Heterologous expression of Vibrio natriegens transporters enhances the growth of Escherichia coli and Bacillus subtilis on glutamate

Abstract

Bacterial growth modulation is crucial in microbial synthetic biology. In this study, we found that glutamate is an extremely poor carbon source for Escherichia coli and Bacillus subtilis. The slow growth on glutamate can be effectively overcome by the heterologous expression of glutamate transporters from Vibrio natriegens. Our results revealed that cross-species substrate transporters could be employed to shift bacterial cellular resource allocation, offering a potential genetic strategy for modulating microbial biomass growth.

Graphical abstract

Glutamate plays a central and multifaceted role in bacterial physiology [[1], [2], [3], [4], [5]]. It connects carbon and nitrogen metabolism and acts as the principal entry point for inorganic nitrogen assimilation, primarily through the Glutamine Synthetase (GS)/ glutamate synthase (GOGAT) cycle, in which ammonia is incorporated into biomolecules [2,3,5]. Therefore, glutamate is an indispensable nitrogen donor for the biosynthesis of most nitrogen-containing biomolecules such as amino acids and nucleotides [2,3,5]. Critically, glutamate serves as a major compatible solute, and its accumulation enables bacteria to maintain turgor pressure to combat osmotic stress [1,4,5]. Concurrently, it acts as the primary counterion to balance the positive charge of potassium ions (K⁺), which are essential for pH homeostasis, membrane potential and acidic stress [1,4,5]. The role of glutamate in metabolic engineering and microbial synthetic biology is profound and multifaceted, acting as a crucial precursor for various nitrogen-containing compounds [6,7].

Cell growth is a fundamental property of bacterial cells and is primarily determined by external nutrient compositions [[8], [9], [10], [11]]. The artificial modulation of bacterial growth is of high value for both fundamental studies and applications in metabolic engineering and synthetic biology [10,12]. Previous attempts to enhance bacterial growth have often relied on adaptive laboratory evolution [13,14], whereas rational engineering strategies remain limited. In this study, we found that glutamate serves as an extremely poor carbon source for the model bacteria Escherichia coli and Bacillus subtilis. The growth-limiting effects of glutamate can be overcome by the heterologous expression of certain substrate transporters from Vibrio natriegens.

In this study, we characterized the growth rates of the model bacteria E. coli and B. subtilis on various carbon sources with defined carbon qualities. As shown in Fig. 1a–h , the growth rates of E. coli and B. subtilis varied significantly in the different carbon sources. Among the carbon sources used, glucose was the richest carbon source while glutamate was the poorest carbon source. Quantitatively, the doubling times of the four bacterial strains on glutamate were 5.7 $\pm$0.2 h (E. coli NCM3722 strain), 60.9 $\pm$3.2 h (E. coli MG1655 strain), 27.2 $\pm$2.8 h (B. subtilis strain 168), and 3.7 $\pm$0.1 h (B. subtilis DK1042 strain), respectively. In stark contrast, the marine bacterium V. natriegens, an emerging next-generation chassis species in biotechnology due to its ultrafast growth performance [[15], [16], [17], [18], [19]], was capable of growing rapidly on glutamate as the sole carbon source, with a doubling time of approximately 44.1 $\pm$0.5 min (Fig. 1i, j).

Fig. 1.

Escherichia coli and Bacillus subtilis show limited growth on glutamate. (a) Growth rates of the E. coli NCM3722 strain on various carbon sources. (b) Typical exponential growth curves of the E. coli NCM3722 strain on various carbon sources. (c) Growth rates of the E. coli MG1655 strain on various carbon sources. (d) Typical exponential growth curves of the E. coli MG1655 strain on various carbon sources. (e) Growth rates of the B. subtilis strain 168 on various carbon sources. (f) Typical exponential growth curves of the B. subtilis strain 168 on various carbon sources. (g) Growth rates of the B. subtilis DK1042 strain on various carbon sources. (h) Typical exponential growth curves of the B. subtilis DK1042 strain on various carbon sources. (i) Growth rates of various bacterial species including E. coli, B. subtilis, and Vibrio natriegens on a medium with glutamate as the sole carbon source. (j) Typical exponential growth curves of various bacterial species on a medium with glutamate as the sole carbon source, corresponding to the conditions in panel I. Error bars denote standard deviations of at least three biological replicates (n$\ge 3$). Significance analysis was based on Dunnett One-Way ANOVA using GraphPad Prism 10. *p < 0.05; **p < 0.01; ***p < 0.001.

