Integrating metabolic and evolutionary engineering for enhanced propionic acid production under cross stress: a multi-omics perspective

Background: Propionic acid (PA), a potential building block for C3-based bulk chemicals, is used as a food preservative and antifungal agent because of the antimicrobial properties of its calcium-, potassium-, and sodium salts, as well as in the manufacture of pharmaceuticals, perfumes, pesticides and fungicides. However, industrial development of PA is seriously inhibited by oxygen stress, acid stress and glucose-induced osmotic stress concentration on account of the characteristic of Propionibacterium acidipropionici . To alleviate inhibition and increase PA production, enhancement P. acidipropionici tolerance to cross stress may be an effective strategy. Results: In this study, we first performed a combination of metabolic engineering (deletion of ldh and poxB and overexpression of mmc ) with evolutionary engineering (selection under oxygen stress, acid stress and osmotic stress) in P. acidipropionici . The results indicated the mutants received superior physiological activity, especially the mutant III exhibited steady 1.5-3.5 folds higher growth property and further 37.1% PA titer and 37.8% PA productivity increase than the wildtype. Moreover, omics analysis revealed the determinants such as Dps , GroES , dnaK , ADI and GAD referred to the acid adaptation of microbes were positively mobilized. ABC-type glycine betaine referred to the adaptability to osmotic stress was detected to be 2-4 folds up-regulated. More than 2-fold down-regulation of NADH oxidase and almost 3-fold up-regulation of SOD and POD were observed in three mutants. Moreover, an approximately 2.5-fold upregulation of mmc was also found. Conclusion: The multi-omics analysis revealed the multidirectional variation tendency of P. acidipropionici under cross stress and provided in-depth insights into

the mechanism of tolerance and high production of PA, which layed the foundation for construction of microbial cell factories.

Background
Currently, the traditional fossil-fuel-based PA synthesis is becoming increasingly less desirable due to energy shortages, environmental pollution and the desire for sustainable development, while the fermentation of PA from renewable resources using Propionibacterium has unique advantages [1]. As the major production strains, the members of the genus Propionibacterium are generally recognized as safe [1]. P. acidipropionici, P. shermanii and P. freudenreichii are capable of utilizing a wide range of carbon sources to produce PA, including soy molasses [2], corn stover hydrolysate [3], sorbitol [4], glycerol and glucose [5] and whey lactose [6]. To date, many strategies have been developed to improve the yield of PA, including high density fermentation [7], immobilization of Propionibacterium cells [8], iNTRODUCTION OF BIOCOMPATIBLE SMALL MOLECULE [6], reduction of by-product [9], controlled pH-shift in fed-batch culture [10] and engineering of metabolic pathways [11]. However, in addition to the common problems of industrial fermentation such as end-product inhibition and by-product accumulation, the anaerobic growth of Propionibacterium spp. means that the culture conditions for PA production must be strictly controlled [12]. Although the creation of an anaerobic environment is no longer as challenging due to the development of production technology, the cost of maintaining anaerobic conditions is high and incompatible with the concept of developing a green economy. PA is synthesized by Propionibacterium spp. via the dicarboxylic acid pathway, which also generates byproducts such as lactic acid (LA) and acetic acid (AA) [9]. Thus, the accumulation of acid products during fermentation easily results in an acidic environment, affecting cell activity and inhibiting metabolism, which precludes industrial-scale PA production. Moreover, hypertonic solutions casused by HIGHER SUBSTRATE CONCENTRATION tend to dehydrate cells and inhibit their growth and reproduction [13,14]. Consequently, enhancing the tolerance of Propionibacterium spp. to oxygen, acid and hypertonic solutions is considered as an effective strategy to alleviate the inhibition and has triggered a research hotspot.
The adaptation of bacteria to a stressful environment can lead to the accumulation of mutations that confer a fitness advantage when exposed to a different stressor, which was named 'cross-stress' protection [15,16]. Therefore, the application of cross stress in adaptive evolution, especially in combination with metabolic engineering, has certain advantages for industrial strain development. Metabolic engineering has been used as an effective strategy for increasing PA production. Liu et al. improved the PA titer of P. jensenii by 34.7% by overexpressing phosphoenolpyruvate carboxylase (ppc) and deleting lactate dehydrogenase (ldh) [11]. Guan et al. obtained a 12.2% increase of PA by constructing a P.
acidipropionici strain with simultaneous deletions of ldh1, ldh2 and pyruvate oxidase (poxB) [17]. However, combining metabolic engineering with laboratory evolution to improve the synthesis of PA has not been published, and may be a successful strategy. Furthermore, omics technologies are able to identify crucial genes and metabolites, providing valuable information that reflects stress-induced changes and the intricate interplay between organisms and the environment [18,19]. For example, comparative genomic and transcriptomic analyses of P.
acidipropionici revealed the molecular mechanisms of acid tolerance, while proteomics analyses revealed how P. acidipropionici respond to propionic acid stress 6 [20]. Therefore, in this study, P. acidipropionici CGMCC 1.2232 was engineered by combining metabolic engineering with evolutionary engineering to adjust the direction of metabolic flow to maximize the production of PA and improve the adaptability to environmental stresses. To probe the variation footprints of determinants involved in the tolerance mechanisms under cross stress and mechanism of high PA production, multi-omics analysis of wild-type P.
acidipropionici and the improved mutants was performed and then integrated, providing valuable guidance for the industrial production of PA.

