The first two authors contributed equally to this work.The ontology of imidazoline receptors was first proposed by Bousquet P in 1980s [1] and was classified as I 1 , I 2 , and I 3 (non-I 1 /I 2 ) subtypes. However, deficiency in high selective ligand restrains the illumination of I 1 imidazoline receptor functions. In 2000, Piletz JE screened human hippocampal expression library and found a strong candidate protein for I 1 imidazoline receptor, named "imidazoline receptor antisera-selected" (IRAS) [2]. Concurrently, Alahari SK discovered a protein with identical sequence as IRAS and named it as "nischarin" [3]. Afterward, this protein was proved to be similar with I 1 imidazoline receptor in tissue distribution, ligand-binding property, and intracellular signallings and mediate many processes such as hypotensive effect of rilmenidine, inhibition of opioid addiction, inhibition of cell migration, antiapoptosis, and so on [4]. Herein, we originally generated IRAS conditional knockout mice (IRAS floxed/floxed ) and IRAS null mice (IRAS À/À ), which might be valuable tools for functional exploration of IRAS/nischarin and I 1 imidazoline receptors.Mouse IRAS gene is located on chromosome 14 and scatters into 21 exons, and we generated IRAS floxed/floxed mice by flanking exon 4 with loxP sites and then obtained IRAS À/À mice by crossing EIIa-Cre mice ( Figure 1A and Data S1 for more details). As for IRAS À/À mice, genotyping with the primer pair 5loxP-f and 3loxP-r could distinguish wild-type, heterozygote, and knockout mice ( Figure 1B). Sequencing result of the PCR product from 5loxP-f and 3flank primer pair revealed that only a loxP site remained between upstream and downstream homologous arms. In RTqPCR, the amplification curve of IRAS À/À was normal when the primer pair was originated from exon 3 and exon 5; however, the product is smaller than the wild type ( Figure 1C). Moreover, the amplification curve of IRAS À/À was null when the primer pair was originated from exon 2 and exon 4 ( Figure 1D). All the results above illustrated the normal transcription and the absence of exon 4 counterpart in the transcription product of IRAS À/À mice, which resulted in a new stop code in the following exon 5. In the Western blot assay, three bands, between 95 and 130 kD, were absent in cerebellum tissue of IRAS À/À mice ( Figure 1E). As several bands that represent functional IRAS have been reported, including 85 and 33 kD in human and 67 kD in bovine, these characteristic bands were proposed to be the splicing products in mouse cerebellum. All the above declared the functional deletion of IRAS gene.The IRAS À/À mice were born small compared with wild-type littermates (Figure 2A, B). Furthermore, weights of knockout embryos at day 12.5 (0.121 g, 0.133 g, and 0.143 g) were smaller than those of wild-type littermates (0.173 and 0.173 g) ( Figure 2C). These results suggested the participation of IRAS in 978 CNS Neuroscience & Therapeutics 19 (2013) 978-981 ª 2013 John Wiley & Sons Ltd prenatal growth and the postnatal growth re...