Molecular and Cellular Neuroscience 45 (2010) 245–257

Contents lists available at ScienceDirect

Molecular and Cellular Neuroscience

j ourna l homepage: www.e lsev ie r.com/ locate /ymcne

A transgenic mouse line for molecular genetic analysis of excitatoryglutamatergic neurons

Lotta Borgius ⁎, C. Ernesto Restrepo, Richardson N. Leao, Noor Saleh, Ole Kiehn ⁎Mammalian Locomotor Laboratory, Karolinska Institutet, Retzius väg 8, 17177 Stockholm, Sweden

⁎ Corresponding authors.E-mail addresses: Lotta.Borgius@ki.se (L. Borgius), O

1044-7431/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.mcn.2010.06.016

a b s t r a c t

a r t i c l e i n f o

Article history:Received 5 January 2010Revised 18 June 2010Accepted 25 June 2010Available online 1 July 2010

Keywords:GlutamateTransgenic mouseCre-recombinaseInterneuronVesicular glutamate transporters

Excitatory glutamatergic neurons are part of most of the neuronal circuits in the mammalian nervous system.We have used BAC-technology to generate a BAC-Vglut2::Cre mouse line where Cre expression is driven bythe vesicular glutamate transporter 2 (Vglut2) promotor. This BAC-Vglut2::Cre mouse line showed specificexpression of Cre in Vglut2 positive cells in the spinal cord with no ectopic expression in GABAergic orglycinergic neurons. This mouse line also showed specific Cre expression in Vglut2 positive structures in thebrain such as thalamus, hypothalamus, superior colliculi, inferior colliculi and deep cerebellar nuclei togetherwith nuclei in the midbrain and hindbrain. Cre-mediated recombination was restricted to Cre expressingcells in the spinal cord and brain and occurred as early as E 12.5. Known Vglut2 positive neurons showednormal electrophysiological properties in the BAC-Vglut2::Cre transgenic mice. Altogether, this BAC-Vglut2::Cre mouse line provides a valuable tool for molecular genetic analysis of excitatory neuronal populationsthroughout the mouse nervous system.

.Kiehn@ki.se (O. Kiehn).

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

Introduction

Excitatory glutamatergic neurons are located throughout themammalian brain and spinal cord and are important components inthe majority of neuronal circuits in the nervous system. Vesicularglutamate transporters are responsible for transport of glutamate intopresynaptic vesicles and so far three different vesicular glutamatetransporters – Vglut1-3 – have been described (Aihara et al., 2000;Bellocchio et al., 2000; Fremeau et al., 2002; Herzog et al., 2001;Takamori et al., 2002; Takamori et al., 2000, 2001; Varoqui et al.,2002). Today these transporters are used as tools for functionalstudies of glutamatergic neurons, since the combined expression ofthe Vglut isoforms seems to cover all known glutamatergic neurons(Fremeau et al., 2004a; Fremeau et al., 2004b; Moriyama andYamamoto, 2004) with a distribution of expression that to a largeextent is complementary. Vglut1 expression is high in the cerebralcortex and hippocampus, while Vglut2 is expressed in thalamus,hypothalamus, inferior and superior colliculi, many nuclei in the mid-and hindbrain, and in the spinal cord (Bai et al., 2001; Fremeau et al.,2001; Varoqui, et al., 2002). Vglut3 is more sparsely expressed inselected nuclei in the midbrain (Fremeau, et al., 2002; Herzog et al.,2004; Schafer et al., 2002). In addition glutamate transporters areexpressed in certain non-glutamatergic neurons that may useglutamate as a co-transmitter such as dopaminergic, GABAergic,

serotoninergic and peptidergic neurons (Dal Bo et al., 2004; Gras et al.,2002; Kawano et al., 2006; Ponzio et al., 2006). All three isoforms havealso been detected in astrocytes (Montana et al., 2006). Concordantwith the expression pattern, previous studies have shown that Vglut1is important for glutamatergic transmission in hippocampus (Fremeau,et al., 2004b; Wojcik et al., 2004), while studies of Vglut2 knockoutshave shown impaired thalamic glutamatergic transmission (Moecharset al., 2006). In addition the Vglut2 knockoutmice die at birth due to thedisruption of the respiratory local network in the brainstem (Moechars,et al., 2006; Wallen-Mackenzie et al., 2006).

Genetically modified mice have become a powerful tool in theanalysis of neuronal networks (Callaway, 2005; Crone et al., 20082009; Dougherty and Kiehn, 2010; Gosgnach et al., 2006; Hinckley etal., 2005; Kullander et al., 2003; Lanuza et al., 2004; Luo et al., 2008;Sohal et al., 2009; Tan et al., 2006; Tan et al., 2008; Thoby-Brisson etal., 2009; Tsai et al., 2009; Wilson et al., 2005; Zhang et al., 2007;Zhang et al., 2008). An important approach in such analyses is the useof cell specific driven expression of Cre recombinase in combinationwith conditionalmouse lines. This strategy canprovide both cell specificlabeling and inactivation/ablation or activation of cells (Callaway, 2005;Crone, et al., 2008; Gosgnach, et al., 2006; Luo, et al., 2008; Sohal, et al.,2009; Tan, et al., 2006; Tan, et al., 2008; Thoby-Brisson, et al., 2009; Tsai,et al., 2009; Zhang, et al., 2007; Zhang, et al., 2008).

Local neuronal networks called central pattern generators, or CPGscontrol rhythmic motor behaviors. For instance, respiration andlocomotion are two biological functions known to be controlled bysuch specialized networks located in the brainstem and spinal cord,respectively (Feldman and Del Negro, 2006; Grillner, 2003; Kiehn,

Fig. 1. (A) Schematic overview of the BAC-Vglut2::Cre construct. A targeting cassettecontaining 5′ and 3′ sequences surrounding the predicted ATG site of the Slc17a6 genetogether with the iCre sequence and the β-lactamase gene flanked by two FRT sites wasinserted into the RP23-84M15 BAC clone using recombination in bacteria. After theremoval of the β-lactamase gene, the purified and linearized 201 kb BAC clone wasinjected into pronuclei obtained from C57BL/6 mice. (B and C) Cre expression in the E12.5 BAC-Vglut2::Cre embryo. (B) Para-sagittal section of a E 12.5 embryo showing Creexpression in the spinal cord (SC), the dorsal root ganglia (DRG), the ventral hindbrain(HB), surrounding themescencephalic (M) and telencephalic (T) ventricles as well as inthe diencephalon (D). (C) Transverse section of the E 12.5 embryonic spinal cord anddorsal root ganglia at the lumbar level. Scale bars: (B) 1 mm; (C) 0.1 mm.

246 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

2006; Kiehn and Butt, 2003). These networks are organized by amixture of excitatory and inhibitory neurons. In the mammalianspinal locomotor CPG, the main excitatory drive appears to beglutamatergic and the majority of the spinal excitatory glutamatergicneurons express Vglut2 (Kiehn et al., 2008).

Here we report the generation and characterization of a transgenicmouse line that shows specific expression of Cre in Vglut2 positiveneurons in the spinal cord and in brain areas known to express Vglut2.This transgenic mouse line provides a valuable tool by itself or incombination with other conditional alleles containing loxP sites, toidentify and analyze the functional role of excitatory glutamatergiccomponents in neuronal circuits throughout the nervous systemincluding the locomotor networks in the spinal cord.

Results

Generation of the BAC-Vglut2::cre transgenic mouse line

We have used bacterial artificial chromosomal (BAC) clone RP23-84M15 to direct the expression of the codon-improved Cre recombi-nase, iCre (Shimshek et al., 2002). This clone contained the completeslc17a6 gene (Vglut2) including a 96 kb sequence upstream and a56 kb sequence downstream of the gene. BAC-technology was used toincrease the probability for a temporal and cell-specific expressionthat does not interfere with the intrinsic gene function. The BAC clonewas modified using a targeting cassette containing 5′ and 3′sequences that are homologous to sequences surrounding thepredicted ATG site of the slc17a6 gene, the Cre recombinase openreading frame and the ampicillin resistance gene (bla) flanked by twoFRT sites in the same orientation (see Fig. 1A). The targeting cassettewas inserted into the BAC clone using recombination in bacteria at aposition that ensured the inactivation of the slc17a6 gene. After theremoval of the ampicillin resistance gene, the purified and linearized201 kb BAC-clone was injected into pronuclei obtained from C57BL/6mice.

Out of sixteen animals born three were positive founders. Thesethree founders all showed similar and restricted Cre expression in thenewborn spinal cord, that corresponded well with the previouspublished Vglut2 mRNA expression pattern. Due to the similaritybetween the founders we have focused on one of these lines.

