Strained Partnership: Biocompatibility and Transcriptomic Responses of Hydrogenotrophic Methanogens to Methyl Diethanolamine in Bio-integrated Carbon Capture and Utilization

Abstract

Biointegrated carbon capture and utilization (BICCU) represents a promising alternative to conventional carbon capture technologies, combining microbial methanogenesis with CO₂ capture agents, such as methyl diethanolamine (MDEA), to produce renewable methane from dilute CO₂ sources. However, the biotoxicity of these solvents poses a major challenge to the viability of the BICCU approach. This study investigates the biocompatibility and physiological responses ofMethanococcus maripaludis and Methanosarcina barkeri to MDEA, through batch experiments and transcriptomic analysis. Both methanogens exhibited inhibited methane production when exposed to MDEA, with M. barkeri displaying higher sensitivity. M. maripaludis maintained the highest CH₄ production rate at 30 mmol/L MDEA, while M. barkeri was almost fully inhibited at this concentration. Transcriptomic analysis revealed upregulation of genes associated with S-layer formation, cell envelope biogenesis, glycosyltransferase activity, and ion transport, alongside downregulation of lysine biosynthesis and secretion system genes. These findings suggest a metabolic trade-off, where cellular energy is reallocated toward stress response mechanisms. Results demonstrate that, while MDEA can be integrated with microorganisms, the exact response will depend on the methanogenic strain. Future research should focus on more biocompatible solvents and the selection of resilient methanogenic strains.

Introduction

Global CO₂ emissions from the energy sector have increased exponentially over the past decade, reaching an annual emission of 37.4 Gt in 2023. These emissions are primary drivers of global warming, climate change, and associated environmental and socio-economic consequences. In recognition of this urgency, the 2015 Paris Agreement set a goal of limiting the increase in global average temperature to 1.5 °C above preindustrial levels by 2030. Achieving this objective requires the development and implementation of sustainable strategies to address the dual challenge of reducing CO₂ emissions while meeting growing global energy demands. Transitioning energy systems toward electrification with renewable energy is a key strategy in achieving carbon neutrality by 2050. However, electrification and decarbonization of hard-to-abate sectors, such as industrial processes and heavy transportation reliant on high-energy-density chemical carriers, remain challenging.

To date, chemical scrubbing with amine-based solvents remains the most mature and widely adopted industrial carbon capture technology. This process involves CO₂ absorption into an amine solution, followed by thermal desorption (>100 °C) to regenerate the solvent and store or utilize the captured CO₂. However, this approach entails substantial energy penalties due to the high-energy demand for CO₂ desorption and solvent regeneration, alongside solvent degradation over time.

Biointegrated carbon capture and utilization (BICCU) represents a recently formulated concept in which CO₂ from dilute sources such as flue gas is captured and utilized as a feedstock for biochemical conversion into value-added products, including renewable fuels. BICCU into methane integrates amine scrubbing with biomethanation, where hydrogenotrophic methanogens act as microbial catalysts to strip captured CO₂ and convert it into methane (CH₄) (eq ). According to Le Chatelier’s principle, microbial CO₂ utilization drives the bicarbonate equilibrium toward CO₂ release (eq ), enabling solvent regeneration at lower temperatures. , Compared to conventional amine scrubbing, BICCU significantly reduces the energy required for CO₂ stripping, which ranges from 2.7 to 4.2 GJ per ton of CO₂. ,

$${\mathrm{CO}}_2+{\mathrm{R}}_3\mathrm{N}(\mathrm{tertiary}\mathrm{amine})+{\mathrm{H}}_2\mathrm{O}\rightleftharpoons {{\mathrm{HCO}}_3}^{-}+{\mathrm{R}}_3{\mathrm{NH}}^{+}$$ 1

$$4{\mathrm{H}}_2+{\mathrm{CO}}_2\to {\mathrm{CH}}_4+2{\mathrm{H}}_2\mathrm{O}$$ 2

Methanogens are a group of anaerobic methane-producing archaea with a massive diversity. Methanogens rely on CO₂ for their metabolism, and its concentration influences their activity. , Methanogens play a key role in biogas reactors, but have also been used in conjunction with exogenous addition of renewable hydrogen for the production of methane from highly concentrated CO₂ sources, a robust power-to-X (PtX) technology. − Methanogens are currently employed in industrial biomethanation systems for both biogas upgrading and the production of renewable methane. Their application is well-established across various process configurations from bench-to-industrial scale, including both ex situ and in situ biomethanation setups, demonstrating technical feasibility and stable process performance. − However, in contrast to biomethanation of gaseous CO₂, methanogens in BICCU are in close contact with chemical capture agents that are not designed for biocompatibility.

For the successful implementation of BICCU into methane, methanogens must maintain robust interactions with the capture agent without compromising growth, physiology, and methane metabolism. Although amine-based solvents such as methyl diethanolamine (MDEA) exhibit high CO₂ absorption capacity and bioavailable CO₂ release without forming a strong covalent bond, their potential biotoxicity remains a key concern. , A recent study demonstrated that methanogens within mixed microbial consortia exhibit varying responses to different capture agents due to their biocidal properties, , which can differentially inhibit methanogenesis. Among several capture agents tested, including monoethanolamine (MEA), polyethylene glycol (PEG-200), and N ¹-dodecyl-N ³,N ³-dimethylpropane-1,3-diamine (C₁₂-DMAPA), MDEA was identified as a promising candidate for BICCU. However, previous studies with mixed microbial consortia showed that methanogenic inhibition by MDEA becomes significant at concentrations between 50 and 100 mmol/L, ,, despite the presence of a diverse microbial community that typically confers resilience and mitigates transient stress. This limitation poses a challenge for industrial scalability, as conventional amine scrubbing operates at capture solvent concentrations of 1.5–2.5 mol/L.

Although many Archaea have evolved to survive in extreme environments, recent studies indicate that some are highly sensitive to changes in their chemical environment induced by MDEA. , Microbial stress responses are triggered when environmental changes exceed tolerance thresholds, activating protective mechanisms to mitigate damage. However, excessive stress can impair metabolic function or lead to cell death, ultimately compromising system performance in BICCU applications. While previous studies have demonstrated methanogenic inhibition by various compounds, including long-chain organic molecules, surfactants, trace elements, and antibiotics. These effects depend on the compound’s toxicity and dosage. , However, the physiological, cellular, and metabolic responses of methanogens to CO₂ capture solvents remain largely unexplored. The precise mechanisms underlying methanogen sensitivity to MDEA and the reasons for metabolic inhibition remain unclear. Furthermore, microbial interactions within mixed inoculum introduce additional complexity, making it challenging to isolate the specific causes of toxicity. To address this gap, it is crucial to investigate the physiological and cellular responses of methanogens to aqueous CO₂-capture agents at the individual strain level. Such research is essential for selecting suitable microbial catalysts and thereby optimizing the BICCU system’s efficiency and viability.

