serum levels of mullerian inhibiting substance in boys from birth to 18 years, as determined by...

0021-972X/90/7001-0011$02.00/0Journal of Clinical Endocrinology and MetabolismCopyright© 1990 by The Endocrine Society

Vol. 70, No. 1Printed in U.S.A.

Serum Levels of Mullerian Inhibiting Substance in Boysfrom Birth to 18 Years, as Determined by EnzymeImmunoassay*MARILYN L. BAKER, SYLVIA A. METCALFE, AND JOHN M. HUTSON

Surgical Research Laboratory, Royal Children's Hospital Research Foundation, Parkville, Victoria; and theDepartment of Paediatrics, University of Melbourne, Melbourne, Victoria, Australia

ABSTRACT. An enzyme immunoassay was set up with theaim of determining the serum levels of Mullerian inhibitingsubstance (MIS) during childhood. A monoclonal antibodyagainst purified bovine MIS was combined with a polyclonalantibody against recombinant human MIS to make a sandwichassay. This assay detected MIS in human serum within thefollowing criteria. Ninety-eight boys, aged between birth and 18yr, who had been admitted to the Royal Children's Hospital,were included. MIS levels were measured in samples taken for

biochemical screening of unrelated disorders. MIS was detectedin the serum up to 16 yr of age, but was low beyond 12 yr andundetectable at 18 yr. High MIS levels were found at 4-12months, consistent with MIS having an important function atthis time. Germ cells undergo an important transformation fromgonocytes to spermatogonia at the same time as the MIS levelspeak, suggesting a possible function for MIS. {J Clin EndocrinolMetab 70: 11,1990)

MULLERIAN inhibiting substance (MIS; anti-Mul-lerian hormone), produced by the Sertoli cells of

the testis, causes regression of the Mullerian ducts in themale fetus during sexual differentiation (1). AlthoughMIS was once thought to be a fetal hormone only (2),MIS production does not cease after this period and hasbeen measured by bioassay in adult roosters (3) and incalves for 2 months (4) and humans for about 2 yrpostnatally (5). RIAs have detected MIS in sera fromneonatal calves (6, 7) and in rete testis fluid from adultbulls (6). Immunohistochemical study of human testic-ular biopsies detected MIS up to 10 yr of age (8). MIS isalso produced by ovarian granulosa cells and has beenmeasured in bovine follicular fluid by RIA (7).

Significant secretion of MIS postnatally long afterregression of the fetal Mullerian ducts suggested thatthis newly described hormone may have other functions.An immunoassay was set up to quantify serum MISconcentrations in the human male and correlate theselevels with possible functions of MIS. Monoclonal anti-bodies to purified bovine MIS (bMIS) were produced andwere found to cross-react with recombinant human MIS(rhMIS; a kind gift from P. K. Donahoe, Massachusetts

Received March 14, 1989.Address all correspondence and requests for reprints to: Sylvia A

Metcalfe, PhD., Department of Surgery, Royal Children's Hospital,Flemington Road, Parkville, Victoria 3052, Australia.

* This work was supported by the National Health and MedicalResearch Council of Australia.

General Hospital, Boston, MA). One of these antibodieswas used to develop a sandwich enzyme immunoassay(ElA) together with polyclonal antibodies to rhMIS.Using this EIA the serum levels of MIS were measuredin males from birth to 18 yr of age.

Materials and Methods

Experimental subjects

Sera from patients at the Royal Children's Hospital, Mel-bourne, were collected from the Biochemistry Department.These samples had been collected for routine biochemical anal-yses and stored for 1 week at -20 C, after which time they arenormally discarded. Sera were used for MIS determination atthe end of the week's storage. The clinical diagnosis and/orreason for blood collection were examined to determine anypossible correlations with MIS levels.

Experimental animals

BALB/c mice and New Zealand White rabbits were immu-nized as described in Materials and Methods and in accordancewith NIH guidelines for the care and use of laboratory animals.

