A blueprint for designing the next-generation of synthetic C₁ microbes

Abstract

Engineering microbes for synthetic one-carbon (C₁) assimilation continues to gain momentum with the expanding demand for sustainable bioprocesses. While most efforts focused on model microbes, non-canonical hosts offer untapped potential due to native metabolic properties, enzyme activities, and substrate tolerance. This perspective outlines key considerations for selecting and engineering such strains, including metabolic modeling, use of native C₁-inducible promoters, and adaptation to anaerobic conditions. Environmental impacts are evaluated through life cycle assessment, identifying substrates with low carbon footprints. Integrating techno-economic and sustainability insights at early stages is essential to guide the development of efficient, scalable C₁-based biomanufacturing systems.

Synthetic one-carbon assimilation could contribute to a more sustainable and circular carbon economy, but much work in this field has focused on model microorganisms. Here the authors provide their perspective on the potential value of non-model microbes, and how that potential could be realised.

Introduction

Recent interest in biomanufacturing focused on finding ways to render bioprocesses increasingly sustainable, especially regarding the choice of carbon substrates. Using sugar‑containing feedstocks (e.g., glucose or sucrose) for bioproduction has become a limiting factor as it competes with food production, taking away space from agriculture that could be used for human consumption or reforestation. In the quest for more sustainable feedstocks, various solutions have been explored, e.g., second-, third- and fourth-generation substrates¹. Of particular interest is the use of next-generation feedstocks, including one‑carbon (C₁) molecules that can be derived from or converted to CO₂, i.e., methanol, methane (CH₄), carbon monoxide (CO), and formate. In light of the increasing levels of greenhouse gases in the atmosphere, the ability to valorize CO₂ would make bioproduction more sustainable, alleviate the pressure of resource supplement, reduce industrial production costs, and—above all—promote a circular carbon economy^2,3. Not all microorganisms use C₁ molecules as a growth substrate; with notable exceptions that include photoautotrophs, e.g., microalgae and cyanobacteria, chemolithoautotrophs (e.g., acetogens and methanogens), and formatotrophic, methylotrophic, and methanotrophic bacteria². Some of these hosts have been adopted for bioproduction—e.g., cyanobacteria for biofuels⁴, acetogens for conversion of syngas into bioethanol, acetic acid and higher‑value chemicals⁵, and methylotrophic bacteria for single‑cell protein⁶. Useful as these microbes are, there is growing interest in engineering versatile, polytrophic microorganisms to efficiently utilize C₁ feedstocks. Here, we use the term polytrophs to refer to microorganisms that naturally grow on a wide variety of substrates (e.g., sugars, organic acids, or aromatic compounds), but do not typically utilize C₁ substrates as a carbon source. Such hosts can be specifically chosen for their native traits—e.g., tolerance to high substrate concentrations, ease of genetic manipulation, or robustness under bioprocess conditions—that are often limited in natural C₁‑trophs. This emerging research area is both challenging and exciting, requiring advanced metabolic engineering to reshape central carbon metabolism. While still in its early stages, this subfield already yielded promising outcomes^7–11, and holds enormous potential to leverage unique metabolic and physiological traits naturally found in polytrophic microorganisms, e.g., native C₁ processing reactions, stress resistance mechanisms, and metabolic flexibility that would be difficult to engineer from first principles. The literature indicates that most synthetic C₁ assimilation strategies focused on model organisms, e.g., Escherichia coli and Saccharomyces cerevisiae—understandably so, given the extensive genetic and engineering toolkits available for these organisms. Recently, the limited number of microorganisms chosen to engineer C₁‑trophy expanded to include Cupriavidus necator¹², Pseudomonas putida¹³, and Corynebacterium glutamicum¹⁴. Yet, this focus overlooks the immense metabolic diversity of the extensive microbial world.

From a broader perspective, despite current efforts, titers, and yields of microbial bioproduction from C₁ compounds are still too low for supporting large-scale applications and widespread economic feasibility, especially when substantial metabolic engineering efforts are still needed. Some notable exceptions include the relatively few bioprocesses adopting native C₁-trophs that have achieved commercial success in specific niches, e.g., single-cell protein production by methylotrophs and syngas conversion by acetogens^15,16. We argue that expanding metabolic engineering efforts toward novel, non-model microorganisms metabolically and physiologically primed for synthetic C₁ assimilation will enhance bioproduction efficiency. Our aim is to highlight synthetic C₁ assimilation in non-natural C₁-trophic bacteria, with the potential to unlock new capabilities and provide additional tools for sustainable bioproduction. This perspective complements the proven value of native C₁-trophs in C₁-based biomanufacturing, which have demonstrated commercial success in some application areas. We provide a practical guideline to engineer polytrophs, considering the entire bioprocess pipeline (from the initial design phase to the sustainability and economical assessment toward commercialization), with emphasis on the criteria for choosing promising microbial hosts and the engineering tools best suited to enhance their performance.

Non-model microbial strains for C₁ assimilation

Bioprocess design

Synthetic C₁ assimilation and bioproduction remain challenging even in microbial hosts with well-characterized metabolic, physiological and adaptive traits. In contrast, non-model hosts are often treated as biological “black boxes” due to the relatively limited information available compared to well-established model microorganisms. Therefore, implementing standardized workflows is essential not only for metabolic modifications but also for the entire bioprocess framework. Adopting non-model organisms for C₁-based bioproduction requires planning across all stages of the engineering workflow (Fig. 1). This effort includes evaluating the substrate and target product(s), fermentation parameters, oxygen (O₂) requirement, bioreactor type, downstream processing, choice of host and pathways, sustainability metrics, cost-effectiveness, and compliance with regulatory and safety standards. Embracing a goal-oriented design mindset (“beginning with the end in mind”) is particularly valuable for this daunting task, considering the stark differences between laboratory setups and industrial-scale fermentations. Although this article will not address each of these factors in depth, we emphasize that their interplay shapes the three core components of the entire roadmap: strain selection, metabolic design and engineering, and scale-up and fermentation optimization (Fig. 1).

