antiequine chorionic gonadotropin (ecg) antibodies generated in goats treated with ecg for the...

Antiequine Chorionic Gonadotropin (eCG) AntibodiesGenerated in Goats Treated with eCG for the Inductionof Ovulation Modulate the Luteinizing Hormone andFollicle-Stimulating Hormone Bioactivities ofeCG Differently

VIRGINIE HERVE, FRANCOIS ROY, JEAN BERTIN, FLORIAN GUILLOU, AND

MARIE-CHRISTINE MAUREL

Laboratoire Mecanismes d’Action des Gonadotropines, Unite Mixte de Recherche 6073, Institut National de la RechercheAgronomique/Centre National de la Recherche Scientifique/Universite de Tours, Station de Physiologie de la Reproductiondes Mammiferes Domestiques, 37380 Nouzilly, France

In dairy goats, treatments associating a progestogen and theequine chorionic gonadotropin (eCG) are the easiest way toinduce and synchronize estrus and ovulation and to permitartificial insemination (AI) and/or out of season breeding.From the first treatment, the injection of eCG induces, in somefemales, the production of anti-eCG antibodies (Abs) that willinterfere with the effectiveness of subsequent treatments.These anti-eCG Abs delay the preovulatory LH surge and theovulation time, leading to poor fertility of the treated females.In this study, by in vitro bioassays, we show that anti-eCG Abscan positively or negatively modulate the LH and/or FSH bio-activities of eCG. Moreover, the modulation level of eCG bio-activity does not depend on the anti-eCG Ab affinity for eCG,as shown by surface plasmon resonance technology. The spec-

ificity of anti-eCG Abs tested by competitive ELISA high-lighted the importance of a glycan environment in the recog-nition mechanism, especially the sialic acids specific to eCG.The different effects of anti-eCG Abs on eCG bioactivitiescould be explained by two hypotheses. First, steric hindrancepreventing the interaction of eCG with its receptors wouldexplain the inhibitory effect of some anti-eCG Abs; second, aconformational change in eCG by anti-eCG Abs could induceinhibition or potentiation of eCG bioactivities. It is significantthat these modulations of eCG bioactivities by anti-eCG Absimpact mainly on the FSH bioactivity of eCG, which is essen-tial for ovarian stimulation and subsequent fertility aftertreatment and AI, and to a lesser extent on LH bioactivity.(Endocrinology 145: 294–303, 2004)

GONADOTROPINS PLAY AN important role in the con-trol of gametogenetic and endocrine activities of go-

nads. These complex glycoproteins are composed of twodissimilar subunits, � and �, whose noncovalent associationis required for their biological activities. Gonadotropins be-long to the glycoprotein hormone family and comprise hy-pophysial LH and FSH, and human (hCG) and equine (eCG)chorionic gonadotropins. Gonadotropins are currently usedin human and veterinary medicine to mimic the endocrinemechanisms of reproductive cycles. In humans, FSH andhCG are used to conduct in vitro fertilization protocols andin anovulation treatment. In ruminants, particularly goatsand ewes, eCG is commonly used in association with a pro-gestogen to induce and synchronize estrus and ovulation.

eCG is a dimeric glycoprotein secreted by trophoblasticcells in the mare between the 36th and 120th d of gestation(1, 2). Interestingly, dimeric eCG binds only to the equine LH

receptor (3, 4), whereas it exhibits pronounced FSH activityin addition to its LH activity in species other than equine (5,6). Each subunit is composed of a peptidic part connected toa glycan moiety (N- and O-chains). eCG is the most heavilyglycosylated glycoprotein hormone, with a majority of bi-antenna glycans ending mainly in sialic acids that play animportant role in the hormone half-life in vivo (7). Glycanchains are also involved in the stability of the heterodimerichormone (8) and are necessary for the efficiency of signaltransduction (9, 10). The dual activity of eCG, its long half-life, and its availability in large quantities make this uniquegonadotropin a convenient exogenous hormone in treat-ments to induce and synchronize estrus and ovulation insmall ruminants.

However, the repeated use of eCG treatments for the in-duction of ovulation is generally followed by a decrease infertility from 60% to 40% (11, 12). This phenomenon has beenexplained by unwanted immunological responses (13). In-deed, the presence of anti-eCG antibodies (anti-eCG Abs) inthe plasma of eCG-treated goats and ewes has been dem-onstrated (11–14). A survey of 2500 females (ewes and goats)demonstrated that a 500-IU eCG injection induced strictlysimilar humoral immune response kinetics, with a maximumanti-eCG Ab concentration between 10–17 d after injection(13, 14). About 60% of treated females developed Abs to eCG,

Abbreviations: Ab, Antibody; AI, artificial insemination; dg, degly-cosylated; eCG, equine chorionic gonadotropin; HBS, HEPES-bufferedsaline; hCG, human chorionic gonadotropin; IgG, immunoglobulin G;mAb, monoclonal antibody; Neu5Gc, N-glycolyl-neuraminic acid; RU,resonance units; SA, sialic acid; SPR, surface plasmon resonance.Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

0013-7227/04/$15.00/0 Endocrinology 145(1):294–303Printed in U.S.A. Copyright © 2004 by The Endocrine Society

doi: 10.1210/en.2003-0595

294

but anti-eCG Ab levels were highly variable among individ-uals. Previous reports showed a significant association be-tween the females displaying high or low anti-eCG responsephenotypes and particular microsatellite alleles located in-side the major histocompatibility complex class II (13, 14).Anti-eCG Abs, already present by the time of eCG injectionand resulting from the previous immune response againsteCG 1 yr earlier, were defined as residual anti-eCG Abs.Females with a high residual anti-eCG Ab concentrationexhibited a significantly lower fertility after AI, because of adelay in both the onset of estrus and the preovulatory LHsurge (13, 14).

Surprisingly, some females displayed significant fertilityafter AI despite high residual anti-eCG Ab levels. Therefore,to address the mechanism of immune interference in eCG-induced reproductive function, we tested the impact of anti-eCG Abs on both LH and FSH bioactivities of the eCG. In thepresent study we show that these polyclonal anti-eCG Abscould modulate one or both bioactivities of eCG differently.These anti-eCG Abs recognized primarily the glycan moietyof eCG. Furthermore, our results demonstrate that the in-terference of anti-eCG Abs with FSH activity and to a lesserextent with LH activity has a physiological impact on thefertility of treated females.

Materials and MethodsPlasma and hormone preparations

Goat plasmas were obtained from a previous study (13) conducted onfemales treated in several herds (n � 350) and on an experimental farmof Institut National de la Recherche Agronomique, Domaine de Galle(Avord, France; n � 200). In all cases, females were treated for 11 d withprogestogen (vaginal sponge impregnated with 45 mg fluorogestoneacetate) and received an im injection of 500 IU eCG (Syncro-part, batch13054A1, CEVA, Libourne, France) 48 h before withdrawal of thesponge. AI was routinely performed 43 h after the sponge was with-drawn. Plasma samples were collected just before eCG treatment (d 0)and 10 d later (d 10). All plasma samples were stocked at �20 C.Anti-eCG Abs, present on d 0 and resulting from the previous immuneresponse against eCG 1 yr earlier, were defined as residual anti-eCG Abs.

