Precise pattern of recombination in serotonergic and hypothalamic neurons in a Pdx1-cre transgenic mouse line

Abstract

Background: Multicellular organisms are characterized by a remarkable diversity of morphologically distinct and functionally specialized cell types. Transgenic techniques for the manipulation of gene expression in specific cellular populations are highly useful for elucidating the development and function of these cellular populations. Given notable similarities in developmental gene expression between pancreatic β-cells and serotonergic neurons, we examined the pattern of Cre-mediated recombination in the nervous system of a widely used mouse line, Pdx1-cre (formal designation, Tg(Ipf1-cre)89.1Dam), in which the expression of Cre recombinase is driven by regulatory elements upstream of the pdx1 (pancreatic-duodenal homeobox 1) gene.

Methods: Single (hemizygous) transgenic mice of the pdx1-creCre/0 genotype were bred to single (hemizygous) transgenic reporter mice (Z/EG and rosa26R lines). Recombination pattern was examined in offspring using whole-mount and sectioned histological preparations at e9.5, e10.5, e11.5, e16.5 and adult developmental stages.

Results: In addition to the previously reported pancreatic recombination, recombination in the developing nervous system and inner ear formation was observed. In the central nervous system, we observed a highly specific pattern of recombination in neuronal progenitors in the ventral brainstem and diencephalon. In the rostral brainstem (r1-r2), recombination occurred in newborn serotonergic neurons. In the caudal brainstem, recombination occurred in non-serotonergic cells. In the adult, this resulted in reporter expression in the vast majority of forebrain-projecting serotonergic neurons (located in the dorsal and median raphe nuclei) but in none of the spinal cord-projecting serotonergic neurons of the caudal raphe nuclei. In the adult caudal brainstem, reporter expression was widespread in the inferior olive nucleus. In the adult hypothalamus, recombination was observed in the arcuate nucleus and dorsomedial hypothalamus. Recombination was not observed in any other region of the central nervous system. Neuronal expression of endogenous pdx1 was not observed.

Conclusions: The Pdx1-cre mouse line, and the regulatory elements contained in the corresponding transgene, could be a valuable tool for targeted genetic manipulation of developing forebrain-projecting serotonergic neurons and several other unique neuronal sub-populations. These results suggest that investigators employing this mouse line for studies of pancreatic function should consider the possible contributions of central nervous system effects towards resulting phenotypes.

Figure 1

Transgenic mouse genotyping using multiplex allele-specific PCR and melting curve analysis. PCR was performed in an optical cycler (ABI 7300) using 1-10 ng genomic DNA from mice of the indicated genotypes and a reagent mix containing SYBR GreenER. Amplification plots (A, C) and melting curves (B, D) are shown. Primers were designed to amplify 2 specific products: a genomic control product, generated from any genomic mouse template; and a transgene-specific product, generated only from genomic templates containing a cre transgene. Normalized fluorescence (y-axis, A &C) is the baseline-subtracted ratio of SYBR signal to ROX (passive reference dye) signal during amplification cycling (A, C). Normalized fluorescence derivative (y-axis, B & D) is the 2^nd derivative of normalized fluorescence during the melting curve step. Dotted lines (A, C) indicate cycle threshold. A & B. Genomic DNA from a wild-type mouse; note robust amplification of the genomic control product with a single melting peak (allowing the distinction of a negative result from a failed PCR). C & D. Genomic DNA from a pdx1-cre^Cre/0 mouse; note robust amplification with 2 distinct melting peaks corresponding to the control and cre-specific products. Arrows indicate presence of the genomic control product (Control) and the transgene-specific product (Cre).

Figure 2

Cre-mediated recombination in the pancreatic primordium and inner ear in the e10.5 embryo. Whole-mount images of a pdx1-cre^Cre/0; rosa26^LacZ/+ embryo (A & B) and a pdx1-cre^0/0; rosa26^LacZ/+ embryo (C) processed for β-galactosidase activity. β-galactosidase activity was observed in the pancreatic primordium (bottom arrow, left panel) and inner ear formation (top arrow, A; region in higher magnification in B). β-galactosidase activity was not evident in pdx1-cre^0/0; rosa26^LacZ/+ control embryos (C). Scale bars, 1 mm (A and C) and 150 μm (B).

