Interlocked feedback loops contribute to the robustness of the Neurospora circadian clock

Interlocked feedback loops may represent a common feature among the regulatory systems controlling circadian rhythms. The Neurospora circadian feedback loops involve white collar-1 (wc-1), wc-2, and frequency (frq) genes. We show that WC-1 and WC-2 proteins activate the transcription of frq gene, whereas FRQ protein plays dual roles: repressing its own transcription, probably by interacting with the WC-1/WC-2 complex, and activating the expression of both WC proteins. Thus, they form two interlocked feedback loops: one negative and one positive. We establish the physiological significance of the interlocked positive feedback loops by showing that the levels of WC-1 and WC-2 determine the robustness and stability of the clock. Our data demonstrate that with WC-1 being the limiting factor in the WC-1/WC-2 complex, the greater the levels of WC-1 and WC-2, the higher the level of the FRQ oscillation and the more robust the overt rhythms. Our data also show that, despite considerable changes in the levels of WC-1, WC-2, and FRQ, the period of the clock has been limited to a small range, suggesting that the interlocked circadian feedback loops are also important for determining the circadian period length of the clock.

Antillatoxin is a marine cyanobacterial toxin that potently activates voltage-gated sodium channels

Antillatoxin (ATX) is a lipopeptide derived from the pantropical marine cyanobacterium Lyngbya majuscula. ATX is neurotoxic in primary cultures of rat cerebellar granule cells, and this neuronal death is prevented by either N-methyl-d-aspartate (NMDA) receptor antagonists or tetrodotoxin. To further explore the potential interaction of ATX with voltage-gated sodium channels, we assessed the influence of tetrodotoxin on ATX-induced Ca2+ influx in cerebellar granule cells. The rapid increase in intracellular Ca2+ produced by ATX (100 nM) was antagonized in a concentration-dependent manner by tetrodotoxin. Additional, more direct, evidence for an interaction with voltage-gated sodium channels was derived from the ATX-induced allosteric enhancement of [3H]batrachotoxin binding to neurotoxin site 2 of the α subunit of the sodium channel. ATX, moreover, produced a strong synergistic stimulation of [3H]batrachotoxin binding in combination with brevetoxin, which is a ligand for neurotoxin site 5 on the voltage-gated sodium channel. Positive allosteric interactions were not observed between ATX and either α-scorpion toxin or the pyrethroid deltamethrin. That ATX interaction with voltage-gated sodium channels produces a gain of function was demonstrated by the concentration-dependent and tetrodotoxin-sensitive stimulation of 22Na+ influx in cerebellar granule cells exposed to ATX. Together these results demonstrate that the lipopeptide ATX is an activator of voltage-gated sodium channels. The neurotoxic actions of ATX therefore resemble those of brevetoxins that produce neural insult through depolarization-evoked Na+ load, glutamate release, relief of Mg2+ block of NMDA receptors, and Ca2 + influx.

Interplay of structure and disorder in cochaperonin mobile loops.

Protein-protein interactions typically are characterized by highly specific interfaces that mediate binding with precisely tuned affinities. Binding of the Escherichia coli cochaperonin GroES to chaperonin GroEL is mediated, at least in part, by a mobile polypeptide loop in GroES that becomes immobilized in the GroEL/GroES/nucleotide complex. The bacteriophage T4 cochaperonin Gp31 possesses a similar highly flexible polypeptide loop in a region of the protein that shows low, but significant, amino acid similarity with GroES and other cochaperonins. When bound to GroEL, a synthetic peptide representing the mobile loop of either GroES or Gp31 adopts a characteristic bulged hairpin conformation as determined by transferred nuclear Overhauser effects in NMR spectra. Thermodynamic considerations suggest that flexible disorder in the cochaperonin mobile loops moderates their affinity for GroEL to facilitate cycles of chaperonin-mediated protein folding.

NO hyperpolarizes pulmonary artery smooth muscle cells and decreases the intracellular Ca2+ concentration by activating voltage-gated K+ channels.

