(1961) for a purer preparation of the hormone, assaying IU/mg. explain the discrepancy.

Transcription

1 Department of Experimental Biochemistry, The London Hospital Medical College, London, E. 1., England ON THE MOLECULAR WEIGHT OF PREGNANT MARE'S SERUM GONADOTROPHIN By C. J. O. R. Morris ABSTRACT The chromatographic behaviour of pregnant mare serum and of a purified preparation of the gonadotrophin from uterine endometrial cup secretion on the cross-linked dextran gel media Sephadex G-100 and G-200 has been examined, and is consistent with a molecular weight greater than the value of usually accepted. The sedimentation coefficient of an active preparation has been determined by sedimentation in a sucrose density gradient and serial bioassay. The value of 3.2 S for SW20 is consistent with current values. The literature data are shown to be most consistent with a molecular weight of for the hormone, and the mechanism of its chromatography on dextran gels is discussed. The molecular weight of a purified preparation of pregnant mare's serum gonadotrophin (PMSG) assaying IU/mg was determined by Bourrillon 8c Got (1959) from measurements of the sedimentation coefficient (3.7 S), the diffusion constant (10.2) and the partial specific volume (0.700) of a solution in water. They calculated the molecular weight as They later (Bourril lon 8c Got 1960), determined the sedimentation and diffusion constants of the same preparation in 0.1 m NaCl solution, and found a considerably lower value (4.2) for the diffusion constant, giving a molecular weight of A sedimentation coefficient of 3.4 S was found by Legault-Demare et al. (1961) for a purer preparation of the hormone, assaying IU/mg. Studies in this laboratory on the purification of PMSG from serum and from uterine endometrial cup secretion by molecular sieve chromatography on cross-linked dextran gels (Sephadex) showed that the hormone behaved as if it had a considerably higher molecular weight. This study was undertaken to explain the discrepancy.

2 - - EXPERIMENTAL Materials gift from Professor Lyophilised pregnant mare's serum assaying 4-5 IU/mg was a A. S. Parkes, Cambridge. Uterine endometrial cup secretion was obtained from preg nant mares at slaughter by Dr. R. V. Short of the School of Agriculture, Cambridge, and transported without delay to this laboratory. The lyophilised secretion had a gonadotrophic activity of IU/mg. Preparation G5D2 was prepared directly from lyophilised secretion by chromatography on diethylaminoethyl cellulose at />jj 4.6. It had a biological activity of IU/mg. Preparation G5DEB was pre pared from G5D2 by chromatography on ECTEOLA cellulose at pu 4.0. It had a gonadotrophic activity of IU/mg. The Chromatographie purification proce dures will be described in detail in a future communication. Methods 1. Chromatography on Sephadex G-100 and G-200 These materials (from Pharmacia A. B., Uppsala, Sweden) were allowed to swell in 0.5 M NaCl for at least 24 h before use to ensure complete equilibration. Columns, usually X 1 cm, were packed by sedimentation and washed with 0.5 M NaCl to constant height and until the effluent had a negligible ultra-violet absorption at 280 m,m. Development of the chromatogram was usually carried out with 0.5 M NaCl at flow rates of 3 5 ml/h, fractions of ml being collected automatically. The protein content of the fractions was determined by measurement of the ultra violet absorption at 280 µ. In some cases the 254 ß absorption of the column effluent was monitored continuously using an LKB Uvicord absorptiometer. 2. Density gradient sedimentation This was carried out by the method of Martin Se Ames (1961), in the SW-39 rotor of the Spinco Model L preparative ultracentrifuge. A linear gradient of 5 to 20 % (w/v) sucrose in 0.1 M NaCl was prepared by means of the apparatus of Martin 8c Ames (1961) and stored at 4 for 18 h before use. The sample (1-2 mg) dissolved in 0.1 M NaCl (0.1 ml) was carefully layered on to the gradient immediately before centrifugation, and the tubes loaded into the precooled rotor. Usually bovine serum albumin and bovine /-globulin (Armour) were used as reference proteins in the other two tubes of the rotor. Typical operating conditions were 15 h at r. p. m. at 4. At the end of the sedimentation period the rotor was allowed to coast to rest without braking, and the tubes successively placed in the fractionator of Martin Se Ames (1961). This allows the bottom of the cellulose acetate tube to be pierced with a needle without loss so that the contents can be collected in fractions from the bottom to the top of the tube. Eighteen 0.25 ml fracions were collected by means of a micro peristaltic pump and fraction collector as described by Morris (1960). The protein content of the individual fractions was determined by measurement of the absorption at 280 µ, while appropriate fractions were bulked for bioassay. Calculation of the sedimentation coefficient of the unknown protein is carried out by comparison of the sedimentation rate of its zone maximum with those of the two reference proteins in the other tubes of the rotor. The distance of migration from the meniscus are then directly proportional to the respective Sw2o values. The S value

