Different D-xylose consumption rates in E. coli strains under anaerobic conditions
E. coli BL21(DE3), BW25113, C strain, and MG1655 strains were anaerobically grown in a fermentation medium supplemented with D-glucose (12.5 mM) and D-xylose (12.5 mM). All four strains consumed D-glucose and appeared to fully consume the D-glucose within 4–6 h (Figure 1). Cells began to uptake D-xylose after depleting the D-glucose; however, the D-xylose consumption rate varied in a strain-dependent manner. BW25113 and C strain fully depleted the D-xylose 4 h after consumption of D-glucose. After depleting the D-glucose, it took 10 and 36 hours for the MG1655 and BL21(DE3) strains to consume the D-xylose, respectively. The BL21(DE3) strain exhibited the longest D-xylose consumption delay after D-glucose depletion. Moreover, even when the cells were cultured in D-xylose-only media, the D-xylose consumption rate of the BL21(DE3) strain was slower than that of the other strains (Figure S1).
Accelerated Anaerobic Growth Of Bl21(De3) Strains Through Adaptive Evolution
Our study sought to obtain BL21(DE3) cells with an increased D-xylose consumption rate through adaptive evolution by serially transferring cultures to fresh fermentation media. BL21(DE3) cells were anaerobically grown in a fermentation medium containing D-glucose (12.5 mM) and D-xylose (12.5 mM). The growth rate and D-glucose consumption rate of the bacteria did not change significantly after several transfers; however, the maximum D-xylose consumption rate increased gradually. The D-xylose consumption rate was 0.6 mM/h in the first passage, but gradually increased to 0.8, 1.3, and 1.4 mM/h in each subsequent passage (Figure 2). Cultures of the 1st and 4th passages were spread on LB agar plates to isolate pure colonies. The progeny strains JH001 and JH019 were obtained after anaerobically culturing the newly obtained colonies in the same medium. Notably, these progeny strains consumed D-xylose faster after full D- glucose depletion compared to the parent strain.
The maximum D-xylose consumption rate of the wild-type BL21(DE3) strain was 1.1 mM/h in D-glucose- and D-xylose-supplemented anaerobic media. In contrast, the D-xylose consumption rates of the JH001 and JH019 rates were 1.9 mM/h and 2.9 mM/h, which represented 1.7- and 2.7-fold increases compared to the BL21(DE3) strain (Table 1). Moreover, the JH001 strain exhibited an increased D-xylose consumption rate but its cell growth was not significantly increased (Figure 3C). However, strain JH019 showed increased cell growth (Figure 3E).
In the medium containing D-xylose only, the adaptively evolved strains JH001 and JH019 exhibited a faster D-xylose consumption compared to BL21(DE3) (Figure 3). Concretely, the BL21(DE3) strain had a maximum D-xylose consumption rate of 1.98 mM/h, whereas the JH001 strain exhibited an increased rate of 3.69 mM/h. Moreover, when the JH001 strain was cultured in D-xylose-supplemented medium, the D-xylose was consumed between 4 and 10 hours, but cell growth was considerably slower (Figure 3D). In contrast, the maximum D-xylose consumption rate of the JH019 strain increased to 7.36 mM/h and there were no cell growth delays.
Table 1
Specific growth rate and xylose consumption rate, and fermentation profiles of wild-type BL21(DE3) strain and adapted strains with D-glucose plus D-xylose, and D-xylose.
Added sugar
(mM)
|
Strain
|
Fermentation
time (h)†
|
Specific growth rate (µ) (fold)
|
Maximum xylose consumption rate (mM/h) (fold)
|
Metabolites (mM)
|
Acetate
|
Ethanol
|
Formate
|
Lactate
|
Succinate
|
Glucose (12.5)
+Xylose (12.5)
|
BL21(DE3)
|
42
|
0.88 ± 0.02 (1.0)
|
1.1 ± 0.0 (1.0)
|
22.7 ± 0.1
|
3.2 ± 0.4
|
33.6 ± 0.1
|
0.2 ± 0.0
|
12.2 ± 0.2
|
JH001
|
12
|
0.85 ± 0.03 (1.0)
|
1.9 ± 0.2 (1.7)
|
22.0 ± 0.9
|
15.2 ± 0.6
|
38.9 ± 1.8
|
0.2 ± 0.0
|
5.3 ± 0.1
|
JH019
|
12
|
0.91 ± 0.07 (1.0)
|
2.9 ± 0.6 (2.7)
|
22.8 ± 0.8
|
14.4 ± 0.9
|
39.4 ± 1.5
|
0.2 ± 0.0
|
4.8 ± 0.3
|
Xylose (25)
|
BL21(DE3)
|
21
|
0.59 ± 0.04 (1.0)
|
2.0 ± 0.0 (1.0)
|
19.4 ± 0.3
|
7.4 ± 0.1
|
23.4 ± 0.4
|
ND*
|
17.9 ± 0.1
|
JH001
|
16
|
0.63 ± 0.03 (1.1)
|
3.7 ± 0.6 (1.9)
|
22.8 ± 0.2
|
15.1 ± 0.2
|
41.5 ± 0.4
|
ND
|
4.3 ± 0.1
|
JH019
|
10
|
0.60 ± 0.02 (1.0)
|
7.4 ± 0.6 (3.7)
|
21.6 ± 0.1
|
13.7 ± 0.3
|
38.2 ± 0.2
|
ND
|
4.3 ± 0.1
|
†Fermentation time (h) when glucose plus xylose or xylose were completely consumed. |
*ND, Not detected. |
Variation In Fermentation Products In Adapted Bl21(De3) Cells
The difference between the wild-type and adaptively evolved strains was confirmed based on their organic acid and ethanol output during fermentation. When provided with both D-glucose and D-xylose, there was no significant difference in the amount of acetate, formate, and lactate produced by the bacterial strains. However, while the wild-type strain produced 3.2 mM of ethanol, the adaptively evolved strains produced 15.2 and 14.4 mM. Conversely, the JH001 and JH019 strains produced 5.3 and 4.8 mM of succinate, respectively, whereas the wild-type strain produced 12.2 mM.
