22
Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site on 09/23/08. For personal use only.

Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

AR300-FM ARI 22 January 2007 14:24

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 2: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

Life Among the AxonsClay M. ArmstrongDepartment of Physiology, University of Pennsylvania, Philadelphia, Pennsylvania19104; email: [email protected]

Annu. Rev. Physiol. 2007. 69:1–18

First published online as a Review inAdvance on October 19, 2006

The Annual Review of Physiology is online athttp://physiol.annualreviews.org

This article’s doi:10.1146/annurev.physiol.69.120205.124448

Copyright c© 2007 by Annual Reviews.All rights reserved

0066-4278/07/0315-0001$20.00

Key Words

ion channel, ion selectivity, voltage gating, inactivation

AbstractA blink in history’s eye has brought us an understanding of electricity,and with it a revolution in human life. From the frog leg twitch ex-periments of Galvani and the batteries of Volta, we have progressedto telegraphs, motors, telephones, computers, and the Internet. Inthe same period, the ubiquitous role of electricity in animal andplant life has become clear. A great milestone in this journey was theelucidation of electrical signaling by Hodgkin & Huxley in 1952.This chapter gives a personal account of a small part of this story,the transformation of the rather abstract electrical conductances ofHodgkin & Huxley into the more tangible gated ion channel.

1

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 3: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

INTRODUCTION: MY LIFE ASAN ASPIRING SCIENTIST

My early years were in some respects quitetragic. I grew up a male in Texas too small toplay football. Perhaps this fact more than anyother destined me to be a scientist. To com-pensate, I struggled for a time on the tennisteam, without notable success. Helpful in tak-ing my mind off these failures were my highschool English teachers, both spinsters, oneshrunken but vital and witty, the other ro-mantic and in thrall to Beowulf. Together theygave me a permanent love for literature. Math,chemistry, and physics scarcely existed in thecurriculum, leaving me with deficits I am stillstruggling to overcome. All in all, it was not apromising start for a scientific career.

Things changed when I got to Rice, wheremy heart belongs, educationally speaking. Atthe time Rice was tough enough that a third ofthe freshman class did not return the secondyear. The idea was to whip those Texas coun-try boys into shape (football players were ex-empt), and for the survivors it seemed to workquite well. My love of literature persisted andgrew; in literature I found a continual awaken-ing. And at Rice I got an excellent groundingin chemistry and physics but demonstrated nogreat talent.

What does a middle-class boy with a goodmemory and no particular aptitudes do withhimself? He goes to medical school, in mycase Washington University. It is and was afine school, but after the first year I was nota fine student. Fatefully, in physiology classI was exposed to the Hodgkin & Huxley (1)formulation and was permanently imprintedby its mysteries. I was also fascinated by brainelectrophysiology, and under the supervisionof Bill Landau, I worked in the lab of thelegendary George H. Bishop, using his hand-made equipment. It was in his darkened, elec-trically shielded room that I discovered justhow remote and specialized science is andlearned many things that I could never explainto my mother. All the big questions about myplace in the universe were at least temporar-

ily shunted aside. In compensation, I foundit exciting to walk mentally in obscure placeswhere there were few footprints. One of myfavorite pieces of work was done in Bishop’slab, deciphering the electrical waves elicited inthe visual cortex by lateral geniculate stimula-tion. The pieces of this puzzle all fell togetherone memorable day as I walked in Forest Park.Alas, only single-cell recordings were consid-ered significant in those days, and my techni-cally unfashionable work sank without a trace(2, 3).

Life, which is a series of chances, changeddirection when the threat of the draft pushedme from Washington University to the NIH,where I worked in the lab of the famous K.S.(Kacy) Cole. Kacy’s lab had a mind-hand dual-ity, and I signed on as a pair of unskilled hands.I learned an enormous amount there, study-ing, as my aunt Roberta called it, little bagsof salt (cells). The focus in Kacy’s lab was, ofcourse, nerve conduction, and as I saw it whenI got my license to think a bit, there werethree questions remaining after Hodgkin &Huxley’s (1) great work. (a) How do ionsget through membranes? (b) How does themembrane distinguish between Na+ and K+?(c) How does the membrane change its electri-cal resistance and ion selectivity in a fractionof a millisecond? The dominant view was thatcarriers, one carrier for Na+ and one for K+

ions, ferry ions across membranes. Channelswere a less popular idea because selectivityseemed harder to explain for a channel. Someguesses were pretty wild. For a time there waswidespread intoxication with lipids and phasetransitions and even a proposal that the con-ducting path in a membrane is formed of lipid,with protein as the insulator. One promi-nent biochemically oriented lab was certainthat axon propagation was mediated by acetyl-choline. Others, certain that ions were boundin cytoplasm, denied the importance of mem-branes. In Kacy’s physically oriented lab I wassafe from many of these turbulent currentsand had excellent exposure to electricity, feed-back, mathematics, and physics.

2 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 4: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

In closing this short personal section, Ihave many debts to mentors, postdocs, andstudents. None is greater than my debt to A.F.Huxley, in whose lab I struggled unsuccess-fully, after my NIH years, to become a worthystudent of muscle. My failure in this regard inno way dampens my admiration and respectfor this great and imaginative scientist.

TEA+ AND THE GROSSARCHITECTURE OF THEK CHANNEL

My first (and perhaps only) good researchidea came as a result of a short visit to A.F.Huxley’s lab, where a talented student, DenisNoble, was modifying the Hodgkin-Huxleyequations to fit cardiac action potentials (4).These equations describe an action potentialthat is sharply depolarized to a peak by an in-crease of Na conductance and then repolar-ized by an increase of K conductance. One ofNoble’s tasks was to explain why the peak of acardiac action potential is followed by a longplateau rather than repolarizing promptly, asin an axon. For this purpose, he invoked the“anomalous rectifier,” recently discovered byKatz (5) in muscle fibers: anomalous in thesense that K permeability decreases with volt-age, rather than increasing like the normalK permeability in axon membrane. On re-turning to the NIH I found that Tasaki &Hagiwara (6) had recently published a papershowing that internal TEA+ causes a cardiac-like plateau action potential when injectedinto squid axons. Their data suggested that to-tal conductance during the plateau was lowerthan at rest, as though TEA+ somehow pro-duced anomalous rectification. This was goodenough to get started, so Leonard Binstockand I spent a summer at Woods Hole in-jecting squid axons with TEA+ and exam-ining the membrane currents under voltageclamp. Neither of us was very good with ax-ons, but luck eventually smiled on us, andwe saw that with ∼40-mM TEA+ inside, theaction potential indeed had a long plateauand that the outward IK measured in volt-

Anomalous orinward rectifier:prechannel namesfor KIR channels

Rest: resting Vm.Loosely, Vm <

−60 mV

IK or INa: currentthrough K or Nachannels,respectively. Inwardcurrent is plotteddown. In prechanneldays, one spoke ofNa conductance orpermeability ratherthan Na channels

Depolarization:driving Vm positiverelative to rest

age clamp was completely suppressed. In con-trast, external TEA+ had almost no effect.Most excitingly, when we raised the extracel-lular [K+], we found that inward K+ currentwas near normal even though there was nooutward current! With intracellular TEA+, Kconductance became a K-anomalous rectifier.Activation of the rectifier on depolarization(in present terms, the gating) seemed to oc-cur normally; i.e., the conducting structures,whatever they were, activated normally but al-lowed only influx of K+ ions and no efflux (7).

Further work on the very giant axons ofChilean squid showed that with low concen-tration, when only half of IK was blocked in thesteady state, the time course of block was slowenough to be directly measurable (8). Thismade it possible to estimate the rate of K+

movement through the conductor simply bymultiplying the TEA+ entry rate by the ra-tio of [K+] to [TEA+], yielding an intriguingnumber, approximately 600 K+ ions ms−1.

