Structural heterogeneity of membrane receptors and GTP-binding proteins and its functional consequences for signal transduction

Differential G protein subunit expression by prostate cancer cells and their interaction with CXCR5

Background

Prostate cancer (PCa) cell lines and tissues differentially express CXCR5, which positively correlate with PCa progression, and mediate PCa cell migration and invasion following interaction with CXCL13. However, the differential expression of G protein α, β, and γ subunits by PCa cell lines and the precise combination of these proteins with CXCR5 has not been elucidated.

Methods

We examined differences in G protein expression of normal prostate (RWPE-1) and PCa cell lines (LNCaP, C4-2B, and PC3) by western blot analysis. Further, we immunoprecipitated CXCR5 with different G protein subunits, and CXCR4, following CXCL13 stimulation. To investigate constitutive coupling of CXCR5 with CXCR4 and PAR-1 we performed invasion assay in PCa cells transfected with G_αq/i2 or G_α13 siRNA, following CXCL13 treatment. We also investigated Rac and RhoA activity by G-LISA activation assay in PCa cells following CXCL13/thrombin stimulation.

Result

Of the 22 G proteins studied, G_αi1-3, G_β1-4, G_γ5, G_γ7, and G_γ10 were expressed by both normal and PCa cell lines. G_αs was moderately expressed in C4-2B and PC3 cell lines, G_αq/11 was only present in RWPE-1 and LNCaP cell lines, while G_α12 and G_α13 were expressed in C4-2B and PC3 cell lines. G_γ9 was expressed only in PCa cell lines. G_α16, G_β5, G_γ1-4, and G_γ13 were not detected in any of the cell lines studied. Surprisingly, CXCR4 co-immunoprecipitated with CXCR5 in PCa cell lines irrespective of CXCL13 treatment. We also identified specific G protein isoforms coupled to CXCR5 in its resting and active states. G_αq/11/G_β3/G_γ9 in LNCaP and G_αi2/G_β3/G_γ9 in C4-2B and PC3 cell lines, were coupled to CXCR5 and disassociated following CXCL13 stimulation. Interestingly, G_α13 co-immunoprecipitated with CXCR5 in CXCL13-treated, but not in untreated PCa cell lines. Inhibition of G_αq/i2 significantly decreased the ability of cells to invade, whereas silencing G_α13 did not affect CXCL13-dependent cell invasion. Finally, CXCL13 treatment significantly increased Rac activity in G_αq/i2 dependent manner, but not RhoA activity, in PCa cell lines.

Conclusions

These findings offer insight into molecular mechanisms of PCa progression and can help to design some therapeutic strategies involving CXCR5 and/or CXCL13 blockade and specific G protein inhibition to abrogate PCa metastasis.

Structural flexibility of small GTPases. Can it explain their functional versatility?

Multiple interactions with many different partners are responsible for the amazing functional versatility of proteins, especially those participating in cellular regulation. The structural properties that could facilitate multiple interactions are examined for small GTPases. The role of cellular constraints, compartmentation and scaffolds on protein-protein interactions is considered.

Opposing changes of trimeric G protein levels during ontogenetic development of rat brain

Developmental changes in the distribution of guanine nucleotide-binding regulatory proteins (G proteins) were investigated in the rat brain during postnatal development. Using a standard or high-resolution urea-SDS-PAGE and specific polyclonal antipeptide antibodies oriented against G(i)alpha1/G(i)alpha2, G(i)alpha3, G(s)alpha, G(o)alpha1/G(o)alpha2, G(q)alpha/G(11)alpha and Gbeta subunit, all these proteins were determined by quantitative immunoblotting in homogenates prepared from cortex, thalamus, hippocampus and pituitary of 1-, 7-, 12-, 18-, 25- and 90-day-old animals. The levels of the majority of G protein alpha subunits, namely G(i)alpha1, G(i)alpha2, G(i)alpha3, G(o)alpha1, G(o)alpha2, G(q)alpha, G(11)alpha and Gbeta, were high already at birth. Whereas the short variant of G(s)alpha, G(s)alphaS, rose sharply in all tested brain regions between postnatal day (PD) 1 and 90, the long variant of G(s)alpha, G(s)alphaL, was unchanged in cortex and thalamus and slightly increased in hippocampus. An increase was observed also in expression of G(i)alpha1/G(i)alpha2 and G(o)alpha1 protein, while G(o)alpha2 remained constant. Minority protein G(o)alpha* dramatically increased in cortex and thalamus, was unchanged in hippocampus and not detectable in pituitary. By contrast, the highest levels of G(i)alpha3 and G(q)alpha/G(11)alpha were detected as early as at PD 1. During the next 90 days, the immunological signal of G(i)alpha3 almost disappeared and G(q)alpha/G(11)alpha continuously declined to the levels corresponding to 50% of the levels determined at birth. Expression of Gbeta subunit was basically unchanged during postnatal development. Our present analysis indicates that G(s)alpha, G(i)alpha/G(o)alpha and G(q)alpha/G(11)alpha proteins are differently expressed in the course of brain development. Differential expression of the individual alpha subunits of trimeric G proteins during postnatal development suggests their different roles in maturation of the brain tissue.

