Specific activation of the paralemniscal pathway during nociception

NEUROSYSTEMS

Specific activation of the paralemniscal pathway duringnociception

Laura Frangeul,1 Cesar Porrero,2 Maria Garcia-Amado,2 Benedetta Maimone,1 Madlyne Maniglier,1

Francisco Clasc�a2 and Denis Jabaudon1,31Department of Basic Neurosciences, University of Geneva, 1 rue Michel Servet, 1211 Geneva, Switzerland2Department of Anatomy & Neuroscience, School of Medicine, Autonoma University, Madrid, Spain3Department of Neurology, Geneva University Hospital, 1211 Geneva, Switzerland

Keywords: mouse, sensory physiology, somatosensory cortex, thalamocortical pathways

Abstract

Two main neuronal pathways connect facial whiskers to the somatosensory cortex in rodents: (i) the lemniscal pathway, whichoriginates in the brainstem principal trigeminal nucleus and is relayed in the ventroposterior thalamic nucleus and (ii) the paralem-niscal pathway, originating in the spinal trigeminal nucleus and relayed in the posterior thalamic nucleus. While lemniscal neuronsare readily activated by whisker contacts, the contribution of paralemniscal neurons to perception is less clear. Here, we function-ally investigated these pathways by manipulating input from the whisker pad in freely moving mice. We report that while lemniscalneurons readily respond to neonatal infraorbital nerve sectioning or whisker contacts in vivo, paralemniscal neurons do not detec-tably respond to these environmental changes. However, the paralemniscal pathway is specifically activated upon noxious stimu-lation of the whisker pad. These findings reveal a nociceptive function for paralemniscal neurons in vivo that may critically informcontext-specific behaviour during environmental exploration.

Introduction

Sensory information generated by contacts of the facial whiskersof the snout is crucial for rodents in exploring and navigatingthrough their environment. This information is transmitted via theinfraorbital branch of the trigeminal nerve (ION) to brainstemtrigeminal nuclei. Trigeminal neurons then project to the contralat-eral thalamus via two main pathways, the so-called lemniscal andparalemniscal pathways (Fig. 1). The lemniscal pathway originatesin the principal trigeminal nucleus (PrV) and projects to the con-tralateral ventroposteromedial nucleus (VPM) of the thalamus.Conversely, the paralemniscal pathway originates in the rostraldivision of the interpolaris spinal trigeminal nucleus (SpVi) andprojects to the contralateral posterior thalamic nucleus (Po) (Pierretet al., 2000). VPM and Po thalamocortical (TC) neurons, in turn,send their axons to the primary somatosensory cortex (S1), wherethey form largely interdigitated circuits: VPM TC axons formsynapses in layer (L)4 and L6, while Po TC axons form synapsesin layers L1 and L5A. In addition, many Po TC neurons sendaxonal branches to the secondary somatosensory cortex (S2) aswell as to the primary motor cortex (M1) (Fig. 1; for recentreviews see Alloway, 2007; Erzurumlu et al., 2010; Bosman et al.,2011; Kleinfeld & Deschenes, 2011; Ohno et al., 2012; Pouchelonet al., 2012).

Lemniscal neurons in the PrV and VPM play a critical role in relay-ing signals generated by whisker contacts and display robust, topo-graphically-organised, fixed-latency responses to whisker deflections,even in head-fixed or anesthetised rats (Simons & Carvell, 1989; Min-nery & Simons, 2003; Minnery et al., 2003). In contrast, Po TC neu-rons only poorly and unreliably respond to whisker deflections, andloss of either SpVi or Po TC neurons does not affect detection ofwhisker movement (Diamond et al., 1992a,b; Sosnik et al., 2001;Alloway, 2007; Nakamura et al., 2009; Pouchelon et al., 2012).Instead, it has been proposed that Po TC neurons act as a ‘coinci-dence detectors’ for descending cortical input and ascending periph-eral input, thereby encoding spatial properties of whisker activity(Ahissar et al., 2000; Alloway, 2007; Groh et al., 2013; but see Masriet al., 2008) and relay input between cortical areas in corticothalamo-cortical loops (Theyel et al., 2010; Sherman, 2012). However, ourunderstanding of the role of ascending paralemniscal input to the Poin these processes remains unclear. Importantly, most studies haveused in vitro, anesthetised, or head-fixed preparations, and may haveoverlooked potential functions of this pathway in other physiologicalcontexts (Diamond et al., 1992a,b; Sosnik et al., 2001; Crochet &Petersen, 2006; Nakamura et al., 2009; Gambino & Holtmaat, 2012).Here, we investigated the function of these pathways by using

two complementary approaches. First, in order to assess the overallsensitivity of Po TC neurons (paralemniscal) to environmental stim-uli, we quantitatively examined the effects of sectioning the ION atbirth on the development of Po TC axon arbors in S1, andcompared them with the effect on VPM (lemniscal) axons. Second,in order to identify potential sensory modality-specific functions of

Correspondence: Denis Jabaudon, 1Department of Basic Neurosciences, as above.E-mail: [email protected]

Received 19 July 2013, revised 10 January 2014, accepted 27 January 2014

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons Ltd

European Journal of Neuroscience, Vol. 39, pp. 1455–1464, 2014 doi:10.1111/ejn.12524

these pathways, we quantified lemniscal and paralemniscal neuronactivation in freely moving mice (i) during tactile exploration oftheir environment and (ii) after noxious stimulation of the whiskerpad, using expression of the immediate–early gene c-Fos to reportneuronal firing (Bisler et al., 2002; Staiger et al., 2002; Schoenen-berger et al., 2009; Wagener et al., 2010).Our results indicate, first, that the effect of ION sectioning from

birth is markedly more severe on VPM axons that on Po axons and,second, that while whisker stimulation during dual whisker environ-mental exploration strongly activates lemniscal neurons, paralemni-scal neurons are selectively activated by noxious stimulation of thewhisker pad. These findings reveal a nociceptive function for para-lemniscal neurons in vivo which may critically contribute to the pro-cessing of whisker input in physiologically realistic scenarios.

Materials and methods

All experimental procedures were approved by the Geneva CantonalVeterinary Authority and the Autonoma University Bioethics Com-mittee and were carried out in accordance with the European Com-munities Council Directive of 24 November 1986 (86/609/EEC).C57BL/6 mice were used.

Infraorbital nerve section

On the day of birth, mice pups were anesthetised by hypothermiaand the right ION was sectioned as previously described (Rhoadeset al., 1990). A small strip of sterile paraffin film was placedbetween the nerve cut ends to prevent axonal regrowth.

