BIOCHEMICAL AND BIOPHYSICALRESEARCHCOMMUNICATIONS Pages 973-980
Vol. 123, No. 3, 1984 September 28, 1984
EFFECT OF HEPARIN MODIFICATION ON ITS CIRCULAR DICHROISM SPECTRUM* German B. Villanueva 1 , and Nancy Allen Department of Biochemistry, New York Medical College Vallhalla, New York 10595 Received July 31, 1984
Summary. The effect of modification of the carboxyl groups of high affinity heparin was investigated. The binding affinity toward antithrombin III decreases in the following order: Heparin > heparin methyl ester > heparinylglycine > heparinylglycine methyl ester. This result agrees qualitatively with the previous studies using unfractionated heparin. Esterification of the carboxyl groups (i.e., HME) does not affect the CD profile of heparin at 210 nm but introduction of a bulkier glycine methyl ester (i.e., HGME) leads to formation of a very intense band at 235 nm. Based on reported CD analyses of uronic acid derivatives and our model building studies, it is concluded that the large difference in CD spectra of HGME as compared to unmodified heparin and HME is due to a change in ring conformation of the uronic acid moiety (i.e. 4C to IC 4 or vice versa). © 1984 Academic Press, Inc.
'
i
Heparin is a highly charged glycosaminoglycan
with various biological
effects,
the most prominent of these being its action on the blood coagulation system. Heparin functions as a blood anticoagulant antithrombin
the latter
the binding of
III with thrombin and a number of other serine proteases of the
coagulation cascade antithrombin
by accelerating
(i).
It was shown that heparin binds to lysine residues
III (2) and this interaction produces a conformational
(3) which accelerates
in
change in
its activity against thrombin.
Heparin has a CD spectrum below 220 nm with a negative trough at 210 nm and a positive peak at the 191-193 nm region. conclusively assigned to, an n ÷ ~
It has been attributed
absorption transition of either an N-acetyl
amide bond or the iduronate carboxyl group or both (4,5). lost upon periodate oxidation
(6).
to, but not
However,
These CD bands are
the ellipticity at 210 nm is almost
doubled upon acidification from pH 7.5 to 2.5 (4,5).
Supported by NIH Grant HL 23265. i
To whom correspondence
regarding this article should be addressed.
0006-291X184 $1.50 973
Copyright © 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 123, No. 3, 1 9 8 4
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
At present, it is not clear whether changes in this parameter are reflections of a conformational transition in the molecule or simply an acidbase property of the carhoxyl groups. use
carboxyl-modified
heparin
The objective of the present study is to
heparin to investigate the conformational
through circular dichroism.
integrity
of
Results suggest that the pH dependent
CD
spectra of heparin is due to both the acid-base property of the carboxyl
groups
and a conformational transition in the molecule. MATERIALS AND METHODS Materials. High affinity heparin was a gift from Dr. G. %Nn Dedem of Diosynth B.V., Holland. Glycine methyl ester was purchased from Sigma and l-ethyl-3-(3dimethylami~opropyl)-carbodlimide hydrochloride from Pierce Chemicals. N-methylN-nitroso-p-toluenesulfonamide was purchased from Aldrich. Uronic acid content was measured by the carbazole method (7). Nitrogen, sulfur and methoxy determinations were performed by Schwarzkopf Microanalytical Laboratory. The carbazole reaction together with nitrogen and sulfur analysis were used as the bases for determining heparin concentration and extent of modification. Heparinylglycine methyl ester (HGME) was prepared by reaction of high affinity heparin with glycine methyl ester as previously reported (8). Percentage analysis found: N, 4.0; S, 10.28; OMe, 3.21 and glycine, 8.25. Heparinyl glycine (HG) was prepared from HGME by saponification cf the latter in 0.1N NaOH under nitrogen atmosphere for 2 days at 4 ° C. Analysis found: N, 3.75; S, 10.53; and glycine, 8.1~ Heparin methyl ester (HME) was prepared according to published procedure (8,9) using ethereal diazomethane generated from alkali-treated N-methyl-N-nitroso-p-toluenesulfonamide. Analysis found: N, 3.33; S, 9.98; and OMe, 3.45. Analysis of untreated high affinity heparin found: N, 2.05; S, 10.34. On the basis of this analysis, the carboxyl groups in heparinylglycine methyl ester, heparinylglycine and heparin methyl ester were esterified to the extent of 94%, 92%, and 75%, respectively. Circular dichroism spectra were recorded in a Cary 60 recording spectropolarimeter equipped with a Cary 6001 circular dichroism attachment. The spectra are reported as mean residue ellipticity, [~], in degrees cm 2 dmole -I and utilizing a disaccharide mean formula weight of 563 (5). RESULTS The uronic acid carboxyl groups in heparin were converted either to the methyl ester or to an amide with glycine and glycine methyl ester as showc schematically below: O II C-OH
II I
O II C-N-CH2-C-O-CH3
OH O , I II C-N-CH 2-C-OH
H I
H I
H I
HA-Heparin
OH
Heparlnylglycine methyl ester (HGME)
Heparinyl glycine (HG)
974
0 II ------~-O-CH3 H i
Heparin methyl ester (}{ME)
Vol.
