HAL Id: hal-01740172
https://hal.sorbonne-universite.fr/hal-01740172
Submitted on 21 Mar 2018
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
The electrical significance of axon location diversity
Maarten H. P. Kole, Romain Brette
To cite this version:
Maarten H. P. Kole, Romain Brette. The electrical significance of axon location diversity. Current
Opinion in Neurobiology, Elsevier, 2018, 51, pp.52-59. �10.1016/j.conb.2018.02.016�. �hal-01740172�
The
electrical
significance
of
axon
location
diversity
Maarten
HP
Kole
1,2and
Romain
Brette
3Theaxoninitialsegment(AIS)isauniquedomainofthe proximalaxonservingcriticalelectricalandstructuralroles includingtheinitiationofactionpotentialsandmaintenanceof cellularpolarity.Recentexperimentalandtheoreticaladvances demonstratethattheanatomicalsiteforinitiationisremarkably diverse.TheAISlocationvariesnotonlyaxially,alongtheaxon, butaxonsalsoemergevariablyfromeitherthesomaor proximaldendrites.Here,wereviewtheevidencethatthe diversityofAISandaxonlocationhasasubstantialimpacton theelectricalpropertiesandspeculatethattheanatomical heterogeneityofaxonlocationsexpandssynapticintegration withincelltypesandimprovesinformationprocessinginneural circuits.
Addresses
1DepartmentofAxonalSignalling,NetherlandsInstitutefor
Neuroscience,RoyalNetherlandsAcademyofArtsandSciences (KNAW),Meibergdreef47,1105BAAmsterdam,theNetherlands
2CellBiology,FacultyofScience,UtrechtUniversity,3584CHUtrecht,
TheNetherlands
3SorbonneUniversite´s,UPMCUnivParis06,INSERM,CNRS,Institut
delaVision,17rueMoreau,75012Paris,France
Correspondingauthor:Kole,MaartenHP(m.kole@nin.knaw.nl)
CurrentOpinioninNeurobiology2018,51:52–59
ThisreviewcomesfromathemedissueonCellularneuroscience EditedbyJuanBurroneandErikaHolzbaur
https://doi.org/10.1016/j.conb.2018.02.016
0959-4388/ã2018TheAuthors.PublishedbyElsevierLtd.Thisisan openaccessarticleundertheCCBY-NC-NDlicense( http://creative-commons.org/licenses/by-nc-nd/4.0/).
Introduction
Neurons,thecellularcomputationalunitsof thecentral nervoussystem,aremorphologically polarizedcells and segregated into diverse functional compartments. A uniqueandcriticalsiteoftheproximalaxonof mamma-lianneuronsistheaxoninitialsegment(AIS),clustering high densities of voltage-gated sodium (Nav) channels [1].The AIS definesthelocationfrom whereall action potentials(APs)are initiatedand propagate bidirection-ally,towardsthepresynapticterminalsandbackintothe soma and dendrites (reviewed in [2,3]).Understanding the structural and molecular properties of the site of initiationisof fundamentalimportance as itaffectsthe electricalfunctionofneurons.ThediscoverythattheAIS cytoskeleton and associated ion channel clustering change in response to network activity has led to the
conceptthattheAISmayservetohomeostaticallyscale intrinsic excitability [4,5,6]. In addition, evidence is accumulating that the morphological diversity of the AIS and its functional impact are greater than appre-ciated; not only is the AIS location variable along the axon branch, but axons themselves may also emerge eitherfromthesomaorfromdendrites,andinsomecell typesevenuptohundredsofmicrometresfromthesoma. Here,wereviewrecentelectrophysiologicaland theoret-icaladvancesintothisanatomicalheterogeneityfuelling theideathatspikeinitiationlocationmaynotonlyimpact excitability,butalsothebackpropagationofaction poten-tialsandsynapticintegration.
Location,
location,
location
ThepositionoftheAISalongtheaxonvaries substan-tiallybetweenneuronalcelltypes[7],withincelltypes andacross early neuronal development [8,9].In addi-tion,alreadysincetheseminalstudiesofRamo´nyCajal, it is known thataxons havevariable sites fromwhere theyemerge, often fromthe soma but also from den-drites,such asthe parallel fibresof cerebellargranule cells or the axons fromthe avian optical lobe neuron [10,11] (Figure 1a). Although neurons with dendritic axonsweregenerallybelievedtoberare,detailed mor-phologicalreconstructionsshowthattheyexistacrossa wide variety of cell types including the hippocampal dentategyrusbasketcellsandsomatostatin interneur-ons[12,13],inapproximately75%ofthepopulationof dopaminergiccellsofthesubstantianigra[14–16]andin 40%ofthecerebellargranulecells[11,17](Figure1a–d). Furthermore, staining for ankyrin-G or bIV-spectrin, markers for the AIS structure, combined with Cre recombinase reporter mice made it possible to obtain detailedinformationabouttheAIS locationwithinthe axon, as well as the axon location itself, across large populationsofidentifiedcelltypes.Alsothisrecentbody of work revealed that dendritic axons are present in approximately60%ofthehippocampalpyramidal neu-rons,30%ofthethick-tuftedsomatosensorylayer5 pyra-midalneurons,40%ofcorticalinterneuronbasketcells and60% ofMartinotti cells [7,18,19] (Figure 1e–f). Giventheabundanceofdendriticallytargetedaxonsthe questionarises:whatcouldbetheroleofsuch asymme-tryinaxonposition,andmorebroadlyofthediversityof AISlocation?Asearlyasin1898Ramo´nyCajalproposed thatincellswithdendriticaxons,“ ... thesoma,orcell bodydoesnotalwaystakepartintheconductionofthenerve impulseswhicharereceived.The afferentwaveis sometimes propagateddirectlyfromthedendritestotheaxon”(reviewed in[20]).Recentpatch-clamprecordings,combinedwith morphologicalreconstructionsandtheoreticaladvances,
are shedding more light on the excitability and the associatedcurrentflow.