Anterograde labelling

We first placed a microdeposit of Fast Blue (Sigma-Aldrich,Switzerland) in the macrovibrisssae zone of the S1 barrel field(S1BF) to retrogradely label S1BF-projecting Po and VPM neuronsand determine their stereotactic coordinates, which were subse-quently used for anterograde labelling experiments. To label TCaxons, lysine-fixable biotinylated dextran amine (BDA), 10 000 MW(Invitrogen, California, USA) was iontophoresed into the left VPMor Po of ION sectioned (n = 22) or intact (n = 10) adult mice. Amicropipette containing 3% BDA in 0.01 M pH 7.4 phosphate buffer(PB) was stereotaxically positioned. For Po injections, coordinates(in mm) were: AP, �2.1; and ML, �1.3 from the bregma; DV, �3.0from the pial surface. For VPM injections, coordinates were: AP,�2.1 and ML, �1.6 from the bregma; DV, �3.0 from the pial sur-face. Current pulses (200–400 nA, 1 s on/off) were delivered for30 min. Brains were perfusion-fixed after 7 days and seriallysectioned on a freezing microtome at 60 lm. To reveal in fine detailthe BDA-labelled axons, all sections were processed using glucose-oxidase and an avidin-biotin-peroxidase kit (ABC; Vector,

California, USA) with nickel sulfate-enhancement. The serial sectionswere then alternately counterstained with thionin or cytochromeoxidase histochemistry. Sections were finally mounted on glass slides,dehydrated, and coverslipped with DePeX (Sigma, Switzerland).

Bicuculline injections

Adult mice (n = 3) were anesthetised with isoflurane. Bicucullinemethiodide (0.3 lL; Ascent Scientific, Bristol, UK) at 15 mM in 0.9%NaCl was stereotaxically injected into the somatosensory cortex usinga nanoinjector (Nanoject II Auto-Nanoliter Injector, DrummondScientific, Pennsylvania, USA). Control animals (n = 3) receivedidentical injections of vehicle solution. Mice were allowed to survivefor 1 h after the injection.

Single-whisker stimulation

Mice were first anesthetised with isoflurane and then with urethane(1.5 g/kg, i.p.). The C2 whisker was stimulated by back and forthdeflections (4 Hz), using piezoelectric ceramic elements attached toa glass pipette, for 1 h (Gambino & Holtmaat, 2012).

Dual-whisker environmental exploration

All but two whiskers (C1 and C2) on the right side of the snoutwere clipped in adult mice (n = 3), after which the animals wereplaced in an large playground box containing plastic balls, maze-likepieces of thread and various small objects, as previously described(Bisler et al., 2002). Mice were kept for 1 h in this enriched envi-ronment and then perfused.

Noxious stimulation

Adult mice (n = 6) were briefly anesthetised with isoflurane, and50 lL of a capsaicin solution [10 mM (Sigma), 100% ethanol and7% Tween-80 in saline] or vehicle solution subcutaneously microin-jected into the whisker pad as previously described (Noma et al.,2008). Mice were then put back into their cages and killed after 1 h.

C-Fos immunohistochemistry

Paraformaldehyde-fixed coronal brain sections from n = 14 micewere cut at 50 lm on a vibrating microtome. Briefly, sections wereincubated in 3% H2O2 and 10% methanol in phosphate-bufferedsaline (PBS) for 5 min, followed by rinsing in 0.1 M PBS, andincubated in 3% BSA and 0.3% triton (in PBS) for 1 h. Thesections were incubated with primary polyclonal antisera overnightat 4 °C against c-Fos (raised in rabbit; 1:5000; sc-52; Santa CruzBiotechnology, California, USA), and then sequentially incubated inbiotinylated goat anti-rabbit IgG (1 : 200; Invitrogen) and in peroxi-dase-conjugated avidin–biotin complex (1 : 200; Vectastain, VectorLaboratories). They were then pre-incubated in 0.05% 3,3-diam-inobenzidine-tetra HCl (DAB; Sigma), 0.025% cobalt chloride and0.02% nickel ammonium sulphate (Sigma) in 0.1 M PB. After a 15-min pre-incubation, the reaction was started by adding 0.01% hydro-gen peroxide. Sections were next dehydrated and mounted in DePeXmedium (Sigma).

Stereological axon quantification

Control VPM-injected (n = 3), ION-sectioned VPM-injected(n = 3), control Po-injected (n = 3) and ION-sectioned Po-injected

Fig. 1. Schematic representation of the lemniscal (blue) and paralemniscal(red) pathways.

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

1456 L. Frangeul et al.

(n = 3) mice were used for quantitative analysis of labelled axonarbors. The absolute axon length within S1BF was determined forVPM and Po injections using the isotropic virtual planes methodfollowing a fractionator sampling scheme (Drøjdahl et al., 2010).The absolute number of axon varicosities in each layer was deter-mined with the optical fractionator method. The laminar distributionof axonal length and varicosity numbers was calculated based onthese values.In order to quantify axonal changes on a per neuron basis, we

normalised for differences in injection size or labelling efficacy byestimating the number of TC neurons effectively labelled withBDA in each experiment. This was performed by counting theBDA-labelled white matter TC axons crossing the striatopallialborder of the internal capsule (range: 10–97 axons; Fig. S1). Weconsidered this to be a reliable proxy for the number of effectivelylabelled TC projection neurons because single-neuron axon tracingstudies of VPM and Po cells in rat and mice have consistentlyshown that these TC cells give off a single axon trunk that extendsall the way along the internal capsule and striatopallial border intothe subcortical white matter (Fig. S1 and Deschenes et al., 1998;Gheorghita et al., 2006; Galazo et al., 2008; Ohno et al., 2012).Po axons often give off perpendicular thalamostriatal branches, butthese are thin, finely branched terminal arborisations and have notbeen observed to exit the striatum towards the cortex. We then cal-culated the per neuron laminar distribution of axon length and vari-cosities using these values. Interbouton distance was determined bydividing varicosity number by axonal length within each layer. Thecoefficient of error due to the sampling method was calculated(Gundersen et al., 1999). Student’s t-test and Kolmogorov–Smirnovtests were used to analyse differences between control andION-sectioned cases. We considered significant differences thosewith P-values < 0.05, and highly significant when P-values were< 0.001.