123, No. 3, 1 9 8 4
Previous studies
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(8,10,11) have monitored the effect of carboxyl group
modification of heparin on antithrombin III binding by using heparin cofactor assay.
Although this measures the immediate neutralization of thrombin by
heparin-antithrombin
III, it is not a direct measure of the binding of heparin
to antithrombin III.
In the present study, the binding of heparin and modified
heparin to antithrombin III is investigated directly by monitoring the intrinsic fluorescence enhancement and ultraviolet difference spectrum which are generated when heparin binds to antithrombin III.
Figure IA shows the intrinsic
fluorescence of antithrombin III as a function of muccpolysaccharide
| 2.0
!
i
i
~
|
i
250
300
i
ii
"
HA-H ~1.5 ~x
/ / / ~
HME
~I.0
0.5
////~
o
•
50 1
100 150 200 Mucopolysaccharide,~m! i i i
F 1.2
1.1
HG
ME-~ ''~'''''~
/
-
.
~
/
1.0 I 10
I I I 20 30 40 Mucopolysaccharide, ~glml
I 50
I 60
Figure I. Fluorescence Enhancement and Ultraviolet Difference Spectral Parameters of Antlthrombin III as a Function of Mucopolysaccharide Concentration. (A) UV difference spectra at 290 nm. (B) Intrinsic fluorescence enhancement at 340 nm with excitatiin at 280 ran. (-41-m--), high affinity heparln; (-O-O-), heparin methyl ester; (-Q-t-), heparinylglycine; and (-O-O-), heparinylglycine methyl ester.
975
Vol. 123, No. 3, 1984
concentration.
BIOCHEMICAL A N D BIOPHYSICAL RESEARCH C O M M U N I C A T I O N S
Ken
the binding is compared based on the amount of
mucopolysaccharide required to reach half saturation, decreases in ~ e
order: Heparin > ~ E
> HG > H ~ E .
the binding affinity
This order of reactivity
agrees with the results obtained by heparin cofactor assay (Ii), namely that H~E
is inactive and M E
is relatively more active than HG.
the fluorescence enhancement due to M E
The magnitude of
and HG, however, suggests that they
should be more potent than the 13% and 2.6% observed respectively by heparin cofactor assay.
The results obtained from the ultraviolet difference spectra
also support the trend observed from the fluorescence data and heparin cofactor assay (Figure IB).
H~E
has minimal difference spectra and H E
approaches the
control ~ - h e p a r i n more closely than HG. ~e were far
circular
dichroism spectra of heparin and carboxyl
modified
heparin
investigated in order to assess the contribution of carboxyl groups to the ultraviolet
heparin a ~
HE
electronic
transition.
The circular
dichroism
at pH 7.4 and 2.5 are compared in Figure 2.
the CD minima of heparin and M E HA-he~arin and ~ E
spectra
of
It can be seen that
at 210 ~. have comparable magnitude at pH
7.5.
have ellipticities of 1600 +_ 150 degrees cm2 dmole-i and 1750
0.5
~
1 . 0
1.5
H A - Heparin p H i 5
x
HME,pH2.
