Site of conjugation of bovine serum albumin to corticosteroid hormones and specificity of antibodies

The influence of the site of attachment of bovine serum albumin (BSA) to corticosteroids on the specificity of the antisera obtained in rabbits was investigated. The steroids and positions studied were cortisol and 11-desoxycortisol at C-3, C-6α, C-6β and C-21 and 21-desoxycortisol and C-21-desoxycortisone at C-6α and C-6β. None of the antisera to cortisol showed high specificity. Similar cross reactions with antisera derived from cortisol coupled at C-6β, C-3 and C-21 to BSA were observed. 11-Desoxycortisol coupled at C-6α to BSA yielded the most specific antisera to this adrenal hormone. Cross reactions of antisera derived from coupling the protein to the extreme ends (C-3 and C-21) of 11-desoxycortisol were similar. The orientation of the conjugate at C-6 in 21-desoxycortisol and in 21-desoxycortisone did not influence the relative specificity of the antisera derived from the epimers. Highly specific antibodies were obtained against 21-desoxy-cortisone. Except tor 15% cross reaction with 17-hydroxyprogesterone, the antibodies to 21-desoxycortisol were relatively specific. It was concluded that the site on the steroid molecule to which BSA is attached influences the specificity of the antisera produced but there are also other factors operative.     

Corticosteroid-calcium complexes

Glucocorticoids and calcium ions are shown to interact to yield a complex with properties that are distinct from those of the reactants. Reaction of steroids with Ca2+ appears to require the dihydroxyacetone side chain, since other structures do not react. Evidence for complex formation are: increased aqueous solubility of cortisol when Ca2+ is added to an aqueous or a biphasic aqueous/chloroform (or ethyl acetate) system; increased rate of migration of cortisol during reversed-phase thin-layer chromatography and HPLC; chromatographic comigration of 45Ca2+ and 3H-labeled cortisol; coprecipitation of 45Ca2+-3H-cortisol complexes. After dissociation of the cortisol-calcium complex, the only steroid recovered was cortisol. By the above criteria, the properties of cortisol were not affected by Sr2+, Ba2+, or Mg2+. The cleavage patterns of cortisol in the mass spectrometer corresponded to that of 11 beta-hydroxyandrostenedione when Ca2+ was present, and to cortisol in its absence. We therefore postulate that the structure of the dihydroxyacetone side chain was transiently altered by Ca2+, resulting in a labile C17-C20 bond. These results support our earlier proposal that the chemical and physico-chemical properties of corticosteroids are modified by calcium ions.     

Sex hormones and corticosteroids in pollen of Pinus nigra

In continuation of earlier investigations on steroid hormones in the pollen of pine, we have isolated and quantitatively determined the following steroids by radioimmunoassay and fluorimetric methods: testosterone, testosterone together with epitestosterone, respectively, and androstenedione; progesterone was determined by radioimmunoassay alone. In addition, we have tried to isolate cortisol, cortisone, 11-deoxycortisol, corticosterone and 11 -deoxycorticosterone and to characterize these compounds by specific reactions. The intensity of these reactions were used to estimate the amounts of corticosteroids in pollen.     

Asymmetric reduction of steroidal 20-ketones: Chemical synthesis of corticosteroid derivatives containing and 20α, 21-diol and 17α, 20α, 21-triol side chains

A method is presented for the chemical synthesis of corticosteroid derivatives containing the 20α, 21-diol and 17α, 20α, 21-triol side chains. The ketol side chains of cortisol, corticosterone, 11-deoxycortisol, and 11-deoxycorticosterone were reduced at C-20 with sodium borohydride in a two-phase system consisting of aqueous calcium chloride and an organic phase of chloroform or ethyl acetate. Stereoselectivity of reduction was 92% α-oriented for cortisol and 79% α-oriented for 11-deoxycortisol at ?27°. The 20α-form diminished relative to the 20β-form with increasing temperature. For the 17-deoxy steroids, reduction to the 20α-form was 23% for 11-deoxycorticosterone and 41% for corticosterone. The $\text{20α}\text{20β}$ ratios of 17-deoxy steroids were unchanged between 0° and ?27°. Calcium ions increased the solubility of corticosteroids in the aqueous phase. We propose that calcium ions affect the stereochemistry of reduction by forming a bidentate complex with the side chains of 17α-hydroxy steroids, fixing them in an orientation favorable to 20α-reduction, and by altering the phase partition of the steroids.     