Fig. 1. Blueprint for designing next-generation synthetic C₁ microbes.

The essential steps are interconnected, beginning with [1] the bioprocess design, emphasizing the choice of C₁ feedstock and target product; [2] the selection of a suitable microbial strain based on traits advantageous for engineered C1 assimilation; followed and connected by strain engineering and bioprocess optimization. Finally, assessing the sustainability of the overall process [3] determines whether adjustments or refinements are needed, potentially requiring a return to earlier steps [1 and 2].

C₁ compounds could support a carbon-neutral bioeconomy, provided they originate from renewable sources (e.g., CO₂ electroreduction or photocatalysis)^17,18. However, the efficiency of these technologies remains limited, contributing to prohibitively high market prices. As a result, the majority of reduced C₁ substrates currently available depend on fossil-derived feedstocks^1,19. Even when generated from CO₂ or syngas using renewable energy, selecting a suitable carbon source for microbial growth and production is not straightforward. Gaseous C₁ compounds—e.g., CO₂, CO, and CH₄—suffer from low solubility, restricting their use in fermentation. Despite the notably higher global warming potential of CH₄ compared to CO₂, it has often been excluded from synthetic C₁ assimilation strategies due to challenges in mass transfer and safety risks^20,21. CO₂, in turn, requires an additional electron donor, e.g., hydrogen (H₂), formate, or methanol, to serve as a reducing equivalent for carbon assimilation and energy conservation^22–24. While green H₂ has potential to serve as an external energy carrier, its production is too energy-intensive and costly for widespread use in bioproduction, especially for low-value commodities^1,17. In contrast, interest has grown around reduced liquid C₁ compounds (i.e., methanol and formate), as they are water-soluble and circumvent gas-liquid mass transfer constraints^25,26. Despite methanol’s drawbacks—i.e., flammability, toxicity, and volatility—this water-soluble carrier remains among the most promising candidates for enabling carbon-neutral bioprocesses. Formate is generally less toxic (compared to methanol’s oxidation to formaldehyde) and benefits from high electrochemical conversion efficiencies²⁷. Still, its high oxidation state means that most of the formate must be catabolized for energy conservation, leading to CO₂ re-emission and reduced carbon efficiency. Hence, the source and nature of the C₁ feedstock must be considered as early as possible to ensure alignment with sustainability goals.

In parallel, local accessibility of the chosen substrate is desirable, as transportation may considerably increase the bioprocess’ carbon footprint. The cost of the substrate, in conjunction with the market value of the target product(s), is a determinant of the bioprocess feasibility, and estimating the product yield required to reach economic viability accelerates the design of microbial strains and pathways. Importantly, preliminary (ex ante) techno-economic analysis (TEA) and environmental evaluations (e.g., life cycle assessment, LCA)—though frequently overlooked—are crucial from the outset to guide engineering efforts²⁸. A more detailed analysis of these elements is presented later in the article.

Hence, we posit that the roadmap for synthetic C₁ metabolism should start by considering the bioprocess context—including carbon and energy inputs, target molecule(s) and candidate pathways (either native or synthetic) for substrate assimilation, and product synthesis (Fig. 1). Furthermore, required titers and yields, alongside initial economic and sustainability benchmarks, should shape the subsequent steps in the design and execution of the workflow.

Strain selection

The selection of a microbial host for C₁-based bioproduction is largely determined by the specific bioprocess demands. For example, the fermentation mode dictates O₂ availability in the bioreactor—anoxic or micro-oxic conditions largely preclude the use of obligate aerobes, and the reverse holds true for anoxic bioprocesses. However, other less obvious considerations are equally critical and are addressed in this section.

Non-model organisms often remain insufficiently characterized, particularly in terms of their metabolic network architecture and regulation. Yet, omics-driven profiling—through metabolomics, fluxomics, transcriptomics, and proteomics—can offer insights into central carbon fluxes with different feedstocks (Fig. 2)²⁹. This knowledge helps determine whether the proposed assimilation routes intersect native metabolic fluxes, relevant in the presence of competing pathways³⁰. For example, native routes, e.g., the serine cycle or the reductive tricarboxylic acid (TCA) cycle, reverse the carbon flow direction through the TCA cycle either partially or entirely³¹. Such metabolic conflicts, although surmountable, pose substantial engineering challenges³². Linear and orthogonal pathways with high flux potential, e.g., the reductive glycine pathway (rGlyP), are typically simpler to implement^33–35. In contrast, circular, autocatalytic cycles offer theoretical benefits due to their topology but require tight control at branch points to prevent intermediate depletion or “bleeding”^25,36. Mapping metabolic fluxes and understanding how the host regulates central metabolism supports both pathway selection and rational strain engineering (Fig. 2).

Fig. 2. A roadmap for efficient engineering of C₁ assimilation in P. putida, progressing from a non-model organism to an autotrophic chassis, adopted here as a case study.

The roadmap outlines key strategies pursued to overcome the challenges inherent to this complex engineering task. These include detailed strain characterization, comprehensive multi-omics analyses performed with both sugar-based and C₁ feedstocks, and the development of a diverse suite of synthetic biology tools. The diagram also highlights ongoing efforts and future directions (indicated in orange) needed to establish robust bioproduction platforms from C₁ feedstocks. WGS whole genome sequencing, CCM central carbon metabolism, GSMM genome-scale metabolic models, EDEMP cycle Entner–Doudoroff–Embden–Meyerhof–Parnas–pentose phosphate cycle, rGlyP reductive glycine pathway, TRYE titers, rates, yields, and emissions. Key (not exhaustive) references are provided as follows: isolation: ref. ¹¹²; WGS: refs. ^{113, 114}; GSMM: ref. ¹¹⁵; EDEMP cycle: ref. ⁷⁰; RNASeq on C₁ substrates: refs. ^{50, 52}; pGNW·pQURE system: ref. ¹¹⁶; CIFR: ref. ¹¹⁷; rGlyP: refs. ^{13, 118}; Serine cycles: ref. ³⁸.

