Transferrin and gonadal dysfunction in man*

FERTILITY AND STERILITY Copyright c 1982 The American Fertility Society

Vol. 38, No.5, November 1982 Printed in U.SA.

Transferrin and gonadal dysfunction in man*

Steve D. Holmes, Ph.D.t Larry I. Lipshultz, M.D. Roy G. Smith, Ph.D. Division of Urology, The University of Texas Medical School at Houston,_ Houston, Texas

Transferrin concentrations were quantitated in the seminal fluid of normal, oligozoospermic, and azoospermic patients and related to other known parameters of testicular function. Transferrin concentration in the semen of patients 2 months after vasectomy (132 ± 1.8 llI!/ml) was significantly less than that obtained from pregnancy-proven donors (65.6 ± 10.1 llI!/ml). This indicates that approximately 80% of the seminal fluid transferrin is derived from the testes. The concentration of transferrin in semen from patients with azoospermia (14.4 ± 1.8 llI!/ml), severe oligozoospermia (17.5 ± 1.7 llI!/ml), and moderate oligozoospermia (21.8 ± 4.3 llI!/ml) was significantly lower than normospermic groups. Serum follicle-stimulating hormone (FSH) was measured in a group of infertile patients; those having an elevated FSH had a significantly lower concentration of semen transferrin, 14.1 ± 1.6 llI!/ml, compared with patients who had FSH levels within the normal range (33.7 ± 5.3 llI!/ml). It is possible that the underlying cause in decreased spermatogenesis associated with both an elevated FSH and a decreased transferrin concentration is impaired Sertoli cell function. Fertil Steril 38:600,1982

Received March 29, 1982; revised and accepted July 12, 1982. *Supported by National Institutes of Health (NIH) Grants DA 02852 and HD 11479. tReprint requests: Steve D. Holmes, Ph.D., Department of Urology, Baylor College of Medicine, Room 440E, 1200 Moursund, Houston, Texas 77030.

ticular cells, has not to date been physiologically assayable with the use of specific cell markers. Rat Sertoli cells in monolayer culture have been shown to secrete a large number of proteins,4, 5 of which only four have been identified. s-9 One of the secreted proteins, androgenbinding protein (ABP), has been well documentedS, 10; however, this protein represents only a very small fraction (0.05% to 0.15%) of the total amount of secreted protein. 7 , 11 Measurement of seminal fluid ABP in man has been reported, but the procedure is time-consuming, and only very low levels (0.1 to 0.5 nmole/100 ml semen) have been detected. 12 Transferrin has recently been detected in the media ofrat7 and human 13 Sertoli cells and represents 1% to 5% of the total newly synthesized protein. High levels of transferrin have also been reported in seminal fluid. 14, 15 This investigation describes the measurement of seminal fluid transferrin as a possible marker of Sertoli cell function.

Holmes et aI. Transferrin and gonadal dysfunction

Fertility and Sterility

Approximately 15% of married couples have primary or secondary infertility.l In one-third of these couples, the male partner will be found to have a severe impairment. 2 Unfortunately, 30% of these infertile men will escape a specific diagnosis. 3 Much of the difficulty can be traced to inadequate markers for specific cellular function within the testes. While the Leydig cells can be evaluated by quantitation of serum luteinizing hormone (LH) and testosterone (T) and the germinal epithelium further defined by examination of seminal fluid and testicular biopsies, the Sertoli cell, one of the most important of the intrates-

600

100

1 .~ -;;

80

80

c:

I!

~

40

'"

20

~

o

o~

~

•"

0,,0"-0

"•

. . . . . 0--0 ___0

~~'--oT:5--------'~---------~rl------~~ Transferrin (ng/tube)

Figure 1 Standard curve of RIA of human 125I-transferrin. A plot of B-BnIBo-Bn x 100 versus log concentration of transferrin, where Bo is the CPM of the zero tube, Bn is the nonspecific binding, and B the cpm of the standard. Human 125I-transferrin, 0; rat transferrin, _; human lactoferrin, .l.

