>

Results

Microbial Library Screening

Individual colonies of each bacterial species were taken from TSA (tryptic soy agar) plates and grown in TSB (tryptic soy broth) overnight to ensure cells were viable. Cells were pelleted at 4500xg for 5 min and suspended in phosphate buffered saline (PBS; pH 7.4) and pelleted again to wash the pellet. The pellet was then suspended in Raymond’s minimal media with no additional carbon source added. This suspension was used to inoculate the following experimental samples:

Raymond’s media + no carbon: negative growth control
Raymond’s media + 1 PPM PFAS (PFOA or PFOS): test sample
Raymond’s media + 100 PPM glucose: positive growth control
Raymond’s media + 1PPM PFAS (PFOA or PFOS) + 100 PPM glucose: toxicity control

Following inoculation, samples were monitored daily/bi-weekly for optical density as a measure of growth for approximately 1 month.

The experiment was setup in such a way that if any particular sample showed no growth in the negative control, and positive growth in all remaining samples, it could be concluded that the added PFAS was likely being used in some way in the cellular metabolism. The toxicity control was included to ensure that the sample was not failing to grow simply due to PFAS cytotoxicity.

It was found for a majority of the samples, either a significant decrease in OD was observed when exposed to PFAS as a sole carbon resource, or no significant OD increase was observed compared to the negative control. A single sample did show a very small OD increase when grown in 1PPM PFOA, along with acceptable results for the controls (no growth in negative control, very strong growth in controls containing glucose). This sample was identified as Delftia acidovorans, and was used in further studies.

Delftia acidovorans Fluoride Studies

In order to determine if this sample was capable of defluorinating PFOA, a fluoride monitoring experiment was performed. Delftia acidovorans was grown in 1 PPM PFOA in an identical fashion to the previous experiment. A larger scale was used to allow frequent aliquot removal to measure fluoride content by selective ion probe. This measurement was performed by removing 1 mL aliquots from growth mixture, and pelleting of cells by centrifugation. The aliquot was then tested directly by the probe for fluoride content. A comparison was performed by growing Pseudomonas fluorescens (a common gram negative soil bacteria) with a similar compound, ethylflouroacetate (EtFA) at 1 PPM. This compound has been shown previously to degrade by defluorinating strains such as Pseudomonas fluorescens. Thus, this serves as a positive control for our measurements.

Figure 1: Fluoride release compared to growth of Delftia acidovorans , measured by OD600. Growth was performed in Raymond’s minimal media with 1PPM PFOA.

Figure 2: Fluoride release compared to growth of Pseudomonas fluorecens, measured by OD600. Growth was performed in Raymond’s minimal media with 1PPM EtFA.

As shown in Fig 1, D. acidovorans does exhibit growth in minimal media with PFOA as the sole carbon resource. However, no changes in fluoride concentration could be seen over the course of the experiment, suggesting that either the fluoride concentration was below the limit of detection of the probe, or metabolism was limited to the terminal carbon of PFOA, where no fluorine is present. Fig. 2 shows that our measurement method does work as expected, as fluoride concentration increases along with the increase in OD for P. fluorescens.

Delftia acidovorans Cell Lysis Studies

To test the possibility that enough fluoride was not liberated to stand out from the background in the previous experiment, a follow up was performed that eliminated monitoring cellular growth. Instead large amounts of cells were grown in TSB and lysed to expose the cytosol content. This crude lysate was then exposed to either PFOA (100 PPM), EtFA (100PPM), or water and fluoride concentrations were monitored over time using the selective ion probe (Figure 3). A positive control using P. fluorescens was run under identical conditions (Figure 4).

Cell lysis was performed by sonication of cell pellets suspended in PBS. In addition, half of the samples were not lysed, and are referred to as cell slurries for this experiment.

Figure 3: Fluoride measurement of D. acidovorans crude lysates. Data collected at each time point was linearly scaled such that the negative control (water) was identical to the starting conditions to account for possible fluctuations not associated with defluorination.

Figure 4: Fluoride measurement of P. fluorescens crude lysates. Data collected at each time point was linearly scaled such that the negative control (water) was identical to the staring conditions to account for possible fluctuations not associated with defluorination.

In both cases, fluoride release was observed in both P. fluorescens and D. acidovorans when exposed to PFOA when cells were lysed. A very small increase in fluoride was observed for PFOA in the cell slurries for both samples as well. Interestingly, it was observed that the fluoride concentrations in the cell lysates did not maintain over time, as seen by the decrease in signal after a peak value. This was hypothesized to be due to fluoride being incorporated into other molecules, resulting in loss of signal. It would be expected that any protein activity in a crude lysate would decrease as active peptidases would be functional in this reaction setup. Therefore we assumed that any defluorinating enzyme was being destroyed throughout the reaction, and that the rate of fluoride production in the system was decreasing until it fell below the rate of fluoride uptake in side reactions, leading the data seen in this experiment.

