Metabolic Profiling Analysis of Cytochrome B5 Production in E. Coli N4830-1 Using GC-MS

Metabolic Profiling Analysis of Cytochrome B5 Production in E. Coli N4830-1 Using GC-MS | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Metabolic Profiling Analysis of Cytochrome B5 Production in E. Coli N4830-1 Using GC-MS Thanyaporn Tengsuttiwat, Adam Burke, Naheed Nazly Kaderbhai, Joseph Gallagher, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9277397/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract Introduction: Recombinant protein production is pivotal across diverse industries, necessitating efforts to enhance both quantity and quality. Objectives: We employed gas chromatography coupled with mass spectrometry (GC-MS) to investigate the metabolic effects of producing mammalian cytochrome b 5 in the bacterial host E. coli N4830-1. The model system studied involved the cyt b 5 gene being introduced on plasmids into the host with varying copy numbers (0-6) under a λP L heat-sensitive promoter. Methods: Metabolic profiling involved GC-MS analysis with quality assessments using pooled QCs and multivariate chemometric analysis. This approach revealed specific metabolic features (identified to Level 2 of the Metabolomics Standards Initaive) correlating significantly with different cultivation conditions and variable copy numbers of the cyt b 5 gene among the examined bacterial strains. Results: Our study shows that CYT b 5 production imposed a substantial energetic burden, reflected by depletion of metabolites associated with the tricarboxylic acid (TCA) cycle, glycolysis/gluconeogenesis, pentose phosphate pathway, and glyoxylate metabolism, indicating increased ATP demand. Concurrent reductions in unsaturated fatty acids and changes in lipid-related metabolites suggest membrane remodeling in response to temperature induction (needed to initiate production of CYT b 5 ), as well as metabolic stress. Alterations in glyoxylate shunt and pentose phosphate pathway intermediates further indicate metabolic reprogramming to maintain carbon homeostasis, while significant changes in nucleotide- and amino acid-related metabolites suggest impacts on biosynthetic capacity associated with recombinant expression and gene copy number variation. Conclusion: Overall, these findings demonstrate that CYT b 5 production induces coordinated metabolic adjustments affecting energy-related metabolic pathways, bacterial cell wall and membrane biosynthesis, nucleotide metabolisms and bacterial stress responses, providing valuable insights that may be useful for the optimisation of recombinant protein production processes. GC-MS metabolic profile recombinant protein production cytochrome b5 chemometrics Figures Figure 1 Figure 2 Figure 3 Figure 4 1 INTRODUCTION Recombinant protein production has played a crucial role for decades in a broad range of industries; i.e., pharmaceuticals, food, and cosmetics (Kirk et al., 2002 ; Singh et al., 2016 ). The recombinant proteins market is expanding rapidly, projected to grow from about USD 3.25 billion in 2024 to nearly USD 8.66 billion by 2034, reflecting strong global demand. This growth underscores the critical role of recombinant protein technologies in modern biopharmaceuticals, including vaccines, therapeutic enzymes, and monoclonal antibodies (BioSpace, 2024 ). Although there are certain known limitations due to the lack of the machinery required for post-translational modification of eukaryotic proteins (Prabhu et al., 2021 ), E. coli is still a popular host. It offers numerous advantages, including ease of handling, genetic tractability, minimal growth requirements, rapid growth, and extensive development over many years making it a versatile host suitable for a wide range of applications (Leis et al., 2013 ). However, the development of improved bacterial hosts for the expression of recombinant proteins more efficiently is still an area of interest with strategies based on the optimisation of culturing conditions and genetic engineering methods (Tripathi and Shrivastava, 2019 ). Over the last two decades, metabolomics has emerged as a valuable tool for refining recombinant expression systems. By investigating the metabolome, which encompasses the fundamental constituents of diverse biomolecules involved in metabolism, including amino acids that are the building blocks of proteins, metabolomics offers insights into the biochemical alterations within the system that may not be fully elucidated through genomic approaches (Fiehn, 2002 ). Metabolomics can be described as the study of low molecular weight compounds which are usually primary metabolites involved in necessary bioprocesses that are essential for the maintenance and growth of cells (Dunn and Ellis, 2005 ; Hollywood et al., 2006 ; Mashego et al., 2007 ). The synthesis of secondary metabolites necessitates energy, such as ATP, as well as precursors, cofactors, and other pertinent building blocks, which are derived from basic primary metabolites (Bailey, 1991 ). Thus, investigating the correlation between primary and secondary metabolites and recombinant protein production levels, along with low molecular weight compounds, is imperative, as the availability of these metabolites may directly impact the biosynthetic rate of the desired recombinant protein (Stephanopoulos and Vallino, 1991). Additionally, given that the presence and maintenance of foreign DNA can influence both the quality and quantity of the recombinant protein, comprehending the metabolic burdens associated with harbouring such DNA material through metabolomics is beneficial for enhancing the recombinant protein production process. To monitor recombinant protein expression, Kaderbhai developed a reporter system in which positive expression within a microbial host is detected spectrophotometrically through a visible transformation to a bright pink colour (Kaderbhai et al., 1992 ). Building on this foundation, our previous study investigated cytochrome b 5 -producing E. coli N4830-1 using metabolomic fingerprinting via Fourier transform infrared (FT-IR) spectroscopy (Tengsuttiwat et al., 2022 ). While FT-IR spectroscopy provides rapid metabolic fingerprints, more detailed metabolite profiling is required to elucidate pathway-level alterations associated with recombinant protein production. Mass spectrometry (MS) – especially when coupled with chromatographic separation methods is frequently used in metabolomics due to its high sensitivity, selectivity, accuracy and resolution (Kaderbhai et al., 2003 ; Dunn and Ellis, 2005 ; Dunn et al ., 2011; Dunn et al., 2013 ). Thus, in this work GC-MS combined with chemometrics approaches was employed to investigate the metabolic alterations in E. coli , induced by overproduction of a specific recombinant protein, CYT b 5 . 2 METHODS 2.1 Bacterial strains and culture conditions Seven strains of Escherichia coli N4830-1 (F − suo thi-1 thr-1 leuB6 lacY1 fhuA21 supE44 rfbD1 mcrA1 his ilv galK8 Δ( hemF-exp ) Δ( bio uvrB ) [λ Δ Bam N + c I857 Δ(CroattR)]) harbouring the pEX-CYT plasmid were kindly provided by Dr Naheed Kaderbhai. The strains carry different copy numbers of cyt b 5 gene, ranging from 0 to 6, which are labelled as N0 to N6. All strains were streaked onto LB agar (Formedium, UK) prior to growing overnight at 37°C with 75 µg/mL ampicillin supplemented LB agar for selection of the desired colonies. The cells were cultured for the CYT b 5 production as described by (Kaderbhai et al., 1992 ; Tengsuttiwat et al., 2022 ). Briefly, 1 mL overnight culture was inoculated into a 250 mL Erlenmeyer flask containing 25 mL of LB medium with ampicillin (75 µg/mL). The cells were first incubated shaking at 150 rpm and 30°C for 2.5 h to reach the mid-log phase, the incubation temperature was then immediately increased to 38.5°C to activate the promoter, λP L . The culture was incubated for 9 h to allow for the protein production at 38.5°C, while a control set (N0-N6) was incubated for the same duration at 30°C. 2.2 Sample preparation After cultivation, each of the bacterial samples were collected as 1 mL aliquots to determine their biomass by measuring the optical density at 600 nm (OD 600nm ). The bacterial samples were also collected for metabolic profiling by immediately halting the metabolic process using solvent-based quenching, following which bacterial metabolites were subsequently extracted from the samples using cold methanol. Derivatisation was performed to maximise the range of metabolites that can be analysed using GC-MS; i.e., lower boiling point and improve thermal stability. The full procedures for quenching, extraction and derivatisation were undertaken as described previously and outlined below (Winder et al., 2008 ). 2.2.1 Quenching and extraction To immediately halt metabolic activities in bacterial cells, the samples were quenched with x2 volume of 60% methanol (at -48°C), where 15 mL sample was quenched with 30 mL of cold 60% methanol. These were immediately centrifuged at 4,347 g for 5 min at -9 ºC. The supernatant was decanted, followed by a 5 min centrifugation to remove residual supernatant via pipetting. The quenched pellets were then kept frozen at -80°C, until further analysis. At the point of analysis, upon defrosting the samples, 500 µL of cold (-48°C) 80% methanol was added to the quenched biomass to resuspend the pellets and these were transferred to 2 mL microcentrifuge tubes. This step was repeated to minimise the loss of biomass during the transfer. To extract the intracellular metabolites, freeze-thaw cycling approach using liquid nitrogen was performed three times. The cell suspension was then centrifuged at 14,500 g , -9°C for 10 min prior to biomass normalisation of the sample extracts according to their OD 600nm . Different volumes of cold (-48°C) 80% methanol were added to the biomass-normalised samples to ensure concentration uniformity of all samples. Following this a 700 µL sample volume was taken from each tube. In addition, an equal volume of these normalised extracted samples (250 µL) was collected from each tube, and pooled together to be used as a biological quality control (QC), as detailed in (Broadhurst et al., 2018 ) and then divided into aliquots of 450 µL. An aliquot (100 µL) of an internal standard (IS, Table S1 ) was added to the samples, followed by drying using a speed-vac centrifuge (ScanVac MaxiVac Alpha & MaxiVac Beta Vacuum Concentrator, LaboGene, Denmark) for 6 h. 2.2.2 Derivatisation Dried extracts were derivatised with 50 µL of 20 mg/mL O-methoxyamine • HCl dissolved in pyridine, and the samples were vortexed for 10 s prior to heating at 65°C for 40 min. This was followed by the addition of 50 µL of N -Methyl- N -(trimethylsilyl)trifluoroacetamide (MSTFA) (Merck Life Science UK Limited, UK) to all the samples, which were then vortexed for 10 s before heating again as described above. Finally, 20 µL of retention index solution (RI, Table S1 ) was added to each sample, followed by vortexing and centrifugation at 17,000 g for 15 min at 4°C (Muhamadali et al., 2015b ; Dunn et al., 2011b ; Begley et al., 2009 ). Aliquots (100 µL) of the derivatised samples were then transferred into GC vials. 2.3 Instrumental setup A J&W DB-5ms GC Column (Agilent Technologies, Inc.) was used in a split mode (ratio of 20:1) with helium as carrier gas in constant flow of 1 mL/min. An Agilent 8890 GC instrument equipped with an Agilent 7250 GC/Q-TOF was operated with an initial temperature program of 70°C for 4 min. The oven temperature was then increased at 20°C/min to 300°C and held for another 4 min before cooling down to 70°C. The transfer line temperature was 280°C and a 25 mL/min gas saver flow was switched on after 3 min. The mass range used was 45–600 Da with an acquisition rate of 10 spectra per second (Begley et al., 2009 ). 2.4 Data analysis Initially, the GC-MS raw data files were converted into mzML format using ProteoWizard software (64-bit version). The transformed data files were subsequently processed for deconvolution and blank subtraction using Mass Spectrometry-Data Independent AnaLysis software (MS-DIAL version 4.9.2). Deconvoluted data were corrected to remove technical variation using succinic acid d 4 , an internal standard, which showed the most reproducibility in terms of standard deviation when compared among the four deuterated standards (mixture of succinic acid d 4 , glycine d 5 , malonic acid d 4 and alanine d 7 , Table S1 ). Signal drift was further corrected using QC-based LOWESS normalisation. For data curation, first, the relative standard deviation (RSD) of each peak area was calculated from normalised pooled QCs and only those metabolite features with RSD ≤ 30% were kept for further analysis. Second, extracted ion chromatograms (EIC) of each peak were investigated manually to remove those with poor alignment or low signal to noise ratio. Furthermore, the features with a Gaussian chromatographic peak shape were separated and further analysed. Among these curated features, some were annotated as known compounds based on the GC-MS DB-Public-KovatsRI-VS3.msp’ spectral library, while others were labelled as unknown. Those features that were annotated based on their matching with the database, including, EI spectra and retention index, were considered to be identified to MSI level 2 based on these two orthogonal properties (Sumner et al., 2007 ). MetaboAnalyst, an online platform for metabolomics data analysis, was used to analyse pre-processed data by using t -test and ANOVA post-hoc analysis. All chemometric analysis including principal component analysis (PCA) and discriminant function analysis (PC-DFA), were applied using MATLAB® software (version r2019b, The MathWorks Inc., UK); our code is available via GitHub ( https://github.com/Biospec/ ). Significant features identified using t -test and ANOVA, together with those from PCA and PC-DFA loadings plots were subsequently combined and investigated. The relationship among those annotated metabolites in biological pathways of E. coli was determined using pathway analysis from the MetaboAnalyst platform as well as individual evaluation using classical biochemistry pathways and the Kyoto Encyclopedia of Genes and Genomes (KEGG). The pathway analysis parameters were set as follows: the visualisation method was a scatter plot based on testing significant features; the enrichment method applied was the Hypergeometric Test; the topology measure used was Relative-betweenness Centrality; and the reference metabolome included all compounds from the selected pathway library-Prokaryotes, Escherichia coli K-12 MG1655 (KEGG). 