Analysis of the reducible components of the muscle protein, connectin: Absence of lysine-derived cross-links

Vol.

96, No.

October

BIOCHEMICAL

3, 1980

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Pages

16, 1980

1240-1247

ANALYSIS OF THE REOXJCIBLB COMPONENTS 03 THEMUSCLEPROTEIN, CONNXCTIN: ABSENCEOF LYSIN&DERIVEDCROSS-LINKS Simon P. Robins and Garry J. Rucklidge Rowett Research Institute, Bucksburn, Aberdeen, AB2 9SB, Scotland, U.K. Received

August

11,198O

After exhaustive salt extractions of rabbit and humanskeletal muscle, the amino acid compositions of the residual proteins were similar to those reported for connectin. Complete removal of collagen contamination was achieved only after treatment of the connectin preparations with bacterial On reduction with K83H4, the small amounts of lysine-derived collagenase. reducible cross-links that were present in the initial connectin preparations were completely absent after treatment with collagenase. Inadulthuman connectin somehexitol-lysine derivatives were present after reduction. These results indicate that, in contrast to previous reports, connectin does not psrticipate in the samelysyl oxidase-mediated cross-linking system that occurs in collagen aa elastin.

INTRODUCTION Connectin

is

an intractable

and cardiac muscle (l-3). filament fibre

protein that has been isolated from skeletal

This protein was proposed as a third

structural

in muscle that is responsible for the mechanical. continuity

by forming a fine network of filaments interconnecting

of the

the Z lines.

These proposals were based on visualization

in the electron microscope of

filamentous networks in ghost muscle fibres

(2,4) that appeared to have

similar properties

and amino acid composition to those of connectin (2).

Analysis of the products of reduction with NaB3H4of urea-SDS extracted connectin from chicken skeletal muscle led to the conclusion that the protein contained lysine-derived

cross-links

oxidase dependent cross-linking

(5)

without the use of denatursnts indicated 0006-291X/80/191240-08$01.00/0 d 1980

by Academic Press, in any form

of reproduction

Inc. reserved.

the samelysyl

system that occurs in collagen and elastin

Subsequently, a study of humanskeletal

(67).

Copyright All rights

and, hence, utilized

1240

muscle connectin prepsred

that the reducible

components

Vol.

96, No.

3. 1980

8lOCHEMlCAL

undergo an age-related determined

for collagen

In view muscle,

decrease

ageing and certain

BIOPHYSICAL

in amount (8),

RESEARCH

similar

COMMUNICATIONS

to that

previously

(v-11).

of the potential

particularly

AND

importance

in relation nutritional

of this

to rqofibriller deficiencies,

reducible

components

of connectin

of which

are presented

in this

from several

type of cross-linking assembly

and the effects

we have investigated different

in

species,

of

the the results

report.

