Team:WPI-Worcester/Collaborations
Collaborations
This year, we collaborated with Harvard Biodesign to determine if our most promising antifreeze proteins were effective in a different biofilm-forming strain of E. coli.
Here is the Biofilm Assay Protocol used by both teams participating in this collaboration! WPI
We completed the biofilm assay linked above with strains of E. coli with modified pili created by the Harvard team. The strains we tested were JW4283, a fim H knockout and negative control for biofilm formation, and C2M+C5 and C9-3+C5, which express modified pili when induced by arabinose and rhamnose. Upon receiving agar stabs from the Harvard team, we streaked plates and incubated them at 37C for 24 hours. Afterwards, we placed the plates in the refrigerator until we received the arabinose and rhamnose needed to induce pili formation in the experimental strains. Once we had all necessary materials, we grew 5mL liquid cultures of each of the 3 strains in LB in accordance with the protocol. The JW4283 culture was not supplemented with antibiotics or inducers, but the C2M+C5 and C9-3+C5 cultures were supplemented with 5 μL of chloramphenicol, 5 μL of ampicillin, 0.01% arabinose, and 0.5% rhamnose. We prepared the 1:100 dilutions in LB broth and M9 minimal media, which we found to support biofilm formation while designing our biofilm assay. The remainder of the protocol was left unchanged. Below are the results of the assay. A photo of the biofilm plate after it was stained with crystal violet on 9/3/15. Although there appears to be some remaining crystal violet, biofilm formation was not robust in any of the strains.
A table of the OD595nm for each well with the blank subtracted. Average ODs for each strain in both LB and M9 are also included.
A graph of average OD595nm for C2M+C5 and C9-3+C5 in M9 minimal media normalized to average OD595nm for JW4283 in M9 minimal media (indicated by the line at 1). Error bars=+/- standard error of the 4 ODs measured for the two strains in M9.
According to the results we collected, the Harvard strains with modified pili do not form biofilms under the conditions we subjected them to. These strains may be able to form biofilms under different conditions. Further investigation would have to be done to determine if different if different media types, incubation periods, or transformation into different strains of E. coli could induce biofilm formation. Determining optimal biofilm formation conditions could allow the Harvard team to use the biofilm assay to gather quantitative results for their various binding assays. However, it is possible that the modified pili interfere with biofilm formation. In that case, the biofilm assay would not be a valid method of quantification. Harvard
The Harvard team tested our TiAFP and BclA-TiAFP antifreeze proteins in the curli strain, a strain of E. coli expressing modified CsgA fusion proteins. This strain, which forms biofilms, was developed by the Joshi lab and therefore readily available to the Harvard team. They followed our general biofilm protocol, linked above, with the exception of the following conditions:
1. Round-bottom plates at the Wyss were not sanitary so we used two flat-bottom plates.
2. Curli strain obtained from Joshi Lab was grown in 1 mL pure YESCA media (100 mg Casamino acids, 10
mg yeast extract dissolved in ddH2O per 1 mL solution).
3. Liquid cultures were also supplemented with 5 μL of carbomycin (strains were grown on chloramphenicol
+ carbomycin resistance plates)
4. Having been informed by the Joshi Lab that the bacteria would almost certainly perish if left in a 96-well
plate for 48 hours at 37 degrees Celsius, the Harvard team followed their advice and incubated the plate
for 48 hours on a room-temperature shaker (25 degrees Celsius)
Following the protocol (except for the above-mentioned differences), the Harvard team grew up the delta FimB strain (negative control), wild type CsgA curli bacteria (positive control), the curli bacteria transformed with the BBa_J23117 plasmid and two separate strains with the curli bacteria transformed with two separate antifreeze protein plasmids developed by WPI (TiAFP and BclA). The plate reader results can be seen below. The loading order in the wells was:
A1 - A4) delta B
B1 - B4) WT CsgA
C1 - C4) Curli #1 BBa_J23117
D1 - D2 and D4) Curli #2 BclA-TiAFP (D3 was left out due to a loading error and likely became contaminated
with other bacteria)
E1 - E4) Curli #3 TiAFP A table of OD595nm for each well including average ODs for each strain.
A graph of average OD595nm for CsgA expressing J23117, BclA-TiAFP, and TiAFP normalized to average OD595nm for WT CsgA in (indicated by the line at 1). Error bars=+/- standard error for the individual wells for each strain.
The Harvard results indicated that on average there was more biofilm formation on the positive control (WT CsgA in row B) than on the curli bacteria strains that had been transformed with the antifreeze protein plasmids developed by WPI (rows D and E), while there was also greater average biofilm formation in the positive control than in the negative control (delta FimB in row A), both of which we expected to see. Repetitions of the assay will have to be done in the future to verify our results.
