Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICALRESEARCH COMMUNICATIONS Pages 1681-1687
August 14, 1992
Modulation of K + channels by hydrogen peroxide E. Vega-Saenz de Micra and B. Rudy* Department of Physiology and Biophysics, and Department of Biochemistry, New York University Medical Center, 550 First Ave., New York, N.Y., 10016
Received July 2, 1992
External application of hydrogen peroxide (H202) was foun~l to inhibit the time.dependent fast inactivation process of three cloned voltage-gated K1- channels expressed in Xenopus oocytes: KShlIIC, KShlIID and HuklI. As expected from kinetic models where some channels are still opening while a significant fraction of channels is already inactivated there was a large increase in current magnitude concomitant to inactivation block. The channels otherwise_ functioned normally. The effects of H 2 0 2 were specific (other cloned voltage-gated K T channels were not affected), and reversible, the currents returned to normal upon removal of the HgO 2. H 2 0 2 is produced du/dng normal metabolism; it could act as a modulator of exci~alSflity through effects on K T channels if effective local concentrations are reached in neuronal regions close to the channel. KShlIIC and KShlIID currents are very similar to an O2-sensitive K + current present in type I cells of the carotid body which is believed to underlie the modulation of excitability of these cells by changes in arterial 0 2 pressure. H 2 0 2 has been proposed as an intermediary between 0 2 and cellular response in the carotid body; our results provide support for this model. ® ~992Ao~demioP..... ~no.
Hydrogen peroxide is produced during normal cellular m~t~bolism. Although in excess or under special circumstances (e. g. the presence of Fe ~T or other transition metal ions) it produces tissue damage, it may also play roles in the regulation of cellular function (1-2). Many of the toxic effects of H 2 O g a r e believed to be the result of the generation of more toxic molecular species sucti as'highly reactive hydroxyl radicals (2-5). H 2 0 2 itself is an oxidizing agent but is not very reactive (3,5). Several enzymatic reactions result in the production of H 2 0 2 ; for example: monoamine oxidase (MAO), the enzyme that catalyzes the oxidati-ve-deamination of dopamine and other catecholamines within nerve terminals (6); superoxide dismutases which reduce superoxide to hydrogen peroxide (1, 7-8) and NAD(P)H oxidases, membrane associated enzymatic activities with ubiquitous distribution (e.g. 9-15, see also 1 & 2 and references therein) which can generate high concentrations of H 2 0 2 , as in macrophages (11-12). The role of HgO 9 production as a microbicidal agent in phagocytosis is well established (8, 11-12)'7. lqeutrophils and other phagocytes generate H 2 0 2 and other oxidants which are important for antimicrobial activity and as a side effeEt ffaay result in some inflammatory tissue damage. Cohen has suggested that high levels of H 2 0 2 resulting from an accelerated turnover of dopamine may have a role in neuronal destruction in Parkinson's disease (6). Oxidants result in damage in other tissues as well
To whom correspondence should be addressed.
1681
0006-291X/92 $4.00 Copyright © 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(4,8). However, it has also been proposed that a regulated production of H 2 0 2 may play non-destructive roles in cellular function (reviewed in 2). For example several investigators (1-2, 14-16) have suggested a role for H 2 0 2 as a participant in the cellular response to insulin. Cross et al., (13) have recentlyproposed that H 2 0 2 is an intermediary in the transduction of arterial 0 2 pressure (PO2) into modtffatlon of neuronal activity in chemosensitive neurons in me carotid body (for a review on sensory transduction in the carotid body see 17). According to this model 0 2 leads to the production of H 2 0 2 by an oxidase present in the plasma membrane of type-I chemoreceptor neurons acting as a pO-, sensor protein. The H-,O-, in turn acts on mechanisms modulating the excitability of type I cells, perhaps a K channel. This example is particularly relevant to the results described here Experiments are presented showing that H-,O-, inhibits, in a reversible and specific fashion, the time-dependent fast reactivation o~ certain voltage-gated K channels with a resultant increase in K + currents. Type I cells of the carotid body have a K channel which responds to changes in pO 2 (18-23). The experiments indicate that this modulation of K + channel activity plays a major role as one of the initial events in the sensory response to changes in pO 2 in the carotid body. Interestingly, the pO 2sensitive current of type I cells (20-23) 15as similar kinetics and voltage dependence to two of the currents studied here (KShIIIC and in particular KShIIID.1); moreover the differences in the currents seen with and without H 2 0 2 are similar to those seen in type I cells in normoxia vs. hypoxia (20-23). z..
+1--
z,
.
.
•
.
.
.
z,
/-.,
.
