Metabolic stress constrains microbial L-cysteine production in Escherichia coli by accelerating transposition through mobile genetic elements

Heieck, Kevin; Arnold, Nathanael David; Brück, Thomas Bartholomäus

doi:10.1186/s12934-023-02021-5

Metabolic stress constrains microbial L-cysteine production in Escherichia coli by accelerating transposition through mobile genetic elements

Research
Open access
Published: 16 January 2023

Volume 22, article number 10, (2023)
Cite this article

Download PDF

You have full access to this open access article

Microbial Cell Factories Aims and scope Submit manuscript

Metabolic stress constrains microbial L-cysteine production in Escherichia coli by accelerating transposition through mobile genetic elements

Download PDF

Kevin Heieck¹,
Nathanael David Arnold¹ &
Thomas Bartholomäus Brück¹

4161 Accesses
8 Citations
1 Altmetric
Explore all metrics

Abstract

Background

L-cysteine is an essential chemical building block in the pharmaceutical-, cosmetic-, food and agricultural sector. Conventionally, L-cysteine production relies on the conversion of keratinous biomass mediated by hydrochloric acid. Today, fermentative production based on recombinant E. coli, where L-cysteine production is streamlined and facilitated by synthetic plasmid constructs, is an alternative process at industrial scale. However, metabolic stress and the resulting production escape mechanisms in evolving populations are severely limiting factors during industrial biomanufacturing. We emulate high generation numbers typically reached in industrial fermentation processes with Escherichia coli harbouring L-cysteine production plasmid constructs. So far no genotypic and phenotypic alterations in early and late L-cysteine producing E. coli populations have been studied.

Results

In a comparative experimental design, the E. coli K12 production strain W3110 and the reduced genome strain MDS42, almost free of insertion sequences, were used as hosts. Data indicates that W3110 populations acquire growth fitness at the expense of L-cysteine productivity within 60 generations, while production in MDS42 populations remains stable. For the first time, the negative impact of predominantly insertion sequence family 3 and 5 transposases on L-cysteine production is reported, by combining differential transcriptome analysis with NGS based deep plasmid sequencing. Furthermore, metabolic clustering of differentially expressed genes supports the hypothesis, that metabolic stress induces rapid propagation of plasmid rearrangements, leading to reduced L-cysteine yields in evolving populations over industrial fermentation time scales.

Conclusion

The results of this study implicate how selective deletion of insertion sequence families could be a new route for improving industrial L-cysteine or even general amino acid production using recombinant E. coli hosts. Instead of using minimal genome strains, a selective deletion of certain IS families could offer the benefits of adaptive laboratory evolution (ALE) while maintaining enhanced L-cysteine production stability.

Glucose consumption rate-dependent transcriptome profiling of Escherichia coli provides insight on performance as microbial factories

Article Open access 14 September 2022

Genome reduction boosts heterologous gene expression in Pseudomonas putida

Article Open access 21 February 2015

Omics-guided bacterial engineering of Escherichia coli ER2566 for recombinant protein expression

Article 21 December 2022

Background

The amino acid L-cysteine, harbouring a thiol group, provides a high redox activity in cell metabolism, plays a crucial role in protein folding, functions as a catalytic residue of several enzymes and serves as a building block of 5-L-glutamyl-L-cysteinylglycine (GSH) and as a donor compound of sulphur, which is required for the synthesis of Fe/S clusters, biotin, coenzyme A and thiamine [1, 2].

Besides the essential function in metabolism, L-cysteine is also of considerable industrial importance, with applications ranging from pharmaceutical products and cosmetics over food production to feed additives in livestock farming [3, 4].

To date, the cheapest and thereby most prevalent means of L-cysteine production involves chemical hydrolysis of—and extraction from—keratinous biomass, such as feathers, pig bristles and animal hair by means of electrolysis [5]. Up to 27 tons of hydrochloric acid are required to obtain 100 kg of a racemic mixture of cysteine from 1.000 kg raw material [5, 6]. In order to circumvent negative impacts upon the environment associated with hydrochloric waste disposal, alternative technologies such as fermentation and enzymatic conversion have been explored and rapidly gained significance since their implementation. In 2004, 12% of the globally manufactured L-cysteine global originated from fermentation [7].

The enzymatic conversion of DL-2-amino-∆2-thiazoline-4-carboxylic acid (D-ATC) to L-cysteine with Pseudomonas spp. derived enzymes is limited by product inhibition [8, 9]. For biotechnological L-cysteine production, the bacteria C. glutamicum and E. coli harbouring optimised plasmids represent the dominant expression organisms. Since titres from C. glutamicum are low (approx. 950 mg/L), E. coli is the preferred host for L-cysteine production by fermentation [10].