运动员剧烈运动后血中应激免疫抑制蛋白的产生

我们曾经报道,大鼠或小鼠在束缚应激后血中产生了一种能抑制免疫功能的应激免疫抑制蛋白,（又称Ｎｅｕ－ｒｏｉｍｍｕｎｅｐｒｏｔｅｉｎ,ＮＩＰ,神经免疫蛋白）。本工作证明,运动员在大运动量的训练后血清中也产生一种能抑制淋巴细胞转化的物质,它的生化特性及分子量与前述大鼠和小鼠中的应激免疫抑制蛋白相同。在体外实验中,应激大鼠的血清培养人淋巴结细胞,获得了与大鼠实验相同的结果,即人淋巴结细胞也能产生应激免疫抑制蛋白。同时小鼠束缚应激的血清和大运动量的人类血清可以分别抑制人正常淋巴细胞和正常小鼠由ＣｏｎＡ诱导的淋巴细胞转化,以上结果表明,这种应激免疫抑制蛋白的种属特异性不强。     

Characterization of prednisone, prednisolone and their metabolites by gas chromatography—mass spectrometry

Human urinary metabolites of the synthetic corticosteroids prednisone and prednisolone were detected in the course of gas chromatographic steroid profiling as methoxime-trimethylsilyl derivatives. Metabolites were provisionaly identified by combined gas chromatography—mass spectrometry. The major metabolites were 11-keto/11-hydroxy conversion products, 20-hydroxy and 4,5-dihydro analogues of the parent drugs. Cortisone, 6-hydroxy and fully saturated A-ring compounds were minor metabolites. Retention indices and mass spectral data are presented.     

Apparent mineralocorticoid excess, pseudohypoaldosteronism, and urinary electrolyte excretion: toward a redefinition of mineralocorticoid action

Patients with apparent mineralocorticoid excess (AME) have low or absent activity of the enzyme 11 beta OH steroid dehydrogenase (11SD), and inappropriately high intrarenal levels of cortisol resulting in Na+ retention and hypertension. Pseudohypoaldosteronism (PHA), in contrast, is characterized by salt wasting despite hyperaldosteronemia, reflecting low or absent mineralocorticoid receptors (MR). Although AME is presumed to reflect inappropriate cortisol occupancy of MR, several features also suggest inappropriate occupancy of glucocorticoid receptors (GR). To test this possibility, we administered carbenoxolone, which is known to block 11SD, to four patients with PHA, and observed marked mineralocorticoid effects, e.g., antinatriuresis and elevated plasma bicarbonate. To further test the possibility that occupancy of renal GR may induce a classical mineralocorticoid response, we administered the highly specific glucocorticoid RU 28362 to adrenalectomized rats and showed that it has profound antinatriuretic effects. Finally, by selectively blocking MR with RU 28318 or GR with RU 38486, we have shown that corticosterone, the physiologic glucocorticoid in rats, has an antinatriuretic effect in adrenalectomized rats via either MR or GR occupancy. Previous studies have clearly shown that MR are inherently nonselective and have equivalent intrinsic affinity for aldosterone, corticosterone, and cortisol. The present studies suggest that this nonselectivity includes the nuclear response element to which either MR or GR may bind to elicit a mineralocorticoid effect, and further underscore the importance of the enzyme 11SD in the specific mineralocorticoid action of aldosterone.     

Urinary cortisol and cortisol metabolites in women with idiopathic hirsutism.

Urinary unconjugated cortisol, total 17-oxogenic steroids and cortisol metabolites (THF, allo-THF, THE, cortol and cortolone) were measured in 90 normal women and in 90 women with idiopathic hirsutism of comparable age group. After carrying out Student's "t"-test, the mean steroid excretion values in hirsute women were significantly higher than those from normal females. The results indicate that, due to stress, these hirsute women do have altered adrenocortical function, as assessed by the estimation of these corticosteroids of which urinary unconjugated cortisol was found to be the most sensitive index.     

