Graphics & Design

A Preponderance of Circulating Basic Isoforms Is Associated with Decreased Plasma Half-Life and Biological to Immunological Ratio of Gonadotropin-Releasing Hormone-Releasable Luteinizing Hormone in Obese Men 1

Description
A Preponderance of Circulating Basic Isoforms Is Associated with Decreased Plasma Half-Life and Biological to Immunological Ratio of Gonadotropin-Releasing Hormone-Releasable Luteinizing Hormone in Obese Men 1
Published
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
   A Preponderance of Circulating Basic Isoforms Is Associated with Decreased Plasma Half-Life andBiological to Immunological Ratio of Gonadotropin-Releasing Hormone-Releasable Luteinizing Hormone inObese Men* C. CASTRO-FERNA ´NDEZ, A. OLIVARES, D. SO¨DERLUND,J. C. LO´PEZ-ALVARENGA, E. ZAMBRANO, J. D. VELDHUIS, A. ULLOA-AGUIRRE,  AND  J. P. ME´NDEZ  Research Units in Developmental Biology (C.C.-F., A.O., D.S., J.P.M.) and Reproductive Medicine(A.U.-A.), Instituto Mexicano del Seguro Social; and Departments of Reproductive Biology (E.Z.) and Endocrinology (J.C.L.-A.), Instituto Nacional de la Nutricio´n SZ, Mexico D.F., Mexico; and Department of Internal Medicine, University of Virginia Health Sciences Center (J.D.V.),Charlottesville, Virginia 22908  ABSTRACT Hormonalabnormalitiesofthereproductiveaxishavebeendescribedin obesity. In men, extreme obesity is associated with low serum tes-tosterone (T) and high estrogen [estrone and estradiol (E 2 )] levels. Aschangesinthesexsteroidmilieumayprofoundlyaffectthecarbohydrateheterogeneityandthussomeofthebiologicalandphysicochemicalprop-erties of the LH molecule, we analyzed the relative distribution of LHisoformscirculatingunderbaselineconditions(endogenousGnRHdrive)as well as the forms discharged by exogenous GnRH stimulation fromputative acutely releasable and reserve pituitary pools in overweightmen. Secondarily, we determined the impact of the changes in LH ter-minal glycosylation on the  in vitro  bioactivity and endogenous half-lifeofthegonadotropin.Sevenobesesubjectswithbodymassindexesrang-ingfrom35.7–45.5kg/m 2 andsevennormalmenwithbodymassindexesfrom22.5–24.2kg/m 2 underwentbloodsamplingat10-minintervalsfora total of 10 h before and after the iv administration of 10 and 90   g GnRH. Basally released and exogenous GnRH-stimulated serum LHisoformswereseparatedbypreparativechromatofocusingandidentifiedby RIA of eluent fractions. Serum pools of successive samples collectedacross2-hintervals(fiveserumpoolspersubject)containingLHreleasedunderbaselineandexogenousGnRH-stimulatedconditionsweretestedforbioactivityemployingahomologous invitro bioassay.MeanserumTand E 2  levels were significantly lower and higher, respectively, in theobese men than in the control group [serum T, 13.5  2.4  vs.  19.4  1.4nmol/L(mean  SEM ;  P  0.01);serumE 2 ,0.184  0.01 vs. 0.153  0.01nmol/L(  P  0.05)].MeanbaselineserumLHlevelsweresimilarinobesesubjectsandnormalcontrols(13.3  1.3and12.2  1.2IU/L).Althoughmultiple parameter deconvolution of the exogenous GnRH-induced LHpulsesrevealedthatthemagnitudeofthepituitaryresponseintermsof secretory burst mass, secretory amplitude, and half-duration of the LHpulses was similar in obese and control subjects, the apparent endoge-nous half-life of LH was significantly (  P    0.05) shorter in the obesegroup (98  11 min) than in the normal controls (132  10 min). Underall conditions studied, the relative abundance of basic isoforms (thosewithpH  7.0)wassignificantly(  P  0.05)increasedintheobesesubjectscompared with the controls (percentages of LH immunoactivity recov-eredatpH  7.0:obesesubjects,34–57%;normalcontrols,22–46%).ThebiologicaltoimmunologicalratioofLHreleasedinbaselineandlowdose(10  g) GnRH-stimulated conditions were similar in obese subjects andnormal controls, whereas LH released by obese subjects in response tothe high (90   g) GnRH dose exhibited significantly lower ratios thanthose detected in normal individuals (0.62  0.07 and 0.45  0.09  vs. 1.01    0.10 and 0.81    0.09 for LH released within 10–120 min and130–240 min after GnRH administration in