Colocalization of Six3-Cre and Tyrosine Hydroxylase

Figure 1.

The following are images from a manuscript in preparation. The two students who undertook this project under my direction, Jenny C. Kam and Matthew T. Hoertkorn, share first author credit. We are grateful to Richard Palmiter for the use of his laboratory space and materials, Glenda Froelick for advice on and assistance with our histology, Y. Furuta for the Six3-Cre mice and J. Stadler and G. Stanley McKnight for the Rosa26R mice. Kristen Nagata assisted with the maintenance of the Six3-Cre, Rosa26R mouse colony. This work was supported by training grant NIH 5 T32 NS 07332 (C.M.C.). Detailed methods are posted below the figures.

Abstract
Expression of the homeobox gene Six3 is necessary for the formation of the eyes and ventral forebrain. We crossed genetically modified Six3-Cre mice, which expressed the bacterial enzyme Cre-recombinase under the direction of the Six3 promoter, with mice that carry the Rosa26 reporter. In those progeny with both modifications (Six3-Cre,Rosa26R), the lacZ gene was transcribed to produce the enzyme β-galactosidase (β-gal) only in cells in which Cre-mediated recombination had occurred. In other words, β-gal was a marker for Six3 gene expression that may have occurred at any time during the life of the mouse. Using double immunohistofluorescence, we colocalized β-gal with the enzyme tyrosine hydroxylase (TH), a marker of catecholaminergic neurons. Specific staining for β-gal was observed in predicted areas such as in the hypothalamus, cortical layers II/III and V and hippocampus. However, the consistent, regionally specific expression pattern of Six3, demonstrated here for the first time in adult mice, included novel areas of Six3 gene expression. For example, we observed specific β-gal staining in midbrain and brainstem areas, particularly the locus coeruleus (A6), where TH and β-gal were highly colocalized. The Six3-Cre mouse may prove to be a useful tool for site-specific genetic manipulation of certain catecholaminergic and non-catecholaminergic cell groups in the mouse brain.

Figure 1.

Figure 1. The expression of β-gal-lir in specific forebrain areas. A composite of 4X images from a single coronal section is shown in the background. A-D. A coronal section through the hippocampus (4X), showing β-gal-lir in CA1-CA3 and dentate gyrus (A), TH-lir terminal distribution (B), DAPI stained cell nuclei (C), and a composite of the three (D). As-Ds. Sagittal section through the hippocampus (10X), with individual (As-Cs) and composite (Ds) images as above. E-H. A sagittal section through the cortex (10X) showing β-gal-lir cells in layers II/III and V (E), TH-lir (F), DAPI (G), and the composite image (H). I-J. Sagittal section through the lateral septal nucleus (10X) showing β-gal-lir (I), and a composite of the same image for β-gal-lir, TH-lir and DAPI (J). K-N. A sagittal section through the caudate putamen (4X), demonstrates that scattered β-gal-lir in this structure is limited at the rostral (left) border with the cortex, and the caudal boundaries with thalamus (dorsal) and pallidum (ventral). This pattern closely approximates the distribution of TH-lir terminal staining (L and N). O-R. Sagittal section through the accumbens nucleus (10X). demonstrates the diffuse pattern of brightly-stained β-gal-lir cells in this structure (O), as well as TH-lir terminals (P), DAPI (Q) and the composite (R). The anterior commissure (ac), is on the caudal boundary (right). S-V. Coronal section through the amygdala (10X) demonstrates individual and composite staining as above. W. A composite image of a sagittal section through the ventral pallidal area or lateral globus pallidus (10X).

Figure 2.

Figure 2. . The expression of β-gal-lir in selected hindbrain areas. A. A composite image from a sagittal section through the mammillary nucleus (10X). B. A composite image (10X) from a coronal section showing the ventral margin of the periaqueductal gray (top), including TH-lir in A10dc, β-gal-lir in the region of the oculomotor nucleus, and TH-lir in A10c (bottom). C. A sagittal section through the motor nucleus of the trigeminal nerve (10X), with β-gal-lir cells 30-50 μm in diameter. D. A composite image of β-gal-lir and DAPI nuclear cell bodies from a sagittal section through the cerebellum (10X). An inset shows large β-gal-lir cells and processes.

Figure 3.