How does TEA+ work? I studied TEA+

with molecular models and found it to beroughly tetrahedral in shape, with little flex-ibility. Moreover, it is approximately the sizeof a K+ ion with one hydration shell (∼8 A).The simplicity of the TEA+ molecule seemedto require an equally simple model, and theidea grew irresistibly in my mind that TEA+

was blocking a channel through the mem-brane that had a wide inner vestibule and anarrower outer section too small for TEA+ topass through. Furthermore, it seemed that agate at the inner end of the channel preventedTEA+ entry when it was closed (at rest). Toinvestigate the features of TEA+ that affectedits entry rate into the channels, I had East-man Kodak synthesize derivatives of TEA+

with one of its ethyl arms replaced by a pro-gressively longer hydrocarbon chain. The ex-periments were performed in Chile, where theocean was inspiring, the facilities inspiringlyprimitive, and the companionship wonderful.Much to my surprise, the blocking potencyincreased with chain length. Kinetically, themain effect was not on the entry but on the exitrate, roughly consistent with the idea that each

www.annualreviews.org • Life Among the Axons 3

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 5: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

additional CH2 group added ∼600 cal mol−1

to the binding energy. When TEA+ was ap-plied at low concentration there was a dra-matic “inactivation” of the current, shown inFigure 1a for C9

+ (nonyltriethylammoniumion) (9). In Figure 1a, the top trace is controlIK, and below this are traces of two familiesof IK taken at low and higher internal C9

+

concentrations. Each family was elicited by aseries of voltage steps to −30 through +90mV, with 2-s recovery intervals between steps.The inactivation rate is faster with high C9

+

and changes little or not at all as the voltage

is driven more positive. This implies that C9+

simply diffuses into the channel at a rate pro-portional to its concentration and that voltagehas little effect on the entry rate.

The cartoons in Figure 1 depict what isoccurring during C9

+ block of IK. At rest agate at the inner end of the channel is closed,and the vestibule just above it is empty. Afterdepolarization, several steps lead to the all-or-none opening of the gate, and internal K+

ions move into the vestibule, with their hydra-tion shell in place. To pass through the chan-nel’s narrow selectivity filter (a term coined

Figure 1C9

+ block of IK resembles inactivation. The cartoons underneath parts a and b depict what is occurringat corresponding stages. (a) IK (in squid axon) normally increases, with a lag after a step depolarization(arrow), to a fixed level (trace labeled control ). With 60 μM C9

+ in the axon, IK increases normally andthen “inactivates.” At rest (−60 mV) a gate at the inner end of the channel protects a relatively widevestibule. When the gate is opened by depolarization, K+ ions enter the vestibule, dehydrate as they gothrough a narrow tunnel (filter), and rehydrate at the outer end of the tunnel. C9

+, present at lowconcentration, takes time to find the channel but then enters the vestibule and gets stuck, with itshydrophobic nonyl chain bound to a hydrophobic region in the vestibule wall. “Inactivation” is faster athigher C9

+ concentrations. A family of depolarizations is shown at each concentration for steps to −30though +90 mV. (b) At the end of a short depolarization, IK is inward (external [K+] is high) but decaysrapidly as the channels deactivate. Deactivation is much slower after a long step because the gate isunlikely to close when the vestibule contains C9

+. In this case, IK initially increases as C9+ is driven from

the vestibule by inward-moving K+, and the gate then closes. (c) The gate can be slammed shut owing torepolarization to −100 mV, trapping C9

+ in the vestibule. Figure modified from Reference 9.

4 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 6: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

by Bert Hille), they must dehydrate and thenrehydrate as they leave the filter. After a timethat depends on concentration, a C9

+ ion dif-fuses into the vestibule and gets stuck becauseit cannot move through the filter. Its hydro-carbon nonyl chain increases affinity by bind-ing to a hydrophobic portion of the vestibule’slining.

On repolarization C9 comes out of thechannels, and there are several points of inter-est, illustrated in Figure 1b. The traces showIK during a depolarizing step, with repolar-ization to −60 mV, for a C9

+-containing axonbathed in high external K+. When the stepis short, there is little block, and the chan-nel gates close relatively quickly, with a sim-ple monotonic time course. After the longerstep, when block is complete, IK on repolar-ization is at first quite small but increases withtime as external K+ ions move inward throughthe channel and push out the C9

+ ions. Inlow external K+ (not shown in Figure 1), thisknock-off effect does not occur, and recoveryis much slower. Even with high K+, closingof a blocked channel is greatly prolonged: C9

has a “foot-in-the-door” effect (an apt phrasefrom Hodgkin) that hinders the gate fromclosing. However, if voltage is made very neg-ative it is possible to close the gate and trapC9

+ in some of the channels for many seconds(Figure 1c).

All these phenomena were so easy to ex-plain with a channel model that I acceptedit and never looked back. At approximatelythe same time, Hille (10) was finding com-pelling reasons to believe that Na+ also trav-eled through channels. The main features of Kchannel architecture seemed as follows. (a) Agate at the inner end was, following Hodgkinand Huxley, fully open or fully closed and forthis reason is portrayed as an open-or-closedtrap door in Figure 1. How voltage operatedthis door remained to be determined. (b) Thegate opened into a vestibule large enough forTEA+ or a hydrated K+ ion. Part of the liningwas hydrophobic. (c) Above the vestibule was anarrow tunnel, permeable to a dehydrated or

Repolarization:restoring Vm toward,to, or negative to rest

Pancho: FranciscoBezanilla

partially dehydrated K+ ion but not to TEA+,with its covalently bound side chains.

TEA+ sites are at both ends of K chan-nels in frog myelinated axons. The inner siteis similar to the one just described for squidchannels, but the outer one is quite differentin character and not sensitive to C9

+ (11).

SELECTIVITY ANDCONDUCTION

With a mental picture of the channel’s archi-tecture, could we imagine how the channeldistinguishes between Na+ and K+ ions? Ac-cording to the gross architectural picture, K+

apparently dehydrates at least partially whilegoing through the filter. Both Na+ and K+

have a charge of 1 e (electronic charge) andare simple hard-sphere cations (12), so whycannot the smaller Na+ (which has a Paul-ing radius of 0.95 A) pass through a filterthat accepts K+ (1.33 A)? This problem be-came acute when Pancho and I discovered thatNa+ in the internal medium partially blockedK channels at high membrane voltage (13).Aided in our thinking by the close-fit hy-pothesis used by Lorin Mullins (14), we pro-posed the scheme given in Figure 2a. Thetop row shows K+ and Na+ ions in water,with the oxygens of water molecules packedclosely around both, binding more tightly tothe smaller Na+ than to K+ (which has a lowerhydration energy). The filter is lined withhydrophilic groups, e.g., carbonyl oxygens,and its diameter makes it a good fit for K+

(Figure 2a, lower left). Thus, K+ is boundabout equally well in water or in the chan-nel’s filter. For Na+, however, binding in thefilter is not favorable because the filter’s wallsare rigidly fixed and prevent close approach ofthe carbonyl oxygens to the ion (Figure 2a,lower right). The result is that Na+ has muchhigher energy in the filter than in water; itis therefore unlikely to leave water and en-ter the filter. This is precisely what is neededfor selective transport through a channel: Se-lectivity is conferred by having a low entry

www.annualreviews.org • Life Among the Axons 5

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 7: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

