Isolation and Characterization of Type V Collagen from Human Post-Burn Granulation Tissues

Yoko Hashimoto, B.S.P., Tien-Yu Shieh, D.D.S., Ph.D . , Hisashi Aoyama, M .D., Ph.D . , Yohei lzawa, M.D., Ph.D., and Taro Hayakawa, M.D., Ph.D. Department of Biochemistry, School of Dentistry, Aichi-Gakuin University (YH, T-YS, TH), Nagoya; Department of Plastic Surgery, Aichi Medical College (HA), Aichi-ken; and Department of Plastic Surgery, C hukyo Hospital (YI), Nagoya, Japan

Sodium dodecyl sulfate (SDS)-polyacrylamide gel electro­phoresis of pepsin-solubilized collagens from post-burn granulation tissues revealed that type V collagen consisted of 3 a chains: a1(V), a2(V), and a3(V). The mean value (0.12 ± 0.01 SD) of the type V Ity pe I ratio in the gran­ulation tissues was significantly higher (p < 0.001) than that (0.03 ± 0.01 SD) of the ratio in normal skin. The average ratio of al (V) :a2(V):a3(V) of type V collagen pu­rified from the granul ation tissues was determined to be about 5:3:1. SDS-polyacrylamide gel electrophoresis pat­terns of 3 a chains were not affected in the presence or absence of 2-mercaptoethanol. Purified type V collagen was degraded by bacterial collagenase, but remained intact

Type V co llagen, which was first isolated from human placenta [1]. occurs with variable chain composition and has been locali zed in a variety of tissues (2) . Re­cently, native human placenta type . V collagen was resolved into 2 fra Ctions, one conta1l1111g al (V) and

a2(V) in a 2:1 ratio, and one con taining a1(V) , ll'2(V), and a3(V) in a 1 :1:1 ratio (3).

In our previous paper [4], interrupted gel electrophoresis pat­terns of pepsin-so lubilized collagens from human post-burn gran­ulation tissues showed the presence of the al chain of type V collagen. Recently 3 a chains of type V coll agen have been iden­tified in experimental granulation tissue in rats [5] , and a high ratio of type V to type I collagen has been noted in the ea rly stage of the hea ling of rabbit tooth-extraction wounds [6].

In this study, we have isolated type V collagen from human post-burn granul ation tissues and determined some of the phys­icochemical properties of its 3 constituent a chains.

MATERIALS AND METHODS

Patient Specimens Specimens of human normal skin and post­burn granulation tissues were obtained during surgica l operation and stored in a freezer (- 20°C) until used . Normal skin samples were all from frontal thighs of patients.

Preparation of Pepsin-Solubilized Collagen Human normal skin and post-burn gra nulation tissues were solubilized by limited

Manuscript received Novcmber 12, 1985; accepted for publication March 28, 1986.

This work was supported in part by a g rant from the Japan Burn Federation .

Reprint requests to: Taro Hayakawa, M .D. , Department of Biochem­istry, School of Dentistry, Ai chi-Gakuin University, 1- 100 Kusumoto­cho, C hikusa-ku, N agoya 464, Japan.

Abbreviation: SDS: sodium dodecyl sulfate

after tadpole collagenase digestion, in contrast to type I and type III collagens. Amino acid analyses of each a chain separated on SDS-gel electrophoresis of type V collagen revealed that all 3 a chains of type V collagen were poor in alanine, rich in hydroxylysine, and had high ratios of hydroxylysine/lysine, which are typical features of type V collagen. The purified type V collagen was further frac­tionated by ammonium sulfate into 2 molecular species, [a1(V)]za2(V) and al(V)a2(V)a3(V).

Our data demonstrate that type V collagen in prepara­tions from human post-burn granulation tissues consists of 3 a chains and can be reso lved into 2 distinct heterotrimers . J fllII est DeY/natol 87:540-543, 1986

pepsin digestion, and then collagen samples were collected by NaC l precipitation [7].