Recovery and expression of messenger RNA from postmortem human brain tissue

The Bryan Alzheimer's Disease Research Center Rapid Autopsy Program at Duke University Medical Center obtains postmortem human brain tissue for experimental investigations. We evaluated 19 brains for RNA integrity and mRNA gene expression. Nine were from patients diagnosed with Alzheimer's disease, and ten were from nondemented controls. In all cases, the following variables were recorded: postmortem procurement delay (range, 1 hour and 10 minutes to 14 hours), pH of cerebrospinal fluid, premortem fever or sepsis, provision of supplemental oxygen in the agonal period, and temporal relation to time of death (either sudden death or protracted illness). Total RNA was extracted, quantified, and evaluated by agarose gel electrophoresis and quantitative gene expression analysis of 18S rRNA and edg-1 using TaqMan technology. All samples appeared to yield intact RNA without significant degradation, and expression of the edg-1 gene was detected by the real time reverse transcriptase polymerase chain reaction in all cases. We conclude that intact RNA can be obtained from postmortem human brain tissue, even in patients with severe premortem illnesses and delayed postmortem tissue procurement intervals. However, we caution that the successful expression of certain genes from postmortem brain tissue may require enhanced procurement efforts to maximize RNA integrity.

Phospholipase C-delta1 is activated by capacitative calcium entry that follows phospholipase C-beta activation upon bradykinin stimulation

To characterize the regulatory mechanism of phospholipase C-delta1 (PLC-delta1) in the bradykinin (BK) receptor-mediated signaling pathway, we used a clone of PC12 cells, which stably overexpress PLC-delta1 (PC12-D1). Stimulation with BK induced a significantly higher Ca(2+) elevation and inositol 1,4,5-trisphosphate (IP(3)) production with a much lower half-maximal effective concentration (EC(50)) of BK in PC12-D1 cells than in wild type (PC12-W) or vector-transfected (PC12-V) cells. However, BK-induced intracellular Ca(2+) release and IP(3) generation was similar between PC12-V and PC12-D1 cells in the absence of extracellular Ca(2+), suggesting that the availability of extracellular Ca(2+) is essential to the activation of PLC-delta1. When PC12-D1 cells were treated with agents that induce Ca(2+) influx, more IP(3) was produced, suggesting that the Ca(2+) entry induces IP(3) production in PC12-D1 cells. Furthermore, the additional IP(3) production after BK-induced capacitative calcium entry was detected in PC12-D1 cells, suggesting that PLC-delta1 is mainly activated by capacitative calcium entry. When cells were stimulated with BK in the presence of extracellular Ca(2+), [(3)H]norepinephrine secretion was much greater from PC12-D1 cells than from PC12-V cells. Our results suggest that PLC-delta1 is activated by capacitative calcium entry following the activation of PLC-beta, additively inducing IP(3) production and Ca(2+) rise in BK-stimulated PC12 cells.