MIS purification

Bovine MIS was purified from newborn calf testes (Metcalfe,S. A., Baker, M.L., Watts, L.M., Hutson, J.M., unpublisheddata). Briefly, minced testes were incubated in Ham's F-10medium (Flow, Melbourne, Australia) containing 20 mMHEPES and 10"5 M phenylmethylsulfonylfluoride. After centri-

11

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 December 2014. at 11:44 For personal use only. No other uses without permission. . All rights reserved.

12 BAKER, METCALFE, AND HUTSON JCE & M • 1990Vol70«Nol

figation the supernatant was precipitated with (NH4)2SO4 (30-45%), and the resulting precipitate was purified further bywheat germ lectin chromatography (Sepharose, Pharmacia,Uppsala, Sweden), followed by hydrophobic interaction chro-matography (phenyl-Superose, Pharmacia). The purified MISfraction was tested for biological activity in a mouse bioassay.This has been developed in our laboratory and is based on therat bioassay previously described (9); the only difference is thatfetal mice are used, at the appropriate time of Mullerian ductregression (10), instead of fetal rats.

Antibodies

Monoclonal antibodies were produced against purified bMIS.Eight-week-old male BALB/c mice were immunized sc with 20/ng bMIS in 0.3 mL of a 1:1 mixture of phosphate-bufferedsaline (PBS; 0.01 M phosphate buffer and 0.15 M NaCl, pH 7.4)and Complete Freund's Adjuvant. After 9 weeks the animalswere boosted with 20 ng bMIS, ip, in 0.3 mL of a 1:1 mixtureof PBS and Incomplete Freund's Adjuvant and again at 24 and27 weeks with 20 and 10 ng bMIS, respectively (ip, diluted in0.2 mL saline); the last immunization was 3 days before death.

Spleen cells from the immunized mice were fused with themyeloma cell line SP2/0-Ag 14 (Murdoch Institute, Melbourne,Australia) as previously described (11). Fused cells were dis-tributed into 96-well (0.2-mL) Linbro plates (Flow) which hadbeen plated with feeder cells (normal mouse spleen cells pre-pared as for the spleen cells from the immunized mouse) 3 daysbefore fusion. Hybrids were screened for antibody productionusing a solid phase enzyme immunoassay. Flat bottomed 96-well Nunc Immunoplates (Nunc, Copenhagen, Denmark) werecoated with purified bMIS in carbonate buffer (50 mM; pH 9.6)overnight at 4 C. After washing (3 times) with PBS containing0.05% Tween-20 (PBST), hybridoma culture media were addedand incubated for 1 h at 37 C. A second wash, as describedabove, was followed by incubation with horseradish peroxidase-conjugated sheep antimouse immunoglobulins (Ig; Silenus,Melbourne, Australia) for 2 h at room temperature (RT). Aftera final wash with PBST, tetramethyl benzidine was used as theenzyme substrate (12). The color reaction was stopped with 2MH2SO4, and the color intensity was recorded at 450 nm on aTitertek Multiscan MCC/340 (Flow). The Ig isotype of positivehybridomas was determined using hybridoma cutlure mediumin a Misotest kit (Commonwealth Serum Laboratories, Mel-bourne, Australia).

Antibody production by positive hybridomas was amplifiedby ascites production in BALB/c mice as described previously(11). The IgG fraction of the ascitic fluid was purified by affinitychromatography on protein-A-Sepharose (Pharmacia) (13). As-citic fluid was bound to the column in 0.1 M phosphate buffer,pH 8.9, containing 0.15 M NaCl and the IgG fraction elutedwith citrate buffer (0.1 M; pH 5.5). After dialysis against PBS,the IgG fraction was concentrated, and the final concentrationwas determined by its molar absorption coefficient.

