Corticotrophin releasing hormone: its potential for a role in human myometrium.

Aside from its role as a hypothalamic stress hormone, corticotrophin releasing hormone (CRH) is also a placental hormone, at least in primates. Although the function of placentally derived CRH remains to be fully elucidated, elevated CRH levels have been associated with premature labour, suggesting that the hormone may be involved in regulating the duration of pregnancy. Indeed, pregnant human myometrium expresses functional CRH receptors (CRH R1 and CRH R2 subtypes) thought to signal predominantly via the second messenger cAMP. Thus, like other cAMP-producing hormones in the myometrium such as beta(2) agonists, CRH may play a part in maintaining uterine quiescence. However, several of the CRH receptor isoforms identified to date have a reduced ability to activate adenylate cyclase, raising the question as to whether they are linked to other signal transduction pathways. Here, we discuss critically the evidence for the peptide's role in regulating contractility, both directly at the myometrium and indirectly via the fetal membranes and decidua. The possibility of a role in myometrial growth modulation is also described. Experimental Physiology (2001) 86.2, 273-281.

Characterization of adenylyl cyclases in cultured human granulosa cells.

Granulosa cells play an essential role in follicular development and formation of corpora lutea. Many functions of granulosa-lutein cells are controlled by activation of G protein-coupled receptors and the formation of cyclic AMP (cAMP) by adenylyl cyclase. There are at least nine mammalian adenylyl cyclase isoenzymes, which show different sensitivities towards other signalling systems. The aim of this study was to identify the types of adenylyl cyclase present in human granulosa cells and to investigate its functional regulation by G proteins, calcium and the protein kinase C and A pathways. Granulosa cells were obtained from women undergoing IVF. The cells were maintained in primary culture and they consistently expressed mRNA coding for adenylyl cyclase I, III, VI, VII and IX. The signals for adenylyl cyclase V and VIII were more variable among patients and there was no signal for adenylyl cyclase II. The expression of multiple adenylyl cyclase proteins was confirmed by immunochemistry with subtype-specific antibodies. The formation of cAMP in cultured cells was stimulated many times by hCG (EC(50) value 4.2 iu ml(-1)) and by prostaglandin E(2) (PGE(2); EC(50) = 0.75 micromol l(-1)) in a concentration-dependent manner, thus confirming the presence of receptors coupled positively to G(s). The diterpene forskolin, which stimulates all isoforms of adenylyl cyclase except for adenylyl cyclase IX, increased cAMP formation to higher levels than hCG or PGE(2). The strong stimulation by forskolin indicates that adenylyl cyclase IX is unlikely to be the major source of cyclase activity in these cells. Basal and forskolin- or PGE(2)-stimulated adenylyl cyclase activity was amplified 1.5-2.0 times by phorbol-12,13-dibutyrate, indicating that protein kinase C-sensitive enzymes (for example, adenylyl cyclase types IV, V, VI or VII) may be active in the cells. In contrast, hCG-stimulated activity was inhibited (76 +/- 6%) by phorbol ester. Stimulation of G(i) with the alpha-adrenoceptor agonist clonidine inhibited hCG-induced cyclase activity. This finding indicates that adenylyl cyclase II and IV subtypes, which are stimulated by betagamma subunits released from G(i), are not predominant. Increases in intracellular free calcium concentrations by the ionophore A23187, the calcium-ATPase inhibitor thapsigargin or by fluprostenol, a selective prostanoid FP receptor agonist, which is known to open calcium channels in granulosa cells, or removal of calcium by EGTA, had no significant effects on basal or forskolin-stimulated formation of cAMP. These results indicate that subtypes adenylyl cyclase I, III and VIII, which are activated by calcium, and adenylyl cyclase V and VI, which are inhibited by calcium, are not dominant isoforms in granulosa-lutein cells. The protein kinase A inhibitor H89 had no effects on formation of cAMP; this finding rules out the involvement of adenylyl cyclase V and VI subtypes, which are subjected to negative feedback by protein kinase A. These results indicate that adenylyl cyclase VII is the dominant functional isoenzyme in human granulosa-lutein cells.

Desensitization of oxytocin receptors in human myometrium.