Results and discussion
Metabolic and evolutionary engineering, and the growth rates of mutant strains The ldh and poxB genes, which are main contributors of the by-products acetate and lactate [21], were deleted. Additionally, the methylmalonyl-CoA carboxyltransferase (mmc) gene from P. acidipropionici¸ which controls the carbon flow into the propionate-producing Wood-Werkman cycle [22] was overexpressed. These manipulations yielded the strains P. acidipropionici-Δldh, P. acidipropionici-Δldh-ΔpoxB, and P. acidipropionici-Δldh-ΔpoxB + mmc, respectively. In evolutionary engineering, specific stress is applied gradually on the basis of the growth rate of the adapted strain, which gives a buffer period so that the strain can adequately mobilize the in vivo response mechanisms to resist stress (Fig. 1). During the evolution of P. acidipropionici-Δldh into the mutant I, the growth of the experimental strain displayed a peak at OD 600 17.1 on the fourth day of culture, followed by a 7 downward trend because of the decline of pH and the increase of glucose concentration. Subsequently, the growth rate of P. acidipropionici-Δldh tended to be stable at OD 600 13.9 and began to descend when the pH was decreasing again (6 to 5), and was ultimately stable at OD 600 15.9. From P. acidipropionici-Δldh-ΔpoxB to the mutant II (deletion of poxB followed by evolutionary engineering), the formation of the subsequent two peaks at OD 600 16.5 and 15.4 on the 6th and 15th day corresponded to the progressive deepening of the levels of single stress factors.
From P. acidipropionici-Δldh-ΔpoxB + mmc to the mutant III (overexpression of mmc followed by evolutionary engineering), the growth rate of the strains started to pick up a steady upward trend, and the strain showed a normal growth curve after supplementation with fresh medium. The highest OD 600 was 14.6, while the lowest OD 600 of 9.8 was observed with pH 4, 120 g·L − 1 glucose and 21% oxygen sparging.
These results indicate that the experimental strain could still maintain a considerable growth rate under the superimposed cross stress, even when exposed to simultaneously increasing stress levels. This indicated that the strain had developed a certain level of 'cross-stress' protection and acquired a fitness advantage based on the gradual application of stress by artificial control so that it could deal with cross stress.
To further investigate the growth rate of the strain obtained by metabolic and evolutionary engineering, the wild type and mutant III were compared. Compared with the 0.09 h − 1 specific growth rate of the wild type under normal fermentation conditions (pH 7), mutant III had a significant improvement at 0.14 h − 1 , which was apparently higher than the 0.131 ± 0.007 h − 1 previously reported for growth on soy molasses and soy molasses hydrolysate [2]. However, when the strains were 8 exposed to low pH, the growth of both the wild type and the mutant III was suppressed accordingly. Nevertheless, the growth of the mutant III was significantly superior to that of the wild type. With the decrease of pH, the growth rate difference between the wild type and mutant III gradually increased, and mutant III could maintain normal growth, while the growth of the wild type was clearly suppressed (Fig. S1A). As the concentration of glucose in the medium increased, the difference of the growth rate between the wild type and mutant III gradually increased. At normal glucose concentrations (30 g·L − 1 glucose), the specific growth rate of the wild type was about 0.09 h − 1 , while that of mutant III reached 0.2 h − 1 , which was 1.4-fold higher than the previously reported value for the same glucose concentrations [23]. While the wild type almost stopped growing at a high glucose concentration of 120 g·L − 1 , mutant III could maintain growth at OD 600 0.4 (Fig.   S1B). Futhermore, the strain exhibited similar growth characteristics when exposed to oxygen stress compared with acid stress and osmotic stress. A 2-fold growth rate difference has observed under anaerobic fermentation conditions (oxygen content 0.001%). Under these conditions, the specific growth rate of mutant III reached 0.18 h − 1 , while that of the wild type was 0.09 h − 1 . By contrast, with 21% oxygen content in the sparged gas mixture, the specific growth rate of the wild type was only 0.03 h − 1 , while that of mutant III was 0.05 h − 1 (Fig. S1C). As a result, the growth performance of mutant III under cross stress was clearly superior to that of the wild type. Importantly, the generation of the mutant III has proved that evolutionary engineering is an effective method for improving the resistance of P.
acidipropionici to cross stress. PA fermentation using the engineered P . a c i d i p r o p i o n i c i 9 Batch fermentations were carried out using wild-type P. acidipropionici as well as the mutants I, II and III to investigate their PA production capacity. As shown in Table 1, under the optimal culture conditions (pH 7, 99% N 2 , 5 g·L − 1 glucose), the PA titer and productivity of mutant I were increased by 12.4 and 13.5% compared to the wild type. Its performance was therefore also significantly better than that of P.
However, there was a 3% increase of AA titer (5.24 ± 0.83 vs. 5.63 ± 0.62 g·L − 1 ), which may be related to the deletion of ldh, which increased carbon flow from pyruvate into the synthesis of LA [21]. The PA production of the mutant II only increased by 3.2%, while the titers of AA and LA were both decreased by more than 70%, which was consistent with the variation tendency of the PA, LA and AA titers of P. acidipropionici-ΔpoxB-Δldh [ 17]. A further 37.1% increase of PA titer (28.1 ± 0.96 g·L − 1 vs. 38.7 ± 1.14 g·L − 1 ) and 37.8% increase of PA productivity (0.216 ± 0.006 g·L − 1 vs. 0.298 ± 0.008 g·L − 1 ·h − 1 ) was achieved in the mutant III. This increase was much more significant than the reported value achieved via the overexpression of mmc in P. freudenreichii in batch fermentation [23]. At the same time, there was a 78.6 and 87.8% decrease of the byproduct yields because of the deletion of ldh and poxB, and the overexpression of mmc drove the carbon flow into the synthesis of PA to the greatest extent. Table 1 Analysis of microbial production using engineered P. acidipropionici in batch fermentation. Comparative omics analysis of the wild type and the mutants I, II, III

