Neuron,
Vol. 2, 129-140,
January,
1990, Copyright
0
1990 by Cell Pres,
The Drosophila Shaker Gene Codes for a Distinctive K+ Current in a Subset of Neurons Keith Baker and Lawrence Salkoff Department of Anatomy and Neurobiology, Washington University School of Medicine St Louis, Missouri 63110
Summary A transient identified by injecting
K+ current coded by the Shaker gene was in muscle and expressed in Xenopus oocytes cRNA transcribed from a cloned cDNA. The
Shaker current has not previously been identified in neurons. Mutational analysis now reveals that in neurons, Shaker is required for expression of a very rapidly inactivating K+ current with a depolarized steady-state inactivation curve. Together, these properties distinguish the Shaker-coded current from similar fast transient K+ currents coded by other genes. The Sh5 mutation further enhanced the depolarization of the Shaker current steady-state inactivation curve. Deletion of the Shaker gene completely removes the transient K+ current from a small percentage of neurons (15%) in a mixed population, and removes a portion of the whole-cell current in about 35% of neurons. The remaining 50% of neurons were apparently unaffected by deletion of the Shaker gene. The unique combination of rapid inactivation and depolarized steady-state inactivation of the Shaker current may reflect a unique functional role for this curent in the nervous system such as the rapid repolarization of action potentials. Introduction Diverse types of K+ channels provide a major mechanism for the generation of complex electrical properties in electrically excitable tissues (Rudy, 1988; Hille, 1984). The molecular mechanisms that produce the diversity of K+ channels, however, is only now becoming known. The Shakergene in Drosophila codes for a transient K+ channel and is expressed in both muscle (Salkoff, 1983; Wu and Haugland, 1985; Sole et al., 1987) and nerve (Jan et al., 1977; Tanouye et al., 1981; Pongs et al., 1988; Barbas et al., 1989). The current produced by the Shaker protein has been extensively studied in Drosophila muscle and in the Xenopus oocyte expression system, in which the transcription products of cDNAs coding for the Shaker proteins have been shown to express transient K+ currents (Timpe et al., 1988b; Iverson et al., 1988). However, the nature of the current in neurons has not yet been reported. The transient K+ currents in both embryonic (Byerly and Leung, 1988) and larval (Sole et al., 1987) neuronal cell bodies are not apparently affected by mutations in the Shaker gene that otherwise eliminate the transient KC current from muscle. 1r-1 contrast, our results indicate that the Shaker protein is present in a subset of cell bodies
in the developing adult nervous system and has distinctive characteristics that allow it to be distinguished from transient K+ currents coded by at least one and perhaps several other genes. Cloning of the Drosophila Shaker locus revealed the primary structure of a voltage-gated K+ channel (Papazian et al., 1987; Temple et al., 1987; Baumann et al., 1987; Kamb et al., 1987). The molecular analysis (Schwarz et al. 1988; Kamb et al., 1988; Pongs et al., 1988) also revealed differential splicing as a mechanism for producing a diversity of gene products from the Shaker locus. Although the alternatively spliced mRNAs produced at the Shaker locus express currents that differ kinetically when expressed in Xenopus oocytes (Timpe et al., 1988a, 1988b; lverson et al., 1988), the biological relevance of these observations as they relate to differential expression of Shaker in different tissues and at different times in development is unknown. It is also unclear how differential splicing relates to possible kinetic diversity of K+ currents. Differential splicing has been suggested as a mechanism for producing diverse K+ currents (Agnew, 1988; Timpe et al., 1988b; lverson et al. 1988). We sought to explore the question of in vivo diversity suggested by molecular studies. Our results indicate that, at least in neuronal cell bodies, the Shakergene codes for a transient K+ current with fairly uniform properties. In contrast, a major contributor to diversity in transient K+ currents appears to be from a gene (or genes) other than Shaker. Three genes that have homology to the Shakergene and possibly code for non-Shaker K+ currents have now been cloned and sequenced from