Construction Of IRAS Knockout Mice And Screening Of Proteins Interactive With Fragments Of IRAS

In1980’s, the ontology of imidazoline receptor was first proposed by Bousquet,after which several indicative investigations were conducted to illuminate thepharmacological property and the molecular nature of the receptor.Pharmacologically, it is different from classical2receptors in ligand bindingand histological distribution. Nowadays, several subtypes of imidazoline receptorhave been identified, among which I1receptor participates in central regulation ofblood pressure, as well as renal sodium excretion, gastric secretion and insulin effect.However, the low specificity of commonly used reference compounds impacts thepharmacological delineation of the receptor.On the other hand, to pursue the entity of the receptor, Piletz screened the humanhippocampus expression library and found the candidate protein for I1receptor namedIRAS. Although many researches reaffirmed the relationship between IRAS proteinand I1receptor, bottleneck still remains on direct experimental evidence. Furthermore,no animal model is presently available for the research of IRAS function, includingthe regulations on blood pressure and drug dependence.Additionally, a cluster of IRAS biological functions, including materialization ofI1receptor, inhibition of cell invasion, inhibition of apoptosis and regulation of opiumreceptor system, were reported, meanwhile a cluster of splicing variants of IRAS weresimultaneously reported. However, the relationship between the functions and thevariants remains unclear.Therefore, we plan to construct IRAS conditional knockout mouse which wouldbe a powerful animal tool, with high selectivity, for I1receptor/IRAS proteininvestigation. Moreover，we plan to initiate a screening, with yeast two hybrid, forproteins interactive to segments of IRAS protein, in order to provide molecular cuesfor IRAS function.I. Construction and phenotyping of IRAS knockout miceMouse IRAS gene locates on chromosome14and scatters into21exons. Wenoticed that only2exons, the fourth and the fifth, are applicable for gene targeting.Accordingly, two targeting vectors were constructed.For vector₁, because the fifth exon and the sixth exon locate too close fromeach other, we targeted them as a whole to reduce the impact on the genome structure.The long homologous arm （HA） was derived from the upstream sequence for its composition of a long intron. The short HA was derived from the downstreamsequence which is composed of the seventh, the eighth and the ninth exons. Inaccordance with the backbone vector, the final vector₁was arranged in the order oflong HA, loxP, neo, target sequence, loxP and short HA. To coordinate with the vector,ES cell screening schema were as the followings: long PCR5’primer was located onintermediate sequence between target sequence and short HA, while3’primer waslocated downstream of the short HA. SpeI was applied in5’ southern blot, with wildband of11kb and recombinant band of7kb. EcoRI was applied in3’southern blot,with wild band of21kb and recombinant band of14kb. Constructed vectors werefirstly subjected to verification, and then lineated for electroporation. As a result,vector₁electroporation yielded304ES cell clones, and208positive clones afterloxP screening, however, no recombinant clone was found in long PCR and southernblot.For vector₂, which targeted the fourth exon, HAs with similar origins werearranged on the backbone in the order of long HA, loxP, target sequence, neo, loxP,short HA.5’primer of long PCR was located on the3’end of neo, while3’primer waslocated downstream of the short HA. Vector₂electroporation yielded366ES cellclones,268positive clones after loxP screening, and luckily, three recombinant clones,2-11B,1-10A and2-3D, were found in long PCR. After that, four chimera wereestablished from2-11B, and from one of which, we ultimately harvested therecombinant heterozygote. After amplified the population with C57/BL mice, wecrossed the recombinant with EII-cre mouse, which yielded the systemic IRASknockout mouse. Meanwhile, we crossed the recombinant with the flper mouse,which can delete the neo sequence between two FRT elements. Therefore, the IRASgene was floxed with two loxP elements in the offspring, from which spatio-temporalknockout mouse could be prepared.Next, we conducted the verification of the recombinant and the knockout mouse.Genomicly, schema for ES cell screening, including the loxP PCR and the long PCR,were again applied. The sequencing result of the long PCR product was identical toour presumption before. However, the electrophoresis result of the loxP PCR productshowed an unexpected band above the recombinant band. Additional experimentsrevealed that it was consisted of recombinant sequence and wild-type sequence. Referto the sequencing result of the recombinant genome, we postulated that it is a nealingproduct of one recombinant strand and two wild-type strands. Transcriptionally, thetotal mRNA from the liver of knockout mouse was extracted and reverse transcribedfor real time PCR. The sense primer was originated from