With its fast growth and simple cultivation Escherichia coli is a widely used microorganism in biotechnological processes and in industrial microbiology. One of the most important applications of recombinant DNA technology is the genetic manipulation of E. coli K-12 for the production of human insulin [1]. Modified E. coli strains are also currently used for the synthesis of different enzymes, amino acids and other peptide hormones. To maximize productivity, i.e., the yield in relation to duration and costs, it is essential to permanently optimize the biotechnological process. One major problem during high density growth of E. coli K-12 is the production and excretion of acetate, which affects growth and recombinant protein expression [2,3]. To circumvent this, different growth strategies [3] have been applied as well as targeted changes in central carbon metabolism [2,4] or control of the glucose transport process has been modified [5,6]. Especially the latter approach seems to be very helpful since acetate excretion mainly occurs when the transport rate exceeds the metabolism which causes a temporal metabolic imbalance.

In E. coli, glucose is taken up by the phosphoenolpyruvate (PEP)-dependent glucose-phosphotransferase-system (Glc-PTS) [7]. Phosphotransferase systems usually consist of two cytoplasmic energy-coupling proteins, Enzyme I (EI, gene ptsI) and Histidine-containing protein (HPr, gene ptsH), and in particular for E. coli K-12 of a range of more than 20 different carbohydrate spe­cific Enzymes II (EIIs), which catalyze concomitant carbohydrate transport and phosphorylation [8]. The first step in the PTS-typical phosphorylation-chain is catalyzed by EI, a PEP-dependent protein-kinase. The use of PEP, an intermediate of glycolysis, as a phosphoryl group donor couples tightly carbohydrate transport and metabolism. In the case of the Glc-PTS, the phosphate group is subsequently transferred from EI~P to HPr, from HPr~P to the soluble EIIA^Glc (sometimes also called EIIA^Crr, gene crr), and finally from EIIA^Glc~P to the glucose-specific membrane protein EIICB^Glc (gene ptsG), which is responsible for glucose uptake and phosphorylation. EIICB^Glc (50.7 kDa) consists of two functional domains, the membrane bound EIIC^Glc domain (41.1 kDa) and the cytosolic EIIB^Glc domain (9.6 kDa). The EIIC^Glc domain forms a stable homodimer in the membrane and is responsible for glucose uptake, whereas the EIIB^Glc is located in the cytoplasm and phosphorylates the glucose [9]. Both domains are connected through a flexible linker. The linker is surface exposed, since a proteolytic cleavage within the linker is possible [10]. Phosphorylation of EIICB^Glc protects against protease cleavage, suggesting a conformational change of this region during glucose uptake [10]. The linker shows the highly conserved amino acid sequence KTPGRED (aa 382-388) which is present in most of the PTS transport proteins of the glucose/N-acetyl-glucosamine family. The function of this motif was unclear so far [7]. This motif appears to be nonessential for transport, since alanine substitutions show no or only a slight effect with the exception of EIICB^GlcG385A which exhibited a highly reduced phosphorylation activity of less than 10% of wild type activity [10,11]. Only a complete deletion of this sequence led to a total loss of transport and phosphorylation activity [12].

Regulation of the ptsG gene for the EIICB^Glc is very complex and occurs both at the levels of transcription and posttranscriptional control. The major specific regulator of ptsG expression is the repressor Mlc (mnemonic for makes large colonies, previously DgsA, gene dgsA), which is inactivated by glucose in the medium. In contrast to other repressors, induction of Mlc is not catalyzed by direct binding of glucose, or by any other small molecular inducer. Instead, as part of an unusual regulatory mechanism, the membrane-bound EIICB^Glc binds Mlc, but only when it is in its dephosphorylated form. Thus, in the absence of glucose, Mlc binds to its target promoter/operator ptsGop, while in the presence of glucose, the dephosphorylated EIICB^Glc sequesters the repressor away from its operator, allowing enhanced ptsG transcription [13,14,15,16]. Besides this main regulation via the glucose repressor Mlc, several other global factors were identified. These are cAMP-CAP [17], ArcA [18], SoxS [19], Fis [20] and two alternating sigma factors σ³² [21] and σ^S [22]. In addition to these transcriptional regulation mechanisms, a posttranscriptional regulation system, the so-called sgrRST-system [23,24], was identified as regulating ptsG mRNA stability as well as transport activity of EIICB^Glc. Accumulation of glucose-6-phosphate (Glc6-P) or fructose-6-phosphate (Fru6-P) in the cell activates the transcriptional activator SgrR which, in turn, is responsible for the activation of the small regulatory sRNA SgrS [24]. SgrS itself has two functions. On the one hand, it is capable of forming Hfq-dependent RNA-RNA hybrids with the ptsG mRNA, a first step in RNaseE dependent specific degradation of this mRNA. On the other hand, the sgrS gene encodes a small protein of 43 amino acids called SgrT which is responsible for downregulation of glucose transport [25].