Drosophila (Butler et al., 1989). These genes have been shown to code for voltage-gated K+ currents that activate and inactivate more slowly than the Shaker-coded currents (Wei et al., submitted). Results Southern Blot and Physiological Analyses Demonstrate the Absence of the Shaker Gene in the Deletion Genotype, ShA deletion of the X chromosome between the X-Y translocation breakpoints B55 and W32 removes the portions of the Shaker gene that code for the regions of the protein presumed to span the membrane (SlS6) and form the channel (Figure IA; Pongs et al., 1988). This deletion was confirmed by Southern blot analysis (data not shown). The male genotype, D~(?,SSSD/ W32p, carrying this deletion was used in our study and is designated as Sh-. Hemizygous male animals carrying this deletion are viable. The stock was balanced by C(l)DX Fw,f Sh+. These attached-X females carry the normal Shaker gene and are designated as Sh+. These female flies were used as controls because they share common autosomes with the males. Ore-
NeWXl 130
A
Df (1) W32P/B55D v
*centromere
140
120
,
-
0 80
100
I
60
1 20 kb
40
4 I Ic excms for membrane spanning
segments
Sl through
S6
Sh+
Sh-
I Figure Shaker
1. Molecular
Map
and
Confirmation
1OOmV of the
Deletion
of
(A) Molecular map showing the proximal and distal molecular limits of the deletion between X chromosome breakpoints 855 and W32. Exons coding for membrane-spanning segments Sl-S6 are removed by this deletion. A description of the genotypes of flies used in this study is given in the text. (B) Dorsal longitudinal flight muscles were assayed for the presence of voltage-gated transient K’ currents in Sh+ and Sh- flies, The control (Sh') shows a transient K+ current, whereas Sh- is missing the transient K+ current. The holding potential was -80 mV. Test pulses were applied in 10 mV increments up to +20 mV.
gon-R wild-type flies were used as a second control; the membrane physiology of this wild-type strain was indistinguishable from C($DX. Figure IB shows the physiological effect, in muscle, of deleting the Shaker gene; Sh- flies have no voltage-gated transient KC current in the pupal dorsal longitudinal flight muscles. In contrast, all stocks carrying Sh+ do have a transient K+ current in the pupal flight muscles. Since both Southern blot and physiological analyses confirm that the Shaker gene is absent from the Sh- flies, transient K+ currents recorded from Sh- flies in this study must be coded by a gene (or genes) other than Shaker.
Ionic Conditions Select for K+ Currents Our recordings were made using a new preparation that consisted of acutely dissociated cells from the thoracic ganglia of very late stage pupae (approximately 12 hr prior to adult eclosion). The cell population contains both motoneurons and interneurons. No attempts were made to distinguish the type of cell being recorded from. K+ currents were isolated by adding tetrodotoxin (100 nM) and CdC12 (200 PM) to the bath solution to block Na+ and Cal+ currents, respectively. Previous studies using embryonic Drosophila neurons have demonstrated that Na+ currents are blocked by tetrodotoxin (O’Dowd and Aldrich, 1988) and Cal+ currents are blocked by CdCIL (Byerly and Leung, 1988). Reversal potentials of the remaining currents followed the bath K+ concentration (Figures 2A and 2B), as expected for K+ selective channels. As a further demonstration that K+ selective channels were passing the majority of the wholecell current, Cs+, a known K+ channel blocker, was substituted for K+ in the patch electrode. A typical recording obtained with K+ inside the cell is shown in Figure 2C, whereas Cs+ inside the cell abolishes the current (Figure 2D). None of the cells sampled with Cs+ inside (n = 28) showed evidence of a transient outward current, thus confirming KC as the main current-carrying ion. Peak Current Density Is Lower after Shaker Is Deleted Peak current was measured for the unidentified population of cells isolated from the thoracic ganglia of late stage pupae. Three genotypes were studied, SW, W, and Sh-. Since cell capacitance is a measure of membrane area, the normalization of the peak current to cell capacitance provided a measure of current density. As shown in Figure 3A, peak current density was 174 pA/pF for Sh+ cells (n = IOI), 183 pA/pF for Sh5 cells (n = 91), and 136 pA/pF for Sh- (n = 113) cells. Peak current density for Sh- was