Comparative transcriptomics and scarless genome editing uncover a periplasmic c-type cytochrome important for extracellular electron uptake in Thioclava electrotropha ElOx9^T

ABSTRACT

Extracellular electron uptake (EEU) is a form of extracellular electron transfer that enables microorganisms to use solid-phase electron donors for respiration and biosynthesis, with important implications for biogeochemical cycling and biotechnology. Thioclava electrotropha is a metabolically versatile marine Alphaproteobacterium capable of autotrophic sulfur oxidation, heterotrophic growth, in addition to cathodic electron uptake despite lacking homologs to known extracellular electron transfer proteins from mineral-reducing or mineral-oxidizing microbes. Here, we explore the genetic basis for EEU in T. electrotropha through integrated transcriptomic, genetic, and electrochemical approaches. RNA sequencing revealed distinct transcriptional profiles for cathode oxidation versus sulfur oxidation, with 584 genes uniquely upregulated during EEU. Among the most highly upregulated genes were putative c-type cytochromes predicted to localize to the periplasm. A putative monoheme c-type cytochrome gene (AKL02_08760) designated pmcA showed >100-fold upregulation during cathode oxidation. We developed a genetic system enabling scarless deletions in T. electrotropha and demonstrated that deletion of pmcA significantly decreased EEU capacity, while complementation restored and enhanced electron uptake beyond wild-type levels. Phylogenetic analysis revealed PmcA homologs in 2,587 bacterial species across multiple phyla, with distribution patterns indicative of horizontal gene transfer. The pmcA gene belongs to a putative four-gene cytochrome c-multicopper oxidase operon, of which all members were upregulated during EEU, with the operon displaying atypical codon usage and elevated GC content consistent with horizontal acquisition. These findings establish a novel and potentially horizontally transferred mechanism for bacterial EEU. Further elucidation of this pathway will provide new targets for engineering enhanced bioelectrochemical systems with EEU capabilities.

IMPORTANCE

Though extracellular electron transfer (EET) has been shown to drive critical biogeochemical processes in a range of environments, there are only a limited number of biomarkers that help us assign the genetic potential for EET to other microbes. This is especially true for organisms that use EET to acquire electrons from external electron donors. This work identifies a novel gene involved in the extracellular electron uptake mechanism employed by the marine sediment chemolithoautotroph and sulfur-oxidizing bacterium, Thioclava electrotropha. Homologs of this novel gene are found in over 2,500 species spanning multiple phyla. Though biochemical characterization is necessary to fully understand the role of this protein in EET, this work supports the potential for a widely distributed and previously uncharacterized mechanism of extracellular electron uptake. As EET has enabled the potential for multiple biotechnological applications, including microbial fuel cells and microbial electrosynthesis, further characterization of this system has implications for improved engineering of bioelectrochemical systems or use of novel electrotrophic microorganisms.

INTRODUCTION

Extracellular electron uptake (EEU) describes the process by which some microorganisms accept electrons for respiration or biosynthesis from external sources, such as conductive minerals, other microorganisms, or in applied systems from cathodes (electrodes poised at electron-donating potentials) (1, 2). This process has important implications for understanding microbial ecology in environments where organic electron donors are limited, such as subsurface sediments. In marine sediments specifically, extracellular electron transfer processes are critical for elemental cycling where electron transfer between microorganisms and mineral phases drives mobilization of energy sources and nutrients that can influence global nutrient cycles and carbon sequestration. Further, electro(auto)trophic organisms hold considerable promise for biotechnological applications, including microbial electrosynthesis, bioremediation, biofuel production, and sustainable manufacturing processes (3–5).

Some of the best-characterized EEU pathways share homology to the MtrCAB porin-cytochrome complex in Shewanella oneidensis MR-1 that catalyzes iron reduction, in which MtrA and MtrC are decaheme c-type cytochromes and MtrB is an outer membrane ß-barrel protein that enables electron transfer across the outer membrane (6). Genetic evidence from Rhodopseudomonas palustris TIE-1 indicates that PioAB, which shares homology with MtrAB, is essential for iron oxidation during photoferrotrophy (7–9). Furthermore, MtoAB, another MtrAB homolog, has been implicated in Fe(II) oxidation in S. lithotrophicus ES-1. Purified MtoA was shown to oxidize soluble Fe(II) in vitro (10), and proteomic analyses demonstrated that MtoAB is specifically expressed during growth on solid Fe(II) minerals, including smectite and magnetite, but not on Fe(II)-citrate (11–13). The MtoAB pathway is thought to be preferentially used for the oxidation of solid, reduced forms of iron compared with soluble Fe²⁺, which is favored by the Cyc2 pathway characterized in the acidophile Acidithiobacillus ferrooxidans and demonstrated in ES-1 (11, 14, 15). A homolog of Cyc2 is the predominant outer-membrane monoheme c-type cytochrome in the marine neutrophilic iron oxidizer Mariprofundus ferrooxydans PV-1 (15, 16). While these organisms may oxidize soluble iron, the resulting insoluble product is often coordinated into remarkable mineral forms, such as the twisted ribbon stalks produced by PV-1 (17). As these interactions center on iron, there is a gap in knowledge for EEU mechanisms in other chemolitho(auto)trophs, such as sulfur oxidizers.

Thioclava electrotropha is a member of the Alphaproteobacteria family Rhodobacteraceae, originally isolated from marine sediments from Catalina Harbor, California, following enrichment on electrodes poised at cathodic potentials (–400 mV vs Ag/AgCl) (18). This organism exhibits remarkable metabolic versatility, capable of both chemolithoautotrophic growth using reduced sulfur compounds or hydrogen, and chemoorganotrophy utilizing a variety of organic carbon sources coupled to oxygen or nitrate reduction (19, 20). T. electrotropha is also capable of oxidizing a cathode poised at an environmentally and physiologically relevant electron-donating potential (18, 19, 21). Notably, this organism engages in interfacial extracellular electron transfer through an unknown mechanism that is independent of hydrogen or formate use (21). Though the genome lacks homologs to known interfacial extracellular electron transfer pathway proteins, it contains 39 genes encoding proteins with canonical “CXXCH” c-type heme-binding motifs (20). This enrichment in putative c-type cytochromes has been observed broadly in electrochemically active microbes (22, 23). Recently, a high-throughput transposon mutagenesis screen identified 54 genes required for growth utilizing a cathode as the sole electron donor in T. electrotropha (20). None of these 54 genes encode proteins with canonical c-type heme-binding motifs. Given the central role of c-type cytochromes in characterized EET pathways, this result raises the question whether the organism contains redundant systems for EEU or whether T. electrotropha utilizes a cytochrome-independent mechanism for EEU. To address this question, we used transcriptomics to identify differentially expressed genes during cathode oxidation, reasoning that genes involved in EEU would be upregulated under cathodic conditions.

Here, we present a transcriptomic, genetic, and electrochemical investigation into the extracellular electron uptake pathway employed by Thioclava electrotropha. This work presents evidence for a novel pathway of extracellular electron uptake that includes a previously uncharacterized putatively periplasmic monoheme c-type cytochrome gene (AKL02_08760), designated pmcA. These results provide new insights into the diversity of strategies employed by electroactive microorganisms and advances our understanding of the genetic determinants for microbe-electrode interactions.

