Tissue and zonal-specific expression of an olfactory receptor transgene

Abstract

Discrimination of odorants is thought to arise from the selective expression of one of a small number of individual receptors in any single olfactory neuron. Receptor genes are expressed in a small subset of neurons throughout a zonally restricted region of the sensory epithelium. We demonstrate that a 6.7 kb region upstream of the M4 olfactory receptor coding region was sufficient to direct expression in olfactory epithelium. Moreover, reporter expression recapitulated the zonal restriction and distributed neuronal expression observed for endogenous olfactory receptors. Transgenic lines were obtained that directed expression in two different receptor zones, one of which was identical to the endogenous M4 receptor. When the reporter was expressed in the same zone as the endogenous M4 receptor, the two expression patterns were, in large part, nonoverlapping. These results suggest a model in which important regulatory elements are located in close proximity to transcription initiation sites of the olfactory receptor genes and receive information defining zonal patterning via long-range processes.

Fig. 1.

M4 is a single copy gene.a, Partial restriction map of the mouseM4 gene locus showing the schematic location of theM4 open reading frame (filled box). Deduced protein sequence shows 317 amino acids, representing the M4 coding region. B,BamHI; H, HindIII;R, EcoRI. b, Five micrograms of mouse liver DNA were digested with the indicated enzymes in each lane. The Southern blot was probed with the³²P-labeled M4 coding sequence. The probe showed a single band hybridizing in each lane. The size shown on theright corresponds to each of the hybridized bands.

Fig. 2.

Endogenous expression of M4. RNA from different tissues was used in a RT-PCR assay with primers PQ1 and PQ2, spanning the M4 coding region. PCR products (1 kb) are shown resolved on a 1% agarose gel. The grouped tissues represent two independent sets of RT-PCR reactions with brain, olfactory bulb, olfactory neuroepithelium (nose), liver, pancreas, and testis. Sections (10 μm) cut through the mouse nasal cavity were used in immunohistochemical staining, using anti-M4 antibody. A restricted distribution of the M4-positive cells is shown confined to the posterior loop, indicated byarrowheads (middle panel). A higher magnification of the section is shown in the bottom panel; M4-expressing neurons are marked witharrowheads.

Fig. 3.

Mapping and sequence analysis of theM4 transcription start site. A schematic diagram shows the M4 ATG and the 5′ proximal region (6.7 kb). Theopen box represents the 1 kb M4 coding region, and the M4-specific (PQ3–PQ4 nested) primers are shown by the two bars used in the RACE reaction. Results from the RACE analysis mapped the transcription start site 4 kb 5′ to the ATG, resulting in 2.7 kb of untranscribed sequences 5′ of the 300 bp exon. The underlined AG and ATGrepresent the splice donor site and the translation start site, respectively. The bottom panel shows the PCR products generated from the genomic DNA as the template, 2.7 kb containing the 300 bp exon (E), lane 2. A 4 kb PCR product represents the intron (I),lane 7. Lanes 3–5 represent the negative controls and lane 6 the positive PCR control (see Materials and Methods). M, Marker lanes.

Fig. 4.

Structure of the M4 transgene construct. The 6.7 kb DNA 5′ to the M4 ATG was fused to the E. coli gene encoding lacZ and the 3′ end of the mouse protamine 1 structural gene (β-galactosidase, mp1, shaded boxes). The mp1 gene provides a poly(A⁺) addition site and a 93 bp intron. Primers designed across the intron are shown by the arrows.Below, An agarose gel showing a 635 bp PCR product generated from the transgene message, lane 3, and a 728 bp from the genomic transgene (T-G) DNA, lane 5. No products are seen in lanes 2 and4, representing the wild-type mouse DNA and samples with no reverse transcriptase (−RT), respectively. Marker lanes are shown as lanes 1 and6.

Fig. 5.

Reporter transgene expression of E5and D5 in different tissues. RT-PCR experiment was performed on the cDNA from seven different tissues, using the primer pair RR318 and RR319. A 728 bp PCR product was generated with the genomic DNA from the transgenic mouse (TG), whereas the cDNA from various tissues generated a 635 bp product that is 93 bp smaller. The right panel represents the β-galactosidase reporter expression in the D5transgenic line. The top panel(+RT) shows PCR products with reverse transcriptase, and the bottom panel(−RT) is shown without reverse transcriptase. Expression in the E5 transgenic line is shown on theleft panel with a 123 bp marker lane represented byM. The tissues are NE, nose;BR, brain; TE, testis; SP, spleen; LU, lung; LR, liver; andKI, kidney.

Fig. 6.

Patterns of expression of β-galactosidase in mouse nasal cavity of D5 and E5transgenic mice. a, Coronal sections of the mouse nasal cavity are shown with the schematic localization of theD5 (left) and the right nasal cavity of an E5 transgenic mouse stained with X-gal at a similar sectioning level. The extents of the transgene expression (which coincided with M4 antibody staining) are indicated inblue. The regions in a comparable section that were not expressing OMP by in situ hybridization are indicated ingreen. The septum (SEPTUM), the positions of ectoturbinates 1 and 2, and the positions of endoturbinates II and IIIare indicated. The dorsal region is displayed at the topof the image. Overlays were produced by scanning images and processing them in Photoshop. Coronal sections through the E5 mouse nasal cavity (b–d) were stained with X-gal. The reporter expression is localized to the mature neurons within the neuroepithelial layer. A low magnification of the E5coronal section (a–c) shows X-gal staining, marked withblue arrowheads. The staining is confined to the most lateral and ventral regions of the epithelium on both sides of the septum (zone IV). Higher magnification of the section (d) shows X-gal-positive neurons over the central portion of the epithelium (blue arrowheads). Double-labeled sections from a D5 transgene animal were post-fixed in Zamboni and stained for activity with X-gal, followed byM4 antibody (e, g). A coronal section at low magnification shows double-stained neuroepithelium with X-gal staining inblue; peroxidase-stained M4-positive cells are represented by brown arrowheads (zone II). Regions of the epithelium showing no X-gal or M4staining are depicted by gray arrows. A high-magnification micrograph (f) showsM4-positive neurons colocalized with neurons expressing β-galactosidase, which is driven by the putative M4promoter region. Approximately 1% of the M4 or β-galactosidase-positive cells showed double staining (black arrowheads) within the same cell with peroxidase and X-gal (g).

Fig. 7.

Disruption of the transgene integration sites inD5 and E5. Five micrograms ofC57BL/6J (Wt) and D5 andE5 mouse liver DNA were digested withPvuII (left) or EcoRI (right), electrophoresed in 0.75% agarose, blotted onto a nylon membrane, and hybridized with the ³²P-labeled probe. The primer pair PQ74/PQ79 generated a D5-specific junction probe used on the left, and PQ 37/PQ39 generated a probe for E5 junction that was used on theright. DNA sizes are indicated on thesides of the panels. Both D5(PvuII-digested) and E5(EcoRI) lanes show an extra band, representing the transgene insertion-induced disruption.