Electrophoresis Interrupted gel electrophoresis was ca rried out by essentially the same method reported by Sykes et al. [7], but with some modifications. The compositions of gel and electrode buffer were the same as those reported by Laemmli [8]. Gel elec­trophoresis was performed on slab gels with stacking gel and runnin g gel acrylamide concentrations of 3% and 8%, respec­tively. The gel dimensions were 135 X 140 X 2.0 111.m with twelve 5 mm-wide sample wells . Stacking was done at a current of 30 mA o Once the dye front had entered the running gel, the current was increased to 50 mA and kept there for 45 n1.in . The current was then switched off and the sample wells were filled with 20% 2-mercaptoethanol. This reducing agent was allowed to diffuse into the gel for 30 min before the current was again switched on; electrophoresis was then resllmed for a further 1.5 h.

SDS-polyacrylamide gel electrophoresis was carried out by the method of Laemmli [8].

Gels were stained with 0.25% Coomassie Brilliant Blue R so­lution for about 1 h and subsequently destained in acetic acid and methanol-destaining solution by diffusion with continllolls stir­ring. The scanning and quantitation of each collagen band were carried out using a Shimadzu CS-910 dual-wavelength TLC scan­ner (sample side, 600 nm; reference side, 400 nm). The relation­ship between sample size and stain uptake was demonstrated to be linear over the range of protein amounts used.

Purification of Type I, III, and V Collagens Pooled post­burn granulation tissues were solubilized by digestion with pep­sin; type I and 1lI collagens, and type V collagen, were purified according to the procedures of Epstein [9] and Ehrlich and White [1 OJ, respectively.

Animal and Bacterial Collagenase Treatments Highly pu­rified tadpole collagenase was a gift from Dr. Y . Nagai, Tokyo

0022-202X/86/S03.50 Copy right © 1986 by Thc Society for Investiga tive Dermatology, Inc.

540

VOL. 87. N O. 4 OCTO BER 1986

CXI (III) ­

CXI(l) ­CX2(1)-

normal skins

2 3 4

post-burn granulation tissues

5 6 7 8

-origin

Figure 1. Interrupted ge l electro phores is patterns of pepsin-solubilizcd co llagens from hum anno rm al skin ( /-4) alld post-burn g ranulatioll tissucs (5-8) with (1-3 and 5-7) or w ithout (4 and 8) reducing agent .

Medica l and Oental University. Bacteria l coll agenase (Clostridiul/l histolYliCIIIII , type IV g rade) was purchased from Worthington Biochemica l Co. and was further purified by gel filtration (Sephacryl S-200) . We confirmed th at the co ll agenase preparation (1 unit), w hi ch was su ffic ient to deg rade co mpletel y 100 J.Lg of type I co ll agen, showed essentia ll y no proteoly tic activ ity on the sa m e amount of bovine serum albumin as that of type I co llagen . The react io n mixture consisted of each type of co llagen (150 J.Lg) in 50 mM Tris-H C I buffer, pH 7.6, contai nin g 0. 1 M N aCl, 0.5 M g lu cose, and 5 mM CaCI2. Incubatio n was done for 17 h at 20°C with animal co llagenase (3 units) and at 3rC wi th the bac­te.- ial o ne (1 unit).

Amino Acid Analysis of Stained Bands from Poly­acrylamide Gels To prepare the sam ples for amino acid anal­ysis, the stained target protein band on th e slab gel w as cut out by a razor blade, put in to a hydro lyzing tube, and washed 5 times with 3 ml of acetone. After the acetone was removed with an aspirator, the sli ce was dried by flushin g w ith N 2 gas and then hydrolyzed under red uced pressure at 110 ± 1°C for 20 h in 5 ILl of thiog lyco lli c acid and 0 .5 ml of 6 N constant boiling H C !. The hyd ro lysa te was transferred to another tube and evaporated at 45°C using a C reig rotary eva porato r. The gel s lice rem ained as a swollen mass in the hyd ro lyzing tube and was washed with 0. 5 ml of distil .led water. This washin g was then combined with the dried hyd ro lysate an d evaporated as described above. The residue was final ly taken up in 45 J.LI of 0.2 M sodium citrate buffer, pH 2.2, and 25 J.LI of the resulting so lu tio n were subj ected to a mino acid analysis. * Amino acid an alysis was ca rried o ut using a S himadzu m odel LC -4A hi g h-perform ance liquid chro mato­g raph p 11 equipped with a fluorescence spectrophotometer

*Y. Hashimoto, S. Yamagata, and T. Hayakawa , manuscript in prep­ara tion .