Figure 3. Forebrain co-localization of TH-lir and β-gal-lir. Dopaminergic cells groups of the hypothalamus including the A14, pariventricular (A), A13, zona incerta (B), and A12, or arcuate (C) regions. A composite image of the area (4X) is shown in the left-most panel, followed by 20X (A and B) or 10X (C) individual and composite images of the detail indicated by the dashed frame. A representative colocalized cell in each area is marked by the arrow, while a cell producing only TH is marked by the solid triangle and a cell producing only β-gal is marked by the empty triangle.

Figure 4.

Figure 4. Colocalization of TH-lir and β -gal-lir in the locus coeruleus (A6) and A4. A. A composite image (4X) of the area of the locus coeruleus, with the area of the detail for subsequent TH-lir, β-gal-lir and composite images (20X), respectively, indicated by the dashed frame. B. A more rostral composite of the A6 region. C. A composite of the A4 cell group demonstrating co-localization of TH- and β-gal-lir.

A primary purpose of posting “preliminary results” here is to provide addition information as a supplement to our published work. As such, the following images, from a powerpoint presentation of this project fro Lab Meeting, are presented below.

TH and Beta-gal Colocolization

Main Points

Key

Staining Guide

A12 (Arcuate)

A12 (Arcuate)

Beta-gal

Forebrain

A16 (Olfactory Lobe)

A16 (Olfactory Lobe)

Forebrain

A14

A14 (Periventricular)

A14 (Periventricular)

A14d (dorsal)

A14d (dorsal)

A14l (lateral)

A13 (Zona Incerta)

A13 (Zona Incerta)

A13 (Zona Incerta)

A13c (caudal)

Hypothalamus Cross Section

A13c (caudal)

A13c (caudal)

A12 (arcuate)

A12 (arcuate)

Hypothalamus Cross Sections

A11

A11

Midbrain

Subdivisions of A10

A10dr (dorsal rostral)

A10dr (dorsal rostral)

A10 (VTA)

A10 (VTA)

Beta-gal

Subdivisions A10

A10c (caudal)

A10c (caudal)

A10dc (dorsal caudal)

A10dc (dorsal caudal)

A9 (SNc)

A9 (SNc)

A9 (SNc)

Beta-gal

A8

A8 (retrorubral field RRF)

A8 (retrorubral field RRF)

Noradrenergic Cell Groups

A6 Locus Coeruleus

A6 Locus Coeruleus

A6 Locus Coeruleus

A6 Locus Coeruleus

A6 Locus Coeruleus

A5 (Noradrenergic Cell Bodies)

TH-ir

Unknown TH cells

Noradrenergic Cell Groups

The End

Cranial Nerve I Olfactory

Anterior Olfactory Nucleus

Anterior Commisure

Caudate Putamen

Accumbens Nucelus

Lateral Septal Nucelus

Lateral Septal Nucelus (dorsal part)

Medial Preoptic Nucleus

Medial Preoptic Nucleus (medial part)

Lateral Globus Pallidus

Lateral Globus Pallidus

Suprachiasmatic Nucleus

Suprachiasmatic Nucleus

Hippocampus

Habenular Nucleus (medial part)