However, there are still major obstacles in upscaling fermentation processes with engineered microorganisms. The stability of strains with synthetic production is highly fragile and presents a challenge when implementing bioprocesses on a large scale [11, 12]. The introduction of designed plasmid constructs and the upregulation of the genetic elements for recombinant L-cysteine production pose a defiance to the tightly regulated homeostasis within host cells [13, 14]. This metabolic load hinders the expression of other genes, thereby negatively affecting growth rate and promoting evolutionary pressure [15,16,17]. Furthermore, L-cysteine has an inhibitory to toxic effect on E. coli cell growth, depending on the concentration present in cells [18]. In microorganisms, several concepts are reported that can lead to a selection advantage and thus to both phenotypic and genotypic variation within populations [19, Full size image

The distribution of the different IS families was very similar for all three production plasmids. This observation was the same for EGPs compared to LGPs (Fig. 6B). Nevertheless, the total number of reads that could be mapped to IS doubled on average when comparing W3110 EGPs with LGPs, while the number remained the same for MDS42 populations. This indicates, that transpositions of different IS families propagated in similar frequencies. For plasmids extracted from W3110 populations, we most frequently identified reads that could be mapped to the IS3 and IS5 family, which supports the result of a high expression of insJK, an IS3 family transposase, in LGPs of W3110_pCYS. In addition, reads mapped to the ISAs1 family were very abundant. Among plasmids from MDS42 populations, the reads could be mapped mainly to sequences belonging to IS200- and IS110 families. Due to deletion of most IS, the few transposition events in MDS42 populations are in line with the stable L-cysteine production and growth rates, rendering E. coli MDS42 as a stable host for industrial L-cysteine fermentations.

Conclusions

The data in this study suggest that insertion sequence (IS) transposition, triggered by the metabolic stress of L-cysteine production, leads to structural genetic rearrangements in production plasmids. Within 60 generations, as achieved in industrial large-scale fermentations, L-cysteine production capacities collapsed by up to 65–85% in E. coli W3110 populations while growth rates increased up to 27% depending on the L-cysteine plasmid construct. Contrarily, E. coli MDS42 populations exhibited nearly stable space–time yields of 75–100% with no considerable effect on growth related fitness.

Metabolic stress of L-cysteine producing populations, cultivated in shake flasks, as illustrated in this study, certainly differs from cultures grown in large-scale bioreactors, e.g. O₂ limitation as reflected in upregulation of anaerobic genes in LGPs. Yet the fundamental issue is the same: Cells are driven to eliminate factors that impair growth fitness by evolutionary adaptation. Frequently, these factors involve metabolically burdensome compounds meant to be produced by engineered E. coli cells, establishing non-producing populations in continuous cultivations [21, 49,50,51,52].

Common industrial practice employs clone banks of production strains stored as frozen aliquots. Although the population appears inconspicuous for a number of divisions, a few cells present in the starting seed may harbour genetic mutations. Cell bank aliquots with rare pre-existing mutations that disrupt production, should therefore be screened beforehand, using deep-sequencing techniques.

Opposed to previous work relying on PCR-based methods for IS identification, we chose a deep sequencing approach on population level circumventing amplification biases [53]. Thereby, an accumulation of predominantly IS3 and IS5 fragments could be detected in W3110 LGPs, whereas IS accumulation was neglectable in MDS42 populations. However, due to read length limitations of the Illumina sequencing technology, map** of IS to the corresponding plasmid reference sequences pose a hindrance. In addition, repetitive sequences within IS complicate distinguishability and read assembly. For this purpose, the Burrows-Wheeler aligner’s Smith-Waterman alignment (BWA-SW) is suitable to capture chimeric reads containing junctions between plasmid and IS sequences with frequent alignment gaps. From a future perspective, we suggest the usage of long read sequencers such as PacBio or Oxford Nanopore for a deeper understanding of IS mediated transposition into plasmid sequences. Full coverage of singular plasmid sequences would provide both the full length IS and its insertion site.

Consequently, the genetically stable minimal genome E. coli strain MDS42 rendered as a promising host for L-cysteine production. However, MDS42, while showing a superior growth rate and stability compared to W3110, exhibits an overall lower L-cysteine space–time yield (Additional file 1: Fig. S1).

This study postulates that selective deletion of IS and the corresponding transposases in genomes of industrially relevant production strains, such as W3110 can be used as a new, more-targeted approach to optimise the E. coli based industrial L-cysteine production. This could provide a trade-off between increased plasmid stability over the entire production phase while maintaining evolutionary adaptability of the cell system linked to high cell density fermentations at maximal growth rates and L-cysteine productivity.