Figure 3

Cre-mediated recombination coincides with serotonergic neurogenesis in the e11.5 embryo. Epifluorescence images of the e11.5 developing hindbrain of embryos, transversely sectioned, immunostained for GFP (green) and 5-HT or Nkx2.2 (red). In pdx1-cre^Cre/0; Zeg^GFP/0 embryos, GFP was always expressed in or adjacent to newborn serotonergic neurons. Both serotonergic and non-serotonergic neurons expressed GFP in the rostral hindbrain (B, C, D) (rhombomere 1, r1; rhombomere 2, r2) and caudal hindbrain (ch) (F). The degree of co-expression of GFP and the serotonergic phenotype was greatest in the rostral hindbrain, with little overlap in the caudal hindbrain and sparse GFP expression in rhombomere 4 (r4) (E). GFP was not expressed in the Nkx2.2⁺ progenitor zone (D). GFP expression was not evident in sections from a pdx1-cre^0/0; Zeg^GFP/0 embryo (A). Scale bars, 100 μm.

Figure 4

Cre-mediated recombination in the hindbrain and diencephalon in the e16.5 embryo. Epifluorescence images from pdx1-cre^Cre/0; Zeg^GFP/0 (B, D, F, H) and pdx1-cre^0/0; Zeg^GFP/0 (A, C, E, G) e16.5 embryos, transversely sectioned, immunostained for GFP (green) and 5-HT (red). GFP was expressed in the dorsal raphe nucleus (dr) (B), caudal linear raphe (clr) (D), caudal hindbrain (ch) (F) and hypothalamus (hp) (H). In the rostral hindbrain, GFP expression occurred in the serotonergic dorsal raphe and caudal linear nuclei (B, D). In the caudal hindbrain, GFP expression was observed in the non-serotonergic inferior olive nucleus, adjacent to serotonergic raphe nuclei (F). GFP expression was not evident in sections from pdx1-cre^0/0; Zeg^GFP/0 control embryos (A, C, E, G). Scale bars, 100 μm.

Figure 5

Cre-mediated recombination in the ventral diencephalon in the e11.5 embryo. Epifluorescence images of the e11.5 diencephalon of pdx1-cre^Cre/0; Zeg^GFP/0 embryos, transversely sectioned, immunostained for GFP (green) and Isl1 (A) or Mash1 (B) (red). GFP was not expressed at the ventral surface near the floor plate adjacent to the Isl1⁺ Mash1⁺ neurogenic zone. Scale bars, 100 μm.

Figure 6

Cre-mediated recombination in forebrain-projecting serotonergic neurons, inferior olive neurons and hypothalamic neurons in the adult brain. A-D: Individual optical sections obtained using confocal imaging of saggital sections from adult pdx1-cre^Cre/0; Zeg^GFP/0 mice. A: Wide-field image of the serotonergic dorsal raphe nucleus (dr) demonstrating extensive and anatomically restricted expression of TPH and GFP in this structure. B: Higher-magnification image of the dorsal raphe nucleus: a large majority of TPH⁺ neurons express GFP and that all GFP⁺ cells in this region are serotonergic neurons. C: In the median raphe nucleus (mr), there was partial overlap between GFP and TPH expression. D: In the caudal hindbrain, GFP expression was observed in the inferior olive nucleus (io), adjacent to but not overlapping with serotonergic raphe nuclei. E-F: Brightfield images of saggital sections obtained from adult pdx1-cre^Cre/0; rosa26^LacZ/+ mice, processed for LacZ activity. E: The inferior olive nucleus was labeled with LacZ. F: Multiple nuclei of the hypothalamus, notably the dorsomedial, lateral and arcuate nuclei, were labeled with LacZ. Scale bars: 80 μm (A); 60 μm (B, C, D); 200 μm (E); 150 μm (F).

Figure 7

Lack of detectable expression of endogenous pdx1 in the mouse hindbrain. Section from the r1 region of a pdx1-cre^Cre/0; Zeg^GFP/0 embryo, immunostained for GFP (A) and Pdx1 (B) using TSA amplification. No detectable expression of Pdx1 was observed, despite robust expression of GFP. Scale bars, 100 μm.