AIS
location
diversity
and
intrinsic
excitability;
the
theory
Inordertounderstandtheelectricalimpactofthespatial variationinaxonlocation,itisimperativetoconsiderthe cellularmorphologyfromabiophysicalperspective.Ina
situationwheretheinitiationsiteoftheactionpotentialis located to a thinaxon,electrotonically closeto amuch larger somatodendritic compartment, resistive coupling theoryapplies[21,22].Thesodium(Na+)current gener-atedintheAISflowsprimarily towardsthesoma,which thus acts as a current sink, and subsequently exits as capacitive current through the large somatodendritic membrane(Figure2a).Becausetheintracellularmedium
TheelectricalsignificanceofaxonlocationdiversityKoleandBrette 53
Figure1
(a)
(d) (e) (f)
(c) (b)
Current Opinion in Neurobiology
Dendriticaxonsareacommonfeatureofvariousneurontypes.(a)Exampleofadendriticorigin(redopenarrow)oftheparallelfibre(greenclosed arrow)fromacerebellargranulecellprojectingtoaPurkinjecell.BlackarrowsdepicttheaxipetalcurrentflowdrawnbyRamo´nyCajal.(b)A substantianigradopaminergicneuronwithdendriticaxonillustratedwithpatchpipetterecordinglocations.Theaxon-somadistanceofthiscellis 215mm.(c)Distalaxonorigin(100mmfromthesoma)inahippocampaldentategyrusbasketcell.(d)Exampleofaratparietalcortexlayer2/3 pyramidalneuronwithaxonemergingfromthickbasaldendrite.(e)Immunofluorescenceofabiocytin-filledratthick-tuftedlayer5pyramidal neuron(red)withdendriticoriginoftheAIS(bIV-spectrinidentified,green).Filledarrowsmarktheaxons.Openarrowsindicatetheaxon-carrying dendrites.(f)ImmunofluorescenceimageofaThy1+mousehippocampalCA1pyramidalneuron(red)withAnkyrin-G(green).Scalebars;b, 100mm;c–f,20mm.
isresistive,theaxialcurrentfollowsOhm’slaw:avoltage gradient develops between the soma and axonal site, equal to the product of the input current and axial coupling resistance (Ra) between the two sites, DV=RaI (Figure 2a) [21]. Ra is furthermore deter-minedbythegeometryoftheaxonalordendriticbranch inbetweenthesomaandAIS,proportionaltothedistance andinverselyproportionaltothecross-sectionarea (diam-eter squared). This simplified model predicts a strong influenceof thegeometryof theAIS andneighbouring dendritic sites and makesseveral predictionsaboutthe functionalimpactthatwewill reviewin detailbelow. The amount of Na+ current needed for agiven axonal depolarizationisinverselyproportionaltoaxialresistance (I=DV/Ra).Therefore,aneuronwithamoredistalAISor thinneraxonshouldeitherbemoreexcitable,reflectedin
ahyperpolarizationofthevoltagethresholdandreduced rheobase (current to threshold) for AP generation or require a lower Na+ channel density for a given level ofexcitability(Figure2bandc).Thistheoretical predic-tionhasalso beenconfirmedin detailed multi-compart-mentalcomputationalmodelsinwhichmembrane prop-erties can be maintained constant while axonal compartments are varied in their location [18,22,23] and is consistent with the notion that spike initiation occurs at the distal end of the AIS due to its larger electrotonicisolation[24,25].