d-
Our experiments raise the possibility that H 2 0 2 may modulate neuronal excitability by affecting K + currents and provide further support to the idea that H 2 0 2 may be the intermediary in the regulation of excitability in response to changes in pO 2. Experimental Procedures Reagents. Several preparations of H 2 0 2 were tested with identical results. These include 3 and 30% solutions from Sigma and'several samples of a 3% solution obtained from the local drugstore. Each dilution of H 2 0 2 was prepared fresh from the stocks which were stored in their original containers. The-efficacy of the experimental solutions was found to decrease with time after dilution and exposure to air. Catalase from beef liver was obtained from Boehringer-Manheim, and Tetraethylammonium (TEA) from Fluka. RNA synthesis utilized Molecular Biology grade reagents from Sigma, BoehringerManheim and Pharmacia. Expression of voltage-gated K + channels in Xenovus oocytes. The Xenopus oocyte expression system (24-26) was used to study the K-~ channels formed by specific K + channel proteins. For this purpose Xenopus oocytes are injected with full-length capped RNA transcripts (cRNA) of cloned K + channel cDNAs and the currents thus induced studied with voltage-clamp methods (27). To produce the cRNAs, recombinant plasmids containing the K T channel cDNA insert were linearized by digestion with the appropriate restriction enzyme and RNA transcripts synthesized with T3 or T7 polymerase depending on the cDNA (28). The following cDNAs were used in this study: 29-4, one of the alternatively-spliced products of the Drosophila Shaker gene (28), KShIIIA.1 (29), KShIIIC (30), KShIIID.1 (31) and HuklI (32). The cRNAs (2-5 ng of cRNA/oocyte) were injected into stage V and VI Xenopus /aev/soocytes as described in Iverson and Rudy (28). The oocytes were incubated for two to three days at 18°C in ND96 solution (96 mM NaC1, 2mM KC1, 1.8 mM CaC12, 1 mM MgC12, 5 mM Hepes, pH 7.5) supplemented with 100 U/ml penicillin and 10Cl , g / m l streptomycin. All electrophysiological recordings were carried out at 21-22°C with a standard two microelectrode voltage-clamp (28) under continuous perfusion with ND96 with or without reagents. The data were low-pass filtered at 3 kHz with an 8-pole Bessel filter and digitized and analyzed using the pCLAMP system (Axon Instruments). Results
H 2 0 2 modulation of KShIIIC and KShIIID currents. The current in chemosensitive neurons of the carotid body that is sensitive to changes in pO 2 is a TEA-sensitive, 1682
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
voltage-dependent, transient K + current which activates when the m e m b r a n e potential is more positive than -20 m V (19-23). Several genes encoding products which express voltage-dependent K + currents in Xenopusoocytes have b e e n identified (reviewed in 3335). A m o n g these the products of F,vo related genes, KShlIIC (30) and KShlI1D (31) express TEA-sensitive, transient K " currents which activate at voltages m o r e positive than -20 mV. KShlIIC and I~ShlIID are two of the four known genes in the Shill or Shaw-related subfamily of K ' - channel genes (31). Each one of these genes generates m o r e than one product by alternative splicing. So far we have been unable to detect any responses to changes in pOp on KShIIIC or KShII1D.1 currents expressed in Xenopus oocytes. The finding that "02 leads to production of H 2 0 2 in carotid body type I cells (13) prompted us to test the effects of H 2 0 p on these currents. We found that external application of H 2 0 2 blocked the fast inactFvation of both KShIIIC (Fig. 1A) and KShIIID.1 (Fig 1B) currents. Concomitant to inactivation block there is a large increase in current magnitude. The effects of H 2 0 p on KShIIIC and KShIIID currents are quickly reversed when the H 2 0 2 is remove~t from the bath (Fig. 1). The observed effects of HpOp solutions on these currents are the result of the H p O 2 present in them and not other gontaminants as they are blocked in the presence of 10007~nits/ml of catalase (data not shown). The time course of inactivation of the pOg-sensitive channel is particularly similar to that of the currents expressed by KShlIID. KShlIIC inactivates much faster. However, since different Shill proteins can form heteromultimeric channels with intermediate inactivation properties (unpublished observations), heteromultimers of KShlIIC with KShlIID or with ShlII proteins expressing non-inactivating channels could also produce currents similar to those of the pOp-sensitive channel. We are presently studying which, if any, Shill channels are expressed iYatype I cells of the carotid body. Not all S h i l l or all transient K + channels are affected by H 2 0 2. The inhibition of channel inactivation by H 2 0 2 is specific to certain K + channel proteins. No effect was
CONTROL
H202
Wash
A
/ ~ ~ -
H202
KShIIIC ~j
~¢
~-
CONTRO[
2 5 0 nA
B H202
KShlIID. 1 - -
~
CONTROL
2 , 0 0 0 nA
ls
Figure 1. Reversible effect of HgO9 on KShIIIC (A) and KShlIID.1 (B) currents in Xenopusoocytes. The currents were recorded during depolarizing pulses to 30 mV from a holding potential of -100 mV, before the application of H202 (CONTROL), in the presence of 800 uM (A) or 500 uM (B) H909 (H~O~), or after removal of the H~O~ (WASH). The three traces from each oocyte ~trF sup~ri?nposed in the rightmost column. 1683
V o l . 186, No. 3, 1992
A
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
B
Shaker 29-4 CONTROL
KShIIIA. 1
CONTROL
H202
L 1,000nA
I 2,000nA
15ms
75ms
H202
Figure 2. Lack of effect of 82O 2 on 29-4 (A) and KShIIIA.1 (B) currents in Xenopus oocytes. The currents were re~or~ed during depolarizing pulses from -50 to + 40 mV in 10 mV intervals from a holding potential of -100 mV before the application of H202 (CONTROL) and in the presence of 1.6 mM H202 (11202).
seen on the transient currents expressed by 29-4 (Fig. 2A), or on the delayed-rectifier type currents expressed by KShIIIA.1, a protein closely related to KShIIIC and KShlIID.1 (Fig. 2B). However similar effects to those seen with KShIIIC and KShlIID.1 were seen with H u k I I (data not shown). H 2 0 2 primarily affects the fast inactivation of K S h l I I C and K S h l I I D channels. The primary effect of H p O 2 appears to be the reversible inhibition of the fast inactivation process. The voltage--dCpendence of channel opening, and the sensitivity of the channels to TEA, are not significantly affected (Fig 3). Concomitant with the removal of inactivation there is an increase in the magnitude of the currents (see Fig. 1, 3 and 4) which is particularly large in the case of KShIIIC. There are also small changes in the rate of rise of macroscopic currents. In the case of KShIIID a slow inactivation process becomes more noticeable during long depolarizing pulses (Fig. 1B and Fig. 4). Although a full kinetic analysis is beyond the scope of this paper these changes in the currents can be explained as a result of fast inactivation block (ref. 36, pp. 490-501).