3 RESULTS 3.1 GC-MS data analysis In this study, the intracellular metabolic profiles of the CYT b 5 -producing E. coli N4830-1 were investigated using GC-MS. After deconvolution, 861 metabolic features were detected and normalised using the internal standard, succinic acid d 4 . These metabolic features were subsequently pre-processed by blank filtration, resulting in 580 features prior to data curation by removal of irreproducible features in the QCs (RSD > 30%). Data were further curated by elimination of those features with non-Gaussian chromatographic peak shape, using the EIC chromatogram, resulting in a total of 340 features being retained. These metabolic features were subsequently subjected to PCA to study and to compare the metabolic changes among the seven bacterial strains in both conditions: under recombinant protein induction at 38.5°C and control groups at 30°C. A PCA scores plot of the GC-MS data collected for all the samples was constructed, which illustrated a separation between the control (circles) and induced (stars) samples (Fig. 1 ). This segregation between the two culture conditions was predominantly according to PC1 axis, which accounted for 41.35% of the total explained variance (TEV). The first 20 PCs, with a TEV of 92.07%, were next analysed using DFA with a priori groups of 14 classes, i.e., seven individual strains and two conditions – as no information on either gene copy or culture conditions were provided to the DFA algorithm, these methods can be considered semi-supervised. This combined processes called PC-DFA resulted in a scores plot (Figure S1 ) of the GC-MS data, which displayed a trajectory similar to that seen in the PCA scores plot (Fig. 1 ), showing the two clusters of CYT b 5 -producing and non-producing E. coli strains. Whereas DF1 axis illustrates the distinct cultivation conditions, DF2 axis provided further informative trend showing the separation of strains N0 to N6 from negative to positive (Figure S1 ). Notably, the results from PC-DFA do not require validation, as the observed trend followed an ordered sequence from strain N0 to N6 without explicitly providing this order to the algorithm. The observed trend is consistent with the increasing number of cyt b 5 gene copies (0–6), which is expected to correlate with target protein expression levels; however, contributions from plasmid-associated metabolic burden cannot be excluded as a potential factor influencing the metabolic profiles. To investigate the metabolic changes within each condition, the data were analysed separately, based on the two culturing conditions (induced cultures or uninduced controls). The PCA scores plot of the induced samples illustrated a trend based on the PC2 axis (14.87% TEV), showing strains N0 and N1 on the negative side, strain N2 on the middle and strain N3-N6 on the positive side (Figure S2b). However, as expected, no clear trends were detected in the PCA scores plot of the control samples (Figure S2a). Furthermore, while the PC-DFA scores plot of the data from the control condition demonstrated no specific trend (Fig. 2 a), that of the induced group illustrated separations according to DF1 axis from strain N0 to N6 as shown in Fig. 2 b. These trends on both PCA and PC-DFA scores plots of induced samples were further evaluated individually and investigated via pathway analysis in MetaboAnalyst. The input for this analysis was the significant metabolic features identified by all chemometrics and statistical tests, as described in the methods section. These are also listed in Tables S2 and S3. MetaboAnalyst enables comprehensive metabolic pathway analysis by integrating pathway enrichment analysis with pathway topology analysis. The results of this analysis suggested that differential levels of CYT b 5 production modulated multiple metabolic pathways, including the TCA cycle, glycolysis/gluconeogenesis, pentose phosphate pathway, glyoxylate and dicarboxylate metabolism, purine metabolism, pyrimidine metabolism, and glycerolipid metabolism. CYT b 5 production is regulated by a λP L promoter that requires an increase in cultivation temperature (30 ◊ 38.5°C) to induce gene expression. Therefore, it was necessary to exclude any metabolic effects attributable solely to this 8.5°C temperature shift, to enable accurate investigation and detection of metabolic changes specifically associated with CYT b 5 production. To assess the impact of temperature induction, the bacterial strain N0 (host containing the plasmid but lacking the cyt b 5 gene) was analysed under the two culture conditions using PCA (Figure S3) and in combination with t -test statistical evaluation. The important annotated metabolites listed in Table S4 were thus identified as being influenced exclusively by temperature variation rather than by CYT b 5 production. 4 DISCUSSION 4.1 GC-MS data interpretation 4.1.1 Effect on energy metabolisms Regarding the chemometric and statistical analyses, significant metabolic features were mapped to various fundamental metabolic pathways in the bacterial host. This includes central metabolic processes such as glycolysis and TCA cycle. These pathways, in conjunction with the electron transport chain, play a crucial role in cellular energy production. Many metabolic features within these pathways were observed in significantly lower levels in CYT b 5 -producing strains compared to the control. This includes metabolites such as pyruvate, oxaloacetate, fumarate, lactate, citrate, isocitrate, gluconic acid, glucuronic acid and 3-phosphoglycerate (Fig. 3 ). The reduced abundance of these metabolites indicates their substantial utilisation during temperature-induced cultivation, likely reflecting the high energetic demands associated with recombinant protein production. Notably, under induction conditions, nicotinamide was identified as a key metabolite whose concentration progressively declined with increasing copy numbers of the cyt b 5 gene across bacterial strains N0 to N4 (Fig. 4 ). Nicotinamide is a precursor of the essential energy-related cofactor nicotinamide adenine dinucleotide (NAD, and it can be biosynthesised from tryptophan, which was likewise identified as a significant metabolite exhibiting differential levels between the two bacterial culture conditions (Fig. 4 ). 4.1.2 Effect on bacterial stress responses Several metabolites associated with the pentose phosphate pathway, the glyoxylate shunt, and glutathione metabolism exhibited significant changes in abundance between the induced and control samples. Specifically, there was a decrease in the levels of pyruvate, oxaloacetate, citric acid, isocitric acid, gluconic acid, glutamate, putrescine, and ribose-5-phosphate in the CYT b 5 -producing condition. Under this condition, induced samples exhibited a pronounced increase in the levels of cadaverine and 5-aminovaleric acid (5-AVA), metabolites that are likely derived from lysine catabolism. Additional metabolites, including the polyamines cadaverine, putrescine, and spermidine, were also detected. These compounds are known to be upregulated in response to a range of physicochemical stresses, such as elevated temperature, osmotic fluctuations, reactive oxygen species, and ultraviolet radiation (Rhee et al., 2007 ). Specifically, putrescine concentrations declined under induction, whereas cadaverine displayed an increasing trend. 4.1.3 Effect on permeability of bacterial host cells Numerous intermediates involved in bacterial cell wall and membrane biosynthesis exhibited significant alterations between the two culture conditions. In particular, glycerol, glycerol-3-phosphate, ethanolamine, O -phosphoethanolamine, and several unsaturated fatty acids, including oleic acid, elaidic acid, and palmitoleic acid were among the metabolites affected by the temperature-induced production of CYT b 5 , serving as intermediates in phospholipid biosynthesis (Sinensky, 1971 ). Moreover, bacterial cell wall synthesis appeared to be impacted, as evidenced by the significantly reduced levels of N -acetylglucosamine in the induced condition compared to the control set. 4.1.4 Effect on nucleotide metabolisms Nucleotide metabolism, specifically purine and pyrimidine metabolism, also displayed alterations under the investigated conditions. While certain metabolites detected in these pathways exhibited a decreasing trend in the CYT b 5 -producing condition, such as 5,6-dihydrouracil, 2-deoxyuridine, glutamine, hypoxanthine, and thymine, others such as orotic acid and 3-aminoisobutyric acid showed a significant increase. However, it is worth noting that cytosine and orotic acid from pyrimidine metabolism, and guanine from purine metabolism, exhibited significant alterations under both cultivation conditions (control and induced samples) and strain effects (inducing strains N0-N6). These findings indicate that the metabolic changes likely reflect the combined effects of heat induction and CYT b 5 production. 4.2 Biochemical discussions Metabolic profiling of the CYT b 5 -producing E. coli N4830-1 was investigated using GC-MS as this method focusses on central carbon and nitrogen metabolism. The samples were cultured under optimal conditions, first at 30.0°C, then using an induction temperature of 38.5°C (for the λP L promoter) and an incubation time of 9 h, as determined in our previous publication (Tengsuttiwat et al., 2022 ). The generated GC-MS data were processed using QC practices (Broadhurst et al., 2018 ) and then subjected to statistical tests, including t -tests and ANOVA, as well as chemometric analyses such as PCA and PC-DFA The clustering patterns obtained from both PCA and PC-DFA were consistent with those previously reported using FT-IR spectroscopy (Tengsuttiwat et al., 2022 ). Notably, the PC-DFA scores plots demonstrated a clear gene copy number-dependent trend among induced samples (N0-N6) across three independent batch cultures, underscoring the reproducibility and robustness of the overall analytical workflow. Importantly, the present GC-MS analysis further provided detailed biochemical insights into specific metabolites that play crucial roles under different cultivation conditions and among the seven bacterial strains. The significant metabolites were mapped onto metabolic pathways, revealing major effects on energy-related mechanisms, bacterial cell envelope biosynthesis, and nucleotide metabolisms. The synthesis of recombinant proteins is an energy-intensive process, especially in these strains where CYT b 5 production reached 0.73 mM calculated from data in our previous publication (Tengsuttiwat et al., 2022 ) using the extinction coefficient value of 171 cm − 1 mM − 1 at 423 nm (Gómez-Tabales et al., 2020 .) Consistent with this, previous studies have reported a significant decrease in intracellular ATP levels following temperature-induced recombinant production of human fibroblast growth factor protein (Templeton et al., 2013 ). Additionally, cells producing recombinant proteins also exhibit increased metabolic flux through the TCA cycle and related pathways to meet the elevated ATP demand required for intensive protein synthesis and stress response mechanisms (Wittmann et al., 2007 ). Consequently, the substantial ATP requirement associated with CYT b 5 production likely contributed to the observed depletion of metabolites involved in key energy-generating pathways, including the TCA cycle, glycolysis/gluconeogenesis, the pentose phosphate pathway, as well as glyoxylate and dicarboxylate metabolism. This observation is also supported by prior studies indicating that glycolytic flux in E. coli is highly responsive to variations in ATP demand (Koebmann et al., 2002 ). During the recombinant protein production, the bacterial host exhibited significant physiological alterations in response to the temperature change between the control and inducing culture conditions. These adaptations prominently involved modifications in cell wall and membrane metabolisms, particularly influencing membrane fluidity, as previously documented in the literature (Russell, 2003 ; Wang et al., 2023 ). Membrane integrity in E. coli is influenced not only by heat stress but also by environmental factors, including temperature and pH fluctuations (Rowlett et al., 2017 ). In the present study, an increase in pH from 6.8 to 8.2 was observed following induction (Table S5). Such alkalinisation may be attributed to the oxidation of carbon sources, depletion of amino acids, and the accumulation of basic metabolic by-products (Sánchez-Clemente et al., 2020 ). Additionally, recombinant protein production imposes a considerable metabolic burden on the host, triggering adaptive responses including rearrangement of lipid composition (Ami et