For the present study we selected 37 goats that exhibited the highestanti-eCG Ab concentration on d 0 from 3–22 �g/ml and on d 10 from6–70 �g/ml. The number of previous treatments per goat varied from2–5, depending on the age of the female. The plasma control corre-sponded to a pool of plasmas from non-eCG-treated goats.

eCG (standard eCG FL 652), eLH, eCG �- and �-subunits, and eCGNZY-01 isoforms (15) were provided by Dr. Y. Combarnous and F.Lecompte (Institut National de la Recherche Agronomique, Nouzilly,France). Totally deglycosylated eCG (GRB-VII-128A) was provided byDr. Georges R. Bousfield (Wichita State University, Wichita, KS).

Anti-eCG Ab affinity purification

Abs from selected plasmas were affinity-purified on a HiTrap Gprotein column (Amersham Pharmacia Biotech, Uppsala, Sweden) dis-playing a high binding capacity for goat immunoglobulins G (IgGs) (16).As previously described, 5 ml of each plasma were briefly submitted toaffinity purification (13). The purified Abs were then stored at �20 C.

Measurement of anti-eCG Ab concentrations by aquantitative ELISA

The anti-eCG Ab concentrations in plasmas or IgG-purified fractionswere measured using a specific quantitative ELISA as previously de-scribed (13). The anti-eCG Ab concentration was expressed as micro-grams per milliliter of plasma.

LH in vitro bioassay

The LH in vitro bioassay was based on the stimulation of testosteroneproduction by isolated rat Leydig cells as previously described (17).Different concentrations of eCG (100 �l) were preincubated overnight atroom temperature, either with or without undiluted plasma (9 �l) or thecorresponding IgG fraction, added to 91 �l Leibowitz L15 medium.Leydig cells (15 � 104 cells in 200 �l L15) were pooled with the prein-cubated samples and incubated for 4 h at 34 C. After centrifugation, thesupernatants were stored at �20 C. Testosterone concentrations wereassayed by a specific RIA (18).

FSH in vitro bioassay

The FSH in vitro bioassay was based on the stimulation of proges-terone production by the Y1 cell line derived from a mouse adrenalcortex tumor stably expressing the human FSH receptor (donated byAres Serono, Geneva, Switzerland) as previously described (19). Dif-ferent concentrations of eCG (500 �l) were preincubated overnight atroom temperature, either with or without 45 �l undiluted plasma or thecorresponding IgG fraction added to 455 �l culture medium. Next, 15 �104 Y1 cells were stimulated with 400 �l preincubated samples for 4 hat 37 C. The media were harvested, boiled, and centrifuged, and thesupernatants were collected and assayed for progesterone by a specificRIA (20).

Sensitivity of LH and FSH in vitro bioassays

The sensitivities of the two biological assays obtained with Y1 cellsand Leydig cells were analyzed with eCG ranging from 6–50 ng/ml. Theprogesterone secretion varied between 6.93–33.49 ng/ml, and the tes-tosterone secretion ranged from 9.34–25.67 ng/ml. Progesterone andtestosterone secretion started with the same concentration of eCG, i.e. 6ng/ml. The same range of eCG stimulation was thus used for both invitro biological assays.

Measurement of anti-eCG plasmas and IgG effects on LHand FSH bioactivities of eCG

In both biological assays we used plasmas collected on d 0 or 10 andtheir corresponding IgG fractions. Two controls were systematicallyperformed. The first determined the basal level of hormone productionby Leydig cells and Y1 cells. The second determined the level of hormoneproduction by Leydig cells and Y1 cells in the presence of control plasmaor of the corresponding IgG fraction devoid of anti-eCG Abs.

The results were expressed as the percentage of eCG bioactivity:100 � [(A � B)/(C � D)], where A is the amount of progesterone ortestosterone secreted during stimulation with a defined concentration ofeCG preincubated with a plasma or IgG of treated females, B is theamount of progesterone or testosterone secreted during stimulation withpreincubated plasma or with IgG of treated females without eCG, C isthe amount of progesterone or testosterone secreted during stimulationwith a defined concentration of eCG preincubated with a control plasmaor with control IgG of untreated females, and D is the amount of pro-gesterone or testosterone secreted during stimulation only with controlplasma or with control IgG of untreated females.

Radioligand receptor assay

Only IgG fractions purified from anti-eCG plasmas collected on d 0were analyzed. The ability of anti-eCG IgG to modulate the bindingactivity of eCG to the rat LH and FSH receptors was analyzed using aradioligand receptor assay, performed as previously described (17).Enriched receptor membrane fractions were prepared from rat testes.For the binding assay, 5 ng/ml to 10 �g/ml eCG were preincubatedovernight at room temperature with 6.5 �g/ml anti-eCG Abs. Eachpreincubated sample (100 �l) was then incubated for 4 h at 34 C, with50 �l buffer containing 36 mm CaCl2, 50 �l radioiodinated hormone(40,000 cpm) and 100 �l testicular membranes. The reaction was stoppedby adding 2 ml cold 10 mm Tris-HCl, pH 7.4, at 4 C, and the tubes werecentrifuged at 3,000 � g for 40 min. The bound radioactivity was mea-sured with a �-counter (Packard Instruments, Perkin-Elmer, Les Ulis,France).

Herve et al. • eCG Modulation by Abs in Treated Goats Endocrinology, January 2004, 145(1):294–303 295

Affinity measurement of anti-eCG Abs by surface plasmonresonance (SPR)

The anti-eCG Ab/eCG interaction was measured by SPR using aBiacore 1000 system (Biacore International, Uppsala, Sweden). Affinitymeasurements were performed with diluted plasmas or correspondingaffinity-purified IgGs collected on d 0 or 10. Sensor chips, amine cou-pling kit, and HEPES-buffered saline (HBS) buffer were supplied byBiacore (21).

eCG (35 �l) prepared at 100 �g/ml in 10 mm sodium acetate, pH 4,was covalently coupled on either a carboxymethylated dextran CM5sensor chip for purified IgGs or an F1 sensor chip for plasma analysis,using an amine coupling kit as described by the manufacturer. F1 sensorchips are well suited for plasma samples because their dextran matrixis shorter than CM5, and this reduces nonspecific binding when dilutedplasmas are injected. The eCG immobilization level was 6500 resonanceunits (RU) on CM5 and 1600 RU on F1 chips.

Kinetic measurements of anti-eCG Ab interaction on coupled eCGwere performed with HBS as running (20 �l/min) and diluting buffersat 23 C. The anti-eCG Ab concentration of injected IgG fraction (35 �l)ranged from 1.2–64 nm. Injected plasmas (35 �l) were diluted from 1:10to 1:80. To prevent nonspecific interactions on the coated surface, pu-rified IgGs and plasmas were preincubated for 30 min at 37 C in HBSto which was added 1 mg/ml CM-dextran (Fluka, Buchs, Switzerland)before injection. The regeneration of the coupled surface was performedby injecting NaOH 10 mm (10 �l). The binding curve obtained withcontrol plasma or purified IgG fraction was used for nonspecific bindingsubtraction.