TY PE V COLLAGEN mOM PO T-BURN GRANULATI O N 541

(Shimadzu RF-540) set at an excita tion waveleng th of 348 nm and em ission wavelength of 450 nm.

Ammonium Sulfate Fractionation of Type V Collagen T he type V collagen preparation isolated from human post-burn g ran­ulation tissues w as dissolved in 0.5 M acetic acid at a concentration of 0.5 m g/ml and then subj ected to am m onium sul fa te fraction­ation as described by Niyibizi et al [3].

RESULTS

Figure 1 shows interrupted gel electrophoresis patterns of pepsin­solubilized collagens from human no rm al sk in samples and post­burn g ranulation tissues . Type V coll agen consists of3 a chains­ad (V), a2(V), and a3(V)-and th ey mi g rated at similar rates with and without reduction and showed up between 0' 1 (I) and al (III) on th e electrophoretogram. N either the prescnce no r absencc of reducing agent made any difference in the mobility of each a chain of type V co llagen. Ta ble I summarizes th e conten ts of different collagen types and their ratio in 3 cases of each human norm al skin and post-burn g ranul atio n tiss lle. The mean va lue (0.12 ± 0.01 SD) of the type V Itype I ratio in th e g ranul ation tissues was signifi can tl y hig her than that (0.03 ± 0.01 SO) of the ratio in normal skin.

As shown in Fig 2, no ne of the a chai ns of type V coll agen was affected by treatment w ith tadpole coll agenase; under the same conditions, however, both type j and type ITI coll agens were completely degraded in to 2 fragments , TC A and TCB 0 " the other hand , treatment with bacterial co llagenase prior to th e elec­trophoresis resulted in almost complete disappearanc.e of all 3 a chains of type V coll agen together with type I and type III col­lagens.

The amin o acid compos itio ns of the 3 a chains of type V coll agen separated on SOS-polyacrylamide gel are g iven in Ta ble II together with those of the 2 a chains of type I co ll agen from hum an no rm al skin.

The average ratio of 0'1 (V):a2(V):a3(V) fro m 5 independent ana lyses was calculated to be abo ut 5:3:1 .

T he purified type V coll agen preparation was furth er subjected to ammonium sulfate fractionation fo llowed by SOS-po ly­acry lam ide gel electrophoresis of the resultin g fractiun. The first pellet of the 2 precipitations consisted m ainly of [0'1 (V)ha2(V) (2, in Fig 3), and the second pellet w as a mi xture of 2 distinct fra ctio ns (3, in Fig 3). The final supernatant was comprised mainl y of 0' 1 (V)a2(V)a3(V) (4, in Fig 3). The chain ratios determined by densitometry were 2: 1 for [0'1 (V)ha2(V) and 1:1:1 for 0'1 (V)a2(V)a3(V).

DISCUSSION

We previously showed that the ratio of type III to type I collagen (0.37-0.72) in post- burn g ranulation tissues is sig nifi ca ntly hig her than that in normal sk in (0. ]7 ± 0.04 SO) [1 2]. T his is also ev ident from the data o n the sa me ratio shown in Tab le I.

Recently, we demonstrated that fa irl y hig h anim.a l collagenase activity can be extracted directl y from human post-burn g ranu-

Table I. Contents of Type I, II I, and V Collagens and Their Ratios in Human N o rm al Skin and Post-Burn Gra nulation T issues

Tissucs

Normal skin I 2 3

Mcan ± SD Granulation 1

2 3

Mean ± SD

Type I

90.7 80.7 84.2

85.2 ± 5. 1 66.0 67.5 64.2

65 .9 ± 1. 7*

Content (%)

Type III Type V

7.7 1. 6 16.4 2.9 12.6 3.2

12.2 ± 4.4 2.6 ± 0.9 27.0 7.0 23.8 8.7 28.3 7.5

26.4 ± 2.3* 7.7 ± 0.9*

*0.001 < P < 0.01, " 1' < 0.00 1, '.'1' < 0. 1 ill co rresponding va lue from normal subjects.

