The amino acid composition of elastin in its soluble and insoluble state

Vol. 33, No. 2, 1968

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

THEAMINO ACID COMPOSITION OF ELASTIN IN ITS SOLUBLEAND INSOLUBLESTATE* Division

by J.A. Petruska and L.B. Sandbergf* of Biology, California Institute of Technology

Received September 3, 1968 Introduction The protein

elastin

is highly

insoluble in normal animals, being depos-

ited in a cross-linked

aggregated state (8).

has yet to be isolated

from normal elastic

Whenelastic

tissues,

ligament are autoclaved,

such as aorta, the elastin

Subsequent extraction

(5). elastin

Its soluble noncross-linked tissues.

nasal and ear cartilage,

is left

leads to somepurification

that pure elastin

amounts of other proteins may remain firmly to the elastin;

in particular,

tic

is ever obtained,

ground substance protein and collagen,

fraction

obtained by neutral-salt

but very little,

In animals raised on copper-deficient soluble protein

fraction

noncross-linked

form (13).

linked, of which

tissues (2,4), extraction

tissues from normal animals is mainly ground substance protein

collagen is also in this fraction,

of the Small

bound, perhaps covalently

there are considerable amounts in the elastic The soluble protein

and nuchal

behind as an insoluble residue

with hot alkali

(S), but it is doubtful

form

diets,

may contain substantial

of elas(4).

Some

if any, elastin, however, it appears that the amounts of elastin

in the

The amino acid compositions reported for soluble and insoluble elastictissue fractions

in normal and copper-deficient

animals are examined here.

making somereasonable assumptions about the protein to extrapolate

to a composition for pure elastin

By

components, we are able

in its

soluble and insoluble

states. *Supported by Public Health Service Grant GM06965. **Established Investigator of the American Heart Association, Present Address: Department of Surgery, University of Utah, Salt Lake City, Utah.

222

Vol. 33, No. 2, 1968

Sources

of

8lOCHEMlCAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Data

The amino acid compositions examined pertain tic

tissue fractions.

and nasal cartilage

The bovine tissues are nuchal ligament (2 , 4 , 5#9#ll), (5,9),

and aorta (5) of normal cattle;

are aortas of normal and copper-deficient are obtained by neutral-salt autoclaving

(2,s).

to bovine and porcine elas-

extraction

swine (13). (4,9,13);

Although data are available

8

7

65432

CONTENT

the porcine tissues

The soluble fractions

the insoluble

fractions

by

for insoluble fractions

8

OF GLY (RES

ear

7

654321

E

PER 1000)

Fig. 1. Amino acid plots against gly, before correction for collagen (dotted lines and open dots), and after correction for collagen (solid lines and closed dots), for normal bovine elastic tissue fractions. 1,2,3, and 4 are insoluble fractions from nuchal ligament of g-year old cow (ll), l-year old cow (11), adult cattle of unspecified age (S), and term-foetal calf (ll), respectively, 5 and 6 are insoluble fractions from aorta and ear cartilage, respectively, of adult cattle of unspecified age (5). 7 is a soluble fraction from autoclaved bovine nuchal ligament (4); 8, the ground substance protein isolated from the soluble fraction of bovine nasal cartilage (9). The straight lines are fitted to the points by the least-squares method.

223

BIOCHEMICAL

Vol. 33, No. 2, 1968

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

cleansed with hot alkali

(1,5),

these data are not used because the alkali

treatment is destructive

to someof the amino acids, particularly

arginine,

Assumptions Two assumptions are made about the protein sues.

First,

components in the elastic

tis-

we assumethat only three components are present in significant

amounts; namely, elastin,

ground substance protein,

and collagen,

suppose that collagen is the only component containing The amino acid contributions of hypro, are calculated

from collagen,

Second, we

hydroxyproline

(hypro),

corresponding to the content

using the composition of purified

32

collagen from calf

I E

50 0 PRO

50 0

a

0

'

'

'

'

'

'1)

CONTENT

1

OF GLY (RES.