To overcome the growth stress of E. coli and B. subtilis on glutamate, we hypothesized that introducing glutamate uptake systems from efficient glutamate-utilizing V. natriegens could enhance their growth. Glutamate uptake in bacteria is often mediated by two types of single-protein transporters: the Na⁺-coupled symporter GltS and the H⁺-coupled symporter GltP [20]. The genome of V. natriegens [16] encodes one GltS_Vnat and three GltP homologs, PN96_RS15775, PN96_RS03935, and PN96_RS10385, which we designated GltP1_Vnat, GltP2_Vnat, and GltP3_Vnat, respectively. We then used the IPTG-inducible Plac or PgracO2 promoter to drive the heterologous expression of various V. natriegens GltS and GltP transporters [18]. The heterologous expression of GltS_Vnat, GltP1_Vnat, and GltP3_Vnat elicited doubling times of 1.29 $\pm$0.02 h, 1.42 $\pm$ 0.03 h and 1.55 $\pm$0.06 h, respectively, thus substantially increasing the growth rates of wild-type E. coli NCM3722 strain on glutamate (doubling time: 5.7 $\pm$0.2 h); in contrast, the effect of the empty plasmid (regardless of whether IPTG was provided or not) was weak (Fig. 2a, b). We noted that GltP2_Vnat had no growth acceleration effect; since the expression of this glutamate transporter was driven by a strong inducible Plac promoter in a medium-copy-number pZA31 vector (p15A replicon, 20–30 copies per cell, no protein tags) [21,22], it is unlikely that the lack of growth effect of GltP2_Vnat was due to its poor expression. Instead, this could be because GltP2_Vnat has poor substrate affinity or intrinsically low catalytic efficiency; alternatively, it might have defects in membrane localization in heterologous bacterial species. We further found that the heterologous expression of GltS_Vnat strongly increased the glutamate uptake rate of the E. coli NCM3722 strain (Fig. 2c, Fig. S1), supporting the hypothesis that substrate uptake limitation accounts for the slow growth of the wild-type strain on glutamate-containing medium. A similar growth acceleration effect was observed in E. coli wild-type MG1655 strains expressing GltS_Vnat, GltP1_Vnat or GltP3_Vnat (Fig. 2d, e). Furthermore, the expression of GltS_Vnat and GltP1_Vnat substantially enhanced the growth of two B. subtilis strains, DK1042 and 168, on glutamate (Fig. 2f–j). Moreover, the heterologous expression of GltS_Vnat strongly increased the glutamate uptake rate of B. subtilis strain 168 (Fig. 2h), consistent with the results for E. coli (Fig. 2c). These findings demonstrate that the heterologous expression of cross-species glutamate transporters successfully alleviates the growth limitations of E. coli and B. subtilis on glutamate and also suggest that the growth bottleneck of E. coli and B. subtilis on glutamate lies in their substrate uptake.

Fig. 2.

Heterologous expression of Vibrio natriegens transporters boosts bacterial growth on glutamate. (a) Growth rates of wild-type Escherichia coli NCM3722 and related strains expressing various V. natriegens glutamate transporters on a medium with glutamate as the sole carbon source. (b) Typical exponential growth curves for the four E. coli NCM3722-derived strains growing on glutamate, as shown in panel A. (c) Glutamate uptake rate (J_glut) of the wild-type E. coli NCM3722 strain and its corresponding GltS_Vnat-expressing strain during growth in glutamate medium. This rate represents the net glutamate consumption derived from the culture medium, assuming negligible extracellular degradation or secretion. (d) Growth rates of wild-type E. coli MG1655 and related strains expressing various V. natriegens glutamate transporters on a medium with glutamate as the sole carbon source. (e) Typical exponential growth curves for the four E. coli MG1655-derived strains growing on glutamate, as shown in panel D. (f) Growth rates of B. subtilis 168 wild-type and related strains expressing various V. natriegens glutamate transporters on glutamate as the sole carbon source. (g) Typical exponential growth curves for the three B. subtilis 168-derived strains growing on glutamate, as shown in panel F. (h) Glutamate uptake rate (J_glut) of wild-type B. subtilis 168 and its corresponding GltS_Vnat-expressing strain during growth in glutamate-containing medium. This rate represents the net glutamate consumption derived from the culture medium, assuming negligible extracellular degradation or secretion. (i) Growth rates of wild-type B. subtilis DK1042 and related strains expressing various V. natriegens glutamate transporters on a medum containing glutamate as the sole carbon source. (j) Typical exponential growth curves for the four B. subtilis DK1042-derived strains growing on glutamate, as shown in panel I. Error bars denote standard deviations of at least three biological replicates (n$\ge 3$). Significance analysis was based on Dunnett One-Way ANOVA using GraphPad Prism 10. *p < 0.05; **p < 0.01; ***p < 0.001. For panels C and H, Significance analysis was based on an unpaired t-test using GraphPad Prism 10. *p < 0.05; **p < 0.01; ***p < 0.001.