Kinetic and affinity constants were calculated using BIA Evaluationsoftware (version 2.2.4). This Biacore software allows kinetic constants(kon and koff) to be calculated with sd and a statistical validation test (�2).The affinity constant was calculated as the kon/koff ratio, and the dis-sociation constant was calculated as the koff/kon ratio.

Specificity of anti-eCG Abs

The eCG regions recognized by anti-eCG Abs were determined usinga competitive ELISA, previously described (22). Before the assay, goatplasmas collected on d 0 or 10 were incubated overnight with increasingconcentrations (from 50 ng/ml to 10 �g/ml) of standard eCG FL652,isolated eCG �- and �-subunits, eLH, totally deglycosylated eCG (dgeCG), eCG treated with neuraminidase (eCG SA4, 1% sialic acid), andthree isoforms of eCG NZY-01 (15): eCG isoform with 17% sialic acid(eCG SA1), eCG isoform with 9.5% sialic acid (eCG SA2), and eCG

isoform with 4.7% sialic acid (eCG SA3). The 100% cross-reaction wasobtained with standard eCG FL652.

Statistical analysis

Average values were represented as the mean � sem. Differences inprogesterone and testosterone secretions were determined by t test orANOVA. Differences were considered significant at P � 0.05.

ResultsPlasmas from eCG-treated goats modulate LH and FSHbioactivities of eCG differently

We analyzed 37 plasmas with high concentrations of anti-eCG Abs that were collected just before eCG injection (d 0)and 10 d after injection (d 10). Figure 1 shows the effects ofplasmas on the LH and FSH bioactivities of eCG, rangingfrom 6–50 ng/ml, incubated with medium, with plasmasfrom untreated females, or with different plasmas from eCG-treated females. For each assay, eCG preincubated with me-dium or control plasma was used as a reference. Both con-trols led to identical testosterone and progesterone secretionwhen stimulation was carried out with a similar concentra-tion of eCG. However, when eCG was preincubated withplasmas containing anti-eCG Abs, three different effects onLH and FSH bioactivities of eCG were observed. First, someplasmas displayed no effect on LH and FSH bioactivities andinduced the same testosterone and progesterone productionas that obtained with the control plasma. Secondly, someplasmas displayed an inhibitory effect on LH and/or FSHbioactivities. These plasmas induced lower progesterone andtestosterone secretion for all eCG concentrations tested com-pared with the control plasma. In the last case, some plasmasdisplayed a hyperstimulatory effect on LH and/or FSH bio-activities, leading to higher testosterone and progesteronesecretion for all eCG concentrations tested compared withcontrol plasma. By contrast, when incubated without eCG, all

FIG. 1. Effects of three plasmas on eCG-induced testosterone (A) and progesterone (B) secretion tested in LH and FSH in vitro bioassays. TheLH in vitro bioassay was based on the stimulation of testosterone production by isolated rat Leydig cells. The FSH in vitro bioassay was basedon the stimulation of progesterone production by the Y1 cell line derived from a mouse adrenal cortex tumor stably expressing the human FSHreceptor. Plasmas were preincubated overnight with eCG ranging from 6–50 ng/ml, and testosterone (A) and progesterone (B) secretion wasmeasured by RIA after 4-h incubation at 34 C for Leydig cells and 37 C for Y1 cells. Representative stimulation-response curves were obtainedwith eCG preincubated with different plasmas, displaying an inhibitory effect on both LH and FSH bioactivities (E), no effect on either LH orFSH bioactivities (�), or a hyperstimulating effect on both LH and FSH bioactivities (�) and with eCG preincubated with control plasma (Œ).

296 Endocrinology, January 2004, 145(1):294–303 Herve et al. • eCG Modulation by Abs in Treated Goats

of these plasmas had no significant effect on basal secretionof progesterone and testosterone.

Among the 37 plasmas tested that contained anti-eCG Abs,10 plasmas had no effect on eCG bioactivities (data notshown), whereas 27 modulated eCG bioactivities. To com-pare the effects of these 27 plasmas, LH and FSH bioactivitieswere determined at eCG concentrations ranging from 12.5–100 ng/ml. The control plasma used as the reference repre-sents 100% of LH and FSH bioactivities. Six types of plasmaswere observed. Three types included plasmas displaying aninhibitory effect on the eCG bioactivities (n � 17): 11 plasmasinhibited up to 10 times both FSH and LH bioactivities (Fig.2A), 2 plasmas inhibited up to 10 times only FSH bioactivity(Fig. 2B), and 4 plasmas inhibited up to two times only LHbioactivity (Fig. 2C). Surprisingly, three other types includedplasmas displaying a hyperstimulatory effect on the eCGbioactivities (n � 10): two plasmas hyperstimulated both LH

and FSH bioactivities (up to 2.3 times for LH bioactivity andup to 1.5 times for FSH bioactivity; Fig. 2D), four plasmashyperstimulated up to 2 times only FSH bioactivity (Fig. 2E),and four plasmas hyperstimulated up to two times only LHbioactivity (Fig. 2F). These results demonstrate that plasmasof eCG-treated females containing anti-eCG Abs can mod-ulate in different ways one or both bioactivities of eCG.

IgGs are responsible for the modulation of LH and FSHbioactivities of eCG

To ensure that the effects were due to the anti-eCG Abs andnot to plasma contaminants, we compared the effects ofdifferent plasmas on the eCG bioactivities with those dis-played by their respective purified IgG fraction. One hyper-stimulatory plasma and one inhibitory plasma of both bio-activities of eCG were analyzed, as well as one control

FIG. 2. Types of anti-eCG plasma effects onthe LH and FSH bioactivities of eCG. Plas-mas were preincubated overnight with arange of eCG (12.5–100 ng/ml) before cellstimulation. Results were expressed as apercentage of LH (�) or FSH bioactivity (f)of eCG. The dark line corresponding to 100%bioactivity was obtained with controlplasma of untreated females, devoid of anti-eCG Abs (�). The different histograms rep-resent a plasma displaying an inhibitory ef-fect on both LH and FSH bioactivities (A), aplasma displaying an inhibitory effect onlyon FSH bioactivity (B), a plasma displayingan inhibitory effect only on LH bioactivity(C), a plasma displaying a hyperstimulatingeffect on both LH and FSH bioactivities (D),a plasma displaying a hyperstimulating ef-fect only on FSH bioactivity (E), and aplasma displaying a hyperstimulating effectonly on LH bioactivity (F). Testosterone andprogesterone secretion were measured asdescribed above. **, Significant difference inbioactivity percentage (P � 0.05).

Herve et al. • eCG Modulation by Abs in Treated Goats Endocrinology, January 2004, 145(1):294–303 297

plasma collected from an untreated female. Plasmas andtheir corresponding IgG fractions were tested at the sameanti-eCG Ab concentrations. Control and tested IgG fractionswere used at a similar concentration of total IgG.

When the inhibitory plasma or its purified IgG fractionwas preincubated with 100 ng/ml eCG, testosterone secre-tion decreased to 28% (Fig. 3A) and 31% (Fig. 3B) of thecontrol level, respectively, and progesterone secretion to 20%(Fig. 3C) and 53% (Fig. 3D) of the control level, respectively.When the hyperstimulatory plasma or its purified IgG frac-tion was preincubated with 100 ng/ml eCG, testosteronesecretion increased to 145% and 156% (Fig. 3, A and B) of thecontrol level, respectively, and progesterone secretion to144% (Fig. 3C) and 171% (Fig. 3D) of the control level, re-

spectively. The non-IgG counterpart of the plasmas did notmodify LH and FSH bioactivities (data not shown).