The ability of archaea to modulate gene expression in response to environmental changes is well established. Among methanogens, Methanococcus maripaludis (order Methanococcales) and Methanosarcina barkeri (order Methanosarcinales) are well-characterized, mesophilic, and fast-growing hydrogenotrophic methanogens with fully sequenced genomes. Both species utilize CO₂ as their sole carbon source, reducing it with H₂ to generate CH₄. − Further, M. maripaludis is a strictly hydrogenotrophic, noncytochrome methanogen with a simple S-layer cell envelope, whereas M. barkeri is a cytochrome-containing, metabolically versatile methanogen with a methanochondroitin-based cell matrix. − Given their metabolic capabilities and physiological and structural differences, M. maripaludis and M. barkeri serve as ideal model organisms for studying the inhibitory effects of MDEA on hydrogenotrophic methanogens. We hypothesize that methanogens exhibit species-specific responses to MDEA exposure, which may be influenced by differences in tolerance levels, cell morphology, and metabolic adaptability. Consequently, gene expression patterns relevant to stress response and methanogenesis could undergo significant changes in response to MDEA exposure.

This study aims to investigate the biocompatibility of M. maripaludis and M. barkeri with MDEA through laboratory-scale batch culture experiments. Additionally, we performed transcriptomic analysis to explore how defined concentrations of 50 and 100 mmol/L of MDEA-induced inhibition affect the physiology and metabolic functions of M. maripaludis, thereby identifying differential gene expression patterns in response to chemical stress. Understanding these responses is crucial for optimizing BICCU performance and ensuring its feasibility for industrial-scale CO₂ capture and conversion of these mitigated emissions to renewable fuel in the form of methane.

Materials and Methods

Chemicals and Gas Composition

The tertiary amine, methyl diethanolamine (MDEA), with a purity of >99% (Sigma-Aldrich), was used as the aqueous scrubbing agent. All chemicals used in the study were of analytical grade. Gases were purchased from Air Liquide ALPHAGAZ (Denmark). The synthetic flue gas composition consisted of CO₂ and N₂ at a 20:80 ratio. All solutions and media were degassed to achieve anoxic conditions using 100% N₂. The headspace of batch reactors was flushed with either 100% H₂ or with a 20:80 H₂/CO₂ mixture.

Methanogenic Strains, Media, and Culture Conditions

The methanogenic strains M. barkeri 227 (DSM 1538) and M. maripaludis OS7 (NBRC 103642) were used as model organisms under mesophilic conditions (37 °C). M. barkeri 227 was obtained from the German Collection of Microorganisms (DSMZ), while M. maripaludis OS7 was obtained from NITE (Biological Resource Center, National Institute of Technology and Evaluation, Japan). M. barkeri 227 was cultivated in Medium 120 (DSMZ) containing (per liter of demineralized water): 0.35 g of K₂HPO₄, 0.23 g of KH₂PO₄, 0.50 g of NH₄Cl, 0.5 g of MgSO₄ × 7H₂O, 0.25 g of CaCl₂ × 2H₂O, 2.25 g of NaCl, 2 mL of FeSO₄ × 7H₂O solution (0.1%, w/v) from medium DSM 119, 1 mL of trace element solution SL-10 from medium 320, 0.5 mL of sodium resazurin (0.1%, w/v), 2 g of NaHCO₃, 1 mL of Wolin’s vitamin solution DSM 141 (10×), and 0.30 g of l-cystine HCl × H₂O (pH 7.0). M. maripaludis OS7 was cultivated in Medium 911 containing (per liter of demineralized water): 5 mL of KP7 buffer (K₂HPO₄, 106 g/L; KH₂PO₄, 53 g/L), 3.05 g of MgCl₂ × 6H₂O, 0.15 g of CaCl₂ × 2H₂O, 0.54 g of NH₄Cl, 0.3 g of Na₂SO₄, 30 g of NaCl, 2 mL of trace element solution from medium 927, 1 mL of sodium resazurin (1 g/L), 1.5 g of Na₂CO₃, and 0.5 g of l-cystine HCl × H₂O (pH 7.1).

Initially, both strains were cultivated in their recommended media at 37 °C under anaerobic conditions. Once growth was confirmed, the strains were adapted to grow in the modified mineral media. Growth monitoring was performed at regular intervals using headspace gas compositions (CH₄, CO₂, and H₂) as described by Ngoumelah et al., confirming that growth in the modified mineral media was comparable to that in the recommended media. Two consecutive transfers in the modified mineral medium were performed before the batch experiments to ensure acclimatization.

Biocompatibility Assay with Aqueous CO₂ Capture Agent

Experimental Setup

The biocompatibility assay was conducted in laboratory-scale batch reactors as described by Sieborg et al., with minor modifications. Briefly, experiments were conducted in 250 mL borosilicate serum bottles under anoxic conditions. A concentration series (0–100 mmol/L), abiotic controls (without inoculum), and positive controls (containing inoculum without MDEA) were prepared in triplicate in a total volume of 45 mL of the modified mineral medium per reactor. The synthetic flue gas (CO₂/N₂, 20:80) was filter-sterilized and bubbled through the liquid phase in all reactors at 37 °C, except for the positive controls, until the pH remained stable for at least 5 min, indicating CO₂ saturation in the liquid phase containing the capture agent MDEA. The reactors were placed in a water bath at 37 °C to maintain mesophilic conditions during the gas flushing. Methanogenic inoculum (5 mL, 10%, v/v) in the exponential phase (OD₆₀₀ ∼ in the range of 0.18–0.25) was added to batch reactors. For M. barkeri, the exponential phase was determined by an average methane concentration of approximately 15–20 mmol/L per culture.

The headspace of the positive controls was flushed with filter-sterilized CO₂/H₂ (20:80) and pressurized to an average total pressure of 1.80 ± 0.04 bar. The headspace of the abiotic controls and treatment reactors was flushed with 100% H₂ and pressurized to an average total pressure of 1.44 ± 0.04 bar, ensuring H₂ partial pressure equivalent to that of positive controls. The gas-to-liquid volume ratio (V _gas/V _liquid) was maintained at 4.0. All batch reactors were incubated at 37 °C with continuous magnetic stirring at 400 rpm. During incubation, headspace pressure was recorded using a pressure sensor (Vernier GO Direct Gas Pressure, U.S.A.). Headspace gas samples (2 mL) were collected aseptically at the lag, midexponential, and final phases of the experiment for gas composition analysis. The final pH of the liquid phase was measured at the end of the experiment. For M. maripaludis OS7, 1 mL liquid samples were collected at the start and end of the experiment to determine cell density via OD₆₀₀ measurements. Dissolved inorganic carbon (DIC; CO_2(aq) and HCO₃ ^–) analysis was conducted to quantify CO₂ absorption in the capture solution at the beginning (t ₀) and the residual CO₂ in the liquid fraction at the end of the experiment (t _end), as described by Oliveira et al. using gas chromatography (Shimadzu GC-2014, Japan).

