Memory in Ion Channel Kinetics.

Ion channels are transport proteins present in the lipid bilayers of biological membranes. They are involved in many physiological processes, such as the generation of nerve impulses, hormonal secretion, and heartbeat. Conformational changes in the ion channel-forming protein allow the opening or closing of pores to control the ionic flux through the cell membranes. The opening and closing of the ion channel have been classically treated as a random kinetic process, known as a Markov process. Here the time the channel remains in a given state is assumed to be independent of the condition it had in the previous state. More recently, however, several studies have shown that this process is not random but a deterministic one, where both the open and closed dwell-times and the ionic current flowing through the channel are history-dependent. This property is called long memory or long-range correlation. However, there is still much controversy regarding how this memory originates, which region of the channel is responsible for this property, and which models could best reproduce the memory effect. In this article, we provide a review of what is, where it is, its possible origin, and the mathematical methods used to analyze the long-term memory present in the kinetic process of ion channels.

Nitric oxide modulates the firing rate of the rat supraoptic magnocellular neurons.

In vitro, nitric oxide (NO) inhibits the firing rate of magnocellular neurosecretory cells (MNCs) of hypothalamic supraoptic and paraventricular nuclei and this effect has been attributed to GABAergic activation. However, little is known about the direct effects of NO in MNCs. We used the patch-clamp technique to verify the effect of L-arginine, a precursor for NO synthesis, and N(omega)-nitro-L-arginine methyl ester hydrochloride (L-NAME), an inhibitor of NOS, on spontaneous electrical activity of MNCs after glutamatergic and GABAergic blockade in Wistar rat brain slices. 6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX) (10 microM) and dl-2-amino-5-phosphonovaleric acid (dl-AP5) (30 microM) were used to block postsynaptic glutamatergic currents, and picrotoxin (30 microM) and saclofen (30 microM) to block ionotropic and metabotropic postsynaptic GABAergic currents. Under these conditions, 500 microM L-arginine decreased the firing rate from 3.7+/-0.6 Hz to 1.3+/-0.3 Hz. Conversely, 100 microM L-NAME increased the firing rate from 3.0+/-0.3 Hz to 5.8+/-0.4 Hz. All points histogram analysis showed changes in resting potential from -58.1+/-0.8 mV to -62.2+/-1.1 mV in the presence of L-arginine and from -59.8+/-0.7 mV to -56.9+/-0.8 mV by L-NAME. Despite the nitrergic modulator effect on firing rate, some MNCs had no significant changes in their resting potential. In those neurons, hyperpolarizing after-potential (HAP) amplitude increased from 12.4+/-1.2 mV to 16.8+/-0.7 mV by L-arginine, but without significant changes by L-NAME treatment. To our knowledge, this is the first demonstration that NO can inhibit MNCs independent of GABAergic inputs. Further, our results point to HAP as a potential site for nitrergic modulation.

Nitric oxide modulates ATP-evoked currents in mouse Leydig cells.

Testosterone synthesis within Leydig cells is a calcium-dependent process. Intracellular calcium levels are regulated by different processes including ATP-activated P2X purinergic receptors, T-type Ca2+ channels modulated by the luteinizing hormone, and intracellular calcium storages recruited by a calcium-induced calcium release mechanism. On the other hand, nitric oxide (NO) is reported to have an inhibitory role in testosterone production. Based on these observations, we investigated the interaction between the purinergic and nitrergic systems in Leydig cells of adult mice. For this purpose, we recorded ATP-evoked currents in isolated Leydig cells using the whole cell patch clamp technique after treatment with L-NAME (300 µM and 1 mM), L-arginine (10, 100, 300, and 500 µM), ODQ (300 µM), and 8-Br-cGMP (100 µM). Our results show that NO produced by Leydig cells in basal conditions is insufficient to change the ATP-evoked currents and that extra NO provided by adding 300 µM L-arginine positively modulates the current through a mechanism involving the NO/cGMP signaling pathway. Thus, we report an interaction between the nitrergic and purinergic systems in Leydig cells and suggest that Ca2+ entry via the purinergic receptors can be regulated by NO.

A quartz cell for studying planar lipid bilayer membranes.

