While some of the other methods are more sensitive and quantifiable than the drop-collapse method, none of them are practical for high-throughput screening. Unfortunately, even the drop-collapse assay involves a number of steps, including growing each strain in broth culture and testing the supernatant for its ability to collapse a water drop on a hydrophobic surface; this can be highly labor- and time-intensive and thus not suitable for a truly high-throughput screen in which thousands of strains would need to be tested. Furthermore, this test is generally used as a qualitative assay only, and a measurement of the collapsed water droplet under a microscope or many serial dilutions of each sample is required to get a semi-quantitative estimate of surfactant abundance. For this reason, high-throughput use of the drop-collapse assay in a mutagenesis screen would not identify strains which have either increased or incomplete loss of surfactant production. A novel biosurfactant detection method was developed here in order to quickly screen large numbers of bacteria for surfactant production directly on an agar plate. This atomized oil method is at least as sensitive as the drop-collapse assay, and was found to be useful for all tested biosurfactant-producing strains as well as synthetic surfactants. Additionally, it is semi quantitative, and is capable of identifying intermediate phenotypes. As an illustration of this method, the atomized oil procedure was used in the context of a high-throughput screen of mutants of P. syringae B728a to identify those altered in surfactant production. This method proved very effective, identifying multiple mutations of the gene cluster encoding the non ribosomal peptide synthetase responsible for syringafactin production,macetas 25 litros as well as several genes involved in its regulation.

The drop-collapse assay was performed as according to Bodour and Miller-Maier. 2 µl 10W-40 Pennzoil® was applied to delimited wells on the lid of a 96-well plate and allowed to equilibrate at room temperature. Next, 5 µl of either diluted surfactant samples or supernatant from bacterial cultures or resuspended bacterial colonies were pipetted onto the oil surface. Drops which retained a spherical shape were scored as negative for surfactant content, while drops which had a visibly-decreased contact angle with the oil and spread were scored as positive for surfactant content. The atomized oil assay was conducted as follows: Bacteria were spotted onto LB or KB agar plates using sterile toothpicks and grown overnight. For more uniform inoculation of plates with cells diluted to a common cell concentration, a colony was resuspended in phosphate buffer, the OD600 determined in a spectrophotometer, and a small volume of suspension containing the desired number of cells was pipetted onto the plate surface and incubated overnight. Alternatively, if visualizing purified surfactant, 5 µl of diluted surfactant was pipetted onto the plate and allowed to equilibrate for 30 minutes before assaying. An airbrush was used to apply a fine mist of mineral oil onto the plate with an air pressure between 15 and 20 psi. Depending on the airbrush and setup used, experimenters will need to optimize the appropriate settings in order to deposit a constant and controlled stream of oil droplets.Halo radii were measured with a ruler from the leading edge of the bacterial colony to the edge of the surfactant halo. A novel surfactant detection assay was developed using P. syringae B728a which produces the lipopeptide surfactant syringafactin as a test organism. Syringafactin has previously been demonstrated to be a surfactant by use of the drop-collapse assay; supernatant from P. syringae DC3000 collapses on a hydrophobic surface, demonstrating the presence of a surfactant, while supernatants from mutant strains which do not produce syringafactin do not.

Although they focused on characterizing the syringafactin extract from P. syringae DC3000, the authors also confirmed that syringafactin is produced in strain B728a. We developed a method of surfactant detection involving the misting of oil droplets onto agar plates, hypothesizing that the presence ofsurfactants would alter the interaction of the oil with the agar surface. When a fine mist of mineral oil was sprayed over the surface of a KB agar plate on which bacterial colonies of P. syringae B728a had grown, a light-diffractive halo was seen around the colonies. In contrast, no such halo was observed around E. coli DH5α , a strain which is not predicted to produce a biosurfactant. Upon microscopic inspection, it was seen that oil droplets on an un-inoculated agar surface and near DH5α were in energetically unfavorable distorted shapes. This was presumably due to random heterogeneity in the hydrophobicity of the agar surface. However, when surfactants spread over the agar surface such as in the vicinity of P. syringae B728a, the droplets assumed a more uniform, energetically favorable hemispherical shape. Furthermore, the light-diffractive halo observed macroscopically was actually caused by the de-wetting, or beading, of the oil droplets near the surfactant-producing bacteria. The oil droplets, which presumably were in contact with the biosurfactant, stood higher on the plate and appeared more spherical than droplets on the agar surface away from surfactant-producing colonies. These raised droplets reflected light at a different angle, making them appear brighter under an indirect source of light. In order to show that this atomized oil assay was indeed detecting biosurfactant, we obtained a variety of strains with characterized biosurfactant production and for which isogenic strains blocked in biosurfactant production are available. In addition to P. syringae DC3000 which produces syringafactin, we tested Pseudomonas fluorescens SS101, Bacillus subtilis 3610,macetas de 9 litros and Pseudomonas aeruginosa PA14. All of the tested biosurfactant-producing bacterial strains produced easily detectable bright halos when sprayed with atomized mineral oil, while none of the biosurfactant mutants exhibited halos in this assay. Thus all biosurfactants tested were readily detected with the atomized oil assay and no evidence of false positive indications of surfactant activity was obtained.

