An Organic Acid Based Counter Selection System for Cyanobacteria

Cyanobacteria are valuable organisms for studying the physiology of photosynthesis and carbon fixation, as well as metabolic engineering for the production of fuels and chemicals. This work describes a novel counter selection method for the cyanobacterium Synechococcus sp. PCC 7002 based on organic acid toxicity. The organic acids acrylate, 3-hydroxypropionate, and propionate were shown to be inhibitory towards Synechococcus sp. PCC 7002 and other cyanobacteria at low concentrations. Inhibition was overcome by a loss of function mutation in the gene acsA, which is annotated as an acetyl-CoA ligase. Loss of AcsA function was used as a basis for an acrylate counter selection method. DNA fragments of interest were inserted into the acsA locus and strains harboring the insertion were isolated on selective medium containing acrylate. This methodology was also used to introduce DNA fragments into a pseudogene, glpK. Application of this method will allow for more advanced genetics and engineering studies in Synechococcus sp. PCC 7002 including the construction of markerless gene deletions and insertions. The acrylate counter-selection could be applied to other cyanobacterial species where AcsA activity confers acrylate sensitivity (e.g. Synechocystis sp. PCC 6803).


Introduction
Current research concerning the physiology of cyanobacteria has led to an increased understanding of photosynthesis, CO 2 fixation, and metabolism [1][2][3]. In addition to basic science, researchers are also exploring the application of cyanobacteria for the production of biofuels and chemicals [4,5]. Continued advances in these areas will require more efficient methods for genetic manipulation and gene expression [6,7]. In particular, counter selection methods that are applicable to multiple strains of cyanobacteria would be useful in dealing with the issues of multiple chromosomes and limited antibiotic resistance cassettes. Currently, a sacB counter selection system was demonstrated in a glucose tolerant strain of Synechocystis sp. PCC 6803 (PCC 6803) and a mazF based system in wild type PCC 6803 [8,9]. Unfortunately, attempts to build a sacB counter selection system have been unsuccessful and mazF has not been demonstrated in Synechococcus sp. PCC 7002 (PCC 7002), another model organism that grows rapidly and is widely tolerant to light intensity, temperature and salt conditions. Due to these advantages, PCC 7002 has recently been explored for the production of α-olefins and sugars via metabolic engineering [10,11].
Since a counter selection system was not available for PCC 7002, we developed a new system based on sensitivity to organic acids. In general, organic acids can cause toxicity by depleting proton motive force, accumulating anions, and inhibiting enzymes [12][13][14]. The organic acids of interest in this work are acrylate, 3-hydroxypropionate, and propionate. Acrylate was shown to inhibit beta oxidation in Pseudomonas and Ralstonia species, 3-hydroxypropionate was shown to inhibit the chorismate pathway in Escherichia coli, and propionate was shown to inhibit gluconeogenesis via metabolism to 2-methylcitrate in Salmonella [15][16][17][18]. Additionally, exposure to propionate was shown to cause a decrease in the free Coenzyme-A (CoA) pool in a species of Synechococcus [19].
This work demonstrates that these three organic acids, acrylate, 3-hydroxypropionate, and propionate, are inhibitory to multiple species of cyanobacteria at low concentrations. This inhibition can be overcome in PCC 7002 by a single loss of function mutation in a gene encoding a CoA-ligase (acsA). These observations were used as a basis for the development of a counter selection system using acrylate sensitivity. Acrylate was chosen amongst the organic acids due to its inhibitory activity in multiple species and dramatically increased tolerance in strains without AcsA activity. This counter selection method was used to directly introduce and select for integrations into the acsA locus, as well as generate a scarless integration in a neutral locus. The ability to rapidly introduce heterologous gene constructs and make markerless deletions will advance both basic science and applied research efforts in cyanobacteria.