MATERIALS AND METHODS

Bacterial strains and culture conditions

All bacterial strains utilized in this work are listed in Table 1. All strains were stored in 40% v/v glycerol at −80°C, unless otherwise stated. Prior to each experiment, Thioclava electrotropha ElOx9^T strains were grown on lysogeny broth plus salt and ions (LBS + Ions) agar plates and incubated at 30°C (25 g/L LB [Fisher Bioreagents], 10 g/L NaCl, 3 g/L MgCl₂*6H₂O, 0.15 g/L CaCl₂*2H₂O, with 15 g/L agar added for solid media). Single colonies were used to inoculate LBS + Ions broth or artificial low-sulfate saltwater base (SWB) (18) amended with 5 mM sodium acetate (200 RPM). For chemolithoautotrophic growth with reduced sulfur species as the electron donors, SWB media were amended with anaerobic polysulfide stock to a final concentration of 22.5 mM and allowed to oxidize overnight in a shaking incubator (200 RPM). The anaerobic polysulfide stock was prepared, as previously described (24). Abiotic oxidation of polysulfide at near-neutral pH produces primarily thiosulfate (S₂O₃²⁻) and elemental sulfur (S⁰), the latter of which precipitates as a white particulate suspension (25). Both S₂O₃²⁻ and S⁰ served as potential electron donors in chemolithoautotrophic growth experiments. The pH of the media was adjusted to 7.2 by adding sterile 1 mM phosphate buffer. Escherichia coli strain WM3064, a diaminopimelic acid (DAP) auxotroph, was used as the donor strain in conjugation experiments. WM3064 was grown in lysogeny broth (LB) supplemented with 300 µM DAP. Kanamycin (50 mg/L) was added to media to maintain plasmids in T. electrotropha and E. coli WM3064 or to select for transconjugants. Bioelectrochemical experiments were carried out in SWB cathode medium comprised of artificial low-sulfate saltwater base lacking organic carbon additions, vitamins, and minerals (18). Unless otherwise specified, all cultures were incubated at 30°C.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Description or genotype	Source or reference
Thioclava electrotropha ElOx9T strains
ARR065	Electrochemically enriched and subsequently isolated from coastal sediments (Catalina Island, CA), wild type	(18, 19)
ARR288	∆soxCD	This study
ARR322	∆pmcA	This study
ARR324	∆14455	This study
ARR333	ARR065 with pBBR1MCS-2 empty vector	This study
ARR334	ARR322 with ppmcA+	This study
E. coli strain
WM3064	Donor strain for conjugation; thrB1004 pro thi rpsL hsdS lacZ∆M15 RP4-1360 ∆(araBAD)567 ∆dapA1341::[erm pir]	W. Metcalf, University of Illinois, Urbana
Plasmids/vectors
pSMV3	9.5-kb mobilizable suicide vector; oriR6K, sacB, Kmr	(26)
pSMV3∆soxCD	2-kb fusion PCR fragment containing ∆soxCD cloned into the SacI and BamHI sites of pSMV3; used to make the ARR288 ∆soxCD strain	This study
pSMV3∆pmcA	2-kb fusion PCR fragment containing ∆pmcA cloned into the SacI and BamHI sites of pSMV3; used to make the ARR322 ∆pmcA strain	This study
pSMV3∆14455	2-kb fusion PCR fragment containing ∆14455 cloned into the SacI and BamHI sites of pSMV3; used to make the ARR324 ∆14455 strain	This study
pBBR1MCS-2	5.1-kb broad-host range plasmid: Kmr, lacZ	(27)
ppmcA+	pmcA PCR fragment cloned into the BamHI and SacI sites of pBBR1MCS-2	This study

Transcriptomics experimental design

We evaluated the transcriptional profiles of Thioclava electrotropha ElOx9^T under four growth conditions: aerobic heterotrophic growth, chemolithotrophic growth with reduced sulfur species, early electron uptake (24 h), and sustained electron uptake conditions (~6 days).

For aerobic heterotrophic growth, cells were grown in LBS + Ions broth overnight in triplicate (~18 h). Then, 500 μL of each replicate culture was combined with 1,000 μL of RNAProtect Bacteria Reagent (Qiagen) in a 1.5-mL tube, the tube was vortexed for 5 s, incubated at room temperature for 5 min, and cells were pelleted via centrifugation (10,000 × g, 10 min). The supernatant was decanted, and cell pellets were immediately frozen at −80°C until RNA purification.

For chemolithoautotrophic growth, cells were grown overnight in LBS + Ions, pelleted via centrifugation (8,000 × g, 5 min), and resuspended in triplicate flasks containing SWB medium with oxidized polysulfide to a final OD₆₀₀ = 0.2. The pH of flasks was monitored hourly to prevent cell death due to acidification resulting from complete S₂O₃²⁻ or S⁰ oxidation. Once flask pH reached ~5.0 (approximately 4 h), 500 μL of each replicate culture was collected and processed with RNAProtect Bacteria Reagent as above. Cell pellets were immediately frozen at −80°C until RNA purification.

For the early extracellular electron uptake condition, both planktonic and electrode-attached cells were collected and preserved for RNA sequencing. Bioelectrochemical cells were set up as described below. To obtain enough biomass for RNA isolation, each replicate was comprised of 6 individual bioelectrochemical cells pooled together; that is, 18 reactors were required to obtain enough biomass for triplicates of this condition. Reactors were individually disconnected from the potentiostat to minimize the possibility of gene expression changes under open circuit conditions during sample processing. Then, 500 μL of the planktonic fraction from each reactor was collected, pooled (six reactors per replicate), and processed with the RNAProtect Bacteria reagent as above while maintaining 1:2 sample:RNAProtect Bacteria reagent volume ratios. To collect cathode-attached biofilms, the remaining planktonic fraction in bioelectrochemical cells was removed by decanting and/or pipetting. Then, 2 mL of RNAProtect Bacteria reagent was immediately added to the electrode surface, and the surface-attached biofilm was scraped using a sterile cell scraper. Biofilm samples were pooled (six reactors per replicate, same provenance as planktonic fractions) and processed as above. Cell pellets were immediately frozen at −80°C until RNA purification

Finally, for the sustained extracellular electron uptake condition, triplicate samples, each comprised of six individual replicate bioelectrochemical cells, were processed as the early EEU condition, and cell pellets were immediately frozen following processing at −80°C until RNA purification.

RNA purification, sequencing, and data analysis

RNA was isolated from frozen, RNAProtect-treated cell pellets following the Enzymatic Lysis and Proteinase K Digestion of Bacteria protocol (Protocol 4) located in the RNAProtect Bacteria Reagent (Qiagen) handbook. RNA was then purified with the RNeasy Mini Kit (Qiagen), with the addition of ß-mercaptoethanol to Buffer RLT immediately before use and inclusion of the optional on-column DNase treatment. RNA concentrations were measured for all extractions using a Qubit fluorometer and Qubit RNA HS Assay Kit (Thermo Fisher Scientific). RNA extracts were frozen at −80°C.