Computational integration of omic data into metabolic models provides further guidance for host and pathway selection strategies. Common modeling tools include flux balance analysis (FBA), enzyme cost minimization (ECM), and minimum-maximum driving force (MDF) models^37,38. FBA predicts steady-state flux distributions that optimize objectives (e.g., biomass formation), assessing both compatibility and energy balance³⁴. ECM estimates optimal enzyme and metabolite concentrations supporting a desired flux distribution while minimizing protein investment, whereas MDF identifies pathways with the highest thermodynamic driving forces, ensuring robust energy profiles^36,39,40. While exploring novel C₁ routes is important, we advise to first evaluate pathways with demonstrated high flux potential to improve the chances of successful implementation in non-model hosts. These include the rGlyP, the ribulose and xylulose monophosphate (RuMP and XuMP) cycles, native and synthetic versions of the serine cycle, and the Calvin–Benson–Bassham (CBB) cycle^41–43. The natural CBB cycle has been implemented in Komagataella phaffii⁴⁴, highlighting its potential applicability beyond traditional systems. Emerging synthetic pathways incorporating new-to-nature enzymes, e.g., formolase⁴⁵ and glycolaldehyde synthase⁴⁶, have shown some promise for C₁ assimilation; however, in vivo implementation remains quite challenging due to the enzymes’ high K_M values and the strict requirement for elevated formaldehyde concentrations that are largely incompatible with most microbial hosts⁴⁷. Ongoing enzyme engineering efforts aim to address these limitations, yet substantial improvements in kinetic and affinity parameters are still needed before these pathways can be applied effectively in microbial systems.

Strain selection must also consider the organism’s endogenous C₁ metabolism—particularly the presence of C₁-relevant enzymes, e.g., carboxylases, carboligases, oxidases, and dehydrogenases—along with substrate and intermediate tolerance. Phosphoenolpyruvate carboxylase (Ppc), widely found across bacterial species, supports gluconeogenic fluxes as a crucial step in the serine cycle⁴⁸. The glycine cleavage system operates counter to the rGlyP in its native configuration, but can be coupled to serine hydroxymethyltransferase to convert tetrahydrofolate-bound carbon moieties into L-serine³⁸. Hence, overlap between the host’s metabolism and synthetic pathways can simplify design efforts when flux distribution and regulatory interference are kept at a minimum.

Tolerance to reduced C₁ compounds is another desirable host trait, as these substrates or their intermediates are often toxic^49–51. Formaldehyde is detrimental at low concentrations (low millimolar range), followed by formate [in the hundred(s) millimolar range] and methanol (molar range)^49–52. Detoxification is generally mediated by oxidation, which yields electrons for growth. However, when the same compound serves as both carbon and energy source, the balance between oxidation and assimilation becomes critical. High tolerance to formaldehyde or formate often correlates with reduced carbon assimilation due to preferential oxidation. Methanol tolerance is especially nuanced: strains with high oxidation capacity frequently produce large amounts of formaldehyde, compromising C₁ assimilation if the aldehyde is not efficiently captured by downstream reactions⁵⁰. Native methylotrophs generally possess complex formaldehyde-sensing regulation systems and tolerate relatively high concentrations of this toxic compound (up to 5-8 mM formaldehyde for Methylobacterium extorquens AM1 or K. phaffii⁵³, compared to ≤ 1 mM of E. coli⁵⁴ or 1–2 mM in P. putida⁵⁰). When comparing heterotrophic organisms, P. putida seems to be less tolerant to methanol than E. coli, due to substantial alcohol oxidation and high energy yield from methanol, rendering formaldehyde readily available for assimilation if efficient pathways are incorporated⁵⁰. P. putida encodes several alcohol dehydrogenases active on methanol that use either pyrroloquinoline quinone (PQQ) or NAD(P)⁺ as cofactors⁵⁵. PQQ-dependent methanol dehydrogenases (MeDHs) are promising for synthetic methylotrophy, as they offer superior thermodynamic and kinetic properties compared to NAD(P)⁺-dependent enzymes²⁶. It should be mentioned, however, that PQQ-dependent Mdhs require O₂ to support PQQ biosynthesis, thus their use is largely limited to aerobic applications. When the host of choice possesses a functional PQQ biosynthesis pathway, engineering efficient PQQ-MeDHs becomes feasible—even if native MeDHs are absent²⁴. While most Pseudomonas species produce PQQ, E. coli does not (Box 1)^24,56. Furthermore, tolerance to C₁ molecules should extend to longer-chain intermediates, e.g., glycolaldehyde or hydroxypyruvate. These reactive molecules arise in synthetic routes, e.g., the synthetic acetyl-coenzyme A (CoA) pathway and the synthetic variants of the serine cycle. Hence, screening for tolerance across multiple toxic metabolites is essential when selecting among non-model organisms.

Lastly, successful strain deployment requires a versatile and easy-to-deploy set of genetic and synthetic biology tools. These include functional promoters (both constitutive and inducible), reliable plasmid and genomic expression systems, codon optimization or harmonization protocols, and the capacity to perform precision genomic edits, including large deletions and insertions⁵⁷. Additional tools that allow knock-downs, base-editing, or in vivo mutagenesis—whether targeted or random—are highly beneficial. In summary, identifying a microbial host with known C₁ metabolic features and compound tolerance, capable of sustaining one or more C₁ assimilation pathways for the desired product(s) and amenable to genetic and metabolic modifications using existing synthetic biology toolkits (Figs. 1 and 2) is the second key step of this roadmap.