Analysis of Cre expression

To determine the pattern of Cre expression, we first looked atembryos from BAC-Vglut2::Cre mice from embryonic day (E) 12.5(Fig. 1B and C). For expression studies, a specific Cre-antibody(Kellendonk et al., 1999) was used. At E 12.5, high Cre expression wasdetected in the spinal cord, in the dorsal root ganglia (DRG) andthroughout various areas of the brain, such as the ventral hindbrain,especially the pons and tegmentum, the diencephalon (thalamus,hypothalamus), the dorsal mesencephalon but also lateral to themesencephalic ventricle and surrounding the fourth ventricle. Theexpression surrounding the mesencephalon and the telencephalonwas always seen in the post-mitotic cell layers and never in theventricular zone immediately bordering the ventricles. In additionhigh expression was seen in the septal area and in the trigeminalganglion (data not shown). No Cre expression was detected outsidethe CNS.

We next looked at Cre expression in the nervous system ofnewborn (age 0–2 days) BAC-Vglut2::Cre mice (Fig. 2). In the lumbarspinal cord (Fig. 2A), there was a distinct pattern of Cre-expressingcells in the ventral and intermediate area, including the area thatcontains the locomotor network (Kjaerulff and Kiehn, 1996; Kremerand Lev-Tov, 1997). This expression pattern corresponds to previousreports of Vglut2 expression in the spinal cord (Crone, et al., 2008;Kullander, et al., 2003).We also found a strong expression of Cre in the

dorsal horn where Cre positive cells constituted 66.8±1.1% (n=2) ofthe total number of Cre-GAD67-glycine positive cells in laminae I–IV.GAD67and glycine-positive cells constituted 28.8±0.8% and 5.0±0.2%,respectively in these laminae. This expression pattern correlates withthe expression pattern found in previous studies showing that 55–70%of the cells in the superficial dorsal horn (laminae I–III) areglutamatergic (Todd et al., 2003; Todd and Sullivan, 1990). A largenumber of Cre positive cells were also found in the DRG (Fig. 2B).

In newborn brain from BAC-Vglut2::Cre mice (Fig. 2C–G), highlevels of Cre expression were found in the thalamus, hypothalamus,superior and inferior colliculi, deep cerebellar nuclei and central grayand in different nuclei in the hindbrain including the formationreticularis and the vestibular nuclei. Lower levels of expression wereseen in the frontal cerebral cortex and hippocampus. In addition,expression was also found in the medial septum and in the cochlearnucleus. All of these areas have previously been shown to expressVglut2 (Fremeau, et al., 2001; Herzog, et al., 2001). No expression wasfound in areas such as the caudate putamen or cerebellar cortex, areasknown to lack Vglut2 expression (Fremeau, et al., 2001; Herzog, et al.,2001).

We finally investigated the Cre expression in 4-wk old BAC-Vglut2::Cre mice (Fig. 3). Overall the Cre expression pattern at thisage resembles the expression pattern seen early postnatally. In thespinal cord (Fig. 3A) there was a distinct pattern of Cre positive cells inthe ventral and intermediate area, together with a strong expressionin the dorsal horn. In the DRG, (Fig. 3B) Cre expression occurredmainly in small and medium sized cells, corresponding to previousstudies in adult rat (Brumovsky et al., 2007). In the brain (Fig. 3C–H),expression was seen in the hindbrain, the formation reticularis, andthe vestibular nuclei. In addition, expression was seen in thalamus,hypothalamus, superior and inferior colliculi, in the deep cerebellar

Fig. 2. Expression of Cre in the spinal cord, DRG and brain of newborn BAC-Vglut2::Cremice. (A) Transverse section of the lumbar spinal cord showing the laminae borders. (B) Transversesectionof the lumbarDRG. (C–G)Coronal sectionsof thebrain showinghighexpressionofCre invariousnuclei in thebrainstem, inferior colliculus (IC) superior colliculus (SC),mammillarybodies (MB), thalamus (Thal), hippocampus (HP), and medial septum (MS). Expression was also found in the amygdala, especially in the medial nucleus (Me), the ventromedialhypothalamic nucleus (VMH), and the frontal cortex (CTX). Cerebellar cortex (CBX), spinal trigeminal nucleus (SPV), deep cerebellar nucleus (DCN), vestibular nucleus (Ve), reticularformation (RF), dorsal cochlear nucleus (DC), dentate gyrus (DG), substantia nigra (SN), caudate putamen (CPu) and claustrum (Cl). Scale bars: (A) 0.2 mm; (B) 0.1 mm; (C–G) 1 mm.

247L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

nuclei and in the medial septum. In hippocampus, expression wasseen throughout the CA3 and in the anterior dentate gyrus. Scatteredexpression was also seen in the frontal cerebral cortex. No expressionwas found in the caudate putamen or cerebellar cortex.

Analyses of Cre recombination

Cre recombination efficiency was evaluated by crossing the BAC-Vglut2::Cre mouse with different reporter lines, such as the Z/EGmouse harboring the GFP transgene (Novak et al., 2000), the ROSA26::lacZ mouse harboring the lacZ transgene (Soriano, 1999) and theROSA26::YFP mouse harboring the YFP transgene (Srinivas et al.,2001). For all recombination studies, no double transgenic mice wereused for breeding, instead we used the offspring from transgenic micecarrying either the BAC-Vglut2::Cre allele or the reporter allele.Embryonic Cre recombination was analyzed using BAC-Vglut2::Cre//Rosa26::lacZ double transgenic embryos. E 12.5 embryos werecollected and Cre recombination was visualized by β-galactosidasestaining (Fig. 4A). At E 12.5, the lacZ expression was strong in thespinal cord but weaker in the brain and DRG. This indicates that theCre expression is either turned on earlier or is higher in the spinal cordat this age compared to other areas of the nervous system although no

obvious differences in expression levels were detected with immu-nostaining against the Cre protein. No recombination was seenoutside the CNS.

In the newborn mice, we used the lumbar spinal cord from doubletransgenic BAC-Vglut2::Cre//Z/EG mice to look for cell-specificrecombination. The analysis showed a clear overlapping expressionof Cre and GFP (Fig. 4B–E). We quantified the overlap in the ventraland intermediate spinal cord (laminae VII–X) where 82.6±3.7%(n=3) of the Cre positive cells also expressed GFP. The recombinationefficiency was similar using the ROSA26::YFP reporter mice where81.7±3.5% (n=3) of the Cre positive cells also expressed YFP. Inaddition, 11.1±4.5% (n=3) of the GFP positive cells lacked Creexpression, indicating a transient expression of Vglut2. A few of theGFP positive but Cre negative cells showed expression of glycine. Thisobservation suggests that some spinal glycinergic cells may have atransient glutamatergic phenotype similar to what has been reportedfor glycinergic cells in retina (Sun and Crossland, 2000).

To evaluate the Cre recombination efficiency in the brain we usedBAC-Vglut2::Cre//ROSA26::lacZ mice (Fig. 4F–H). Recombination wasfound in the majority of areas where Cre was expressed, such as thethalamus, hypothalamus, superior and inferior colliculi and deepcerebellar nuclei. Recombination was also evident in the hippocampus

Fig. 3. Expression of Cre in the spinal cord, DRG and brain of 4-wk old BAC-Vglut2::cre mice. (A) Transverse section of the lumbar spinal cord. Dorsal horn (DH), ventral horn (VH).(B) Transverse section of lumbar DRG. (C–H) Coronal sections of the brain showing high expression of Cre in various nuclei in the brainstem, superior and the inferior culliculus (SC, IC),medialmammillarynucleus (MM), thalamus (Thal) andmedial septum(MS). Expressionwas also found in the amygdala, especially the ventromedial hypothalamicnucleus (VMH) and inthe frontal cortex (CTX). Cerebellar cortex (CBX), spinal trigeminal nucleus (SPV), reticular formation (RF), deep cerebellar nucleus (CBN), vestibular nucleus (Ve), dorsal cochlear nucleus(DC), pons (P), hippocampal formation (HPF), substantia nigra (SN), caudate putamen (CPu). Scale bars: (A and B) 0.2 mm; (C–H) 1 mm.

248 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

and amygdala together with different nuclei in the hindbrain includingthe formation reticularis and the vestibular nuclei. No recombinationwas seen in the cerebellar cortexor caudate putamen.Althoughwehavenot quantitatively studied the cell specific recombination in thenewborn brain, it appears that the overall recombination patternoverlaps with the Cre protein expression pattern.

We further analyzed the efficiency of the Cre-mediated recombi-nation, by crossing the BAC-Vglut2::Cre mouse with a Vglut2flox/flox

mouse line that has conditionally targeted the Slc17a6 gene byinsertion of loxP sites surrounding exon 2 (Hnasko et al., 2010). SinceVglut2 knockout mice die at birth, brains and spinal cords weresampled from E 18.5 mice and analyzed for Vglut2 protein content byWestern blot analysis. As shown in Fig. 4I, double transgenicBACVglut2::Cre//Vglut2flox/flox mice lack the Vglut2 protein, showingthat the Cre expression was effective enough to drive a strong anduniversal recombination of the floxed Slc17a6 alleles in these animals.

Finally, we determined the potential germline expression of Cre bycrossing double transgenic BAC-Vglut2::Cre//ROSA26::lacZ maleswith wild-type females. The pattern of recombination was investi-gated at embryonic day E 12.5 (n=8). Fifty percent of the offspringcarrying both transgenic alleles showed specific recombination onlyin brain and spinal cord, overlapping with the Cre expression pattern.Fifty percent of the offspring showed recombination both in brain andspinal cord but also outside the CNS. None of the offspring carryingonly the ROSA26::lacZ allele showed any recombination.