Chemicals and 3-hydroxypropionate quantification
Acrylate was purchased from MP Biomedicals (ICN211387). 3-hydroxypropionate was purchased from TCI America (H0297) as a ca. 3.6 mol/L solution in water. Additional chemicals and reagents were purchased from Fisher Scientific and Sigma Aldrich. The 3-hydroxypropionate solution purchased from TCI had an unknown purity, so the solution was quantified with an enzyme assay using propionyl-CoA synthase. Escherichia coli (E. coli) BL21 harboring pKS1 (expressing propionyl-CoA synthase) was grown in lysogeny broth (LB) to an OD 600 of 0.4 and induced with 0.4 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). The induced culture was incubated at 30 °C for 4 h before centrifugation. The resulting cell pellet was resuspended in twice the volume of 100 mM Tris pH 7.8 and lysed by sonication. The cell debris was removed by centrifugation at 16,000 x g for 10 min at 4 °C. The supernatant was incubated at 63 °C for 10 min and the precipitated protein was removed by centrifugation at 20,000 x g. The resulting protein fraction was concentrated 2X using an Amicon Ultra-4 centrifugation column to increase activity. The resulting enzyme was used in an assay to quantify 3-HP by NAPH reduction as described in Schneider et al [20].

Strains, culturing conditions, and media
All strains used in this study are listed in Table 1. PCC 7002, Synechococcus sp. PCC 7942 (PCC 7942), and PCC 6803 were obtained from the Pasteur Culture Collection. PCC 7002 was grown in Medium A + pH 8.0 [21]. PCC 7942 and PCC 6803 were grown in BG-11 [22]. Unless otherwise noted, liquid cultures were grown in 10 mL volumes at a light intensity of 140 µE/m 2 /s at 35 °C and bubbled with air. Agar plates were incubated under the same light and temperature conditions. Cultures used to observe the initial mutant population ( Figure  1) and harvest RNA were grown in 50 mL volumes without agitation or supplementation with air. Cultures were inoculated using patches grown from a single colony on solid media under the same conditions. Unless otherwise noted, cell growth was monitored by measuring the optical density at 730nm (OD 730 ) using a Spectrophotometer 20 (Milton Roy).

Determining minimum inhibitory concentrations
Inocula were prepared by growing cultures of wild type and mutant cyanobacteria to an OD 730 of 0.3. These cultures were used to inoculate, to an initial OD 730 of 0.03, a 96 well plate containing buffered media with increased concentrations of the compound of interest. Stocks of the compounds of interest were prepared in Medium A + and adjusted to a pH of 8.0. Wild type PCC 7002 was challenged with organic acid concentration increments of 5 µM for acrylate and 1 mM for 3hydroxypropionate and propionate. Mutant strains with increased tolerance were further challenged with concentration increments of 10 mM acrylate, 5 mM 3-hydroxypropionate, and 100 mM propionate. Plates were incubated at 35 °C at a light intensity of 140 µE/m 2 /s. OD 730 was measured periodically using a plate reader (Tecan M1000). The minimum inhibitory concentration (MIC) was determined as the concentration at which no increase in OD 730 was observed after 48 h.

Quantification of mutation frequency
Wild type PCC 7002 was grown to an OD 730 of 1.0 and serially diluted in triplicate on both unmodified and acrylatecontaining plates. Acrylate concentrations of 50 µM, 500 µM, 5 mM, and 50 mM were used. Plates were incubated for 5 days and counted for colonies. Mutation frequency is defined as the number of acrylate tolerant colony forming units, divided by the total number of colony forming units.

RNA-sequencing and analysis
Cultures of PCC 7002 were grown in 5 mM acrylate or unmodified medium A + in quadruplet to an OD 730 of 0.5-0.6. Duplicate cultures from each condition were harvested for RNA sequencing. RNA was isolated from PCC 7002 using the Trizol 95 method and a RNeasy kit (Qiagen) [23]. RNA purity was verified using an Agilent RNA 600 Nano Kit and Agilent 2100 Bioanalyzer. Ribosomal RNA was removed using the Epicentre rRNA removal kit for Gram negative bacteria (RZNB1056). RNA was sequenced by the University of Wisconsin Gene Expression Center using an Illumina HiSeq 2000. The resulting RNA sequencing reads were aligned to the PCC 7002 genome (Genbank ID CP000951.1-CP000957.1) and analyzed using CLC Genomics Workbench software. Reads per kilobase per million total reads (RPKM) was used to quantify expression values [24]. Genes identified as being differentially regulated were defined as having greater than 2 fold changes in RPKM and a P-value less than 0.05. Single nucleotide polymorphisms (SNPs) were identified using the CLC Genomics Workbench software. Genes of interest were identified as having SNPs in acrylate cultures but not control cultures. These data have been deposited in NCBI's Gene Expression Omnibus (REF) and are accessible through the GEO Series accession number GSE48981 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc=GSE48981).

