This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Molecular Structure of Ion Channels In the previous sections, we showed how voltage gated sodium and potassium channels work and how the cooperative operation of those channels generates and shapes the action potential. We showed that depolarization opens Na+ and K+ channels, but that Na+ channels automatically shut off, i.e., inactivate, while K+ channels do not, and only close when the cell repolarizes. These are the features that characterize ion channels as shown by electrophysiological studies. Electrophysiological studies, however, cannot show the structural underpinnings of the channel. In other words, voltage clamp and other electrophysiological studies show what the channel does and its electrical behavior, but they cannot reveal its structure nor how the particular structure it possesses produces each aspect of its behavior. The questions that we explore in this chapter concern the molecular structures of channels and how those structures impart activation, inactivation, and selectivity for passing only one type of ion. Since each channel is a protein, we will examine the molecular structure of a channel, its complement of amino acids, and how the arrangement of amino acids in different parts of the channel enables the channel to operate in a way consistent with its behavior revealed in voltage clamp studies. Before turning to the structures of ion channels, a brief review of amino acids and proteins will be helpful. A brief review of amino acids and proteins Individual amino acids are linked together to from polypeptide chains by peptide bonds in which the carboxyl group (COOH) of one amino acid is attached to the amino group (NH2) of the next. Each polypeptide thus has a free NH2 group at one end (the amino or N terminus) and a free COOH group at the other (the carboxyl or C terminus). The side groups of the amino acids (R1 and R2 in the example below) protrude from the peptide backbone and vary with the amino acid. Twenty amino acids are commonly used in the construction of proteins. Each amino acid within a polypeptide or protein is called a residue because amino acids lose a few atoms (usually hydrogen and oxygen) when they are polymerized into a larger molecule. Figure 1. Amino acids are linked by peptide bonds to form polypeptides. Amino acids are connected by a peptide bond formed between the COOH (carboxyl) group of one residue and the NH2 + (amino) group of the next. R1 and R2 indicate the amino acid side chains. Some amino acids are acidic and positively charged, some are basic and negatively charged, some are polar but uncharged whereas others are nonpolar. Nonpolar residues are hydrophobic (water hating) and stretches of hydrophobic residues therefore tend to be found in the portions of ion channels that are embedded in the lipid membrane. The portions of the protein that are embedded in the membrane are called 1 the transmembrane segments of the protein. The protein in this case is an ion channel. In contrast, charged and polar residues, which are hydrophilic or water loving, are more likely to be located in regions of the protein that extend into the extracellular or intracellular fluids. Finally, some amino acids have additional features that are important for channel functioning. Serine, threonine and tyrosine residues, for example, can be modified by the attachment of a negatively charged phosphate group (these residues can be phosphorylated). This changes the electrostatic properties of the amino acid, alters the local structure and thus changes the function of the protein. As we shall see in later sections, phosphorylation modulates the properties of many ion channels. For example, phosphorylation of some channels increases its conductance, thereby allowing more ions to flow through the channel than when it is not phosphorylated. Proteins assume one of two main types of structures: the helix and the sheet. Both result from hydrogen bonding between the NH and C=O groups of the polypeptide backbone. The helix is the most prevalent form and is particularly important for determining which portions of the protein comprise the transmembrane segments (which parts of the protein are embedded in the membrane). The polypeptide backbone is twisted into a helix such that about 20 amino acid residues comprise a length of about 30 angstroms along the helix. Since the cell membrane is also about 30 angstroms wide, a stretch of about 20 amino acids is just sufficient to span the membrane. Below we illustrate other factors which further show that the transmembrane segments of ion channels are formed from helices comprised of ~20 amino acid residues . The molecular structure of the Na+ channel The first voltage gated ion channel whose structure was determined was the Na+ channel. Na+ channels are particularly rich in skeletal muscle and in the