Organism and culture conditions
Axenic cultures of the cyanobacterium Synechocystis sp. PCC 6803 were obtained from the Pasteur Culture Collection (Paris, France). All cultures were grown photoautotrophically under continuous illumination of 150 μmol photons/m2 s (warm white fluorescent tubes, Osram L 32) at 29 °C. High-density cultures (optical density at 750 nm—OD750 of approximately 2.0) were grown in BG11  with different NaCl concentrations (ranging from 0 to 4 %) and bubbled with CO2-enriched air (5 %, v/v). Cultures with lower cell densities (OD750 of approximately 0.5) were grown in Erlenmeyer flasks in BG11 medium, which were shaken continuously at 120 rpm. For isoprene production studies, cultures were pre-cultivated at high CO2 in presence of different NaCl concentrations. After 24 h, the pre-cultures were used to inoculate the main cultures at OD750 of approximately 1 in 50 ml of BG11 with different NaCl concentrations. Isoprene production was induced by adding IPTG (1 mM final concentration) in the strains # 643 and # 704, in which ispS is under the control of P
. During the cultivation in closed Schott flasks, which allow the sampling of the headspace via sampling ports in the closing caps, 50 mM of NaHCO3 as an inorganic carbon source was added to the medium. The cultures were incubated at 30 °C, with an illumination of approximately 150 µmol photons/m2 s under continuous stirring at 150 rpm. After 24 h, samples of 500 µl of the head space were taken and injected manually in the GC–MS system. After analyzing the isoprene amounts, samples for GG determination and transcriptomic and metabolomic analyses were taken.
Growth analysis and pigment determination
Growth curves of Synechocystis wild type and the isoprene-producing strains were recorded over 24 h of incubation. The optical density of the culture was determined at 720 nm. To show the correlation of the optical density to the dry cell biomass, 5–10 ml of the culture was collected by filtration on MF Nitrocellulose Membrane Filters (0.45 µm) (Millipore, Darmstadt, Germany). Each sample was dried at 90 °C for 12 h and the dry cell weight was measured.
The chlorophyll a, phycocyanin, and carotinoide values were measured spectrophotometrically. These values were corrected according to Sigalat/de Kuckowski  and the chlorophyll a/phycocyanin and chlorophyll a/Carotinoide ratio was determined.
Synthesis of codon-optimized ispS gene
The isoprene synthase (ispS) cDNA sequence of Pueraria montana (kudzu vine) was obtained from the NCBI database (Acc. No. AY315652). To ensure efficient expression of the plant cDNA in the cyanobacterial host, the codon usage was adopted to that of Synechocystis. Rare codons in the kudzu ispS sequence, i.e., codon-usage frequency below 10 % in Synechocystis, were changed to more frequently used codons. The chloroplast import sequence was removed from the ispS gene. The optimized ispS sequence is shown in Additional file 2. The optimized ispS coding sequence flanked by the engineered P
promoter upstream as well as the oop terminator downstream was obtained via gene synthesis service (GeneArt® Gene Synthesis, Life Technologies).
Plasmid construction and conjugation of Synechocystis with the isoprene synthase gene
The synthetic P
-ispS-oop DNA fragment contained a SalI restriction site upstream and PstI site downstream, which facilitated subsequent insertion into the shuttle vector pVZ325. In addition, an NdeI restriction site overlapping with the start codon of ispS gene was inserted, which enabled subsequent promoter swaps. The synthetic DNA fragment was provided in a standard cloning vector, pMA (GeneArt® Gene Synthesis, Life Technologies). The P
-ispS-oop fragment was excised from the pMA vector via SalI/PstI digestion and then cloned into the SalI/PstI-cut pVZ325 vector (Additional file 3). To assess the ispS expression under different promoters, the psaA* promoter sequence was removed by SalI/NdeI and replaced by alternative promoter fragments with compatible cohesive ends. The rbcL promoter was obtained from Synechocystis and includes the native upstream region −260 to +1 bp relative to the rbcL start codon, while for the psbA2 promoter, the upstream region was chosen from −559 to +1 bp relative to the psbA2 start codon. The Ptac/lacI promoter was amplified by PCR from the E. coli cloning vector pGEX-6K-1 (Acc.Nr. U78872.1) and encompassed a 2142 bp DNA fragment that also contains the lacI repressor gene and the LacI-binding operator region of Ptac. The dxs gene was PCR-amplified from Synechocystis genomic DNA (sll1945). The 1923 bp DNA sequence for DXS was fused upstream with the psbA2 or the rbcL promoter via NdeI, and the oop terminator sequence was added downstream of the dxs stop codon. Respective dxs expression cassettes were cloned into pVZ325a via SalI/XmaI. pVZ325 derivative plasmids harboring an ispS expression cassette were transferred into Synechocystis cells by conjugation according to Zinchenko et al. . Exconjugants were selected on BG11 agar plates containing 10 μg/ml gentamycin.