Materials and methods

Strains and plasmids

Two parental E. coli K12 strains were used to construct the strains with three specific plasmids (Table 3):

Table 3 Plasmids used in this study: All plasmids are derived from pACYC184 with a p15A origin of replication and approx. 15 copies/cell

Full size table

E. coli W3110: F- λ- rph-1 INV(rrnD, rrnE).

E. coli MDS42 (Scarab Genomics): MG1655 genome almost free from any ISs, fhuACDB, endA and more [48].

For introduction of the plasmids into the strains, standard transformations with chemical competent cells were performed. For the long-term cultivation experiments, single colonies of freshly transformed cells were used to inoculate precultures.

PCYS_i was generated by ligation of four amplified and digested amplicons in an equimolar ratio in a 40 µl reaction. The fragments contained the metabolic pathway genes of pCYS (ydeD, cysE-XIV and serA317) and the empty vector as a backbone. The digestions were conducted with SacI, XhoI, BamHI, PacI, and NcoI. Fast digest restriction enzymes, buffer, the T4 DNA Ligase as well as the Phusion DNA polymerase were used from Thermo Fisher Scientific. For amplification of different fragments, specific PCR primer were used (Additional file 1: Table S1). PCYS_m was generated by introducing the stationary phase promoter pfic and the cysM gene into the backbone of pCYS via Gibson cloning standard procedure (NEB). PCRs were performed with specifically designed primers (Additional file 1: Table S1).

Media

A custom cultivation medium for the production of L-cysteine was used for all cultivations. This medium consisted of 10 g/L glucose, 5 g/L KH₂PO₄, 5 g/L (NH₄)₂SO₄, 1 g/L Na₃Citrat × 2 H₂O, 0.9 g/L L-isoleucine, 0.6 g/L D, L-methionine, 0.5 g/L NaCl, 2 g/L ammonium thiosulphate, 1.2 g/L MgSO₄ × 7 H₂O, 18 mg/L thiamine-HCl, 9 mg/L pyridoxine–HCl, 15 mg/L tetracycline and 100 ml/L LB-medium. Additionally, 10 ml/L of a trace element solution was added. This trace element solution included 3.75 g/L H₃BO₄, 1.55 g/L CoCl₂ × 6 H₂O, 0.55 g/L CuSO₄ × 5 H₂O, 3.55 g/L MnCl₂ × 4 H₂O, 0.65 g/L ZnSO₄ × 7 H₂O, and 0.33 g/L Na₂MoO₄ × 2 H₂O. The trace element solution is adjusted to pH 4.0 with HCl before autoclaving. Everything else got sterile filtrated after pH adjustment to 7–7.05. Tetracycline, vitamins, CaCl₂ and MgSO₄ were added immediately before inoculation.

Simulated long-term cultivation

The simulated long-term cultivation method was adapted from Rugbjerg et al. [21]. Single colonies of freshly transformed cells were inoculated into 250 ml baffled shaking flasks containing 25 ml medium. W3110 strains were cultivated at 32 °C for 10 h and horizontal shaking at 150 r.p.m (New Brunswick Innova 44). Each strain harbouring one out of three plasmids got cultivated in triplicates. Due to a higher growth rate of the MDS42 strain, cultures were kept at 32 °C just for 5 h to always maintain an exponential growth phase. After each time point cultures were inoculated into 25 ml fresh medium (starting OD600 = 0.05) and incubated under the same conditions for another 10 h or 5 h. At each passage, sample’s OD₆₀₀ were recorded to determine the accumulated generations (Additional file 1: Table S2) and 1 ml got snap-frozen with 1 ml 50% glycerol in liquid N2 and stored at − 80 °C. Subsequently the sampled cryo-cultures were re-cultivated for 72 h with a starting OD₆₀₀ of 0.01 into 25 ml fresh medium under above mentioned conditions. Growth rates were monitored every 30 min. L-cysteine yield determination was performed after 72 h. RNA extraction was carried out at OD₆₀₀ values of 0.6–0.8. Extraction of plasmid DNA was performed from the same culture (Fig. 2).

Determination of L-cysteine yields by spectrophotometry

Each population sample from the glycerol stock was used for inoculation of 25 ml medium and the culture got cultivated at 32 °C at 150 r.p.m for 72 h (New Brunswick Innova 44). After incubation, 1 ml culture was centrifuged at 15.000 × g for 1 min. Both the pellet and the supernatant were further treated according to a modified method of Gaitonde to determine the L-cysteine yields (Additional file 1: Note 1) [54].

Measurement of population growth rates

In order to measure population growth rates, the OD₆₀₀ values of cultures grown for L-cysteine productivity analysis (as described in the previous section) were recorded every hour. Un-inoculated medium was used for background subtraction. Calculation of growth rates were performed with the following formula:

$$r= {\frac{N(t)}{N(0)}}^\frac{1}{t}-1$$

where: N(t): cell number at time t, N (0): cell number at time 0, r: growth rate and t: time passed. Values of measured growth rates can be found in Additional file 1: Table S3. Population growth rates were normalised based on growth rates of the corresponding non-producing strains harbouring empty vectors.

