studies on the mechanism of protein adsorption on

0 downloads 0 Views 3MB Size Report
In an incugation mixtu~~ ot 60 ml(2,4) the ~~action waA Ata~t~d gy addition ot 1.5 ml pack~d gutyl S~pha~o~~ (ca. 20 ~mol/ml pack~d g~l) co~~~~ponding to a ...
T.C.J. Gribnau, J. Visser and R.J.F. Nivard (Editors), Affinity Chrorrw.tography and Related Techniques © 1982 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

39

STUDIES ON THE MECHANISM OF PROTEIN ADSORPTION ON HYDROPHOBIC AGAROSES H.P. JENNISSEN*, A. DEMIROGLOU* and E. LOGEMANN§

* Institut

fur Physiologische Chemie der Ruhr-Universitat Bochum and

§Institut fur Rechtsmedizin der Universitat Freiburg, West Germany

ABSTRACT Although phosphorylase

~

is strongly adsorbed to butyl agarose at

high ionic strength attempts to detect a significant binding of 14C_bu_ tylamine and 14C-hexylamine in solution were unsuccessful. We conclude that there are no accessible, high-affinity alkyl binding sites or pockets on the phosphorylase

~

molecule at high ionic strength. However

a newly discovered temperature dependent sol-gel transition of the enzyme indicates the presence of a number of low-affinity hydrophobic . binding areas on the surface. Kinetic studies of the adsorption of phosphorylase

~

to immobilized butyl residues support a biphasic binding

mechanism. Conformational changes may be associated with the adsorption of the enzyme. This is indicated by the finding that freeze-inactivation of phosphorylase

~

is prevented by the adsorption to butyl agarose.

INTRODUCTION Previous studies have shown that phosphorylase

~

is cooperatively

adsorbed to butyl agarose (1) and exhibits adsorption hysteresis (2). A multivalent binding mechanism has been proposed

(1,3). Desorption

kinetics have supported the model of negative cooperative protein ption

adso~

(4). Recently (5) a rationale for the synthesis 'of butyl agaroses

with finite distribution coefficients has been described. However a number of questions have remained unanswered, e.g.: How many hydrophobic binding sit~s exist on phosphorylase ~ in solution? Can they be detected by binding studies with soluble alkylamines? What information can be gained from an analysis of the adsorption kineticsl And finally what role do conformational changes of the enzyme play? Experiments devised to answer some of these questions will be described in this paper.

40

MATERIALS AND METHODS Activation of Sepharose 4B with CNBr and the coupling of n- 14 C-alkylamines was either performed according to ref. 6 (freeze-inactivation experiments) or according to the modified procedure, ref. 7 (all other experiments). 14C-Methylamine and 1- 14 C-ethylamine were obtained from New England Nuclear, n-1- 14 C-hexylamine was obtained from Amersham. n-1- 14 C-butylamine was synthesized from n-1- 14 C-butanol as described ~). The degree of substitution has been determined in our laboratory (6) by the addition of radioactive alkylamine tracers to the 2M solution o~ of the amine to be coupled. Since at that time n-1- 14 C-butylamine was not available the degree of substitution of butyl agaroses was determined with the tracer 1- 14 C-ethylamine (6). However the question remained if different 14C-alkylamines added as tracer to the n-butylamine solution yield identical results. Therefore the degree of substitution of agarose with butylamine was determined with the three tracers 14C_ methylamine, 1- 14 C-ethylamine and n-1- 14 C-butylamine (Table 1). If the

7ABU 1 Ditte~entiae

eageeeing ot gutye 14C-aekyeamine4

BrCN

Sepha~04e

with

t~ace~4

ot homoeogou4

IMMOBILIZED RESIDUE CONCENTRATION

mg/ml

~MOL/ml

PACKED GEL 14C

14C-METHYLAMINE

14C-ETHYLAMINE

14C-BUTYLAMINE

1

14C

2

--uc --uc4 2

8

82.3

±

0.3 ( 3)

18.3

± 1. 2 (3)

4.6

15

130.9

±

6.0 ( 3)

30.9

±

0.4 (3)

4.2

30

180;0

±

8.2 (3)

38.8

±

2.0 (3)

4.6

8

22.4

±

0.5 (3)

18.3

±

0.3 (3)

1.2

15

30.9

±

1.2 (.2)

25.5

±

0.2 (!2)

1.2

30

42.9

±

1.4 . (3)

35.6

±

0.3 (3)

1.2

Activation and coupeing we~e pe~to~m~d acco~ding to ~et. 7. the anaey4i4 ot the gee4 i4 d~4c~iged in ~et. 6. 14C1. 14C2 and 14C4 denote the t~ace~4 ot methyeamine. ethyeam.ine and gutyeamine ~e4pectiveey. 1he numge~ ot anaey4e4 i4 given in pa~enthe4e4. ro~ tu~the~ detaie4 4ee text.