exon3, and theamplification curve was normal and null when the antisense primer was originatedexon5and exon4, respectively, which illustrated the normal transcription and theabsence of exon4counterpart in the transcription product. Based on the sequence, a new stop code was created just downstream of the exon3, and the knockout peptidehas only125amino acids. Furthermore, proteins from cerebellum of knockout mouseand wild-type mouse were prepared and the western blot result revealed threediscrepant bands between95kD and130kD and one discrepant band at130kD,detected by C-terminal antibody and N terminal antibody respectively, whichsuggested that the physiological form of IRAS in mice were a product of alternativesplicing.Phenotyping of the knockout mice included the following modules: fertility andreproduction, development, and neuro-behavior character:a) IRAS knockout had no influence on the gamete formationb) IRAS knockout resulted in remarkable prenatal and postnatal growth retardation.c) IRAS knockout mice were susceptive to death. Typical lobar pneumonia wasfound in the necropsy.d) No difference in the body temperature between the wild type and the knockoutmouse, however, the food consumption of the knockout mice is significantly low.e) IRAS knockout resulted in hyperglycemia and low glucose tolerance.f) No difference in the locomotive activity and the result of open field test betweenthe wild type and the knockout mouse.g) Grip strength measurement and rolling rod test showed no difference in the motorability between the wild type and the knockout mouse.h) Water maze test revealed no difference in the learning and memory between thewild type and the knockout mouse.i) PPI test showed no difference in the susceptibility to schizophrenia between thewild type and the knockout mouse.j) Tail suspension test and forced swim test revealed that the knockout mouse is lesssusceptible to depression.k) In hot plate and tail flick test, knockout mice were more sensitive to thermalinduced acute pain.l) IRAS knockout mouse showed reduced response to morphine and methadoneanalgesia however, was more susceptible to multi-methadone induced toleranceand the subsequent physical dependence.Conclusively, we constructed two targeting vectors and achieved therecombinant ES cell clones, the recombinant mouse, the systemic knockout mouse, aswell as the floxed mouse. The phenotype of the systemic knockout mouse includedthe followings: growth retardation, hyperglycemia, modulation of pain and morphineresponse, modulation the susceptibility to depressionII. Protein interactionIn yeast two hybrid, human IRAS protein was divided into six segments, and five of which were cloned into pDBLeu vector as the baits. As the result of screening,totally three interaction were discovered from human fetus brain expression library.First, IRAS （621aa-870aa） corresponded to SERTAD1whole protein, NCBI gene ID:29950, which was reported to be an oncogene and actively takes part in cellproliferation and apoptosis suppression. Secondly, IRAS （431aa-620aa） correspondedto SORL1protein segment （started from1700aa）, NCBI gene ID:6653. SORL1isreported to regulate the trafficking of the precursor of amyloid-beta peptides, andhighly associate with Alzheimer’s disease. Thirdly, IRAS （621aa-870aa） correspondedto USP11protein segment （started from654aa）, NCBI gene ID:8237. USP11is akind of deubiquitinating enzyme. Afterwards, we focused the interaction verificationon SERTAD1. In yeast two hybrid and co-immunoprecipitation （co-ip）, we duplicatedthe interaction between IRAS （621aa-870aa） and SERTAD1. And then, we found thatthe whole IRAS protein and SERTAD1could bind to each other in co-ip, and aftertagged with fluorescent marker, they co-located in cell plasma of HEK293underconfocal inspection, with FRET value of0.16. Because IRAS is a candidate for I1receptor, moxonidine and agmatiine were tested for their influences on co-ip betweenIRAS and SERTAD1. The results showed that the interaction was intact from theeffects of these two drugs. IRAS was reported to involve two presumed splicing site,512and1152. In co-ip, SERTAD1could interact with IRAS （512-1504）, but not withIRAS （1152-1504）. Furthermore, we mapped the IRAS sequence with mammaliantwo-hybrid and found that IRAS （621aa-700aa）, which is a putative coiling coildomain, was the key position for the interaction.In conclusion, three IRAS interactive proteins were discovered by yeasttwo-hybrid, and we verified the interaction between SERTAD1and IRAS, then,located the key interactive element of IRAS on621aa-700aa. For the linkage tocarcinogenesis, SERTAD1has been considered as an oncogene. Therefore, our workmay be a cue for the association between IRAS and cancer.On the whole, in this study, we constructed IRAS conditional knockout mice,which is a high-quality animal model for I1receptor/IRAS research. Meanwhile, wefound four phenotypes in the knockout mice, including alterations in drug dependence,depression, development and diabetes. On the other hand, we found the interactionbetween IRAS and SERTAD1, which was a hint of IRAS function in carcinogenesis.