Cre is expressed specifically in Vglut2 positive cells in the spinal cordand DRG

The cell specificity of Cre expression was evaluated in detail inlumbar spinal cords from newborn BAC-Vglut2::Cre mice by combiningin situ hybridization against Vglut2 mRNAwith immunohistochemistryagainst the Cre protein (Fig. 5A–C). Overall, Cre expression co-localizedwithVglut2mRNAexpression. In theventral spinal cord 86±2% (n=2)of Cre positive cells also expressed Vglut2mRNA. Half of the Cre positiveand Vglut2 mRNA negative cells were located close to the central canal.In the dorsal horn, despite theweaker in situ hybridization signal, a clearcellular co-localization could be seen between Cre and Vglut2 mRNA(Fig. 5B). The weaker in situ hybridization signal in cells located in thedorsal horn and close to the central canal is most likely due to the smallsoma sizes of these cells, leading to lower mRNA levels that are close toor below the limits of detection using in situ hybridization.

We also looked at the cell-specific expression and recombination inDRG from newborn double transgenic BAC-Vglut2::Cre//Rosa26::YFPmice (Fig. 5D–G). First, Cre-stainingwas comparedwith YFP expressionand revealed that 83.6±13.0% (n=3) of the Cre positive cells in DRGalso expressed YFP. Secondly, since the Vglut2 protein in the DRG cellscan be detected in the cell soma (Brumovsky, et al., 2007; Tong et al.,2001) we could identify Vglut2-expressing cells using a Vglut2 specificantibody. This analysis showed that 90.6±11.3% (n=3) of the Crepositive and YFP positive cells also were Vglut2 positive.

Fig. 4. Analysis of Cre-mediated recombination in E 12.5 embryos and newborn mice. (A) Para-sagittal section from E 12.5 BAC-Vglut2::Cre//ROSA26::YFP embryo showingrecombination, visualized by X-gal staining in the spinal cord (SC), dorsal root ganglia (DRG), hindbrain (HB), midbrain (M), diencephalon (D), and telencephalon (T). Norecombination was seen outside the CNS. (B–E) Recombination in the spinal cord from a newborn BAC-Vglut2::Cre//Z/EG double transgenic mouse. The overlays (B and E) showthat the majority of the Cre positive cells also express GFP. The arrowhead points to a Cre expressing, GFP positive cell, while the arrow points to a GFP positive cell lacking Creexpression. (F–H) Recombination in the brain from newborn BAC-Vglut2::Cre//ROSA26::YFP double transgenic mice. Horizontal (F) and coronal (G and H) sections showingrecombination visualized by X-gal staining, in the inferior colliculus (IC), deep cerebellar nucleus (DCN), vestibular nucleus (Ve), reticular formation (RF), dorsal cochlear nucleus(DC), hippocampus (HP), thalamus (Thal), medioventral hypothalamic nucleus (VMH), and the medial nucleus of the amygdala (Me). No recombination was seen in the striatum(ST). (I) Cre expression resulted in a total recombination event in brain and spinal cord shown by the lack of Vglut2 protein in the BAC-Vglut2::Cre//Vglut2flox/flox mice. Western blotanalysis was performed using crude synaptosomal fractions prepared from whole brains and spinal cords from E 18.5 BAC-Vglut2::Cre//Vglut2flox/flox (lane 1) or from E 18.5Vglut2flox/flox mice (lane 2). Antibodies for Vglut2 and synaptophysin (Syp) were used to detect the two proteins. The same amount of total protein was loaded in lanes 1 and 2. Scalebars: (B) 0.1 mm; (C–D) 50 μm; (G) 0.5 mm; (A, F and H) 1 mm.

249L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

To further evaluate the Cre expression specificity we assessed thepossible expression in inhibitory neurons and the expression inknown Vglut2 positive neurons in the lumbar spinal cord. Putative

glycinergic neurons were detected with an antibody against glycine(Fig. 6A–C) while putative GABAergic neurons were detected in BAC-Vglut2::Cre mice crossed with a GAD67::GFP knock-in transgenic

Fig. 5. Specific expression of Cre in Vglut2 positive cells in newborn BAC-Vglut2::Cre mice. (A–C) shows in situ hybridization against Vglut2 mRNA in combination withimmunohistochemistry against the Cre protein. (A) shows a hemisegment of a lumbar spinal cord with close-ups from the dorsal (B) and the ventral horn (C). First column Creimmunoreactivity (red), second column Vglut2 in situ hybridization reaction (black) and third column merged picture. (D–G) In DRG, both Cre expression and recombination,indicated by YFP expression, are restricted to Vglut2 positive cells. DRG from BAC-Vglut2::Cre//ROSA26::YFP double transgenic mice were immunostained for Cre (D) and Vglut2(E) and compared with the recombined YFP expression (F). (G) Overlays of the three previous panels. Lower panels D2–G2 are close ups from the panels above. Scale bars: (A, D1–G1)0.1 mm; (B and C) 50 μm; (D2–G2) 25 μm.

250 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

mouse line (Tamamaki et al., 2003) (Fig. 6D–F). The GAD67::GFPmouse line expresses GFP in large proportions of GABAergic neuronsthroughout the CNS (Tamamaki, et al., 2003), and GFP is reliablyexpressed in GABAergic neurons in the spinal cord (Restrepo et al.,2009). No ectopic Cre expressionwas observed in these populations ofcells. In contrast, Cre expression was pronounced in a number ofmolecularly identified Vglut2 positive neuronal populations in theventral spinal cord, such as Chx10-expressing neurons (Crone, et al.,2008; Lundfald et al., 2007) (Fig. 6G–H) and Hb9-expressing neurons(Wilson, et al., 2005) (Fig. 6K and L). Thus, 95±4% (n=4) of theChx10 neurons also co-expressed Cre, while 94±6% (n=4) of theHb9 cells showed overlapping although weak Cre expression. Incomparison, no expression of Cre (Fig. 6I and J; n=3) was seen in the

Fig. 6. Restricted expression of Cre in molecularly identified excitatory subpopulations buVglut2::Cre mice. (A–C) Immunohistochemistry against glycine (green), used to visualize aa lumbar spinal cord with close-ups from the dorsal (B) or the ventral horn (C). Note thVglut2::Cre//GAD67::eGFP was used to reveal possible co-expression of Cre (red) in GABAthe dorsal (E) and the ventral horn (F) respectively. Note that there was no expression offactor Chx10 expressed in Vglut2 positive V2a neurons (green) and Cre (red) shows pronoV2b neurons. (K and L) Vglut2 positive Hb9-expressing neurons express Cre. (H1–3, J1–3, L1H1–3, J1–3, L1–3) 50 μm.

inhibitory Gata2/3-expressing cells (Crone, et al., 2008; Lundfald, etal., 2007). Overall, this quantitative analysis in the spinal cord andDRGs, showed that Cre expression was specific to glutamatergic cellswith no or little ectopic expression.

Interestingly, in crosses between BAC-Vglut2::Cre mice and Z/EGmice we sometimes observed fluorescent reporter protein in the skin(data not shown). This skin expression of GFP is most likely due toexpression of fluorescent reporter in the peripheral axon of Crepositive cells located in the DRG. Previous reports have shown Vglut2immunoreactivity originating from DRG Vglut2 positive cells both indeep dermal nerve bundles and in close relation to the epidermis(Brumovsky, et al., 2007). No Cre expression was detected in the skinin the BAC-Vglut2::Cre mice at any age (data not shown).

t no expression in inhibitory neurons in the lumbar spinal cord from newborn BAC-ll putative glycinergic neurons in the spinal cord, and Cre (red). (A) Hemisegment ofat there was no co-expression of Cre and glycine immunoreactivity. (D–F) The BAC-ergic neurons (green). (D) Hemisegment of a lumbar spinal cord, with close-ups fromCre in putative GABAergic cells. (G and H) Antibody staining against the transcriptionunced co-localization. (I–J) No Cre expression was seen in inhibitory Gata2/3 positive–3) Close-ups from the panels above. Scale bars: (A, D, G, I and K) 0.1 mm; (B, C, E, F,

251L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

Fig. 7. Electrophysiological characterization of GFP and Cre positive neurons in the anteroventral cochlear nucleus (AVCN) (A–F) and themedial nucleus of the trapezoid body (MNTB)(G–K) from P 13–P 17mice. GFP positive bushy cells (A) in the AVCNof BAC-Vglut2::Cre//Z/EGmicewere filledwith Alexa 633 (B). (C) shows the overlay picture, and (D) shows theGFPpositive Calyx ofHeld terminals in theMNTB originating fromGFPpositive bushy cells. Arrows in (A–C) showaGFP labeled bushy cell. (E) Intracellular recordings fromGFP positive bushycells show outward rectification and an initial bursting in response to positive current steps. The GFP positive bushy cells do not fire action potentials at rest (−59.8±0.4 mV, n=5).(F) Elongated, GFP negative AVCN cells show tonic frequency and sag voltage response in response to positive current steps. (G) Recording fromCalyx ofHeld synapse located in theMNTBin BAC-Vglut2::Cremouse showing high frequencyfiring in response to current injection. (H) Single burst firing inMNTBprinciple cell. (I–J) Calyxof Held synapses stained for Vglut2 protein(I) and MNTB principle cells (J) filled with Alexa 633. (K) shows the overlay picture of (I–J). Arrows in panels (I–K) point to a Calyx of Held synapse. Scale bars: (C, D and K) 50 μm.