41

ethylamine tracer

~s

employed the apparent degree of sUbstitution is CL

1.2-fold higher than the value obtained with the butylamine

tracer~

The

significance of this difference is being examined with higher members of the homologous series. However addition of the methylamine tracer leads to ca. 5.5-fold higher results than expected. The calculated ratios (see Table 1) are independent of the CNBr concentration employed for activation. The basis for the differential reactivity of the alkylamines is still unclear since the basicity of the amine does not

cor~

relate with the enhanced labelling effect (8). It maybe that steric or apolar effects play an important role. All methods pertaining to the preparation and determination of activity of phosphorylase ~ (ca. 80 U/mg) and the radioactive labelling of the enzyme (3H-phosphorylase ~r' ca. 45 U/mg) have been extensively described (1, 2). The high ionic strength buffer A employed contains 10 mM tris(hydroxymethyl) aminomethane/maleate, 5 mM dithioerythritol, 1.1 M ammonium sulfate, 20% sucrose, pH 7.0.

Binding measurements of alkylamines to phosphorylase

~

were per-

med in a flow dialysis apparatus of Feldmann (9) according to the method of Colowik and Womak (10)~ The upper chamber of the cell was filled with 0.5 ml buffer A with or without the enzyme (&0-100 mg/ml) to which the 14C-alkylamine was added to a final concentration of 1-2 mM and ca. 10 6 cpm/ml at 5 0 C. The lower chamber (ca~ 0.03 ml) was perfused with buffer A (18 ml/hr) and 0.3 ml fractions were collected every minute. For radioactivity measurements 0.2 ml aliquots were taken from the fractions and counted in 2 ml scintillation fluid (Quickscint 212, Zinsser, see ref. 2). The methods for measuring sorption kinetics have been previously described (4). Phosphorylase ~ is irreversibly freeze-inactivated (11) under standard conditions by freezing

the

enzyme

(0.3-3 mg/ml) in buffer B

(20 mM sodium B-glycerophosphate,

mM EDTA, 20 mM mercaptoethanol,

pH 7.0) for 3 hours at -18 to _20 0

in a freezer and thawing the enzyme

for 5 min at 30 0



Slow thawing at 50

leads to the same result. For

the preparation of solute free phosphorylase mg/ml) was dialyzed for 24

hr~

~

the native enzyme (10

against four changes of the 600-fold

volume of double distilled water in a nitrogen atmosphere. In spite of the precautions the specific activity of the enzyme decreased to ca. 50 U/mg. The freeze-inactivation experiment with rylase

~

adsorbed

phospho-

was performed on two columns in the following way: A sample

of 1 ml phosphorylase ~ (3.5 mg/ml) was applied to 1 ml packed butyl Sepharose (activated with 20 mg/ml CNBr and coupled according to ref. 6)

42

on a small column (0.8 cm i.d. x 12 cm) in buffer B. After washing the gel with 5 ml buffer B the column was allowed to run dry by gravity. The control column remained at 5 0 C. The gel of the second. column was removed~ frozen at - 20 0 C for 3 hours and then thawed at 50 after ad-

'ding 1 ml buffer B. After filling the thawed gel back into the column both columns were eluted with buffer B in which the pH had been lowered to pH 5.6 with HCI (6). The eluted enzyme was then analyzed. RESULTS AND DISCUSSION Binding Studies Binding of immobilized alkyl residues to protein. One of the most \

important parameters governing the adsorption of a protein to immobilized alkyl residues is the surface concentration of residues (5.6). If the protein contains more than one available binding site for the immobilized residue a sigmoidal binding curve (adsorbed protein vs. immobilized residue concentration) will be obtained (for review see ref. 4.5). The adsorption of phosphorylase E on butyl Sepharose is a good example for the cooperative binding of immobilized alkyl residues to a protein • Binding of soluble alkyl residues to protein. The qualitatively different aspect in the interaction, of a protein with immobilized or soluble residues lies in the fact that in the former case sional concentration

(mol/m 2 )

a two-dimen-

and in the latter case a three-dimensional

(mol/I) concentration of residues is effective. This difference becomes fully evident in the case of phosphorylase

E.

