Voltage Gated Calcium Channel

Embed Size (px)

Citation preview

  • 8/12/2019 Voltage Gated Calcium Channel

    1/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 1

    NCBI Bookshelf. A serv ice of the Nationa l Libra ry of Medicine, Na tiona l Instit ut es of Healt h.

    Madame Curie Bioscienc e Database [Internet]. A ustin (TX): Landes Bioscience ; 2000 -.

    Molecular Properties of Voltage-Gated Calcium Channels

    Terrance P. Snutch, Jean Peloquin, Eleanor Mathews, and John E. McRory.

    Native Voltage-Gated Ca Channels

    Early electro physiologicalrecordings from neurons, muscle and endocrine cells revealed voltage-activated calcium (Ca )

    currents with distinct characteristics, suggesting the existence of two major classes of Ca channels based upon the

    membrane potentials at which they first open (see c hapterby Tsien); low-v oltage ac tiv ated (LV A) and high-voltageactivated (HVA). The LVA (or T-type) channels, typically hav e a small conductance (8-12 pico Siemens (pS)), open inresponse to small changes from the resting membrane potential and inactivate rapidly. In c ontrast, the HVA currentsgenerally possess larger conduc tances (15-25 pS), are activ ated by stronger depolarizations and display v ariableinactivation kinetics. To date, multiple types of HVA Ca channels (L-, N-, P/Q- and R-type) have been categor ized on the

    basis o f anumber o f criteria including, single c hannel conductance, kinetics, pharmacology , and c ellular distribution.

    High Voltage-Activated Ca Channels

    L-Type Channels

    L-type Ca channels were initially described in peripheral neurons and cardiac c ells, but appear to be pr esent in all

    exc itable as well as many ty pes of non-ex citable c ells. In certain ce lls, L-ty pe channels hav e been sho wn to bepreferentially loc alized to specific subc ellular regions. For example, the L-type channels responsible for skeletal musclecontraction areconc entrated on the transverse tubule membrane, while neuronal L-type channels are located primarily

    on cell bodies and prox imal dendrites. The L-type channel is the primary route for Ca entry into cardiac, skeletal, and

    smooth muscles. The skeletal muscle L-type channel acts as a vo ltage sensor for ex citation-contraction (E-C) co upling in

    skeletal muscle, presumably linking membrane depolarization to Ca release from intracellular stores. While Ca entry

    through this channel is not required for the initiation of contraction in skeletal muscle, it may prov ide a source of Ca to

    replenish internal stores. There is some evidence that L-type channels are involved in exocy totic release from

    endocrine cells and some neurons and the localization of L-type channels on the cell soma has also implicated these

    channels in the regulation of gene ex pression.

    Much is known about the pharmacological properties of L-type Ca channels. The three main classes of organic L-type

    channel blockers are the phenylalkylamines (verapamil), benzothiazapines (diltiazem), and 1,4-dihydropyridines (DHPs)(e.g., nitrendipine, nifedipine, nimodipine). The DHP antagonists bind preferentially to channels in the active conformation,

    a state favored by depolarization (producing more potent inhibition at depolarized potentials). A number o f DHP agonistshave also been dev eloped, the mo st highly utilized of which is (-)-Bay K 8644 which increases both the o pen time and thesingle channel conductance (see chapter by Striessnig for mor e detail). L-type c hannels are also blocked by certain nativepeptide toxins such as -agatoxin IIIA (-Aga IIIA ), isolated from the ve nom of the funnel web spiderAgelenopsisaperta. -Aga II IA reduces the current amplitude without affecting the time co urse and unlike the DHPs, -Aga IIIA

    inhibition is vo ltage-independent and blocks L-type channels at all potentials (see chapter by Adams and Lewis for more

    detail).

    L-type channels have a unitary c onductance ranging from 20 and 27 pS using 110 mM barium (Ba ) as the charge carrier.

    L-type channels require large departures from resting potential to bec ome activated and ty pically begin to open atpotentials positive to -10 mV, although they can activ ate at significantly mo re negative potentials in chro maffin cells,sensory neurons, and cardiac cells. In the presence of Ba as the charge carrier, once open, L-type channels do not

    inactivate significantly during depolarizations of hundreds of milliseconds. Compared to Ba currents, using Ca as

    the charge carrier L-type currents are smaller and inactivate rapidly. This Ca -dependent inactivation has a number of

    characteristic properties and inactivation attributable to Ca influx is greatest at depolarizations at which Ca entrythrough the channel is maximal. While the degree of inactivation is slowed by the addition of BAPTA and other Ca

    chelators, it is not co mpletely abolished. Ca -dependent inactivation can however be eliminated by intracellular

    applications of trypsin, suggesting that the mechanism through which Ca acts to inactivate the channel is in close

    proximity to, if not part of, the channel complex i tsel f. Because the rate of Ca -dependent inactivation does not

    change with channel density, Neely et al (1994)22 proposed a local domain hypothesis, in which Ca affects only the

    channel through which it enters (see chapter by Lee and Catterall for mor e detail). Moreov er, curr ent models v iewcalmodulin to be constitutively bo und to the C-terminus forming part of the Ca sensing machinery, ultimately leading to

    the signal transduction of Ca -dependent inactivation.

    N-Type Channels

    In addition to L- and T-type Ca channels, recordings from chick dorsal root ganglion (DRG) cells revealed a third type o f

    2 +

    2 +

    2 +

    1 -4

    2+

    2 +

    4

    5

    6 2 +

    2

    2 + 2 +

    2 +

    2 , 4 , 7 , 8

    9 -1 3 6

    1 4 - 1 6

    2 +

    1 7 , 1 8

    2

    2 +

    2 +

    2 ,3 ,1 9 2 + 2 +

    2 +

    2 + 2 +2 0 2 +

    2 +

    2 +

    2 ,3 ,2 1 ,2 2 2 +

    2 +

    2 +

    2 + 2 3

    2 +

  • 8/12/2019 Voltage Gated Calcium Channel

    2/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 2

    single channel Ca conductance of 13 pS (in 11 0 mM Ba ), intermediate between that of the T- (8 pS) and L- (25 pS) type

    channels. Although this conductance shares some general electrophysiological characteristics with currents

    through both T- and L-type channels, it could not be attributed to either. Consequently, the c orresponding channel wasdesignated as N (neither)-type.

    Although first identified in c hic k DRG neuro ns, N-ty pe channels hav e also been detec ted in mammalian DRG cells,

    mammalian and amphibian sympathetic neurons, and other cells of the peripheral and central nervous systems.

    N-type channels appear to be expressed only in neuronal tissues, although an N-type current has been reported in rat

    thyro id C-cell line. Electrophy siologically, N-type channels are most easily distinguished from L-type channels by their

    inactivation properties. Unlike L-type channels, N-type channels display time-dependent inactivation (with Ba as the

    charge carrier). N-type c urrents decay with a time co nstant () ranging from 5 0 to 11 0 ms, significantly slower than the

    rapid ( = 20 -50 ms) inactivation of the LVA T-type channels, but much faster than the non-inactivating L-type channels. Inchick DRG neurons, the N-type current de cay s almost completely during a test depolarization of 140 ms, while L-typecurrent shows little inactivation over the same period of time. Howev er, N-type c urrents do not always inactivate rapidly.

    In sympathetic neurons, the decay rate of N-type c urrents is much slower ( = 500 to 800 ms) and can be incomplete, ev enover depolarizations lasting longer than one second. Thus, there appears to be at least two distinct components to N-

    type current inactivation. These differences in inactivation kinetics c ould reflect different subty pes of N-type channel.Alternativ ely , a single N-ty pe channel c ould suppor t both currents by switching between the s low- and fast-inac tiv atingstates.

    In addition to the time-dependence parameter, there is also a vo ltage-dependent aspect to N-type channelinactivation. Holding the cell membrane at potentials between -60 and -40 mV results in significant inactivation

    of the N-type current, and strongly negative potentials are required to reprime the channels. N-type channels are markedlymore sensitive to the effects o f holding potential on inactivation than are L-type c hannels. At resting membrane potentialsof -20 mV, N-type channels are completely inactivated while L-type c hannels remain available for o pening.

    Theoretically, the different inactivation proper ties of N- and L-type c hannels prov ides two parameters that can be used todissect the relative contributions of the two channel types to the whole cell HVA c urrent. One method takes advantage

    of the different inactivation rates. The co mponent of whole c ell current that dec ays during a pro longed depolarization canbe attribute d to the inac tiv ating N-ty pe channels , while the non-inac tiv ating po rtion is identified as L-ty pe current. Thesecond appro ach ex ploits the different ranges ov er which v oltage-dependent inactivation takes place. The co ntribution ofeach ty pe to the whole ce ll current may be determined by analyzing the differences in whole cell curre nts elicited bydepolarizations from resting potentials of -40 and -90 mV. Because L-type channels are relatively resistant to the effects o fholding potential on inactivation while N-type channels inactivate at depo larized membrane potentials, the difference underthese two co nditions should reflect the contribution of N-type channels to the whole-cell current. However , neither methodmay be adequate to pro perly distinguish these currents. Some N-type c hannels can inactivate quite slowly and inactivationmay not be c omplete. In addition, voltage-dependent inactivation of N-type c hannels can be highly v ariable and takes placeover a wide range of holding potentials between -80 to -20 mV. If N-type channels predominate in a cell, the residual

    current through incompletely inactivated N-type channels may b e significant. Thus definitions of N- and L-type currentbased solely on these c riteria may not be v alid.

    Pharmacolo gically, N-type channels are sensitive to inhibition by a class of native peptide tox ins called the ?-conoto xins,whic h are a family of small (1 3-29 amino ac id) peptides fo und in the v enom of predator y marine snails of the genusConus. All known -conotoxins inhibit N-type Ca channels, although their specificities and blocking affinities for

    this particular c hannel vary significantly. To date, -co notox in GVI A (-CgTx), a 27 -amino acid peptide from Conusgeographus is the most spec ific -conoto xin peptide for N-type c hannel inhibition. -CgTx produc es co mplete and

    irrever sible inhibition of N-type c urrents in DRG, hippocampal, sympathetic, and sensory neurons at c oncentrations ofapproximately 100 nM to 1 M. At higher concentrations (5-15 M), -CgTx also inhibits L- and T-type currents,

    although unlike N-type channels, the effects are incomplete and reversible (see chapter by Adams and Lewis for more

    detail).