significantly less than that for either SW or ShS (P < 0.05; Student’s t-test). The peak current density was not significantly different between Sh+ and Sh5. Thus, in Sh- cells, in which Shaker is deleted, the peak current density is reduced 22% or 26% when compared with Sh+ or Sh5 flies, respectively. This indicates that Sh- cells are missing a component of the peak whole-cell K+ current. These findings were further investigated by measuring the steady-state current, which should be relatively unaffected by the presence or absence of the Shaker gene and is thus an internal control. Shaker codes for transient K+ channels that inactivate given a sustained depolarization. Once the channels inactivate, they no longer pass current and their presence can no longer be detected using a voltage clamp. Thus, Shaker should affect only peak current and not sustained current. This expectation is fulfilled, since the current density measured after 80 ms of depolarization is not affected by deleting the Shaker gene. As shown in Figure 3A, the current density measured 80
Maker 131
K+ (:urrent\
I” Neuron‘
012345676
9
10
Time (ms*c)
Command
Voltage
(mV)
D -1
Tlma (msec)
Figure
2. Currents
Are Carried
Tlms
(mmc)
by K’ Ions
(A) Current records obtained by giving a 2 ms depolarization to +50 mV followed by repolarization to various potentrals ranging from -100 to 0 mV in 10 mV increments (28 mM K+ in the bath). (6) Current amplitude immediately following the repolarization is plotted as a function of the potential to which the cell was repolarized. The experiment is repeated for three different K+ bath concentrations (2 mM, filled squares; 28 mM, open circles; 140 mM, filled diamonds). The solid line drawn through the data is a least-squares fit. The intersection of the fitted line with the abscissa indicates that the current reverses direction at -30 mV with 28 mM K+ in the bath. Perfect K+ selective channels would have reversed at -40 mV with 28 mM K+ in the bath. When the K+ concentration in the bath is lowered to 2 mM (different cell), the reversal potential moves rn the hyperpolarized direction (solid squares), and when the K+ concentration in the bath is rarsed to 140 mM (different cell), the reversal potential moves in the depolarized direction (solid triangles). Thus, as expected for K+ currents, the reversal potential follows the K+ concentration in the bath. A histogram of reversal potentials obtained from different cells with 28 mM K+ in the bath is shown in the lower part of (B) (n = 106). Only cells with greater than 100 pA of peak current and greater than 70% inactivation were considered. Currents used to determine reversal potentials were obtained within 3 min of establishing the whole-cell recording configuration. The means from Sh+, Shi, and Sh- cells were not significantly different and so were grouped together with a mean of -30 mV (SD 3.6 mV, n =
106).
(C) Transient outward current when 140 mM K+ is inside the cell and 28 mM K+ is in the bath. (D) Representative current traces obtained when 140 mM Cs+, a known K+ channel blocker, is inside the cell (same in [Cl). The cell was held at -90 mV, and a series of voltage steps ranging from +50 to -50 mV elicited no iransient
ms after the voltage jump was 39 pA/pF for pA/pF for SW, and 37 pA/pF for Sh-. These densities are not significantly different from other.
Sh+, 43 current one an-
A Small Percentage of Cells Are Completely Missing a Transient K+ Current after Shaker Is Deleted Transient K+ currents were found in almost all Sh+ cells examined (9&T%, n = 142 ). In contrast, in Shflies approximately 15% of cells were missing a transient K+ current. To determine the relative amount of transient K+
current
scale
as
current.
current in each cell, the current amplitudes were measured at the peak (total current) and at 80 ms (steady-state current), and the amount of inactivation was expressed as a percentage of the peak current. Using this operational definition, a current that shows no inactivation is assigned 0% inactivation; a current that inactivates completely is assigned 100% inactivation. Representative transient K+ currents from 3 different Sh- cells are shown in Figure 36 along with the percent inactivation derived using this operational definition. Figure 3C shows a histogram of the amount of tran-
NetIrOn
132
U Wild Type (.Sh*)
i g
Deletion
Mutant
(Sh-)
II; Wild Type (Sh’)
0.30
q n
iI!