Notably, some simulation studies showed that a distal location of the AIS can reduce neuronal excitability [5,23,26].Theoretically,thiscanhappen iftheproximal axonand thesomatodendritic compartment are compa-rableinsize,sothatthesomaisnolongeracurrentsink fortheAIS[22].However,incorticalpyramidalneurons, theaxon diameter scaleswith soma size,and isalways much smaller compared to the soma [27,28]. Another situation where excitability could be reduced with increasingsoma-AIS distance iswhenahyperpolarizing currentislocallystronglyactivatedattheAIS,for exam-ple by Kv1 or Kv7 channels [21,26]. Direct recording fromtheAISandestimatesofconductancedensityshow thatbelow thespike threshold theslowly-activated M/ Kv7currentandfast-activatingKv1currentare,however, relativelysmallcomparedtotheinwardNa+current[29– 31].Similar differencesbetween Kv7and Nav conduc-tancedensitieswerefoundintheaviannucleus magno-cellularis[32].Thusforthesecellstheorypredictsthatif allmembranepropertiesremainconstant,adistalshiftof theAISmakesthecellmoreexcitable.Finally,forin vivo-like fast excitatory postsynaptic potentials (EPSPs), a significant frequency-dependent voltage attenuation couldoccuralongtheaxon.Inthiscase,largedistalshifts of the AIS could reduce excitability, resulting in a U-shapedrelationshipbetweenAISpositionandexcitability [33].Whatis theexperimentalevidence for the theo-reticalpredictions?
AIS
location
diversity
and
intrinsic
excitability;
the
experimental
observations
Inagreementwiththerole oftheAISin intrinsic excit-ability,geneticallydisruptingthedevelopmentofanAIS or pharmacologically inactivatingthe AISNa+ channels substantially depolarizes the AP threshold (>10mV), increases the rheobase and reduces the firing rate [30,34,35].In accord,anumberof experimentalstudies showedthatincreasingordecreasingtheAISlengthraises or reduces the intrinsic excitability, respectively, by affecting voltage thresholdand firing rate (reviewed in [36,37]).Ontheotherhand,experimentallyisolatingand interpretingthespecific roleoftheAISoraxonlocation hasproventobesubstantiallymorechallengingandthe interpretationof the results iscontentious.The studies observingadistalrelocationoftheAISfoundadecreased
Figure2 1200 20 –64 –50 800 400 0 0 0 μm Axon ΔV = Ra x I 5 μm 45 μm V Soma (mV) I axial (nA) ICap Ra VSoma dV/ d t (V/s) V Soma (mV) 0 60 –60 –400 Dendrites (a) (b) (c)
Current Opinion in Neurobiology
Resistivecouplingbetweentheaxonandsoma.(a)Left,Circulationof currententeringtheaxonemergingfromadendrite.Thecurrent mainlyflowstothesoma,acurrentsink,alongaresistivepath(Ra:
axialresistance).Right,ByOhm’slaw,avoltagegradientforms betweensomaandinjectionsite,proportionaltoRa.(b)Phase-plane
plot(dV/dtversusV)ofsomaticactionpotentialsinasimplemodel withtheAISat3differentlocations(0,5and45mm),revealsalower voltagethresholdwithincreasingsoma-axondistance.(c)Axialcurrent enteringthesomaasafunctionofsomaticpotential:thethresholdis lowerandthecurrentissmallerforlargersoma-axondistance. bandc,Adaptedfrom[22].
intrinsicexcitabilitywhichinsomecasescorrelatedwith theextentof thedistalrelocation[5,26,38,39].A reduc-tion in intrinsic excitability is opposite to the above described theoretically predicted lower AP voltage threshold. One possibleexplanationfor thediscrepancy couldbethattheexperimentalconditionsusedtoinduce AIS/axon location changes, suchas chronicelevation of neuronalactivitybypharmacologicalapproachesor opto-geneticstimulation, wereinall cases accompaniedbya reduction in the neuronal input resistance (RN), which directly implies an increase in rheobase [5,26,38,39]. Other changesthat havean impact onexcitability may alsooccur,suchasphosphorylationofNa+channels[6]. As mentioned above, another possibility is that a large Kv7conductanceispresentintheAIS,whichcouldlead to decreased excitability with more distalAIS locations [21,26].
An alternative strategy to empirically investigate the electricalroleoftheAIS-somaseparationisacomparative analysisbetweensubpopulationsofneuronswith identi-fiedanatomical locationof theAIS[14,16,18,19,33]. Inlayer5ofthesomatosensorycortex,pyramidalneurons withaxon-carryingdendriteshavea3mVmore hyper-polarizedactionpotentialvoltagethresholdcomparedto neurons with somatic axons [18]. However, detailed geometrical quantification and correlation analysis showed that the AIS-soma distance was not a strong predictor of the voltage threshold and neither for the firing rate. Furthermore, comparative electrophysiologi-calrecordingsfromCA1pyramidalneuronsalsofailedto showdifferencesinactionpotentialvoltagethresholdasa functionofdistancebetweentheAISandthesoma,across a range of 40mm [19]. Interestingly, recent in vivo recordingsfromdopaminergiccellsinthesubstantianigra followedbyreconstructionandcomputationalmodelling found thatalow firingratein neuronswithadistalAIS was bestexplained bythenegativecorrelation between thesoma-AISdistanceandtheAISlength[40].Together, theexperimentalstudiesillustratethatdisentanglingand understandingtheroleofAISlocationinintrinsic neuro-nalexcitabilityrequires extensiveandjointquantitative assessments of multiple factors. However, the general
problem is that correlation does not imply
causation. Conclusively demonstrating the specific impact of AIS/axon location on intrinsic excitability throughexperimentalmeansawaitsfurtherinvestigation and methodologiesto manipulateeach factorseparately and selectively. Perhaps optogenetic tools to control organelle positions [41] mayin thefuture beleveraged to selectivelyrelocatetheinitialsegment.
