Co-Consumption of Methanol and Succinate by Methylobacterium extorquens AM1

Methylobacterium extorquens AM1 is a facultative methylotrophic Alphaproteobacterium and has been subject to intense study under pure methylotrophic as well as pure heterotrophic growth conditions in the past. Here, we investigated the metabolism of M. extorquens AM1 under mixed substrate conditions, i.e., in the presence of methanol plus succinate. We found that both substrates were co-consumed, and the carbon conversion was two-thirds from succinate and one-third from methanol relative to mol carbon. 13C-methanol labeling and liquid chromatography mass spectrometry analyses revealed the different fates of the carbon from the two substrates. Methanol was primarily oxidized to CO2 for energy generation. However, a portion of the methanol entered biosynthetic reactions via reactions specific to the one-carbon carrier tetrahydrofolate. In contrast, succinate was primarily used to provide precursor metabolites for bulk biomass production. This work opens new perspectives on the role of methylotrophy when substrates are simultaneously available, a situation prevailing under environmental conditions.


Introduction
Bacteria often live in environments containing diverse substrates [1]. One of such habitat is the phyllosphere, where facultative methylotrophic bacteria are found to be ubiquitous and abundant [2,3,4]. These methylotrophic bacteria belong to the genus Methylobacterium and are known to metabolize methanol but also a limited number of alternative carbon substrates, such as organic acids and alcohols. Plant leaf surfaces release diverse carbon sources, mainly sugars and organic acids, at low amounts (mM range) [5,6,7], and these sources are heterogeneously located and result of leaching through the cuticle [8]. In addition to these substrates, volatile carbon substrates, particularly the plant cell wall metabolism byproduct methanol, are released via the stomata. Methanol emission peaks in the morning, when the stomata first open [9]. There is evidence that methanol is consumed by Methylobacterium and contributes to the epiphytic fitness of the organism [6,10]. However, in addition to the peak of methanol emission in the morning, Methylobacterium should adapt its metabolism to use additional carbon sources during the rest of the day when methanol emission is low or during the night when stomata are closed and methanol is consequently no longer available [9]. Accordingly, additional carbon sources were suggested to be relevant for the efficient colonization of plant surfaces in situ [10].
M. extorquens AM1 is a model methylotrophic organism, and a number of novel enzymes and pathways involved in methanol dissimilation and assimilation were shown to operate in this organism [11,12,13,14]. In the past, a number of studies reported metabolic differences between methylotrophic growth conditions (i.e., methanol as sole source of carbon and energy) and multicarbon growth conditions (i.e., succinate as sole source of carbon and energy). These investigations include transcriptomic [15], proteomic [16,17], and metabolomic studies [18,19]. The core of the central metabolism of M. extorquens AM1 was described to encompass 85 biochemical reactions that are strongly reprogrammed upon adaptation to nutrient changes [20]. Indeed, metabolic pathways such as the tetrahydromethanopterin-dependent oxidation pathway, the serine cycle, and the ethylmalonyl-CoA pathway are essential during growth on methanol but dispensable for growth on organic acids (although some individual enzymes may still be required). In contrast, a complete TCA cycle and pyruvate dehydrogenase, which provide energy during growth on organic acid, are not required during C1 growth [20,21,22].
All of these metabolic pathways are strongly connected via a dense network of reactions that interconvert key C2-C3-C4 metabolites (i.e., acetyl-CoA, phosphoenolpyruvate, pyruvate, oxaloacetate, and malate). Due to the occurrence of sequences of reactions that generate substrate cycles, this network of reactions is highly flexible and allows the efficient switching of the central metabolism towards the utilization of alternative substrates [20,21,22]. Recently, the adaptation involved in the transition from succinate to methanol utilization via a systemslevel approach was investigated [23]. The study revealed that a significant amount of methanol is quickly oxidized to formate in the early stage of the transition but that the first steps of the assimilation processes are repressed. Thereafter, assimilation starts only when the entire set of required enzymes is expressed [23]. Such a transition between succinate and methanol [23] might mimic a diauxic shift, which is a well-described mechanism of catabolic repression during mixed substrate conditions. Indeed, diauxic growth is based on an important genetic regulation phenomenon that was uncovered decades ago as a strategy for bacteria to address the availability of two substrates [24,25,26]. In essence, one substrate is utilized exclusively, and it is only upon exhaustion of the ''preferred'' substrate, often the substrate supporting the higher growth rate, that the genes for enzymes required for the second substrate are induced. Several regulation mechanisms were described to trigger catabolic repression, like the phosphotransferase system (PTS), riboswitches, or regulators such as Crp/CyaA [26].