Box 1 P. putida as a case study for synthetic C₁ metabolism: from non-canonical polytroph to C₁ assimilator.

P. putida is a versatile organism that has gained attention for biotechnological purposes owing to its increased tolerance to several physicochemical stresses, flexible metabolism, and potential for bioproduction of complex molecules. P. putida has been characterized to display increased C₁ oxidation capacity, connected to the presence of PQQ-dependent alcohol dehydrogenases that act on methanol. Furthermore, this non-pathogenic bacterium is endowed with an inherent potential to hosting the profound metabolic rearrangements that support C₁ assimilation. Figure 2 outlines a roadmap with the key essential steps and methodological foundation (gathered over the past two decades) that enabled synthetic C₁ metabolism in this host. Even though P. putida has been adopted as a prototypical bacterium for some biotechnological approaches (e.g., bioremediation), it is seldom regarded as a model for engineering C₁-trophy pathways despite all its merits. The extensive physiological and metabolic studies across biotechnological contexts conducted in the past years have hardly included its worth as a host for synthetic C₁-trophy relative to more traditional microbes, e.g., E. coli and S. cerevisiae. Synthetic biology tool development and a growing body of multi-omic data have recently enabled engineering synthetic serine cycles¹³ and the rGlyP¹¹⁸ under mixotrophic and autotrophic conditions, respectively. These C₁ assimilation engineering efforts underscore the importance of a priori strain characterization, undertaking multi-omic analyses, and developing efficient metabolic and genetic engineering tools to maximize chances of success (Fig. 2). Further steps, besides the examples included in this article, must further characterize the minimal set of mutations required for autotrophic growth for a specific C₁ pathway and, most importantly, promote robust culturing potential for bioproduction at high volumetric capacities (see Bioproduction by Engineered C₁-Trophic Strains).

Engineering non-model organisms for C₁ assimilation

Several reviews address metabolic engineering strategies for implementing C₁-trophy^58–60, including a recent account of the range of genetic and metabolic tools applicable to this goal⁶¹. Readers are encouraged to consult these sources for deeper guidance on specific steps, e.g., growth-coupled selection, adaptive laboratory evolution (ALE), hypermutagenesis protocols, preparation of expression libraries, and implementation of isotopic labeling. To illustrate how this roadmap can be deployed to a heterotrophic host, we present our efforts to engineer P. putida for C₁ assimilation as a case study (Box 1). We also place the focus on discussing strategies that remain largely underexplored in the context of synthetic C₁ assimilation, which includes the use of inducible promoters responsive to C₁ compounds, combinatorial pathway design, and omics-informed identification of metabolic bottlenecks.

Inducible promoters sensing C₁ substrates and intermediates

Current synthetic efforts predominantly rely on constitutive promoters to modulate expression of both native and heterologous genes. These strategies have enabled the identification of pathway limitations at the transcriptional level, evaluation of enzyme variants, and evolution-driven determination of genetic modifications required for establishing C₁-trophy^{13,16,35,41,62}. Nevertheless, specific growth rates (μ) and production yields remain consistently low, with only a few recent exceptions—e.g., implementing the rGlyP in C. necator³⁵ or the RuMP cycle in E. coli^16,63. In C. necator, for instance, protein levels from the rGlyP were downregulated, which unexpectedly improved yields compared to the native CBB cycle³⁵. Still, the engineered strain grew at nearly half the μ of the wild-type bacterium³⁵. In the case of E. coli, the RuMP cycle enabled doubling times of 3.5 h under optimal conditions after evolution—surpassing some native methylotrophs⁶³. This engineering goal was aided by a copy number variation strategy that mitigated formaldehyde-induced DNA-protein cross-linking (DPC)⁶³. Despite these advances, to the authors’ knowledge, none of the current synthetic C₁-trophic systems have produced commercially viable titers of added-value products¹⁶. Major gaps remain in our understanding of how organisms adapt to synthetic pathways or the extent of their plasticity in response to such rewiring. In nature, microbes adjust their metabolic programs based on external cues, enabling efficient resource use and avoiding the accumulation of toxic intermediates. We argue that synthetic C₁ assimilation must move beyond static expression designs and integrate dynamic regulatory systems, either leveraging native inducible elements or engineered regulators derived from other organisms.

Several inducible regulatory systems responsive to C₁ substrates have been identified in C. necator, E. coli, and P. putida, including the FdsR-P_fdsR, FrmR-P_frm, and EfgA systems^{38,50,54,64–67}. For instance, a formaldehyde biosensor was introduced in E. coli to monitor optimal expression levels of MeDH genes and the formaldehyde-assimilation modules within the RuMP cycle, effectively minimizing formaldehyde build-up⁶⁵. However, these genetic modules were expressed under control of IPTG- and tetracycline-inducible systems, respectively⁶⁵. In follow-up work, the FrmR-P_frm system was used to enhance regeneration of ribulose-5-phosphate via the pentose phosphate pathway, supporting methanol assimilation under mixotrophic conditions⁶⁴. Still, the inducible system was not systematically benchmarked against constitutive expression of the same genes⁶⁴. In P. putida, overexpression of regulator genes associated with the ped cluster led to increased activity of native PQQ-dependent MeDHs, though regulatory mechanisms governing the cluster remain poorly understood³⁸. In addition to bacterial systems, eukaryotic methylotrophs (e.g., K. phaffii) have also evolved highly regulated C₁-responsive promoters. The alcohol oxidase I (AOX1) promoter, strongly induced by methanol and tightly repressed in the presence of other carbon sources, has been extensively used in both native and heterologous contexts to achieve methanol-inducible gene expression⁶⁸. To date, no synthetic C₁-trophic strain employs a C₁-responsive inducible device as its core gene expression system.