Energy

K+

Na+

b

K-

-

++

++

K-+

+

- ++

-+

+

-+

+KKK

-+

+

-+

+

FilterVestibuleOutside

C

-

C

-

C

-

C

-

K+ in H2O

K+ in filter

--

--

+ +

++

+

+

+

+

Na+ in filter

Na+ in H2O

a

--

+

+

- +

+

+

+

-

+

+

KK

K

C

-

C

-

C

-

C

-

Na

Na

Figure 2(a) A theory for K+ selectivity. K+ and Na+ ions are shown in water and in a carbonyl-lined filter that is agood fit for K+. The energy of K+ in water and in the filter is approximately the same. The energy ofNa+ in the filter is much higher than in water because the filter walls are too rigid to allow the carbonylsto close around this small ion. Through the use of an elementary calculation, the situation drawn wouldprovide a selectivity that is orders of magnitude greater than observed, implying that the walls are notcompletely rigid. Modified from Reference 13. (b) K+ is hydrated in the vestibule, dehydrates to enterthe filter, and rehydrates at the outer end. Modified from Reference 16. The energy diagram (lowerdiagram) shows that K+ faces no large barriers in transit through the channel. Na+ is excluded by a highbarrier (cf. the appendix to Reference 13).

barrier for the selected ion rather than an en-ergy well or, put another way, by selective ex-clusion rather than selective binding (see ap-pendix in Reference 13).

It is useful to think of the energy of a K+

ion in five positions as it passes through thefilter (Figure 2b, right to left): (1) hydratedin the water of the vestibule, (2) in the tran-sition state as it dehydrates to enter the fil-ter, (3) in the filter, (4) as it leaves the filterand rehydrates, and (5) rehydrated in the ex-ternal solution. Because the transport rate ofK+ through the channel is fast, K+ must en-counter no large energy barriers as it passesthrough. Dehydration at both ends of the fil-ter must be performed quite efficiently to keepthe barrier low. Additionally, the energy of K+

in the filter must be similar to the energy ofK+ in H2O, where the ion is bound tightly(hydration energy 79.3 kCal mol−1; see Ref-erence 15). If so, K+ binding in the selectiv-ity filter must be approximately as tight as inH2O. A clever experiment verifies this: K+ re-

mains bound in the filter for minutes whenK+ is removed from internal and external so-lutions while the channel gate is closed (17).Ion replacement, on the other hand, is rapidbecause the energy barrier for an entering, re-placement ion is low (Figure 2b).

GATING

To me the most fascinating aspect of chan-nels has always been gating: how the chan-nels respond to voltage and how their gatesopen and close in fractions of milliseconds.Any student of gating willing to study thepapers of Hodgkin & Huxley (1) has a bighead start. Here I greatly simplify theirwork. They showed that axons conduct ac-tion potentials in much the same way thatsubmarine telephone cables (the first trans-Atlantic telephone cable was laid in 1956)conduct telephone messages. The objectiveis to get high-frequency transmission alongan axon or cable that consists of a conductor

6 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 8: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

surrounded by an insulator, both immersedin salt water. The signal current flowingalong an axon/telephone cable continuallyweakens as current leaks out of the con-ductor (axoplasm/copper) through the resis-tance and capacitance of the insulator (mem-brane/insulator). These losses are particularlysevere in axons, in which the signal attenu-ates in millimeters, versus miles for a tele-phone cable. In both cases the solution is toput in boosters, which detect the weakenedsignal coming in from upstream and amplifyit back to full strength. The boosters in axonsare Na channels, and in telephone cables theyare electronic amplifiers. Axons also have Kchannels to get them repolarized and readyfor the next action potential.

Hodgkin & Huxley (1) put all this togetherin a masterful summary that deserves to liveforever. They quantitatively described sepa-rate Na and K conductances (now known tobe mediated by channels). Both conductancescan be activated over approximately the samevoltage range; activation begins near −50 mVand is complete near 0 mV. The activationprocess for both is similar, but Na conduc-tance is faster, and the action potential is wellestablished by Na+ influx before the K con-ductance turns on. Activation of the K con-ductance allows an efflux of K+ that pullsmembrane voltage back toward the restinglevel. This is made easier by a second gate onthe Na channel, termed the inactivation gate,which shuts down the Na conductance, facili-tating repolarization. Hodgkin & Huxley didnot and could not know the mechanisms in-volved in selectivity, activation, and inactiva-tion. They seemed to favor the idea that Na+

and K+ each had a selective carrier that ferried(conducted) ions across the membrane whenthe carrier was activated by control particlessensitive to membrane voltage. K+ carriers,e.g., were switched on or off by the movementof four hypothetical particles termed n, eachwith a charge of 1–2 e. The particles couldbe either positive and move outward to acti-vate the carrier, or negative and move inward;Hodgkin & Huxley had no way to know. All

four particles had to be activated to switch onthe carrier/conductance. The Na+ carrier wassimilar but with three control particles (m)rather than four. Movement of a single controlparticle (h) with a charge of ∼2 e “inactivated”the Na carrier.

The Hodgkin & Huxley scheme accountednicely for the kinetics of activation and deac-tivation of the carriers/conductances. At restall the control particles were deactivated onone side of the membrane, and they activatedone by one following depolarization, movingto the other side. Because all had to be ac-tivated to switch on the carrier, there was apronounced lag in activation, making both Naand K conductance increase sigmoidally. Incontrast, on repolarization a conducting car-rier was switched off by deactivation of a sin-gle one of its control particles; the time coursewas exponential (for sufficiently negative volt-age) and had no initial lag. The requirementfor multiple control particles to explain thekinetics gave a premonition of the multisub-unit (or multidomain) nature of the moleculesunderlying the conductances.

DESTROYING NaINACTIVATION

The most accessible of the gating functionsturned out to be Na inactivation. The findingthat K channels could be given a semblanceof inactivation by C9

+ raised the question ofwhether inactivation of Na channels had asimilar inactivation mechanism, a terrifyingdeparture from the formulation of Hodgkin& Huxley. The question received additionalimpetus from the finding that Na inactivationcould be removed without much effect on ac-tivation by perfusing an axon internally withpronase, a proteolytic enzyme mix first used inaxons to facilitate internal perfusion. An over-dose of pronase caused changes in the voltageclamp currents that were first interpreted as aneffect on K channels (18). I thought the effectlooked more like removal of Na inactivation,so Pancho, Eduardo Rojas, and I set out to see,and got, the result shown in Figure 3 (19).

www.annualreviews.org • Life Among the Axons 7

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 9: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

Figure 3Pronase destroys inactivation when applied internally. Families of superimposed INa traces are shownbefore and after pronase application, in an axon with IK blocked by TEA+. (Left) Before pronaseapplication INa increases after depolarization as the channels activate, then decreases as inactivationoccurs. On repolarization there is a small inward tail of INa through the small fraction of channels thatdid not inactivate. (Right) After pronase application, inactivation does not occur. On repolarization thereis a large inward current through the still-active channels. This current decays rapidly as the channelsdeactivate. Each trace shown represents a depolarization from −70 mV to the indicated voltage, with a2-s interval between each depolarization. INa is outward at 60 and 80 mV where the voltage drives Na+outward through the channels. Modified from Reference 19.

Activation gate:gate at the inner endof a KV or NaVchannel, which canbe open (activated)or closed(deactivated). S4segment chargemoves outward asthe gate opens,inward as it closes

After blocking of the K channels of an axonwith TEA+, the Na channels fail to inactivateafter pronase treatment, even though activa-tion and deactivation occur normally. Pronaseremoves inactivation from a channel in a sin-gle clip: The inactivation rate of those chan-nels not yet touched by pronase is completelynormal (not shown in Figure 3). The pictureof a ball-and-chain mechanism for Na inac-tivation was almost irresistible: A single in-ternal C9

+-like particle (ball) diffuses in andblocks a Na channel after its internally locatedactivation gate is partially or fully activated.Pronase cuts a peptide chain that links the ballto the inner end of a channel, leaving the acti-vation gate untouched. This seemed nice butwas distinctly unlike the model of Hodgkin &Huxley, who had never been wrong! Furtherelucidation had to wait on other evidence.