The metabolic profile of M. extorquens AM1 in the presence of a multicarbon compound(s) in addition to methanol has not been thoroughly investigated. Preliminary enzyme activity data in cell extracts indicated that methanol dehydrogenase and enzymes of the assimilatory serine cycle were detectable in M. extorquens AM1 cell extracts when cells were incubated overnight with methanol plus succinate [27]. An intermediate level of some of the enzymes activities compared to their levels under pure methanol or succinate conditions suggests the presence of a dedicated metabolism adapted to mixotrophic conditions. However, no information on cell growth or substrate utilization during the incubation with both substrates was included in this earlier study. From another perspective, the co-consumption of methanol and thiosulfate by several Methylobacterium species, including M. extorquens AM1, has been reported [28]. Thiosulfate was effectively used as an additional energy source and appeared to enhance growth capacity. These results indicate that M. extorquens AM1 is able to utilize at least two very different energy sources of inorganic and organic nature.
In this study, we investigated the metabolism of M. extorquens AM1 under a mixed carbon substrate condition, i.e., methanol plus succinate, to address whether diauxic growth or coconsumption occurs.

Characterization of M. extorquens AM1 Growth on Methanol Plus Succinate
To infer the adaptation of M. extorquens AM1 to conditions under which both succinate and methanol are available, growth experiments were performed on minimal medium with equivalent C-mol of methanol and succinate, i.e., 60 mM and 15 mM, respectively. A growth rate of 0.1860.01 h 21 was observed under this condition, which is roughly similar to that under either pure succinate growth (0.2060.01 h 21 ) or pure methanol growth conditions (0.1760.01 h 21 ) (Table 1). Both methanol and succinate were consumed during the exponential growth phase ( Figure 1A) indicating the co-consumption of the two compounds. For each substrate, the consumption rate under mixed substrate conditions was lower than that observed under pure culture conditions. Succinate utilization dropped by 34%, and methanol utilization dropped by 70% (Table 1). Notably, the sum of the two consumption rates (in moles of carbon) was similar to the amount of substrate consumed under pure conditions, approximately 17 C-mmol?g 21 ?h 21 ( Table 1). The relatively higher contribution of succinate to growth under mixed substrate conditions (approximately 72% C-mol consumed) shows that succinate was the predominant substrate when both substrates were simultaneously available. The biomass yield obtained was similar, i.e., approximately 10.6 g.C-mol 21 , under the different conditions (Table 1). Growth ceased once succinate was fully consumed, and growth resumed after a transition phase of approximately 1.5 h.
To follow the metabolic fate of methanol and succinate during mixed substrate conditions, we performed an experiment with 13 C-labeled methanol (.99% 13 C) and succinate at a natural abundance of 13 C (1.1%). The determination of 13 C-CO 2 production in the exhaust gas of the bioreactor revealed that the methanol consumed was almost entirely converted to CO 2 ( Figure 1B and Table 1), indicating that the methanol was used mainly for catabolic purposes. A portion of the released CO 2 was produced from succinate; 41% of the CO 2 was 12 C labeled and represents 29% of the carbon of the consumed succinate. The proportions of both 13 C-and 12 C-CO 2 produced were stable over time until all succinate was depleted. The 12 C-CO 2 production was then abolished, but the dissimilation of methanol remained; consequently, the 13 C-CO 2 production became exclusive.