252 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

253L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

Functional characterization of known Vglut2 positive neurons

To test that Cre expression and activity did not interfere withneuronal activity, we performed intracellular recordings from knownVglut2 positive neurons with well-characterized firing properties.Although there is an abundance of Vglut2 positive neurons in thespinal cord, none of them have firing properties that can berecognized as specific for Vglut2 positive neurons (see Doughertyand Kiehn 2010). We therefore looked at cell-types in the brain thatfulfill these criteria. Bushy cells located in the anteroventral cochlearnuclear (AVCN) express Vglut2 and they project either ipsilaterally tothe lateral superior olive or contralaterally to the medial nucleus ofthe trapezoid body (MNTB) (Billups, 2005). In the MNTB, bushy cellsgive rise to the Calyx of Held synapses, one of the most comprehen-sively studied glutamatergic synapse in the brain. In the AVCN of BAC-Vglut2::Cre//Z/EG mice (ages ranging from P13–P17), we observed alarge number of spherical GFP positive neurons with morphologyresembling bushy cell morphology (Fig. 7A–C). In the MNTB of thesemice, we found GFP positive Calyx of Held synapses (Fig. 7D).Recordings from GFP-positive bushy cells showed electrophysiolog-ical properties, including outward rectification and single spiking(Fig. 7E), similar to that previously demonstrated for GFP-negativebushy cells (Leao et al., 2006). GFP negative elongated non-bushy cellsin the AVCN showed tonic firing (Fig. 7F) and a depolarizing sag inresponse to large hyperpolarizing current pulses also similar to whathas been shown before (Leao, et al., 2006). In the MNTB of BAC-Vglut2::Cre mice, Vglut2 positive Calyx of Held synapses (Fig. 7I–K)showed high frequency firing in response to current steps (Fig. 7G)and had a firing threshold of 83.3±16.7 pA (n=3) in agreement withthe known firing properties of Calyx of Held synapses in Cre-negativemice (Leao et al., 2005). The MNTB principle cells showed single burstfiring and a firing threshold of 170±12.3 pA (n=5) (Fig. 7H). Restingmembrane potentials in Calyx of Held synapses and MNTB principleneurons were −73.9±1.5 mV (n=3) and −62.5±0.8 mV (n=5).

Together our electrophysiological recordings from GFP and Crelabeled cells show that GFP or Cre expression does not interfere withnormal electrophysiological properties.

Discussion

The ability to perform genetic manipulations is fundamental forthe analysis of neuronal networks, and the most common techniqueused to target specific cell types involves cell specific expression of Crerecombinase combined with mice carrying conditional targetedalleles.

One strategy to obtain a reproducible and efficient cell-specificgene expression in vivo is to use BAC technology to create transgenicmice. (Giraldo andMontoliu, 2001; Gong et al., 2003; Heintz, 2001). Inaddition, this strategy limits the effect on the endogenous gene ex-pression since both endogenous alleles are left intact and the expres-sion from the BAC clone is disrupted. This is of particular interestwhen looking at neurotransmitter function where knock-inapproaches will interfere with the normal function even if only oneallele is affected (Heisler et al., 1998; Moechars, et al., 2006).

In this paper we describe the generation and characterization of atransgenic mouse line based on the modified BAC-clone encompass-ing the complete slc17a6 gene that expresses Cre recombinase underthe control of the regulatory elements for the Vglut2 protein. Wedemonstrate that this transgenic mouse line had a specific andrestricted expression of Cre in Vglut2 positive cells in the spinal cordand that there was no ectopic expression in inhibitory neurons. TheBAC-Vglut2::Cre mice also showed specific Cre expression in DRGand in known Vglut2 positive structures in the brain.

In BAC-Vglut2::Cre mice, cell specific recombination was demon-strated by reporter protein expression in the spinal cord, DRG and thebrain. The expression of reporter protein overall correlated with the

Cre expression pattern. Furthermore, the double transgenic mice BAC-Vglut2::Cre//Vglut2lox/lox were completely lacking the Vglut2 pro-tein, demonstrating the efficiency of the recombinase.

In the spinal cordwe identified a proportion of Cre negative cells inwhich recombination had occurred as revealed by the expression ofGFP or YFP. This might be due to ectopic expression of the reporter,however, the consistency between animals and reporters insteadindicates an early and transient expression of Vglut2 in certain cells inthe spinal cord, something that has not been reported before. Afraction of these potentially transiently Cre expressing cells showedco-expression of glycine suggesting either that some glycinergic cellstransiently co-express Vglut2 in the spinal cord similar to what hasbeen reported in cells in retina (Sun and Crossland, 2000) or thatVglut2 positive neurons switch to an inhibitory phenotype. We alsodetected areas in the brain such as the frontal part of thehypothalamus and the CA1 area of the hippocampus that had verylow Cre expression in combination with a strong recombinationevent. Although we did not look for it, other areas of the brain mightshow a similar pattern that probably reflects a transient Vglut2expression during embryonic development that is later down-regulated (Fremeau, et al., 2004b; Boulland et al., 2004). For example,some phenotypes seen in Vglut1 knockout animals in the earlydevelopmental stages are believed to partially be caused by acompensation by transiently expressed Vglut2 in neurons that inadulthood predominantly express Vglut1 (Takamori, 2006). Thephenomenon of transient expression is not specific to Vglut2 or tothe BAC-Vglut2::Cre mouse line but has been seen in other studies(Gong et al., 2007; Sahly et al., 2007; Sato et al., 2007; Xu et al., 2008)and is an unavoidable consequence of dynamic gene regulation duringembryonic life.

A proportion of the Cre positive cells in the spinal cord did notrecombine. Again, thismight correspond to a temporal aspect of Slc17a6driven Cre expression where a sudden strong Cre expression gives riseto a weak reporter expression since the recombination event has notyet occurred. We also identified areas in the brain with strong Creexpression but with a comparably weaker reporter expression, such asthe more dorsal areas of the medial nucleus of the amygdala. It is,however, important to remember that the level of expression from thereporter is not directly correlated to the level of Vglut2 expression butinstead reflects the intensity of the ROSA26 or Z/EG promoter itself. Byusing the BAC-Vglut2::Cre mice, the combined analysis of Creexpression and the recombination event together with neurotransmit-ter phenotyping will make it possible to dynamically monitor Vglut2regulationduringdevelopmentof thenervous systemorwhenplasticityis induced.

One obvious feature of our transgenic mouse line is the largenumber of Cre expressing cells, both in the spinal cord but alsothroughout the brain often with a broader expression than previouslyrevealed by in situ hybridization against Vglut2 mRNA. One reason forthis discrepancy is probably that the antibody we use is very sensitiveand detects low amounts of Cre protein with high specificity. ThisBAC-Vglut2::Cre transgenic mouse line will, therefore, be of greatassistance in identifying Vglut2 expressing neurons throughout theCNS.

Importantly, our electrophysiological studies demonstrated thatthe firing and electrical properties of the Cre or GFP positive bushycells in the AVCN and their terminals in the MNTB (Calyx of Held)were identical to the neuronal properties found in non Cre- and GFP-expressing mice.

Functional use

The reliable and widespread expression of Cre in known Vglut2expressing cells in the rodent nervous system makes the presentmouse line an asset for functional studies of neuronal networks andcellular properties. Targeting of fluorescent proteins to Vglut2

254 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

expressing cells will immediately facilitate electrophysiological andanatomical studies of glutamatergic neurons throughout the brainand spinal cord (Al-Mosawie et al., 2007; Crone, et al., 2008).Genetically driven activation and inactivation studies of molecularlydefined neuronal populations that have a combination of Vglut2expressing cells and other neurotransmitter phenotypes (Lanuza et al.2004; Goulding, 2009) will also be possible with the present mouseline. Finally, the BAC-Vglut2::Cre line is ideal to be used together withviral vectors that carries floxed gene constructs (Abbott et al., 2009;Adamantidis et al., 2007; Cardin et al., 2009; Gradinaru et al., 2009;Petreanu et al., 2009; Sohal, et al., 2009; Zhang, et al., 2007). Thepromotor specificity provided by both the viral vector and thetransgenic mouse will allow intersectional specificity of paramountimportance for attempts to decipher the function of complex neuronalcircuitries in the mammalian brain and spinal cord.