MATERIALS AND METHODS

tions were at 22° C). The tubes were then centrifuged at 2500 x g for 15 minutes, the supernatant was aspirated, and the pellet was counted. Nonspecific binding (in the presence of 1000 ng transferrin) was less than 4% of the total counts. Unknown samples were extrapolated from a logit-log plot of the standards. Transferrin concentration was linear over the range 0.5 to 50 ng per tube. Follicle-stimulating hormone (FSH) was measured by RIA from kits provided by Amersham Corporation, Arlington Heights, IL. Our results are expressed as ng/ml FSH, where 1 mlU/ml Second International Reference Preparation-human menopausal gonadotropin (IRP-hMG) = 23 ng/ml LER-907. Seminal fluid was obtained by masturbation after 2 to 3 days of sexual abstinence. Following liquefaction, the volume of the samples was measured, sperm density was determined with a hemocytometer, and motility as well as forward progression was estimated. 16 Patients from our infertility clinic were classified into four groups on the basis of sperm concentration: (A) azoospermia, defined as no spermatozoa present in a centrifuged sample of seminal fluid; (B) severe oligozoospermia, < 10 x 106 spermlml; (C) moderate oligozoospermia, 10 to 20 X 106 spermlml; and (D) normospermia, > 20 x 106 spermlml. Seminal fluid was also obtained from patients 2 months after vasectomy, but only ifthere were no spermatozoa present in a centrifuged sample. The

Human transferrin, lactoferrin, rat transferrin, and goat anti-rabbit immunoglobulin G (lgG) were obtained from Cappel Laboratories, Cochranville, PA. Rabbit anti-human transferrin was from Bio-Rad, Richmon
Figure 2 Chromatographic separation of transferrin. A 100-....1 aliquot (10 ng) of 125I-transferrin was applied to a 70 x 0.5-mm column of Sephadex G100 equilibrated with phosphate-buffered saline with 0.1% albumin (PBSA). Each 300 ....1 fraction collected was measured for radioactivity (___). Seminal fluid from a pregnancy-proven donor was diluted 11500 in PBSA and 100 ....1 chromatographed as above. The immunoreactivity in the fractions was determined by RIA (0- -0).

Vol. 38, No.5, November 1982

Holmes et al. Transferrin and gonadal dysfunction

100

10

80

8

60

6

C. >< c:

'"c: '" ~

C[

a: >-

.Q

c:

t

40

""'-'

20

'"'"

c: ~

>=

::;;

2

'"

c:

9

X

0 1

2

3

4

5

6

7

8

9

10

Fraction

601

Table 1. Transferrin Concentration in Seminal Fluid Study group

Sperm density

Seminal fluid transferrin

x l(/llml

Wflml

Postvasectomy Azoospermic Severely oligozoospermic Moderately oligozoospermic Normospermic Pregnancy-proven donors

o

<10

13.2 ± 1.8a (n = 6) 14.4 ± 1.8 (n = 9) 17.5 ± 1.7 (n = 21)

10-20

21.8 ± 4.3 (n = 6)

> 20 > 65

42.5 ± 5.2 (n = 15) 65.6 ± 10.1 (n = 12)

o

aMean ± standard error of the mean. 0.5

10

50

ng Transferrin

Figure 3 . Dose-response curves for transferrin standards aJ?-d seminal fluid from a pregnancy-proven donor. Transfernn, ~. Seminal fluid diluted serially from 1/500 to 118000 (~- -~) In PBSA, 100 ILl assayed in a total volume of 300 ILl (see Materials and Methods).

specimens were centrifuged at 1000 x g for 10 minutes, and the separated seminal plasma was stored at - 20° C.

RESULTS The RIA for transferrin (Fig. 1) was very reproducible, with an intraassay coefficient of variation (CV) of 3.0% and an interassay CV of 6.4%. Specificity of the antibody for transferrin was demonstrated by immunoelectrophoresis. Only one precipitin arc corresponding to transferrin was observed with human serum (data not shown). Lactoferrin is also an iron-binding protein present in semen 17 and has a molecular weight similar to that of transferrin. 18 However, lactoferrin and rat transferrin, tested in concentrations up to 500 ng per tube, did not affect the binding of 125I-transferrin (Fig. 1). An aliquot of seminal fluid from a pregnancy-proven donor was chromatographed on Sephadex G100, and the immunoreactivity in the fractions was compared with the chromatographic profile of 125I-transferrin (Fig. 2). There was a close agreement between the RIA-determined transferrin and radiolabeled transferrin. The dose-response curves of transferrin and a serial dilution of seminal fluid were also compared, as shown in Figure 3. Both curves were found to be linear and parallel, (parallelism t-test = 0.84; see Rodbard I9 ). Similar parallelism was observed for serum transferrin (data not shown). The concentration of transferrin in seminal fluid from pregnancy-proven donors (65.6 ± 10.1 602

