DispatchR441

Dispatches

Optogenetics: Illuminating Sources of Locomotor Drive

A recent study has used optogenetics to identify the source of excitatory drivefor locomotion in zebrafish, revealing unexpected differences in the commandsignals from hindbrain to spinal cord.

Embryonic spinal cord

V0

Dd

TFs

V1V2MNV3

V2a(Chx10)En1

Dbx1

Hb9Sim1

Ventral

Current Biology

CaudalRostral

Dorsal

Figure 1. Development of spinal cord.

Schematic cut away of embryonic zebrafishspinal cord illustrating the major dorso-ventral postmitotic divisions that form duringspinal cord development. These include fourventral divisions (V0–V3), a motoneurondomain (MN), and numerous dorsal domains(Dd), collapsed here into one region forsimplicity. DNA transcription factors (TFs)uniquely identify neurons originating fromthe different ventral domains. The V2aneurons marked by the transcription factorChx10 are the focus of this study. For moredetails, see main text.

David L. McLean

As the saying goes, seeing is believing.If this is true, then optogeneticanalysis of neural circuitry intransparent zebrafish should converteven the most skepticalneuroscientists. The arrival ofoptogenetic technology hasrevolutionized functional studies ofthe nervous system [1,2]. Neuralactivity can now be precisely controlledin freely behaving animals by simplyilluminating light-sensitive actuatorsexpressed in different groups ofneurons. In this issue, Kimura andcolleagues [3] report how the use ofoptogenetics in zebrafish allowedthem to identify a population ofneurons in the hindbrain that enablelocomotion. They went on to useelectrophysiology to demonstrateunexpected heterogeneity in thelocomotor command signals fromneurons within this population tospinal cord.

Before launching into theirexperiments, it will help to brieflycover some basic principles of spinalcord development. All vertebratelocomotor networks are assembledfrom cells arising from one of fourpostmitotic domains in spinal cord,numbered V0–V3 [4,5]. Neurons in eachdomain can be identified by theexpression of specific transcriptionfactors that regulate gene expression(Figure 1). The new work describedhere focuses on cells arising from theV2 region, more specifically thosemarked by the transcription factorChx10 (V2a neurons). In the zebrafishspinal cord, these neurons providerhythmic excitatory drive tomotoneurons on the same side of thebody during locomotion [6,7]. However,V2a neurons also extend from thespinal cord well into the brain [6].Although hindbrain V2a cells share anumber of morphological features withspinal ones [8], until now theircontribution to locomotion was largelya mystery.

In all vertebrates, including zebrafish,spinal networks generate locomotion,but it is descending commands fromthe brain that decide when to move,where to move and for how long [9].Kimura et al. [3] took advantage of theconserved genetic coding of neuronalidentity to explore whether hindbrainV2a cells represent a well-knownsource of descending ‘reticulo-spinal’drive during locomotion. There aretwo obvious predictions if hindbrainV2a cells are responsible foractivating and sustaining locomotionin zebrafish: if you stimulate hindbrainV2a cells, the fish should swim; ifyou silence the cells, they shouldstop. These predictions are tailor-madefor optogenetic evaluation andreflect the gold standard forassessing the contribution of neuronsto behavior, namely sufficiency andnecessity.

Kimura et al. [3] began by creating aheroic number of stable transgenic fishlines using the Gal4:UAS system. Thisapproach provides a greater deal offlexibility for expressing different DNAconstructs in the same population ofneurons [10]. To ensure sufficientexpression levels of early-onsetChx10-dependent constructs, theauthors performed their experimentsat the earliest point zebrafish beginswimming spontaneously (about threedays old). To test the idea thatactivation of hindbrain V2a cells wouldevoke swimming, they selectivelyexpressed channelrhodopsin (ChR) inV2a neurons. ChR is an ion channelthat generates inward current flow inresponse to blue light [11,12], whichmakes neurons fire action potentials or‘spike’. Different regions of the nervoussystem were then illuminated fromabove using custom designed opticalequipment and the response of thefreely moving tail of head-fixed fishwas monitored from below using ahigh-speed camera.