Tight regulation of assimilation pathway genes is a defining trait of native C₁-utilizing organisms. In M. extorquens AM1, a model methylotroph, not only is the assimilatory network dynamically regulated, but formaldehyde-induced translation arrest also serves as a protective mechanism against DPCs and metabolic waste^48,54. In addition to C₁ substrates and their oxidized derivatives, intermediate metabolites from specific pathways may serve as regulatory nodes, especially pertinent for CO₂-fixation routes⁶⁹. Dynamic control systems have been largely absent from synthetic C₁-trophic designs, despite their potential to alleviate formaldehyde toxicity, reduce metabolic burden and improve resource allocation—factors that are critical under bioproduction conditions.

Combinatorial assimilatory pathways

Glucose assimilation may proceed through its direct phosphorylation to glucose-6-phosphate or via oxidation to derivatives such as gluconate⁷⁰. These intermediates feed into central metabolism through the Embden–Meyerhof–Parnas, Entner–Doudoroff or pentose phosphate pathways⁷⁰. Effective coordination across these routes requires multilayered and finely tuned regulation. We propose that C₁ assimilation need not be restricted to a single, discrete biochemical sequence. Rather, a modular combination of multiple C₁-assimilatory routes—when advantageous to the host—can be harnessed simultaneously. Formaldehyde-assimilating pathways must compete with detoxification systems that scavenge electrons while converting formaldehyde into formate. When this balance is disrupted, carbon flux may be diverted inefficiently, leading to formate accumulation or complete oxidation to CO₂, both of which diminish substrate utilization and can introduce toxicity.

We demonstrated that combining synthetic variants of the serine cycle in P. putida led to improved growth metrics relative to the individual parental pathways³⁸. This configuration, termed enhanced serine-threonine cycle (STC), incorporated the serine aldolase reaction to expand formaldehyde-derived C₁ assimilation, while maintaining native formate assimilation through the tetrahydrofolate branch of the STC³⁸. Similarly, recent studies showed that introducing the bacterial RuMP cycle into K. phaffii, which natively runs a XuMP cycle, boosted erythritol production and reduced by-product generation⁷¹. While expressing multiple pathway genes may impose an additional metabolic load on the cell, we argue that the physiological gains—when coupled with precise, dynamic regulation—of combinatorial pathway integration could enhance the performance of engineered synthetic C₁-trophs.

Omics-guided bottleneck identification

Even when incorporating all previously discussed strategies to enable C₁-trophy, engineering such a metabolic configuration demands extensive proteome reallocation within the host cell. Consequently, complete autotrophy typically arises only after extended ALE campaigns—if at all. ALE enables the progressive emergence of a desired phenotype through gradual genetic changes, circumventing the need for multiple mutations to occur simultaneously⁷². Conversely, if extensive and simultaneous genomic alterations are required for achieving methylotrophy, the likelihood of ALE success diminishes⁷³. Systems-level analyses, informed by omic technologies and executed with tailored synthetic biology toolsets^74,75, can assist in identifying bottlenecks that hinder the evolutionary progression toward synthetic autotrophy.

Comparing ECM predictions—i.e., protein demand per reaction to sustain desired fluxes—with proteomics datasets can reveal limiting enzyme levels, whether from the host or within a synthetic pathway. Isotopic labeling, fluxomics and metabolomics further enable the localization of pathway bottlenecks, particularly at branching points where flux competition may occur. Additionally, omic-guided assessments can help identify co-substrates that replicate flux patterns of the desired phenotype, accelerating the development of a functional C₁-assimilating strain. In the case of rGlyP engineering in C. necator, proteomics revealed a downregulation across all constitutively-expressed modules required for growth on formate, indicating a misalignment between expression levels and metabolic demand³⁵.

Finally, deregulated synthetic systems often produce and excrete by-products as a result of unresolved metabolic constraints, which can impair growth and product formation⁷⁶. Analyzing the secretome of engineered strains can thus uncover specific failure points, enabling targeted interventions. Collectively, these high-throughput datasets inform and refine engineering while identifying interventions to enhance the effectiveness of ALE in driving the evolution of an autotrophic phenotype.

Engineering O₂-dependent lifestyles

As noted in the bioprocess design, O₂ availability during fermentation must be defined from the outset. Anaerobic C₁ assimilation pathways, including the ancestral Wood–Ljungdahl pathway and the reductive TCA cycle, involve O₂-sensitive enzymes, e.g., CO dehydrogenase and ferredoxin-dependent carboxylases⁷⁷. Anoxic growth inherently links biomass formation to by-product synthesis⁷⁸. For instance, acetogenic bacteria reduce part of the carbon substrate to acetate to dissipate excess reducing equivalents. Acetogenesis through the Wood–Ljungdahl pathway operates near the thermodynamic limit of life, which ensures highly efficient C₁-to-product conversion^79,80. However, transferring such systems into alternative hosts remains largely unexplored due to obstacles that include energy conservation, cofactor redox balance and proper maturation of O₂-sensitive enzymes. Native acetogens currently represent the only industrial-scale non-phototrophic C₁-trophic systems for CO₂ fixation.

Engineering obligate aerobes to function under anaerobic conditions has proven challenging⁸¹. Recently, however, E. coli was engineered to perform oxic fermentation (respirofermentation)⁸². The required deletion of genes encoding seven NAD(P)⁺/quinone oxidoreductases and twelve quinone-linked dehydrogenases to redirect metabolism toward a fermentative regime enabled lactate and isobutanol production under oxic conditions. To mitigate redox imbalance, specific respiratory modules were introduced to allow controlled use of O₂ for redox balancing⁸². Such seminal work demonstrates that fermentative metabolism can be imposed under O₂-rich environments through extensive rewiring, bypassing the limitations of anaerobiosis while supporting robust product formation—including from C₁ substrates. Although the product portfolio from anaerobic C₁ assimilation remains relatively narrow due to energetic constraints, synthetic and native systems offer promising routes for the sustainable synthesis of platform chemicals^83–85.