Holmes et al. Transferrin and gonadal dysfunction

f.Lg/ml) was five times greater than from postvasectomy patients (13.2 ± 1.8 f.Lg/ml; Table 1), indicating that approximately 80% of the seminal fluid transferrin is derived from the testes. The transferrin concentrations in the semin~l fluid of patients with azoospermia (14.4 ± 1.8 f.Lg/ml) was not significantly different from the concentration in postvasectomy patients (13.2 ± 1.8 f.Lg/ml; Table 1). As with the azoospermic group, the patients with severe (17.5 ± 1.7 f.Lg/ ml) and moderate oligozoospermia (21.8 ± 4.3 f.Lg/ml) had significantly decreased transferrin concentrations, compared with the pregnancyproven donors (P < 0.001). Those patients who were classified as normospermic had transferrin levels (42.5 ± 5.2 f.Lg/ml) not significantly different from the pregnancy-proven donors but greater than the other groups (P < 0.01). The pregnancy-proven donors had a sperm motility of ~ 60% and a forward progresion of ~ 2 + (scale 0 to 4). The infertility patients were similarly divided on the basis of sperm viability and forward progression. As shown in Table 2, there was no statistical difference in seminal fluid transferrin concentrations in either the division of motility « 60% or ~ 60%) or forward progression « 2+ or ~ 2+). Of the 51 patients in this study, it was found that 26 had had their serum FSH evaluated. It Table 2. Transferrin Concentration in the Seminal Fluid of Infertility Patients Compared with Motility and Forward Progression Seminal fluid transferrin Wflml

Forward progression < 2 + Forward progression "" 2 +

27.6 ± 3.2a (n = 28) 33.3 ± 6.2 (n = 17)

Motility < 60% Motility "" 60%

29.4 ± 4.2 (n = 22) 30.1 ± 4.5 (n = 23)

aMean ± standard error of the mean.

Fertility and Sterility

Table 3. Transferrin Concentration in Seminal Fluid from Patients with Normal or Elevated Serum FSH Study group

Seminal fluid transferrin IJIIlml

Normal FSH (50-350 ng/ml) Elevated FSH (> 350 ng/ml)

33.7 ± 5.3a (n = 16) 14.1 ± 1.6 (n = 10)

aMean ± standard error of the mean.

was of interest to compare the seminal fluid transferrin to the serum FSH in these patients (Table 3). Those patients with a normal FSH (50 to 350 ng/ml) had a significantly greater (P < 0.01) transferrin concentration (33.7 ± 5.3 Ilg/ml) than those patients with an elevated FSH (14.1 ± 1.6 Ilg/ml). The sperm concentration in the normal FSH group varied from 1 to 150 X 106 sperm/ml. In the group with elevated FSH, one patient had 17 x 106 sperm/ml, while the other nine had less than 4 x 106 sperm/ml. The serum transferrin was also determined in these patients, and no difference was found between the groups with normal or elevated FSH. DISCUSSION

The RIA for seminal fluid transferrin was precise, reproducible, and comparable to values obtained by radial immunodiffusion. 14 , 15 Chromatographic identity of radioimmunoassayed transferrin with radiolabeled transferrin was demonstrated, and dose-response curves for standards and seminal fluid were both linear and parallel. This is the first report that we can find in the literature of the quantitation of transferrin by RIA. The concentration of transferrin in seminal fluid from pregnancy-proven donors averaged 65.6 Ilg/ml (Table 1). Tauber et al. 14 measured transferrin by radial immunodiffusion and reported similar values (70 Ilg/ml) in their group of donors. By examining split ejaculates of human seminal fluid, they demonstrated that transferrin and spermatozoa had similar distribution patterns. This is indirect evidence that both components originate from the same part of the reproductive tract. We found that the transferrin concentration in seminal fluid from postvasectomy patients was 13.2 Ilg/ml, confirming the above report and indicating that 80% of the transferrin was derived from the testes. We have recently shown that human Sertoli cells in monolayer culture secreted transferrin, 13 which constituted 1% to 3% of the total proteins Vol. 38, No.5, November 1982