Synechocystis 6803 cells were collected by centrifugation (4000 rpm, 4 °C, 4 min), and the cells were suspended in 500 µl PGTX solution  [39.6 % (w/v) phenol, 7 % (v/v) glycerol, 7 mM 8-hydroxyquinoline, 20 mM EDTA, 97.5 mM sodium acetate, 0.8 M guanidine thiocyanate, 0.48 M guanidine hydrochloride]. The suspensions were incubated for 15 min at 65 °C and then incubated on ice for 5 min. After addition of 500 μl chloroform/isoamyl alcohol (24:1), samples were incubated at room temperature for 10 min and centrifuged at 6000 rpm at 20 °C for 10 min. The upper aqueous phase was transferred into a new tube, and the same volume of chloroform/isoamyl alcohol (24:1) was added. After mixing, samples were centrifuged as described above, and the aqueous phase was removed again and combined with an equal volume of isopropanol. After gently inverting the tube, RNA was precipitated overnight at −20 °C. RNA was pelleted through centrifugation (13,000 rpm, 4 °C, 30 min). The pellet was washed with 1 ml of 70 % ethanol (13,000 rpm, 20 °C, 5 min), allowed to air dry for approximately 10 min and resuspended in 30 μl RNase-free distilled water.
cDNA synthesis, semi-quantitative RT-PCR and qRT-PCR
DNA-free RNA was reverse transcribed into cDNA using RevertAid H Minus reverse transcriptase (Fermentas, St. Leon-Rot, Germany) according to the manufacturer’s protocol. Before RT-PCR analysis, cDNA amounts were calibrated using the constitutively expressed rnpB gene. RT-PCR of ispS (primer sequences in Additional file 9) was performed using the Biometra Personal Cycler and PCR Master Mix (Qiagen) as described previously in more detail .
Calibrated cDNA was also used for qPCR analysis using the LightCycler 1.5 system (Roche, Basel, Switzerland) and SYBR Green fluorescence (Roche) for detection. To normalize gene expression, the constitutively expressed reference gene was amplified, and the average cycle threshold at each time point (n = 3) was used to calculate relative expression values. The expression of the selected genes at the different NaCl concentrations was normalized by subtraction of their cycle threshold values from the mean of the control gene, setting the respective value at 0 mM NaCl arbitrarily to 1.
A new high-resolution microarray was designed based on two recent RNAseq studies [64, 65]. RNA was directly labeled with the Kreatech ULS labeling kit for Agilent gene expression arrays with Cy3 according to the manufacturer’s protocol. Fragmentation and hybridization were performed following the manufacturer’s instructions for Agilent one-color microarrays. Feature extraction was performed with the Agilent “feature extraction” software and protocol “GE1_107_Sep09.” The raw data were analyzed with the R package limma . Raw data were normexp background subtracted and quantile normalized. All probes of one RNA feature were summarized, and control features were deleted. p values were adjusted for multiple testing after using the Benjamini–Hochberg procedure. The threshold for significant differentially expressed genes was log2 fold change ≥ 1 and adjusted p value ≤ 0.05. The data have been deposited in the GEO database under the accession number GSE74940.
