A novel autolysis system controlled by magnesium and its application to poly (3-hydroxypropionate) production in engineered Escherichia coli

ABSTRACT

The release of intracellular products, especially polyhydroxyalkanoates, is still a great challenge in industry. To solve this bottleneck, a novel autolysis system strictly controlled with magnesium was constructed and applied to poly(3-hydroxypropionate) production in engineered Escherichia coli. The autolysis system was constructed by inserting the 5′untranslated region (5′UTR) behind promoter PmgtA with lysis genes (S, R, and Rz, from E. coli) overexpressed. The autolysis system functioned well (lysis efficiency of more than 90%) in the P3HP producer with double plasmids containing lysis genes and P3HP biosynthesis genes, whereas the P3HP production was reduced due to plasmid losses. After the autolysis genes and P3HP biosynthesis genes were integrated into one plasmid, the P3HP content of 72.7% (2.4 times of the control) and the plasmid stability of 79.8 ± 3.1% were achieved in strain Q2646 with promoter PmgtA-UTR. However, the strain Q2647 with promoter PmgtA could not accumulate P3HP because of rapid cell lysis. The novel autolysis system activated in Mg²⁺-depleted conditions proves to be feasible for polyhydroxyalkanoates production, which may have great application potential for other intracellular products.

Introduction

The release of intracellular products from microbial cells is an essential process for industrial production of bio-based chemicals. Organic solvent, alkali or detergent and ultrasonication are widely used for cell disruption and intracellular substances extraction in industry.¹ However, these extraction processes usually account for as high as 70∼80% of the total production cost ² and the cell disruption is a significant cost factor in industry. To develop a cost-effective recovery process for the intracellular products, many attempts have previously been made to disrupt the cells by constructing autolysis systems. Liu et al. developed an auto-inducible lysis system in E. coli using the lysis genes from Salmonella phage P22 and promoter PnrsB, which activated by addition of nickel.³ Zhang et al. constructed an autolysis E. coli system regulated by Mg²⁺ concentration, which consisted of the lysis genes from λ bacteriophage and promoter Pmgt from Salmonella typhimurium.⁴ Hajnal et al. designed a synthetic ribosome binding site for the autolysis of E. coli and Halomonas campaniensis under environmental stresses.⁵

Compared with the autolysis systems induced by external chemical or physical stimuli such as IPTG, L-arabinose, nickel and heat shock, the autolysis system controlled by magnesium depletion is more feasible in microbial production due to zero-added chemicals and no additional energy consumption. However, the previous study on the auto-inducible lysis systems controlled by magnesium suggested that the lysis genes (S, R, and Rz) could still be activated at Mg²⁺ concentration of 50 μM or so.⁴ Because the lysis proteins can degrade the peptidoglycan and cause cell wall disruption,^6,7 the lysis genes must be strictly regulated, otherwise even low level of protein expression can seriously influence the cell growth and the biosynthesis of intracellular products. Therefore, the regulation of the lysis genes in the present Mg²⁺- controlled system should be further improved.

P3HP, a novel polyhydroxyalkanoates, is regarded as a promising alternative to the fossil fuel-based materials. The microbial production of P3HP has been achieved by genetically engineered strains so far.^8,9 However, few efforts have been devoted to the application of the autolysis system to P3HP fermentation. To overcome the present problems, we designed and constructed a new autolysis system by introducing the mgtA-UTR as a Mg²⁺-responsive riboswitch. Furthermore, the effects of this novel autolysis system on the production of P3HP were investigated in the present paper.

Results and Discussion

Construction of optimized autolysis E. coli system

To verify the function of the lysis genes, the strain Q2537 was constructed by transforming the pY01 into E. coli DH5α, which contained the lysis genes (S, R, and Rz) controlled by L-arabinose promoter P_BAD. The bacterial-growth curves of Q2537 and E. coli DH5α containing empty plasmid pBAD18 were shown in Fig. 1. After the expression of lysis genes were induced by L-arabinose, the OD₆₀₀ of Q2537 fell sharply from 0.61 to 0.19 within 1 h. Meanwhile, the cell growth of uninduced Q2537 showed the same increasing trend as the control strain E. coli DH5α. The results suggested that the lysis genes were efficient on the cell disruption.

Figure 1.

Effects of lysis genes expression on the growth of E. coli. E. coli DH5α with empty plasmid pBAD18-kan and Q2537 with lysis genes were used as the control strains. Q2537 with lysis genes was induced with L-arabinose at 0.6∼0.8 OD₆₀₀.