AIS
location
diversity
and
backpropagation
AcriticalfunctionoftheAISisthegeneration ofanAP backpropagating into the soma and the dendritic tree. SomaticallyrecordedAPsacquiredwithsufficientlyhigh temporalresolution show, in mostmammalianneurons,anearlycomponentreflectingthespikegeneratedinthe AISandalate componentreflectingspikegenerationin the somatodendritic compartment[30,42].According to resistive couplingtheory,thecurrenttransmittedto the soma bytheAISspikeisinversely proportional toaxial resistance[18]andinparticularthesomashould there-forereceivelesscurrentfromadendriticaxoncompared toasomaticaxon(Figure2).Surprisingly,heterogeneity inAISandaxonpositionisnotalwaysreflectedinchanges intheinitialrisingphaseoftheAP.InhippocampalCA1 pyramidal neurons and thick-tufted layer 5 pyramidal neuronsthesomaticallyrecordedAPisindistinguishable betweendendriticandsomaticaxons[18,19].In thick-tufted layer 5 cells detailed geometric analysis showed thatthesmallerantidromiccurrent,withamoredistalAIS position, matches the smaller somatodendritic capaci-tance of neurons with axon-carrying dendrites, which have a thinner apical dendrite [18]. As a resultof the cable properties between the axon and dendrites, the antidromiccurrentflowmaintainsaconstantrise-timeof the somatodendritic spike. Whether a tuning between AIS location and the dendritic tree is a generalrule in building neuronsremainsto befurtherinvestigatedbut evidenceforcovarationisindeedseeninothercelltypes. Forexample,intheaviannucleuslaminaris(NL)ofthe chick,aregioninvolvedinauditory processing,neurons tunedtoahighcharacteristicfrequencyhavelarge soma-AISdistancesbut alsoshorterdendriteswhencompared tothosewithlowcharacteristicfrequency,whichhavea short soma-AIS distancesand largerdendrites[33,43]. In addition, ion channel properties also correlate with frequency tuning, including expression of dendritic T-type Ca2+ channels[44]. However,unlikecortical pyra-midalneurons,NLneuronsdonotregeneratetheAPat the soma with Na+ current and the AP amplitudes between theNL subpopulations differ as afunction of AISpositionandfrequencytuning [33].
Further examples of covariation between AIS location and dendritesare foundin midbraindopaminergic neu-rons withdendriticaxons(Figure 1b).Inthesecellsthe axon-carryingdendriteis2-foldlargerindiameter com-paredtonon-axoncarryingdendrites,whichwillfacilitate backpropagationtowardsthesoma[15,16].Despitesuch morphologicaladaptations,thedistallocationoftheAIS seems to be electrically significant: in these cells the rising phase of the somatic AP is slow, unreliable and canevenfailduringtheintegrationoffluctuatingin vivo-like synaptic inputs [15,45]. Collectively, the findings fromvariousstudiesshowthatthedistancebetweensoma andAISimpactsAPbackpropagationbutthefunctional consequencesarehighlyvariablebetweencell-types.
Dendritic
axons
enrich
synaptic
integration
According to resistive coupling theory, when synaptic currententerstheaxon-carryingdendrite,itflowstowards the soma and thereby produces a voltage gradientbetweenthesomaandthepointwheretheaxonemerges fromthedendrite,whichaddsuptothesomaticpotential, producing a larger synaptic potential in the axon. In accordwiththetheory,compartmentalsimulationsbased on substantia nigra dopaminergic neurons show EPSP amplification in axon-carrying dendrites (Figure 3a and b).Thesamecurrentinputwillproduceasmaller synap-ticdepolarizationattheaxonwhenenteringthecellatthe non-axoncarrying dendritesor when theaxon emerges from the soma. The difference is theoretically propor-tionalto theaxialresistance betweenthesoma andthe branchingpointoftheaxon,inparticularinverselyrelated todendritediameter(Figure3c).Experimentsaddressing synapticintegration inthesecells havebeen performed with simultaneous dual or triple patch-clamp recording fromtheaxon-carryingandnonaxon-carryingdendrites [14–16]. In support of the theoretical predictions that EPSPsare amplifieddue to resistive coupling,barrages of simulated EPSPs mimicking in vivo activity were observed to trigger APs more reliably when generated intheaxon-carryingdendritecomparedtothenon-axon carrying dendrite [15]. Recently, increased coupling of EPSP input to AP output has also been demonstrated with two-photon uncaging in CA1 pyramidal neurons comparingexcitatorysynapsesontheaxon-carrying den-drites with other dendrites [19]. The difference in
integration is reflectedby a somaticallyrecorded lower AP voltage threshold when uncaging glutamate at the axon-bearingdendrite,thedifferencebeingagaindueto thevoltagegradient betweensoma and axonbranching pointduringstimulation(Figure3d).