CONTROL
H20 2
H202 + TEA
-.a
300 nA
30 ms
Figure 3. Voltage-dependence and TEA-block of KShIIIC currents in the presence of H202. The currents were recorded during depolarizing pulses from -50 to +40 mV in 10 mAz intervals from a holding potential of -100 mV before the application of tI202 (CONTROL), in the presence of 500 uM H~O2 (11~O2), and in the presence of 500 ulVI H20 2 plus 1 mM TEA (t120~_ +TEA). Similar rest~ts were obtained in oocytes injected wgthKShIIID.1 cRNA. - 1684
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Vol. 186, No. 3, 1992
100 ~.tM
50 ~M
200 ~M
H202 H202 CONTROL
r"--
CONTROL
2,000 nA 1S
Figure 4. Effect of increasing concentrations of H202 on KShlIID.1 currents in Xenopus ~hown are superimposed records of the currents recorded during a depolarizing pulse to +30 mV from a holding potential of -100 mV, before the application of H202 (CONTROL) and in the presence of the indicated concentration of H202 (H202). Similar results were obtained in oocytes injected with KShlIIC cRNA.
Dose Response. Variations among oocytes made the determination of a dose response difficult. However, for a given oocyte, as shown in Fig. 4 for KShlIID.1, the effects of H 2 0 2 are clearly concentration dependent. The effects of intermediary concentrations are a graded version of the effects seen at high concentrations. Similar results were obtained with KShlIIC (data not shown). Significant inactivation block and current increase were observed with concentrations as low as 25 uM. Full removal of fast inactivation occurred usually with concentrations between 600 uM to 1.5 mM. Discussion
Mechanism of inactivation removal by H 2 0 9. We have shown here that external HgOp blocks, specifically and reversibly, the inactivation of three cloned voltage-gated-K ~" channels (KShlIIC4 KShlIID.1 and HuklI). Inactivation block is accompanied by an increase in total K current, but the channels otherwise functioned normally. Not all K + channels tested were affected by H 2 0 2. Another ShIII channel (KShlIIA.1) and an inactivating channel (29-4) were not hff~cted. The three channels affected by H 2 0 2 have in common a cysteine residue in a similar amino acid sequence context in the hm'ino end portion of the protein. The amino end of Sh proteins, believed to be intracellular (33-35), contains a domain which is responsible for fast channel inactivation perhaps, by acting as a tethered "ball and chain" which occludes the internal mouth of the channel preventing ion flow (37-38). In KShlIIC the cysteine is in position 6, and is present in the sequence SSVCVSS (30); in KShlIID.1 there is a cysteine in the same position in the sequence SSVCVWS (31) and in HuklI there is a cysteine in position 13 in the sequence SSGCNSH (32). In a recent paper, Ruppersberg et al., (39) showed that cysteine oxidation results in inactivation removal of Raw3 and RCK4, homologues of KShIIIC and HukII respectively. These investigators found that in inside-out patches of oocyte membrane, where the internal surface of the membrane is exposed to the outside, inactivation of these two channels disappears. They concluded that this effect was due to oxidation since it was reversed by the addition of reduced glutathione or DTT. Moreover, they showed that if cysteine 13 in RCK4 is replaced with serine inactivation is not affected in insideout patches. Given that only channels containing cysteine in the N-terminal inactivation gate were affected by H 2 0 2 , that H 2 0 2 is a much more reactive oxidizing agent than 02, and 1685
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that H202 is highly permeable through membranes (3,5), it is very likely that the effects shown'faef-ewere the result of the oxidation of the cytoplasmic inactivation-gate cysteines in a manner similar to that discovered by Ruppersberg et al., (39). Could 1-1202 act as a modulator of neuronal excitability? K + channels modulate the ex~itabilffy 5f neurons (36). When open they drive the membrane potential towards the K +equilibri~r~ potential and oppose the effects of depolarizing and signal generating Na and Ca + currents. The effects of H-~O-, shown here if they were to take place in neurons would result in a local net incr~as~e of K + current and hence decreased excitability of that cell. Man)( neurotransmitters and other stimuli modulate neuronal excitability by affecting K T channels, typically by changes in phosphorylation (36, 40-42). Could H.,Oo, by affecting the redox state of cysteines in certain K ÷ channel proteins, also b~ a~n intermediary involved in the modulation of excitability? Several examples of modulation of protein activity by changes in the redox state of cysteines have been proposed (reviewed by Ziegier in 43). However, as pointed out by Ziegler, in contrast to protein phosphorylation it is more difficult to conceive how specificity is achieved in this case. Concentrations of external H202 in the lower range of ~hose affecting other systems (2) were found here to have sf-gn~icant effects on several K 1- channels. However, it is difficult to ascertain the potential physiological significance of these concentrations given that the cytoplasm of the oocyte, as that of other cells, contains a significant reducing power (which is likely to be the mechanism by which inactivation is restored when the H202 is removed from the bath). Therefore. the effective intracellular concentrations in contact with the channel protein are expected to be smaller than bath concentrations. Specificity of H202 oxidation of cysteines could be achieved if the sequence surrounding the cystein6- affects its sensitivity to oxidation. In this regard it is very notable that the cysteines believed to be involved in the effects seen here are present in a very similar sequence context in the three channels that were affected by H202. Specificity could also be achieved if the H202 producing system is co-localized wlhh the channel, such that effective local concentrations are achieved before the H202 is destroyed by catalase and other H202 scavenger systems. Acknowledgments We thank C. Kentros for comments on the manuscript. This work was supported by a Grant-In-Aid from the American Heart Association.