    -CgTx binding sites (and by ex tension N-type Ca channels) are distributed throughout the PNS and CNS, including the

    corte x, hippocampus, olfactory b ulb, and cerebellar cor tex, and appear especially co ncentrated in regions of high synapticdensity. Although N-type channels were first identified by single channel recordings from the cell bodies of DRG

    neurons, they appear to be mor e abundantly loc alized on dendrites and axon terminals. In muscle, -CgTx binding occurs

    at the active zones of presynaptic cells in spatial register with the postsynaptic acety lcholine (ACh) receptor s on the muscle.Labeling is rarely found between active zones, nor is it localized to areas o f the presynaptic membrane that do not face themuscle. N-type channels have also been observ ed to cluster in areas of synaptic c ontact on hippocampal CA1 neurons.

    The presence of N-type c hannels on the presynaptic membrane suggests that Ca entry through these channels is

    responsible for triggering neurotransmitter release. An early study demonstrated that -CgTx blocks electrically-

    induced release from the frog NMJ and numerous subsequent studies have demonstrated that the application of -CgTxinhibits neurotransmitter release in the central and peripheral nervous system. Furthermore, biochemical studies

    indicate that N-type c hannels are physically associated with proteins such as sy naptotagmin and syntax in which are part ofthe exocytotic machinery. There appear to be species- and cell-specific differences in N-type-channel-regulated

    neurotransmission. For example, while -CgTx completely abolishes neurotransmission at the avian and amphibian NMJ, ithas no effect on the mammalian motor nervous system. The abil ity of this toxin to inhibit neurotransmission

    also varies depending on the type of synapse within a given species. For example, inhibitory synaptic transmission

    in hippocampal CA1 neuro ns is strongly reduced by the application of -CgTx, whereas the tox in blocks exc itatory

    2 + 2 +

    1 , 2 , 1 9 , 2 4 , 2 5

    2 6 - 2 9

    3 0-3 3 3 4 -3 9

    3 1 , 4 0

    4 1

    2 +

    1

    2 , 4 1- 4 2

    4 0 , 4 3

    1 , 2 4 , 2 5 , 2 9

    2 , 2 4

    2 4 , 2 5

    4 4 , 4 5 2 +

    4 6

    4 7 - 4 9

    4 7 - 5 2

    2 +

    2 , 5 3 - 5 8

    1

    5 4

    2 +

    5 9

    60-67

    6 8 - 70

    5 9 , 6 5 , 7 1 - 7 3

    5 2 , 6 5 - 6 7

  • 8/12/2019 Voltage Gated Calcium Channel

    3/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 3

    transmissions to a muc h lesser ex tent. In addition, while -CgTx inhibits release o f ACh from b oth autonomic and centralneurons in the rat, release from c entral neurons is approximately 20-fold less sensitive .

    In spite of the c omplete and irrev ersible inhibition of N-type channels produced by -CgTx, application of the to xin tomany types of neurons only partially inhibits neurotransmitter release, suggesting that other type s of Ca channels

    contribute to neurotransmitter release from both central and peripheral neurons. In fact, while regulation of

    transmitter release from peripheral neurons appears to predominantly invo lve N-type channels, release in the centralnervous system appears to be co ntrolled primarily by other types of Ca channels that are insensitive to both -CgTx and

    DHPs.

    The presence o f N-type c hannels in regions other than the sy napse indicates that these c hannels have other functions in

    addition to neuro transmitter release. N-type channels localized to dendritic branch points may be involv ed in integrationor amplification of neural inputs. N-type channels may also play a role in nervo us system development as ev idenced by

    the ex pression of N-type channels on postmitotic ce rebellar granule cells. These cells only begin migration after theappearance of N-type channels and -CgTx c auses a ce ssation o f migration.

    Other HVA Ca Channels: P-, Q-, and O-Types

    The original classification system of Ca channels, which was ex panded from the simple LVA/ HVA dichotomy to enc ompassT-, L- and N-channels, was subsequently found to be too restrictive to adequately describe all types of Ca conductances.

    The availability o f blocking agents that target L- and N-type c hannels revealed other HVA c urrents that could not be definedaccording to this scheme. These no vel channel ty pes, var io usly nam ed P-, Q-, O-, and R-, have pr imarily

    been defined on the basis o f their distinctiv e pharmac ological pro perties rather than elec trophy siological characteristic s.

    The P-type current was originally identified as an HVA current in Purkinje c ells that is insensitive to the agents typicallyused to inhibit L- and N-type channels. These channels are thought to support the Ca -dependent action potentials in

    the dendrites of c erebellar Purkinje c ells, which are unaffected by DHPs and -CgTx, but are po tently b locked bycompo nents of the venom o f the funnel web spiderAgelenopsis aperta.

    Whole cell recordings fro m Pur kinje cells reveal an HVA c urrent that peaks at voltages b etween -30 and -20 mV andinactivates s lowly over the duration of the depolarization. Single channel analysis of P-type channels reveals

    conduc tances in ranges similar to those of N- and L-type channels. Multiple unitary c onductance lev els of 9, 14, 1 9 pS in 110mM Ba have been reported for P-type c hannels in the Purkinje cell soma and dendrites, and a P-type current in

    hypo glossal motorneurons has a unitary c onductance o f 20 pS.

    The venom of the funnel web spider, like that of the cone snail, is a cocktail of toxins that target different elements of thesynaptic machinery . FTX, a no n-peptide c omponent o f the v enom (arginine poly amine and a synthetic analog o f FTX, sFTX),was initially repo rted to be specific b lockers of P-ty pe channels , but subsequently sho wn to pro duce inhibit ion of

    other Ca currents in conjunction with the P-type block.

    -Aga IV A, a 48-amino acid peptide also found in the venom o fA. apertapotently inhibits P-type Ca channels.

    In Purkinje cells, co mplete inhibition is observed at co ncentrations below 200 nM, with half-maximal block produced atconc entrations between 2 and 10 nM. Inhibition is rapid, occ urring within two minutes of application, and while theinhibition is poorly rev ersible by wash-out, it can be remov ed by a series of strong depolarizations (i.e., to +7 0 mV ). Blockof P-type currents by ?-Aga IV A in other neurons is qualitatively similar to that in Purkinje cells although the kinetics v aryslightly. For e xample, while inhibition of P-type current in spinal cord interneurons and neurons in the visual cortex occ ursas rapidly as that in the Purkinje cells, the rate o f block is several times slower in CA1 and CA3 hippocampal neurons.

    A peptide to xin iso lated from the c one snail Conus magushas also been shown to inhibit P-type channels. This toxin, -

    CgTx MVIIC, blocks P-type channels with an IC of 1-10 M. Howev er, -CgTx MVIIC also inhibits N-type channels as well

    as the Q- and O-type conductances (see chapter by Adams and Lewis for more detail).

    P-type channels do not account for all of the DHP- and -CgTx-resistant current in neurons since a substantial fraction ofcurrent remains even after ex posure to saturating conc entrations of DHPs, -CgTx, and -Aga IV A. Cultured rat c erebellargranule cells expre ss an HVA current that is unaffected by these inhibitors at co ncentrations which block L-, N- and P-typechannels, respectively. However, the channels supporting this novel current (termed Q-type) are partially blocked

    by ?-Aga IVA at c oncentrat ions 1 0 to 1 00 times that required for P-ty pe inhib itio n and are c ompletely blo cked by -CgTxMVIIC (IC = 30 to 300 nM). In addition to differing sensitivities to these tox ins, Q-type channels partially rec ov er from -

    Aga I VA-induc ed inhibitio n within minutes o f tox in washo ut. Q-ty pe channels also disp lay ele ctrophy siological pr opertiesdistinct from tho se of P-type channels. While P-type currents in Purkinje c ells and cere bellar granule cells show noinactivation over a 10 0 ms test depolarization, the Q-type current in granule cells decay s to approx imately 65% of the peakcurrent ov er the same time period.

    The existence of O-type channels has been inferred solely from pharmacological studies. O-type channels were identified ashigh affinity -CgTx MVII C binding sites. These channels are significantly more sensitive to the toxin than are other -

    CgTx MVIIC-sensitive channels It is possible that P-, Q-, and O-type channels are members of the same channel family

    whic h po ssess slightly different pharm acological and elec trophy siological pr operties as a r esult o f alternativ e sp licing o f the subunit gene and/or different complements of auxiliary subunits. While O-type c hannels have no t been loc alized

    immunohistochemically, evidence from binding studies suggests that they are widely distributed in the mammalian CNS.

    Estimates of O-type bindings sites in rat brain preparations suggest that O-type channels are more prevalent than N-type

    6 5

    2 +

    5 8 , 7 4 , 7 5

    2 +

    7 6 , 7 7

    5 6

    7 8

    2 +

    3 6 , 3 7 , 4 5 , 7 9 - 8 1

    8 2 2 +

    82-85

    3 7 , 5 2 , 8 1 , 8 6

    8 7

    3 9

    3 , 8 2 , 8 3

    2 + 3 ,4 5

    2 + 5 2 ,8 4 ,8 6

    86

    8 8

    5 08 8

    4 5 , 8 9 , 9 0

    5 0

    9 1

    4 5

    14 5

  • 8/12/2019 Voltage Gated Calcium Channel

    4/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 4

    channels in the CNS, and it has been pro posed that O-type channels are localized exc lusively to sy naptic termini, whichwould largely prev ent their detectio n thro ugh ele ctrophy siological me ans.

