The Structure And Function Of φC31 Integrase And HIV Integrase Were Analyzed

Site-specific integrases were widely used in gene therapy and transgenic studies in pursuit of safe and long-term gene expression. Recently, Streptomyces phage OC31integrase was reported to be a potential tool to mediated site-specific recombination in many species. However its gene transfer efficiency was low compared to that of virus-based vectors. The lentivirus vector system origined from HIV-1was the most often used virus-based vectors in present, but its random integration caused serious safe problem. So a gene transfer system in possession of both high efficiency and safety could be ideal for gene therapy. In this study, we resolved the structural foundation of ΦC31integrase-mediated site-specific recombination to its recognition sequences and investigated cellular proteins which influenced the HIV-1integrase mediated gene transfer efficiency and gene expression, tried to provide earlier necessary theoretics groundwork for integrase’s structure and function relationship study and also for exploiting the above mentioned ideal gene transfer system.Firstly, the function of ΦC31integrase’s C terminal was studied. Previously, DAXX was found interacting with ΦG31integrase and inhibiting the integration efficiency dramatically.451RFGK454of ΦDC31integrase was identified as the interacting motif. Here we reported that the motif was necessary for the integrase’s activity. Bioinformatics analysis indicated a probable DNA binding motif from407-517just around this four amino acids. EMSA was carried out to study their DNA binding properties. The results showed that deletion of the tetramer or replacement of Arg with His both reduced the integrase’s binding ability and binding specificity dramatically. However, both truncated Int407-517and IntΔ407-517couldn’t bind to attB sequence. These suggest that this motif is necessary for its specific DNA binding character, but it is not sufficient for the binding and there might be more than one binding related motifs in the integrase.Afterward we focus on HIV-1integrase and studied the cellular proteins which can interact with the integrase and influence the integration and gene expression of lentivirus.We found the distribution of HIV-1integrase changed from an even diffused state to a typically PML-NB-like punctate state when SUMO1/Ubc9or SUMO2/Ubc9were co-expressed in293T cells. Actually the HTV-1integrase had a strong co-localization with PML when they were co-transfected in293T cells. The above phenomena made us to speculate that interacting with HIV-1integrase, SUMO1/Ubc9 or SUMO2/Ubc9were just like a bridge to bring the integrase to PML-NB. Based on this supposition, yeast mating assay, subcellular localization assay and Co-Immunoprecipitation assay were carried out and the results told us that HIV-1integrase could interact with SUM01, SUM02and Ubc9directedly and with PML indirectedly. Further investigation showed that HIV-1integrase could be covalently modified by SUM01and SUMO2.Then the influence of these proteins on the integration and gene expression of lentivirus were detected. Through Flow Cytometry the percentage of EGFP positive cells and mean EGFP intensity were examined to indicate the integration and gene expression of lentivirus respectively. The rusults showed that when Ubc9was overexpressed, the percentage of EGFP positive cells decreased from37.29%±7.54%to25.21%±2.63%(p<0.05).The percentage declined even dramatically when SUM01was co-expressed, to be16.46+3.02%(p<0.01), or when SUMO2was co-expressed, to be22.89%±4.49%(p<0.05).The above results indicated that SUMO1, SUMO2and Ubc9could inhibit the integration of lentivirus evidently and there was an adding effect between SUMO1/SUMO2and Ubc9. A more accurate method named Alu-LTR Real-Time Nested PCR was also carried out to confirm the results. The trends were almost the same as the former method. Besides, the results of the mean EGFP intensity showed us that the expression of EGFP was also inhibited due to the existence of Ubc9in a certain extent, comparatively only80%to that of the control. At this time only Ubc9could influence the report gene expression. Though there was no detected inhibitory effect of PML itself to the integration and gene expression of lentivirus in our experiments, other members of PML-NB, such as sp100, DAXX could prevent the infection of lentivirus. These proteins were either modified by SUMO or had a non-covalent interaction with SUMO and could all interact with HIV-1integrse directly. Combining our results in this section, a model called SUMO pathway medicated inhibiting infection of lentivirus by PML-NB was brought forward:HIV-1integrse interacted with Ubc9and SUMO1/2and the three proteins formed a complex. Because Ubc9is an SUMO conjugating enzyme and PML-NB is a highly SUMO-modified structure, naturally, the Ubc9-integrse-SUMO1/2complex transferred to PML-NB. Through the SUMO pathway, PML-NB captured the HIV-1integrse and members of it interacted with the integrase and finally inhibit the integration and/or gene expression of lentivirus. On the other hand, SUMO1/2modified integrase could also localize to PML-NB.