When provided with D-xylose only, neither the wild-type nor the adaptively evolved strains produced lactate, and acetate production was not significantly different. Moreover, similar to the D-glucose + D-xylose condition, ethanol production was further increased and succinate decreased in the adaptively evolved JH001 and JH019 strains (Table 1).
Identification of adaptive mutations in the evolved strains via genome sequencing
Whole-genome sequencing analysis of the adaptively evolved strains with increased D-xylose consumption identified a C91A point mutation (Q31K missense in XylR protein) in the xylR gene of JH001 strain, as well as a C740T substitution (A247V) in the xylR gene and IS (insertion sequence) insertion in the open reading frame of the carB gene of the JH019 strain (Table 2). Given that D-xylose cannot be consumed in a xylR null mutation background (Figure S3), we assumed that the xylR point mutations represented a gain of function mutation responsible for faster D-xylose uptake in the adaptively evolved strains. Since xylR encodes a transcriptional activator, the expression of the xylose operon was also investigated (see below).
Table 2
Genomic analysis of E. coli BL21(DE3) adapted strains.
Strain
|
Genotype
|
Reads
|
Bases
|
Reads (trimmed)
|
Bases (trimmed)
|
Avg. length (trimmed)
|
Reads matched
|
% Reads matched
|
Fraction of
reference covered
|
Avg. coverage
|
JH001
|
xylR C91A (Q31K)
|
24,611,910
|
6,177,589,410
|
21,909,198
|
3,353,116,156
|
153
|
19,789,604
|
90
|
1
|
664.35
|
JH019
|
xylR C740T (A247V), carB::IS1
|
17,170,757
|
3,823,379,807
|
14,945,996
|
2,196,575,011
|
147
|
14,674,322
|
98
|
1
|
473.06
|
Transcript analysis of the xylose operon in adaptively evolved BL21(DE3) strains
qRT-PCR was conducted to confirm whether the expression of the xylose operon was enhanced in adapted cells carrying xylR point mutations. The expression levels of the xylA and xylF genes (which encode xylose isomerase and xylose ABC transporter, respectively) were compared between the wild-type and adaptively-evolved BL21(DE3) strains grown in a fermentation medium containing both D-glucose (12.5 mM) and D-xylose (12.5 mM). Compared to BL21(DE3), the expression of the xylA and xylF genes in the JH001 strain were upregulated 11- and 3-fold, respectively. Similarly, in the case of the JH019 strain, the expression levels of the xylA and xylF genes increased 5- and 2-fold compared to the wild-type strain, respectively (Figure 4A). When each strain was grown in fermentation media containing only D-xylose (25 mM), the JH001 and JH019 cells exhibited significantly elevated transcript levels of the xylA and xylF genes, which were at least 5 times higher than those in the wild type BL21(DE3) strain (Figure 4B). These results suggest that D-xylose transporting and metabolizing enzymes are highly expressed in adapted BL21(DE3) cells carrying xylR adaptive mutations.
Increased D-xylose consumption rate by xylR single point genome editing
The single point mutation (C91A) of the xylR gene identified in the adaptively evolved JH001 strain was introduced into the genome of the BL21(DE3) and MG1655 wild-type strains via the target-mismatched CRISPR/Cas9 method [30]. When BL21(DE3) XylRQ31K cells were grown in fermentation media containing both D-glucose and D-xylose, D-xylose was completely consumed 2 h after D-glucose depletion (Figure 5A). In xylose-only fermentation medium, it took 21 h for the BL21(DE3) strain to fully deplete the D-xylose, whereas BL21(DE3) XylRQ31K cells completely consumed the D-xylose in 6 h (Figure S1A, Figure 5B). Similar results were also observed in MG1655 wild-type and MG1655 XylRQ31K cells (Figure 5C and D). These results indicate that the XylRQ31K mutation is responsible for the faster growth of the evolved BL21(DE3) cells in the fermentation medium through enhanced D-xylose transport and metabolism.
Additionally, to confirm whether the carB gene inactivation also identified in the JH019 strain affects the D-xylose consumption rate, a carB deletion mutation was introduced into the BL21(DE3) wild-type strain and the adaptively evolved strain JH001. D-xylose consumption rates were accelerated when the carB gene deletion was introduced into the BL21(DE3) cells (Figure S4A and S4B). The carB deletion mutation improved the D-xylose consumption rate of the JH001 strain when the cells were grown in a fermentation medium containing only D-xylose (Figure S4D). These results indicated that the carB mutation enhanced anaerobic cell growth in the D-xylose medium.