electric organ of some eels, the eels that generate hundreds of volts used to stun or even kill their prey. In all cases, a very large glycoprotein was purified and was called the subunit. Messenger RNA was obtained and the full amino acid sequence of the subunit of the Na+ channel was deduced from the sequence of nucleotides of the complementary DNA generated from the mRNA. The Na+ channel protein of the subunit is comprised of about 2000 amino acids and is the product of a single gene. But so what? What can be learned about the structure of the Na+ channel by simply knowing its sequence of amino acids? Luckily there are a few rules that provide major insights into which regions of the protein lie within the lipid membrane and how the protein itself is structurally and functionally organized. Of importance in this regard is the hydrophobicity plot. Each amino acid is placed in the plot as it occurs in the protein and each is assigned a hydrophobicity value, which reflects its ability to interact with water. Recall from the previous discussion of amino acids that nonpolar amino acids are hydrophobic (water hating) and are assigned a high positive value, whereas polar amino acids are hydrophilic and are given a negative value. A running average of these values over several amino acids is then calculated around each amino acid in the protein sequence, and plotted as a function of position along the protein chain. Also recall from the previous description that a sequence of about 20 hydrophobic helictial amino acids would span the cell membrane. Thus, sequences of ~20 hydrophobic amino acids in the hydrophobicity plot represent a transmembrane segment. Let's examine the hydrophobicity plot of a Na+ channel shown in the top panel of Fig. 2 in greater detail. Look first at the pattern of the initial 400 amino acids. Beginning at amino acid #75 and ending at about #225, there is a stretch of about 150 amino acids that are all hydrophobic. Since 30 hydrophobic amino acids would span the membrane, this suggests that there should be 5 membrane spanning segments from 75225. If you look carefully, and use a little imagination, you can also see 5 little peaks in this region. Each "peak" is about 20 amino acids long, and each is indicated by a colored box above that span from amino acid 75225. We refer to these peaks as a segment (S), and they are labeled S1S5, as 2 shown above each colored box and in the middle panel of Fig. 2. This is followed by a short span of 100 hydrophilic residues, from about #225325, followed by about more 30 hydrophobic residues. The last set of hydrophobic residues thus forms a sixth transmembrane segment and is labeled S6 (magenta box) in Fig. 2. These 6 transmembrane segments are collectively called domain I, and it is apparent that three other domains, labeled domains II, III and IV, closely resemble domain I, in that domains IIIV also contain six transmembrane segments. Figure 2. Top panel hydrophobicity plot of the amino acid sequence of a Na+ channel. The colored regions show the hydrophobic amino acids. The hydrophobic amino acids show a series of peaks, where each peak is assigned a colored box, where each colored box represents a transmembrane segment. There is a repeating pattern such that series of 6 peaks is repeated four times. Each series of 6 peaks is a domain of the channel. Middle panel: linear arrangement of transmembrane segments from amino acid sequence and hydrophobicity plot. Lower panel: top view of Na+ channel showing how four domains are actually arranged around a central pore. 3 The liner arrangement of the segments is shown in the middle panel of Fig. 2. Each segment in a domain is linked to its neighboring segment by an intra or extracellular linker of hydrophilic amino acids. In addition each of the four domains is separated from its neighboring domain by a short intracellular hydrophilic segment. Notice that the hydrophilic stretches of amino acids are not embedded in the membrane but rather are in the extracellular or intracellular fluids and link the transmembrane segments. The arrangement of the segments and domains is not stretched out linearly in the membrane. Rather the four domains are folded, each next to its neighbor, to form a circular ring with a central pore. This arrangement would be seen if you could look down on the channel from the extraceullar space, as shown in the bottom panel of Fig. 2. In short, the sodium channel is a single, large protein composed of four domains, where each domain is, in turn, composed of 6 transmembrane segments. The transmembrane segments, together with the intracellular and extracellular linking segments, the hydrophilic linking regions, play specific roles in the channel operation. The above section showed the molecular arrangement of the Na+ channel. Another important channel that conducts cations is the calcium channel, which has not been discussed so far in the course but is exceptionally important and will be considered later in the course. The only point I wish to make here is that the structure of the calcium channel is strikingly similar to that of the Na+ channel (Fig. 3). It too is constructed from a single protein, has four domains each with six transmembrane segments. Figure. 