Partial Characterization of Somatomedin Bioactivity in Term Human Amniotic Fluid R. MARVIN BALA AND GLEN R. SMITH Division of Endocrinology and Metabolism, Department University of Calgary, Calgary, Alberta ABSTRACT. Somatomedin (SM) activity in human amniotic fluid (AF) was partially characterized. Aliquots of pooled AF, obtained at term gestation, were acidified to pH 2.3 and sequentially studied by ultrafiltration on various membranes, Sephadex gel filtration, and starch gel electrophoresis (SGE). SM activity was measured by a rat cartilage bioassay. SM activity in AF, under these conditions, was present in several stable indicated molecular size (IMS) forms; large (>45,000) and small (near 10,000).

of Medicine, Faculty of Medicine,

The small SM was found in basic and acidic forms after SGE. These studies suggest that SM activity in term human AF is somewhat similar to SM in unextracted plasma. In addition, immunoreactive GH and insulin were measured in AF but at lower concentrations than usually found in plasma. After chromatography of AF, GH was found in large and small IMS forms, whereas most of the insulin was found in the monomer insulin IMS area. (J Clin Endocrinol Metab 43: 907, 1976)


OMATOMEDIN (SM) activity in plasma has been extensively studied, as summarized in several recent reviews (1-3). SM has been found to be present in, or produced by, tissues such as muscle (4), liver (5), and kidney (6). Cell receptors for SM have been demonstrated in various tissues including cartilage, liver and placenta (1,2). SM activity has been reported in other body fluids such as cerebrospinal fluid (7), urine (8), and in amniotic fluid (9,10). In this study, we have attempted to partially characterize SM activity in pooled human amniotic fluid (AF) obtained at normal term delivery. Our results, utilizing ultrafiltration, molecular sieve gel filtration, starch gel electrophoresis, and SM bioassay techniques suggest that SM in human AF occurs in several different molecular size and charge forms. These results are partially analogous to previous studies of SM in plasma (11-15). In addition, we have measured GH and insulin in AF fractions. Received February 17, 1976. These studies were presented in part at the Canadian Society for Clinical Investigation Meeting, Quebec, 1976. This work was supported by Medical Research Council of Canada Grant MA4107. hGH was kindly provided by the National Pituitary Agency, N.I.H.

Materials and Methods Amniotic fluid (AF) was obtained, at the time of rupture of the membranes, during delivery of normal women at full term. Only AF obtained from normal pregnancies and without obvious contamination with blood was used for these studies. The AF was centrifuged, to remove cellular contaminants, then stored at —40 C. Assays SM activity in the AF or various fractions was measured by an in vitro SM bioassay, utilizing hypophysectomized rat costal cartilage segments, as previously reported (16). All samples were assayed in quadruplicate (16) using a minimum of 3 dose levels. Unfractionated AF was assayed at l U, Vs, and Viefinalincubate dilutions. Estimations of SM activity, at each dose level of sample assayed, were considered to be valid if the mean was at least 2.5 times its SD. Samples which showed inhibitory or significantly nonparallel dose responses compared to the reference pooled normal male serum (greater than 25% difference in sample SM activity as calculated after adjustment for the different dose levels) were reassayed at different dose levels. One unit of SM activity was designated as the SM activity in 1 ml of pooled normal male reference serum. The SM activity in lyophilized fractions was expressed as units per mg of dry weight or protein. Growth hormone (GH) (2) and insulin (IRI) were measured by radioimmunoassays, as previously described for GH (17), and a standard double antibody method for IRI (18).


The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.