ACTIVATION

The experiments described above providedevidence that ions go through channels thathave gates at their cytoplasmic end. It wasa necessity from the physicist’s point of viewthat the gate be controlled by something anal-ogous to the charged particles (m, n) postu-

lated by Hodgkin & Huxley (see above sec-tion, “Gating”). A magnetic sensor, for exam-ple, would be sensitive only to current or todVm/dt and could be excluded. Whatever theidentity of the charges, they inevitably pro-duce a current as they move through the mem-brane. The control, or gating, current wouldpresumably be quite small and of short dura-tion because the charges would at most movefrom one side of the membrane to the other.It would be outward as the gate opens andinward as it closes. But could it be detected?With stimulus from Rojas, Pancho and I setout to find it. We suppressed all the ioniccurrent, subtracted out the linear portion ofthe capacitive current, and turned up theamplification.

After intense effort and much sortingthrough artifacts, we were rewarded by therecordings in Figure 4 (20). Trace i was invery low external Na+ and shows inward INa

preceded by a transient outward current. Af-ter adding tetrodotoxin (TTX) to block ionmovement through Na channels, we got traceii. The peak of the current is the same as inthe absence of TTX, and the current’s de-cay toward zero is visible thanks to the ab-sence of INa. Trace ii has the time course

8 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 10: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

expected for gating current: It starts early anddies away as the channels open. The currentis outward on depolarization and inward onrepolarization (not shown in Figure 4), as ex-pected for the passive movement of charge inthe electric field of the membrane. There is noway to tell from the current itself whether thegating charge is positive and moving outwardon depolarization, or whether it is negativeand moving inward.

From trace ii (Figure 4) it seems that thegate opens normally in the presence of TTX.Hille (21) had postulated that TTX binds to areceptor near the outside end of the Na chan-nel. The absence of an effect of external TTXon gating current is consistent with an inter-nal activation gate that is unaffected by TTXon the other side of the membrane. This fur-ther supports the idea that both the activationand the inactivation gates of Na channels areinside. Above all, we were happy that gatingcurrent, which had to exist, not only existedbut was detectable.

Gating current is small and sometimesfrustrating to work with. One must contin-ually remember that it is obtained by sub-tracting out the best estimate of the linearcapacitive current and that this estimate cannever be perfect. Furthermore, Ig is a com-bined signal generated by all the steps re-quired to open or close a gate, and the com-ponents are temporally overlapped and ofteninseparable. Nonetheless, there were two im-mediate lessons, one about activation and oneabout inactivation (the latter is described be-low). Before inactivation occurs, the (inward)gating current on repolarization decays at ap-proximately the same rate as INa as the acti-vation gate closes. The Hodgkin & Huxleymodel of three independent gating particlespredicted that it should be three times faster.This made it clear that information on acti-vation/deactivation would have to be assem-bled empirically and that no preexisting the-ory could help much (22, 23). If the channelwere made of protein, as seemed likely fromthe pronase results, one could think of the gat-

Figure 4Gating current (Ig) and sodium current (INa). Ig is generated by thevoltage-driven movement within the membrane of charged “particles,” nowknown to be arginine and lysine residues. This movement forces theconformational changes that open and close the activation gate. (Trace i) IKwas eliminated by removing all K+, and INa was reduced by lowering[Na+]. Capacitive current was removed by subtraction. INa is preceded byan outward gating current that starts very quickly after depolarization.(Trace ii) After adding TTX to block INa movement, Ig is seen in isolation.The same result can be achieved by complete removal of Na+. Modifiedfrom Reference 20.

Ig: gating current,generated byoutward or inwardS4 movement

Activation/deactivation: (a)Driving theactivation gatetoward open/closed.(b) Driving S4segmentsoutward/inward

KV or NaV channel:voltage-dependentK+- orNa+-selectivechannel made of fouridentical subunitpeptides (K) or onepeptide with foursimilar domains(Na). Each subunitpeptide (K) ordomain (Na) has sixtransmembranesegments, S1–S6

ing transitions not in terms of gating parti-cles but as transitions among conformationalstates of a channel protein.

A lover of speculation, I proposed themodel channel shown in Figure 5 (24). Thehypothetical protein had (a) four subunits sur-rounding a pore and (b) a gate that could openonly when all four subunits were activated.Activation and deactivation for each subunitwere governed by a zipper-like arrangementof charges (Figure 5b). Making the inside ofthe membrane positive caused relative mo-tion between the two halves of the zipper,driving a conformational change that, by anunknown mechanism, activated the subunit.This made it possible to activate the next sub-unit and so on until all subunits were acti-vated and the gate could open. When Na (25)and K channels (26, 27) were cloned someyears later, definitively proving at last that thechannels were proteins, I was pleased to seesome echoes of these speculations evident inthe amino-acid sequence.

www.annualreviews.org • Life Among the Axons 9

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 11: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

Figure 5An early speculation about channel gating. The model channel had four subunits and a gate. In eachsubunit, or in the junction between, was a zipper-like arrangement of charge (b). The small relativemotion shown (deactivated < > activated) transferred one full electronic charge across the membranebecause all the charges moved in the membrane field. This motion drove a conformational change(details unknown but easy to imagine) (a) that activated the zipper’s subunit, with activation pictureddiagramatically as outward rotation. This made it possible for the next subunit to be activated by itszipper and so on, until the gate was free to open. Modified from Reference 24.

KIR channel: a Kchannel that acts likea one-way valve,blocking outwardcurrent. KIRchannels are openedby K+ influx, causedby negative Vmand/or highextracellular K+

S4: the fourthtransmembranesegment, containingfour to eightpositively chargedamino acids

THE SEQUENCING OF Na ANDK CHANNELS

Once channels were sequenced they becamemuch more tangible. Abandoning chronologyfor simplicity, let us begin with the two fam-ilies of K channels, KV and KIR. Membersof both families are made of four identicalsubunits (peptides); KIR peptides are approxi-mately one-fourth as large as those of the KV

family. The N and C termini are inside forboth. For each KIR subunit there are 2 trans-membrane crossings (T1 and T2), yielding atotal of 8 crossings for a KIR channel, whereasa KV subunit has 6 (S1–S6), and a KV channel,thus, 24. T1 and T2 are analogous to S5 and

S6 and are now known to form the ion conduitof the channel. Segments S1–S4 are foundin voltage-sensitive channels. The very obvi-ous voltage sensor is the S4 segment, whichcontains seven positively charged amino acidsin some KV peptides, much like the positiveside of the zipper. The complementary nega-tive charges, which are an energetic necessity,are still something of a mystery, as discussedbelow.

Na channels seem to have come laterin evolution. They involve mutations thatstitched the subunit peptides together to forma very large peptide containing four domains.As in the KV family, the Na channel apparently

10 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 12: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

has an ion conduit made by apposition ofS5 and S6 from the four domains, forminga central core and four subdomains (S1–S4)surrounding the core (details are still uncer-tain). After the stitching, the domains differ-entiated, presumably to confer Na selectivity,faster gating, and inactivation.

THE DISCOVERY OF THEPORE REGION

A prescient Na channel model from Guy &Seetharamulu (28) aided the effort to asso-ciate regions of the sequence with function.These researchers identified the region be-tween S5 and S6 (in present terminology) asthe likely pore and selectivity filter. A searchfor the binding site of the potent K channelblocker charybdotoxin (29) led within a fewyears to the identification of the pore regionin K channels (30, 31) and of a signature se-quence (32) that made it possible to identifypotential K channels in the protein database.Among these was a bacterial KIR family chan-nel, KcsA (33).