In conclusion, IgGs preincubated with eCG provoked thesame effects on LH and FSH bioactivities as their corre-sponding plasmas. These results clearly demonstrate thatanti-eCG IgGs present in plasma are responsible for inhibi-tion or hyperstimulation of LH and FSH bioactivities of eCG.

Modulation of LH or FSH bioactivity of eCG by anti-eCGAbs affects the fertility of treated goats differently

It has been demonstrated that only residual anti-eCG Absoriginating from the previous treatment impact on the de-crease in fertility. Females with high residual anti-eCG con-centration exhibit a lower fertility after AI because of a delayin the preovulatory LH surge. As demonstrated above, theseresidual Abs displayed an effect on the LH and/or FSHbioactivities of eCG.

For this study we analyzed anti-eCG plasmas (n � 37)collected on d 0, just before the administration of a new eCGtreatment. All exhibited a high anti-eCG Ab concentration.Our aim was to discriminate which effect of anti-eCG plas-mas on LH and FSH bioactivities had the greatest impact onthe fertility of treated goats after AI. We compared the effectsof the plasmas on 12.5 ng/ml eCG, which roughly corre-sponds to the quantity of injected eCG (50 �g), taking intoaccount the blood volume of a goat (4 liters). Table 1 showsthe effects of anti-eCG plasmas on the eCG bioactivities andfertility results of females after the newly administered eCGtreatment and AI. Eleven goats were not pregnant, and 26were pregnant and kidded. First, we observed that 100% ofgoats with anti-eCG Abs with a hyperstimulatory effect onFSH bioactivity were pregnant and kidded, as did 76% ofgoats with anti-eCG Abs with no effect on FSH bioactivity.By contrast, none of the goats with anti-eCG Abs inhibitoryof FSH bioactivity was pregnant. However, among the fe-males with anti-eCG Abs inhibitory of LH bioactivity, only39% were pregnant and kidded. This result contrasts with91% of pregnant females in which anti-eCG Abs had a hy-perstimulatory effect on LH bioactivity and 85% of pregnantfemales in which anti-eCG Abs had no effect on LH bioac-tivity. For comparison, the normal pregnancy rate when

FIG. 3. Effects of a hyperstimulatory and an inhibitory plasma andtheir corresponding IgG fractions on the LH and FSH bioactivities ofeCG. IgG fractions were obtained by affinity chromatography on pro-tein G as described in Materials and Methods. Plasmas and the IgGfractions prepared at the same anti-eCG Ab concentration were pre-incubated overnight with culture medium alone (�) or added to 100ng/ml eCG (f). Testosterone (A and B) and progesterone (C and D)secretion was measured by RIA. **, P � 0.05; *, P � 0.1 (significantdifference in testosterone or progesterone secretion).

TABLE 1. Modulation of LH or FSH bioactivity of eCG byanti-eCG plasmas affects the fertility of treated females afterAI differently

Effects of anti-eCG Abs oneCG bioactivities

Kidding rate afterAI (%) No. of females

FSH (�) 100 10FSH (0) 76 21FSH (�) 0 6LH (�) 91 11LH (0) 85 13LH (�) 39 13

For this study, anti-eCG plasma samples from 37 previouslytreated goats were used. Plasmas were collected on d 0, just before theeCG injection, and exhibited a high anti-eCG Ab concentration. Theireffects on LH or FSH bioactivities of eCG were analyzed according tothe fertility result of females after AI: 11 goats were not pregnant, and26 were pregnant and kidded. Anti-eCG Ab effects on LH or FSHbioactivities of eCG are symbolized by (0) for no effect, (�) for hy-perstimulating effect, and (�) for inhibitory effect.

298 Endocrinology, January 2004, 145(1):294–303 Herve et al. • eCG Modulation by Abs in Treated Goats

goats are treated with eCG for the first time is, on the average,65%. Our results demonstrate that plasmas displaying a hy-perstimulatory effect on both bioactivities only come frompregnant goats, and plasmas displaying an inhibitory effecton both bioactivities only come from nonpregnant goats. Thisstudy explains the lack of pregnancy of treated females dueto an inhibition of eCG bioactivities by anti-eCG Abs, espe-cially Abs inhibitory of FSH bioactivity. These results clearlydemonstrate that the FSH bioactivity of eCG and, to a lesserextent, LH bioactivity are necessary to induce ovulation withaccurate timing to achieve significant fertility of treated andinseminated females.

Effects of anti-eCG Abs on eCG binding to the LH andFSH receptors

To determine whether the modulation of LH and FSHbioactivities was due to inhibition of hormone binding to itsreceptors, we explored the effects of two IgG fractions in-hibitory of both bioactivities on the binding of eCG to rat FSHand LH receptors (Fig. 4, A and B). The first inhibitory IgGfraction elicited 48% and 45% inhibition of binding to FSHand LH receptors, respectively. The second inhibitory IgGfraction did not alter the binding of eCG to either receptor,as the binding curve obtained was similar to that of eCGpreincubated with medium alone. Similarly, we investigatedthe effect of an IgG fraction hyperstimulatory of both eCGbioactivities, which did not change the binding of eCG to theFSH and LH receptors.

Affinity for eCG of the plasmas containing anti-eCG Abs orIgG fractions

To investigate whether the anti-eCG Ab effects on eCGbioactivities correlated with their affinity for eCG, the affinityfor eCG of plasmas of eCG-treated females or of their puri-fied IgG counterpart was determined by SPR. Affinity wasmeasured with either purified IgG fractions or diluted plas-mas when sample volumes were limited. First, we verifiedthat plasmas and their corresponding purified IgG fractionsyielded similar kinetic constants (data not shown).

The results obtained showed that plasmas or IgG fractionsdisplayed strikingly different interaction kinetics whatevertheir effects on the eCG bioactivities. For example, whensamples were analyzed at 13.3 nm anti-eCG Ab, the maximalbinding level, measured just after the injection of samples,varied from 150-1200 RU. Likewise, the rate of dissociationmeasured 6 min later varied from 10–70% (data not shown).

The affinity measurements of four types of anti-eCG Absare summarized in Table 2. The results obtained clearly showa high variability in the affinity (Ka) and dissociation constant(Kd) values whatever the effects of anti-eCG Abs on eCGbioactivities. The dissociation constants varied from 0.06–13.3 nm. It should be noted that most of the tested plasmasor IgGs displayed a high affinity for eCG, with the Ka valueranging from 1010–108 m�1. Significantly, these results clearlydemonstrated that anti-eCG Abs displaying a strong effect oneCG bioactivities did not have the highest affinity constant.Likewise, the lack of effect on eCG bioactivities was not dueto a low affinity of anti-eCG Abs, but, rather, resulted from

the localization of the epitopes they recognized on the hor-mone surface.