Experimental methods

Generation of the BAC-Vglut2::Cre mouse

The mouse BAC clone (RCPI23-84M15) encompassing the slc17a6gene from the female C57BL/6 J mouse BAC library was obtained fromCHORI-USA and was modified by homologous recombination in EL250bacteria. A modified pConst plasmid was used as a targeting constructcontaining the open reading frame of iCre followed by the ampicillinresistance gene (bla), flanked by two FRT-site in the same orientation.The targeting cassette was created by first generating the 5′ and 3′Vglut2 homology arms by PCR using primers 5′-GTTTAAACCGGGAATT-TACGTTGTTTTTCG-3′ and 5′-ATCCATTCTTGTAAAGACTGGTGTCCAG-3′or 5′-CATGTCTAGAGTAAGAGACTGCAACCCCATGCT-3′ and 5′-CATGGC-TAGCGTTTAAACAAAGACAGGCTTTATGAGCAG-3′. The 5′ homology armwas inserted upstream of the ATG codon for the iCre recombinase gene.The 3′ homology arm was inserted downstream of the ampicillinresistance gene. The targeting cassette was excised by PmeI digestion,and inserted by homologous recombination into the BAC clone. Therecombinationwas performed as previously described (Lee et al., 2001).Briefly, EL250 bacteria, containing the BAC84M15 clonewere incubatedfor 15 min at 42 °C to induce recombination enzymes, followed bytransformation of the targeting cassette and selection on ampicillinplates. Correctly targetedBAC clones, identifiedby restrictiondigest, stillin EL250, were incubated with 0.1% L-arabinose for 1 h to induce Flperecombinase expression, allowing the excision of the ampicillinresistance gene. Again, restriction pattern analyses were used to verifythe recombination. The 200 kb-insertwas excised byNot1 digestion andpurified using Sepharose CL4b (Amersham) before injection intopronuclei of fertilized C57BL/6 zygotes.

Additional mice lines used

Three well-characterized reported lines where used for therecombination studies 1) the Z/EGmouse harboring the GFP transgene(Novak, et al., 2000), 2) the ROSA26::lacZ mouse harboring the lacZtransgene (Soriano, 1999) and3) theROSA26::YFPmouse harboring theYFP transgene (Srinivas, et al., 2001). All of these lines carry promotorsthat are ubiquitously expressed in the nervous system. A Vglut2flox/flox

mouse (kindly provided by Drs. T Hnasko and R. Palmiter) (Hnasko,et al., 2010) was used for the additional recombination test of the BAC-Vglut2::cremouse. For the specificity studies, the GAD67::GFP knock-intransgenic mouse was used (Tamamaki, et al., 2003). The geneticbackground strain for all mice lines used was C57BL/6.

Mice breeding

Animals were bred and raised under standard animal housingconditions in a 12 h light/dark cycle. Food and water were availablead libitum. All animal experiments were performed in accordance

with the European Communities Council Directive (86/609/EEC) andthe National Institute of Health guidelines, and were approved by thelocal Animal Care and Use Committee.

Genotyping

BAC-Vglut2::Cre transgenic founders andoffspringwere identifiedbygenotyping PCR using primers 5′-GAAGCTAACAGCAGCCGCTC-3′, 5′-AGGAGCATCTTCCAGGTGTG-3′, 5′-TTGCATCGCATTGTCTGAGTAG-3′ and5′-TTCCCACACAAGATACAGACTCC-3′. The Z/EG transgenic mice weregenotypedusingprimers5′-GTCAATCCGCCGTTTGTTCCCACGG-3′ and5′-GCGTGTACCACAGCGGATGGTTCGG-3′, the ROSA26::YFP transgenicmicewere genotyped (500 bp wildtype, 250 bp mutant fragments) using theprimers, 5′-AAAGTCGCTCTGAGTTGTTAT-3′, 5′-GCGAAGAGTTTGTCCTC-AACC-3′ and 5′-GGAGCGGGAGAAATGGATATG-3′ and the Vglut2flox/flox

micewere genotypedusingprimers 5′-CAGTGTGCTGTAACTGAGATAGT-3′ and 5′-TCTTTTGGGGTGCCATTTCAACACT-3′. The BAC-Vglut2::Cretransgenic lines were maintained on a C57BL/6 background.

Dissections

Embryonic day (E) 12.5, neonatal (postnatal day (P) 0–2) or 4-wkold mice were used. The morning of plug confirmation was termed E0.5. Pregnant mice were anaesthetized deeply with isofluran (AbottScandinavia, Solna, Sweden) before performing the caesarian section.The mother was decapitated and the embryos removed. Embryoswere dissected immediately in ice-cold oxygenated (5% CO2 in O2)10% calcium Ringer's solution composed of (in mM): NaCl, 111,1; KCl,3.1; NaHCO3, 25; KH2PO4, 1.1; MgSO4, 3.7; CaCl2, 0.25; and glucose,11.1. Neonatal mice were anaesthetized with isofluoran, decapitatedand eviscerated before the spinal cord or brain was dissected out inice-cold-oxygenated ringer (Kiehn and Kjaerulff, 1996). 4-wk oldmice were anaesthetized with intra-peritoneal injection of sodiumpentobarbital (3 mg/35 g, Apoteket, Stockholm, Sverige). Animalswere perfused trans-cardially with 4% paraformaldehyde (PFA) inphosphate-buffered saline (PBS), where after the spinal cord, dorsalroot ganglia, and brains were dissected out.

Immunohistochemistry and X-gal staining

Spinal cords from neonatal mice were fixed for 20 min, brains for2 h and embryos for 1 h on ice in 4% PFA-PBS and cryoprotected in 20%sucrose overnight prior to final embedding in Tissue-Tek (SakuraFinetek, Zoeterwounde, the Netherlands) and stored at −80 °C.Brains, spinal cords, and dorsal root ganglia from 4-wk old micewere post-fixed in 4% PFA-PBS before cryo-protected and stored at−80 °C. Tissues were sectioned on a cryostat (spinal cords 12 μmsections, brains and embryos 20 μm sections). Immunohistochemistrywas performed as described before (Lundfald, et al., 2007; Restrepo,et al., 2009). To block nonspecific binding sites, slices were incubatedwith 1% Donkey Serum (Jackson), 1% Goat Serum (Jackson) and 0.1%Triton-X (Sigma) for 1 h in PBS before processing with primaryantibodies against glycine (Immunosolution (Jesmond, Australia), ratpolyclonal, IG1002, 1:7000), Chx10 (guinea pig polyclonal, 1:20,000,gift from Dr. Kamal Sharma, see Peng et al. 2007), Gata2/3 (guinea pigpolyclonal, 1:6-8000, gift from Dr. Kamal Sharam, Peng et al. 2007),Hb9 (rabbit polyclonal, 1:16,000, gift from Dr. Thomas Jessell), Vglut2(guinea pig polyclonal, 1:2500, gift from Drs. M. Watanabe and T.Hökfelt, see Brumovsky et al 2007,) GFP (chicken polyclonal, 1:2000,Abcam #13970) and Cre (rabbit polyclonal, 1:6000, a kind gift fromDr. G. Shutz, see Kellendonk et al JMB 1999). The specificity of theprimary antibodies except the Cre antibody has been tested before(Brumovsky, et al., 2007; Crone, et al., 2008; Lundfald, et al., 2007;Restrepo, et al., 2009; Wilson, et al., 2005). The Cre antibody wastested in brain and spinal cord tissue from wild-type mice embryosand newborns where it showed no staining. Pre-absorption with the

255L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

bovine serum albumin did not diminish the Cre-staining in BAC-Vglut2::Cre mice. Species-specific secondary antibodies from JacksonImmuno Research (FITC, CY2, CY3 or CY5) were used as recom-mended by the manufacturers. For X-gal staining, sections wereincubated at room temperature for 4–5 h in 0.01 M PBS pH 7.4containing 10 mM K4Fe(CN)6, 10 mM K3Fe(CN)6, 2 mM MgCl2, 0.1%Tween 20 and 1 mg/ml X-gal.

In situ hybridization for Vglut2

Digoxygenin in situ hybridization was performed essentially aspreviously described (Restrepo, et al., 2009) on 12 μm cryosectionsusing digoxygenin-labeled probe antisense to vesicular glutamatetransporter 2 (using a Vglut2 probe spanning the base pairs 540–983.;produced by Dr. L. Borgius). Hybridizationwas carried out at 60 °C anddigoxygenin probes were visualized by development with BCIP/NBTovernight. No specific signal was detected using the sense probe. Thesections were then counter-stained with rabbit anti-Cre antibodyfollowed by anti-rabbit fluorophore conjugated secondary antibody.

Cell counts and analysis

All images were obtained using either a Zeiss LSM 510 laserscanningmicroscope equippedwith Argon 488, helium-neon 543 nm,helium-neon 633 nm laser for excitation and appropriate filters fordetection, or with an Olympus BX51 X-Cite 120 fluorescence and lightmicroscope with appropriate filter settings for excitation anddetection of the different flourophores and bright light.