MATERIALS andMEl'HODS Connectin nrensration. Mixed skeletal muscles from 2-year-old rabbits, breast muscle from l&week-old chickens and cardiac muscle from a 2-year-old steer were freshly obtained at slaughter and human psoas muscle (from JV-year-old, male) was obtained within 24 h post mortem. The extraction procedures were based on those of Fujii and Kurosu (8) and were all carried out at 4OC. Samples (100 g wet w-t) were minced and homogenized in a Waring blender with 5 vols (v/w) of 25mM-KCl-@v&borate buffer containing 15mM-mercaptoethanol. After centrifuging at 2000 Q for 30 min, the residue was resuspended in IO volumes of O.lM-KCl-@r&borate (pH 7.0)-15mM mercaptoethanol and stirred for 16 h. Fibrous material was removed by passing through cheesecloth and the myofibrillar suspension was centrifuged at 5000 g for 20 min. The residue was extracted by stirring with 0.5M-KC1-1OmM-Na~P207-5OmM-NeH2P0~(pH 6) containing 15mM-mercaptoethanol for 3 h and was centrifuged at 5000 g for 20 min. The residue was subjected to three further extractions with the 0.5M-KC1 buffer and sny fibrous material that aggregated during this procedure was removed by filtration. Finally, the residue was stirred with 0.6M-KI-6OmM-Na2S203for 12 h snd after centrifugation this extraction was repeated three times. For csrdiac muscle a further extraction with l.lM-KI-6Om%Na2S205 was carried out. The insoluble material was washed exhaustively with water and freeze-dried. The collagenous material removed at various stages of the extraction was combined and washed successively with 0.5M-KC1 buffer, pH 6, water, 0.6M-KI-6Om%Na2S203and finally O.l5M-NaCl-IOmM-phosphate, pH 7.5. Treatment of connectin with collsgenase (Sigma, Type V) at sn enzyme: substrate ratio of I:20 was csrried out in @mM-lYis/HC!l (pH 7.6)-lxM-CaCl2 -lOmM-N-ethylmaleimide maintained for 4 h at 37OC in sn orbital shaker. The insoluble material was washed successively with incubation buffer and water and was then freeze-dried. Borohydride reduction. Samples of connectin and intramuscular collagen were eauilibrated with O.l5M-NaCl-IOmM-uhosnhate (DH 7.5) and were reduced $$ Kf3H4 at a I:30 boroh&ide:protein radio (w&) as-described previously . The reduced proteins were hydrolysed in 6M-HCl for 24 h at 108OC. Amino acid compositions were determined with a Analytical procedures. Locsrte snalyser after hydrolysis of the samples in 6M-HCl in sealed, evacuated tubes at 108OC for 24 h. For Nf-methyl-histidine analyses, interference by the relatively large amounts of histidine was obviated by preliminary treatment or a portion of the hydrolysate with histidine decarboxylase for 1 h in acetate buffer, pH 4.5.

1241

Vol.

96, No.

3. 1980

Table

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Amino acid compoeitionea of connectin preparations

1

Rabbit

, Untreated

Thr Ser Glu

Collagenasetreated

:i

7; 22

5:

fgc u-8 His WC

77

;:

il 82

5:

5;

22

25

84 36

28 2.2

37 -0.7

:2 :;2.8

65

77

77

i;

Pi

69

E

96

70

82

m/2 Vdl Met Ile

c011e&Xlaeetreated

124

137

GUY Ala

Untreated

59

116 69

R-0

,

2: 78

;: 0.0 79

ressed as residues per 1000 residues without correction aEScp hydrolytic recoveries: tryptophen wasnot determined.

for

b @droxyproline was determined separately by a sensitive calorimetric procedure (12). C Analyses of bydroxylysine were carried out on en extended basic columnof the enalyser using high loadings corresponding to about 2 mgof protein. Analyses of the labelled compoundsin the I(B3Hd-reduced proteins were performed by fractionation of the hydzolysates on an extended basic column of a Locarrte snalyser, previously calibrated with authentic collagen cross-linking amino acids (13). The identities of the isolated components were confirmed by molecular weight estimation on a 2 x 140 cm column of Biogel-P2 and by snalysis of the products of degradation by periodate (14).

The amino acid compositions of the material remaining after extractions

of myofibrillar

proteins from rabbit

psoas muscle sze shown in Table

1.

skeletal

exhaustive

muscle and human

Attempts to completely remove the

collagen contamination by purely physical mesns (8) were unsuccessful but treatment of the extracted

protein with bacterial

removal of the hydroxyproline collsgenous material, lower after

(Table 1).

the relative

1242

total

Consistent with the removal of

proportions

treatment with collagenase.

collagenase effected

of glycine and proline were

For humanpsoas muscle,

Vol.

96, No.

3, 1980

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

IL LP ‘2_ iP

3-

i

2-

l-

/ 12 -

e

rz --

10 b 1

&+JLL

I.5

1L

t

12

8-

!

10 I 1

L-

0.5 20 Effluent

Fig.