While this new assay readily detected a variety of both lipopeptides and glycolipids of bacterial origin, we tested the behavior of other types of surfactants with this procedure. All of a variety of commercially available surfactants were detectable by this assay. Many of the surfactants behaved similarly to the biosurfactants, causing the oil droplets to assume raised hemispherical shapes that appeared bright when illuminated. However, a few of the surfactants created a less obvious “dark halo” in which the oil droplets still assumed a circular form, but were less hemispherical and had an increased contact with the water-agar surface. These “dark halo” droplets, in contrast to the raised droplets in “bright halos,” were flat and appeared less bright than the surrounding surfactant-free droplets at certain angles. Interestingly, when the surfactants were ranked by their hydrophilic-lipophilic balance values, a common value used to describe surfactants in industry, it was found that surfactants with low HLB values all yielded bright halos while those with higher HLB values resulted in dark halos. The sensitivities of the atomized oil and drop-collapse assays to detect a variety of surfactants were compared. Using a range of dilutions of a given surfactant, we determined the lowest concentration of that surfactant that was still detectable by a given assay. Additionally, crude extracts of surfactin, rhamnolipid and syringafactin were prepared, and their limits of detection by the two assays were compared. For all tested surfactants and biosurfactants, the atomized oil assay was found to be more sensitive than the drop-collapse assay. In general, the atomized oil assay detected surfactant at concentrations more than 10-fold lower than that of the drop-collapse assay. In order to relate the size of the observed halo around a source of surfactant to the amount of that surfactant, different dilutions of a syringafactin-containing extract were tested with the atomized oil assay and halo diameters were measured. A log-linear relationship between the amount of surfactant applied to plates and the diameter of the halo was observed. Thus a quantitative estimate of the relative difference in amounts of surfactant in different prepared samples can be readily estimated. For each 10-fold increase in concentration of the spotted surfactant, the radius of oil drop alteration increased by about 1.7 mm. Because halo sizes were very consistent for a given amount of surfactant, with standard deviations rarely above 0.25 mm, careful replicate measurements of halos should easily enable the distinction in amounts of surfactant that differ by three-fold or more.

However, it must be emphasized that such semi quantitative estimates are only relevant when comparing samples of the same surfactant on a single medium, since different surfactants will diffuse at different rates.While it may be possible to quantify the surfactant in a prepared sample by measuring halos, calculating the surfactant produced by a bacterial colony is confounded by the additional parameter of time. The prepared samples discussed above were applied at a distinct time and measured one hour later, but bacterial colonies could produce surfactant over many hours of growth. Given that the distance over which a specified amount of surfactant will spread across an agar surface would be expected to be somewhat dependent on time, we determined the extent to which this factor would influence estimates of surfactant concentration using the atomized oil assay. A fixed concentration of a syringafactin-containing extract from P. syringae B728a was applied to agar plates and destructively analyzed by the atomized oil assay at various times after application. Halo radii continued to increase with time, although the rate slowed considerably after about two hours. Because of this, in addition to the fact that the bacteria continue to multiply and that the production of many biosurfactants is regulated by cell density , we concluded that halo measurements could not be used to calculate the absolute amount of surfactant produced by a colony without further investigation. Fortunately for screening purposes, relative amounts of surfactant production should be readily assessed using the atomized oil method unless the growth rate of the strains being compared differs greatly. Given that a consistent estimate of surfactant production from a given bacterial strain would be needed to compare strains in a high-throughput survey, we estimated variance in estimates of syringafactin production in replicate cultures of wild-type P. syringae B728a. Replicate cultures of P. syringae were established on plates by toothpick inoculation. On average, about 2.3 ± 0.6 X 106 bacteria were applied to a plate using this technique. Radii of halos from the resulting syringafactin production after colony formation were 8.8 ± 0.8 mm. To determine if variations in the number of cells initially deposited to establish spots affected the apparent surfactant production, a defined number of cells were applied in replicate spots onto the plate and oil was sprayed onto the plates after incubation overnight as in the toothpick-inoculated plates. The radii of oil drop halos around these replicate spots exhibited a similarly small variation as those around colonies established by toothpick inoculation. Application of cells by toothpick therefore results in inconsequential variations in eventual surfactant production as measured by this assay. Due to this limited variation, any strains displaying a halo that differed in radius by 20% or more than a reference strain would likely be significantly different in surfactant production. However, it is important to later confirm the regulation transcriptionally, in the event that a smaller halo is the result of a slower growth rate in a mutant strain. The atomized oil assay was used to individually screen a library of about 7,700 transposon mutants of P. syringae for surfactant production. Mutants with a halo radius that differed by more than 1.5 mm from that of wild-type colonies were identified in an initial assessment; this should correspond to an approximate 10-fold increase or decrease in surfactant production. Mutants with large growth defects were discarded based on the logic that fewer cells will produce less total surfactant, although three mutants with slight growth defects were saved for further testing, which includes a cell-normalized measurement of surfactant production. These mutants with visible growth defects were later determined to have insertions in the suhB homolog Psyr_1233, the secA homolog Psyr_4094, and a PhoH-like protein Psyr_4346. No mutations were observed to cause visible increases in the growth rates. 28 total mutants with significantly altered surfactant production were identified after replicate tests.