Polyclonal antibodies to rhMIS were raised in two femaleNew Zealand White rabbits. Fifty micrograms of rhMIS wasinjected sc at 7 weeks of age in 2 mL of a 1:1 mixture of rhMISin saline and Complete Freund's Adjuvant. A further 20 jugwere given sc at 10 weeks in a 1:1 mixture of rhMIS in saline

and Incomplete Freund's Adjuvant and again at 23 weeks sc in2 mL saline.

The IgG fraction of rabbit antiserum to rhMIS was purifiedon protein-A-Sepharose as described above, except serum wasbound to the column in 0.1 M phosphate buffer, pH 8.0, and0.15 M NaCl, and the IgG fraction was eluted with citrate buffer(0.1 M; pH 3.5).

EIA

Flat-bottomed 96-well Nunc Immunoplates (Nunc) wereused as a solid support. Wells were coated with the IgG fractionof hybridoma 22F61 murine ascites (2.5 Mg/mL in 50 mMcarbonate buffer, pH 9.6) overnight at 4 C. After washing withPBST (three times), standard and samples were incubated for4 h at 37 C. Both standard and samples were diluted in amixture of adult female sera and PBST, so that the finalconcentration of serum was 50% in all cases. rhMIS was usedto construct the standard curve. After a second wash withPBST (three times), the IgG fraction of the rabbit antisera torhMIS diluted in PBST (5 fig/mh) was added and incubatedovernight at RT. A third wash with PBST (three times) wasfollowed by incubation with horseradish peroxidase-conjugatedswine antirabbit Ig (Dakopatts, Glostrup, Denmark) for 1 h atRT. After a final wash with PBST, tetramethyl benzidinesubstrate was added, and the color absorbance was recorded asdescribed above. A standard curve was constructed for eachassay using serial dilutions of rhMIS starting at 400 Mg/L. Eachsample was measured in duplicate at one or more dilutionsdepending on the amount of sample.

The specificity of the assay was tested using the followinghormones: inhibin and activin (Dr. D. Robertson, MonashUniversity, Melbourne, Australia), GH and insulin-like growthfactor-I (Dr. G. Warne, Royal Children's Hospital), and FSHand LH (Department of Biochemistry, Royal Children's Hos-pital). Each hormone was serially diluted in 50% female serumto give the following concentration ranges: inhibin, 3.9-2000Mg/L; activin, 3.9-2000 ^g/L; GH, 1.9-1000 /ug/L; insulin-likegrowth factor-I, 1.9-1000 Mg/L, FSH, 0.19-100 IU/L; LH, 0.98-500 IU/L. These concentrations were selected to include atleast the normal physiological values found in serum (14, 15).

Statistics

The standard curve for the EIA was calculated by linearregression analysis, and slope and midrange concentrationswere computed. Analysis of variance statistics were performedto compare the MIS value at birth with that at various ages.

Results

Antibodies

Two monoclonal antibodies were successfully pro-duced from one fusion; one was an IgM isotype (26H121),and the other was an I g d (22F61). When tested forcross-reactivity with rhMIS in the solid phase EIA (seeMaterials and Methods), both monoclonals were positive.This was unexpected, as other groups working with bMIS


SERUM MIS LEVELS 13

had not been able to produce monoclonal antibodies thatrecognized MIS in serum (16). 22F61 hybridoma culturemedium also detected rhMIS by Western blotting tech-niques (data not shown). The IgM proved to be difficultto purify by various techniques, whereas the IgGx couldbe purified on protein-A-Sepharose and so was chosenfor development of the EIA. We wanted to develop asandwich assay, as this would be more specific, and weused polyclonal antiserum as the second antibody, sincethis would recognize several epitopes on MIS.