3 - under the conditions of sedimentation may also be obtained from the absolute distance of migration by the method of Martin Se Ames (1961). 3. Biological assay Two methods for assessing the gonadotrophic potency of test materials were used. Absolute potencies were determined by augmentation of the ovarian weight in im mature rats. The animals received a single 1 ml injection and were autopsied 48 h later. A 2 X 2 assay design was used with two levels of the International Standard Preparation. The precision of the assay was greatly improved by distribution of litter mates within the groups. The biological activity of multiple fractions was followed by a rapid quantitative ovarian hyperaemia method developed in this laboratory (Dedman Se Morris, un published). This makes use of the facilitation of the hyperaemia response by hyaluronidase described by Serment Se Gnauli (1955). Female rats of the Wistar strain, 28 days old, weighing g were injected with the test material in a volume of 1 2 ml, together with 1500 IU of a commercial hyaluronidase preparation (Hyalase, Evans). The animals were killed after 6 h, and the degree of ovarian hyperaemia assessed on a graded scale. Standards prepared from the International Standard PMSG Preparation were included for comparison. The threshold for response for PMSG on our strain of rats lies at about 3-5 IU. RESULTS 1. Molecular sieve chromatography of pregnant mare's serum Fig. 1 A shows the Chromatographie behaviour of lyophilised pregnant mare's serum (40 mg in 1.0 ml 0.5 m NaCl) on a 100 X 1 cm column of Sephadex G-100. The biological activity (shaded area) is confined to peak I, and is almost excluded from the dextran gel particles. Peak II corresponds to the equine serum albumin zone. Fig. 1 illustrates the chromatography of the same material (35 mg in 0.5 ml 0.5 m NaCl) on a 95 X 1 cm column of Sephadex G-200. The figure shows the three zones found with most mammalian sera on Sephadex G-200 (Flodin 1962; Flodin 8c Killander 1962). Peak I contains the 19 S macroglobulins, peak II mainly the 7 S /-globulins, while peak III contains the serum albumin and some low molecular «-globulins. It will be seen that the gonadotrophic activityis confined to the 6-7 S region, and as in the case of G-100, is faster moving than the serum albumin. 2. Molecular sieve chromatography of purified PMSG preparations Fig. 2 A shows the Chromatographie behaviour of the purified endometrial secretion preparation G5D2 on Sephadex G-100. Three zones are partially resolved, peak III corresponding to the serum albumin (4-5 S) region. As in the case of serum (Fig. 1 A) the biological activity (shaded) is almost ex cluded from the gel and is present in both peaks I and II. Fig. 2 shows the complex Chromatographie pattern given by the same

4 40 50 Fig. 1 A. Chromatography of pregnant mare's serum on Sephadex G X 1 cm column, flow rate 5 ml/h, 1.1 ml fractions. Eluent 0.5 M NaCl Fig. 1 B. Chromatography of pregnant mare's serum on Sephadex G X 1 cm column, flow rate 5 ml/h, 1.1 ml fractions. Eluent 0.5 M NaCl.

5 Fig. 2 A. Chromatography of preparation G5D2 on Sephadex G X 0.9 cm column, flow rate 3 ml/h, 1.3 ml fractions. Eluent 0.5 m NaCl Fig. 2 B. Chromatography of preparation G5D2 on Sephadex G X 0.9 cm column, flow rate 4.6 ml/h».1.3 ml fractions. Eluent 0.5 M NaCl.

6 IO 20 Fig. 3. Chromatography of preparation G5D2S on Sephadex G X 1 cm column, flow rate 3 ml/h, 1.6 ml fractions. Eluent water. 30 preparation G5D2 on Sephadex G-200. Four distinct zones are evident, the gonadotrophic activity (shaded) being confined to zone III. Owing to the good recovery of biological activity, the method is very suitable for preparative purposes, a threefold enhancement of specific activity being readily obtained. Since Bourrillon 8c Got (1960) have suggested that the molecular weight of PMSG is lower in water than in 0.1 m NaCl, a fraction prepared from peak III of Fig. 2 was subjected to chromatography in water on a Sephadex G-100 column prepared in water, with the result shown in Fig. 3. The figure shows that although resolution into two zones did occur, the gonadotrophic activity was confined to the excluded zone I, as in the ex periment shown in Fig. 2 A, so that no change in the Chromatographie be haviour of the hormone was produced by the change to water as developing medium. 3. Density gradient sedimentation of a purified PMSG preparation Fig. 4 (full line) shows the protein concentration profile of preparation G5DEB after 90 min sedimentation at r.p.m. at 4. The direction of sedi at fraction 18. The broken mentation is from right to left, the meniscus being line shows the corresponding profile of bovine serum albumin centrifuged