References
1. Ramasarma, T. (1982) Biochim. Biophys. Acta 694, 69-93. 2. Ramasarma, T. (1990) Indian J. of Biochem. and Biophys. 27, 269-274. 3. Halliwell, B. and Gutteridge, JMC (1985) Free radicals in Biology and Medicine. Oxford: Clarendon Press. 4. Halliwell, B. (ed). (1988) Oxygen radicals and tissue injury. Proccedings of an Upjohn Symposium. FASEB, Maryland, USA. 5. Halliwell, B. (1989) Acta Neurol. Scand. 126, 23-33. 6. Cohen, G. (1988) in Oxygen radicals and tissue injury. Proccedings of an Upjohn Symposium. Halliwell, B. (ed). FASEB, Maryland, USA. pp. 130-135. 7. Fridovich, I. (1986) Adv. Enzymol. 58, 61-97. 8. Cerutti, P.A., Fridovich, I., and McCord, J.M. Eds. (1988) Oxy-Radicals in Molecular Biology and Pathology. Alan R. Liss, Inc., New York USA. 9. Deme D., Virion, A., Hammon, N.A. and Pommier, J. (1985) FEBS Lett. 186, 107-110. 10. Heinecke, J.W. and Shapiro, B.M. (1989) Proc. Natl. Acad. Sci. 86, 1259-1263. 11. Iyer, G.Y.N., Islam, D.M.F., and Quastel, J.H. (1961) Nature 192, 535-541. 12. Hurst, J.K. and Barrett, W.C, (1989) Crit. Rev. Biochem. 24, 271-328. 1686
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13. Cross, A.R., Henderson, L., T. G. Jones, O., Delpiano M.A. and Hentschel J. and Acker, H. (1990) Biochem. J. 272, 743-747. 14. Krieger-Brauer, H. and Kather, H. (1992) J. Clin. Invest. 89, 1006-1013. 15. Mukherjee, S.P. and Lynn, W. S. (1977) Arch. Biochem. Biophys. 184, 69-76. 16. Muchmore, D.B., Little, S.A., and deHaen, C. (1982) Biochemistry21, 3886-3892. 17. Fidone S. and Gonzalez C. (1986) In Handbook of Physiology. The Respiratory System II. A.P. Fishman ed. American Phys. Soc., Bethesda, MD pp. 247-312. 18. Lopez-Barneo J., Lopez-Lopez J.R., Urena J., Gonzalez C. (1988) Science 241, 580582. 19. Hescheler J., Delpiano M.A., Acker, H. and Pietruschka, F. (1989) Brain Res. 486, 7988. 20. Delpiano M.A. and Hescheler J. (1989) FEBS letL 249, 195-198. 21. Lopez-Lopez J., Gonzalez C., Urena J. and Lopez-Barneo J. (1989) J. Gen. Physiol. 93, 1001-1015. 22. Urena J., Lopez-Lopez J., Gonzalez C., and Lopez-Barneo J. (1989) J. Gen. Physiol. 93, 979-999. 23. Ganfornina M.D. and Lopez-Barneo J. (1991) Proc. Natl. Acad. Sci. USA. 88, 29272930. 24. Lester H,A. (1988) Scienc~ 241, 1057-1063. 25. Snutch T.P. (1988) Trends Neurosci. 11, 250-256. 26. Soreq, H., and Seidman, S. (1992) Methods in Enzymol. 207, 227-225. 27. Stuhmer, W. (1992) Methods in Enzymol,207, 319-338. 28. Iverson, L. E. & Rudy, B. (1990) J. NeuroscilO, 2903-2916. 29. McCormack, T. Vega-Saenz de Miera, E., & Rudy, B. (1990). Proc. Natl. Acad. Sci. 87, 5227-5231. 30. Rudy, B., Sen, K., Vega-Saenz de Miera, E., Lau, D., Ried, T. and Ward, D.C. (1991) J. Neurosci. Res. 29, 401-412. 31. Vega-Saenz de Miera, E., Moreno, H., Fruhling, D., Kentros, C. and Rudy, B. (1992) Proc. R. Soc. Lond. B. 248, 9-18. 32. Ramaswami, M., Gautam, M., Kamb, A., Rudy, B., Tanouye, M.A., & Mathew, K.M. (1990) J. Molec. Cell. Neurosci. 1, 214-223. 33. Jan, L.-Y. and Y.-N. Jan (1990) Trends in Neurosci. 13, 415-419. 34. Perney, T.M. and Kaczmarek, L.K. (1991) Current Opinion in Cell Biology3, 663-670. 35. Rudy, B., Kentros, C., and Vega-Saenz de Miera, E. (1991) Molec. CeU.Neurosci. 2, 89-102. 36. Hille, B. (1992) Ionic Channels of Excitable Membranes. Second Edition (Sinauer, Sunderland, MA). 37. Hoshi, T., Zagotta, W. N., and Aldrich, R. W. (1990) Science250, 533-538. 38. Zagotta, W. N., Hoshi, T., and Aldrich, R. W. (1990) Science250, 568-571. 39. Ruppersberg, J.P., Stocker, M., Pongs, O., Heinemann, S. H., Frank, R., and Koenen, M. (1991) Nature352, 711-714. 40. Kaczmareck, L. K. and Levitan, I. B. (1987) Neuromodulation: The Biochemical Control of Neuronal Excitability. Oxford Univ. Press. N. Y. 41. Levitan I.B. (1988) Ann. Rev. Neurosci. 11, 119-136. 42. Rudy, B. (1988) Neur0science 25, 729-750. 43. Ziegler, D.M. (1985) Ann. Rev. Biochem. 54, 305-329.