PER 1000)

Fig. 2. Amino acid plots against gly, before correction for collagen (dotted lines and open dots), and after correction for collagen (solid lines and closed 1 is the insoluble fraction; 2 and dots), for normal porcine aortic fractions. of the soluble fraction left 3 are the residue and supernatant, respectively, The straight lines were standing in the cold (13), as a step in purification. fitted by least squares.

224

Vol. 33, No. 2, 1968

(6)

BIOCfiEMICAL

and pig (3) skins,

AND BfOfWYSICAL RESEARCH COMMUNICATIONS

In correcting

for collagen, we subtract

these contri-

butions from the observed amino acid contents in the fractions, before and after total

correction

The contents

for collagen are expressed in residues per 1000

amino acid residues and plotted

as described below,

The content of one of the amino acids showing a large net change, specifically

gly,

ala, or val,

is plotted

the ordinate,

is then plotted

amino acids.

When the data for all

Against this,

along

the corresponding content for each of the other the fractions

are so plotted,

the points

/'

200

. / /'I

150

100 l ’ 50;i'^ ’

along the abscissa.

/

,

HO

i

HIS

/ I

I

0 50 0 50 0 150 IOQ 50 0 200

CONTENT

OF GLY (RES.

PER 1000)

Fig. 3. Amino acid plots against gly, before correction for collagen (dotted lines and open dots), and after correction for collagen (solid lines and closed dots), for Cu-deficient porcine aortic fractions. 1 and 2 are residues from soluble fractions coacervated at room temperature (13); 3 and 4 are supernatant and residue, respectively, from soluble fraction left standing in the cold (13) 5 and 6 are residue and supernatant, respectively, from normal aortic soluble fraction left standing in the cold (same as 2 and 3, respectively, in Fig, 2). The straight lines are the least-squares fit.

225

Vol. 33, No. 2, 1968

BIOCHEMICAL

AND BIOPHYSIC~L RESEARCH COMMUNICATIONS

for each amino acid should lie on a straight substance protein

and ground

This is the line connecting pure elastin

are present,

pure ground substance protein,

and

If collagen is present in the amounts calcu-

lated from hypro, there should be a better rection

line if only elastin

fit

to a straight

line after

cor-

for collagen,

Results The results

are very similar

Here we described the results rection

for collagen,

whether one plots against gly,

of plotting

line indicates

against gly before and after

To each set of points plotted,

by least squares (Figs. 1,2,3).

of the standard deviation

a straight

The standard deviation

the goodness of fit

ala, or val.

correction

line is fitted

of the points from the

in each case (Table I),

is smaller after

cor-

The average value

for collagen than before

(Table I). Table I Average Value of the Standard Deviation Between the Points and the Lines For the Sets of Plots in Figs, 1 to 3 Average Standard Deviation in Residues per 1000 Residues After Correction Before Correction For Collagen Figure For Collagen 1 2 3

After

4.4 2.3 4.7

correction

axis at the samepoint, lrE1l

for collagen,

Moving vertically

several of the lines intersect

“E” (Figs, 1,2,3).

corresponds to pure elastin

sent.

3.9 2.0 3.1

The simplest interpretation

is that

in which these amino acids are completely ab-

along the dashed line at “E,” we can read off

corresponding contents of other amino acids from the lines. arrive

the gly

at the compositions for pure elastin

the

In this manner we

as shown in Table II.

Discussion The computed compositions for soluble elastin and insoluble

elastin

in normal swine are identical

226

in copper-deficient within

swine

the errors of amino

Vol. 33, No. 2, 1968

acid analysis,

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

except for the content of lysine.

bers noted between these two elastins

(Table II)

sine lost during the formation of cross-links, in the tables.