2.5
~..j.qF~----- HA-He~rin, pH 2,5 I
200
I
2~0
I
220 in nm
I
230
240
Figure 2. Circular Dischroism Spectra of High Affinity Heparin and Heparin Methyl Ester.At Neutral and Acid pH. m e solution at neutral pH is in water while the solution at acid pH is in 0.5M NaCI.
976
Vol. 123, No. 3, 1984
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
/
\.
2
o~0 "o
%
HG, pH 7.5
~2 HG, pH 2.5 r~
4
x
I HGME, pH Z5
101-
\
/ ~ F - - H G M E
210
220
230 Xin
, pH2.5
240
250
260
nm
Figure 3. Circular Dichroism Spectra of Heparinylglycine and Heparinylglycine Methyl Ester. All other conditions are the same as in Figure 2.
150 degrees
cm 2 d m o l e -I ,
ellipticities enhancement HME
in H A - h e p a r i n
cannot
It is evident
derivatives
minimum
centered
(+250
are enhanced 150 degrees
carboxyl
groups,
are
at 215 nm
~ 45 degrees at pH 2.5
this
cm 2 dmole -I to +500
~I00
from heparin.
degrees
) has appeared
~ 45 degrees
cm 2 dmole -I at 215 nm.
more
At p H
cm 2 dmole -I
Since CD
interesting of these
two
7.5, HG has a
) and a new positive
at 235 nm.
These
two bands
cm 2 dmole -I at 235 nm and
At pH 7.5, HGME has CD bands w h i c h
977
The
groups.
3 that the CD ellipticities
different
in HME.
the
the pH dependent
of the carboxyl
of HG and HGME provides
in Figure
(-i000
that
enhances
the same extent.
to 1.4 times
indicates
to the p r o t o n a t i o n
totally
Acidification
to a p p r o x i m a t e l y
times as compared
of the CD spectra
heparin
band
is 1.6
be due solely
Examination observation.
at 210 nm.
of both m u c o p e l y s a c c h a r i d e s
has no ionizable
spectra
respectively,
-3800 are more
Vol. 123, No. 3, 1 9 8 4
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
than an order of magnitude
larger than those of HG at these wavelengths,
ellipticity being +6000 degrees cm2 d~cle-1
the
~ 500 degrees cm 2 dmole -I at 235 nm and Ii,000 ~ 700
at 215 nm.
Within the range of experimental
bands in HGME are only slightly sensitive
to pH.
error,
these
Similar studies done on HGME
prepared from commercial heparin gave spectra which were only five times greater than those of HG at 235 r~n (Data not shown). DISCUSSION The results of the present studies confirm previous virtually no antithrombin groups,
III activity and that ~ E ,
with no free carboxyl
although less active than unmodified heparin,
than HG which has free carboxyl groups.
findings that H~ME has
is relatively more active
Although it has Dot been so
demonstrated
that carboxyl groups are essential for the binding of heparin and
antithrombin
III, these results suggest that the relative charge orientation of
heparin must play an important role, because esterification displacement of this charge by one methylamino group
(i.e. HME) and
(i.e., HG) causes dramatic
loss of heparin cofactor activity. Comparison cf the CD spectra of unmodified heparin and HME provides interesting observations. spectra
(Figure 2).
some
HME and unmodified heparin have very similar CD
The pH dependence of the CD spectrum of heparin has been
known for a long time (4,5).
From titration studies
(5) it was suggested that
the variation of optical parameters of heparin with pH was due to the acid-base property of the carboxyl group rather than a conformational contrast with these previous groups still exhibits
change.
However,
in
studies, HME, which has no ionizable carboxyl
this pH dependence.
The acetamido n ÷ fi transition in
heparin is not expected to be pH dependent.
Thus,
it is concluded from this
study that the pH dependence of the CD spectra of heparin must be predominantly due to conformational
changes,
of the carboxyl groups. studies
and to a lesser extent to the acid-base property
This is in agreemeet with results from hydrodynamic
(12) which indicate that the heparin molecule behaves as a random coil
at pH 2.5 and as a fully extended rod at neutral pH. that these two phenomena are inseparable,
978
It is possible,
however,
i.e., that the conformation of heparin
Vol. 123, No. 3, 1 9 8 4
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
may be a function of the integrity of the carboxyl groups.