Suppression of mitogen- and tumor-cell-induced lymphocyte stimulation by tumor-associated fetal antigens

The ability of tumor-associated fetal antigens (TAFA) to suppress mitogen and tumor-cell-induced blastogenesis was investigated. Three different TAFA (I, III, and IV) were tested in PHA and Con A lymphocyte proliferation assays. Spleen cells from New Zealand black rats (NBR) were fractionated over nylon wool; and nonadherent (NA) and adherent (AD) cells were compared with unfractionated (UF) cell responses. Preincubation of NA cells with TAFA-I was followed by addition of phytohemagglutinin (PHA) elicited suppression in a 3-to 4-day assay. UF cells were unresponsive to TAFA and/or PHA at all concentrations tested. TAFA dose—response titration curves in Con A proliferation assays revealed that all TAFA tested (TAFA I, -III, and -IV from fibrosarcomas; TAFA-I and -III from osteosarcomas) were suppressive. For some TAFA, nanogram quantities produced significant suppression. In mixed leukocyte tumor cell assays (MLTC) both UF and NA normal rat spleen cells were tested for proliferative responses to syngeneic tumor cells. Four distinct TAFA, purified by high-pressure liquid chromatography, suppressed lymphocyte proliferation when added to MLTC cultures in 5-day assays. Specificity experiments demonstrated that trinitrophenol-bovine serum albumin did not produce similar immunosuppression. TAFA did not block recognition of tumor antigen nor produce nonspecific cytotoxicity of the spleen cells. Significant suppression of DNA synthesis was produced by TAFA-1 following cocultivation with spleen and tumor cells for 1, 2, and 3 days, compared to no suppression when spleen and tumor cells were washed free of TAFA-I prior to tumor cell addition at Day 0. Similar experiments using rat embryo fibroblasts (REF) as stimulators demonstrated that pre-REF cocultivation treatment of lymphocytes with TAFA-I significantly reduced subsequent blastogenic responses. This effect was not reversible; however, if TAFA-I was added to responders previously stimulated by REF, a suppressive response was not seen. Experiments were also carried out to determine the reversibility of TAFA-I effects. Cells were treated with TAFA-I from 1 to 5 days, followed by determination of lymphocyte blastogenesis. TAFA-I effects are reversible and antigen presence is required to completely suppress (or inhibit) stimulation by tumor cells.     

Spectrophotometric study of cytochrome P-450 (11 beta) interaction with physiological effectors

Using the optical absorbance spectroscopy method, the interaction of a number of biospecific ligands (steroids, adrenodoxin) with homogeneous cytochrome P-450 (11 beta) from bovine adrenal mitochondria was investigated. The parameters of the steroid-protein interaction in a number of substrates and products of the 11 beta- and 18 (19)-hydroxylation with the active site of cytochrome P-450 (11 beta) were determined. A sharp decrease in the cytochrome affinity for steroids upon the insertion of the first hydroxy group was observed, which provides for a predominant formation of monohydroxylated products from the substrate and minimum amounts of dihydroxylated ones, despite the presence of more than one position for the substrate hydroxylation by cytochrome P-450 (11 beta). Some structural elements of the steroid molecule were determined as any alterations in these strongly affect the enzyme affinity for the steroid. These structures are: 1) delta 4-3-oxo structure; 2) either 21-hydroxy group of pregnen steroids or the one fulfilling its functions, 17 beta-hydroxy or 17-oxo group of androsten steroids, and 3) the 11th position of all the substrates under study. It was shown that the binding of various substrates into stoichiometric (1:1) steroid-protein complexes provides a transition to high spin state from 30-40% (cortisol, corticosterone) to 90-95% (11-deoxycorticosterone) of hemoprotein iron. Using the experimental system containing individual cytochrome P-450 (11 beta) and adrenodoxin, as well as the steroid and nonionic detergent Tween 20, it was shown that the parameters of substrate binding and hemoprotein spin equilibrium did not differ from the corresponding parameters of the cytochrome-adrenodoxin dienzyme complex. The peculiarities of the multiligand interactions in the 11 beta-hydroxylase system, involving cytochrome, substrates and ferredoxin demonstrate some analogy with a bacterial camphor hydroxylase system and some differences from the mitochondrial system for the side chain cleavage of cholesterol.     

Synthesis and structure-activity relationships of a series of increasingly hydrophobic cationic steroid lipofection reagents