al., 2009 ), an increase in saturated fatty acids involved in phospholipid metabolism (Sinensky, 1971 ), and elevated fluxes of pyruvate, acetate, and lactate formation (Wittmann et al., 2007 ). Collectively, these adaptations reflect coordinated cellular strategies to maintain membrane stability and metabolic balance under the combined stresses of temperature induction and recombinant protein synthesis. Furthermore, stress-related proteins are often synthesised during recombinant protein production, leading to an increased ATP demand (Hoffmann and Rinas, 2004 ). Heat shock proteins, for example, can be triggered to cope with the stressful environment (Schumann, 2016 ). Both the membrane adaptation mentioned above, and the synthesis of stress-responsive proteins are energy-intensive processes. This could lead to significant depletion of glucose for energy production, thereby promoting the utilisation of alternative carbon sources such as acetate commonly associated with recombinant protein synthesis. This utilisation was evidenced by the observed impact on the glyoxylate shunt due to CYT b 5 production. However, glucose levels remained relatively stable among all the samples. This may be explained by the fact that glucose can also be synthesised from fatty acids through gluconeogenesis and glyoxylate cycle (Cordero et al., 2008 ). In this process, the glyoxylate cycle provides succinate, which enters the TCA cycle and subsequently feeds into gluconeogenesis, thereby maintaining the detected glucose level. This observation also supports the finding of reduction in unsaturated fatty acids, such as oleic acid, elaidic acid, and palmitoleic acid, in the induced samples. The glyoxylate shunt, along with the pentose phosphate pathway, plays a crucial role in carbon homeostasis. The significant alterations observed in gluconic acid and pyruvate within the pentose phosphate pathway suggest metabolic adaptation to balance carbon flux. When preferred carbon sources like glucose become insufficient to meet increased energetic requirements, E. coli can activate the glyoxylate shunt to utilise simpler carbon sources such as acetate and fatty acids. Additionally, oxidative stress is a common consequence of recombinant protein production due to increased cellular energy demand (Templeton et al., 2013 ). In response, bacterial cells may utilise pathways such as the pentose phosphate pathway, glyoxylate shunt, and glutathione metabolism to mitigate the effects of such stress (Stincone et al., 2015 ; Ahn et al., 2016 ; Ozkul and Kocak, 2020 ). The pentose phosphate pathway is essential for supplying precursors for nucleotide and amino acid biosynthesis. Correspondingly, significant alterations in metabolites such as cytosine, orotic acid, O -phosphoethanolamine, myristic acid, heptanoic acid, serine, and tryptamine across the seven CYT b 5 -producing strains suggest that these biosynthetic pathways were affected by heat-induced CYT b 5 expression and potential differences in cyt b 5 gene copy number (Fig. 4 ). Furthermore, the untargeted metabolomics approach used in this study does not allow absolute quantification of metabolites (rather a peak ratio against an internal standard), and some detected compounds, such as amino acids, may originate from either endogenous synthesis or uptake from the nutrient-rich LB medium. Future studies incorporating metabolic footprinting would help clarify metabolite origins and consumption dynamics. 5 CONCLUSIONS In this study, an untargeted metabolomic approach was employed utilising GC-MS to examine the metabolic profiles of seven strains of E. coli N4830-1, in the context of temperature-induced production of CYT b 5 . The most notable alterations were observed in the TCA cycle, where key metabolites such as pyruvate, oxaloacetate, citrate, isocitrate, and fumarate exhibited marked depletion in the induced condition. This observation indicates the heightened energy requirements inherent to recombinant protein production systems. The imposition of heat stress and the demands of recombinant protein production necessitate metabolic adaptations within bacterial hosts. Significant alterations in detected metabolites suggest responses aimed at fortifying bacterial cell wall and membrane biosynthesis, as well as deploying defensive mechanisms against oxidative stress through pathways such as the pentose phosphate pathway, glyoxylate shunt, and glutathione metabolism. The aforementioned mechanisms all entail energy expenditure, thereby increasing the demand for energy within host cells. However, to gain a comprehensive understanding of the metabolic burden associated with recombinant CYT b 5 production among the seven strains, future studies could explore a combination of metabolic profiling (intracellular metabolites), footprinting (extracellular metabolites), and lipidomics (membrane lipids). Additionally, targeted metabolomics approach could be employed to facilitate quantitative analysis. Since energy consumption played a crucial role among all the samples, adenylate energy charge is worth considering as it represents the available metabolic energy of a host at the time of sampling (Atkinson and Walton, 1967 ; Atkinson, 1968 ). In addition, fluxomics is a valuable metabolomic approach for monitoring metabolic pathways (Emwas et al., 2022 ); for instance, the analysis of metabolic flux using 13 C-labeled molecules. Such investigations may reveal potential bottlenecks in the recombinant protein production process and support optimisation of the production system, but only really work in highly defined medium with a sole carbon supply which may affect recombinant protein production as current production systems are replete with nutrients. Subsequent genetic engineering strategies or media optimisation, for example, could be proposed to enhance the efficiency of the production system. In conclusion, this study shows that CYT b 5 production triggers coordinated metabolic changes that impact energy metabolism, cell wall and membrane biosynthesis, nucleotide metabolism, and bacterial stress responses. These insights may in the future offer valuable guidance for optimising recombinant protein production processes in prokaryotic hosts. Declarations Supplementary Information The online version contains supplementary material. Acknowledgements We thank the University of Liverpool and the Royal Thai Government for financial support. R.G. thanks UK BBSRC for funding metabolomics (BB/W019558/1). Author Contributions Thanyaporn Tengsuttiwat: experimental design, sample collection and preparation, GC-MS data analysis, data interpretation and manuscript preparation. Adam Burke: GC-MS data analysis and manuscript preparation. Naheed Nazly Kaderbhai and Joe Gallagher: genetic engineering of the strains, growth condition optimisation, and manuscript preparation. Royston Goodacre: co-principal investigator, experimental design, data interpretation, and manuscript preparation. Howbeer Muhamadali: co-principal investigator, experimental design, data interpretation, and manuscript preparation. Declarations RG is EiC of Metabolomics , he was not involved in the review process for this paper. All authors declare that they have no conflict of interest. Data availability GC-MS data is freely available at MetaboLights repository with the unique identifier REQ20250520210627 (www.ebi.ac.uk/metabolights/). References Ahn, S., Jung, J., Jang, I.-A., Madsen, E.L., and Park, W. (2016). Role of Glyoxylate Shunt in Oxidative Stress Response. Journal of Biological Chemistry 291(22), 11928-11938. doi: 10.1074/jbc.M115.708149. Ami, D., Natalello, A., Schultz, T., Gatti-Lafranconi, P., Lotti, M., Doglia, S.M., et al . (2009). Effects of recombinant protein misfolding and aggregation on bacterial membranes. Biochim Biophys Acta 1794(2), 263-269. doi: 10.1016/j.bbapap.2008.10.015. Atkinson, D.E. (1968). The energy charge of the adenylate pool as a regulatory parameter. Interaction with feedback modifiers. Biochemistry 7(11), 4030-4034. doi: 10.1021/bi00851a033. Atkinson, D.E., and Walton, G.M. (1967). Adenosine triphosphate conservation in metabolic regulation. Rat liver citrate cleavage enzyme. J Biol Chem 242(13), 3239-3241. Bailey, J.E. (1991). Toward a Science of Metabolic Engineering. Science 252(5013), 1668-1675. Begley, P., Francis-McIntyre, S., Dunn, W.B., Broadhurst, D.I., Halsall, A., Tseng, A., et al . (2009). Development and performance of a gas chromatography-time-of-flight mass spectrometry analysis for large-scale nontargeted metabolomic studies of human serum. Anal Chem 81(16), 7038-7046. doi: 10.1021/ac9011599. BioSpace, 2024. Recombinant proteins market size to hit USD 8.66 billion by 2034 . Available at: https://www.biospace.com/press-releases/recombinant-proteins-market-size-to-hit-usd-8-66-billion-by-2034 (Accessed: 17 February 2026). Broadhurst, D., Goodacre, R., Reinke, S. N., Kuligowski, J., Wilson, I. D., Lewis, M. R., & Dunn, W. B. (2018). Guidelines and considerations for the use of system suitability and quality control samples in mass spectrometry assays applied in untargeted clinical metabolomic studies. Metabolomics, 14(6), 72. https://doi.org/10.1007/s11306-018-1367-3 Cordero, P., Campion, J., Milagro, F.I., Marzo, F., and Martinez, J.A. (2008). Fat-to-glucose interconversion by hydrodynamic transfer of two glyoxylate cycle enzyme genes. Lipids in Health and Disease 7(1), 49. doi: 10.1186/1476-511X-7-49. Dunn, W.B., Broadhurst, D.I., Atherton, H.J., Goodacre, R., and Griffin, J.L. (2011a). Systems level studies of mammalian metabolomes: the roles of mass spectrometry and nuclear magnetic resonance spectroscopy. Chemical Society Reviews 40(1), 387-426. doi: 10.1039/B906712B. Dunn, W.B., Broadhurst, D., Begley, P., Zelena, E., Francis-McIntyre, S., Anderson, N., et al . (2011b). Procedures for large-scale metabolic profiling of serum and plasma using gas chromatography and liquid chromatography coupled to mass spectrometry. Nature Protocols 6(7), 1060-1083. doi: 10.1038/nprot.2011.335. Dunn, W.B., and Ellis, D.I. (2005). Metabolomics: Current analytical platforms and methodologies. TrAC Trends in Analytical Chemistry 24(4), 285-294. doi: 10.1016/j.trac.2004.11.021. Dunn, W.B., Erban, A., Weber, R.J.M., Creek, D.J., Brown, M., Breitling, R., et al . (2013). Mass appeal: metabolite identification in mass spectrometry-focused untargeted metabolomics. Metabolomics 9(1), 44-66. doi: 10.1007/s11306-012-0434-4. Emwas, A.H., Szczepski, K., Al-Younis, I., Lachowicz, J.I., and Jaremko, M. (2022). Fluxomics - New Metabolomics Approaches to Monitor Metabolic Pathways. Front Pharmacol 13, 805782. doi: 10.3389/fphar.2022.805782. Fiehn, O. (2002). Metabolomics — The Link Between Genotypes and Phenotypes. Plant Molecular Biology, v.48, 155-177 (2002) 48. doi: 10.1007/978-94-010-0448-0_11. Gómez-Tabales, J., García-Martín, E., Agúndez, J.A.G., and Gutierrez-Merino, C. (2020). Modulation of CYP2C9 activity and hydrogen peroxide production by cytochrome b(5). Sci Rep 10(1), 15571. doi: 10.1038/s41598-020-72284-0. Han, Q., and Eiteman, M.A. (2019). Acetate formation during recombinant protein production in Escherichia coli K-12 with an elevated NAD(H) pool. Eng Life Sci 19(11), 770-780. doi: 10.1002/elsc.201900045. Hoffmann, F., and Rinas, U. (2004). "Stress Induced by Recombinant Protein Production in Escherichia coli," in Physiological Stress Responses in Bioprocesses: -/-. (Berlin, Heidelberg: Springer Berlin Heidelberg), 73-92. Hollywood, K., Brison, D.R., and Goodacre, R. (2006). Metabolomics: Current technologies and future trends. PROTEOMICS 6(17), 4716-4723. doi: 10.1002/pmic.200600106. Kaderbhai, N.N., , G., JOSEPH, HE, MINGYUE, and and KADERBHAI, M.A. (1992). A Pink Bacterium as a Reporter System Signaling Expression of a Recombinant Protein. DNA and Cell Biology 11(7), 567-577. doi: 10.1089/dna.1992.11.567. Kaderbhai, N.N., Broadhurst, D.I., Ellis, D.I., Goodacre, R., and Kell, D.B. (2003). Functional genomics via metabolic footprinting: monitoring metabolite secretion by Escherichia coli tryptophan metabolism mutants using FT-IR and direct injection electrospray mass spectrometry. Comparative and functional genomics 4(4), 376-391. doi: 10.1002/cfg.302. Kirk, O., Borchert, T.V., and Fuglsang, C.C. (2002). Industrial enzyme applications. Current Opinion in Biotechnology 13(4), 345-351. doi: https://doi.org/10.1016/S0958-1669(02)00328-2. Koebmann, B.J., Westerhoff, H.V., Snoep, J.L., Nilsson, D., and Jensen, P.R. (2002). The glycolytic flux in Escherichia coli is controlled by the demand for ATP. J Bacteriol 184(14), 3909-3916. doi: 10.1128/jb.184.14.3909-3916.2002. Leis, B., Angelov, A., and Liebl, W. (2013). Screening and expression of genes from metagenomes. Adv Appl Microbiol 83, 1-68. doi: 10.1016/b978-0-12-407678-5.00001-5. Mashego, M.R., Rumbold, K., De Mey, M., Vandamme, E., Soetaert, W., and Heijnen, J.J. (2007). Microbial metabolomics: past, present and future methodologies. Biotechnology Letters 29(1), 1-16. doi: 10.1007/s10529-006-9218-0. Muhamadali, H., Xu, Y., Ellis, D., Trivedi, D., Rattray, N., Bernaerts, K., et al . (2015a). Metabolomics investigation of recombinant mTNFα production in Streptomyces lividans . Microbial Cell Factories 14. doi: 10.1186/s12934-015-0350-1. Muhamadali, H., Xu, Y., Ellis, D.I., Allwood, J.W., Rattray, N.J., Correa, E., et al . (2015b). Metabolic Profiling of Geobacter sulfurreducens during Industrial Bioprocess Scale-Up. Appl Environ Microbiol 81(10), 3288-3298. doi: 10.1128/aem.00294-15. Muhamadali, H., Xu, Y., Morra, R., Trivedi, D.K., Rattray, N.J.W., Dixon, N., et al . (2016). Metabolomic analysis of riboswitch containing E. coli recombinant expression system. Molecular BioSystems 12(2), 350-361. doi: 10.1039/C5MB00624D. Ozkul, C., and Kocak, E. (2020). Metabolic response of Escherichia coli to subinhibitory concentration of ofloxacin. Journal of Research in Pharmacy 24, 593-601. doi: 10.35333/jrp.2020.207. Prabhu, S.K., Yang, Q., Tong, X., and Wang, L.X. (2021). Exploring a combined Escherichia coli-based glycosylation and in vitro transglycosylation approach for expression of glycosylated interferon alpha. Bioorg Med Chem 33, 116037. doi: 10.1016/j.bmc.2021.116037. Rhee, H.J., Kim, E.J., and Lee, J.K. (2007). Physiological polyamines: simple primordial stress molecules. J Cell Mol Med 11(4), 685-703. doi: 10.1111/j.1582-4934.2007.00077.x. Rowlett, V.W., Mallampalli, V., Karlstaedt, A., Dowhan, W., Taegtmeyer, H., Margolin, W., et al . (2017). Impact of Membrane Phospholipid Alterations in Escherichia coli on Cellular Function and Bacterial Stress Adaptation. J Bacteriol 199(13). doi: 10.1128/jb.00849-16. Russell, A.D. (2003). Lethal effects of heat on bacterial physiology and structure. Sci Prog 86(Pt 1-2), 115-137. doi: 10.3184/003685003783238699. Sánchez-Clemente, R., Guijo, M. I., Nogales, J., & Blasco, R. (2020). Carbon Source Influence on Extracellular pH Changes along Bacterial Cell-Growth. Genes , 11 (11), 1292. https://doi.org/10.3390/genes11111292 Schumann, W. (2016). Regulation of bacterial heat shock stimulons. Cell Stress and Chaperones , 21 (6), 959–968. https://doi.org/10.1007/s12192-016-0727-z Sinensky, M. (1971). Temperature control of phospholipid biosynthesis in Escherichia coli . J Bacteriol 106(2), 449-455. doi: 10.1128/jb.106.2.449-455.1971. Singh, R., Kumar, M., Mittal, A., and Mehta, P.K. (2016). Microbial enzymes: industrial progress in 21st century. 