Frozen RNA extracts were shipped on dry ice to SeqCenter (Pittsburgh, PA) for cDNA synthesis, library preparation, and RNA sequencing. At SeqCenter, samples were treated with RNase-free DNase (Invitrogen), RNA was converted to cDNA, and sequencing libraries were prepared with Illumina’s Stranded Total RNA Prep Ligation with Ribo-Zero Plus kit with 10 bp unique dual indices. Sequencing was carried out on a NovaSeq X Plus sequencer, which produced 2 × 151 bp reads. Demultiplexing, quality control, and adapter trimming were performed with bcl-convert v4.1.5.

Sequencing read quality was assessed with FastQC 0.12.1 (28). Reads were then quality filtered using Trimmomatic 0.39 (29), with the following options: LEADING:20 TRAILING:20 HEADCROP:10 SLIDINGWINDOW:4:20 MINLEN:20. Trimmed reads were mapped to the Thioclava electrotropha ElOx9^T reference genome (NCBI Accession no. GCF_002085925.2) with bowtie2 2.5.1 (30) in sensitive mode. SAMtools 1.19.2 was used for sam/bam file conversions, sorting, and indexing (31). Mapped reads were then quantified to estimate RNA abundance using the featureCounts command in the Subread package version 2.0.6 (32) with arguments -p and -s 2. Transcriptomic sequencing statistics and read mapping rates are presented in Table S1. Resultant count matrices were then imported into R Studio for normalization and differential expression analysis (33) with DESeq2 version 1.42.1 (34). In pairwise comparisons between experimental conditions, genes with adjusted P-values (Benjamini-Hochberg method) ≤ 0.01 and |log₂FC| ≥ 1 were considered differentially expressed. Transcriptomic data were visualized in R using packages ggplot2 version 3.5.1 (35) and ComplexHeatmap version 2.18.0 (36). For additional context, the genome was queried for putative c-type cytochromes (those genes containing putative c-type heme-binding motifs (CX_2-4CH)) and were identified using Prosite (37). PSORTb 3.0.3 was used to predict the subcellular localization of all protein-coding genes in the genome (38).

Reactor setup and bioelectrochemical measurements

Bioelectrochemical experiments were performed in standard three-electrode bioelectrochemical cells as previously described (18). The reactor was comprised of a 10.68 cm² indium-tin oxide coated glass electrode (Delta Technologies), a platinum wire counter electrode (Sigma-Aldrich), and an Ag/AgCl (1M KCl) reference electrode (CH Instruments). Reactors were filled with 20 mL of cell resuspensions and bubbled with 0.2 μm-filtered room air to provide oxygen as the terminal electron acceptor. Amperometric and voltammetric measurements were made with either a 16-channel MPG-2 potentiostat (BioLogic) or a 4-channel Squidstat Prime potentiostat (Admiral Instruments).

For transcriptomic experiments, wild-type Thioclava electrotropha ElOx9^T cells were grown overnight in LBS + Ions broth, pelleted via centrifugation (8,000 × g, 5 min), and resuspended in SWB cathode medium (SWB lacking organic carbon, vitamin, and mineral solutions) to a final OD₆₀₀ of 0.2. The cell resuspension was transferred to reactors (20 mL/reactor), and reactors were connected to the potentiostat. For the early extracellular electron uptake condition, the working electrode was poised at an electron-donating potential of −278 mV vs SHE for 24 h. A cathodic potential of −278 mV vs SHE was chosen as this is just below the calculated redox potential for S⁰ oxidation in seawater at pH 8 but well above the potential necessary for H₂ evolution to prevent H₂ from serving as an electron donor to the cells (18, 19). For the sustained extracellular electron uptake condition, the working electrode was poised at −278 mV vs SHE for ~6 days, and current was measured (chronoamperometry [CA]). After 60 h of chronoamperometry, cyclic voltammetry (CV) was performed to assess reactor health. The working potential was cycled between −378 and 422 mV vs SHE three times at a scan rate of 1 mV/s for the Biologic potentiostat or 2.4 mV/s for the Squidstat Prime potentiostat.

For experiments determining the capacity for various Thioclava electrotropha strains to oxidize a cathode poised at an electron-donating potential, strains were grown overnight in SWB + Acetate medium. Kanamycin (50 mg/L) was added to the medium for plasmid maintenance in strains harboring pBBR1MCS-2 vectors. Cell-free bioelectrochemical cells containing sterile SWB cathode medium, and kanamycin as required, were set up as described above. CA was performed as above until current measurements stabilized, typically 30 min. Linear sweep voltammetry (LSV) was then performed by scanning the working potential from +422 mV to −578 mV vs SHE at a scan rate of 1 mV/s. Three sweeps were performed. The abiotic LSVs were used for background subtraction. After abiotic LSVs were complete and current stabilized, cells from overnight cultures were pelleted via centrifugation (8,000 × g, 5 min) and resuspended in the bioelectrochemical cells to a final OD₆₀₀ of 0.2. CA was performed for approximately 6 days followed by LSV.

To determine the amount of current consumption that was due to biological processes, NaCN was added to a final concentration of 10 mM at the end of CA experiments. We define biological current as the difference in average current consumption in the 30 min preceding NaCN addition and average current consumption in the 30 min following NaCN addition. A minimum of four replicates were performed for each strain. Student’s t-tests were used to determine if the average biological currents of deletion and complementation strains differed from wild type.

Construction of gene deletion and complementation mutants

T. electrotropha in-frame gene deletion strains ARR288, ARR322, and ARR324 were constructed following Saltikov and Newman (26) with modifications to improve merodiploid resolution rates. Briefly, 1-kb regions flanking the target gene(s) were PCR amplified using the primers listed in Table S2 using NEB Q5 high-fidelity DNA polymerase. PCR amplicons were size selected via agarose gel electrophoresis (0.8% w/v agarose) and gel extracted using the NEB Monarch gel extraction kit. The pSMV3 vector was isolated from overnight cultures of E. coli WM3064 using the NEB Monarch plasmid miniprep kit, digested with SacI-HF and BamHI-HF (NEB), size selected, and gel extracted. DNA concentrations of PCR fragments and digested, cleaned pSMV3 vector backbone were quantified with a Qubit fluorometer and 1× dsDNA HS assay kit (Invitrogen) and assembled with NEB HiFi DNA Assembly kit per the manufacturer’s guidelines. DNA assembly reactions were then processed through the NEB Monarch gel extraction kit directly to remove reaction components and other inhibitors that inhibit transformation. Approximately 10 ng of the cleaned assembly was transformed into E. coli WM3064 via electroporation (1-mm cuvette gap, 2.1 kV). Transformants were verified for correct assembly by colony PCR with M13 primers by determining insert length. Confirmed plasmids were transferred to T. electrotropha by conjugation (1:1 donor to recipient ratio, 16 h mating on an LBS + Ions + DAP plate, 30°C). Transconjugants containing the integrated mutagenesis vector were selected for on LBS + Ions agar plates supplemented with 50 mg/L kanamycin. Transconjugants were then streaked onto SWB plates containing decreased NaCl (0.5% NaCl versus 2% in standard SWB) and grown heterotrophically with 10% w/v sucrose as the sole carbon source at 22°C for 7–10 days, or until colonies arose. Colonies were then replica plated onto LBS + Ions agar and LBS + Ions + kanamycin agar. Kanamycin-sensitive colonies were then screened by PCR for their respective gene deletions. Gene deletion mutants were then verified via whole-genome sequencing as described below.