To gain insight into the effects of this transporter on growth acceleration (Fig. 3a), we employed 4D label-free mass spectrometry to quantitatively analyze the proteomes of E. coli and B. subtilis expressing GltS_Vnat, given the pivotal role of proteome allocation in governing bacterial growth control [10,11,23,24]. Heatmap analysis revealed that GltS_Vnat expression substantially reshaped the global gene expression profiles of E. coli and B. subtilis during growth on glutamate (Fig. 3b). Proteome allocation analysis using the Proteomaps website [25] further showed that GltS_Vnat expression triggered a marked upregulation of protein translation and biosynthetic pathways, along with the downregulation of carbon catabolic pathways (Fig. 3c). Specifically, the ribosome content, assessed via both the RNA/protein ratio and ribosomal protein abundance [18], increased strongly in both E. coli and B. subtilis strains expressing GltS_Vnat (Fig. 3d, e). We also observed an increased expression of other biosynthetic pathways essential for growth in minimal media, such as amino acid and nucleotide biosynthesis, in strains expressing GltS_Vnat (Fig. 3f, g ) [26]. Additionally, the GS-GOGAT pathway was upregulated, suggesting that nitrogen assimilation was enhanced by GltS_Vnat expression (Fig. 3h). Conversely, the expression of the nutrient catabolic and assimilation pathways was strongly reduced (Fig. 3i, j), indicating a lower metabolic cost for carbon assimilation. In addition, the abundance of various stress-responsive pathways decreased (Fig. 3k and l). Since carbon limitation typically upregulates catabolism and stress responses [27], these results suggest that GltS_Vnat expression alleviates the severe carbon limitation experienced by wild-type cells growing on glutamate-containing media. The shift in the bacterial proteome upon GltS_Vnat expression could be related to the regulatory effect mediated by CRP-cAMP, as increased carbon uptake might inhibit the CRP-cAMP signaling pathway by reducing cAMP levels so that cells could shift cellular resources from catabolic proteins to ribosomes and anabolic proteins to achieve faster growth [10,21]. In B. subtilis, a similar resource allocation strategy might be achieved with the implementation of the HPr-CcpA signaling pathway, which could directly repress the expression of catabolic proteins in response to more favorable carbon sources [28].

Fig. 3.

The effect of GltS_Vnat expression on proteome allocation in Escherichia coli and Bacillus subtilis during growth on glutamate as the sole carbon source. Proteomic data includes two biological replicates for each condition. (a) The four strains analyzed via proteomics, including wild-type E. coli NCM3722, wild-type B. subtilis 168, and their corresponding GltS_Vnat-expressing counterparts, growing on a medium containing glutamate as the sole carbon source. (b) Heatmap analysis of the proteomes of wild-type and GltS_Vnat-expressing strains of E. coli and B. subtilis grown on glutamate-containing medium.(c) Proteome allocation of the wild-type and GltS_Vnat-expressing strains of E. coli and B. subtilis grown on glutamate-containing medium as visualized using the Proteomaps website. (d) The RNA/protein ratio and ribosomal protein (r-protein) abundance of wild-type E. coli NCM3722 and its GltS_Vnat-expressing counterpart. The error bar denotes the standard deviations of three biological replicates (n = 3). (e) The RNA/protein ratio and r-protein abundance of wild-type B. subtilis 168 and its GltS_Vnat-expressing counterpart. The error bar denotes the standard deviations of three biological replicates (n = 3). (f) The proteome fraction of various biosynthetic pathways in wild-type E. coli NCM3722 and its GltS_Vnat-expressing counterpart. (g) The proteome fraction of various biosynthetic pathways in wild-type B. subtilis 168 and its GltS_Vnat-expressing counterpart. (h) The relative abundances of GS-GOGAT pathways in wild-type GltS_Vnat-expressing strains. Data of one biological replicate of the wild-type strain was set as “1.” (i) The proteome fraction of proteins involved in carbon assimilation and amino acid and peptide uptake in wild-type E. coli NCM3722 and its GltS_Vnat-expressing counterpart. (j) The proteome fraction of various nutrient catabolic and uptake proteins in wild-type B. subtilis 168 and its GltS_Vnat-expressing counterpart. (k) The proteome fraction of various stress-responsive pathways in wild-type E. coli NCM3722 and its GltS_Vnat-expressing counterpart. (l) The proteome fraction of various stress-responsive pathways in wild-type B. subtilis 168 and its GltS_Vnat-expressing counterpart. Abbreviations: AA, amino acid; NT, nucleotide.