Mutant construction
All plasmids used in this study are listed in Table 2. A PCC 7002 ΔacsA mutant (BPSyn_006) was constructed by double homologous recombination as previously described [25]. Briefly, a linear DNA fragment was constructed containing a streptomycin resistance marker derived from pSRA81 (aadA) flanked by approximately 600 bp regions that are homologous to regions directly 5' and 3' of the acsA (SYNPCC7002_A1838) gene. The acsA upstream and acsA downstream pieces were individually amplified using the primers in Table 3, digested, and then ligated together with the digestion fragment of pSRA81. The resulting fragment was PCR amplified and transformed into wild type PCC 7002. A similar protocol was used to replace acsA with a 40 bp barcode sequence, resulting in strain BPSyn_014. A linear fragment was created with 600 bp homologous regions and a 40 bp barcode sequence in between. This fragment was transformed into PCC 7002 and mutants were selected on plates containing 50 µM acrylate. Positive clones were streaked out on plates containing 10 mM acrylate to achieve complete chromosomal segregation. The resulting mutant had a deletion of acsA without a residual antibiotic resistance cassette. Complementation of acsA was performed through the introduction of acsA under the native acsA promoter into glpK (SYNPCC7002_A2842). Due to a frameshift mutation, glpK is a pseudogene in PCC 7002. For this reason, glpK was chosen as a neutral insertion site. Plasmids pGLPK_acsA_Sp R and pGLPK_acsAW49L_Sp R were constructed with homologous flanking regions to replace glpK with either (1) a wild type copy of acsA or (2) acsAW49L under the native acsA promoter along with a streptomycin resistance marker, using the in vitro recombination method using primers in Table 3 [26]. Plasmids pACSA_pcpcB_YFP and pGLPK_pcpcB_YFP were constructed in a similar manner using the cpcB promoter from PCC 6803 and yellow florescent protein (YFP) gene from plasmid pAQ1_Exp_YFP built by Xu et al. [27]. All cyanobacterial strains were screened by colony PCR for proper integration and full segregation of all chromosomes.

AcsA activity assay
Wildtype acsA and the acsAW49L mutant were cloned into the pET-28a plasmid (Novagen) with an N-terminal His-tag. Strains of E. coli DH10B harboring these plasmids were grown in 50 mL LB at 37 °C to an OD 600 of 0.6-0.8 and induced with 1 mM IPTG. Induced cultures were incubated at 37 °C for 3 h. The cultures were centrifuged at 3,000 x g for 15 min and the resulting cell pellets were frozen at -20 °C. BugBuster reagent (Novagen) was used to lyse cells and collect the soluble protein fraction. AcsA protein was purified from the soluble fraction using Ni-NTA His-tag affinity chromatography. Reactions containing 500 nM enzyme and 2 mM organic acid were performed at 37 °C for 6 min. Specific activity was determined by measuring the loss of free Coenzyme A over time using Ellman's reagent. Specific activity is defined as µmol free CoA min -1 µmol -1 AcsA. 3-hydroxypropionate is abbreviated to 3-HP.

Counter selection
PCC 7002 was transformed by adding 3 µg of linearized plasmid to 2 mL of culture at OD 730 1-1.5. The cultures were incubated without agitation overnight under the previously mentioned light and temperature conditions. Transformants were selected by plating both 100 µL and the remaining 1.9 mL transformation suspension, concentrated, on solid media supplemented with 50 μM acrylate. Plates were incubated in the light for 2-3 days at 35 °C. A successful transformation usually had significantly more colonies than the no-DNA, negative control. Single colonies were patched onto solid media supplemented with 50 μM acrylate. This intermediate step was used to avoid colony PCR contamination by excess DNA used in the transformation. After 1 day of growth at standard conditions, the presence of the insertion was confirmed by colony PCR. Single colonies of positive transformants were isolated on solid media containing 10 mM acrylate. Incubation for 3-5 days allowed for complete segregation of the transformants. Segregation was confirmed by colony PCR, using primers flanking the insertion site, specific to the insertion sequence, and specific to the sequence being replaced. Confirmed mutants were maintained on unmodified medium.