NO causes pulmonary vasodilation in patients with pulmonary hypertension. In pulmonary arterial smooth muscle cells, the activity of voltage-gated K+ (Kv) channels controls resting membrane potential. In turn, membrane potential is an important regulator of the intracellular free calcium concentration ([Ca2+]i) and pulmonary vascular tone. We used patch clamp methods to determine whether the NO-induced pulmonary vasodilation is mediated by activation of Kv channels. Quantitative fluorescence microscopy was employed to test the effect of NO on the depolarization-induced rise in [Ca2+]i. Blockade of Kv channels by 4-aminopyridine (5 mM) depolarized pulmonary artery myocytes to threshold for initiation of Ca2+ action potentials, and thereby increased [Ca2+]i. NO (approximately 3 microM) and the NO-generating compound sodium nitroprusside (5-10 microM) opened Kv channels in rat pulmonary artery smooth muscle cells. The enhanced K+ currents then hyperpolarized the cells, and blocked Ca(2+)-dependent action potentials, thereby preventing the evoked increases in [Ca2+]i. Nitroprusside also increased the probability of Kv channel opening in excised, outside-out membrane patches. This raises the possibility that NO may act either directly on the channel protein or on a closely associated molecule rather than via soluble guanylate cyclase. In isolated pulmonary arteries, 4-aminopyridine significantly inhibited NO-induced relaxation. We conclude that NO promotes the opening of Kv channels in pulmonary arterial smooth muscle cells. The resulting membrane hyperpolarization, which lowers [Ca2+]i, is apparently one of the mechanisms by which NO induces pulmonary vasodilation.

Elevation of intracellular calcium by muscarinic receptor activation induces a block of voltage-activated rat ether-à-go-go channels in a stably transfected cell line.

We have studied the properties of r-eag voltage-activated potassium channels in a stably transfected human embryonic kidney cell line. It was found that r-eag channels are rapidly and reversibly inhibited by a rise in intracellular calcium from 30 to 300 nM. The inhibition does not appear to depend on the activity of calcium-dependent kinases and phosphatases. The effect of calcium on r-eag channel activity was studied in inside-out membrane patches. Calcium inhibited r-eag channel activity with a mean IC50 of 67 nM. Activation of muscarinic receptors, generating calcium oscillations in the transfected cells, induced a synchronous inhibition of r-eag mediated outward currents. This shows that calcium can mediate r-eag current inhibition following muscarinic receptor activation. The data indicate that r-eag channels are calcium-inhibitable voltage-activated potassium channels.

Lambert-Eaton sera reduce low-voltage and high-voltage activated Ca2+ currents in murine dorsal root ganglion neurons.

Voltage-gated Ca2+ channels are categorized as either high-voltage activated (HVA) or low-voltage activated (LVA), and a subtype (or subtypes) of HVA Ca2+ channels link the presynaptic depolarization to rapid neuro-transmitter release. Reductions in transmitter release are characteristic of the autoimmune disorder, Lambert-Eaton syndrome (LES). Because antibodies from LES patients reduce Ca2+ influx in a variety of cell types and disrupt the intramembrane organization of active zones at neuromuscular synapses, specificity of LES antibodies for the Ca2+ channels that control transmitter release has been suggested as the mechanism for disease. We tested sera from four patients with LES. Serum samples from three of the four patients reduced both the maximal LVA and HVA Ca2+ conductances in murine dorsal root ganglion neurons. Thus, even though LES is expressed as a neuromuscular and autonomic disorder, our studies suggest that Ca2+ channels may be broadly affected in LES patients. To account for the specificity of disease expression, we suggest that incapacitation of only a fraction of the Ca2+ channels clustered at active zones would severely depress transmitter release. In particular, if several Ca2+ channels in a cluster are normally required to open simultaneously before transmitter release becomes likely, the loss of a few active zone Ca2+ channels would exponentially reduce the probability of transmitter release. This model may explain why LES is expressed as a neuromuscular disorder and can account for a clinical hallmark of LES, facilitation of neuromuscular transmission produced by vigorous voluntary effort.

Common molecular determinants of local anesthetic, antiarrhythmic, and anticonvulsant block of voltage-gated Na+ channels.

