Differential impact of Lhx2 deficiency on expression of class I and class II odorant receptor genes in mouse

Abstract

Odorant receptor (OR) genes can be classified into two types: fish-like class I OR genes and mammalian-specific class II OR genes. We have previously shown that Lhx2, a LIM-homeodomain protein, binds to the homeodomain site in the promoter region of mouse M71, a class II OR, and that a knockout mutation in Lhx2 precludes expression of all tested class II OR genes including M71. Here, we report that most class I OR genes, which are expressed in a dorsal region of the olfactory epithelium, are still expressed in Lhx2-deficient embryos. There are two exceptions: two class I OR genes, which are normally expressed in a more ventral region, are no longer expressed in Lhx2 mutant mice. Lhx2 is transcribed in olfactory sensory neurons irrespective of expression of class I or class II OR genes. Thus, a deficiency of Lhx2 has a differential impact on class I and class II OR gene expression.

Introduction

The olfactory system of mammals is capable of detecting millions of different types of chemicals in the environment. Olfactory sensory neurons (OSNs), the primary sensory neurons in the olfactory system, initiate the sense of smell by detecting odorants through odorant receptors (ORs). These ORs are G-protein coupled receptors with a putative seven-transmembrane domain structure. OR genes form the largest gene superfamily present in any genome analyzed so far. Since their initial discovery in rat (Buck and Axel, 1991), OR genes have been identified in several vertebrate species (Mombaerts, 1999).

In mouse, ∼ 1375 OR genes have been identified from the genomic sequence (Zhang et al., 2007). Phylogenetic analysis separates OR sequences into two broad but distinct classes of OR genes, referred to as class I and class II ORs (Glusman et al., 2001, Zhang and Firestein, 2002). Class I ORs are phylogeneticaly more ancient (Glusman et al., 2001) and resemble the family of ORs first reported in fish (Ngai et al., 1993) and then in frog (Freitag et al., 1995). There are 158 class I genes in the mouse genome, all of which are located in a single cluster on Chromosome 7 that is interrupted by the β-globin cluster (Bulger et al., 1999, Tsuboi et al., 2006, Zhang et al., 2007). By contrast, class II genes, the terrestrial-specific OR genes, are distributed throughout the mouse genome (Zhang et al., 2007). An OSN most likely expresses a single functional allele (Chess et al., 1994) of a single OR gene (Malnic et al., 1999). The mechanisms underlying this feature of ‘singular’ OR gene expression are not understood, but irreversible DNA rearrangements have been excluded by nuclear transfer experiments (Eggan et al., 2004, Li et al., 2004). Deletion of an OR coding region reveals that OSNs can co-express two OR loci, one of which can produce a functional OR (Feinstein et al., 2004, Lewcock and Reed, 2004, Serizawa et al., 2003, Shykind et al., 2004).

Short transgenes of two class II genes, MOR23/MOR267-13 and M71/MOR171-2, replicate many of the intricate features of OR gene expression (Rothman et al., 2005, Vassalli et al., 2002). Genomic sequences comprising as little as 405 and 161 bp upstream of the putative transcription start sites (TSSs) of the MOR23 and M71 OR genes, respectively, are capable of directing zonal restriction of transgene expression and axonal projection of transgene-expressing OSNs to the endogenous MOR23 or M71 glomeruli. In both cases, the sequences upstream of the TSS contain a homeodomain site (TATTXX) and an O/E-binding site (Y3CAR4 in which at least one of the pyramidines is C and at least one purine is G) within 37 base pairs (bp) of each other. These motifs can be discerned within the putative promoter regions of other class II genes in mouse, rat and human (Hirota and Mombaerts, 2004, Hoppe et al., 2003, Hoppe et al., 2006, Rothman et al., 2005, Vassalli et al., 2002). Mutagenesis of these motifs in the M71 promoter revealed that combined mutations in both sites can abolish transgene expression (Rothman et al., 2005). However, the introduction of identical mutations into the endogenous M71 locus by gene targeting did not preclude expression, but resulted in decreased numbers of M71-expressing OSNs and ventralization of the expression pattern in the olfactory epithelium (Rothman et al., 2005). It thus appears that homeodomain and O/E sites are involved in the control of class II gene expression, but they are not the sole determinants. For class I genes, neither the promoter regions nor the motifs have been functionally tested.

The LIM-homeodomain protein Lhx2 and O/E-1 bind to the homeodomain site and O/E sites, respectively, of the M71 class II gene promoter region in vitro (Hirota and Mombaerts, 2004, Rothman et al., 2005). Lhx2-deficient embryos, which die late in utero and are severely anemic (Porter et al., 1997), do not express class II genes (Hirota and Mombaerts, 2004, Kolterud et al., 2004). Expression of class I genes has not been examined in Lhx2-deficient embryos.

Here, we document the expression of class I genes in Lhx2-deficient embryos. We find that most class I genes are expressed in the dorsal region of the olfactory epithelium of Lhx2-deficient embryos. There are two exceptions: two class I genes, which are expressed ventrally in wild-type mice, are no longer expressed in Lhx2-deficient embryos. It appears that these two genes have a class II-like promoter and a class I coding region. Our findings suggest that distinct regulatory pathways underlie the control of expression of class I and class II genes.