PORE, FILTER, AND GATE: THEKcsA CRYSTAL STRUCTURE

MacKinnon and colleagues (34) then beganthe arduous task of crystallizing the KcsAchannel and subjecting it to X-ray analysis.The following is a brief summary of theirwork, but the original papers are a pleasurethat should not be missed. In each of the foursubunits that comprise the channel, there aretwo membrane-crossing helices, T1 and T2,that correspond to S5 and S6 in the KV familyand are so labeled here. S5 and S6 are joinedby a long linker that dips into the membrane tomake a pore helix that supports the selectivityfilter. Figure 6 shows the KcsA structure withS5 omitted. The S6 helices form a cone thatcontains the pore helix and the filter, form thewalls of a cavity, and line the inner part of thepore. They converge toward the inner side ofthe membrane; the gate region is at the con-vergence. The gate is closed in KcsA: There

Figure 6The core of a bacterial K channel in cutaway showing, from bottom to top,the gate region (closed), a hydrated K+ ion in the cavity, two dehydratedK+ ions in the filter, and a partially hydrated K+ ion at the outer mouth ofthe filter. After Reference 34.

are three points in the convergence zone atwhich methylene groups crowd together tooclosely for even a dehydrated K+ ion to pass.Because these groups are hydrophobic, eachof the three seals makes a very high energybarrier to ion movement, a very effective gate.The KV family also has a gate in this part ofthe S6, although the structure of this family’sS6, based on the sequence, is probably quitedifferent, as Yellen (35) has pointed out. TheKV gate closes tightly enough to exclude smallcations (36). The open state of KcsA has notbeen crystallized, but a reasonable model isprovided by another bacterial channel, MthK(37), in which the convergence zone is widelyspread apart and more than wide enough toadmit a hydrated K+ ion.

www.annualreviews.org • Life Among the Axons 11

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 13: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

A K+ ion going from axoplasm to externalfluid passes through the gate region, entersthe water-filled cavity, and approaches the in-ner end of the selectivity filter. To enter thefilter, the ion must dehydrate and the carbonylgroups at both ends of the filter must be care-fully arranged to lower the energy barrier fordehydration/rehydration. Once in the filter,the dehydrated ion is surrounded by carbonylsof the filter wall. Starting inside, there are fourbinding sites—4, 3, 2, and 1 (38)—and prob-ably only two are simultaneously occupied, 4and 2, or 3 and 1. An ion approaching fromthe cavity would first repel the occupying ionsfrom positions 4 and 2 to positions 3 and 1.When the entering ion occupies site 4, the ionin 3 would move to 2, and the ion in 1 wouldrehydrate as it left the filter. The occupyingions at this point are again in 4 and 2, and thecycle can repeat. The figures of Reference 39show dehydration in progress.

INACTIVATION OFNa CHANNELS

After the gating current measurements (22,23), it was reasonable to believe that Na inac-tivation resembled C9

+ block in that the volt-age dependence of block/inactivation arosefrom coupling to activation: Both C9

+ andthe inactivation particle simply diffused intotheir blocking site when the activation pro-cess made the site available. However, therewere clear differences. The Na channel canclose readily when inactivated, useful func-tionally in preventing leak during recoveryfrom inactivation. Related to this, approxi-mately one-third of gating charge remainsmobile after Na inactivation. This is unlikein TEA+ block, which immobilizes all gatingcharge in K channels. The not-immobilizedcharge moves on repolarization to close thechannel promptly (no foot in the door), avert-ing leak during recovery. Also, Na channelsneed not open fully before inactivating (40),although most inactivation in an action poten-tial occurs from the open state. There is much

recent evidence on Na inactivation, which willbe the subject of a future communication.

INACTIVATION OFK CHANNELS

Some K channels, notably those termedShaker B, inactivate rapidly. Yellen and col-leagues (40a) found that K inactivation resem-bles C9

+ block in that it occurs only when thechannel is open. Furthermore, like C9

+, theinactivation particle is subject to knock off byK+ ions moving inward through the filter af-ter repolarization, and it acts like a foot inthe door, preventing closing of the activationgate. Hoshi et al. (41) identified the inacti-vation particle as a “ball” at the N terminusof the Shaker B peptide. Four balls are at-tached by flexible chains to the inner surfaceof the channel and diffuse into the channelmouth when the gate is open. Mutations thatshorten the chain reduce the number of pos-sible conformations of the chain and thus in-crease the inactivation rate. Decreasing thenumber of balls by mutation (42) or enzy-matically (43) reduces the rate of inactiva-tion, proving that there are four inactivationparticles rather than one, as in the Na chan-nel (19). MacKinnon and colleagues (44) haveX-ray pictures of the ball within the cavity.

Aldrich and colleagues (45) described asecond type of inactivation, referring to it as Cinactivation to distinguish it from N-type (orball-and-chain) inactivation. The mechanismfor inactivation is not completely clear, butthe experiments of Yellen and colleagues (46)strongly suggest that it involves a constrictionnear the outer mouth of the selectivity filter.

S4 MOTION DURINGACTIVATION

As a voltage-dependent channel activates, gat-ing charge must move; as noted above, thereis no physical alternative. Common sense andsome evidence tie this to motion of the S4segments (47, 48). The original estimate ofHodgkin and Huxley was that approximately

12 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 14: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

six electronic charges move to activate a chan-nel, but current estimates for the Shaker Kchannel are more than twice that (49, 50). Pre-cisely how the motion occurs is not yet clear.Simplest conceptually are the zipper model(Figure 5), its helical screw variant (51), andthe paddle model (52). All three models haveproblems. The negative counterions for thefirst two models have never been identified insufficient number. The absence of counteri-ons in the paddle model raises, in my mind, theinsurmountable energetic problem of puttingmultiple arginines (charged and hydrophilic)into lipid, at a cost of approximately 15 kCalmol−1 for each arginine (53).

A more complex model postulates crevicesthat penetrate around the S4 segments and gopart way through the membrane from bothsides (54, 55). These crevices admit anionsfrom the internal and external solution toserve as counterions. A new crystal structureof Kv1.2 is helpful but not definitive because itshows the S4 only in one state, probably open(56). There are no obvious crevices. How theS4 moves during gating remains a challengingquestion.

HOLDING THE ACTIVATIONGATE OPEN

Yellen and colleagues (57) found a thought-provoking way of holding open the gate of aKV channel even when Vm dictates it shouldbe closed. They introduced a cysteine residue(replacing a valine) in the gating region andadded cadmium. The cadmium binds to theintroduced cysteine and forms a bridge to anative histidine 10 residues further down andin a different subunit. Remarkably, the gate istied open, unable to close on repolarization.

HOLDING THE GATE CLOSEDWITH 4-AMINO PYRIDINE

4AP+ is a K channel blocker with an inter-esting mode of action (58, 59). It enters andleaves the vestibule from inside when the ac-tivation gate is open, as do TEA+ and C9

+.