Phosphatidylinositol 4-kinase of Torpedo californica electrocytes: physico-chemical characterization and regulation by calcium and vicinal molecules of phosphatidylinositol

A phosphatidylinositol 4-kinase (Ptdlns 4-kinase, M(r) approximately 95,000) from the membranes of the electric organ of Torpedo californica was purified to apparent homogeneity. The Michaelis constant for ATP (KM = 280 +/- 60 microM at 20 degrees C) and the inhibition constant for adenosine (Ki = 0.4 mM at 20 degrees C) qualify the electrocyte Ptdlns 4-kinase as a type III kinase. The Ptdlns 4-kinase phosphorylates preferentially exogenous Ptdlns, added in the form of mixed Ptdlns/Triton X-100 micelles, whereas endogenously bound Ptdlns in the membrane fragments of electrocytes is a very poor substrate. It is important that the enzyme and the substrate Ptdlns are situated in different lipid bilayers. The catalytic turnover constant for exogenous Ptdlns is k = 55.3 +/- 6 min-1 at 20 degrees C and the molar Triton X-100/Ptdlns ratio of 16:1. For the substrate Ptdlns in the 'micellar solvent' Triton X-100, steady state kinetics were analysed in terms of the mole fraction X = n(Ptdlns)/[n(Ptdlns) + n(Triton X)] yielding the characteristic Michaelis mole fraction XM = 0.019 +/- 0.005 at 20 degrees C. The activity of the enzyme was enhanced about 5-fold in the presence of Triton X-114, yielding k = 277 +/- 30 min-1 at 20 degrees C. Triton X-114 has a shorter head-group, indicating that the vicinity of the Ptdlns head group in the mixed micelles should not be screened by bulky neighbours. The inhibition of the enzyme activity by Ca2+ is highly cooperative yielding the Hill inhibition constant Ki = 0.47 +/- 0.1 mM and the Hill coefficient h = 3.6 +/- 0.5. The enthalpy of activation is 100 +/- 10 kJ/mol between 0 degree C and 20 degrees C. Although the Ptdlns 4-kinase can be affinity-chromatographically copurified with the nicotinic acetylcholine (AcCho) receptor, suggesting tight association between the two proteins. AcCho does not affect the activity of the Ptdlns 4-kinase in the presence of the AcCho receptor.

Phospholipase C-delta1 and oxytocin receptor signalling: evidence of its role as an effector

Although the oxytocin receptor modulates intracellular Ca2+ ion levels in myometrium, the identities of signal molecules have not been clearly clarified. Our previous studies on oxytocin receptor signalling demonstrated that 80 kDa Ghalpha is a signal mediator [Baek, Kwon, Lee, Kim, Muralidhar and Im (1996) Biochem. J. 315, 739-744]. To elucidate the effector in the oxytocin receptor signalling pathway, we evaluated the oxytocin-mediated activation of phospholipase C (PLC) by using solubilized membranes from human myometrium and a three-component preparation containing the oxytocin receptor-Ghalpha-PLC-delta1 complex. PLC-delta1 activity in the three-component preparation, as well as PLC activity in solubilized membranes, was increased by oxytocin in the presence of Ca2+ and activated Ghalpha (GTP-bound Ghalpha). Furthermore the stimulated PLC-delta1 activity resulting from activation of Ghalpha via the oxytocin receptor was significantly attenuated by the selective oxytocin antagonist desGly-NH2d(CH2)5[Tyr(Me)2,Thr4]ornithine vasotocin or GDP. Consistent with these observations, co-immunoprecipitation and co-immunoadsorption of PLC-delta1 in the three-component preparation by anti-Gh7alpha antibody resulted in the PLC-delta1 being tightly coupled to activated Ghalpha on stimulation of the oxytocin receptor. These results indicate that PLC-delta1 is the effector for Ghalpha-mediated oxytocin receptor signalling.

Is signal transduction modulated by an interaction between heterotrimeric G-proteins and tubulin?