The effects of the lysis systems controlled by P_BAD, PmgtA and PmgtA -UTR in the strains Q2537, Q2549, and Q2582 were further analyzed. The corresponding lysis efficiencies were 91.5 ± 1.0%, 96.5 ± 1.5%, and 97.0 ± 1.0%, respectively (Fig. 2). The results showed that both promoter PmgtA and PmgtA -UTR could successfully regulate the expression of lysis genes under Mg²⁺ limited condition. The mgtA 5′-UTR functions as a Mg²⁺-responsive riboswitch in Salmonella and E. coli.^10,11 This mgtA riboswitch can cause premature termination of the mgtA transcription at high external Mg²⁺ levels.^10,12 When the Mg²⁺ concentration reduced to μM level,¹³ the mgtA transcription reactivated and caused cell lysis. Thus, the tight control of the lysis genes can be achieved by inserting the 5′-UTR behind promoter mgtA.

Figure 2.

The Alysis efficiencies of different lysis systems controlled by P_BAD, PmgtA and PmgtA -UTR in the strains Q2537, Q2549, and Q2582.

The lysis system was also integrated into the Escherichia coli chromosome. Unexpectedly, the lysis efficiency of the resulting strain was just 10.1 ± 1.2%. It was speculated that PmgtA was too weak to induce the expression of the lysis genes after integrated into the chromosome. So, promotors P_lac1−6 and P_T7 were also tested. However, the lysis efficiencies were still quite low as 12 ± 0.7% and 18 ± 1.8%. The results suggested that the expression of the lysis genes was not enough to disrupt the cells in a short time after the lysis genes were integrated into the chromosome.

The expression of the lysis genes (S, R, and Rz) in E. coli is lethal due to the degradation of the peptidoglycan of the cell wall. Even low level of lysis genes expression is likely to influence the biosynthesis of intracellular products. Therefore the effects of autolysis systems with PmgtA and PmgtA -UTR on P3HP production should be further investigated.

P3HP production by E. coli with double plasmids containing autolysis genes and P3HP biosynthesis genes

To test the influence of the lysis systems on P3HP accumulation, the strains Q2588 (containing plasmids pY02 and pWQ02) and Q2572 (containing plasmids pY03 and pWQ02) were used for P3HP production in shake flask fermentation. As shown in Table 1, P3HP production of Q2572 was decreased in contrast with the control strain Q1638. Both the P3HP production and content of strain Q2588 were much lower than the control. Furthermore, the lysis efficiencies of the strains Q2588 and Q2572 were also tested. The lysis efficiencies of higher than 90% were achieved through the double plasmids system containing autolysis genes and P3HP biosynthesis genes. Therefore, the autolysis system functioned well in the P3HP producer with double plasmids system, whereas the P3HP production was reduced.

Table 1.

P3HP production, contents and plasmids stability of different strains with double plasmids.

				Plasmids stability (%)
Strains	CDW(g/L)	P3HP (g/L)	P3HP content (%)	pWQ02	pY02	pY03
Q1638	6.6 ± 0.11	2.1 ± 0.01	31.8	83.1 ± 3.2	—	—
Q2572	5.4 ± 0.16	1.7 ± 0.03	31.4	30.3 ± 0.5	—	20.7 ± 1.2
Q2588	5.2 ± 0.21	1.4 ± 0.04	26.9	40.0 ± 2.2	17.1 ± 3.2	—

(Q1638 without lysis system was used as the control. P3HP content was calculated using the ratio of P3HP weight to cell dry weight.)

Introduction of exogenous plasmid can increase metabolic burden and lead to plasmid losses during fermentation. To study whether this potential cause existed in this system, plasmid stability was tested. Comparing with Q1638, the stabilities of the plasmid pWQ02 in Q2572 and Q2588 were reduced dramatically from 83.1% to 30.3 ± 0.5% and 40.0 ± 2.2%. And the stabilities of the plasmids pY02 and pY03 were just 17.1 ± 3.2% and 20.7 ± 1.2%. Therefore, the strains containing 2 plasmids system were not stable enough and needed to be improved for the P3HP production.

P3HP production by E. coli with single plasmid containing autolysis genes and P3HP biosynthesis genes

Due to the large losses of the plasmids during the fermentation with the double plasmids strains, the autolysis systems containing PmgtA and PmgtA-UTR were respectively integrated into the plasmid pWQ02 originally containing the P3HP biosynthesis genes (Fig. 3). The resulting plasmids pY04 and pY05 were separately transformed into the strain Q1463¹⁴ to generate Q2647 and Q2646. The resulting strains containing single plasmid were tested for P3HP accumulation.

Figure 3.

The backbones of plasmids constructed in this study.

