Expression patterns of promoters for NPY Y₁ and Y₅ receptors in Y₅RitTA and Y₁RVenus BAC-transgenic mice

Animal experiments

Animal care was in compliance with the European Community Council Directive of November 24, 1986 (86/609/EEC) and in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publication no. 85–23, revised 1996) and the European Community guidelines for the use of experimental animals. The pronucleus injections were performed by Frank Zimmermann at the IBF, INF 347, D-69120 Heidelberg, Germany, according to a license (35–9185.81/G-4/02) of the Regierungspräsidium Karlsruhe, Germany. The experimental protocol was approved by the animal investigation committee of the Ministero dell'Istruzione dell'Università e della Ricerca, Italy.

Y₁R gene BAC screening

A mouse 129 BAC library (BAC Mouse ES release I, BAC4921, Genome Systems, St Louis, MO, USA) was screened at Genome Systems with a 2-kb DNA probe obtained from XbaI digestion of a pBSY6-7 plasmid containing the Y₁R gene (Eva et al., 1992). Two clones were obtained: 22673 and 22674, both of which were fingerprinted with several restriction enzymes and hybridised with the probe mentioned above. All the work was carried out on clone number 22674.

Plasmid construction of pBS.htTApA.2frtneo

The neomycin resistance of the pCR4-TOPO (Invitrogen, Karlsruhe, Germany) was amplified with overhang primer which flanked the resulting PCR product with the 34-bp minimal flippase recombinase (Flp) target (Frt) sequences in the same orientation and direction. This fragment was digested with BglII and EcoRI and inserted downstream of the human growth hormone polyadenylation (pA) signal of the plasmid pBS.itTApA (Krestel et al., 2004) to create the plasmid pBS.itTApA.2frtneo.

Plasmid construction of pBS-Venus 2frtneo

The Venus gene was excised with HindIII-EcoRV digestion from plasmid pCS2-Venus (Nagai et al., 2002) and subcloned in HindIII-EcoRV, in pBS.iCrepA.2frtneo plasmid in place of the iCre gene. Correct insertion of the Venus gene was confirmed by sequence analysis using oligonucleotides sitting outside of the insertion site.

Recombinogenic targeting of BAC

The Y₁R BAC clone containing the Y₁R and the Y₅R genes in opposite orientation and, in between, their potential 25-kb regulatory region, was modified using the E.coli-based chromosome system adapted for recombinogenic targeting and subcloning of BAC DNA (Lee et al., 2001). The EL250 strain (kindly provided by Dr Neal Copeland) used in this study, originating from DH10B E.coli cells, has homologous recombination (λ red genes) and site-specific recombination (Flp) functions with the former controlled by temperature and the latter by arabinose.

Preparation of electrocompetent cells and generation of recombinants

For BAC modification, 100 mL LB broth recipe medium was inoculated with 2 mL of overnight culture from single colonies of bacteria containing the BAC and grown to 0.5–0.7 optical density 600. Cultures were then induced for Beta, Exo and Gam expression by shifting the cells to 42 °C for 15 min and immediately chilling for 20 min on ice. Cells were then centrifuged for 5 min at 5500 g at 4 °C and washed three times with 100 mL of ice-cold sterile 10% glycerol. Electrocompetent cells were then resuspended in 1 mL of ice-cold sterile water and electroporated.

Cell transformation was performed by electroporation of 100–300 ng linear DNA into 50 µL of ice-cold competent cells in cuvettes (0.1 cm) using a Bio-Rad gene pulser set at 1.75 kV, 25 mF, with a pulse controller set at 200 Ω. One millilitre of LB broth recipe medium was added after electroporation. Cells were incubated at 32 °C for 1.5 h with shaking, and then spread on appropriate selective agar media.