Although it is generally accepted that tubulin plays an important role in G-protein-mediated signal transduction in a variety of systems, the mechanism of this phenomenon is not completely understood. G-protein-tubulin interaction at the cell membrane and the cytosol, and the influence of such an interaction on cellular signaling are discussed in this review article. Because the diameter of a microtubule is 25 nm and the plasma membrane is 9-11 nm thick, it is not possible for membrane-associated tubulin to assemble into a complete microtubule in the membrane environment. However, tubulin heterodimers may be able to function in the membrane environment as individual heterodimers or as polymers arranged into short protofilaments. At the cell membrane, membrane-associated tubulin may influence hormone-receptor interaction, receptor-G-protein coupling, and G-protein-effector coupling. Structural proteins, such as tubulin, can participate in cellular signaling by communicating through physical forces. By virtue of its interaction with the submembranous network of cytoskeletal proteins, tubulin, when perturbed in one locus, can transmit large changes in conformations to other points. Thus, GTP binding to membrane-associated tubulin might lead to a conformational change in either receptors or G proteins. This may, in turn, influence the binding of an agonist to its receptor. On the other hand, in the cell cytosol, subsequent to agonist-induced translocation of G-proteins from the membrane compartment to the cytosol, G-proteins may affect microtubule formation. In GH3 and AtT-20 cells (stably expressing TRH receptor), transiently transfected with Gq alpha cDNA, soluble tubulin levels decreased in Gq alpha-transfected GH3 and AtT-20 cells, by 33% and 52%, respectively. These results suggest that G-proteins may have a direct effect on the microtubule function in vivo. Because tubulin and G-protein families are ubiquitous and highly conserved, an interaction between these two protein families may occur in vivo, and this, in turn, can have an impact on signal transduction. However, the physiological significance of this interaction remains to be demonstrated.

G-protein-coupled receptor kinases

beta-Adrenergic receptors are prototypes of the many G-protein-coupled receptors. Activation and inactivation of these receptors are regulated by multiple mechanisms which can affect either their function or their expression. The most obvious changes of such receptor systems are induced by activation of the receptors themselves by their respective agonists, and this process is called receptor desensitization. One of these mechanisms of desensitization is due to the actions of specific receptor kinases, termed the G-protein-coupled receptor kinases (GRKs). These kinases specifically phosphorylate only the agonist-occupied form of such receptors. This phosphorylation is then followed by binding of inhibitor proteins, called arrestins, to the receptors. Binding of arrestins results in displacement of the G-proteins from the receptors and hence causes uncoupling of receptors and G-proteins. Recent data indicate that the function and subcellular distribution of GRKs is itself subject to regulation. Various mechanisms have evolved to anchor the different GRKs to the plasma membrane. In addition, recent data indicate that GRKs can also associate with intracellular membranes where they may exert as yet unknown functions. A pathophysiological role for GRKs can be inferred from recent studies on heart failure as well as the observation that chronic treatment with various agonists or antagonists for G-protein-coupled receptors results in alterations of GRK expression.

Mechanisms of beta-adrenergic receptor desensitization: from molecular biology to heart failure

beta-Adrenergic receptors are often studied as prototypes of the large family of G-protein-coupled receptors, which includes many other well-known members such as the muscarinic acetylcholine receptors, but also the receptors for light, taste and olfaction. These receptors are regulated by multiple mechanisms which can affect either their function or their expression to a rapidly changing environment. The most obvious changes are effected by receptor agonists, and this process is called receptor desensitization. On the functional level, the most intriguing and important mechanism of desensitization involves the phosphorylation of beta-adrenergic and homologous receptors by specific receptor kinases, termed the G-protein-coupled receptor kinases (GRKs). This phosphorylation is followed by binding of arrestins to the receptors, which causes uncoupling of receptors and G-proteins and thus results in a loss of receptor function. On the expression level, there appear to be two major pathways leading to a reduction of the receptor number: degradation of the receptors themselves, or reduced receptor synthesis brought about by reduced receptor mRNA levels. Heart failure is accompanied by a markedly reduced responsiveness of the beta-adrenergic receptor system, which in many ways resembles the phenomena seen in agonist-induced receptor desensitization. The levels of beta 1-adrenergic receptors are reduced, and this reduction is paralleled by similar decreases in the levels of the corresponding mRNA. At the same time, the activity and the mRNA levels of one of the GRK-isoforms, GRK2 (which is identical to the beta-adrenergic receptor kinase 1) are increased. These alterations may contribute to the loss of beta-adrenergic receptor responsiveness in heart failure and result in further impairment of cardiac function.