For cell counting, estimates of labeled cells (labeled by immuno-reactivity or in situ hybridization) were obtained from 3–5 slices from2–4 animals. Sections were taken more than 50 μm apart. The ventralspinal cord was defined as the part of the cord ventral to a horizontalline extending laterally from the dorsal border of the central canal.The ventral and intermediate area was defined as the part of the cordventral to a horizontal line extending from the ventral border of thedorsal funiculus. Co-localization was determined from high-resolu-tion pictures inspected in Adobe Photoshop. For co-localizationstudies in DRG, Cre staining was used to identify neuronal profiles.YFP labeling and Vglut2 staining overlapping that neuronal profilewas counted as co-expression. For the newborn spinal cord, To-Pro-3that stains nuclei was used together with Cre staining to identifyneuronal profiles. Overlapping labeling (GFP, YFP, Chx10, Hb9 orVglut2 mRNA) of that neuronal profile was counted as co-expression.Any cases in which more than one nucleus was evident within whatappeared to be a single neuronal profile was discarded from thecounting. This procedure is similar to what has been used in previouspublications in our lab (see Restrepo et al. 2009). For reproduction,quality of images was improved in Adobe Photoshop using onlybrightness and contrast scaling.

Western analysis

Spinal cords and brains from E 18.5mice, either BAC-Vglut2::Cre//Vglut2flox/flox mice or Vglut2flox/flox mice were homogenized in50 mM Tris–HCl pH 7.5 containing 0.32 M sucrose and 1 mM EDTAsupplemented with protease inhibitor cocktail (Roche #693159).After mild centrifugation at 2000 rpm for 10 min, the clearedsupernatant was saved and centrifuged at 20,000 rpm for 20 min.The pellets were re-suspended in ice-cold 50 mM Tris–HCl pH 7.5containing 1 mM EDTA. 20 μg of the crude synaptosomal fractionswere separated by SDS-PAGE and transferred onto polyvinylidenedifluoride membrane. The membrane was sequentially probed withantibodies against Vglut2 (rabbit polyclonal, 1:5000, #135403,Synaptic Systems, Goettingen) and synaptophysin (mouse monoclo-nal, 1:500, #S5768, Sigma), followed by secondary antibody AlexaFlour 680 (rabbit or mouse, 1:10,000, Molecular Probes) and

developed using Odyssey Infrared Imaging System ver. 2.1 from Li-Cor Biosciences.

Electrophysiological recordings from Vglut2 positive neurons

BAC-Vglut2::Cre//Z/EG, BAC-Vglut2::Cre or wild-type P 13-P 17mice were anaesthetized with isofluoran before being decapitated.The brains were rapidly removed and placed in ice-cold artificialcerebrospinal fluid (ACSF, in mM: KCl, 2.49; NaH2PO4, 1.43; NaHCO3,26; glucose, 10; sucrose, 252; CaCl2, 1.0; MgCl2. 4.0). Horizontal sliceswere prepared containing the brainstem and cerebellum and kept inice-cold hypertonic sucrose solution. The cerebellum ⁄ brainstemblock was fixed to the stage of a vibrating tissue slicer to obtain200 μm slices containing the anteroventral cochlear nucleus (AVCN)and the medial nucleus of the trapezoid body (MNTB). Slices werekept in a submerged holding chamber containing ACSF (in mM: NaCl,124; KCl, 3.5; NaH2PO4, 0.25; MgCl2, 1.5; CaCl2, 1.5; NaHCO3, 30;glucose, 10), bubbled with 95% O2/5% CO2, and incubated at 35 °C for1 h. For recordings, slices were transferred to a submerged recordingchamber under an upright microscope equipped with infrared anddifferential interference optics. Whole-cell recordings were per-formed using either an Axopatch 200B or an Axoclamp 2A amplifier(Molecular Devices). Glass pipettes were pulled from borosilicateglass and filled with an internal solution containing (in mM): KCl,17.5; K-gluconate, 122.5; NaCl, 9; MgCl2, 1; Mg-ATP, 3; GTP-Tris, 0.3;HEPES, 1; EGTA, 0.2 to obtain a final tip resistance between 3 and 7MΩ pH was adjusted to 7.2 using KOH. Recordings with seriesresistance N15 MΩ were discarded. Data was digitized usingWinWCP (Dr. John Dempster, http://spider.science.strath.ac.uk/sipbs/software_ses.htm) and analyzed in Matlab (Mathworks).Firing properties and membrane resting potential were obtained(in fast current clamp mode in the Axopatch 200B or bridge mode inthe Axoclamp 2A) by applying 200 ms-current steps ranging from−50to 350 pA. After recordings, slices were rescued and fixed in 4% PFA inPBS overnight for immunohistochemistry assessment. After fixation,slices were washed in 0.1 M PBS, and blocked using PBS containing 1%normal donkey serum and 0.1% Triton for 30 min. Sections wereincubated with antibodies against Cre (see above) and/or GFP (chickenpolyclonal, 1:2000, Abcam #13970). Species-specific secondary anti-bodies from Jackson Immuno Research were used as recommended bythe manufacturers. Slices were then washed in PBS and mounted forconfocal microscopy.

Acknowledgments

This work was supported by NIH R01NS040795-09 (OK), EU-grant(SPINAL CORD REPAIR, OK), Swedish Medical Research Council (OK),Söderbergs Foundation (OK), Friends of Karolinska Institutet (OK) andHjärnfonden (OK). RNL was supported by a Long Term Fellowship fromthe International Human Frontier Science Program. The Vglut2flox/flox

mice were kindly provided by Drs. T. Hnasko and R. Palmiter. TheGAD67::eGFP mice were a kind gift from Dr. N. Tamamaki. We thankDrs. M. Watanabe and T. Hökfelt for the Vglut2 antibody and Dr. D.Engblom for material and technical support regarding the BAC cloningand Dr. G. Schütz for the Cre antibody.

References

Abbott, S.B., Stornetta, R.L., Socolovsky, C.S., West, G.H., Guyenet, P.G., 2009.Photostimulation of channelrhodopsin-2 expressing ventrolateral medullaryneurons increases sympathetic nerve activity and blood pressure in rats. J. Physiol.587, 5613–5631.

Adamantidis, A.R., Zhang, F., Aravanis, A.M., Deisseroth, K., de Lecea, L., 2007. Neuralsubstrates of awakening probed with optogenetic control of hypocretin neurons.Nature 450, 420–424.

Aihara, Y., Mashima, H., Onda, H., Hisano, S., Kasuya, H., Hori, T., Yamada, S., Tomura, H.,Yamada, Y., Inoue, I., Kojima, I., Takeda, J., 2000. Molecular cloning of a novel brain-

256 L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

type Na(+)-dependent inorganic phosphate cotransporter. J. Neurochem. 74,2622–2625.

Al-Mosawie, A., Wilson, J.M., Brownstone, R.M., 2007. Heterogeneity of V2-derivedinterneurons in the adult mouse spinal cord. Eur. J. Neurosci. 26, 3003–3015.

Bai, L.Q., Xu, H., Collins, J.F., Ghishan, F.K., 2001. Molecular and functional analysis of anovel neuronal vesicular glutamate transporter. J. Biol. Chem. 276, 36764–36769.

Bellocchio, E.E., Reimer, R.J., Fremeau, R.T., Edwards, R.H., 2000. Uptake of glutamateinto synaptic vesicles by an inorganic phosphate transporter. Science 289,957–960.

Billups, B., 2005. Colocalization of vesicular glutamate transporters in the rat superiorolivary complex. Neurosci. Lett. 382, 66–70.

Boulland, J.L., Qureshi, T., Seal, R.P., Rafiki, A., Gundersen, V., Bergersen, L.H., Fremeau Jr.,R.T., Edwards, R.H., Storm-Mathisen, J., Chaudhry, F.A., 2004. Expression of thevesicular glutamate transporters during development indicates the widespreadcorelease of multiple neurotransmitters. J. Comp. Neurol. 480, 264–280.

Brumovsky, P., Watanabe, M., Hokfelt, T., 2007. Expression of the vesicular glutamatetransporters-1 and -2 in adult mouse dorsal root ganglia and spinal cord and theirregulation by nerve injury. Neuroscience 147, 469–490.

Callaway, E.M., 2005. A molecular and genetic arsenal for systems neuroscience. TrendsNeurosci. 28, 196–201.

Cardin, J.A., Carlen, M., Meletis, K., Knoblich, U., Zhang, F., Deisseroth, K., Tsai, L.H.,Moore, C.I., 2009. Driving fast-spiking cells induces gamma rhythm and controlssensory responses. Nature 459, 663–667.

Crone, S.A., Quinlan, K.A., Zagoraiou, L., Droho, S., Restrepo, C.E., Lundfald, L., Endo, T.,Setlak, J., Jessell, T.M., Kiehn, O., Sharma, K., 2008. Genetic ablation of V2a ipsilateralinterneurons disrupts left-right locomotor coordination in mammalian spinal cord.Neuron 60, 70–83.

Crone, S.A., Zhong, G., Harris-Warrick, R., Sharma, K., 2009. In mice lacking V2ainterneurons, gait depends on speed of locomotion. J. Neurosci. 29, 7098–7109.

Dal Bo, G., St-Gelais, F., Danik, M., Williams, S., Cotton, M., Trudeau, L.E., 2004. Dopamineneurons in culture express VGLUT2 explaining their capacity to release glutamateat synapses in addition to dopamine. J. Neurochem. 88, 1398–1405.

