Conversion of CO2 into organic acids by engineered autotrophic yeast

Significance Industrial biotechnology bears great potential to reduce CO2 emissions by producing chemicals from renewable agricultural feedstocks, however at the risk to compete with food production. While most industrially relevant organisms are heterotrophs there has been recent progress to equip some with a CO2 fixation pathway leading to autotrophic growth. We have recently developed a synthetic autotrophic strain of the industrial yeast Komagataella phaffii. Here we integrated the pathways to lactic and itaconic acid (two chemical building blocks) into this strain. Up to 2 g L−1 of itaconic acid were produced from CO2 as the only carbon source. This work paves the way toward net CO2 capturing into long living chemical products based on a synthetic autotrophic chassis strain.


Supplementary
. Lactic acid production was not improved by using the reverse engineered K. phaffii strain (1). (A) Growth profile, (B) lactic acid production profile. Time axis corresponds to the production phase under autotrophic conditions. At least three biological replicates were used in the screening to monitor the producing strains. Shades represent the standard deviations (±).eng strain: reverse engineered autotrophic K. phaffii strain. Figure S4. Control cultivations performed with unlabeled glycerol and methanol in the 13 C labeling experiment showed similar profiles to the labelled cultures. (A) Growth profile of unlabeled and labeled itaconic acid production and non-grower strain, (B) growth profile of unlabeled and labeled lactic acid production and non-grower strain, (C) itaconic acid production profile of unlabeled and labeled producer and non-grower strain, (D) lactic acid production profile of unlabeled and labeled producer and non-grower strain, (E) carbon isotopologue distribution in the produced itaconic acid, (F) carbon isotopologue distribution in the produced lactic acid. M# denotes the number of 13 C carbons in the respective organic acid and 13 C-deg indicates the 13 C labeling degree of every sample. Time axis corresponds to the production phase under autotrophic conditions. At least three biological replicates were used in the screening to monitor the strains. Time axis corresponds to the production phase under autotrophic conditions. Shades represent the standard deviations (±). Figure S5. Isotopologue distribution of the carbon atoms of the reversed labeled samples in the produced (A) itaconic acid and (B) lactic acid. M# denotes the number of 13 C carbons in the respective organic acid and 13 C-deg indicates the 13 C labeling degree of every sample. Error bars indicate the standard deviation of three biological replicates. Figure S6. Production of organic acids in the synthetic autotrophic K. phaffii strain depends on elevated CO2 concentration. (A) Setup 1: cultures are incubated after the start of the production phase at 5% CO2 and after the sampling point at 115h put to ambient CO2 concentration. Setup 2: cultures are incubated after the start of the production phase at ambient CO2 concentration and after the sampling point at 115h put to 5% CO2. (B) Growth profiles of cultures cultivated according to setup 1 are shown with a solid line and cultures cultivated according to setup 2 are shown with a dashed line. (C) Organic acid production profiles of cultures cultivated using setup 1 (solid line) and cultures cultivated using setup 2 (dashed line). Time axis corresponds to the production phase under autotrophic conditions. Shades represent the standard deviation of 3 biological replicates (±). Figure S7. Modification of process conditions shows different results for the different products. (A) Growth profiles of the itaconic acid production strain and the control using different starting biomass at 5% CO2 concentration. (B) Growth profiles of the itaconic acid production strain and the control using different starting biomass at 10% CO2 concentration. (C) Growth profiles of the lactic acid production strain using different starting biomass at 5% CO2 concentration. (D) Growth profiles of the lactic acid production strain using different starting biomass at 10% CO2 concentration. (E) Glycolic acid production profiles of the different starting biomasses of the lactic acid production strain at 5% CO2 concentration. (F) Glycolic acid production profiles of the different starting biomasses of the lactic acid production strain at 10% CO2 concentration. Time axis corresponds to the production phase under autotrophic conditions. Shades represent the standard deviations of biological replicates (±). Table S1. Specific growth rates, specific production rates, and yields of the strains used in this study. The values given in ranges belong to multiple screenings.  (2).

Optimization of GC-EI-TOFMS data evaluation for isotopologue distribution analysis
For each analyte two fragments, namely M-57 (= [M-C4H9] + ) and M-15 (= [M-CH3] + ), were evaluated in both profile and centroid mode using a symmetric extraction window of ± 50 ppm, resulting in a total of four different evaluation methods.
For the assessment of the different evaluation methods, the 13 C labeling degree was calculated based on the isotopologue fractions as defined by Mairinger et al. (3)  For samples cultivated on nat C media (media with a natural carbon isotope distribution), data for a total of 15 biological replicates were available, more specifically 3 replicates each for 2 time points for the knock out strain and 3 time points for the producing strain with the RuBisCO gene, resulting in a total of 5 groups of n=3 replicates.
The measured 13 C labeling degree of these 15 nat C samples was compared to the expected 13 C labeling degree deduced from the natural carbon isotope distribution (see Supplementary Figure S8 for itaconic acid and Supplementary Figure S10 for lactic acid).
For itaconic acid the evaluation of the M-57 in profile and M-15 in centroid mode shows no significant difference to the expected value, nevertheless M-57 in profile mode had to be rejected due to the high standard deviation. Generally, all results show higher standard deviation for nat C samples, as higher mass isotopologues are of low intensity in these samples.
For the 13 C samples (see Supplementary Figure S9), whose 13 C labeling degree cannot be compared to any expected values as it is the case for the nat C samples, Analysis of Variance (ANOVA) was applied for each replicate group in order to check for statistically significant differences between the results obtained with the different evaluation methods (α=0.05, p-values in Figure legend). For all 5 replicate groups of the 13 C samples no significant difference could be observed for the four different evaluation methods.
In summary, the evaluation of the M-15 itaconic acid fragment in centroid mode gave best results in terms of deviation from the expected value and precision under repeatability conditions of measurement (see Supplementary Figure S8). Thus all data presented for itaconic acid in this paper are based on that data evaluation method (see Experimental).
Supplementary Figure S8. 13 C labeling degree of itaconic acid (average ± standard deviation) obtained using four different data evaluation methods (orange bars) for n=15 12 C samples (5 groups of 3 biological replicates each). The blue line depicts the expected value for the nat C samples.
Supplementary Figure S9. 13  The same approach was used for the comparison of the evaluation methods for lactic acid.
For the nat C samples (n=15, see Supplementary Figure S10) both data evaluation methods based on fragment M-15 showed comparable deviation from the expected value and variance.
In contrast to itaconic acids evaluation of the 13 C samples (Supplementary Figure S11) shows no significant differences between the four sample evaluation methods (p-values for ANOVA in Figure  legend) for lactic acid.
As both data evaluation methods based on the fragment M-15 can be regarded as not significantly different in terms of deviation from the expected value and precision under repeatability conditions of measurement, data evaluation for lactic acid was also carried out in centroid mode evaluating the fragment M-15.
Supplementary Figure S10. 13 C labeling degree of lactic acid (average ± standard deviation) obtained using four different data evaluation methods (orange bars) for n=15 12 C samples (5 groups of 3 biological replicates each). The blue line depicts the expected value for the nat C samples.