Specificity of plasmas of eCG-treated goats

To determine which part of eCG is recognized by each typeof anti-eCG Ab effect, 21 plasmas of eCG-treated femalesdisplaying a clear effect on eCG bioactivities were analyzed.Each plasma was preincubated with native eCG, eCG �-sub-unit, eCG �-subunit, chemically dg eCG, or eLH. Cross-reaction was calculated at 50% of the displacement obtainedwith purified native eCG as competitor by competitiveELISA.

Table 3 summarizes the specificity of the plasmas in re-lation to their effects on eCG bioactivities. Most plasmasweakly recognized isolated eCG �- and �-subunits and dg

FIG. 4. Effects of purified anti-eCG IgG on the binding of eCG to theFSH and LH receptors. Increased concentrations of eCG, rangingfrom 5 ng/ml to 10 �g/ml, were incubated overnight at room temper-ature without anti-eCG IgG fractions (�) or with two representativepurified IgG fractions that had an inhibitory effect on both bioactiv-ities of eCG (Œ and f) and one IgG fraction hyperstimulatory of bothbioactivities (E). After 4-h incubation, the radioactivity bound to theFSH receptor (A) or to the LH receptor (B) was measured.

Herve et al. • eCG Modulation by Abs in Treated Goats Endocrinology, January 2004, 145(1):294–303 299

eCG. Nineteen of 21 plasmas cross-reacted with eCG� lessthan 5%, and the other 2 cross-reacted less than 9%. Eighteenof 21 plasmas cross-reacted with eCG� less than 1.5%, andthe other 3 cross-reacted between 20–44.7%. Eighteen of 21plasmas did not recognize dg eCG or recognized it only veryweakly, with a cross-reaction of less than 5%, and the other3 had a cross-reaction of less than 20%. Thus, plasmas con-taining anti-eCG Abs preferentially recognized the eCG��dimer and seemed to be influenced by the glycan chains ofthe dimeric hormone.

No plasma with a hyperstimulatory effect on the FSHbioactivity of eCG recognized dg eCG, but they weakly (4%)recognized eCG� and clearly cross-reacted with eCG �-sub-unit (between 20–44.7%). Plasmas with an inhibitory effecton both eCG bioactivities weakly recognized dg eCG andeCG �-subunit. No plasma displaying no effect on either eCGbioactivity recognized eCG�. For the other types of plasmas,we did not observe a correlation between the specificity andthe modulation of eCG bioactivities by anti-eCG Abs.

To confirm these observations, we analyzed the cross-reaction with eLH, which has the same peptidic structure aseCG, but differs in its glycan composition. We observed thatanti-eCG Abs did not recognize eLH. These results confirmthat most anti-eCG Abs recognize the glycan moiety of eCG.

The glycan part of eCG is mainly composed of biantennaglycans ending in sialic acids, in contrast with eLH endingmainly in sulfates. Therefore, we next tested plasma speci-ficity toward totally dg eCG and different dg isoforms ofeCG. We found that the sialic acid composition of eCG mod-ulated the plasma specificity for eCG (Table 4), which de-creased in line with the sialic acid content. Only 5 of the 10plasma samples used in this study were tested against dgeCG (1% of sialic acids), and none of these five recognized thedesialylated eCG. These data demonstrate the important roleof sialic acids in the immunogenicity of eCG. They reinforce

TABLE 2. Calculation of kinetic constants for the interaction between eCG and several anti-eCG plasmas or purified IgGs displayingdifferent effects on LH and FSH biological activities of the hormone

Anti-eCG Ab effects on eCG bioactivities kon (M�1, sec�1) koff (sec�1) ka (M�1) Kd (nM)

FSH (0), LH (0) 1.66, 107 1.13, 10�3 1.46, 1010 0.06(n � 3) 1.33, 106 1.41, 10�3 9.4, 108 1.06

2.25, 105 1.63, 10�3 1.38, 108 7.24

FSH (�), LH (�) 1.65, 106 9.16, 10�5 1.8, 1010 0.05(n � 4) 1.89, 105 1.54, 10�4 1.63, 109 0.61

8.67, 105 1.74, 10�3 4.98, 108 22.32, 105 7.31, 10�4 3.17, 108 3.15

FSH (�), LH (�) 4.87, 105 3.37, 10�4 1.44, 109 0.69(n � 2) 2.01, 105 9.96, 10�4 2, 108 4.95

FSH (�), LH (0) 1.43, 106 3.48, 10�4 4.1, 109 0.24(n � 5) 1.12, 106 1.36, 10�3 8.23, 108 1.21

8.99, 105 3.73, 10�3 2.41, 108 4.146.11, 10

53.29, 10�4 1.85, 109 5.38

9.67, 104 1.29, 10�3 7.49, 107 13.3

Affinity (Ka) and dissociation constants (Kd) were calculated from association (kon) and dissociation rates (koff), which were analyzed usingBIA evaluation software. Anti-eCG Ab effects on eCG bioactivities are symbolized by (0) for no effect, (�) for hyperstimulating effect, and (�)for inhibitory effect.

TABLE 3. Characterization of the eCG regions recognized by theplasmas containing anti-eCG Abs (n � 21), which display differenteffects on eCG bioactivities

Plasmasize (n)

Plasma effects oneCG bioactivities

Percentage of crossed-reaction withthe competitors

dg eCG eCG � eCG � eLH

1 LH (�) FSH (�) 0 4 44.7 01 LH (0) FSH (�) 0 4 20 05 LH (�) FSH (0) �1 �1 �1 02 LH (0) FSH (0) �1 �3 0 01 LH (�) FSH (0) 20 0 0 02 LH (�) FSH (�) �20 �9 �1.5 08 LH (�) FSH (�) �5 �5 �1.5 01 LH (�) FSH (�) 0 1 20 0

For each plasma, the percentage of cross-reactivity was measuredtoward standard eCG, eCG�, eCG�, dg eCG, and eLH, using a com-petitive ELISA. The 100% cross-reaction was determined with stan-dard eCG. The inhibition of anti-eCG Ab binding to coated eCG wasmeasured by increased concentrations of each competitor, rangingfrom 50 ng/ml to 10 �g/ml. Anti-eCG Ab effects are symbolized by (0)for no effect, (�) for hyperstimulating effect, and (�) for inhibitoryeffect.

TABLE 4. Specificity of anti-eCG plasmas (n � 10) towarddifferent forms of desialylated eCG

Plasma no.Competitors

dg eCG eCG SA117% SA

eCG SA29.5% SA

eCG SA34.7% SA

eCG SA41% SA

1 0 80 35 14 ND2 0 74 19 7 ND3 0 49 29 23 ND4 0.5 100 83 48 ND5 1.5 74 65 9 06 1.6 56 43 32 07 2 52 40 7 08 3 83 72 60 09 17 80 61 10 ND

10 20 87 87 31 ND

For each plasma, the percentage of cross-reactivity was measuredtoward standard eCG, eCG SA1 (eCG isoform with 17% sialic acids),eCG SA2 (eCG isoform with 9.5% sialic acids), eCG SA3 (eCG isoformwith 4.7% sialic acids), eCG treated with neuraminidase (eCG SA4with 1% of sialic acids), and dg eCG. The 100% of cross-reaction wasobtained with standard eCG. The inhibition of anti-eCG Ab bindingto coated eCG was measured by a competitive ELISA, after incubationof plasmas with increasing concentrations of each competitor, rangingfrom 50 ng/ml to 10 �g/ml. SA, Sialic acids; ND, not determined.