obese and controls, respec-tively;  P  0.05).Collectively,theseresultsindicatethatthealteredsexsteroid hormone milieu characteristic of extreme obesity provokes aselective increase in the release of less acidic LH isoforms, which maypotentially modify the intensity and duration of the blood LH signaldelivered to the gonad. Altered glycosylation of LH may thereforerepresent an additional mechanism modulating the hypogonadalstate prevailing in morbid obesity. (  J Clin Endocrinol Metab  85: 4603–4610, 2000) Received January 26, 2000. Revision received June 23, 2000. Rerevi-sion received August 21, 2000. Accepted September 2, 2000.Address all correspondence and requests for reprints to: Juan PabloMe´ndez, M.D., or Alfredo Ulloa-Aguirre M.D., D.Sc., Coordinacio´n deInvestigacio´n Me´dica, Unidad de Investigacio´n Me´dica en Biologı´a delDesarrollo, Centro Me´dico Nacional Siglo XXI, Instituto Mexicano delSeguro Social, Avenue Cuauhte´moc 330, Apdo. Postal 73–032, Col. Doc-tores, 06725 Me´xico D.F., Mexico. E-mail: jpmb@servidor.unam.mx andaulloaa@buzon.main.conacyt.mx.* This work was supported by Grant 28589N from the Consejo Na-cional de Ciencia y Tecnologı´a (CONACyT), Mexico (to A.U.-A.); GrantFP 0038/426 from the FOFOI-Instituto Mexicano del Seguro Social,Mexico City, Mexico (to J.P.M.); and NIH General Clinical ResearchCenter Grant RR-00847 and NIH U-54 Specialized Cooperation CentersProgram in Reproductive Research (NICHHD HD-28934; to J.D.V.). S EVERAL STUDIES have demonstrated that plasma totaltestosterone (T) concentrations are decreased in obesemen compared with those found in nonobese men of similarage (1–5). Moreover, there is decreased binding capacity of thesexhormone-bindingglobulin(SHBG)inobesity(2,4–6).Accordingly, plasma free T concentrations in these subjectsare less decreased than total T levels (3, 5); in fact, somestudies report normal free T levels (1, 7). These changes areinversely correlated with indexes of body weight, represent-ing a continuum across varying degrees of obesity (3).Weightlosscanreversethesechanges(7).Likewise,elevatedestrone and estradiol (E 2 ) concentrations have been docu-mented in some obese men (4, 5, 7). 0021-972X/00/$03.00/0 Vol. 85, No. 12The Journal of Clinical Endocrinology & Metabolism  Printed in U.S.A. Copyright © 2000 by The Endocrine Society 4603  In obese men, basal gonadotropin levels have been re-portedasbeingeitherwithin(1,2)orbelowthenormallimits(5, 8). Vermeulen  et al.  (5) demonstrated that integrated se-rum LH concentrations are significantly lower in extremelyobeseindividualsandthatmeanLHpulseamplitudeaswellas the mean sum of all diurnal LH pulse amplitudes are alsodecreased.The mechanism(s) responsible for the changes in hor-monal concentrations in obese men have not been clarified,although relative hyperinsulinism promotes a reduction inSHBG (4). Whether the relative hypoandrogenism has a pri-mary testicular or a hypothalamo-pituitary srcin remainsunknown. However, it has been reported that the responseof Leydig cells to hCG stimulation is normal in obese men,suggesting that the primary cause of the decreased T con-centrations does not reside within the testes (1, 9). On theotherhand,theassociatedhyperestrogenemiamayinfluencegonadotropin regulation (5).Carbohydrates in LH and other glycoprotein hormonesplay a major role in structure and function. Oligosaccharidesinfluence not only intracellular folding of the subunits andsecretion of the glycoprotein heterodimer, but also its circu-latory survival and capacity to evoke signal transduction atthe receptor level (10–13). Several studies indicate that gly-cosylation of anterior pituitary glycoprotein hormones isregulated by hypothalamic inputs and/or by end productsfrom the target glands under the control of these trophichormones (14–18). The cellular mechanisms by which rele-vantfeedbackendproductscontrolglycosylationareknownin part. For example,  in situ  hybridization studies haveshown that the messenger ribonucleic acid levels of severalglycosyltransferases (such as   -2,6-sialyltransferase,   -1,4-galactosyltransferaseand  -mannosidaseII)aresignificantlyincreasedinthyrotrophsderivedfromhypothyroidmice(16,17). In the case of gonadotropins, Dharmesh and Baenziger(18) observed that the activity of both pituitary  N  -acetyl-galactosamine transferase and sulfotransferase increasedseveral-fold after ovariectomy, thereby favoring the produc-tion of more LH oligosaccharides terminating with sulfate-GalNAc residues. Estrogen administration returned the ac-tivities of these transferases to basal levels (18).Considering the influence of sex steroids on the carbohy-drate structure of both gonadotropins, we attempted to de-fine the impact of the changes in the androgen to estrogenratio in male obese subjects on the glycoprotein isoformdistribution and  in vitro  biological activity of the LH signalsynthesized and secreted by the anterior pituitary gland. Forthis purpose, we investigated the charge distribution (whichin glycoprotein hormones is primarily dependent on thepresence of terminal sulfate and/or sialic acid residues) and in vitro  biological to immunological relationships of the LHisoforms released during baseline and GnRH-stimulatedconditionsinextremelyobesecomparedwithnormalweightindividuals. We assumed that consecutive administration of a low (10  g) and a high (90  g) GnRH dose would facilitatethe identification of the various types of LH isoforms dis-charged from an intracellular, biologically enriched, LH re-leasable pool (19–21) as well as forms of this gonadotropinthat were newly synthesized in that particular endocrinemilieu. The qualitative changes represented within the LHpulses released before and after GnRH administration fromtheanteriorpituitaryglandsofobeseandnormalindividualswere then correlated with the corresponding LH  in vitro  biological to immunological ratio (B/I ratio) as well as withLH secretory activity and half-lives, as resolved by multipa-rameter deconvolution analysis. Subjects and Methods Seven obese men, aged 24–32 yr (median, 27 yr), with body massindexes (BMI) ranging from 35.7–45.5 kg/m 2 (median, 39.7 kg/m 2 ) andseven normal men [21–28 yr old (median, 24 yr); BMI, 22.5–24.2 kg/m 2 (median,23.5kg/m 2 )]agreedtoparticipateinthestudy.Approvalfromthe human ethics committee of the institute and informed written con-sent from the volunteers were obtained. All subjects were in goodgeneral health; physical examination, including testicular size, was nor-mal, as were the results of routine laboratory test of liver, kidney, andthyroid.Allobesevolunteersweresingleandmanifestedonlydecreasedlibido,butnotimpotence;noneofthesubjectsinthisgrouphadreceivedany dietary treatment or anorexigenic drugs during the previous 6months.Volunteerswereadmittedtothemetabolicresearchwardoftheinstituteat0730h,andanindwellingheparinizedivcatheterwasplacedin an antecubital vein. Starting at 0800 h, blood samples were obtainedevery10minfor10h;atthebeginningofthethirdhour,subjectsreceivedarapidivbolusof10  gGnRH(SeronodeMexico,MexicoD.F.,Mexico)and,4hlater,asecondGnRHboluscontaining90  gofthedecapeptide.Subjects were recumbent during the study and consumed light meals at0900 and 1400 h. Blood samples were allowed to clot at room temper-ature for 30 min, then were centrifuged at 1000   g . Sera were separatedinto three aliquots and stored frozen at  20 C until assay.  LH, FSH, T, and E  2  immunoassays The RIA of LH was performed employing  125 I-labeled LH-I3 as thetracer (SA, 70–90   Ci/  g protein), the reference LH preparation LER-907 as the standard, and the antihuman LH-2, at a final dilution of 1:800,000,astheantiserum(22).Thesensitivityoftheassaywas0.7IU/L[1 mg LER-907  277 IU of the Second International Reference Prepa-ration of human menopausal gonadotropins (2nd IRP-HMG)]. Eachsubject’s set of samples was processed in duplicate for LH determina-tions in a single RIA run. The intra- and interassay coefficients of vari-ationweredeterminedusingmultiplereplicates(n  3/dose)ofaserumpool collected from postmenopausal women, assayed at dose levels thatdisplaced [ 125 I]LH from the antibody at 15–23%, 45–59%, and 75–84%total binding; these