1687
BIOCHEMICAL AND BIOPHYSICALRESEARCH COMMUNICATIONS Pages 1681-1687
August 14, 1992
Modulation of K + channels by hydrogen peroxide E. Vega-Saenz de Micra and B. Rudy* Department of Physiology and Biophysics, and Department of Biochemistry, New York University Medical Center, 550 First Ave., New York, N.Y., 10016
Received July 2, 1992
External application of hydrogen peroxide (H202) was foun~l to inhibit the time.dependent fast inactivation process of three cloned voltage-gated K1- channels expressed in Xenopus oocytes: KShlIIC, KShlIID and HuklI. As expected from kinetic models where some channels are still opening while a significant fraction of channels is already inactivated there was a large increase in current magnitude concomitant to inactivation block. The channels otherwise_ functioned normally. The effects of H 2 0 2 were specific (other cloned voltage-gated K T channels were not affected), and reversible, the currents returned to normal upon removal of the HgO 2. H 2 0 2 is produced du/dng normal metabolism; it could act as a modulator of exci~alSflity through effects on K T channels if effective local concentrations are reached in neuronal regions close to the channel. KShlIIC and KShlIID currents are very similar to an O2-sensitive K + current present in type I cells of the carotid body which is believed to underlie the modulation of excitability of these cells by changes in arterial 0 2 pressure. H 2 0 2 has been proposed as an intermediary between 0 2 and cellular response in the carotid body; our results provide support for this model. ® ~992Ao~demioP..... ~no.
Hydrogen peroxide is produced during normal cellular m~t~bolism. Although in excess or under special circumstances (e. g. the presence of Fe ~T or other transition metal ions) it produces tissue damage, it may also play roles in the regulation of cellular function (1-2). Many of the toxic effects of H 2 O g a r e believed to be the result of the generation of more toxic molecular species sucti as'highly reactive hydroxyl radicals (2-5). H 2 0 2 itself is an oxidizing agent but is not very reactive (3,5). Several enzymatic reactions result in the production of H 2 0 2 ; for example: monoamine oxidase (MAO), the enzyme that catalyzes the oxidati-ve-deamination of dopamine and other catecholamines within nerve terminals (6); superoxide dismutases which reduce superoxide to hydrogen peroxide (1, 7-8) and NAD(P)H oxidases, membrane associated enzymatic activities with ubiquitous distribution (e.g. 9-15, see also 1 & 2 and references therein) which can generate high concentrations of H 2 0 2 , as in macrophages (11-12). The role of HgO 9 production as a microbicidal agent in phagocytosis is well established (8, 11-12)'7. lqeutrophils and other phagocytes generate H 2 0 2 and other oxidants which are important for antimicrobial activity and as a side effeEt ffaay result in some inflammatory tissue damage. Cohen has suggested that high levels of H 2 0 2 resulting from an accelerated turnover of dopamine may have a role in neuronal destruction in Parkinson's disease (6). Oxidants result in damage in other tissues as well
To whom correspondence should be addressed.
1681
0006-291X/92 $4.00 Copyright © 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(4,8). However, it has also been proposed that a regulated production of H 2 0 2 may play non-destructive roles in cellular function (reviewed in 2). For example several investigators (1-2, 14-16) have suggested a role for H 2 0 2 as a participant in the cellular response to insulin. Cross et al., (13) have recentlyproposed that H 2 0 2 is an intermediary in the transduction of arterial 0 2 pressure (PO2) into modtffatlon of neuronal activity in chemosensitive neurons in me carotid body (for a review on sensory transduction in the carotid body see 17). According to this model 0 2 leads to the production of H 2 0 2 by an oxidase present in the plasma membrane of type-I chemoreceptor neurons acting as a pO-, sensor protein. The H-,O-, in turn acts on mechanisms modulating the excitability of type I cells, perhaps a K channel. This example is particularly relevant to the results described here Experiments are presented showing that H-,O-, inhibits, in a reversible and specific fashion, the time-dependent fast reactivation o~ certain voltage-gated K channels with a resultant increase in K + currents. Type I cells of the carotid body have a K channel which responds to changes in pO 2 (18-23). The experiments indicate that this modulation of K + channel activity plays a major role as one of the initial events in the sensory response to changes in pO 2 in the carotid body. Interestingly, the pO 2sensitive current of type I cells (20-23) 15as similar kinetics and voltage dependence to two of the currents studied here (KShIIIC and in particular KShIIID.1); moreover the differences in the currents seen with and without H 2 0 2 are similar to those seen in type I cells in normoxia vs. hypoxia (20-23). z..
+1--
z,
.
.
•
.
.
.
z,
/-.,
.