The P3HP production of 1.6 ± 0.01 g P3HP l⁻¹ was achieved in the recombinant strain Q2646 with the P3HP content of 72.7%, which was 2.4 times of the control (Table 2). Furthermore, the plasmid stability was 79.8 ± 3.1%, which is similar to the control. However, the P3HP did not accumulate in the recombinant strain Q2647 during the fermentation (Table 2) and the cell lysis happened at early stage. The results showed that strictly controlled autolysis system was obtained in strain Q2646 with promoter PmgtA -UTR, while the autolysis system controlled by promoter PmgtA in Q2647 relatively expressed too early to synthesize the P3HP under the fermentation conditions.

Table 2.

P3HP production, contents and plasmids stability of different strains with single plasmid.

Strains	CDW(g/L)	P3HP (g/L)	P3HP content (%)	Plasmids stability (%)
Q1638	6.6 ± 0.51	2.0 ± 0.02	30.3	84.1 ± 2
Q2646	2.2 ± 0.04	1.6 ± 0.01	72.7	79.8 ± 3.1
Q2647	1.8 ± 0.02	0	0	8.4 ± 0.9

(Q1638 without lysis system was used as the control. P3HP content was calculated using the ratio of P3HP weight to cell dry weight.)

Conclusions

A novel autolysis system strictly controlled by magnesium was constructed by inserting the 5′-UTR behind promoter PmgtA. Products like P3HP can accumulated continuously in the cells and released automatically by this autolysis system. The P3HP content reached 72.7% in engineered E. coli with promoter PmgtA-UTR, which is 2.4 times of the control strain. The novel autolysis system activated in Mg²⁺-depleted conditions proves to be feasible for polyhydroxyalkanoates production, which may have great application potential for other intracellular products.

Materials and methods

Bacterial strains and growth conditions

E. coli DH5α was used for plasmid cloning and maintenance. Luria-Bertani medium (LB medium), and N-minimal medium (M9 medium) (per liter: 15.2 g Na₂HPO₄·12H₂O, 3.0 g KH₂PO₄·3H₂O, 0.5 g NaCl, 1.0 g NH₄Cl, 20 g glucose, and 1 ml trace elements stock solution (per liter: 3.7 g (NH₄)₆Mo₇O₂₄·4H₂O, 24.7 g H₃BO₃, 15.8 g MnCl₂·4H₂O, 2.9 g ZnSO₄·7H₂O, 2.5 g CuSO₄·5H₂O), pH 7.0) were used for bacterial culture and verification of lysis efficiency. For P3HP fermentation, a modified minimal medium containing the following components (per liter): 1.5 g KH₂PO₄, 3 g (NH₄)₂SO₄, 1.9 g KCl, 1.09 g citric acid, 1.14 g sodium citrate, 0.138 g FeSO₄·7H₂O, 3 g glucose, 20 g glycerol, and 1ml trace elements solution was used. 50 μg/ml kanamycin or 100 μg/ml ampicillin was added to keep the recombinant plasmid when necessary. The strains were cultivated at 37°C with 180 rpm unless specific.

Construction of recombinant plasmids and strains

The lysis genes S, R, and Rz were amplified from the genome of E.coli BL21 (DE3) by PCR with the primers 5′-CGAGCTCAAGGAGATATAATGCCAGAAAAACATGACCT-3′ and 5′-AAAACTGCAGCTATCTGCACTGCTCATTAAT-3′, then cloned into pBAD18 vector between SacI and PstI sites to get plasmid pY01. The single PmgtA gene was amplified with primers 5′-CGAGCTCCTTTCGTTATTCAGCACCCG-3′ and 5′-CGAGCTCGCGATATAATACCTGCTGGC-3′, the PmgtA-UTR was amplified with primers 5′-CGAGCTCCTTTCGTTATTCAGCACCCG-3′ and 5′-CGAGCTCAA GGAGTCCCTCCGCACTGT-3′, then cloned into SacI site of pY01 to obtain pY02 and pY03, respectively. The fragment PmgtA -SRRz was amplified from pY02 and PmgtA-UTR -SRRz was amplified from pY03, with the same primers 5′- CCGCTCG AGCTTTCGTTATTCAGCACCCG-3′ and 5′- CCCAAGCTTAAAAGCCTCCGGT CGGAGGCTTTTCTATCTGCACTGCTCATTAAT-3′, then cloned into XhoI and HindIII sites of pWQ02¹⁵ to get pY04 and pY05, respectively. The backbone of the plasmids constructed in this study was shown in Fig. 3. All the plasmids were verified by PCR and gene sequencing.

The plasmids pY01, pY02, and pY03 were transformed into E. coli DH5α to generate the strains Q2537, Q2549, and Q2582. The pY02 and pY03 were also transformed into Q1638 to obtain the P3HP producing strains Q2588 and Q2572, respectively. The plasmids pY04 and pY05 were transformed into the strain Q1463 to get the P3HP producing strains Q2647 and Q2646, respectively. Q1463 and Q1638 were constructed in our previous study.¹⁴ All the recombinant strains and plasmids are listed in Table 3.