synthesized. Thus, the Sertoli cells are probably the major source of seminal fluid transferrin. The transferrin possibly serves as an iron source for the developing germ cells. When transferrin concentrations were determined in infertile patients on the basis of sperm motility or forward progression, no differences were noted (Table 2); only sperm density was directly related to transferrin levels. Seminal fluid transferrin concentrations in patients with azoospermia and postvasectomy patients were not significantly different. Transferrin concentrations in seminal fluid of patients with oligozoospermia were significantly less than in pregnancy-proven donors or the normospermic group. It appears, therefore, that azoospermia and oligozoospermia are directly correlated with decreased transferrin concentration, which perhaps is indicative of Sertoli cell dysfunction. In patients exhibiting no hypothalamic-pituitary disorders, elevated serum FSH has been associated with spermatogenic damage. 2o This observation has been credited to the lack of a testicular protein, termed inhibin, capable of FSH suppression. In the rat, the Sertoli cell has been demonstrated to be a source of this FSH-inhibiting factor.21 We have shown that decreased spermatogenesis is associated with elevated serum FSH, in accordance with the inhibin hypotheSiS. 21 , 22 Since serum LH was not measured, we cannot rule out Leydig cell dysfunction in the group of patients with elevated FSH. 23 However, our results concerning the inverse relationship between FSH and transferrin indicate a possible abnormality of Sertoli cell function. The measurement of transferrin in seminal fluid has the potential of being a clinical marker of Sertoli cell function. REFERENCES 1. Simmons FA: Human infertility. N Engl J Med 255:1140, 1956 2. MacLeod J: Human male infertility. Obstet Gynecol Surv 26:335, 1971 3. Greenberg SH, Lipshultz LI, Wein AJ: Experience with 425 subfertile male patients. J Uro1119:507, 1978 4. Wright WW, Musto NA, Mather JP, Bardin CW: Sertoli cells secrete both testis-specific and serum proteins. Proc Natl Acad Sci USA 78:7565, 1981 5. Koskimies AI, Kormano M: The proteins in fluids from the seminiferous tubules and rete testis of the rat. J Reprod Fertil 34:433, 1973 6. Steinberger A, Heindel JJ, Lindsey IN, Elkington JSH, Sanborn MB, Steinberger E: Isolation and culture of FSH responsive Sertoli cells. Endocr Res Commun 2:261, 1975 Holmes et aI. Transferrin and gonadal dysfunction

603

7. Skinner MK, Gris~old MD: Sertoli cells synthesize and secrete transferrin-like protein. J ,BioI Chem 255:9523, 1980 8. Lacroix M, Smith FE, Fritz IB: Secretion of plasminogen activator by Sertoli cell enriched cultures. Mol Cell Endocrinol 9:227, 1977 9. Feig LA, Bellve AR, Erickson NH, Klagsbrun M: Sertoli cells contain a mitogenic polypeptide. Proc Nat! Acad Sci USA 77:4774, 1980 10. Fritz IB, RommertB FG, Louis BG, Dorrington JH: Regulation by FSH and dibutyryl cyclic AMP of the formation of androgen binding protein. J Reprod Fertil 46: 17, 1976 11. Wilson RM, Griswold MD: Secreted proteins from rat Sertoli cells. Exp Cell Res 123:127,1979 12. Plymate SR, Furiss BL, Smith ML, Jacob WH, Matez LA: Seminal fluid androgen binding protein. Andrologia 13: 308, 1981 13. Holmes SD, Lipshultz LI, Smith RG: Human Sertoli cell function and transferrin secretion (Abstr). Presented at the Seventy-Seventh Annual Meeting of the American Urology Association, May 16 to 20, 1982, Kansas City, Missouri 14. Tauber PF, Zaneveld LJD, Propping D, Schumacher GFB: Components of human split ejaculates. J Reprod Fertil 43:249, 1975

15. Schill WB, Krizic A, Iljosk H: Determination of various semen parameters and sex hormone levels in subfertile men during Kallikrein therapy. Adv Exp Med BioI 120:537, 1979 16. Amelar RD, Dubin L, Walsh PC: Male infertility. Philadelphia, W. B. Saunders Co, 1977, p 105 17. Heckman A, Rilmke P: The antigens of human seminal plasma with special reference to lactoferrin as a spermatozoa-coating antigen. Prot BioI Fluids 16:549, 1969 18. Aisen P, Listowsky I: Iron transport and storage proteins. Annu Rev Biochem 49:357,1980 19. Rodbard D: Statistical quality control and routine data processing for radioimmunoassays and immunoradiometric assays. Clin Chem 20:1255,1974 20. de Kretser DM: Endocrinology of male infertility. Br Med Bull 35:187, 1979 21. Steinberger A, Steinberger E: Secretion of an FSH-inhibiting factor by cultured Sertoli cells. Endocrinology 99:918, 1976 22. Smith KD, Steinberger E: What is oligospermia? In The Testis in Normal and Infertile Man, Edited by P Troen, HR Nankin. New York, Raven Press, 1977, p 489 23. Rodriguez-Rigau LJ, Smith KD, Steinberger E: A possible relation between elevated FSH levels and Leydig cell dysfunction in azoospermic and oligospermic men. J Androl 1:127,1980

Holmes et al. Transferrin and gonadal dysfunction

Fertility and Sterility

604

1