Multi-step pathway engineering in probiotic Saccharomyces boulardii for abscisic acid production in the gut

Abstract

The plant hormone abscisic acid (ABA) has gained attention for its role in animals and humans, particularly due to its protective effects in various immune and inflammatory disorders. Given its high concentrations in fruits like figs, bilberries and apricots, ABA shows promise as a nutraceutical. However scalability, short half-life and cost limit the use of ABA-enriched fruit extracts and synthetic supplements. In this study, we propose an alternative ABA administration method to overcome these challenges. We genetically engineered a strain of the probiotic Saccharomyces boulardii to produce and deliver ABA directly to the gut of mice. Using the biosynthesis pathway from Botrytis cinerea, four genes (bcaba1-4) were integrated into S. boulardii, enabling ABA production at 30 °C, as previously described in Saccharomyces cerevisiae. Introducing an additional cytochrome P450 reductase gene resulted in a 7-fold increase in ABA titers, surpassing previous ABA-producing S. cerevisiae strains. Supplementation of the ABA-producing S. boulardii in the diet of mice (at a concentration of 5 × 10⁸ CFU/g) led to effective gut colonization but resulted in low serum ABA levels (approximately 1.8 ng/mL). The absence of detectable serum ABA after administration of the ABA-producing probiotic through oral gavage, prompted further investigation to determine the underlying cause. The physiological body temperature (37 °C) was identified as a major bottleneck for ABA production. Modifications to enhance the mevalonate pathway flux improved ABA levels at 37 °C. However, additional modifications are needed to optimize ABA production before testing this probiotic in disease contexts in mice.

Keywords: Abscisic acid; In situ production; Metabolic engineering; Probiotic engineering; Saccharomyces boulardii.

Fig. 1

ABA production in S. boulardii and S. cerevisiae strains in vitro at 30°C. a ABA titer in supernatant of Enterol and SbP strains containing bcaba1234. Strains were cultivated for 24 h at 30 °C in 50 mL of YPD medium. Average ABA titers were calculated from 5 independent biological replicates. b ABA titer in supernatant of strains E-ABA1 and Table A3. E-ABA1 is based on the genetic background of S. boulardii strain Enterol® and contains bcaba1234, Table A3 (Otto et al., 2019) is based on S. cerevisiae strain CENK.PK113-5D and contains bcaba1234, bccpr1 and tHMG1. Strains were cultivated for 12 h or 24 h at 30 °C in 20 mL of YPD medium. Average ABA titers were calculated from 3 independent biological replicates. Data are represented as mean ± SEM and statistically analyzed using a Student's t-test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig. 2

ABA production in E-ABA strains containing a copy of a cytochrome P450 reductase gene. (a–b) All strains were based on the E-ABA1 strain. Average ABA titers were calculated from 4 to 6 independent biological replicates. Effect of bccpr1 (E-CABA1), fusion of bccpr1 to bcaba1 (E-CABA2) or bcaba2 (E-CABA3), fusion of bmr to bcaba1 (E-CABA4) or bcaba2 (E-CABA5) and combined fusion of bccpr1 to bcaba1 and bmr to bcaba2 (E-CABA6) or bccpr1 to bcaba2 and bmr to bcaba1 (E-CABA7) on ABA titers. a Strains were cultivated for 24 h at 30 °C in 3 mL of YPD medium in 24-well deep well plates. b Strains were cultivated for 24 h at 30 °C in 3 mL of YPD medium in test tubes. Data are represented as mean ± SEM and statistically analyzed using a one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig. 3

Determination of CFU count in the feces of conventional mice that received either the yeast supplemented diet or were administered the yeast through oral gavage. a Schematic representation showing treatment of the mice. First, conventional mice were fed a yeast-supplemented diet containing 5 × 108 CFU/g of food. The diet included either E-ABA1 or SbP-ABA1 and was administered for 3 consecutive days (day 0, day 1, day 2) (yellow dots). Second, conventional mice received 3 doses of 5 × 108 CFU of SbP-ABA1 through oral gavage (day 0, day 1, day2) (yellow dots). The time points for collection of feces are depicted by brown dots. b Fecal carriage of yeast in both set-ups. Fecal samples were collected as shown in the timelines, plated on YPD media and kept at 30 °C for 2 days. Data are represented as mean ± SEM (n = 3 mice).