An emerging strategy has further pushed the boundaries of anoxic limitations by enabling CO₂ reduction to formate under oxic conditions and at atmospheric CO₂ levels—reactions typically exclusive to anaerobes using the Wood–Ljungdahl route or the rGlyP⁸⁶. This goal was achieved through the so-called CORE pathway, which harnesses O₂-tolerant CO₂ reduction via a novel reaction catalyzed by a β-ketoacid cleavage enzyme, requiring only one molecule of ATP and NADPH each⁸⁶. This case exemplifies how pathway conceptualization combined with the discovery or engineering of enzymes for new-to-nature chemistries can unlock biochemical transformations traditionally restricted to anoxic conditions.

How biofoundries can accelerate C₁ metabolic engineering

Biofoundries are advanced, automated laboratories that integrate robotics, synthetic biology and computational tools to accelerate biological engineering⁸⁷. These platforms enable high-throughput design, construction and testing of engineered strains by automating critical steps, e.g., DNA assembly, strain development, ALE, phenotype screening, and multi-omic data integration. Each stage of the workflow discussed in this perspective can be supported by biofoundry capabilities, creating iterative feedback loops to enhance overall process efficiency. Additionally, these platforms can incorporate artificial intelligence and machine learning tools to predict and optimize fluxes, as well as to engineer enzymes with improved activity or stability for enhanced pathway performance⁸⁸. The use of biofoundries in both academic and industrial settings can accelerate the development of microbial strains for sustainable C₁-based bioproduction, while also reducing associated time and costs. Fulfilling this milestone is particularly relevant when selecting novel hosts for C₁ engineering, capable of growing on C₁ substrates and sufficiently robust to support bioproduction at competitive performance levels.

Bioproduction by engineered c₁-trophic strains

Achieving efficient bioproduction from C₁ substrates remains a challenge, as microbial systems often prioritize biomass over product formation under these conditions, necessitating precise resource allocation. Hence, we posit that aligning the target compound(s) with the chosen assimilation pathway—or vice versa—is critical for synthetic C₁-based manufacturing. For instance, C₃-yielding pathways, e.g., the rGlyP and the RuMP cycle, are optimal for producing lactate, whereas acetyl-CoA–based routes, e.g., the serine cycle, are more suitable for generating acetyl-CoA-derived products, e.g., poly(3-hydroxybutyrate). While examples of synthetic C₁-trophic bioproduction remain limited, several studies stand out as proof-of-principle of what is possible. In one case, engineering E. coli with the RuMP cycle followed by ALE enabled the strain to reach an optical density of ~100 units in fed-batch cultivations, producing lactate at ~3 mM^10,16. Beyond strictly C₁-trophic systems, C₁ compounds could serve as co-substrates to enhance redox balance and yields. For example, an engineered E. coli using methanol alongside establishing a synthetic RuMP cycle achieved a yield of succinic acid on biomass of 0.98 g g^–1, 0.5 mM D-allulose per 1 mM methanol with xylose co-feeding, and 2 g L^–1 n-butanol and 4.6 g L^–1 ethanol from methanol and xylose^89–91. Similarly, C. glutamicum engineered with the RuMP cycle improved cadaverine production when fed ribose and methanol⁹². These examples stress the value of targeting C₃-C₄ molecules, relevant for downstream polymerization into carbon-storing materials⁹³. In principle, C₁-based microbial processes for materials and protein production are economically viable⁹⁴. Production of bulk fuels and chemicals may be more feasible using native C₁-trophs—whereas synthetic systems may prove more competitive for high-value molecules.

Metabolic engineering strategies developed for traditional systems can be extended to C₁-based production, and a growing body of literature provides valuable guidance in this area^95–97. A challenge particular to C₁ utilization is minimizing by-product formation and feedstock over-oxidation, both essential for improving carbon efficiency. Engineering microbial metabolism for C₁ assimilation will most certainly perturb native flux distributions, often prompting the cell to generate unwanted by-products to restore carbon and redox homeostasis. These side products drain mass and energy, and identifying and redirecting these fluxes is key for increasing yields. At the same time, by-products may present opportunities for valorization by selecting final products that can derive from them.

Also, translating laboratory success to industrial settings is a key (and challenging) step in advancing C₁-based biomanufacturing. Some natural C₁-trophic systems have already achieved relatively high technological readiness, e.g., the pilot-scale gas fermentation of acetone and isopropanol using engineered acetogens⁹⁸. These advances show that C₁-driven processes can be scaled, provided that engineered strains are integrated into robust bioreactor systems. Scale-up efforts require addressing general process variables, e.g., nutrient balance, pH control, feeding strategy, gas solubility and substrate delivery. However, in C₁ fermentations, additional emphasis must be placed on securing a cost-effective, sustainable feedstock supply⁹⁹—particularly when using industrial off-gases or waste streams—and on developing robust, stable strains with fine-tuned flux control (Fig. 1). Dynamic regulation and predictive flux modeling are particularly valuable for managing carbon assimilation and optimizing yields, as previously discussed.