Microbial Uptake Kinetics of CO₂ and H₂

The gas composition in the headspace (H₂, CO₂, CH₄, and N₂) was analyzed using gas chromatography (Shimadzu GC-2014, Japan) equipped with a thermal conductivity detector (TCD) and two packed columns. For the separation of H₂, argon (Air Liquide ALPHAGAZ) was used as carrier gas through a packed column (Restek ShinCarbon ST, Bellefonte, U.S.A.). Helium (Air Liquide ALPHAGAZ) was used as the carrier gas through a Porapak Q column (CS-Chromatography Service GmbH, Germany) to separate CO₂, CH₄, and N₂. Methanogenic conversion of H₂ and CO₂ to CH₄ results in a 4-fold reduction in headspace pressure, which was used to model microbial uptake kinetics via the modified Gompertz equation. Methane production was calculated following eqs –:

Methane partial pressure at time “n” (PM _n , in bar)

$${\mathrm{PM}}_n={\mathrm{P}}_n\times x{\mathrm{CH}}_{4,n}$$ 3

Methane production at time “n” (MP _n , in bar)

$${\mathrm{MP}}_n={\mathrm{PM}}_{\mathrm{pre‐sampling},n}-{\mathrm{PM}}_{\mathrm{post‐sampling},n-1}$$ 4

Cumulative methane production at time “t ₀ to t _n ” (MP_cum,n , in bar)

$${\mathrm{MP}}_{\mathrm{cum},n}=\sum _{t=0}^{t=n}{\mathrm{MP}}_{\mathrm{total}}$$ 5

Molar methane production at time “t ₀ to t _n ” (nCH₄, _n , in mol)

$$n{\mathrm{CH}}_{4,n}=({\mathrm{MP}}_{\mathrm{cum},n}\times V_{\mathrm{headspace}})/(R\times T)$$ 6

where P _n is the internal pressure at time “n” (bar), xCH_4,n is the methane fraction in headspace at time “n” (dimensionless), PM_{pre‑sampling,n} and PM_{post‑sampling,n–1} are the methane partial pressure (bar) measured before gas composition analysis at the sampling time “n” and after gas composition analysis at the sampling time “n – 1”, respectively, V _headspace is the reactor headspace volume (L), R is the universal gas constant (0.082 atm L/mol·K), and T is the incubation temperature (K). Maximum methane production rates (RCH₄, mmol/h) and methane production potential (PCH₄, mmol/L per culture) were estimated by fitting the modified Gompertz model to the cumulative methane production data using OriginPro 2018 (Origin-Lab Corporation, Northampton, MA, U.S.A.).

CO₂ mass balances were calculated following eq . Microbial conversion of captured CO₂ was estimated from cumulative methane (nCH₄) at the end of incubation (t _end), considering 1:1 CO₂/CH₄ stoichiometry. The residual CO₂ pool partitioned in the gaseous fraction and liquid fraction was quantified from CO₂ remaining in the headspace and DIC (t _end), respectively. Any carbon not accounted for as CH₄ formation, headspace CO₂, or liquid-phase DIC was reported as “biomass and other”, which may include dissolved CH₄ in the liquid phase, carbon assimilated into biomass, and potential sampling losses from the total absorbed CO₂.

CO₂ absorbed (mmol) by MDEA after saturation (t ₀)­

$$n{\mathrm{CO}}_2\mathrm{absorbed}(\mathrm{mmol})={\mathrm{DIC}}_{t_0}\times V_{\mathrm{L}}$$ 7

The overall CO₂ mass balance

$$n{\mathrm{CO}}_2,\mathrm{absorbed}=n{\mathrm{CH}}_4(t_{\mathrm{end}})+n_{\mathrm{liquid\; phase}}\mathrm{DIC}+n_{\mathrm{headspace}}{\mathrm{CO}}_2+n_{\mathrm{biomass\; and\; other}}$$ 8

RNA Extraction and Purification

Based on methane production results, RNA was extracted from positive controls of M. barkeri and M. maripaludis, as well as treatments where M. maripaludis was exposed to 50 and 100 mmol/L MDEA, and M. barkeri to 10 mmol/L MDEA. RNA extraction was performed in four biological replicates per condition. For each sample, 100 mL of the liquid sample (50 mL per subreplicate) was collected during the exponential phase. Cells were harvested by centrifugation at 5,000× g for 5 min at 4 °C. The supernatant was removed, and the resulting cell pellet was snap-frozen in liquid nitrogen and stored at −80 °C until further processing.

Total RNA was extracted using a modified RNeasy Mini Kit (Qiagen, Germany). Cell pellets were resuspended in Buffer RLT with 1% (v/v) β-mercaptoethanol. The cell suspension, containing approximately 10⁹ cells, was transferred into chilled Power Beads Tubes (Qiagen, Germany) and subjected to bead-beating with a FastPrep-24 homogenizer (MP Biomedicals, U.S.A.) at 4 m/s for 60 s. The manufacturer’s protocol for RNA extraction was followed thereafter. RNA purification was performed using the riboPOOL cleanUP module (siTOOLs Biotech, Germany), and RNA was eluted in 40 μL of nuclease-free water. RNA concentration was measured using the QubitTM RNA BR Assay Kit (Invitrogen, U.S.A.) on a Qubit 4 fluorometer. RNA integrity was assessed via RIN values obtained from a bioanalyzer with RNA ScreenTape (Agilent Technologies, U.S.A.). Throughout the procedure, samples were kept on ice-cold storage blocks to prevent RNA degradation.

cDNA Library Preparation and Sequencing

Purified total RNA was polyadenylated using E. coli poly­(A) polymerase. Approximately 6 μg of total RNA was incubated at 37 °C for 30 min with 1 mM ATP and 5 U of E. coli poly­(A) polymerase (NEB, England). RNA was subsequently purified using the riboPOOL cleanUP module. Ribosomal RNA (rRNA) was depleted using the riboPOOL Probes Protocol (siTOOLs Biotech, Germany), as per the manufacturer’s guidelines. The success of rRNA depletion was verified using a bioanalyzer. cDNA synthesis and library preparation were performed using the Direct cDNA Sequencing V14 kit (SQK-LSK114, Oxford Nanopore Technologies, U.K.), with SuperScript IV Reverse Transcriptase (Thermo Fisher Scientific, U.S.A.), according to the manufacturer’s recommendations. Adaptor ligation and barcoding were conducted using the Native Barcoding Kit 24 V14 (Oxford Nanopore Technologies, U.K.). Approximately 100 fmol of prepared, barcoded, and adaptor-ligated library was loaded onto a PromethION R10.4.1 flow cell, and sequencing was performed for 72 h.