    Many studies implicate P-, Q, and O-type channels in neurotransmitter release. While N-type channels mediate release at

    some sy napses in the mammalian CNS, the -Aga IVA -sensitive P- and Q-type channels appear to play a more prominentrole. -Aga IVA potently blocks Ca uptake into synaptosomes84 and partially inhibits the release of dopamine and

    glutamate from sy napto so mes and at CA 1 -CA 3 sy napses in the hippo campus. I n the per ipheral nervo us

    system, -Aga IVA has little or no effect on the autonomic nervous system, but P-type channels are probably responsible

    for neurotransmitter release at the mammalian NMJ. As -Aga IVA blocks both P- and Q-type channels, it is possible

    that both c hannel types are inv olve d in neurotransmission. Wheeler and co lleagues found that the pharmacolo gicalproperties of the -Aga I VA -sensitive channels supporting neurotransmission in the hippocampus and for -Aga IV A-

    induced bloc k of inhibitory po stsynaptic potentials in the cerebellum were more similar to Q- than P-type channels.Finally, O-type channels may also mediate neurotransmission at c ertain synapses, as norepinephrine release in thehippocampus is inhibited by subnanomolar conc entrations of -CgTx MVII C.

    R-Type Channels

    A co mpo nent o f the HV A current in c erebellar granule c ells remains ev en after the applicatio n of nimo dipine, -CgTx , -Aga V IA , and -CgTx MVII C. This c urrent, categorized as R (residual or resistant)- ty pe, comprises appro ximately 15% of

    the HVA c urrent in these cells. R-type current may not necessarily reflect a single channel type, but a family of molecularlydistinct channels with similar pharmacological and electrophy siological characteristics.

    R-type currents begin to ac tivate around -40 mV and reach a peak amplitude at 0 mV . The current inactivates rapidly, andthe increased rate o f inactivation with Ca as the c harge carrier suggests that the channels supporting the R-type currentinactivate in a Ca -dependent manner. R-type channels are equally sensitive to block by Cd and Ni ions. The exact

    nature of the c hannels supporting this current is currently unknown. See the below sec tion on Class E channels for furtherdiscussion.

    Cloned Calcium Channels

    HVA Ca Channels Are Multi-Subunit Complexes

    Biochemical studies have established that high threshold vo ltage-gated Ca channels are multi-subunit complexes. Taking

    advantage of the high-affinity binding of organic antagonists, several groups purified the L-type channel from skeletalmuscle. Four distinct polypeptides, designated (17 5-kDa), (17 0-kDa), (52-kDa), and (32-kDa), co-migrate with

    the ligand-binding activity. A minor 212-kDa band also co-purified and was shown to represent a larger, much less

    abundant form of the skeletal muscle subunit. Similar approaches have been used to isolate the cardiac L-type

    and brain N-type channels These complexes also consist of an a1 subunit associated with and subunits. The and

    - are highly similar, if not identical to, the subunits associated with the skeletal muscle . However, unlike the

    skeletal muscle L-type c hannel, no subunits appeared as part of either complex . A no ve l 95-kDa polypeptide was found tocomigrate with the N-type c hannel although it is unclear whether this represents a bona fide channel subunit or aproteoly tic fragment.

    While subunit c ompo sitio n differ s slightly depending on channel ty pe, a general mo del has been pro po sed for HVA channelsin which four to five proteins form a multisubunit complex (fig.1A). In this model, the subunit forms the channel proper,

    comprising both the voltage-sensing mechanism and the Ca selective pore , and the remaining proteins interact with the

    subunit to modulate activity.

    Primary Structure and Properties of Ca Channel a1 Subunits

    The first cDNAs encoding Ca channel subunits were isolated from rabbit skeletal muscle. The L-type subunit is

    an 187 3-residue protein that bears a high degree of amino acid similarity to the voltage-gated Na and potassium (K )

    channels (see fig.1). The subunit is predicted to c onsist of four homologous, mainly hy drophobic do mains (designated

    domains I, II , II I and IV ). Each o f the four domains is comprised o f six putative membrane-spanning segments (S1-S6). TheS4 segment in each domain contains positive ly-charged residues ev ery third or fourth position and is believed to form partof the v oltage-sensing mechanism of the channel. Between the S5 and S6 segments of each do main are two hy drophobicsegments, SS1 and SS2, which are predicted to form the channel pore (fig.1B).

    Based upon similarity to v oltage-gated Na channels, Tanabe et al (1987 ) speculated that the subunit may form both the

    Ca -selective pore and the voltage sensor of the channel complex. This hypothesis was supported by studies

    demonstrating that expre ssion of the in myo tubes from dysgenic mice restored normal skeletal muscle-type E-C

    coupling and the slow Ca current absent in these cells. In addition, expression in dysgenic myotubes restored the

    charge movement observed in normal myotubes upon membrane depolarization. These results indicated that the

    skeletal muscle a1S subunit acts both as a vo ltage-sensor, prov iding a physical co nnection between membranedepolarization and Ca -release from intracellular stores for the initiation of muscle contraction, and is also part of a

    functional VGCC.

    Using the skeletal muscle clone as a probe, cDNAs encoding homologous L-type subunits have been subsequently c loned

    4 5

    5 2 ,9 2 2 +

    9 3 ,9 4 5 8 ,7 4 ,7 6

    9 5

    9 6 , 9 7

    5 8 , 9 8

    4 5

    8 9

    2 + 2 + 2 +

    2+

    2 +

    1 29 9 - 10 2

    11 03 1 04

    1 052

    21 0 6 , 1 0 7

    1 05

    12 +

    1

    2+

    2 +1

    5 , 1 0 81 s

    + +

    1

    + 51

    2 +

    1 S2 + 7

    1 S1 09

    2 +

    1

    http://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30873/?report=objectonlyhttp://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30873/?report=objectonlyhttp://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30873/?report=objectonly
  • 8/12/2019 Voltage Gated Calcium Channel

    5/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 5

    from cardiac and smooth muscle. Injection of the cardiac subunit into dysgenic myotubes resulted in the

    expre ssion of a VGCC which differed markedly in terms of activation rate, Ba permeability, and E-C coupling from the

    current conducted through channels formed by the skeletal muscle clone. Expression of cardiac and smooth muscle

    subunit clones inXenopuso oc ytes resulted in large inward c urrents that were sensitiv e to the organic

    channel agonists and antagonists, thereby identifying them as L-type channels. Co-expr ession of skeletal muscle-derived - and subunits, while affecting the amplitude and vo ltage dependence of the currents, was not required for c hannel

    activity or drug binding, suggesting that the subunit is capable of forming a functional channel in the absence of the other

    subunits. Howev er, because some V GCC subunits may b e endogenously e xpressed by Xenopusoocy tes, it is possible

    that the a1 prote in forms a co mplex with these endogenous auxiliary subunits. This prompted seve ral groups to ex amine theproperties of the a1 subunit in cells lacking these proteins. Murine L-cells and Chinese Hamster Ovary (CHO) cells

    stably transformed with subunits express voltage activated Ca currents sensitive to L-type channel blockers. WhileCa currents in cells expre ssing the smooth muscle a1 subunit displayed similar drug sensitivities and kinetics to the native

    currents, the currents supported by activated considerably more slowly than currents recorded from skeletal muscle

    cells.

    At least nine different sub unit genes are no w kno wn to be expressed in the mammalian nerv ous sy stem (see fig.2, Table

    1). Initially, four distinct classes of subunits were isolated from a rat brain library on the basis of their homology to the

    rabbit skeletal muscle subunit. Each cDNA hybridized to one of four distinct banding patterns on Northern blots of

    rat brain mRNA, allowing them to be grouped into four classes, designated , , and . Subsequently, a fi fth

    subunit ( ) was isolated from rat brain. Southern blot analysis and DNA sequencing indicated that the five classes are

    separate members of a multigene family with the , , and channels being more similar to one another than they are

    to the c lass C and D channels (fig.2). The class C clo ne is almost identical to the car diac a1 subunit, suggesting that the c lass Cand D clo nes represent members of the DHP-sensitive L-ty pe channels, while the A, B, and E clo nes are DHP resistant. Fourother VGCC subunit genes have been identified in the mammalian genome. The shares the most sequence

    identity with the L-type channels. The , , and clones represent the LVA branch of the VGCC family.

    The individual subunit clones share the most homology in the transmembrane domains with the majority of sequence

    divergence occ urring in the putative cy toplasmic regions of the channels. The loop between domains II and III, and thecytoplasmic tail vary in size as well as sequence. The DHP-sensitive channels (classes C, D, F, and S) have relatively short(130 amino ac id) sequences linking domains II and III while the analogous region in the c lass A and B channels aresignificantly larger ( 430 amino acid). Howev er, despite the size similarity between the linkers, the c lass A and B clo nesshow little sequence homology in this region. While the LVA channel clones (classes G, H, and I) share much less

    sequence identity with the other c lasses, the v oltage-sensing S4 region and the loo p that forms the channel pore are wellco nserv ed. Other motifs, such as the -subunit binding site and the E-F hand, which are found in the HVA classes o f VGCCsare absent in the LVA channels. Whole ce ll and single channel electrophy siological tec hniques have pro vided informationabout the functional and pharmacological properties of the c loned channels and allowed researchers to assign the individualclones to c hannel types (Table 1).