Vm: voltage at theinside of amembrane, withoutside defined as0 mV

Its blocking affinity is very low when the gateis open (KD ∼ 55 mM) but high when it isclosed (KD ∼ 30 μM). Presumably, 4AP+ fitsloosely in the vestibule in the open state butthe vestibule contracts around it in the closedstate. 4AP+ then serves as an adhesive to keepthe vestibule contracted. The gate can open,but slowly and with low probability. In short,4AP+ enters through the open gate and pullsthe gate closed behind it. This is the oppositeof TEA+, which tends to hold the gate open,like a foot in the door. Interestingly, when4AP+ is trapped in the vestibule by the closedgate, the S4 segments of all subunits can movein and out almost normally in response tovoltage changes (58, 60). Total charge move-ment, however, is reduced by 5% or 10% be-cause the final opening step and its associatedcharge movement are unlikely to occur. Also,the Q-V curve, which relates gating chargemovement and Vm, is shifted to the right by5–10 mV, reflecting the small amount of extrawork that is required to move the S4s outwardwith 4AP+ trapped in the vestibule. This facttells a good deal about the coupling of S4 tothe gate.

COUPLING OF S4 TO THE GATE

Open probability Po can vary from ∼10−9 at−80 mV (49, 50) to ∼0.7 at +20 mV. Do theS4s pull the gate open when they are acti-vated? Or do they lock the gate shut when theyare deactivated? Or both? TEA+ and 4AP+

interact with the gate, and studying these in-teractions is quite helpful in deciding suchquestions. First, consider deactivation of thechannel. TEA+ in the vestibule holds the gateopen and freezes S4 motion (61). Figure 7a

shows a mechanical explanation: Flanges onthe S6 segments prevent the S4s from mov-ing downward as long as the gate is held openby TEA+. (K+ and Rb+ in the vestibule alsoimpede closing, suggesting that the vestibulecontains only water when the gate closes nor-mally.) Apparently there is rigid coupling be-tween S4 deactivation and gate closing: TheS4 cannot move inward while the gate is open.

www.annualreviews.org • Life Among the Axons 13

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 15: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

++++++

+

++

++

++

+

++

++

++

+

++++++

+

:4AP+

Lax gating spring Tensed

b

S4

S6

++++++

+

++

++

++

+

Outside

++++++

+

++

++

++

+

:TEA+a

Flange

Figure 7Thinking about the coupling of S4 and the activation gate. (a) When TEA+ is in the vestibule, thechannel gate is held open, and the S4s are immobilized in activated position. Projections (flanges) on theS6 prevent the S4s from moving inward as long as the gate is open. When TEA+ leaves the vestibule, thegate can close, and negative internal voltage drives the S4s inward to the deactivated position. Thedeactivated S4s lock the gate shut. (b) 4AP+ in the vestibule stabilizes the closed state by an energyroughly equivalent to that of a hydrogen bond, thus making the gate unlikely to open. Experimentally,with 4AP+ in the vestibule the S4s can move almost normally, from deactivated (left) to activated (right).Thus, an elastic connection such as the gating spring diagrammed must be very compliant. Arrows showthe forces exerted on the gate by the gating springs (via imaginary pulleys) when they have been tensedby outward movement of the S4s. The forces are quite small. Modified from Reference 63.

Does S4 motion pull the S6 open by meansof some so-far-unknown gating spring (62)?4AP+ holds the gate closed but impedes S4motion only slightly (see above). This sug-gests that any link between the S4 and thegate is very weak and easy to stretch (63). InFigure 7b, a hypothetical S4-S6 connection isdrawn as an S4-linked gating spring attachedvia a pulley (to get the force in the right direc-tion) and an inelastic cord to S6. If the springis very stiff, the S4s cannot move when 4AP+

holds the gate shut. At the other extreme, if thespring is completely compliant, no extra workwill be required to move the S4s, and the cen-ter point of the Q-V curve will be unaffected.As noted above, the experimental curve showsonly a small displacement, signifying enoughenergy to change Po by a factor of less than10, many orders of magnitude less than theexperimental range of Po. This strongly sug-gests that the main function of the S4 is tolock the gate closed and that S4 makes onlya minor contribution to pulling it open. Thisidea gains some support from the crystal struc-ture of Kv1.2 (64), which shows that the S4-S5 linker is in a position that would appear

to lock the gate shut when S4 is deactivated.Mutational alterations in this region of the S6and the facing region of the S4-S5 linker pro-foundly alter gating (65, 66).

WHY IS THE ACTIVATION GATEALL OR NONE (USUALLY)?

The gate-opening step is all or none and mustinvolve coordinated movement in the S6 seg-ment of all four subunits. This is clear fromsingle-channel records, in which the openingis a single jump that cannot be resolved intosmaller jumps. One factor almost surely is thatthe gate region is either large enough to ad-mit a hydrated K+ ion, or it is not: Even par-tially dehydrating a K+ in a hydrophobic en-vironment like the gate region would createa large energy barrier and make ion flux toosmall to detect. A related but more difficultquestion is why all four S4s must be activatedto open the gate. A possible answer is thatthe gate region’s hydrophobic lining must bepulled in four directions to effectively breakthe hydrophobic bonds that almost certainlyhelp to hold it closed. An analogy would be

14 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 16: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

the need to pull in at least two directions toopen a plastic sack whose sides are clingingtogether electrostatically (63).

SUMMARY AND CONCLUSIONS

Ion channels have gone from concept toreality in the past 50 years. They havebeen probed, sequenced, and crystallized, and

much is now understood about their generalproperties, selectivity, and mechanisms of gat-ing. They are essential in virtually all physi-ological regulatory mechanisms, perception,memory, and thought. Their genetic defectsare involved in an ever-growing number ofdiseases, and they offer enormous possibilitiesfor therapeutic intervention. Ah, to be able tostart over again!

LITERATURE CITED

1. The foundationof ion conduction.Difficult to readbecause of olderconventions butrewarding.

1. Hodgkin AL, Huxley AF. 1952. A quantitative description of membrane cur-rent and its application to conduction and excitation in nerve. J. Physiol. (Lond.)

117:500–442. Armstrong CM. 1968. Monosynaptic activation of pyramidal cells in area 18 by optic

radiation fibers. Exp. Neurol. 21:413–283. Armstrong CM. 1968. The inhibitory path from the lateral geniculate body to the optic

cortex in the cat. Exp. Neurol. 21:429–414. Noble D. 1962. A modification of the Hodgkin-Huxley equations applicable to Purkinje

fiber action and pace-maker potentials. J. Physiol. (Lond.) 160:317–525. Katz B. 1949. Les constants electriques de la membrane du muscle. Arch. Sci. Physiol.

2:285–996. Tasaki I, Hagiwara S. 1957. Demonstration of two stable potential states in the squid

giant axon under tetraethylammonium chloride. J. Gen. Physiol. 40:859–857. Armstrong CM, Binstock L. 1965. Anomalous rectification in the squid giant axon injected

with tetraethylammonium chloride. J. Gen. Physiol. 48:859–728. Armstrong CM. 1968. Time course of TEA+-induced anomalous rectification in squid

giant axons. J. Gen. Physiol. 21:413–503

9. Elucidates thegross architectureof K channels.Among my papers,this is my favorite.

9. Armstrong CM. 1971. Interaction of tetraethylammonium ion derivatives with thepotassium channels of giant axons. J. Gen. Physiol. 59:413–37

10. Unrivaled as aninstructional andreference book onchannels.

10. Hille B. 2001. Ion Channels of Excitable Membranes. Sunderland, MA: Sinauer.722 pp. 3rd ed.

11. Armstrong CM, Hille B. 1972. The inner quaternary ammonium ion receptor in potas-sium channels of the node of Ranvier. J. Gen. Physiol. 59:388–400

12. Cotton FA, Wilkinson G, Murillo CA, Bochmann M. 1999. Advanced Inorganic Chemistry.New York: John Wiley

13. An awful title,but the selectivitytheory is good.

13. Bezanilla F, Armstrong CM. 1972. Negative conductance caused by the entry ofsodium and cesium ions into the potassium channels of squid axons. J. Gen. Physiol.60:588–608