Production of BAC-transgenic mice

Modified BAC was digested by NotI to free the 150-kb insert, containing the Y₅RitTA/Y₁RVenus sequence, from the BAC backbone. The insert was purified on a Sepharose CL-4B gradient (Amersham, Milano, Italy), using a microinjection buffer containing (in mm) Tris-HCl, pH 7.6, 10; EDTA, 0.1; and NaCl, 100. BAC insert (∼ 0.5–1 ng) was microinjected into the pronucleus of (C57BL/6 × DBA2)F1 zygotes. Founder mice were analysed by PCR using primers Y₁51 (5′-CTGCTGTTGACACCATTTGTCTAACACTTGG-3′) and Venus2 (5′-CTTCAGCTCGATGCGGTTCAC-3′), and by Southern blot analysis of phenol–chloroform-extracted tail or liver DNA, digested with MscI or with BglII and hybridised with a DNA probe obtained by PCR on BAC Y₅RitTA/Y₁RVenus using Y₁51 and Venus2 primers.

Generation of double-transgenic mice and genotyping

Transgenic mice Tg^{Y5RitTA/Y1RVenus} were crossed with Tg^LC1 (Schonig et al., 2002). Pups were genotyped by PCR of tail DNA with specific primers. Tg^LC1 rspCre1 (5′-ACCAGG TTC GTT CAC TCA TGG-3′) and rspCre2 (5′-AGG CTA AGT GCC TTC TCT ACAC-3′), 200 bp; Tg^{Y5RitTA/Y1RVenus}, itTA2RT (5′-CAACTCAATAGCTTGCCGCAG-3′) and itTA (5′-TGCCTTGGAGCTCCTGAATGAAGTTG-3′), 500 bp.

Regulation with Dox treatment

In double-transgenic mice, Tg^{Y5RitTA/Y1Rvenus}- and Tg^LC1-positive Cre recombinase expression was suppressed from conception until postnatal day (P)0 with Dox (Sigma-Aldrich, Milano, Italy) in the drinking water (which contained 1% sucrose and 20 (line 4 and 36) and 50 mg (line 25) Dox/L) of pregnant females; it was subsequently induced by transferring pups to Dox-naïve foster mothers (Krestel et al., 2001).

After PCR tail genotyping, positive pups were analysed for Cre immunoreactivity at P20, P30, P40 and P52.

Immunohistochemistry

Mice at 4–5 weeks of age were given an anaesthetic overdose (Avertin; 0.5 mg per gram of body weight, i.p.) and transcardially perfused with 4% paraformaldehyde in PBS. Brains were removed, postfixed for 2 h and placed in 2% agarose in 1 × PBS. Cre immunostaining was carried out using the ABC Vectastain kit (Vector Laboratories, Burlingame, CA, USA) on 50-µm vibratome coronal sections. After endogenous peroxidase blocking in 0.5% H₂O₂ in 0.1 m PBS for 15 min, sections were blocked with 10% normal goat serum in DayI buffer (PBS containing 0.3% Triton X-100 and 1% bovine serum albumin) for 1 h and incubated overnight at room temperature with primary Cre antibody, a polyclonal rabbit anti-Cre recombinase antiserum (Novagene, Madison, WI, USA) at 1 : 3000 dilution in DayI solution. After washing in DayII buffer (1 : 3 dilution of DayI in 0.1 m PBS), sections were incubated for 1 h with a secondary biotinylated goat antirabbit antibody (1 : 500; Vector) followed by the ABC reagent (Vector). Peroxidase was reacted with 0.05% diaminobenzidine tetrahydrochloride (DAB; Sigma-Aldrich, Milano, Italy), and 0.003% hydrogen peroxide, mounted on slides and air-dried overnight. DAB-developed slices were coverslipped with DPX mounting medium and analysed with a Leica microscope. For colocalization studies, Cre immunofluorescence detection was performed using the above protocol with some modifications: the primary Cre antibody dilution was lowered to 1 : 1500, the secondary antibody incubation was performed in a modified DayI solution containing 0.15% Triton X-100 in PBS, and was followed by 1 h incubation with Avidin Texas Red (1 : 600; Vector). The extension of the immunostaining was independently evaluated by two observers on an ordinal scale, using four different levels: high (+ + +) when > 50% of cellular elements of a brain region were positive to the histochemical reaction; intermediate (+ +) when between 20 and 50% of the elements were stained; low (+) when < 20% of the elements were stained; and very low (+/–) when only sparse positive elements were detected (Table 1).