Data Processing and RNA-Seq Analysis

Basecalling and demultiplexing of raw reads were conducted using Dorado v.0.6.0 with sup v.4.3.0 (https://github.com/nanoporetech/dorado), without trimming barcodes but requiring barcodes at both ends. Full-length cDNA reads were identified and trimmed using Pychopper v.2.7.9 (https://github.com/epi2me-labs/pychopper). Poly­(A) tails were trimmed using Cutadapt v.4.0. Reads were mapped to the M. maripaludis OS7 reference genome (NCBI accession no. NZ_AP011528.1) using Minimap2 v.2.26 with settings: −ax splice −p 0.99 −k14 −MD. Transcript quantification was performed using Oarfish v.0.4.0. Differential gene expression analysis was conducted with DESeq2 v.1.44.0. Genes in MDEA-treated samples were considered significantly differentially expressed compared to untreated controls with a cutoff of log₂ fold change of > 0.585 and < –0.585 and an adjusted p value (p _adj) of ≤ 0.05, as determined by the Benjamini–Hochberg procedure. Only transcripts detected in a minimum of three out of four replicates were included in the analysis. Gene ontology and metabolic pathway enrichment analyses were performed using EggNOG v.4.5.1 and KEGG pathway analysis.

Scanning Electron Microscopy (SEM)

Scanning electron microscopy (SEM) was performed at the end of the experiment to examine potential morphological damage caused by MDEA to both M. barkeri and M. maripaludis. Briefly, 50 mL of liquid samples were aseptically obtained from batch reactors containing M. barkeri with 10 mmol/L MDEA and M. maripaludis with 100 mmol/L MDEA, and their respective positive controls without MDEA were included in sterilized anaerobic culture vials. Cells were harvested by centrifugation at 5,000× g for 5 min. The resulting pellet was thoroughly washed twice with sterile Milli-Q water. Samples were then placed on a silicon wafer and left to dry before sputter coating with 8 nm of platinum (Leica EM SCD 500). The acquisition parameters included high vacuum, 5.00 kV, a working distance of 7.4 mm, and a current of 27 pA on a FIB/SEM 3D Versa (Thermo Fisher).

Data Analysis

Statistical analyses were performed using Minitab statistical software version 21.1.0. One-way ANOVA was performed to determine significant differences among individual means of measured parameters across different concentrations of MDEA, followed by Tukey’s post hoc test. A p value of < 0.05 was considered statistically significant.

Results

Biocompatibility of MDEA: Effects on Methane, Growth, and pH

Methanogenesis was observed under varying concentrations of the aqueous CO₂ capture agent methyl diethanolamine (MDEA) in both cultures with M. maripaludis and M. barkeri 227. Compared to the culture without MDEA (positive control, with CO₂/H₂, 20:80 in the headspace), MDEA supplementation led to quantifiable methane (CH₄) production, but overall methanogenic activity was substantially reduced relative to the control, indicating an inhibitory effect of MDEA (Figure ). All CH₄ measurements and potential production estimations were conducted in triplicate and showed relatively small standard deviations. The two strains exhibited distinct responses in both CH₄ production rate and CH₄ production potential when exposed to MDEA. The maximum CH₄ production rate reflects the kinetic activity of methanogenesis, whereas the CH₄ production potential represents the overall conversion capacity over the course of the experiment. Furthermore, trends in kinetic activity and cumulative conversion differ with MDEA concentration and between the two species, indicating that higher CH₄ production rates do not directly correspond to enhanced CH₄ production potential. This suggests that although methanogenic activity may be initially stimulated, it does not sustain, likely due to MDEA-mediated inhibition over time, which limits overall conversion efficiency despite higher CO₂ availability at higher MDEA concentrations.

1.

: Biocompatibility of the methanogenic strains M. maripaludis and M. barkeri 227 under different concentrations of the CO₂ capture agent methyl diethanolamine (MDEA). All experiments were conducted with 100% H₂ in the headspace, except for the positive control, which contained a 20:80 CO₂/H₂ mixture. (A) Maximum methane (CH₄) production rate, representing the kinetic activity of methanogenesis, and (B) CH₄ production potential, representing the overall conversion capacity over the course of the experiment, for both strains, M. maripaludis (blue) and M. barkeri (orange). All data are shown as mean ± SD (n = 3). Error bars indicate the standard deviation of three independent measurements. Different letters denote statistically significant differences among MDEA concentrations within each species (uppercase A–F for M. maripaludis, lowercase a–f for M. barkeri; p < 0.05).

For M. maripaludis, the highest CH₄ production rate of 0.068 mmol of CH₄/L·h occurred at 30 mmol/L MDEA, whereas M. barkeri was strongly inhibited, with the highest CH₄ production rate of only 0.001 mmol of CH₄/L·h at 10 mmol/L MDEA. The highest CH₄ production potential for M. maripaludis was observed at 50 mmol/L MDEA, reaching 26.92 mmol of CH₄/L per culture, after which it declined. Considering the CH₄ production potential, the highest value was observed at 50 mmol/L MDEA; however, beyond this concentration, the CH₄ potential declined markedly at 70 and 100 mmol/L, indicating a threshold above which MDEA exerts increasing inhibitory effects. In M. barkeri, the CH₄ production potential was significantly lower at 1.001 mmol of CH₄/L per culture, representing a 97% reduction relative to the positive control. One-way ANOVA indicated statistically significant differences (p < 0.05) in the maximum CH₄ production rates for M. maripaludis between the positive control and treatments with all MDEA concentrations. Similarly, the CH₄ production potential at both 50 and 100 mmol/L MDEA was significantly lower than the control (p < 0.05).

The percentage increase in cell density of M. maripaludis at the end of incubation (t _end) relative to the initial inoculated cell quantity (mean ± standard deviation), determined based on optical density measurements at 600 nm (OD₆₀₀) (Table ). This data did not show a clear correlation with the maximum CH₄ production rate across different MDEA concentrations. At 100 mmol/L MDEA, precipitate formation in the medium interfered with spectrometric readings, making it difficult to obtain reliable cell abundance data. This precipitate formation is likely due to changes in the medium’s pH. As microbes metabolized CO₂, the pH-dependent chemical equilibrium between CO₂ and MDEA [HCO₃ ^– + R₃NH⁺ (MDEA⁺) ⇌ CO₂ + R₃N (MDEA) + H₂O] shifted, increasing the concentration of un-ionized MDEA and leading to a rise in pH. The final recorded pH values of the liquid phase at the end of incubation are represented in Table .

1. Initial pH after CO₂ Saturation, Final pH at the End of Incubation, and Percentage Increase in Optical Density (OD₆₀₀) relative to the Initial OD₆₀₀ at t ₀ .