~

Fig. 1 shows

flow di-

alysis expiriment in which 1-2 mM n-1- 14 C-hexylamine in buffer A is incubated with phosphorylase

E

(0.85 mM monomer units). A similar experi-

ment was conducted with n-1- 14 C-butylamine (not shown). In contrast to the experiments with immobilized butyl residues (1.2) no significant binding of soluble butyl- or hexylamine to phosphorylase

E

can be detec-

ted (see Fig. 1). Thus the binding constant of the protein-hexylamine interaction must be so low that a mixture of mM concentrations of the binding partners does not lead to a significant saturation of putative hydrophobic areas or pockets on phosphorylase

E.

In adffition no binding

was observed at higher temperatures. We therefore conclude that there are no high-affinity alkyl binding sites on phosphorylase

E.

Specific

high-affinity binding sites for small alkanes have however been found on a variety of other proteins e.g. hemoglobin. myoglobin and B-lactoglobulin (12). For the binding of pentane the binding constant (K 1 ) of the specific. high-affinity binding site was between 1-8 x 10 3 M- 1 for

43

E E a. u

~

I.) BUFFER (0)

6000

BUFFER+ ENZYME

~

l-

>

5000



l-

u

0

«

4000

0 « a::

3000

UJ

::::>

0

0

q

0

0

0



I-

Z

•••••

2000

0

...J

u. u.

UJ

1000

10

5

20

15

TIME. min

fig. 1. Binding analy.6i.6 o/- n-1- 14 C-h€.x.ylamin€. to p!t.o.6phoJ1.ylaA€.! at high ionic .6tJ1.€.ngthin a /-low-dialY.6i.6 c€.le. 7h€. aJ1.J1.OW.6 indicat€. th€. addition 01- 10 ~l 0.05 ~ n-1- 14 C-h€.x.ylamin€. (5 x. 10 7 cpm/ml) to th€. upp€.J1. chamg€.J1. containing 0.5 me .6olution (.},(O). Sampl€..6 W€.J1.€. tak€.n at th€. indicat€.d tim€..6 /-J1.om €./-/-lu€.nt /-J1.action.6 o/- th€. low€.J1. chamg€.J1. (1 /-J1.action/min). 7h€. J1.adioactivity o/- th€. amin€. in th€. /-J1.action.6 i.6 giv€.n on lh€. oJ1.dinat€.. fOJ1. /-uJ1.th€.J1. d€.tail.6 .6€.€. J1.€./-. 9, 10 and ~at€.J1.ial.6 and ~€.thod.6. the three mentioned proteins (12) whereas the low affinity binding sitffi showed constants (K 2

)

in the range of 1-6 x 10 2 M- 1



From the apparent

a.ssociation constants of half-maximal saturation (K o • 5) for the binding of phosphorylase

£

to butyl agaroses (20-30 ~mol/ml packed gel) of

9-16 x 10 4 M- 1 (1,2) and the estimated minimum number (i.e. 3-4) of butyl residues interacting with the enzyme during nucleation (2,3) apparent association constants for a monovalent interaction of the enzyme with ~ne butyl residue can be calculated according to (9-16 x 10 4 to be of the magnitude 17-54 M- 1



)1/3-4

Binding constants in this range are

below our detection limit (see Fig. 1) and also demonstrate that the binding of phosphorylase

£ to butyl Sepharose requires a multivalent

type of mechanism. Detection of low-affinity hydrophobic binding sites. For soluble, globular proteins the existence of hydrophobic areas on the accessible surface has been suggested by studies on protein structure (13,14). Such information is as yet unavailable for phosphorylase

£.

We could·

44

fig.

2.

7empe4aiu~e

induced 60e-gee

i~an6iiion

ot ph06ph04yea6e

A. yee-6iaieot ph06pho~yea6e ~ aeiquoi atie~ wa~mbng io 30 0 B. Soe-6iaie ot 6econd aeiquoi ot enzyme kepi ai 0 C. C. Spi~ii-eevee. D. ceamp.

~.

c.

however obtain strong evidence for the existence of such hydrophobic surface sites on phosphorylase £ through the discovery of a fully reversible,temperature dependent sol-gel transition of the enzyme at high ionic strength (see Fig. 2): 30 0 phosphorylase £-sol

< 0

>

phosphorylase £-gel

(1 )

0

Since the gel developes after increasing the temperature from 0 0 to 30 0 endothermic, hydrophobic interactions apparently underly gel formation. This indicates the presence of numerous low-affinity hydrophobic surface sites on the. enzyme which allow polymerization under these conditions.

