Overexpression of the Did2 gene promoted β-carotene synthesis
The optimization of metabolic pathways indirect related to β-carotene synthesis improves the production of β-carotene. Engineered Y. lipolytica strain is oleaginous yeast, has excellent potential for β-carotene production, and a high capacity to store β-carotene [17]. To further explore the potential of engineered Y. lipolytica strain for producing β-carotene, we studied the effects of the indirectly related metabolic pathways on the β-carotene synthesis pathway. Several genes outside the β-carotene synthesis pathway have been shown to affect the production of β-carotene in S. cerevisiae [12, 18]. For example, the incorporation of the Did2 gene, a member of the MVB sorting pathway, led to the most significant improvement of β-carotene yield by engineered S. cerevisiae [13]. The YL-C1 strain is an engineered Y. lipolytica with basal β-carotene producing capacity. We integrated the Did2 gene into YL-C1 at the Gut2 site, resulting in stain YL-C31 to explore the effect of the Did2 gene on β-carotene synthesis in the engineered Y. lipolytica strain. The strain YL-C2, which the Gut2 gene was knocked out, was used as second control since the integration of the Did2 gene interrupted the Gut2 gene. β-carotene content was analyzed using HPLC after 96 h of fermentation. The β-carotene content in YL-C1, YL-C2, and YL-C31 strains was 9.85, 8.87, and 15.63 mg/g DCW, respectively; and the β-carotene concentration was 49.86, 51.45, and 80.65 mg/L, respectively (Fig. 1a, 1b). Both the content and concentration of β-carotene in the YL-C31 strain are highest. These results indicate that overexpressing the Did2 gene increased the β-carotene production in the engineered Y. lipolytica strain.
ATP and NADPH decreased after the overexpression of the Did2 gene
Coenzyme factors are essential for β-carotene synthesis. Synthesis of one mole of β-carotene requires 8 moles of ATP, 8 moles of CTP, and 16 moles of NADPH [11]. The levels of ATP and NADPH, coenzyme factors during the logarithmic growth phase in the engineered Y. lipolytica strains were examined to explore the reason that the overexpression of the Did2 gene improved the β-carotene synthesis. For ATP, the YL-C1, YL-C2, and YL-C31 strains produced 19.25, 138.34, and 45.19 nmol/g protein of ATP, respectively (Fig. 2a). For NADPH, the YL-C1, YL-C2, and YL-C31 strains produced 207.65, 162.35, and 127.49 nmol/g protein of NADPH, respectively (Fig. 2b). The ATP concentration was elevated (P < 0.05) in the YL-C2 with the Gut2 gene knocked out. The improved ATP status indicates that the interruption of the Gut2 gene might have diverted more G3P to the glycolytic pathway, which produced more ATP. Compared to YL-C2, both ATP and NADPH concentrations were lower (P < 0.05) in the YL-C31. This result indicates that the overexpression of the Did2 gene led to a decrease in ATP and NADPH. The lower concentration might be attributed to that the ATP and NADPH were consumed for the synthesis of β-carotene.
Overexpression of the Did2 gene improved the mRNA level of the genes in the β-carotene synthesis pathway
We measured the relative mRNA level of Thmg, Ggs1, CarRA, and CarB, key genes in the β-carotene synthesis pathway, to explore the reason that the overexpression of the Did2 gene improved the production of β-carotene in Y. lipolytica. The actin gene was used as an internal reference [10]. The mRNA of key genes in YL-C1 was set as 1. For Thmg, the mRNA in YL-C31 was increased (P < 0.05) by 17% compared to YL-C1, increased (P < 0.05) by 30% compared to YL-C2 (Fig. 3a). For Ggs1, the mRNA in YL-C31 was increased (P < 0.05) by 45% compared to YL-C1, increased (P < 0.05) by 35% compared to YL-C2 (Fig. 3b). For CarRA, the mRNA in YL-C31 was increased (P < 0.05) by 78% compared to YL-C1, increased (P < 0.05) by 97% compared to YL-C2 (Fig. 3c). For CarB, the mRNA in YL-C31 was increased (P < 0.05) by 55% compared to YL-C1, increased (P < 0.05) by 91% compared to YL-C2 (Fig. 3d). So, the mRNA of Thmg, Ggs1, CarRA, and CarB genes were all higher (P < 0.05) in the YL-C31 strain with the overexpressed Did2 gene.
Overexpression of the Did2 gene improved the utilization of precursors in the β-carotene synthesis pathway
The utilization of the precursors is directly linked to β-carotene synthesis. The sesquiterpenes, diterpenes, triterpenes, and tetraterpenes all compete with β-carotene for the precursor FPP [19]. The lycopene is the direct precursor of β-carotene. To ascertain the effect of overexpression of the Did2 gene on the utilization of precursors, we measured the concentration of FPP and lycopene. The FPP and lycopene levels in YL-C1 were considered as 1. For FPP, the utilization of FPP in YL-C31 was increased (P < 0.05) by 34% compared to YL-C1, increased (P < 0.05) by 37% compared to YL-C2 (Fig. 4a). For lycopene, the utilization in YL-C31 was increased (P < 0.05) by 8.1% compared to YL-C1, increased (P < 0.05) by 7.9% compared to YL-C2 (Fig. 4b). The utilization of both FPP and lycopene in YL-C31 was highest (P < 0.05) among strains YL-C1, YL-C2, and YL-C31.
