ENDOCRINE RESEARCH COMMUNICATIONS, 4 ( 3 & 4 ) , 205-215 (1977)


V.G. Hum, H.G. Batting" and K.F. Mori

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

Drug Research Laboratories, Health Protection Branch, Health & Welfare Canada, Ottawa, Ont., Canada K1A OL2 SUMMARY

Native human chorionic gonadotropin (hCG) was resistant to carboxypeptidase digestion even in the presence of urea. Isolated a subunit of the hormone (hCG-a), though unreactive to enzyme treatment in the absence of denaturant, released up to four amino acid residues from the C-terminus on incubation with a mixture of carboxypeptidases A and B in urea. While an hCG-a product which lacked Ser-92 recombined completely with intact hCG-B, hCG-a from which Ser-92 and Lys-91 were removed showed only partial recombination. The two recombinants were devoid of any in vivo biologic activity, but retained some of the immunologic activity of the native recombinant. These findings indicate that the integrity of the C-terminal residue of serine in hCG-a is essential for the expression of & vivo biologic activity of the native hormone. The primary structure of hCG1 has recently been established and thus selective chemical modification of the hormone permits the correlation of the biologic and/or immunologic activities with


Bureau of Nutritional Sciences. Abbreviations used are: hCG, hCG-a and hCG-6, human chorionic gonadotropin and its a and 8 subunits; CPA and CPB, carboxy peptidases A and B; PAGE, polyacrylamide gel electrophoresis; and LH, luteinizing hormone. 205 Copyright 0 1977 by Marcel Dekker, Inc. All Rights Reserved. Neither this work nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage and retrieval system, without permission in writing from the publisher.



certain specific functional molecular groups.


Ideally, a chemical

modifier affects its designated group(s) without introducing any side reactions andlor conformational changes.

However, in our

experience we have not achieved such an ideal (1-3).

On the other

hand exopeptidases, such as carboxypeptidase, have proven extremely useful in protein modification work because o f the mild conditions

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

of incubation for the reaction and the specificity of its action. In order to probe the role(s) of the C-terminal regions of hCG in subunit interactions and in biologic and immunologic activities, we digested hCG and its isolated a subunit with carboxypeptidases under various conditions.

The results are presented in this communication.

Such an approach was successfully used by Cheng et al. ( 4 ) for other glycoprotein hormones, LH and throtropin.

MATERIALS AND METHODS Human CG, 12,200 IU/mg, was prepared from a commercial preparation (2650 IU/mg) as described previously (5) and its subunits The following were obtained by the method of Morgan et al. (6). commercial enzyme preparations were used: CPA and CPB (both treated with diisopropyl phosphorofluoridate, Worthington). The concentrations of hCG and its subunits, intact or modified, were estimated from the weights of lyophilized material andlor by amino acid analysis. The methods of PAGE, in vivo bioassay and determination of immunologic activity by complement fixation were given in an earlier publication (3). Recombination of intact or carboxypeptidasedigested hCG-a with intact hCG-B was examined as follows: each subunit was dissolved in 0.01 M phosphate buffer, pH 7.4, to a concentration of 1 mg/ml and equal volumes of these two subunit solutions were mixed. The mixture was incubated at 370 for 18 hr and an aliquot was subjected to PAGE as described before (2). The remaining mixture was lyophilized and the lyophilizate was chromatographed on a precalibrated column of Sephadex G-100 in 1% ammonium bicarbonate. A protein peak, as determined by optical density at 225 nm, appearing at the same elution position as that of native hCG, was collected to serve as a recombinant formed.



Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

Carboxypeptidase digestion Prior to use, CPA was washed and solubilized according to method i) described by Ambler (7). The concentrations of CPA and CPB were determined spectrophotometrically ( 7 ) . In a few preliminary experiments varying amounts of carboxypeptidases relative to substrates (hCG and hCG-aup to 1/20 (w/w) were used. However, hCG or hCG-adid not release any free amino acids after incubation for 18 hr with CPA or CPB alone (at 1/20) or a mixture of CPA (1120) and CPB (1120) in the absence of denaturant. Hence in preparative experiments all digestion was carried out in 6 M urea (ultra pure, Schwarz-Mann, freshly prepared and deionized on a mixed bed ion-exchange resin) not only because carboxypeptidases are almost fully active in this urea concentration (8) but also because hCG-a after exposure to urea at 3 7 O for 18 hr appeared to renature nearly completely (e.g., See Fig. 1 and Table 111). Samples of hCG and hCG-cL, 5 to 15 mg, were taken in 0.2 M N-ethylmorpholine acetate buffer, pH 8.5, in 6 M urea to a concentration of 5 mg/ml and incubated with either CPA or a mixture of CPA and CPB for 6 and 18 hr at 37O. In each digestion, proper substrate controls (the hormone or its subunit with heat-inactivated carboxypeptidase(s)) and enzyme controls were included. At the end of the incubation a small aliquot was taken and tested for free amino acids by high voltage paper electrophoresis at pH 1.6 (formic acid) and also by amino acid analysis. Another aliquot was shaken with Dowex 50-X8 resin (hydrogen form, 20-50 mesh) to adsorb liberated amino acids which were eluted from the resin with three 2-volume portions of 5 M ammonia. The combined eluates were lyophilized and free amino acids in the lyophilizate were identified by high voltage paper electrophoresis at pH 1.6 and/or paper chromatography in the systems of methanol-pyridine-water (v/v, 201115) and 2-propanolformic acid-water (vlv, 4012150). The rest of the digestion mixture was dialyzed against water and lyophilized. The lyophilized products were tested for biologic and immunologic activities and/or recombination property. In addition, they were subjected to analytical exclusion chromatography on Sephadex G-100 in 1% ammonium bicarbonate and to the N-terminal group analyses by the dansyl chloride procedure (9). '

RESULTS The conditions for the digestion of hCG and its subunit with the various carboxypeptidases and the free amino acids liberated under these conditions are summarized in Table I.

In no case was

free amino acid found in any enzyme or substrate controls.


hCG appeared resistant to digestion with any of the carboxypeptidase



Table I Digestion of hCG and hCG-a with carboxypeptidases i n 6 M urea

Substrate hCG

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

hCG- a

Incubation time (hr)


CPA (1/20) + CPB (1/20)

6 18 6

CPA (1/30)

18 6

CPA (1/60) + CPB (1/30)

18 6

CPA (1/20)

18 CPA (1/30) + CPB (1/30)



Free amino acid detecteda) 0 0 0 0

Ser (0.92) Ser(l.03) Ser (0.94), Lys (1.10) Ser(0.89), Lys(0.98) Ser (0.91), Lys (1.05) His( 0.85), Tyr ( 0 . 8 4 ) Ser(0.86), Lys(0.92) His(0.96), Tyr(l.09)

a) Values in the parenthesis show the average amount of amino acid (moles per mole of substrate) obtained from three independent experiments. The molecular weight of hCG-a was assumed to be 15,000.

preparations in the presence of denaturant.

On the other hand, up

to four amino acids were released from hCG-a after incubation with the carboxypeptidases in urea. single amino acid, serine.

The use of CPA alone released a

The second amino acid, lysine, was

liberated only after incubation with a mixture of CPA and CPB, as had been expected from the known substrate specificities of the enzymes.

Release of the third and subsequent amino acids required

a similar mixture but containing a higher amount of CPA.

There was

no difference in the extent of digestion between 6-hr and 18-hr incubation. When a) the’established sequence at the C-terminus of this subunit, -Cys-Tyr-Tyr-His-Lys-Ser-COOH,

b) the known speci-



ficities of CPA and CPB, and c) the order of the amino acids liberated under the digestion conditions are considered together, it must be concluded that these amino acids were released from the C-terminus by the action of the carboxypeptidases. All of the digestion products showed essentially the same elution profile as that of intact hCG-a in analytical exclusion chromatography and yielded a single residue of alanine in the N-terminal group analysis, the same Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

as that observed for the intact subunit.

In other words, no in-

ternal peptide bond was cleaved and no other peptide.was formed as a result of incubation with the caboxypeptidases.

The amino acid

composition of the lyophilized digestion products is shown in Table 11, the results of which confirm those of Table I. When various hCG-ci products obtained after incubation with the carboxypeptidases were tested for recombination ability (Fig. 1)- the product that lacked one serine residue recombined with intact hCG-B, but the one lacking both a serine and lysine molecule did


only partially (about 30% as estimated from the gel

filtration pattern).