Analysis of Venus expression

Vibratome coronal sections (50 µm) of brains from perfused mice were mounted on slides and used directly for fluorescence distribution analysis using a Leica DM LB2 fluorescence microscope equipped with a DFC 320 Leica camera and an Olympus Bx51WI or a Zeiss LSM5 Pascal confocal laser-scanning microscope. The intensity of fluorescence was independently evaluated by two observers on an ordinal scale as described for immunostaining of Cre.

Y₁R and Y₅R promoter-controlled reporter gene expression in four transgenic lines

The Y₅RitTA/Y₁RVenus BAC, containing transcription units for Venus and itTA driven by the Y₁R and the Y₅R regulatory regions, respectively, was used to generate Tg^{Y5RitTA/Y1RVenus}-transgenic mice (Fig. 1B). Six founder lines were identified by Southern blot analysis of genomic DNA, and five founders showed the expected hybridization pattern (data not shown). To monitor Y₅R promoter activity using itTA, the founder mice were crossed with Tg^LC1-transgenic mice that carry P_tet-bi controlled Cre recombinase and Luciferase as tTA reporter genes (Schonig et al., 2002). P_tet-bi is a bidirectional tTA-responsive promoter composed of an array of seven Tet operator sequences flanked by two human cytomegalovirus promoter IE minimal promoters. P_tet-bi is activated by tTA and initiates transcription of the Cre recombinase and the Luciferase transgenes in opposite directions (Fig. 1B). Analysis of Cre expression was performed by immunohistochemical staining of coronal brain sections from 4-week- to 2-month-old transgenic mice. Specific Cre expression was detected in offspring from founders 4, 25, 36 and 39 but with different reporter levels. Cre and Venus expression levels were correlated with the relative copy number of the transgene, which was found in Southern blots to be 1 : 2 : 2 : 2 : 6 between the lines 27, 39, 4, 36 and 25. In low-expressing mice of founder 39, Cre immunoreactivity was observed in some dispersed cells of the cerebral cortex and hippocampus (data not shown). Lines 4 and 36 showed similar Cre-expression patterns. In both lines, moderate to high Cre expression was present in the olfactory system, cerebral cortex, hippocampus, basal ganglia and suprachiasmatic nuclei (SCh; the data are summarised in Table 1 and in Fig. 2). Lower Cre expression was found in the medial amygdala (Me) and central amygdala, and weak Cre immunostaining was observed in a few thalamic and hypothalamic nuclei.