Florescence Detection
Cultures of PCC 7002 containing YFP were grown in triplicate to an OD 730 0.5-1. Cultures were normalized to 4 OD 730 * mL and centrifuged at 5,000 x g for 10 min. The resulting pellets were aspirated, resuspended in 300 µL BugBuster Protein Extraction Reagent (Novagen), rocked at room temperature for 30 min, and centrifuged at 16,000 x g for 25 min at 4 °C. The florescence of the resulting supernatant was measured (excitation 514 nm, emission 527 nm) using a Tecan M1000 plate reader.

Quantification of acetate
After 75 h of growth, 1 mL samples taken from triplicate cultures were centrifuged at 16,000 x g for 10 min. Each supernatant was passed through a 0.22 µm filter prior to analysis via gas chromatography and flame ionization detection using a Shimadzu GC-2010 oven with a Restek Stabilwax-DA fused silica column (11052) and AOC-20i auto-injector. The limit of detection for acetate was 10 µM.

Results and Discussion
Cyanobacteria are sensitive to several organic acids Two components are necessary for a successful counter selection method. First, a compound must be identified that is inhibitory at low concentrations. Second, this inhibition must be removed by the loss of a genetic element. Here, the toxicity, as judged by MIC, of organic acids towards PCC 7002 PCC 6803, and PCC 7942 was assessed by exposing each strain to increasing concentrations of organic acids in buffered media. The organic acids acrylate, 3-hydroxypropionate, and propionate were each inhibitory to all three cyanobacteria at relatively low concentrations (Table 4). Acrylate generated the lowest MIC for each strain (<50 µM) followed by propionate (0.25-4 mM) and 3-hydroxypropionate (2-70 mM).

Mutations increase organic acid tolerance
Wild type PCC 7002 was inoculated into medium containing 5 mM acrylate, a concentration several orders of magnitude above the MIC. The OD 730 was observed relative to control cultures for a period of 200 h. For the first 50 h of incubation, no increase in OD 730 was observed in acrylate containing cultures. After this long lag the OD 730 increased at rate similar to the control culture ( Figure 1A). Sub-culturing of the acrylate exposed cultures into fresh medium containing 5 mM acrylate resulted in no growth inhibition ( Figure 1B). Additionally, acrylate exposed cultures that were serially sub-cultured in medium without acrylate for four iterations maintained the ability to grow in an acrylate concentration above the wild type MIC. These data suggested that a mutation arose within the population that resulted in increased tolerance to acrylate. When wild type PCC 7002 was plated onto solid media containing 5 mM acrylate, mutant colonies were observed at a frequency of around 10 -6 . Mutants were observed on plates containing 500 µM and 50 µM acrylate at 5 and 10 times the frequency of 5 mM, respectively. No mutants were observed on plates containing 50 mM acrylate. This rate of mutation frequency was suggestive of a loss of function mutation [28]. Additionally, the correlation between mutation frequency and concentration of acrylate suggested that gene dosage or the presence of partial loss of function mutations existed. Interestingly, mutants obtained on acrylate containing medium were also more tolerant of 3-hydroxpropionate and propionate.

RNA-Sequencing was used to identify mutations
RNA was harvested from the cultures described in Figure  1A. Samples were processed and sequenced to identify  (Table 5) compared to cells grown in unsupplemented media. In particular, isiA and sufA were down-regulated >10-and >2-fold, respectively. This was not surprising, as acrylate was shown to act as an antioxidant in eukaryotic algae [29]. Due to the starting OD 730 (0.03) and the OD 730 at which the cultures were harvested (0.5), it was assumed that the majority of each population carried a mutation for acrylate tolerance. An analysis of single nucleotide polymorphisms (SNPs) relative to the published genome sequence (GenBank CP000951.1) was performed (Table 6). Four genes were identified with SNPs in all acrylate samples but not present in any of the control samples. The gene acsA (GenBank NC_010475.1) was of interest because some acetyl-CoA ligases have been shown to have activity towards acrylate [30]. In acrylate exposed samples, the reads spanning bp 152 of acsA, 60% (32/53) had a point mutation of a G to a T. The point mutation results in an amino acid change of Trp49 to Leu. This amino acid is part of a conserved region in the acsA of PCC 7002, PCC 6803 (GenBank AP012278.1) and Escherichia coli K12 (GenBank NP_418493.1), suggesting it is integral to a functional protein.
These results led to the hypothesis that loss of function of acsA would result in the observed increase in organic acid tolerance.