300 Endocrinology, January 2004, 145(1):294–303 Herve et al. • eCG Modulation by Abs in Treated Goats

the importance of the specific glycan sequences of eCG, andespecially sialic acids, in the humoral immune response in-duced by eCG treatment.

Discussion

The present report demonstrates the ability of polyclonalanti-eCG Abs secreted by eCG-treated goats to modulatepositively or negatively LH and/or FSH bioactivities of in-jected eCG. We clearly demonstrated that the purified IgGfraction of eCG-treated female plasmas is responsible formodulation of eCG bioactivities. It was shown that anti-oLHor anti-hCG anti-idiotypic Abs can mimic LH activity andstimulate progesterone production in granulosa cells in theabsence of gonadotropins (23, 24). In our study we ruled outthe presence of anti-eCG antiidiotypic Abs, because plasmasdid not modify basal levels of steroids on their own.

Modulation of the biological activity of hormones by Abs,especially with monoclonal Abs (mAbs), has been reportedin several in vitro or in vivo studies. In superovulation treat-ments, the administration of an anti-eCG mAb (Neutra-preg-nant mare’s serum gonadotropin) shortly after the preovu-latory LH surge synchronized final follicular maturation andshortened the period of multiple ovulations by neutralizingeCG (25). Other reports have shown an enhancement ofbioactivity for human, bovine and porcine GH (26–28) andhuman TSH (29) by complexing with certain mAbs. In vivo,the potentiation of GH (26) and TSH (29) activity by mAbshas been demonstrated in hypopituitary Snell dwarf mice aswell as in normal sheep for GH (30). Our results are rein-forced by a previous study (31) with 14 anti-eCG mAbsshowing that some anti-eCG mAbs modulated LH and/orFSH bioactivities of eCG either positively or negatively. ThemAbs without an effect on eCG bioactivities had the lowestaffinity for the hormone, and the degree of inhibition of eCGbioactivities was correlated to increasing mAb affinity foreCG. By contrast, our results showed that the level of mod-ulation of eCG bioactivities did not correlate with the affinityof polyclonal anti-eCG Abs present in treated goats. Thus, thehighest affinity for eCG was observed for anti-eCG Ab with-out an effect on eCG bioactivities.

In the present report we show that the modulating effectof anti-eCG Abs on eCG bioactivities can be explained by twomechanisms. Some anti-eCG Abs blocked hormone interac-tion with its specific receptors, probably because the anti-eCG Ab epitopes overlap the receptor-binding site on theeCG surface, inducing a steric hindrance that prohibits hor-mone binding. Other anti-eCG Abs did not disturb the eCGbinding, probably because they could induce a conforma-tional change in the three-dimensional structure of the hor-mone, leading to a specific enhancement or inhibition of itsbiological activity. Indeed, the enhancement of human TSHbioactivity can be explained by a modification of the hor-mone conformation by mAbs, which fits into the receptorbetter (29), whereas the enhancement of human GH bioac-tivity can be explained by an alteration of the nature or of theduration of the interaction between the mAb-hormone-receptor complex (32).

The different anti-eCG Ab effects on one or another eCGbioactivity suggest the existence of different molecular de-

terminants responsible for each of the LH and FSH bioac-tivities. This view reinforces the hypothesis of negative spec-ificity (33, 34), which assumes that there is a high affinity sitecarried on the �-subunit of all glycoprotein hormones andinhibitory sites located on the �-subunit in position 94–96 forLH bioactivity and in position 102–103 for FSH bioactivity,which prevents the binding of dimeric hormones to otherreceptors (35, 36). Therefore, it is possible that some poly-clonal anti-eCG Abs could modify the folding of the �-sub-unit sequences of eCG and thus modulate LH and/or FSHbioactivities.

Our results concerning anti-eCG Ab specificity demon-strated that anti-eCG Abs from treated goats weakly recog-nized eCG �- and �-subunits and did not display commonspecificity profiles according to their effects on eCG bioac-tivities. Surprisingly, our study showed that none recognizednative eLH. However, eCG and eLH have identical polypep-tidic sequences, encoded by the same �- and �-subunit genes(37, 38). The major structural differences are the glycosyla-tion of the N- and O-glycan chains linked to the �- and�-subunits of eCG. Glycan chains on eCG are mainly termi-nated by sialic acids, whereas more than 72% of glycan chainson eLH end in sulfates and a few in sialic acids (39, 40). Wedemonstrated that sialic acids are very important in the im-munogenicity of eCG, because the cross-reactivity of plasmascontaining anti-eCG Abs decreased in line with the sialic acidcontent of eCG. The majority of animals possess two types ofsialic acids, N-acetyl-neuraminic acid and N-glycolyl-neura-minic acid (Neu5Gc). However, Neu5Gc is absent only inadult birds and humans (41). Injections of glycoconjugatescontaining Neu5Gc are highly immunogenic in adult hu-mans, and exposure of humans to horse serum results in aserum sickness reaction (42, 43). Two other types of sialicacids (4-O-acetyl-9-O-lactyl-N-acetyl-neuraminic acid and4-O-acetyl-9-O-lactyl-N-glycolyl-neuraminic acid) have onlybeen identified in horses (41). This difference in sialic acidscould explain the immunogenic properties of eCG wheninjected in goats. Thus, anti-eCG polyclonal Abs largely seemto recognize glycan chains, but we cannot exclude that anti-eCG Abs recognize a peptidic moiety of the hormone onlywhen modeled by the glycan environment. Indeed, an mAbspecific for the hCG� 109–145 region recognized a delineatedepitope strictly dependent on a glycan environment in region115–145 (44). This epitope was recognized to an extent onnative hCG, whereas it was not recognized on desialyatedhCG.

Numerous studies have reported the importance of glycanchains in the hormonal signal transduction (45–47). For ex-ample, the presence of Asn52 glycan in the hCG dimer isnecessary for inducing and stabilizing a conformationalchange in hCG upon binding to the receptor, resulting inactivation of signal transduction pathways (48, 49). Thus, adirect interaction of Abs on one or several glycan chainscould provoke a total inhibition of the bioactivity by sterichindrance. Another possibility would be that the binding ofAbs to one of the glycan chains would freeze the latter in agiven position and induce a stimulation of the hormonalsignal by transconformation of the hormone. Such Abswould be very valuable tools for studying the interference

Herve et al. • eCG Modulation by Abs in Treated Goats Endocrinology, January 2004, 145(1):294–303 301

mechanism of the hormone glycan chains in the transductionof the hormonal signal.

In the present study we show that residual polyclonalanti-eCG Abs could modulate one or both eCG bioactivitiesdifferently. Both LH and FSH bioactivities of eCG areimportant for the efficacy of the treatment, because FSHin mammals induces follicular growth and maturation,whereas LH leads to terminal follicular growth and ovula-tion. The real importance of FSH bioactivity in ovarian stim-ulation and fertility is supported by transgenic mice whoseFSH �-subunit or FSH receptor genes have been invalidated,presenting a blockade of folliculogenesis before the pre-antrum stage, resulting in female sterility (50, 51). The com-parison of anti-eCG Ab effects in relation to the fertility oftreated goats after AI proves that FSH bioactivity and, to alesser extent, LH activity are indispensable for efficient ovar-ian stimulation and subsequent fertility.