The use of cholesterol-based cationic lipids and the ability of glucocorticoids to reduce local inflammatory response to lipoplexes motivated an investigation of structure-activity relationships for cationic steroids. A one-step synthetic scheme using iminothiolane was developed to link spermine to the 21-OH position of steroids via an amidine linkage. Five steroids (cortisol, dexamethasone, corticosterone, 11-deoxycortisol, and 11-deoxycorticosterone) with increasing hydrophobicity of the parent steroid (Log P(ster) from 1.51 to 3.01) were conjugated with spermine, formulated with dioleoylphosphatidylethanolamine (DOPE) at DOPE : steroid mole ratios (R) of R = 0.5 to 2, and then complexed with 1 microg enhanced green fluorescent protein (EGFP) plasmid DNA at charge ratios (CR) = 2 to 24 amines per phosphate (0.5 to 6 steroids per phosphate). The resulting 105 different formulations of the cationic steroid series were used to lipofect bovine aortic endothelial cells. Transgene expression data at either 24 or 48 h post-lipofection for all formulations was collapsed onto master curves when plotted against a single empirical dimensionless parameter, the lipofection index (LI) = CR (Log P(liposome))(Log P(ster)/|DeltaLog P|) [R/(R + 1)] where DeltaLog P = Log P(DOPE)- Log P(ster) and Log P(liposome) is a mole-weighted average of the DOPE/cationic steroid liposome hydrophobicity. For 7 < LI < 29, the EGFP expression at 24 or 48 h post-lipofection increased linearly with LI (EGFP approximately 0 for LI < 7), but did not increase further for LI > 29, thus providing a predictive design rule based on Log P of the hydrophobic moiety of the cationic steroid lipid.     

HIV inhibits the early steps of lymphocyte activation, including initiation of inositol phospholipid metabolism

Mechanisms accounting for HIV-associated suppression of lymphocyte proliferation were investigated. In previous work we demonstrated that purified and inactivated HIV-suppressed lymphoid cell proliferation. In this report we used an inactivated preparation of HIV obtained from infected CEM cells grown in serum free media and demonstrated that this HIV-associated suppression acted in the early steps of activation to inhibit the incorporation of radiolabeled phosphorus into phosphatidylinositol 4,5-bisphosphate and phosphatidic acid. Initially we showed that both purified CD4 and CD8 T lymphocyte subsets were affected and HIV-associated inhibition did not require the CD4 molecule. Impaired lymphocyte blastogenesis (decreased size and granularity and decreased expression of receptors to IL-2 and transferrin) in response to PHA indicated an effect of inactivated HIV on the early steps of activation. This was confirmed by time studies where 1) a 2 min HIV-pretreatment followed by washing before stimulation was sufficient to inhibit PHA induced proliferation of normal lymphocytes, and 2) addition of HIV to PHA prestimulated lymphocytes failed to inhibit proliferation, e.g., there was no effect on preactivated lymphocytes. HIV was mainly inhibitory of lymphocyte proliferation induced by PHA or mAb to the CD3 receptor. In contrast to the effect on the CD3/TiR, responses via the CD2 receptor were not suppressed, e.g., stimulation with the monoclonal antibodies T11(2) + T11(3). Inasmuch as responses by direct A23187 + PMA stimulation of intracellular pathways were also inhibited, it appears that the HIV-induced defect was not (or not only) membrane receptor mediated. The earliest (min) measurable event after stimulation was the initial increase in intracellular Ca2+ which was unaffected by HIV pretreatment. The next measurable event (min to h) of stimulation is a sustained increase in inositol phospholipid turnover. Pretreatment of mononuclear cells with inactivated HIV resulted in a decreased inositol phospholipid turnover as judged from decreased 32P incorporation into phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate. This led to decreased generation of DAG as reflected in the reduced radiolabeling of its metabolite PA. Reduced availability of DAG presumably interferes with pkC activation and leads to decreased expression of receptors for IL-2 and transferrin and impaired proliferation.     

Modulation of the adherence of human lymphocytes to measles-infected cells by corticosteroids

Corticosteroids have been shown to have an inhibitory effect on many aspects of the human inflammatory and immunologic response. It is largely unknown by what mechanism corticosteroids alter the biologic function of lymphoid cells. One possible mechansim by which steroids could alter lymphocyte function is by modulation of surface receptors. The present study was designed to examine this possibility by the study of the effect of hydrocortisone (HC) and other steroids on the lymphocyte surface binding site for measles virus antigens. These receptors were detected by an assay involving the rosetting of lymphocytes to measles-infected Hela cells (Hela-K11). We showed that HC and several other steroids inhibit the adherence of monocyte-depleted peripheral blood lymphocytes to Hela-K11 cells (HC inhibition, 34 ± 12.1%, mean ± 1 SD). Inhibition could only be observed after the removal of glass-adherent cells. Modulation of lymphocyte functional surface determinants may be important in explaining some of the pharmacologic effects of corticosteroids on the immune system.     

Does season alter responsiveness of the reproductive neuroendocrine axis to the suppressive actions of cortisol in ovariectomized ewes?