AcsA has activity towards acrylate and other organic acids
To examine the function of AcsA, His-tagged versions of the wild type gene and an acsAW49L mutant were heterologously expressed in E. coli. The wild type AcsA protein was found in the soluble protein fraction and was purified using Ni-NTA Histag affinity chromatography. Expression of the acsAW49L mutant gene generated an insoluble protein. Attempts to purify the AcsAW49L protein from the soluble fraction with Ni-NTA affinity chromatography failed. The activity of purified AcsA protein towards various organic acids was assessed in vitro. CoA ligase activity was observed for acetate, acrylate, 3-   hydroxypropionate, and propionate ( Figure 2). No activity was observed for succinate under these assay conditions (data not shown). The activity of AcsA towards these organic acids supports the results that a single mutation increases tolerance to all three compounds. While the presence of acyl-CoA thioesters may be toxic, we hypothesize that the observed toxicity of acrylate and other organic acids was caused because the CoA thioesters were metabolic dead-ends and the continuous activity of AcsA depleted the free CoA pool. This hypothesis is supported by prior work in which a species of Synechococcus grown in the presence of propionate was no longer able to import acetate and was observed to have decreased levels of free CoA [19].

Loss of function of acsA increases organic acid tolerance
A PCC 7002 acsA deletion mutant, BPSyn_006, was constructed by double homologous recombination to replace acsA with the streptomycin resistance marker aadA. BPSyn_006 had dramatically higher MIC values for acrylate, 3hydroxypropionate, and propionate (Table 4). Specifically, this mutant had 2,800-fold, 26-fold, and >100 fold increase in MIC to acrylate, 3-hydroxypropioante, and propionate, respectively. A similar observation was made in PCC 6803 by replacing acsA (sll0542) with a kanamycin resistance cassette, resulting in strain BP6803_002. BP6803_002 has a 3,500-fold increased tolerance to acrylate, but no increased tolerance to 3hydroxypropionate. This suggests that the mechanism of toxicity of 3-hydroxypropionate is different in PCC 7002 and PCC 6803. This difference between species and the order or magnitude increase in acrylate tolerance for both species makes acrylate a more attractive compound for a universal counter selection.

The organic acid tolerance phenotype was complemented in PCC 7002
Attempts to complement the PCC 7002 acsA deletion strain (BPSyn_006) by introducing and expressing acsA on the native plasmid pAQ1 or in a chromosomal loci with a kanamycin resistance cassette resulted in heterogeneous strains that could not be made homozygous, and thus could not be used in MIC determination experiments. In order to reuse the aadA cassette, a markerless deletion of the native acsA locus via counter selection and then reinsert acsA and the aadA resistance marker into a newly validated neutral insertion site. The acsA markerless deletion strain (BPSyn_014) was created by transforming wild type PCC 7002 with a linear DNA fragment to replace acsA with a 40 bp barcode ( Figure 3A). Transformants were selected on solid medium containing 50 µM acrylate, a low concentration that allows for the growth of heterozygous transformants. Mutants positive for the barcode insertion were streaked onto medium containing 10 mM acrylate. The increase in acrylate concentrations allowed for selection of homozygous clones. Single colonies were demonstrated to be homozygous via colony PCR ( Figure 3B). This strain, BPSyn_014, had MIC values identical to BPSyn_006. When compared to wild type PCC 7002, strain BPSyn_014 did not have a significant defect in growth up to 75 h ( Figure 4A), but was observed to secrete acetate (370 ± 20 µM for BPSyn_014 vs. none detected for wild type). The markerless deletion allowed for the reuse of the aadA selection marker to introduce acsA into another locus on the chromosome.
The locus glpK was chosen as a neutral insertion site for introduction of genes for complementation of the increased organic acid phenotype. In PCC 7002, glpK has a frameshift mutation and is thus considered a pseudogene. Since glpK cannot produce a functional protein, it was hypothesized that replacement of glpK would not affect physiology. Strain BPSyn_014 was complemented by introducing either (1) a wild type copy of acsA or (2) a W49L acsA mutant under the native acsA promoter along with the aadA selection cassette, into the glpK locus of the chromosome, resulting in strains BPSyn_022 and BPSyn_026, respectively. The MIC for acrylate and 3hydroxypropionate for BPSyn_022 were 10µM and 15mM, respectively, similar to wild type PCC 7002 ( Table 4). The MIC for acrylate for BPSyn_026 is 7mM, which suggests that the W49L mutant results in only a partial loss of function (Table 4). Additionally, the frequency of acrylate tolerance mutations of BYSyn_022 and BPSyn_026 observed on 50 µM acrylate plates were the same as wildtype PCC 7002.