The present results lead us to consider the use of anti-eCGAbs to modulate, either positively or negatively, the biolog-ical activity of exogenously injected hormones. These Abscould be valuable in diminishing the quantity of injected eCGwhile still obtaining significant fertility in ovulation induc-tion treatments. Such Abs are particularly interesting andhave implications for animal husbandry as well as poten-tially for human fertility treatment and assisted reproductiontechniques, where exogenous gonadotropins are intensivelyused.

Acknowledgments

We warmly thank Dr. Y. Combarnous and F. Lecompte (InstitutNational de la Recherche Agronomique, Nouzilly, France) for havingprovided eLH, standard eCG FL 652, �-eCG and �-eCG subunits, andeCG NZY-01 isoforms and for their critical comments and many helpfuldiscussions. We are grateful to Dr. G. Bousfield (Wichita University,Wichita, KS) for providing totally deglycosylated eCG GRB-VII-128A.We thank A. Ythier (Ares Serono, Geneva, Switzerland) for the gift ofY1 cells stably expressing the human FSH receptor. We acknowledge F.Bouvier (Institut National de la Recherche Agronomique, Domaine deGalles, Avord, France) and all the breeders involved in this study. Wealso thank Dr. P. Crepieux for critical reading and for reviewing theEnglish of the manuscript, and Dr. H. Watier for scientific comments.

Received May 14, 2003. Accepted September 26, 2003.Address all correspondence and requests for reprints to: Dr.

M. C. Maurel, Laboratoire Mecanismes d’Action des Gonadotropines,Unite Mixte de Recherche 6073, Institut National de la RechercheAgronomique/Centre National de la Recherche/Universite de Tours,Station de Physiologie de la Reproduction des Mammiferes Domes-tiques, 37380 Nouzilly, France. E-mail: maurel@tours.inra.fr.

This work was supported by funds from the Region Centre and theInstitut National de la Recherche Agronomique, France.

Present address for F.R.: International Agency for Research onCancer, 150 Cours Albert Thomas, 69008 Lyon, France.

References

1. Allen WR, Moor RM 1972 The origin of the equine endometrial cups. I.Production of PMSG by fetal trophoblast cells. J Reprod Fertil 29:313–316

2. Papkoff H 1974 Chemical and biological properties of the subunits of pregnantmare serum gonadotropin. Biochem Biophys Res Commun 58:397–404

3. Guillou F, Combamous Y 1983 Purification of equine gonadotropins andcomparative study of their acid-dissociation and receptor-binding specificity.Biochim Biophys Acta 755:229–236

4. Combarnous Y, Guillou F, Martinat N, Cahoreau C 1984 Origine de la doubleactivite FSH et LH de la choriogonadotropine equine (eCG/PMSG). AnnEndocrinol (Paris) 45:261–268

5. Combarnous Y, Hennen G, Ketelslegers JM 1978 Pregnant mare serum go-

nadotropin exhibits higher affinity for lutropin than for follitropin receptorsof porcine testis. FEBS Lett 90:65–68

6. Licht P, Gallo AB, Aggarwal BB, Farmer SW, Castelino JB, Papkoff H 1979Biological and binding activities of equine pituitary gonadotrophins and preg-nant mare serum gonadotrophin. J Endocrinol 83:311–322

7. Matzuk MM, Hsueh AJ, Lapolt P, Tsafriri A, Keene JL, Boime I 1990 Thebiological role of the carboxyl-terminal extension of human chorionic gonad-otropin �-subunit. Endocrinology 126:376–383

8. Heikoop JC, van den Boogaart P, de Leeuw R, Rose UM, Mulders JW,Grootenhuis PD 1998 Partially deglycosylated human choriogonadotropin,stabilized by intersubunit disulfide bonds, shows full bioactivity. Eur J Bio-chem 253:354–356

9. Matzuk MM, Keene JL, Boime I 1989 Site specificity of the chorionic gonad-otropin N-linked oligosaccharides in signal transduction. J Biol Chem 264:2409–2414

10. Sairam MR, Jiang LG 1992 Comparison of the biological and immunologicalproperties of glycosylation deficient human chorionic gonadotropin variantsproduced by site directed and chemical deglycosylation. Mol Cell Endocrinol85:227–235

11. Baril G, Remy B, Leboeuf B, Beckers JF, Saumande J 1993 Synchronizationof estrus in goats: the relationship between eCG binding in plasma, time ofoccurrence of estrus and fertility following artificial insemination. Therio-genology 40:621–628

12. Baril G, Remy B, Vallet JC, Beckers JF 1992 Effect of repeated use of pro-gestagen-PMSG treatment for estrus control in dairy goats out of breedingseason. Reprod Dom Anim 27:161–168

13. Roy F, Maurel MC, Combes B, Vaiman D, Cribiu EP, Lantier L, Pobel T,Deletang F, Combarnous Y, Guillou F 1999 The negative effect of repeatedequine chorionic gonadotropin treatment on subsequent fertility in alpinegoats is due to a humoral immune response involving the major histocom-patibility complex. Biol Reprod 60:805–813

14. Roy F, Combes B, Vaiman D, Cribiu EP, Pobel T, Deletang F, CombarnousY, Guillou F, Maurel MC 1999 Humoral immune response to equine chorionicgonadotropin in ewes: association with major histocompatibility complex andinterference with subsequent fertility. Biol Reprod 61:209–218

15. Lecompte F, Roy F, Combarnous Y 1998 International collaborative calibrationof a preparation of equine chorionic gonadotrophin (eCG NZY-01) proposedas a new standard. J Reprod Fertil 113:145–150

16. Akerstrom B, Bjorck L 1986 A physicochemical study of protein G, a moleculewith unique immunoglobulin G-binding properties. J Biol Chem 261:10240–10247

17. Guillou F, Martinat N, Combarnous Y 1985 Rapid in vitro desensibilizationof the testosterone response in rat Leydig cells by sub-active concentrations ofporcine luteinizing hormone. FEBS Lett 184:6–9

18. Hochereau-de-Reviers MT, Copin M, Seck M, Monet-Kuntz C, Cornu C,Fontaine I, Perreau C, Elsen JM, Moomarov O 1990 Stimulation of testos-terone production by PMSG injection in the ovine male: effect of breed and ageand application to males carrying or not carrying the “ F ” Booroola gene. AnimReprod Sci 23:21–32

19. Kelton CA, Cheng SVY, Nugent NP, Scheweickhardt RL, Rosenthal JL,Overton SA, Wands GD, Kuzeja JB, Luchette CA, Chappel SC 1992 Thecloning of the human follicle stimulating hormone receptor and its expressionin COS-7, CHO and Y1 cells. Mol Cell Endocrinol 89:141–151

20. Saumande J, Tamboura D, Chupin D 1985 Changes in milk and plasmaconcentrations of progesterone in cows after treatment to induice superovu-lation and their relationships with number of ovulations of embryos collected.Theriogenology 23:719–731