Complementation plasmids for each gene deletion were constructed by cloning the ORF(s) in the direction of the lac promoter in the broad host complementation vector pBBR1MCS-2 (27). pBBR1MCS-2 was isolated from overnight cultures of E. coli WM3064, as described above. PCR was used to generate gene fragments containing the deleted gene(s) and flanked with BamHI and SacI sites on the 5′ and 3′ ends, respectively. Isolated vector and PCR products were restriction digested with BamHI-HF and SacI-HF (NEB) and subsequently cleaned and concentrated with the NEB Monarch gel extraction kit. PCR products were then ligated into the digested vector using T4 DNA Ligase (NEB). Ligations were cleaned and concentrated using the NEB Monarch gel extraction kit and transformed into WM3064 via electroporation as described above. Transformants were selected for on LBS + Ions + DAP + 50 mg/L kanamycin plates. Successful ligations were verified by PCR using M13 primers. The verified complementation vectors were conjugated into their respective T. electrotropha deletion strain and transconjugants were selected by plating onto LBS + Ions + 50 mg/L Kanamycin agar plates.

Whole genome sequencing and analysis

Deletion strains were subjected to whole genome sequencing to verify successful gene deletion. Each deletion strain was grown in LBS + Ions broth overnight. DNA was extracted from overnight cultures using the Qiagen DNeasy UltraClean Microbial Kit according to the manufacturer’s instructions. DNA extracts were sent to SeqCenter (Pittsburgh, PA) for library preparation and sequencing. Briefly, sequencing libraries were prepared using the tagmentation-based and PCR-based Illumina DNA Prep kit using custom 10 bp unique dual indexes and selecting a target insert size of 280 bp. No additional DNA fragmentation or size selection steps were performed. The sequencing libraries were sequenced on an Illumina NovaSeq X Plus sequencer, which produced 2 × 151 bp paired-end reads. Demultiplexing and initial quality control and adapter trimming was performed with bcl-convert v4.2.4 (Illumina).

Sequencing read quality was assessed with FastQC 0.12.1 (28). Reads were then quality filtered using Trimmomatic 0.39 (29) with the following options: LEADING:20 TRAILING:20 HEADCROP:10 SLIDINGWINDOW:4:20 MINLEN:100. Trimmed reads were mapped to the Thioclava electrotropha ElOx9^T reference genome (NCBI Accession no. GCF_002085925.2) with bowtie2 2.5.1 (30) in sensitive mode. Sequencing statistics are presented in Table S3. SAMtools 1.19.2 was used for sam/bam file conversions, sorting, and indexing (31). Sorted and indexed BAM files were visualized against the reference genome using the Integrative Genomics Viewer web application available at igv.org (39). Plots displaying genome annotations and read depth coverage were created using a custom JavaScript-based interactive tool developed with React and HTML5 Canvas (Fig. S1).

Ion chromatography for anion (sulfate) analysis

Medium samples from the abiotic control, triplicate wild-type sulfur oxidation transcriptome samples, and triplicate ∆soxCD (ARR288) sulfur oxidation assays were collected to determine sulfate concentrations via ion chromatography. One milliliter of media from triplicate cultures was collected and passed through 0.2-μm polyethersulfone filters at the time of inoculation (t = 0 h) and 4 h post-inoculation (t = 4 h). An additional sampling was conducted 24 h post-inoculation for ∆soxCD replicates. Samples were diluted 1:5 in MilliQ water and analyzed with a Dionex Aquion Ion Chromatograph (Thermo Fisher Scientific).

Phylogenetic analysis and codon usage bias analysis

Homologs to PmcA were identified by querying the NCBI blastp nonredundant database (search date: 23 October 2025, Tables S6 and S7). Two phylogenetic trees were constructed: one containing ~40 manually selected representatives from broad phylogenetically distinct lineages and one containing the top ~50 homologs. All sequences used in phylogenetic analyses had >50% amino acid identity and >69% (100 amino acids minimum) query coverage to PmcA. Amino acid sequences were aligned with MUSCLE as implemented in EMBL EBI’s Job Dispatcher (40). RAxML v. 8.2.12 was used to infer phylogenetic relationships using the maximum likelihood method and PROTGAMMAWAG amino acid substitution model. Bootstrap values were calculated based on 100 trees generated. The resulting phylogenetic trees were visualized and annotated with the Interactive Tree of Life (iTOL) program (41).

We evaluated codon usage bias for genes pmcA (AKL02_08760), AKL02_08765, AKL02_08770, and AKL02_08775, which corresponds to T. electrotropha’s putative cytochrome c-multicopper oxidase operon. Relative synonymous codon usage (RSCU) values for each gene and for the genome were calculated using the following formula:

$${\displaystyle {\displaystyle \mathrm{R}\mathrm{S}\mathrm{C}\mathrm{U}=\frac{\mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d}\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}\mathrm{o}\mathrm{f}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{o}\mathrm{n}}{\mathrm{e}\mathrm{x}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{e}\mathrm{d}\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}\mathrm{o}\mathrm{f}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{o}\mathrm{n}\mathrm{u}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{r}\mathrm{u}\mathrm{n}\mathrm{i}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{u}\mathrm{s}\mathrm{a}\mathrm{g}\mathrm{e}}}}$$

The observed frequency of a codon is defined as follows:

$${\displaystyle {\displaystyle \mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d}\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}=\frac{\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}\mathrm{s}\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{o}\mathrm{n}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}}{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{o}\mathrm{a}\mathrm{c}\mathrm{i}\mathrm{d}}}}$$

The expected frequency of a codon under uniform usage is defined as follows:

$${\displaystyle {\displaystyle \mathrm{e}\mathrm{x}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{e}\mathrm{d}\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}=\frac{1}{\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{s}\mathrm{y}\mathrm{n}\mathrm{o}\mathrm{n}\mathrm{y}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{c}\mathrm{o}\mathrm{d}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{h}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{o}\mathrm{a}\mathrm{c}\mathrm{i}\mathrm{d}}}}$$

For each gene, coding sequences were parsed into non-overlapping triplets starting from the start codon (ATG). Codons were grouped by their corresponding amino acids according to the standard genetic code. For each amino acid with multiple synonymous codons, the observed usage of each codon was compared with the expected usage under random selection.

RESULTS AND DISCUSSION

Gene expression profiles differ based on electron source

A transcriptomic experiment was conducted to evaluate T. electrotropha gene expression under three conditions, each utilizing a different electron source (Fig. 1A). The aerobic heterotrophic pre-growth condition utilized acetate as an organic electron donor, whereas the two chemolithotrophic conditions utilized inorganic electron donors: either reduced sulfur species or an indium-tin oxide electrode poised as a cathode. Transcriptome profiles from each condition formed discrete experimental clusters (Fig. 1A). The two chemolithotrophic conditions were chosen as they utilize two different inorganic electron donors, though they create contrasting chemical microenvironments. Complete oxidation of thiosulfate or elemental sulfur produces sulfuric acid as a byproduct, resulting in acidic conditions (42, 43). In cathode systems, electron uptake likely consumes protons, as one would expect for electrochemical oxygen reduction that generates water, resulting in a localized alkaline microenvironment around the electrode (44, 45). Given the environmental differences between these two conditions, it is not surprising that the overall expression patterns differ distinctly (Fig. 1A).

