Potential for ambiguity in the interpretation of immunochemical studies on multienzymic proteins

Biochimica et Biophysica Acta, 707 (1982) 193-198

193

Elsevier Biomedical Press

BBA 31301

POTENTIAL FOR AMBIGUITY IN THE INTERPRETATION OF IMMUNOCHEMICAL STUDIES ON MULTIENZYMIC PROTEINS THE CASE OF URIDYLATE SYNTHASE R O N A L D W. M c C L A R D * and M A R Y ELLEN JONES

Department of Biochemistry, University of North Carolina, School of Medicine, Chapel Hill, NC 27514 (U.S.A.) (Received April 16thl 1982)

Key words: Uridylate synthase," Multienzyme complex; Immunoreaetivity

Rabbit antibodies directed against homogeneous uridylate synthase multienzyme from mouse Ehrlich ascites carcinoma precipitate both the orotidine-5'-monophosphate decarboxylase (EC 4.1.1.23) and orotate phosphoribosyltransferase (EC 2.4.2.10) activities of mouse and human erythrocyte uridylate synthase. When the partially purified human enzyme is used as antigen the two activities coprecipitate with the same apparent titer; however, when the mouse carcinoma protein was studied under the same conditions the decarboxylase activity immunoprecipitated with significantly higher avidity than did the transferase activity. Since the mouse multienzyme has been shown to be a single polypeptide that contains both activities (McClard, R.W., Black, M.J., Livingstone, L.R. and Jones, M.E. (1980) Biochemistry 19, 4699-4706), these results were, at face value, surprising. However, when the mouse orotate phosphoribosyltransferase activity (which is largely lost upon dilution into the immunoassay medium) was stabilized with 5-phosphoribosyl 1-pyrophosphate, both enzyme activities displayed the same apparent antibody titer. The immunochemical studies indicate that the antibodies, as a population, preferentially bind to a form or forms of the enzyme which contain(s) denatured transferase domains. A calculation based on a simple model yields a value of approximately 100 for the relative selectivity of the antibody for the denatured form of uridylate synthase. These results illustrate an ambiguity that is inherent in the interpretation of immunochemical studies on such multienzymic proteins; that is, it is possible to conclude incorrectly that two enzyme activities are not functionally associated if one of the catalytic domains is particularly unstable and thereby displays greater immunoreactivity for the specific antiserum.

Uridylate synthase is a single polypeptide that contains both of the last two enzyme activities for the de novo synthesis of UMP [1]. The protein has a monomeric subunit molecular weight of approx. 51000 and undergoes dimerization [2] to become the fully active form [3]. The orotate phosphoribosyltransferase domain is notably unstable and disproportionate losses in this activity have been consistently observed [1,4-10]. * Present address: Department of Chemistry, Boston College, Chestnut Hill, MA 02167, U.S.A. 0167-4838/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

Rabbit antibodies were recently developed against this bifunctional polypeptide purified from Ehrlich ascites cells, and the antibody-antigen reactions can be conveniently monitored using Staphylococcus aureus (Cowan) cells [11]. The antibodies also recognize the human erythrocyte protein with greatly reduced affinity [12]. Both enzyme activities of the human enzyme co-precipitate with identical apparent titer upon addition of anti-IgG-S, aureus cell complexes [13] to partially purified enzyme protein; this result is consistent

194

with the hypothesis that the human enzyme is also bifunctional [14]. Such a finding could be regarded as convincing evidence for the existence of a functional complex comprising these two catalytic activities, and such an experiment should, in general, be applicable to any protein that may be considered a putative enzyme complex. However, when the same experiment was performed using the mouse Ehrlich ascites protein and the antiserum directed specifically against it, we obtained the unexpected result that the orotidine-5'-monophosphate decarboxylase activity (on the basis of percent of control) was immunoprecipitated with much lower amounts of anti-lgG-S. a u r e u s cell complexes than was the orotate phosphoribosyltransferase activity. At face value this was surprising in view of the fact that the two activities are associated on a single polypeptide molecule [1]. However, this seemingly ambiguous finding can be satisfactorily explained by a model that assumes that the antibody population has a preference for those enzyme species that contain denatured transferase domains. A simple model can be constructed in order to estimate the magnitude of this selectivity. A preliminary account of this work has been presented [12].

