Mutant plasmids were constructed with a single core sequence (CM, HM, OM), a double mutant plasmid with simultaneous mutations of c-CMV and c-HSV and a plasmid with all three sites mutated (AM) (Additional file 1: Fig. S5A–C)5. In three different strains of competent E. coli cells, DH5α, FastT1 and BL21, the transformation efficiency of plasmids CM, HM and DM were significantly higher compared with the unmutated plasmids (Fig. 5A–C). The transformation efficiency of the OM plasmid was not significantly different from the unmutated plasmid (Additional file 1: Fig. S5K). These results indicate the presence of a host defense effect that depends on pEGFP-N1 containing the core sequence in E. coli cells.
Considering the E. coli DH5α strain as an example, 100 pg, 10 pg and 1 pg of plasmid pEGFP-N1 were transformed into the same volume of competent cells. For100 pg plasmid there was a marked effect but for 10 pg or 1 pg plasmid, too few transformed clones were produced, resulting in a large variation in transformation efficiency (Additional file 1: Fig. S5I, J).
Another plasmid, pXT7-myh9a, with a lower transfection efficiency was selected. The plasmid contained two core sequences, one in ori and the other in the myh9a coding sequence. The core sequences at these two sites were mutated according to the previous strategy and plasmids with an ori mutation, myh9a mutation or with both sites mutated were constructed (Additional file 1: Fig. S5L). Transformation efficiency was increased by 1119% ((119.5–10.67)/10.67 = 11.19) when the core sequence in ori was mutated in the OM plasmid (Fig. 5D, E). By contrast, mutating the core sequence in myh9a derived from the zebrafish genome had no effect on the transformation efficiency of the plasmid. We conclude that the core sequence within ori is essential for resisting plasmid invasion but core sequences within intrinsic coding regions have no impact on the defense response in cells.
The impact of plasmid size on the differences in transformation efficiency of mutations in ori were investigated. Plasmids containing the same mutations as DM and MM were constructed and the length of myh9a truncated to produce plasmids of different sizes: 3 kbp (3027 bp), 4 kbp (4103 bp), 6 kbp (6294 bp) and 8.8 kbp (8913 bp). We found that the number of transformed clones of the 3 kbp unmutated plasmid (224.5 ± 52.6) was not significantly higher than that of the mutant plasmid (193 ± 42.6). However, the number of transformed clones of the unmutated 4 kbp plasmid was 89.5 ± 11.5and increased significantly to 170.8 ± 37.1, with a growth rate of 90.9%, when ori was mutated. Moreover, the growth rate of the 6 kbp plasmid increased by 384% ((42.8–8.8)/8.8) after mutation and that of the 8.8 kbp plasmid increased by 738% ((40.5–4.8)/4.8) (Fig. 5F). These results indicate that the core sequence improves transformation efficiency for large plasmids and also indicates the role that the core sequence plays in protecting cells from foreign plasmid invasion.
The sequence of 7 bases in the core sequence position of ori in different sized plasmids was investigated by randomly annealing bases at seven core positions, putting the sequences into the pEGFP-N1 and pXT7-myl7 plasmids before adding ampicillin for resistance screening. After 4 h, the plasmids were extracted for deep sequencing of ori. The results show that the core sequence, 5ʹ-GTTTGTT-3ʹ, accounted for 19.06% of the total of 371 sequences in pEGFP-N1 (Fig. 5G). However, for the pXT7-myl7 plasmid, 5ʹ-GTTTGTT-3ʹ only accounted for 2.24% of a total of 501 sequences and no influence was shown of the position of the core sequence (Fig. 5H). The above data shows that for larger plasmids, the core sequence 5ʹ-GTTTGTT-3ʹ is not conducive to the survival of the plasmid but may become the target of attack. Mechanisms involved in the host defense against the core sequence merit further study.
Moreover, we have shown that the ds-CMV fragment with the core sequence was cleaved when delivered to the zebrafish zygote, indicating that the core sequence is recognized and attacked by the defense system of eukaryotic cells. In addition, transcriptome sequencing analysis showed that the transcription level of E.coli DH5α transformed by c-CMV or by the c-CMV mutant, pEGFP-N1, was altered. Thus, we suspect that the core sequence may stimulate a new unknown defense response (data not shown). Besides, we also found that the expression of GFP on the plasmid was affected when the core sequence was mutated. In HEK293T cells, compared with cells transfected with pEGFP-N1 (Additional file 1: Fig. S7A–A’’’), photographed using the same parameters, it was found that HM (Additional file 1: Fig. S7C–C’’’) and OM (Additional file 1: Fig. S7D–D’’’) transfected cells had no GFP fluorescence, while DM (Additional file 1: Fig. S7E–E’’’) and AM (Additional file 1: Fig. S7F–F’’’) transfected cells showed attenuation of the fluorescence intensity. But in E. coli, the mutation of the core sequence does not affect the yield of the plasmid (Additional file 1: Fig. S6K). This result indicates that the mutation of core sequence will induce different defense mechanisms and affect the activity of different elements in the plasmid in eukaryotic cells. Our findings merit further study to illuminate the mechanism responsible.
To summarize, core sequences present in the CMV promoter, HSV poly (A) signal and ori are central to the resistance of bacteria to infection by foreign plasmids or DNA fragments. Moreover, such mechanisms may be exploited to improve techniques of transformation efficiency for large plasmids.