Cell number counts were calculated with Agilent’s online tool “E. coli Cell Culture Concentration from OD₆₀₀ Calculator” (https://www.agilent.com/store/biocalculators/calcODBacterial.jsp, accessed 11/14/2022).

RNA sequencing and analysis

Sampled cryo-cultures of the simulated long-term cultivation were re-cultivated for 72 h with a starting OD₆₀₀ of 0.01 into 25 ml fresh medium under above mentioned conditions. One sample per early/late generation population, plasmid and strain (12 in total; strains: W3110, MDS42; plasmids: pCYS, pCYS_i, pCYS_m) were chosen for RNA extractions using a standard RNA extraction kit (Promega SV total RNA isolation system). Depletion of ribosomal RNA, transcriptome library construction, and sequencing of the corresponding samples were performed by Eurofins Genomics. Sequencing was conducted with the technology of Illumina NovaSeq 6000 with a 150 bp paired-end reading. Raw read counts were created using featureCounts (version 1.5.1) [55]. Only reads overlap** “CDS” features were counted. All reads map** to features with the same meta-feature attribute were summed. Only reads with unique map** positions and a map** quality score of at least 10 were considered for read counting. Supplementary alignments were ignored. Instead of counting paired-end reads twice, they were counted only once, i.e. as single fragment. Reads map** to multiple features were assigned to the feature that has the largest number of overlap** bases. A Trimmed Mean of M-values (TMM) normalization was performed using the edgeR package (version 3.16.5) [56, 57]. Map** of reads to reference sequences of E. coli K-12 W3110 and E. coli K-12 MDS42 was performed using BWA-NEM (version 0.7.12-r1039).

To determine which features are significantly differentially expressed, a small threshold (< < 0.01) was applied to the false discovery rate (FDR) values, which is the p-value adjusted for multiple testing using Benjamini–Hochberg procedure. Fold changes were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP).

The metabolic gene clustering was carried out with the EcoCyc E. coli database regarding all differentially expressed features with p-values < 0.05. Map** results and expression profiling statistics are shown in Additional file 1: Tables S4, S5.

Plasmid analysis of early and late generation population of E. coli W3110 and MDS42

Sampled cryo-cultures of the simulated long-term cultivation were re-cultivated for 72 h with a starting OD₆₀₀ of 0.01 into 25 ml fresh medium under above mentioned conditions. One sample per early/late generation population, plasmid and strain (12 in total; strains: W3110, MDS42; plasmids: pCYS, pCYS_i, pCYS_m) were chosen for plasmid extraction, using a standard kit (ThermoFisher Scientific GeneJET Plasmid Miniprep Kit). In order to test for potential plasmid loss during cultivation, we determined the plasmid DNA content in early and late generation populations (Additional file 1: Table S12). For The subsequent library preparation and deep sequencing was performed by Eurofins Genomics for Illumina NovaSeq 6000 S4 paired-end 2 × 150 bp. The per base coverage depth was > 140.000x. Adapter trimming, quality filtering and per-read pruning was done to retain only high quality bases. Reads were then mapped to the corresponding reference plasmid sequence, followed by a single nucleotide variant calling. Reads which could not be mapped to the plasmid reference sequences, got then mapped against an insertion sequence database (ISfinder_Nucl) with the Burrows-Wheeler Aligner. The reads, which could then be mapped to insertion sequences, were blasted against each insertion sequence family in the E. coli W3110 genome with NCBI’s megablast algorithm. Only highly similar sequences were selected and displayed.

Availability of data and materials

All data and materials are available as described in the study and its additional Additional file 1.

Abbreviations

ALE:: Adaptive laboratory evolution
C. Glutamicum :: Corynebacterium glutamicum
DEG:: Differentially expressed gene
E. coli :: Escherichia coli
EGP:: Early generation population
IS:: Insertion sequence
LGP:: Late generation population
NGS:: Next generation sequencing
R.P.M:: Rounds per minute
SNP:: Single-nucleotide polymorphism