3 Biotech 6(2), 174. doi: 10.1007/s13205-016-0485-8. Stephanopoulos, G.N., and Vallino, J. (1991). Network Rigidity And Metabolic Engineering In Metabolite Overproduction. Science (New York, N.Y.) 252, 1675-1681. doi: 10.1126/science.1904627. Stincone, A., Prigione, A., Cramer, T., Wamelink, M.M.C., Campbell, K., Cheung, E., et al . (2015). The return of metabolism: biochemistry and physiology of the pentose phosphate pathway. Biological Reviews 90(3), 927-963. doi: https://doi.org/10.1111/brv.12140. Sumner, L.W., Amberg, A., Barrett, D., Beale, M.H., Beger, R., Daykin, C.A., et al . (2007). Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3(3), 211-221. doi: 10.1007/s11306-007-0082-2. Templeton, N., Dean, J., Reddy, P., and Young, J.D. (2013). Peak antibody production is associated with increased oxidative metabolism in an industrially relevant fed-batch CHO cell culture. Biotechnology and Bioengineering 110(7), 2013-2024. doi: https://doi.org/10.1002/bit.24858. Tengsuttiwat, T., Kaderbhai, N.N., Gallagher, J., Goodacre, R., and Muhamadali, H. (2022). Metabolic Fingerprint Analysis of Cytochrome b 5 -producing E. coli N4830-1 Using FT-IR Spectroscopy. Frontiers in Microbiology 13. doi: 10.3389/fmicb.2022.874247. Tripathi, N.K., and Shrivastava, A. (2019). Recent Developments in Bioprocessing of Recombinant Proteins: Expression Hosts and Process Development. Frontiers in Bioengineering and Biotechnology 7. doi: 10.3389/fbioe.2019.00420. Wang, L., Li, A., Fang, J., Wang, Y., Chen, L., Qiao, L., et al . (2023). Enhanced Cell Wall and Cell Membrane Activity Promotes Heat Adaptation of Enterococcus faecium. Int J Mol Sci 24(14). doi: 10.3390/ijms241411822. Winder, C.L., Dunn, W.B., Schuler, S., Broadhurst, D., Jarvis, R., Stephens, G.M., et al . (2008). Global Metabolic Profiling of Escherichia coli Cultures: an Evaluation of Methods for Quenching and Extraction of Intracellular Metabolites. Analytical Chemistry 80(8), 2939-2948. doi: 10.1021/ac7023409. Wittmann, C., Weber, J., Betiku, E., Krömer, J., Böhm, D., and Rinas, U. (2007). Response of fluxome and metabolome to temperature-induced recombinant protein synthesis in Escherichia coli . J Biotechnol 132(4), 375-384. doi: 10.1016/j.jbiotec.2007.07.495. Additional Declarations No competing interests reported. Supplementary Files supplementaryinformation.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 18 May, 2026 Reviews received at journal 23 Apr, 2026 Reviewers agreed at journal 16 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers invited by journal 13 Apr, 2026 Editor assigned by journal 01 Apr, 2026 Submission checks completed at journal 01 Apr, 2026 First submitted to journal 31 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9277397","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":624190951,"identity":"4e78aa5e-e101-420c-9bcc-2c0e1364ad55","order_by":0,"name":"Thanyaporn Tengsuttiwat","email":"","orcid":"","institution":"National Center for Genetic Engineering and Biotechnology","correspondingAuthor":false,"prefix":"","firstName":"Thanyaporn","middleName":"","lastName":"Tengsuttiwat","suffix":""},{"id":624190952,"identity":"6cb22045-c7d1-4ddf-8f32-96b7ce9fc887","order_by":1,"name":"Adam Burke","email":"","orcid":"","institution":"University of Liverpool","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"Burke","suffix":""},{"id":624190955,"identity":"83a2feed-d4de-4d7b-abd3-25ae6b9976df","order_by":2,"name":"Naheed Nazly Kaderbhai","email":"","orcid":"","institution":"Aberystwyth University","correspondingAuthor":false,"prefix":"","firstName":"Naheed","middleName":"Nazly","lastName":"Kaderbhai","suffix":""},{"id":624190961,"identity":"a85e1080-7ae1-4061-b019-bf783e6b6ffe","order_by":3,"name":"Joseph Gallagher","email":"","orcid":"","institution":"Aberystwyth University","correspondingAuthor":false,"prefix":"","firstName":"Joseph","middleName":"","lastName":"Gallagher","suffix":""},{"id":624190962,"identity":"dbb1978f-9844-4a7d-b481-708c1154cc75","order_by":4,"name":"Howbeer Muhamadali","email":"","orcid":"","institution":"University of Liverpool","correspondingAuthor":false,"prefix":"","firstName":"Howbeer","middleName":"","lastName":"Muhamadali","suffix":""},{"id":624190963,"identity":"764e06c1-8753-4e9a-aa67-29b46499c592","order_by":5,"name":"Royston Goodacre","email":"data:image/png;base64,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","orcid":"","institution":"University of Liverpool","correspondingAuthor":true,"prefix":"","firstName":"Royston","middleName":"","lastName":"Goodacre","suffix":""}],"badges":[],"createdAt":"2026-03-31 09:09:59","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9277397/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9277397/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107489640,"identity":"edfafc2c-c3be-4625-917b-f5910be41960","added_by":"auto","created_at":"2026-04-22 02:48:28","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":575785,"visible":true,"origin":"","legend":"\u003cp\u003ePCA scores plot of the GC-MS metabolic profiles of CYT b\u003csub\u003e5\u003c/sub\u003e-producing (stars, n\u0026nbsp;=\u0026nbsp;6 biological replicates) and non-producing (circles, n\u0026nbsp;=\u0026nbsp;3 biological replicates) E. coli N4830-1 strains from GC-MS metabolic profiles. N0–N6 indicate strains containing 0 to 6 copies of the cyt b\u003csub\u003e5\u003c/sub\u003e gene. The number in brackets describes the percentage of the total explained variance (TEV) associated with each PC.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/b4769a2ef47a143d6c173634.png"},{"id":107468425,"identity":"7a6dc707-6c38-4124-9870-4adf3481f080","added_by":"auto","created_at":"2026-04-21 19:15:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":60999,"visible":true,"origin":"","legend":"\u003cp\u003ePC-DFA scores plots of E. coli N4830-1 strains in (A) control and non-induced conditions (circles, n\u0026nbsp;=\u0026nbsp;3 biological replicates) with 96.53 % TEV using PCs 1-15 with the a prior class membership based on the biological replicates, and (B) CYT b\u003csub\u003e5\u003c/sub\u003e -producing condition (stars, n\u0026nbsp;=\u0026nbsp;6 biological replicates) with 90.60 % TEV using PCs 1-15 with the a prior class membership based on the biological replicates. The black arrow shows the trend of induced samples from strains N0-N6.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/5e3990078b4556b38b71ccc8.png"},{"id":107468426,"identity":"c3cc1472-c8b0-4f65-a373-3ff97eab39d7","added_by":"auto","created_at":"2026-04-21 19:15:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":160718,"visible":true,"origin":"","legend":"\u003cp\u003eSignificant annotated metabolites identified by t-test analysis (comparing 2 classes: induced N0-N6 strains (n = 6 biological replicates) labelled as 1-7 and control N0-N6 strains (n = 3 biological replicates) labelled as 8-14) and chemometrics analyses (PCA and PC-DFA with 14 classes). The bar represents the means and standard deviations are represented by error bars.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/024df87a4d8004ddbec74617.png"},{"id":107488820,"identity":"c062de79-808b-4983-8604-579f7eadf0ce","added_by":"auto","created_at":"2026-04-22 02:45:54","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":136942,"visible":true,"origin":"","legend":"\u003cp\u003eSignificant annotated metabolites identified by ANOVA (7 classes of 7 strains, n\u0026nbsp;=\u0026nbsp;6 biological replicates) and chemometrics analyses (PCA and PC-DFA with 7 classes, n\u0026nbsp;=\u0026nbsp;6 biological replicates). These plots show box-whiskers where the line within the box is the median value, the top and bottom of the boxes are the 25\u003csup\u003eth\u003c/sup\u003e and 75\u003csup\u003eth\u003c/sup\u003e percentiles, the size of the box is the interquartile range (IQR), and the whiskers extend to the most extreme data points.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/6fe06415345eeeb55c75a1b3.png"},{"id":107704506,"identity":"c948844f-aa38-481b-ba90-73188f4cfae8","added_by":"auto","created_at":"2026-04-24 08:45:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1125006,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/284bf14e-579d-41ed-8954-e608bd794ca3.pdf"},{"id":107468423,"identity":"af28a8e3-c949-48b6-8b29-907eecd506a0","added_by":"auto","created_at":"2026-04-21 19:15:34","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":595857,"visible":true,"origin":"","legend":"","description":"","filename":"supplementaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-9277397/v1/5cc7662a6e0e84d47431076d.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eMetabolic Profiling Analysis of Cytochrome B5 Production in E. Coli N4830-1 Using GC-MS\u003c/p\u003e","fulltext":[{"header":"1 INTRODUCTION","content":"\u003cp\u003eRecombinant protein production has played a crucial role for decades in a broad range of industries; i.e., pharmaceuticals, food, and cosmetics (Kirk et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Singh et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The recombinant proteins market is expanding rapidly, projected to grow from about USD 3.25\u0026nbsp;billion in 2024 to nearly USD 8.66\u0026nbsp;billion by 2034, reflecting strong global demand. This growth underscores the critical role of recombinant protein technologies in modern biopharmaceuticals, including vaccines, therapeutic enzymes, and monoclonal antibodies (BioSpace, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Although there are certain known limitations due to the lack of the machinery required for post-translational modification of eukaryotic proteins (Prabhu et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), E. \u003cem\u003ecoli\u003c/em\u003e is still a popular host. It offers numerous advantages, including ease of handling, genetic tractability, minimal growth requirements, rapid growth, and extensive development over many years making it a versatile host suitable for a wide range of applications (Leis et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). However, the development of improved bacterial hosts for the expression of recombinant proteins more efficiently is still an area of interest with strategies based on the optimisation of culturing conditions and genetic engineering methods (Tripathi and Shrivastava, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOver the last two decades, metabolomics has emerged as a valuable tool for refining recombinant expression systems. By investigating the metabolome, which encompasses the fundamental constituents of diverse biomolecules involved in metabolism, including amino acids that are the building blocks of proteins, metabolomics offers insights into the biochemical alterations within the system that may not be fully elucidated through genomic approaches (Fiehn, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Metabolomics can be described as the study of low molecular weight compounds which are usually primary metabolites involved in necessary bioprocesses that are essential for the maintenance and growth of cells (Dunn and Ellis, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Hollywood et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Mashego et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). The synthesis of secondary metabolites necessitates energy, such as ATP, as well as precursors, cofactors, and other pertinent building blocks, which are derived from basic primary metabolites (Bailey, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e1991\u003c/span\u003e). Thus, investigating the correlation between primary and secondary metabolites and recombinant protein production levels, along with low molecular weight compounds, is imperative, as the availability of these metabolites may directly impact the biosynthetic rate of the desired recombinant protein (Stephanopoulos and Vallino, 1991). Additionally, given that the presence and maintenance of foreign DNA can influence both the quality and quantity of the recombinant protein, comprehending the metabolic burdens associated with harbouring such DNA material through metabolomics is beneficial for enhancing the recombinant protein production process.