Fig. 4

ABA titers in serum of mice fed or gavaged with an ABA-producing yeast for 1 week. a Mice were either fed a yeast-supplemented diet containing ABA-producing strain SbP-ABA1 (5 × 108 CFU/g of food) or received oral gavage with this strain (5 × 108 CFU). b Mice were fed a yeast-supplemented diet containing either wild-type Enterol (5 × 108 CFU/g of food), the ABA-producing Enterol strain E-ABA1 (5 × 108 CFU/g of food) or the ABA-producing S. cerevisiae strain Table A3 (2 × 108 CFU/g of food). c Mice were fed a yeast-supplemented diet containing either wild-type Enterol (5 × 108 CFU/g of food), the ABA-producing Enterol strains E-ABA1 (5 × 108 CFU/g of food) or E-CABA1 (5 × 108 CFU/g of food) or a diet supplemented with synthetic ABA (100 mg/kg of food). Serum was taken at sacrifice after 1 week of treatment. Each dot represents one mouse. Data are represented as mean ± SEM and statistically analyzed using a one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig. 5

Effect of temperature on ABA production levels and RNA expression levels of bcaba1234. a ABA titer in supernatant of strains E, E-ABA1, E-CABA1 and SABA3. E-ABA1 and E-CABA1 are based on the genetic background of S. boulardii strain Enterol and contain bcaba1234, and bcaba1234 and bccpr1 respectively. SABA3 (Otto et al., 2019) is based on S. cerevisiae strain CENK.PK113-5D and contains various modifications as shown in Table 1. Strains were cultivated for 24 h in 3 mL of YPD medium in 24-well deep well plates. Average ABA titers were calculated from 3 independent biological replicates and error bars indicate the standard error of the mean. b OD600 of strains analyzed in Fig. 5A c mRNA expression of aba1, aba2, aba3 and aba4 in the pellet of a 24 h culture of E-ABA1 grown at 30 °C or 37 °C in 3 mL of YPD in test tubes. Average expression was calculated from 3 independent biological replicates. Data are represented as mean ± SEM and statistically analyzed using a Student's t-test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig. 6

ABA production in E-ABA strains containing an additional copy of aba1 and aba3 genes expressed under various promoters. All described strains were based on the E-ABA1 strain. Average ABA titers were calculated from 3 independent biological replicates. Effect of an additional copy of aba1 and aba3 expressed under control of the TDH3 (E-ABA2), TEF1 (E-ABA3), RPL18B (E-ABA4) or SAC6 (E-ABA5) promoter on ABA titers. Strains were cultivated for 24 h in 3 mL of YPD medium in 24-well deep well plates. a Cultures were grown at 30 °C. b Cultures were grown at 37 °C. Data are represented as mean ± SEM and statistically analyzed using a one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig. 7

ABA production in E-ABA strains containing a copy of tHMG1 and/or cpr1. a-c The effect of adding tHMG1 to E-ABA2, E-ABA3, E-ABA4 and E-ABA5 on ABA titers during cultivation at 30 °C (a) or 37 °C (b,c). d-e The effect of adding cpr1 to E-Table A2 and E-Table A3 on ABA titers during cultivation at 30 °C (d) or 37 °C (e). Strains were cultivated for 24 h in 3 mL of YPD medium in 24-well deep well plates. Average ABA titers were calculated from 3 to 5 independent biological replicates. Data are represented as mean ± SEM and statistically analyzed using a one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.