In the present study, we investigated the possible mechanisms by which oxytocin might regulate oxytocin receptor (OTR) density. Exposure of cultured myometrial cells to oxytocin for a prolonged period caused desensitization: the steady-state level of oxytocin binding was 210 x 10(3) binding sites/cell, but this was time-dependently reduced to 20.1 x 10(3) sites/cell by exposing the cells to oxytocin for up to 20 h. In contrast, Western blotting data showed that the total amount of OTR protein was not affected by oxytocin treatment for up to 24 h. Flow cytometry experiments demonstrated that OTRs were not internalized during this treatment. However, RNase protection assays and Northern analysis showed that in cultured myometrial cells OTR mRNA was reduced by oxytocin treatment to reach a new low steady-state concentration. Analysis of this mRNA in myometrial biopsies from 17 patients undergoing emergency Caesarean section showed how it decreased with advancing labour. Samples obtained after 12 h of labour contained approximately 50 times less OTR mRNA than samples obtained from patients in labour for less than 12 h. We speculate that this decrease in OTR mRNA represents in-vivo OTR desensitization.

The desensitization of oxytocin receptors in human myometrial cells is accompanied by down-regulation of oxytocin receptor messenger RNA.

We have recently provided evidence for the desensitization of oxytocin receptors in human myometrial cells. In the present study, we have investigated the possible mechanisms by which oxytocin (OT) might regulate OT receptor density. The steady state level of OT binding in cultured myometrial cells was 220 x 10(3) binding sites/ cell, but this was time-dependently reduced to 27 x 10(3) sites/cell by exposure to OT for up to 20 h. Similarly, OT exposure decreased the binding of OT to cell membranes. In contrast, Western blotting data showed that the total amount of OT receptor protein was not affected by OT treatment for up to 48 h. Flow cytometry experiments demonstrated that OT receptors are not internalized during prolonged exposure of the cells to OT. However, RNase protection assays and Northern analysis showed that OT receptor mRNA was reduced by OT treatment to reach a new low steady state level with a time course similar to that of the disappearance of cell surface OT binding sites. Possible mechanisms involved in mRNA down-regulation include transcriptional suppression and destabilization of mRNA by RNA binding proteins.

Prostaglandin E receptors in myometrial cells.

Prostaglandins (PGs) exert their effects via binding to specific cell surface receptors and influencing second messenger systems through G-proteins. PGE2 may interact with at least four receptor subtypes (EP1, EP2, EP3, EP4), each showing different pharmacological profiles. The second messengers calcium, inositol phosphates (InsPs) and cyclic nucleotides play decisive roles in uterine contractility. The question in this investigation was, which EP receptors, G-proteins and second messenger systems transmit PGE2 induced signals in human myometrium. We have measured changes in InsPs and cAMP formation and also in intracellular calcium concentration ([Ca2+]i) induced by PGE2 and receptor subtype selective analogues in cultured human myometrial cells. PGE2 increased cAMP level and this effect was shared by the EP2 receptor subtype selective agonist Butaprost and by Misoprostol (EP3 > EP2 > EP1). Sulprostone (EP3 > EP1) did not stimulate adenylyl cyclase activity per se, but inhibited forskolin-stimulated adenylyl cyclase in a pertussis toxin (PT) sensitive way. PGE2, GR63799X (EP3 selective), Sulprostone and Misoprostol activated phospholipase-C (PLC), this effect was resistant to PT treatment. PGE2 also elevated [Ca2+]i from the resting level of 60-90 nM up to 350 nM. Low concentrations (1-300 nM) of PGE2 increased [Ca2+]i without PLC activation. The selective EP1 inhibitor AH6809, Nifedipine, Verapamil and PT treatment inhibited this effect of PGE2. In cultured human myometrial cells PGE2 interacts with EP1 receptors, which elevate [Ca2+]i independently from PLC, but involving a Gi protein and plasmamembrane calcium channels; EP2 receptors which stimulate adenylyl cyclase; EP3A receptors, which inhibit adenylyl cyclase activity through Gi activation and EP3D receptors which activate PLC through a PT-insensitive pathway and also elevate [Ca2+]i.

Activation of the prostaglandin FP receptor in human granulosa cells.