References

Cicchillo RM, Baker MA, Schnitzer EJ, Newman EB, Krebs C, Booker SJ. Escherichia coli L-serine deaminase requires a [4Fe-4S] cluster in catalysis. J Biol Chem. 2004;279(31):32418–25.
Article CAS Google Scholar
Johnson DC, Dean DR, Smith AD, Johnson MK. Structure, function, and formation of biological iron-sulfur clusters. Annu Rev Biochem. 2005;74:247–81.
Article CAS Google Scholar
Barrose JM. Commission Implementing Regulation (EU) No 1006/2013 concerning the authorisation of L-cystine as a feed additive for all animal species. Official J European Union. 2013;59–60.
Clemente Plaza N, Reig Garcia-Galbis M, Martinez-Espinosa RM. Effects of the usage of l-Cysteine (l-Cys) on human health. Molecules. 2018. https://doi.org/10.3390/molecules23030575.
Article Google Scholar
Ismail NI. Production of cysteine: approaches, challenges and potential solution. Int J Biotechnol Wellness Indus. 2014;3:95–101.
Article Google Scholar
Renneberg R. High grade cysteine no longer has to be extracted from hair. Biotechnology for beginners. Amsterdam: Academic Press; 2008. p. 132.
Google Scholar
Berehoiu RMT, Popa CN, Popescu S. Assessment of the E 920 additive (L-cysteine) in relation to some problems of modern food industry. Economic Eng Agric Rural Devel. 2013;13(1):413–8.
Google Scholar
Shiba T, Takeda K, Yajima M, Tadano M. Genes from Pseudomonas sp. strain BS involved in the conversion of L-2-amino-Delta(2)-thiazolin-4-carbonic acid to L-cysteine. Appl Environ Microbiol. 2002;68(5):2179–87.
Article CAS Google Scholar
Nam KH. Effects of anoxic conditions on the enzymatic conversion of D, L-2-amino-thiazoline-4-carboxylic acid to L-cystine. Acta Biotechnol. 1997;17:185–93.
Article CAS Google Scholar
Wei L, Wang H, Xu N, Zhou W, Ju J, Liu J, et al. Metabolic engineering of Corynebacterium glutamicum for L-cysteine production. Appl Microbiol Biotechnol. 2019;103(3):1325–38.
Article CAS Google Scholar
Nielsen J, Keasling JD. Engineering cellular metabolism. Cell. 2016;164(6):1185–97.
Article CAS Google Scholar
Borkowski O, Ceroni F, Stan GB, Ellis T. Overloaded and stressed: whole-cell considerations for bacterial synthetic biology. Curr Opin Microbiol. 2016;33:123–30.
Article CAS Google Scholar
Carbonell-Ballestero M, Garcia-Ramallo E, Montanez R, Rodriguez-Caso C, Macia J. Dealing with the genetic load in bacterial synthetic biology circuits: convergences with the Ohm’s law. Nucleic Acids Res. 2016;44(1):496–507.
Article CAS Google Scholar
Lynch M, Marinov GK. The bioenergetic costs of a gene. Proc Natl Acad Sci USA. 2015;112(51):15690–5.
Article CAS Google Scholar
Ceroni F, Algar R, Stan GB, Ellis T. Quantifying cellular capacity identifies gene expression designs with reduced burden. Nat Methods. 2015;12(5):415–8.
Article CAS Google Scholar
Glick BR. Metabolic load and heterologous gene expression. Biotechnol Adv. 1995;13:247–61.
Article CAS Google Scholar
Matthew Scott CWG, Mateescu Eduard M, Zhang Zhongge, Hwa Terence. Interdependence of cell growth and gene expression, origins and consequences. Science. 2010;330:1099–102.
Article Google Scholar
Ohtsu I, Wiriyathanawudhiwong N, Morigasaki S, Nakatani T, Kadokura H, Takagi H. The L-cysteine/L-cystine shuttle system provides reducing equivalents to the periplasm in Escherichia coli. J Biol Chem. 2010;285(23):17479–87.
Article CAS Google Scholar
Susann Müller HHaTB. Origin and analysis of microbial population heterogeneity in bioprocesses. Anal Biotechnol. 2010;21:100–13.
Google Scholar
**ao Y, Bowen CH, Liu D, Zhang F. Exploiting nongenetic cell-to-cell variation for enhanced biosynthesis. Nat Chem Biol. 2016;12(5):339–44.
Article CAS Google Scholar
