separating type-V from type-IV collagen, and high- mol.-wt. (HMW) aggregates from 7 S collagen. Uterus and maternal villi contain 2 forms of type-V collagen --.
493
Bioscience Reports 2, 493-502 (1982) Printed in Great Britain
I s o l a t i o n and n a t i v e c h a r a c t e r i z a t i o n of c y s t e i n e - r i c h collagens from bovine placental tissues and uterus and t h e i r r e l a t i o n s h i p to types IV and V collagens M. Z. ABEDIN, Shirley AYAD, and 3acqueline B. WEISS* Department of Rheumatology, University of Manchester Medical School, Manchester, MI3 9PT, U.K. (Received 24 May 1982)
A s i m p l i f i e d p r o c e d u r e for the f r a c t i o n a t i o n and p u r i f i c a t i o n of different collagen types from various tissues is described which is particularly efficient in s e p a r a t i n g type-V f r o m type-IV collagen, and highmol.-wt. (HMW) aggregates from 7 S collagen. Uterus and maternal villi contain 2 forms of type-V collagen -{~I(V)}2c~2(V) and {al(V)~2(V)cx3(V)}- which have been s e p a r a t e d on DEAE-cellulose. Uterus however a p p e a r s to be the r i c h e s t s o u r c e of both HMW aggregates and the {cxl(V)a2(V)c~3(V)} collagen, and a p r o b a b l e r e l a t i o n s h i p b e t w e e n t h e s e c o l l a g e n s is discussed. The collagen molecule is known to exist in at least five genetically distinct forms (types l-V) of tool. wt. approximately 300 000 (Bornstein & Sage, 1980). Although types Ill and IV collagens contain cysteine the amount is small and of the order of 3-4 residues per I000. Recently two new cysteine-rich collagenous aggregates have been isolated. Both have a similar high tool. wt. (350-360 000), consist of several subunits, and are extensively disulphide bonded (containing over 15 residues cysteine/1000). The first of these, 7 S collagen (Risteli et al., 1980)9 has recently been established as the crosslinking domain of type-IV collagen (Kuhn et al., 19gl) and is therefore derived from basement-membrane structures. 7 S collagen has so far only been isolated by virtue of its stability to bacterial collagenase digestion at 20~ (long form) and consists of subunits of approximately 27 000. Treatment with bacterial collagenase at 37~ converts the long form to a form which has a tool. wt. of approximately 200-225 000 consisting of subunits of about 17 000 mol. wt. (short form) (Risteli et al., 19g0). The second collagenous aggregate [high-molecular-weight (HMW) aggregate] consists of three major subunits with tool. wts. of 3555 000 (Chung et al., 1976; Furuto & Miller, lg0; Laurain et a l , 1980; lander et al., 1951). The larger of the three subunits is more acidic, consists of alternating collagenous and non-collagenous regions (Furuto & Miller, 1980), and has a non-helical conformation, whereas the other two basic subunits are responsible for the helical nature of the native aggregates (Furuto & Miller, 1981). Unlike 7 S collagen, this collagen is generally believed not to be of basement-membrane origin. *To whom correspondence should be addressed. 01982
The Biochemical Society
494
ABEDIN ET AL.