3. Molecular structures of Na+ and Ca++ channels are similar. In the sections below we consider three features of the various segments; 1) which segment "senses" voltage (charge on the inside of the cell) and how that segment responds to membrane depolarization to open activation gates; 2) which portions form the inactivation gate and how the inactivation gate is coupled to the voltage sensor; 3) the pore region through which the ions travel and how the structure of the pore region 4 confers exquisite selectivity for one particular ion and excludes other ions. As a preview, the S4 segment is the voltage sensor, the intracellular linkage region between segments III and IV form the inactivation gate, and segments S56 form the pore. We first consider the voltage sensor in S4 and then turn to the inactivation gate. In the final portion we consider the ion selectivity filter in the pore, which is one of the great stories in molecular neurobiology. The S4 segment is the voltage sensor and opens the activation gate For an ion channel to respond to changes in the electric charge across the membrane, some charged part of it must act as a voltage sensor and must undergo some movement that triggers the opening of the activation gate. This idea was originally postulated by Hodgkin and Huxley in 1952. They proposed that the gating of a channel was accompanied by movement of a charge within the membrane. Such a movement of charge within the protein constitutes a small current, which is now called the gating current (Fig. 5). It should be noted that the gating current is very small because it is only due to the small movement of a charged element of the protein and it occurs before the channel actually opens, which then allows a much larger current, carried by Na+ ions, to flow through the pore of the channel. The S4 segment has the exact properties required for a voltage sensor. In the S4 segment, basic (positively charged) amino acids occur at every third residue and form a spiral of positive charge on the outside of an helical segment (an helix makes a complete turn for each 3.4 amino acids). In the S4 transmembrane segment, each of these positive charges is paired to negative charges in the helices of neighboring segments, thereby helping to stabilize the segment or helix (Fig.4). The positive charges on the outside of the cell also aid in stabilizing the configuration of the helix at the normal resting potential. Figure 4. The positively charged amino acids in the S4 segment are attracted to negatively charged amino acids (E&D) in S2 and S3. This holds them in place at rest. Depolarization of the cell overcomes this electrostatic attraction, thereby repelling the positively charged S4 segment, which then initiates channel opening. When the cell is depolarized, and the inside of the cell gains positive charges, the positive charges act to repel the positively charged helix and displace it upward into the extracellular space. Each charge on the outside of the helix is shifted upward to the position of the next charged neighbor where it is stabilized. The net effect is to cause the helix to rotate upwards into the extracellular fluid and thereby add positive charges to the outside of the membrane. This movement of positive charges in the membrane repels positive ions in the extracellular fluid and generates a small current, called the gating current. The gating current is evoked before the channel opens and thus it occurs before the ionic current that flows through the open channel. The upward displacement of the S4 segment is coupled to and thus opens the activation gate, as shown in Fig. 5. Recall that there the sodium channel has 4 domains, and each domain has an S4 segment. The S4 segments in domains I, II and III have to move upward to open the activation gate, which allows ionic current to flow through the pore of the channel (Fig. 7, top panel). 