21. Jonsson U, Fagerstam L, Lofas S, Stenberg E, Karlsson R, Frostell A, MarkeyF, Schindler F 1993 Introducing a biosensor based technology for real-timebiospecific interaction analysis. Ann Biol Clin 51:19–26

22. Maurel MC, Ban E, Bidart JM, Combarnous Y 1992 Immunochemical studyof equine chorionic gonadotropin (eCG/PMSG): antigenic determinants on �-and �-subunits. Biochim Biophys Acta 1159:74–80

23. Sairam MR, Bhargavi GN, Downey BR 1992 Biological mimicry of gonado-tropin action by antiidiotypic antibodies to luteinizing hormone: character-ization and biological properties. Endocrinology 131:1212–1222

24. Nau EA, Epsaro EM, Scherer MS, Abel JH 1996 Monoclonal antibodies spe-cific for the binding site on the LH receptor alter progesterone production incultured granulosa cells. J Endocrinol 150:9–16

25. Dieleman SJ and Bevers MM 1987 Effects of monoclonal antibody againstPMSG administered shortly after the preovulatory LH surge on time andnumber of ovulations in PMSG/PG-treated cows. J Reprod Fertil 81:533–542

26. Holder AT, Aston R, Preece MA, Ivanyi J 1985 Monoclonal antibody-medi-ated enhancement of growth hormone activity in vivo. J Endocrinol 107:R9–R12

27. Aston R, Holder AT, Ivanyi J, Bomford R 1987 Enhancement of bovine growthhormone activity in vivo by monoclonal antibodies. Mol Immunol 24:143–150

28. Wang BS, Szewczyk E, Shieh HM, Hart IC 1990 Potentiation of the growth-promoting activity of porcine growth hormone (pGH) with an antibody gen-erated in rabbits to the peptide sequence pGH110–118. J Endocrinol 127:481–485

29. Holder AT, Aston R, Rest JR, Hill DJ, Patel N, Ivanyi J 1987 Monoclonal

302 Endocrinology, January 2004, 145(1):294–303 Herve et al. • eCG Modulation by Abs in Treated Goats

antibodies can enhance the biological activity of thyrotropin. Endocrinology120:567–573

30. Pell JM, Johnsson ID, Pullar RA, Morrell DJ, Hart I, Holder AT, Aston R 1989Potentiation of growth hormone activity in sheep using monoclonal antibod-ies. J Endocrinol 120:R15–R18

31. Chopineau M, Maurel MC, Combarnous Y, Durand P 1993 Topography ofequine chorionic gonadotropin epitopes relative to the luteinizing hormoneand follicle-stimulating hormone receptor interaction sites. Mol Cell Endocri-nol 92:229–239

32. Holder AT, Blows JA, Aston R, Bates PC 1988 Monoclonal antibody enhance-ment of the effects of human growth hormone on growth and body compo-sition in mice. J Endocrinol 117:85–90

33. Combarnous Y, Salesse R, Garnier J 1981 Physico-chemical properties ofpregnant mare serum gonadotropin. Biochim Biophys Acta 667:267–276

34. Combarnous Y 1992 Molecular basis of the specificity of binding of glyco-protein hormones to their receptors. Endocr Rev 13:670–691

35. Moyle WR, Campbell RK, Myers RV, Bernard MP, Han Y, Wang X 1994Co-evolution of ligand-receptor pairs. Nature 368:251–255

36. Chopineau M, Martinat N, Galet C, Guillou F, Combarnous Y 2001 �-Subunit102–104 residues are crucial to confer FSH activity to equine LH/CG but arenot sufficient to confer FSH activity to human CG. J Endocrinol 169:55–63

37. Sugino H, Bousfield GR, Moore WT, Ward DN 1987 Structural studies onequine glycoprotein hormones. Amino acid sequence of equine chorionicgonadotropin �-subunit. J Biol Chem 262:8603–8609

38. Bousfield GR, Liu WK, Sugino H, Ward DN 1987 Structural studies on equineglycoprotein hormones. Amino acid sequence of equine lutropin �-subunit.J Biol Chem 262:8610–8620

39. Matsui T, Sugino H, Miura M, Bousfield GR, Ward DN, Titani K, MizuochiT 1991 �-Subunits of equine chorionic gonadotropin and luteinizing hormonewith an identical amino acid sequence have different asparagine-linked oli-gosaccharide chains. Biochem Biophys Res Commun 174:940–945

40. Smith PL, Bousfield GR, Kumar S, Fiete D, Baenziger JU 1993 Equinelutropin and chorionic gonadotropin bear oligosaccharides terminating withSO4-4-GalNAc and Sia�2,3Gal, respectively. J Biol Chem 268:795–802

41. Varki A 1992 Diversity in the sialic acids. Glycobiology 2:25–4042. Merrick JM, Zadarlik K, Milgrom F 1978 Characterization of the Hanga-

nutziu-Deicher (serum-sickness) antigen as gangliosides containing n-glyco-lylneuraminic acid. Int Arch Allergy Appl Immunol 57:477–480

43. Fujii Y, Higashi H, Ikuta K, Kato S, Naiki M 1982 Specificities of humanheterophilic Hanganutziu and Deicher (H-D) antibodies and avian antiseraagainst H-D antigen-active glycosphingolipids. Mol Immunol 19:87–94

44. Krichevsky A, Birken S, O’Connor J, Acevedo H, Bikel K, Lustbader J,Hartree A, Canfield R 1994 Development, characterization, and application ofmonoclonal antibodies to the native and synthetic �COOH-terminal portionof human chorionic gonadotropin (hCG) that distinguish between the nativeand desialylated forms of hCG. Endocrinology 134:1139–1343

45. Sairam MR, Manjunath P 1983 Hormonal antagonist properties of chemicallydeglycosylated human choriogonadotropin. J Biol Chem 258:445–449

46. Sairam MR, Barghavi GN 1985 A role for glycosylation of the � subunit intransduction of biological signal in glycoprotein hormones. Science 229:65–67

47. Min KS, Hattori N, Aikawa J, Shiota K, Ogawa T 1996 Site-directed mu-tagenesis of recombinant equine chorionic gonadotropin/luteinizing hor-mone: differential role of oligosaccharides in luteinizing hormone- and follicle-stimulating hormone-like activities. Endocr J 43:585–593

48. Matzuk MM, Keene JL, Boime I 1989 Site specifity of the chorionic gonad-otropin N-linked oligosaccharides in signal transduction. J Biol Chem 264:2409–2414

49. Erbel PJ, Haseley SR, Kamerling JP, Vliegenthart JF 2002 Studies on therelevance of the glycan at Asn-52 of the �-subunit of human chorionic go-nadotropin in the �� dimer. Biochem J 364:485–495

50. Kumar TR, Wang Y, Lu N, Matzuk MM 1997 Follicle stimulating hormone isrequired for ovarian follicle maturation but not male fertility. Nat Genet15:201–204

51. Dierich A, Sairam MR, Monaco L, Fimia GM, Gansmuller A, LeMeur M,Sassone-Corsi P 1998 Impairing follicle-stimulating hormone (FSH) signalingin vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesisand hormonal imbalance. Proc Natl Acad Sci USA 95:13612–13617

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

Herve et al. • eCG Modulation by Abs in Treated Goats Endocrinology, January 2004, 145(1):294–303 303