Control and treatments	Initial pH (mean ± SD) after CO2 saturation	Final pH (mean ± SD) at the end of incubation	Percentage (%) increase in OD600
Positive control (without MDEA)	7.03	7.01	72.20 ± 4.51
10 mmol/L	7.01	6.99	19.71 ± 5.59
20 mmol/L	7.04	7.25	11.91 ± 2.09
30 mmol/L	7.06	7.91	3.32 ± 1.07
50 mmol/L	7.10	8.00	43.57 ± 6.60
70 mmol/L	7.15	8.25	26.25 ± 3.32
100 mmol/L	7.21	8.40	no data (precipitate formed)

^aThe positive control (without MDEA) was not subjected to CO₂ saturation.

CO₂ Capture and Mass Balance

CO₂ conversion into methane in the BICCU process involves two stages: chemical capture of CO₂ into MDEA solvent and microbial conversion of captured CO₂ to CH₄. The dissolved inorganic carbon (DIC) content measured after CO₂ absorption increased proportionally as MDEA increased from 10 to 100 mmol/L, confirming efficient CO₂ loading (Figure A). The carbon mass balance at the end of incubation differed between the two methanogens (Figure B). At 50 mmol/L, M. maripaludis converted approximately 50% of the captured CO₂ to CH₄, with only residual pools in liquid-phase DIC and biomass/other, indicating the highest bioconversion performance. However, at 70–100 mmol/L, liquid-phase DIC remained high, consistent with biocompatibility becoming limiting at higher MDEA. In contrast, the cumulative CH₄ of M. barkeri remained low, while most carbon remained as residual DIC, indicating that CO₂ remained available in the medium. The small fraction of CO₂ in the headspace supports that it is retained in the MDEA-containing liquid.

2.

CO₂ capture and carbon mass balance under different concentrations of the CO₂ capture agent methyl diethanolamine (MDEA). (A) Dissolved inorganic carbon (DIC; mmol/L) in the capture solution at t ₀ after CO₂ absorption, once pH stabilized. (B) Carbon mass balance (mmol) at the end of incubation (t _end) for M. maripaludis and M. barkeri. The converted CO₂ was reflected in CH₄, while the residual fraction was detected in the liquid-phase as DIC and in the gas-phase as headspace CO₂, and the unaccounted carbon fraction was assumed to represent biomass and other. All data are shown as the mean ± SD (n = 3). Error bars indicate the standard deviation of three independent measurements.

Transcriptomic Response of M. maripaludis to MDEA Exposure

The control samples of M. barkeri produced high-quality RNA (RIN > 8.5) at appropriate concentrations (mean: 147 ng/μL) for further processing. However, RNA quantity and quality from MDEA-treated M. barkeri samples were insufficient for further analysis (RIN < 8.5, mean: 37 ng/μL), even with a total liquid sample volume of 100 mL. This is consistent with the pronounced inhibition observed in M. barkeri under MDEA exposure, resulting in very low biomass.

Consequently, sequencing and further analysis proceeded only with M. maripaludis, as RNA from these samples was of high quality (RIN > 8.5) and obtained in sufficient yield across all conditions. Control samples yielded >1000 ng/μL, while 50 mmol/L MDEA-treated samples yielded 611 ng/μL, and 100 mmol/L MDEA-treated samples yielded 210 ng/μL. The final cDNA library was prepared equimolarly from each sample, with the barcoded library reaching a final concentration of 9.14 ng/μL at ∼100 fmol.

Sequencing results, presented in Table S1, show that the raw sequencing throughput yielded 3.67 Gb from 2.34 million reads, with an N50 read length of 2,138 bp. After filtering, full-length cDNA reads mapped to both the genome and transcriptome of M. maripaludis (NCBI accession number NZ_AP011528.1), resulting in 349.8 Mb from 461 924 reads, with an N50 read length of 1378 bp. The median Phred quality score for filtered reads was 17.1, corresponding to a base call accuracy of 98.18%. The mapping depth and coverage of full-length cDNA reads to their reference genome are visualized in the Supporting Information, showing nonuniform gene expression across the genome and distinct regions of up- and downregulation in response to MDEA exposure.

To assess the tolerance limits of M. maripaludis under MDEA exposure, differential gene expression analysis was conducted at 50 and 100 mmol/L MDEA by comparison to the positive control without MDEA. A total of 32 genes (1.9%) were downregulated, while 87 genes (5.1% of total genes) were upregulated in M. maripaludis exposed to MDEA compared to the positive control, using a cutoff of log₂ fold change of > 0.585 and < –0.585, respectively, with a p _adj ≤ 0.05 (Benjamini–Hochberg procedure) (Figure ). At 50 mmol/L MDEA, 6 genes were uniquely downregulated, while 24 genes were upregulated. Exposure to 100 mmol/L MDEA resulted in 20 uniquely downregulated and 42 uniquely upregulated genes. The circos plot (Figure S1) illustrates nonuniform gene expression across the genome, highlighting regions of down- and upregulation in response to MDEA exposure.

3.

Volcano plot of differentially expressed genes between the control and MDEA treatments. Red dots represent genes downregulated in both 50 and 100 mmol/L MDEA when compared to the control. Green dots indicate genes upregulated in response to both 50 and 100 mmol/L MDEA. Light green and dark green dots represent genes uniquely upregulated at 50 and 100 mmol/L, respectively. Black dots denote genes that were not significantly differentially expressed.

Functional classification of the differentially expressed genes using cluster of orthologous groups (COG) analysis grouped them into 19 categories (Table S2), with the highest number of DEGs assigned to “unknown function”. Genes involved in energy production, nucleotide metabolism, carbohydrate metabolism, coenzyme metabolism, cell envelope biogenesis, and protein turnover were exclusively upregulated at both 50 and 100 mmol/L MDEA concentrations. Downregulated genes primarily associated with inorganic ion transport and intracellular trafficking, including FeoA family protein and ATPase gene belonging to the T2SS/T4P/T4SS family associated with the function of type II secretions, were unique to the 100 mmol/L MDEA treatment. Interestingly, genes related to signal transduction and defense mechanisms, such as the type II toxin–antitoxin system, the HicB family antitoxin, and the CRISPR-associated helicase/endonuclease Cas3, were exclusively upregulated under 100 mmol/L MDEA treatment, suggesting a stress-induced response specific to higher MDEA concentrations (Table ). The key metabolic and physiological response mechanisms of M. maripaludis under 100 mmol/L MDEA exposure are illustrated in Figure . These integrated responses reflect a complex cellular response strategy employed to support survival and activity under MDEA-induced stress rather than a defined mechanistic pathway.

2. Selected Up- and Downregulated Genes in M. maripaludis OS7 Following Exposure to 100 mmol/L MDEA, Including Gene Function, Differential Expression (Log₂ Fold Change) and Adjustedp Value, and COG Classification, Indicating Pronounced Physiological and Stress-Related Responses at Higher MDEA Concentrations.