Isoproterenol-induced subcellular redistribution of G-protein beta subunits in S49 lymphoma cells demonstrated by a novel competitive ELISA

A novel competitive ELISA has been developed for the determination of levels of the beta subunit of guanine-nucleotide-binding protein (G-protein) using antipeptide antibodies directed against the amino terminus of the beta subunit. Because beta subunits form highly hydrophobic.heterodimeric complexes with gamma subunits of G-proteins, specific assay conditions were required. Optimal concentrations of antibodies, detergents, Mg2+ as well as ionic strength were determined. In addition, we found that an effective binding of the used antibodies to the beta subunit was ensured only after denaturation of the beta gamma complexes. Subsequently, this ELISA was used for quantitation of the beta subunit in subcellular fractions of S49 lymphoma cells during isoproterenol-mediated desensitization of beta-adrenergic controlled transmembrane signalling system. A 10 min as well as 60 min treatment of the cells with isoproterenol (1 nmol/ml) resulted in a significant shift of G-protein beta subunits (presumably as beta gamma complexes) from the plasma membrane fractions to low-density microsomal fractions. No significant change was detected after the hormone action in the distribution of plasma membrane constitutive enzymes. In conclusion, the developed ELISA helped us to reveal that beta-adrenergic stimulation can induce redistribution of the beta gamma dimer from plasma membranes to low-density microsomes.

The stimulus-sensitive H2O2-generating system present in human fat-cell plasma membranes is multireceptor-linked and under antagonistic control by hormones and cytokines

Previous work demonstrated that human fat-cells possess a plasma-membrane-bound H2O2-generating system that is activated by insulin. Here we show that this system is under antagonistic control by various hormones and cytokines that typically act through several distinct receptor families. Similarly to insulin, oxytocin and tumour necrosis factor alpha acted as stimulators of NADPH-dependent H2O2 generation, whereas isoprenaline, a beta-adrenergic agonist, had inhibitory effects. Surprisingly, the acidic and basic isoforms of fibroblast growth factor as well as homodimeric platelet-derived growth factor AA and BB had antagonistic stimulatory and inhibitory effects on NADPH-dependent H2O2 generation. The agents tested acted at discrete ligand-specific receptors and their mechanisms of action were membrane-delimited and occurred in the absence of ATP. These findings implied that established pathways of signal transduction, including receptor kinases or second-messenger-dependent protein kinases A and C, were not involved and placed the stimulus-sensitive H2O2-generating system in a position comparable with adenylate cyclase. It was concluded that the stimulus-sensitive H2O2-generating system of human fat-cells meets all criteria of a universal signal-transducing system for hormones and cytokines that may link ligand binding to cell-surface receptors to changes in the intracellular redox equilibrium.

Rapid kinetics of G protein subunit association: a rate-limiting conformational change?

G protein subunit association and dissociation are thought to play an important role in signal transduction. We measured alpha beta gamma heterocomplex formation using resonance energy transfer. Fluorescein-labelled alpha(F-alpha) emission was quenched approximately 10% on mixing with eosin-labelled beta gamma(E-beta gamma). Unlabelled beta gamma did not quench F-alpha fluorescence. Stopped-flow kinetics showed a t1/2 ranging from 2.5 s to 0.25 s for 50 nM to 1200 nM E-beta gamma. The rate saturated at high E-beta gamma concentrations consistent with a two-step mechanism. We report the first rapid-mix studies of G protein subunit association kinetics which suggest that alpha and beta gamma combine by a two-step process with a maximal rate of 4.1 +/- 0.4 s-1.

The relationship between receptor-effector unit heterogeneity and the shape of the concentration-effect profile: pharmacodynamic implications

The apparent concentration-effect relationship is the ensemble of many effector units (such as individual cells or channels) that do not always exhibit a uniform stimulus-effect relationship. This concept is substantiated by many observations of heterogeneity in receptor-effector populations including hormone secreting cells, response to hormonal stimuli, activity pattern of second messengers, stimulus-evoked synaptic currents, and single ion channels. The relationship between drug concentration and magnitude of pharmacologic response is commonly described by the sigmoidal Emax model which was derived from the Hill equation. The sigmoidicity factor (N) in this model is assumed to be a pure mathematical parameter without physiological connotations. This work demonstrates that the numerical value of N (measured empirically) is the product of two factors: (i) the degree of heterogeneity of the effector subunits, i.e., the elemental component that upon drug stimulus contributes its pharmacological effect independently and does not interact with other subunits (it could range from a single receptor up to a whole tissue), and (ii) value of N*--the shape factor of the subunits' concentration-effect relationship. A special case of this approach occurs when N* > 5, which is an on-off case. Here N is determined by the distribution (density equation) of the subunit values. In case of heterogeneity of the microparameters of the effector subunits the apparent N will always have a lower value than N*. According to this theory it can be concluded that without knowledge of the distribution of the microparameters no mechanistic interpretation can be deduced from the apparent N value. If in the future N* can be determined by theoretical or experimental methods, the distribution function relating N* to N can be calculated. The relevance of this theory is increased in view of the progress being made in advanced research techniques which may enable us to determine the concentration-effect relationship at the level of the individual effector unit.