For the soluble elastin,

The small difference can be attributed

in num-

to the ly-

These have not been included

there are 46 residues of lysine per

Table II Amino Acid Contents Assigned to Pure Elastin Amino Acid

Amino Acid Residues per 1000 Residues Normal Bovine Normal Porcine Cu-Deficrent Porcine Insoluble Elastin Insoluble Elastin Soluble Elastin 334 236 127 154 27 67 34 0 0 0 5.7 9.0 0 0 0 0 6 5.9

GUY Ala Pro Val Ilu Leu Phe Met CYS Ser

Thr Tyr TV His Arg Ls + Gi::

328 250 109 147 19 57 36 1.1 0 7.2 13 20 0 0 1.3 2.1 0 10

320 236 106 138 21 53 32 0 0 5.3 13 18 0 0 0.4 % 12

Each value is the average of the values determined from the plots xnst gly, ala, and val after correction for collagen, The determination is made at the point where the line for Asx intercepts the ordinate (the point labelled “Eft in Figs. l-3). The content of lysine is underlined to emphasize its much larger value in copper-deficient porcine elastin. Note :

1000 residues of amino acid.

With the insoluble material,

tent has dropped to two residues per 1000 residues.

the lysine con-

This difference

Iysine content is in part due to the production of the cross-linking stances desmosine, isodesmosine, and lysinonorleucine.

in sub-

However, these

known derivatives

of lysine,

for approximately

one-half of the lysine present in the insoluble material,

Recent reports

which occur in insoluble elastin,

(7,10) have indicated

account only

the presence of aldehydes in elastin

derived from the enzymatic conversion of lysine to alpha-amino adipic-

227

Vol. 33, No. 2, 1968

semialdehyde.

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

This aldehyde content of elastin

seventh of the total would therefore

accounts for about one-

lysine in the noncross-linked

soluble material,

seem that the remaining lysine residues are either

during the conversion of soluble elastin

to the insoluble fiber

cess of enzymatic peptide cleavage, or more likely, cross-linking

It lost

by somepro-

they are converted to

substances which have not as yet been identified.

has already been presented in favor of the latter

Someevidence

argument (12).

The good agreement between the two computed compositions for soluble and insoluble porcine elastin like protein ally,

(Table II),

reported by Smith, -et al,

the elastin

precursor

substantiates

that the soluble elastin-

(13) is closely related

(tropoelastin).

Further,

it can be said that the

coacervated product reported by these workers is a fairly impurities

to, if not actu-

pure protein;

the

being collagen and ground substance proteins present in not greater

than 15%concentration, Minimum molecular size of both bovine and porcine soluble elastin indicated

by amino acid composition is predicted

approximately

30,000 molecular polypeptide

as

to be about 350 residues or

weight.

References

1. 2.

43: 5. 6. 7. 8. 9.

10. 11. 12. 13.

Cleary, E.G., L.B. Sandberg, and D.S. Jackson, Biochemistry and Physiology of Connective Tissue, P. Comte, editor, Societe Ormeco et Imprimerie du Sud-Est Lyons, France, 1966, p, 167. Cleary, E.G., L.B. Sandberg, and D.S. Jackson, J, Cell Biol., z (1967) 469. Eastoe, J.E., Biochem. J., 61 (1955) 589. and V. Moret, J, Atheroscler. Res., 2 Gotte, L., Serafini-Fracas&ii, (1963) 244, Gotte, L., P. Stern, D.F. Elsden, and S.M. Partridge, Biochem, J., g (1963) 344. Heidrich, H.G. and L.K. Wynston, 2. Physiol. Chem., 342 (1965) 342, Miller, E.J. and H.M. Fullmer, J. Exp, Med., 123 (19m 1097. Partridge, Fed. Proc., 25 (1966) 1023 Partridge, S.M., H.F. DGis, and G.S. Adair, Biochem. J., 79 (1961) 15. Science, 161 v968) 475. Pinnell, S.R., G.R. Martin, and E.J. Miller, Sandberg, L.B., Ph.D. Thesis, Grad, Div. of the Univ. of Ore. Med. School, 1966. Sandberg, L.B. and E.G. Cleary, Biochim. Biophys. Acta, 154 (1968) 411. N. Weissman, and W.H. Carnes, Biochem, BiopF. Res. COPIOUS., Smith, D.W., 31 (1968) 309. 228