Table 3.

Bacterial strains and plasmids used in this study.

Strains and plasmids	Relevant properties	Source or reference
Strains
Escherichia coli DH5α	F−,φ80dlacZ△M15,△(lacZYA-argF)U169, deoR, recA1, endA1, hsdR17(rk-,mk+)	Invitrogen
Escherichia coli BL21(DE3)	F−, ompT, hsdSB (rB−mB−), ga, dcm, λ(DE3)	Invitrogen
Q1463	△prpR::lacI PT7gdrAB PT7dhaB123	Ref. 14
Q1638	Q1463 carrying pWQ02	Ref. 14
Q2537	Escherichia coli DH5α carrying pY01	This study
Q2549	Escherichia coli DH5α carrying pY02	This study
Q2572	Q1638 carrying pY03	This study
Q2582	Escherichia coli DH5α carrying pY03	This study
Q2588	Q1638 carrying pY02	This study
Q2646	Q1463 carrying pY05	This study
Q2647	Q1463 carrying pY04	This study
Plasmids
pBAD18	KmR, cloning vector, pBR322 origin, araBAD promotor	NBRP-E. coli at NIG
pET21a	AmpR, pBR322 origin, lacI PT7	Novagen
pACYCDuet-1	CmR, p15A origin, lacI PT7	Novagen
pWQ02	AmpR, pET21a carrying polyhydroxyalkanoate synthase phaC and propionaldehyde dehydrogenase pduP	Ref. 15
pWQ04	CmR, pACYCDuet-1 carrying glycerol dehydratase dhaB123 and its reactivating factor gdrAB	Ref. 15
pY01	KmR, pBAD18 carrying lysis genes S, R, and Rz	This study
pY02	KmR, pBAD18 carrying lysis genes S, R, and Rz and promoter PmgtA	This study
pY03	KmR, pBAD18 carrying lysis genes S, R, and Rz, and promoter PmgtA-UTR	This study
pY04	AmpR, pET21a carrying phaC, pduP, lysis genes S, R, and Rz, and promoter PmgtA	This study
pY05	AmpR, pET21a carrying phaC, pduP, lysis genes S, R, and Rz, and promoter PmgtA-UTR	This study

Detection of the lysis efficiency

Single colony of each strain was cultivated overnight in LB medium to get seed solution, then inoculated at 10% into M9 medium, 10 mM MgSO₄ was added to prevent earlier cell lysis. The control strains were E. coli DH5α with empty vector pBAD18 and Q2537 without induction. To activate promotor P_BAD, 0.1% L-arabinose was added into Q2537 broth cells culture at OD₆₀₀ 0.6∼0.8 under 30°C. To evaluate the activity of PmgtA, the cell growth (OD₆₀₀) was monitored at different time intervals.

The lysis efficiency was detected by plate colony-counting methods. 1 ml of the cell-culture LB medium was centrifuged and washed twice with sterile water, then re-suspended with M9 medium without Mg²⁺. The cell suspensions were separated equally into 2 tubes. One tube, as control group, was supplemented with 10 mM MgSO₄ (L- arabinose instead of MgSO₄ was added for Q2537). After 1 h of incubation at 37°C, the cells suspension was diluted 10⁻⁶ times and spreaded on the LB agar plates with or without Mg²⁺ respectively. The monoclonal cells on each plate were counted after cultured overnight. The lysis efficiency was analyzed by comparing the colony-forming units (CFU) on Mg²⁺-depletion LB agar plates and CFU on LB agar plates containing Mg²⁺. The lysis efficiencies of other strains were tested using the same method as described above.

Flask fermentation for P3HP production

Shake flask fermentation was performed in a 500mL baffled-flask containing 100 mL modified minimal medium at 37°C shaking at 200 rpm, the initial concentration of MgSO₄ in the medium was 10mM. Gene expression of cultures were induced with 0.05mM isopropyl-β-d-thiogalactoside at OD₆₀₀ 0.6∼0.8. After induction, the cultures were further incubated at 30°C for 48 h. The cells were harvested by centrifugation and washed with cool absolute ethanol, then dried in an oven at 80°C. P3HP was extracted from the cells with hot chloroform in a Soxhlet apparatus, the content was calculated as described in our pervious paper.¹⁴

Funding

This work was supported by100-Talent Project of CAS (to GZ), Key Program of CAS (ZDRW-ZS-2016–3M and KGZD-EW-606–1–3), National Natural Science Foundation of China (31670089, 31670493), Taishan Scholars Climbing Program of Shandong (No.tspd20150210), and Shandong Province Postdoctoral Innovation Foundation (201603076).