From
single
axons
to
neural
circuits
Thestudiesreviewedsofarshowthatbiophysical differ-encesbetweentheaxon-carryingandnon-axoncarrying dendritesmostlyleadstodifferentialcomputational pro-cessing of synaptic inputs. This raises the tantalizing possibility that variation of axon location in single cell classesenriches therepertoire of synapticintegrationat thecircuit level.Even further,withinanindividualcell with dendritic axons, thedistinct branch types maybe receiving specific afferent inputs. Evidence for branch-specificsynapticclusteringcomesfromsingle-cell analy-sisofdopaminergic neurons[46].Anatomical identifica-tionofinputsrevealedahigherdensityofglutamatergic synapsesonthe dendriteslocated near theaxon in the substantianigraparscompacta (SNc)compared to den-dritesinthesubstantianigraparsreticulata(SNr),making the SNc inputs more likely to drive the activity of dopaminergicneurons[46].Notably,inthehippocampus, pyramidalneuronswithdendriticaxonsarepreferentially observed in the superficial layers of the stratum
Figure3
Recording site
Axon
EPSP at axon origin
X1 X1 20 ms non AcDendrites 0.5 mV X2 X2 X3 X3 4 4 4 2 0 -2 -4 -6 -8 0 5 10 15 5 3 EPSP ratio Δ AP threshold (mV)
Dendritic diameter (μm) distance: stim. branch to soma (μm) 2 2 3 1 1 0 0 AcDendrites (a) (b) (c) (d)
Current Opinion in Neurobiology
Differentialsynapticintegrationindendriticaxons.(a)Simulatedmorphologywithadendriticaxonemerging50mmfromthesoma,andsynaptic inputsatdifferentlocations;onadendrite(X1)andontheaxon-carryingdendrite(locationsX2andX3).(b)SimulatedEPSPsmeasuredattheaxon
originforthethreelocationsshownina,showinglargerandmorerapidEPSPswhensynapsesontheaxon-carryingdendriteareactivated.(c)The EPSPamplificationvariesinverselywiththediameteroftheaxon-carryingdendrite(decreasingRa).(d)Somaticrecordingsofactionpotentialsin
CA1pyramidalcellsrevealamorehyperpolarizedvoltagethresholdwhentheaxon-carryingdendrite(AcDendrite,red)isstimulatedcomparedto non-axoncarryingdendrites(black),withlowerthresholdpotentialswhentheAISisfurtherfromthesoma(equivalentwithincreasingRa).
pyramidale[19].Asubdivisionofdendriticand somatic-axon pyramidal neuronsalong thesuperficial-to-deeper layer axis is in accord with experimentally observed gradients at the level of genetics, type of inhibitory inputs,sharpwaveripples,andactivityrecruitment dur-ing behavioural tasks [47,48]. In addition, recent data showed that the layer 5 neurons with dendritic axons haveadistinct dendriticmorphology, putatively reflect-ingamorphologicalsubtype[18].Whetherthe subdivi-sion has a geneticbasis is not known but classification basedonsingle-celltranscriptomeanalysisrevealedthree distincttypesinthedeeperlayersoflayer5[49].Recent retrogradetraceranalysisshowedthatthick-tufted layer 5 neurons can be subdivided based on their projection area,dendritic morphologyand invivofiringproperties, with cells projecting to theposterior medial divisionof thethalamus(POm) showingreduceddendritic branch-ing in thelayer 5 lamina and higherspontaneous firing rate [50].WhetherthePOm projectinglayer5 cells are characterizedbydendriticaxons,however,isnotknown and remainstobeexamined.
Interestingly,reconstructionsandmodellingofthe cere-bellargranulecells(CGCs)showedthatCGCswith axon-carrying dendrites are characterized by more complex claw-shapedendingsreceivingagreaterexcitatoryinput frommossy-fibres[17].Thesignificanceforprocessingin theCGClayerremainstobefurtherexaminedandmay be different from the examples described above since CGCs are electrotonically extremely compact, showing uniformEPSPsalongtheirdendriteandsoma[51]. Nev-ertheless, it is tempting to speculate thatresistive cou-plingofaxonswithinthethinCGCdendritesimpactsthe temporal processing of the high-frequency information from mossy fibre inputs and expands the integrative capabilities of the CGC layer to relay information to the Purkinje cells. Both theoretical and experimental studiesthus convergeto theideathatdiverse dendrite-axonarchitecturesinneuronsexpandsynapticintegration and theencoding capabilitiesof cellularcircuits.
