Antimicrobial genes from Allium sativum and Pinellia ternata revealed by a Bacillus subtilis expression system

Antimicrobial genes are found in all classes of life. To efficiently isolate these genes, we used Bacillus subtilis and Escherichia coli as target indicator bacteria and transformed them with cDNA libraries. Among thousands of expressed proteins, candidate proteins played antimicrobial roles from the inside of the indicator bacteria (internal effect), contributing to the sensitivity (much more sensitivity than the external effect from antimicrobial proteins working from outside of the cells) and the high throughput ability of screening. We found that B. subtilis is more efficient and reliable than E. coli. Using the B. subtilis expression system, we identified 19 novel, broad-spectrum antimicrobial genes. Proteins expressed by these genes were extracted and tested, exhibiting strong external antibacterial, antifungal and nematicidal activities. Furthermore, these newly isolated proteins could control plant diseases. Application of these proteins secreted by engineered B. subtilis in soil could inhibit the growth of pathogenic bacteria. These proteins are thermally stable and suitable for clinical medicine, as they exhibited no haemolytic activity. Based on our findings, we speculated that plant, animal and human pathogenic bacteria, fungi or even cancer cells might be taken as the indicator target cells for screening specific resistance genes.

. Allium sativum cDNA library insert sizes. (a) The cDNA library of E. coli expression system was inserted between 500 -2000 bp. (b) The cDNA library of Bacillus subtilis expression system was inserted between 600 -1500 bp. Figure S2. Nematicidal activities of test proteins. (a-b) Nematode repellence assay. The nematodes of the fourth stage larvae (L4) were placed on the NGM medium (nematode growth medium) between the test protein and the control protein (the distance was very important and must be equally spaced), and the condition of the nematode was continuously observed under microscope after 6 -12 h. (a) More nematode individuals were present far away from the antimicrobial protein, (b) less number of nematodes within the range of the resistant protein. (c) CI (choice index) = (number of nematodes in lawn A) -number of nematodes in lawn B) / total number of nematodes. CI value < 0, nematode preferred strain B; CI value > 0, nematodes preferred strain A; CI value = 0, no preference for nematodes. Strain B stands for 11 different antimicrobial strains, including PtR1259, PtR280, PtR325, PtR743, PtR594, PtR840, PtR857 and PtR1776, and strain A is the control. Experiments were repeated three times under the same conditions and similar results were observed. Vertical bars showed SD.

Construction of E. coli system cDNA library
For the construction of cDNA library by using E. coli expression system, the difference from the B. subtilis secretory protein expression system is that the cDNA and the vector pET22-(b) were cleaved with EcoR I and Not I restriction endonucleases (Table S5). After cDNA and vector ligation, the final products were first transformed into HST08 competent cells and then transformed into E.
coli DE3 competent cells. The plasmid was extracted and stored at -20 °C .

Screening of E. coli system cDNA library
After the plasmid was transformed into DE3, the bacterial cells were diluted 10 3 fold. Then, 100 μl of bacteria was taken and inoculated onto a membrane (0.45 μm pore size) fixed on LB plates containing appropriate amount of ampicillin. concentration was determined with a spectrophotometer and adjusted with PBS solution. The extracted protein was stored at 4 °C for further use.

Protein extraction of E. coli expression system
Transformed E. coli cells were growth on the membrane over LB plates containing ampicillin. Then, plates were incubated at 37 °C for 24 h. After 24 h, membrane was transferred to fresh LB plates containing ampicillin with and without IPTG (as control). After 3 h of culture, staining was performed with staining agar medium. If the plate is full of blue single colonies that proved the gene is induced and expressed. Then colonies on the LB/AMP (control) and LB/AMP/IPTG plates were scraped off with sterile blades and collected into two 1.5 ml tubes, respectively. Volume of each culture was recorded and stored at -20 °C. The frozen samples were thawed on ice and equal volume of lysate (20 mM HEPES, pH 7.6, 500 mM NaCl, 10 % glycerol, 1 mM EDTA, 1 Mm PMSF, 5 µg/ml leupeptine, 1 % v/v Aprotinin, and 0.1 % NP40) as the sample volume was added. After vortexing and stirring, the sample was put into liquid nitrogen for several seconds and thawed in 4 °C water. Above procedure was repeated 6 times to completely break the cells. After centrifugation for 10 minutes, 4 °C and 12000 rpm, and then supernatant was collected into 1.5 ml tubes and stored at 4 °C.

Antibacterial bioassay
Indicator bacteria were shaken for 8 -10 h and concentrations were adjusted with spectrophotometer. The NA plates were prepared and 300 µl of indicator bacteria mixed with 4 ml of semi-solid NA medium poured into upper layer of the NA plates. All protein samples were filtered with a 0.22 µm bacterial filter and 20 mg aliquots of protein were dropped onto the filter paper (0.5 cm in diameter

Percent haemolytic analysis
Briefly, haemolytic activities were tested with freshly prepared sheep red blood cells. Sheep red blood cells were collected then washed three times with PBS

Western blot
The isolated proteins were electroblotted onto PVDF membranes and the protein transfer was confirmed by using a color pre-stain marker. PVDF membranes were blocked in 2.5 % nonfat dry milk for 2.5 h, transfer to nonfat dry milk mouse containing mouse primary antibody overnight, washed three times in phosphate-buffered saline with 0.1 % Tween-20 (PBST) for 10 -15 minutes. After incubation with peroxidase-conjugated goat anti-mouse antibody for 3 h followed by three washes with PBST and detection by enhanced chemiluminescence (ECL) western blot assay. Gel imager was used to detect the bands.