In mice of the highest-expressing line (line 25), the Cre expression pattern overlapped the one reported in the literature for the rat and mouse Y₅R in most of the brain regions (Larsen & Kristensen, 1998; Naveilhan et al., 1998; Parker & Herzog, 1998, 1999; Xin & Huang, 1998; Broberger et al., 1999; Wolak et al., 2003). The highest Cre expression was found in the olfactory system, layers II/III, IV and VI of the cerebral cortex, stratum pyramidale of the CA1-CA3 fields, granule cell layer of the dentate gyrus (GrDG), basolateral amygdala (BLA), caudate–putamen (CPu), nucleus accumbens, ventral pallidum, tenia tecta, lateral habenula and the ependymal–subependymal cell layer of the third ventricle (3V; Table 1, 1, 2). Moderate to low Cre expression was also found in the layer V of the cerebral cortex, Me, central amygdala and anterior cortical amygdaloid nucleus, claustrum, bed nucleus of stria terminalis, septum, islands of Calleja, medial preoptic area and lateral habenula. In the hypothalamus, Cre-immunoreactive staining is also similar to endogenous rat Y₅R gene expression. High levels of Cre expression were found in SCh, paraventricular (PVN) and dorsomedial (DM) hypothalamic nuclei and mammillary nuclei (MM), whereas low levels of Cre immunostaining were observed in the lateral hypothalamic area (LH), anterior hypothalamus (AH), ventromedial (VMH), arcuate (Arc) and periventricular (Pe) hypothalamic nuclei (Fig. 2 and Table 1). Conversely, in the thalamus, moderate to low Cre expression was detected in the anterodorsal, paraventricular, reuniens and lateral geniculate nuclei, while there was no Cre expression in the central, anteromedial, reticular or mediodorsal thalamic nuclei, in the ventral group or in the zona incerta, all thalamic regions in which the Y₅R is expressed in the rat (Wolak et al., 2003).

Although we did not map Cre expression in the brainstem as rigorously as we did in the higher structures, we noted a low to moderate expression of Cre immunoreactivity confined to the substantia nigra, colliculi and a few metencephalic nuclei.

Conditional Cre expression in forebrain

P_tet-bi controlled Cre and Luciferase expression can be prevented by Dox treatment which inhibits the binding of itTA to P_tet-bi, permitting a temporal control of Cre-mediated recombination (for review see Sprengel & Hasan, 2007). Therefore, we investigated the Tet sensitivity of Y₅R-itTA-activated Tg^LC1 mice by Dox treatment of lines 25, 4 and 36. Our results demonstrate that the extent of Cre expression could be restricted by Dox. When double-transgenic mice were raised with Dox from conception, itTA induction was prevented, keeping Cre at undetectable levels (Fig. 3). If Dox was applied from conception until birth (P0), the Cre-specific immunosignal slowly increased over time, reflecting a slow unsilencing of the P_tet-bi of the Tg^LC1 transgene (Mack et al., 2001; Krestel et al., 2004; Zhu et al., 2007). In mice with Dox suppression during intrauterine development, a weak expression of Cre was observed at P20 when compared with the untreated lines 25 and 36 at P40–60 in histochemical stains (Fig. 3) indicating that at P20 the Tg^LC1 responder genes were not yet fully activated. In all lines, Cre expression levels similar to those observed in untreated mice were reached by P40 in the olfactory system, cerebral cortex, GrDG and CA1-CA3, basal ganglia, septal area, medial preoptic area, SCh, DM and the ependymal–subependymal cells of the 3V (Table 1 and Fig. 3). Further analysis of mice from line 25 at longer time windows for Cre reactivation showed that, in the BLA, in the reuniens and paraventricular thalamic nuclei, and in most of the hypothalamic nuclei (AH, LH, Pe, PVN, Arc and MM), the reactivation of Cre was reached by P52 but Cre expression was still limited to a minor subset of neurons as compared to Dox-naïve mice (Table 1 and Fig. 3). Furthermore, no Cre induction was evident in the VMH, lateral habenula and geniculate thalamic nucleus at this age of mice.

Expression of the Y₁R promoter as monitored by Venus in five transgenic lines

As mentioned above, BAC copy number-dependent Venus expression could be detected in all five transgenic lines. According to their location and aspect and to the lack of glial fibrillary acidic protein immunoreactivity (data not shown), the majority of the examined cells were neurons and Venus fluorescence was present in both cell bodies and cell processes. In offspring from founders 27 and 39, Venus-fluorescent cell bodies and fibres were observed in the cerebral cortex, central amygdala and hypothalamus (data not shown). In mice of lines 4, 25 and 36, Venus fluorescence was intense and widely expressed through the brain, with an overlapping pattern of distribution among the three transgenic lines (Table 2, 1, 4). We found that, in general, the Venus fluorescence recapitulated the endogenous Y₁R expression in mouse, although some discrepancies were observed in the cerebral cortex and in the hippocampal formation (Naveilhan et al., 1998; Kopp et al., 2002; Fetissov et al., 2004; Kishi et al., 2005; Fig. 4 and Table 2).