Dougherty, K., Kiehn, O., 2010. Firing and cellular properties of V2a interneurons in therodent spinal cord. J. Neurosci. 30, 24–37.

Feldman, J.L., Del Negro, C.A., 2006. Looking for inspiration: new perspectives onrespiratory rhythm. Nat. Rev. Neurosci. 7, 232–242.

Fremeau, R.T., Troyer, M.D., Pahner, I., Nygaard, G.O., Tran, C.H., Reimer, R.J., Bellocchio,E.E., Fortin, D., Storm-Mathisen, J., Edwards, R.H., 2001. The expression of vesicularglutamate transporters defines two classes of excitatory synapse. Neuron 31,247–260.

Fremeau Jr., R.T., Burman, J., Qureshi, T., Tran, C.H., Proctor, J., Johnson, J., Zhang, H., Sulzer,D., Copenhagen, D.R., Storm-Mathisen, J., Reimer, R.J., Chaudhry, F.A., Edwards, R.H.,2002. The identification of vesicular glutamate transporter 3 suggests novel modes ofsignaling by glutamate. Proc. Natl. Acad. Sci. U.S.A. 99, 14488–14493.

Fremeau Jr., R.T., Kam, K., Qureshi, T., Johnson, J., Copenhagen, D.R., Storm-Mathisen, J.,Chaudhry, F.A., Nicoll, R.A., Edwards, R.H., 2004a. Vesicular glutamate transporters1 and 2 target to functionally distinct synaptic release sites. Science 304,1815–1819.

Fremeau Jr., R.T., Voglmaier, S., Seal, R.P., Edwards, R.H., 2004b. VGLUTs define subsetsof excitatory neurons and suggest novel roles for glutamate. Trends Neurosci. 27,98–103.

Giraldo, P., Montoliu, L., 2001. Size matters: use of YACs, BACs and PACs in transgenicanimals. Transgenic Res. 10, 83–103.

Gong, S., Zheng, C., Doughty, M.L., Losos, K., Didkovsky, N., Schambra, U.B., Nowak, N.J.,Joyner, A., Leblanc, G., Hatten, M.E., Heintz, N., 2003. A gene expression atlas of thecentral nervous system based on bacterial artificial chromosomes. Nature 425,917–925.

Gong, S., Doughty, M., Harbaugh, C.R., Cummins, A., Hatten, M.E., Heintz, N., Gerfen, C.R.,2007. Targeting Cre recombinase to specific neuron populations with bacterialartificial chromosome constructs. J. Neurosci. 27, 9817–9823.

Gosgnach, S., Lanuza, G.M., Butt, S.J., Saueressig, H., Zhang, Y., Velasquez, T.,Riethmacher, D., Callaway, E.M., Kiehn, O., Goulding, M., 2006. V1 spinal neuronsregulate the speed of vertebrate locomotor outputs. Nature 440, 215–219.

Goulding, M., 2009. Circuits controlling vertebrate locomotion: moving in a newdirection. Nat. Rev. Neurosci. 10, 507–518.

Gradinaru, V., Mogri, M., Thompson, K.R., Henderson, J.M., Deisseroth, K., 2009. Opticaldeconstruction of parkinsonian neural circuitry. Science 324, 354–359.

Gras, C., Herzog, E., Bellenchi, G.C., Bernard, V., Ravassard, P., Pohl, M., Gasnier, B., Giros,B., El Mestikawy, S., 2002. A third vesicular glutamate transporter expressed bycholinergic and serotoninergic neurons. J. Neurosci. 22, 5442–5451.

Grillner, S., 2003. The motor infrastructure: from ion channels to neuronal networks.Nat. Rev. Neurosci. 4, 573–586.

Heintz, N., 2001. BAC to the future: the use of bac transgenic mice for neuroscienceresearch. Nat. Rev. Neurosci. 2, 861–870.

Heisler, L.K., Chu, H.M., Brennan, T.J., Danao, J.A., Bajwa, P., Parsons, L.H., Tecott, L.H.,1998. Elevated anxiety and antidepressant-like responses in serotonin 5-HT1Areceptor mutant mice. Proc. Natl. Acad. Sci. U.S.A. 95, 15049–15054.

Herzog, E., Bellenchi, G.C., Gras, C., Bernard, V., Ravassard, P., Bedet, C., Gasnier, B., Giros,B., El Mestikawy, S., 2001. The existence of a second vesicular glutamate transporterspecifies subpopulations of glutamatergic neurons. J. Neurosci. 21, RC181.

Herzog, E., Gilchrist, J., Gras, C., Muzerelle, A., Ravassard, P., Giros, B., Gaspar, P., ElMestikawy, S., 2004. Localization of VGLUT3, the vesicular glutamate transportertype 3, in the rat brain. Neuroscience 123, 983–1002.

Hinckley, C.A., Hartley, R., Wu, L., Todd, A., Ziskind-Conhaim, L., 2005. Locomotor-likerhythms in a genetically distinct cluster of interneurons in the mammalian spinalcord. J. Neurophysiol. 93, 1439–1449.

Hnasko, T.S., Chuhma, N., Zhang, H., Goh, G.Y., Sulzer, D., Palmiter, R.D., Rayport, S.,Edwards, R.H., 2010. Vesicular glutamate transport promotes dopamine storageand glutamate corelease in vivo. Neuron 65, 643–656.

Kawano, M., Kawasaki, A., Sakata-Haga, H., Fukui, Y., Kawano, H., Nogami, H., Hisano, S.,2006. Particular subpopulations of midbrain and hypothalamic dopamine neuronsexpress vesicular glutamate transporter 2 in the rat brain. J. Comp. Neurol. 498,581–592.

Kellendonk, C., Tronche, F., Casanova, E., Anlag, K., Opherk, C., Schutz, G., 1999. Induciblesite-specific recombination in the brain. J. Mol. Biol. 285, 175–182.

Kiehn, O., 2006. Locomotor circuits in the mammalian spinal cord. Annu. Rev. Neurosci.29, 279–306.

Kiehn, O., Butt, S.J., 2003. Physiological, anatomical and genetic identification of CPGneurons in the developing mammalian spinal cord. Prog. Neurobiol. 70, 347–361.

Kiehn, O., Kjaerulff, O., 1996. Spatiotemporal characteristics of 5-HT and dopamine-induced rhythmic hindlimb activity in the in vitro neonatal rat. J. Neurophysiol. 75,1472–1482.

Kiehn, O., Quinlan, K.A., Restrepo, C.E., Lundfald, L., Borgius, L., Talpalar, A.E., Endo, T.,2008. Excitatory components of the mammalian locomotor CPG. Brain Res. Rev. 57,56–63.

Kjaerulff, O., Kiehn, O., 1996. Distribution of networks generating and coordinatinglocomotor activity in the neonatal rat spinal cord in vitro: a lesion study. J. Neurosci.16, 5777–5794.

Kremer, E., Lev-Tov, A., 1997. Localization of the spinal network associated withgeneration of hindlimb locomotion in the neonatal rat and organization of itstransverse coupling system. J. Neurophysiol. 77, 1155–1170.

Kullander, K., Butt, S.J., Lebret, J.M., Lundfald, L., Restrepo, C.E., Rydstrom, A., Klein, R.,Kiehn, O., 2003. Role of EphA4 and EphrinB3 in local neuronal circuits that controlwalking. Science 299, 1889–1892.

Lanuza, G.M., Gosgnach, S., Pierani, A., Jessell, T.M., Goulding, M., 2004. Geneticidentification of spinal interneurons that coordinate left-right locomotor activitynecessary for walking movements. Neuron 42, 375–386.

Leao, R.M., Kushmerick, C., Pinaud, R., Renden, R., Li, G.L., Taschenberger, H., Spirou, G.,Levinson, S.R., von Gersdorff, H., 2005. Presynaptic Na+channels: locus, develop-ment, and recovery from inactivation at a high-fidelity synapse. J. Neurosci. 25,3724–3738.

Leao, K.E., Leao, R.N., Sun, H., Fyffe, R.E., Walmsley, B., 2006. Hyperpolarization-activated currents are differentially expressed in mice brainstem auditory nuclei.J. Physiol. 576, 849–864.

Lee, E.C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D.A., Court, D.L., Jenkins, N.A.,Copeland, N.G., 2001. A highly efficient Escherichia coli-based chromosomeengineering system adapted for recombinogenic targeting and subcloning of BACDNA. Genomics 73, 56–65.

Lundfald, L., Restrepo, C.E., Butt, S.J., Peng, C.Y., Droho, S., Endo, T., Zeilhofer, H.U.,Sharma, K., Kiehn, O., 2007. Phenotype of V2-derived interneurons and theirrelationship to the axon guidance molecule EphA4 in the developing mouse spinalcord. Eur. J. Neurosci. 26, 2989–3002.

Luo, L., Callaway, E.M., Svoboda, K., 2008. Genetic dissection of neural circuits. Neuron57, 634–660.