The recent discoveries of these two posttranscriptional regulatory mechanisms provide new approaches to control glucose uptake under various growth conditions. For example Negrete et al. managed to reduce acetate excretion of glucose fermenting E. coli cells by overexpressing SgrS [26]. Likewise, exclusive overproduction of SgrT led to a drastic reduction of cell growth in minimal medium with glucose as a sole carbon source [25]. This gave the first hint that the Glc-PTS might be a direct target of SgrT and that the functions of SgrS and SgrT are redundant. Subsequently, Gabor et al. [27] repeated this growth experiment using minimal medium with sucrose as a single carbon source. For this experiment, an E. coli derivative was used, which shares all the components of the typical PTS cascade with the Glc-PTS (EI, HPr, EIIA^Glc) with the exception of the sucrose specific EIIBC^Scr [28]. This transporter protein like the EIICB^Glc belongs to the glucose/N-acetyl-glucosamine–sucrose/ß-glucosides superfamily of EII proteins [29]. However, the sucrose specific transporter has a different order of the two functional domains and lacks a conserved KTPGRED motif in the linker region between these two sites. Overproduction of SgrT did not interfere with cell growth in minimal medium with sucrose providing a first hint that indeed EIICB^Glc and no other component of the Glc-PTS might be the SgrT target [27].

In this study we focused our interest on the regulation of EIICB^Glc activity by SgrT. We identified the SgrT target sequence within EIICB^Glc and characterized the interaction between the glucose transporter and the small regulatory peptide in great detail. This may eventually lead to novel approaches to minimize metabolic overflow and thus improve the feasibility of the use of E. coli in biotechnological applications.

Media and growth conditions. Cells were grown routinely either in Luria broth without glucose and calcium ions (LB₀), or in 2xTY medium as described [55]. Antibiotics were used at the following concentrations: tetracycline (Tc) 10 mg/L, kanamycin (Kn) 25 mg/L and ampicillin (Ap) 50 mg/L, respectively. Minimal medium supplemented with 0.2% carbon source was used as indicated [56]. IPTG was used at concentrations of 100µM-500µM for induction of protein production in growth inhibition assays and at a concentration of 1 mM for induction of protein production for crosslinking experiments. Cells were incubated at 37 °C with shaking.

Bacteria strains and plasmids. All strains used were E.coli K-12 derivates. Table 1 and Table 2 list the genotypes and sources of the relevant bacterial strains and plasmids. Oligonucleotides are listed in Table 3 in the supplemental materials. Alleles were moved between strains by P1-transduction (performed as described previously [47]) or inserted via λ Red recombination [44].

Isolation of chromosomal and plasmid DNA, restriction analysis, PCR and DNA sequencing. All manipulations of chromosomal or recombinant DNA were carried out using standard procedures as described previously [55]. Plasmid DNA was prepared using QIAprep Spin Miniprep Kit (Qiagen, Hilden, Germany). Restriction enzymes were purchased from New England Biolabs (Schwalbach, Germany) or Fermentas (St. Leon-Rot, Germany) and used according to supplier recommendations. Oligonucleotides for PCR were purchased from Thermo Fisher Scientific (Ulm, Germany). DNA sequencing was commissioned to Scientific Research and Development (Bad Homburg, Germany). Polymerase chain reactions (PCR) were performed as described by [57] using TaKaRa DNA polymerase from Lonza (Köln, Germany).