In the first of a number of technicallydemanding experiments, Kimura et al.[3] confirmed that briefly shining blue

light on ChR–V2a cells in hindbrainconsistently evokes swimmingbehavior. The authors then activatedmore local regions of hindbrain, whichrevealed that some V2a neurons weremore effective at evoking swimmingthan others. In particular, the caudalhindbrain was consistently the mostreliable location (Figure 2A). The nextstep was to silence the V2a cells. To dothis, the authors used either eNpHR3.0(Halo3) or archaeorhodopsin-3 (Arch),both of which generate outward currentflow in response to green light andsilence neurons [13–15]. As expected,whole hindbrain illumination of Halo3–or Arch–V2a neurons, and selectiveillumination of the special caudalregion, prematurely terminatedspontaneously generated swimming(Figure 2B). Are these neurons bothsufficient and necessary formaintaining locomotion? The answerwas a resounding yes.The regional differences in the ability

to start or prematurely stop swimmingbehavior raised another question.Could differences in the projectionpatterns of hindbrain V2a neurons

a b c d e

a b c d e

a b c d e

A

B

C

Rhythmic

Rhythmic

Rhythmic

Tonic

MiV1

V2a ChR–transgenic fish

V2a Arch– or Halo3–transgenic fish

V2a GFP–transgenic fish

Rh

Rh

T

Rh

Rh

T

yt

yt

on

yth

yth

on

m

m

c

m

m

c

c

c

c

c

hb-V2a

sc-V2a

Swimactivity

Shine blue light on region ‘d’

Swimming starts

Shine green light on region ‘d’

Swimming stops

Caudal

Current Biology

Figure 2. Optogenetic and electrophysiolog-ical evaluation of hindbrain V2a neurons.

(A) Top down view of a three-day-oldzebrafish illustrating the activation ofhindbrain V2a neurons with light. The brainand spinal cord were divided into five equalregions (a–e) and blue light was shone ontransgenic fish expressing channelrhodopsinin V2a neurons (V2a ChR-transgenic fish).Illumination of the caudal hindbrain region(d) was the most effective at activatingswimming. (B) Same as in A, but this timeillumination of the caudal hindbrain regionwith green light in transgenic fish expressingeither archaeorhodopsin-3 (Arch) oreNpHR3.0 (Halo3) prematurely terminatesswimming. (C) Schematic representations ofelectrophysiological recordings from therostrally located MiV1, a caudal hindbrainV2a cell (hb-V2a), a V2a cell in the spinal

Current Biology Vol 23 No 10R442

explain the discrepancies? Perhapsmore rostral V2a cells do not innervatespinal cord. A creative use of thephotoconvertible protein, Kaede [16],along with more standard retrogradefilling of V2a neurons confirmed thisidea. In the caudal hindbrain region alarge proportion of relatively small V2acells project to spinal cord, while in lessreliable rostral regions, only a fewrelatively large reticulo-spinal neuronsare found.

At this point, the story was alreadypretty convincing. Hindbrain V2aneurons provide a crucial source ofexcitatory drive to spinal locomotorcircuits. Also, the relative ability ofdifferent regions of hindbrain to startor stop locomotion matches theirextent of spinal innervation.Nonetheless, Kimura et al. [3] thentook their study a step further. Arethese neurons even active duringlocomotion? If so, are there anyregional differences in spikingbehavior that match the optogeneticobservations? To answer thesequestions, the authors turned toelectrophysiology.

The ability to record electricalsignals from hindbrain neurons relieson stability, so the recordings had tobe performed in larvae immobilized bya plant toxin that blocks neuromusculartransmission. In this situation, therhythmic motor output that wouldnormally generate swimmingmovements can be monitored using asuction electrode, which picks up theelectrical signals from motoneuronaxons that innervate the tail muscles(Figure 2C). Hindbrain neuron activitywas monitored using either whole-cellpatch clamp recordings of membranepotential or cell-attached recordingsof spikes.

Kimura et al. [3] focused on twohindbrain regions that had thelowest and the highest influence onlocomotion. In the rostral region, theytargeted a large reticulo-spinal neuronthat is readily identifiable based on itsmorphology [17,18], known as MiV1(middle rhombencephalon, ventral,level 1). In the caudal region, there were

cord (sc-V2a), and swimming activity (bottomtrace). Shaded grey boxes highlight thetiming of spiking activity related to swimactivity. MiV1 and sc-V2a cells fire at reliablelocations in the cycle (rhythmic), whilehb-V2a cells do not (tonic). In reality theserecordings were performed separately, butthey are combined here for illustrationpurposes.