5 Figure 5. Schematic arrangement of the voltage sensor in S4 segment. In the resting state, the S4 segment is positioned in the membrane, attracted by the negative charges inside. When + charges enter the cell causing depolarization the + changes in the sensor are repelled by the + charges inside the depolarized cell. The repulsive + charges "push" the S4 segment upward into the extracellular space, which in turn repels + charges in the extracellular fluids, thereby generating the "gating current". The gating current is transient, and lasts only as long as long as the S4 segment moves. The upward movement of the S4 segment is coupled to other segments and thereby opens the activation gate, allowing ions to flow through the pore. The flow of ions through the pore is called the ionic current. The effects of repolarization are exactly the opposite: the increase in negative charges on the inside (and/or the increase in positive charges on the outside) then attract the positive charges on the S4 helix, causing it to move downward back to its normal position. That movement, in turn, is also coupled to and closes the activation gate. Inactivation Activation of Na+ channels, the opening of the channel so that ions flow through the pore, is always shortly followed by inactivation, where a second gate automatically closes and prevents ions from flowing through the pore, even though the cell is depolarized. One might then ask whether the voltage sensor also comprises a portion of the inactivation gate? The answer is no; the voltage sensor and activation gates are not part of the inactivation gate and are independent of it. This was shown in the 1970s when investigators injected the enzyme, pronase, into the axoplasm of the squid giant axon. When an experiment was conducted in voltage clamp, applying positive current to the inside of the cell caused an opening of the activation gate, and thus an inward current carried by Na, and when the applied current was stopped, the activation gate closed. Thus the operation of the activation gate was not affected by pronase. What was affected was inactivation; with pronase, there was no inactivation and inward current, carried by Na+, flowed so long as the cell was clamped at a positive membrane potential. In other words, the activation gate behaved normally but the inactivation gate was eliminated by intracellular pronase. This experiment showed something more; since the pronase was injected into the axon, the inactivation gate must be on the cytoplasmic side of the channel protein. Recent evidence shows that the inactivation gate is on the intracellular loop that connects the III and IV transmembrane segments, as shown in Fig. 6. Presumably pronase acted by cutting some site or sites on that intracellular loop. 6 Figure 6. Two views of the operation of the inactivation gate. The top panel shows the ball and chain view, where the ball (shown as an oval) is attached by two linkers to two segments of the channel. The idea is that when the voltage sensor in S4 moves in response to depolarization and opens the activation gate (open state in middle), ions flow through the pore. A moment later the ball is pulled into the pore, blocking it and preventing any further flow of ions, even though the cell is still depolarized and the activation gate (pore) is "open". The bottom panel shows this in more detail. Specifically, the linkers connect to only two domains, domain III and domain IV (domain II is not shown for clarity). The ball is formed by the IFM motif of amino acids, with one side linked to domain III and the other to domain IV (IFM is a 3 amino acid sequence, isoleucine (I), phenylalanine (F), and methionine (M)). When the cell is depolarized, the linker in domain IV is pulled upward and causes the ball, formed by the IFM sequence, to plug the channel thereby producing channel inactivation. Thee four domains are shown, but the segments in each domain are not shown. The linkage from domain III is on S6 while the linkage on domain 4 is on S1. The linkages to each segment in domains III and IV can be seen in Fig. 2 and 7. There is only one inactivation gate on sodium channels. The gate consists of a ball and two chains. The ball has a unique amino acid sequence, the IFM motif, that is connected to intracellular linkages, the chains (Fig. 6). One chain connects the ball to the S6 segment in domain III while the other connects the ball to the S1 segment in domain IV (see Figs. 6, 7). Here is how it works. Upon depolarization, the voltage sensors (S4 segments) in domains I, II and III move and open the activation gate. Na+ ions now flows through the channel pore, from the outside of the cell to the inside, thereby generating the upstroke of the action potential. The voltage sensor (S4 segment) in domain IV, however, is slower than the voltage sensors in the other domains, and begins to move about a millisecond after the other voltage sensors have already moved. When the voltage sensor in domain IV is displaced upward, it moves the ball and causes it to block the channel pore. Thus, although the cell remains depolarized and the activation gates remain open, current now stops because the pore is blocked by the inactivation ball. The channel is now inactivated. Thus, in the Na+ channel the four domains do not behave equally. The inactivation complex interacts with the segments of domains III and IV but is not affected by domains I and II. The activation gate is controlled entirely by the S4 segments in domains I, II and III, while the inactivation gate is controlled by the S4 segment in domain 4. 