Construction of plasmids. The open reading frame of sgrT (sgrT ORF) was amplified using chromosomal DNA of LJ110, which is closely related to wild type E. coli K-12, with forward primer sgrT+, containing a PstI restriction site and reverse primer sgrT-, which had a HindIII restriction site. This PCR product was purified with Wizard DNA purification system (Promega) and cloned into the vector pTM30, resulting in plasmid pTM30sgrT. The expression plasmid pTM30 provides a tac-promoter, an artificial start codon and and an artificial ribosomal binding site as described before [54]. pTM30sgrT3HA carries additional sequences that encode a 3xHA tag (received from pFA6a3HA, Oligonucleotides HA+/-) fused to the C-terminus of SgrT. For pACYC184sgrT3HA the tac_PO and sgrT3HA sequence from pTM30sgrT3HA was amplified with oligonucleotides TacPO+ and HA- and cloned into the vector pACYC184. For the bimolecular fluorescence complementation assay the sgrT ORF was amplified by PCR (Oligonucleotides pETS+/-) and cloned into the vector pET11a-link-NGFP [52] using the restriction enzymes XhoI/BamHI. The genes encoding EIIB^Glc (aa 389-477), Linker-EIIB^Glc (aa 380-477), Linker-EIIB^Glc-P384R (aa 380-477), EIIC^Glc (aa 1-381), EIIC^Glc-Linker (aa 1-396), EIIC^Glc-Linker-P384R (aa 1-396) and EIICB^Glc (aa 1-477) were also amplified by PCR and purified. The oligonucleotides were used as follows: EIIB^Glc (pMRB+/pMRG-), Linker-EIIB^Glc (pMRLB+/pMRG-), Linker-EIIB^Glc-P384R (pMRLB-P384R+/pMRG-), EIIC^Glc(pMRG+/pMRC-), EIIC^Glc-Linker (pMRG+/pMRCL-), EIIC^Glc-Linker-P384R (pMRG+/pMRCL-P384R-) and EIICB^Glc (pMRG+/-). PCR products were cloned into the vector pMRBAD-link-CGFP [52] using the restriction enzymes NcoI/AatII or SphI/AatII, respectively. For fluorescence microscopy, an SgrT-GFP fusion protein was created. The sgrT ORF was amplified with oligonucleotides SgrT+/SgrT2- and a gfp gene was amplified using pBLP2 as template and oligonucleotides Gfp2+/-. Both PCR products were purified and cloned together into pTM30, resulting in pTM30sgrT-gfp. All oligonucleotide sequences used are listed in Table3 in the supplemental materials.

Site-directed mutagenesis. We generated defined mutations in the ptsGHis gene using the pRR48GH plasmid as a template and the Phusion Site-directed Mutagenesis Kit according to standard protocol (Finnzymes, Vantaa, Finland). The oligonucleotides used are listed in Table 3 in the supplemental materials (K150E+/-, V12F+/-, K382A+/ktpg-, T383A+/ktpg-, P384A+/ktpg-, P384R+/ktpg-, G385A+/ktpg-, R386A+/red-, E387A+/red-,D388A+/red-).

Construction of strains. JKA1: In order to characterize the consequence of a sgrRST-deletion we disrupted the sgrRST-wildtype genes via λ Red recombination as described in the protocol of Datsenko and Wanner [44]. Briefly, a chloramphenicol resistance selection marker with flanking regions that are homologous to chromosomal sequences at the 5’ and 3’ end of the sgrRST gene region was amplified from template pKD3 [44] by using the primer pair SgrR+ and SgrS-. The PCR product was purified with the Wizard DNA purification system (Promega), treated with DpnI and further enriched by ethanol precipitation. Subsequently, the DNA-fragment was integrated into the chromosome of BW25113/pKD46 [44] via λ Red recombination, resulting in BW25113ΔsgrRST::cat. JKA1 was created by the transfer of the deletion cassette of BW25113ΔsgrRST::cat into LJ110 via P1-transduction. PCR of flanking regions was used to confirm the correct integration of the desired gene disruption.

JKA12: For crosslinking analysis a strain with a deletion of ptsG and sgrRST gene regions was created. The sgrRST gene region in BW25113/pKD46 was disrupted with a kanamycin resistance cassette via λ Red recombination, using the protocol of Datsenko and Wanner [44]. The deletion cassette of BW25113ΔsgrRST::kan was then inserted via P1-transduction into LJ120, resulting in JKA12. PCR was used to confirm correct integration of the desired gene disruption.

JKA17: To characterize the interaction of SgrT and EIICB^Glc with bimolecular fluorescence complementation a BL21 (λDE3) strain with a ptsG-deletion was created via P1-transduction (ΔptsG::cat from LJ120). PCR was used to confirm correct integration of the desired gene disruption.

JKA18: For fluorescence microscopy one strain with a deletion of sgrRST gene region and a chromosomal point mutation in ptsG was constructed. The deletion cassette of BW25113ΔsgrRST::kan was inserted via P1-transduction into LJB5, resulting in JKA18. PCR was used to confirm correct integration of the desired gene disruption. The oligonucleotides are listed in Table 3 in the supplemental materials.

Western blot analyses. For Western blot analysis, bacterial cells were treated as described in the section “crosslinking with paraformaldehyde” or grown overnight in LB₀ with ampicillin and inoculated in fresh medium to an OD₆₅₀ = 0.1. The cultures were grown to early-log phase (OD₆₅₀ = 0.2) and induced with IPTG. Cells were harvested at an optical density at 650 nm of 1 by centrifugation and resuspended in 100 µL sterile water and 100 µL of SDS-PAGE loading buffer (125 mM Tris-HCl (pH 6.8), 2% sodium dodecyl sulphate (SDS), 10% glycerol, 5% ß-mercaptoethanol, 0.01% bromophenol blue). If not described otherwise, samples were heated at 95 °C for 10 min. 15 µL of total cellular proteins were separated by electrophoresis on 0.1% SDS-containing 15% polyacrylamide gels and transferred to a Nitrocellulose membrane (Schleicher & Schuell, Dassel, Germany). For the detection of EIICB^Glc-His protein derivatives, we used a Penta-His antibody (Qiagen, Hilden, Germany). SgrTec3HA was detected with HA-antibody (kindly provided by Anja Lorberg, University of Osnabrück). Detection of antibody binding was performed using infrared-labeled second antibodies (LI-COR Biosciences, Bad Homburg, Germany).Visualization and quantification were done using an Odyssey infrared imager (LI-COR Biosciences, USA) and the software provided by the supplier (Odyssey 2.1).