At this point, we decided to attempt to identify a possible dehalogenase from Delftia acidovorans. The genome of this microbe is known, and two haloacid dehalogenases were identified (PZP66635.1& WP_011137954.1), referred to as type I and type II. Both sequences were used in modeling using Phyre as well as used to construct biobricks for testing in E. coli.

Delftia acidovorans Dehalogenases

Naming Scheme

Naming notation for Dehalogenase genes and expression blocks

Plasmid notation:
Represented by pXX_DeHa
Where XX can represent:
C: pSB1C3
A: pSB1A3
K: pSB1K3
KB: pSRKBB

Dehalogenase gene & Expression Block Notation
Represented by DeHaX.Y.Z
Where,
X is the gene type (1 or 2 ; referencing the two dehalogenase genes in D. acidovorans)
Y is the variant (native variant = 1, E. coli codon optimized = 2, etc.)
Z shows the tags/modifications on the protein (0 = no modifications, HisC = 6xHis tag on carboxyl terminal, HisN = 6xHis tag on amine terminal, gene, etc.)
Note: If Z is replaced by the 'gene' modifier, the Dehalogenase gene sequence is being referenced. In all other cases, the fully assembled expression block is referenced. Example: DeHa1.1.gene references BBa_K3347000 while DeHa1.1.0 references BBa_K3347001.

Example notations
pKB_DeHa2.2.HisN: Dehalogenase type II, variant 2 (E. coli optimized) with a N terminal 6xHis tag all inside the pSRKBB vector
pC_DeHa1.1.0: Dehalogenase type I, variant 1 with no further tags or modifications all inside the pSB1C3 vector

Expression in E. coli

We expressed several variants of the dehalogenase genes in E. coli, including the native Dehalogenase type I and II under the control of the lac promoter in DH5a, as well as the E. coli optimized variants for both types I and II under the control of the T7 promoter in BL21 DE3. Among these optimized variants, His tagged versions were also expressed, ultimately leading to the expression of the following proteins in E. coli:

pC_ DeHa1.1.0 in DH5a
pC_ DeHa2.1.0 in DH5a
pKB_DeHa1.2.0 in BL21 DE3
pBK_DeHa1.2.HisC in BL21 DE3
pBK_DeHa1.2.HisN in BL21 DE3
pBK_DeHa2.2.0 in BL21 DE3
pBK_DeHa2.2.HisC in BL21 DE3
pBK_DeHa2.2.HisN in BL21 DE3

Each of these strains were grown in 50 mL LB + appropriate antibiotic selector at 37 degrees C in a shaking incubator until OD reached 0.6, at which point cultures were split into 2x25mL aliquots. One aliquot received IPTG (1mM) to induce protein production. Cultures were shaken for an additional 6 hrs before harvesting. E. coli were lysed by pelleting at 5000xg for 10 min, and suspending in lysis buffer (PE LB from G-Biosciences + C0mplete protease inhibitor cocktail + 0.4 mg/mL Lysozyme), and shaken for 1 hr at 37 degrees C. Soluble proteins were separated from crude cell extract by centrifugation (5000xg for 30 min at 4 degrees C) and transferred to clean micro-centrifuge tube. Soluble protein extracts were placed on ice when in use and frozen at 0C for storage.

Initial Analysis of DeHa1.2 and DeHa2.2

To determine if catalytic activity is seen on small scale, a simple experiment was devised to monitor fluoride as a degradation product of PFAS hydrolysis. To a reaction vessel a sample of soluble protein extract was added to either PFOA, EtFA, or water. A time 0 measurement was made immediately upon addition and fluoride monitored over the course of 200 minutes. REsults can be seen below in figure 5. From this simple experiment it appears that both the positive control, EtFA, as well as PFOA-containing samples exhibit rapid fluoride concentration increase followed by a large period of unchanging concentrations when the blank was subtracted. The blank (not shown) exhibited a small increase in fluoride upon addition to a much lower extent than either samples. A cause of concern is the abundance of fluoride released throughout the course of the experiment. No greater than 400 ppb fluoride was measured in any sample, suggesting that while defluorination is occuring, the equilibration point is very low, ultimately resulting in very little free fluoride in solution. This analysis is not fully conclusive, and our team is not comfortable at this time making the claim that we have identified a PFAS defluorinating enzyme; however we are currently working to optimize reaction conditions for these enzymes in order to perform more in-depth analysis. Please reference our future work page for more detail.

Figure 5: Time-correlated fluoride concentration monitoring of A: DeHa1.2 and B: DeHa2.2 upon exposure to PFOA and EtFA. Fluoride concentration was measured by a selective ion probe.

Team:USAFA/Results