Protein extraction and immune-blotting
For Western Blot analysis, cyanobacterial cells were collected by centrifugation (4000 rpm, 4 °C, 4 min), and suspended in Tris–EDTA-NaCl (TEN) buffer (50 mM Tris–HCl, pH 8; 5 mM EDTA, 100 mM NaCl) with addition of 100 μM PMSF. Cells were disrupted by sonication and insoluble material was removed by centrifugation. The supernatant was collected as the soluble protein fraction and the protein concentration was determined by Bradford . The protein samples were separated by 12 % SDS-PAGE and transferred to a PVDF membrane (GE Healthcare, Freiburg, Germany). For immunodetection, the rabbit serum-containing specific polyclonal antibodies against IspS  and horseradish peroxidase-conjugated secondary antibodies were used. Peroxidase activity was detected by chemiluminescence.
Analysis of low molecular mass organic solutes
Low molecular mass solutes were extracted from freeze-dried cell pellets with 80 % ethanol (HPLC grade, Roth, Germany) at 68 °C for 2 h. For GC analysis, a defined amount of sorbitol was added as an internal standard. The extracts were centrifuged (13,000g, 5 min, 20 °C) and the supernatant was lyophilized. The dry extract was resuspended in 500 μl ethanol (99 % HPLC grade, Roth, Germany) and centrifuged. The subsequent supernatant was again dried and then resuspended in 500 μl deionized water (HPLC grade, Carl Roth, Karlsruhe, Germany). After drying, the final extract was dissolved in pyridine, silylated, and analyzed by gas chromatography (GC) according to Hagemann et al. .
GC—MS analyses of isoprene
GC—MS analyses were performed using a GC—MS-QP 5000 (Shimadzu) comprising a Tri Plus auto-sampler. Analytes were ionized by an electrospray ionization (ESI) system, which operated in electron impact mode with ionization energy of 70 eV. Helium gas (99.999 %) was used as a carrier gas at a constant flow rate of 75 ml/min, and an injection volume of 0.5 μl was employed (split-injection). The injector temperature was maintained at 150 °C, the ion-source temperature was 180 °C, and the oven temperature was programmed from 135 °C (isothermal). Mass spectra were taken at 70 eV in a full scan mode and for fragments from 50 to 280 m/z. The mass-detector used in this analysis was Turbo-Mass Gold-Perkin-Elmer, and the software used to handle mass spectra and chromatograms was a GC—MS solution system 1.2.
Single photon ionization time-of-flight mass spectrometry (SPI-MS)
SPI-MS has already been shown to be well suited for the fast, time resolved, online analysis of coffee roasting products [69, 70], cigarette smoke [71, 72], and waste incineration plant fumes [73, 74]. For isoprene production studies using the SPI-MS, cultures were pre-cultivated at high CO2 to an optical density (OD750) of approximately two. Then, cultures were supplemented with 50 mM NaHCO3 and shifted to different culturing conditions [dark, high light, salt (NaCl) etc.]. The cultures were maintained in hybridization vessels (Glasgerätebau Ochs GmbH, Bovenden-Lenglern) equipped with silicon septa at an ambient temperature of 30 °C. By using deactivated gas chromatography capillaries (TSP-fused silica deactivated with DPTMDS, ID 150 μm, OD 375 μm; BGB, Rheinfelden), a stream of compressed air with a constant flow rate of 10 ml/min was maintained. The sample inlet was a metal capillary (Hydroguard MXT, ID 0.28 mm; Restek, Bad Homburg) placed in the center of the septum as well as the upper part of the cultivation vessel gas compartment. The capillary ran through a heatable transfer line (length 2.0 m), which was constantly heated to 220 °C. Its end was aligned with the tip of an also heated, hollow, stainless steel needle, which was pointed to the center of the ion source.
For ionization, UV light was generated by frequency tripling of the given Nd:YAG laser (Surelite III, Continuum, Santa Clara, USA) signal (wavelength 1064 nm, pulse duration 5 ns, repetition rate 10 Hz). As a consequence of repeated frequency tripling of the UV laser pulse within a xenon-filled collision cell, VUV photons with a wavelength of exactly 118 nm, which is equivalent to energy of 10.49 eV, were generated. For a detailed description of the formation process, see Mühlberger et al. . The given photons are transferred to the ionization chamber, focused on the inlet needle and absorbed by gaseous (analyte) molecules. When the ionization energy (IE) of these is exceeded, ions are produced. Therefore, all species with an IE below 10.49 eV, most organic compounds, are accessible, and as a positive side effect, signals originating from matrices, such as oxygen (IE 12.06 eV), nitrogen (IE 15.58 eV), or water (IE 12.62 eV), are suppressed. Transferring only low excess energy, the soft ionization process, leads to an inhibited fragmentation and less complex spectra and facilitates rapid data interpretation.
