 IlIId 2 represent sa mples from patient no. 1; /tlll es 3 O/rd 4 from patient no. 3; 2. Juva K, Prockop OJ; M o difi ~ d procedure for th e assay of H3_ Or and lalles 5 a/ld 6 from patient no. 5. Llllles 1, 3, all d 5 arc from nonaffccted Cr·-labeled hydroxyprolinc. An al Biochem 15:77-83, 1966 skin and lall es 2, 4, alld 6 from affected skin. 3. Bra'dford MM: A ra pid and scnsiti ve mcthod of quantitation of mi- 232 O IKARINEN ET AL THE JO UnNAL OF INVESTIGATIVE DEnMATOLOG'r

crogram quantities of protein utilizing the principle of protein­ 17. Werb Z, Gordon S: Elastase secretion by stimulated macrophages, binding dye. Anal Biochem 72:248-254, 1976 C haracteriza tion and regulation. J Exp Med 142:361 - 377, 1975 4. Brunk CF, Jones KC, James TW: Assay for nanogram quantities of 18. Myllyla R, Seppa H: Studies on enzymes o f collagen bio sy nth esi~ DN A in cellular homogenates. Anal Biochem 92:497-500, 1979 and the synthesis of hydroxyproline in macrophages and l11 as~ 5. Laemmli UK: C leavage of structural proteins during assembly of cells. Biochem J 183:311-316, 1979 the head of bacteriophage T 4. Nature 227:680-685, 1970 19. Vitto J , Hannuksela M , Rasmussen 'OG: Protocoll agen proline h y,\ 6. Bonner W M, Laskey RA: A film detection method for tritium­ droxylase activity in scleroderma and other connective tissue dis" labeled proteins and nucleic acids in polyacrylamide gels. Eur J orders. Ann C lin Res 2:235-239, 1970 Biochem 46:83-92, 1974 20. Kuutti-Savolainen E-R, Kero M: Enzymes of collagen biosynthesis 7. Oikarincn J, Ryhanen L: Cortisol decreases the concentration of in skin and serum in dermatologica l . I. Enzym es of the translatable type I procolla gen mRNA species in the developing skin. C lin C him Acta 96:43-51 , 1979 chick-embryo ca lva ri a. Biochem J 198:519-524, 1981 21. Jo hnson RL, Z iff M : Lymphokine stimulati on of collagen accu_ 8. T ho mas P: Hybridization of denatured RNA and sm all DNA frag­ mulation. J C lin Invest 58:240-252, 1976 m ents transferred to nitrocellulose. Proc N at! Acad Sci U SA 22. Jimenez SA, McArthur W , Rosenbloom J: Inhibition of coll agen 77:5201-5205, 1980 synthesis by mononuclear cell supernates. J Exp Med 150:1421-1431, 9. C hu M-L, M yers J C , Bernard MP, Ding J-F, Ramirez F: C loning 1979 and characteriza ti on of five overlapping cDN As specific for the 23. Postlethwaite AE, Kang AH: Inductio n of fibroblast pro liferation human proal (I)collagen chain. Nucleic Acids Res 10:5925-5934, by human m ononuclear leucocyte-derived proteins. Arthritis Rheurn 1982 26:22-27, 1983 10. M aniatis T, Jeffrey A, Kleid DG: Nucleotide sequence of ri ghtward 24. Duncan MIl., Per/ish J S, Fleischmajer R: Lymphokin e/mo nokine operati on of phage. Proc N atl Acad Sci USA 72:1184-1188,1975 inhibition of fibroblast proliferation and coll agen production : role 11 . Kivirikko KI, M yll yla R: Post-translational enzymes in the bi osyn­ in progressive systemic sclerosis. J In vest Dermarol 83:377-384, thesis of collagen: intracellular enzym es, M ethods in Enzymo logy, 1984 vol 82, Stru ctural and Contractile Proteins, Part A. Edited by LW 25. Abergel RP, Pizzurro D, Meeker C A, Las k G, Matsuo ka LY, Minor C unningham, DW Fredriksen. N ew York, Aca demic Press, 1982, RR, C hu M-L, Uitto J: Biochemical composition of the connective pp 245-304 tissue in and analysis of coll agen m etabolism in 12. Ryhanen L, Rantala-Ryhanen S, Tan E ML, Uitro J: Assay of col­ fibroblast cultures. J Invest Dermatol 84:384-390, 1985 lagenase activity by a rapid, sensitive and specific method. Col­ 26. K ii hiiri VM, Vuorio T, N anto-Salo nen K, Vuorio E: Increased type lagen Relat Res 2:1 17-130,1982 I collagen mRNA levels in cultured scleroderma fibroblasts. Biochim 13. Kivirikko KI, Laitinen 0, Prockop DJ: Modifica tion of a specific Biophys Acta 781:183-186,1984 assay for hyd roxyproline in urine. Anal Biochem 19:249-255, 1962 27. Rowe DW, Moen RC, Davidson JM, Byers PM, Bornstein P, Pal­ 14 . Sykes B, Puddle B, Francis M, Smith R: T he estimation of two miter RD: Correlation of pro collagen mRNA levels in normal and from human dermis by interrupted gel electrophoresis. transformed chick embryo fibroblasts with different rates of pro­ B iochcm Biophys Res Commun 72:1478-1480, 1976 coll agen synthesis. Biochemistry 17:1581-1590, 1978 15. Uitto J : Coll agen biosynthesis in . A review with em­ 28. O hlsson K, Olsson I: The neutral pro teinases of human granulocyte: phas is on scleroderma. Ann C lin Res 3:250-258, 1972 isolati on and partial characterization of two granulocyte collagen­ 16. Sa ndberg LB, Soskel NT, Leslie JG: Elastin structure, bi osynthesis ases . Eur J Biochem 36:473-481, 1973 and its relationship to disease states. N Eng l J Med 304:566-579, 29. Reynolds J): The molecular and cellular intera cti ons involved in con­ 1981 necti ve tiss ue destruction. Br J Dermatol 11 2:715-723, 1985