Gene ID	Description	Log2 fold change	Adjusted p value	Cluster of orthologous groups (COG)
gene MMOS7_RS03865	UDP-N-acetylglucosamine 2-epimerase (non-hydrolyzing); WP_119720821.1	2.2269	0.0343	M: cell wall/membrane/envelope biogenesis
gene MMOS7_RS06065	glycosyltransferase family 4 protein; WP_119721061.1	3.0634	0.0037	M: cell wall/membrane/envelope biogenesis
gene MMOS7_RS02090	type II toxin–antitoxin system HicB family antitoxin; WP_119720650.1	2.9014	0.0138	S: secondary metabolites biosynthesis, transport, and catabolism
gene MMOS7_RS05290	CRISPR-associated helicase/endonuclease Cas3; WP_013999400.1	2.9078	0.0188	V: defense mechanisms
gene MMOS7_RS00525	ammonium transporter; WP_119720503.1	3.2203	0.0423	P: inorganic ion transport and metabolism
gene MMOS7_RS03490	FeoA family protein; WP_013999027.1	–1.7204	0.0149	P: inorganic ion transport and metabolism
gene MMOS7_RS02960	sulfate permease; WP_013999114.1	3.7907	0.0055	P: inorganic ion transport and metabolism
gene MMOS7_RS04790	cation/proton antiporter; WP_119720921.1	3.6570	0.0077	P: inorganic ion transport and metabolism
gene MMOS7_RS00385	ATPase%2C T2SS/T4P/T4SS family; WP_119720492.1	–2.0045	0.0111	U: intracellular trafficking, secretion, and vesicular transport
gene MMOS7_RS01265	molybdopterin molybdotransferase MoeA; WP_013998652.1	4.1500	0.0000	H: coenzyme transport and metabolism
gene MMOS7_RS09050	coenzyme F420-0/l-glutamate ligase; WP_119721348.1	3.0268	0.0498	H: coenzyme transport and metabolism
gene MMOS7_RS03955	4Fe–4S binding protein; WP_013998912.1	4.6233	0.0201	C: energy production and conversion
gene MMOS7_RS01345	metalloregulator ArsR/SmtB family transcription factor; WP_011170161.1	4.0887	0.0036	K: transcription
gene MMOS7_RS09080	FlaD/FlaE family flagellar protein; WP_119721351.1	3.2633	0.0138	N: cell motility
gene MMOS7_RS03205	4-hydroxy-tetrahydrodipicolinate synthase; WP_013999071.1	–1.8959	0.0020	E: amino acid transport and metabolism

4.

Overview of the physiological and metabolic processes in M. maripaludis most influenced by exposure to 100 mmol/L MDEA. The up- and downregulated genes are indicated in red and blue arrows, respectively. Abbreviations of the hydrogenotrophic methanogenesis pathway of M. maripaludis showed are as follows: Enzymes: Fwd, formylmethanofuran dehydrogenase; Ftr, formylmethanofuran-H₄MPT formyl transferase; Mch, methenyl-H₄MPT cyclohydrolase; Mtd, F₄₂₀-dependent methylene-H₄MPT dehydrogenase; Mer, methylene-H₄MPT reductase; Mtr, methyl-H₄MPT coenzyme M methyltransferase; Mcr, methyl coenzyme M reductase; Hdr, heterodisulfide reductase; Eha/Ehb, energy-conserving hydrogenases; Fru, F₄₂₀-reducing hydrogenase (selenocysteine containing); and Frc, F₄₂₀-reducing hydrogenase (cysteine containing). Metabolites: Fd_red, reduced ferredoxin; Fd_ox, oxidized ferredoxin; MFR, methanofuran; HS-CoM, coenzyme M; and HS-CoB, coenzyme B.

Methanogenesis and Regulatory Responses

Several genes associated with methanogenesis were differentially expressed under MDEA exposure. Upregulation of genes involved in molybdopterin biosynthesis, iron–sulfur cluster assembly, cobalamin biosynthesis, and coenzyme synthesis were observed, indicating metabolic responses to MDEA stress. In contrast, the downregulation of lysine biosynthesis and secretion system genes at 100 mmol/L MDEA suggested an energy trade-off prioritizing stress responses over energy metabolism and growth of the methanogen.

SEM Image Analysis

Scanning electron microscopy (SEM) images show a distinct difference in cell morphology between the MDEA-treated sample and the MDEA-untreated control for M. barkeri, whereas M. maripaludis shows no obvious differences compared with the MDEA-untreated control. Control cells exhibited the characteristic coccoid shape of typical M. barkeri, cells appearing intact, smooth-surfaced, and closely associated in dense, tetrad-like arrangements (Figure A). In contrast, cells exposed to MDEA exhibited clear morphological damage, including disrupted cell structure and a marked loss of membrane rigidity (Figure B). SEM images of M. maripaludis under untreated MDEA conditions showed a compact cellular morphology with well-defined cell boundaries (Figure C). Following exposure to MDEA, the overall morphology of M. maripaludis was less affected by MDEA (Figure D) than M. barkeri, which showed clear structural disintegration (Figure B).

5.

Scanning electron microscopic images of (A) M. barkeri in MDEA untreated control (10,000×), (B) M. barkeri in 10 mmol/L MDEA (8,000×) following 7 days of incubation, (C) M. maripaludis in MDEA untreated control (32,500×), and (D) M. maripaludis in 100 mmol/L MDEA (32,500×) following 5 days of incubation. Red arrows in panel B indicate the disrupted cell structure of M. barkeri, showing the loss of cell integrity. Red arrow in panel D indicates less pronounced morphological changes in M. maripaludis compared to the control shown in panel C.