\u003c/p\u003e \u003cp\u003eTo monitor recombinant protein expression, Kaderbhai developed a reporter system in which positive expression within a microbial host is detected spectrophotometrically through a visible transformation to a bright pink colour (Kaderbhai et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). Building on this foundation, our previous study investigated cytochrome \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e-producing \u003cem\u003eE. coli\u003c/em\u003e N4830-1 using metabolomic fingerprinting via Fourier transform infrared (FT-IR) spectroscopy (Tengsuttiwat et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). While FT-IR spectroscopy provides rapid metabolic fingerprints, more detailed metabolite profiling is required to elucidate pathway-level alterations associated with recombinant protein production. Mass spectrometry (MS) \u0026ndash; especially when coupled with chromatographic separation methods is frequently used in metabolomics due to its high sensitivity, selectivity, accuracy and resolution (Kaderbhai et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Dunn and Ellis, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Dunn \u003cem\u003eet al\u003c/em\u003e., 2011; Dunn et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Thus, in this work GC-MS combined with chemometrics approaches was employed to investigate the metabolic alterations in \u003cem\u003eE. coli\u003c/em\u003e, induced by overproduction of a specific recombinant protein, CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e.\u003c/p\u003e"},{"header":"2 METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Bacterial strains and culture conditions\u003c/h2\u003e \u003cp\u003eSeven strains of \u003cem\u003eEscherichia coli\u003c/em\u003e N4830-1 (F\u003csup\u003e\u0026minus;\u003c/sup\u003e\u003cem\u003esuo thi-1 thr-1 leuB6 lacY1 fhuA21 supE44 rfbD1 mcrA1 his ilv galK8\u003c/em\u003e Δ(\u003cem\u003ehemF-exp\u003c/em\u003e) Δ(\u003cem\u003ebio uvrB\u003c/em\u003e) [λ Δ\u003cem\u003eBam\u003c/em\u003e N\u003csup\u003e+\u003c/sup\u003e\u003cem\u003ec\u003c/em\u003eI857 Δ(CroattR)]) harbouring the pEX-CYT plasmid were kindly provided by Dr Naheed Kaderbhai. The strains carry different copy numbers of cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e gene, ranging from 0 to 6, which are labelled as N0 to N6. All strains were streaked onto LB agar (Formedium, UK) prior to growing overnight at 37\u0026deg;C with 75 \u0026micro;g/mL ampicillin supplemented LB agar for selection of the desired colonies. The cells were cultured for the CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production as described by (Kaderbhai et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Tengsuttiwat et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Briefly, 1 mL overnight culture was inoculated into a 250 mL Erlenmeyer flask containing 25 mL of LB medium with ampicillin (75 \u0026micro;g/mL). The cells were first incubated shaking at 150 rpm and 30\u0026deg;C for 2.5 h to reach the mid-log phase, the incubation temperature was then immediately increased to 38.5\u0026deg;C to activate the promoter, \u003cem\u003eλP\u003c/em\u003e\u003csub\u003eL\u003c/sub\u003e. The culture was incubated for 9 h to allow for the protein production at 38.5\u0026deg;C, while a control set (N0-N6) was incubated for the same duration at 30\u0026deg;C.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Sample preparation\u003c/h2\u003e \u003cp\u003eAfter cultivation, each of the bacterial samples were collected as 1 mL aliquots to determine their biomass by measuring the optical density at 600 nm (OD\u003csub\u003e600nm\u003c/sub\u003e). The bacterial samples were also collected for metabolic profiling by immediately halting the metabolic process using solvent-based quenching, following which bacterial metabolites were subsequently extracted from the samples using cold methanol. Derivatisation was performed to maximise the range of metabolites that can be analysed using GC-MS; i.e., lower boiling point and improve thermal stability. The full procedures for quenching, extraction and derivatisation were undertaken as described previously and outlined below (Winder et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.2.1 Quenching and extraction\u003c/h2\u003e \u003cp\u003eTo immediately halt metabolic activities in bacterial cells, the samples were quenched with x2 volume of 60% methanol (at -48\u0026deg;C), where 15 mL sample was quenched with 30 mL of cold 60% methanol. These were immediately centrifuged at 4,347 \u003cem\u003eg\u003c/em\u003e for 5 min at -9 \u0026ordm;C. The supernatant was decanted, followed by a 5 min centrifugation to remove residual supernatant via pipetting. The quenched pellets were then kept frozen at -80\u0026deg;C, until further analysis. At the point of analysis, upon defrosting the samples, 500 \u0026micro;L of cold (-48\u0026deg;C) 80% methanol was added to the quenched biomass to resuspend the pellets and these were transferred to 2 mL microcentrifuge tubes. This step was repeated to minimise the loss of biomass during the transfer. To extract the intracellular metabolites, freeze-thaw cycling approach using liquid nitrogen was performed three times. The cell suspension was then centrifuged at 14,500 \u003cem\u003eg\u003c/em\u003e, -9\u0026deg;C for 10 min prior to biomass normalisation of the sample extracts according to their OD\u003csub\u003e600nm\u003c/sub\u003e. Different volumes of cold (-48\u0026deg;C) 80% methanol were added to the biomass-normalised samples to ensure concentration uniformity of all samples. Following this a 700 \u0026micro;L sample volume was taken from each tube. In addition, an equal volume of these normalised extracted samples (250 \u0026micro;L) was collected from each tube, and pooled together to be used as a biological quality control (QC), as detailed in (Broadhurst et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and then divided into aliquots of 450 \u0026micro;L. An aliquot (100 \u0026micro;L) of an internal standard (IS, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) was added to the samples, followed by drying using a speed-vac centrifuge (ScanVac MaxiVac Alpha \u0026amp; MaxiVac Beta Vacuum Concentrator, LaboGene, Denmark) for 6 h.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.2.2 Derivatisation\u003c/h2\u003e \u003cp\u003eDried extracts were derivatised with 50 \u0026micro;L of 20 mg/mL O-methoxyamine \u0026bull; HCl dissolved in pyridine, and the samples were vortexed for 10 s prior to heating at 65\u0026deg;C for 40 min. This was followed by the addition of 50 \u0026micro;L of \u003cem\u003eN\u003c/em\u003e-Methyl-\u003cem\u003eN\u003c/em\u003e-(trimethylsilyl)trifluoroacetamide (MSTFA) (Merck Life Science UK Limited, UK) to all the samples, which were then vortexed for 10 s before heating again as described above. Finally, 20 \u0026micro;L of retention index solution (RI, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) was added to each sample, followed by vortexing and centrifugation at 17,000 \u003cem\u003eg\u003c/em\u003e for 15 min at 4\u0026deg;C (Muhamadali et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015b\u003c/span\u003e; Dunn et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2011b\u003c/span\u003e; Begley et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Aliquots (100 \u0026micro;L) of the derivatised samples were then transferred into GC vials.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Instrumental setup\u003c/h2\u003e \u003cp\u003eA J\u0026amp;W DB-5ms GC Column (Agilent Technologies, Inc.) was used in a split mode (ratio of 20:1) with helium as carrier gas in constant flow of 1 mL/min. An Agilent 8890 GC instrument equipped with an Agilent 7250 GC/Q-TOF was operated with an initial temperature program of 70\u0026deg;C for 4 min. The oven temperature was then increased at 20\u0026deg;C/min to 300\u0026deg;C and held for another 4 min before cooling down to 70\u0026deg;C. The transfer line temperature was 280\u0026deg;C and a 25 mL/min gas saver flow was switched on after 3 min. The mass range used was 45\u0026ndash;600 Da with an acquisition rate of 10 spectra per second (Begley et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Data analysis\u003c/h2\u003e \u003cp\u003eInitially, the GC-MS raw data files were converted into mzML format using ProteoWizard software (64-bit version). The transformed data files were subsequently processed for deconvolution and blank subtraction using Mass Spectrometry-Data Independent AnaLysis software (MS-DIAL version 4.9.2). Deconvoluted data were corrected to remove technical variation using succinic acid \u003cem\u003ed\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e, an internal standard, which showed the most reproducibility in terms of standard deviation when compared among the four deuterated standards (mixture of succinic acid \u003cem\u003ed\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e, glycine \u003cem\u003ed\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e, malonic acid \u003cem\u003ed\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e and alanine \u003cem\u003ed\u003c/em\u003e\u003csub\u003e7\u003c/sub\u003e, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Signal drift was further corrected using QC-based LOWESS normalisation. For data curation, first, the relative standard deviation (RSD) of each peak area was calculated from normalised pooled QCs and only those metabolite features with RSD\u0026thinsp;\u0026le;\u0026thinsp;30% were kept for further analysis. Second, extracted ion chromatograms (EIC) of each peak were investigated manually to remove those with poor alignment or low signal to noise ratio. Furthermore, the features with a Gaussian chromatographic peak shape were separated and further analysed. Among these curated features, some were annotated as known compounds based on the GC-MS DB-Public-KovatsRI-VS3.msp\u0026rsquo; spectral library, while others were labelled as unknown. Those features that were annotated based on their matching with the database, including, EI spectra and retention index, were considered to be identified to MSI level 2 based on these two orthogonal properties (Sumner et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2007\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMetaboAnalyst, an online platform for metabolomics data analysis, was used to analyse pre-processed data by using \u003cem\u003et\u003c/em\u003e-test and ANOVA post-hoc analysis. All chemometric analysis including principal component analysis (PCA) and discriminant function analysis (PC-DFA), were applied using MATLAB\u0026reg; software (version r2019b, The MathWorks Inc., UK); our code is available via GitHub (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/Biospec/\u003c/span\u003e\u003cspan address=\"https://github.com/Biospec/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Significant features identified using \u003cem\u003et\u003c/em\u003e-test and ANOVA, together with those from PCA and PC-DFA loadings plots were subsequently combined and investigated. The relationship among those annotated metabolites in biological pathways of \u003cem\u003eE. coli\u003c/em\u003e was determined using pathway analysis from the MetaboAnalyst platform as well as individual evaluation using classical biochemistry pathways and the Kyoto Encyclopedia of Genes and Genomes (KEGG). The pathway analysis parameters were set as follows: the visualisation method was a scatter plot based on testing significant features; the enrichment method applied was the Hypergeometric Test; the topology measure used was Relative-betweenness Centrality; and the reference metabolome included all compounds from the selected pathway library-Prokaryotes, \u003cem\u003eEscherichia coli\u003c/em\u003e K-12 MG1655 (KEGG).\u003c/p\u003e \u003c/div\u003e"},{"header":"3 RESULTS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.1 GC-MS data analysis\u003c/h2\u003e \u003cp\u003eIn this study, the intracellular metabolic profiles of the CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing \u003cem\u003eE. coli\u003c/em\u003e N4830-1 were investigated using GC-MS. After deconvolution, 861 metabolic features were detected and normalised using the internal standard, succinic acid \u003cem\u003ed\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e. These metabolic features were subsequently pre-processed by blank filtration, resulting in 580 features prior to data curation by removal of irreproducible features in the QCs (RSD\u0026thinsp;\u0026gt;\u0026thinsp;30%). Data were further curated by elimination of those features with non-Gaussian chromatographic peak shape, using the EIC chromatogram, resulting in a total of 340 features being retained. These metabolic features were subsequently subjected to PCA to study and to compare the metabolic changes among the seven bacterial strains in both conditions: under recombinant protein induction at 38.5\u0026deg;C and control groups at 30\u0026deg;C. A PCA scores plot of the GC-MS data collected for all the samples was constructed, which illustrated a separation between the control (circles) and induced (stars) samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). This segregation between the two culture conditions was predominantly according to PC1 axis, which accounted for 41.35% of the total explained variance (TEV).