We describe a simplified procedure for the isolation and purification of the different collagen types which has been applied to the separation and characterization of placental and uterine collagens. Materials
and Methods
Preparation of collagen types
Bovine amnion, chorion, foetal and maternal villi, and pregnant uterus were digested with pepsin (Sigma P5264) in 0.5 M acetic acid, pH 2.8, at #~ for 36 h at an enzyme(dry w t . ) - t o - s u b s t r a t e ( w e t wt.) ratio of l:100. The pepsin digests were then fractionated and purified as outlined in Fig. 1. Chromatography in the native state
CM-cellulose chromatography of the 1.2 M and 2 M NaCI (pH 2.g) precipitates was carried out on a CM52 (Whatman) column (1 x 12 cm) equilibrated with 0.0t~ M sodium a c e t a t e (pH tt.8) containing 2 M urea at 12~ (Kresina & Miller, 1979; Ayad et al., 1980). Fractions were eluted usng a linear salt gradient (0-0.2 M NaCI in a total volume of 600 ml)~ dialysed against 0.l M acetic acid, and freezedried. DEAE-cellulose chromatography of the 1.2 M and 2 M NaC1 (pH 2.8) precipitates was carried out on a DE52 (Whatman) column (1 x 15 cm) equilibrated with 0.05 M Tris/HCl (pH 8.6) containing 0.02 M NaC1 and 2 M urea at 12~ (Bentz et al., 1978). Fractions were eluted with a linear salt gradient (0.02 M-0.32 M NaCI in a total volume of 800 ml), dialysed against 0.1 M acetic acid, and freezedried. Bacterial collagenase digestion
C o l l a g e n s a m p l e s were incubated with bacterial collagenase E. (Calbiochem) in 0.02 M Tris/HCl (pH 7.t~) containing 0.5 M NaCI, 5 mM CaCI 2 at 20~ or 37~ for 24 h at an enzyme-to-substrate ratio of 1:50. The protease inhibitors N-ethylmaleimide (I0 mM) and phenyl methane sulphonyl fluoride (2 raM) were included to eliminate any non-specific proteolysis. Analytical methods
Electrophoresis on SDS/polyacrylamide gels (7% v/v) was performed by the method of Laemmli (1970) but without the use of a spacer gel. Samples for amino acid analysis were hydrolysed in constant boiling HCI at l l 0 ~ for 24 h under nitrogen and analysed on a 3eol 6AH amino acid analyser. Quantitation of the various collagens was assessed by hydroxyproline analysis of the various precipitates and supernatants obtained during purification (\Voessner~ 1961). Results
and Discussion
Preparation and purification of collagen types
A simplified procedure for the separation of various collagens has been developed (Fig. l) and its efficiency assessed by several criteria.
CYSTEINE-RICH
COLLAGENS
t~95
Three major precipitates were obtained by differential salt precipitation at pH 2.g: at 0.7 M, 1.2 M, and 2 M NaCI. The 0.7 M precipitate contained types I and III collagens accounting for 87-93% of the total pepsin-solubilized collagen and was not examined further. The 1.2 M precipitate was further fractionated by dialysis against 0.02 M Na2HPO ~ (pH 9.2), which easily separated types IV and V collagens. SDS/gel electrophoresis (Fig. 2A, lanes a and b) and amino acid analyses (Table I) of the phosphate supernatant were characteristic of type-IV collagen (Glanvitle et al., 1979; Kresina & Miller, 1979; Sage et al., 1979). The phosphate precipitate contained type-V collagen and occasionaIIy traces of types I and III collagens which were removed by s e l e c t i v e l y p r e c i p i t a t i n g t h e type-V c o l l a g e n by dialysis against phosphate-buffered saline (PB5) at 4~ This method was first used to p r e c i p i t a t e type-V c o l l a g e n in a form which induced platelet aggregation (Chiang et al., 1980). The pure PBS precipitate from placental membranes and foetal villi contained the a l ( V ) and c*2(V) chains only (lane c), but in the case of maternal villi and uterus an additional c~3(V) chain was observed (lane d; see also Fig. 3B inset). Pepsin digest (pH 2.8) from amnion, chorion, villi or uterus
0.7M NaCI
1 0.TMppt.
t 0.7M Sup.
(Types I and Ill)
1.2M NaCl
i 1.2M Sup.
1.2M p p t (1) dissolve in 0.1M acetic acid
2M NaCI (2) d i a l y s e v s 0.02M Na 2 HPO4 (pH 9 . 2 )
t phosphate
(I)
[ phosphate sup. (Type IV)
ppt.
I 2M p p t .
dissolve in 0.1M acetic
I 2M Sup.
(1) dissolve in 0.2M acetic acid
acid
(2)
1 PBS s u p (traces of types I/IIl)
(2) dialyse vs O.02M Na 2 HPO4 (pH 9 . 2 ) or PBS (pH 7.2)
dialyse vs PBS (pH 7 . 2 ) I PBS p p t (Type V)
I phosphate o r
I phosphate or
PBS p p t . (HMW a g g r e g a t e )
PBS Sup. (7S collagen)
Fig. i. Isolation and purification of collagens from pepsin digests of bovine placental tissues and pregnant uterus.