Ratio

111 /1 VII V / 111

0.08 0.02 0.2 1 0.20 0.04 0. 18 0. 15 0.04 0.25

0. 14 ± 0.06 0.03 ± 0.01 0.2 1 ± 0.04 0.41 0.11 0.26 0.35 0.13 0.37 0.44 0. 12 0.27

0.40 ± 0.05* 0. 12 ± 0.01** 0.30 ± 0.06***

542 II AS HIMOTO ET AL

type V collagen

tadpole collagenase + bacterial collagenase - +

crl cr2

type I coIIoQen

+ + -

-

type III collagen

+ + ... -

-Figure 1.. Susccptlbllltles of type V collagen to both animal and bacten"l co llagcn3ses and their comparison with those of type I and III colla gcns. Elec trophoresis was run w ith 8% pol yacrylamide gel by the Laemmli mcthod . T ype III co llagen was subjected to gel electroph oresis after 2-mcrca ptoethan ol reduction.

la tio n tissues, and this collagenase breaks down type I co ll agen preferentiall y over type III. Thi s coll agena se activity was fur­thermore shown to decrease signifi cantly, keepin g a good co r­relatio n to th e decrease of the rati o o f type III co ll agen to type I durin g 2 weeks followin g thin skin autograph [13]. T hese findin gs suggest that the pre fe rentia l deg radation of type I co llagen by g ranulation-t issue coll agen ase mi ght be at leas t o ne of th e factors that a ffects th e elevatio n of th e type II I/type I ratio . As shown in Fi g 2, type V coll agen was not affected at all by animal coll agenase treatment und er th e sa m e conditio ns in which bo th type I and III colla gens were completely deg raded. Putting all th ese results to­geth er, there mi ght possibly be a relative in crease in type V col­lagen in human post-burn granulation tissues together with an absolu te in crease.

Table II. A min o A cid Compos itions of Three a C hain s of Type V Collagen and Two a C hains of Type I Coll agen

Residues per I (JOO Residucs

Type V Type I Collagen Collagen

Amin o Acids 0'1 0'2 a3 a l a2

Hyp 95 96 103 104 94 Asx 54 55 50 43 47 T hr 20 28 21 16 19 Ser 20 3 1 26 30 3 1 Glx 99 86 94 78 7 1 Pro 128 109 110 129 11 8 Cys 0 0 0 0 0 Gly 343 350 334 336 330 Ala 41 59 42 110 103 Va l 19 28 25 19 35 Met 5 5 4 7 8 lie 17 14 16 7 16 Leu 38 33 48 18 3 1 Tyr 2 1 7 1 3 Phe 12 11 10 12 12 His 5 7 12 2 7 Hyl 39 20 38 9 8 Lys 17 14 15 28 20 Arg 45 54 45 49 49

THE JOU I{NAL OF IN VEST IG ATI VE DERMA·r O LOGY

CXJ(V) ,

CX3(V)­CX2(V) -

2 3 4

Figure 3. SOS-5% polyacrylamide gel electropho resis patten IS of 0' 1 (V)a2(V)a3( V) and [a l (V) ba2(V) fra ctionated by ammonium ,;.lIfate precipitation . . ' , Type V co llagen used for the fra ctionation ; 2, la I (V) ba2(V); 3, a mi xture of both molecular spccies; 4, a l (V)a2(V)a3(V).

All 3 a chains o f type V co ll agen are poo r in A la, rich in Hyl , and have hi g h ratios of H yl/Lys, which are typica l features of type V co ll agen.

It has been known th at type V coll agen contains 3unique ch ains: a1 (V), a2(V), and a3(V) . C urrent data indi ca te that the chains m ay be co mbin ed to form 2 m ost prevalent heterotrimeri c mol­ecules, lex1 (V)ha2(V) and a1 (V)a2(V)a3(V), as well as a ho m o ­trimeri c molecule composed solel y of al (V) ch ain s 12·1. Recentl y, Niyibi zi et al 13] confirmed th e ex isten ce of 2 maj o r molecul a r assemb lies of hum an p lacenta l type V co ll agen w hich could b e reso lved into 2 fraction s witho ut any denaturin g agent like 2 M

un.:a which was used in prev io us works [1 4,15] . U s in g a1111110-

nium su lfate fractionation, we have shown that a type V collagen preparation from human post-burn g ranul ation tissues ca n also be resolved into 2 fractions, onc conta inin g a l (V) and a2(V) in a 2:1 ratio and one conta inin g al(V), a2(V), and a3(V) in a 1:1:1 ratio.