Season can profoundly influence activity of the hypothalamic-pituitary-adrenal axis and alter reproductive neuroendocrine responsiveness to stress and gonadal steroids. Here we tested the hypothesis that the inhibitory effect of a stress-like increment in plasma concentration of the adrenal steroid cortisol on pulsatile LH secretion varies with season. LH pulse patterns were monitored prior to and during the administration of cortisol in the same seven ovariectomized ewes during three stages of the yearly breeding cycle: breeding season, transition to anestrus, and midanestrus. The elevation in cortisol mimicked the rise in plasma level of cortisol in response to an immune/inflammatory stress. During all three seasons, cortisol acutely suppressed the pulsatile release of LH. This inhibition reflected a marked reduction of LH pulse amplitude and a minimal suppression of LH pulse frequency. Of interest, the suppressive effect of this physiologic increment in cortisol did not vary across seasons. This provides initial evidence that, in ovariectomized ewes, cortisol-induced suppression of pulsatile LH secretion differs from that of gonadal steroids in that it is not profoundly influenced by season.     

Human T cell activation by OKT3 is inhibited by a monoclonal antibody to CD44.

The CD44 molecule, also known as Hermes lymphocyte homing receptor, human Pgp-1, and extracellular matrix receptor III, has been shown to play a role in T cell adhesion and activation. Specifically, anti-CD44 mAb block binding of lymphocytes to high endothelial venules, inhibit T cell-E rosetting, and augment T cell proliferation induced by the CD2 or CD3-TCR pathways. We have characterized an anti-CD44 mAb (212.3) which immunoprecipitates a 90-kDa protein and is specific for CD44 as shown by peptide mapping and antibody competition studies. Interestingly, our studies with 212.3 demonstrate that this CD44-specific mAb completely inhibits T cell proliferation stimulated by the anti-CD3 mAb, OKT3. Inhibition is not a result of reduced cell viability, but is associated with 1) inhibition of IL-2 production, 2) inhibition of IL-2R expression, and 3) inhibition of OKT3-mediated increases in intracellular Ca2+ levels. In addition, 212.3 does not inhibit proliferation by the T cell mitogens PHA or PWM nor does it inhibit proliferation in a mixed lymphocyte reaction. Similar to other anti-CD44 mAb, 212.3 also augments T cell proliferation induced by mAb directed against the T11(2) and T11(3) epitopes of CD2. Thus, these studies describe a novel CD44-specific mAb (212.3) that inhibits T cell activation by OKT3 by blocking early signal transduction. Furthermore, these studies suggest that "receptor cross-talk" between the CD3-TCR complex and CD44 may regulate T cell activation.     

Steroid inhibition of calcitriol-induced prolactin production in GH4C1 cells. Specificity and sensitivity.

The induction of prolactin (PRL)-gene expression by calcitriol (1,25-dihydroxyvitamin D3, 1,25-dihydroxycholecalciferol) in clonal rat pituitary tumour (GH4C1) cells was selectively inhibited by cortisol [IC50 (concentration causing 50% inhibition) = 3.2-4.1 nM]. The steroid specificity of this effect was investigated and various steroids were found to inhibit calcitriol-stimulated PRL production with the following relative potencies: cortisol, 1; dexamethasone, 8; 11-deoxycortisol, 0.5; corticosterone, 0.4; aldosterone, 0.07; testosterone and oestradiol, less than 0.003. The steroid antagonist RU 38486 did not affect basal or calcitriol-stimulated PRL production, but antagonized the effect of 10 nM-cortisol in a concentration-dependent manner. Neither progesterone nor 11-deoxycortisol antagonized the effect of 10 nM-cortisol. Calcitriol-induced PRL production was 14 times more sensitive to dexamethasone inhibition than was non-stimulated PRL production. Growth-hormone production was stimulated by dexamethasone, in the presence or absence of calcitriol, with a concentration-dependence similar to that of dexamethasone inhibition of basal PRL production. These data indicate that steroid inhibition of calcitriol-stimulated PRL production is a specific glucocorticoid effect. The sensitivity of calcitriol-stimulated PRL production to dexamethasone was 14-26-fold greater than that of other measured responses in the same cells. Two of the possible explanations for this selectively increased sensitivity to glucocorticoids are: amplification of the glucocorticoid effect via an induced mediator; and the presence of very-high-affinity glucocorticoid-receptor-binding sites on DNA.     