JCE & M • 1976 Vol 43 • No 4


Protein content was measured by the method of Lowry et al. (19). Electrolytes and urea were quantitated by a standard autoanalyzer method. Ultrafiltration (UF) A group of the frozen stored AF samples were thawed and centrifuged. The pooled supernatants were acidified with 23M formic acid, to a 1% v/v final concentration in attempts to dissociate protein aggregates and prevent aggregation during processing. After centrifugation, 37 ml aliquots of the acidified AF were added to 4 separate 400 ml UF cells (Amicon) containing XM 100, XM 50, UM 10 or UM 2 membranes. The approximate indicated molecular size (IMS) retention by these membranes is 100,000, 50,000, 10,000, and 1,000 daltons, respectively. UF was carried out at 4 C with the recommended psi positive pressure with nitrogen, until the volumes were reduced to approximately 5 ml. 100 ml of 1% formic acid was then added to the UF cells and volume again reduced to 5 ml; this process was repeated 3 times. The final retentates and filtrates were lyophilized and stored at — 20 C until assayed. Plasma, obtained from normal volunteers, who were taking no medication, was pooled and similarly processed. Sephadex gel filtration Another group of the frozen stored AF samples was thawed, pooled and centrifuged. 860 ml of the AF was then concentrated to 45 ml by UF on a UM 05 membrane (IMS retention of 500). The retentate was acidified to pH 2.3 by addition of formic acid, then chromatographed on a 10 x 100 cm column (Pharmacia) of Sephadex G75 with 1% formic acid, pH 2.3, eluent buffer, in a reverse flow manner at 4 C. This column was previously calibrated with proteins of known molecular size. The eluate was collected in small consecutive volumes, then pooled into 0.1 Kav intervals, lyophilized, and stored at —20 C. Quantitative starch gel electrophoresis (SGE) The fractions eluted in the 12,000-7700 IMS area after Sephadex gel filtration of AF were subjected to quantitative SGE. SGE was carried out at pH 8.3, using the partially discontinuous method described by Ferguson (20), with minor modifications. Starch (Connaught) was solubilized at 97 C (18 gm/100 ml) and poured into

enclosed glass plate molds. The gel slab dimensions were 0.33 x 38 x 58 cm. Human serum albumin and porcine insulin were applied as reference mobility proteins in one gel channel. SGE was carried out for approximately 4 h, at 1200 V and 100 mA, with continuous cooling of the starch gel plates at 5 C. After the electrophoresis was completed, the gel channel with the reference proteins and one gel channel with the proteins being separated were sliced off as a strip and stained. The remaining gel slab was sliced across the gel channels, parallel to the sample application slots, at standardized distances relative to the reference protein band mobility, and frozen at - 6 0 C. The gel slices were then thawed and eluted by pressure in a syringe packed with glass wool at the outlet. The eluates were centrifuged at 25,000 x g and the supernatants were lyophilized and weighed. Approximately 50% of the total applied proteins (measured by the method of Lowry) were recovered after quantitative SGE. Control starch gel eluate had no effect on the SM bioassay.

Results Amniotic fluid SM activity The mean SM activity of 26 separate normal term AF samples was 1.17 ±0.09 (SE) X the reference serum compared on the basis of SM activity per mg of protein. By extrapolation from the mean AF protein concentration, the mean SM activity of term AF was 0.07 units/ml. Ultrafiltration of AF and plasma The dry weight (lyophilized) and the protein content was 8.3 and 4.3 mg/ml for the pooled AF, and 86.5 and 76 mg/ml for the plasma, respectively. After ultrafiltration on the various membranes (Fig. 1), the mean overall recoveries of dry weight and SM activity were 98% (range 97-99) and 86% (range 69-98) for AF; and 100% (range 99-102) and 94% (range 67-115) for plasma, respectively. As shown in Fig. 1, of the total recovered SM activity in AF and plasma, approximately 60 and 30% had an IMS smaller than 50,000, whereas only 12 and 25% respectively, had an IMS less than 10,000 despite the acidic

The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.




a 20£ 40 u

fc!6 0 80-



XM 50

UM 10

UM 2



FIG. 1. Ultrafiltration of similar aliquots of acidified term amniotic fluid (AF) or normal pooled plasma (P) on various Diaflo (Amicon) membranes. The XM 100, XM 50, UM 10 or UM 2 membranes have designated indicated molecular size retention of 100,000, 50,000, 10,000 or 1000 daltons, respectively. The amounts of SM activity recovered in the retentates and filtrates are shown as a percentage of the total recovered SM. The mean SM potency ± SE is shown in units/mg dry weight.