Voltage-gated Na+ channels are the molecular targets of local anesthetics, class I antiarrhythmic drugs, and some anticonvulsants. These chemically diverse drugs inhibit Na+ channels with complex voltage- and frequency-dependent properties that reflect preferential drug binding to open and inactivated channel states. The site-directed mutations F1764A and Y1771A in transmembrane segment IVS6 of type IIA Na+ channel alpha subunits dramatically reduce the affinity of inactivated channels for the local anesthetic etidocaine. In this study, we show that these mutations also greatly reduce the sensitivity of Na+ channels to state-dependent block by the class Ib antiarrhythmic drug lidocaine and the anticonvulsant phenytoin and, to a lesser extent, reduce the sensitivity to block by the class Ia and Ic antiarrhythmic drugs quinidine and flecainide. For lidocaine and phenytoin, which bind preferentially to inactivated Na+ channels, the mutation F1764A reduced the affinity for binding to the inactivated state 24.5-fold and 8.3-fold, respectively, while Y1771A had smaller effects. For quinidine and flecainide, which bind preferentially to the open Na+ channels, the mutations F1764A and Y1771A reduced the affinity for binding to the open state 2- to 3-fold. Thus, F1764 and Y1771 are common molecular determinants of state-dependent binding of diverse drugs including lidocaine, phenytoin, flecainide, and quinidine, suggesting that these drugs interact with a common receptor site. However, the different magnitude of the effects of these mutations on binding of the individual drugs indicates that they interact in an overlapping, but nonidentical, manner with a common receptor site. These results further define the contributions of F1764 and Y1771 to a complex drug receptor site in the pore of Na+ channels.

Changes in voltage activation, Cs+ sensitivity, and ion permeability in H5 mutants of the plant K+ channel KAT1.

KAT1 is a voltage-dependent inward rectifying K+ channel cloned from the higher plant Arabidopsis thaliana [Anderson, J. A., Huprikar, S. S., Kochian, L. V., Lucas, W. J. & Gaber, R. F. (1992) Proc. Natl. Acad. Sci. USA 89, 3736-3740]. It is related to the Shaker superfamily of K+ channels characterized by six transmembrane spanning domains (S1-S6) and a putative pore-forming region between S5 and S6 (H5). The 115 region between Pro-247 and Pro-271 in KAT1 contains 14 additional amino acids when compared with Shaker [Aldrich, R. W. (1993) Nature (London) 362, 107-108]. We studied various point mutations introduced into H5 to determine whether voltage-dependent plant and animal K+ channels share similar pore structures. Through heterologous expression in Xenopus oocytes and voltage-clamp analysis combined with phenotypic analysis involving a potassium transport-defective Saccharomyces cerevisiae strain, we investigated the selectivity filter of the mutants and their susceptibility toward inhibition by cesium and calcium ions. With respect to electrophysiological properties, KAT1 mutants segregated into three groups: (i) wild-type-like channels, (ii) channels modified in selectivity and Cs+ or Ca2+ sensitivity, and (iii) a group that was additionally affected in its voltage dependence. Despite the additional 14 amino acids in H5, this motif in KAT1 is also involved in the formation of the ion-conducting pore because amino acid substitutions at Leu-251, Thr-256, Thr-259, and Thr-260 resulted in functional channels with modified ionic selectivity and inhibition. Creation of Ca2+ sensitivity and an increased susceptibility to Cs+ block through mutations within the narrow pore might indicate that both blockers move deeply into the channel. Furthermore, mutations close to the rim of the pore affecting the half-activation potential (U1/2) indicate that amino acids within the pore either interact with the voltage sensor or ion permeation feeds back on gating.

Adjacent pore-lining residues within sodium channels identified by paired cysteine mutagenesis.

The pores of voltage-gated ion channels are lined by protein loops that determine selectivity and conductance. The relative orientations of these "P" loops remain uncertain, as do the distances between them. Using site-directed mutagenesis, we introduced pairs of cysteines into the P loops of micro1 rat skeletal muscle sodium channels and sought functional evidence of proximity between the substituted residues. Only cysteinyl residues that are in close proximity can form disulfide bonds or metal-chelating sites. The mutant Y401C (domain I) spontaneously formed a disulfide bond when paired with E758C in the P loop of domain II; the same residue, when coupled with G1530C in domain IV, created a high-affinity binding site for Cd2+ ions. The results provide the first specific constraints for intramolecular dimensions of the sodium channel pore.