C-Fos quantification

For dual-whisker environmental exploration, c-Fos+ neurons werecounted on 10 9 images in both hemispheres. The number ofc-Fos+ cortical neurons was quantified in C1 and C2 corticalcolumns on five slices per animal. Quantification of laminar distri-bution was performed by dividing cortical thickness into ten equalbins. We used cytochrome oxidase and DAPI staining to determinethe distribution of the distinct cortical layers across these bins: bin1 corresponded to L1, bins 2 and 3 to L2/3, bins 4 and 5 to L4,bin 6 to L5A, bin 7 to L5B and finally bins 8, 9 and 10 to L6(Fig. S2A). Statistical analysis across bins was performed withone-way ANOVA with significance level set at P < 0.05. Ten slicesper animal were used to quantify the number of c-Fos+ neurons inVPM, Po and SpVi, while five slices per animal were examined toquantify PrV c-Fos+ cells, encompassing the whole rostrocaudalextent of these nuclei. Using the Image J Cell Counter plug-in(NIH), cell numbers in the ‘active’ and ‘silent’ sides werecompared. Statistical comparisons were done using paired Student’st-tests.In the noxious stimulation experiments, data were compared

between capsaicin- and vehicle-injected mice. C-Fos+ neurons werequantified at the level of C1 and C2 cortical columns on five sec-tions per animal. Quantification of laminar distribution was per-formed using ten equal bins as described above. Statistical analysisacross bins was performed with one-way ANOVA with significancelevel set at P < 0.05. Ten slices per animal for the VPM, Po andSpVi, and five slices for the PrV, were used to quantify the number

of c-Fos+ neurons, encompassing the whole rostrocaudal extent ofthese nuclei. Laminar cellular densities in the cortex weredetermined using DAPI (40,6-diamidino-2-phenylindole) staining.Statistical comparisons were done using unpaired Student’s t-tests.Colour-coded density plots of the distribution of c-Fos+ cells wereperformed using the 2-D Histogram Calculation (hist2) function ofMATLAB.

Results

Po TC axonal arbors are insensitive to ION sectioning duringdevelopment

During the first few postnatal days, VPM TC neurons require nor-mal mechanoreceptive input from the whisker pad in order to arbor-ise into discrete whisker-related L4 domains known as barrels,defining the S1BF. In the absence of whisker pad input, such asproduced by neonatal section of the ION, axonal arborisation ofVPM TC neurons is dramatically perturbed (Van der Loos & Wool-sey, 1973; Jensen & Killackey, 1987; Rebsam et al., 2005). In orderto investigate whether Po TC neurons likewise require whisker padinput during development, we sectioned the ION in newborn miceand anterogradely labelled Po and VPM neurons to investigatepotential cell type-specific, input-dependent changes in their axonalarbors. As the Po is a cellularly heterogeneous nucleus (Clasc�aet al., 2012; Ohno et al., 2012), we first identified the sub-nuclearlocation of S1BF-projecting Po neurons using retrograde labellingby focal microinjections of Fast Blue (FB) in S1BF (Fig. 2A, leftpanel). The stereotaxic coordinates of FB retrogradely-labelled Po(and VPM) neurons were determined (Fig. 2A, right panel), andused to define the location of S1BF-projecting Po (and VPM) neu-rons for anterograde labelling using micro-iontophoresis of BDA(Fig. 2B and C, Fig. S1).Following ION section, input-deprived VPM axons still branched

and extended throughout L4, but they lost their barrel compartmen-talisation confirming previous findings (Fig. 2D and E, Table 1; alsoVan der Loos & Woolsey, 1973; Jensen & Killackey, 1987; Wim-mer et al., 2010). The overall lamina-specific distribution of axonsand boutons (i.e. putative synapses) was unchanged (Fig. S1). How-ever, quantifications at the single-cell level, performed by normalis-ing these values to the number of labelled neurons for eachinjection, revealed an L4-specific decrease in bouton numbers andcorresponding increase in interbouton distance for individual VPMTC neurons following ION section (Fig. 2F and Table 1; L4boutons/neuron, ION sectioning vs. control, �43.9%, t4 = 3.67,P = 0.021; L4 interbouton distance, ION sectioning vs. control,1.5-fold increase, t4 = �8.56, P = 0.001).In stark contrast, Po TC neurons were unaffected by ION

sectioning, as the laminar distribution of axons and boutons, andinter-bouton distance, were not significantly changed, even whenvalues were normalised for individual cells (Fig. 2G–I and Table 1).The only evident effect was a more homogenous distribution ofunbranched L1-directed Po axons within L4, related to the loss ofcompartmental separation of septae and barrels in this layer(Fig. 2G). As the time course of axonal extension of VPM and PoTC neurons in the cortex largely overlaps (Galazo et al., 2008;Kichula & Huntley, 2008), the different outcome of ION sectioningis unlikely to reflect distinct stage-specific susceptibilities to inputmodification but rather reflects overall differences in inputsensitivities. Together, these results indicate that, unlike VPM TCprojections, Po TC projections are morphologically largely insensi-tive to whisker pad input during development, suggesting that the

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

Paralemniscal transmission of nociceptive input 1457

lemniscal and paralemniscal pathways may serve distinct sensoryfunctions.

Dual whisker environmental exploration strongly activates thelemniscal pathway

We next investigated whether lemniscal and paralemniscal neuronsare differentially activated by whisker contacts. For this purpose, wequantified the pathway-specific expression of c-Fos, an immediate–early gene that is expressed following plastic events or intensespiking, as demonstrated in vivo and using optogenetic approaches(Bisler et al., 2002; Staiger et al., 2002; Schoenenberger et al.,2009; Wagener et al., 2010; Kim et al., 2013).We first examined whether lemniscal and paralemniscal neurons

expressed c-Fos upon deflection of single whiskers. For this purpose,

we performed back-and-forth deflections of the C2 whisker at 4 Hzusing a piezoelectric ceramic element in lightly anesthetised, head-fixed mice (Gambino & Holtmaat, 2012). This controlled stimulationprotocol lead to weak activation of neurons in L4 of S1BF and in theVPM, indicating activation of the lemniscal pathway, while Po neu-rons and their laminar cortical targets (L5A) remained inactive(Fig. 3A). Lack of c-Fos expression in Po neurons did not, however,reflect an inability of these neurons to express this transcription factorwhen activated. Injection of the GABA receptor antagonist bicucul-line into deep cortical layers of S1BF led to strong increases in c-Fosexpression in the ipsilateral Po, reflecting synaptic responses to disin-hibition of corticothalamic S1BF neurons (Fig. S2B; Theyel et al.,2010). These findings validate c-Fos as a sensitive and specific repor-ter of neuronal activity in our system and suggest differential activa-tion of lemniscal and paralemniscal pathways by whisker contacts.