On the other hand, the product lacking four

amino acids did not show any recombination ability at all. When the two recombinants were tested for in vivo biologic activity (Table III), both were inactive.

However, as judged by its complement

fixation (Fig. 2 ) , the first recombinants showed about 6 0 % , whereas the second about 20% of the immunologic activity displayed by the native recombinant or a recombinant of control hCG-a with intact hCG-8.

The complement fixation curve obtained with the second

recombinant and anti-hCG serum displayed, in addition to a vertical



Table I1 Amino acid composition of products obtained from hCG-a by carboxypeptidase digestion in 6 M ureaa)

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

Intact hCG-a Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-cystineb) Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine

5.83 3.23 3.26

6.41 7.70 7.65 9.13 6.81 4.42 5.07 9.69 6.69 2.92 1.18

4.08 3.74 4.06

CPA(1/30) 18 hr

After incubation with CPA(1/60) + CPB CPA(1/30) + CPB (1/30) 18 hr (1/30), 18 hr

5.73 2.68 3.20 6.47 6.95 6.59 9.72 7.14 3.75 5.22 9.35 7.19 2.73 1.29 4.07 3.88 4.22

5.12 1.76 3.37 5.71

5.27 3.08 2.79 5.69 7.52 6.82 8.75 6.70 3.98 4.82 9.75 6.81 2.82 1.02 3.90 3.60 4.01

a) Given as residues per molecule (M.W. = 15,000). destruction or incomplete hydrolysis.


6.42 9.16 6.62 4.07 4.92 9.48 7.23 2.69 1.23 3.58 2.88 3.74

Corrected for

b, Determined as cysteic acid,

shift, a slight lateral shift to the right.

It should be noted

that any one of the modified hCG-a products or intact hCG-a or hCGB alone did not fix complement with rabbit anti-hCG serum within the antigen concentration range shown in Fig. 2. DISCUSSION Judging from a compact conformation of the native hCG molecule, as inferred from a high disulfide content as well as from a few hydrodynamic properties of this hormone (lo), we had anticipated




Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

b Fig











Schematic presentation of PAGE patterns for'hCG and hCG-a after carboxypeptidase digestion in 6 M urea. Samples of 100 ug were applied to 7.5% PAGE and run at pH 8.1, 4 mA for 1.5-2 hr. a. native hCG; b. intact hCG-a; c. intact hCG-8; d. native recombinant; e. hCG-a lacking one serine (treated with CPA); f. e + c; g. hCG-a lacking one serine and one lysine (treated with CPA + CPB); h. g + c; i. hCG-a lacking four amino acids, Ser, Lys, His and Tyr (treated with CPA CPB); j . i + c; k. control hCG-ci (treated with heatinactivated CPA and CPB) + c.

a considerable high degree of resistance of the hormone to a class

of enzymes including the two carboxypeptidases used in this study which act under rather mild conditions.

Indeed, native hCG was

resistant to the action of CPA or CPA plus CPB even in the presence


of denaturant. Theoretically, carboxypeptidase digestion of isolated hCG-a should have proceeded to the half-cystine residue at position 87, releasing two tyrosyl residues. However, since only one tyrosyl was liberated, the disulfide bridge at position 87 may have imposed some steric constraint to interfere with enzymic hydrolysis of Tyr88. Of the three digestion products, only Ser(92)-less




Table I11 Biologic activity of recombinants of carboxypeptidase-treated hCG-a with intact hCG-B


Biologic activity (IU/mg)*

Intact hCG Intact hCG-a


Control hCG-aa)

Intact hCG-f3


Intact hCG-8


8,860 (6,420



Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

Ser (92)-less hCG-ab) + Intact hCG-8 Ser (92), Lys (91) -less hCGyaC) Intact hCG-8


(3,813 -

11,480 (8,142


+ Od)

Mean, followed by 95% confidence limits in parenthesis.

a) Treated with heat-inactivated CPA(1/30) and CPB(1/30) 37O for 18 hr.

b, Product of incubation with CPA(1130)

Product of incubation with CPA(1/60)



in 6 M urea at

for 18 hr. i-

CPB(1/30) for 18 hr.