mouse Ehrlich ascites carcinoma uridylate synthase, used in immunoprecipitation reactions, was the fraction obtained after precipitation by (NH4)2SO 4 and dialysis [1]. The specific activity of orotidine-5'-monophosphate decarboxylase was about 10 m u n i t / m g and the ratio of orotidine-5'monophosphate d e c a r b o x y l a s e / o r o t a t e phosphoribosyltransferase was typically about 2. Assays of orotate phosphoribosyltransferase and orotidine-5'-monophosphate decarboxylase were performed as described elsewhere [1]. In experiments where the transferase activity was a small fraction of the decarboxylase activity, the transferase activity was linear with respect to protein concentration. Results and Discussion Rabbit antiserum directed against mouse uridylate synthase [11] recognizes human uridylate synthase (Fig. 1); both the orotidine-5'-monophosphate decarboxylase and orotate phosphoribosyltransferase activities are precipitated with the same apparent titer. This observation is consistent with the hypothesis that the human enzyme is also a bifunctional polypeptide [1,14]. In contrast, these two activities of the mouse polypeptide display an

Experimental Antiserum directed against homogeneous uridylate synthase [1] was raised as previously described [11]. Immunoprecipitation assays were performed essentially as described elsewhere [11]. Briefly, the assays were performed in 20 mM Tris-HC1, pH 7.5, plus 10 mM sodium ethylenediaminetetraacetate, 150 mM NaC1, 0.1% Nonidet P-40 and 0.01% NaN 3 (total volume, 0.5 ml). Each assay contained 50 /zg non-immune rabbit IgG and 1 munit uridylate synthase (measured as orotidine-5'-monophosphate decarboxylase activity). Immune complexes were precipitated by the addition of 50 ~1 of washed S. a u r e u s cells [11]. Partially purified human erythrocyte uridylate synthase was obtained by DEAE-cellulose chromatography of hemolysates as described in the literature [10,15]. The specific activity was 0.8 n m o l / m i n per mg protein and the ratio orotidine5'-monophosphate decarboxylase/orotate phosphoribosyltransferase was 1.7. Partially purified

•

I

5-

10

100

.E E

50

\,

\

\\

00 HI

antiserum

added

Fig. 1. lmmunoprecipitation of human erythrocyte uridylate synthase by rabbit anti-mouse uridylate synthase serum adsorbed to S. a u r e u s cells. The titration was performed as described in the Experimental section. O, orotidine-5'-monop h o s p h a t e decarboxylase; ©, orotate p h o s p h o r i b o syltransferase.

195 i

i

100(

\ D\O \ \ c

T/D = 0.067

S. oureu$ plus

antibodies \ \

cI in solution (T/D = 0.133)

>

0

1

Fig. 4. Proposed model to explain the differential precipitation of orotidine-5-monophosphate decarboxylase and orotate phosphoribosyltransferase activity by rabbit anti-mouse uridylate synthase serum. C ~ , native uridylate synthase ( T / D =0.2); C I , uridylate synthase with denatured orotate phosphoribosyltransferase domain ( T / D = 0).

2

JJI antiserum

added

Fig. 2. I m m u n o p r e c i p i t a t i o n o f mouse uridylate synthase by rabbit anti-mouse u r i d y l a t e synthase serum adsorbed to S. a u r e u s cells. The titration was performed as described in the

Experimental section. O, orotidine-5'-monophosphate decarboxylase; (3, orotate phosphoribosyltransferase.

i

i

i

i

I

i

80 ( 60

40 i

c t~

Ief

cL 0

8

~

6

©

F 0

--©

© 4

i

0

L

i

i

I

,

J

,

50 % OMPDase

,

in pellet

I

100 removed

Fig. 3. Re-plot of immunotitrations; the plot is expressed as the ratio OMPDase/OPRTase (orotidine-5'-monophosphate decarboxylase/orotate phosphoribosyltransferase) remaining in the supernatant during the course of immunoprecipitation of the orotidine-5'-monophosphate decarboxylase activity (abscissa). Q, data shown in Fig. 2; &, re-plot of a similar experiment with a second enzyme preparation; C), re-plot of the data shown :~n Fig. 1.