A

D E F

G H I

B C

Fig. 2. Po TC axonal arbors were insensitive to ION sectioning during development. (A) The sub-nuclear location of S1BF-projecting VPM and Po neuronswas determined by retrograde labelling from S1BF. Location of the injection site is shown on a ‘flat’ serial reconstruction of S1BF on the left; right: cameralucida reconstruction of the soma of retrogradely-labelled VPM and Po neurons. (B and C) Representative examples showing location of VPM and Po injectionsites used for BDA anterograde labelling. The BDA microinjection centre is shown on cytochrome oxidase-counterstained thalamus sections. The interauralcoronal level of each section (in mm) is indicated on the upper right corner. (D) Section of the ION at birth perturbed the characteristic discontinuous distribu-tion of VPM TC arbors in L4 of S1BF. Camera lucida drawings made from three consecutive sections. The dashed line in the photomicrograph delineates theborder between L4 and L5A. Arrowheads indicate L4 barrels. (E) Flat S1BF reconstructions showing increased tangential dispersion of BDA-labelled VPMaxons following ION section. (F) Stereological quantification of axonal length, bouton number and interbouton distance per VPM neuron. Values aremean � SEM; *P < 0.05; **P < 0.001. (G) Section of the ION at birth did not affect Po TC arbors in S1BF. Dashed line delineates the border between L4and L5A. Blue asterisk indicates location of a barrel (unlabelled). (H) The tangential distribution of BDA-labelled Po axons was unchanged following IONsection. (I) Stereological quantification of axonal length, bouton number and interbouton distance per Po neuron. Values are mean � SEM. Scale bars, 300 lm(A, E and H), 500 lm (B and C), 250 lm (D and G).

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

1458 L. Frangeul et al.

In order to examine these two pathways in a more physiologi-cal setting, we next determined the activation of lemniscal andparalemniscal neurons in freely moving mice exploring an enrichedenvironment. In order to quantify whisker-specific activation allalong the whisker-to-cortex pathway, we silenced input on one sideof the snout by clipping all whiskers, while only two adjacentwhiskers (C1 and C2) were left contralaterally (Wagener et al.,2010). Under these conditions, only these two whiskers are used forexploration (the active whisker pair), while the contralateralcorresponding silent whiskers are used as internal controls (silentwhisker pair).As predicted, lemniscal neurons were strongly activated by dual-

whisker environmental exploration: neurons in the cortical columnscorresponding to whiskers C1 and C2 strongly expressed c-Fos(Fig. 3B), and were located mostly within L4, the principal laminartarget of VPM TC neurons, with an average four-fold increase in c-Fos+ neurons compared to the corresponding contralateral silentwhiskers (Fig. 3C; number of c-Fos+ neurons: L4 silent, 250 � 27;L4 active, 957 � 23; P = 0.005).Accordingly, C1C2 whisker-specific neurons of the VPM and in

the PrV strongly expressed c-Fos (Fig. 3D–F), with an average six-fold increase in c-Fos+ neurons compared to the contralateral, silentside (number of c-Fos+ neurons: VPM silent, 24 � 8; VPM active,216 � 57; P = 0.029; PrV silent, 11 � 2; PrV active, 74 � 6;P = 0.016). Similarly, when normalised to the total number of neu-rons in each layer, about two-thirds of L4 neurons in active columnsexpressed c-Fos (47.6 � 1.1%; colour-coded in Fig. 3C; silent col-umns, 12.5 � 1.3%; P = 0.0016).In striking contrast, paralemniscal neurons of the SpVi, Po and

L5A were not detectably activated by dual whisker environmentalexploration (Fig. 3C–F), as their expression of c-Fos was unchangedcompared to the silent side (number of c-Fos+ neurons: L5A silent,57 � 5; L5A active, 95 � 11; P = 0.12; Po silent, 68 � 11; Poactive, 62 � 16; P = 0.68; SpVi silent, 87 � 8; SpVi active,77 � 11; P = 0.07). Pathway-specific activation of lemniscalneurons did not result from the acute clipping of neighbouringwhiskers and lack of activation of paralemniscal neurons did notresult from the lack of multi-whisker input, as unclipped miceexploring their environment showed similar laminar distribution ofc-Fos+ neurons when quantified within a single barrel column (Fig.S2C). Therefore, together with the developmental insensitivity of PoTC neurons to whisker input, the findings above indicate that

lemniscal and paralemniscal pathways are sensitive to distinctplasticity paradigms.

The paralemniscal pathway is specifically activated by noxiousstimulation of the whisker pad

Given the lack of c-Fos expression in paralemniscal pathway neu-rons during dual-whisker environmental exploration, we hypothe-sised that these neurons might be activated by distinct plasticityparadigms, including non-mecanoreceptive sensory input, such asnoxious stimuli. To investigate this possibility we performed a non-tactile noxious stimulation of the whisker pad in freely movingmice. For this purpose, we unilaterally microinjected capsaicin, analgogenic chemical that activates TRPV1 receptors on nerve termi-nals (Le Bars et al., 2001; Noma et al., 2008), into the whisker pad.We next examined paralemniscal and lemniscal neuron activation inresponse to capsaicin or vehicle microinjections using c-Fos expres-sion. Consistent with our hypothesis, single capsaicin microinjec-tions produced a strong contralateral increase in the number ofc-Fos-expressing neurons in L5A, the main laminar target of Po TCneurons in S1BF (Fig. 4A and B; proportional change 6.6 � 2.3compared to vehicle: L5A c-Fos+ neurons, vehicle, 24 � 6; capsai-cin, 127 � 2; P < 0.001). Accordingly, c-Fos-expressing neurons inPo and SpVi were increased approximately six-fold compared tovehicle-injected mice and were evenly distributed throughout thisnucleus (Fig. 4C–F; number of c-Fos+ neurons: Po vehicle,173 � 48; Po capsaicin, 1505 � 425; P = 0.03; SpVi vehicle,137 � 30; SpVi capsaicin, 495 � 73; P = 0.01). When normalisedto the total number of neurons in each layer, 16.5 � 0.3% of L5Aneurons expressed c-Fos after capsaicin injection (color-coded inFig. 4B; vehicle, 3.1 � 0.8%; P < 0.001).In contrast, c-Fos expression remained unchanged in L4 and in

lemniscal neurons of the VPM and PrV, indicating that activation ofparalemniscal neurons by noxious whisker pad stimulation is a path-way-specific effect (Fig. 4B–F; number of c-Fos+ neurons: L4 vehi-cle, 144 � 53; L4 capsaicin, 98 � 28; P = 0.49; VPM vehicle,165 � 98; VPM capsaicin, 132 � 39; P = 0.76; PrV vehicle:71 � 10; PrV capsaicin, 58 � 8; P = 0.95). Together, these resultsindicate that sensory information from the whisker pad reaches thesomatosensory cortex via distinct and complementary channels, andthat the paralemniscal pathway can be activated for transmission ofnociceptive input.