No response at doses up to 20 ug modified hCG-a and 20 vg intact hCG-8.

exhibited complete recombination with intact hCG-B, although the resulting recombinant showed no in vivo biologic activity. As seen from some reduction in the biologic activity for the recombinant of control hCG-a with intact hCG-6 (Table 111), some effect due to urea was evident. Nonetheless, the loss of the activity that can be accounted for by the release of Ser-92 was more severe.


on these findings, it appears that the C-terminal part of the hCG-a molecule is important in interactions between the two subunits and that the terminal residue of serine is essential for the expression of in vivo biologic activity of the hormone.

It remains to be

resolved whether the serine residue is directly involved in these





anti - hCG serum 1/1200


Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.




z 0 0



I00 200 ANTIGEN (ng)


Fig. 2 Complement fixation curves for recombinants of carboxypeptidasetreated hCG-a with intact hCG-6. Anti-hCG serum (rabbit) was used at a concentration of 1/1200. Unfilled circles: native hCG;. unfilled triangles: control hCG-a (see Table 111) + intact hCG-B; filled circles: hCG-a treated with CPA (1/30, 18 hr) + intact hCG-6; filled triangles: hCG-a treated with CPA (1/60) plus CPB (1/30) for 18 hr + intact hCG-B.

processes or whether its role in either process is related to conformation or both.

The removal of this serine residue from the

hCG molecule resulted in about 40% reduction of the complementfixing property and the removal of an additional Lys-91 reduced it



further, suggesting that the C-terminal region of hCG-a constitutes one of the antigenic determinants of the native hormone.


further reduction in the complement fixation activity after removal of the penultimate lysyl residue may have resulted at least in part from some general conformational changes, since both vertical and lateral shifts were seen in the corresponding complement fixation curve (11).

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.

It is pertinent to compare the present results with those of recent studies with LH and its subunits, since hCG and LH share many common biologic, chemical and immunologic characteristics. Further, the sequence of seven amino acids at the C-terminal region is identical among porcine LH's.

subunits of hCG and human, bovine, ovine and

According to Cheng et al. ( 4 ) , four to five amino

acids were released from the C-terminus of isolated bovine LH-a with CPA plus CPB in the absence of urea.

The ease of carboxypeptidase

digestion of LH-cxrelative to hCG-a may be attributable to the known differences in the carbohydrate content (including sialic

acid) and a small variance in the primary structure of the a subunit between hCG and LH, since such differences signify dissimilar tertiary structures.

These authors have also noted some conformational

changes as measured by circular dichroism in the LH-cxas a result of carboxypeptidase modification.

On the other hand, the inaccessibility

of the bond between Lys-95 and Ser-96 of porcine LH-a (corresponding to Lys-91 and Ser-92 of hCG-a) to CPA and CPB, being in agreement in general with our results, has been inferred by Combarnous and Maghuin-Rogister (12).

They have also postulated that the C-




chain in native porcine LH is at the surface

of the molecule but that its carboxyl end is probably shielded. This conclusion is compatible with the present findings. REFERENCES 1.


Hum, V.G., Knipfel, J.E. and Mori, K.F., Biochemistry, 2359, 1974.


Hum, V.G., Botting, H., and Mori, K.F., Endocrine Res. Commun., 145, 1976.

Endocr Res Downloaded from informahealthcare.com by McMaster University on 12/25/14 For personal use only.


3. 4.

Mori, K.F., Hum, V.G., and Botting, H.G., Mol. & Cel. Endocrinol., 6 : 181, 1977. Cheng, K-W., Glazer, A.N., and Pierce, J.G., J. Biol. Chem., 7930, 1973.

248: -



Mori, K.F., Endocrinology,


Morgan, F.J., Canfield, R.E., Vaitukaitis, J.L., and Ross, G.T., Endocrinology, 94: 1601, 1974.


Ambler, R.P., Methods in Enzymol., 11: 155, 1967.

97, 1970.

8. Halsey, Y.D., and Neurath, H., J. Biol. Chem., 1955. 9. Gray, W.R., Methods in Enzymol.,




139, 1967.

10. Mori, K.F., and Hollands, T.R., J. Biol. Chem., 1971.


11. Levine, L., and Van Vunakis, H., Methods in Enzymol., 928, 1967.



12. Combarnous, Y., and Maghuin-Rogister, G., Eur. J. Biochem., 42: 13, 1974. -

Carboxypeptidase digestion of human chorionic gonadotropin.

415KB Sizes 0 Downloads 0 Views