Prostaglandin F2 alpha (PGF2 alpha) has regulatory (mainly luteolytic) effects in the ovary but the mechanism of action is not completely understood. Reverse transcriptase-polymerase chain reaction (RT-PCR) techniques were used to demonstrate the presence of mRNA encoding the PGF2 alpha receptor (FP receptor) in human granulosa-lutein cells. Specific primers for the amplification of cDNA were designed and yielded a single product of 696 bp corresponding to the FP receptor. The identity of this product was verified by sequencing. Fluprostenol, a selective FP receptor agonist, activated phospholipase C (PLC) and increased intracellular free calcium concentration, confirming the functional activation of the receptor. We have demonstrated by Western blotting that granulosa cells express PLC-beta and PLC-gamma isoforms. The cells responded to pervanadate with increased PLC activity and increased tyrosine phosphorylation, demonstrating a functional PLC-gamma tyrosine kinase pathway. However, fluprostenol did not provoke any detectable tyrosine phosphorylation. Moreover, the effect of fluprostenol was inhibited through protein kinase C stimulation by phorbol 12, 13-dibutyrate, and was not affected when cells were treated with phenylarsine oxide, which blocks tyrosine phosphorylation. These results suggest that the FP receptor activates PLC-beta rather than PLC-gamma isoforms. Fluprostenol-induced activation was pertussis toxin resistant. Granulosa cells express G proteins of the Gq family (resistant to pertussis toxin) and mRNA for both G alpha q and G alpha 1 l has been identified by RT-PCR. In conclusion, human granulosa cells have a functional FP receptor the effects of which are mediated through PLC-beta activation probably via Gq/1 l.

Fluprostenol activates phospholipase C and Ca2+ mobilization in human myometrial cells.

The objective of this study was to investigate the mechanism of action of PGF2 alpha in cultured human myometrial cells. We measured the effects of PGF2 alpha and fluprostenol, a selective PGF2 alpha receptor (FP receptor) agonist, on phospholipase C(PLC) activation, on changes in the intracellular free calcium concentration ([Ca2+]i), and on protein tyrosine phosphorylation. PGF2 alpha and fluprostenol activated PLC (determined by measuring the formation of inositol phosphates) and increased [Ca2+]i in a concentration-dependent manner. The apparent affinity of the FP receptor for fluprostenol was higher than that for PGF2 alpha when measuring PLC activation, but the receptor displayed similar affinities for both agonists when measuring increases in [Ca2+]i. These effects were not altered by treating the cells with pertussis toxin (PT), suggesting that the FP receptor is linked to PLC activation by a G protein of the Gq family. By contrast, the effect of oxytocin on PLC activation involved both PT-resistant and PT-sensitive pathways. Human myometrial cells responded to pervanadate and epidermal growth factor with increased PLC activity and increased tyrosine phosphorylation, demonstrating a functional PLC-gamma tyrosine kinase pathway. However, neither fluprostenol nor oxytocin stimulated tyrosine phosphorylation, but the effects of both agonists were inhibited after protein kinase C stimulation. These data suggest that fluprostenol and oxytocin activate PLC-beta rather than PLC-gamma isoforms. The effect of fluprostenol is Ca2+ dependent, but is unlikely to involve a direct effect of Ca2+ on PLC activity.

Prostaglandin E2 activates phospholipase C and elevates intracellular calcium in cultured myometrial cells: involvement of EP1 and EP3 receptor subtypes.

PGE2 is a powerful modulator of uterine contractility, but there is uncertainty as to which receptor subtypes (EP1, EP2, EP3, or EP4), G proteins, and second messenger systems are activated by PGE2 in myometrium. Here we show that in cultured human myometrial cells, PGE2 (1-100 microM) activates phospholipase C (PLC) up to 500% over the control level and elevates intracellular calcium ([Ca2+]i) from the resting level of 60-90 nM up to 350 nM in a concentration-dependent manner. Stimulation by the receptor subtype-selective analogs GR63799X (EP3), sulprostone (EP3 > EP1), and misoprostol (EP3 > EP2 > EP1) indicates that these effects are transmitted through EP3 receptors. Both effects are resistant to pertussis toxin (PT). Lower concentrations of PGE2 (1-300 nM) increase [Ca2+]i via a PT-sensitive pathway, without PLC activation. This [Ca2+]i increase occurs after an inverse dose-related delay and is inhibited by the selective EP1 antagonist AH6809 and calcium channel blockers. By comparison, oxytocin stimulates PLC up to 1000% over the control level and elevates [Ca2+]i up to 800 nM in a concentration-dependent manner without any measurable delay; both effects are partly sensitive to PT. These data provide functional evidence for the presence of different stimulatory mechanisms for PGE2 in myometrium: 1) a low affinity receptor (probably EP3D) that activates PLC through a PT-insensitive pathway; and 2) a high affinity receptor (probably EP1), independent from PLC and involving a PT-sensitive G protein (G(i)?). Both pathways lead to elevation of [Ca2+]i.