    / Ca 2.1Class A subunits have been cloned from both rabbit (BI-1, BI-2) and rat (rbA-I) brain andDrosophila

    melanogaster(Dmca1A). Northern blot analysis identified a single RNA transcript of 9.4 kb in rabbit brain, while two

    transcripts of 8.3 and 8.8 kb were detected in rat brain. transcripts are widely distributed throughout the nervous

    system, as well as being present in the heart and pituitary, but not in skeletal muscle, stomach, or kidney. I n brain, thehighest lev els of class A transcripts were found in the cerebe llum, suggesting that this clone might encode a P-type c hanneland initial expression studies supported this hypothesis. Studies showed that clones expressed inXenopus oocytes

    supported HVA currents which were insensit ive to DHPs and -CgTx but inhibited by -Aga VIA. However, a

    number of discrepancies between currents elicited in ooc ytes ex pressing class A clones and native P-type curre nts havecalled this into question. The currents display prominent time- and voltage-dependent inactivation, yet P-type

    currents show little time-dependent inactivation and are relatively insensitive to holding potential. Furthermore, thepharmacological sensitivities of channels are quite different from those of P-type c hannels. While these currents are

    blo cked by -A ga IVA, they are appro ximately 20 0-fold less sensit ive to the tox in (IC 200 nM) than are P-ty pe

    currents (IC 2-10 nM). In addition, channels are markedly more sensitive to block by the snail toxin -CgTx MVIICthan P-type channels (IC 150 nM vs. 1-10 M for P-type channels). Sather et al (1993) noted that the kinetic and

    electrophysiological features of the current were more similar to the Q-type current described in cerebellar granule

    cells by Randall et al (1993). Based on these results, some researchers have suggested that the class A clones represent

    Q-type channels, and that P-type channels are the product of a different gene. However, Stea et al (1994) noted that the

    high cor relation between the localization of transcripts and P-type channel immunoreactivity implied a possible

    structural similarity between P- and Q-type channels and the gene product. They further proposed that the functional

    differences between the two channel types may arise as a result of differential post-translational processing of the proteins(which co uld affect toxin binding), subunit composition of the channel complex , and/or alternative splicing of the gene.

    The auxiliary subunits of the VGCC complex are known to modulate the properties of the subunit (see below). The

    inactivation kinetics of the subunit are dramatically affected by the type of subunit with which it is associated.

    Expression of the subunit from rat brain in the absence of the subunit results in a current that inactivates

    considerably (40% remained after a 400 ms test pulse). Co-expression of either the or subunit increased current

    1 1 0 1 1 1 ,1 1 21

    2 +

    1 1 31

    1 1 0 , 1 1 1 , 1 1 4 , 1 1 5

    2

    11 1 6 , 1 1 7

    1 1 8 ,1 1 9 1 1 4

    1

    2 +

    2 +

    1 S

    1

    1

    1 S1 2 0

    1 A 1 B 1 C 1 D 1

    1 E1 2 1

    1 A 1 B 1 E

    1 1 F1 2 2 , 1 2 3

    1 G 1 H 1 I 1 2 4 - 1 2 6

    1

    1 2 7

    1A v

    11 2 8 1 2 9

    1 3 0

    1 A

    1 A1 2 8 , 1 3 1 , 1 3 2

    1 3 1 , 1 3 21 A

    1 A

    5 0

    5 0 1 A 5 0

    1 3 1

    1 A1 3 3

    1 3 2

    1 A

    1 A

    1

    1

    1 A1 3 2

    1 A

    1 b 3 1 A

    http://www.ncbi.nlm.nih.gov/books/NBK6181/table/A30875/?report=objectonlyhttp://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30874/?report=objectonlyhttp://www.ncbi.nlm.nih.gov/books/NBK6181/table/A30875/?report=objectonlyhttp://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30874/?report=objectonly
  • 8/12/2019 Voltage Gated Calcium Channel

    6/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 6

    inactivation to a rate similar to that of the Q-type current. In contrast, currents recorded from oocy tes expressing the +

    combination show significantly slower inactivation kinetics, such that the wave form is more similar to that of native P-

    type currents. The subunit also appears to affect voltage-dependent inactivation of the subunit. The subunit

    shifted the steady-state inactivation of the approximately 15 to 20 mV more depolarized, thus reducing the sensitivity

    of the c hannel to holding potential.

    Multiple isoforms derived from the alternative splicing of transcripts have been detected by sev eral groups.

    Bourinet et al (1999) isolated an variant which possessed an number of sequence differences when compared to

    the rbA -1 c lone and ex amined the functional implications of these splicing ev ents. A v aline insertion in the I-II linker bothslowed time-dependent inactivation and altered steady-state inactivation. variants containing this valine have

    inactivation properties similar to P-type channels, while v aline-less isoforms, such as r bA-1, appeared more Q-like. A

    seco nd splice site c onsisted o f the insertion of an asparagine-proline (N-P) pair in the IV S3-IVS4 linker. This affects theelectrophy siological properties of the channel by pro ducing a depolarizing shift in the current-vo ltage relationship. The

    N-P insertion also had the e ffect o f decreasing the affinity of the c hannel for -A ga IV A b y decreasing the on-rate of thetoxin and increasing the off-rate. Thus, it is likely that the gene encodes both P- and Q-type channels, and the distinct

    channel properties reflect differences b oth in subunit composition and alternative splicing. P-type channels may becomprised o f splice variants that contain the v aline insertion in the I-II linker, but not the N-P pair in the IV S3-IV S4 loop.Conversely, Q-type c urrents may be produce d by channels lacking the valine, but containing the N-P insertion. In addition,the association of different subunits may also be an important determinant of the P- ver sus Q-type pheno type.

    / Ca 2.2

    Class B subunits hav e been cloned from rat (rbB-I) or ( ), human ( , ), and rabbit brain

    (BIII), as well as from the forebrain of the marine ray Disco pyge ommata(doe-4). The various clones encode

    proteins of 2336 to 2339 amino acids with predicted molecular weights of 260 to 262 kDa. The amino acid

    sequence is more similar to that of the , with the majority of the sequence divergence occ urring in the cytoplasmic loopbetween domains II and II I and in the cy to plasmic carbox y l tai l.

    Initial indications that this class of subunit corresponds to N-type channels came from the work of Dubel et al (1992)

    who sho wed that a poly clonal antib ody (CNB-1) raised against the II -II I lo op region of the rbB-I clone immunoprec ipitatedalmost 5 0% of the high-affinity -CgTx b inding sites, but none of the DHP-binding sites from rat brain. Furthermor e,Northern blot analysis of expe rimental cell lines showed that rbB-I expre ssion was co rrelated with the presence of N-typec hanne ls in ne rv e tissue s and ce ll line s that ex pre ss N-ty pe channe ls. No rthe rn b lo tting and in situ

    immunohistochemistry e xperiments have localized the rbB-I transcripts to the cere bral cor tex, hippocampus, forebrain,midbrain, ce rebellum, and brainstem. At the subc ellular lev el, rbB-I pro tein is found on dendrites, at presynaptic terminalsand, to a lesser extent, neuronal cell bodies. The localization pattern of the compares well with that observe d with a

    monoclonal antibody against -CgTx, although the -CgTx antibody staining was more widely distributed.

    Molecular cloning and biochemical studies have also prov ided evidence for the existence o f multiple isoforms of the

    subunit. A t least two of these iso fo rms represent channels with d ifferentially sp liced carbo x y l tails, and theinability of CNB-1 to immuno prec ipitate all of the -CgTx binding sites might suggest the ex istence o f additional isoformswith dist inct II-I II loop seq uences. In additio n, clones with small insertio ns and de let ions sc attered througho ut the

    channel have been identified, and expre ssion studies indicate that these sequence v ariations have a profound influence onthe properties of the channel (see below). These include a variant of the human N-type calcium channel that lacks the

    synaptic protein interaction site in the domain II-III linker and can therefore not associate with synaptic proteins such

    as syntaxin 1A and SNAP-25.

    Transient expression of both human in HEK cells and rabbit brain BIII in dysgenic myotubes produced HVA

    currents that first activated be tween -10 and -30 mV and reac hed a maximum between +1 0 and +30 mV. Currents partiallyinactivated ov er the time c ourse o f the depolarization and were sensitive to holding potential (showing 50% curre ntinactivation at approximately -60 mV). A t a holding potential of -40 mV , the bulk of the c urrent (90%) was inhibited. Inagreement with binding studies, the -induced curre nts were irreversibly bloc ked by 1 M -CgTx and were insensitive to

    DHPs.

    The properties of curre nts generated inXeno pus ooc ytes by expre ssion of the rbB-I clone agreed well with those seen withthe and BIII clones in terms of pharmacological sensitivities and voltage-dependence of activation. However, there

    were some notable discrepancies in other pro pert ies. For example, the rbB-I channel was less sensitiv e to holding po tentia l,and the rates of activation and inactivation of the rbB-I clone were markedly slower, resulting in significantly differentcurrent waveforms. Co-expression of the subunit shifted the voltage-dependence of inactivation to more negative

    potentials similar to those observ ed with the human and rabbit clones. The subunit also increased the rate of activationsuch that the current attained peak magnitude in approximately 1 20 ms (co mpared to 1 50-25 0 ms for the rbB-I subunitalone), and increased the rate of inactivation of the rbB-I current. A fter 800 ms, cur rent through rbB-I alone had decre asedby 15-20 %, whereas co-ex pression of the subunit re sulted in a 6 5-7 0% reduct ion in pe ak current. Despite the r ateincreases produced by subunit co-expre ssion, these parameters remained dramatically different from those displayed bythe other clones. The rate o f activation of rbB-I (11 0 ms to peak) was still significantly slower than that of a1B-1 c urrents (10ms). In addition, the clone showed biphasic inactivation; the first, rapid phase had a of 46-105 ms. The of the slow

    phase ranged between 291 and 453 ms. In co ntrast, decay of rbB-I cur rents was monophasic and much slower (= 7 00 ms).