Table 1. Quantification of TC axon morphology in control and ION-sectioned mice

Layer

Axon length/neuron (mm) Boutons/neuron Interbouton distance (lm)

Control ION-section P Control ION-section P Control ION-section P

VPM1 0.05 � 0.01 0.04 � 0.03 0.581 5.38 � 1.18 5.70 � 2.86 0.922 10.12 � 0.00 4.18 � 3.00 0.1862/3 2.21 � 1.15 0.57 � 0.18 0.10 467.16 � 275.66 143. 34 � 17.27 0.361 5.34 � 0.87 3.91 � 1.08 0.3604 8.23 � 0.69 7.05 � 1.23 0.452 1464.99 � 79.23 820.65 � 156.64 0.021* 5.60 � 0.19 8.67 � 0.30 0.001**5A 1.01 � 0.11 1.28 � 0.32 0.478 115.68 � 34.62 91.54 � 30.75 0.630 9.68 � 1.54 17.33 � 7.89 0.9965B 1.99 � 0.52 1.38 � 0.21 0.339 339.18 � 114.77 142.26 � 39.60 0.180 6.18 � 0.69 12.41 � 5.15 0.3496 2.85 � 0.97 1.57 � 0.30 0.273 258.85 � 120.36 115.89 � 4.54 0.357 22.29 � 12.14 13.36 � 2.16 0.540

Po1 1.52 � 0.11 3.32 � 1.38 0.319 258.91 � 35.22 692.93 � 179.30 0.132 6.03 � 0.80 4.43 � 0.79 0.2282/3 1.07 � 0.46 0.55 � 0.55 0.518 179.42 � 86.64 253.99 � 127.61 0.654 7.21 � 1.45 2.73 � 1.37 0.104 0.96 � 0.30 1.79 � 0.76 0.369 115.96 � 30.51 386.47 � 178.58 0.210 7.95 � 0.72 5.13 � 0.58 0.105A 6.03 � 1.46 9.55 � 2.20 0.253 1276.14 � 297.55 1883.25 � 319.56 0.237 4.82 � 0.57 5.13 � 0.83 0.7735B 1.40 � 0.24 2.56 � 1.41 0.996 254.30 � 52.63 326.14 � 122.27 0.618 5.65 � 0.48 7.92 � 2.20 0.5186 1.61 � 0.36 1.29 � 0.22 0.213 180.47 � 40.78 292.29 � 140.07 0.410 8.96 � 1.14 6.70 � 2.48 0.329

Significant (*P < 0.05) and highly significant (**P < 0.001) statistical differences between control and ION-sectioned mice. Mean values are � SEM.

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

Paralemniscal transmission of nociceptive input 1459

A

B

D

E F

C

Fig. 3. Environmental exploration strongly activated the lemniscal pathway. (A) Back-and-forth mechanical deflections of the C2 whisker induced c-Fosexpression in corresponding L4 neurons of the contralateral C2 barrel and VPM (blue arrowheads). (B) Environmental exploration (EE) using the C1–C2 whis-ker pair strongly activates contralateral L4 neurons. Blue arrowheads delineate the C2 barrel. Control: C2 barrel on the contralateral silent (i.e. lacking whiskers)side. (C) Quantification of bin-specific (empty circles) and layer-specific (bars) proportional changes in c-Fos+ neuron numbers in response to EE. Graph barsare color-coded to indicate the percentage of c-Fos+ neurons in each corresponding layer for active whiskers. (D) EE induced C1–C2 whisker-specific expres-sion of c-Fos in lemniscal neurons of the PrV and VPM. ‘PrV active’ inset: 50 lm. (E) Quantification of c-Fos+ neurons in the thalamus and brainstem foractive and silent whiskers. (F) c-Fos+ neuron ratio in the lemniscal and paralemniscal pathways between active and silent sides (values: mean � SEM,*P < 0.05; **P < 0.001 for C, E and F). Scale bars, 100 lm (main panels), 50 lm (insets in D for PrV).

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

1460 L. Frangeul et al.

A

C

D

E F

B

Fig. 4. Paralemniscal pathway neurons were specifically activated by noxious stimuli. (A) L5A and L1 neurons in S1BF (arrowheads), which receive inputfrom Po TC neurons, were specifically activated by unilateral noxious stimulation of the whisker pad with capsaicin. (B) Quantification of bin- (empty circles)and layer-specific (bars) proportional changes in c-Fos+ neuron numbers following capsaicin injection. Graph bars are colour-coded to indicate the percentage ofc-Fos+ neurons in each corresponding layer for capsaicin. (C) Noxious stimulation of the whisker pad induced c-Fos expression in paralemniscal neurons of theSpVi and Po. (D) Colour-coded density plots of the distribution of c-Fos+ cells at distinct anteroposterior levels of the thalamus (pixel side, 50 lm). (E) Quanti-fication of c-Fos+ neurons in the thalamus and brainstem for capsaicin and vehicle-injected mice. (F) Capsaicin/vehicle c-Fos+ neuron ratio in lemniscal andparalemniscal pathways (mean � SEM. *P < 0.05; **P < 0.001 for B, E and F). Scale bars, 100 lm.

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

Paralemniscal transmission of nociceptive input 1461

Discussion

Our results indicate that the lemniscal and paralemniscal pathwayshave distinct and complementary functional properties. While whis-ker input critically controls the developmental connectivity of VPMTC neurons, the development of Po TC neurons is markedly lessdependent on whisker pad input, suggesting fundamental differencesin the sensory functions of the two pathways in adult circuits. Sup-porting such a distinction, exploration of the environment robustlyactivates lemniscal neurons while paralemniscal neurons in thebrainstem, thalamus and cortex are not detectably activated usingc-Fos as a marker of neuronal activity. In striking contrast, paralem-niscal neurons, but not lemniscal neurons, are specifically androbustly activated by nociceptive whisker pad stimuli, revealing acritical sensory modality-specific function for the paralemniscalpathway in vivo.Supporting a role for paralemniscal neurons in the detection and

processing of painful stimuli, previous studies in the cat using fieldrecordings have showed that regions of the posterior thalamus canrespond to pinches and painful electrical stimulation of the limbs(Poggio & Mountcastle, 1960; Guilbaud et al., 1977; Guilbaud,1985). Furthermore, type C (i.e. nociceptive) fibres are present inthe ION and project to the SpVi (Hayashi, 1985; Dallel et al.,1988). In addition, electrophysiological observations have suggestedthat at least subsets of Po TC neurons can be activated by noxiousstimulation (Sewards & Sewards, 2002; Masri et al., 2008; Masri &Keller, 2012), and it has been proposed that Po overactivity plays arole in central pain syndrome (Masri et al., 2009). Our study bringsa systemic understanding to these independent anatomical and physi-ological observations by demonstrating the coordinated activation ofneurons along the paralemniscal pathway upon noxious stimulation.This activation of the paralemniscal pathway by noxious stimula-