7 The pore region The pore, the part of the channel through which ions flow from the extracellular to the intracellular fluids, is formed by S56. The amino acids linking the two segments actually dip into the pore and form the selectivity filter, the structure that enables the channel to selectively pass Na+ ions and excludes all other ions. The detailed arrangement of the pore and the way the selectivity filter operates has been worked out for K+ channels and is considered in a later section. Figure 7. Differential movements of S4 segments in the four domains of the Na+ channel produce activation and inactivation. The structure of potassium channels The first potassium channel to be sequenced came from the fruit fly, Drosophila melanogaster. Drosophila has been an invaluable resource for genetic studies and has made critically important contributions to our understanding of how K+ channels operate. Of special importance is the Shaker mutant. Researchers use ether as an anesthetic to count various mutants during their screens (otherwise the it just flies away). Occasionally, instead of just going to sleep, some flies shake their legs and wings when exposed to ether. This turns out to be due to mutations in specific genes. Among these strains of 8 mutant flies are some that have defects in a particular gene, a piece of DNA that has been termed the Shaker locus (because these flies shake when anesthetized). The defect (shaking) was due to the loss of a particular K+ channel (as we shall see below, there are hundreds of different types of K+ channels). Recall, that the function of K+ channels is to repolarize action potentials and thereby shorten their durations. The loss of this K+ channel caused abnormally long action potentials in some, but not all, axons of Shaker flies (but only when they were anesthetized). When the prolonged action potentials arrived at muscle fibers, they cause unusually prolonged transmitter release, prolonged muscular contractions and hence shaking. When the sequence of the Shaker K+ channel was determined, protein encoded by the gene was termed the subunit. The subunit is in some ways strikingly similar to the Na+ channels described previously, and in other ways is strikingly different. The features that are similar are that the subunit has six membrane spanning regions, S1S6, with one S4 region that closely resembled the S4 voltage sensor in the Na+ channel. What is different is that the subunit is the size of the Na+ channel but with a structure similar to one domain of a Na+ channel. Recall that all 4 domains of the Na+ channel are produced by a single gene. In contrast, genes that code for K+ channels only transcribe mRNA that produce one subunit. An individual subunit cannot by itself form a channel. Rather, the Shaker K+ channel is formed by combining four subunits into a tetramer. Indeed, it is now known that all K+ channels are formed from four subunits, where each subunit is encoded by one gene. In short, the K+ channel, like the Na+ channel, has four domains, with each domain composed of six transmembrane segments, S16. However, unlike Na+ channels where the entire channel is encoded by one gene, only one domain (subunit) of the K+ channel is produced by single gene. The entire K+ channel is composed of four separate subunits that combine to resemble the Na+ channel (Fig. 9). Like the Na+ channel, each subunit has 6 transmembrane segments, where the S4 segment is the voltage sensor, as in Na channels. Figure 8. Comparison of subunit composition in Na+ compared to K+ channels. 9 Figure 9. Top view showing how subunits tetramerize to form a functional K+ channel. The 4 subunits that compose the channel correspond to the four domains in Na+ channels. Potassium channel diversity There are a huge variety of voltage gated K+ channels, far more than Na+ channels and far more than were ever suspected. Each type has its unique features. Some potassium channels, such as the Shaker K+ channel, have fast activation and have inactivation gates, similar to those on Na+ channels. Other K+ channels are slower (activate slower) and do not have inactivation gates. The K+ channel that we discussed in previous chapters that acts to shape the action potential in the squid giant axon, as one example, does not have an inactivation gate and is slower than the Shaker K+ channel. Some of the diversity is a consequence of different genes that encode different K+ channels. The K+ channels in the squid axon, for example, are produced by a different gene than the Shaker K+ channel. But there are also two other factors that account for K+ channel diversity. One factor is that there is substantial alternative splicing of the RNAs transcribed by potassium channel genes. The DNA of all genes contain expressed sequences called exons, which code for the protein, and noncoding sequences called introns. Because RNA is synthesized from this DNA, the RNA, called premRNA, also contains the exons as well as noncoding introns. During the production of the mature mRNA that encodes the protein, the introns are cut out and the exons are spliced together to produce a competent mRNA that is translated in the cytoplasm. If the splicing results in the incorporation of different combinations of exons, then different mRNAs will be produced, and different proteins will be expressed. The idea is that a K+ channel gene, in this case the gene for Shaker, can produce several different subunits through alternative splicing. 