14. Mullins LJ. 1960. An analysis of pore size in excitable membranes. J. Gen. Physiol. 43:105–17

15. Robinson RA, Stokes RH. 1965. Electrolyte Solutions. London: Buttersworth16. Armstrong CM. 1975. Ionic pores, gates and gating currents. Quart. Rev. Biophys. 7:179–

21017. Gomez-Lagunas F. 1997. Shaker B K+ conductance in Na+ solutions lacking K+ ions: a

remarkably stable nonconducting state produced by membrane depolarizations. J. Physiol.(Lond.) 499:3–15

www.annualreviews.org • Life Among the Axons 15

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 17: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

18. Rojas E, Atwater I. 1967. Blocking of potassium currents by pronase in perfused giantaxons. Nature 215:850–52

19. Armstrong CM, Bezanilla FM, Rojas F. 1973. Destruction of sodium conductance inac-tivation in squid axons perfused with pronase. J. Gen. Physiol. 62:375–91

20. Armstrong CM, Bezanilla F. 1973. Currents related to movement of the gating particlesof the sodium channels. Nature 242:459–61

21. Hille B. 1975. The receptor for tetrodotoxin and saxitoxin. A structural hypothesis.Biophys. J. 15:615–19

22, 23. Thesepapers provide astill-useful theoryof coupledactivation-inactivation.

22. Bezanilla F, Armstrong CM. 1977. Inactivation of the sodium channel. I. Sodiumcurrent experiments. J. Gen. Physiol. 70:549–66

23. Armstrong CM, Bezanilla F. 1977. Inactivation of the sodium channel. II. Gatingcurrent experiments. J. Gen. Physiol. 70:567–90

24. Armstrong CM. 1981. Sodium channels and gating currents. Physiol. Rev. 61:645–83

25. The Holy Grail,part one: theamino-acidsequence of a Nachannel.

25. Noda M, Shimuzu S, Tanabe T, Takai T, Kayano T, et al. 1984. Primary structureof Electrophorus electricus sodium channel deduced from cDNA sequence. Nature

312:121–1726. Tempel BL, Papazian DM, Schwarz TL, Jan YN, Jan LY. 1987. Sequence of a probable

potassium channel component encoded at Shaker locus of Drosophila. Science 237:770–75

27. Pongs O, Kecskemethy N, Muller R, Krah-Jentgens I, Baumann A, et al. 1988. Shaker en-codes a family of putative potassium channel proteins in the nervous system of Drosophila.EMBO J. 7:1087–96

28. Guy HR, Seetharamulu P. 1986. Molecular model of the action potential sodium channel.Proc. Natl. Acad. Sci. USA 83:508–12

29. Miller C, Moczydlowski E, LaTorre R, Phillips M. 1985. Charybdotoxin, a protein in-hibitor of single Ca2+-activated K+ channels from mammalian skeletal muscle. Nature313:316–18

30. Yellen G, Jurman ME, Abramson T, MacKinnon R. 1991. Mutations affecting inter-nal TEA blockade identify the probable pore-forming region of a K+ channel. Science251:939–42

31. Hartmann HA, Kirsch GE, Drewe JA, Tagliatella M, Joho RH, Brown AM. 1991. Ex-change of conduction pathways between two related K+ channels. Science 251:942–44

32. Heginbotham L, Lu Z, Abramson T, MacKinnon R. 1994. Mutations in the K+ channelsignature sequence. Biophys. J. 66:1061–67

33. Schrempf H, Schmidt O, Kummerlen R, Hinnah S, Muller D, et al. 1995. A prokaryoticpotassium ion channel with two predicted transmembrane segments from Streptomyceslividans. EMBO J. 14:5170–78

34. The Holy Grail,part two: thestructure of asimple K channel.

34. Doyle DA, Carbal JM, Pfuetzner RA, Kuo A, Gullbis JM, et al. 1998. The structureof the potassium channel: molecular basis of K+ conduction and selectivity. Science

280:69–7735. Yellen G. 2002. The voltage-gated potassium channels and their relatives. Nature 419:35–

4236. del Camino D, Yellen G. 2001. Tight steric closure at the intracellular activation gate of

a voltage-gated K+ channel. Neuron 32:649–5637. Jiang Y, Lee A, Chen J, Cadene M, Chait BT, MacKinnon R. 2002. The open pore

conformation of potassium channels. Nature 417:523–2638. Morais-Cabral JH, Zhou Y, MacKinnon R. 2001. Energetic optimization of ion conduc-

tion rate by the K+ selectivity filter. Nature 414:37–42

16 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 18: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

39. Unbelievable!You can see K+ions disrobing.

39. Zhou Y, Morais-Cabral JH, Kaufman A, MacKinnon R. 2001. Chemistry of ioncoordination and hydration revealed by a K+ channel-Fab complex at 2.0 A reso-lution. Nature 414:43-48

40. Bean BP. 1980. Sodium channel inactivation in the crayfish giant axon. Must channelsopen before inactivating? Biophys. J. 35:595–614

40a. Demo SD, Yellen G. 1991. The inactivation gate of the Shaker K+ channel behaves likean open-channel blocker. Neuron 7:743–53

41. Hoshi T, Zagotta WN, Aldrich RW. 1990. Biophysical and molecular mechanisms ofShaker potassium channel inactivation. Science 250:533–38

42. MacKinnon R, Aldrich RW, Lee AW. 1993. Functional stoichiometry of Shaker potas-sium channel inactivation. Science 262:757–59

43. Gomez-Lagunas F, Armstrong CM. 1995. Inactivation in Shaker K+ channels: a test forthe number of inactivating particles on each channel. Biophys. J. 68:89–95

44. Zhou M, Morais-Cabral JH, Mann S, MacKinnon R. 2001. Potassium channel receptorsite for the inactivation gate and quaternary amine inhibitors. Nature 411:657–61

45. Hoshi T, Zagotta WN, Aldrich RW. 1991. Two types of inactivation in Shaker K+

channels: effects of alterations in the carboxy-terminal region. Neuron 7:547–5646. Yellen G, Sodickson D, Chen TY, Jurman ME. 1994. An engineered cysteine in the

external mouth of a K+ channel allows inactivation to be modulated by metal binding.Biophys. J. 66:1068–75

47. Stuhmer W, Conti F, Suzuki H, Wang XD, Noda M, et al. 1989. Structural parts involvedin activation and inactivation of the sodium channel. Nature 339:597–603

48. Yang N, Horn R. 1995. Evidence for voltage-dependent S4 movement in sodium chan-nels. Neuron 15:213–18

49. Hirschberg B, Rovner A, Lieberman M, Patlak J. 1995. Transfer of twelve charges isneeded to open skeletal muscle Na+ channels. J. Gen. Physiol. 106:1053–68

50. Islas L, Sigworth F. 1999. Voltage sensitivity and gating charge in Shaker and Shab familypotassium channels. J. Gen. Physiol. 114:723–41

51. Catterall WA. 1995. Structure and function of voltage-gated ion channels. Annu. Rev.Biochem. 64:493–531

52. Jiang Y, Ruta V, Chen J, Lee A, MacKinnon R. 2003. The principle of gating chargemovement in a voltage-dependent K+ channel. Nature 423:42–48

53. Radzicka A, Wolfenden R. 1988. Comparing the polarities of the amino-acids: side-chaindistribution coefficients between the vapor phase, cyclohexaon, 1-octanol, and neutralaqueous solution. Biochemistry 27:1670–77

54. Ahern CA, Horn R. 2004. Specificity of charge-carrying residues in the voltage sensor ofpotassium channels. J. Gen. Physiol. 123:205–16