Compared with poor carbon sources, a preferred carbon source (e.g., glucose) enables faster growth by increasing cellular resource allocation to biosynthetic pathways, especially for ribosome and amino acid biosynthesis, following the linear bacterial growth law [10,21,26]. Our results indicated that GltS_Vnat expression converts glutamate from a poor carbon source into a relatively favorable carbon source. To further validate that the shift in proteome allocation upon GltS_Vnat expression follows the linear bacterial growth law, we quantitatively compared the proteomes of cells grown on different carbon sources. We found that the abundance of ribosomal/translation machinery and amino acid biosynthetic pathways increased linearly with increasing growth rate (R-line and A-line; Fig. 4a–d), which is consistent with the known linear bacterial growth law established in E. coli [10,21,26]. The data points corresponding to the wild-type and GltS_Vnat-expressing strains for glutamate (colored symbols) coincided with the linear R-line and A-line for both species (Fig. 4a–d), confirming that GltS_Vnat expression increases glutamate utilization efficiency and biosynthetic flux (Fig. 4e), which is consistent with the effect of a better carbon source.

Fig. 4.

Verification of the proteome allocation shifts upon GltS_Vnat expression according to the linear growth laws of bacteria. (a) The growth dependence of ribosome and translation-affiliated proteins for Escherichia coli NCM3722 cells growing on different carbon sources. In addition to the two conditions in Fig. 3 (red and blue symbols), data of the wild-type E. coli NCM3722 strain growing on mannose (man), glycerol (gly) and glucose (glu) were also included. The R-line denotes the best-fit of simple linear regression (r² = 0.99). The slope is significantly non-zero (p< 0.001).(b) The growth dependence of amino acid (AA) biosynthetic pathways for E. coli NCM3722 cells growing on different carbon sources. Growth conditions are the same as panel A. The A-line denotes the best-fit of simple linear regression (r² = 0.96). The slope is significantly non-zero (p< 0.001).(c) The growth dependence of ribosome and translation-affiliated proteins for Bacillus subtilis 168 cells growing on different carbon sources. In addition to the two conditions in Fig. 3 (orange and purple symbols), data of the wild-type B. subtilis 168 strain growing on ribose (rib), mannose (man) as well as the spo0A-null strain growing in mannose medium were also included (see detailed information in Ref. [24]). Note that the spo0A-null strain grows faster than the wild-type strain due to increased cellular resources allocated to biosynthetic pathways, which increases the nutrient flux of B. subtilis growing in mannose medium (see detailed information in Ref. [24]). The R-line denotes the best-fit of simple linear regression (r² = 0.96). The slope is significantly non-zero (p< 0.001).(d) The growth dependence of amino acid (AA) biosynthetic pathways in B. subtilis 168 cells growing on different carbon sources. The growth conditions are the same as in panel C. The A-line denotes the best-fit of simple linear regression (r² = 0.96). The slope is significantly non-zero (p< 0.001).(e) Schematics showing that the heterologous expressions of Vibrio natriegens glutamate transporters increase the glutamate utilization efficiency and strongly enhance the biosynthesis flux during growth on glutamate-containing medium for the two model bacterial species. Dataset description: the data of E. coli on glucose, glycerol and mannose in panels A and B were re-analyzed from the dataset in Ref. [18]. The data of B. subtilis 168 (both wild-type and spo0A-null strains) on mannose originated from the dataset in Ref. [24]. The data on the B. subtilis 168 strain on ribose originated from the dataset in Ref. [35].