d-
Our experiments raise the possibility that H 2 0 2 may modulate neuronal excitability by affecting K + currents and provide further support to the idea that H 2 0 2 may be the intermediary in the regulation of excitability in response to changes in pO 2. Experimental Procedures Reagents. Several preparations of H 2 0 2 were tested with identical results. These include 3 and 30% solutions from Sigma and'several samples of a 3% solution obtained from the local drugstore. Each dilution of H 2 0 2 was prepared fresh from the stocks which were stored in their original containers. The-efficacy of the experimental solutions was found to decrease with time after dilution and exposure to air. Catalase from beef liver was obtained from Boehringer-Manheim, and Tetraethylammonium (TEA) from Fluka. RNA synthesis utilized Molecular Biology grade reagents from Sigma, BoehringerManheim and Pharmacia. Expression of voltage-gated K + channels in Xenovus oocytes. The Xenopus oocyte expression system (24-26) was used to study the K-~ channels formed by specific K + channel proteins. For this purpose Xenopus oocytes are injected with full-length capped RNA transcripts (cRNA) of cloned K + channel cDNAs and the currents thus induced studied with voltage-clamp methods (27). To produce the cRNAs, recombinant plasmids containing the K T channel cDNA insert were linearized by digestion with the appropriate restriction enzyme and RNA transcripts synthesized with T3 or T7 polymerase depending on the cDNA (28). The following cDNAs were used in this study: 29-4, one of the alternatively-spliced products of the Drosophila Shaker gene (28), KShIIIA.1 (29), KShIIIC (30), KShIIID.1 (31) and HuklI (32). The cRNAs (2-5 ng of cRNA/oocyte) were injected into stage V and VI Xenopus /aev/soocytes as described in Iverson and Rudy (28). The oocytes were incubated for two to three days at 18°C in ND96 solution (96 mM NaC1, 2mM KC1, 1.8 mM CaC12, 1 mM MgC12, 5 mM Hepes, pH 7.5) supplemented with 100 U/ml penicillin and 10Cl , g / m l streptomycin. All electrophysiological recordings were carried out at 21-22°C with a standard two microelectrode voltage-clamp (28) under continuous perfusion with ND96 with or without reagents. The data were low-pass filtered at 3 kHz with an 8-pole Bessel filter and digitized and analyzed using the pCLAMP system (Axon Instruments). Results
H 2 0 2 modulation of KShIIIC and KShIIID currents. The current in chemosensitive neurons of the carotid body that is sensitive to changes in pO 2 is a TEA-sensitive, 1682
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
voltage-dependent, transient K + current which activates when the m e m b r a n e potential is more positive than -20 m V (19-23). Several genes encoding products which express voltage-dependent K + currents in Xenopusoocytes have b e e n identified (reviewed in 3335). A m o n g these the products of F,vo related genes, KShlIIC (30) and KShlI1D (31) express TEA-sensitive, transient K " currents which activate at voltages m o r e positive than -20 mV. KShlIIC and I~ShlIID are two of the four known genes in the Shill or Shaw-related subfamily of K ' - channel genes (31). Each one of these genes generates m o r e than one product by alternative splicing. So far we have been unable to detect any responses to changes in pOp on KShIIIC or KShII1D.1 currents expressed in Xenopus oocytes. The finding that "02 leads to production of H 2 0 2 in carotid body type I cells (13) prompted us to test the effects of H 2 0 p on these currents. We found that external application of H 2 0 2 blocked the fast inactFvation of both KShIIIC (Fig. 1A) and KShIIID.1 (Fig 1B) currents. Concomitant to inactivation block there is a large increase in current magnitude. The effects of H 2 0 p on KShIIIC and KShIIID currents are quickly reversed when the H 2 0 2 is remove~t from the bath (Fig. 1). The observed effects of HpOp solutions on these currents are the result of the H p O 2 present in them and not other gontaminants as they are blocked in the presence of 10007~nits/ml of catalase (data not shown). The time course of inactivation of the pOg-sensitive channel is particularly similar to that of the currents expressed by KShlIID. KShlIIC inactivates much faster. However, since different Shill proteins can form heteromultimeric channels with intermediate inactivation properties (unpublished observations), heteromultimers of KShlIIC with KShlIID or with ShlII proteins expressing non-inactivating channels could also produce currents similar to those of the pOp-sensitive channel. We are presently studying which, if any, Shill channels are expressed iYatype I cells of the carotid body. Not all S h i l l or all transient K + channels are affected by H 2 0 2. The inhibition of channel inactivation by H 2 0 2 is specific to certain K + channel proteins. No effect was
CONTROL
H202
Wash
A
/ ~ ~ -
H202
KShIIIC ~j
~¢
~-
CONTRO[
2 5 0 nA
B H202
KShlIID. 1 - -
~
CONTROL
2 , 0 0 0 nA
ls
Figure 1. Reversible effect of HgO9 on KShIIIC (A) and KShlIID.1 (B) currents in Xenopusoocytes. The currents were recorded during depolarizing pulses to 30 mV from a holding potential of -100 mV, before the application of H202 (CONTROL), in the presence of 800 uM (A) or 500 uM (B) H909 (H~O~), or after removal of the H~O~ (WASH). The three traces from each oocyte ~trF sup~ri?nposed in the rightmost column. 1683
V o l . 186, No. 3, 1992
A
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
B
Shaker 29-4 CONTROL
KShIIIA. 1
CONTROL
H202
L 1,000nA
I 2,000nA
15ms
75ms
H202
Figure 2. Lack of effect of 82O 2 on 29-4 (A) and KShIIIA.1 (B) currents in Xenopus oocytes. The currents were re~or~ed during depolarizing pulses from -50 to + 40 mV in 10 mV intervals from a holding potential of -100 mV before the application of H202 (CONTROL) and in the presence of 1.6 mM H202 (11202).
seen on the transient currents expressed by 29-4 (Fig. 2A), or on the delayed-rectifier type currents expressed by KShIIIA.1, a protein closely related to KShIIIC and KShlIID.1 (Fig. 2B). However similar effects to those seen with KShIIIC and KShlIID.1 were seen with H u k I I (data not shown). H 2 0 2 primarily affects the fast inactivation of K S h l I I C and K S h l I I D channels. The primary effect of H p O 2 appears to be the reversible inhibition of the fast inactivation process. The voltage--dCpendence of channel opening, and the sensitivity of the channels to TEA, are not significantly affected (Fig 3). Concomitant with the removal of inactivation there is an increase in the magnitude of the currents (see Fig. 1, 3 and 4) which is particularly large in the case of KShIIIC. There are also small changes in the rate of rise of macroscopic currents. In the case of KShIIID a slow inactivation process becomes more noticeable during long depolarizing pulses (Fig. 1B and Fig. 4). Although a full kinetic analysis is beyond the scope of this paper these changes in the currents can be explained as a result of fast inactivation block (ref. 36, pp. 490-501).