coefficients ranged from 4.1–6.2% and 6.1–11.3%,respectively. All LH RIA reagents were provided by the NIDDK (Be-thesda, MD) through Dr. A. F. Parlow from the National Hormone andPituitary Program (Torrance, CA). Serum FSH was measured in the firstand third baseline samples by an enzyme-linked immunoassay method(23), employing a FSH standard preparation provided by the WHOCollaborating Center for Research and Reference Services in the Immu-noassay of Hormones in Human Reproduction (London, UK) and cal-ibrated against the International Reference Preparation 78/549; the sen-sitivity of the assay was 0.5 IU/L as expressed in terms of the 2ndIRP-HMG, and the intra- and interassay coefficients of variation at theED 50  level were less than 7% and less than 13% respectively. Results of the LH and FSH immunoassays are expressed in international unitsaccording to the 2nd IRP-HMG. Reference ranges for the LH and FSHassays were 5–15 and 3–12 IU/L, respectively. Serum T and E 2  concen-trations were determined in each sample collected during the baselineperiod by RIA after solvent extraction (recoveries   90% for both sexsteroids) using antisera provided by the WHO Matched Reagent Pro-gram (Geneva, Switzerland) as previously described (24, 25). Referenceranges for the T and E 2  assays were 15–30 and 0.036–0.165 nmol/L,respectively. Intra- and interassay coefficients of variation for both as-says (at 45–55% total binding) were less than 5% and less than 8%,respectively. Chromatofocusing of serum samples Serum samples from all subjects were subjected to concentration bydialysis and freeze-dried. LH isoforms were then separated on the basis 4604 SERUM LH ISOFORMS IN MALE OBESITY   JCE & M  •  2000 Vol. 85  •  No. 12  ofchargeaspreviouslydescribed(22).Briefly,foreachindividualseries,samples corresponding to the (2-h) baseline period and to the low andhigh GnRH-stimulated study periods (4 h each) were separately pooled(threeserumpoolspersubject;pools1,2–3,and4–5inFig.1),transferredto dialysis membrane tubing (mol wt cut-off, 12,000–14,000; SpectrumMedical Industries, Los Angeles, CA), dialyzed at 4 C for 24 h againstdeionized water and thereafter against 0.01 mol/L ammonium carbon-ate (pH 7.5), and freeze-dried. Lyophilates were redissolved to 1/10ththesrcinalvolumeinPharmalyte(pH8–10.5)-HCl(PharmaciaBiotech,Piscataway, NJ; 1:45 dilution in deonized water, pH 7.0), and the sus-pensionwasthenappliedtothetopofa20  1-cmcolumnofpolybufferexchange resin (PBE-118, Pharmacia Biotech), previously equilibratedfor 18–24 h with 25 mmol/L triethylamine-HCl (pH 11.0) and chro-matofocusedat4C.Eluatefractions(2-mLeach)werecollectedataflowrate of 1 mL/4 min. The pH of each fraction was then measured, andwhen a limiting pH of 7.0 had been reached, the eluent buffer (Phar-malyte-HCl) was changed by Polybuffer-74 (Pharmacia Biotech) diluted1:8 in deonized water (pH 4.0) to elute proteins bound at pH 7.0–4.0.ProteinsboundatthelowerlimitingpH(pH  4.0;saltpeak)werefinallyrecovered by the addition of 1.0 mol/L NaCl to the chromatofocusingcolumn. Sets of fractions corresponding to eight 0.99 pH units wereseparately pooled, concentrated by dialysis and freeze-drying as de-scribed above, and stored frozen at  20 C until determination of its LHcontent by RIA. Each specimen was redissolved in phosphate (0.05mol/L)-bufferedphysiological(0.15mol/L)saline(pH7.4),suchthatthemajority of the dose levels fell on the linear portion of the LH RIAstandard curve (Fig. 2). To avoid interassay variations, all pooled frac-tions from the chromatofocusing separations were assayed in triplicateincubations in a single assay run.  