55. Chanda B, Asamoah OK, Blunck R, Roux B, Bezanilla F. 2005. Gating charge displace-ment in voltage-gated ion channels involves limited transmembrane movement. Nature436:852–56

56. Long SB, Campbell EB, MacKinnon R. 2005. Crystal structure of a mammalian voltage-dependent Shaker family K+ channel. Science 309:897–903

57. Holmgren M, Shin KS, Yellen G. 1998. The activation gate of a voltage-gated K+ channelcan be trapped in the open state by an intersubunit metal bridge. Neuron 21:617–21

58. Armstrong CM, Loboda A. 2001. A model for 4-aminopyridine action on K channels.Similarities to TEA+ action. Biophys. J. 81:895–904

59. Del Camino D, Kanevsky M, Yellen G. 2005. Status of the intracellular gate in theactivated-not-open state of shaker K+ channels. J. Gen. Physiol. 126:419–28

www.annualreviews.org • Life Among the Axons 17

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 19: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

ANRV300-PH69-01 ARI 8 January 2007 14:52

60. Loboda A, Armstrong CM. 2001. Resolving the gating charge movement associated withlate transitions in K channel activation. Biophys. J. 81:905–16

61. Perozo E, Papazian DM, Stefani E, Bezanilla F. 1992. Gating currents in Shaker K+

channels. Implications for activation and inactivation models. Biophys. J. 62:160–6862. Howard J, Hudspeth AJ. 1987. Mechanical relaxation of the hair bundle mediates adap-

tation in mechanoelectrical transduction by the bullfrog’s saccular hair cell. Proc. Natl.Acad. Sci. USA 84:3064–68

63. Armstrong CM. 2003. Voltage-gated K channels. Sci. STKE 2003(188):RE1064. Long B, Campbell EB, MacKinnon R. 2005. Voltage sensor of Kv1.2: structural basis of

electromechanical coupling. Science 309:903–865. Lu Z, Klem AM, Ramu Y. 2002. Mechanism of rectification in inward-rectifier K+ chan-

nels. J. Gen. Physiol. 120:663–7666. Hackos DH, Chang TH, Swartz KJ. 2002. Scanning the intracellular S6 activation gate

in the shaker K+ channel. J. Gen. Physiol. 119:521–32

18 Armstrong

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 20: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

AR300-FM ARI 22 January 2007 14:24

Annual Review ofPhysiology

Volume 69, 2007Contents

FrontispieceClay M. Armstrong � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �xx

PERSPECTIVES, David L. Garbers, Editor

Life Among the AxonsClay M. Armstrong � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1

CARDIOVASCULAR PHYSIOLOGY, Jeffrey Robbins, Section Editor

Mitochondrial Ion ChannelsBrian O’Rourke � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 19

Preconditioning: The Mitochondrial ConnectionElizabeth Murphy and Charles Steenbergen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 51

CELL PHYSIOLOGY, David E. Clapham, Section Editor

Iron HomeostasisNancy C. Andrews and Paul J. Schmidt � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 69

Transporters as ChannelsLouis J. DeFelice and Tapasree Goswami � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 87

ECOLOGICAL, EVOLUTIONARY, AND COMPARATIVEPHYSIOLOGY, Martin E. Feder, Section Editor

Hypoxia Tolerance in Mammals and Birds: From the Wildernessto the ClinicJan-Marino Ramirez, Lars P. Folkow, and Arnoldus S. Blix � � � � � � � � � � � � � � � � � � � � � � � � �113

Hypoxia Tolerance in Reptiles, Amphibians, and Fishes: Life withVariable Oxygen AvailabilityPhilip E. Bickler and Leslie T. Buck � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �145

ENDOCRINOLOGY, Kathryn B. Horwitz, Section Editor

Integration of Rapid Signaling Events with Steroid Hormone ReceptorAction in Breast and Prostate CancerCarol A. Lange, Daniel Gioeli, Stephen R. Hammes, and Paul C. Marker � � � � � � � � � �171

xiii

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 21: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

AR300-FM ARI 22 January 2007 14:24

Nuclear Receptor Structure: Implications for FunctionDavid L. Bain, Aaron F. Heneghan, Keith D. Connaghan-Jones,

and Michael T. Miura � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �201

GASTROINTESTINAL PHYSIOLOGY, John Williams, Section Editor

Regulation of Intestinal Cholesterol AbsorptionDavid Q.-H. Wang � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �221

Why Does Pancreatic Overstimulation Cause Pancreatitis?Ashok K. Saluja, Markus M. Lerch, Phoebe A. Phillips, and Vikas Dudeja � � � � � � � � � �249

NEUROPHYSIOLOGY, Richard Aldrich, Section Editor

Timing and Computation in Inner Retinal CircuitryStephen A. Baccus � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �271

Understanding Circuit Dynamics Using the Stomatogastric NervousSystem of Lobsters and CrabsEve Marder and Dirk Bucher � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �291

RENAL AND ELECTROLYTE PHYSIOLOGY, Gerhard H. Giebisch,Section Editor

Molecular Mechanisms of Renal Ammonia TransportI. David Weiner and L. Lee Hamm � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �317

Phosphatonins and the Regulation of Phosphate HomeostasisTheresa Berndt and Rajiv Kumar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �341

Specificity and Regulation of Renal Sulfate TransportersDaniel Markovich and Peter S. Aronson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �361

RESPIRATORY PHYSIOLOGY, Richard C. Boucher, Jr., Section Editor

Overview of Structure and Function of Mammalian CiliaPeter Satir and Søren Tvorup Christensen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �377

Regulation of Mammalian Ciliary BeatingMatthias Salathe � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �401

Genetic Defects in Ciliary Structure and FunctionMaimoona A. Zariwala, Michael R. Knowles, and Heymut Omran � � � � � � � � � � � � � � � � � �423

SPECIAL TOPIC, β-ARRESTINS, Robert J. Lefkowitz, Special Topic Editor

Regulation of Receptor Trafficking by GRKs and ArrestinsCatherine A.C. Moore, Shawn K. Milano, and Jeffrey L. Benovic � � � � � � � � � � � � � � � � � � � �451

β-Arrestins and Cell SignalingScott M. DeWire, Seungkirl Ahn, Robert J. Lefkowitz, and Sudha K. Shenoy � � � � � �483

xiv Contents

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.

Page 22: Annu. Rev. Physiol. 2007.69:1-18. Downloaded from ...AR300-FM ARI 22 January 2007 14:24 Annu. Rev. Physiol. 2007.69:1-18. Downloaded from arjournals.annualreviews.org by CNRS-multi-site

AR300-FM ARI 22 January 2007 14:24

Physiological Roles of G Protein–Coupled Receptor Kinases andArrestinsRichard T. Premont and Raul R. Gainetdinov � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �511

Stop That Cell! β-Arrestin-Dependent Chemotaxis: A Tale ofLocalized Actin Assembly and Receptor DesensitizationKathryn A. DeFea � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �535

Regulation of Receptor Tyrosine Kinase Signaling by GRKs andβ-ArrestinsChristopher J. Hupfeld and Jerrold M. Olefsky � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �561

Indexes

Cumulative Index of Contributing Authors, Volumes 65–69 � � � � � � � � � � � � � � � � � � � � � � � �579

Cumulative Index of Chapter Titles, Volumes 65–69 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �582

Errata

An online log of corrections to Annual Review of Physiology chapters (if any, 1997 tothe present) may be found at http://physiol.annualreviews.org/errata.shtml

Contents xv

Ann

u. R

ev. P

hysi

ol. 2

007.

69:1

-18.

Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

CN

RS-

mul

ti-si

te o

n 09

/23/

08. F

or p

erso

nal u

se o

nly.