gb
0.25
Mutant (Sh5) Deletion Mutant
(Sh‘)
z
;
0.20
z 0
0.15
z
0.10
,g
0.05
2 LL
Peak
80
msec
0
0
10
20
30
40
% Inactivation Figure
0 % Inactivation
L
92
% Inactivation
t-
+50
3. Pedk Current
Density
50
60
of Peak
Is Reduced
70
80
90
100
Current by 23% in Sh
Cell
Vol. 2, 129-140,
January,
1990, Copyright
0
1990 by Cell Pres,
The Drosophila Shaker Gene Codes for a Distinctive K+ Current in a Subset of Neurons Keith Baker and Lawrence Salkoff Department of Anatomy and Neurobiology, Washington University School of Medicine St Louis, Missouri 63110
Summary A transient identified by injecting
K+ current coded by the Shaker gene was in muscle and expressed in Xenopus oocytes cRNA transcribed from a cloned cDNA. The
Shaker current has not previously been identified in neurons. Mutational analysis now reveals that in neurons, Shaker is required for expression of a very rapidly inactivating K+ current with a depolarized steady-state inactivation curve. Together, these properties distinguish the Shaker-coded current from similar fast transient K+ currents coded by other genes. The Sh5 mutation further enhanced the depolarization of the Shaker current steady-state inactivation curve. Deletion of the Shaker gene completely removes the transient K+ current from a small percentage of neurons (15%) in a mixed population, and removes a portion of the whole-cell current in about 35% of neurons. The remaining 50% of neurons were apparently unaffected by deletion of the Shaker gene. The unique combination of rapid inactivation and depolarized steady-state inactivation of the Shaker current may reflect a unique functional role for this curent in the nervous system such as the rapid repolarization of action potentials. Introduction Diverse types of K+ channels provide a major mechanism for the generation of complex electrical properties in electrically excitable tissues (Rudy, 1988; Hille, 1984). The molecular mechanisms that produce the diversity of K+ channels, however, is only now becoming known. The Shakergene in Drosophila codes for a transient K+ channel and is expressed in both muscle (Salkoff, 1983; Wu and Haugland, 1985; Sole et al., 1987) and nerve (Jan et al., 1977; Tanouye et al., 1981; Pongs et al., 1988; Barbas et al., 1989). The current produced by the Shaker protein has been extensively studied in Drosophila muscle and in the Xenopus oocyte expression system, in which the transcription products of cDNAs coding for the Shaker proteins have been shown to express transient K+ currents (Timpe et al., 1988b; Iverson et al., 1988). However, the nature of the current in neurons has not yet been reported. The transient K+ currents in both embryonic (Byerly and Leung, 1988) and larval (Sole et al., 1987) neuronal cell bodies are not apparently affected by mutations in the Shaker gene that otherwise eliminate the transient KC current from muscle. 1r-1 contrast, our results indicate that the Shaker protein is present in a subset of cell bodies
in the developing adult nervous system and has distinctive characteristics that allow it to be distinguished from transient K+ currents coded by at least one and perhaps several other genes. Cloning of the Drosophila Shaker locus revealed the primary structure of a voltage-gated K+ channel (Papazian et al., 1987; Temple et al., 1987; Baumann et al., 1987; Kamb et al., 1987). The molecular analysis (Schwarz et al. 1988; Kamb et al., 1988; Pongs et al., 1988) also revealed differential splicing as a mechanism for producing a diversity of gene products from the Shaker locus. Although the alternatively spliced mRNAs produced at the Shaker locus express currents that differ kinetically when expressed in Xenopus oocytes (Timpe et al., 1988a, 1988b; lverson et al., 1988), the biological relevance of these observations as they relate to differential expression of Shaker in different tissues and at different times in development is unknown. It is also unclear how differential splicing relates to possible kinetic diversity of K+ currents. Differential splicing has been suggested as a mechanism for producing diverse K+ currents (Agnew, 1988; Timpe et al., 1988b; lverson et al. 1988). We sought to explore the question of in vivo diversity suggested by molecular studies. Our results indicate that, at least in neuronal cell bodies, the Shakergene codes for a transient K+ current with fairly uniform properties. In contrast, a major contributor to diversity in transient K+ currents appears to be from a gene (or genes) other than Shaker. Three genes that have homology to the Shakergene and