    Evidence obtained by Stea et al (1999) from a second rat brain clone (rbB-II or ) found that the differences in

    1 A

    2 a

    1 A 2 a

    1 A

    1 A1 2 8 , 1 3 4 -

    1 3 6 1 3 51 A

    1 A

    1 A

    1 A

    1B v

    11 3 7

    1 B-11 3 8

    1 B-1 1 B-21 3 9

    1 4 0 1 4 1

    1 4 21 B

    1 A

    11 3 7

    1 3 7 , 1 3 9 , 1 4 0 , 1 4 3

    1 B5 6

    1 B1 3 8 - 1 4 0 , 1 4 3 - 1 4 6

    1 B

    1 3 2

    1 4 5

    1 B-11 4 7 1 4 0

    1 B

    1 B-1

    1 4 81 b

    1 B-1

    1 3 81 B-II

  • 8/12/2019 Voltage Gated Calcium Channel

    7/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 7

    channel kinetics were the result o f small amino acid alterations that are mo st likely the pro duct o f alternative splicingand/or cDNA cloning artifacts. While differs from in four regions, they found that the substitution of a glycine for

    a glutamate in transmembrane segment I S3 was sufficient to speed the activation and inactivation kinetics. It has bee nnoted that, while N-type channels are ty pically descr ibed as having fast kinetics, this is not always the c ase (see N-typechannels, above). It may be that isoforms containing a glutamate in IS3, such as the rbB-I ( ) clone may acc ount for the

    slow and incomplete inactivation of N-type c urrent that has been described in sympathetic neurons. A s is the case with the gene, alternative splicing and differential subunit composition may comb ine to produce slight modifications in channel

    characteristics with tissue or dev elopmental specificities. Just recently , a newly discove red tissue specific splice isoformvariant o f the N-ty pe channel has been discov ered in do rsal ro ot ganglia (DRG). This sp lice v ariant arising from the

    presence o f exon 37 a has the unique property of being expressed exc lusive ly in nociceptiv e neurons of the DRG andultimately may serve as a nove l target for pain management.

    / Ca 1.2

    The first complete class C ( ) subunit to be cloned was isolated from cardiac muscle (CARD1). The cardiac and

    skeletal muscle L-type VGCCs arise from separate genes and are approximately 66% identical at the amino acid level.Subsequently, a1C clones were isolated from rabbit lung (pSCaL) and rat aorta (990: VSM ) which shared 95%

    identity with the cardiac clone. Class C clones later isolated from rat (rbC-I, rbC-II) and mouse (mbC) brain are also

    more closely related to the cardiac and smooth muscle subunits than to the skeletal muscle clone. The clones code

    for proteins of 2140 to 217 1 amino acids with predicted molecular masses of 235 to 239 kDa. Antibodies directed againstthe II-II I loo p of the neuro nal class C channel also identify a truncated form with an approx imate mass of 195 kDa.

    The high degree o f similarity amongst these pro teins suggest that they are produc ts of a single gene, and this is supported bygenomic Southern blotting experiments. Howev er, there are regions of considerable diversity in these clones which are theresult of a lternative splicing of the primary transcript. pSCaL, the smooth muscle channel isolated by Biel et a l

    (1990) differs from the cardiac form in the amino terminus, the IS6 and IVS3 transmembrane segments, and by a 25-amino acid insertion in the I-II linker. In co ntrast, the VSM clone, also isolated from smooth muscle, co ntains a cardiac

    channel-like I S6 segment and a 68 residue substitution in the carb oxy l tail. This sequence is also found in the neuronalclones, rbC-I and II. Finally, rbC-I and rbC-II contain regions of identity with both the smooth muscle and cardiac

    clones, but also contain many substitutions, primarily in cy toplasmic r egions of the protein and the II IS2 transmembranesegment. Notably, many of these substitutions are localized to the c yto plasmic linker between domains II and III, and mayreflect cell-specific functions of the channels. The truncated form of the neuronal protein may also be the result ofalternative splicing, or may be due to po st-translational proce ssing as is the case with the skeletal muscle channel.

    As wo uld be ex pected conside ring the diverse nature of the tissues from whic h class C cDNAs hav e been cloned, the

    gene has a widespread pattern of expression. Transcripts of 8.9 kb have been detected in heart. In smooth muscle and

    brain, hy br idizing transc ripts were sl ightly smaller (8.6 and 8 kb, respec tiv ely ). Additio nal transcripts o f 15 .5 kb

    (cardiac) and 12 kb (smooth muscle and brain) were also detected which were pr oposed to represent stable proc essingintermediates. Northern blot analysis indicate that the class C gene is expressed in heart, smooth muscle (e.g., uterine,

    lung, stomach, and intest ine), and throughout the CNS. Within the brain, high expression levels are detected inthe olfactory bulb, cer ebellum, striatum, thalamus, hypothalamus and cortex , and at much lower lev els in the pons/medullaand spinal cord. Thus far, there is no evidence for exclusive expression of splice variants in specific tissues. cDNAs

    containing both variants of the IV S3 transmembrane segment have been isolated from heart, smooth muscle, andbrain and the alternate c arb ox y l tail is ex pressed in bo th smooth muscle and neuro nal t issues.

    Howev er, a mo re detailed study of the expr ession pattern of class C variants in rat has rev ealed tissue-specific differences inexpression of the rbC-I and rbC-II proteins. Overall, rbC-II is the more abundant form, and generally more prevalent in

    any give n tissue, although the relative amounts of the two transcripts vary between brain regions and tissue types.

    The subcellular localization of class C subunits was studied using the polyclonal antibody , CNC1. Immunoprecipitation

    and Western blotting ex periments indicate d that class C a1 subunits compr ise approx imately 7 5% of the DHP binding sites inrat cereb ral cortex and hippocampus. CNC1 immunoreactiv ity was distributed at low levels on c ell bodies and prox imaldendrites, with staining diminishing along the length of the dendrite. In addition, clusters of high levels of immunoreactivitywere ob ser ved on the sur face o f cells (as opposed to represent ing a c y to plasmic po ol of channels).

    Expression of clones inXenopusooc ytes re sulted in currents with electrophy siological and pharmacological propertiesc harac te ristic of L-ty pe c hanne ls. I n Ba , de po larizatio ns to -1 0 to -3 0 mV elic ite d large inwar d c ur re nts

    that inactivated slowly , if at all, over the course of a several hundred millisecond the test pulse. The curr ents peakedbetween + 10 and +30 mV and were inhib ited by Cd (100 -20 0 M) and were se nsitiv e to DHPs. Like nativ e L-ty pe

    currents, the c loned L-type c hannels showed little sensitivity to holding potential. At holding potentials as high as -20 mV,half of the channels remained available for opening.

    The class C channels were shown to be mo dulated by the auxiliary subunits in much the same manner as the class A and Bchannels. Co-expression of with and - significantly increased the magnitude of the whole cell currents. This

    increase appeared to be mediated primarily through interaction with the subunit, while the addition of - had a slight

    synergistic effect. In addition, co-expression of rbC-II with the auxiliary subunits and/or - caused a small

    hyperpolarizing shift in the voltage dependence of activation of the channel and altered channel kinetics. The rate of

    activation for rbC-II varied substantially among o ocy tes. Time co nstants of activation ranged between 4 and 50 ms with anaverage of approximately 1 0 ms. Coexpression with and - both increased the rate of activation and reduced the

    1 B-II 1 B-I

    1 B-I

    1 A

    1 4 9

    1C v

    1 1 C1 1 0

    1 1 11

    1 1 2

    1 5 0 1 5 1

    1 1 C

    6

    1 5 0 , 1 5 2 , 1 5 3

    1 1 1

    1

    1 5 0

    6

    1 C1 1 0

    1 1 2 , 1 5 0

    1 1 2

    1 1 2 , 1 5 0 , 1 5 1

    6 , 1 5 01 C

    1 1 2 ,1 5 0,1 5 2 1 1 2 ,1 5 0

    1 5 0

    16

    1 C1 1 0,1 1 1 ,1 1 5 ,1 5 4 2 +

    2 +

    1 C 1 b 2

    2

    1 b 21 5 4

    1 b 2

  • 8/12/2019 Voltage Gated Calcium Channel

    8/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 8

    degree of variability. Furthermore, the and - subunits increased the rate of inactivation. Unlike the rbB-I

    channel (see abov e), neither auxiliary subunit had a significant effect on the vo ltage dependence of inactivation. With theexc eption of the voltage dependence of channel activation, the modulatory effects on the kinetics and voltage dependentparameters of the channel appear to be mediated primarily through interaction with the subunit, while the addition of -

    had slight synergistic effects.

    Whole cell recording using Ca as the c harge c arr ier resulted in c urrent trac es with marke dly different wavefo rms. The

    magnitude of the whole cell current was significantly smaller in Ca than Ba , indicating that the channels are more

    permeable to Ba than to Ca . In addition, instead of eliciting currents that are essentially non-inactivating, depolarizing

    pulses produce currents that decay rapidly by more than 50%. The increase in inactivation seen with Ca as the

    permeant ion retains all the hallmarks of Ca -dependent inactivation (see above).

    Calcium dependent inactivation (CDI) is predominantly linked to two regions of the Alpha1 subunit C-terminus. The EFhand, located from aa 1526 to1554, is responsible for CDI and may also regulate voltage-dependent inactivation

    (VDI). The second region (found downstream of the EF hand) is comprised of three distinct binding motifs. Peptide A; (aa

    1588 to 1610) and Peptide C (1615 to 1636 aa), and IQ(1649 to 1669), all work together to form a calcium bound

    Calmodulin binding site. In the absence o f calcium, Ca free calmodulin (ie Apo Cam) is pre-associated with the channel at

    a site localized between the EF motif and IQ region. Calcium entering through the channel binds to calmodulin, thus

    inducing a conformational change that relieves an inhibitory action of the c almodulin/C-terminus complex on the v oltage-dependent inactivation machinery.

    /Ca 1.3

    A class o f VGCC cDNAs sharing about 7 0% amino acid identity with the cardiac clones has b een cloned fro m a v ariety ofspec ies inc luding rat (RB ; rCA CN4 A, rCA CN4 B ), human ( ; neuroendocrine or -cell o r hCA CN4 :

    HCa3a ), and chicken. A cDNA encoding an invertebrate ortholog of the class D subunit has been isolated fromDrosophila me lanogaster (Dmca 1 D). These clones retained little similarity (~40% amino acid identity) with the non-

    DHP-sensitive class A clones, but were almost identical to the partial rat brain clone designated class D. In spite of

    the sequence divergence between c DNAs generated from class C and class D genes, the two c hannel types are r emarkablysimilar in certain regions. As with all VGCC subunits discussed thus far, the transmembrane regions tend to be highly

    conserved, while the intracellular loop sequences are much more divergent. In addition to these regions, the clones are

    almost identical to the DHP-sensitive c lass C and S clones in the segments predicted to form the DHP and pheny lalkylaminebinding site s, suggesting that the c lass D subunit cDNAs also enc ode members o f the DHP-sensitiv e L-ty pe family of

    VGCCs. The ex ception to this lies in the DHP-binding region in theDrosophilaDmca 1 D clone which co ntains a number ofnon-conserv ed c hanges. This finding, howev er, is consistent with the pharmacology of phenylalkylamine and DHP bindinginDrosophilahead membranes, and provides further support for the role of these regions in drug binding.