no easily identifiable cells, so insteadthey consistently targeted ventrallylocated neurons, which they term‘small V2a cells’. In the latter case, theytook care to demonstrate that the cellsprojected to spinal cord.The first observation was that MiV1

and small V2a cells are indeed activeduring locomotion. What followed,however, was more surprising. Circuitsin the spinal cord generate rhythmicpatterns of activity in response tounpatterned, ‘tonic’ excitatory drive[19]. The idea is that reticulo-spinalneurons are a major source of thistonic drive. Given that V2a neuronsform a continuous column from spinalcord into hindbrain, the expectationwas that V2a neurons graduallytransition from highly rhythmic cells inspinal cord to less rhythmic ones inmore rostral regions. The assumptionhere is that more rostral cells wouldbe higher up in the chain of command.Instead, what Kimura et al. [3]

observed was an unexpectedtransition from rhythmic, to tonic, backto rhythmic drive as you move fromspinal cord to rostral hindbrain(Figure 2C). The tonic activity in smallV2a cells is certainly consistent withprevailing views of reticulo-spinal driveand the importance of this region insustaining locomotion. However, thehighly rhythmic activity from therostral MiV1 cells is more difficult toexplain, especially as MiV1 cells havebeen implicated in turning [20], whichis not necessarily a rhythmic behavior.Clearly, when it comes to thefunctional organization of V2aneurons in zebrafish hindbrain, thisfinding suggests there is more to itthan meets the eye.So, what did we learn? Using

optogenetics, Kimura et al. [3] haveunambiguously identified the sourceof locomotor drive from hindbrain tospinal cord. The relevance of thisfinding will likely extend beyondzebrafish, given the common geneticorigin of brainstem and spinal circuitryin vertebrates [4,5]. The authors havealso demonstrated that spinal circuitsreceive both tonic and rhythmic signalsfrom hindbrain neurons. While thisobservation alone is not new [9], whatis novel is that both signals originatefrom a single genetically identifiedpopulation, and not in a way you mightpredict based on anatomy. Obviously,there are still many open questions. Arethere differences in the spinal neuronstargeted by rhythmic versus tonic

DispatchR443

excitatory drive? Also, what are therest of the hindbrain V2a cells doing ifnot controlling locomotion?Convincing answers to thesequestions, and many more, are surelynot far off if this technical tour deforce is anything to go by.

References1. Gross, M. (2011). Shining new light on the brain.

Curr. Biol. 21, R831–R833.2. Deisseroth, K., Feng, G., Majewska, A.K.,

Miesenbock, G., Ting, A., and Schnitzer, M.J.(2006). Next-generation optical technologies forilluminating genetically targeted brain circuits.J. Neurosci. 26, 10380–10386.

3. Kimura, Y., Satou, C., Fujioka, S., Shoji, W.,Umeda, K., Ishizuka, T., Yawo, H., andHigashijima, S. (2013). Hindbrain V2a neuronsin the excitation of spinal locomotor circuitsduring zebrafish swimming. Curr. Biol. 23,843–849.

4. Arber, S. (2012). Motor circuits in action:specification, connectivity, and function.Neuron 74, 975–989.

5. Goulding, M. (2009). Circuits controllingvertebrate locomotion: moving in a newdirection. Nat. Rev. Neurosci. 10, 507–518.

6. Kimura, Y., Okamura, Y., and Higashijima, S.(2006). alx, a zebrafish homolog of Chx10,marks ipsilateral descending excitatoryinterneurons that participate in the regulation ofspinal locomotor circuits. J. Neurosci. 26,5684–5697.

7. Eklof-Ljunggren, E., Haupt, S., Ausborn, J.,Dehnisch, I., Uhlen, P., Higashijima, S., and El

Manira, A. (2012). Origin of excitationunderlying locomotion in the spinal circuit ofzebrafish. Proc. Natl. Acad. Sci. USA 109,5511–5516.

8. Kinkhabwala, A., Riley, M., Koyama, M.,Monen, J., Satou, C., Kimura, Y.,Higashijima, S., and Fetcho, J. (2011). Astructural and functional ground plan forneurons in the hindbrain of zebrafish. Proc.Natl. Acad. Sci. USA 108, 1164–1169.

9. Le Ray, D., Juvin, L., Ryczko, D., and Dubuc, R.(2011). Supraspinal control of locomotion: themesencephalic locomotor region. Prog. BrainRes. 188, 51–70.

10. Scott, E.K. (2009). The Gal4/UAS toolbox inzebrafish: new approaches for definingbehavioral circuits. J. Neurochem. 110,441–456.

11. Boyden, E.S., Zhang, F., Bamberg, E.,Nagel, G., and Deisseroth, K. (2005).Millisecond-timescale, genetically targetedoptical control of neural activity. Nat. Neurosci.8, 1263–1268.

12. Wang, H., Sugiyama, Y., Hikima, T., Sugano, E.,Tomita, H., Takahashi, T., Ishizuka, T., andYawo, H. (2009). Molecular determinantsdifferentiating photocurrent properties of twochannelrhodopsins from Chlamydomonas.J. Biol. Chem. 284, 5685–5696.

