Abstract
Tityustoxin block of wild-type and mutant SgkV1A channels from
Loligo opalesceus neurons was studied in Xenopus oocytes using the
two electrode voltage clamp method. The mutant channel was
created by replacing a histidine residue at position 351 on the outer
vestibule with a glycine. Wild-type channels show greatly reduced
sensitivity to tityustoxin block at pH 6. The glycine mutant's
sensitivity to tityustoxin was not pH-dependent. Block of both the
wild type and mutant channels proved to be voltage dependent. Kp
values ranged from 14 nm at 0 my to 70 nm at 100 my for both
channels. K rates of H351 channels were not effected by voltage,
whereas Ka rates of wild-type channels decreased slightly with
voltage. Kg rates increased as voltage increased in both types of
channels. Results suggests that a charged residue on the toxin
enters the channel pore when tityustoxin blocks these channels,
making the toxin sensitive to the membrane voltage.
Introduction
Tityustoxin is a small peptide of 30 amino acids found in
scorpion venom and has been shown to be a potent K“ channel blocker
(Werkman et al. 1993). Though much is known about other scorpion
toxins, both functionally and structurally, tityustoxin is relatively
new and unstudied. Understanding tityustoxin better can help in
underständing the structure of the channels they bind to.
Scorpion toxins have become important tools in understanding ion
channels because they provide a rough map of the channel. Scorpion
toxins can bind to channels at 8 or more positions. By knowing
exactly which toxin residues bind to which channel residues, and
what the structure of the simple toxin peptide is, a general map of a
channel can be made. Fixing 8 or more residues in space sets
dimensional limits on proposed structures of the channel.
A great deal of work on understanding the structure and blocking
mechanism of charybdotoxin, another scorpion toxin closely related
to tityustoxin, has already been done. The accepted mechanism for
charybdotoxin block is that the peptide binds to the outer vestibule
of the channel and a lysine at position 27 enters the pore, thus
blocking ion flow (Mackinnon and Miller, 1988). It is plausible that
tityustoxin blocks by the same mechanism due to similarities in the
peptide structure. However, there are structural differences in the
proposed binding regions of these two toxins which could lead to
different blocking mechanisms. The primary purpose of this paper is
to determine whether tityustoxin blocks by the same mechanism as
other scorpion toxins or if some other mechanisms needs to be
invoked.
A secondary goal is to show that a histidine at position 351 of
SgkviA channels is necessary for the pH dependence of tityustoxin
block of these channels. It has previously been shown that changing
the homologous site in Shaker channels (position 425) from wild
type phenylalanine to histidine causes the channel to become pH
sensitive to charybdotoxin binding. Changing pH from 7-5.5 causes
the channel to become 200 fold less sensitive to charybdotoxin
(Perez-Cornejo et. al, 1998) It has also been shown by previous
spring students, Sky Pittston (1996) and Uma Sanghvi (1997), that
tityustoxin block of the wild type SgkviA channels in squid neurons
is pH dependent. It was hypothesized that a histidine at position
351 is responsible for this pH dependence because its charge
changes over the biological pH range. DEPC treatments were used to
chemically alter the histidine residue, but did not conclusively show
that histidine is responsible for pH sensitivity to tityustoxin. This
paper will show more convincingly that this is in fact true by using
a mutant channel in which this histidine is replaced by a glycine.
Materials and Methods
SgkviA channels containing a G87R mutation where expressed in
Xenopus oocytes by cRNA injection as previously described by
Rosenthal and Bezanilla(1997). After 2-3 days currents between 5¬
30 uA could be recorded from wild type channels. Currents from
H351G mutants were a little smaller. Currents were recorded by the
two electrode whole cell voltage clamp method using p/4 leak
subtraction. Only oocytes that had a small leak (under -0.2 uA at -
60 mv holding potential) and stable K“ currents of at least 4-5 uA
were used.
Electrodes were filled with 3 M KCl, and the external bath
solution was either ND-96: 96 mM Nacl, 2 mM KCI, 1 mM MgCl,, 1.8
mM Cacl,, and 5 mM Hepes or a solution containing 90mM Nacl, 10