496
ABEDIN
ET
Fig. 2. A. SDS/polyacrylamide-gel (7% v/v) electrophoresis of various collagen fractions purified according to Fig. I. Reduction was carried out with the addition of dithiothreitol (DDT) prior to electrophoresis. (a) 1.2 M NaCI precipitate-phosphate supernatant (type-IV collagen) from villi (unreduced). (b) As (a) but reduced. (c) 1.2 M NaCI precipitate-PBS precipitate (type-V collagen) from placental membranes. (d) As (c) but from maternal villi. (e) 2 M NaCI precipitate-phosphate/PBS precipitate (HMW aggregates) from uterus (unreduced). (f) As (e) but reduced. (g) 2 M NaCI precipitate-phosphate/PBS supernatant (7 S collagen) from uterus (unreduced). (h) As (g) but reduced. B. SDS/polyacrylamide-gel (7%) electrophoresis of the 2 M NaCI precipitate-phosphate supernatant (7 S collagen) treated with bacterial collagenase at 20 ~ (a,b) and 37~ (c,d). Samples (b) and (d) were reduced with DTT prior to electrophoresis.
AL.
CYSTEINE-RICH
COLLAGENS
t497
l o ~, o~ o'~
e,t
.,4
la
ca em ~a r I> ! ll}
.,.4
oe~ ca
o~ ~
cag v ~ v
v
O
,la
,.-4
~
~ .,.4
o
49g
ABEDIN
ET AL.
The amino acid analyses (Table 1) of the PBS precipitate were typical of type-V collagen (Burgeson et al., 1976; Bentz et al., 1975; Sage & Bornstein, 1979). The 2 M NaCI (pH 2.8) precipitate could also be fractionated by dialysis against either 0.02 M Na2HPO # (pH 9.2) or PBS. On SDS/gel electrophoresis in the unreduced state both the phosphate (or PBS) precipitate (Fig. 2A~ lane e) and supernatant (lane g) hardly penetrated the 7% gel. On reduction, however, the two fractions were distinct~ the phosphate (or PBS) precipitate (lane f) and supernatant ( l a n e h) resembling~ respectively~ the HMW aggregate (Furuto & Miller~ 19g0~ 198t; Laurain et al., 1980; 3ander et a l , 1981) and the long form of 7 S collagen (Risteli et al., 1980). The three major components of the HMW aggregate exhibited the characteristic tool. wts. of 35 000~ 45 000, and 55 000~ whereas the 7 S collagen showed the expected pattern of several chains with tool. wts. ranging from 27 000 to 250 000 (representing dimers, trimers, etc., of the 27 000mol.-wt, subunit). The amino acid analyses of the 2 M NaC1 phosphate precipitate and supernatant (Table 1) also agree with those reported for the HMW aggregate and 7 S collagen respectively (Furuto & Miller, 1980; Laurain et at., 1980; 3ander et al., 1981; Risteli et al., I980). This m e t h o d of p r e p a r a t i o n of 7 S collagen is simple. Previously it had been prepared by a lengthy and tedious procedure involving d i f f e r e n t i a l s a l t p r e c i p i t a t i o n at acid and neutral pH~ bacterial collagenase digestion (which destroys other collagens and is wasteful), and ion-exchange chromatography (Risteli et al., 1980). Bacterial collagenase digestion
The collagenous components of the 2 M NaCl phosphate precipitate were unaffected by bacterial collagenase digestion at either 20~ or 37~ in agreement with the results of Furuto and Miller (1981) for t h e HMW a g g r e g a t e . In t h e c a s e of the 2 M NaCI phosphate supernatant, bacterial coIlagenase had no e f f e c t at 20~ (Fig. 25, lanes a,b). However~ t r e a t m e n t at 37~ produced two components, of 150 000 and ~0 000 mol. w t , which on reduction gave rise to a range of subunits 17-200 000 (Fig. 2B, lanes c~d)~ characteristic of the short form of 7 S collagen (Risteli et al., 1980). Ion-exchange chromatography in the native state