    The cloned subunits range in size from the 1634-amino acid (187 -kDa) rat brain isoform to the 2516-amino acid (27 6-

    kDa)Drosophilachannel clone. The range in protein sizes is due primarily to the truncated carbox yl terminal ends of the

    RB , HCa3a, and rCACN4B clones. The rCACN4B clone is a full 535 residues smaller than its rCACN4Acounterpart and is proposed to result from the use of an alternative splice acceptor site. In addition to the truncation of

    the carbo xy l tail, a number of other re gions have been identified in which v ariants have b een produc ed through alternativesplic ing. These regio ns inc lude insertio ns in the cy to plasmic linker between domains I and I I, the

    extrac ellular linker connecting IV S3 and IVS4, and the transmembrane segments IS6 and IVS3. In addition, Kollmar et al(1997 ) have reported that the chicken brain and cochlear proteins differ in the IIIS2 segment and IVS2-IVS3 loops,

    as well as in the carboxyl tail. Presumably, these splice variants impart functional differences to the channel. While it is noty et clear what these func tio nal d ifferences may be , Ihara et al (1 995) note that RBa1, HCa3a and rCACN4B are al l

    truncated at different sites. Furthermore, a number of potential PKA sites are eliminated by the truncations, which mayresult in the differential regulation of these isoforms b y phosphory lation.

    The class D channels, often termed neuro endocrine be cause of their presence in brain and pancreatic cells, have also be endetected in the retina, ov aries, and cochlear hair cells, but not in heart, skeletal muscle, spleen, colon, o r liver. Reportsdiffer on whether transcripts are present in kidney. Within the CNS, class D expression is found in the

    hippocampus, habenula, basal ganglia, and thalamus. The subcellular localization of the class D subunit wascharacterized using the poly clonal antisera anti-CND1. Anti-CND1 was generated against a peptide homologous to the

    unique II-III loop of the rat brain clone (Hui et al, 1991). Class D channels appear to be far less abundant in the rat

    CNS than class C channels. The sera labeled the cell bodies and pro ximal dendrites of both projec tion neurons andinterneurons throughout the brain. In contrast to the punctate staining pattern seen with the class C antibody, anti-CND1staining was evenly distributed o ve r the cell bo dy. The staining pattern of anti-CND1 was typical for neurons in all regions ofthe CNS with the notable ex ception o f the cerebe llar Purkinje c ells. While the cell bodies o f these neurons were labeled, therewas a marked absence o f staining o n the Purkinje c ell dendrites.

    Transient ex pressio n o f hum an inXenopusoo cy tes and stable expr ession of the rat CACN4A and CACN4B clones

    (Ihara et al, 1995) in CHO cells give s rise to DHP-sensitive curr ents, confirming the notion that class D channels are

    members of the L-type family. I n both systems, functional expression of the subunit required co-expression of the

    subunit. In Xenopus oocy tes, transient expression of with the and subunits yielded larger currents than those

    produced by expression of plus alone.51 Ba currents in oocytes expressing + + first activated upon

    1 b 2 1 B

    2

    2 +

    2 + 2 +

    2 + 2 +

    1 5 4 ,1 5 5 2 +

    2 +

    1 5 6

    1 5 7

    1 5 8 1 5 9

    2 +

    2 3 , 1 6 0

    1 6 1

    1D v

    11 6 2 1 6 3

    1 D5 1

    11 6 4

    1 6 5 1 6 6 11 6 7

    5 1 ,1 6 4 1 2 0

    1

    1 D

    1

    1 6 7

    1 D

    1 1 6 2 , 1 6 3 , 1 6 51 6 3

    5 1 , 1 5 2 , 1 6 2 , 1 6 3

    1 6 61 D

    1 6 3

    1 D1 6 4 , 1 6 6

    5 116

    1 D1 6 2

    1 D5 1

    1 6 3

    1 D

    1 D 2

    1 D2 +

    1 D 2

  • 8/12/2019 Voltage Gated Calcium Channel

    9/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 9

    depolarizations positive to -30 mV and peak current attained with depolarizations to 0 mV , thus the current-vo ltagerelationship of the is somewhat more negative than that of the (see above). channels activated rapidly and

    inactivated little over depolarizations lasting as long as 700 ms. channels inactivate to a considerably lesser degree

    over long test pulses than do channels.

    As indicated, c lass D channels fall unde r the heading of DHP-sensitiv e L-ty pe channels. Cd produc es substantia l bloc k,

    while Ni has a minimal effect on the current. The DHP agonist Bay K8644 inc reases current magnitude and shifts the

    voltage-dependence o f activ atio n by appro ximately -1 0 mV. In addition, the c urrent is inhibited by the DHP antagonistnifedipine. However, the affinity of DHPs for channels is generally lower than that observed with .

    Moreover, unlike other DHP-sensitive channels, the cloned is partially and reversibly blocked by high concentrations of

    -CgTx (10-15 M).

    The predominance of the a1D subunit-containing VGCCs in the cochlear hair cells and in the -cells of the pancreas suggestthat these c hannels may be inv olv ed in tonic ex oc ytotic release in these c ells Kollmar et al (1 997 )

    suggest that the electrophy siological properties of the subunit, such as its lack of inactivation during depolarizations

    may render it suitable for mediating tonic release. In addition, as suggested by the localization of channels on the cell

    bo dy and at the b ase of de ndrites of neuro ns in the CNS, these c hannels may be inv olved in integr ating signals impingingupon the neuron from multiple sources.

    / Ca 2.3

    The class E gene encodes a VGCC subunit (a1E) that does not fall neatly into either the HVA or LVA c ategories. cDNAs

    have been isolated from rabbit (BII-1, BII-2) and rat (rbE-II) brain, and fromDyscopyge o mmata. These clones

    code for proteins between 217 8 and 2259 amino acids with predicted molecular masses of approximately 25 0 kDa. Splicevariants of the r abbit brain c hannel, BII -1 and BII-2, differ from one another in their c arbo xy l tai ls, result ing in the additio n

    of a putative PKA site.

    The class E clones appear to be mo re c losely related to the DHP-insensitive non-L-type channels (54-60% amino acididentity) than to the L-type channels (less than 45% similarity). However, class E channels are less similar to either class Aor B channels than these two classes are to one another, suggesting that the class E channels are members o f a novel, moredistantly related subgroup of DHP-insensitive channel (fig.2).

    Northern blotting studies have identified transcripts ranging in size from 10.5 to 12 kb in the mammalian CNS. High

    levels of ex pression was identified in the cerebral co rtex, hippocampus, and striatum, while lower lev els were detec ted inthe olfactory bulb, midbrain, and Purkinje and granule cell layers of the cerebellum. While appears abundant in brain,

    none was detected in skeletal muscle, heart, stomach or kidney. A t the subcellular level, protein was localized nearly

    exc lusive ly to the cell body of neurons throughout the CNS. Dendritic staining varies acro ss brain regions. For ex ample, inthe co rtex and hippocampal formation there is barely perceptible staining of the dendritic branches, while in Purkinje cells, antibodies labeled the distal dendritic branches, but not the main dendritic trunks.

    The channel was initially reported to be a novel member of the LVA family of Ca channels. Expression of rbE-II inXenopusooc ytes pro duced a channel that activated rapidly at low membrane potentials (threshold -50 mV ) andinactivated significantly dur ing the depolarization. Other v oltage-dependent parameters of this channel (curr ent-voltagerelationship, voltage-dependence o f inactivation) were also considerably mo re negative than those of other c loned HVAchannels. The rbE-II c urrent magnitude increased steeply with increasing depolarizations, peaking at aro und 10 mV, andsteady state inactivation analysis indicated that the c hannels were inactivated near the resting membrane potential of thecell. In addition, rbE-II channels were equally permeable to Ca and Ba , a property reported to be unique to the LVA

    channels. Another similarity with LVA c hannels was the high sensitivity of the current to block by Ni. Furthermore, thechannel was found to b e ex pressed in many of the cells that have be en shown to possess T-type curr ents. Howev er, Soong etal (1993) noted a number of discrepancies between rbE-II and native T-type currents. For example, although the

    voltage-dependent pro perties o f rbE-I I c urrents were more negativ e than those o f the other c loned HVA Ca channels, the

    activation and peak current potentials were not as hyperpolarized as for typical T-type channels. Analysis of the

    electrophysiological properties of other class E channels have produced some results that contradict those of Soong

    et al (1993). In these studies, the clones formed HVA c hannels, activating at approximately -10 mV and peaking at

    +30 mV. The single channel conductance of channels is also much larger than that of T-type c hannels (12-14 pS vs. ~8pS). As channels share propert ies with LVA as well as HVA channels, the detection of pure currents in native

    cells may difficult.

    It has been suggested that the class E channels may b e one of a group o f channels comprising the R-type c urrent.

    The two currents share some elec trophy siological and pharmacological characteristics, such as strong vo ltage-dependenceof activation and insensitivity to DHPs, -CgTx, and -Aga IV A. However, the R-type c urrent is smaller in Ca than in

    Ba , whereas the channels support the two currents equally . Mo st re levant, mice lacking the gene entirely

    still exhibit significant R-type c urrent. Thus, while class E channels may comprise a component of R-type current in

    cerebellar granule cells, the R-type current may actually result from incomplete block of other Ca channels by applied

    pharmacological agents, the expression of additional splice variants of already identified Ca channel subtypes, or as ye t

    to-be-identified subunits.