conditions used. The increase in SM potency (SM units/mg dry weight) of the UM 2 retentates compared to the starting materials could be accounted for by salt removal, since the SM activity was calculated on the basis of dry weight of samples. The differences in the percentage of total recovered dry weight and SM retained by the various UF membranes are compared in Fig. 2. 48% of the total recovered SM activity in AF had an IMS between 50,000 and 10,000. Whereas 70% of the SM activity in plasma had an IMS greater than 50,000, with another peak of activity in the 10,0001000 IMS area. Compared to plasma, AF had a relatively greater proportion of SM activity with an IMS less than 1000. Sephadex gel filtration After gel filtration of the UM 05 retentate of AF on Sephadex G75 under acidic conditions (Fig. 3), the overall recoveries of dry weight and SM activity were 93 and 94%,

respectively. Approximately one-third of the total recovered dry weight and SM activity was eluted in the Kav interval less than 0.15 with an IMS greater than 45,000 daltons. 9% of the total recovered dry weight and 45% of the SM activity was eluted in the 0.550.75 Kav intervals with an IMS between 12,000-5500. The mean SM potency of these fractions was 0.39 units/mg dry weight compared to 0.09 units/mg for the UM 05 retentate applied. The SM activity in the Kav intervals on each side of this peak may indicate zone spreading with gel filtration. Immunoreactive GH (IR-GH) was detected in all lyophilized fractions with Kav intervals less than 0.85. More than one half of the total recovered IR-GH was eluted in the Kav intervals corresponding to an IMS area greater than 22,500, however, the 22,500-16,500 IMS area fraction showed the maximal IR-GH potency (2.0 ng IR-GH/mg dry weight). 75% of the total recovered IRI was eluted in the 5500-4000 IMS area fraction (0.7 ng IRI/mg dry weight).


K >

>. 32. o

id (K

,c. ID




O I-










FIG. 2. Per cent of the total recovered dry weight and SM, in various indicated molecular size areas, after ultrafiltration of amniotic fluid (AF) or plasma (P) on various membranes. The results are based on calculated differences in the per cent of the total recovered weight or SM in the retentates or filtrates on various membranes.

The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.

JCE & M • 1976 Vo) 43 • No 4




FiG. 3. Gel filtration of the UM 05 retentate of pooled amniotic fluid on Sephadex G75 with 1% formic acid, pH 2.3, buffer eluent.











Starch gel electrophoresis (SGE)

Molecular size of SM

When the 12,000-7700 IMS area proteins eluted from Sephadex were fractionated by Quantitative SGE (Fig. 4), 39% of both protein and SM was recovered. 81% of the recovered protein and 55% of the SM was negatively charged and migrated toward the anode, however, the proteins with the greatest SM potency migrated toward the cathode.

After UF of plasma approximately 70% of the total SM activity had an IMS greater than 50,000 despite the acidic conditions used. These results are similar to previous findings after gel filtration of plasma (11-15) and the UF studies of SM activity in liver perfusates (5). Whereas, after UF of bovine plasma, Liberti (22) noted that 80% of the SM activity appeared to have an IMS smaller than 50,000. Estimation of molecular size by ultrafiltration is not precise; in particular, the XM 100 membrane appears to retain albumin as similarly noted by Van Wyk et al. (12). The percentages of the total recovered dry weight in the corresponding IMS areas (other than in the IMS area greater than 100,000) after UF (Fig. 2) or Sephadex gel filtration (Fig. 3) are comparable when the prior removal of the salt from the AF before gelfiltrationis taken into account. In contrast to plasma, only 35-40% of the total recovered SM activity in AF appeared to have an IMS greater than 45,000-50,000. Comparison of Figs. 2 and 3 would suggest that most of the SM activity, with an IMS between 10,000 and 50,000 on UF, had an IMS close to 10,000. This would suggest that a relatively greater proportion of the total SM activity in AF (45%) has an IMS near 10,000 than that found in plasma (25%). Non-

Discussion SM activity in AF SM activity in normal term AF was less than one-tenth times normal reference plasma in terms of SM activity per volume. However, the SM activity in AF is equal to or greater than plasma if expressed in terms of SM activity per mg of protein. Fetal urine may be a source of SM in AF, even though in previous studies we were unable to detect SM activity in highly concentrated fractions of adult urine (unpublished observations related to the same fractions used for IR-hGH studies (21)). The possible role of the placenta in SM production requires investigation. The relationship of SM activity in AF to fetal growth and development has only been preliminarily explored to date (9,10).