45

0..0. 91 0

0..0. a -1

0..07 -2 N

0.0.6



kt2

E-Aga

;,==='

(2 )

Elf -Aga

(E * ) denotes a different species of bound 3H-phosphorylase £r due to an increase in the number of butyl residues interacting with the enzyme (2) or possibly to an altered conformation. Preliminary values for the magnitude of the rate constants derived from Fig. ding reaction of 3H-phosphorylase £r are k+1

=

4

for the fast bin-

13.5 mM- 1 min- 1 and k_1

0.332 min-I. The respective constants for the second slow phase were calculated to be k+2 = 0.068 min- 1 and k_2 = 0.0]8 min-l. Further propagation steps (2) have as yet not been amenable to study. In other experiments it has been found that the initial adsorption rate is a hyperbolic function of the free enzyme concentration

(4.16).

This is especially interesting since the binding curves of phosphory-

45

0..0. g"

0.0.8 -1

om

-2 N

0..0.6



E

0.0.4

-5

u 0.03

10

20

30

40

50

60

70

t.min

0.0.2

x~ )(

0.Q1

10.

20.

~)(--)(---i.'"

3D

40.

50. t. min

60.

70

80

go.

rig. 3. 7im~ d~p~ndence ot th~ ad~o~ption ot 3H-pho~pho~yla~~ g on gutyl S~pha~o~~ in gutt~~ A. -~ In an incugation mixtu~~ ot 60 ml(2,4) the ~~action waA Ata~t~d gy addition ot 1.5 ml pack~d gutyl S~pha~o~~ (ca. 20 ~mol/ml pack~d g~l) co~~~~ponding to a tinal conc~nt~ation ot ginding unit~ ot ca. 10 n~. 7h~ initial conc~nt~ation (e) ot enzym~ waA ca. 0.09 mg/ml (0.9 n~). At th~ indicat~d tim~A Aampl~~ w~~e tak~n and count~d (2). InA~~t: S~miloga~ithmic plot ot the data. ro~ detinition ot in 6Z ~e~ ~et. 15. 10~ tu~the~ detailA Aee ~et. 2,4 and the text. (I) and (X) denote two Aepa~at& ~xp~~imentA und~~ identical conditionA. Kinetic Studies Adsorption kinetics. Fig. 3 shows the adsorption Df 'H-phosphorylase ~r on butyl Sepharose under pseudo-first order conditions i. e. 10

[El,

where Aga denotes binding units

(3.4) and

E the

[Aga] ~

enzyme ligand.

The semilogarithmic plot in the insert shows that adsorption occurs in a biphasic manner: a fast phase (t 1j2

",

5 min) and a slow phase (t 1j2

15 min). Previously only the first phase was considered

'"

(4). From these

phasss the relaxation times (15) were calculated at increasing concentrations of butyl agarose (see Fig.

4).

The reciprocal relaxation times

of the fast phase appear to depend linearly on the concentration of butyl Sepharose, however those of the slow phase are practically independent of the concentration of excess component. The concentration independence of the latter relaxation times is strong evidence against a

47

lase

B show

negative cooperativity (1). On the other 'hand the result is

in agreement with a saturable rate-limiting step of adsorption (see eq. 2.

Role of Enzyme Conformation In the studies of alkane . binding to proteins (17) it was found that conformational changes of the protein can exert a strong influence on the binding of hydrocarbons. Similarly conformational changes may be associated with the adsorption of proteins on alkyl agaroses and possibly with the phenomenon of adsorption hysteresis. Freeze-inactivation of phosphorylase

B also

appears to be related to conformational changes.

Freeze-inactivation of phosphorylase b. It has long been known that phosphorylase ture to ca. 0 0 natured

B can

be reversibly inactivated by lowering the

tempera~

(18). On freezing however the . enzyme is irreversibly de-

and precipitated

(11). This freeze-inactivation can be preven-

ted if certain cryoprotective solutes e.g. glucose. betain or hydroxprolin are added to buffer B in moderate concentrations (50-100 mM) (11). The mechanism of the inactivation and cryoprotection however remained obscure until it was found that the buffer constituents (i.e. buffer salts and mercaptoethanol) are responsible for the inactivation. Freeze-inactivation is almost entirely eliminated after removal of all buffer components from the enzyme solution as is shown in Table 2.

7a£.le. 2 lntlue.nce. ot t~e.e.zing on and me.~captoe.thanol.

Unfrozen control

pho~pho~yla~e. ~

in the.

a£.~e.nce.

ot

£.utte.~

Betain

Phosphorylase b

mM

% activity

0

100

1

0

2

0.1

84 88

3

1

85

4

10

91

~alt

Frozen samples

conce.nt~ation ot ~olute. t~e.e. e.nzyme. p~e.pa~e.d a~ de.~c~i£.e.d in Mate.~~ and Me.thod~ wa~ 0.3 mg/ml in £.utte.~ B. De.tail~ ot the. t~e.e.ze.-inac­ tivation p~oce.du~e. a~e. give.n in the. ~ame. ~e.ction.