Fig 1.

Transcriptome analysis. (A) Principal component analysis of RNA-seq data for all biological replicates reveals distinct gene expression profiles for the aerobic heterotrophic pre-growth condition (gray), sulfur oxidation condition (red), and the four EEU sample types (early EEU biofilm – yellow, early EEU planktonic fraction – green, sustained EEU biofilm – blue, sustained EEU planktonic fraction – orange). (B) UpSet plot showing the distribution of differentially expressed genes (DEGs) unique to and shared among the sulfur oxidation condition and all EEU conditions collectively, with respect to the pre-growth condition. Upregulated genes are in red, and downregulated genes are in blue (log₂FoldChange ≥1 or ≤1, respectively, P_adj ≤ 0.01).

Differential expression analysis, comparing all transcriptome profiles against the aerobic heterotrophic pre-growth condition, identified more than one thousand significantly differentially expressed genes (DEGs, log₂FC ≥ |1|, P_adj ≤ 0.01) in each condition (Table S4; Fig. S2). In line with the clustering patterns observed in the PCA analysis, 584 genes were differentially expressed under all EEU conditions but were not differentially expressed in the sulfur oxidation condition (Fig. 1B). Conversely, 559 DEGs were unique to the sulfur oxidation condition, which highlights the unique transcriptional profiles of these two lithotrophic conditions. These results may suggest that cells employ different mechanisms or strategies for the oxidation of these two inorganic electron donors and/or different strategies for dealing with acid production resulting from sulfur oxidation compared with the alkaline and reducing environment at the cathode. Supporting this is the fact that most of the sox sulfur oxidation pathway was upregulated during sulfur oxidation, but only the periplasmic c-type cytochrome encoded by soxX was upregulated in cathode-oxidizing conditions (all comparisons with respect to pre-growth, Table S4). All other genes in the sox pathway were not differentially expressed during cathode oxidation with respect to pre-growth. Previous whole-genome mutational analysis in Thioclava electrotropha found that the sox sulfur oxidation pathway was not required for cathode oxidation (20), providing further evidence in support of a unique metabolic pathway for cathode oxidation independent of the sox system.

All cathode oxidation samples clustered together regardless of experiment duration (early vs sustained) and cell fraction (biofilm vs planktonic cells) (Fig. 1A). The early EEU condition refers to bioelectrochemical cells that were operated for 24 h. The sustained EEU condition refers to bioelectrochemical cells that were operated for approximately 6 days, in line with previous experiments (20). Biofilm samples were taken from biomass attached to the electrodes as opposed to the cells collected from the media suspension. Differences observed in these sample groups were intended to inform gene expression differences that result from transitioning to cathodic conditions, forming a biofilm on the electrode, and/or sustaining these activities.

Overall, we observed modest differences in gene expression profiles across experiment durations and cell fractions, particularly regarding the early EEU planktonic fraction. This sample had the highest number of unique DEGs (399 genes differentially expressed only in this condition, Fig. 1B). Compared with all other cathode samples, the early EEU planktonic fraction had more upregulated genes for transcription (COG category K, 111 genes vs 66–79) and replication, recombination, and repair (L, 90 vs 62–73) (Fig. S3A). The early EEU planktonic fraction exhibited more downregulated genes for translation (J, 119 vs 36–40), energy production (C, 102 vs 68–74), and amino acid transport and metabolism (E, 115 vs 53–76) compared with all other cathode samples (Fig. S3B). Given that planktonic T. electrotropha cells are not known to directly oxidize a cathode (21), differences in early EEU biofilm and planktonic gene expression patterns are expected as cells transition to biofilm formation and utilization of a solid-phase electron source. Notably, these differences diminished during sustained EEU, whereby the biofilm and planktonic gene expression profiles were nearly indistinguishable. This potentially indicates that either planktonic cells may be participating in EEU with the cathode-attached biofilms or that biofilms are releasing daughter cells into the planktonic phase. These differences in gene expression patterns during early EEU may inform early adaptations to cathodic conditions; however, given the tight clustering of transcriptome profiles of all cathode samples and low baseMean and/or log₂FC values for the unique DEGs in the early EEU planktonic sample, these gene expression differences are modest relative to the overall similarity in cathodic transcriptional profiles. Nonetheless, these early EEU samples provide an additional time point in our data set to investigate gene expression dynamics during EEU.

Putative c-type cytochrome gene expression is upregulated during cathode oxidation

In bacteria that engage in interfacial extracellular electron transfer, c-type cytochromes, particularly those containing multiple heme cofactors, make up the majority of characterized proteins involved in these processes (3, 46, 47). These proteins typically span the inner to outer membrane in Gram-negative bacteria (3, 46). As Thioclava electrotropha is a Gram-negative bacterium that has been shown to require direct contact with a cathode for EEU (21), we searched the list of DEGs (relative to aerobic heterotrophic pre-growth) for upregulated genes that are predicted to localize to the cellular envelope and are predicted to contain c-type heme binding motifs (Fig. 2; Table S5). To be broadly inclusive of putative cytochromes, genes containing both canonical (CXXCH) and non-canonical (CX_3-4CH) c-type heme binding motifs were included as studies have shown that cytochromes containing atypical c-type heme-binding motifs can facilitate EET (48–51).

Fig 2.

Putative c-type cytochromes are upregulated during EEU. Heatmap of DEGs containing putative c-type heme-binding motifs predicted to localize to the cellular envelope or with unknown localizations. The log₂FC values for all genes under all lithotrophic conditions with respect to the pre-growth condition are indicated by cell color, with upregulated genes shown in red and downregulated genes shown in blue. Cells containing an asterisk (*) were statistically significantly differentially regulated (log₂FoldChange ≥1 or ≤1, P_adj ≤ 0.01). GenBank locus tag numbers and annotations are listed for all genes. The two genes selected for bioelectrochemical investigation, pmcA and hypothetical protein AKL02_14455, are shown in bold. Further details can be found in Table S5.

Gene expression patterns for genes that fit these criteria were remarkably similar between all EEU conditions regardless of time point or cell fraction (Fig. 2). In biofilms from the sustained EEU condition, we identified thirteen genes predicted to contain c-type heme-binding motifs that were significantly upregulated and predicted to localize to the cellular envelope, most of which were also upregulated in all other EEU samples. Among all upregulated genes, AKL02_08760 demonstrated the highest fold-change in expression, which increased over 100-fold with respect to pre-growth. Based on RefSeq annotations, this gene encodes a 144 amino acid c-type cytochrome that is predicted by pSORTb to localize to the periplasm. Sequence analysis indicates this gene contains a single CXXCH heme-binding motif. Hereafter, we designate AKL02_08760 as pmcA for periplasmic monoheme c-type cytochrome. A second gene was highly upregulated in all EEU conditions was AKL02_14455, a hypothetical protein containing a single CXXCH heme-binding motif that is also predicted to localize to the periplasm. This gene was upregulated ~8-fold in early EEU biofilms and was upregulated ~30-fold in sustained EEU biofilms with respect to pre-growth. Although the potential role of these genes in EEU is unclear based on sequence analysis alone, the fact that they are two of the most upregulated genes under EEU conditions, they are predicted to localize to the cell envelope, and each contains a putative c-type heme-binding motif encouraged us to further investigate the role of these proteins in EEU.