Conclusion
In summary, recent studies are slowly shedding more lightonboththeabundanceandsignificanceofdendritic axons but manyquestionsremain to be resolved.First, whileinsomecelltypestheAIS/axonlocationco-varies with the structure of the dendritic tree, it is not well understood whether this is a generic phenomenon. Addressing this would require an integrated imaging approach for both theaxonaland dendritic treein con-junctionwithelectricalrecordings.Secondly,whatarethe molecularmechanismsinvolvedinestablishingtheaxon originandareneuronswithdendriticaxonsrepresenting genetic subtypes? Recently developed patch-clamp RNA-sequencingapproaches[52,53]couldinprinciple beemployedtoharvestthecytoplasmiccontentof iden-tifiedneuronswithorwithoutdendriticaxonsandlinkthe
morphological types to the single-cell transcriptome to resolve whether they represent genetic subtypes. Fur-thermore, the emerging theoretical and experimental evidencethatisreviewedindicatesthatalthoughintrinsic excitability isonly marginallydifferentin neuronswith dendritic axons, the neuronal architecture provides a unique integrative pathway for synaptic potentials. Addressing the question of significance of axon place-mentwillthereforerequirelinkingthecellularproperties to connectivity and resolving its role at the level of information processing in the circuit. Ultimately, com-biningcomputational,anatomical andfunctional record-ingapproacheswillanswerthelong-standingquestionof the functional significance of AIS/axon location variability.
Conflicts
of
interest
statement
Nothingdeclared.Acknowledgements
ThisworkwassupportedbytheAgenceNationaledelaRecherche [ANR-14-CE13-0003]andaprogramgrant[16NEPH02]oftheresearch programmeoftheFoundationforFundamentalResearchonMatter(FOM), whichispartoftheNetherlandsOrganisationforScientificResearch (NWO).WearethankfultothemembersoftheKoleandBrettelaboratories fortheircontributionsanddiscussion.
References
and
recommended
reading
Papersofparticularinterest,publishedwithintheperiodofreview, havebeenhighlightedas:
ofspecialinterest ofoutstandinginterest
1. RasbandMN:Theaxoninitialsegmentandthemaintenanceof neuronalpolarity.NatRevNeurosci2010,11:552-562.
2. KoleMHP,StuartGJ:Signalprocessingintheaxoninitial segment.Neuron2012,73:235-247.
3. BenderKJ,TrussellLO:Thephysiologyoftheaxoninitial segment.AnnuRevNeurosci2012http://dx.doi.org/10.1146/ annurev-neuro-062111-150339.
4. KubaH,OichiY,OhmoriH:PresynapticactivityregulatesNa(+) channeldistributionattheaxoninitialsegment.Nature2010, 465:1075-1078.
5. GrubbMS,BurroneJ:Activity-dependentrelocationoftheaxon initialsegmentfine-tunesneuronalexcitability.Nature2010, 465:1070-1074.
6.
EvansRapidMD,modulationDumitrescuofaxonAS,KruijsseninitialsegmentDLH,TaylorlengthSE,influencesGrubbMS: repetitivespikefiring.CellRep2015,13:1233-1245.
ThisstudyshowsthattheeffectofAISstructuralplasticityonexcitability iscompensatedbydephosphorylationofNavchannels.
7. HofflinF,JackA,RiedelC,Mack-BucherJ,RoosJ,CorcelliC, SchultzC,WahleP,EngelhardtM:Heterogeneityoftheaxon initialsegmentininterneuronsandpyramidalcellsofrodent visualcortex.FrontCellNeurosci2017,11:332.
8. GutzmannA,Ergu¨lN,GrossmannR,SchultzC,WahleP, EngelhardtM:Aperiodofstructuralplasticityattheaxoninitial segmentindevelopingvisualcortex.FrontNeuroanat2014, 8:11.
9. GalianoMR,JhaS,HoTS-Y,ZhangC,OgawaY,ChangK-J, StankewichMC,MohlerPJ,RasbandMN:Adistalaxonal cytoskeletonformsanintra-axonalboundarythatcontrols axoninitialsegmentassembly.Cell2012,149:1125-1139.
10. CajalSRY:RecollectionsofMyLife.MITPress;1989.
11. CajalSRY:Estructuradelkiasmao´pticoyteorı´ageneraldelos entrecruzamientosdelasvı´asnerviosas.Rev.Trim.Microgra´fı´ca; 1898.
12. AmaralDG:AGolgistudyofcelltypesinthehilarregionofthe hippocampusintherat.JCompNeurol1978,182:851-914.
13. MartinaM,VidaI,JonasP:Distalinitiationandactive propagationofactionpotentialsininterneurondendrites. Science2000,287:295-300.
14. Ha¨usserM,StuartGJ,RaccaC,SakmannB:Axonalinitiation andactivedendriticpropagationofactionpotentialsin substantianigraneurons.Neuron1995,15:637-647.
15. GentetLJ,WilliamsSR:Dopaminegatesactionpotential backpropagationinmidbraindopaminergicneurons.J Neurosci2007,27:1892-1901.