A quartz chamber is proposed for use in experiments with planar lipid bilayer membranes. Membranes are formed in a hole made on the lateral wall of a fused quartz test tube, immersed in an electrolyte solution. The quartz cell is easy to clean, chemically inert and easily made. Membranes formed in this chamber had specific resistances higher than 10(8) omega.cm2 and excellent mechanical stability.

Short term regulation of cell-cell communication in TM3 Leydig cells. A perforated patch study.

Determination of the junctional conductance (g(j)) in TM3 Leydig cells by the dual whole cell patch clamp technique (DWCPC) shows that coupling undergoes a rapid and irreversible run down. Addition of ATP or cAMP derivatives to the pipette solution has been shown to prevent this phenomenon in several tissues, but this same treatment is unable to inhibit run down in Leydig cells. Because the run down in junctional conductance may pose serious problems to the interpretation of results, we also measured g(j) by using the double perforated patch clamp technique (DPPT). Access to the cell interior was achieved by adding 200 microgram/ml of nystatin to the pipette solution. With this method, run down in g(j) was greatly reduced, amounting to no more than 5% of the initial value. Exposure of the cells, under DWCPC or DPPT, to dibutyryl cAMP or to tumor promoting agent (TPA) led to a decrease in cell to cell communication. Staurosporine, a PKC inhibitor, increased g(j) and was able to prevent and reverse the uncoupling action of cAMP or TPA. Our results indicate that cell-cell communication in Leydig cells is down regulated by both protein kinases A and C, interacting in a complex manner.

Neurotensin modulates synaptic transmission in the nucleus of the solitary tract of the rat.

Whole-cell patch clamp recordings were made from neurons of the rat subpostremal region of the nucleus tractus solitarius (NTS) in transverse brainstem slices. Neurotensin (NT) enhanced the firing rate of action potentials from 0.8 +/- 0.4 Hz in control to 1.9 +/- 1.3 Hz (n = 9) and increased their decay time. The peak amplitude of the after-hyperpolarization was decreased by 34+/-5% (n = 9). These effects were associated with a depolarization of 4 +/- 1 mV (n = 10) in the resting membrane potential and an increase in the input resistance (from 768 +/- 220 MOmega to 986+/-220 MOmega; n = 5) and were compensated by manually hyperpolarizing the cell to control values. In voltage clamp experiments NT decreased an outward current (from 488 +/- 161 to 340 +/- 96 pA at +40 mV; n = 5) which reversed near the potassium equilibrium potential. In addition, NT increased the frequency of both excitatory and inhibitory spontaneous synaptic currents, an effect blocked by tetrodotoxin, and did not change the evoked excitatory or inhibitory postsynaptic currents. The selective NTR1 receptor antagonist SR48692 reversibly blocked the effects of NT on both action potentials and spontaneous synaptic currents. Our results suggest that NTR1 receptors can modulate post-synaptic responses in neurons of the subpostremal NTS by increasing cell excitability as a result of blockade of a potassium conductance.

Electrophysiological evidence for the presence of NR2C subunits of N-methyl-D-aspartate receptors in rat neurons of the nucleus tractus solitarius.

The nucleus tractus solitarius (NTS) plays an important role in the control of autonomic reflex functions. Glutamate, acting on N-methyl-D-aspartate (NMDA) and non-NMDA ionotropic receptors, is the major neurotransmitter in this nucleus, and the relative contribution of each receptor to signal transmission is unclear. We have examined NMDA excitatory postsynaptic currents (NMDA-EPSCs) in the subpostremal NTS using the whole cell patch clamp technique on a transverse brainstem slice preparation. The NMDA-EPSCs were evoked by stimulation of the solitary tract over a range of membrane potentials. The NMDA-EPSCs, isolated pharmacologically, presented the characteristic outward rectification and were completely blocked by 50 microM DL-2-amino-5-phosphonopentanoic acid. The I-V relationship of the NMDA response shows that current, with a mean (+/- SEM) amplitude of -41.2 +/- 5.5 pA, is present even at a holding potential of -60 mV, suggesting that the NMDA receptors are weakly blocked by extracellular Mg2+ at near resting membrane potentials. This weak block can also be inferred from the value of 0.67 +/- 0.17 for parameter delta obtained from a fit of the Woodhull equation to the I-V relationship. The maximal inward current measured on the I-V relationship was at -38.7 +/- 4.2 mV. The decay phase of the NMDA currents was fitted with one exponential function with a decay time constant of 239 +/- 51 and 418 +/- 80 ms at a holding potential of -60 and +50 mV, respectively, which became slower with depolarization (e-fold per 145 mV). The biophysical properties of the NMDA receptors observed in the present study suggest that these receptors in the NTS contain NR2C subunits and may contribute to the synaptic signal integration.