Cloning of the rat edg-1 immediate-early gene: expression pattern suggests diverse functions

The edg-1 immediate-early gene encodes a G-protein-coupled receptor homolog implicated in endothelial cell differentiation. We report the cloning of the rat edg-1 gene. Our Northern analyses indicate that edg-1 is much more widely expressed than previously thought. edg-1 mRNA was found in many organs at several stages of development with relatively high levels present in adult brain. edg-1 transcripts were also detected in several cell lines. Expression of edg-1 mRNA in the PC12 cell model of neuronal differentiation was unaffected by agents that cause PC12 cells to differentiate or proliferate. Therefore, edg-1 may play a cell-type-specific role in differentiation and also participate in neurotransmission.

The complex of human Gs protein with the beta 3 adrenergic receptor: a computer-aided molecular modeling study

Three-dimensional (3D) models of the human Gs protein, the human beta 3 adrenergic receptor and their complex are constructed using computer-aided molecular modeling techniques. The structures of bacterial EF-tu 200 and bacteriorhodopsin were used as starting points for modeling the Gs protein and beta 3 receptor, respectively. Experimental data are used as constraints to guide the modeling. The resulting 3D structures of the Gs protein, the beta 3 receptor and their complex are in accord with the experimental data. It is found that the third intracellular loop of the beta 3 receptor as well as its C-terminus are involved in the binding. Various residues of N-terminus and C-terminus of the Gs protein also participate in the binding. The model of the complex suggests that the Gs protein binds to the beta 3 receptor in such a way that it will be placed in the interface of membrane and intracellular space. This orientation is supported by experimental data. It is concluded that the modeled structure of the complex of the alpha subunit of the human Gs protein and the beta 3 adrenergic receptor is in agreement with the experimental data and it can provide a basis for understanding the way these proteins interact.

Age-dependent changes in transmembrane signalling: identification of G proteins in human lymphocytes and polymorphonuclear leukocytes

In human neutrophils (PMNLs) we found that in the elderly IP3 formation was significantly decreased compared to that of young subjects. For FMLP receptor binding affinity and number no measurable differences occurred upon ageing, studying both the low or the high affinity receptors. The amount of ADP-ribosylated G proteins, catalysed by pertussis toxin (PT) or cholera toxin (CT), was significantly increased in PMNLs of the elderly. In lymphocytes, the PT-catalysed ADP ribosylation of G proteins was also increased with ageing, while the CT-catalysed ribosylation was decreased. The autoradiogram of [32P]ADP-ribosylated proteins by CT in lymphocytes of young individuals showed a major polypeptide of 40,000 M(r). In contrast, in lymphocytes of the elderly, the major polypeptide was 45,000 M(r). In PMNLs, CT labelled quite strongly the 45,000 M(r) band, mainly in the elderly. When PT was used, no age-related pattern changes could be demonstrated, while differences could be observed between the two types of cells. The use of antiserum P680 (G alpha common) showed no age-related pattern changes, while the intensity of the labelled proteins varies with age and cell type. The antiserum U46 (Go alpha) could identify in lymphocytes of young subjects two polypeptides 68,000 and 41,000 M(r). The prominent polypeptide in lymphocytes of the elderly was the 70,000 M(r) and no other polypeptides could be recognized. In PMNLs of young subjects the U46 and serum identified a range of species. In PMNLs of the elderly all these bands were weakly labelled. The present data indicate changes in the pattern and the quantity of G proteins in lymphocytes and PMNLs of elderly subjects.

Signaling mechanisms that regulate saliva formation

The precipitating event in the formation of saliva is the binding of neurotransmitter molecules to cell surface receptor proteins. The principal neurotransmitters involved are acetylcholine and norepinephrine that bind, respectively, to muscarinic-cholinergic, and alpha- and beta-adrenergic receptors. The transduction of the extracellular signal requires an integral membrane protein capable of binding GTP, a G protein, that specifically interacts with the receptor. The components of G protein transduction systems are fairly well studied, but the pathways by which signals are routed are just being recognized. Delineation of such routing pathways is essential to understanding the regulation of saliva formation.