CONTROL
H20 2
H202 + TEA
-.a
300 nA
30 ms
Figure 3. Voltage-dependence and TEA-block of KShIIIC currents in the presence of H202. The currents were recorded during depolarizing pulses from -50 to +40 mV in 10 mAz intervals from a holding potential of -100 mV before the application of tI202 (CONTROL), in the presence of 500 uM H~O2 (11~O2), and in the presence of 500 ulVI H20 2 plus 1 mM TEA (t120~_ +TEA). Similar rest~ts were obtained in oocytes injected wgthKShIIID.1 cRNA. - 1684
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Vol. 186, No. 3, 1992
100 ~.tM
50 ~M
200 ~M
H202 H202 CONTROL
r"--
CONTROL
2,000 nA 1S
Figure 4. Effect of increasing concentrations of H202 on KShlIID.1 currents in Xenopus ~hown are superimposed records of the currents recorded during a depolarizing pulse to +30 mV from a holding potential of -100 mV, before the application of H202 (CONTROL) and in the presence of the indicated concentration of H202 (H202). Similar results were obtained in oocytes injected with KShlIIC cRNA.
Dose Response. Variations among oocytes made the determination of a dose response difficult. However, for a given oocyte, as shown in Fig. 4 for KShlIID.1, the effects of H 2 0 2 are clearly concentration dependent. The effects of intermediary concentrations are a graded version of the effects seen at high concentrations. Similar results were obtained with KShlIIC (data not shown). Significant inactivation block and current increase were observed with concentrations as low as 25 uM. Full removal of fast inactivation occurred usually with concentrations between 600 uM to 1.5 mM. Discussion
Mechanism of inactivation removal by H 2 0 9. We have shown here that external HgOp blocks, specifically and reversibly, the inactivation of three cloned voltage-gated-K ~" channels (KShlIIC4 KShlIID.1 and HuklI). Inactivation block is accompanied by an increase in total K current, but the channels otherwise functioned normally. Not all K + channels tested were affected by H 2 0 2. Another ShIII channel (KShlIIA.1) and an inactivating channel (29-4) were not hff~cted. The three channels affected by H 2 0 2 have in common a cysteine residue in a similar amino acid sequence context in the hm'ino end portion of the protein. The amino end of Sh proteins, believed to be intracellular (33-35), contains a domain which is responsible for fast channel inactivation perhaps, by acting as a tethered "ball and chain" which occludes the internal mouth of the channel preventing ion flow (37-38). In KShlIIC the cysteine is in position 6, and is present in the sequence SSVCVSS (30); in KShlIID.1 there is a cysteine in the same position in the sequence SSVCVWS (31) and in HuklI there is a cysteine in position 13 in the sequence SSGCNSH (32). In a recent paper, Ruppersberg et al., (39) showed that cysteine oxidation results in inactivation removal of Raw3 and RCK4, homologues of KShIIIC and HukII respectively. These investigators found that in inside-out patches of oocyte membrane, where the internal surface of the membrane is exposed to the outside, inactivation of these two channels disappears. They concluded that this effect was due to oxidation since it was reversed by the addition of reduced glutathione or DTT. Moreover, they showed that if cysteine 13 in RCK4 is replaced with serine inactivation is not affected in insideout patches. Given that only channels containing cysteine in the N-terminal inactivation gate were affected by H 2 0 2 , that H 2 0 2 is a much more reactive oxidizing agent than 02, and 1685
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that H202 is highly permeable through membranes (3,5), it is very likely that the effects shown'faef-ewere the result of the oxidation of the cytoplasmic inactivation-gate cysteines in a manner similar to that discovered by Ruppersberg et al., (39). Could 1-1202 act as a modulator of neuronal excitability? K + channels modulate the ex~itabilffy 5f neurons (36). When open they drive the membrane potential towards the K +equilibri~r~ potential and oppose the effects of depolarizing and signal generating Na and Ca + currents. The effects of H-~O-, shown here if they were to take place in neurons would result in a local net incr~as~e of K + current and hence decreased excitability of that cell. Man)( neurotransmitters and other stimuli modulate neuronal excitability by affecting K T channels, typically by changes in phosphorylation (36, 40-42). Could H.,Oo, by affecting the redox state of cysteines in certain K ÷ channel proteins, also b~ a~n intermediary involved in the modulation of excitability? Several examples of modulation of protein activity by changes in the redox state of cysteines have been proposed (reviewed by Ziegier in 43). However, as pointed out by Ziegler, in contrast to protein phosphorylation it is more difficult to conceive how specificity is achieved in this case. Concentrations of external H202 in the lower range of ~hose affecting other systems (2) were found here to have sf-gn~icant effects on several K 1- channels. However, it is difficult to ascertain the potential physiological significance of these concentrations given that the cytoplasm of the oocyte, as that of other cells, contains a significant reducing power (which is likely to be the mechanism by which inactivation is restored when the H202 is removed from the bath). Therefore. the effective intracellular concentrations in contact with the channel protein are expected to be smaller than bath concentrations. Specificity of H202 oxidation of cysteines could be achieved if the sequence surrounding the cystein6- affects its sensitivity to oxidation. In this regard it is very notable that the cysteines believed to be involved in the effects seen here are present in a very similar sequence context in the three channels that were affected by H202. Specificity could also be achieved if the H202 producing system is co-localized wlhh the channel, such that effective local concentrations are achieved before the H202 is destroyed by catalase and other H202 scavenger systems. Acknowledgments We thank C. Kentros for comments on the manuscript. This work was supported by a Grant-In-Aid from the American Heart Association.