Development of a genetic system in Thioclava electrotropha

To investigate the role of these highly upregulated c-type heme-binding motif-containing genes in EEU, we first established a genetic deletion system in Thioclava electrotropha. While we have previously demonstrated genetic tractability in this organism (20), clean gene deletion mutants had not yet been constructed. We employed the pSMV3 allelic exchange vector—a suicide plasmid commonly used to generate scarless deletions in other electroactive organisms, such as Shewanella species (26). However, we discovered that successful sucrose counterselection was only achievable when T. electrotropha was grown heterotrophically in low-salt (0.5% NaCl) medium lacking organic carbon sources other than 10% sucrose, and when grown below the organism’s optimal growth temperature at ≤22°C. Cells consistently evaded counterselection in both complex media (LBS + Ions + 10% sucrose) and defined media (SWB + acetate + 10% sucrose), even when NaCl was adjusted to 0.5% and cells were grown at ≤22°C. These requirements are consistent with established constraints on SacB counterselection: elevated sodium chloride concentrations have been shown to lower SacB activity and/or attenuate sucrose toxicity, and decreased temperatures are thought to improve SacB stability in some bacteria (52). However, the necessity to remove alternative carbon sources is poorly understood. We speculate that access to preferred carbon substrates may decrease sucrose uptake and/or catabolism, potentially limiting the production and accumulation of levan to toxic levels, although this was not directly tested.

We first tested this gene deletion system by targeting part of T. electrotropha’s sulfur oxidation system: soxCD. SoxCD is a sulfur dehydrogenase that catalyzes the six-electron oxidation from protein-bound sulfane to sulfone (42). Without SoxCD, complete oxidation of S₂O₃²⁻ and/or S⁰ would be blocked. This target was selected because the deletion phenotype could be readily screened for in the laboratory, providing a clear and practical phenotypic marker to validate our gene deletion system. Successful deletion of soxCD was verified by whole-genome sequencing and alignment to the Thioclava electrotropha reference genome (Fig. S1A). Phenotypic screens indicated our scarless ∆soxCD deletion mutant was unable to fully oxidize S₂O₃²⁻ and/or S⁰ (sulfate concentrations and pH of the medium remained unchanged after 24 h of growth in SWB containing reduced sulfur species), whereas we measured a decrease in pH from 7.2 to ~4 and an increase in sulfate concentrations from 2.0 ± 0.0 mM to 10.6 ± 0.5 mM after just four hours of growth of wild-type cells (data not shown). It is unclear whether the failure of the ∆soxCD mutant to produce sulfate reflects a block in S⁰ oxidation, S₂O₃²⁻ oxidation, or both. Nevertheless, this proof-of-concept demonstrated that our genetic system could generate scarless deletions in T. electrotropha, and we therefore applied this validated approach to construct deletion mutants of the highly upregulated c-type heme-binding motif-containing genes previously highlighted (Table 2).

TABLE 2.

Characteristics of genes targeted for deletion and subsequent bioelectrochemical characterization^a

	soxCD (AKL02_13800-13805)	pmcA (AKL02_08760)	AKL02_14455
RefSeq annotation	c-type cytochrome soxD; sulfite dehydrogenase soxC	Cytochrome c	Hypothetical protein
Predicted localization (score)	Periplasmic (10); unknown (4.9)	Periplasmic (9.83)	Periplasmic (9.84)
Heme-binding motifs	CAACH, CKACH; n.d.	CAACH	CHDCH
COG category	C; S	C	G
KEGG Ontology	K22622; K17225	n.a.c	K03431
Log2FC in Biofilm 24h	0.03; 0.27	7.10b	2.96
Log2FC in Planktonic 24h	−1.29; −1.10	6.45	2.66
Log2FC in Biofilm 6d	−0.26; −0.12	6.68	4.92
Log2FC in Planktonic 6d	−0.90; −0.90	6.62	3.98
Log2FC in S Oxidation	1.40; 0.98	0.56	2.20

^a Log₂FC values are with respect to pre-growth.

^b Bolded log₂FC values have P_adj values ≤ 0.01.

^c n.d., none detected; n.a., none assigned.

Electrochemical measurements confirm role of PmcA, a putative periplasmic c-type cytochrome, in EEU

We constructed scarless in-frame gene deletion mutants for pmcA (Fig. S1B) and AKL02_14455 (Fig. S1C) and tested these mutants electrochemically compared with wild type (Fig. 3). To quantify the amount of current consumption that was due to biological processes in both the mutants and wild-type strains, cyanide was added to reactors to measure abiotic current. We refer to the difference in current consumption pre- and post-cyanide addition as biological current (Fig. S4). Each mutant was tested in at least four replicate experiments.

Fig 3.

Biological current measurements and catalytic activity for all strains. (A) Biological current was determined for all strains (n ≥ 4, all strains) after approximately 6 days of continuous chronoamperometry at −278 mV vs SHE (see Fig. S4 for chronoamperometry plots). Biological currents that were significantly different from wild type (P < 0.05) are indicated with an asterisk (*). (B–E) Representative background-subtracted linear sweep voltammograms (LSVs) of all deletion and complementation strains compared with wild type. (B) WT vs ∆pmcA vs pmcA⁺ LSVs show a decrease in catalysis in the deletion mutant that is rescued upon complementation, outperforming the catalytic activity of wild type. (C) WT vs ∆soxCD LSV shows decreased catalysis in the deletion mutant. (D) WT vs ∆14455 LSVs show no difference in catalysis between the two strains. (E) WT vs pBBR1MCS-2 empty vector LSV shows no difference in catalysis between the two strains.

In the pmcA deletion strain, a statistically significant decrease in biological current consumption was measured compared with wild type (Fig. 3A). Linear sweep voltammetry (LSV) was then performed to investigate the electrochemical behavior of each strain, by showing the onset potential of electron uptake for each mutant and the degree of change in electron uptake capacity with increasing voltage (metrics of catalytic activity). Background-subtracted LSV measurements showed that capacity for catalysis was nearly abolished in the ∆pmcA strain (Fig. 3B). Interestingly, upon complementation, we measured a statistically significant increase in current consumption (Fig. 3A) and increased catalysis in LSV analysis (Fig. 3B). The enhanced EEU capacity observed upon complementation, exceeding wild-type levels, suggests that altering pmcA expression results in enhanced extracellular electron transfer. These findings warrant further investigation into if and how protein abundance is altered under these conditions, and whether modulating expression of pmcA affects EEU rates in T. electrotropha.