tion was remarkably specific, as neurons of the lemniscal pathwaywere not detectably activated. This specificity was robust even forthe distinct cortical laminar targets of Po and VPM neurons, as L5Aneurons contralateral to the noxious stimulus expressed c-Fos, butL4 neurons did not; these latter are the main VPM targets in S1BF.Interestingly, using extracellular recordings in anesthetised rats,Guilbaud and collaborators (Lamour et al., 1983) have reported thatdeep cortical layers in S1 respond to painful stimulation while moresuperficial ones mostly do not. Furthermore, the secondary somato-sensory cortex (S2) and primary motor cortex (M1) contralateral tothe painful stimulus, which are also targets of Po TC neurons(Galazo et al., 2008; Ohno et al., 2012), were also activated in ourstudy (data not shown), providing additional indication of paralem-niscal specificity. Interestingly, c-Fos expression was also increasedin the VPL (Fig. 4D), consistent with the findings of Hains and col-laborators in chronic neuropathic pain after spinal cord lesion (Hainset al., 2006). Furthermore, SpVc neurons, which are activated bynoxious craniofacial stimuli (Noma et al., 2008) and project to othertrigeminal nuclei and to pain-processing stations such as the parabra-chial nucleus rather than to the thalamus (Furuta et al., 2008), werealso activated by whisker deflections in our studies (Fig. S2D), sug-gesting an integrative function of SpVc neurons within the trigemi-nal nuclear complex. Central sensitisation (i.e. hyperactivity ofnociceptive neurons in the central nervous system) may have con-tributed to the neuronal activation reported here. Interestingly, ourresults indicate that this process is neuron type-specific, as paralem-niscal pathway neurons are preferentially activated.As S1BF L5B corticofugal neurons, which send strong projec-

tions to the Po (Killackey & Sherman, 2003; Theyel et al., 2010;Sherman, 2012), were not activated by noxious stimulation, Po

responses very probably reflected activation via ascending paralem-niscal input rather than descending corticothalamic input. Lack ofc-Fos expression in paralemniscal neurons during dual-whiskerenvironmental exploration is not at odds with the well-establishedelectrophysiological activation of Po neurons during whisker move-ment (Ahissar et al., 2000; Alloway, 2007; Kleinfeld & Deschenes,2011). Indeed, levels of c-Fos expression largely follow the totalnumber of spikes per neuron (Schoenenberger et al., 2009), andchanges in the timing of whisker-evoked spiking, which are thoughto be key in transmission along this pathway (Oberlaender et al.,2011), might not be detected using this indicator. Besides, overalldisinhibition of corticothalamic input with bicuculline markedlyincreased activity of Po, in line with the strong ‘top-down’ controlsover Po activity (Theyel et al., 2010; Sherman, 2012). Althoughmotor activity could in principle have contributed to neuronal acti-vation in capsaicin-injected mice, these mice were in fact strikinglyless active than in the dual-whisker exploration condition, remainingmostly immobile and showing decreased or minimal whisking activ-ity following the injection compared to vehicle-injected mice.As the Po is a cellularly heterogeneous nucleus (Clasc�a et al.,

2012; Ohno et al., 2012), it is possible that distinct neuronal sub-types in this nucleus have distinct functions, some specialising innociception while others are involved in corticothalamocorticalloops. In our study, activated Po neurons were present within theS1BF-projecting region, but also more caudally, suggesting thatseveral Po regions may be secondarily activated following paralem-niscal activation. Therefore, the function of Po TC neurons inencoding spatial properties of whisker activity (Ahissar et al., 2000;Alloway, 2007) and relaying input between cortical areas (Theyelet al., 2010; Sherman, 2012) may be modulated by ascending para-lemniscal input in response to noxious stimuli under physiologicalconditions (Groh et al., 2013).Finally, evidence from comparative neuroanatomy strongly sug-

gests that the paralemniscal pathway is evolutionarily older than thelemniscal pathway (Butler, 2008). Early vertebrates were thus proba-bly endowed with a single principal somesthetic pathway, the para-lemniscal precursor, which conveyed both noxious and non-noxiousfacial stimuli. The sensory modality-specific activation of these twopathways reported here suggests that, while whisker-bearing mam-mals such as rodents developed the circuitry for high-resolutionanalysis of facial hair contacts via elaboration of the lemniscal path-way, detection of noxious stimuli, a primitive and fundamental abil-ity, remained confined to the paralemniscal pathway.

Supporting Information

Additional supporting information can be found in the online ver-sion of this article:Fig. S1. Microphotographs showing location of all VPM (A) and Po(B) BDA injection sites. The number of labeled axons counted asthey cross the striato-pallial border is indicated for each experiment.Scale bars: 100 lm. (C, D) Quantification of VPM and Po total axo-nal length and boutons following ION section. Values:mean � SEM. (E, F) Total individual axons effectively labeled bythe BDA microinjections was estimated by counting the axonaltrunks (arrowheads) at their point of crossing of the striato-pallialborder. Cytochrome oxidase counterstain. Scale bars E and F: 500and 100 lm respectively. (G) High-magnification of labeled terminalcortical branches of VPM axons containing bouton-like varicosities.Scale bar: 5 lm.Fig. S2. C-Fos expression reflects genuine activation of Po andVPM neurons. (A) Cytochrome oxidase staining showing the

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

1462 L. Frangeul et al.

distribution of the distinct cortical layers across bins: bin 1 corre-sponds to L1, bin 2 and 3 to L2/3, bin 4 and 5 to L4, bin 6 to L5A,bin 7 to L5B and bin 8, 9 and 10 to L6. Scale bar: 100 lm. (B)Disinhibition of corticothalamic neurons by bicuculline injection indeep layers of S1BF induces ipsilateral expression of c-Fos in Poneurons. Scale bar: 100 lm. (C) L4 and VPM neurons are stronglyactivated by exploration of the environment in unclipped mice. Bluearrowheads delineate a barrel. Quantification of bin- (empty circles)and layer-specific (bars) fold-changes in c-Fos+ neuron numbers inresponse to EE with all whiskers (unclipped mouse). Graph bars arecolor-coded to indicate the percentage of c-Fos+ neurons in eachcorresponding layer for active whiskers. Scale bars: 100 lm. (D)Dual whisker environmental exploration and noxious stimulationinduces c-Fos expression in SpVc neurons. Scale bars: 100 lm.

Acknowledgements

Work in the Jabaudon laboratory is supported by the Swiss National ScienceFoundation (PP00P3_146337), the Velux Foundation, the 3R Foundation andthe NARSAD Foundation. Work in the Clasc�a laboratory is supported bySpain’s MICINN (2010-19695). We are thankful to Florian Smets, AudreyBenoit, and Bego~na Rodr�ıguez for technical assistance and to the membersof our laboratories and Eiman Azim for helpful comments on the manuscript.We are thankful to Fr�ed�eric Gambino for the piezoelectric stimulation ofwhiskers and to Daniel Huber for help with MATLAB visualisations. Theauthors declare no conflict of interest.