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe first 20 PCs, with a TEV of 92.07%, were next analysed using DFA with \u003cem\u003ea priori\u003c/em\u003e groups of 14 classes, i.e., seven individual strains and two conditions \u0026ndash; as no information on either gene copy or culture conditions were provided to the DFA algorithm, these methods can be considered semi-supervised. This combined processes called PC-DFA resulted in a scores plot (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) of the GC-MS data, which displayed a trajectory similar to that seen in the PCA scores plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), showing the two clusters of CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing and non-producing \u003cem\u003eE. coli\u003c/em\u003e strains. Whereas DF1 axis illustrates the distinct cultivation conditions, DF2 axis provided further informative trend showing the separation of strains N0 to N6 from negative to positive (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Notably, the results from PC-DFA do not require validation, as the observed trend followed an ordered sequence from strain N0 to N6 without explicitly providing this order to the algorithm. The observed trend is consistent with the increasing number of cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e gene copies (0\u0026ndash;6), which is expected to correlate with target protein expression levels; however, contributions from plasmid-associated metabolic burden cannot be excluded as a potential factor influencing the metabolic profiles.\u003c/p\u003e \u003cp\u003eTo investigate the metabolic changes within each condition, the data were analysed separately, based on the two culturing conditions (induced cultures or uninduced controls). The PCA scores plot of the induced samples illustrated a trend based on the PC2 axis (14.87% TEV), showing strains N0 and N1 on the negative side, strain N2 on the middle and strain N3-N6 on the positive side (Figure S2b). However, as expected, no clear trends were detected in the PCA scores plot of the control samples (Figure S2a). Furthermore, while the PC-DFA scores plot of the data from the control condition demonstrated no specific trend (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), that of the induced group illustrated separations according to DF1 axis from strain N0 to N6 as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb. These trends on both PCA and PC-DFA scores plots of induced samples were further evaluated individually and investigated via pathway analysis in MetaboAnalyst. The input for this analysis was the significant metabolic features identified by all chemometrics and statistical tests, as described in the methods section. These are also listed in Tables S2 and S3. MetaboAnalyst enables comprehensive metabolic pathway analysis by integrating pathway enrichment analysis with pathway topology analysis. The results of this analysis suggested that differential levels of CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production modulated multiple metabolic pathways, including the TCA cycle, glycolysis/gluconeogenesis, pentose phosphate pathway, glyoxylate and dicarboxylate metabolism, purine metabolism, pyrimidine metabolism, and glycerolipid metabolism.\u003c/p\u003e \u003cp\u003eCYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production is regulated by a \u003cem\u003eλP\u003c/em\u003eL promoter that requires an increase in cultivation temperature (30 \u0026loz; 38.5\u0026deg;C) to induce gene expression. Therefore, it was necessary to exclude any metabolic effects attributable solely to this 8.5\u0026deg;C temperature shift, to enable accurate investigation and detection of metabolic changes specifically associated with CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production. To assess the impact of temperature induction, the bacterial strain N0 (host containing the plasmid but lacking the cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e gene) was analysed under the two culture conditions using PCA (Figure S3) and in combination with \u003cem\u003et\u003c/em\u003e-test statistical evaluation. The important annotated metabolites listed in Table S4 were thus identified as being influenced exclusively by temperature variation rather than by CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4 DISCUSSION","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.1 GC-MS data interpretation\u003c/h2\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e4.1.1 Effect on energy metabolisms\u003c/h2\u003e \u003cp\u003eRegarding the chemometric and statistical analyses, significant metabolic features were mapped to various fundamental metabolic pathways in the bacterial host. This includes central metabolic processes such as glycolysis and TCA cycle. These pathways, in conjunction with the electron transport chain, play a crucial role in cellular energy production. Many metabolic features within these pathways were observed in significantly lower levels in CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing strains compared to the control. This includes metabolites such as pyruvate, oxaloacetate, fumarate, lactate, citrate, isocitrate, gluconic acid, glucuronic acid and 3-phosphoglycerate (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe reduced abundance of these metabolites indicates their substantial utilisation during temperature-induced cultivation, likely reflecting the high energetic demands associated with recombinant protein production. Notably, under induction conditions, nicotinamide was identified as a key metabolite whose concentration progressively declined with increasing copy numbers of the cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e gene across bacterial strains N0 to N4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Nicotinamide is a precursor of the essential energy-related cofactor nicotinamide adenine dinucleotide (NAD, and it can be biosynthesised from tryptophan, which was likewise identified as a significant metabolite exhibiting differential levels between the two bacterial culture conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e4.1.2 Effect on bacterial stress responses\u003c/h2\u003e \u003cp\u003eSeveral metabolites associated with the pentose phosphate pathway, the glyoxylate shunt, and glutathione metabolism exhibited significant changes in abundance between the induced and control samples. Specifically, there was a decrease in the levels of pyruvate, oxaloacetate, citric acid, isocitric acid, gluconic acid, glutamate, putrescine, and ribose-5-phosphate in the CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing condition. Under this condition, induced samples exhibited a pronounced increase in the levels of cadaverine and 5-aminovaleric acid (5-AVA), metabolites that are likely derived from lysine catabolism. Additional metabolites, including the polyamines cadaverine, putrescine, and spermidine, were also detected. These compounds are known to be upregulated in response to a range of physicochemical stresses, such as elevated temperature, osmotic fluctuations, reactive oxygen species, and ultraviolet radiation (Rhee et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Specifically, putrescine concentrations declined under induction, whereas cadaverine displayed an increasing trend.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003e4.1.3 Effect on permeability of bacterial host cells\u003c/h2\u003e \u003cp\u003eNumerous intermediates involved in bacterial cell wall and membrane biosynthesis exhibited significant alterations between the two culture conditions. In particular, glycerol, glycerol-3-phosphate, ethanolamine, \u003cem\u003eO\u003c/em\u003e-phosphoethanolamine, and several unsaturated fatty acids, including oleic acid, elaidic acid, and palmitoleic acid were among the metabolites affected by the temperature-induced production of CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e, serving as intermediates in phospholipid biosynthesis (Sinensky, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1971\u003c/span\u003e). Moreover, bacterial cell wall synthesis appeared to be impacted, as evidenced by the significantly reduced levels of \u003cem\u003eN\u003c/em\u003e-acetylglucosamine in the induced condition compared to the control set.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003e4.1.4 Effect on nucleotide metabolisms\u003c/h2\u003e \u003cp\u003eNucleotide metabolism, specifically purine and pyrimidine metabolism, also displayed alterations under the investigated conditions. While certain metabolites detected in these pathways exhibited a decreasing trend in the CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing condition, such as 5,6-dihydrouracil, 2-deoxyuridine, glutamine, hypoxanthine, and thymine, others such as orotic acid and 3-aminoisobutyric acid showed a significant increase. However, it is worth noting that cytosine and orotic acid from pyrimidine metabolism, and guanine from purine metabolism, exhibited significant alterations under both cultivation conditions (control and induced samples) and strain effects (inducing strains N0-N6). These findings indicate that the metabolic changes likely reflect the combined effects of heat induction and CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Biochemical discussions\u003c/h2\u003e \u003cp\u003eMetabolic profiling of the CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing \u003cem\u003eE. coli\u003c/em\u003e N4830-1 was investigated using GC-MS as this method focusses on central carbon and nitrogen metabolism. The samples were cultured under optimal conditions, first at 30.0\u0026deg;C, then using an induction temperature of 38.5\u0026deg;C (for the \u003cem\u003eλP\u003c/em\u003e\u003csub\u003eL\u003c/sub\u003e promoter) and an incubation time of 9 h, as determined in our previous publication (Tengsuttiwat et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The generated GC-MS data were processed using QC practices (Broadhurst et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and then subjected to statistical tests, including \u003cem\u003et\u003c/em\u003e-tests and ANOVA, as well as chemometric analyses such as PCA and PC-DFA The clustering patterns obtained from both PCA and PC-DFA were consistent with those previously reported using FT-IR spectroscopy (Tengsuttiwat et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Notably, the PC-DFA scores plots demonstrated a clear gene copy number-dependent trend among induced samples (N0-N6) across three independent batch cultures, underscoring the reproducibility and robustness of the overall analytical workflow. Importantly, the present GC-MS analysis further provided detailed biochemical insights into specific metabolites that play crucial roles under different cultivation conditions and among the seven bacterial strains. The significant metabolites were mapped onto metabolic pathways, revealing major effects on energy-related mechanisms, bacterial cell envelope biosynthesis, and nucleotide metabolisms.\u003c/p\u003e \u003cp\u003eThe synthesis of recombinant proteins is an energy-intensive process, especially in these strains where CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production reached 0.73 mM calculated from data in our previous publication (Tengsuttiwat et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) using the extinction coefficient value of 171 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e mM\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e at 423 nm (G\u0026oacute;mez-Tabales et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e.) Consistent with this, previous studies have reported a significant decrease in intracellular ATP levels following temperature-induced recombinant production of human fibroblast growth factor protein (Templeton et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Additionally, cells producing recombinant proteins also exhibit increased metabolic flux through the TCA cycle and related pathways to meet the elevated ATP demand required for intensive protein synthesis and stress response mechanisms (Wittmann et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Consequently, the substantial ATP requirement associated with CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e\u003cem\u003e5\u003c/em\u003e\u003c/sub\u003e production likely contributed to the observed depletion of metabolites involved in key energy-generating pathways, including the TCA cycle, glycolysis/gluconeogenesis, the pentose phosphate pathway, as well as glyoxylate and dicarboxylate metabolism. This observation is also supported by prior studies indicating that glycolytic flux in \u003cem\u003eE. coli\u003c/em\u003e is highly responsive to variations in ATP demand (Koebmann et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2002\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDuring the recombinant protein production, the bacterial host exhibited significant physiological alterations in response to the temperature change between the control and inducing culture conditions. These adaptations prominently involved modifications in cell wall and membrane metabolisms, particularly influencing membrane fluidity, as previously documented in the literature (Russell, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Membrane integrity in \u003cem\u003eE. coli\u003c/em\u003e is influenced not only by heat stress but also by environmental factors, including temperature and pH fluctuations (Rowlett et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In the present study, an increase in pH from 6.8 to 8.2 was observed following induction (Table S5). Such alkalinisation may be attributed to the oxidation of carbon sources, depletion of amino acids, and the accumulation of basic metabolic by-products (S\u0026aacute;nchez-Clemente et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Additionally, recombinant protein production imposes a considerable metabolic burden on the host, triggering adaptive responses including rearrangement of lipid composition (Ami et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), an increase in saturated fatty acids involved in phospholipid metabolism (Sinensky, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1971\u003c/span\u003e), and elevated fluxes of pyruvate, acetate, and lactate formation (Wittmann et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Collectively, these adaptations reflect coordinated cellular strategies to maintain membrane stability and metabolic balance under the combined stresses of temperature induction and recombinant protein synthesis.