EIA

The characteristics of the standard curve are shown inFig. 1. Five dilutions, starting from 100 Mg/L were fit toa straight line with correlation coefficients (from nineassays) ranging from 0.993-1.0. The mean of the slopeswas 0.719, with a SD of 0.096. The lowest amount of MISthat could be detected was 0.325 ng (6.25 Mg/L), and themidrange of the assay was 2.66 ng (53.2 /ug/L) Thewithin-assay variability ranged from 8.4-16.0% (mean,11.8%), with the greatest errors occurring at the higherconcentrations. The between-assay variability for nineassays was 11.2%.

MIS could not be detected in female serum of any age.When rhMIS was diluted with female serum or PBST,different standard curves were obtained. Dilution in neo-natal or adult female serum caused a shift (i.e. a loweringof absorbance values) in the standard curve comparedwith dilution in PBST alone. However, as we requirerelatively large volumes of serum samples to detect smallquantities of MIS, it is necessary to use serum as a

450 nm

1 00 1 0 0 0

FIG. 1. EIA standard curve; assay conditions are described in text.The five points from 6.25-100 jtg/L were fit to a straight line forcalculation of data. The A450 is plotted against the log10 concentrationof rhMIS in micrograms per L.

diluent in this assay. The inhibitory effects of serum inimmunoassays is a common finding and is probably dueto nonspecific protein interactions (17). Thus, adult fe-male serum was used as this was easily available, and allsamples and standards were diluted so that a constantamount of serum was always present.

All samples were measured in duplicate and at one,two or three dilutions, depending on amount of serumavailable. Diluted sera showed good parallelism with thestandard, and the SD as a percentage of the mean from88 samples ranged from 1.8-19.7%, with a mean of 11.6%.As the mouse bioassay (which requires microgramamounts of MIS) is not as sensitive as the EIA, andthere is no other immunoassay available yet for humanMIS, we could not compare our results with any others.

None of the hormones used to check for specificity ofthe assay was positive at any of the concentrations tested,indicating no cross-reactivity of the antibodies to MISwith any of these hormones. In addition, since the slopesof the assay curves constructed with MIS-positive serumdiluted with adult female serum consistently paralleledthe slope of the standard curve, this observation supportsthe specificity of the EIA for MIS.

MIS serum data

The means of MIS serum concentrations from differ-ent age groups are shown in Fig. 2. The concentration ofMIS in male serum appears to rise after birth untilapproximately 1 yr of age and then decreases until after16 yr of age, when it is undetectable. There was nocorrelation with the clinical diagnoses, which included alarge number of medical and surgical conditions, none ofwhich was a genital anomaly,

The difference in MIS levels at 0-2 months and 4-12months was found to be statistically significantly differ-ent, with P < 0.005. The MIS levels after 14 yr of age(when 14-16 yr are compared with 0-2 months and when

4 m 6 m 1 y 2 y 4 y 6 y 8 y l O y 1 2 y 1 4 y 1 6 y

AGE RANGE (months/years)

FIG. 2. Mean serum MIS levels in boys as a function of age. Note thatthe age ranges alter after 6 months, 1 yr, and 2 yr of age. The barsindicate the SEM, with the values above representing the number ofsubjects in each group.


14 BAKER, METCALFE, AND HUTSON JCE & M • 1990Vol70«Nol

16-18 yr are compared with 0-2 months) were alsostatistically significantly different from the levels at 0-2months (P < 0.025).

Discussion

We have developed a specific and sensitive sandwichEIA for detection of human serum MIS using a monoclo-nal antibody raised against bMIS and polyclonal anti-bodies to rhMIS.

A sandwich assay was chosen to increase the sensitiv-ity of the assay, and as only one suitable monoclonalantibody was available, this was used as the primaryantibody to increase the specificity of the assay. We wereunable to use affinity-isolated antibodies due to thelimited amount of purified MIS available. However, thisshould not be a problem in the future, and this shouldallow us to improve the sensitivity of the assay. We wereable to measure MIS in sera from boys aged from 1 dayto 16 yr. It appears that serum MIS is negligible in malesaged 16-18 yr. This age group presumably includes post-pubertal males, although no information was availableon their pubertal status. It would be interesting in afuture study to compare pubertal status with serum MISlevels.