Thalamus

Hypothalamus

Hypothalamus

Hypothalamus

Hypothalamus

Supramammillary Nuclei

Periaqueductal Gray

Midbrain

Superior Colliculi

Optic Nerve

Inferior Colliculi

Inferior Colliculi

Raphe Cap

Raphe Cap

Pons

Pons

Cranial Nerve VI abducens nucleus

Lateral Parabranchial nucleus

Raphe

Raphe

Cranial Nerve VII – facial nucleus

Cranial Nerve VII – facial nucleus

Cranial Nerve VIII – Vestibulocochlear nucleus

Experimental procedures
Transgenic mice
F4 generation Six3-Cre mice (Furuta et al., 2000), backcrossed onto the congenic C57BL/6 strain from the mixed 50% C57Bl/6, 50% 129SV/J background (a generous gift of Y. Furuta, University of Texas, Houston, Texas), were crossed with mice that carry the Rosa26 reporter (R26R) allele (Soriano, 1999). In the R26R construct, the lac-Z gene was modified by the insertion of a neo expression cassette, which prevents the transcription of the lac-Z gene product, β-gal. Because loxP sites flank the neo cassette, the bacterial enzyme Cre-recombinase (Cre) can excise it, allowing β-gal to be transcribed (Soriano, 1999). By viewing β-gal-lir in adult Six3-Cre,Rosa26R mouse brains with immunohistofluoresence (IHF), we were able to infer where in the brain Six3-Cre was expressed both during development and adult life.
Brain extraction and sectioning
The Six3-Cre,R26R mice were given an overdose of the anesthetic sodium pentobarbital. They were perfused transcardially with cold, heparinized PBS (15 ml) followed by 4% paraformaldehyde (PFA, 5 ml). The brains were removed and submerged in 4% PFA overnight, then transferred to 30% sucrose for 48 hours. They were then frozen by immersion in Isopentane and stored at –80 C° until they were cryo-sectioned. Brains were sectioned into either coronal or sagittal sections to a thickness of 30 µm and suspended in PBS containing 0.1% sodium azide (preservative) in glass scintillation vials.
Immunofluorescent colocalization of TH and β-gal
Brains were rinsed three times for 5 minutes (3X5) in PBS, and then suspended in stock (0.03%) hydrogen peroxide (H2O2) for 20 minutes. After another wash (3X5 in PBS), they were suspended in 8% Normal Donkey Serum (NDS) in PBS with 0.1% Triton X-100 (TX-100; detergent) for at least one hour, after which the sections were suspended with the primary antibodies for at least 48 hours in the same solution. Primary antibody concentrations, determined by titration experiments, were as follows: TH-antibody (1:750, mouse monoclonal anti-tyrosine hydroxylase, CHEMICON, Temecula, CA); β-gal antibodies (1:6,000 rabbit IgG fraction to β-galactosidase, ICN Pharmaceuticals, Inc, Aurora, OH or 1:5,000 goat polyclonal Anti-β-galactosidase from Biogenesis, Brentwood, NH). Sections were rinsed (3X5) in PBS with 3% NDS and 0.1% TX-100, and then suspended with both secondary antibodies for at least two hours: 1:200 donkey anti-rabbit and 1:200 donkey anti-goat IgG (II+L) conjugated to Cy2TM (cyanine, a green fluorophore with an excitation range between 425-525 nm with a peak at 492 nm, and emission range between 490-600 nm with a peak at 510 nm) or donkey anti-mouse IgG (II+L) conjugated to Cy3TM (indocarbocyanine, a red fluorophore with an excitation range between 475-575 nm with a peak at 550 nm, and emission range between 550-650 nm with a peak at 570 nm) affinity purified secondary (1:200, Jackson ImmunoResearch, West Grove, PA) in 3% NDS and PBS with 0.1% TX-100 in all vials. After several test sections were examined with a fluorescent microscope (Nikon Eclipse TE 200), the remaining sections were rinsed (2X5) in PBS and 4µl stock DAPI (3µM, DAPI dihydrochloride D-1306 from Molecular Probes, Eugene, OR) was added for 5 minutes. Sections were rinsed (3X5) in PBS, mounted on Superfrost plus glass slides, dried for no more than 4 hours, and briefly cleared in EtOH and xylene before being coverslipped with DPX. Each fluorophore was visualized using a specific Nikon filter: Cy2 (β-gal): cube no. 96170, “EF-4 FITC HYQ 25mm”, excitation 460-500, barrier 510-560 nm; Cy3 (TH): cube no. 96157, “EF-4 G2E/C”, excitation 528-553, barrier 590-650 nm; DAPI: cube no. 96101, “EF-4 UV-2EC Dapi/Hoechs”, excitation 340-380, barrier 435-485 nm. A Photometrics Cool SNAP fx camera and RS Image 1.7.3, (Roper Scientific, Trenton, NJ) were used to acquire the images, and composite images were created in Adobe Photoshop.
Control experiments
One set of sagittal sections was processed with secondary antibodies alone to control for non-specific staining. In addition, we examined two brains from Six3-Cre mice that did not carry the Rosa26R, under identical conditions, and observed no β-gal-lir. In pilot studies, we observed very faint fluorescence in all TH-lir cells of areas A9 and A10 with the 96170 filter using the digital camera (although not by visual inspection of the sections). Therefore, we evaluated the specificity of the fluorescence by comparison with sections stained only for TH, or only for β-gal-lir. The faint fluorescence in TH-lir areas detected under the 96170 filter on sections stained only for TH was equivalent to that from sections stained for both TH and β-gal. In addition, no faint fluorescence was observed under the 96170 filter on sections stained only for β-gal-lir. Thus, we did not consider faintly fluorescent cells under the 96170 filter to be β-gal-lir.