Strain
The adaptation of microbes to environmental stress can produce beneficial phenotypes through random genomic mutations and subsequent positive selection [24]. In this study, the resistance mechanism of the strains to cross stress may be Combined with the sample correlation analysis (Fig. 4), the data revealed that The distribution of DEPs quantified in the GO secondary annotation was statistically analyzed to gain insights into the potential functional implications for stress tolerance (Fig. 7).  (Fig. 8A), we observed that upregulated DEPs were involved in glutathione transferase activity and FAD binding in the mutant I, glycerol metabolic process and alditol metabolic process in the mutant II, while they were involved in cysteine biosynthetic process and sulfate reduction in the mutant III. In the KEGG enrichment (Fig. 8B), the microbial metabolism in diverse environment-related pathways were enriched (1.64-fold, 1.79-fold, 1.66-fold) in the mutants I, II and III, respectively. Degradation of aromatic compounds and bacterial chemotaxis were detected to be significantly enriched in the mutants I and II, while sulfur metabolism was found in the mutant III, demonstrating that this mutants developed positive responses to the adverse environment. As can be seen in Fig. 9, at the metabolic level, carboxylic acids metabolism such as butanoate and pyruvate, methane metabolism, amino acid metabolism such as glycine, serine and threonine, as well as cysteine and methionine in the three mutants were all significantly affected when subjected to cross stress. The pathway analysis results implied the mutants might modulate the stress response mechanisms against to abiotic stress.
Differentiation analysis of the determinants related to cross stress of the wild type and the mutants I, II and III In order to understand the detailed resistance mechanism of P. acidipropionici subjected to cross stress, we investigated the variation footprints of the changes from the wild type to mutant III, from the DNA to metabolite levels. Maintaining pH homeostasis is the main challenge under acid stress, which may be supported by 14 ATP metabolism and active transport of protons. The ΔpH in the cell can drive the generation of proton motive force (PMF) to enhance cellular protonation [27].
Previous studies have shown that specific genes are commonly expressed by Grampositive neutrophilic bacteria during acid stress [28]. In this study, genes related to the ATPase-coupled transmembrane transporter activity were confirmed to be downregulated to retard protonation based on the transcriptome analysis in the three mutants. The interconversion between NADH and NAD is associated with ATP metabolism in the coupling of proton levels. The NADH levels were reduced 2.97-, 2.80-, and 3.54-fold, while those of NAD were reduced 8.11-, 6.50-, and 5.81-fold in three mutants, respectively. Moreover, the determinants of the regulation of pH homeostasis, such as cations and inorganic ions, were also discovered in the three mutants. ABC transporters were transcriptionally upregulated in the three mutants, which may mediate the translocation of large amounts of substrate across the cell membranes [29]. The Dps protein, which protects cells against acid damage [30], was expressed 3.47-fold more at the transcriptional level in the mutant I, as well as 2.15-, 1.89-, and 3.49-fold more at the proteome level in three mutants compared to the wild type, respectively. The genes GroES and dnaK related to the acid adaptation of bacteria were also differentially expressed at the protein level [31].