References
1. Ramasarma, T. (1982) Biochim. Biophys. Acta 694, 69-93. 2. Ramasarma, T. (1990) Indian J. of Biochem. and Biophys. 27, 269-274. 3. Halliwell, B. and Gutteridge, JMC (1985) Free radicals in Biology and Medicine. Oxford: Clarendon Press. 4. Halliwell, B. (ed). (1988) Oxygen radicals and tissue injury. Proccedings of an Upjohn Symposium. FASEB, Maryland, USA. 5. Halliwell, B. (1989) Acta Neurol. Scand. 126, 23-33. 6. Cohen, G. (1988) in Oxygen radicals and tissue injury. Proccedings of an Upjohn Symposium. Halliwell, B. (ed). FASEB, Maryland, USA. pp. 130-135. 7. Fridovich, I. (1986) Adv. Enzymol. 58, 61-97. 8. Cerutti, P.A., Fridovich, I., and McCord, J.M. Eds. (1988) Oxy-Radicals in Molecular Biology and Pathology. Alan R. Liss, Inc., New York USA. 9. Deme D., Virion, A., Hammon, N.A. and Pommier, J. (1985) FEBS Lett. 186, 107-110. 10. Heinecke, J.W. and Shapiro, B.M. (1989) Proc. Natl. Acad. Sci. 86, 1259-1263. 11. Iyer, G.Y.N., Islam, D.M.F., and Quastel, J.H. (1961) Nature 192, 535-541. 12. Hurst, J.K. and Barrett, W.C, (1989) Crit. Rev. Biochem. 24, 271-328. 1686
Vol. 186, No. 3, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13. Cross, A.R., Henderson, L., T. G. Jones, O., Delpiano M.A. and Hentschel J. and Acker, H. (1990) Biochem. J. 272, 743-747. 14. Krieger-Brauer, H. and Kather, H. (1992) J. Clin. Invest. 89, 1006-1013. 15. Mukherjee, S.P. and Lynn, W. S. (1977) Arch. Biochem. Biophys. 184, 69-76. 16. Muchmore, D.B., Little, S.A., and deHaen, C. (1982) Biochemistry21, 3886-3892. 17. Fidone S. and Gonzalez C. (1986) In Handbook of Physiology. The Respiratory System II. A.P. Fishman ed. American Phys. Soc., Bethesda, MD pp. 247-312. 18. Lopez-Barneo J., Lopez-Lopez J.R., Urena J., Gonzalez C. (1988) Science 241, 580582. 19. Hescheler J., Delpiano M.A., Acker, H. and Pietruschka, F. (1989) Brain Res. 486, 7988. 20. Delpiano M.A. and Hescheler J. (1989) FEBS letL 249, 195-198. 21. Lopez-Lopez J., Gonzalez C., Urena J. and Lopez-Barneo J. (1989) J. Gen. Physiol. 93, 1001-1015. 22. Urena J., Lopez-Lopez J., Gonzalez C., and Lopez-Barneo J. (1989) J. Gen. Physiol. 93, 979-999. 23. Ganfornina M.D. and Lopez-Barneo J. (1991) Proc. Natl. Acad. Sci. USA. 88, 29272930. 24. Lester H,A. (1988) Scienc~ 241, 1057-1063. 25. Snutch T.P. (1988) Trends Neurosci. 11, 250-256. 26. Soreq, H., and Seidman, S. (1992) Methods in Enzymol. 207, 227-225. 27. Stuhmer, W. (1992) Methods in Enzymol,207, 319-338. 28. Iverson, L. E. & Rudy, B. (1990) J. NeuroscilO, 2903-2916. 29. McCormack, T. Vega-Saenz de Miera, E., & Rudy, B. (1990). Proc. Natl. Acad. Sci. 87, 5227-5231. 30. Rudy, B., Sen, K., Vega-Saenz de Miera, E., Lau, D., Ried, T. and Ward, D.C. (1991) J. Neurosci. Res. 29, 401-412. 31. Vega-Saenz de Miera, E., Moreno, H., Fruhling, D., Kentros, C. and Rudy, B. (1992) Proc. R. Soc. Lond. B. 248, 9-18. 32. Ramaswami, M., Gautam, M., Kamb, A., Rudy, B., Tanouye, M.A., & Mathew, K.M. (1990) J. Molec. Cell. Neurosci. 1, 214-223. 33. Jan, L.-Y. and Y.-N. Jan (1990) Trends in Neurosci. 13, 415-419. 34. Perney, T.M. and Kaczmarek, L.K. (1991) Current Opinion in Cell Biology3, 663-670. 35. Rudy, B., Kentros, C., and Vega-Saenz de Miera, E. (1991) Molec. CeU.Neurosci. 2, 89-102. 36. Hille, B. (1992) Ionic Channels of Excitable Membranes. Second Edition (Sinauer, Sunderland, MA). 37. Hoshi, T., Zagotta, W. N., and Aldrich, R. W. (1990) Science250, 533-538. 38. Zagotta, W. N., Hoshi, T., and Aldrich, R. W. (1990) Science250, 568-571. 39. Ruppersberg, J.P., Stocker, M., Pongs, O., Heinemann, S. H., Frank, R., and Koenen, M. (1991) Nature352, 711-714. 40. Kaczmareck, L. K. and Levitan, I. B. (1987) Neuromodulation: The Biochemical Control of Neuronal Excitability. Oxford Univ. Press. N. Y. 41. Levitan I.B. (1988) Ann. Rev. Neurosci. 11, 119-136. 42. Rudy, B. (1988) Neur0science 25, 729-750. 43. Ziegler, D.M. (1985) Ann. Rev. Biochem. 54, 305-329.
1687