After ionization, a time-of-flight mass analyzer, which is capable of separating a large amount of ions in very short time intervals, is the next step. In principle, the separation is based on temporal differences of ions with various m/z values traveling along a field-free drift path, from ion source to detector. Therefore, ions are accelerated and equipped with a specific amount of kinetic energy. Depending on their m/z as well as the resulting velocities, the ions reach the detector at different times. In this case, the use of a reflector TOF analyzer additionally enhances the mass resolution due to a temporal focusing of ions with different kinetic energies. The detection unit is represented by a microchannel plate (MCP, 1.6–1.65 kV). The detailed experimental setup was described elsewhere .
Data acquisition was carried out by a LabVIEW routine (National Instruments, Austin, USA) based on custom-written software , whereby the spectra were recorded by two transient recorder cards (DP 210, Aquiris, Switzerland) with different gain settings and a signal resolution of eight bit. The processing was also performed by a LabVIEW routine, customized by Photonion GmbH (Schwerin, Germany). In particular, the data from both recorder cards were merged, while the threshold was set to 0.0006 up to 0.02 V, depending on the signal of a single ion event and the noise level. For converting the independent dimension ‘flight time’ into the crucial variable ‘m/z’, a standard gas mixture of 1,3-butadiene (concentration 10.20 ppm), acetone (9.58 ppm), isoprene (11.50 ppm), and styrol (9.69 ppm) from Linde (Oberschleißheim) was used. For each standard gas measurement, 150 successive single laser shots, in this case 150 spectra, were recorded and averaged, which equals a duration of 15 s. Using the known m/z for 1,3-butadiene and styrol as well as the resulting mass spectra, the flight time was transformed and the spectrum mass calibrated, respectively. Depending on the expected measuring time and data amount for each isoprene sample analysis, the number of recorded spectra was adjusted by presetting the average number for raw data recording (one stored spectra per 0.1 s up to per 10 s). The isoprene signal (m/z 68) was extracted from raw data as the peak area (a.u.) per given time period. For medium blank values (BG11), net 150 s and for culture samples, net 1500 s were averaged. For quantification, the resulting values were determined relative to those of the standard gas measurements (11.5 ppm).
Metabolomic profiling analysis
Cyanobacteria were grown in liquid media in a closed flask system, in the presence of 50 mM NaHCO3. After 24 h, 10 ml of cells (OD750 of approximately 1.5) was harvested by fast filtration in the light and immediately frozen in liquid nitrogen. Metabolite profiles were determined by gas chromatography electron ionization time-of-flight mass spectrometry (GC-EI-TOF-MS) as described previously [33, 78, 79]. The extraction protocol was slightly modified to enable comparison of high and low salt (NaCl) samples. Frozen samples were incubated in 630 μl of precooled methanol and extracted for 1 h at 4 °C with a final 15 min extraction at 70 °C. After centrifugation, 500 µl of extract was transferred into a new microfuge tube, and 200 µl of chloroform and 200 µl of diethylamine were added. After a 5 min incubation at 37 °C, 500 µl of water was added for phase separation. After phase separation by centrifugation, 600 µl of the upper aqueous phase was dried in a speed vacuum concentrator and further processed for GC—MS measurements as was described previously [33, 78, 79]. Metabolite responses were calculated and normalized to an internal standard, U-13C-sorbitol, and biomass using the optical density at 750 nm (OD750) of each sample [33, 79]. In this study, relative changes of metabolite pools were routinely assessed as response ratios, i.e., as x-fold changes of metabolite pools of isoprene producers in comparison to the WT pools. All experiments were repeated using three independent cell cultures.
The means of biological repeats, standard errors, and heteroscedastic Student’s t test were calculated using Microsoft Excel. One-way analysis of variance (ANOVA) was performed using multi-experiment viewer software, MeV (Version 4.6.2; http://www.tm4.org/mev/; ).