The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.



FlG. 4. Quantitative starch gel electrophoresis (SGE) of the 12,000-7700 indicated molecular size area proteins after Sephadex gel filtration (Fig. 3). SGE was carried out at pH 8.3 using a partially discontinuous buffer system. Human serum albumin and porcine insulin are diagrammatically shown as anodal migrating protein bands.



suppressible insulin-like activity-soluble (NSILA-s) and SM-C purified from plasma extracts appear to be similar (1-3). Chochinov et al. (23) have preliminarily reported the presence of significant amounts of NSILA-s, in AF, as measured by a placental cell membrane radioreceptor assay for insulin (RRI). Using approximately similar gel filtration conditions to those reported here they found that the majority of the RRI activity was eluted with an IMS of 10,000 or less, with a small amount in the larger IMS area. Different molecular forms of SM These studies suggest that SM in AF occurs in a large (>45,000) and small (approximately 10,000) IMS forms which are stable in acidic conditions. In addition, our findings after SGE suggest that the SM activity found in the AF fractions with an IMS near 10,000 is composed of several groups of proteins of similar molecular size but differing charge. These findings are similar to our previous studies of SM in plasma (15). Our studies are only in partial agreement with Chochinov et al. (23) who found that the RRI activity in AF, in the small IMS, was mainly acidic in nature. It is possible that the basic SM in AF is similar to SM-C purified from plasma by Van Wyk et al. (2). However, it is unlikely that the non-basic SM is similar to the neutral SM-A




^ 8 127 16 albumin insulin

2 0 cm from origin

purified by Hall (3,24), in view of the anodal mobility compared to porcine insulin. The acidic forms of SM in AF were assayed by the rat cartilage assay and therefore cannot be similar to SM-B as purified by Uthne (3,25). Other hormones in AF Similar to Chochinov et al (23), we have found the majority of IRI in a monomer form in AF. However, by extrapolation to unconcentrated AF, we have found lower levels of IRI in AF (0.04 ng/ml). Our findings of IR-GH in AF in the various IMS size forms is somewhat similar to our previous studies of IR-GH in plasma (11,26), even though the amount of IR-GH in AF was much less. A proportion of the IR-GH measured may be due to cross-reaction with the relatively large amounts of human prolactin in AF (27). Further studies of SM, as well as other hormones, in AF are required. Acknowledgments The authors thank A. Bardai, M. Scace, and C. Wright for their technical assistance, and J. Gayford for her secretarial assistance.

References 1. Van Wyk, J. J., L. E. Underwood, R. L. Hintz, D. R. Clemmons, S. J. Voina, and R. P. Weaver, The somatomedins: A family of insulinlike hormones under growth hormone control, Recent Prog Hormone Res 30: 259, 1974.

The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.