In the olfactory region, strong Venus fluorescence was monitored in neurons of the anterior olfactory nucleus, medial part (AOM) and in fibres of the AOM and olfactory tubercle. Prominent fluorescence was present in neocortical areas, exhibiting a clear laminated distribution of individually labelled cells. High levels of Venus expression were seen in neurons and fibres of cingulate cortex, piriform cortex and layers II/III and VI of the cerebral cortex. However, in layer IV of the cerebral cortex, a cortical region which contains high levels of Y₁R immunoreactivity and mRNA, Venus expression was limited to a small number of fibres.

In the hippocampus, the highest number of Venus-fluorescent cell bodies was found in the GrDG with densely stained axon fibres projecting to CA3. Conversely, we did not note the presence of Venus-fluorescent cell bodies in the CA2 and CA3 fields, although these areas contain a low to moderate number of Y₁R-immunoreactive neurons. In the amygdaloid area, moderate to high expression of Venus was seen in cell bodies and fibres of BLA, central amygdala, Me and anterior cortical amygdaloid nucleus. Densely stained fibres were seen in the basal ganglia and fluorescent cell bodies were seen in claustrum, endopyriform nucleus and the diagonal band of Broca, but not in the bed nucleus of stria terminalis, a region where Y₁R-immunoreactive neurons have been identified (Kopp et al., 2002; Kishi et al., 2005). In the septal area and in the thalamus we found the presence of fluorescent fibres, whereas cell bodies were devoid of or exhibited only weak fluorescent signal. In the hypothalamus, moderate to high Venus fluorescence was observed in cell bodies and fibres of several hypothalamic nuclei (SCh, AH, LH, Pe, PVN, DM, Arc and MM), with the exclusion of VMH, where fibres but not cell bodies were observed (Fig. 4 and Table 2).

Coexpression of Cre and Venus in the mouse brain

As described above, most brain regions positive for Venus also expressed Cre (Table 3). To take the colocalization analysis of Venus and Cre to a cellular level, we analysed brains derived from line 25 using confocal microscopy. The highest degree of colocalization was found in the GrDG, PVN (both in the parvocellular and magnocellular division) and subependymal layer of the 3V, where ∼ 80–90% of cells were positive for Venus and Cre (5, 6) whereas in the AOM, BLA and the reuniens thalamic nucleus a smaller cell population (∼ 60%) displayed colocalization of Venus and Cre (Fig. 7). In other brain regions there were regional differences in the relative number of cells that coexpressed Cre and Venus. For instance, analysis of sections from different animals suggests that, in layers II/III and VI of the cerebral cortex and in the CA1 region of the hippocampus, most of the Venus-fluorescent cells coexpressed Cre whereas only 40–50% of Cre-immunofluorescent cells in these locations were Venus-positive (Fig. 5). Conversely, in the AH, DM and ventrolateral part of SCh, 50–70% of Cre-immunoreactive cells were also positive for Venus while only a subpopulation of Cre-immunoreactive cells coexpressed Venus (6, 7).

No colocalization of Venus and Cre was found in the olfactory tubercle, layers IV and V of the cerebral cortex, CA2 and CA3 regions, CPu, nucleus accumbens, ventral pallidum, bed nucleus of stria terminalis, septum, islands of Calleja, VMH or anterodorsal, paraventricular or lateral geniculate thalamic nuclei, which displayed immunoreactive staining for Cre but were devoid of Venus-fluorescent cell bodies. In contrast, in the amygdaloid hippocampal area, the diagonal band of Broca and the vascular organ of the lamina terminalis only single-labelled Venus-fluorescent cells were observed.