apparent difference in their respective avidities for antibodies in the antiserum, as shown by the example in Fig. 2. The orotidine-5'-monophosphate decarboxylase activity appears to be precipitated by the IgG-S. a u r e u s cell complex by lower amounts of specific antiserum (on the basis of percent activity of control). The precipitation of the orotate phosphoribosyltransferase activity appears to lag behind that of the decarboxylase (Fig. 2). This unexpected result can be better understood by replotting the data of the experiment shown in Fig. 2. Fig. 3 shows a plot of the ratio of orotidine-5'-monophosphate decarboxylase/ororate phosphoribosyltransferase that remains in solution as a function of the amount of orotidine5'-monophosphate decarboxylase activity that has been immunoprecipitated. It is apparent that the ratios of orotidine-5'-monophosphate decarboxylase/orotate phosphoribosyltransferase activities in the control (0% precipitation) samples are much greater than the ratio of 2 (as mentioned in the Experimental section) measured prior to dilution of the enzyme into the immunoassay mixture. As expected, the rise in this ratio is caused by significant inactivation of the orotate phosphoribosyltransferase activity upon dilution into the immunoassay medium [9]. Curve A is a replot of the data of Fig. 2. Curve B depicts a separate experiment (like that depicted by curve A) and illustrates the variability among enzyme preparations of the control (0% precipitation) activities ratio (ordinate). The data for the human enzyme from

196

[

]

I

I

I

ff--

I

--

100= 80 ~= >

60

c 0 o

~

40

~

0

0.1

0.2

0.3

0.4

1.0

pl antiserum

Fig. 5. Collateral immunoprecipitation of mouse uridylate synthase activities in the presence of P-Rib-PP, a stabilizer of the orotate phosphoribosyltransferase activity. This experiment is identical to that shown by Fig. 2 except that 1 mM P-Rib-PP was included in the immunoassay medium. ~ , orotate phosphoribosyltransferase; II, orotidine-5'-monophosphate decarboxylase.

Fig. 1 are also replotted in Fig. 3. It is evident from Fig. 3 (and inferred from Fig. 2) that as the mouse orotidine-5'-monophosphate decarboxylase activity is progressively immunoprecipitated, the ratio orotidine-5'-monophosphate decarboxylase/orotare phosphoribosyltransferase approaches an extrapolated (at 100% precipitation) value near 5. Essentially the same result was obtained when uridylate synthase, obtained directly from a fresh Ehrlich ascites cell homogenate, was used as the antigen. The activities ratio is approx. 5 for the human enzyme throughout the course of the immunoprecipitation.

The most rational explanation for these observations is that the anti-uridylate synthase antibodies preferentially bind to enzyme molecules that have a denatured orotate phosphoribosyltransferase domain. Such denatured species must exist, as evidenced by the variable but high values of orotidine-5'-monophosphate decarboxylase/orotate phosphoribosyltransferase in the control assays (0% precipitation). A diagram which illustrates this hypothesis is presented in Fig. 4. If this model were correct, one could predict that the orotidine-5'-monophosphate decarboxylase and orotate phosphoribosyltransferase activities would coimmunoprecipitate under conditions where the phosphoribosyltransferase activity has been selectively stabilized. Indeed, when stabilization is accomplished by the addition of 1 mM P-Rib-PP, the ratio of orotidine-5'-monophosphate decarboxylase/orotate phosphoribosyltransferase rises to about 2.5 throughout the course of immunoprecipitation and both activities indeed co-precipitate (Fig. 5). The populations of anti-uridylate synthase IgG species apparently partition themselves between reactions with native (Cff]) or denatured ((211) species (Fig. 6). The formation of either complex

I

A

ii

100c

*

--

~._

j, 01

~,i".¸

•

•

•" ~

~ 60

\ !X

\

f 01

\

- - ,

~ 40 ~ 2o

_ _

"". . . . .

~!.