16. BlytheSN,WokosinD,AthertonJF,BevanMD:Cellular mechanismsunderlyingburstfiringinsubstantianigra dopamineneurons.JNeurosci2009,29:15531-15541.
17. HoustonCM,DiamantiE,DiamantakiM,KutsarovaE,CookA, SultanF,BrickleySG:Exploringthesignificanceof
morphologicaldiversityforcerebellargranulecellexcitability. SciRep2017,7:46147.
18.
HamadaCovariationMS,ofGoethalsaxoninitialS,desegmentVriesSI,locationBretteR,andKoledendriticMHP: tree normalizesthesomaticactionpotential.ProcNatlAcadSciUS A2016,113:14841-14846.
In this study theoretical analysis, electrophysiological patch-clamp recordinganddetailedidentificationofaxonsanddendritesare com-bined,providingatheoreticalbasisforthediverseAISlocationsacross neocorticalpyramidalneurons.
19.
ThomeCambridgeC,KellySB,T,BothYanezM,DraguhnA,SchultzA,C,BeckEngelhardtH,EgorovM,AV: Axon-carryingdendritesconveyprivilegedsynapticinputin hippocampalneurons.Neuron2014,83:1418-1430.
This study describes a surprisingly vast abundance of dendritically targetedaxonsinpopulationsofhippocampalpyramidalneuronsand demonstratesaroleinsynapticintegrativefunctions.
20. TriarhouLC:Axonsemanatingfromdendrites:phylogenetic repercussionswithCajalianhues.FrontNeuroanat2014,8:133.
21.
Brettecompartmentalization.R:SharpnessofspikePLoSinitiationComputBiolinneurons2013,9:e1003338.explainedby
Thisstudyintroducesresistivecouplingtheory,anapplicationofcable theorytothecasewhenthesomaactsasacurrentsinkfortheaxonal initiationsite.
22. TelenczukM,FontaineB,BretteR:Thebasisofsharpspike onsetinstandardbiophysicalmodels.PLOSONE2017,12: e0175362.
23. GulledgeAT,BravoJJ:Neuronmorphologyinfluencesaxon initialsegmentplasticity.eNeuro2016:3.
24. BaranauskasG,DavidY,FleidervishIA:Spatialmismatch betweentheNa+fluxandspikeinitiationinaxoninitial
segment.ProcNatlAcadSciUSA2013,110:4051-4056.
25. EyalG,MansvelderHD,deKockCPJ,SegevI:Dendritesimpact theencodingcapabilitiesoftheaxon.JNeurosci2014, 34:8063-8071.
26. LezmyJ,LipinskyM,KhrapunskyY,PatrichE,ShalomL,PeretzA, FleidervishIA,AttaliB:M-currentinhibitionrapidlyinducesa uniqueCK2-dependentplasticityoftheaxoninitialsegment. ProcNatlAcadSciUSA2017http://dx.doi.org/10.1073/ pnas.1708700114.
27. SloperJJ,PowellTP:Astudyoftheaxoninitialsegmentand proximalaxonofneuronsintheprimatemotorandsomatic sensorycortices.PhilosTransRSocLondBBiolSci1979, 285:173-197.
28. PetersA,ProskauerCC,Kaiserman-AbramofIR:Thesmall pyramidalneuronoftheratcerebralcortex.Theaxonhillock andinitialsegment.JCellBiol1968,39:604-619.
29. HallermannS,deKockCPJ,StuartGJ,KoleMHP:Stateand locationdependenceofactionpotentialmetaboliccostin corticalpyramidalneurons.NatNeurosci2012,15:1007-1014.
30. KoleMHP,StuartGJ:Isactionpotentialthresholdlowestinthe axon? NatNeurosci2008,11:1253-1255.
31. BattefeldA,TranBT,GavrilisJ,CooperEC,KoleMHP: HeteromericKv7.2/7.3channelsdifferentiallyregulateaction potentialinitiationandconductioninneocorticalmyelinated axons.JNeurosci2014,34:3719-3732.
32. KubaH,YamadaR,IshiguroG,AdachiR:RedistributionofKv1 andKv7enhancesneuronalexcitabilityduringstructuralaxon initialsegmentplasticity.NatCommun2015,6:8815.
33.
KubaenhancesH,IshiiauditoryTM,OhmoricoincidenceH:Axonaldetection.siteofspikeNatureinitiation2006, 444:1069-1072.
Thestudypresentsoneofthefirstpiecesevidenceforafunctionalroleof AISlocationdiversityinsensoryintegration.Basedoncellularrecordings fromaviannucleuslaminariscellsthatareorganizedintonotopicmaps theauthorsdemonstratehowdistalAISlocationsoptimizeneuronsforthe integrationofhigh-characteristicfrequencies.
34. JenkinsPM,KimN,JonesSL,TsengWC,SvitkinaTM,YinHH, BennettV:Giantankyrin-G:acriticalinnovationinvertebrate evolutionoffastandintegratedneuronalsignaling.ProcNatl AcadSciUSA2015,112:957-964.