Effects of the phospholipid environment in the plasma membrane on receptor interaction with the adenylyl cyclase complex of intact cells

In this study we have examined the effects of variations of the plasma membrane phospholipid and cholesterol content on the metabolic functions of the adenylyl cyclase complex in intact cells. Exposure of cells to 0.1 U/ml of sphingomyelinase led to the degradation of 75, 55 and 40% of the cellular total sphingomyelin mass in human skin fibroblasts (HSF), Chinese hamster lung fibroblasts (CHLF) and rat liver hepatocytes (RLH), respectively. Degradation of sphingomyelin in native cells led in turn to a reduction (within 60 min) of the plasma membrane cholesterol content (by 25, 15 and 10%, respectively). This manipulation of the plasma membrane lipid content did not affect the forskolin or prostaglandin E1-induced activation of adenylyl cyclase (as measured from the conversion of [3H]adenine via [3H]ATP to [3H]cAMP). These manipulations did, however, increase the basal rate of [3H]cAMP formation in rat liver hepatocytes (but not in the fibroblast cell types). With Chinese hamster lung fibroblasts, transfected to express an alpha 2-adrenergic receptor, it was observed that the alpha 2-adrenergic receptor-induced inhibition of adenylyl cyclase activity was slightly (but significantly) diminished in sphingomyelin and cholesterol-depleted cells. With isolated rat liver hepatocytes it was observed that the glucagon (receptor) mediated activation of adenylyl cyclase was also reduced in sphingomyelinase-treated cells. In another set of experiments, CHLF and RLH cells were exposed for 2 h to vesicles prepared from dilauroylphosphatidylcholine, to increase the lateral packing density in the outer leaflet of the plasma membrane. In such treated cells, the receptor-coupling to adenylyl cyclase was markedly reduced both in CHLF (the alpha 2-adrenergic receptor) and RLH (the glucagon-receptor) cells. We conclude that the direct activation of adenylyl cyclase (i.e., by forskolin) is not markedly affected by manipulations outer leaflet phospholipid composition (either reduction of sphingomyelin or increase of phosphatidylcholine), whereas receptor-coupled events clearly are.

Bioamine receptors: evolutionary and functional variations of a structural leitmotiv

Bioamines act as neurohormonal messengers through their binding to receptors which belong to the largest membrane protein family known so far: the seven spanning membrane receptors. This class of receptors transmits the effect of agonist binding to intracellular effectors by interacting with an intermediary G-protein. The diversity of receptor subtypes inside the protein family, observed in many animal species, is the result of a long evolutionary process. The tendency to protein diversification depends upon gene duplications and upon the continuous accumulation of mutations. The maintenance of vital functions in organisms, however, strictly requires enough structural conservation to ensure the functionality of the corresponding proteins. Both forces cooperate to ensure the adaptation of organisms to a changing environment. We have reviewed here the main conformational and functional constraints exerted on the structure of the bioamine receptors. They are mainly the transmembrane conformation of the receptors, their ability to bind ligands, to interact with G-proteins and to desensitize. The molecular basis of the biochemical and pharmacological differences used to classify the members of the receptor family have also been examined. Interestingly, this classification is very close to that obtained by the molecular phylogeny methods, used to elucidate the evolutionary relationships between bioamine receptors. However, this latter classification allows to accurately distinguish between different receptor subtypes (paralogous genes) and species homologous (orthologous genes). In addition, the calculation of phylogenetical distances reveals two main periods of diversification: the first one occurred before the separation of arthropods from vertebrates, in the Precambrian, and corresponds to the appearance of the main subtypes of the bioamine receptors. The second one, which occurred about 400 million years ago, might accompany the cephalization of the CNS in vertebrates.

Molecular characterization of a human brain adenosine A2 receptor

A cDNA encoding a G protein-coupled receptor of unknown ligand specificity was isolated from a human hippocampal cDNA library by virtue of the high degree of structural homology between members of this receptor family. The cloned receptor DNA was transfected into human embryonic kidney 293 cells. Stably transfected cell lines bound a variety of adenosine agonists and antagonists with affinities characteristic of a brain adenosine A2a receptor. The A2a specific agonist CGS21680 stimulated cAMP production but did not alter intracellular calcium concentrations in transfected 293 cells.