To guide feedstock selection, in silico simulations can offer insights into strain performance across various C₁ substrates. As an example, we conducted a genome-scale metabolic modeling analysis to evaluate the sustainability of different carbon sources, comparing predicted carbon emissions to produce lactate and 3-hydroxybutyrate using glucose, acetate (e.g., from acetogenesis) and reduced, water-soluble C₁ substrates (methanol and formate). Our simulations included wild-type P. putida and strains engineered with the rGlyP, RuMP cycle, homoserine cycle (HSC) or STC. Methanol yielded the highest production rates and the least CO₂ emissions, followed by glucose and acetate—with acetate resulting in the lowest metabolic activity and largest CO₂ release. Formate performed poorly due to its extensive oxidation (~70–80%) solely to supply energy. Notably, under maximum production rates without biomass formation, synthetic methylotrophy could become net carbon-negative—i.e., capturing CO₂ while producing the desired compound—depending on the oxidation state of the final product (Table 1). Thus, in terms of both carbon efficiency and sustainability, reduced substrates (e.g., methanol) may outperform more oxidized alternatives including acetate and formate (Fig. 3). While in silico modeling highlights the theoretical advantages of certain synthetic pathways and feedstocks, translating them into functioning in vivo systems presents additional challenges. Physiological and metabolic limitations (e.g., redox imbalance), cofactor availability and cycling, and cellular stress often reduce the overall pathway efficiency, as observed in several recent implementations of synthetic C₁ assimilation^100,101.

Table 1.

Metabolic flux distribution of C₁-assimilating pathways compared to wild-type P. putida grown on glucose or acetate as the main substrate

Condition and pathway	WT		HSC	rGlyPb		RuMP cycle	STC		Comments
Carbon source	Glucose	Acetate	Methanol	Methanol	Formate	Methanol	Methanol	Formate
Product: L-Lactate
qPmax (mmol gCDW–1 h–1)	11	9	11	14	4	12	9	3	No growth
Max. emitted CO2 (%)a	33.66	46.17	23.09	16.52	72.44	6.81	39.36	79.99	No production
Min. emitted CO2 (%)a	3.02	25.01	4.11	–22.47	61.32	–2.62	23.12	74.57	No growth
µmax(h–1)	0.59	0.48	0.69	0.75	0.25	0.83	0.54	0.18	No production
Product: 3-hydroxybutyrate [poly(3-hydroxybutyrate) monomer]
qPmax (mmol gCDW–1 h–1)	7	6	9	9	3	8	7	2	No growth
Max. emitted CO2 (%)a	33.66	46.17	23.13	16.52	72.44	6.81	39.36	79.99	No production
Min. emitted CO2 (%)a	22.11	26.54	–3.11	–3.38	65.41	1.19	19.71	74.37	No growth
µmax(h–1)	0.59	0.48	0.69	0.75	0.25	0.83	0.54	0.18	No production

Carbon yields, growth and production rates were determined using flux balance analysis (FBA) for production of L-lactate or 3-hydroxybutyrate. FBA was performed using an updated version³⁸ of the P. putida metabolic model iJN1463 that harbors the C₁ assimilatory pathways indicated.

Maximum specific production rate (q_P^max), maximum specific growth rate (μ_max), maximum (Max.), minimum (Min.), homoserine cycle (HSC), reductive glycine pathway (rGlyP), ribulose monophosphate (RuMP) cycle, serine-threonine cycle (STC), and wild-type (WT).

^aPercentage of CO₂ emissions normalized on a C-mol basis (i.e., carbon yield).

^bThe reductive glycine pathway requires a CO₂-enriched atmosphere to support microbial growth.

Fig. 3. Circular carbon economy based on phasing out fossil fuels and revalorization of CO₂ waste streams through a reduced C₁ bioeconomy, or acetate derived from anaerobic acetogenesis using syngas or green H₂.

Emitted CO₂ and product yields were predicted using genome-scale metabolic modeling and are represented as traffic lights, with green indicating increased suitability, yellow intermediate, and red disadvantageous properties. Glucose was used as model sugar for predictions, though other sugar feedstocks are expected to yield comparable outcomes. Biomass waste, bio-based recycling and upcycling of plastics or other highly recalcitrant carbon sources are also envisioned as substrates for future carbon-neutral bioprocesses.

Hybrid chemical-biological systems are being explored to integrate synthetic C₁ assimilation with industrial operations. Power-to-X, for instance, couples renewable electricity-driven CO₂ electro-reduction with microbial fermentation. Another strategy involves carbon capture and utilization, feeding CO₂ from flue gases into bioreactors for its transformation into value-added products. These approaches introduce additional engineering challenges, including gas–liquid transfer efficiency and cost-effective media design—e.g., incorporating renewable ammonia (NH₃) as a nitrogen source¹⁰². From a sustainability perspective, these integrated operations offer promises by enabling conversion of renewable electricity and captured CO₂ into added-value compounds, supporting a truly circular carbon economy (Fig. 3).

Integrated sustainability assessment

When designing C₁-based bioprocesses, the choice of feedstock is a pivotal factor. While cost and local availability remain critical constraints, the environmental performance of the feedstock must also be weighed in the context of sustainable production. Next-generation C₁ feedstocks include CO₂, CO, CH₄, methanol, and formate, with C₂ compounds (e.g., ethanol and acetate) also relevant in broader applications¹⁸. CO₂ is abundant in the atmosphere and can be captured from industrial emissions. CO, commonly found in syngas, and CH₄, widely produced by agriculture and waste decomposition, can also be obtained renewably^103,104. Methanol production is feasible through bio-based or (electro)chemical methods, while formate can be derived from photo- or electrochemical CO₂ reduction¹⁰⁵. Although C₂ substrates are not the primary focus of this work, compounds such as acetate and ethylene glycol can also be synthesized from CO₂ through microbial or chemical electrosynthesis routes^106–108. Electrochemical CO₂ reduction to other feedstocks (e.g., CO, CH₄, methanol, and CH₂O₂) is a rapidly evolving technology with significant potential for enabling a circular carbon economy when powered by renewable energy. Despite increasing research and notable advances in CO₂ reduction, the scalability of these technologies depends on resolving high energy demands, improving catalyst durability, and overcoming economic limitations. Nonetheless, several companies are actively developing electrochemical CO₂ conversion platforms that may soon make renewable C₁ feedstocks accessible at the relevant scale, enhancing the viability of C₁-based biomanufacturing.