The external nutrient composition is a primary factor that modulates bacterial growth, imposing fundamental physiological constraints. Our study provides a simple strategy for converting glutamate from an extremely poor to a relatively favorable carbon source for E. coli and B. subtilis. The strong growth enhancement achieved through the heterologous expression of ultrafast-growing V. natriegens glutamate transporters indicates that the major bottleneck for growth on glutamate is the limitation of substrate uptake. Thus, the use of cross-species nutrient transporters may represent an effective approach to shift bacterial cellular resource allocation and enhance the growth capacity of bacteria on specific substrates, especially uptake-limited substrates. In addition, it should be noted here that although overcoming the uptake limitation of glutamate is likely to be the major effect of GltS_Vnat expression, other effects such as changes in the Na⁺ gradient, membrane potential, or secondary energy metabolism might also contribute to the observed growth improvement effects since GltS is a Na⁺-coupled glutamate symporter.

While our primary focus was overcoming the fundamental growth limitation of bacteria on glutamate, we acknowledge that further optimization of downstream metabolic flux and redox balancing is required to increase growth and substrate utilization efficiency for improved bioproduction. For example, achieving a high flux towards specific glutamate-related chemicals (e.g., GABA) will likely require further metabolic engineering, such as the downregulation of competing pathways and enhancement of target pathways [29,30]. In addition, redox cofactor balancing, such as optimization of the NAD(P)H/NAD(P)⁺ supply [31], may also be required to further improve bioproduction. Nevertheless, proteome shifts from stress/catabolism towards biosynthesis pathways during the overexpression of V. natriegens glutamate transporters (Fig. 3) established a more favorable baseline metabolic state for further engineering.

Mechanistically, the poor glutamate utilization efficiency of E. coli and B. subtilis can be attributed to multiple factors. First, it is possible that the expression levels of native glutamate transport systems in these species were not sufficient to support high growth rates on media containing glutamate as the sole carbon source. Alternatively, their glutamate transporters may have lower catalytic efficiency (lower V_max) and/or lower substrate affinity (higher K_m) than the glutamate transporters in V. natriegens, which could be verified by further biochemical and structural characterization. Another potential issue is whether the heterologous expression of V. natriegens glutamate transporter imposes significant physiological costs. For example, the heterologous expression of an Na⁺-coupled glutamate symporter (GltS_Vnat) could lead to the cellular accumulation of Na⁺, inducing osmotic stress [32], while the heterologous expression of an H⁺-coupled glutamate symporter (GltP) might cause the direct consumption of the proton motive force and lead to additional cellular ATP cost [33,34]. Nevertheless, our proteomic data (Fig. 3k, l) showed that various cellular stress-responsive pathways, including the osmotic stress response pathway, were downregulated during the heterologous expression of V. natriegens glutamate transporters, suggesting that the potential physiological cost is much lower than the benefit of overcoming the severe carbon limitation of glutamate.

Data Availability Statement

All data generated or analysed during this study are included in this published article and its supplementary information files or are available upon request. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the iProX partner repository with the dataset identifier: PXD069464.

CRediT authorship contribution statement

Fei Han: Methodology, Investigation, Formal analysis, Data curation. Qian Wang: Methodology, Investigation, Formal analysis. Yiheng Wang: Investigation. Xingyu Liu: Investigation. Yifan Zhang: Investigation. Kunlong Liang: Investigation. Xiongfeng Dai: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition, Conceptualization. Manlu Zhu: Writing – review & editing, Writing – original draft, Supervision, Funding acquisition, Conceptualization.

Declaration of Competing Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Xiongfeng Dai reports financial support was provided by National Natural Science Foundation of China. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix. Supplementary materials

Fig. S1. Representative data of the net consumption of external glutamate in the medium of both wild-type and GltS_Vnat-expressing E. coli strains. The line denotes the best-fit of a simple linear regression. The slope of the linear plot of external glutamate concentrations versus the OD₆₀₀ was further multiplied with the exponential growth rate of each strain to obtain the glutamate uptake rate.