GroES were upregulated in all three mutants. DnaK was downregulated in mutant I, but upregulated in the mutants II and III. Furthermore, GroES was found downregulated in the mutants I, II according to transcriptional analysis. Amino acid metabolism plays an important role in the acid tolerance of bacteria [32]. The arginine deiminase (ADI) system, which consists of arginine deiminase, carbamate kinase (CK) and ornithine transcarbamoylase, has been proved to exist in the genus Propionibacterium [ 33,34]. This system is activated under acidic conditions because of its contribution to pH homoeostasis [28]. In this study, the expression of ADI and CK were not detected in RNA-seq and metabolome analysis, but the proteins were detected in the proteome. Moreover, the glutamate decarboxylase (GAD) pathway, which can consume H + [ 35], did not show significant differential expression in the mutant I, II, and III, respectively. Therefore, the trajectory of change of ADI and GAD remained consistent, whereby mutant II showed more positive regulation of ADI and GAD, while mutant I was more positive in the regulation of CK.
We observed that the ABC-type glycine betaine transport system, which is related to the adaptability to osmotic stress [36], was not observed transcriptionally regulated in the three mutant, but upregulated 2.81-, 2.06-, 4.51-fold on the proteomic level in the mutant I, II, and III. Additionally, it was reported that high osmolarity can induce the oxidative stress regulons [37,38]. It has been reported that H 2 O 2 can be produced by NADH oxidase in the cell membrane when anaerobes are exposed to oxygen, posing a threat to cell survival [39]. Interestingly, we only observed a significant upregulation of NADH oxidase by proteome analysis with 2.00-, 2.05-, 2.05-fold in three mutants. In addition, the expression of superoxide dismutase (SOD) and peroxidase (POD) was upregulated in the protein level. Notably, there was a 2.99-fold upregulation of SOD expression in the mutant III and 3.18-fold upregulation of POD expression in the mutant II. These results indicate that the mutants also adapted to oxidative stress. In addition, as a bioprotective agent, trehalose exhibits prominent biological functions against adverse abiotic stress, such as oxygen radicals and acid stress [6]. Found by metabolomics, the level of trehalose was a persistent upward trend from the wild type to the mutant III, especially the mutant III with the most significant level in the content compared to the wild type. It verified that the mutant III was subjected to the relatively large degree of cross stress. Some metabolites such as serine, aspartic acid, glycine and trimethylol propane were only observed in the mutant II and III, while methanediimine, glycol acid and ethylaminewhich were only observed in the wild type.
Pyruvate from glucose or glycerol is a vital precursor in the metabolic pathway of propionic acid synthesis in Propionibacterium spp. However, the pyruvate is also converted into byproducts, most importantly into acetate by poxB and lactate by ldh [18]. Therefore, poxB and ldh were selected to be deleted to reduce the metabolic flux into the formation of acetate and lactate. However, another source of acetate synthesis is the conversion of pyruvate to acetyl-CoA. This indicated that the deletion of poxB did not completely block the carbon flux into acetate synthesis, which was also confirmed by the fermentation data. Unsurprisingly, no differences of ldh and poxB were found in the omics results. The lactate permease involved in lactate transport driven by a proton motive force was respectively upregulated 3.6-, 2.6-and 1.88-fold in the mutants according to the proteomic analysis. Therefore, decreased lactate synthesis promoted lactate permease activity to maintain the intracellular metabolic homeostasis [40]. For PA synthesis, pyruvate is converted to oxaloacetate, which finally flows to PA via mmc through the transfer of one carbon from methylmalonyl-CoA to pyruvate [41]. The overexpression of mmc led to 1.

