Samples for this study come from CORK observatories installed at IODP Holes U1382A and U1383C as described elsewhere (Edwards et al., 2012b). In brief, instrument strings containing OsmoSampler systems (Wheat et al., 2011) were deployed at different depths within the holes from 2011-2017. Additional OsmoSampler systems were deployed at the wellhead of the CORKs in 2014 and were inoculated with bottom seawater, making them useful for identifying differences between crustal subsurface and bottom seawater inoculated microbial communities. OsmoSampler systems included Flow-through Osmo Colonization Systems (FLOCS) for mineral colonization experiments, as described elsewhere (Ramrez et al., 2019). Each FLOCS contained sterile (autoclaved and ethanol-rinsed) substrates including crushed basalts, pyrite, pyrrhotite, or inert glass beads housed within polycarbonate cassettes and sleeves. Fluids were introduced into the FLOCS via the OsmoSampler pumps, allowing fluid microbial communities to colonize the substrates.

All FLOCS were recovered in October 2017 during cruise AT39-01 of the RV Atlantis with the ROV Jason (Woods Hole Oceanographic Institution) following methods described elsewhere (Ramrez et al., 2019). In brief, the instrument strings with the downhole FLOCS were pulled up to the ship on a wire, then immediately disassembled to store the FLOCS in a cold room. Wellhead FLOCS incubated in milkcrates attached to crustal fluid umbilicals at the seafloor were also collected on this cruise with the ROV. FLOCS contents were distributed inside an ethanol-rinsed and UV-irradiated HEPA-filtered hood using sterile instruments. Substrates for the experiment were stored cold (2-4C) and in the dark in sterile centrifuge tubes with ultrafiltered (0.22 m-mesh Millipore Sterivex and 0.02 m-mesh Whatman Anotop 25 filters) crustal fluid until incubation at the shore-based laboratory. Parallel samples for initial characterization of the substrate biofilms were transferred to sterile cyrovials, flash frozen with liquid nitrogen, and stored at -80C until analysis. In addition to the FLOCS samples, 4-10L samples of raw and unfiltered crustal fluids were collected into ethanol-rinsed cubitainers after collection into gamma-irradiated bags on the seafloor using a Mobile Pumping System. Cubitainers were stored cold (2-4C) and in the dark for approximately one year before beginning the experiments.

A detailed description of the cathodic poised potential protocol used is available elsewhere (Jones and Orcutt, 2019). In brief, glass two-cell, three-electrode MFC systems (Adams and Chittenden, CA, USA) were used as incubations chambers, run in parallel with a multichannel potentiostat (model CHI1030C, CHI Instruments, TX, USA). Nafion 117 proton exchange membrane (Fuel Cell Store, TX) separated the half-cells. Cells were filled with distilled water and autoclaved prior to filling with media and sample inocula. Ag/AgCl reference electrodes (Gamry Instruments, PA, USA part 932-00018 and/or Analytical Instrument Systems, NY, USA) were calibrated before each run by immersion in 3 M NaCl and compared against a known electrode kept only for that purpose (max +/- 20 mV drift from the value of lab master electrode). Working electrodes (WEs) were 2 4 cm2 Indium Tin Oxide (ITO) coated glass slides (Delta Technologies Ltd, CO, USA part CB-50IN-1111), constructed for MFC as described elsewhere (Rowe et al., 2015). Counter electrodes (CEs) were carbon cloth with a 4 4 cm2 surface area. Electrodes were sterilized by rinsing with 80% ethanol, air-drying, then exposing to UV radiation for 15 minutes per side.

Each experiment consisted of five treatments: Three MFCs filled with buffered and double autoclaved crustal fluid (cool and oxic), inoculated with samples, and incubated with a WE set poised at -200 mV versus a standard hydrogen electrode (SHE) referred to as the Echem treatments; one MFC filled with the same fluid and WE but without any sample inoculum referred to as the Fluid treatment; and one glass bottle with a mixed sample inoculum and ITO electrode without any voltage applied referred to as the Offline treatment. Samples were transferred into the MFCs in HEPA-filtered, UV-irradiated biosafety cabinet or hood using sterile tools. An aliquot of each sample was also collected into a sterile plastic tube for microbial community analysis representing a time zero (T0) condition that may have changed from the Shipboard (SB) initial condition. For the Echem and Fluid treatments, the counter cell redox couple was H2O/O2. Testing (data not shown) determined that a strong kill control method was necessary to achieve sterility of the crustal fluid, consisting of 1 h autoclaving at 121C, incubation at room temperature in the dark for ~24 h to allow for spore germination, and then another 1 h autoclaving at 121C before cooling down to 4C for the addition of substrate. Sodium bicarbonate was added to the media (0.1 M) as a buffer. MFCs were incubated at 4C in the dark.

Once the MFCs were constructed, a cyclic voltammetry (CV) sweep was performed for each cell with the potentiostat before each poised potential experiment at a scan rate of 0.1 V s-1 and sample interval of 0.001 V, across -1 to 1 V range. Then the chronoamperometry (CA) experiments began by applying a constant voltage (-200 mV versus SHE) to the Echem and Fluid MFCs. Current generation was monitored and recorded until current changes began to decline after a period of approximately 15 days. At the end of each CA experiment, two further CV sweeps were performed: one with the incubated ITO electrode (Tend i) and one with a fresh ITO electrode dipped into the incubated media (Tend ii).