mM KCI, 2 mM Cacl,, 5 mM MgCl,, 10 mM Hepes at pH 7.15. For the
pH-dependence experiments, solutions of 85 mM Nacl, 10 mM KCl, 2
mM Cacl,, 5 mM MgCl,,10 mM MES at pH 6 and 90 mM Nacl, 10 mM
KCl, 2 mM Cacl,, 5 mM MgCl,, 10 mM MES at pH 8.2 were used.
loxin was diluted in the same solution used in the bath to a
concentration of 100, 50, or 10 nM. Solution changes were made by
injecting 1.5-2 ml of the new solution into the bath while a vacuum
removed the previous solution. Complete solution change could be
done in under a minute, and leak generally improved during this
period.
Currents were recorded during 50 and 500 ms pulses to -40,-20.
0, 20, 40, 60, 80, and 100 my in toxin free-solution. The toxin
solution was then washed in, and currents were again recorded for
the same duration and voltages.
Time constants were obtained by fitting a single exponential to
the slow rise of current traces in toxin. The exponential was fit
from the time at which the control current reached a maximum to
the time at which the current in toxin reached a maximum (Fig. 1).
This was done to avoid problems caused by rapid activation of the
unblocked current.
Results:
SqkviA Channels
SqkviA is a K“ channels cloned from a CDNA library prepared
from the stellate ganglia of Loligo opalescens (Rosenthal et al.
1996). It was previously shown that this channel shares many
homologous regions with Shaker B channels including the outer pore
region. A point mutation was made to the channel at position 351 by
replacing a histidine with a glycine. This position is homologous to
position 425 in ShakerB, a proposed binding site for charbdotoxin
(Goldstein and Miller, 1992). In all experiments both the wild-type
and H351G channels were used.
pH-dependence of block by tityustoxin
pH-dependence of toxin block was shown by comparing Kt current
before and after application of 50 nM tityustoxin at pH 8.2 and pH 6.
The effects of toxin on H351G channels showed no detectable pH¬
dependence. At pH 8.2, 51.4% of the current was blocked. Similarly,
at pH 6, 57% of the current was blocked. (see figure 2 A,B) However,
the wild-type channels did display pH-dependent block by toxin. 57%
of the current was blocked at pH 8.2(fig. 2 C), whereas only .05% or
less was blocked at pH 6 (fig. 2 D). Similar resusits were obtained
in one other experiment.
Voltage Dependence of Toxin Binding
The Kp of toxin binding was calculated at each voltage using the
equation:
Ko - ioxinl Toxin/(contro + toxin)
Where lg represents the current when no toxin is present after a
500 ms pulse and lgn represents the current in the presence of toxin
after a 500ms pulse. The K, of toxin binding increases as voltage
increases, as is shown in fig. 3. The increase in K, is similar in both
the wild type and H351G channels.
Change in Ke and Ke rates
The time constant (t) of current increase in the presence of toxin
was measured by fitting a single exponential to the curve. Using the
1= 1/(Kltoxin) + Kor)
equations,
and Ko- KoKon
it is possible to solve for Kgg and Kan. It was not possible to fit an
exponential to current traces from voltage steps below 40 my for
the H351G. Using wild-type channels, exponentials could not be
fitted to voltage steps below 60 mv. In H3516 channels, K...
increases from .006 at 40 my to .027 at 100 my, whereas K.,
remains relatively constant (Fig. 4). In wild-type channels, K.,
decreases slightly from .00023 at 60 my to .00018 at 100 my, and
Ky increases from .021 to .029 over the same range(Fig. 5).
Discussion
It is clear from the results of this study that a histidine at
position 351 is necessary and sufficient to confer pH sensitivity to
toxin block in SgkviA channels. This is likely due to the fact that at
low pH histidine is a positively charged residue, whereas glycine is
always neutral. However, it is also possible that steric differences
between histidine and glycine could account for the differences in
pH-sensitivity.
Results obtained in this study also seem to indicate that
tityustoxin blocks SgkviA channels by a mechanism similar to that
shown by other scorpion toxins, at least in the H351G mutant.
Changes in Kp, Kan, and Kay values are consistent with those found
using charybdotoxin. Goldstein and Miller graphed the Ky values of
charybdotoxin binding to Shaker channels by use of the same log
scale that was used in this paper (Goldstein and Miller, 1993). In
their data the K, increases by about a factor of .063-.079/my