Environmental impact assessment throughout the lifecycle of a product or process is typically performed via LCA¹⁰⁹. LCA helps decision-makers identify environmental hotspots across all production stages. In parallel, TEA evaluates the technical and financial viability of a process, informing early-stage design and commercial feasibility. A recent review comprehensively compared traditional, second-, third-, fourth-generation, and C₁ feedstocks using both LCA and TEA methodologies¹. The authors reported that several C₁ substrates demonstrated better environmental and economic sustainability than sugar-based feedstocks. CO₂ was identified as the lowest-cost feedstock, though this advantage applies when derived from fossil sources, which compromises sustainability. Conversely, renewable CO₂-based substrates yield the lowest climate impacts but are more expensive, underscoring the need for policy and infrastructure support to drive down their costs and make truly sustainable options economically viable.

Currently, only a limited number of bioprocesses using C₁ substrates have reached sufficient maturity for LCA, which is expected given that LCA is typically performed in later development stages. However, assessing sustainability prior to scale-up is increasingly recognized as critical (Fig. 1). The scarcity of early-stage LCAs is due in part to their non-mandatory nature and the high costs involved, which may be prohibitive for startups. Nevertheless, growing demand for sustainable technologies and the role of LCA in supporting Environmental Product Declarations¹¹⁰—particularly under EU policy—are making these assessments more common and relevant. One notable example is LanzaTech’s use of the natural autotroph Clostridium autoethanogenum for gas fermentation⁹⁸. Their LCA compares the greenhouse gas emissions of acetone and isopropanol production from fossil sources (natural gas, oil and coal) to those produced using waste carbon emissions from industrial gases, municipal waste or biomass. The analysis, based on a gate-to-gate boundary, focuses on in-facility processes and considers biomass and biogas as co-products. The results showed net-negative emissions: –1.78 kg CO₂e per kg of acetone and –1.17 kg CO₂e per kg of isopropanol⁹⁸. These outcomes are driven by avoided emissions from waste gases and minimal process emissions, making the system effective at carbon capture. However, these carbon savings apply only to the production phase; emissions from product use or disposal remain unaccounted for. While bioproduction from C₁ substrates can be part of a carbon-negative strategy, such framing is incomplete without considering full cradle-to-grave emissions. Most C₁-derived products are eventually incinerated, degraded or landfilled, thus releasing carbon. Therefore, C₁ bioproduction is more accurately positioned within the framework of a circular carbon economy, where emissions are reabsorbed into the production cycle. True carbon-negative technologies must result in permanent atmospheric CO₂ removal, not merely delay or displace emissions.

TEA is the preferred tool to evaluate the economic viability of scale-up efforts. TEAs assess emerging technologies at low readiness levels, helping to shape investment and design strategies. In contrast to LCAs, TEAs prioritize cost metrics and commercial potential. As with LCAs, most available TEAs pertain to natural C₁-trophs due to the limited scale of synthetic systems. Several TEAs have analyzed algal biofuels, microbial electrosynthesis and gas fermentation. In this sense, the economic prospects of microbial electrosynthesis and gas fermentation were recently examined, identifying key cost barriers¹¹¹. The promise of these technologies is self-evident, particularly when integrated into existing industrial infrastructures such as steel mills, which can provide low-cost waste gases. However, achieving broader adoption will require significant cost reduction and technical improvement.

Together, LCA and TEA offer complementary information about C₁-based bioproduction. LCA confirms the environmental benefits of these systems, while TEA stresses the need for further development to ensure economic feasibility. Beyond environmental and economic considerations, intellectual property frameworks can also impact the advancement of synthetic C₁ platforms. Patent protections, licensing constraints and freedom-to-operate issues can shape the adoption of novel strains, pathways and tools. Social acceptance will likewise influence regulatory and market uptake. Public concerns about ecological safety and industrial impacts must be addressed through transparent communication that underscores environmental benefits and long-term sustainability. Engagement with policymakers, consumers, and communities is essential for building support.

Looking ahead, sustainability assessment frameworks must expand to include broader impact categories and more detailed scenario modeling. A promising addition is absolute sustainability assessment, which evaluates whether a process is sustainable within planetary boundaries. Unlike LCA, which enables relative comparisons, absolute assessments define thresholds beyond which systems become environmentally unsustainable. Adopting this methodology in C₁-based bioprocessing will be crucial to ensure genuine sustainability and avoid superficial claims of environmental benefit.

Outlook

Expanding synthetic C₁ assimilation to non-model microbial strains offers multiple advantages over traditional model systems. Leveraging the inherent metabolic versatility of these organisms—including their adaptability, tolerance to C₁ substrates and native enzymatic capabilities for C₁ processing—can enhance the efficiency and sustainability of microbial bioproduction. Despite persistent challenges such as limited genetic accessibility and incomplete metabolic characterization, these gaps represent fertile ground for future innovation. Progress in this field will depend on the continued development of genetic and computational tools tailored to non-model hosts, supported by systematic application of omics-informed metabolic engineering. Advancing TRY metrics (titer, rate, yield), alongside innovations in feedstock utilization and bioreactor design, will be essential for achieving economic viability in C₁-based biomanufacturing. Importantly, TRY optimization should incorporate emissions (E) to enable a TRYE framework—capturing both productivity and environmental performance. Policy mechanisms—carbon pricing, emissions credits and targeted investments in low-carbon technologies—will also play a central role in enabling large-scale deployment relative to carbon-intensive alternatives. Finally, early implementation of LCA and TEA will be key to informing microbial host selection, engineering priorities and bioprocess configurations aligned with both environmental and economic sustainability goals.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

G.F. and O.P.: Methodology; validation; formal analysis; investigation; visualization; writing – original draft. P.I.N.: Conceptualization; funding acquisition; project administration; supervision; writing – review and editing.

Peer review

Peer review information

Nature Communications thanks Matthias Steiger, who co-reviewed with Roghayeh Shirvani, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-64483-y.