[Isolation and functional analysis of tobacco MARs]

Two new MAR segments (M14 and M17) were cloned from tobacco genome. Both of the sequences contained several typical consensus sequences of MARs, which were different from the original MAR sequence, such as 90%AT-box, A-box, T-box, the base unpairing regions (BUR), autonomously replicating sequences (ARS), the consensus sequence for topoisomerase II, MAR recognition sequence (MRS), origin of replication (ORI), curved DNA motifs and ATATTT et al. To investigate the effects of these two sequences on gene expression in transgenic plants, 3 plant expression vectors were constructed with uidA gene coding beta-glucuronidase (GUS) which were flanked on one side and on both sides by the MARs we obtained. These plant expression vectors with one or two MARs were transformed into tobaccos via Agrobacterium-mediated transformation method, with the plant expression vector pCAMBIA2301 without MAR and wild type tobacco as controls. GUS histochemical staining results showed that the uidA gene expressed stably in transgenic tobaccos. Quantitative detection of GUS activity showed that the MARs could increase GUS expression levels in vivo in contrast to the controls, wherever they were flanked on one side or both sides of uidA gene. The vector ligated with MARs in the same direction on both sides of uidA could increase the GUS expression level much better than both vectors which just ligated with single MARs on one side. The former one increased the average GUS activity for 3.14 folds, but 1.56 and 2.43 folds for the latter two vectors with single MARs respectively contrasting to the pCAMBIA2301 control. But the expression differences among individual transformants were still obvious. Therefore, it was concluded that the DNA sequences we obtained in this experiment were two novel MARs and could enhance gene expression in vivo. In the meanwhile, although the numbers of the MARs typical motifs in M14 were more than in M17, especially the 90% AT box which had been considered to be the highest correlative motif with binding strength in vitro, the enhancement of gene expression was lower yet, which implied no correlation between improvement of gene expression and binding strength between MARs and nuclear matrix in vitro.