Metabolic effects of C21 steroids in female Epinephelus akaara (Teleostei: Serranidae)

Cortisol, 11-deoxycorticosterone, progesterone, 17 alpha-hydroxy-progesterone, 17 alpha-hydroxy-20 beta-dihydroprogesterone, or a combination of the last 2 steroids, were injected into different groups of immature female Epinephelus. None of the steroids tested had significant effects on serum electrolyte level, and hepatosomatic and gonadosomatic indices. Serum glucose concentration was elevated after treatment with cortisol, 11-deoxycorticosterone, 17 alpha-hydroxy-20 beta-dihydroprogesterone or a combination of 17 alpha-hydroxy-20 beta-dihydroprogesterone and 17 alpha-hydroxyprogesterone. Muscle protein concentration was lowered after treatment with cortisol, 17 alpha-hydroxy-20 beta-dihydroprogesterone or a combination of 17 alpha-hydroxyprogesterone and 17 alpha-hydroxy-20 beta-dihydroprogesterone. Liver protein was significantly elevated after treatment with progesterone but lowered after cortisol treatment. The results suggest that oocyte maturation is an energy consuming process, and that steroid hormones regulating these processes, including 17 alpha-hydroxy-20 beta-dihydroprogesterone and 11-deoxycorticosterone adjust metabolism to provide energy for these processes.     

Regulation of lymphocyte proliferation by uterine serpin: interleukin-2 mRNA production, CD25 expression and responsiveness to interleukin-2

During pregnancy, the endometrium of the ewe secretes large amounts of a progesterone-induced protein of the serpin superfamily of serine proteinase inhibitors called ovine uterine serpin (OvUS). This protein inhibits lymphocyte proliferation in response to concanavalin A (ConA), phytohemagglutinin (PHA), or mixed lymphocyte reaction. The purpose of these experiments was to characterize the mechanism by which OvUS inhibits lymphocyte proliferation. Ovine US caused dose-dependent inhibition of lymphocyte proliferation induced by phorbol myristol acetate (PMA), an activator of protein kinase C. The PHA-induced increase in CD25 expression was inhibited in peripheral blood mononuclear leukocytes (PBML) by OvUS. However, no effect of OvUS on Con A-induced expression of CD25 was observed. Further analysis using two-color flow cytometry revealed that OvUS inhibited ConA-induced expression of CD25 in gammadelta-TCR- cells but not gammadelta-TCR+ cells. Stimulation of PBML for 14 hr with ConA resulted in an increase in steady state amounts of interleukin-2 (IL-2) mRNA that was not inhibited by OvUS. Ovine US was also inhibitory to lymphocyte proliferation induced by human IL-2. Results suggest that OvUS acts to inhibit lymphocyte proliferation by blocking the upregulation of the IL-2 receptor and inhibiting IL-2-mediated events. Lack of an effect of OvUS on ConA-stimulated CD25 expression in gammadelta-TCR+ cells may reflect a different mechanism of activation of these cells or insensitivity to inhibition by OvUS.     

Social interactions and cortisol treatment increase the production of aggressive electrocommunication signals in male electric fish,Apteronotus leptorhynchus

Brown ghost knife fish, Apteronotus leptorhynchus, continually emit a weakly electric discharge that serves as a communication signal and is sensitive to sex steroids. Males modulate this signal during bouts of aggression by briefly (approximately 15 ms) increasing the discharge frequency in signals termed "chirps." The present study examined the effects of short-term (1-7 days) and long-term (6-35 days) male-male interaction on the continuous electric organ discharge (EOD), chirping behavior, and plasma levels of cortisol and two androgens, 11-ketotestosterone (11KT) and testosterone. Males housed in isolation or in pairs were tested for short-term and long-term changes in their EOD frequency and chirping rate to standardized sinusoidal electrical stimuli. Within 1 week, chirp rate was significantly higher in paired fish than in isolated fish, but EOD frequency was equivalent in these two groups of fish. Plasma cortisol levels were significantly higher in paired fish than in isolated fish, but there was no difference between groups in plasma 11KT levels. Among paired fish, cortisol levels correlated positively with chirp rate. To determine whether elevated cortisol can cause changes in chirping behavior, isolated fish were implanted with cortisol-filled or empty Silastic tubes and tested for short-term and long-term changes in electrocommunication signals and steroid levels. After 2 weeks, fish that received cortisol implants showed higher chirp rates than blank-implanted fish; there were no difference between groups in EOD frequency. Cortisol implants significantly elevated plasma cortisol levels compared to blank implants but had no effect on plasma 11KT levels. These results suggest that male-male interaction increases chirp rate by elevating levels of plasma cortisol, which, in turn, acts to modify neural activity though an 11KT-independent mechanism.