Overexpression of the Did2 gene increased the protein level of the key enzymes in the β-carotene synthesis pathway
We performed the Western blot assay to measure whether the amount of the key enzymes in the β-carotene synthesis pathway was affected by the overexpression of the Did2 gene. A fusion StrepII tag was chosen for co-expression of tHMG, carRA, and carB proteins. The fusion StrepII tag has been successfully used for analyzing the expression of carotenoid synthesis enzymes in Rb. Sphaeroides [10]. Plasmids pJN44-tHMG-s, pJN44-carRA-s, pJN44-carB-s were separately transformed into YL-C1, YL-C2, and YL-C31, respectively, resulted in YL-C1ts, YL-C2ts, YL-C31ts, YL-C1as, YL-C2as, YL-C31as, YL-C1bs, YL-C2bs, and YL-C31bs. tHMG-strepII (Fig. 5a) and carB-strepII (Fig. 5b) are identified by Western blotting. The protein bands were scanned. The expression amounts of the key enzymes in YL-C1ts, YL-C1bs were regarded as 1. For tHMG-strepII, the protein level in YL-C31ts was increased (P < 0.05) by 37% compared to YL-C1ts, increased (P < 0.05) by 92% compared to YL-C2ts (Fig. 5c). For carB-strepII, the protein level in YL-C31bs was increased (P < 0.05) by 17% compared to YL-C1ts, increased (P < 0.05) by 25% compared to YL-C2ts (Fig. 5d). tHMG-strepII protein has the highest amount when the pJN44-tHMG-s was expressed in YL-C31. carB-strepII protein has the highest amount when the pJN44-carB-s was expressed in YL-C31. These results demonstrate that the overexpression of the Did2 gene increased (P < 0.05) the level of key enzymes (tHMG-strepII, carB-strepII). For carRA-strepII, the exact bands cannot be identified by Western blotting (As shown in Fig. S1). The reason may be that the stability of carRA-strepII is weak, and the protein was degraded during the extraction.
Overexpression of the Did2 gene increased the transcription level of the Vps4 gene in the MVB sorting pathway
The Did2 gene is a positive regulator of the MVB sorting pathway [20]. The Did2 protein, one subunit of the ESCRT protein complex, recruits Vps4 protein to bind ESCRT. Meanwhile, the Vps4 protein is a core factor of the MVB sorting pathway [21]. To explore the effect of the overexpression of the Did2 gene on the MVB sorting pathway, we measured the mRNA levels of the Vps4 gene and the Did2 gene. The mRNA level of the Did2 gene increased (P < 0.05) by 43% compared to YL-C1, and increased (P < 0.05) by 68% compared to YL-C2 (Fig. 6a). The mRNA level of the Vps4 gene increased (P < 0.05) by 28% compared to YL-C1, and increased (P < 0.05) by 47% compared to YL-C2 (Fig. 6b). These results demonstrate that the mRNA level of the Vps4 gene was increased (P < 0.05) by the overexpression of the Did2 gene.
Overexpression of two copies of the Did2 gene further stimulated the MVB sorting pathway but reduced the protein level of key enzymes in the β-carotene synthesis pathway
To further explore the effect of the MVB sorting pathway on the protein level of key enzymes in the β-carotene synthesis pathway, the MVB sorting pathway was promoted by overexpressing two copies of the Did2 gene. Two copies of the Did2 gene were integrated into the engineered Y. lipolytica strain (YL-C1) genome at the Gut2, resulted in YL-C32. The mRNA level of the Vps4 and Did2 genes in YL-C32 was further elevated (P < 0.05) by 50% and 33% compared to YL-C31 (Table 1), respectively. Meanwhile, compared to the overexpression of one copy of Did2 gene, the mRNA level of the Thmg, Ggs1, CarRA, and CarB, the key genes in β-carotene synthesis pathway, increased (P < 0.05) by 63%, 57%, 54%, and 85% (Table 1), respectively. However, the protein level of tHMG-strepII in YL-C32ts was reduced compared to YL-C31ts (Fig. 7). For carB-strepII, the protein level in YL-C32bs was also reduced compared to YL-C31bs (Fig. 7). These results indicate that the protein level of key enzymes (tHMG-strepII, carB-strepII) in the β-carotene synthesis pathway was lower in the YL-C32 strain with overexpressing two copies of the Did2 gene. Furthermore, the β-carotene content was reduced by 25% (P < 0.05) (Table 1).