Discussion

MDEA-Induced Methanogenesis Inhibition and Strain-Specific Responses

The inhibition of methanogenesis in M. maripaludis and M. barkeri upon exposure to MDEA highlights distinct species-specific responses to chemical stress. The observed peak in CH₄ production in M. maripaludis at 30 mmol/L MDEA, followed by a yield increase at 50 mmol/L, suggests a dual effect in which CO₂ released from MDEA may initially provide a substrate advantage. However, at higher concentrations, inhibitory effects become dominant, leading to a decline in methane production. The end-point carbon mass balance at higher MDEA concentrations also reveals that the limitation is biocompatibility, not CO₂ availability. Further, the absence of a clear correlation between the cell density of M. maripaludis and CH₄ production suggests that several factors may influence CH₄ production and growth. In contrast, M. barkeri was strongly inhibited, with minimal CH₄ production and more carbon as residual DIC, even at low MDEA concentrations. Differences in cell morphology and envelope composition between M. barkeri and M. maripaludis could contribute to the observed differences in tolerance between the two strains. , SEM images of M. barkeri suggest that MDEA exposure compromises cell integrity, potentially leading to the observed inhibition of methane production and poor RNA quantity in MDEA-treated samples. M. barkeri prefers low-osmolarity environments to effectively produce the methanochondroitin layer, a unique extracellular polysaccharide that could potentially strengthen the cell’s resistance to direct chemical interference. In contrast, high-osmolarity conditions in the medium, such as those associated with MDEA, could potentially affect the synthesis of methanochondroitin. Consequently, M. barkeri could become more susceptible to MDEA-induced toxicity, allowing MDEA to interfere with the cytoplasmic membrane and penetrate the cell, ultimately compromising cell integrity as observed in Liu et al., who reported that excess NH₃ can induce structural alterations in the cytoplasmic membrane of M. barkeri including reduced structural stability, increased permeability, loss of membrane potential, destabilized extracellular polymeric substances (EPSs), and decreased lipid content, particularly affecting polyunsaturated carbon–carbon double bonds and longer lipid chains. In addition, although CO₂ reduction to methane with H₂ in both species involves membrane-associated protein complexes, their energy-conservation strategies differ depending on the availability of cytochrome. Cytochrome-lacking M. maripaludis conserves energy predominantly via the cytosolic HdrABC–Mvh electron bifurcation system, which couples H₂ oxidation to the reduction of the heterodisulfide (CoM–S–S–CoB) and ferrodoxin reduction. However, the membrane-associated hydrogenase Eha/Ehb complex is also involved in generating reduced ferrodoxin, whereas ATP conservation occurs via chemiosmotic ion translocation, particularly through the membrane-associated Mtr complex. In contrast, cytochrome-containing M. barkeri relies more strongly on a proton-motive force (PMF)-linked membrane-bound electron-transport chain, including Ech (ferredoxin-dependent hydrogenase) for ferredoxin reduction and Vho (methanophenazine-dependent hydrogenase) for H₂ oxidation, with electron flow coupled to the reduction by HdrDE (membrane-bound CoB–S–S–CoM reductase). , Our transcriptomic analysis of M. maripaludis under MDEA stress showed increased expression of Fe–S cluster genes associated with HdrABC at 100 mmol/L, suggesting that the organism reallocates resources toward maintaining bifurcation-dependent redox coupling under stress. Accordingly, under membrane-active stressors such as MDEA, a membrane-coupled energy conservation may be more sensitive to cell envelope disturbances than bifurcation-centric energy conservation. If MDEA compromises the cell envelope of M. barkeri, as observed in SEM images, it would be expected to preferentially impair cytochrome- and methanophenazine-linked electron transfer associated with the membrane-bound HdrDE, thereby reducing methanogenesis. These differences in energy-conservation pathways could also be another reason for the observed higher inhibition of methanogenesis in M. barkeri than in M. maripaludis at low MDEA concentrations.

Gene Expression and Stress Response

The transcriptomic response to MDEA exposure provided valuable insights into both direct inhibition and indirect stress response mechanisms. Downregulated genes likely represent pathways directly inhibited by MDEA, whereas upregulated genes indicate physiological responses aimed at mitigating cellular damage and maintaining cellular function.

Functional classification of differentially expressed genes (DEGs) revealed significant impacts on membrane integrity and homeostasis. The upregulation of genes related to S-layer formation, cell envelope biogenesis, and glycosyltransferase activity, including PEGA domain-containing protein, glycosyltransferase family 4 protein, and DUF2334 domain-containing protein, suggests a cellular response to reinforce membrane integrity against MDEA-induced toxicity. These responses were particularly pronounced at 100 mmol/L MDEA, indicating a dose-dependent adaptation. The need for structural reinforcement could result from increased membrane permeability due to MDEA exposure, leading to ionic imbalances and cellular stress.

MDEA increases the liquid pH to 8.2 and 8.5 at 50 and 100 mmol/L, respectively, before CO₂ saturation. Upon CO₂ saturation, the pH reduces to 7.1 and 7.2, respectively. Since cells are inoculated only after CO₂ saturation, they are not exposed to the initial high pH levels. With the microbial conversion of CO₂ to CH₄, the pH levels of the incubations again increased to 8.0 and 8.4 for 50 and 100 mmol/L MDEA, respectively, which could potentially impact cellular pH homeostasis. M. maripaludis exhibits optimal growth within a pH range of 6.8 to 7.2, although it can tolerate pH levels ranging from 6.5 to 8.0. The increased pH observed at higher MDEA concentrations may have contributed to reduced cell growth and metabolic activity. Therefore, the observed stress responses could result from the confounded effects of MDEA toxicity and pH-induced stress. However, the disintegration of the cell wall structure of M. barkeri, even without the pH changes induced by CO₂ conversion, indicates that MDEA has a strong effect on cellular integrity at neutral pH (Table ). The observed pH shift at higher MDEA concentrations could impair enzyme activity and destabilize the intracellular environment. , Genes associated with ion transport were differentially expressed, with upregulation of cation/proton antiporters and ABC transporters suggesting active efforts to counteract ionic imbalances. The exclusive upregulation of sulfate and ammonium transporters at 100 mmol/L MDEA further indicates a cellular shift toward survival strategies, potentially to maintain intracellular ion balance under alkaline stress. − Additionally, the increased expression of ArsR/SmtB metalloregulatory transcription factors implies a role in metal ion regulation, likely as a response to disruptions in metal homeostasis caused by MDEA. −

Methanogenesis under MDEA Stress

Despite the inhibitory effects of MDEA, the upregulation of methanogenesis-related genes suggests that the M. maripaludis cultures attempted to sustain methane production under stress. At 100 mmol/L MDEA, the increased expression of genes involved in molybdopterin and iron–sulfur cluster biosynthesis suggests a potential cellular response for maintaining redox balance and cofactor availability under stress conditions. , This upregulation coincides with a marked downregulation of genes involved in lysine biosynthesis and T2SS/T4P/T4SS secretion systems, which are associated with ATP-dependent assembly of surface structures, suggesting a possible shift in metabolic priorities. Together, these transcriptional changes point toward a resource reallocation strategy in which stress-responsive pathways, such as detoxification, are favored over biosynthesis pathways associated with growth.

MDEA may interfere with methanogenesis by inhibiting key enzymes. The differential expression of genes involved in coenzyme biosynthesis, including coenzyme B, F₄₂₀, and cobalamin, suggests that MDEA exposure disrupts the synthesis or function of essential cofactors required for methane metabolism. Such inhibition would significantly impact the efficiency of methanogenesis, as these cofactors play vital roles in electron transfer and enzymatic catalysis within the methanogenesis pathway. , The upregulation of genes associated with cobalt/nickel transport under MDEA-induced stress may indicate a cellular response to maintain or increase the availability of these metal ions, which are required for the biosynthesis of several methanogenic cofactors, given the established role of cobalt and nickel uptake systems in supporting cofactor biosynthesis in methanogens.