Abbreviations

BDA, biotinylated dextran amine; FB, Fast Blue; ION, infraorbital branch ofthe trigeminal nerve; L, layer; M1, primary motor cortex; Po, posteriornucleus; PrV, principal trigeminal nucleus; S1, primary somatosensory cor-tex; S1BF, S1 barrel field; S2, secondary somatosensory cortex; SpVi, spinaltrigeminal nucleus interpolaris; TC, thalamocortical; VPM, ventroposterome-dial nucleus.

References

Ahissar, E., Sosnik, R. & Haidarliu, S. (2000) Transformation from temporalto rate coding in a somatosensory thalamocortical pathway. Nature, 406,302–306.

Alloway, K.D. (2007) Information processing streams in rodent barrel cortex:the differential functions of barrel and septal circuits. Cereb. Cortex, 18,979–989.

Bisler, S., Schleicher, A., Gass, P., Stehle, J.H., Zilles, K. & Staiger, J.F.(2002) Expression of c-Fos, ICER, Krox-24 and JunB in the whisker-to-barrel pathway of rats: time course of induction upon whisker stimulationby tactile exploration of an enriched environment. J. Chem. Neuroanat.,23, 187–198.

Bosman, L.W.J., Houweling, A.R., Owens, C.B., Tanke, N., Shevchouk,O.T., Rahmati, N., Teunissen, W.H.T., Ju, C., Gong, W., Koekkoek, S.K.E.& De Zeeuw, C.I. (2011) Anatomical pathways involved in generating andsensing rhythmic whisker movements. Front. Integr. Neurosci., 5, 53.

Butler, A.B. (2008) Evolution of the thalamus: a morphological and func-tional review. Thalamus Relat. Syst., 4, 35–58.

Clasc�a, F., Rubio-Garrido, P. & Jabaudon, D. (2012) Unveiling the diversityof thalamocortical neuron subtypes. Eur. J. Neurosci., 35, 1524–1532.

Crochet, S. & Petersen, C.C.H. (2006) Correlating whisker behavior withmembrane potential in barrel cortex of awake mice. Nat. Neurosci., 9,608–610.

Dallel, R., Raboisson, P., Auroy, P. & Woda, A. (1988) The rostral part ofthe trigeminal sensory complex is involved in orofacial nociception. BrainRes., 448, 7–19.

Deschenes, M., Veinante, P. & Zhang, Z.W. (1998) The organization of cor-ticothalamic projections: reciprocity versus parity. Brain Res. Brain Res.Rev., 28, 286–308.

Diamond, M.E., Armstrong-James, M. & Ebner, F.F. (1992a) Somatic sen-sory responses in the rostral sector of the posterior group (POm) and in

the ventral posterior medial nucleus (VPM) of the rat thalamus. J. Comp.Neurol., 318, 462–476.

Diamond, M.E., Armstrong-James, M., Budway, M.J. & Ebner, F.F. (1992b)Somatic sensory responses in the rostral sector of the posterior group(POm) and in the ventral posterior medial nucleus (VPM) of the rat thala-mus: dependence on the barrel field cortex. J. Comp. Neurol., 319, 66–84.

Drøjdahl, N., Nielsen, H.H., Gardi, J.E., Wree, A., Peterson, A.C., Nyeng-aard, J.R., Eyer, J. & Finsen, B. (2010) Axonal plasticity elicits long-termchanges in oligodendroglia and myelinated fibers. Glia, 58, 29–42.

Erzurumlu, R.S., Murakami, Y. & Rijli, F.M. (2010) Mapping the face in thesomatosensory brainstem. Nat. Rev. Neurosci., 11, 252–263.

Furuta, T., Timofeeva, E., Nakamura, K., Okamoto-Furuta, K., Togo, M.,Kaneko, T. & Deschenes, M. (2008) Inhibitory gating of vibrissal inputsin the brainstem. J. Neurosci., 28, 1789–1797.

Galazo, M.J., Martinez-Cerde~no, V., Porrero, C. & Clasc�a, F. (2008) Embry-onic and postnatal development of the layer I-directed (“matrix”) thalamo-cortical system in the rat. Cereb. Cortex, 18, 344–363.

Gambino, F. & Holtmaat, A. (2012) Spike-timing-dependent potentiation ofsensory surround in the somatosensory cortex is facilitated by deprivation-mediated disinhibition. Neuron, 75, 490–502.

Gheorghita, F., Kraftsik, R., Dubois, R. & Welker, E. (2006) Structural basisfor map formation in the thalamocortical pathway of the barrelless mouse.J. Neurosci., 26, 10057–10067.

Groh, A., Bokor, H., Mease, R.A., Plattner, V.M., Hangya, B., Stroh, A.,Deschenes, M. & Acs�ady, L. (2013) Convergence of cortical and sensorydriver inputs on single thalamocortical cells. Cereb. Cortex, doi: 10.1093/cercor/bht173. [Epub ahead of print].

Guilbaud, G. (1985) Thalamic nociceptive systems. Philos. T. Roy. Soc. B.,308, 339–345.

Guilbaud, G., Caille, D., Besson, J.M. & Benelli, G. (1977) Single unitsactivities in ventral posterior and posterior group thalamic nuclei duringnociceptive and non nociceptive stimulations in the cat. Arch. Ital. Biol.,115, 38–56.

Gundersen, H.J., Jensen, E.B., Kieu, K. & Nielsen, J. (1999) The efficiencyof systematic sampling in stereology–reconsidered. J. Microsc., 193, 199–211.

Hains, B.C., Saab, C.Y. & Waxman, S.G. (2006) Alterations in burst firingof thalamic VPL neurons and reversal by Nav1.3 antisense after spinalcord injury. J. Neurophysiol., 95, 3343–3352.

Hayashi, H. (1985) Morphology of terminations of small and large myelin-ated trigeminal primary afferent-fibers in the cat. J. Comp. Neurol., 240,71–89.

Jensen, K.F. & Killackey, H.P. (1987) Terminal arbors of axons projectingto the somatosensory cortex of the adult rat. II. The altered morphology ofthalamocortical afferents following neonatal infraorbital nerve cut. J. Neu-rosci., 7, 3544–3553.

Kichula, E.A. & Huntley, G.W. (2008) Developmental and comparativeaspects of posterior medial thalamocortical innervation of the barrel cortexin mice and rats. J. Comp. Neurol., 509, 239–258.

Killackey, H.P. & Sherman, S.M. (2003) Corticothalamic projections fromthe rat primary somatosensory cortex. J. Neurosci., 23, 7381–7384.

