Thirty-seventh Annual Meeting February 14-18, 1993 Washington Convention Center Washington, D.C. Wednesday Symposia and Posters, Part II

1993 Biophysical Journal  
0654. cAMP-dependent protein kinase (cAPK) is one of the simplest members of a family of enzymes that play critical regulatory roles in the eukaryotic cell. cAPK, being one of the smallest protein kinases, has served as a prototype for the entire family of over 200 enzymes. The catalytic (C) subunit of cAPK is itself also subject to posttranslational modifications, both phosphorylation and myristyladon. The crystal structure of a ternary complex containing the Csubunit, an inhibitor peptide,
more » ... MgATP shows how key conserved residues converge at the active site. It also describes the sites of posttranslational modification. Two stable phosphorylation sites exist and replacing these phosphates leads to a destabilized or inactive enzyme that, in some cases, is unable to recognize the regulatory (R) subunit. Phosphorylation at Thrl97 is particularly critical and a homologous residue, critical for activation, is found in other protein kinases such as the cell division cycle kinases, cdc2. Myristylation stabilizes the enzyme by folding into a hydrophobic pocket. The inhibition of the C subunit by either the type I R-subunit or a heat stable protein kinase inhibitor (PKI) depends on the synergistic high affinity binding of MgATP. Mutant C-subunits that are catalytically intact but unable to be regulated by either PKI or R, we can demonstrate that different regions of the enzyme surface are important for the high affinity binding of these two physiological inhibitom The dissociated C-subunit is free to migrate between the cytoplasm and the nucleus while association with either R or PKI prevents access to the nucleus. Microinjecting directly into the nucleus shows that free C-subunit can move in both directions while R and the holoenzyme cannot escape from the nucleus. Multifunctional Ca2t/calmodulin-dependent protein kinase (CaM kinase) is one of the major protein kinases coordinating celular responses to honnones and neurotransmitters. CaM kinase has structuralfunctional properties that may facilitate its rsponse to changes in Call, that are transient or pulsatfle. Substrate phosphorylation and autophoaphorylation are stimulated when calmodulin binds and disrupts the autoinhibitory/calmodulin-binding domain. After calmodulin dissociates from an autophosphorylated subunit, the kinase no longer deactivates because its autoinhibitory domain is disrupted by the anionic phosphate moiety. The effects of transient rises in Ca2 are potentiated by the conversion of CaM kinase to this Ca2+-independent form. The frequency of Ca2+ oscillations or spikes may be decoded by CaM kinase. The affinity of CaM kinase for calmodulin increases more than 100-fold by autophosphorylation which traps calmodulin by increasing its off-rate from less than a second to several seconds. Trapping is a cooperative process that is inefficient at low occupancy of calmodulin on the multimeric kinase. Cooperativity results from the mechanism of autophosphorylation which involves a subunit phosphorylatng its neighbor in the holoenzyme, each with bound calmodulin. Simulations of kinase activation at limiting calmodulin show that repetitive Ca2+ pulses lead to recruitment of calmodulin to the holoenzyme, which further stimulates autophosphorylation and trapping. This cooperative positive feedback loop potentiates the response of the kinase to sequential calcium spikes and establishes a theshold frequency at which the enzyme becomes highly active. The biochemical properties of CaM kinase provide for molecular potentiation of calcium signals and frequency detection. Voltage and calcium-gated potassium channels are very sensitive to blockade by internal Ba+. The similarity in dehydrated ionic radius of Ba++ and K+ suggests that Ba+ binds to at least one of the sites where K+ binds as it moves through the pore. In order to identify the part of the pore which interacts with Ball, we have made a chimeric channel in which parts of the NGK2 (Kv3. 1) channel were transplanted into a noninactivating N-terminal deletion mutant of Shaker B, since the two channels were found to exhibit different conductances to K+ and sensitivities to block by internal Ba+. A chimera, in which most of the S6 and several amino acids on its carboxyl end were transplanted, displayed a sensitivity to Ba+ different from Shaker B. Multiple point mutations have been made to identify the responsible residue(s). We are examining the voltage-dependence of the Ba++ block in order to estimate the location along the length of the pore where the relevant residue(s) exert their effect. We are also examining Ba'" on and off rates in order to gauge whether the effect of Ba"+ affmity can be ascribed to changes in the Ba++ binding site directly or in Ba+ access to the site. W4PMWA5 A loucine zipper motif, five lucines spaced seven residues apart, is closely associated with the S4 domain of most known K-channels. The S4 domain Is the putative voltage sensor of the molecule. We found that in RCK1 K-channels, mutation of leudnes 2 and 4 into valines (V2, V4) causes a shift in the activation curve of +70 mV and -45 mV, respectively. In these experiments, the shift was measured at an open probability of 0.1. Dimeric constructs containing one wild-type and one mutant channel (WTV2, WTV4) In a single reading frame show intermediate phenotypes with shifts of +47 mV and -20 mV. However, data from monomeric and dimeric constructs alone do not allow us to postulate a model which describes the role