In vitro differentiation between oxytocin- and vasopressin-secreting magnocellular neurons requires more than one experimental criterion.

The phenotypic differentiation between oxytocin (OT)- and vasopressin (VP)-secreting magnocellular neurosecretory cells (MNCs) from the supraoptic nucleus is relevant to understanding how several physiological and pharmacological challenges affect their electrical activity. Although the firing patterns of OT and VP neurons, both in vivo and in vitro, may appear different from each other, much is assumed about their characteristics. These assumptions make it practically impossible to obtain a confident phenotypic differentiation based exclusively on the firing patterns. The presence of a sustained outward rectifying potassium current (SOR) and/or an inward rectifying hyperpolarization-activated current (IR), which are presumably present in OT neurons and absent in VP neurons, has been used to distinguish between the two types of MNCs in the past. In this study, we aimed to analyze the accuracy of the phenotypic discrimination of MNCs based on the presence of rectifying currents using comparisons with the molecular phenotype of the cells, as determined by single-cell RT-qPCR and immunohistochemistry. Our results demonstrated that the phenotypes classified according to the electrophysiological protocol in brain slices do not match their molecular counterparts because vasopressinergic and intermediate neurons also exhibit both outward and inward rectifying currents. In addition, we also show that MNCs can change the relative proportion of each cell phenotype when the system is challenged by chronic hypertonicity (70% water restriction for 7 days). We conclude that for in vitro preparations, the combination of mRNA detection and immunohistochemistry seems to be preferable when trying to characterize a single MNC phenotype.

Modulation of gap junction mediated intercellular communication in TM3 Leydig cells.

Long-term modulation of intercellular communication via gap junctions was investigated in TM3 Leydig cells, under low and high confluence states, and upon treatment of the cells for different times with activators of protein kinase A (PKA) and protein kinase C (PKC). Cells in low confluence were readily coupled, as determined by transfer of the dye Lucifer Yellow; on reaching confluence, the cells uncoupled. Western blots and RT-PCR revealed that connexin 43 (Cx43) was abundantly expressed in TM3 Leydig cells and its expression was decreased after the cells achieved confluence. Stimulation of PKA or PKC induced a decrease in cell-cell communication. Staurosporin, an inhibitor of protein kinases, increased coupling and was able to prevent and reverse the uncoupling actions of dibutyryl cAMP and 12-O-tetradecanoyl-phorbol-13-acetate (TPA). Under modulation by confluence, Cx43 was localized to the appositional membranes when cells were coupled and was mainly in the cytoplasm when they were uncoupled. In addition, cAMP and TPA reduced the surface membrane labeling for Cx43, whereas staurosporin increased it. These data show a strong correlation between functional coupling and the membrane distribution of Cx43, implying that this connexin has an important role in intercellular communication between TM3 cells. Furthermore, increased testosterone secretion in response to luteinizing hormone was accompanied by a decrease in intercellular communication, suggesting that gap junction mediated coupling may be a modulator of hormone secretion in TM3 cells.

TsTX-IV, a short chain four-disulfide-bridged neurotoxin from Tityus serrulatus venom which acts on Ca2+-activated K+ channels.

The primary structure of TsTX-IV, a neurotoxin isolated from Tityrus serrulatus scorpion venom, is reported. Its amino acid sequence was determined by automated Edman sequential degradation of the reduced and carboxymethylated toxin and of relevant peptides obtained by digestion with Staphylococcus aureus strain V8 protease or trypsin and cleavage by CNBr. The complete sequence showed 41 amino acid residues, which account for an estimated molecular weight of 4520, and eight half-cystine residues which cross-link the toxin molecule with four disulfide bonds. The molecular weight determined by mass spectrometry was 4518. Comparison of this sequence with those from other scorpion toxins showed a resemblance with toxins which act on different types of K+ channels. TsTx-IV was able to block Ca2+-activated K+ channels of high conductance. TsTX-IV is the first four-disulfide-bridged short toxin from T. serrulatus so far completely sequenced.