1

collagenase rabbit the

muscle complete

60

LO

1

80

I

,b. 20

LO

volumeOnlt

Effluent

60

80

volume(mi)

Chromatograw of acid hydrolyaatee of the KB'I!Q-reduced components of collagen aud connectin on a 0.9 x 30 cm column of a Lccarte analyzer using 0.35I%sodium citrate buffer, pE 5.25 The radioactive profile6 are of (a) intra(CO~GTI temp. 56"~). muscular collqen, (b) conuectin prepared by salt extractiona end (c) collwe-treated connectin from rabbit skeletal muscle, and (d) intramuscular collagen, (e) connectin prepared by salt extractions end (f) collwe-treated connectin from humen peoaa muscle. Each set of samples, a-c aud d-f, were reduced concurrently with the same l&q solution for direct comparison. The chromato~aphic positions indicated are A, dihydroqlysinonorleucine; B, mannitol-lyeine; C, glucitol-lyeine which E and G, co-chromatogrsphs with D, hydroxylyainonorleuctie; anhydrc derivatives of hexitol-lysine; F, hietidinohydrogmerodeemosine and Ii, lysinonorleucine.

treatment connectin, absence

resulted a small

in the

total

amount

of hydroxy-proline

removal

of hydroxylysine

of hydroxylysine (Table

1243

1).

remained

Analyses

of

but despite

for

Vol.

96, No.

BIOCHEMICAL

3, 1980

ti-methyl-histidine

revealed

AND

BIOPHYSICAL

the presence

RESEARCH

COMMUNICATIONS

of about 0.05 residues

per

1000

When residues in hydrolysates of collagenase-treated connectin. collagenase-treated connectin was subjected to SDS-gel electrophoresis (5% acrylsmide),

sample buffer

the small amount of material

exhibited

that dissolved in the 2% SDS

a very high molecular wei&t.

Analyses of the components labelled by reduction with KB3H4of the connectin preparations

and the intramuscular

samemuscles sxe illustrated three major labelled

in Figure

In rabbit

being an artefact

connectin,

the

collagen were identified

hydro~lysinonorleucine

hydroxymerodesmosine, the latter

from the

In the young growing rabbit,

1.

compoundsin the intrsmusculm

as dihydroxylysinonorleucine,

procedure (6).

collagens isolated

and histidinoof the reduction

the only reducible

components in the

basic region of the chromatogrsm were shown to be derived from collsgen contamination, as the collagenase-treated sample was completely devoid of labelled

components in this region (Fig.

1).

Further analysisd

labelled material that eluted with the acid and neutral 15 ml) revealed several unidentified

precursors,

dihydroqynorleucine

and hydroxynorleucine,

increase with age (9, 15).

and their

amino acids (5

-

componentsbut the reduced cross-link

In adult human intramuscular collagen, maskedby hexitol-lysines

the

the cross-links

were entirely

absent.

were partially

snhydro derivatives

that are known to

These components were also detected in the

connectin prepsxation and very small amounts appeared to remain after collagenase treatment. lysine-derived

cross-links

In the latter

sample, however, none of the

were present.

Analyses (not shown) of connectin isolated

from chicken breast muscle

and bovine cardiac muscle gave similxc results to those obtained for the rabbit

in that none of the known lysine-derived

reducible

cross-links

were

detected. DISCUSSION The complete absence of lysine-derived,

reducible

cross-links

conuectin from the four species snalysed in this study refutes

1244

in

the suggestion

Vol.

96, No.

(5,8)

3, 1980

that

linking

BIOCHEMICAL

thi s muscle protein

AND

BIOPHYSICAL

shares the samelysyl

system as collagen and elastin.

treatment lysinonorleucine

RESEARCH

oxidase mediated cross-

In connectin extracted

produced during acid hydrolysis

these anhydro derivatives The hexitol-lysine

(15).

shown to be degraded by periodate whereas lysinonorleucine Reducible lysine-hexose

of humsnconnectin extracted labile

al&nines,

but their similar

intermolecular

reducible

involvement in cross-linking extraction

were

is stable under

adducts have been shown to be

cross-links.

without denaturing conditions

three additional

are

derivatives

membranesand a number of other proteins

these compoundsdo not constitute

(5).

chromatographs close to one of the

anhydro forms of the hexitol-lysines:

present in erythrocyte

by urea-SDS

was proposed as the major reduced cross-link

As shown in Fig. 1, lysinonorleucine

these conditions.

COMMUNICATIONS

but

(16)

In a hay to presence

compoundswere detected (8),

was not established.