7he.

al~

After freezing there is only a small ca. 15% loss in activity and therefore no significant cryoprotective effect of betain •. The highest concen-

48

tration of betain was only 10 mM since the cryoprotective effect increases when the concentration of buffer B components is reduced. Since it has been shown (18) that phosphorylase

£

can bind cystein in stoichio-

metric a~ounts and that the bound cystein can be exchanged by mercaptoethanol, freeze-inactivation may be due to a direct interaction of mercaptoethanol and possibly buffer salts with the enzyme. This may lead _ to conformational changes of the enzyme. The following reaction scheme is proposed for the freeze inactivation of phosphorylase

£:

Phosphorylase -b( cryos t a bl e ) ~ ~ Phosphorylase -b( cryo 1 a,bOlo ) 1 e

0)

Thus the equilibrium of eq. 3 may be shifted from the cryostable form to the cryolabile form through the buffer constituents. Alternatively this might result at the high local solute concentrat,ions occurring during freezing. This shift can be reversed by addition of cryoprotectants o~ by the removal of the inactivating buffer solutes (Ta~le 2). Further evidence for the importance of enzyme conformation in inactivation is the finding that phosphorylase

£

freez~

can ~e protected by

adsorption to butyl Sepharose(Table 3). 7agle 3 P~otection ot phohpho~ylahe! on gutyl Sepha~ohe.

agdinht

t~eeze

Eluted enzyme U/mg Control gel (50) Frozen gel

(_20 0 )

inactivation gy

adho~ption'

Yield

%

84

83

69

84

7he amount ot enzyme adho~ged wah ca. 3 mg/ml packed gel which wah taken ah 100~ in calculating the yield. 7he eluted t~actionh we~e pooled and analyzed. tO~ tu~the~ detailh hee ~ate~ialh and ~ethodh. Table 3 demonstrates that the enzyme adsorbed on butyl Sepharose in buffer B retains 82% of its initial specific activity, which it would have fully lost by freezing in solution alone. We therefore conclude that the binding of immobilized butyi residues to phosphorylase

£

stabilizes the enzyme in a conformation not liable to freeze-inactivation.

49 ACKNOWLEDGMENTS We thank Mrs. G. Botzet and Mrs. I. Bichbaumer for excellent technical assistence. This work was supported by Grant Je 84/6-5 from the Deutsche Forschungsgemeinschaft and by the Fonds der Chemie. REFERENCES

1• 2.

H.P. Jennissen, Biochemistry, 15 (1976).5683-5692 H.P. Jennissen and G. Botzet, Int. J. Biolog. Macromolecules,

3. 4. 6.

H.P. H.P. Vol. H.P. H.P.

7.

H.P. Jennissen, Protides BioI. Fluids Proc. Colloq., 23 (1976)

8.

E. Logemann and H.P. Jennissen, Hoppe-Seyler's Z. Physiol. Chern.,

5~

(1979) 171-179

Jennissen, J. Chromatogr., 159 (1978) 71-83 Jennissen, in G. Weber (Ed.), Advances in Enzyme Regulation 19, Pergamon Press, New York, 1981 in press Jennissen, J. Chromatogr., (1981) in press Jennissen and L.M.G. Heilmeyer, Jr., Biochemistry, 14 (1975)

754-760

675-679

361, (1980) 295-296

K. Feldmann, Anal. Biochem., 88 (1978) 225-235 10. S.P. Colowik and F.C. Womak, J. BioI. Chern., 244 (1969) 774-777 11. B. Schobert and H.P. Jennissen, Hoppe-Seyler's Z. Physiol. Chern.,

9.

12. 13. 14. 15.

361 (1980) 329-330

A. Wishnia, Biochemistry, 8 (1969) 5064-5070 I.M. Klotz, Arch. Biochem. Biophys., 138 (1970) 704-706 C. Chotia, J. Mol. Biol., 105 (1976) 1-14 C.F. Bernasconi, Relaxation Kinetics, Academic Press, New York, 1976, pp. 21-29, 141-147 16. H.P. Jennissen, J. Solid-Phase Biochem., 4 (1979) 151-165 17. A. Wishnia and T.W. Pinder, Biochemistry, 3 (1964) 1377-1384 18. S. Shaltiel, J.L. Hedrick and E.H. Fischer, Biochemistry, 8 (1969)

2429-2436