Rugbjerg P, Myling-Petersen N, Porse A, Sarup-Lytzen K, Sommer MOA. Diverse genetic error modes constrain large-scale bio-based production. Nat Commun. 2018;9(1):787.
Article Google Scholar
Podlesek Z, Zgur BD. The DNA damage inducible SOS response is a key player in the generation of bacterial persister cells and population wide tolerance. Front Microbiol. 2020;11:1785.
Article Google Scholar
Napolitano R, Janel-Bintz R, Wagner J, Fuchs RP. All three SOS-inducible DNA polymerases (Pol II, Pol IV and Pol V) are involved in induced mutagenesis. EMBO J. 2000;19:6259–65.
Article CAS Google Scholar
Loffler M, Simen JD, Jager G, Schaferhoff K, Freund A, Takors R. Engineering E. coli for large-scale production—strategies considering ATP expenses and transcriptional responses. Metab Eng. 2016;38:73–85.
Article Google Scholar
Harder BJ, Bettenbrock K, Klamt S. Model-based metabolic engineering enables high yield itaconic acid production by Escherichia coli. Metab Eng. 2016;38:29–37.
Article CAS Google Scholar
Leinfelder W, Heinrich P. European Patent Office. Process for preparing O-Acetylserine, L-cysteine and L-cysteine-related products. EP0858510B1 (Patent) 2001.
Bell JK, Pease PJ, Bell JE, Grant GA, Banaszak LJ. De-regulation of D-3-phosphoglycerate dehydrogenase by domain removal. Eur J Biochem. 2002;269(17):4176–84.
Article CAS Google Scholar
Daßler T. Identification of a major facilitator protein from Escherichia coli involved in efflux of metabolites of the cysteine pathway. Mol Microbiol. 2000;36:1101–12.
Article Google Scholar
Winterhalter C, Leinfelder W. European Patent Office. Microorganisms and process for the fermentative production of L-Cysteine, L-Cystine, N-Acetyl-Serine or thiazolidin-derivates. EP0885962B1 (Patent) 1998.
Tobias Daßler TM. Christoph Winterhalter and August Böck Identification of a major facilitator protein from Escherichia coli involved in efflux of metabolites of the cysteine pathway. Mol Microbiol. 2000;36:1101–12.
Article Google Scholar
Umenhoffer K. Reduced evolvability of Escherichia coli MDS42, an IS-less cellular chasis for molecular and synthetic biology applications. Microb Cell Fact. 2010. https://doi.org/10.1186/1475-2859-9-38.
Article Google Scholar
Csorgo B, Feher T, Timar E, Blattner FR, Posfai G. Low-mutation-rate, reduced-genome Escherichia coli: an improved host for faithful maintenance of engineered genetic constructs. Microb Cell Fact. 2012;11:11.
Article CAS Google Scholar
Liu H, Wang Y, Hou Y, Li Z. Fitness of chassis cells and metabolic pathways for l-cysteine overproduction in Escherichia coli. J Agric Food Chem. 2020;68(50):14928–37.
Article CAS Google Scholar
Mamum AAMA. Identity and function of a large gene network underlyiing mutagenic repair of DNA breaks. Sci Rep. 2012;338:1344–8.
Google Scholar
Loddeke M, Schneider B, Oguri T, Mehta I, Xuan Z, Reitzer L. anaerobic cysteine degradation and potential metabolic coordination in Salmonella enterica and Escherichia coli. J Bacteriol. 2017. https://doi.org/10.1128/JB.00117-17.
Article Google Scholar
Nonaka G, Takumi K. Cysteine degradation gene yhaM, encoding cysteine desulfidase, serves as a genetic engineering target to improve cysteine production in Escherichia coli. AMB Express. 2017;7(1):90.
Article Google Scholar
Berteau O. A new type of bacterial sulfatase reveals a novel maturation pathway in prokaryotes. J Biol Chem. 2006;281:22464–70.
Article CAS Google Scholar
Ploeg JRvd. Identification of sulfate starvation-regulated genes in escherichia coli, a gene cluster involved in the utilization of taurine as a sulfur source. J Bacteriol. 1996. https://doi.org/10.1128/jb.178.18.5438-5446.1996.