Calcium-activated potassium channels are involved in the response of mouse Leydig cells to human chorionic gonadotropin.

The patch-clamp technique was used to investigate the involvement of ion channels in the response of Leydig cells to gonadotropic hormones (viz. hCG). Recordings in the cell-attached configuration (pipette containing 140 mM KCl) showed unitary events with conductance of 187.9 +/- 5.2 pS (N = 24 patches) in about 70% of the cells. These channels were potassium selective and the open channel probability (Po) was always about 1% for displacement of potential from the resting value in the range of -20 to +60 mV. Treatment of the cells with hCG (2 ng/ml) led to a large increase in the frequency of openings, concomitant with a reduction in the mean closed time and there was essentially no effect on the mean open time of the channel. Dibutyryl cAMP (100 microM) produced an effect similar to that of hCG and both required external calcium for their action. No direct effect of either dibutyryl cAMP or hCG were observed in inside-out patches. Reversal potential measurements on excised inside-out patches demonstrated that the channels were highly potassium selective with unitary conductance of about 206.8 +/- 6.36 pS (mean +/- SEM of 6 measurements), and an estimated permeability of 3.6 x 10(-13) +/- 0.2 x 10(-13) cm3/s (mean +/- SEM for 6 measurements), in symmetrical 140 mM KCl. The activity of the channel in excised patches was very sensitive to the free-calcium concentration on the intracellular surface of the channel. Po evaluated at +60 mV increased from 3% at 10 nM to 47% at 100 nM free calcium. The Hill coefficient under these conditions was 1.1. These results demonstrate that Leydig cells have a Ca2(+)-activated K+ channel of large unitary conductance, which can be activated upon the binding of hCG to receptors in the cell membrane.

Patch-clamp detection of macromolecular translocation along nuclear pores.

The present paper reviews the application of patch-clamp principles to the detection and measurement of macromolecular translocation along the nuclear pores. We demonstrate that the tight-seal 'gigaseal' between the pipette tip and the nuclear membrane is possible in the presence of fully operational nuclear pores. We show that the ability to form a gigaseal in nucleus-attached configurations does not mean that only the activity of channels from the outer membrane of the nuclear envelope can be detected. Instead, we show that, in the presence of fully operational nuclear pores, it is likely that the large-conductance ion channel activity recorded derives from the nuclear pores. We conclude the technical section with the suggestion that the best way to demonstrate that the nuclear pores are responsible for ion channel activity is by showing with fluorescence microscopy the nuclear translocation of ions and small molecules and the exclusion of the same from the cisterna enclosed by the two membranes of the envelope. Since transcription factors and mRNAs, two major groups of nuclear macromolecules, use nuclear pores to enter and exit the nucleus and play essential roles in the control of gene activity and expression, this review should be useful to cell and molecular biologists interested in understanding how patch-clamp can be used to quantitate the translocation of such macromolecules into and out of the nucleus.

Ca(2+)-induced down-regulation of epithelial amiloride-sensitive channels reconstituted into planar lipid bilayers.

Multichannel experiments were carried out to investigate the Ca(2+)-induced down-regulation of epithelial Na+ channels reconstituted into planar lipid bilayer membranes. Reconstitution was achieved by fusion of vesiculated apical membrane fragments to solvent-free planar lipid bilayers. We found that the presence of micromolar concentrations of Ca2+ on the side to which the vesicles were added substantially lowered the channel-mediated current. The inhibition was strongly influenced by pH. At pH 8.0, all the current was blocked by 1 mM calcium, whereas at pH 7.1 the inhibition was about 80%. The blocking kinetics was clearly voltage-dependent. The mechanism of blocking cannot be explained either in terms of interactions with a single site, or by a model in which two blocking sites are assumed.

Hurst analysis in the study of ion channel kinetics.