Kim, T.I., McCall, J.G., Jung, Y.H., Huang, X., Siuda, E.R., Li, Y., Song,J., Song, Y.M., Pao, H.A., Kim, R.H., Lu, C., Lee, S.D., Song, I.S., Shin,G., Al-Hasani, R., Kim, S., Tan, M.P., Huang, Y., Omenetto, F.G., Rog-ers, J.A. & Bruchas, M.R. (2013) Injectable, cellular-scale optoelectronicswith applications for wireless optogenetics. Science, 340, 211–216.

Kleinfeld, D. & Deschenes, M. (2011) Neuronal basis for object location inthe vibrissa scanning sensorimotor system. Neuron, 72, 455–468.

Lamour, Y., Guilbaud, G. & Willer, J.C. (1983) Rat somatosensory (SmI)cortex: II. Laminar and columnar organization of noxious and non-noxiousinputs. Exp. Brain Res., 49, 46–54.

Le Bars, D., Gozariu, M. & Cadden, S.W. (2001) Animal models of nocicep-tion. Pharmacol. Rev., 53, 597–652.

Masri, R. & Keller, A. (2012) Chronic pain following spinal cord injury.Adv. Exp. Med. Biol., 760, 74–88.

Masri, R., Bezdudnaya, T., Trageser, J.C. & Keller, A. (2008) Encoding ofstimulus frequency and sensor motion in the posterior medial thalamicnucleus. J. Neurophysiol., 100, 681–689.

Masri, R., Quiton, R.L., Lucas, J.M., Murray, P.D., Thompson, S.M. & Kel-ler, A. (2009) Zona incerta: a role in central pain. J. Neurophysiol., 102,181–191.

Minnery, B.S. & Simons, D.J. (2003) Response properties of whisker-associ-ated trigeminothalamic neurons in rat nucleus principalis. J. Neurophysiol.,89, 40–56.

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

Paralemniscal transmission of nociceptive input 1463

Minnery, B.S., Bruno, R.M. & Simons, D.J. (2003) Response transformationand receptive-field synthesis in the lemniscal trigeminothalamic circuit. J.Neurophysiol., 90, 1556–1570.

Nakamura, S., Narumi, T., Tsutsui, K.-I. & Iijima, T. (2009) Difference inthe functional significance between the lemniscal and paralemniscal path-ways in the perception of direction of single-whisker stimulation examinedby muscimol microinjection. Neurosci. Res., 64, 323–329.

Noma, N., Tsuboi, Y., Kondo, M., Matsumoto, M., Sessle, B.J., Kitagawa,J., Saito, K. & Iwata, K. (2008) Organization of pERK-immunoreactivecells in trigeminal spinal nucleus caudalis and upper cervical cord follow-ing capsaicin injection into oral and craniofacial regions in rats. J. Comp.Neurol., 507, 1428–1440.

Oberlaender, M., Boudewijns, Z.S.R.M., Kleele, T., Mansvelder, H.D., Sak-mann, B. & de Kock, C.P.J. (2011) Three-dimensional axon morphologiesof individual layer 5 neurons indicate cell type-specific intracortical path-ways for whisker motion and touch. Proc. Natl. Acad. Sci. USA, 108,4188–4193.

Ohno, S., Kuramoto, E., Furuta, T., Hioki, H., Tanaka, Y.R., Fujiyama, F.,Sonomura, T., Uemura, M., Sugiyama, K. & Kaneko, T. (2012) A mor-phological analysis of thalamocortical axon fibers of rat posterior thalamicnuclei: a single neuron tracing study with viral vectors. Cereb. Cortex, 22,2840–2857.

Poggio, G.F. & Mountcastle, V.B. (1960) A study of the functional contribu-tions of the lemniscal and spinothalamic systems to somatic sensibility.Central nervous mechanisms in pain. B. Johns Hopkins Hosp., 106, 266–316.

Pouchelon, G., Frangeul, L., Rijli, F.M. & Jabaudon, D. (2012) Patterning ofpre-thalamic somatosensory pathways. Eur. J. Neurosci., 35, 1533–1539.

Rebsam, A., Seif, I. & Gaspar, P. (2005) Dissociating barrel developmentand lesion-induced plasticity in the mouse somatosensory cortex. J. Neuro-sci., 25, 706–710.

Rhoades, R.W., Bennett-Clarke, C.A., Chiaia, N.L., White, F.A., Macdonald,G.J., Haring, J.H. & Jacquin, M.F. (1990) Development and lesion

induced reorganization of the cortical representation of the rats body-sur-face as revealed by immunocytochemistry for serotonin. J. Comp. Neurol.,293, 190–207.

Schoenenberger, P., Gerosa, D. & Oertner, T.G. (2009) Temporal control ofimmediate early gene induction by light. PLoS One, 4, e8185.

Sewards, T.V. & Sewards, M. (2002) Separate, parallel sensory and hedonicpathways in the mammalian somatosensory system. Brain Res. Bull., 58,243–260.

Sherman, S.M. (2012) Thalamocortical interactions. Curr. Opin. Neurobiol.,22, 1–5.

Simons, D.J. & Carvell, G.E. (1989) Thalamocortical response transformationin the rat vibrissa/barrel system. J. Neurophysiol., 61, 311–330.

Sosnik, R., Haidarliu, S. & Ahissar, E. (2001) Temporal frequency of whis-ker movement. I. Representations in brain stem and thalamus. J. Neuro-physiol., 86, 339–353.

Staiger, J.F., Masanneck, C., Bisler, S., Schleicher, A., Zuschratter, W. &Zilles, K. (2002) Excitatory and inhibitory neurons express c-Fos in bar-rel-related columns after exploration of a novel environment. Neurosci-ence, 109, 687–699.

Theyel, B.B., Llano, D.A. & Sherman, S.M. (2010) The corticothalamocorti-cal circuit drives higher-order cortex in the mouse. Nat. Neurosci., 13, 84–88.

Van der Loos, H. & Woolsey, T.A. (1973) Somatosensory cortex: struc-tural alterations following early injury to sense organs. Science, 179,395–398.

Wagener, R.J., David, C., Zhao, S., Haas, C.A. & Staiger, J.F. (2010) Thesomatosensory cortex of reeler mutant mice shows absent layering butintact formation and behavioral activation of columnar somatotopic maps.J. Neurosci., 30, 15700–15709.

Wimmer, V.C., Broser, P.J., Kuner, T. & Bruno, R.M. (2010) Experience-induced plasticity of thalamocortical axons in both juveniles and adults. J.Comp. Neurol., 518, 4629–4648.

© 2014 Federation of European Neuroscience Societies and John Wiley & Sons LtdEuropean Journal of Neuroscience, 39, 1455–1464

1464 L. Frangeul et al.