Incorporation of 13 C Methanol into Amino Acids and Selected Metabolites during Mixed Substrate Conditions Revealed by LC-MS As outlined above, the majority of methanol was found to be catabolized to CO 2 . To confirm that little methanol was assimilated into biomass and to elucidate the metabolic fate of the one-carbon substrate, we analyzed the labeling pattern of intracellular metabolites by LC-MS. Sampling for metabolites was performed at three different timepoints during growth with 13 C methanol and naturally labeled succinate. The first sample was collected in the middle of the first exponential growth phase, and the second sample was harvested just at the complete exhaustion of succinate in the medium. A third sample was harvested to monitor the extent of methanol assimilation 90 min after succinate depletion. The average 13 C labeling (AL 13C ), which corresponds to the percentage of 13 C carbon incorporated into the metabolites, was calculated for each compound (see Materials and Methods). This value reflects the mean proportion of methanol-derived carbon in the molecule: an AL 13C of 1.1% means that all carbon atoms originate from succinate and an AL 13C of 99% means that all carbon atoms originate from methanol. The AL 13C values at the first two timepoints were identical, indicating that the metabolism was stable until the depletion of succinate ( Figure 2). The AL 13C values ranged from 1% to 27% depending on the metabolite, indicating that the dominant carbon source for biosynthesis was indeed succinate, consistent with the measured consumption and production rates (Table 1). However, remarkable differences in the 13 C fraction of the metabolites were observed. Whereas the AL 13C of most amino acids was approximately 2.0%, hexose-phosphate (5%), phenylalanine (6%), tyrosine (6%), and especially methionine (19%) showed higher AL 13C values ( Figure 2). An analysis of the mass isotopomer distribution of methionine revealed that 90% of the methionine contained one labeled carbon ( Figure 3 and Figure S1). Methionine is generated from aspartate plus 5-methyltetrahydrofolate (M-THF). Because no label incorporation was observed in aspartate (1% 62), the 13 C label found in methionine indicates that up to 90% 62 of its precursor, M-THF, was produced from assimilated methanol. As mentioned above, phenylalanine, tyrosine and hexose-phosphate also showed a small but significant incorporation of the label from methanol, which resulted in an increased abundance of the mass isotopomers M1 and M2. These compounds are synthesized by gluconeogenesis. Four metabolites generated from succinate and methanol represent potential precursors for gluconeogenesis: oxaloacetate, pyruvate, glycine, and 5,10-methylenetetrahydrofolate (Me-THF). Because no significant labeling was found in alanine (a derivative of pyruvate), aspartate (a derivative of oxaloacetate) or glycine, Me-THF (which condenses with glycine to form serine) can explain the introduction of the label into the gluconeogenesis pathway. Indeed, monolabeled phosphoglycerate was generated, which can lead to the incorporation of the label into phosphoenolpyruvate, the precursor of phenylalanine and tyrosine, and/or via gluconeogenesis into hexose-phosphate. An analysis of the hexose-phosphate mass isotopomer fractions indicated that 20% of monolabeled 2phosphoglycerate is generated from the serine cycle, i.e., glycine plus Me-THF condensation ( Figure 4). This observation indicates that serine is produced at least partially (.20%) from the Monitoring of oxygen partial pressure (pO 2 ) (white circle), 13 C (black square) and 12 C (white square) CO 2 production in exhaust gas. Metabolite sampling was performed at three timepoints as indicated in A: mid-co-consumption phase (Sampling time 1, S1), end co-consumption phase Two additional metabolites involving the incorporation of C1 precursors showed elevated AL 13C values: pantothenate (1162%) and adenosine monophosphate (AMP) (2762%). Pantothenic acid is generated from one Me-THF and pyruvate, and in the case of AMP biosynthesis, two formyl-THF molecules and one CO 2 are incorporated into the purine part of the molecule. To investigate  whether the observed AL 13C can be explained by the incorporation of C1 compounds originating from methanol, AL 13C values were calculated assuming that: i. tetrahydrofolate activated C1 precursors were mainly derived from methanol (90% 13 C according to the labeling state of methionine, see above); ii. CO 2 contains 59% 13 C (Table 1); and iii. all other carbon atoms originate from naturally labeled succinate (1.1% 13 C). The calculated AL 13C values were 19% (methionine), 11% (pantothenic acid) and 25% (AMP), and in agreement with the experimentally determined values. Thus, these results indicate that the incorporation of 13 C is limited to metabolites that require C1 precursors for formation.
To validate the approach taken and to demonstrate the enhanced assimilation of methanol into intracellular metabolites after succinate depletion, additional samples were taken at 90 minutes after succinate was consumed. Indeed, 13 C incorporation increased in all amino acids ( Figures S2, S3), indicating that cells began to use methanol as a carbon sources in all biosynthesis processes after succinate depletion.

Incorporation of 13 C Methanol into CoA Esters and their Precursors during Mixed Substrate Conditions
The ethylmalonyl-CoA pathway is critical for providing anaplerotic support to the serine cycle during growth on methanol by glyoxylate regeneration [12,29]. To monitor the operation of the ethylmalonyl-CoA pathway during growth on methanol plus succinate, the CoA-esters were extracted and analyzed by LC-MS at different timepoints. Only four of the 12 CoA esters involved in the ethylmalonyl-CoA pathway, as well as free coenzyme A (CoA), could be detected. The key intermediates crotonyl-CoA and ethylmalonyl-CoA were not detectable. Free CoA and all detected CoA esters showed significant incorporation of the label (Figure 2). The highest AL 13C was found for free Coenzyme A (22%), followed by Acetyl-CoA (20%). All CoA esters of C2 and C4 organic acids had AL 13C values of approximately 19%. In addition, all CoA esters showed very similar mass isotopomer distributions ( Figure S4). The decrease of the average labeling with increasing number of carbon atoms in the organic acid moiety indicates that labeled carbon atoms are located in the CoA part rather than in the esterified acids. C4 b-hydroxybutyryl-CoA and C4 methylmalonyl-CoA showed very similar AL 13C , and no significant change in the MID of methylmalonyl-CoA due to the incorporation of 13 C-labeled CO 2 via the ethylmalonyl-CoA pathway (note that the CO 2 in the reactor is 59% labeled) was observed. Thus, it can be deduced that the ethylmalonyl-CoA pathway was not operating during the co-consumption of succinate and methanol, which is consistent with the failure to detect key intermediates of the pathway that are detectable during methylotrophic growth [12].
The incorporation of 13 C label from 13 C methanol seems to be restricted to metabolites for which biosynthesis requires C1 precursors. We used simulations to determine whether the observed 13 C labeling into the CoA moiety also originated from the C1 precursors and was thus consistent with the CoA biosynthetic pathway ( Figure S5). The carbon atoms of CoA originate from AMP, pantothenic acid, and carbon 2 and 3 of cysteine, which are derived from carbon 2 and 3 of serine. As mentioned above, two metabolic origins of the serine carbons are possible, the condensation of glycine plus a C1 precursor (in position 3 of serine) via the serine cycle or a C3 precursor from gluconeogenesis. The AL 13C values calculated based on the measured AL 13C of AMP and pantothenic acid were 18.6% if 20% of serine had incorporated 13 C carbon into position 3 and 21.9% if 90% of serine had incorporated one-13 C carbons (the labeling state of the C1 precursor). The measured AL 13C of coenzyme A was 21.6%, indicating that almost all serine molecules might have contained 13 C carbon from C1 units. These results support the theory that serine is predominantly produced from glycine and Me-THF during growth in the presence of methanol plus succinate and that one-carbon-units for biosynthetic purposes are produced from methanol.

Energetic Contributions of Methanol and Succinate into Metabolism
The above data showed that methanol and succinate were coconsumed but that their carbon atoms had distinct metabolic fates. To better understand the cell physiology and especially the energetic contributions of both substrates during co-consumption, we performed a flux variability analysis using the genome-scale network available for M. extorquens AM1 [20] to determine the flux solution space through the network during co-consumption. Details of the simulations are given in Tables S1, S2. A solution could be found under the standard deviation of the measured constraints (growth rate, substrate uptake rates and CO 2 production rate), indicating that the carbon balance is closed. The calculated methanol and succinate contributions to ATP, NADH, and NADPH flux production for a feasible solution are displayed in Figure 5, and detailed values can be found in Table S3. Note that no single solution exists; rather, multiple satisfactory flux distributions can be predicted. Consequently, a flux variability analysis was performed to analyze the flux solution space under the measured constraints (see Table S4 for detailed results). The calculated flux distribution predicted that methanol oxidation provided 58-74% of the ATP, showing that methanol is the main energy source. ATP is produced exclusively by oxidative phosphorylation, which is fueled by NADH for 73% of reactions, methanol dehydrogenase (cytochrome-dependent) for 17% of reactions, and succinate dehydrogenase (ubiquinone-dependent) for 10% of reactions (see Table S5 for detailed calculations). Methanol dissimilation contributes to 66% of NADH generation (min 0.46 and max 0.76), whereas succinate oxidation, i.e., oxidative operation of the TCA cycle plus pyruvate dehydrogenase, and the pentose phosphate pathway supply only 9% of the NADH needs. Nonetheless, an additional 25% of NADH molecules are produced from succinate during biomass precursor formation, which is higher than the 7% required for their production and thus results in net energy production. Interestingly, some of the NADH can be converted to NADPH via transhy-drogenase (0-3.28 mmol?g 21 ?h 21 ), indicating that an excess of NADH may result from co-consumption.

Discussion
The data reported here show that M. extorquens AM1 coconsumes succinate and methanol under mixed substrate conditions. This result is something of a surprise in light of prior knowledge about the adaptation capacity of this model methylotroph. Indeed, the two pure culture conditions that were studied intensively in the past, methanol versus succinate, rely on the use of specific metabolic pathways on one hand and on driving metabolic fluxes in opposite direction through common enzyme steps, like the serine cycle, on the other hand. Consequently, rerouting the metabolic fluxes in a backward direction through some reactions involves a strong reprogramming of the central metabolism to achieve the metabolic switch between succinate and methanol [23]. In the past, this fundamental difference in central metabolic pathway usage between succinate growth and methanol growth conditions provided the basis of the successful identification of the genes involved under both conditions [30], and these became the two standard conditions under which to study the metabolism under methylotrophic and non-methylotrophic conditions in multiple biochemical and omics experiments [15,17,18,19,31]. Here, we showed that M. extorquens AM1 coconsumes methanol in addition to succinate, and a specific metabolic network takes place that relies on the previously demonstrated modularity [14] and flexibility [20,21,22] of the central metabolism of the model methylotroph. 13 C-labeling revealed the partitioning of methanol and succinate to specific metabolic pathways and functions upon co-consumption. Methanol was mainly used to fulfill energy requirements; however, a portion of the consumed methanol entered biosynthetic pathways. The most promiscuous pathways were the biosynthetic pathways linked to one-carbon metabolism, such as purine biosynthesis from tetrahydrofolate derivatives; in addition, few carbons entered gluconeogenesis via C1 precursors and the first step of the serine cycle. In contrast, under mixed substrate conditions, succinate was used primarily to fulfill the carbon requirements of the cell and to supply roughly one-third of the energy requirements. Half of this energy is produced concomitantly with biomass precursor formation.