2. Van Wyk, J. J., and L. E. Underwood, Relation between growth hormone and somatomedin, Ann Rev Med 26: 427, 1975. 3. Luft, R., and K. Hall (eds.), Advances in Metabolic Disorders, vol. 8, Academic Press, New York, 1975. 4. Hall, K., A. Holmgren, and U. Lindahl, Purification of a sulphation factor from skeletal muscle of rat, Biochim Biophys Ada 201: 398, 1970. 5. Dehnel, J. M., P. D. McConaghey, and M. J. O. Francis, Fractionation of liver somatomedin activity by ultrafiltration, J Endocrinol 62: 355, 1974. 6. McConaghey, P. D., and J. Dehnel, Preliminary studies of 'sulphation factor' production by rat kidney, J Endocrinol 52: 587, 1972. 7. Beaton, G. R., J. Sagel, and L. A. Distiller, Somatomedin activity in cerebrospinal fluid, J Clin Endocrinol Metab 40: 736, 1975. 8. Yalow, R. S., K. Hall, and R. Luft, Immunoreactive somatomedin B in urine,/ Clin Endocrinol Metab 41: 638, 1975. 9. Andersen, H. J., K. W. Kastrup, and P. E. Lebech, The possible role of somatomedin in the growth of the human fetus, Ada Paediatr Scand 63: 328, 1974 (Abstract). 10. Wright, C , and R. M. Bala, Somatomedin levels in amniotic fluid, maternal and cord plasma Clin Res 23: 620A, 1975 (Abstract). 11. Bala, R. M., K. A. Ferguson, and J. C. Beck, Plasma biological and immunoreactive human growth hormone-like activity, Endocrinology 87: 506, 1970. 12. Van Wyk, J. J., K. Hall, J. L. Van den Brande, and R. P. Weaver, Further purification and characterization of sulfation factor and thymidine factor from acromegalic plasma, J Clin Endocrinol Metab 32: 389, 1971. 13. Bala, R. M., Human growth hormone and somatomedin, In Lederis, K., and K. E. Cooper (eds.), Recent Studies of Hypothalamic Function, S. Karger, Basel 1974, p. 114. 14. BaJa, R. M., and J. C. Beck, Fractionation studies on plasma of normals and patients with laron dwarfism and hypopituitary gigantism, Can J Physiol Pharmacol 51: 845, 1973. 15. Bala, R. M., G. R. Smith, E. D. Anderson, and D.

JCE & M • 1976 Vol 43 • No 4

Birnie, Heterogeneity of somatomedin in unextracted serum, Clin Res 22: 728A, 1974 (Abstract). 16. Bala, R. M., C. Hankins, and G. R. Smith, A somatomedin assay using normal rabbit cartilage in clinical studies, CanJ Physiol Pharmacol 53: 403, 1975. 17. Bala, R. M., K. A. Ferguson, and J. C. Beck, Modified solid-phase (tube) radioimmunoassay of human growth hormone, CanJ Physiol Pharmacol 47: 803, 1969. 18. .Morgan, C. R., and A. Lazarow, Immunoassay of insulin: Two antibody system. Plasma insulin levels of normal, subdiabetic, and diabetic rats, Diabetes 12: 115, 1963. 19. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, Protein measurement with the folin phenol reagent,/ Biol Chem 193: 265, 1951. 20. Ferguson, K. A., and A. L. C. Wallace, The characterization of pituitary hormones by starch gel electrophoresis, Recent Prog Hormone Res 19: 1, 1963. 21. Bala, R. M., and J. C. Beck, Human growth hormone in urine, J Clin Endocrinol Metab 33: 799, 1971. 22. Liberti, J. P., Partial purification of bovine sulfation factor, Biochem Biophys Res Commun 39: 356, 1970. 23. Chochinov, R. H., I. K. Mariz, and W. H. Daughaday, Characterization of the receptor reactive insulin (RRI), immunoreactive insulin (IRI) and immunoreactive glucagon (IRG) content of amniotic fluid, Abstract No. 129, 57th Annual Meeting of the Endocrine Society, New York, 1975. 24. Hall, K., Human somatomedin: Determination, occurrence, biological activity and purification, Ada Endocrinol [Suppl] (Kbh) 163 70: 1, 1972. 25. Uthne, K., Human somatomedins purification and some studies on their biological actions, Ada Endocrinol [Suppl] (Kbh) 175, 73: 1, 1973. 26. Bala, R. M., K. A. Ferguson, and J. C. Beck, Heterogeneity of human growth hormone, In Raiti, S. (ed.) Advances in Human Growth Hormone Research, N.I.H. Symposium, Baltimore, 1973, p. 494. 27. Ben-David, M., and A. Chrambach, Isolation of isohormones of human prolactin from amniotic fluid, Endocrine Res Commun 1 (2): 193, 1974.

The Endocrine Society. Downloaded from by [${individualUser.displayName}] on 13 November 2015. at 17:21 For personal use only. No other uses without permission. . All rights reserved.

Partial characterization of somatomedin bioactivity in term human amniotic fluid.

Partial Characterization of Somatomedin Bioactivity in Term Human Amniotic Fluid R. MARVIN BALA AND GLEN R. SMITH Division of Endocrinology and Metabo...
642KB Sizes 0 Downloads 0 Views