It is only in this
context that the results of the titration studies and chemical modifications can be reconciled.
The second aspect of the present study deals with the CD spectra of HG and HGME.
While esterification of heparin (i.e., }{ME) does not have a significant
effect on the CD of heparin, attachment of a glycine moeity or a glycine methyl ester (i.e. HG, HGME) produces a new CD spectrum at 235 nm.
This new band
cannot be due entirely to the glycine amide bond transition because while this transition is obviously present in both HG and HGME, the CD maximum at 235 nm is more than ten times stronger in HGME than in HG.
This long wavelength band has
been previously observed in a number of disymmetric a-hydroxy carboxylic acids (13).
It was observed to occur in copper complexes of heparin to a lesser
extent and opposite in sign (14).
On the monomer level, it has been observed in
glucuronoside but not in iduronoside dependence,
(15) and because of its marked solvent
it was assigned to the n ÷ D
transition of unsolvated molecules and
the band at 210 nm to the same transition as molecules hydrogen bonded to water. In the present case however,
solvent interaction can be ruled out because this
band is not affected by the presence of 6M GdmCl in HGME.
Listowsky et. al. (15)
and Morris et, al. (13) independently observed that the common feature exhibited by uronic acid derivatives with
peculiar long wavelength band at 234 nm is the
presence of a 0-(4) equatorial orientation.
Thus, only the "normal" 210 nm n ÷ ~
transition is observed for uronic acids with the 0-(4) axial orientation while the 210 nm and 234 nm bands are seen with uronic acids containing the 0-(4) equatorial orientation.
Model building studies done in our laboratory indicate
that the methyl ester group
(i.e., }{ME) can easily be accommodated in the sugar
ring structure, but the glycine and glycine methyl ester units
(i.e., HG, HGME)
cannot be oriented in space without interfering with the highly charged sulfate groups in the molecule.
We therefore believe that the appearance of this long
wavelength band in HG and HGME is due to ring inversion (i.e., IC 4 to 4C I or vice versa) of the uronic acid moiety.
It is only through this kind of rotation
of the chromophore against a disymmetric environment of the rest of the molecule
979
Vol. 123, No. 3, 1 9 8 4
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that large changes in the CD spectra can be effected. course, tentative at this time.
This conclusion is, of
Additional physical monitoring methods,
involving nuclear magnetic resonance and X-ray studies, will be necessary to identify the specific conformers of uronic acids which predominate in these modified heparins. REFERENCES i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15.
Rosenberg, R.D. (1977) Sem. in Hematol. 14, 427-440. Rosenberg, R.D. and Damus, P.S. (1973) J. Biol. Chem. 248, 6490-6505. Villanueva, G.B. and Danishefsky. I. (1977) Biochem. Biophys. Res. Commun. 74, 803-809. Stone, A.L° (1977) Fed. Proc., 36, 101-106. Park, J.W. and Chakrabarti, B. ~-[977) Biochem. Biophys. Res. Commun. 78, 604-608. Ching Ming Chang, M. and Ellerton, N.F. (1976) Biopolymers 15, 1409-1423. Bitter, T., Muir, H.M. (1962) Anal. Biochem. ~, 330-334. Danishefsky, I. and Siskovic, E. (1972) Thromb. Res. i, 173-182. Fales, H.M., Jaouni, T.M. and Babashak, J.F. (1973) Anal. Chem. 45, 2302-2303. Danishefsky, I. and Siskovic, E. (1971) Carbohyd. Res. 16, 199-205. Danishefsky, I., Ahrens, M. and Klein, S. (1977) Biochem. Biophys. Acta. 498, 215-222. Lasker, S.E. and Stivala, L. (1966) Arch. Biochem. Biophys. 115, 360-372. Morris, E.R., Ress, D.A., Sanderson, G.R. and Thom, D. (1975) J. Chem. Soc. [Perkin II] 1418-1425. Mukherjee, D.C., Park, J.W. and Chakrabarti, B. (1978) Arch. Biochem. Biophys. 191, 393-399. Listowsky, I., Englard, S. and Avigard, G. (1969) Biochemistry 8, 1781-1785.
980