Moechars, D., Weston, M.C., Leo, S., Callaerts-Vegh, Z., Goris, I., Daneels, G., Buist, A., Cik,M., van der Spek, P., Kass, S., Meert, T., D'Hooge, R., Rosenmund, C., Hampson, R.M.,2006. Vesicular glutamate transporter VGLUT2 expression levels control quantalsize and neuropathic pain. J. Neurosci. 26, 12055–12066.

Montana, V., Malarkey, E.B., Verderio, C., Matteoli, M., Parpura, V., 2006. Vesiculartransmitter release from astrocytes. Glia 54, 700–715.

Moriyama, Y., Yamamoto, A., 2004. Glutamatergic chemical transmission: look! Here,there, and anywhere. J. Biochem. 135, 155–163.

Novak, A., Guo, C., Yang, W., Nagy, A., Lobe, C.G., 2000. Z/EG, a double reporter mouseline that expresses enhanced green fluorescent protein upon Cre-mediatedexcision. Genesis 28, 147–155.

Peng, C.Y., Yajima, H., Burns, C.E., Zon, L.I., Sisodia, S.S., Pfaff, S.L., Sharma, K., 2007. Notchand MAML signaling drives Scl-dependent interneuron diversity in the spinal cord.Neuron 53, 813–827.

Petreanu, L., Mao, T., Sternson, S.M., Svoboda, K., 2009. The subcellular organization ofneocortical excitatory connections. Nature 457, 1142–1145.

Ponzio, T.A., Ni, Y., Montana, V., Parpura, V., Hatton, G.I., 2006. Vesicular glutamatetransporter expression in supraoptic neurones suggests a glutamatergic pheno-type. J. Neuroendocrinol. 18, 253–265.

Restrepo, C.E., Lundfald, L., Szabo, G., Erdelyi, F., Zeilhofer, H.U., Glover, J.C., Kiehn, O.,2009. Transmitter-phenotypes of commissural interneurons in the lumbar spinalcord of newborn mice. J. Comp. Neurol. 517, 177–192.

Sahly, I., Fabre, V., Vyas, S., Milet, A., Rouzeau, J.D., Hamon, M., Lazar, M., Tronche, F.,2007. 5-HT1A-iCre, a new transgenic mouse line for genetic analyses of theserotonergic pathway. Mol. Cell. Neurosci. 36, 27–35.

Sato, S., Inoue, T., Terada, K., Matsuo, I., Aizawa, S., Tano, Y., Fujikado, T., Furukawa, T.,2007. Dkk3-Cre BAC transgenic mouse line: a tool for highly efficient gene deletionin retinal progenitor cells. Genesis 45, 502–507.

Schafer, M.K., Varoqui, H., Defamie, N., Weihe, E., Erickson, J.D., 2002. Molecular cloningand functional identification of mouse vesicular glutamate transporter 3 and itsexpression in subsets of novel excitatory neurons. J. Biol. Chem. 277, 50734–50748.

Shimshek, D.R., Kim, J., Hubner, M.R., Spergel, D.J., Buchholz, F., Casanova, E., Stewart, A.F.,Seeburg, P.H., Sprengel, R., 2002. Codon-improved Cre recombinase (iCre) expressionin the mouse. Genesis 32, 19–26.

Sohal, V.S., Zhang, F., Yizhar, O., Deisseroth, K., 2009. Parvalbumin neurons and gammarhythms enhance cortical circuit performance. Nature 459, 698–702.

Soriano, P., 1999. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat.Genet. 21, 70–71.

257L. Borgius et al. / Molecular and Cellular Neuroscience 45 (2010) 245–257

Srinivas, S., Watanabe, T., Lin, C.S., William, C.M., Tanabe, Y., Jessell, T.M., Costantini, F.,2001. Cre reporter strains produced by targeted insertion of EYFP and ECFP into theROSA26 locus. BMC Dev. Biol. 1, 4.

Sun, H., Crossland,W.J., 2000. Quantitative assessment of localization and colocalizationof glutamate, aspartate, glycine, and GABA immunoreactivity in the chick retina.Anat. Rec. 260, 158–179.

Takamori, S., 2006. VGLUTs: ‘exciting’ times for glutamatergic research? Neurosci. Res.55, 343–351.

Takamori, S., Rhee, J.S., Rosenmund,C., Jahn, R., 2000. Identificationof a vesicular glutamatetransporter that defines a glutamatergic phenotype in neurons. Nature 407, 189–194.

Takamori, S., Rhee, J.S., Rosenmund, C., Jahn, R., 2001. Identification of differentiation-associated brain-specific phosphate transporter as a second vesicular glutamatetransporter (VGLUT2). J. Neurosci. 21, RC182.

Takamori, S., Malherbe, P., Broger, C., Jahn, R., 2002. Molecular cloning and functionalcharacterization of human vesicular glutamate transporter 3. EMBO Rep. 3, 798–803.

Tamamaki, N., Yanagawa, Y., Tomioka, R., Miyazaki, J., Obata, K., Kaneko, T., 2003. Greenfluorescent protein expression and colocalization with calretinin, parvalbumin, andsomatostatin in the GAD67-GFP knock-in mouse. J. Comp. Neurol. 467, 60–79.

Tan, E.M., Yamaguchi, Y., Horwitz, G.D., Gosgnach, S., Lein, E.S., Goulding, M., Albright, T.D.,Callaway, E.M., 2006. Selective and quickly reversible inactivation of mammalianneurons in vivo using the Drosophila allatostatin receptor. Neuron 51, 157–170.

Tan, W., Janczewski, W.A., Yang, P., Shao, X.M., Callaway, E.M., Feldman, J.L., 2008.Silencing preBotzinger complex somatostatin-expressing neurons induces persis-tent apnea in awake rat. Nat. Neurosci. 11, 538–540.

Thoby-Brisson, M., Karlen, M., Wu, N., Charnay, P., Champagnat, J., Fortin, G., 2009.Genetic identification of an embryonic parafacial oscillator coupling to thepreBotzinger complex. Nat. Neurosci. 12, 1028–1035.

Todd, A.J., Sullivan, A.C., 1990. Light microscope study of the coexistence of GABA-likeand glycine-like immunoreactivities in the spinal cord of the rat. J. Comp. Neurol.296, 496–505.

Todd, A.J., Hughes, D.I., Polgar, E., Nagy, G.G., Mackie, M., Ottersen, O.P., Maxwell, D.J.,2003. The expression of vesicular glutamate transporters VGLUT1 and VGLUT2 in

neurochemically defined axonal populations in the rat spinal cord with emphasison the dorsal horn. Eur. J. Neurosci. 17, 13–27.

Tong, Q., Ma, J., Kirchgessner, A.L., 2001. Vesicular glutamate transporter 2 in the brain-gut axis. NeuroReport 12, 3929–3934.

Tsai, H.C., Zhang, F., Adamantidis, A., Stuber, G.D., Bonci, A., de Lecea, L., Deisseroth, K.,2009. Phasic firing in dopaminergic neurons is sufficient for behavioral condition-ing. Science 324, 1080–1084.

Varoqui, H., Schafer, M.K., Zhu, H., Weihe, E., Erickson, J.D., 2002. Identification of thedifferentiation-associated Na+/PI transporter as a novel vesicular glutamatetransporter expressed in a distinct set of glutamatergic synapses. J. Neurosci. 22,142–155.

Wallen-Mackenzie, A., Gezelius, H., Thoby-Brisson, M., Nygard, A., Enjin, A., Fujiyama, F.,Fortin, G., Kullander, K., 2006. Vesicular glutamate transporter 2 is required forcentral respiratory rhythm generation but not for locomotor central patterngeneration. J. Neurosci. 26, 12294–12307.

Wilson, J.M., Hartley, R., Maxwell, D.J., Todd, A.J., Lieberam, I., Kaltschmidt, J.A., Yoshida,Y., Jessell, T.M., Brownstone, R.M., 2005. Conditional rhythmicity of ventral spinalinterneurons defined by expression of the Hb9 homeodomain protein. J. Neurosci.25, 5710–5719.

Wojcik, S.M., Rhee, J.S., Herzog, E., Sigler, A., Jahn, R., Takamori, S., Brose, N., Rosenmund,C., 2004. An essential role for vesicular glutamate transporter 1 (VGLUT1) inpostnatal development and control of quantal size. Proc. Natl Acad. Sci. USA 101,7158–7163.

Xu, Q., Tam, M., Anderson, S.A., 2008. Fate mapping Nkx2.1-lineage cells in the mousetelencephalon. J. Comp. Neurol. 506, 16–29.

Zhang, F., Aravanis, A.M., Adamantidis, A., de Lecea, L., Deisseroth, K., 2007. Circuit-breakers: optical technologies for probing neural signals and systems. Nat. Rev.Neurosci. 8, 577–581.

Zhang, Y., Narayan, S., Geiman, E., Lanuza, G.M., Velasquez, T., Shanks, B., Akay, T., Dyck,J., Pearson, K., Gosgnach, S., Fan, C.M., Goulding, M., 2008. V3 spinal neuronsestablish a robust and balanced locomotor rhythm during walking. Neuron 60,84–96.