We also measured a modest decrease in electrode oxidation along with decreased catalysis via LSV in the ∆soxCD mutant (Fig. 3A and C). While this decrease in biologic current consumption was not statistically significant, it contrasts with observations in S. oneidensis, where deletion of periplasmic cytochromes not directly involved in EET led to increased EET rates due to reduced competition for electrons (53). Whether this reflects a minor contribution of SoxCD to EEU, decreased fitness of the mutant, or a change in the architecture of the periplasm that impacts EEU negatively remains unclear. Given that soxCD was not differentially expressed with respect to pre-growth and the EEU phenotype of the ∆soxCD mutant, ∆soxCD was not investigated further. Similarly, we measured no decrease in electrode oxidation capacity or catalysis for the AKL02_14455 deletion strain (Fig. 3A and D). Although this gene was upregulated during EEU, bioelectrochemical measurements indicated that this gene is unlikely to be directly involved in extracellular electron uptake under the experimental conditions tested.

PmcA homologs are broadly distributed throughout the Alphaproteobacteria and in other bacterial phyla

Phylogenetic analysis of PmcA revealed remarkably broad conservation across diverse bacterial lineages (Fig. 4A; Tables S6 and S7). Homologs are widely distributed among the Alphaproteobacteria, other classes of Pseudomonadota, and multiple other phyla. In fact, a protein BLAST search of NCBI’s nonredundant protein database identified 2,587 homologs with >50% amino acid sequence identity and >69% coverage (100 amino acids minimum) to PmcA (Table S6). A subset of these sequences was selected to construct a phylogenetic tree showcasing the phylogenetic diversity of organisms harboring homologs to PmcA (Fig. 4; Table S7). Importantly, the region surrounding the heme-binding motif was conserved across all sequences used in phylogenetic reconstruction, suggesting functional constraint on this critical domain.

Fig 4.

PmcA phylogeny and organization of the putative cytochrome c-multicopper oxidase operon. (A) The phylogenetic tree was constructed from a manually curated representative selection of homologs to PmcA in the NCBI blastp nonredundant database (queried 23 October 2025, sequences and metadata available in Table S7). All sequences selected for inclusion had ≥50% amino acid identity and ≥69% query coverage (100 amino acids) compared with PmcA. Amino acid sequence alignments were generated using MUSCLE. A best scoring maximum likelihood tree was generated with RAxML v 8.2.12 from 100 bootstrap replicates to identify the optimal tree. The tree was annotated using the interactive tree of life tool (https://itol.embl.de). Branch thickness corresponds to bootstrap support, with thick branches exhibiting 75%–100% bootstrap support. Leaf labels include accession number and lowest taxonomic assignment. Colors in the ring surrounding the tree indicate class-level taxonomic assignments for representatives within the Pseudomonadota when possible, and phylum-level phylogeny for all others. Thioclava electrotropha is indicated by yellow highlighting. (B) Gene diagram showing the organization of the putative cytochrome c-multicopper oxidase operon in Thioclava electrotropha. The operon encodes a putative periplasmic monoheme c-type cytochrome PmcA (AKL02_08760), a multicopper oxidase family protein (AKL02_08765), and two additional putative c-type cytochromes (AKL02_08770 and AKL02_08775). Scale bar represents 500 bp.

The phylogenetic distribution of PmcA homologs is inconsistent with vertical inheritance and instead supports horizontal gene transfer. Notably, the gene shows a patchy distribution within major taxonomic groups. For example, homologs are present among some Alphaproteobacteria, but their distribution is inconsistent with their phylogeny. PmcA homologs in other EET-capable Thioclava species (T. atlantica and T. dalianensis) (19) showed lower amino acid sequence similarity to PmcA in T. electrotropha than several more phylogenetically distant genera (Table S6). Further, homologs appear sporadically across phylogenetically distant lineages of both Gram-positive and Gram-negative bacteria. This distribution pattern, where closely related species differ in gene presence while distantly related organisms share the gene, is characteristic of acquisition by horizontal gene transfer rather than vertical inheritance from a common ancestor. The maintenance of this gene across such diverse recipients suggests that it may confer a selective advantage in specific environmental niches or metabolic contexts.

Horizontal gene transfer of electron transport components, including those involved in EET, has been documented in other electroactive bacteria. For example, comparative genomic analysis revealed that key components of the mtr pathway in Shewanella show evidence of both vertical and horizontal transmission across diverse bacterial lineages (54). Similarly, the extracellular electron transfer machinery in Geobacter species shows phylogenetic patterns consistent with horizontal acquisition of cytochrome genes (23). These observations may provide additional evidence that lateral gene transfer is an important mechanism shaping the distribution of EET capabilities across bacterial lineages. The broad conservation of PmcA across 2,587 bacterial genomes provides a foundation for investigating EEU capacity in other lineages and could reveal whether this cytochrome represents a widespread mechanism for extracellular electron uptake.

Horizontally transferred genes often exhibit suboptimal expression in their new host due to incompatibility between the foreign gene’s regulatory sequences and the recipient’s transcriptional machinery, as well as codon usage bias that mismatches with the host’s tRNA pool (55). Over time, the process of amelioration, the gradual evolutionary process by which horizontally acquired genes adapt to their new genomic context, would be expected to optimize expression of proteins important for function (56). The atypical codon usage patterns observed in pmcA (discussed in detail below) suggest that this gene has potentially been recently acquired and has not fully ameliorated to T. electrotropha’s biosynthetic context, which may result in decreased transcriptional/translational efficiency and lower steady-state protein levels. Increased electron transfer rates measured in the pmcA⁺ complementation strain provide further evidence of suboptimal native expression, which may be due to incomplete amelioration of this potentially recently acquired gene. However, we cannot rule out other explanations, and further work is needed to characterize the relationship between pmcA expression and EEU activity.

Evidence for horizontal transfer of the putative cytochrome c-multicopper oxidase operon

MetaCyc analysis (57) predicted that pmcA belongs to a four-gene operon (AKL02_08760 through AKL02_08775) that encodes a cytochrome c-multicopper oxidase system (Fig. 4B). In addition to PmcA, this operon is also predicted to encode a multicopper oxidase family protein and two additional putative c-type cytochromes. Transcriptional analysis supported this operon prediction, as expression of all four genes was significantly upregulated and showed coordinated upregulation across experimental conditions specifically during EEU relative to pre-growth (Fig. 2; Table S4). However, since the pSORTb predicted localizations for putative c-type cytochromes AKL02_08770 and AKL02_08775 were unknown, these targets were excluded from further genetic and bioelectrochemical analysis in this study. Also noteworthy is a MerR transcriptional regulator located upstream of the cytochrome c-multicopper oxidase operon (AKL02_08750), which was strongly upregulated during EEU (log₂FC = 4.3–5.5 with respect to pre-growth) but also during sulfur oxidation (log₂FC = 2.3). MerR family regulators typically respond to metal ions or redox signals (58, 59), suggesting that expression of the operon may be controlled by a metal- or redox-sensitive regulatory circuit. However, as the operon was specifically upregulated only during EEU, MerR upregulation alone is insufficient to explain operon induction, suggesting that EEU-specific regulatory inputs are required, the nature of which remains an important open question.