13. Chow, B.Y., Han, X., Dobry, A.S., Qian, X.,Chuong, A.S., Li, M., Henninger, M.A.,Belfort, G.M., Lin, Y., Monahan, P.E., et al.(2010). High-performance genetically targetableoptical neural silencing by light-driven protonpumps. Nature 463, 98–102.

14. Zhang, F., Wang, L.P., Brauner, M.,Liewald, J.F., Kay, K., Watzke, N., Wood, P.G.,Bamberg, E., Nagel, G., Gottschalk, A., et al.(2007). Multimodal fast optical interrogation ofneural circuitry. Nature 446, 633–639.

15. Gradinaru, V., Zhang, F., Ramakrishnan, C.,Mattis, J., Prakash, R., Diester, I., Goshen, I.,Thompson, K.R., and Deisseroth, K. (2010).Molecular and cellular approaches fordiversifying and extending optogenetics. Cell141, 154–165.

16. Ando, R., Hama, H., Yamamoto-Hino, M.,Mizuno, H., and Miyawaki, A. (2002). Anoptical marker based on the UV-inducedgreen-to-red photoconversion of a fluorescentprotein. Proc. Natl. Acad. Sci. USA 99,12651–12656.

17. Kimmel, C.B., Powell, S.L., and Metcalfe, W.K.(1982). Brain neurons which project to thespinal cord in young larvae of the zebrafish.J. Comp. Neurol. 205, 112–127.

18. Metcalfe, W.K., Mendelson, B., andKimmel, C.B. (1986). Segmental homologiesamong reticulospinal neurons in the hindbrainof the zebrafish larva. J. Comp. Neurol. 251,147–159.

19. Grillner, S., and Jessell, T.M. (2009). Measuredmotion: searching for simplicity in spinallocomotor networks. Curr. Opin. Neurobiol. 19,572–586.

20. Orger, M.B., Kampff, A.R., Severi, K.E.,Bollmann, J.H., and Engert, F. (2008). Control ofvisually guided behavior by distinct populationsof spinal projection neurons. Nat. Neurosci. 11,327–333.

Department of Neurobiology, NorthwesternUniversity, Evanston, IL 60208, USA.E-mail: david-mclean@northwestern.edu

http://dx.doi.org/10.1016/j.cub.2013.04.015

Bacterial Discrimination:Dictyostelium’s Discerning Taste

New research indicates that the social amoeba Dictyostelium discoideumrecognizes distinctions between Gram(-) and Gram(+) bacterial prey andresponds discriminately to these two groups of bacteria. These findings maylend insight to the origins of microbial pattern recognition in innate immunity.

Michelle L.D. Snyder

Innate immune cells in organisms asdiverse as fruitflies and humansuse conserved pattern recognitionmechanisms to differentiate microbialinvaders from self by detectingmicrobe-associated molecularpatterns (MAMPs) present on fungi,viruses and bacteria but absent fromhosts [1]. In this issue of CurrentBiology, Nasser et al. [2] show thatupon phagocytosis of bacterial preythe social amoeba Dictyosteliumdiscoideum not only discriminatesbetween different species of bacteria,but also responds differentiallyto Gram(-) and Gram(+) groups ofbacteria. The mechanisms by whichD. discoideum discriminates between

Gram(-) and Gram(+) bacteria maybe shared by phagocytes in othereukaryotes and may play roles in theregulation of innate immune activityin other organisms.

Living within the soil, D. discoideumphagocytoses bacteria for nutritionalpurposes. Within this environment,bacteria that have evolvedmechanisms to evade amoeboidphagocytosis and killing would enjoya selective advantage [3]. As it turnsout, various bacterial species haveevolved mechanisms to survivepredation and infect amoebae,promoting the use of D. discoideumas a model to study host–pathogeninteractions [4,5].

Recent evidence indicates thatD. discoideum does not remain

defenseless against infection bybacteria and has evolved mechanismsto efficiently detect and respond tobacteria. Exposure to bacteriaupregulates the expression of genespotentially involved in bacterialrecognition and killing [6–8]. Amongthese are genes homologous toknown pattern recognition moleculesinvolved in innate immunity in otherorganisms, including one that encodesthe Toll/interleukin-1 receptor (TIR)domain-containing protein TirA [9]. TIRdomain-containing proteins playintegral roles in MAMP-recognitionpathways in innate immune systemsof various organisms [1], and inD. discoideum TirA is required forefficient phagocytosis of Gram(-)bacteria [9,10].Nasser et al. [2] hypothesized that

if amoebae can recognize microbialpatterns then, given the differences instructure and molecular compositionof Gram(-) and Gram(+) cell walls,D. discoideum may responddiscriminately to these two groups ofbacteria. Drawing on transcriptomeanalysis coupled with results frommutational screening, Nasser et al. [2]