The un fractionated type V colla gen preparatio n , o n the o the r hand , showed the ave ra ge ratio of a1(V):a2(V):a3(V) to be ab out 5:3:1. Judg in g fro m this rati o togethe r with the ratios of the 2 fraction s m entio ned above, we can ass um e that type V co ll agen in hunl3n post-burn g ranulation tissucs contains lal (V)ha2(V) and al (V)a2(V)a3(V) in an approximate m o lar ratio o f 2 :1.

REFEREN CES

I. Burgeso n RE, EI Ad li FA , Kaitila II , Hollister OW: Feta l membrane coll agens: idcntifica tion of two new collagen alpha chains. Proc Natl Acad Sci USA 73:2579-2583, 1976

2. Millcr EJ: Reccnt information on the chemistry of the co llagens, The Chemistry and Biology of Minerali zcd T issues, Edited by WT Butler. Birmingham , AL, Ebsco Med ia, Inc, 1985, pp BO-93

3. Ni yib izi C, Fictzek PP, va n der Rest M: Hum an placentJ type V collagens. J Bioi C hcm 259: 14170- 14174, 1984

4. Ha yakawa T, Hashim oto Y, Myokei Y, Aoyama H, Izawa Y: The effect of skin grafts on the ratio of coll agen types in human pOSt­burn wound tissues. Connect Tissue Rcs 9:249-252, 1982

5. Hering TM, Marchant RE, Anderson JM : T ype V collagen during grallu lat ion ti ssue development. Exp Mol PathoI 39:219-229, 1983

6. Kurita K. Hashimoto Y, Takci T , Ka wa i T , Ha yakawa T: C hanges in co llagen types durin g the healing of rabbit tooth ext raction wounds. J Dent Res 64:28-32, 1985

7. Sykcs B, Puddle B, iOran cis M, Smith R: The estimation of two coll agens from hum an dermis by intcrrupted gel electrophoresis. Biochcm Biophys Res Commun 72:"1472- 1480, 1976

8. Lacmmli UK: C leavage of stru ctura l proteins during the asse mbly of the head of bacte ri ophage T4. Nature 227:680-685, 1970

9. Epstcin EH Jr: la l (111)[3 Human skin collagen. Release by pepsin

VO L. R7. N O.4 OCT O UEI1 1')86

d igestion and preponderance in fetal life. J Bioi C hem 249:3225-323 1, 1974

10. E hr lich HI' , White BS: T he identifi ca tio ll of OIA and OlE co llagen chains in h ype rt ro phi c sca r. Exp Mo l Pa tho l 34: 1-8, 1981

11 . Ishida Y, Fujita T, Asai K: N ew detect io n all d sepa ratio ll m ethod for amill o ac ids by high- perfo rmance liq uid chro m atography. J C h ro m atog r 204: 143- 148. 198 1

12. Hayakawa T , H ashimoto Y, M yokei Y, Aoya ma H , Izawa Y: C hanges in ty pe of co llagen durin g the deve lop m ent o f hum an pos t- bul'll h ypert ro ph ic sca rs. C lin C h il1l Acta 93: 11 9-'123, 1979

TYPE V C OLLAGEN FRO M POST-I3UR N GRANU LATION 543

13. Kishi J , H ashim oto Y, Aoya m a H , Izawa Y. H ayakawa T: Direc t ex tractio ll of co llagenase fro m hum all pos t-burn wou nd tissues. B io m ed Res 5: 149- 156, 1984

14. Rhodes RK, Mi ller EJ: Evidell ce (o r the c:o-;s tence of an 01 1 (V)0I2(V)0I3(V) collagen molecule in hum an placental tissue. Coil Rei n cs 1:337-343, 198 1

15. M adri J A, Focl lm cr HG, Furthmayr H : Type V co llagens of the hum an placenta: trim cr OI-chain compos itio n, ultras tructural m o r­pho logy and peptide analysis. Coil Rei Res 2: 19-29. 1982