    1 D 1 C1 5 4

    1 D

    1 D

    1 C1 5 4

    2 +

    2 +

    5 1 , 1 6 31 D 1 C

    1 6 8 , 1 6 9

    1 D5 1

    1 6 3 ,1 6 4 ,1 6 6 ,1 7 0,1 7 1 1 6 6

    1 D

    1 D

    6

    1E v

    1 1 E1 7 2 1 2 1 1 4 1

    1 2 1 , 1 4 1 , 1 7 2

    1 2 1 , 1 7 2

    1 E

    1 E

    1 E1 7 3

    1 E2 + 1 2 1

    2 + 2 +

    1 2 1

    2 +

    1 2 1

    1 7 4 - 1 7 6

    1 2 11 E

    1 E1 4 1 , 1 7 7

    1 E 1 E

    8 9 , 1 7 8 , 1 7 9

    2 +

    2 +1 E

    8 9 , 1 7 71 E

    1 8 0

    2 +

    2 +

    1

    http://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30874/?report=objectonly
  • 8/12/2019 Voltage Gated Calcium Channel

    10/28

  • 8/12/2019 Voltage Gated Calcium Channel

    11/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 11

    membrane-spanning domains. Most of the amino ac id changes in these regions are conserv ative, thereb y maintainingstructural elements co mmon to v oltage-gated ion c hannels. The charges located in the fourth transmembrane segment ofeach domain are c onserve d, as are the pore-forming loops b etween the fifth and sixth transmembrane segments. In HVAchannels, a glutamate residue located in each of these four loops is believed to determine the ion selectivity of thechannels. All T-type channels cloned thus far contain aspartate residues instead of glutamates in the domain III and

    IV P-regions. This difference may account for the difference in the permeation properties seen between high and low

    voltage ac tiv ated channels. The intr a- and ex tracellular linkers joining the transmembrane do mains share litt le homologywith either HVA channels or with other T-type channels. Furthermor e, the T-ty pe channels do not seem to possess spec ificfunctional motifs that have been identified in HVA channels, including the binding site in the I-II linker or the putative EF-hand motif in the c arbox yl tail.

    The three classes of T-type channel have been localized using Northern blotting, in situ hybridization, and RT-PCRtechniques. The subunit appears to be expressed abundantly throughout the brain and to a

    lesser degree in heart. Low lev els have also been detec ted in placenta, lung, and kidney. High levels of transcript areobserv ed in the cereb ellum, hippocampus, thalamus, and olfactory bulb, with lesser amounts localized to the cer ebralcortex and septal nuclei. Initially, the was detected only in cardiac tissue, kidney, and liver, with very little, if any,

    expression in the brain. However, a subsequent study suggests that the subunit may be responsible for a large

    proportion of the T-type current in sensory neurons, and another study indicates the expression of in all areas in the rat

    brain. transc ripts have o nly been detec ted in brain, with one study sho wing specific expression in the

    striatum of adult rats.

    Expression o f these three subunits inXenopus oocy tes and HEK-293 cells demonstrates

    that they support currents with most of the characteristics expec ted of T-type channels. Currents activated upo n weakdepolarizations from negative holding potentials. In one study, the three T-type channel classes had differing permeabilityproperties. As has been noted for classic T-type currents, channel was more permeable to Ca than to Ba. However,

    channels were more permeable to Ba than Ca , while channels were equally permeable to the two ions. In mostcases, currents were inhibited by mibefradil and Ni, although the IC of each T-type class varied significantly. The

    activation and inactivation kinetics of the T-type channels are strongly vo ltage-dependent. While rates of activ ation andinactivation are slow near threshold potentials, they accelerate as the strength of depo larization increases. Deactivation isalso v oltage-dependent, increasing at more hyper polarized po tentials. Steady-state inactivation analysis indicates that themajority of the c hannel population would b e inactivated at the resting potential of most cells. However, because all thechannels are not inactivated at the resting potential and the threshold o f activation is so negative, a small proportion o fchannels are capable o f opening at the resting potential, thus producing a window curr ent. The window current re fers to asmall, but sustained influx of Ca that occurs ev en when the cell is ostensibly at rest. This current can contribute to the

    ov erall exc itability of the membrane and may c ontribute to the bur sting and pacemaker activities attributed to the T-typechannels. Finally, as anticipated for T-type channels, the three ex ogenously expre ssed channels have small single channelc onduc tanc es o f 5 ( ), 7 . 5 ( ), and 1 1 ( ) pS.

    Similar to the different c lasses of c hannels within the HVA subfamily, the biophysical prope rties of the three T-type channelsvary co nsiderably . The and po ssess v ery similar activ atio n and inac tiv atio n po tentials, while tho se o f the

    appear to be slightly more negative. Rates of activation and inactivation of and currents also are quite similar.In contrast, activation and inactivation rates for currents are significantly slower. In addition, the activation threshold

    of channels also differs from the values obtained for and channels. However, varying results have been reported

    concerning the direction of the observed voltage shift. The rat brain clone studied by J.-H. Lee et al (1999) activated

    at more posit ive potentia ls than did the and channels , while McRory et al (1999) reported current

    activation at considerably more negative potentials. That the properties of the channel differ from those of the and

    is not entirely unexpected when one takes into account the degrees of similarity seen amongst the three channels.

    Furthermore, multiple splice variants of the have been identified and may account for the contrasting results

    reported for currents. Finally, the properties of the three LVA c hannel clones do not account for all of the T-type

    characteristics in native cells and there may be additional classes of T-type c hannels and/or a set of as y et unknownauxiliary subunits specific to LVA channels which further modulate the properties of the LVA subunit. For example,

    robust alternate splicing for channels has been reported and shown to result in significantly altered biophy sical

    characteristics.

    Auxiliary Ca Channel Subunits

    Biochemical studies have shown that in addition to the pore-forming subunit, HVA Ca channel complexes include two

    or three o ther proteins: a subunit, an - subunit, and in some cases, a subunit (fig.1).

    Subunits

    The subunit is the most ex tensively studied of the auxiliary subunits and appears to have the most profound effects on thefunctional properties of the subunit. In mammals, there are at least four different subunits ( , , , and ) which

    are encoded by distinct genes. The transcripts of at elast two of these genes, the and , are alternatively spliced to give

    rise to , and and and .

    Biochemical and primary seque nce analyses indicate that the subunits are hydrophilic with no transmembrane segments

    1 9 7 , 1 9 8

    1 2 4

    1 2 4

    1 2 4 - 1 2 6 , 1 9 3 , 1 9 5 , 2 0 0 , 2 0 11 G

    1 H1 2 5 1 9 9

    1 H

    1 H2 01

    1 I1 2 6 , 1 9 5

    2 00

    1 2 4 ,1 2 6 1 2 5 ,1 9 3 ,1 9 5 ,1 9 6 ,2 00-2 03

    2 001 G

    1 H

    2 + 2 +

    1 I5 0

    1 9 3 , 2 0 0

    2 +

    1 H 1 G 1 I1 2 6

    1 2 6 , 2 0 01 G 1 H

    1 H 1 G 1 H

    1 I

    1 I 1 G 1 H

    1 I1 2 6

    1 G 1 H2 00

    1 I

    1 I 1 G

    1 H

    1 I1 9 4 , 1 9 5

    1 I

    1

    1 G2 03

    2+

    12 +

    2

    1 1 2 3 4

    1 2

    1 a 1 b 1 c 2 a 2 b

    http://www.ncbi.nlm.nih.gov/books/NBK6181/figure/A30873/?report=objectonly
  • 8/12/2019 Voltage Gated Calcium Channel

    12/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 12

    or glycosylation sites. The subunits contain potential phosphorylation sites for both protein kinase C and cAMP-

    dependent protein kinase. The modulatory effects o f these enzymes on V GCC function may, in part, be the r esult of theiractions on this auxiliary subunit.

    Although the spec ific e ffects o f channel mo dulatio n depend upo n the subunit isoform, all subunits appear to hav e thesame general impact on the properties of HVA subunits. Coexpression of the a1 and subunits in both L-cells and

    Xenopusooc ytes incre ased whole-cell currents and DHP binding without affecting the leve l of message. This suggests that

    rather than enhancing expre ssion of the subunit, the subunit may promo te insertion of the subunit into the

    membrane and/or stabilize a specific conformation of the protein have proposed that the subunit potentiates

    coupling of the gating-charge movement c aused by changes in membrane potential with the opening of the pore therebyincreasing the pro bability of channel activation and, in turn, increasing the peak current.

    In addition to increasing the magnitude of the current through the subunit, co-expression of the subunit alters channel

    kinetics. For most subunits, the rate of inactivation is increased and there is a shift in the v oltage-dependence of activationto more negative potentials. The effect on kinetics of inactivation, howev er, varies depending

    upon the class of subunit expressed. The and proteins increase the rate on inactivation, while the subunit

    significantly slows inactivation.

    These modulatory effects are observ ed regardless of which and subunits are co -expressed, suggesting that the

    mechanism through which the subunit acts is common to all HVA VGCCs. The region required for subunit modulation ofthe subunit has been localized to a stretch of 30 amino acids at the amino-terminal side of the second of two conserv ed

    domains. This region, known as the BID ( subunit interaction domain) is also responsible for anchoring the subunit to

    the . The subunit has been shown to bind to a conserve d motif of 18 amino acids in the intracellular loop between

    domains I and II of the subunit (the AID: subunit interaction domain). The observation that the subunit from

    skeletal muscle dramatically increases the magnitude of the current through brain subunits when co-expr essed in

    Xenopusoocy tes further supports the idea of a common mechanism of - subunit interaction. The subunits exhibithomology with the Src homology 3-guanylate kinase domain of membrane associated guanylate kinases and this regionappears to regulate inactivation of HVA VGCCs. A more detailed account of subunit physiology is provided in the

    Chapter from the Charnet lab.

    - Subunits

    Purification studies indicate that the VGCC - subunit consists of two distinct subunits ( and that are disulfide-bonded

    in the native state. The and subunits are derived from a single gene product that is proteolytically cleaved to form

    1 43 -kDa and 27 - kDa subunits. There are currently fo ur genes that code fo r the - fam ily ; - , - , - ,

    and - .