Article Google Scholar
van Der Ploeg JR, Iwanicka-Nowicka R, Bykowski T, Hryniewicz MM, Leisinger T. The Escherichia coli ssuEADCB gene cluster is required for the utilization of sulfur from aliphatic sulfonates and is regulated by the transcriptional activator Cbl. J Biol Chem. 1999;274(41):29358–65.
Article Google Scholar
Sirko A. Sulfate and thiosulfate transport in Escherichia coli K-12: nucleotide sequence and expression of the CysTWAM gene cluster. J Bacteriol. 1990. https://doi.org/10.1128/jb.172.6.3351-3357.1990.
Article Google Scholar
Kertesz MA. Proteins induced by sulfate limitation in Escherichia coli, Pseudomonas putida, or Staphylococcus aureus. J Bacteriol. 1993. https://doi.org/10.1128/jb.175.4.1187-1190.1993.
Article Google Scholar
Leyh TS, Vogt TF, Suo Y. The DNA sequence of the sulfate activation locus from Escherichia coli K-12. J Biol Chem. 1992;267(15):10405–10.
Article CAS Google Scholar
Ostrowski J, Wu JY, Rueger DC, Miller BE, Siegel LM, Kredich NM. Characterization of the cysJIH regions of Salmonella typhimurium and Escherichia coli B. J Biol Chem. 1989;264(26):15726–37.
Article CAS Google Scholar
Tanaka Y. Crystal structure of a YeeE-YedE family protein engaged in thiosulfate uptake. Sci Adv. 2020. https://doi.org/10.1126/sciadv.aba7637.
Article Google Scholar
Haas M, Rak B. Escherichia coli insertion sequence IS150: transposition via circular and linear intermediates. J Bacteriol. 2002;184(21):5833–41.
Article CAS Google Scholar
Chen CS, Korobkova E, Chen H, Zhu J, Jian X, Tao SC, et al. A proteome chip approach reveals new DNA damage recognition activities in Escherichia coli. Nat Methods. 2008;5(1):69–74.
Article CAS Google Scholar
Kumar A, Beloglazova N, Bundalovic-Torma C, Phanse S, Deineko V, Gagarinova A, et al. Conditional epistatic interaction maps reveal global functional rewiring of genome integrity pathways in Escherichia coli. Cell Rep. 2016;14(3):648–61.
Article CAS Google Scholar
Posfai G, Plunkett G 3rd, Feher T, Frisch D, Keil GM, Umenhoffer K, et al. Emergent properties of reduced-genome Escherichia coli. Science. 2006;312(5776):1044–6.
Article CAS Google Scholar
Kumar PK. Strategies for improving plasmid stability. Trends Biotechnol. 1991;9:279–84.
Article CAS Google Scholar
Ikeda M. Amino acid production process. In: Robert Faurie JT, Bathe B, Debabov VG, Huebner S, Ikeda M, Kimura E, Marx A, Möckel B, Mueller U, Pfefferle W, editors. Advances in Biochemical Engineering/Biotechnology. Berlin: Springer; 2003.
Google Scholar
Stanbury PF, Whitaker A, Hall SJ. Principles of fermentation technology. Amsterdam: Elsevier; 2013.
Google Scholar
Park MK, Lee SH, Yang KS, Jung SC, Lee JH, Kim SC. Enhancing recombinant protein production with an Escherichia coli host strain lacking insertion sequences. Appl Microbiol Biotechnol. 2014;98(15):6701–13.
Article CAS Google Scholar
Zhong S, Dean AM. Rapid identification and map** of insertion sequences in Escherichia coli genomes using vectorette PCR. BMC Microbiol. 2004;4:26.
Article Google Scholar
Gaitonde MK. A spectrophotometric method for the direct determination of Cysteine in the presence of other naturally occuring amino acids. Biochem J. 1967;104:627–33.
Article CAS Google Scholar
Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30.
Article CAS Google Scholar
Robsinson MD. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010. https://doi.org/10.1186/gb-2010-11-3-r25.
Article Google Scholar
Robinson MD, McCarthy DJ, Smyth GK. edgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–40.
Article CAS Google Scholar