G protein expression and second messenger formation in human granulosa cells.

The expression of heterotrimeric (alpha beta gamma subunits) GTP-binding regulatory proteins (G proteins) and the activation of G protein-linked receptors in human granulosa cells were investigated. The cells were obtained from stimulated follicles in women undergoing in vitro fertilization and were cultured in serum-supplemented medium. Immunoblotting with specific antibodies showed that granulosa cell membranes express alpha s, alpha i3 alpha i1,2, alpha q,11 and beta subunits. Three antibodies against alpha o failed to detect this protein. The cells responded to hCG and to prostaglandin E2 with a dose-dependent increase in cAMP formation, confirming the functional activation of G alpha s. The alpha 2 adrenoceptor agonist, clonidine, inhibited hCG-stimulated cAMP formation and this effect was blocked with pertussis toxin, thus involving a Gi-type protein, most likely G alpha i2. Oxytocin provoked an increase in formation of inositol phosphates and intracellular calcium concentration, which was partly pertussis toxin resistant, providing evidence of G alpha q,11 activation. However, a significant component of the response to oxytocin could be blocked by pertussis toxin, indicating Gi-mediated phospholipase C activation (by either alpha i or beta gamma subunits). These data demonstrate the presence of G proteins in granulosa cells and suggest a complex regulation of hormonal signalling. The concentration of cAMP in these cells depended on the balance of G alpha s:G alpha i activation, whereas activation of the inositol phospholipid pathway and rises in intracellular calcium involved both Gq,11 and Gi pathways.

Parturition: activation of stimulatory pathways or loss of uterine quiescence?

Parturition results from the establishment of phasic regular uterine contractions. Contractility in myometrial smooth muscle is stimulated by an increase in intracellular calcium ([Ca2+i]) which activates myosin light chain phosphorylation leading to increased myosin ATPase activity and enhanced rate of acto-myosin cross bridge formation. G proteins play a pivotal role in smooth muscle activation and relaxation by coupling cell membrane receptors to effector enzymes and ion channels. G alpha(s) and G alpha(i) stimulate and inhibit adenylyl cyclase, respectively and control cAMP formation. G alpha(q) stimulates phospholipase C resulting in the formation of two second messengers: inositol 1,4,5-trisphosphate (InsP3) which releases Ca2+ from the sarcoplasmic reticulum, and 1,2-diacylglycerol which activates protein kinase C. The oxytocin receptor stimulates myometrial contractility by increasing [Ca2+i] through both pertussis toxin resistant (G alpha(q)) and pertussis toxin sensitive (?G alpha (i)) pathways. beta-Adrenoceptors and prostaglandin EP2 receptors promote relaxation via G alpha(s)-adenylyl cyclase. The concentration of myometrial oxytocin receptors is five-times higher in pregnant compared to non-pregnant myometrium but decreases in samples obtained during labour. When myometrial slices are challenged with oxytocin there is a rapid increase in InsP3 levels with a time course which is similar to the rise in [Ca2+i] provoked by oxytocin in cultured myometrial cells. The formation of InsP3 in response to oxytocin in myometrial tissue at term is similar in samples obtained before and after the onset of labour. G alpha(q) and G alpha(i) are expressed at similar levels in non-pregnant and in pregnant myometrium obtained before or during labour. By contract, G alpha(s) levels are higher in pregnant compared to non-pregnant myometrium and decrease in samples obtained during labor. These changes in G alpha(s) are paralleled by prostaglandin E2-induced adenylyl cyclase activity in the same tissues. Parturition may be the consequence of downregulation of pathways that favour uterine quiescence by increasing cAMP formation, resulting in a relative dominance of stimulatory receptors that increase InsP3/Ca2+ availability.

Effect of oxytocin antagonists on the activation of human myometrium in vitro: atosiban prevents oxytocin-induced desensitization.