Voltage gating and permeation in a gap junction hemichannel.

Gap junction channels are formed by members of the connexin gene family and mediate direct intercellular communication through linked hemichannels (connexons) from each of two adjacent cells. While for most connexins, the hemichannels appear to require an apposing hemichannel to open, macroscopic currents obtained from Xenopus oocytes expressing rat Cx46 suggested that some hemichannels can be readily opened by membrane depolarization [Paul, D. L., Ebihara, L., Takemoto, L. J., Swenson, K. I. & Goodenough, D. A. (1991), J. Cell Biol. 115, 1077-1089]. Here we demonstrate by single channel recording that hemichannels comprised of rat Cx46 exhibit complex voltage gating consistent with there being two distinct gating mechanisms. One mechanism partially closes Cx46 hemichannels from a fully open state, gammaopen, to a substate, gammasub, about one-third of the conductance of gammaopen; these transitions occur when the cell is depolarized to inside positive voltages, consistent with gating by transjunctional voltage in Cx46 gap junctions. The other gating mechanism closes Cx46 hemichannels to a fully closed state, gammaclosed, on hyperpolarization to inside negative voltages and has unusual characteristics; transitions between gammaclosed and gammaopen appear slow (10-20 ms), often involving several transient substates distinct from gammasub. The polarity of activation and kinetics of this latter form of gating indicate that it is the mechanism by which these hemichannels open in the cell surface membrane when unapposed by another hemichannel. Cx46 hemichannels display a substantial preference for cations over anions, yet have a large unitary conductance (approximately 300 pS) and a relatively large pore as inferred from permeability to tetraethylammonium (approximately 8.5 angstroms diameter). These hemichannels open at physiological voltages and could induce substantial cation fluxes in cells expressing Cx46.

Antigenic peptides containing large PEG loops designed to extend out of the HLA-A2 binding site form stable complexes with class I major histocompatibility complex molecules.

Recognition of peptides bound to class I major histocompatibility complex (MHC) molecules by specific receptors on T cells regulates the development and activity of the cellular immune system. We have designed and synthesized de novo cyclic peptides that incorporate PEG in the ring structure for binding to class I MHC molecules. The large PEG loops are positioned to extend out of the peptide binding site, thus creating steric effects aimed at preventing the recognition of class I MHC complexes by T-cell receptors. Peptides were synthesized and cyclized on polymer support using high molecular weight symmetrical PEG dicarboxylic acids to link the side chains of lysine residues substituted at positions 4 and 8 in the sequence of the HLA-A2-restricted human T-lymphotrophic virus type I Tax peptide. Cyclic peptides promoted the in vitro folding and assembly of HLA-A2 complexes. Thermal denaturation studies using circular dichroism spectroscopy showed that these complexes are as stable as complexes formed with antigenic peptides.

The role of surface loops (residues 204-216 and 627-646) in the motor function of the myosin head.

A characteristic feature of all myosins is the presence of two sequences which despite considerable variations in length and composition can be aligned with loops 1 (residues 204-216) and 2 (residues 627-646) in the chicken myosin-head heavy chain sequence. Recently, an intriguing hypothesis has been put forth suggesting that diverse performances of myosin motors are achieved through variations in the sequences of loops 1 and 2 [Spudich, J. (1994) Nature (London) 372, 515-518]. Here, we report on the study of the effects of tryptic digestion of these loops on the motor and enzymatic functions of myosin. Tryptic digestions of myosin, which produced heavy meromyosin (HMM) with different percentages of molecules cleaved at both loop 1 and loop 2, resulted in the consistent decrease in the sliding velocity of actin filaments over HMM in the in vitro motility assays, did not affect the Vmax, and increased the Km values for actin-activated ATPase of HMM. Selective cleavage of loop 2 on HMM decreased its affinity for actin but did not change the sliding velocity of actin in the in vitro motility assays. The cleavage of loop 1 and HMM decreased the mean sliding velocity of actin in such assays by almost 50% but did not alter its affinity for HMM. To test for a possible kinetic determinant of the change in motility, 1-N6-ethenoadenosine diphosphate (epsilon-ADP) release from cleaved and uncleaved myosin subfragment 1 (S1) was examined. Tryptic digestion of loop 1 slightly accelerated the release of epsilon-ADP from S1 but did not affect the rate of epsilon-ADP release from acto-S1 complex. Overall, the results of this work support the hypothesis that loop 1 can modulate the motor function of myosin and suggest that such modulation involves a mechanism other than regulation of ADP release from myosin.