The amounts of MIS given are based on Lowry (18)protein measurements of rhMIS and so may not beabsolutely accurate values, since they are relative to BSA,used as protein standard. However, the qualitative pic-ture of the ontogeny of MIS in the postnatal boy is stillvalid. The high levels of MIS in the serum of boys aged4-12 months suggests that MIS may be involved in animportant function at this time. In addition, the MISpeak levels occur just after the postnatal secretion oftestosterone, which is documented at 0-4 months, witha peak at 2 months (19). Beyond 3-6 months of age,sexual development is in a quiescent phase, with fewimportant events until the onset of puberty toward theend of the first 10 yr of life. However, one significantevent is the transformation of gonocytes to type A sper-matogonia in the postnatal testis, which occurs mostrapidly at 3-5 months in the human (20) and at about 1week of age in the rat (21). This transformation isessential for normal spermatogenesis after puberty, andwithout it, infertility may occur (20). Since MIS concen-trations normally are high during this period, it is pos-sible that MIS may be directly or indirectly involvedwith gonocyte development. Hadziselimovic et al. (20)have postulated that gonocyte transformation is andro-gen dependent, but this remains speculative.

Germ cell maturation is inhibited in cryptorchid testes,with the first sign of abnormality between 12-24 monthsof age (22). Lack of germ cell maturation in cryptorchidtestes may be associated with low levels of MIS, whichhave been found in biopsies of undescended testes (5).

This study, by documenting high levels of MIS at 4-12 months postnatally, provides evidence that MIS hasa function in the male beyond the fetal period. Thetemporal association between MIS secretion and keyevents in germ cell development may be a clue to discov-ering the predicted functions of MIS after birth.

AcknowledgmentsWe would like to thank Sharyn Robinson and Sophie Gazeas for

help with immunization and ascites production. For technical adviceon tissue culture methods we are grateful to Kerry Fowler and MarcelaPotenza, and to Simone Richardson for advice on EIA development.We thank Elizabeth Vorrath for typing the manuscript.

References1. Donahoe PK, Cate RL, McLaughlin DT, et al. Mullerian inhibiting

substance: gene structure and mechanism of action of a fetalregressor. Recent Prog Horm Res. 1987;43:431-67.

2. Josso N, Picard J-Y, Tran D. The anti-Mullerian hormone. RecentProg Horm Res. 1977;33:117-60.

3. Hutson JH, Ikawa H, Donahoe PK. The ontogeny of Mullerianinhibiting substance in the gonads of the chicken. J Pediatr Surg.1981;16:822-7.

4. Donahoe PK, Ito Y, Price JH, Hendren III WH. Mullerian inhib-iting substance activity in bovine fetal, newborn and prepubertaltestes. Biol Reprod. 1977;16:238-43.

5. Donahoe PK, Ito Y, Morikawa Y, Hendren WH. Mullerian inhib-iting substance in human testes after birth. J Pediatr Surg.1977;12:323-30.

6. Vigier B, Tran D, Du Mesnil Du Buisson F, Heyman Y, Josso N.Use of monoclonal antibody techniques to study the ontogeny ofbovine anti-mullerian hormone. J Reprod Fertil. 1983;69:207-14.

7. Necklaws EC, La Quaglia MP, McLaughlin D, Hudson P, Mudgett-Hunter M, Donahoe PKD. Detection of Mullerian inhibiting sub-stance in biological samples by a solid phase sandwich radioim-munassay. Endocrinology. 1986;118:791-6.

8. Josso N, Tran O, Picard J-Y, Vigier B. Physiology of anti-Mullerianhormone: in search of a new role for an old hormone. In: TsafririA, Eshkol A, eds. Development and function of reproductive or-gans. New York: Raven Press; 1986;73-84.