OBJECTIVE: Our purpose was to investigate whether the sensitivity of myometrial cells to oxytocin is affected by prolonged exposure to oxytocin antagonists. STUDY DESIGN: Tissue slices or cultured myometrial cells were exposed to peptides in vitro. Myometrial activation was studied by measuring the formation of inositol phosphates and the changes in intracellular calcium. Oxytocin binding was measured by saturation analysis. RESULTS: Atosiban and related peptides inhibited oxytocin-induced myometrial activation as pure antagonists (inhibition constant 10 nmol/L) but had no effect on prostaglandin E2-induced activation. Long-term (> or = 24 hours) exposure to atosiban had no residual effect on oxytocin sensitivity. However, long-term exposure to oxytocin resulted in homologous desensitization and loss of oxytocin receptors. Oxytocin-induced desensitization was prevented by coincubation with atosiban. CONCLUSIONS: Atosiban is a pure oxytocin antagonist and has a specific, reversible effect on myometrial cells in vitro. Its potential use for the management or even prevention of idiopathic preterm labor or to reverse uterine hypertony during oxytocin-induced labor should be tested in controlled clinical trials.

The relation between thromboxane and prostaglandin synthesis in human decidua tissue: a comparison of eicosanoid synthesis in minced tissue with that in a cell-free preparation.

Radiotracer studies and radioimmunoassay measurements demonstrate that minced tissues of human decidua produce chiefly thromboxane B2 (TxB2) (70% of total eicosanoids) and small amounts of prostaglandin F2 alpha (PGF2 alpha) (13%) PGD2 (8%), 6-keto-PGF1 alpha (5%) and PGE2 (4%). Inhibition of thromboxane synthesis with a specific inhibitor (OKY-1581: sodium (E)-3-[4(-3-pyridylmethyl)-phenyl]-2-methyl propenoate) increased prostaglandin formation in general, with the main product being PGF2 alpha (38%), a nonenzymic derivative of PGH2. Crude particulate fractions prepared from the same tissue synthesized two major products from [3H]arachidonate, TxB2 and 6-keto-PGF1 alpha (54 and 30%, respectively) and some PGF2 alpha and PGE2 (8-8%). However, in the presence of reduced glutathione (GSH), PGE2 became the main product (81%) (TxB2, 15%; PGF2 alpha, 2%; and 6-keto-PGF1 alpha, 2%). Half-maximal stimulation of PGE2 synthesis occurred at 46 microM GSH. The GSH concentration of tissue samples was found to be 110 +/- 30 microM. We conclude that human first trimester decidua cells possess the key enzymes of prostaglandin and thromboxane synthesis. Apparently, the production of these compounds is controlled by a specific mechanism in the tissue, which keeps PGE and prostacyclin synthesis in a reversibly suppressed state, whereas the formation of thromboxane is relatively stimulated.

PGE2 binding, synthesis, and distribution in hen oviduct.

Prostaglandin E2 (PGE2) bound specifically to particulate fractions prepared from the vagina and uterus (shell gland) portions of the hen oviduct in a time and temperature dependent fashion. Scatchard analysis indicated a single class of high-affinity binding sites in the vagina (Kd congruent to 1 nM), whereas the myometrium exhibited two kinds of binding site populations (Kd1 congruent to 1 nM, Kd2 congruent to 20 nM). It is suggested that these binding sites represent specific PGE2 receptors mediating the effects of PGE2 in oviductal smooth muscle. Vaginal particulate fractions produced approximately four times more prostanoids from [3H]-arachidonate than did uterine preparations. In the presence of epinephrine both tissues synthesized mainly thromboxane (TxB2), PGE2, and significantly less prostaglandin F2 alpha (PGF2 alpha). Addition of glutathione (GSH) or cytosol prepared from the oviduct markedly increased the yield of PGE2 at the expense of TxB2. Of the five morphologically discrete regions of the oviduct the vagina, infundibulum, and uterus contained the highest amounts of PGE and PGF, whereas the magnum and isthmus portions contain the least. TxB2 and 6-keto PGF1 alpha could not be detected in significant quantities in either region. These studies support the notion that PGE2 play a key role in the physiology of oviposition.

Turnover of lipid-bound arachidonate and biosynthesis of prostanoids in the endometrium and myometrium of the laying hen.