7 - _? U 6 8 IO / Fig. 4. Density gradient sedimentation of hormone preparation G5DEB in a 5 gradient. 900 min at r. p. m. Temperature ml fractions. -Preparation G5DEB. -Crystalline bovine serum albumin. 20 /o sucrose simultaneously in another tube in the SW-39 rotor. The region of gonadotro phic activity coincided well with the maximum protein concentration. The sedimentation coefficient determined from the absolute rate of migration of the zone of biological activity was 3.0 S, while the value of SW2o for PMSG determined from the rate of sedimentation relative to the rates of serum albumin and y-globulin under the same conditions was 3.2 S. DISCUSSION The value of the sedimentation coefficient, S determined from the density gradient sedimentation is in reasonable agreement with the values of Legault-Demare et al. (1961) and of Bourillon 8c Got (1959; 1960). The ex periment also demonstrates unequivocally that PMSG activity sediments slower than serum albumin.

8 On the other hand the molecular sieve chromatography experiments de monstrate equally conclusively that the gonadotrophic hormone both from serum and from endometrial cup secretion travels faster through the column than serum albumin, i. e. behaves as if it were a larger molecule. Part of this anomaly can be explained by a reconsideration of the results of Bourillon 8c Got (1959; 1960). Their 1959 determinations of the sedimenta tion and diffusion coefficients were carried out in water solution, a procedure which is known to give invalid results for non-isoelectric proteins owing to the presence of a electrostatic charge effect. These effects have been discussed for diffusion by Gosting (1956) and for ultracentrifugation by Schachman (1959). In brief the condition of electro-neutrality and the presence of counter ions in a non-isoelectric protein results in an anomalously high diffusion co efficient and an anomalously low sedimentation coefficient in the absence of added electrolytes, and it is generally accepted practice to overcome these charge effects by the addition of electrolytes to m concentration. PMSG as an acidic protein with an isoelectric point of 1.8 would be expected to be particularly subject to charge effects in the absence of added electro lytes. In fact Bourillon 8c Got (1959) found the remarkably high value of 10.2 for the diffusion coefficient of PMSG in water. This value must regarded as erroneous, and their 1960 value of 4.2, measured in solution in 0.1 m NaCl as correct. The molecular sieve chromatography experiment shown in Fig. 3 pro vides additional evidence that a PMSG preparation made by chromatography in 0.5 m NaCl does not change its molecular properties on solution in water. The value of 4.2 for the diffusion coefficient combined with the value of 3.7 S for the sedimentation coefficient in 0.1 m NaCl leads to a molecular weight of for PMSG, almost identical with that of serum albumin. Although the discrepancy between Chromatographie and molecular kinetic data is greatly reduced by these considerations, it still remains, since serum albumin enters the gel granules more readily than the hormone, even though they are of similar molecular size and shape. This remaining discrepancy can probably be ascribed to the large negative charge carried by the PMSG mole cule over most of the pu range. This interacts with the few ionised carboxyl groups present in the dextran gel particles, producing an electrostatic repul sion which results in a partial ion exclusion mechanism hindering entry of the hormone molecule into the gel particles and accelerating its passage through the column (see Flodin 1962 for a discussion of this effect). ACKNOWLEDGEMENTS The author is indebted to Miss M. L. Dedman for carrying out bioassays, and to Professor A. S. Parkes and to Dr. R. V. Short for gifts of materials.

9 REFERENCES Bourrillon R. Se Got R.: Acta endocr. (Kbh.) 31 (1959) 559. Bourrillon R. Se Got R.: Acta endocr. (Kbh.) 35 (I960) 221. Flodin P. In: Dextran gels and their applications in gel filtration. Pharmacia, Uppsala (1962). Flodin P. 8c Killander J.: Biochim. biophys. Acta 63 (1962) 403. Gosting L. J.: Advanc. Protein Chem. 11 (1956) 429. Legault-Demare J., Clauser H. Se Jutisz M.: Bull. Soc. Chim. biol. (Paris) 43 (1961) 897. Martin R. G. Se Ames B. N.: J. biol. Chem. 236 (1961) Morris C. J. O. R.: Z. physiol. Chem. 321 (1960) 87. Schachman H. K. In: Ultracentrifugation in Biochemistry, Academic Press, New York (1959) 225. Serment H. 8c Giranti.: Compt. rend. Soc. Biol. 149 (1955) 767.