Crosslinking with paraformaldehyde. For crosslinking of proteins with paraformaldehyde the general procedure from [30] was followed. Cells were grown overnight in LB₀ media with ampicillin and tetracycline and inoculated in 200 mL fresh medium to an OD₆₅₀ = 0.1. The cultures were grown for one hour at 37 °C and induced with 1mM IPTG. After one hour 0.2% glucose was added to cultures when indicated and cultures were incubated for another hour. Then paraformaldehyde solution (4% in PBS (136 mM NaCl, 2.7 mM KCl, 1.8 mM KH₂PO₄, 10 mM Na₂HPO₄) was added in a concentration of 0.3%. Cultures were incubated for 20 min at 37 °C while shaking and cells were harvested via centrifugation. The pellet was washed in a lysis buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM Imidazol, pH 8.0) and finally resuspended in 5 mL of lysis buffer. 1mM AEBSF was added and cells were disrupted by sonification. Cell debris was removed via centrifugation and the supernatant was used for solubilization of membrane proteins. Therefore 2% triton X-100 was added to the supernatant and incubated at room temperature (RT) for 30 min while mixing. Membranes were removed via ultracentrifugation. The supernatant was then used for protein purification with Ni-NTA Agarose (Qiagen, Hilden, Germany). 1.25 mL Ni-NTA agarose was mixed with 5 mL protein suspension and incubated for one hour at RT. Supernatant was removed via centrifugation and unbound protein was removed using wash buffer (50 mM NaH₂PO₄, 300 mM NaCl, 20 mM Imidazol, pH 8.0) twice. 625 µL (1/8) Elution buffer (50 mM NaH₂PO₄, 300 mM NaCl, 250 mM Imidazol, pH 8.0) was used to elute purified protein. The same amount SDS sample buffer was added, proteins heated to 95 °C for 10 min to destroy protein complexes, and equal amounts of proteins were analyzed with Western blot analysis.

Bimolecular fluorescence complementation. For bimolecular fluorescence complementation strain JKA17 was used. Protocol and plasmids were used as described in [31,52]. The cells were inoculated in rich medium with 100 µM IPTG, 0.4% arabinose and 0.2% glucose and incubated for three days at 25 °C while shaking. Cells were harvested via centrifugation and resuspended in 1 mL of lysis buffer [52]. OD₄₂₀ was determined and equal amounts of cells used for measuring fluorescence activity in a fluorimeter (Fluorometer Fluostar Optima, BMG LABTECH GmbH, Ortenberg, Germany).

Fluorescence microscopy. For standard microscope examination, cells were grown to early logarithmic phase in minimal medium complemented with 0.2% glucose. After 2 hours, 5 µM IPTG or 100 mM arabinose was added, incubated for two hours and cells were fixed on a slide with polylysin. The setup used for fluorescence microscopy consisted of a Zeiss Axioplan2e (Carl Zeiss, Jena, Germany) equipped with a 100× alpha-Plan Fluar objective (NA 1.45) and differential interference contrast (DIC). Images were acquired using a Photometrics CoolSNAP HQ Camera (Roper ScientiWc, Tucson, USA). Fluorescence was excited with a helium lamp and appropriate filter sets were used to adjust excitation and emission wavelengths. The setup was controlled by the Metamorphs v6.2 program (Universal Imaging Corporation, Downingtown, USA). Bright field images were acquired as single planes using t DIC. All fluorescence images were taken from single focal planes and scaled using Metamorphs scale image command. All GFP fusions were taken with 1 sec acquisition time. From all cultures, at least 100 cells were controlled. For unspecific cell wall staining, the cells were incubated with 4µM FM4-64 for 10 min at RT.

E. coli tends to produce acetate during high cell density fermentation. Acetate production takes place when the rates of carbohydrate transport and glycolysis exceed a critical value. Many attempts have been performed to couple carbohydrate uptake rates to metabolic flux in order to avoid overflow mechanisms. The sgrRST system provides new regulatory tools to artificially modify glucose uptake rates according to biotechnological needs. Clearly, further fundamental research efforts are necessary to adapt and optimize the sgrRST system as an instrument for fine-tuning carbohydrate uptake in biotechnological applications.