10 Figure 10. Schematic showing how alternative splicing of the RNA transcribed from a single gene can produce a variety of different proteins. In A, a gene is transcribed to produce an RNA with 4 exons and 3 introns. Following transcription, the introns are cut out and the all four exons are spliced together. The mRNA composed of the 4 exons is then translated into a polypeptide. In B, the same gene is transcribed but two different mRNAs are formed due to alternative splicing. In the top, exons 1, 3 and 4 are spliced together and that mRNA is translated into polypeptide 1, having an amino acid sequence different from the mRNA in A, which contained all 4 exons. Yet a different splice variant, and hence a different polypeptide, is formed by splicing exons 1,2 and 4 together. A second factor generating K+ channel diversity is that the K+ channel can be constructed from four of the same subunit type, called a homomeric channel, or the channel can be formed by combing two or more different types of subunits. Channels composed of different subunit types are called heteromeric channels. Each combination of subunit types produces a K+ channel with unique functional properties. Figure 11. Homomeric K+ channels are composed of 4 idential subunits. Heteromeric K+ channels are formed by different types of subunits. The different subunit types are due to alternative splicing of the K+ gene. Channels composed of different subunits have different electrical properties. 11 The pore and the selectivity filter The operation of the pore and how the selectivity filter in the pore endows the channel with its ability to pass only one type of ion, K+ in this case, was shown by Rod MacKinnon and his colleagues. This was a stunning achievement and MacKinnon was awarded the Nobel Prize in Chemistry in 2003. They did this by actually seeing the movement of K+ ions in the channel with Xray crystallography. To obtain enough channels to form a crystal, they used a bacterial K+ channel called KcsA from the bacterium, Steptomyces lividans. It is important to note that the KcsA channel has a simple topology with only two membrane spanning segments per subunit. The two segments correspond to the Shaker K+ channel without S1 through S4 (the KcsA channel has no voltage sensor and is gated by acidity rather than by voltage). Four subunits, each with only two membrane spanning segments, comprise the channel (Fig. 12). Each subunit, in turn, has two protein helices, an outer helix that faces the lipid membrane and an inner helix that faces the central pore (Fig. 12). The inner and outer helices are separated by a series of amino acids that form the pore helix and the selectivity filter. Figure 12. Each subunit of the KcsA K+ channel is composed an outer and inner helix connected by a group of amino acids that form the pore region. The pore region has two major components: 1) the pore helix; and 2) the selectivity filter (not shown). Lower left panel: Top view of a K+ channel composed of four subunits, where each subunit is shown as a different color. The inner helices of the blue and red subunits are indicated by arrows while the outer helices of the green and yellow subunits are indicated by arrows. Lower right: The inner helices of four subunits are arranged as an inverted teepee. The four pore helices are shown in white. The sequence of amino acids in the selectivity filter, the sequence important for K+ selectivity, is the same in the pores of all K+ potassium channels. The entire sequence is composed of four pore loops inside the pore, one loop from each of the four subunits. MacKinnon called these amino acids the K+ signature sequence (Fig. 13). Mackinnon was profoundly struck by this universal feature of K+ channels and in his 2003 Nobel Address he wrote, "The conservation of the signature sequence amino acids in K+ channels throughout the tree of life, from bacteria to higher eukaryotic cells, implies that nature had settled upon a very special solution to achieve, rapid, selective K+ conduction across cell membranes." 