In vitro bioassay of human LH  For each individual series of samples, those corresponding to thecomplete (2-h) baseline period and to the low and high GnRH-stimu-lated study periods were separately pooled, as shown in Fig. 1 (fiveserum pools per subject; pools 1–5 in Fig. 1), and assayed for cAMPproduction and immunoactive LH content. The capacity of each pool toprovokecAMPproductionwastestedbyahomologous invitro  bioassay,whichemploysthehumanembryonickidney-derived293celllinetrans-fected with the human LH receptor complementary DNA (provided byDr. Aaron J. W. Hsueh, Stanford University, Stanford CA). The srcin,handling, ligand specificity, and biochemical properties of the full-length recombinant human LH receptor expressed by this cell line have been described previously (26). Cells were cultured in DMEM (LifeTechnologies, Inc., Gaithersburg, MD), pH 7.3, supplemented with 2.5%FCS, 2 mmol/L  l -glutamine, 100 mg/mL geneticin (Life Technologies,Inc.), 50 U/mL penicillin, and 100   g/mL streptomycin (Sigma, St.Louis, MO) and grown in 162-cm 2 flasks (Costar, Cambridge, MA).Confluentcellswerescrapedandplatedin24-wellcultureplatesfor24hat 37 C in 5% CO 2 . Cells (5  10 4 cells/culture dish) were then washedand exposed to increasing doses (12.5–50   L) of each serum pool orLER-907 in the presence of 0.125 mmol/L 1-methyl-3-isobutylxanthine(Sigma) dissolved in 450   L supplemented DMEM for 24 h at 37 C.Samples (unknowns and standards) were diluted with serum fromwomentreatedwithoralcontraceptivesthatcontainedLHimmuno-and bioactivities not distinguishable from the zero dose, such that the finalconcentration of human serum in each sample did not exceed 10% (50  L/culture well). After incubation, the media and cells were boiled at90 C for 3 min and stored frozen at  20 C. The sensitivity of the assaywas 0.075 mIU LER-907/tube. All pools from a single subject were bioassayed in triplicate incubations in a single assay run. The inter- andintraassay coefficients of variation at the ED 50  dose level were less than18% and less than 10%, respectively. Aliquots (12.5–100   L) of samplesfrom each serum pool series bioassayed were additionally analyzed forimmunoreactive LH content as described above. cAMP RIA Total (intra- plus extracellular) cAMP levels were determined by RIAafter acetylation of the samples and cAMP standards. The RIA of cAMPwas performed as previously described (27), employing 2- O -monosuc-cinyl cAMP tyrosylmethyl ester (Sigma) iodinated by the chloramine-Tmethod and the CV-27 cAMP antiserum (NIDDK) at a final dilution of 1:50,000.Afterincubationat4Cfor24h,antibody-boundandfreecAMPwere separated by ethanol precipitation followed by centrifugation at1,200  g at 4 C. The sensitivity of the assay was 4 fmol/tube and theinter- and intraassay coefficients of variation ranged from 8–12% andfrom 4–6%, respectively.The relative  in vitro  biological activity of LH was calculated by in-terpolation. Data are expressed as the mean B/I activity ratio, the ratioof LH activity exhibited by serum pools 1–5 in the  in vitro  bioassayrelative to that yielded by the immunoassay, calculated after conversionof the results to milliinternational units per mL 2nd IRP-HMG. F IG . 1. Serum LH concentration responses to 10  g (  first black arrow  in each graph) and 90  g ( second arrow ) iv injections of exogenous GnRHin obese and control subjects ( symbols ). The  continuous lines  represent the group means. Aliquots of samples from each 120-min study period(no.1–5,delineatedbythe verticalbrokenlines )werepooledandanalyzedforLHisoformdistributionand invitro biologicalactivityasdescribedin  Subjects and Methods . CASTRO-FERNA ´NDEZ  ET AL.  4605   Deconvolution analysis of GnRH-induced LH pulses Deconvolution analysis was applied to compute the amplitude andmass of significant LH secretory bursts as well as to estimate the ap-parent endogenous LH half-life (t 1/2 ) in blood sampled at 10-min in-tervals after the administration of each of the two consecutive pulses of exogenous low and high dose GnRH (28, 29). Each LH pulse was fitindependently to allow for possible differences in the mode of LHsecretion to the two doses of exogenous GnRH. Based on equivalentsecretory burst half-durations and half-lives at the two GnRH doses, theoptimized estimates of these parameters were determined from bothpulse episodes considered jointly for statistical purposes (30, 31). Themass of each GnRH-stimulated LH secretory burst is the analyticalintegral of the corresponding deconvolution-resolved secretoryimpulse.  