possibly code for non-Shaker K+ currents have now been cloned and sequenced from Drosophila (Butler et al., 1989). These genes have been shown to code for voltage-gated K+ currents that activate and inactivate more slowly than the Shaker-coded currents (Wei et al., submitted). Results Southern Blot and Physiological Analyses Demonstrate the Absence of the Shaker Gene in the Deletion Genotype, ShA deletion of the X chromosome between the X-Y translocation breakpoints B55 and W32 removes the portions of the Shaker gene that code for the regions of the protein presumed to span the membrane (SlS6) and form the channel (Figure IA; Pongs et al., 1988). This deletion was confirmed by Southern blot analysis (data not shown). The male genotype, D~(?,SSSD/ W32p, carrying this deletion was used in our study and is designated as Sh-. Hemizygous male animals carrying this deletion are viable. The stock was balanced by C(l)DX Fw,f Sh+. These attached-X females carry the normal Shaker gene and are designated as Sh+. These female flies were used as controls because they share common autosomes with the males. Ore-
NeWXl 130
A
Df (1) W32P/B55D v
*centromere
140
120
,
-
0 80
100
I
60
1 20 kb
40
4 I Ic excms for membrane spanning
segments
Sl through
S6
Sh+
Sh-
I Figure Shaker
1. Molecular
Map
and
Confirmation
1OOmV of the
Deletion
of
(A) Molecular map showing the proximal and distal molecular limits of the deletion between X chromosome breakpoints 855 and W32. Exons coding for membrane-spanning segments Sl-S6 are removed by this deletion. A description of the genotypes of flies used in this study is given in the text. (B) Dorsal longitudinal flight muscles were assayed for the presence of voltage-gated transient K’ currents in Sh+ and Sh- flies, The control (Sh') shows a transient K+ current, whereas Sh- is missing the transient K+ current. The holding potential was -80 mV. Test pulses were applied in 10 mV increments up to +20 mV.
gon-R wild-type flies were used as a second control; the membrane physiology of this wild-type strain was indistinguishable from C($DX. Figure IB shows the physiological effect, in muscle, of deleting the Shaker gene; Sh- flies have no voltage-gated transient KC current in the pupal dorsal longitudinal flight muscles. In contrast, all stocks carrying Sh+ do have a transient K+ current in the pupal flight muscles. Since both Southern blot and physiological analyses confirm that the Shaker gene is absent from the Sh- flies, transient K+ currents recorded from Sh- flies in this study must be coded by a gene (or genes) other than Shaker.
Ionic Conditions Select for K+ Currents Our recordings were made using a new preparation that consisted of acutely dissociated cells from the thoracic ganglia of very late stage pupae (approximately 12 hr prior to adult eclosion). The cell population contains both motoneurons and interneurons. No attempts were made to distinguish the type of cell being recorded from. K+ currents were isolated by adding tetrodotoxin (100 nM) and CdC12 (200 PM) to the bath solution to block Na+ and Cal+ currents, respectively. Previous studies using embryonic Drosophila neurons have demonstrated that Na+ currents are blocked by tetrodotoxin (O’Dowd and Aldrich, 1988) and Cal+ currents are blocked by CdCIL (Byerly and Leung, 1988). Reversal potentials of the remaining currents followed the bath K+ concentration (Figures 2A and 2B), as expected for K+ selective channels. As a further demonstration that K+ selective channels were passing the majority of the wholecell current, Cs+, a known K+ channel blocker, was substituted for K+ in the patch electrode. A typical recording obtained with K+ inside the cell is shown in Figure 2C, whereas Cs+ inside the cell abolishes the current (Figure 2D). None of the cells sampled with Cs+ inside (n = 28) showed evidence of a transient outward current, thus confirming KC as the main current-carrying ion. Peak Current Density Is Lower after Shaker Is Deleted Peak current was measured for the unidentified population of cells isolated from the thoracic ganglia of late stage pupae. Three genotypes were studied, SW, W, and Sh-. Since cell capacitance is a measure of membrane area, the normalization of the peak current to cell capacitance provided a measure of current density. As shown in Figure 3A, peak current density was 174 pA/pF for Sh+ cells (n = IOI), 183 pA/pF for Sh5 cells (n = 91), and 136 pA/pF for Sh- (n = 113) cells. Peak current density for Sh- was significantly less than that for either SW or ShS (P < 0.05; Student’s t-test). The peak current