12 An important feature of charged ions in solution, which has not been mentioned previously, is that water molecules surround each ion. We refer to such ions as being hydrated. The main idea of the selective filter is that it strips the water molecules from the hydrated K+ ion and conducts dehydrated K+ through the filter. The dehydration is a critical step and why it is relevant for the operation of the selectivity filter will be explained below. The structure of the selectivity filter is complex, but the general idea is that its molecular structure has the shape of an oxygenlined tunnel with four K+ binding sites (Fig. 13A and B2). The oxygen linings at the binding sites are formed by negatively charged carbonyl groups at four sites, sites 14, of the filter (a carbonyl group in organic compounds is a carbon atom doubly bonded to an oxygen atom). At each site, negative carbonyl groups protrude into the tunnel, one carbonyl from each of the four subunits (Fig. 13B4). The arrangement is such that a dehydrated K+ fits precisely into a site, embraced by the negativity at each site. Thus, each site in the tunnel interior mimics the embrace of the water molecules that surround the ion in solution. Importantly, a hydrated K+ or a hydrated Na+ ion does not fit into the sites because they are too big. The precise match in the size of the dehydrated K+ ion and the size of the binding sites is the key factor that imparts selectivity to the filter. Figure 13. Detailed view of KcsA K+ channel features. A: Two subunits show major portions of channel including the outer and inner helices, the central cavity the pore helix (red) and the selectivity filter. The inner helix lines the interior of the central cavity and forms part of the activation gate. The selectivity filter is formed by a sequence of amino acids that link the lower portion of the pore helix with the upper portion of the inner helix. The four small black lines protruding into the selectivity filter are binding sites for dehydrated K+ ions and are composed of the signature sequence of amino acids that characterize the pores of all K+ channels. B1: Inverted teepee arrangement of helices in selectivity filter. The selectivity filter is shown from a side view. Two subunits of the channel are shown in side view. Four K+ ions (green balls) are in the selectivity filter and a single K+ ion is in the central cavity. B2: Expanded side view of selectivity filter 13 with four K+ ions (green balls) located at each of the four K+ binding sites. B3: Top view looking into the filter. A K+ ion is in one binding site and is stabilized by the four subunits. B4: cartoon drawing of a hydrated K+ ion as it would occur in solution. The water is stripped as the ion enters the selectivity filter and is stabilized in one binding site by four negatively charged carbonyl groups, one from each subunit. The negatively charged oxygens in each of the carbonyl groups in the filter replicate exactly the negatively charged oxygen groups to which the K+ ion is bound in solution. In its basic form, the channel has three important parts, as illustrated in Fig. 13A. The first is the gate, located on the intracellular part of the channel. The second is a relatively wide central cavity located just above the gate. The third is the selectivity filter, located above the central cavity. The end of the selectivity filter faces the extracellular fluids. The selectivity filter can exist in one of two configurations. In the open configuration the filter conducts ions while in the closed configuration it is "pinched" so that it does not conduct ions (Fig. 14). Whether the filter is in the open or closed configurations is determined by whether the gate is open or closed, for reasons explained below. Lets follow the exact sequence of events to see how the selectivity filter works by first considering the conditions that occur when the gate in closed and then when it is open. Events when the gate is closed When the gate is closed, one K+ ion is located in the central cavity (Fig. 14). That K+ ion is surrounded by water molecules and is stabilized in the center because the lining of the central cavity is hydrophobic (water hating) and the four negatively charged ends of the helix that faces the cavity act as four, equidistant weak attractants that sort of hold the K+ molecule in its place (Fig. 13B3 and B4). There are two K+ ions in the selectivity filter, each separated by a water molecule, and oscillating between positions 13 and 24 of the filter. The K+ ion in the central cavity cannot enter the selectivity filter because of K+ in the selectivity filter repel it through electrostatic forces. K+ ions cannot move out of the filter because it is in the closed position. Thus, there is no flow of ions through the filter when the gate is closed. Figure 14. K+ ion channel in closed state. Note that the selectivity filter is pinched at the top thereby closing the filter to the extracellular fluids. One hydrated K+ ion is in the central cavity, stabilized by the four negative carbonyl charges from each of the four pore helices (only two pore helicies are shown). See text for further explanation. 