\u003c/p\u003e \u003cp\u003eFurthermore, stress-related proteins are often synthesised during recombinant protein production, leading to an increased ATP demand (Hoffmann and Rinas, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Heat shock proteins, for example, can be triggered to cope with the stressful environment (Schumann, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Both the membrane adaptation mentioned above, and the synthesis of stress-responsive proteins are energy-intensive processes. This could lead to significant depletion of glucose for energy production, thereby promoting the utilisation of alternative carbon sources such as acetate commonly associated with recombinant protein synthesis. This utilisation was evidenced by the observed impact on the glyoxylate shunt due to CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production. However, glucose levels remained relatively stable among all the samples. This may be explained by the fact that glucose can also be synthesised from fatty acids through gluconeogenesis and glyoxylate cycle (Cordero et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). In this process, the glyoxylate cycle provides succinate, which enters the TCA cycle and subsequently feeds into gluconeogenesis, thereby maintaining the detected glucose level. This observation also supports the finding of reduction in unsaturated fatty acids, such as oleic acid, elaidic acid, and palmitoleic acid, in the induced samples. The glyoxylate shunt, along with the pentose phosphate pathway, plays a crucial role in carbon homeostasis. The significant alterations observed in gluconic acid and pyruvate within the pentose phosphate pathway suggest metabolic adaptation to balance carbon flux. When preferred carbon sources like glucose become insufficient to meet increased energetic requirements, \u003cem\u003eE. coli\u003c/em\u003e can activate the glyoxylate shunt to utilise simpler carbon sources such as acetate and fatty acids.\u003c/p\u003e \u003cp\u003eAdditionally, oxidative stress is a common consequence of recombinant protein production due to increased cellular energy demand (Templeton et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). In response, bacterial cells may utilise pathways such as the pentose phosphate pathway, glyoxylate shunt, and glutathione metabolism to mitigate the effects of such stress (Stincone et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Ahn et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Ozkul and Kocak, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The pentose phosphate pathway is essential for supplying precursors for nucleotide and amino acid biosynthesis. Correspondingly, significant alterations in metabolites such as cytosine, orotic acid, \u003cem\u003eO\u003c/em\u003e-phosphoethanolamine, myristic acid, heptanoic acid, serine, and tryptamine across the seven CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e-producing strains suggest that these biosynthetic pathways were affected by heat-induced CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e expression and potential differences in cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e gene copy number (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Furthermore, the untargeted metabolomics approach used in this study does not allow absolute quantification of metabolites (rather a peak ratio against an internal standard), and some detected compounds, such as amino acids, may originate from either endogenous synthesis or uptake from the nutrient-rich LB medium. Future studies incorporating metabolic footprinting would help clarify metabolite origins and consumption dynamics.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 CONCLUSIONS","content":"\u003cp\u003eIn this study, an untargeted metabolomic approach was employed utilising GC-MS to examine the metabolic profiles of seven strains of \u003cem\u003eE. coli\u003c/em\u003e N4830-1, in the context of temperature-induced production of CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e. The most notable alterations were observed in the TCA cycle, where key metabolites such as pyruvate, oxaloacetate, citrate, isocitrate, and fumarate exhibited marked depletion in the induced condition. This observation indicates the heightened energy requirements inherent to recombinant protein production systems. The imposition of heat stress and the demands of recombinant protein production necessitate metabolic adaptations within bacterial hosts. Significant alterations in detected metabolites suggest responses aimed at fortifying bacterial cell wall and membrane biosynthesis, as well as deploying defensive mechanisms against oxidative stress through pathways such as the pentose phosphate pathway, glyoxylate shunt, and glutathione metabolism. The aforementioned mechanisms all entail energy expenditure, thereby increasing the demand for energy within host cells. However, to gain a comprehensive understanding of the metabolic burden associated with recombinant CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production among the seven strains, future studies could explore a combination of metabolic profiling (intracellular metabolites), footprinting (extracellular metabolites), and lipidomics (membrane lipids). Additionally, targeted metabolomics approach could be employed to facilitate quantitative analysis. Since energy consumption played a crucial role among all the samples, adenylate energy charge is worth considering as it represents the available metabolic energy of a host at the time of sampling (Atkinson and Walton, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e1967\u003c/span\u003e; Atkinson, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1968\u003c/span\u003e). In addition, fluxomics is a valuable metabolomic approach for monitoring metabolic pathways (Emwas et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); for instance, the analysis of metabolic flux using \u003csup\u003e13\u003c/sup\u003eC-labeled molecules. Such investigations may reveal potential bottlenecks in the recombinant protein production process and support optimisation of the production system, but only really work in highly defined medium with a sole carbon supply which may affect recombinant protein production as current production systems are replete with nutrients. Subsequent genetic engineering strategies or media optimisation, for example, could be proposed to enhance the efficiency of the production system.\u003c/p\u003e \u003cp\u003eIn conclusion, this study shows that CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production triggers coordinated metabolic changes that impact energy metabolism, cell wall and membrane biosynthesis, nucleotide metabolism, and bacterial stress responses. These insights may in the future offer valuable guidance for optimising recombinant protein production processes in prokaryotic hosts.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eSupplementary Information\u003c/strong\u003e The online version contains supplementary material.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003eWe thank the University of Liverpool and the Royal Thai Government for financial support. R.G. thanks UK BBSRC for funding metabolomics (BB/W019558/1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThanyaporn Tengsuttiwat:\u003csup\u003e\u0026nbsp;\u003c/sup\u003eexperimental design, sample collection and preparation, GC-MS data analysis, data interpretation and manuscript preparation.\u003c/p\u003e\n\u003cp\u003eAdam Burke: GC-MS data analysis and manuscript preparation.\u003c/p\u003e\n\u003cp\u003eNaheed Nazly Kaderbhai and Joe Gallagher: genetic engineering of the strains,\u0026nbsp;growth condition optimisation, and manuscript preparation.\u003c/p\u003e\n\u003cp\u003eRoyston Goodacre: co-principal investigator, experimental design, data interpretation, and manuscript preparation.\u003c/p\u003e\n\u003cp\u003eHowbeer Muhamadali: co-principal investigator, experimental design, data interpretation, and manuscript preparation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclarations\u0026nbsp;\u003c/strong\u003eRG is EiC of \u003cem\u003eMetabolomics\u003c/em\u003e, he was not involved in the review process for this paper.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eAll authors declare that they have no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u0026nbsp;\u003c/strong\u003eGC-MS data is freely available at MetaboLights repository with the unique identifier REQ20250520210627 (www.ebi.ac.uk/metabolights/).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAhn, S., Jung, J., Jang, I.-A., Madsen, E.L., and Park, W. (2016). Role of Glyoxylate Shunt in Oxidative Stress Response. Journal of Biological Chemistry 291(22), 11928-11938. doi: 10.1074/jbc.M115.708149.\u003c/li\u003e\n\u003cli\u003eAmi, D., Natalello, A., Schultz, T., Gatti-Lafranconi, P., Lotti, M., Doglia, S.M., \u003cem\u003eet al\u003c/em\u003e. (2009). Effects of recombinant protein misfolding and aggregation on bacterial membranes. Biochim Biophys Acta 1794(2), 263-269. doi: 10.1016/j.bbapap.2008.10.015.\u003c/li\u003e\n\u003cli\u003eAtkinson, D.E. (1968). The energy charge of the adenylate pool as a regulatory parameter. Interaction with feedback modifiers. Biochemistry 7(11), 4030-4034. doi: 10.1021/bi00851a033.\u003c/li\u003e\n\u003cli\u003eAtkinson, D.E., and Walton, G.M. (1967). Adenosine triphosphate conservation in metabolic regulation. Rat liver citrate cleavage enzyme. J Biol Chem 242(13), 3239-3241.\u003c/li\u003e\n\u003cli\u003eBailey, J.E. (1991). Toward a Science of Metabolic Engineering. Science 252(5013), 1668-1675.\u003c/li\u003e\n\u003cli\u003eBegley, P., Francis-McIntyre, S., Dunn, W.B., Broadhurst, D.I., Halsall, A., Tseng, A., \u003cem\u003eet al\u003c/em\u003e. (2009). Development and performance of a gas chromatography-time-of-flight mass spectrometry analysis for large-scale nontargeted metabolomic studies of human serum. Anal Chem 81(16), 7038-7046. doi: 10.1021/ac9011599.\u003c/li\u003e\n\u003cli\u003eBioSpace, 2024. \u003cem\u003eRecombinant proteins market size to hit USD 8.66 billion by 2034\u003c/em\u003e. Available at: https://www.biospace.com/press-releases/recombinant-proteins-market-size-to-hit-usd-8-66-billion-by-2034 (Accessed: 17 February 2026).\u003c/li\u003e\n\u003cli\u003eBroadhurst, D., Goodacre, R., Reinke, S. N., Kuligowski, J., Wilson, I. D., Lewis, M. R., \u0026amp; Dunn, W. B. (2018). Guidelines and considerations for the use of system suitability and quality control samples in mass spectrometry assays applied in untargeted clinical metabolomic studies. Metabolomics, 14(6), 72. https://doi.org/10.1007/s11306-018-1367-3\u003c/li\u003e\n\u003cli\u003eCordero, P., Campion, J., Milagro, F.I., Marzo, F., and Martinez, J.A. (2008). Fat-to-glucose interconversion by hydrodynamic transfer of two glyoxylate cycle enzyme genes. Lipids in Health and Disease 7(1), 49. doi: 10.1186/1476-511X-7-49.\u003c/li\u003e\n\u003cli\u003eDunn, W.B., Broadhurst, D.I., Atherton, H.J., Goodacre, R., and Griffin, J.L. (2011a). Systems level studies of mammalian metabolomes: the roles of mass spectrometry and nuclear magnetic resonance spectroscopy. Chemical Society Reviews 40(1), 387-426. doi: 10.1039/B906712B. \u003c/li\u003e\n\u003cli\u003eDunn, W.B., Broadhurst, D., Begley, P., Zelena, E., Francis-McIntyre, S., Anderson, N., \u003cem\u003eet al\u003c/em\u003e. (2011b). Procedures for large-scale metabolic profiling of serum and plasma using gas chromatography and liquid chromatography coupled to mass spectrometry. Nature Protocols 6(7), 1060-1083. doi: 10.1038/nprot.2011.335.\u003c/li\u003e\n\u003cli\u003eDunn, W.B., and Ellis, D.I. (2005). Metabolomics: Current analytical platforms and methodologies. TrAC Trends in Analytical Chemistry 24(4), 285-294. doi: 10.1016/j.trac.2004.11.021.\u003c/li\u003e\n\u003cli\u003eDunn, W.B., Erban, A., Weber, R.J.M., Creek, D.J., Brown, M., Breitling, R., \u003cem\u003eet al\u003c/em\u003e. (2013). Mass appeal: metabolite identification in mass spectrometry-focused untargeted metabolomics. Metabolomics 9(1), 44-66. doi: 10.1007/s11306-012-0434-4.\u003c/li\u003e\n\u003cli\u003eEmwas, A.H., Szczepski, K., Al-Younis, I., Lachowicz, J.I., and Jaremko, M. (2022). Fluxomics - New Metabolomics Approaches to Monitor Metabolic Pathways. Front Pharmacol 13, 805782. doi: 10.3389/fphar.2022.805782.