density was not significantly different between Sh+ and Sh5. Thus, in Sh- cells, in which Shaker is deleted, the peak current density is reduced 22% or 26% when compared with Sh+ or Sh5 flies, respectively. This indicates that Sh- cells are missing a component of the peak whole-cell K+ current. These findings were further investigated by measuring the steady-state current, which should be relatively unaffected by the presence or absence of the Shaker gene and is thus an internal control. Shaker codes for transient K+ channels that inactivate given a sustained depolarization. Once the channels inactivate, they no longer pass current and their presence can no longer be detected using a voltage clamp. Thus, Shaker should affect only peak current and not sustained current. This expectation is fulfilled, since the current density measured after 80 ms of depolarization is not affected by deleting the Shaker gene. As shown in Figure 3A, the current density measured 80
Maker 131
K+ (:urrent\
I” Neuron‘
012345676
9
10
Time (ms*c)
Command
Voltage
(mV)
D -1
Tlma (msec)
Figure
2. Currents
Are Carried
Tlms
(mmc)
by K’ Ions
(A) Current records obtained by giving a 2 ms depolarization to +50 mV followed by repolarization to various potentrals ranging from -100 to 0 mV in 10 mV increments (28 mM K+ in the bath). (6) Current amplitude immediately following the repolarization is plotted as a function of the potential to which the cell was repolarized. The experiment is repeated for three different K+ bath concentrations (2 mM, filled squares; 28 mM, open circles; 140 mM, filled diamonds). The solid line drawn through the data is a least-squares fit. The intersection of the fitted line with the abscissa indicates that the current reverses direction at -30 mV with 28 mM K+ in the bath. Perfect K+ selective channels would have reversed at -40 mV with 28 mM K+ in the bath. When the K+ concentration in the bath is lowered to 2 mM (different cell), the reversal potential moves rn the hyperpolarized direction (solid squares), and when the K+ concentration in the bath is rarsed to 140 mM (different cell), the reversal potential moves in the depolarized direction (solid triangles). Thus, as expected for K+ currents, the reversal potential follows the K+ concentration in the bath. A histogram of reversal potentials obtained from different cells with 28 mM K+ in the bath is shown in the lower part of (B) (n = 106). Only cells with greater than 100 pA of peak current and greater than 70% inactivation were considered. Currents used to determine reversal potentials were obtained within 3 min of establishing the whole-cell recording configuration. The means from Sh+, Shi, and Sh- cells were not significantly different and so were grouped together with a mean of -30 mV (SD 3.6 mV, n =
106).
(C) Transient outward current when 140 mM K+ is inside the cell and 28 mM K+ is in the bath. (D) Representative current traces obtained when 140 mM Cs+, a known K+ channel blocker, is inside the cell (same in [Cl). The cell was held at -90 mV, and a series of voltage steps ranging from +50 to -50 mV elicited no iransient
ms after the voltage jump was 39 pA/pF for pA/pF for SW, and 37 pA/pF for Sh-. These densities are not significantly different from other.
Sh+, 43 current one an-
A Small Percentage of Cells Are Completely Missing a Transient K+ Current after Shaker Is Deleted Transient K+ currents were found in almost all Sh+ cells examined (9&T%, n = 142 ). In contrast, in Shflies approximately 15% of cells were missing a transient K+ current. To determine the relative amount of transient K+
current
scale
as
current.
current in each cell, the current amplitudes were measured at the peak (total current) and at 80 ms (steady-state current), and the amount of inactivation was expressed as a percentage of the peak current. Using this operational definition, a current that shows no inactivation is assigned 0% inactivation; a current that inactivates completely is assigned 100% inactivation. Representative transient K+ currents from 3 different Sh- cells are shown in Figure 36 along with the percent inactivation derived using this operational definition. Figure 3C shows a histogram of the amount of tran-
NetIrOn
132
U Wild Type (.Sh*)
i g
Deletion
Mutant
(Sh-)
II; Wild Type (Sh’)
0.30
q n
iI!
gb
0.25
Mutant (Sh5) Deletion Mutant
(Sh‘)
z
;
0.20
z 0
0.15
z
0.10
,g
0.05
2 LL
Peak
80
msec
0
0
10
20
30
40
% Inactivation Figure
0 % Inactivation
L
92
% Inactivation
t-
+50
3. Pedk Current
Density
50
60
of Peak
Is Reduced
70
80
90
100
Current by 23% in Sh
Cell