Statistical analysis Significant differences in mean serum E 2 , T, LH, and FSH concen-trations; LH secretory measures (mass, amplitude, half duration, andapparentendogenoushalf-lifeofGnRH-stimulatedLHsecretorybursts) between obese and control subjects were determined by BonferroniprotectedStudent’sunpairedtwo-tailed t test.Within-groupdifferencesin LH isoform distribution and LH B/I ratio before and after GnRHadministration were determined by ANOVA followed by Student’spaired t test.Allvaluesarereportedasthemean  sem ,unlessspecified. P  0.05 was considered statistically significant. Results  Baseline serum T, E  2  , and gonadotropin levels andsecretory LH response to exogenous GnRH  Mean serum T and E 2  levels were significantly lower andhigher, respectively, in the obese than in the control group[serum T, 13.5  2.4  vs.  19.4  1.4 nmol/L ( P  0.01); serumE 2 , 0.184    0.01  vs.  0.153    0.01 nmol/L ( P    0.05)]. Mean baseline serum LH and FSH concentrations were similar inobese and normal subjects (LH, 13.3    1.3 and 12.2    1.2IU/L; FSH, 7.0    3.2 and 6.8    3.4 IU/L, respectively; P  NS).In all subjects, a significant rise in serum LH levels wasobserved after administration of both low and high GnRHdoses (Fig. 1). Maximal LH concentrations occurred 10–50min(median,20mininbothobeseandcontrolsubjects)afterthelowGnRHpulseandbetween10–30min(median,30and20 min in obese and controls, respectively) after the highGnRH dose. In both groups, the burst mass of LH secretedwas significantly higher after the 90-  g GnRH dose (obesegroup, 43.8  9.0  vs.  61.5  11.5 IU/L; control group, 31.6  3.8  vs.  49.2  10.2 IU/L for the low and high GnRH doses,respectively;  P  0.01 for both groups). Although the meanmagnitude of LH response to both GnRH doses was higherin the obese subjects than in the control group (see above),the differences did not reach statistical significance. Like-wise, there were no differences between the obese and con-trol subjects in the magnitude of LH secretory amplitude(12.7  5.16 vs. 7.4  2.45IU/L  min)andhalf-duration(6.4  2.0  vs.  6.7  1.1 min) of the GnRH-provoked LH bursts. Theapparent endogenous LH half-lives of the deconvolvedGnRH-induced LH pulses were significantly ( P    0.05)shorter in the obese group (98    11 min) than in controls(132  10 min).  Distribution of serum LH isoforms Chromatofocusing disclosed consistent LH immunoactiv-ity within a pH range of 4.0–9.9 as well as in those fractionsrecovered after the addition of 1.0 mol/L NaCl to the chro-matofocusing columns. In both groups, serum LH isoforms F IG . 2. The standard curves of the LH RIA system employed in the present study ( dashed line ) and the positions of the unknown samples( symbols ). Each value represents the mean  SD  of the dose and percent binding observed for each basal and GnRH-stimulated set of samples(n  7 different sample pools for each pH boundary; data for pools with pH 9.0–9.99 and  10.0 were combined and presented as pH  9.0). 4606 SERUM LH ISOFORMS IN MALE OBESITY   JCE & M  •  2000 Vol. 85  •  No. 12  presentinbasalconditionswerepredominantlyrecoveredatpH values of 6.99 or less (Fig. 3). The percentage of LHimmunoactivity recovered in basic pH values (pH  7.0) wassignificantly ( P    0.05) higher in obese than in control in-dividuals. This increase in basic LH isoforms mainly oc-curred at the expense of the more basic isotypes recoveredwithin the pH range of 8.0–8.9 (not shown). In both groups,administrationof10and90  gGnRH,provokedasignificantincrease in the secretion of basic LH isoforms. Although therelative abundance of more basic isoforms released in re-sponsetobothGnRHchallengeswashigherintheobesethanin the control group, the difference reached statistical sig-nificance only for those isoforms secreted in response to thehigh dose (Fig. 3). The proportion of basic LH isoforms re-leased in response to the high GnRH dose also exceeded thatof their acidic (pH   6.99) counterparts in the overweightindividuals, but not in the nonobese control subjects (Fig. 3, lower panel ).  