\u003c/li\u003e\n\u003cli\u003eFiehn, O. (2002). Metabolomics \u0026mdash; The Link Between Genotypes and Phenotypes. Plant Molecular Biology, v.48, 155-177 (2002) 48. doi: 10.1007/978-94-010-0448-0_11.\u003c/li\u003e\n\u003cli\u003eG\u0026oacute;mez-Tabales, J., Garc\u0026iacute;a-Mart\u0026iacute;n, E., Ag\u0026uacute;ndez, J.A.G., and Gutierrez-Merino, C. (2020). Modulation of CYP2C9 activity and hydrogen peroxide production by cytochrome b(5). Sci Rep 10(1), 15571. doi: 10.1038/s41598-020-72284-0.\u003c/li\u003e\n\u003cli\u003eHan, Q., and Eiteman, M.A. (2019). Acetate formation during recombinant protein production in Escherichia coli K-12 with an elevated NAD(H) pool. Eng Life Sci 19(11), 770-780. doi: 10.1002/elsc.201900045.\u003c/li\u003e\n\u003cli\u003eHoffmann, F., and Rinas, U. (2004). \u0026quot;Stress Induced by Recombinant Protein Production in Escherichia coli,\u0026quot; in Physiological Stress Responses in Bioprocesses: -/-. (Berlin, Heidelberg: Springer Berlin Heidelberg), 73-92.\u003c/li\u003e\n\u003cli\u003eHollywood, K., Brison, D.R., and Goodacre, R. (2006). Metabolomics: Current technologies and future trends. PROTEOMICS 6(17), 4716-4723. doi: 10.1002/pmic.200600106.\u003c/li\u003e\n\u003cli\u003eKaderbhai, N.N., , G., JOSEPH, HE, MINGYUE, and and KADERBHAI, M.A. (1992). A Pink Bacterium as a Reporter System Signaling Expression of a Recombinant Protein. DNA and Cell Biology 11(7), 567-577. doi: 10.1089/dna.1992.11.567.\u003c/li\u003e\n\u003cli\u003eKaderbhai, N.N., Broadhurst, D.I., Ellis, D.I., Goodacre, R., and Kell, D.B. (2003). Functional genomics via metabolic footprinting: monitoring metabolite secretion by Escherichia coli tryptophan metabolism mutants using FT-IR and direct injection electrospray mass spectrometry. Comparative and functional genomics 4(4), 376-391. doi: 10.1002/cfg.302.\u003c/li\u003e\n\u003cli\u003eKirk, O., Borchert, T.V., and Fuglsang, C.C. (2002). Industrial enzyme applications. Current Opinion in Biotechnology 13(4), 345-351. doi: https://doi.org/10.1016/S0958-1669(02)00328-2.\u003c/li\u003e\n\u003cli\u003eKoebmann, B.J., Westerhoff, H.V., Snoep, J.L., Nilsson, D., and Jensen, P.R. (2002). The glycolytic flux in Escherichia coli is controlled by the demand for ATP. J Bacteriol 184(14), 3909-3916. doi: 10.1128/jb.184.14.3909-3916.2002.\u003c/li\u003e\n\u003cli\u003eLeis, B., Angelov, A., and Liebl, W. (2013). Screening and expression of genes from metagenomes. Adv Appl Microbiol 83, 1-68. doi: 10.1016/b978-0-12-407678-5.00001-5.\u003c/li\u003e\n\u003cli\u003eMashego, M.R., Rumbold, K., De Mey, M., Vandamme, E., Soetaert, W., and Heijnen, J.J. (2007). Microbial metabolomics: past, present and future methodologies. Biotechnology Letters 29(1), 1-16. doi: 10.1007/s10529-006-9218-0.\u003c/li\u003e\n\u003cli\u003eMuhamadali, H., Xu, Y., Ellis, D., Trivedi, D., Rattray, N., Bernaerts, K., \u003cem\u003eet al\u003c/em\u003e. (2015a). Metabolomics investigation of recombinant mTNF\u0026alpha; production in \u003cem\u003eStreptomyces lividans\u003c/em\u003e. \u003cem\u003eMicrobial Cell Factories\u003c/em\u003e 14. doi: 10.1186/s12934-015-0350-1.\u003c/li\u003e\n\u003cli\u003eMuhamadali, H., Xu, Y., Ellis, D.I., Allwood, J.W., Rattray, N.J., Correa, E., \u003cem\u003eet al\u003c/em\u003e. (2015b). Metabolic Profiling of \u003cem\u003eGeobacter sulfurreducens\u003c/em\u003e during Industrial Bioprocess Scale-Up. Appl \u003cem\u003eEnviron Microbiol\u003c/em\u003e 81(10), 3288-3298. doi: 10.1128/aem.00294-15.\u003c/li\u003e\n\u003cli\u003eMuhamadali, H., Xu, Y., Morra, R., Trivedi, D.K., Rattray, N.J.W., Dixon, N., \u003cem\u003eet al\u003c/em\u003e. (2016). Metabolomic analysis of riboswitch containing \u003cem\u003eE. coli\u003c/em\u003e recombinant expression system. Molecular BioSystems 12(2), 350-361. doi: 10.1039/C5MB00624D.\u003c/li\u003e\n\u003cli\u003eOzkul, C., and Kocak, E. (2020). Metabolic response of Escherichia coli to subinhibitory concentration of ofloxacin. Journal of Research in Pharmacy 24, 593-601. doi: 10.35333/jrp.2020.207.\u003c/li\u003e\n\u003cli\u003ePrabhu, S.K., Yang, Q., Tong, X., and Wang, L.X. (2021). Exploring a combined Escherichia coli-based glycosylation and in vitro transglycosylation approach for expression of glycosylated interferon alpha. Bioorg Med Chem 33, 116037. doi: 10.1016/j.bmc.2021.116037.\u003c/li\u003e\n\u003cli\u003eRhee, H.J., Kim, E.J., and Lee, J.K. (2007). Physiological polyamines: simple primordial stress molecules. J Cell Mol Med 11(4), 685-703. doi: 10.1111/j.1582-4934.2007.00077.x.\u003c/li\u003e\n\u003cli\u003eRowlett, V.W., Mallampalli, V., Karlstaedt, A., Dowhan, W., Taegtmeyer, H., Margolin, W., \u003cem\u003eet al\u003c/em\u003e. (2017). Impact of Membrane Phospholipid Alterations in Escherichia coli on Cellular Function and Bacterial Stress Adaptation. J Bacteriol 199(13). doi: 10.1128/jb.00849-16.\u003c/li\u003e\n\u003cli\u003eRussell, A.D. (2003). Lethal effects of heat on bacterial physiology and structure. Sci Prog 86(Pt 1-2), 115-137. doi: 10.3184/003685003783238699.\u003c/li\u003e\n\u003cli\u003eS\u0026aacute;nchez-Clemente, R., Guijo, M. I., Nogales, J., \u0026amp; Blasco, R. (2020). Carbon Source Influence on Extracellular pH Changes along Bacterial Cell-Growth. \u003cem\u003eGenes\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(11), 1292. https://doi.org/10.3390/genes11111292\u003c/li\u003e\n\u003cli\u003eSchumann, W. (2016). Regulation of bacterial heat shock stimulons. \u003cem\u003eCell Stress and Chaperones\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(6), 959\u0026ndash;968. https://doi.org/10.1007/s12192-016-0727-z\u003c/li\u003e\n\u003cli\u003eSinensky, M. (1971). Temperature control of phospholipid biosynthesis in \u003cem\u003eEscherichia coli\u003c/em\u003e. J Bacteriol 106(2), 449-455. doi: 10.1128/jb.106.2.449-455.1971.\u003c/li\u003e\n\u003cli\u003eSingh, R., Kumar, M., Mittal, A., and Mehta, P.K. (2016). Microbial enzymes: industrial progress in 21st century. 3 Biotech 6(2), 174. doi: 10.1007/s13205-016-0485-8.\u003c/li\u003e\n\u003cli\u003eStephanopoulos, G.N., and Vallino, J. (1991). Network Rigidity And Metabolic Engineering In Metabolite Overproduction. Science (New York, N.Y.) 252, 1675-1681. doi: 10.1126/science.1904627.\u003c/li\u003e\n\u003cli\u003eStincone, A., Prigione, A., Cramer, T., Wamelink, M.M.C., Campbell, K., Cheung, E., \u003cem\u003eet al\u003c/em\u003e. (2015). The return of metabolism: biochemistry and physiology of the pentose phosphate pathway. Biological Reviews 90(3), 927-963. doi: https://doi.org/10.1111/brv.12140.\u003c/li\u003e\n\u003cli\u003eSumner, L.W., Amberg, A., Barrett, D., Beale, M.H., Beger, R., Daykin, C.A., \u003cem\u003eet al\u003c/em\u003e. (2007). Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3(3), 211-221. doi: 10.1007/s11306-007-0082-2.\u003c/li\u003e\n\u003cli\u003eTempleton, N., Dean, J., Reddy, P., and Young, J.D. (2013). Peak antibody production is associated with increased oxidative metabolism in an industrially relevant fed-batch CHO cell culture. Biotechnology and Bioengineering 110(7), 2013-2024. doi: https://doi.org/10.1002/bit.24858.\u003c/li\u003e\n\u003cli\u003eTengsuttiwat, T., Kaderbhai, N.N., Gallagher, J., Goodacre, R., and Muhamadali, H. (2022). Metabolic Fingerprint Analysis of Cytochrome \u003cem\u003eb\u003csub\u003e5\u003c/sub\u003e\u003c/em\u003e-producing \u003cem\u003eE. coli\u003c/em\u003e N4830-1 Using FT-IR Spectroscopy. Frontiers in Microbiology 13. doi: 10.3389/fmicb.2022.874247.\u003c/li\u003e\n\u003cli\u003eTripathi, N.K., and Shrivastava, A. (2019). Recent Developments in Bioprocessing of Recombinant Proteins: Expression Hosts and Process Development. Frontiers in Bioengineering and Biotechnology 7. doi: 10.3389/fbioe.2019.00420.\u003c/li\u003e\n\u003cli\u003eWang, L., Li, A., Fang, J., Wang, Y., Chen, L., Qiao, L., \u003cem\u003eet al\u003c/em\u003e. (2023). Enhanced Cell Wall and Cell Membrane Activity Promotes Heat Adaptation of Enterococcus faecium. Int J Mol Sci 24(14). doi: 10.3390/ijms241411822.\u003c/li\u003e\n\u003cli\u003eWinder, C.L., Dunn, W.B., Schuler, S., Broadhurst, D., Jarvis, R., Stephens, G.M., \u003cem\u003eet al\u003c/em\u003e. (2008). Global Metabolic Profiling of Escherichia coli Cultures:\u0026thinsp; an Evaluation of Methods for Quenching and Extraction of Intracellular Metabolites. Analytical Chemistry 80(8), 2939-2948. doi: 10.1021/ac7023409.\u003c/li\u003e\n\u003cli\u003eWittmann, C., Weber, J., Betiku, E., Kr\u0026ouml;mer, J., B\u0026ouml;hm, D., and Rinas, U. (2007). Response of fluxome and metabolome to temperature-induced recombinant protein synthesis in \u003cem\u003eEscherichia coli\u003c/em\u003e. J Biotechnol 132(4), 375-384. doi: 10.1016/j.jbiotec.2007.07.495.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"metabolomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mebo","sideBox":"Learn more about [Metabolomics](http://link.springer.com/journal/11306)","snPcode":"11306","submissionUrl":"https://submission.nature.com/new-submission/11306/3","title":"Metabolomics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"GC-MS, metabolic profile, recombinant protein production, cytochrome b5, chemometrics","lastPublishedDoi":"10.21203/rs.3.rs-9277397/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9277397/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIntroduction: Recombinant protein production is pivotal across diverse industries, necessitating efforts to enhance both quantity and quality.\u003c/p\u003e\n\u003cp\u003eObjectives: We employed gas chromatography coupled with mass spectrometry (GC-MS) to investigate the metabolic effects of producing mammalian cytochrome \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e in the bacterial host \u003cem\u003eE. coli\u003c/em\u003e N4830-1. The model system studied involved the cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e gene being introduced on plasmids into the host with varying copy numbers (0-6) under a \u003cem\u003eλP\u003c/em\u003e\u003csub\u003eL\u003c/sub\u003e heat-sensitive promoter.\u003c/p\u003e\n\u003cp\u003eMethods: Metabolic profiling involved GC-MS analysis with quality assessments using pooled QCs and multivariate chemometric analysis.\u0026nbsp; This approach revealed specific metabolic features (identified to Level 2 of the Metabolomics Standards Initaive) correlating significantly with different cultivation conditions and variable copy numbers of the cyt \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e gene among the examined bacterial strains.\u003c/p\u003e\n\u003cp\u003eResults: Our study shows that CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production imposed a substantial energetic burden, reflected by depletion of metabolites associated with the tricarboxylic acid (TCA) cycle, glycolysis/gluconeogenesis, pentose phosphate pathway, and glyoxylate metabolism, indicating increased ATP demand. Concurrent reductions in unsaturated fatty acids and changes in lipid-related metabolites suggest membrane remodeling in response to temperature induction (needed to initiate production of CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e), as well as metabolic stress. Alterations in glyoxylate shunt and pentose phosphate pathway intermediates further indicate metabolic reprogramming to maintain carbon homeostasis, while significant changes in nucleotide- and amino acid-related metabolites suggest impacts on biosynthetic capacity associated with recombinant expression and gene copy number variation.\u003c/p\u003e\n\u003cp\u003eConclusion: Overall, these findings demonstrate that CYT \u003cem\u003eb\u003c/em\u003e\u003csub\u003e5\u003c/sub\u003e production induces coordinated metabolic adjustments affecting energy-related metabolic pathways, bacterial cell wall and membrane biosynthesis, nucleotide metabolisms and bacterial stress responses, providing valuable insights that may be useful for the optimisation of recombinant protein production processes.\u003c/p\u003e","manuscriptTitle":"Metabolic Profiling Analysis of Cytochrome B5 Production in E. Coli N4830-1 Using GC-MS","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-21 19:15:18","doi":"10.21203/rs.3.rs-9277397/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-18T15:46:31+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-23T06:23:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"124512300523643844689979321453844569569","date":"2026-04-16T05:19:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"219475955820278509973571550585678769502","date":"2026-04-14T07:50:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-14T00:42:26+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-01T13:39:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-01T13:38:43+00:00","index":"","fulltext":""},{"type":"submitted","content":"Metabolomics","date":"2026-03-31T08:57:52+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"metabolomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mebo","sideBox":"Learn more about [Metabolomics](http://link.springer.com/journal/11306)","snPcode":"11306","submissionUrl":"https://submission.nature.com/new-submission/11306/3","title":"Metabolomics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"4359ba33-165d-463a-a36d-87f464e351cd","owner":[],"postedDate":"April 21st, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-18T15:46:31+00:00","index":27,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-21T19:15:18+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-21 19:15:18","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9277397","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9277397","identity":"rs-9277397","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}