Bacterial growth in 3000 ug/ml carbenicillin (antibiotic)


Optical Density (OD)


UV Absorbance Spectroscopy


In vivo screen (growth in media containing reporter antibiotic) was used to identify slow growth (less stable) variants.
Screening of the randomized library entailed primary, secondary and tertiary steps (see section Materials and methods) due to the fact that the difference in the growth rate for the most stable variant (MonA; a Gb1 variant that has a Tm value >100 deg C) and the least stable variants is greater than but still similar to natural variations in bacterial growth rates. The primary screen entailed picking slow-growing colonies identified on plates that contained reporter antibiotic. These colonies were then grown in the liquid media for the purpose of amplification and isolation of plasmids. In the second screening step, the purified plasmids were transformed into the reporter strain and separately plated to confirm slow-growth phenotype. For colonies that passed the primary and secondary steps an additional step entailed growing the bacteria in a 96-well plate format while monitoring growth rates over 24h. Finally, the slow-growth phenotype was measured and further confirmed in larger volume (10 ml) cultures.

Primary screening of the MonB library on CKC 1000 plates
Fifty microliters of BacterioMatch two-hybrid system repor- ter strain competent cells (Stratagene) were incubated with the MonB library (random codons at positions 23, 27 and 45). Following standard transformation protocols, 90 ml of the transformation reaction were plated on CKC 1000 plates. The nomenclature ‘CKC 1000’ corresponds to plates that contain chloramphenicol (12.5 ug/ml ), kanamycin (50 ug/ ml), and carbenicillin (1000 ug/ml). Chloramphenicol is used to maintain the reporter construct on the F’ episome, and kanamycin is used for selection of bacteria that contain the plasmid that expresses the three-domain chimeric con- struct. Upon expression, the three-domain chimeric construct functions to up-regulate the reporter genes located on the F’ episome (i.e. b-lactamase and b-galactosidase). The nomen- clature ‘CK plates’ corresponds to plates that contain just chloramphenicol and kanamycin but no reporter antibiotic (i.e. carbenicillin). The CK plates were used as controls for colony counts and to estimate the library size (total number of colonies). Small, slow-growing colonies were picked from the CKC 1000 plates and inoculated into 5 ml LB media for the purpose of overnight growth and plasmid recovery.

Secondary screening of the MonB library on CKC 1000 plates
MonB library plasmids that were isolated from slow-growing colonies (the primary step described above) and control plas- mids (the genes for MonA, MonB-WT and MonB-ORDES in the chimeric construct) were transformed into the reporter strain competent cells. Transformation reactions were plated onto small CK plates and small CKC 1000 plates. The result- ing colonies were compared with both the control CK plates and the control variants (MonA, MonB-WT and MonB-ORDES). This comparison was used to identify var- iants that grew more slowly than the MonB-WT colonies. Successful variants were used in the tertiary screening process (i.e. the 96-well plate format described next) to elim- inate false positives.

Tertiary screening of the MonB library in 96-well plates using CKC 2000
The MonB library variants that passed the primary and sec- ondary screens were subjected to a final step of screening using 96-well plates that contained carbenicillin at a concen- tration of 2000 ug/ml. For each of the selected MonB library variants, six colonies were picked form CK plates and inoculated in 1-ml CK media. Cultures were grown overnight and diluted the next day to an OD 600 of 0.1 in a volume of 180 ml of LB containing carbenicillin at a concentration of 2000 ug/ml. The 96-well plates were incubated in a micro-plate reader for 24 h, at 37 degC, with continual shaking and OD 600 readings recorded every 15 min. Growth curves data for the six trials for each variant were averaged and compared with the average growth curves of MonA, MonB-WT and MonB-ORDES. The MonB library variants with growth rates equal to or less than MonB-WT were selected for sequencing and for sub-cloning into the expression vector for the purpose of producing pure protein for circular dichroism (CD) analysis and melting temperature determination.
Large-scale screening in 10-ml cultures containing CKC 3000
In addition to the three screening step described above, the variants that met the slow-growth criteria were further ana- lyzed in a large volume format to confirm the slow-growth phenotype. All Gb1 and MonB library variants in the chi- meric construct were transformed into BacterioMatch two- hybrid system reporter strain competent cells and plated on CK plates. Colonies were picked and grown overnight in 20-ml CK media. The next day the overnight cultures were diluted to OD 600 of 0.1 in 10-ml volume of CK media and allowed to grow for 15 min. The OD 600 readings were measured to insure that all cultures had the same OD 600 before adding carbenicillin. Carbenicillin (3000 ug/ml) was then added to all cultures and incubated at 37degC, 275 rpm for 4 h. The OD 600 readings were measured for all cultures after 4 h. The OD readings illustrated in Fig. 2 were obtained from this large volume growth format.