Survival Mechanisms under MDEA Stress

The strong upregulation of the CRISPR-associated helicase/endonuclease Cas3 and type II toxin–antitoxin system genes at 100 mmol/L MDEA suggests active DNA repair and stress defense responses. , The upregulation of toxin–antitoxin system genes further suggests a regulatory mechanism that may modulate cellular metabolism in response to prolonged MDEA exposure.

The increased expression of stress response genes and inorganic metal ion transporters suggests that MDEA exposure triggers general cellular stress responses, including responses to oxidative and ionic stress, potentially leading to damage to essential proteins and nucleic acids. , The observed regulatory shifts toward metal homeostasis mechanisms suggest that the organism may be attempting to mitigate MDEA-induced toxicity by sequestering or redistributing metal ions required for enzymatic function and cellular stability.

The upregulation of genes related to the cell wall of M. maripaludis under both 50 and 100 mmol/L MDEA suggests that MDEA increases energy demands for maintaining cell membrane integrity. Interestingly, the gene expression of these increased further at 100 mmol/L MDEA, except for the S-layer protein. Furthermore, the upregulation of the UDP-N-acetylglucosamine 2-epimerase gene and two unknown membrane-associated genes at 100 mmol/L MDEA suggests a heightened response to increased toxicity, likely supporting the maintenance of their structural integrity. The absence of pronounced morphological alterations in M. maripaludis, as observed in SEM images, may be linked to the upregulation of genes involved in membrane structure, which could support the prevention of immediate physical damage to the cell. This response differs from that of M. barkeri, which exhibited distinct alterations in cell morphology upon MDEA exposure, and such a response is consistent with differences in membrane composition and energy metabolism between the two methanogens, indicating that survival strategies under MDEA stress may occur at the molecular or physiological level. However, SEM observations alone are insufficient to infer membrane disruption or changes in cell permeability and are therefore not interpreted as standalone evidence but are considered in conjunction with the pronounced inhibition of methane production and reduced growth under MDEA exposure. Future studies, including additional microscopy and imaging, such as live/dead viability staining, are needed to elucidate the specific effects of MDEA-induced stress on cellular integrity.

Considering limitations, our findings provide cell-level MDEA toxicity data and biological constraints on BICCU performance based on two trait-contrasting methanogenic genera studied under axenic batch conditions. The mechanistic hypothesis for MDEA-induced inhibition in M. barkeri is supported by methane production and scanning electron microscopy (SEM) observations, which are interpreted in the context of known differences in cell envelope architecture and energy conservation pathways between M. barkeri and M. maripaludis. By contrasting these two trait-selected model strains, our findings provide unprecedented, strain-resolved insight into methanogen responses to MDEA and represent a first step toward elucidating species-specific responses, while future studies will be needed to determine whether similar or different mechanisms apply across other methanogenic strains. Although responses to MDEA-induced toxicity may differ among microbes in mixed consortia, mechanistic insights into cell envelopes, methane metabolism, and energy-conservation pathways in the two methanogenic strains provide a basis for decoupling the direct and indirect chemical effects of MDEA on methanogens from community-level interactions in mixed consortia. In addition, the present work does not address sequential dynamics or long-term adaptation, which will require continuous reactor operation.

Perspectives for Combining MDEA in Biointegrated Carbon Capture

Conventional capture agents rely on amine scrubbing, such as alkanolamines, which are not designed for biocompatibility or CO₂ bioavailability. Compared to primary and secondary amines, tertiary amines such as MDEA form bicarbonate with high CO₂ loading (up to 1 mol_CO2 /mol_amine). The formed bicarbonate allows microbes to utilize captured CO₂ making MDEA a potentially suitable capture agent for BICCU into methane in terms of CO₂ bioavailability. However, the toxicity of the capture agent may still constrain methanogenic conversion, and its use will, according to the results presented here, depend on the methanogenic species employed. Our findings demonstrate that two methanogens, M. maripaludis and M. barkeri, exhibit different levels of susceptibility to MDEA-induced biotoxicity, with M. barkeri being more sensitive even at low MDEA concentrations. These pronounced differences could be associated with their distinct morphological and physiological characteristics. In addition, the concentration of the capture agent is vital. While MDEA provides a high CO₂ absorption capacity and bioavailable CO₂ release, its biotoxicity at concentrations above 50 mmol/L poses challenges for its integration into biomethanation processes. In particular, the poor survival of M. barkeri, even at MDEA concentrations tolerated by M. maripaludis, emphasizes the importance of selecting the ideal microbial species or consortia for BICCU systems. Thus, the findings highlight that the effectiveness of the process depends on three parameters: the methanogenic species, the capture agent employed, and its concentration. The challenge lies in balancing these three parameters. Therefore, achieving the right balance that supports both the physiological needs of microbes and CO₂ absorption efficiency will be essential for successfully integrating MDEA-based carbon capture with biological conversion processes.

One of the key challenges in this integration remains the avoidance of microbial toxicity, which can compromise system stability despite the promising chemical properties of capture agents. The molecular structure of MDEA, characterized by its two ethanol groups and relatively small size, may enable it to interact easily with microbial membranes, potentially disrupting their cellular integrity, as evidenced by the transcriptomic analysis and scanning electron microscopy. To address this mismatch between CO₂ capture liquids and microbes, future research should focus on designing less toxic capture agents and on identifying or engineering microbial catalysts with enhanced resistance and metabolic compatibility to the designed capture agents. Some methanogenic strains have shown the ability to adapt and develop resistance to such chemical stress , highlighting the importance of systematic screening and selection to identify optimized, resilient microbial partners for BICCU applications. The stress-responsive strategies activated in M. maripaludis may also serve as a basis for engineering more tolerant strains and guiding future screening efforts by indicating which methanogenic traits to prioritize. Additionally, understanding the long-term impacts of MDEA on mixed microbial communities in continuous reactor systems is crucial for the industrial-scale implementation of BICCU technology. By addressing these challenges, biointegrated carbon capture using methanogens could become a viable and sustainable solution to reduce industrial CO₂ emissions while producing renewable methane as an energy carrier.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.6c01434.

• Circos plot showing the genome-wide distribution of differentially expressed genes of M. maripaludis OS7 (Figure S1), sequencing read statistics (Table S1), and list of differentially expressed genes in M. maripaludis OS7 under exposure to 50 and 100 mmol/L MDEA with their predicted gene functions, differential expression, and COG classifications (Table S2) (PDF)

†.

Madhuni M. Wijesooriya and Mathias Eskildsen contributed equally to this work.

This work was supported by the Novo Nordisk Foundation grant number NNF21SA0072700, as part of the Novo Nordisk Foundation CO₂ Research Center (CORC).

The authors declare the following competing financial interest(s): Michael V. W. Kofoed has a patent (EP4555050A1) on bio-integrated carbon capture and utilization.