In vitro incorporation of [3H]arachidonate into various lipid fractions of minced avian endometrium and myometrium has shown that (i) phospholipids and triacylglycerols are labeled preferentially, whereas labeling of cholesteryl esters is only marginal; (ii) acylation reactions with arachidonoyl residues are very rapid in these tissues; (iii) incorporation of arachidonate into phosphatidylcholine and phosphatidylethanolamine seems to be a one-step reaction, whereas the transfer of arachidonate to phosphatidylinositol and phosphatidylserine proceeds after a time lag, suggesting that intermediate reaction steps are involved; and (iv) under conditions where tissue differences in free arachidonate levels are diminished by adding radioinert arachidonate to the incubation medium, endometrium incorporates more than twice as much [3H]arachidonate/unit of phospholipids than does the myometrium, whereas there is no such tissue-specific difference in the labeling of triacylglycerols. Furthermore, after a 90-min incubation phosphatidylserine and phosphatidylinositol not only are labeled markedly higher than either phosphatidylcholine or phosphatidylethanolamine in both tissues but the tissue-specific difference is also the highest (2.4- to 2.8-fold in the case of these phospholipids). Prostanoid synthesis from [3H]arachidonate in a crude membrane preparation in vitro has demonstrated that the myometrium possesses a significantly higher prostanoid-synthesizing capacity than the endometrium. This difference is particularly apparent (about 2.5-fold) in the case of prostaglandin E2. The results suggest that the endometrium is endowed with a special capacity to respond to signals with a rapid alteration of the level of free arachidonic acid, whereas the myometrium has the specific capacity to amplify its own contractions by an increased production of prostaglandins.

Conversion of arachidonic acid to prostanoids in the avian uterus.

In an effort to obtain information on the possible source of prostaglandins which have been shown to play an important role in oviposition we examined the metabolism of arachidonic acid in microsomal preparations of both the muscular and the glandular tissue of the hen uterus. We found that adrenaline and tryptophan (but not hydroquinone) were effective stimulators of prostanoid synthesis. On incubation with [3H]arachidonic acid we identified, using TLC radiochromatography and several solvent systems, prostaglandins F2 alpha and E2 and, predominantly, thromboxane B2 which could not be attributed to platelet contamination. Addition of reduced glutathione increased prostaglandin E2 formation at the expense of thromboxane B2 and at 1 mM concentration suppressed adrenaline-promoted prostanoid synthesis. While the former effect has been documented in many other systems and could be ascribed to the activation of prostaglandin H2 to prostaglandin E2 isomerase, the latter effect is postulated to be due to an inhibition of cyclooxygenase. Interestingly, this inhibitory effect was shared by a number of reducing agents. Although the subcellular preparations were derived from structurally and functionally different tissues, there was no qualitative difference with respect to prostanoid synthesis. Our data support the role of locally produced primary prostaglandins in the regulation of oviposition and raise the question of a potential role for thromboxane in this process.

Selective binding of prostaglandin F2 alpha to membrane fractions enriched in 5-nucleotidase, Ca2+-ATPase, and Mg2+-(Na+ + K+)ATPase.

Homogenates of uterine smooth muscle of laying hens were fractionated by differential centrifugation. Three of the five fractions thus obtained were further separated on sucrose gradient into four subfractions. By employing various enzyme markers, such as 5'-nucleotidase, Mg2+-(Na+ + K+)-ATPase, alkaline phosphatase, Ca2+-ATPase, cytochrome oxidase, acid phosphatase, and N-beta-acetyl glucosaminidase, the major subcellular fractions have been tentatively identified. Using gel filtration to separate bound and free prostaglandin F2 alpha (PGF2 alpha) we found that the specific uptake of PGF2 alpha was highest in the subfractions which also exhibited the highest enrichment in enzymes viewed generally as markers for cell membrane. The results suggest that PGF2 alpha-induced contractile activity in this tissue is initiated by the specific interaction of this agonist with discreet receptors in the sarcolemma, rather than by binding of PGF2 alpha to intracellular organelles or to a cytosolic receptor.

Substrate-induced dissociation of glycerol-3-phosphate dehydrogenase and its complex formation with fructose-bisphosphate aldolase.

A threefold decrease in specific activity of glycerol-3-phosphate dehydrogenase was found on going from 800 nM to 10 nM enzyme concentration. According to ultracentrifugal analyses the dimeric glycerol-3-phosphate dehydrogenase (molecular weight 78,000) dissociates into monomers in the equilibrium mixture of its substrates and products. The concentration-dependent decrease in the specific activity is interpreted as a consequence of subunit dissociation and the estimated dissociation constants are 0.7 micro M and 3.5 micro M at 38 degrees C and 20 degrees C respectively. According to active-enzyme-band centrifugation experiments and kinetic analysis aldolase forms a complex with glycerol-3-phosphate dehydrogenase and this complex formation influences the specific activity of the dehydrogenase. The interaction between glycerol-3-phosphate dehydrogenase and aldolase can provide a regulatory mechanism at the branching point of glycolytic and lipid metabolic pathways.