"

~nl~

Y

~

1000

f100,1000

~

0 0

Y

"V

~ Kt

K /

K Fig. 6. Operational definition of o, the relative selectivity vs. uridylate synthase. Y, antibody; CC], native uridylate synthase; CU, denatured uridylate synthase.

05

1 JJl

0.5

1

antiserum

Fig. 7. Experimental data, (open circles) and theoretical curves (defined by dots) generated by Eqn. 5 of the text. The open circles are the values of "orotidine-5'-monophosphate decarboxylase activity remaining in solution upon the addition of anti-uridylate synthase serum (data depicted by curves • and • in Fig. 3). The theoretical points (dots) are values of orotate phosphoribosyhransferase total domains (CC] + C U ) remaining in solution. The points were computed using Eqn. 5 and by varying the value of o, the relative selectivity (see Fig. 6), from 0.1 to 1000.

197

can be assigned an equilibrium constant and the ratio of these constants is termed the relative selectivity, o (Fig. 6). In this case, the higher the value of o, the greater the relative affinity of the antibody populations for the denatured enzyme species. An approximate value of o can be calculated from the following treatment. The initial (0% precipitation) fraction of native transferase domains, F0, is given by C)D

F°=5(T/D)°-- o : ~ + ~

(1)

(T/D refers to the ratio orotate phosphoribosyltransferase/orotidine-5'-monophospha.te decarboxylase and is inverted at this point for the sake of computational simplicity. Up to this point the ratio D / T was more convenient for the graphical presentation of Fig. 3. The change to the use of T/D for the ensuing calculations is necessary to keep the numbers finite. That is, completely denatured orotate phosphoribosyltransferase would otherwise yield D / T = oo. The coefficient value of 5 represents the limiting value of D / T that native uridylate synthase displays according to the data of Fig. 3.) In addition, a selectivity function is defined as o

q'= o+1

(2)

This hyperbolic function, 4', approximates the extent of competition of the two classes of enzyme molecules for the whole population of antibody molecules (Fig. 6). Thus the fraction of denatured enzyme species ( C I ) precipitated at any point along the immunotitration curve would be given by q, X ~ pt = f r a c t i o n o f C I

precipitated

(3)

since such species exhibit only decarboxylase activity. Values of X ppt, determined experimentally, represent the fraction of control decarboxylase activity immunoprecipitated at any point. The fraction of native transferase species ( C ~ ) immunoprecipitated would then by approximated by ( 1 -- q~ ) F o X~. vt = f r a c t i o n o f C ~

precipitated

(4)

where XTppt is the fraction of control transferase

activity immunoprecipitated. These values are also determined experimentally. The factor F0 is included to reflect the fact that only that fraction of enzyme species contains measurable transferase activity in the control experiment. Thus the fraction of total transferase domains (C£2 + C I ) that still remain in the supernatant at any point along the immunotitration progress curve can be approximated by: X~U p = 1 - q) X ~ p' - ( 1 -- q, ) F 0 X:~ p'

(5)

Various values of o can now be assumed in order to estimate the magnitude of the relative selectivity, o, required to cause the immunotitration curves of decarboxylase activity (experimental) and transferase (calculated) to superimpose. The result of such a theoretical generation is shown in Fig. 7. Curves for both experiments A and B (of Fig. 3) are displayed. The method of estimation cannot adequately distinguish a selectivity value above 102; however, it is clear that the selectivity of the antibody population for denatured transferase domains must be at least of that magnitude to explain the differential immunoprecipitation of the decarboxylase and transferase activities on the basis of percent of control activity. The fact that the decarboxylase activity can be maintained on a single polypeptide, even though the transferase activity has been mostly lost, supports the hypothesis [14] that these two separate catalytic functions of this bifunctinal polypeptide might well be housed within distinctly separated regions of the polypeptide chain. The recent report by Floyd and Jones [16] indicates that partial proteolytic digestion of homogeneous mouse uridylate synthase yields distinctly sized fragments which include one 28 500 dalton species that comprises the decarboxylase domain. The apparent ease with which the putative transferase domain is denatured in the absence of P-RJb-PP is consistent with the facts that the denatured species is preferentially immunoprecipitated (this work), and that upon partial proteolysis in the absence of P-Rib-PP the transferase activity is rapidly lost [16]. The example examined here provides an interesting caution to those who study the immunochemical properties of multienzymes or putative multienzymes. The diversity of species within

