Evaluating the Impact of Hydraulic Fracturing on Streams using Microbial Molecular Signatures

This protocol can be used to answer the question of whether and how hydraulic fracturing affects bacteria in nearby streams and by extension the streams themselves. The main advantage of this technique is its holistic nature, as it takes the researcher from sample collection all the way through data analysis. Demonstrating the procedure will be Jeremy Chansee, a bioinformatician in my laboratory, Sydney Regal, an undergraduate student at Juniata College, and Gillian Leister, my laboratory manager.

To collect sediment samples for nucleic acid extraction, use gloves to submerge a capped, sterile 50 milliliter conical tube into the stream water from the shore. While the tube is submerged, remove the cap and use the cap to scoop approximately three milliliters of sediment from a depth of one to three centimeters into the tube. After collecting the sample, dump all but approximately one milliliter of water out of the tube and use a 1, 000 microliter pipette to add three milliliters of DNA/RNA preservative to the sample.

Swirl the capped conical tube for five seconds to thoroughly mix the preservative and sample and store the sample on ice. Upon return to the lab, store the sample at minus 20 degrees Celsius for 16S DNA analysis or minus 70 degrees Celsius for metatranscriptomics RNA analysis. For filter collection, completely fill and empty an entire sterile one-liter bottle with stream water three times to condition the bottle before filling the bottle one final time.

On a stable surface, draw a full volume of stream water into a sterile luer lock syringe and connect the syringe to a sterile and DNA/RNA free 1.7 centimeter diameter polyethersulfone filter with a 0.22 micron pore size. Flush the entire volume of stream water through the filter. When the entire sample volume has been filtered in the same manner, draw approximately 20 milliliters of air into the syringe and push the air through the filter to remove any excess water from the filter.

Next, use a P1000 micropipette to add two milliliters of DNA/RNA preservative to the larger opening of the filter while holding the filter horizontally with the tip of the pipette in the barrel of the filter to ensure that the preservative enters the filter. Then seal the filter with tightly wrapped squares of paraffin film around each opening and place the filter into a sterile sample bag on ice. Upon returning from sampling, store the filters for 16S or for metatranscriptomic analysis as demonstrated.

Before beginning a sample transfer, clean the work area with 10%bleach and 70%ethanol. For nucleic acid extraction from a sediment sample, use a flame and ethanol sterilized metal tool to transfer approximately 250 milligrams of sample into a microcentrifuge tube. For nucleic acid extraction from a filter sample, use a 70%ethanol and flame sterilized vice grip to break open the filter casing on the sterile surface and remove the core from the casing.

Use a sterile scalpel to slice at the top, bottom, and along the seam of the core and use sterile tweezers to fold the filter paper before cutting it into small pieces with the scalpel. Then carefully place the filter pieces into a microcentrifuge tube without contacting any unsterilized surfaces. For lysis of the cells within either type of sample, subject the tube to a cell disruptor at high speed.

After at least five minutes, centrifuge the samples and transfer the supernatant to a new sterile microcentrifuge tube. Add lysis buffer to the supernatant at a one-to-one ratio and transfer the solution to the provided filter. Place the filter into a new microcentrifuge tube and add 400 microliters of preparation buffer to the tube.

After centrifugation, add 700 microliters of wash buffer to the tube and centrifuge the filter again. After discarding the flow-through, add 400 microliters of wash buffer to the tube for an additional centrifugation and transfer the filter to a new sterile microcentrifuge tube. To elute the DNA, treat the filter with 50 microliters of DNase/RNase-free water for five minutes at room temperature.

In the meantime, place a three HRC filter into a collection tube and add 600 microliters of HRC prep solution. After centrifugation, transfer the filter into a new sterile microcentrifuge tube and transfer the eluted DNA to the filter for centrifugation. The flow-through contains the extracted DNA.

To create a DNA 16S RNA library, first use the freshly extracted the DNA product for 16S ribosomal RNA amplification with a standard PCR protocol. Mix seven microliters of the resulting PCR product and 13 microliters of DNase-free water and load the PCR solution onto a 2%agarose gel. Then run the gel at 90 volts for 60 to 90 minutes to check for a band size of 386 base pairs as evidence of a successful amplification.

To purify the DNA 16S ribosomal RNA library, pool 10 microliters of the PCR products that yielded bright bands and 13 microliters of the samples that yielded faint bands in a sterile microcentrifuge tube and load around 150 to 200 nanograms of each pooled sample into individual wells of a new 2%gel. After running the gel as demonstrated, excise the 386 base pair bands from the gel and use a commercial kit to purify the DNA. Then elute the purified DNA with 30 microliters of 10 millimolar Tris hydrochloride and pack the purified libraries in dry ice before shipping for next generation sequencing.

In this representative analysis, all extractions except for one would be dubbed successful. Bright bands observed following PCR amplification indicate success for the 16S protocol. 16S samples should have a minimum of 1, 000 sequences with at least 5, 000 being ideal, while metatranscriptsomic samples should have a minimum of 500, 000 sequences with at least 2 million being ideal.

Sediment samples from 21 different sites at 13 different streams for 16S and metatranscriptomics analysis are shown. Of those 21 sites, 12 were downstream of fracking activity and classified as hydraulic fracturing positive, and nine were either upstream of fracking activity or in a watershed in which fracking was absent and classified as hydraulic fracturing negative. As assessed by PERMANOVA analysis, the separation observed between the hydraulic fracturing positive and negative data suggests that the hydraulic fracturing positive samples were impacted by fracking.

The most important random forest predictors would reveal which features were most essential for correctly differentiating samples. If a taxon is identified as important by the random forest model, its antimicrobial resistance profile in hydraulic fracturing positive samples could be compared to its profile in hydraulic fracturing negative samples. If they differ greatly, that could suggest that fracking fluid containing biocides entered the stream.

Microbes are everywhere, so contamination is a significant potential issue with this type of work. The initial sample transfer steps are especially prone to this. If one is interested in microbial biodegradation or other metabolisms, shotgun sequencing of RNA can be used to investigate active microbial gene expression.

This technique allows the standardization of molecular techniques for investigating in-situ bacterial communities, as well as bioinformatics analyses of bacterial sequence data.