Download references

Acknowledgements

Not applicable

Funding

Open Access funding enabled and organized by Projekt DEAL. The study was financed by the industrial Contract No #2018112610005964.

Author information

Authors and Affiliations

Technische Universität München, Lichtenbergstraße 4, 85748, Garching, Germany
Kevin Heieck, Nathanael David Arnold & Thomas Bartholomäus Brück

Authors

Kevin Heieck
View author publications
You can also search for this author in PubMed Google Scholar
Nathanael David Arnold
View author publications
You can also search for this author in PubMed Google Scholar
Thomas Bartholomäus Brück
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

TBB: Funding acquisition, Project administration, Conceptualisation, Writing- Reviewing & Editing KH: Conceptualization, Methodology, Investigation, Data curation, Writing—Original draft preparation and reviewing & editing NDA: Investigation, Writing—Reviewing & Editing. TB and KH conceived and designed the study. KH carried out the experiments and drafted the manuscript. KH, NA and TB edited and finalised the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Thomas Bartholomäus Brück.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

There are no conflicts to declare.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1

: Table S1. Oligo sequences used for assembly of plasmids pCYS_i and pCYS_m. Table S2. Average number of generations after each passage of the simulated fermentation of W3110 and MDS42 with integrated pCYS, pCYS_i and pCYS_m. Each strain was cultivated in biological triplicates. Table S3. Average growth rates of samples taken after each passage of the simulated long-term cultivation of W3110 and MDS42 with integrated pCYS, pCYS_i and pCYS_m. Each strain was cultivated in biological triplicates. Growth rates were calculated according to the formula found in the manuscripts’ methods section of “Measurement of population growth rates”. Table S4. Map** statistics overview. For each sample, the following statistics are provided: Reads mapped: the total number of reads mapped to the reference genome. Unique: number of uniquely mapped reads, i.e. read can only be mapped to one reference locus. Reference covered: reference bases covered by at least one read. Mean read coverage: average read coverage of the reference sequence. HGP: high generation population, LGP: low generation population. Table S5. Expression profiling statistics. For each sample, the following statistics are provided: Effective library size: The total number of reads mapped to reference features. Normalized library size: The total number of reads mapped to reference features normalized by the associated normalization factor, which can be derived by dividing the normalized library size by the effective library size. No feature: The number of reads map** to the reference sequence that could not be assigned to any annotated feature, i.e. map** positions and feature positions do not overlap. Filtered: The number of reads that were filtered due to insufficient map** quality or ambiguous map** location. These reads were ignored for read counting. Table S6. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of W3310_pCYS with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). *: Genes were excluded because they did not fall within the FC range of the metabolic cluster. Table S7. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of MDS42_pCYS with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). Table S8. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of W3110_pCYS_i with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). Table S9. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of MDS42_pCYS_i with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). Table S10. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of W3110_pCYS_m with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). Table S11. Table showing accession numbers, gene names, features, logarithmic fold changes (logFC) and p-values of all differentially expressed genes (DEGs) of MDS42_pCYS_m with p-values <0.05. LogFC and logCPM were calculated by dividing values of the later generation population (LGP) by values of the early generation population (EGP). Table S12. Values of calculated plasmid DNA contents extracted from early and late generation populations (EGP, LGP). Plasmid DNA was extracted from 10 ml cultures and eluted in 50 µl each. Cell dry weights were extrapolated with the factor of 0.33 g/L/OD_{600= 1.0} for E. coli K-12 MG1655 cells according to Sauer et al. (2). Table S13. Single-nucleotide polymorphism (SNP) variant table. The SNP calling was done using VarSCan2 (3). Allele frequency cut-off used for variant calling was 1%. For each sample, the following variant summary is provided: Strain, plasmid, population. Additionally, POS: Position at which the variant was observed, BB/ORF: Location affected by the mutation (BB: backbone, ORF: open reading frame), REF: Reference base, ALT: Alternative base, Allele Freq: Variant allele frequency in percentage, Alt Depth: Depth of variant-supporting bases, total depth: Depth of variant-supporting bases and reference-supporting bases. Mutations which showed allele-frequencies >95% were assumed to be originally present in the plasmids. Figure S1. Plots of total L-cysteine yields in mg/OD₆₀₀ = 1.0. Cultivation and subsequent L-cysteine yields determination was carried out in biological triplicates of W3110 (A) and MDS42 (B) with the three different plasmids pCYS, pCYS_i and pCYS_m. Figure S2. Per base coverage depths (x) of sequenced pCYS extracted from early and late generation populations (EGPs and LGPs) of E. coli W3110 and MDS42. Sequencing was conducted with Illumina Novaseq paired end 2x150bp. Figure S3. Per base coverage depths (x) of sequenced pCYS_i extracted from early and late generation populations (EGPs and LGPs) of E. coli W3110 and MDS42. Sequencing was conducted with Illumina Novaseq paired end 2x150bp. Figure S4. Per base coverage depths (x) of sequenced pCYS_m extracted from early and late generation populations (EGPs and LGPs) of E. coli W3110 and MDS42. Sequencing was conducted with Illumina Novaseq paired end 2x150bp. Note 1. Spectrophotometric protocol for L-Cysteine determination adapted from Gaitonde (1).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article

Heieck, K., Arnold, N.D. & Brück, T.B. Metabolic stress constrains microbial L-cysteine production in Escherichia coli by accelerating transposition through mobile genetic elements. Microb Cell Fact 22, 10 (2023). https://doi.org/10.1186/s12934-023-02021-5

Download citation

Received: 20 October 2022
Accepted: 09 January 2023
Published: 16 January 2023
DOI: https://doi.org/10.1186/s12934-023-02021-5

Metabolic stress constrains microbial L-cysteine production in Escherichia coli by accelerating transposition through mobile genetic elements

Abstract

Background

Results

Conclusion

Similar content being viewed by others

Glucose consumption rate-dependent transcriptome profiling of Escherichia coli provides insight on performance as microbial factories

Genome reduction boosts heterologous gene expression in Pseudomonas putida

Omics-guided bacterial engineering of Escherichia coli ER2566 for recombinant protein expression

Background

Conclusions

Materials and methods

Strains and plasmids

Media

Simulated long-term cultivation

Determination of L-cysteine yields by spectrophotometry

Measurement of population growth rates

RNA sequencing and analysis

Plasmid analysis of early and late generation population of E. coli W3110 and MDS42

Availability of data and materials

Abbreviations

References

Acknowledgements

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Additional information

Publisher's Note

Supplementary Information

Additional file 1

Rights and permissions

About this article

Cite this article

Share this article

Keywords

Search

Navigation