    Both proteins are heavily glyc osylated, supporting the prediction that the - subunit is predominantly extracellular

    and a recent study has determined that no more than five residues comprise the cytoplasmic portion of the protein. The

    complex is anchored in the membrane by a single transmembrane segment formed by a portion of the subunit. Thetransmembrane domain is thought to interact with other subunit(s) in the V GCC complex while the extrac ellular domain isresponsible for the modulatory effects.

    There are seve ral splice variants of the - auxiliary subunit, due primarily to the alternative splicing of three specific

    regions. The total splicing combination of these three re gions reveals 5 unique isoforms which are all ex pressed in a tissuespecific manner. Skeletal muscle and brain express the isoforms - and - respectively . The heart expresses

    mainly - and - , and smo oth musc le ex presses - and - . I nterestingly the cardio vascular sy stem

    expre sses all five isoforms.

    The - subunit is expressed in many different tissues including heart, brain, pancreas, testis lung, liver, kidney. The

    - subunit has two regions of alternative splicing. The first region is found on the a2 subunit between residues 661/663

    and involve s the addition of eight amino acid residues. The second is located on the pro tein and is characterized by theaddition of three different residues. The resulting three isoforms of the - subunit are all expressed in hMTC (human

    medullary thyroid cells.) Alternately, in the human heart, only - is expressed. cDNA cloning of the - subunit did

    not uncov er additional splice variants, and this particular - gene appears to be expressed exc lusive ly in the brain.

    In 2002, a new member of the - auxiliary subunit family was described. - demonstrated a unique property in that it

    was shown to b e expressed primarily in endo crine tissues. Immunohistoc hemical studies r ev ealed that the - subunit

    has limited distribution in special cell types of the pituitary, adrenal gland, colon, and fetal liver. Whether the a -d

    subunit plays a physiological role in certain endocrine tissues remains to be seen. Alternative splicing of the - gene

    gives rise to four potential variants (called a-d.

    The functional effects of the - are thought to be more subtle that those of the subunit and are highly dependent on the

    class of subunit and the cell type used for exogenous expression. For example, Singer et al (1991) found that the Ca

    current inXenopusoocytes expressing the cardiac protein was greatly enhanced by co-expression of the - subunit

    from skeletal muscle. In addition, the rates of activation and inactivation were increased and the v oltage dependence o finactivation was shifted to more negative potentials. In contrast, Varadi et al (1991) did not observe these effects when

    2 0 4 - 2 10

    2 1 1

    1

    1

    1 12 1 2 - 2 1 5

    1

    1 4 8 , 1 5 4 , 2 0 7 , 2 0 9 , 2 1 6 - 2 1 9

    1 3 2

    1

    12 2 0

    1

    1 12 2 1

    11 2 8

    1

    2 8 9

    2

    2 28 , 2 2 2

    2

    22 2 3

    2 2 1 2 2 2 3

    2 4

    2 2 42

    2 2 5

    8 , 1 4 2 , 2 2 3 - 2 2 5

    2 1

    2 2 62 1 A 2 1 B

    2 1 C 2 1 D 2 1 D 2 1 E2 2 7

    2 22 2 7

    2 2

    2 2 82 2

    2 2 A 2 32

    2 2 4

    2 42 2 9

    2 4

    2 42 2 9

    2

    12 1 7 2 +

    1 C 2

    2 1 8

  • 8/12/2019 Voltage Gated Calcium Channel

    13/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 13

    they co-expressed the skeletal muscle subunit with the - subunit in L cells. UnlikeXenopus ooc ytes, L cells do not

    express endogenous calcium channel subunits. Thus, it is possible that the - subunit acts synergistically with other

    auxiliary subunits to modulate the properties of the . Finally, there is some evidence that the - subunit is required for

    efficient expression and/or trafficking of the subunit to the cell membrane, an idea that is supported by recent studies

    in tsA-201 cells showing that - subunits are key modulators of current densities of , and channels.

    Subunits

    (Cac ng1)

    The VGCC subunit protein was first identified in guinea pig skeletal muscle during the purification o f 1,4-dihydrpy ridinerecepto rs. This VGCC heteromultimeric pro tein consisted of five different subunits. The subunit, a 28-35 kDa on SDS-

    Page proved to be one of them. Several years later Jay et al (1990) isolated rabbit skeletal muscle cDNA and

    uncov ered the primary sequence the subunit. The cDNA was a 666-nucleotide clone, with a reading frame that would

    y ield a 222 amino acid gly copro tein containing four transmembrane do mains. has b een sho wn to be expressed primar ily

    in skeletal muscle, but recently has also been shown to be weakly expressed in the aorta. Mice lacking the subunit

    display altered skeletal muscle calcium cur rent. Functional effects of the protein include a hyperpolarized shift in the

    steady state inactivation prope rties in skeletal muscleVGCC.

    Stargazin or (Cacng2)

    For nearly a decade the pre sence of other like subunits remained undetected. It was the discov ery of the mouse stargazingene Cacng2 by Letts et al (1998) that ultimately led to the description of seven new subunits. The stargazin gene and

    protein were named after the Stargazer mouse, a mouse strain that is prone to absence seizures including an upward nec k tilt

    and prolonged gaze. These mice had acquired a transposon in a 1.5kb region on c hromosome 1 5 and were hetero zygousrecessive. The stargazin gene, renamed , had inherited a stop codon rendering the protein inactive and

    truncated. The etiological consequences o f a mutated Cacng gene responsible for the absenc e epilepsy phenoty pe of theallelic stargazer (stg) and waggler (wag) mutant mice.

    , has been c lassified as a subunit based on its struc tur al similarity to , despite hav ing o nly weak protein seq uence

    identity (25%). At the tissue expression level, unlike , the subunits are found to be expressed in the brain. Letts et

    al (1998) found mouse mRNA to be expressed in adult mouse brain abundantly with highest expression in

    cereb ellum, olfactory bulb, cereb ral cortex , thalamus, and CA3 and dentate gyrus regions of the hippocampus.

    / (Cac ng3/4)

    Despite the low sequence homology of with its original counterpart, many of the newly discovered subunits

    demonstrate high similarity with the subunit. , and are of closest similarity and make up a subfamily of neuronal

    sub units. Klugbaue r et al (2 00 0) used Norther n blots to sho w that and ex pressions we re ex clusively restr ictedto the brain. However, Chu et al (2001) found to be additionally present in the atrium, aorta, and lung.

    (Cac ng5)

    The subunit is unique in that it is expressed not o nly in the brain and skeletal muscle, but also in different types of

    endocrine tissues, primarily the liver, kidney, heart, lung and testes. There is some disagreement on the categorization ofthis subunit as a bona fide subunit, and it is also referred to as the pr protein. In ex ogenous ex pression experiments themouse pr protein has been shown to the modulate properties of the T-type channel.

    . ,

    In 2001 Chu et al described three new human and rat like subunits (human , , and ). was found in the

    atrium, ventricle, skeletal muscle and a short splice v ariant of its kind was found in atrium, v entricle, aorta, br ain, and lung.

    is ex pressed in all tissues ex ce pt aorta, kidne y , liv er, and spleen. was fo und only in brain and testis.In summary, the eight subunits are technically c onsidered to b e part o f the Claudin family of proteins, and aredifferentially e xpressed among a v ariety of different tissues. They all display four membrane-spanning regions with boththeir C and N termini located intracellularly, while all extrac ellular re gions display N-glyc osylation sites. All the Cacnggenes have four exons (with the exception of Canga5 and 7 which have five exons), and exon1 is predominately the

    largest. The carboxyl terminus is of particular interest given that , , and have a PDZ binding motif in this

    region. The C-termini of the and are longer and lack this consensus motif, however it is interesting to note

    that they have inherited an SS/TSPC site, probably designated for protein interactions. Of particular note, stargazin and

    other member of the subunit family ( , and ) have recently been shown to define a novel family of transmembrane

    AMPA r ec eptor regulatory proteins (TA RPs). The TA RPs appear to regulate two dist inct ro les in A MPA r ec eptor

    signaling: trafficking of AMPA receptors to the plasma membrane and an agonist-mediated dynamic interaction that maycontribute to sy naptic plasticity.

    2 22 3 0

    2 1 92

    1 2

    11 1 0

    2 1 C 1 B 1 E2 3 1

    1

    11

    2 3 2 2 3 3

    1

    12 3 4 2 3 5

    1

    12 3 6 , 2 3 7

    2

    2 3 8

    2 3 8 , 2 3 92

    2 3 8 , 2 4 0

    2 12 3 8

    1 22 3 8

    2

    3 4

    2 1

    2 2 3 42 4 1

    3 42 3 54

    5

    5

    1 G2 4 1

    6 7 8

    2 3 56 7 8 6

    7 8

    2 4 2

    2 3 4 82 4 3 , 2 4 4

    7 52 3 5

    3 4 82 9 0

    2 9 1

  • 8/12/2019 Voltage Gated Calcium Channel

    14/28

    3/26/2014 Molecular Properties of Voltage-Gated Calcium Channels - Madame Curie Bioscience Database - NCBI Bookshelf

    http://www.ncbi.nlm.nih.gov/books/NBK6181/?report=printable 14

    1 .

    2.3.

    4.

    5.

    6.

    7 .

    8.

    9.

    10.

    11 .

    12.

    13.

    14.

    15.

    16.

    17 .

    Effects of Subunits on Channel Biophysics

    Ove r the ye ars, many different combinations of all of the hetermultimer subunits with different acco mpanying subunitshave been tested. Wei et al (1991) co-expressed rabbit Cardiac with , and observed increased peak currents. Letts

    et al, (1998) demonstrated the modulatory effects of neuronal , by co-expressing the subunit with - - -. The

    effects of included a hyperpo larizing shift of the steady state inactivation properties of the channel. Similar results were

    obtained by Klugbauer et al (2002) with and . The same hyperpolarizing effects on channel SSI properties were

    observed with , but were not with . It is interesting to note that shifted the activation potential to a more

    depolarized level, when co-transfec