Ion channels are protein molecules which can assume distinct open and closed conformational states. The transitions between these states can be controlled by the electrical field, ions and/or drugs. Records of unitary current events show that short open-time intervals are frequently adjacent to much longer closed-time intervals, and vice-versa, suggesting that the kinetic process has memory, i.e., the intervals are correlated in time. Here the rescaled range analysis (R/S Hurst analysis) is proposed as a method to test for correlation. Simulations were performed with a two-state Markovian model, which has no memory. The calculated Hurst coefficients (H) presented a mean +/- SD value of 0.493 +/- 0.025 (N = 100). For the Ca(2+)-activated K+ channels of Leydig cells, H was equal to 0.75, statistically different (1% level) from that calculated for the memoryless process. Randomly shuffling the experimental data resulted in an H = 0.55, not significantly different (1% level) from that found for the two-state Markovian model. For a linear three-state Markovian model, H was equal to 0.548 +/- 0.017 (N = 15), again not significantly different (1% level) from that of the memoryless process. Although the three-state Markovian model adequately describes the open- and closed-time distributions, it does not account for the correlation found in this Ca(2+)-activated K+ channel. Our results illustrate the efficacy of the R/S analysis in determining whether successive opening and closing events are correlated in time and can be of help in deciding which model should be used to describe the kinetics of ion channels.

Glomerular permeability to macromolecules in gentamicin-treated rats.

1. To determine the effect of gentamicin on the functional properties of the glomerular barrier, 44 Wistar rats received daily doses of 80 mg/kg body weight for 6 days. Glomerular permeability to neutral dextrans and albumin was evaluated by day 6 and albuminuria was determined on the 1st, 3rd and 5th days of treatment. 2. Treatment induced an intense increase in albuminuria from 74 micrograms/24 h to 11.5 mg/24 h on the 5th day of treatment (N = 11). This increase was associated with the presence of large amounts of albumin in elements of the glomerular filter and in the apical region of the proximal tubular cells (N = 4). Fractional clearances of neutral dextrans having molecular radii in the range of 18-41 A were not significantly different in control (N = 5) and gentamicin-treated rats (N = 7). 3. These results show that gentamicin, a polycation at pH 7.4, produces an increase in the glomerular permeability to negatively charged macromolecules in rats, probably due to interaction of the polycation with negative changes in the glomerular filter.

Effects of nitric oxide on magnocellular neurons of the supraoptic nucleus involve multiple mechanisms.

Physiological evidence indicates that the supraoptic nucleus (SON) is an important region for integrating information related to homeostasis of body fluids. Located bilaterally to the optic chiasm, this nucleus is composed of magnocellular neurosecretory cells (MNCs) responsible for the synthesis and release of vasopressin and oxytocin to the neurohypophysis. At the cellular level, the control of vasopressin and oxytocin release is directly linked to the firing frequency of MNCs. In general, we can say that the excitability of these cells can be controlled via two distinct mechanisms: 1) the intrinsic membrane properties of the MNCs themselves and 2) synaptic input from circumventricular organs that contain osmosensitive neurons. It has also been demonstrated that MNCs are sensitive to osmotic stimuli in the physiological range. Therefore, the study of their intrinsic membrane properties became imperative to explain the osmosensitivity of MNCs. In addition to this, the discovery that several neurotransmitters and neuropeptides can modulate their electrical activity greatly increased our knowledge about the role played by the MNCs in fluid homeostasis. In particular, nitric oxide (NO) may be an important player in fluid balance homeostasis, because it has been demonstrated that the enzyme responsible for its production has an increased activity following a hypertonic stimulation of the system. At the cellular level, NO has been shown to change the electrical excitability of MNCs. Therefore, in this review, we focus on some important points concerning nitrergic modulation of the neuroendocrine system, particularly the effects of NO on the SON.

Functional and structural study comparing the C-terminal amidated ß-neurotoxin Ts1 with its isoform Ts1-G isolated from Tityus serrulatus venom.

Mature Ts1, the main neurotoxin from Tityus serrulatus venom, has its C-terminal Cys amidated, while the isolated isoform of Ts1, named Ts1-G, keeps the non-amidated Gly residue at the C-terminal region, allowing the study of the comparative functional importance of amidation at the C-terminal between these two native toxins. Voltage dependent sodium current measurements showed that the affinity of Ts1-G for sodium channels is smaller than that of the mature Ts1, confirming the important role played by the C-terminal amidation in determining Ts1 activity.

Hypertonicity increases NO production to modulate the firing rate of magnocellular neurons of the supraoptic nucleus of rats.