In vitro B/I ratio of serum LH  Five serum LH pools from each subject were tested at twoor three dose levels for LH bioactivity employing the ho-mologous  in vitro  bioassay system. Incubation of HEK-293cells expressing the recombinant human LH receptor withincreasingamountsofLER-907ortheserumpoolsfromeachset of samples induced significant and parallel dose-depen-dent cAMP accumulation (Fig. 4). In two subjects from eachgroup, baseline levels of   in vitro  bioactive LH present in pool1 were undetectable by the bioassay, whereas in four obeseandtheremainingcontrolsonlythelargestserumaliquot(50  L) showed measurable LH bioactivity. The baseline B/I LHratios were 0.30  0.1 and 0.35  0.1 for obese and controlsubjects,respectively( P  NS).TheB/ILHratiosforallotherserum pools are shown in Table 1. Ratios of B/I LH presentinpools2and3(collectedfromsamplesobtained10–120minand 130–240 min after low dose GnRH administration, re-spectively; see Fig. 1) were similar in obese and controls.However, LH released by obese subjects in response to thehigh GnRH dose exhibited significantly lower ratios thanthose detected in normal individuals. In both groups, LHpresent in pools 3 and 5 (last 2 h after the low and high doseGnRH challenges) exhibited lower B/I ratios than moleculesreleased immediately (first 2 h) after GnRH administration.The lowest B/I LH ratios were detected in pool 5 from theobese subjects (Table 1). Discussion The present study reveals that circulating LH in obeseindividuals contains an increased proportion of more basicisoforms compared with that in men with normal BMI. Thisdistinction in LH species between normal and obese subjectswas not significantly modified by exogenous GnRH admin-istration, which also allowed analysis of isoforms presump-tively present in distinct intracellular pools in gonadotropecells (19–21). Several mechanisms may explain this shift to-wardmorebasicLHisoformsinobesemen.Some invitro and in vivo  studies indicate that GnRH may regulate posttrans-lationalLHglycosylation(14,15,20,21,32).Accordingtothisreasoning, in obese subjects the higher proportion of basicisoforms may result from increased exposure to endogenousGnRH subserved by the reduced serum T levels (33, 34). Ourstudystrategy,however,didnotallowustoclearlyrecognizeany significant alteration in LH release in either basal orGnRH-stimulated conditions in obese subjects, which con-trasts with those observed in normal women during the latefollicular phase and in patients with polycystic ovarian syn-drome, in whom an association between the secretion of morebasicLHisoformsandaccentuatedsignalingofand/orheightenedpituitaryresponsivenesstoGnRH(presumptive-ly facilitated by an enhanced estrogen milieu) has been ob-served (22, 24, 35, 36). Alternatively, elevated proportions of more basic LH isoforms in obese men may be due solely toincreased pituitary exposure to endogenous estrogens. Ev-idence derived from clinical and experimental studies sup-ports this possibility. Serum LH from postmenopausalwomen, which is more negatively charged than at any stageof the menstrual cycle (36), becomes less acidic during long-term estrogen treatment (37, 38). In castrated female rats,sulfation and  N  -acetylgalactosamine incorporation into theoligosaccharides of LH is significantly reduced by estrogentreatment (18). More recently, we have shown that messen-ger ribonucleic acid expression of anterior pituitary   -2,3-sialyltransferase [one of the enzymes that incorporate sialic F IG . 3. Percentrecoveries(normalizedtothetotalLHrecoveredfromeach chromatofocusing run) of immunoactive LH in pH segments of 7.0 or more and 6.99 or less in serum pools from basal and GnRH-stimulated samples in obese and normal weight controls. Data arepresented as the mean  SD .  Different letters above each bar  indicatesignificant(  P  0.05)differenceswithinthesamesubjectgroupinthesame pH region ( i.e.  basal  vs.  10   g GnRH  vs.  90   g GnRH). *,  P  0.05,pH  7.0 vs. pH  6.99withinthesamesubjectgroup;&,  P  0.05 vs.  controls in the same pH area. CASTRO-FERNA ´NDEZ  ET AL.  4607
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks