3.1. HT cultivation bioprocess parameters
The series of previously available 3-HP-producing P. pastoris strains[4, 5] were grown in glycerol batch cultures at pH 5 using a HT 13C-fluxomics platform [35]. The stoichiometry of the engineered reactions and corresponding enzymes in the 3-HP-producing P. pastoris strains are depicted in Fig. 1. Strain engineering strategies were aimed at increasing the delivery of the substrates of the malonyl-CoA to 3-HP pathway (i.e., cytosolic acetyl-CoA, malonyl-CoA, and NADPH), and the reduction of the production of the main by-product (D-arabitol).
<Suggested position for Fig. 1>
Measurement of the extracellular metabolites and the biomass concentration allowed calculating the bioprocess parameters (specific substrate consumption and (by)products production rates (q-rates) and µmax) for each strain (Fig. 2), which is a prerequisite for metabolic flux calculations. Results were consistent with previously reported cultivation data [4, 5]. In a nutshell, comparison between strains PpHP1 and PpHP2 reveals that dissection of MCR from C. aurantiacus into each of its two subunits led to a 10-fold increase in the 3-HP yield, whereas the overexpression of the genes encoding for an acetyl-CoA carboxylase from Yarrowia lipolytica (ACC1) and a cytosolic version of the mitochondrial NADH kinase from Saccharomyces cerevisiae (cPOS5) in strain PpHP6 led to a further increase in product yield. In addition, the expression of a second copy of the gene encoding the C-terminal domain of MCR yielded the highest 3-HP-producing strain (PpHP8), which was also the slowest growing strain and the strain producing the largest amount of D-arabitol in batch cultures. As previously reported, the overexpression of the genes encoding for the cytosolic acetyl-CoA synthesis pathway (i.e. acsSeL641P and PDC1, Fig. 1) in the strain PpHP8 (yielding strain PpHP18) restored the growth rate, but the 3-HP yield dropped drastically [5]. These results pointed at a limitation of resources in PpHP8 when growing at maximal growth rate [5], thereby resulting in an opposite trend between biomass and product yields.
Results in Fig. 2 also show that not only the growth rate of strain PpHP8 was remarkably lower than the growth rate of the reference strain (0.13 h− 1 and 0.22 h− 1, respectively), but also the substrate uptake rate for PpHP8 was lower than the uptake rate of strain X-33 (2.10 and 3.37 mmol gCDW− 1 h− 1). Noticeably, the glycerol uptake rate for all the recombinant strains grown at pH 5 fell within the same range (2.90 to 3.37 mmolGlyc gCDW− 1 h− 1), except for strain PpHP8, which was remarkably lower (2.1 mmolGlyc gCDW− 1 h− 1). A high activity of the malonyl-CoA to 3-HP pathway probably led to a reduced availability of acetyl-CoA for this strain. Acetyl-CoA plays a central role on the biosynthesis of precursors, and it plays a key role in physiological regulation processes, such as the acylation of histones [45]. Growth defects have also been reported in S. cerevisiae strains harbouring a high acetyl-CoA carboxylase activity, where depletion of acetyl-CoA was described as the most likely cause of hampered growth [46]. Moreover, the observed increase in the growth rate and the uptake rate when acsSeL641P was overexpressed support this hypothesis.
Strikingly, the biomass yield for PpHP8-derived strains overexpressing acsSeL641P (PpHP13, ppHP17), and PDC1 plus acsSeL641P (PpHP18) were lower than the biomass yield of strain PpHP8.
<Suggested position for Fig. 2>
3.2. Fluxome of P. pastoris 3-HP-producing strains at pH 5
The bioprocess parameters of the series of strains obtained from batch cultivations and the amino acids isotopologue distributions derived from the corresponding labelling experiments were used to calculate the intracellular fluxes using influx_si. The fluxome of each strain can be found in the Supplementary File S12. A summary of the results can be found in Supplementary File S13.
The metabolic flux profile of the reference strain (X-33) obtained using the HT robotic platform is comparable to the previously reported results for the same strain growing in glycerol chemostats on a similar medium [22]. Noticeably, the fluxes of the upper glycolysis (UG) and pentose phosphate pathway (PPP) in our batch experiments (i.e. at µmax) were higher than the ones observed in chemostat cultures at lower growth rates (0.05, 0.10 and 0.16 h− 1), coherent with the positive correlation between growth rate and the UG and PPP fluxes previously observed in glycerol chemostats (Supplementary Fig. 1). Similarly, fluxes of the lower glycolysis (LG) and the tricarboxylic acid (TCA) cycle reactions were lower than the ones observed in chemostat cultures at lower growth rates, also coherent with the reported inverse correlation between growth rate and LG and TCA cycle fluxes.
To facilitate comparison of metabolic flux distributions amongst strains, a heat map illustrating the fold-change between the relative fluxes (i.e. normalised to the specific glycerol uptake rate) of each recombinant strain compared to the relative fluxes of the reference strain cultivated at pH 5 is shown in Fig. 3. The most drastic changes were observed in the relative fluxes through the UG and PPP reactions. First, when the MCR activity was increased by expressing separately the two MCR domains (i.e. PpHP2 compared to PpHP1), the fluxes of the UG and PPP increased noticeably (10∼25%). Such trend can be explained by increased NADPH requirements, as NADPH is used as the electron donor for the two consecutive reactions catalysed by MCR. When the gene encoding the heterologous cytosolic NADH kinase (cPOS5Sc) was overexpressed (PpHP5), the observed fluxes of the UG and PPP reactions were 15∼30% lower compared to PpHP2. It is well known that the NADPH/NADP+ ratio controls the fluxes towards the oxidative branch of the PPP [47]. In addition, overexpression of cPOS5 in P. pastoris provides the cell with an additional source of NADPH, leading to a higher NADPH/NADP+ ratio [48]. Therefore, decrease of UG and PPP fluxes in PpHP5 compared to PpHP2 was consistent with an increased NADPH/NADP+ ratio.
No major changes were observed when ACC1Yl was overexpressed (i.e. in strain PpHP6, compared to strain PpHP5). For the strain PpHP8, which harboured an additional copy of the gene encoding for MCR-CCa, the highest fluxes towards 3-HP production were observed, while the UG and PPP fluxes were the lowest among all strains. Such observation agrees with the results for the strain PpHP5, as the increase in the NADPH requirements due to production of 3-HP followed independent trends with the fluxes of the oxidative branch of the PPP.
Heterologous expression of acsSeL641P in strain PpHP8 (i.e. generating strain PpHP13) led to a drastic switch in the strain’s fluxome. The relative fluxes through the UG and PPP increased remarkably in PpHP13, while showing a lower relative flux towards 3-HP production compared to PpHP8 or PpHP6 strains. Moreover, the biomass yield of PpHP13 was lower. Therefore, considering the NADPH requirements for biomass and 3-HP production, increased production of NADPH through the PPP seems unfounded. Deletion of the gene encoding for the NADPH-dependent arabitol dehydrogenase enzyme (ArDH) in strains PpHP8 and PpHP13 (i.e. obtaining strains PpHP15 and PpHP17, respectively) led to minor changes in the strains’ fluxomes under the tested growth conditions. Finally, overexpression of PDC1 in PpHP17 (resulting in PpHP18) led to the highest relative flux through the UG and PPP. However, neither the biomass yield nor the 3-HP yield were affected (Fig. 2).
Changes in the fluxes through LG and the TCA cycle reactions followed the opposite trend to the UG and the PPP. Small differences in the fluxes through the glyoxylate cycle were also observed. However, as the absolute values of these fluxes were low (below 0.025 mmol mmol− 1 h− 1), absolute changes of these fluxes did not have an impact on the strain’s biomass and product yields.
<Suggested position for Fig. 3>
Further comparison of the absolute flux distributions (i.e. non-normalised to the glycerol specific uptake rate) in the reference and the 3-HP-producing strains PpHP8 and PpHP18 (Fig. 4) provided additional insights.
<Suggested position for Fig. 4>
First, overexpression of the 3-HP production pathway led to a higher pyruvate decarboxylase flux at the pyruvate node in PpHP8 compared to the reference strain (0.30 mmol gCDW−1 h− 1 and 0.12 mmol gCDW−1 h− 1, respectively). The overexpression of ACSSeL641P and Pdc1 in strain PpHP8 did not result in a higher flux towards cytosolic acetyl-CoA in strain PpHP18, as the absolute flux values for the two strains were identical (0.30 mmol gCDW−1 h− 1). Compared to the reference strain, overexpression of the cytosolic acetyl-CoA production pathway in strain PpHP18 did not increase the LG fluxes. All these findings agree with previous results in S. cerevisiae, where the overexpression of PDC1 led to a higher flux towards this pathway without increasing the glycolytic flux [49]. It is well described that the glycolytic flux is tightly controlled in yeast S. cerevisiae, and the glycolytic flux cannot be increased by overexpressing individual enzymes [50]. In the case of P. pastoris, increased glycolytic fluxes have only been described under low oxygen availability [51] or when a transcription factor controlling the expression of all the glycolytic genes was overexpressed [52]. Therefore, as pyruvate is pulled into the production of 3-HP but the glycolytic flux and the uptake of glycerol do not increase, the overall ATP yield of the 3-HP-producing strains decreases.
Strain PpHP18 has remarkably higher UG and PPP fluxes than strain PpHP8 (Fig. 4). The UG and PPP have a low carbon and energy yield. Therefore, while PDC1 is being overexpressed, the energy requirements in strain PpHP18 sinked the pyruvate into the TCA cycle for ATP generation, hampering the flux toward cytosolic acetyl-CoA and, ultimately, reducing the 3-HP yield. On the contrary, as PpHP8 grew at a lower rate, the energy requirements of the strain were reduced, leaving more substrate available to produce 3-HP.
Overall, comparison of absolute flux distributions suggests that, to further increase the 3-HP yield in strain PpHP8, the glycolytic fluxes would need to be significantly increased. Moreover, results also point to high ATP requirements in strain PpHP18 are the cause of the differences between these two strains.
3.3. ATP and NADPH producing/consuming fluxes of the 3-HP-producing P. pastoris strains
The results from the previous section show that the fluxes through the PPP decreased when a heterologous cytosolic NADH kinase was expressed. Consistently, NADPH production rates calculated from the 13C-flux data (Fig. 5A) indicate that the NADPH produced by the glucose-6-phosphate dehydrogenase (G6PDH) reaction was lower than the actual NADPH requirements in some of the strains overexpressing the cPOS5 gene (i.e. strains PpHP5, PpHP6, andPpHP8), supporting that the NADH kinase reaction contributed to cover the cell’s NADPH requirements.
Strains PpHP13, PpHP17, and PpHP18, which overexpress the cytosolic acetyl-CoA production pathway (i.e. Pdc1 and ACSSeL641P), also produced more NADPH through the oxidative branch of the PPP than the actual cell requirements. The substantial increase in the fluxes through UG and PPP for such strains coincided with an increase in the specific glycerol uptake rate and the growth rate (Fig. 5A, and Fig. 2), but the biomass yield decreased (Fig. 2). Despite the large standard deviation of the fluxes through the UG and PPP fluxes in strains PpHP13, PpHP17, and PpHP18, it can be concluded that these strains did not benefit from the heterologous expression of the cPos5, while PpHP5, PpHP6, and PpHP8 do. NADPH production by NADH kinase is more efficient in terms of both carbon and ATP conservation than the use of the UG and PPP, which would explain the reduction in the biomass and 3-HP yield in the strains PpHP13, PpHP17, and PpHP18.
To corroborate the consistency of the observed metabolic fluxes, FBA was used to verify the redox and energy conservation of the 13C-MFA results. The maximization of the flux through the ATP sink reaction (ATPM) was used as an objective function. This is the function best describing the intracellular fluxes of a cell culture growing on a batch culture with an excess of substrate [44]. Consequently, the resulting flux values are based on in vivo data and, at the same time, they fulfil the biological function of optimizing the biomass yield. FBA results confirmed that both redox and energy balances could be conserved at the given experimental fluxes. The FBA results were also used to calculate the ATP balance (Fig. 5).
A higher ATP of maintenance for strains PpHP13, PpHP17, and PpHP18 was observed, as depicted in Fig. 5B, being PpHP18 the strain with the highest ATP requirements among all strains. Conversely, the calculated ATP requirement for strain PpHP8 was the lowest, particularly due to the lowest growth associated maintenance energy (GAME) requirements, as it is the slowest growing strain.
Figure 5B shows there is a direct correlation between the overall ATP requirements of the strain and the fraction of pyruvate that was directed into the TCA cycle for ATP production. While 47.1% of the pyruvate was channelled into the mitochondria in the reference strain, 32.1% was directed through the same pathway in strain PpHP8. The mitochondrial transport of pyruvate raised to 40.2% and 41.9% when the cytosolic acetyl-CoA pathway was expressed (strains PpHP13 and PpHP18, respectively) to compensate for the higher ATP requirements.
<Suggested position for Fig. 5>
Altogether, these results show the correlation between acetyl-CoA availability, growth rate, ATP requirements, and flux distribution at the pyruvate node, and their impact on the 3-HP production yield. Acetyl-CoA depletion in strain PpHP8 hampered growth rate. When the cytosolic acetyl-CoA biosynthetic pathway was overexpressed (strains PpHP13, PpHP17, and PpHP18), growth rate increased. Increase in the growth rate increased ATP requirements, which increased channelling of pyruvate into the TCA cycle, hampering the 3-HP product yield.
3.4. Fluxome of P. pastoris strains at pH 3.5
It has been reported that a pH of 3.5 (i.e., 1 unit below the pKa of 3-HP) was optimal for the further downstream processing of 3-HP by solvent extraction [12]. Therefore, the reference strain (X-33) as well as some 3-HP-producing strains (i.e. PpHP1, PpHP6, PpHP8, PpHP15, and PpHP18) were further tested at pH 3.5.
No remarkable differences in the fluxome between the two conditions were observed for the reference strain (Fig. 6A). In contrast, higher UG and PPP fluxes were observed at pH 3.5 than at pH 5 most strains PpHP1, PpHP6, PpHP8, PpHP15 (See Fig. 6B). In addition, a higher production of D-arabitol was also observed in most strains when growing at low pH (Fig. 6B and Supplementary Figure S2).
The fold-change of the relative fluxes of each recombinant strain compared to the reference strain (Fig. 6C) showed that the impact of each genetic modification in the flux of each strain followed a similar trend at pH 5 and pH 3.5 (Fig. 3 and Fig. 6B, respectively). 3-HP was produced at pH 3.5, but the yield was slightly lower than the one achieved at pH 5 for all the tested strains. For instance, the highest 3-HP producing strain at both pH values was PpHP8, which produced 3-HP at a yield of 0.084 ± 0.007 Cmol Cmol− 1 at pH 3.5 (0.081 ± 0.006 g g− 1), which is 23% lower than the product yield of the same strain at pH 5. See Supplementary Figure S3 for the comparison of the fluxes at pH 5 and 3.5 for this strain.
Regarding the NADPH production and consumption fluxes for each strain, the same trends were also observed at pH 3.5 (Supplementary Figure S4), that is, NADPH requirements in strain PpHP8 exceeded NADPH production from the PPP, meaning the flux through the cytosolic NADH kinase reaction compensated for that difference. In contrast, the NADPH production through the PPP in strain PpHP18 greatly exceeded the requirements. Moreover, the substrate uptake rate also followed the same trend for all the strains in both pH conditions (Supplementary Figure S4).
Altogether, these results contribute to the understanding of the adaptation of this yeast to a low pH at a fluxome level. Production of D-arabitol at acidic pH was increased for all the 3-HP-producing strains (from 2 to 20-fold). P. pastoris produces D-arabitol under several stress conditions, such as under hypoxia or osmotic stress [51, 53]. Thus, higher D-arabitol production at a lower pH is probably due to a stress response. Moreover, the biomass yield at pH 3.5 was lower for most strains (Fig. 6B), indicating a higher ATP requirement for maintenance. Such results have already been described in other yeasts grown at lower pH, such as S. cerevisiae, where the decrease in the biomass yield was also attributed to an increase in the ATP of maintenance [54]. Moreover, the 3-HP yield was lower than the one of the same strains at pH 5, consistently with previous studies describing 3-HP production in S. cerevisiae grown at pH 3.5 [55]. The observed decrease in the product yield when the ATP usage increases confirms that ATP is a limiting factor towards increasing the 3-HP yield. Similarly, increased D-arabitol production, which is a NADPH sink, can also explain the decrease in the 3-HP yield observed at low pH.
Still, as metabolic flux profiles at pH 3.5 remained mostly unchanged compared to those at pH 5, it is likely that strain engineering strategies at both pH will have the same outcome.
<Suggested position for Fig. 6>