Increases in plasma osmolality enhance nitric oxide (NO) levels in magnocellular neurosecretory cells (MNCs) of the supraoptic nucleus (SON) and modulate the secretion of both vasopressin (VP) and oxytocin (OT). In this paper, we describe the effects of hypertonicity on the electrical properties of MNCs by focusing on the nitrergic modulation of their activity in this condition. Membrane potentials were measured using the patch clamp technique, in the presence of both glutamatergic and GABAergic neurotransmission blockers, in coronal brain slices of male Wistar rats. The recordings were first made under a control condition (295 mosm/kg H2O), then in the presence of a hypertonic stimulus (330 mosm/kg H2O) and, finally, with a hypertonic stimulus plus 500 µM L-Arginine or 100 µM N-nitro-L-Arginine methyl ester hydrochloride (L-NAME). Hypertonicity per se increased the firing frequency of the neurons. L-Arginine prevented the increase in fire frequency induced by hypertonic stimulus, and L-NAME (inhibitor of nitric oxide synthase) induced an additional increase in frequency when applied together with the hypertonic solution. Moreover, L-Arginine hyperpolarizes the resting potential and decreases the peak value of the after-hyperpolarization; both effects were blocked by L-NAME and hypertonicity and/or L-NAME reduced the time constant of the rising phase of the after-depolarization. These results demonstrate that an intrinsic nitrergic system is part of the mechanisms controlling the excitability of MNCs of the SON when the internal fluid homeostasis is disturbed.

Purification and characterization of Ts15, the first member of a new α-KTX subfamily from the venom of the Brazilian scorpion Tityus serrulatus.

Voltage-gated potassium channel toxins (KTxs) are basic short chain peptides comprising 23-43 amino acid residues that can be cross-linked by 3 or 4 disulfide bridges. KTxs are classified into four large families: α-, ß-, γ- and κ-KTx. These peptides display varying selectivity and affinity for K(v) channel subtypes. In this work, a novel toxin from the Tityus serrulatus venom was isolated, characterized and submitted to a wide electrophysiological screening on 5 different subtypes of Na(V) channels (Na(V)1.4; Na(V)1.5; Na(V)1.6; Na(V)1.8 and DmNa(V)1) and 12 different subtypes of K(V) channels (K(V)1.1 - K(V)1.6; K(V)2.1; K(V)3.1; K(V)4.2; K(V)4.3; Shaker IR and ERG). This novel peptide, named Ts15, has 36 amino acids, is cross-linked by 3 disulfide bridges, has a molecular mass of 3956 Da and pI around 9. Electrophysiological experiments using patch clamp and the two-electrode voltage clamp techniques show that Ts15 preferentially blocks K(V)1.2 and K(V)1.3 channels with an IC50 value of 196 ± 25 and 508 ± 67 nM, respectively. No effect on Na(V) channels was observed, at all tested concentrations. Since Ts15 shows low amino acid identity with other known KTxs, it was considered a bona fide novel type of scorpion toxin. Ts15 is the unique member of the new α-Ktx21 subfamily and therefore was classified as α-Ktx21.1.

P2X4 receptors interact with both P2X2 and P2X7 receptors in the form of homotrimers.

BACKGROUND AND PURPOSE: The P2X receptor family consists of seven subunit types - P2X1-P2X7. All but P2X6 are able to assemble as homotrimers. In addition, various subunit permutations have been reported to form heterotrimers. Evidence for heterotrimer formation includes co-localization, co-immunoprecipitation and the generation of receptors with novel functional properties; however, direct structural evidence for heteromer formation, such as chemical cross-linking and single-molecule imaging, is available in only a few cases. Here we examined the nature of the interaction between two pairs of subunits - P2X2 and P2X4, and P2X4 and P2X7. EXPERIMENTAL APPROACH: We used several experimental approaches, including in situ proximity ligation, co-immunoprecipitation, co-isolation on affinity beads, chemical cross-linking and atomic force microscopy (AFM) imaging. KEY RESULTS: Both pairs of subunits co-localize upon co-transfection, interact intimately within cells, and can be co-immunoprecipitated and co-isolated from cell extracts. Despite this, chemical cross-linking failed to show evidence for heteromer formation. AFM imaging of isolated receptors showed that all three subunits had the propensity to form receptor dimers. This self-association is likely to account for the observed close interaction between the subunit pairs, in the absence of true heteromer formation. CONCLUSIONS AND IMPLICATIONS: We conclude that both pairs of receptors interact in the form of distinct homomers. We urge caution in the interpretation of biochemical evidence indicating heteromer formation in other cases.