198 a p o p u l a t i o n of a n t i b o d i e s is a well-established generality, and it is not surprising that one p o r t i o n of a complex molecule might be selectively sought out by a highly specific sub-set of i m m u n o g l o b u l i n molecules. In this case, the high-titer a n t i b o d y p o p u l a t i o n a p p e a r s to have been raised against molecules in which part (the transferase d o m a i n ) has been d e n a t u r e d . I n d e e d this idea is perfectly consistent with the overwhelming consensus that the transferase d o m a i n is far m o r e labile than the rest of the molecule [1,4-10]. Clearly one would p r o b a b l y even predict a priori that when uridylate synthase is h o m o g e n i z e d with a d e n a t u r a n t such as F r e u n d ' s a d j u v a n t (or simply b y dilution [9]), one p r e d o m i n a n t species with d e n a t u r e d transferase is p r o d u c e d . It then would be expected that the i m m u n e response w o u l d be directed most p o t e n t l y against the a b n o r m a l protein. I n d e e d the hisB m u l t i f u n c t i o n a l gene p r o d u c t of Salmonella typhimurium l i k e w i s e d i s p l a y s e n h a n c e d i m m u n o r e a c t i v i t y when d e n a t u r e d with s o d i u m d o d e c y l sulfate [ 17].

Acknowledgements This work was s u p p o r t e d , in part, by grants from the N a t i o n a l Science F o u n d a t i o n ( P C M 7902623) a n d the N a t i o n a l Institute of Child Health and Human Development, National Institutes of H e a l t h (2 R01 HD12787). Dr. McC l a r d was a Fellow of the N a t i o n a l C a n c e r Institute, N a t i o n a l Institutes of H e a l t h (3 F32

CA06386). The a u t h o r s wish to thank D r . . M i c h a e l C a p l o w for his helpful suggestions with regard to this manuscript.

References 1 McClard, R.W., Black, M.J., Livingstone, L.R. and Jones, M.E. (1980) Biochemistry 19, 4699-4706 2 Traut, T.W. and Jones, M.E. (1979) J. Biol. Chem. 254, 1143-1150 3 Traut, T.W. and Payne, R.C. (1980) Biochemistry 19, 6068 6074 4 Kasbekar, D.K., Nagabhushanam, A. and Greenberg, D.M. (1964) J. Biol. Chem. 239, 4245-4249 5 Appel, S.H. (1968) J. Biol. Chem. 243, 3924 3929 6 Grobner, W. and Kelley, W.N. (1975) Biochem. Pharmacol. 24, 379-384 7 Reyes, P. and Guganig, M.E. (1975) J. Biol. Chem. 250, 5097-5108 8 Brown, G.K., Fox, R.M. and O'Sullivan, W.J. (1975) J. Biol. Chem. 250, 7352-7358 9 Kavipurapu, P.R. and Jones, M.E. (1976) J. Biol. Chem. 251, 5589-5599 10 Brown, G.K. and O'Sullivan, W.J. (1977) Biochemistry 16, 3235-3242 11 McClard, R.W. (1982) Appl. Biochem. Biotechnol. 7, 307313 12 McClard, R.W. (1981) Fed. Proc, Fed. Am. Soc. Exp. Biol. 40, 1795 (abstract) 13 Kessler, S.W. (1975)J. lmmunol. 115, 1617-1624 14 Jones, M.E. (1980) Annu. Rev. Biochem. 49, 253 279 15 Hatfield, D. and Wyngaarden, J.B. (1964)J. Biol. Chem. 239, 2580 2586 16 Floyd, E.E. and Jones, M.E. (1982) Fed. Proc., Fed. Am. Soc. Exp. Biol. 41, 1026 (abstract) 17 Staples, M.A. and Houston, L.L. (1979) J. Biol. Chem. 254, 1395 1401