The mechanism allowing the observed partitioning through the central metabolism involved the repression of methanol assimilation, not at the level of C1 activation but at the first step of the serine cycle. In addition, CO 2 assimilation via the phosphoenolpyruvate carboxylase and ethylmalonyl-CoA pathways, which represent 50% of the carbon assimilated into the biomass during pure methylotrophic growth conditions [20], appears to be blocked. Establishing this state of the network requires the coordinated control of several key points of the central metabolism by the cell. The regulation at the enzyme activity level was in part revealed by the enzyme activity measurements performed by Dunstan et al. [27]. In this study, the methanol dehydrogenase activity of M. extorquens AM1 grown in the presence of methanol and succinate was found to be significant, and no catabolic repression of the encoding gene occurred. Secondly, lower activities for the first enzymes of the anabolic serine cycle (i.e., serine-glyoxylate aminotransferase, hydroxypyruvate reductase and glycerate kinase) compared to methylotrophic growth were found [27], although its activities were significantly higher than under pure succinate incubation. These observations corroborate and explain the observed flux partitioning at the first step of the serine cycle, where methanol assimilation flow is mainly, but not entirely, restricted to C1-precursor biosynthesis. This specific regulation, which takes place during co-consumption of methanol plus succinate, is likely to rely on an active process for the activation and repression of a specific set of enzymes. The regulation of genes for enzymes involved in C1 dissimilation and C1 assimilation are at least partially known and are dependent on distinct regulators [32]. Some C1 dissimilation genes appear to be under the control of the methanol concentration [33], indicating that methanol utilization is not catabolically repressed by the presence of succinate. Note that this mechanism is crucial for methanol and formaldehyde detoxification. The genes responsible for C1 assimilation are clustered in 2 main operons and are regulated by a Lys-R-type transcriptional regulator (QscR) [32,34,35]. QscR is known to be expressed at low levels under both methanol and succinate conditions and responds to physiological signals. Indeed, the DNA binding of the regulator is enhanced by the presence of formyl-H 4 F [35] but reduced by NADP + , acetyl-CoA, and weakly reduced by glyoxylate [34]. On one hand, the formyl-H 4 F concentration is expected to increase under methanol utilization, which could enhance C1 precursor biosynthesis; on the other hand, a change in the energetic state of the cell (NADP+/NADPH) [23,31] and the concentrations of acetyl-CoA and glyoxylate, two key metabolites of the serine cycle, could induce a decrease in enzyme activities [31]. In addition, the regulator of the main enzyme of the ethylmalonyl-CoA pathway, crotonyl-CoA carboxylase/reductase, was recently identified [36]. A complex mechanism of regulation of the genes encoding the metabolic enzymes involved in C1 dissimilation and assimilation allows these processes to be decoupled, for instance, during co-consumption (this study) or during a substrate switch [23]. In fact, the mechanism of C1 pathway regulation in M. extorquens AM1, in addition to the high modularity of the central metabolic pathways, is crucial to enable the co-consumption of a C1 substrate (methanol) and a C4 organic acid (succinate).
The metabolic strategy of co-consumption contrasts with the intensively studied phenomenon of diauxic and catabolic repression [24,26,37]. Several hypotheses could explain co-consumption by M. extorquens AM1 rather than the successive use of succinate and methanol. First, it could be a strategy of resource management under environmental conditions. Indeed, methanol is only transiently released by plants during the diurnal cycle [9], and it is a volatile compound; therefore its co-consumption would lead to a delay in the exhaustion of succinate (or other organic acids) when methanol is available and assure the cell of prolonged substrate availability. In addition, except during the peak of emission in the morning, methanol is emitted at a low rate during the day. This might not be sufficient to sustain growth, such that an additional carbon substrate would consequently be required. Another point of consideration is that the growth rates during pure succinate and pure methanol growth are relatively similar, and mixed substrate conditions result in comparable growth. Likewise, Corynebacterium glutamicum exhibits identical growth rates during growth on glucose and acetate and co-consumes both substrates at the same growth rate under mixed substrate conditions [38]. Thus, both Methylobacterium and Corynebacterium reduce the uptake rates of both substrates under mixed substrate conditions. In contrast, Bacillus subtilis, which is able to grow on a mixture of glucose and malate, exhibits a higher growth rate during co-consumption than during pure culture conditions [39]. Interestingly, for Bacillus subtilis, both glucose and malate catabolically repress several other substrates [39]. Thus, similar growth rates on different substrates might lead to the convergence of substrate utilization rather than diauxie. The latter is the case for Escherichia coli and Azotobacter vinelandii, which exhibit different growth rates on glucose and acetate, respectively [40,41]. Mixed substrate conditions result in a diauxic shift in which by E. coli consumes glucose first and A. vinelandii switches from acetate to glucose metabolization.
Nonetheless, the co-consumption of methanol and succinate in M. extorquens could be a strategy to optimize substrate utilization because methanol theoretically has a higher C-mol yield of energy units (ATP, NADH, and NADPH) as well as C1 units (, 150%) compared to succinate but only a slightly higher C-mol yield for other biomass production (115%), based on yield calculations using a genome-scale model [20]. Thus, the partitioning observed constrains methanol to its most efficient utilization pattern. However, similar biomass yields were observed experimentally under each condition (methanol, succinate, and methanol plus succinate), indicating that no substantial gain resulted from the substrate partitioning.
Addressing the role of co-consumption challenges our current ability to measure cell physiology in situ (in planta). The results presented here for the co-consumption of methanol plus a C4 organic acid by M. extorquens AM1 might suggest that heterotrophy could be as important as methylotrophy for leaf colonization capacity by Methylobacterium and is in line with the finding that methylotrophic bacteria in the phyllosphere are facultative methylotrophs rather than obligate methylotrophs [3,42]. Thus, co-consumption might be a prevailing strategy under environmental conditions, where substrate availability is scarce, transient and diverse.  The partial pressure of dissolved oxygen was monitored using polarographic oxygen sensors (InPro 6800, Mettler-Toledo). The methanol concentration was determined by GC-flame ionization detection (GC-FID) (GC 6850, Agilent Technologies; column: DB-Wax, J&W Scientific). The succinate concentration was determined by HPLC-UV-DAD (column: Phenomenex Rezex ROA-organic acid H+7.8 mm) using tartaric acid as an internal standard. The 13 C enrichment of CO 2 in the exhaust gas was measured using two infrared sensors (BCP-CO2, BlueSens), one sensitive to 12 C CO 2 and the other sensitive to 13 C CO 2 . The calibration of each sensor and the specific correction of 12 C and 13 C signals were performed as recommended by the company. Cell dry weight (CDW) was determined upon growth on each substrate (methanol, methanol plus succinate, succinate). The results of 7 cultures were not significantly different, and an overall CDW value was found to be 0.26960.013 (2s).

Sampling, Quenching, and Extraction of Intracellular Metabolites
CoA-ester sampling was performed as follows: a volume of 1 mL of culture was directly injected into 4.5 mL of 220uC cold acidified acetonitrile containing 0.1 M formic acid on a Vortex [12]. The extraction was performed with the sample incubated for 15 min on ice and subsequently freeze-dried and stored at 220uC until analysis. Prior to analysis, dried samples were dissolved in 300 mL of 25 mM ammonium formate buffer (pH 3.5, 2% MeOH). The suspension was centrifuged (14,000 g, 2 min, 25uC), and the supernatant was filtered through a Sartorius Minisart filter (pore size 0.2 mm) before analysis. Amino acids and central metabolites were sampled as described previously [19]. In brief, 1 mL of culture was sampled by fast filtration and washed with 5 ml medium. The filters (RC Sartorius Minisart, pore size 0.2 mm) were directly transferred into shot bottles containing 8 ml of boiling water for quenching and extraction. The extracts were cooled on ice and filtered via a RC Sartorius Minisart filter (pore size 0.2 mm) and then chilled with liquid nitrogen. All samples were lyophilized immediately and stored at 220uC. Dried samples were dissolved in 100 ml double-distilled water and diluted 30/70 (v/v) with acetonitrile prior to analysis.

LC-MS Analysis
LC-MS analysis was performed using a Rheos 2200 HPLC system (Flux Instruments) coupled to an LTQ Orbitrap mass spectrometer (Thermo Fisher Scientific) and equipped with an electrospray ionization probe. CoA-thioesters were separated using a previously described procedure [12], with slight modifications [43]. Polar intracellular metabolites were separated on a pHILIC column (15062.0 mm, particle size 5 mm; Sequant, Umea, Sweden) as previously described [19]. Phosphorylated hexoses were not separated; thus, the data given are an average of the hexose-phosphate pool.