14 Events when the gate is open Next, let us see what happens when the gate opens (Fig.15). Now the high intracellular concentration of K+ forces additional K+ into the central cavity. The repulsive force of the high K+ concentration in the cavity forces the K+ in the center of the cavity towards the selectivity filter. The negatively charged helices now strip the water molecules from the K+ ion allowing it to fit perfectly into site 4 of the filter (Fig. 15, panels 2 and 3). This is followed by a second K+ ion from the cavity into the filter. Each time a K+ is forced into the filter, it occupies one of the filter sites, and the full occupancy of the sites, with four dehydrated K+, changes the filter conformation to the open state. Now the electrostatic forces of the K+ ions in the filter force the K+ ion in site 4 out of the channel into the extracellular space (Fig. 15, panel 3). This vacant site is then filled by the K+ ion that was in site 3, which is turn is filled by the K+ ion that was in site 2, and so on. The K+ ion in site 1 is constantly replenished by the entering K+ ions that are repelled from the central cavity by the high K+ concentration in intracellular fluid that is "seen" or sensed in the central cavity (Fig. 15, panel 4). In this way, there is a "bucket brigade" type of movement of K+ ions through the selectivity filter into the extracellular space. When the gate closes, the central cavity no longer senses a high K+ concentration, and thus there is no longer a sufficient force on the lone K+ in the cavity to force it into the filter. As soon as the last K+ ion is forced from the filter into the extracellular space, leaving two K+ ions in the filter, the filter conformation changes, thereby closing the filter and preventing any of further flux of K+ ions out of the filter. In this condition, ionic current through the filter stops. Figure 15. K+ ion channel in the open state; both the activation gate (bottom) and the selectivity filter are open. Panels 14 show how the hydrated K+ in the central cavity is dehydrated and enters site 4 of the selectivity filter, thereby exerting electrostatic repulsive forces on the K+ ions in the sites above it, and forcing the K+ in site 1 into the extracellular space where it is then hydrated. This sequence is then repeated as a new K+ enters the filter. The K+ ions in each binding site in panel 1 are shaded so that each can be followed as each is successively driven upward and ultimately out of the filter into the extracellular space. 15 ! Why can't Na+ ions flow through the filter? The hallmark of the selectivity filter is that it passes only one type of ion, K+ in this case, and excludes all other ions. As described previously, the reason that K+ passes easily through the filter is because a dehydrated K+ fits precisely into a site, embraced by the negativity at each site, which mimics the embrace of the water molecules that surround the ion in solution (Fig.16). As far as the K+ ion is concerned, being hydrated or being in the filter are energetically equal, and thus the two states are in equilibrium. Thus as water is stripped from the K+ ion, it just slides into a binding site and can move through the binding site as readily as it diffuses in solution. But dehydrated Na+ ions are smaller than K+ ions, so why can't the smaller Na+ ions pass through the selectivity filter? The answer is that Na+ ions do not fit well in the sites of the selectivity filter that stabilize K+ ions, and thus the hydrated Na+ ion is far more stable than a dehydrated Na+ that entered the filter. The probability of the two states greatly favors hydration, as shown by the arrows in the right panel of Fig. 16. Let us explore this in a bit more detail. Imagine that the selectivity filer is open, and thus the external face of the selectivity filter is open to the extracellular fluid where there is a high concentration of Na+ that should drive Na+ into and through the filter. However, the dimensions of the segments fit K+ ions stripped of their surrounding water perfectly. Hydrated Na+ ions are too large to enter the pore. Dehydrated Na+ ions, although smaller than dehydrated K+ ions, fit poorly into the filter, and thus are far more likely to flip back to a hydrated state. The precise fit of K+ ions in the filter sites is the key feature that endows the K+ channel with its ion selectivity. H H H H H H Figure 16. Hydrated K+ and K+ ions are H H shown in the top panels. The dehydrated ions are shown in the bottom panel. Notice that the dehydrated K+ ion is larger than the dehydrated Na+ ion, but that the dehydrated Na+ ion fits poorly in the binding sites of the selectivity filter, while the K+ ion fits perfectly. H H H H H H H H 16 ...
View Full Document
- Spring '08