Three-dimensional structure for the beta 2 adrenergic receptor protein based on computer modeling studies

Computer-aided model building techniques have been used to construct three-dimensional model structures for hamster beta 2 adrenergic receptor. Experimental data were used as constraints to guide the model building procedure, and a number of rather strict criteria were applied to assess the physical plausibility of model structures. We present details of our best model structure to date, which is consistent with a large body of experimental data. We also discuss in detail our model building procedures and evaluation criteria, which we believe may be of general utility in modeling projects.

Guanyl nucleotides modulate binding to steroid receptors in neuronal membranes

The recently characterized corticosteroid receptor on amphibian neuronal membranes appears to mediate rapid, stress-induced changes in male reproductive behaviors. Because the transduction mechanisms associated with this receptor are unknown, we performed radioligand binding studies to determine whether this steroid receptor is negatively modulated by guanyl nucleotides. The binding of [3H]corticosterone to neuronal membranes was inhibited by nonhydrolyzable guanyl nucleotides in both equilibrium saturation binding and titration studies. The addition of guanyl nucleotide plus unlabeled corticosterone induced a rapid phase of [3H]corticosterone dissociation from membranes that was not induced by addition of unlabeled ligand alone. Furthermore, the equilibrium binding of [3H]corticosterone and the sensitivity of the receptor to modulation by guanyl nucleotides were both enhanced by Mg2+. These results are consistent with the formation of a ternary complex of steroid, receptor, and guanine nucleotide-binding protein that is subject to regulation by guanyl nucleotides. Therefore, rapid signal transduction through corticosteroid receptors on neuronal membranes appears to be mediated by guanine nucleotide-binding proteins.

Subunit interactions of GTP-binding proteins

Fluorescence energy transfer [cf. Förster, T. (1948) Ann. Phys. 6, 55-75] was tested for its suitability to study quantitative interactions of subunits of G0 with each other and these subunits or trimeric G0 with the beta 1-adrenoceptor in detergent micelles or after reconstitution into lipid vesicles [according to Feder, D., Im, M.-J., Klein, H. W., Hekman, M., Holzhöfer, A, Dees, C., Levitzki, A., Helmreich, E. J. M. & Pfeuffer, T. (1986) EMBO J. 5, 1509-1514]. For this purpose, alpha 0- and beta gamma-subunits and trimeric G0 purified from bovine brain, the beta gamma-subunits from bovine rod outer segment membranes and the beta 1-adrenoceptor from the turkey erythrocyte were all labelled with either tetramethylrhodamine maleimide or fluorescein isothiocyanate under conditions which leave the labelled proteins functionally intact. In the case of alpha 0- and beta gamma-interactions, specific high-affinity binding interactions (Kd approximately 10 nM) and nonspecific low-affinity binding interactions (Kd approximately 1 microM) could be readily distinguished by comparing fluorescence energy transfer before and after dissociation with 10 microM guanosine 5'-O-[gamma-thio]triphosphate and 10 mM MgCl2 where only low-affinity binding interactions remained. Interactions between alpha 0- and beta gamma-subunits from bovine brain or from bovine retinal transducin did not differ much. The beta gamma-subunits from bovine brain were found to bind with high transfer efficiency and comparable affinities to the hormone-activated and the nonactivated beta 1-receptor reconstituted in lipid vesicles: Kd = 100 +/- 20 and 120 +/- 20 nM, respectively; however, beta gamma-subunits from transducin appeared to bind more weakly to the beta 1-adrenoceptor than beta gamma-subunits from bovine brain. Separated purified homologous alpha 0- and beta gamma-subunits from bovine brain interfered mutually with each other in binding to the beta 1-adrenoceptor presumably because they had a greater affinity for each other than for the receptor. These findings attest to the suitability of fluorescence energy transfer for studying protein-protein interactions of G-proteins and G-protein-linked receptors. Moreover, they supported the previous finding [Kurstjens, N. P., Fröhlich, M., Dees, C., Cantrill, R. C., Hekman, M. & Helmreich, E. J. M. (1991) Eur. J. Biochem. 197, 167-176] that beta gamma-subunits can bind to the nonactivated beta 1-adrenoceptor.