The LC-MS system was equilibrated for 6 min under initial elution conditions between two successive analyses. The LC was coupled to the mass spectrometer. The sheath gas flow rate was 40, the auxiliary gas flow rate was 30, the tube lens was 80 V, the capillary voltage was 35 V, and the ion spray voltage was 4.3 kV. MS analysis was performed in the FTMS positive mode to analyze CoA esters and amino acids and in the negative FTMS mode for all other compounds at a resolution of 60,000 (m/z 400).

Data Analysis
The incorporation of 13 C label into metabolites during the 13 Clabeling experiment was calculated from the analysis of the mass isotopomer distribution (MID) in the mass spectra. The resolution of 60,000 (m/z 400) allowed the separation of carbon, nitrogen and oxygen mass isotopomers; therefore, only carbon MIDs were considered, and correction for naturally occurring isotopes of the other elements was not required for most metabolites. However, the mass resolution of the Orbitrap decreases with increasing m/z values. In case of co-esters (m/z .800), analysis yielded mass resolution below 42,000, resulting in incomplete separation of the isotopes requiring correction for the contributions of N, O and S to MID. The standard deviations (STD) of the measurements were considered to be at least 2%, higher than found over the 3 technical replicates (average STD in amino acids: 0.53%). This was likely due to systematic error resulting from lower linearity of the LTQ-Orbitrap, which was evaluated to be lower than 2% in the intensity range considered. The average 13 C labeling AL 13C was calculated as follows: AL 13C~P i~n i~1 (H i |i) n n = number of carbon atoms, H i relative abundance of the monoisotopic mass + i fraction.

Calculation of Methanol and Succinate Contribution to Energetics
Feasibility analysis and Flux variability analysis was performed using the genome-scale metabolic network of M. extorquens AM1 [20]. Calculations were performed using CellNetAnalyser [44] based on the M. extorquens AM1 biomass composition determined during methylotrophic growth [20] and under the constraints (upper and lower flux limits were fixed corresponding to 2 standard deviations) of the measured fluxes (uptake and production rate, growth rate); see Table S1 for details. All exchange fluxes of carbon sources except methanol and succinate were set to 0, and a network reduction step was applied as described by Peyraud et al. [20] to make accurate predictions. Therefore, the identified methylotrophic sub-network [20] was amended by the opening reactions specifically used under succinate conditions (Table S2). The contributions of methanol and succinate to cell energetics were calculated from a feasible solution (Table S3) under the given constraints, and their deviation was performed by Flux Variability Analysis (Table S4); see Table S5 for calculation results. Figure S1 Mass isotopomers distribution of central metabolites measured by LC-MS of M. extorquens AM1 upon co-consumption with 13 C (.99%) methanol and natural abundance (1.1% 13 C) succinate. Mass isotopomer data in black correspond to samples collected during mid-co-consumption phase (Sampling time 1, see Fig. 1), and in red to samples collected at the end of the co-consumption phase (Sampling time 2). (PDF)  Figure S4 Mass isotopomers distribution in CoA thioesters measured by LC-MS during growth of M.extorquens AM1 upon co-consumption with 13 C (.99%) methanol and natural abundance (1.1% 13 C) succinate. Metabolite quenching, extraction and measurements were per-formed as described in material and methods. Mass isotopomer data correspond to sample collected during mid-co-consumption phase (Sampling time 1, see Fig. 1). (PDF) Figure S5 Scheme of the Coenzyme A biosynthesis in M. extorquens AM1. Identified 13 C carbon entry points are indicated by colored cycles. Red, C1-precursor from tetrahydrofolate pathway; yellow, CO 2 ; green, C3 carbon of serine. (PDF) Table S1 Lists of parameters used for flux balance analysis and flux variability analysis. (PDF)  Author Contributions