of these residues in gating. Using a sequential model which assumes four voltage sensors per channel and cooperative subunit interactions, the data can be fitted by adjusting the amount of gating charge (and/or Introducing charge separation), the midpoint of activation, the degree of cooperativity, or a combination of the three. We are currently engineering and measuring various tetrameric constructs containing one, two, or three mutant channels In a single reading frame. This should enable us to determine more accurately what role the hydrophobic residues play in the process of voltage-dependent gating. Delayed recfier curres wer measued in the extensor digitonam longus muscle (EDL) of contl (CS7BUL10Sn) and mdx mice (an animal model of muscular distrophy) using the cell-attached patch clamp tecniue (see Mathes et al., 1991, Am. J. Physiol, 30:C718) with physiological Kvconcentratons (bath 150 mM, pipette 2-3 mM). cunents were meaured from macro patches (up to SIL diameter pipetes). Outward currens could be detected at potenials positive to -60 mV, with a V Of -37 mV in contl muscle. In mdx both the probability of opening (Vur-32 mV) and the t of activation as function of V seemed to be shifted to the righL Tail currents (IT) were Hlnear in the range of -60 mV to 20 mV. Below -60 mV, a rctfication was apparn reversed arund -100 mV. Meanvarance analysis estimates of single channel cument at V=0 yielded a normalized conductnce of 20 pS for the line portion of the 17V curve. hacivation was at potnts posiive to -10 mV with puses as shortja,18 msec in durtion. Prepulses of 500 msec yielded incomplete inactiation, with V.;=-55 mV. At VO, inactvaon dispays two t's, of 70 msec and 600 msec. Althwugh single channel data confirmed the rsults obtained with the macroscopic data, a smaller ampliuide channel was also observed with less frequency. Upon excision, the delayed rctfier tended to nun down with a time constant of tens of minutes. ATP In the bath solution only marginally incrased the probability of opening of the channels after rundown, and protein kiase A with ATP did not seem better than ATP sbus. These findings indicate that the delayed rctfier is under complex inrcellular conol, and may be difficult to characterize in excised patches. The characteristics of 4-aminopyridine (4-AP) block of voltage-gated r channes vary widely among channel subtypes and the 4-AP site in the channels is not knowvn We examined 4-AP's mechanism and blocking site In Kv2.1 and Kv3.1 r channels cloned from rat brain and expressed in Xenopus oocytes. Patch clamp analysis of Kv3.1 showed that 4-AP ON and OFF rates were at 1ea 100 times faster in open: channeis than in dosed channels. An intacellular oute to the bloccing site Is Ubdyl because the bath-applied, trtary amine readily blocks channels in cell-attaded patches while its quatemary derivative selectively blocks when applied to the intracellular surface of excised patches Whole cel analysis showed at 4-AP blocked Kv3.1 with ICn. more than 300-fold lower than Kv2.l. But unlike extwnl investigate more in detail the gating m When activation curovs were fitted by a sequential model that includes 4 voltage sensors per channel, phenotypes could be predicted correctly if cooperative subunit interactions were assumed for WT, R2N and R31 subunits, in contrast to non-cooperative bebaviour for KSI and R6N subunits. Gating valences for K5I and R6N subunits were not different from WI, suggestng that KS and R6 are not part*f the voltage sensor. Phenotypes for K currents expressed from tetramers with different combinatons of WT and RIQ (1 Arg in S4 neutralized) subunits could he explained if RlQ subunits were gatng in a non-cooperative way. In conclusion, the tetrameric cDNA approac has enabled us to invstigate the contnbution of all charged ammo acId residues in S4 to the total gating valence of RCK1 K channels. A mechanism of cooperative and non-cooperative interactions between WT and mutant subunits helps to explain the voltage-dependence of the activation process. We have identified a rapidly inactivating voltage-gated K' current (I,f,) in smooth muscle ceiasof the guinea-pig proximal colon. Smooth muscle cell were bathed in a Ca2+-ontaining HEPES-buffered solution which also contained tetraethylammnium (IEA; 5-12 mM) and either Cd2' (0.1 mM) or nifedipine (2 gl) to blockcaldium channels. Cell were internaly perfused with a HEPESbuffered Ka solution containing 3 mM EGTA. At room temperature (2?C) activation of Iy" by a test depolarization to +40 mV from a holding potential of -80 mV, was compet within 2-S ms, and the current inactivated with two time constants: 19 ± 1.8 ms and 86 ± 12 ms (n=6). Noradrenaline (NA5OgiM) reduced In. slowly over 10-15 min: at a test depolarization of +40 mV, ly. wa reduced significantly (p<0.05, paired t-test) from 563 ± 17 pA to 453±: 13 pA (n-4). Ioprenaline (ISO; 50 pM) also significantly decreased (p<0.0S, paired t-test) Ix, from 468 t 96 pA to 391 ± 81 pA (n=10), an action that was blocked by propranolol (10 pIM). Replacement of Ca2W% with Mg2e+ decreased the lateny of ISO-mediated reduction in IKur The voltage of half availability (Vu) was littlc affected by ISO (81 ± 3 mV to -84 ± 4 mV, n=6). Substitution of Na+, with W* abolished the ISO-mediated reduction in Iro Forskolin (100 lM) significantly decrased (p< 0.05, paired t-test) lr, from 383 ± 91 pA to 267 ±Sl1 pA (n=4) byshiftingV.. negative from-1.8 ± 5 mV to-89± 6 mV. Ihese data suggest that the 8-adrenoreceptor activation of adenylate cydase in colonic myocy causes inhibition of I.,. A226
doi:10.1016/s0006-3495(93)81409-7 fatcat:4gzbrrjo2jc53cloy25uxl7yxq