Conclusions
Rational P. acidipropionici mutants were constructed by metabolic engineering to promote carbon flow to the PA production. Subsequently, adaptive laboratory evolution of the mutants was conducted by the application of cross stress. The mutant III had a 1.5 to 3.5-fold higher growth rate, as well as a 37.1% increase of PA titer and increase of 37.8% PA productivity compared to the wild type. Omics analysis revealed that the determinants related to the tolerance mechanism and PA production were actively mobilized through increased expression. These results offer a better understanding of the mechanism of increased PA production, which lays a foundation for the construction of advanced microbial cell factories for the industrial fermentation of PA.

Materials and methods
Microorganisms and culture media 18 All strains, plasmids and primers used in this study are listed in

Availability of data and materials
All data generated or analyzed in this study are available in this article and its additional file. 20

Ethics approval and consent to participate
Not applicable.

Consent for publication
All authors provide consent for publication of the manuscript in Biotechnology for Biofuels.

Figures
26 Figure 1 Metabolic and evolutionary engineering of Propionibacterium acidipropionici. The sample correlation analysis in the wildtype, the mutant Ⅰ, Ⅱ and Ⅲ.
30 Figure 5 The GO enrichment analysis of the DEGs in three mutants a wildtype-vs-Mutant I; b wildtype- Figure 6 The KEGG enrichment analysis of the DEGs in three mutants a wildtype-vs-Mutant I; b wildtyp 31 Figure 7 The GO secondary annotation of the DEPs in three mutants a wildtype-vs-Mutant I; b wildtype The GO enrichment A and the KEGG enrichment B of the DEPs in three mutants a and d wildt 32 Figure 9 Pathway analysis of the metabolites of the wild type and the mutants I, II and I Additional file 1: Table S1. Descriptions and Sources of strains, plasmids and primers  The sample correlation analysis in the wildtype, the mutant Ⅰ, Ⅱ and Ⅲ .
44 Figure 5 The GO enrichment analysis of the DEGs in three mutants a wildtype-vs-Mutant I; b wildtype-vs-Mu Figure 6 The KEGG enrichment analysis of the DEGs in three mutants a wildtype-vs-Mutant I; b wildtype-vs 45 Figure 7 The GO secondary annotation of the DEPs in three mutants a wildtype-vs-Mutant I; b wildtype-vs-M Figure 8 The GO enrichment A and the KEGG enrichment B of the DEPs in three mutants a and d wildtype-v 46 Figure 9 Pathway analysis of the metabolites of the wild type and the mutants I, II and I

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.
Supplementary Material.doc