Within this operon, the multicopper oxidase gene belongs to a protein family that has been implicated in reductive and oxidative EET in other prokaryotes. Multicopper oxidases are well-documented catalysts of Mn²⁺ and Mn³⁺ oxidation in a variety of bacteria (60–62). In Geobacter sulfurreducens, multicopper oxidases may be important for reductive EET to iron (III) oxides (63, 64). In Sideroxydans lithotrophicus ES-1, three copper-containing proteins, including two encoding multicopper oxidase motifs, were highly expressed during magnetite oxidation (13). In the archaean Metallosphaera yellowstonensis, a putative multicopper oxidase gene was highly expressed during growth on Fe(II)-ferrihydrite and pyrite, suggesting a possible role of this gene in Fe(II) oxidation (65). More recently, a recombinant FtsP/CopA-like multicopper oxidase from Georhizobium sp. MAB10 was shown to oxidize dissolved Mn²⁺ in vitro and was predicted to be involved in oxidative EET in vivo, although this has yet to be confirmed genetically (66). However, whether this putative multicopper oxidase contributes to cathode oxidation in T. electrotropha remains to be determined. The two monoheme c-type cytochromes downstream of the multicopper oxidase lack characterized homologs and do not belong to well-studied cytochrome families previously implicated in extracellular electron transfer (20), highlighting the need for further investigation into the role of this entire operon in EEU.

As with pmcA discussed in the previous section, several lines of evidence suggest that this entire operon may have been acquired through horizontal gene transfer. In T. electrotropha, this operon exhibited a higher GC content compared with the genome average (66.22% vs 63.87%, respectively), an increase that falls within the range typically associated with HGT events where horizontally transferred genes often display atypical GC content relative to the recipient genome (67, 68). However, GC content alone is not a strong indicator of HGT events, as intrinsic genome variability (e.g., non-uniform GC content) can often explain deviations in GC content of individual genes, operons, or regions of the genome. Therefore, additional analysis evaluating codon usage was performed.

Relative synonymous codon usage (RSCU) analysis provided additional evidence for the potential foreign origin of this operon. While codon usage-based methods for horizontal gene transfer (HGT) detection are known to generate both false positives and negatives (69), our analysis revealed several signatures consistent with foreign gene origin. RSCU analysis quantifies deviations from expected codon frequencies relative to random (equal) usage, where values >1.0 indicate overrepresentation and values <1.0 indicate underrepresentation of specific codons (70). Large deviations in RSCU values between individual genes and the whole genome can indicate HGT events or selection pressure. RSCU analysis (Table S8) revealed significant deviations from genome-wide codon preferences across all operon genes. pmcA showed the strongest evidence for foreign origin, with large RSCU differences for 14 codons and an average absolute RSCU difference of 0.563 among all codons. The two other cytochrome c genes in the operon (AKL02_08770 and AKL02_08775) exhibited high overall RSCU deviations (0.499 and 0.483, respectively), indicative of potential foreign origin or HGT. However, the multicopper oxidase gene (AKL02_08775) had a low overall absolute RSCU difference (0.289) compared with the cytochrome c genes, and no codons were designated as “Strongly Over/Underused,” which may suggest that this gene is subject to stronger selection or functional constraints. However, the average absolute RSCU deviation of the operon was 0.459, which provides strong evidence for HGT of this operon.

The coordinated regulation, atypical codon usage patterns, and functional coherence of this four-gene operon, along with the phylogenetic incongruence of PmcA, provide evidence to suggest that the entire cytochrome c-multicopper oxidase system in Thioclava electrotropha was acquired through horizontal gene transfer. While we propose that this putative periplasmic c-type cytochrome plays a role in EEU in T. electrotropha, whether acquisition of this operon alone confers capacity for EEU is unclear. Although 2,587 instances of homologs to PmcA have been found across broad phylogenetic groups, the capacity for extracellular electron uptake has only been confirmed in two additional species beyond T. electrotropha: T. atlantica and T. dalianensis. Whether their PmcA homologs also confer EEU capacity remains to be determined. It is plausible that PmcA participates in other periplasmic electron transfer processes beyond EEU, as has been shown for small tetraheme cytochromes (STCs) and FccA in Shewanella oneidensis MR-1, where both proteins serve as shared electron carriers across multiple anaerobic respiratory pathways, including EET (71, 72). We note that Thioclava indica is capable of EEU despite lacking this cytochrome c-multicopper oxidase operon, suggesting that EEU capacity may include other components more broadly conserved in the Thioclava. Further work is needed to evaluate whether homologs of PmcA are associated with EEU capacity in other organisms, and to also investigate the role of all genes in the cytochrome c-multicopper oxidase operon in EEU in T. electrotropha.

Conclusion

Extracellular electron uptake enables some microorganisms to access energy from solid-phase electron donors, yet the molecular machinery underlying this process remains poorly understood outside of a few model systems. This work establishes PmcA, a putative periplasmic monoheme c-type cytochrome, as an important component of EEU in Thioclava electrotropha — an organism that lacks homologs to characterized EET proteins. The enhanced EEU capacity measured upon complementation, exceeding wild-type levels, suggests that native expression of this cytochrome may limit EEU rates, though further investigation is needed to fully characterize this relationship. However, the retention of residual EEU activity in the pmcA deletion mutant suggests that PmcA is important but not solely responsible for EEU in T. electrotropha. Additional potentially redundant mechanisms, such as other uncharacterized periplasmic c-type cytochromes or pathways that are independent of c-type cytochromes entirely, may facilitate EEU in this organism. As industrial applications of microbial electrosynthesis and bioremediation expand, understanding and optimizing the factors that control EEU rates will be essential for improving efficiency and expanding the diversity of organisms suitable for these technologies.

The broad phylogenetic distribution of PmcA homologs across more than 2,500 species spanning multiple bacterial phyla raises the possibility that EEU capacity may be more widespread than currently appreciated. Combined with evidence supporting acquisition of the cytochrome c-multicopper oxidase operon through horizontal gene transfer, these findings suggest that lateral genetic exchange may serve as an important mechanism for disseminating EEU capacity across diverse bacterial lineages — paralleling observations for other EET pathways. These findings have potential implications for understanding biogeochemical cycling in electron donor-limited environments, where solid-phase minerals, conductive networks, and even other species of bacteria may serve as critical energy sources or conduits. However, whether the presence of this operon alone confers EEU capacity in other organisms remains to be determined.

Several key questions remain for future investigation. First, biochemical characterization of PmcA is needed to determine its redox properties and electron transfer partners. Second, the remaining genes in the cytochrome c-multicopper oxidase operon warrant investigation into their potential roles in EEU. Third, testing whether organisms harboring PmcA homologs have the capacity for EEU would clarify the ecological and biotechnological implications of this gene’s broad distribution. Finally, understanding how this pathway integrates with T. electrotropha’s central metabolism may reveal new strategies for optimizing microbial electrosynthesis and other bioelectrochemical applications.

DATA AVAILABILITY

All raw DNA and RNA sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive and are publicly available under BioProject PRJNA1293320. The Thioclava electrotropha ElOx9^T reference genome is available from GenBank under accession number GCA_002085925.2.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00250-26.

Supplemental figures. aem.00250-26-s0001.pdf.

Fig. S1 to S5.

Supplemental tables. aem.00250-26-s0002.xlsx.

Tables S1 to S8.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.