9. Donahoe PK, Ito Y, Hendren WH. A graded organ culture assayfor the detection of Mullerian inhibiting substance. J Surg Res.1977;23:141-8.

10. Newbold RR, Carter DB, Harris SE, McLaughlan JA. Moleculardifferentiation of the mouse genital tract: serum-free organ culturesystem for morphological and biochemical correlations. In Vitro1981;17:51-4.

11. Cotton RGH, Jennings IJ, Choo KH, Fowler K. Isolation andcharacterisation of a myeloma-spleen-cell hybrid producing anti-body to phenylalanine hydroxylase. Biochem J. 1980;191:777-83.

12. Beards GM, Campbell AD, Cottrell NR, et al. Enzyme-linkedimmunosorbent assays based on polyclonal and monoclonal anti-bodies for Rotavirus detection. J Clin Microbiol. 1984; 19:248-54.

13. Ey PL, Prowse SJ, Jenkins CR. Isolation of pure IgGb IgG2a andIgG2b immunoglobulins from mouse serum using protein A-Seph-arose. Immunochemistry. 1978;15:429-36.

14. Burger HG, McLachlan RI, Bangah M, et al. Serum inhibin con-centrations rise throughout normal male and female puberty. JClin Endocrin Metab. 1988;67:689-94.

15. Daughaday WH, Rotwein P. Insulin-like growth factors I and II.Peptide, messenger ribonucleic acid and gene structures, serum,and tissue concentrations. Endocr Rev. 1989;10:68-91.

16. Josso N, Picard J-Y. Anti-Mullerian hormone. J Physiol Rev.1986;66:1038-90.

17. Ohtaki S, Endo Y. Sandwich enzyme immunoassay of thyroglobulinin human serum. In: Ishikawa E, Kawai T, Miyai K, eds. Enzymeimmunoassay. Tokyo: Igaku-Shoin; 198l;198-207.

18. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein meas-


SERUM MIS LEVELS 15

urement with the Folin phenol reagent. J Biol Chem. 1951;193:265-75.

19. Job JC, Toublanc JE, Chaussin JL, Grendel D, Roger M, CanlorbeP. The pituitary-gonadal axis in cryptorchid infants and children.Eur J Pediatr. 1987;146(Suppl 2):S2-5.

20. Hadziselimovic F, Thommen L, Girard J, Herzog B. The signifi-cance of postnatal gonadotropin surge for testicular development

in normal and cryptorchid testes. J Urol. 1986;136:274-6.21. Constantinople NL, Walsh PC. Activity of Mullerian-inhibiting

substance in postnatal testes. Surg Forum. 1973;24:538-9.22. Huff DS, Hadziselimovic F, Duckett JW, Elder JS, Snyder HM.

Germ cell counts in semithin sections of biopsies of 115 unilaterallycryptorchid testes. The experience from the Children's Hospital ofPhiladelphia. Eur J Pediatr. 1987;146(Suppl 2):S25-7.

American Society of Andrology—15th Annual Meeting and Postgraduate CourseTo be held April 6 to 9, 1990, Downtown Marriott Hotel, Columbia, SC. The American Society ofAndrology is a Society whose members include physicians, veterinarians, and basic scientists interested inthe entire field of reproduction, including normal and abnormal aspects of clinical andrology, animalreproduction, and fertility and infertility. The theme for the Postgraduate Course, to be held on Friday,April 6,1990 will be "1mmunologic Aspects of Infertility." The scientific program will feature major lecturesby prominent investigators on aspects of pediatric andrology, techniques in andrology research, the role ofgrowth factors and oncogenes in reproduction, and in vitro fertilization and embryo transfer. Short oraland poster presentations of new research will also be delivered.

For further information please contact:Tu Lin, M.D. Program Chairman - (803) 776-4000orHoward Nankin, M.D., Local Arrangements Chairman - (803) 733-3112


serum levels of mullerian inhibiting substance in boys from birth to 18 years, as determined by...

Documents