35. ZontaB,DesmazieresA,RinaldiA,TaitS,ShermanDL,NolanMF, BrophyPJ:AcriticalroleforNeurofascininregulatingaction potentialinitiationthroughmaintenanceoftheaxoninitial segment.Neuron2011,69:945-956.
36. KubaH:Structuraltuningandplasticityoftheaxoninitial segmentinauditoryneurons.JPhysiol2012,590:5571-5579.
37. JamannN,JordanM,EngelhardtM:Activity-dependentaxonal plasticityinsensorysystems.Neuroscience2018,368:268-282.
38. WefelmeyerW,CattaertD,BurroneJ:Activity-dependent mismatchbetweenaxo-axonicsynapsesandtheaxoninitial segmentcontrolsneuronaloutput.ProcNatlAcadSciUSA 2015,112:9757-9762.
39. HatchRJ,WeiY,XiaD,Go¨tzJ:Hyperphosphorylatedtaucauses reducedhippocampalCA1excitabilitybyrelocatingtheaxon initialsegment.ActaNeuropathol2017,133:717-730.
40. MezaRC,Lo´pez-JuryL,CanavierCC,HennyP:Roleoftheaxon initialsegmentinthecontrolofspontaneousfrequencyofnigral dopaminergicneuronsinvivo.JNeurosci2018,38:733-744.
41. vanBergeijkP,AdrianM,HoogenraadCC,KapiteinLC: Optogeneticcontroloforganelletransportandpositioning. Nature2015,518:111-114.
42. CoombsJS,CurtisDR,EcclesJC:Theinterpretationofspike potentialsofmotoneurones.JPhysiol1957,139:198-231.
43. KubaH,YamadaR,FukuiI,OhmoriH:Tonotopicspecialization ofauditorycoincidencedetectioninnucleuslaminarisofthe chick.JNeurosci2005,25:1924-1934.
44. FukayaR,YamadaR,KubaH:TonotopicvariationoftheT-type Ca2+currentinavianauditorycoincidencedetectorneurons.J
Neurosci2018,38:335-346.
45. GraceAA,BunneyBS:Intracellularandextracellular electrophysiologyofnigraldopaminergicneurons—2.Action potentialgeneratingmechanismsandmorphological correlates.Neuroscience1983,10:317-331.
46. HennyP,BrownMTC,NorthropA,FaunesM,UnglessMA, MagillPJ,BolamJP:Structuralcorrelatesofheterogeneousin vivoactivityofmidbraindopaminergicneurons.NatNeurosci 2012,15:613-619.
47. DanielsonNB,ZarembaJD,KaifoshP,BowlerJ,LadowM, LosonczyA:Sublayer-specificcodingdynamicsduringspatial navigationandlearninginhippocampalareaCA1.Neuron 2016,91:652-665.
48. ValeroM,CidE,AverkinRG,AguilarJ,Sanchez-AguileraA, VineyTJ,Gomez-DominguezD,BellistriE,laPridadeLM: DeterminantsofdifferentdeepandsuperficialCA1pyramidal
celldynamicsduringsharp-waveripples.NatNeurosci2015, 18:1281-1290.
49. TasicB,MenonV,NguyenTN,KimTK,JarskyT,YaoZ,LeviB, GrayLT,SorensenSA,DolbeareTetal.:Adultmousecortical celltaxonomyrevealedbysinglecelltranscriptomics.Nat Neurosci2016,19:335-346.
50. Rojas-PiloniG,GuestJM,EggerR,JohnsonAS,SakmannB, OberlaenderM:Relationshipsbetweenstructure,invivo functionandlong-rangeaxonaltargetofcorticalpyramidal tractneurons.NatCommun2017,8:870.
51. DelvendahlI,StraubI,HallermannS:Dendriticpatch-clamp recordingsfromcerebellargranulecellsdemonstrate electrotoniccompactness.FrontCellNeurosci2015,9:93.
52.
FuzikLinnarssonJ,ZeiselS,HarkanyA,Ma´te´ Z,T:IntegrationCalvigioniD,ofYanagawaelectrophysiologicalY,Szabo´ G, recordingswithsingle-cellRNA-seqdataidentifiesneuronal subtypes.NatBiotechnol2016,34:175-183.
Bothstudies introduce apowerful strategytoobtain single-cellRNA sequencingdatafrompatch-clamprecordings(Patch-seq)tolink mor-phologyandelectrophysiologicalproperties,enablingacomprehensive classificationofneuronalsubtypes.
53.
CadwellReimerJ,CR,ShenPalasantzaS,BethgeA,M,JiangToliasX,BerensKFetal.:P,DengQ,YilmazM, Electrophysiological,transcriptomicandmorphologic profilingofsingleneuronsusingPatch-seq.NatBiotechnol 2016,34:199-203.
SeeannotationtoRef.[52].