O n CM-cellulose (Fig. 3A) the 1.2 M NaCI precipitate (from maternal villi) was resolved into two peaks Muting at 0.04 M and 0.i M NaCl which were identified as types IV and V collagens, respectively (Fig. 2A, lanes a, b, and d). Any traces of types I and III collagens co-purifed with the type-V-collagen peak. The 2 M NaCI precipitate (uterus) also gave two peaks at 0.04 M and 0.i M NaCI~ which were identified as 7 S collagen and the H M W aggregate respectively (Fig. 2A, lanes g~h and e,f). DEAE-cellulose chromatography (Fig. 3B) of the 1.2 M precipitate (uterus) resulted in an unbound peak containing type-IV collagen (Fig. 2A, lanes a~b) and occasionally traces of types I and Ill collagens, and a bound peak which was partially resolved into two fractions (I and If). Gel electrophoresis (Fig. 3B, inset) indicated that fraction I
CYSTEINE-RtCH
COLLAGENS
499
Fig. 3. A. CM-cellulose chromatography of the 1.2 M NaCI precipitate from maternal villi ( ) and 2 M NaCI precipitate from uterus (..... ). The column (i x 12 cm) was equilibrated with 0.04 M sodium acetate, pH 4.8~ c o n t a i n i n g 2 M urea at 12OC. Fractions were eluted using a linear salt gradient 0-0.2 M NaCI over 600 ml. B. DEAE-cellulose chromatography of the 1.2 M NaCI precipitate ( ) and 2 M NaCI precipitate (..... ) from uterus. The column (I x 15 cm) was equilibrated with 0.05 M Tris/HCl (pH 8.6) containing 0.02 M NaCI and 2 M urea at 12~ Fractions were eluted using a linear salt gradient 0.02-0.32 NaCI over 800 ml. Inset shows SDS gels of starting material (S) and peaks I and II.
contained only czl(V) and ~2(V) chains (in an approximately 2:1 ratio) whereas fraction H contained approximately equal amounts of all three c~-chains ~ l ( V ) , cz2(V), and c~3(V). This observation agrees with the
500
ABEDIN
ET AL.
p r o p o r t i o n s of t h e s e chains found in human placental villi using phosphocellulose chromatography (Rhodes & Miller, 1981). The amino acid analyses of fractions I and II were very similar (Table 1). Only peak I was o b s e r v e d in the case of the 1.2 M precipitate from placental membranes and foetal villi (not shown). Chromatography of the 2 M precipitate (uterus) resulted in an unbound peak containing the 7 S collagen (Fig. 2A, lanes g,h) and a bound peak, containing the HMW aggregate, which eluted in a similar position to type-V collagen containing all three m-chains. The elution positions of the various collagens on CM- and DEAEcellulose are in agreement with the observations of other workers (Kresina & Miller, 1979; Bentz et al., 197g; Furuto & Miller, 1981) but the behaviour of 7 S collagen on CM-cellulose and HMW aggregates on DEAE-cellulose has not previously been studied. Quantitation of collagen types Table 2 shows the amounts of each collagen (including the two forms of type-V collagen) isolated from various bovine tissues. The highest recovery of the {~I(V)}2~2(V) form of type-V collagen was achieved from the amnion. Uterus and maternal villi also contained t h e {~I(V)~2(V)c~3(V)} form and a g r e a t e r proportion of HMW aggregates, but uterus was the richer source of both these collagens. On the o t h e r hand placental villi provided the better source for type-IV and 7 S c o l l a g e n s . Contrary to some reports (Sage & Bornstein, 1979), we consistently isolated small amounts of type-IV collagen from placental membranes (Ayad el al., 1980).
Table 2.
Quantitation of collagens in bovine tissues
The amount of each collagen was determined by hydroxyproline analysis and is expressed as a percentage of total pepsin-solubilized collagen. The remaining 87-93% consisted of types I and III collagens~ type I:III being 1:2 and constant for each tissue.
Tissue Amion Chorion Foetal villi Maternal villi Pregnant uterus
Type-IV collagen
Type-V collagen {~I(V)}2~2(V) {~I(V)~2(V)~3(V)}
7 S collagen
HMW aggregates
2.0 1.0
8.0 5.0
nd nd
1.0 0.5