Using

procedures for humanconnectim, our analyses showed

that only small amounts of hexitol-lysine basic region of the chromatogrsm.

derivatives

were present in the

In these experiments it was found

essential to remove collagen contamination by treatment with collsgenase: immunological studies showed that the untreated connectin preparations contained types I,

III

and V (AB2) collagens (G.J.R.,

The amino acid analyses (Table 1) acre similar for connectin (1,5). characteristic

unpublished results).

to the values reported

These compositions are not, however, distinctive

or

and the question remains as to whether the isolated material

represents a discrete molecular species or is a cross-linked number of proteins.

aggregate of a

If myosin and actin are the only methylated proteins

in muscle, then the amounts of fl-methyl-histidine

detected corresponds to

about a 10% contamination by myosin or less than 5% contamination by actin. The virtual indicative

absence of these proteins in the gel electrophoretogrems is of someform of covalent bonding in the protein

More recently, salt extraction

aggregate.

a number of other muscle proteins that sxe resistant have been described.

These include the intermediate

1245

to high

Vol.

96, No.

filament

BIOCHEMICAL

3, 1980

proteins,

interconnection

AND

desmin and vimentin, between Z discs

protein distributed

(17),

BIOPHYSICAL

that

appear to provide

and titin,

mainly within the Z lines

the procedures described here msy be identical

RESEARCH

COMMUNICATIONS

lateral

a high molecular weight (18).

Connectin isolated by

to or be contaminated by a

number of these other proteins. Thus, although further

characterization

necesssry, the present results components does not involve

of the isolated protein is

show that cross-linking

of these structural

the copper-dependent enzyme, lysyl

oxidase.

This enzyme is und0ubtealy present in muscle initiating cross-linking of extracellular collagen but its activity within an intracellular compartment containing high concentrations

of csrnosine and anserine is unlikely,

as

these peptides are tiown to chelate copper (19).

We sre indebted to Dr J. L. McPhie of the Depsrtment of Pathology, University of Aberdeen for help in obtaining muscle samples.

6.

Msruyama, K., Nonomura, Y. and Natori, R. (1976) Nature &, 58-59. Maruysma, K., Matsubara, S., Natori, R., Nonomura, Y., Kimura, S., Ohashi, K., Muraksmi, F., Handa, S. and E&u&i, G. (1977) J. Biochem. 82, 317-337. Msmyama, K., Kimura, S., Kuroda, M. and Handa, S. (1977) J. Biochem. 82, 347-350. Remedies, C. G. and Gilmour, D. (1978) J. Biochem. 3, 239-241. Fujii, K., Kimura, S. an& Mszuysma, K. (1978) Biochem. Biophys. Res. Commun.8J 1248-1253. Bailey, A. J., Robins, S. P. and Balisn, G. (1974) Nature a,

7. 8.

Siegel,R. C. (1979) Int. Rev. Connect. Tissue Res. f3, 73-118. Fujii, K. and Ku~osu, H. (1979) Biochem. Biophys. Res. Commun.2,

1. 2. 3. ;:

105-109. 1026-1032. 9. 10. 11. 12. 13. 14. 15.

Bailey, A. J. and Shimokomaki, M. S. (1971) FESS Lett. 16, 86-88. Robins, S. P., Shimokomski, M. S. and Bailey, A. J. (19737 Biochem. J. 111, 771-780. Fujii, K. and Tanser, M. L. (1974) FEBS Lett. 4, 300-302. Firschein, H. E. and Shill, J. P. (1966) Anal. Biochem. & 286-304. Robins, S. P. (1976) Methodology of Connective Tissue Reseezch, PP 37-52. Robins, S. P. and Bailey, A. J. Biochem. J. 14p, 381-385. Robins, S. P. and Bailey, A. J. Biochem. Biophys. Res. Commun.$3, 76-84.

1246

Vol.

96, No.

16. 17. 18. 19.

3, 1980

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Bailey, A. J.,Robins, S. P. and Tanner, M. J. A. (1976) Biochim. Biophys. Acta 424, 51-57. Nature 28 , 249-256. Lazaraides, E. (1980) Proc. Natl. Acad. Sci. Wang, K., McClure, J. and Tu, A.-t 1979) USAs, 3698~3702. Brown, C. E. and Antholine, W. E. (1980) Biochem. Biophys. Res. commun.2, 470-477.

1247