Stimulation of growth factor receptor signal transduction by activation of voltage-sensitive calcium channels.

To understand the mechanisms by which electrical activity may generate long-term responses in the nervous system, we examined how activation of voltage-sensitive calcium channels (VSCCs) can stimulate the Ras/mitogen-activated protein kinase (MAPK) signaling pathway. Calcium influx through L-type VSCCs leads to tyrosine phosphorylation of the adaptor protein Shc and its association with the adaptor protein Grb2, which is bound to the guanine nucleotide exchange factor Sos1. In response to calcium influx, Shc, Grb2, and Sos1 inducibly associate with a 180-kDa tyrosine-phosphorylated protein, which was determined to be the epidermal growth factor receptor (EGFR). Calcium influx induces tyrosine phosphorylation of the EGFR to levels that can activate the MAPK signaling pathway. Thus, ion channel activation stimulates growth factor receptor signal transduction.

Membrane transport mechanisms probed by capacitance measurements with megahertz voltage clamp.

We have used capacitance measurements with a 1-microsecond voltage clamp technique to probe electrogenic ion-transporter interactions in giant excised membrane patches. The hydrophobic ion dipicrylamine was used to test model predictions for a simple charge-moving reaction. The voltage and frequency dependencies of the apparent dipicrylamine-induced capacitance, monitored by 1-mV sinusoidal perturbations, correspond to single charges moving across 76% of the membrane field at a rate of 9500 s-1 at 0 mV. For the cardiac Na,K pump, the combined presence of cytoplasmic ATP and sodium induces an increase of apparent membrane capacitance which requires the presence of extracellular sodium. The dependencies of capacitance changes on frequency, voltage, ATP, and sodium verify that phosphorylation enables a slow, 300- to 900-s-1, pump transition (the E1-E2 conformational change), which in turn enables fast, electrogenic, extracellular sodium binding reactions. For the GAT1 (gamma-aminobutyric acid,Na,Cl) cotransporter, expressed in Xenopus oocyte membrane, we find that chloride binding from the cytoplasmic side, and probably sodium binding from the extracellular side, results in a decrease of membrane capacitance monitored with 1- to 50-kHz perturbation frequencies. Evidently, ion binding by the GAT1 transporter suppresses an intrinsic fast charge movement which may originate from a mobility of charged residues of the transporter binding sites. The results demonstrate that fast capacitance measurements can provide new insight into electrogenic processes closely associated with ion binding by membrane transporters.

DNA loops induced by cooperative binding of transcriptional activator proteins and preinitiation complexes.

DNA conformational changes are essential for the assembly of multiprotein complexes that contact several DNA sequence elements. An approach based on atomic force microscopy was chosen to visualize specific protein-DNA interactions occurring on eukaryotic class II nuclear gene promoters. Here we report that binding of the transcription regulatory protein Jun to linearized plasmid DNA containing the consensus AP-1 binding site upstream of a class II gene promoter leads to bending of the DNA template. This binding of Jun was found to be essential for the formation of preinitiation complexes (PICs). The cooperative binding of Jun and PIC led to looping of DNA at the protein binding sites. These loops were not seen in the absence of either PICs, Jun, or the AP-1 binding site, suggesting a direct interaction between DNA-bound Jun homodimers and proteins bound to the core promoter. This direct visualization of functional transcriptional complexes confirms the theoretical predictions for the mode of gene regulation by trans-activating proteins.