compared to .072/my for tityustoxin acting on SgkvlA channels.
However, Goldstein and Miller only showed data from -20 to 20 my.
it is not known how the Kp of charybdotoxin changes above 20 mv.
Goldstein and Miller also showed that Kg of charybdotoxin increases
with voltage by about a factor of .079/my, while Kay remains
constant. This is similar to the results found in this study with the
H351G mutant. The apparent decrease in Ka, in the wild-type
channels is likely due to the small sample size of wild-type
experiments; this anomalous result merits further investigation.
The similarity between results presented in this paper and those
found for charybdotoxin suggest that tityustoxin is binding to
SgkviA channels by the same mechanism by which Charybdotoxin
binds to Shaker channels. In this model, the peptide binds to the
outer vestibule with lysine-27 entering the pore, inhibiting ion flow.
At more positive membrane voltages, the voltage drop across the
membrane and Kt ions trying to exit the channel force the toxin
peptide off the channel. This increases the Kp and Ky values at that
voltage.
Literature Cited:
Goldstein, S.A.N. and Miller, C. 1992. A point mutation in a Shaker Kt
channel changes its charybdotoxin binding site from low affinity to
high affinity. Biophys. J. 62:5-7.
Goldstein, S.A.N. and Miller, C. 1993. Mechanism of charybdotoxin
block of a voltage-gated K“ channel. Biophys. J. 65:1613-1619.
Mackinnon, R. and Miller, C. 1988. Mechanism of charybdotoxin block
of Ca2t activated K“ channels. J. Gen. Physiol. 91:335-349.
Rosenthal, J.J.C., and Bezanilla, F. 1997. Evidence for RNA editing in
squid giant axon. Am. Zool. 37:1 90a.
Rosenthal, J.J.C, Vickery, R.G., and Gilly, W.F. 1996. A candidate for a
delayed Rectifier K“ channel in squid giant axon. J. of Gen. Phy.
108(3): 207-219.
Perez-Cornejo, P., Stampe, P, and Begenisich, T. 1998. Proton probing
of the charybdotoxin binding site of Shaker K“ channels. J. Gen.
Physiol. 111:441-450.
Pittston, S. 1996. Unpublished results. Hopkins Marine Station spring
student.
Sanghvi, Uma. 1997. Unpublished results. Hopkins Marine Station
spring student.
Werkman, T.R., Gastason, T.A., Rogowski, R.S., Blanstein, M.P., and
Rogawski, M.A. 1993. Tityustoxin -K-Alpha, a structurally novel
and highly potent K“ channel peptide toxin, interacts with the
alpha-denrotoxin binding site on the cloned Kv1.2K“ channel. Mol.
Pharm. 44(2): 430-436.
Fig. 1- Upper trace is Kt current before tityustoxin aplication.
Lower trace is K“ current in the presence of tityustoxin. The
current in toxin reaches a maximum much more slowly. A single
exponential was fitted to this slow rise between the arrows. 10
nm P15A tityustoxin was used.
Fig. 2- (A) H351G at pH 8.2. The top trace is the control current, the
bottom trace is the current in the presence of tityustoxin. (B)
H3516 at pH 6. The amount of current blocked by tityustoxin is
similar to (A). (C) Wild type channels at pH 8. In the presence of
toxin, current reduction is similar to H351G channels. (D) Wild type
at pH 6. Almost no current is blocked by tityustoxin. In all cases
50 nM tityustoxin was used. The H351G traces are from a 500 ms
voltage step to +40 mv. The wild type traces are from a 500 ms
step to +30 mv.
Fig. 3- Wild-type Ky values are the average of three experiments.
H351G Kp values are the average of 5 experiments. Kp's are graphed
as the log of the ratio of the K, at each voltage divided by the K, at
0 mv. The Kp of toxin binding increases with voltage in both wild
type and H351G channels.
Fig. 4- Average of five Ka, and Kgy rates of tityustoxin acting on
H3516 channels. Ka and Kgy are graphed as the log of the ratio of
Kopof at each voltage divided by Kepr at 40 mv.
Fig. 5- Average of two Ka and Kgn rates of tityustoxin acting on
wild-type channels. Ko and Kgr are graphed as the log of the ratio
of Konof at each voltage divided by Kpof at 60 mv.
Fig.1- Current traces with and without tityustoxin present
15 uA
150 ms



aaataaa-
A
H3516
-60 mV
Wild-type
Fig.2- pH dependence of toxin block
PH 6.0
pH8.2
B


aenen eneit
—
5 UA
m t
200 ms
—+40-


senennan
5 UA
-40
Fig. 3- Wild-type and H351G Kd values
0.8
0.6

0.4 -
0.2 -
-20
20 40 60 80 100

-0.4 -
Voltage (my)
—°— Wild type
—H3516
0.7
0.6
0.5
0.4
0.3
§ 02
0.1

40
-0.1
50
Fig. 4- Kon and Koff of H3516
60 70
80
Voltage (my)
90 100
—Koff
— Kon
0.15 -
0.1
0.05
O 2
60
5 -0.05
-0.1
-0.15 -
-0.2
Fig. 5-Wild-type Kon and Koff values

90
70
80
Voltage (my)
100
— 2 — Koff
4 Kon