Acrylate counter selection can be used for heterologous gene expression
To optimize the counter selection method, the 40 bp barcode sequence used to create strain BPSyn_014 and a linearized plasmid containing His-tagged yellow fluorescent protein (YFP) under the high expression cpcB constitutive promoter (pACS_ p cpcB_YFP) were transformed into wild type PCC 7002 and selected for on 50, 100, and 500 µM acrylate. After 3 days, colonies containing integrations were only observed on plates containing 50 µM acrylate. The percentage of colonies that were positive for integrations was 49% (n = 42) for the barcode and 30% for YFP (n = 39). A t-test of these data shows no  significant difference in integration frequency (p = 0.11) indicating that insertion size had little impact on integration efficiency. Colonies not containing integrations were assumed to be spontaneous mutants of acsA. The number of background colonies is consistent with the loss of function mutation frequency previously observed. A homozygous strain for the integration of YFP (BPSyn_015) was obtained by streaking onto plates containing 10 mM acrylate and confirmed by colony PCR ( Figure 3B). As expected, strain BPSyn_015 was highly fluorescent ( Figure 4B). YFP was successfully purified from BPSyn_015 using Ni-NTA affinity chromatography and visualized with PAGE. Next, the acrylate counter selection method was used to introduce heterologous genes into other loci. Strain BPSyn_022 (acsA::BC glpK::acsA aadA) was transformed with the linearized plasmid pGLPK_ p cpcB_YFP to replace acsA and aadA with YFP under the expression of the cpcB promoter in the glpK locus ( Figure 3C). 29% (n=14) of colonies were identified as positive transformants via colony PCR, which is similar to transformations into the acsA locus. Positive transformants (BPSyn_027) were verified as homozygous using the same protocol outlined for integration directly into the acsA locus ( Figure 3D). As expected, YFP was successfully purified from BPSyn_027 and visualized using Ni-NTA affinity chromatography and PAGE. When compared to BPSyn_015, no significant difference was observed in the level of YFP fluorescence ( Figure 4B). These results demonstrate that acrylate counter selection can be used to make modifications beyond the native acsA locus and that the glpK locus has utility as a neutral insertion site. The counter selection methodology is summarized in Figure 5.

Conclusion
Counter selection is an important tool for genetic manipulation of microorganisms. Counter selection is of particular importance in cyanobacterial systems due to the presence of multiple copies of the chromosome and the difficultly of achieving homozygous strains using antibiotic resistance genes. The results presented in this work describe the basis and methodology for using organic acid sensitivity for counter selection in the cyanobacterium PCC 7002. It was shown that the organic acids acrylate, 3-hydroxypropionate, and propionate cause growth inhibition and that this inhibition can be overcome by loss of the gene acsA. Acrylate was further pursued for use in counter selection because of the low concentration required for sensitivity, the dramatic increase in tolerance due to loss of acsA, and the potential to be used in other cyanobacteria (e.g. PCC 6803). A method was optimized for introducing genes of interest directly into the acsA locus. Loss of acsA did not result in a growth defect under the conditions used in this experiment, but did result in an increase in the secretion of acetate by PCC 7002. A neutral site, glpK, was identified as a chromosomal locus that is amenable to acrylate counter selection and can be used for expression of heterologous genes. These results suggest that acrylate counter selection could also be used to make markerless deletions or insertions elsewhere on the chromosome. Additionally, a loss of function of acsA in PCC 6803 was shown to dramatically increase the tolerance to acrylate, suggesting that this method has utility in multiple species. Application of the method presented in this work will be used in future physiology studies and metabolic engineering efforts.