High-CO2 Requirement as a Mechanism for the Containment of Genetically Modified Cyanobacteria

Abstract

As researchers engineer cyanobacteria for biotechnological applications, we must consider potential environmental release of these organisms. Previous theoretical work has considered cyanobacterial containment through elimination of the CO₂-concentrating mechanism (CCM) to impose a high-CO₂ requirement (HCR), which could be provided in the cultivation environment but not in the surroundings. In this work, we experimentally implemented an HCR containment mechanism in Synechococcus sp. strain PCC7002 (PCC7002) through deletion of carboxysome shell proteins and showed that this mechanism contained cyanobacteria in a 5% CO₂ environment. We considered escape through horizontal gene transfer (HGT) and reduced the risk of HGT escape by deleting competence genes. We showed that the HCR containment mechanism did not negatively impact the performance of a strain of PCC7002 engineered for L-lactate production. We showed through coculture experiments of HCR strains with ccm-containing strains that this HCR mechanism reduced the frequency of escape below the NIH recommended limit for recombinant organisms of one escape event in 10⁸ CFU.

Keywords: CO2-concentrating mechanism; biocontainment; carboxysome; cyanobacteria; horizontal gene transfer; natural competence.

Figure 1

Growth of Wild Type PCC7002 (circles) and PCC7002 ΔccmK2K1LMN (triangles) at varying $P_{{\mathrm{CO}}_2}$. Data is given as the average specific growth rate of three replicates (error bars smaller than the data symbols). Values in parentheses are the best fit parameters of the model in eq 1 and are represented by the solid curves.

Figure 2

RuBisCO localization in CCM mutants of PCC7002. Pigment fluorescence represents the shape of the cell. sfGFP fluorescence shows the distribution of RbcL-sfGFP in the cell. Wild Type PCC7002 is included to show minimal background fluorescence. The final column reports growth on Media A plates in ambient air.

Figure 3

Transformation efficiency of Wild Type PCC7002 and competence gene deletion mutants. (a) To determine transformation efficiency, samples of culture were incubated for 24 h in the light with 500 ng of plasmid DNA containing a gentamicin resistance gene (GmR) with regions of homology to PCC7002’s glpK locus. Dilutions of these samples were then plated on Media A with or without gentamicin and grown for 4 days before colonies were counted to determine transformation efficiency. (b) Colony Forming Units (CFU) mL⁻¹ recovered for transformation samples plated on Media A with (Gray) or without (black) gentamicin. Error bars represent the standard error of at least three biological replicates. For PCC7002 ΔcomEC, no colonies were recovered, so the transformation efficiency is reported as less than the inverse of the CFU plated on gentamicin.

Figure 4

HCR containment mechanism has no effect on L-lactate productivity. Wild Type PCC7002, PCC7002 clac143-ldh*, and PCC7002 clac143-ldh* ΔccmK2K1LMN ΔA1643 were grown in bubble tubes with 5% CO₂ to determine the impact of the HCR containment mechanism on growth (a) and L-lactate productivity (b). The dashed lines represent the linear growth rate (LGR) and linear production rate of L-lactate (LPR) reported in (c). Error bars and values in parentheses represent the standard error of the mean of three biological replicates. ρ is the growth associated productivity calculated from the ratio of LPR to LGR. ^a[mg DW L⁻¹ h⁻¹]. ^b[μM L-lactate h⁻¹]. ^c[mmol L-lactate (g DW)⁻¹].

Figure 5

HGT in simulated contamination experiments. Cultures containing the listed strains of PCC7002 were grown in liquid culture and then plated on Media A with gentamicin in ambient air to measure the frequency of the pictured HGT events. Monocultures of each strain were evaluated for their spontaneous escape rate, all of which were near or below the limit of detection for the assay (a–c). The escape rate of 7002_RLC28 (d) or 7002_RLC29 (e) cocultured with 7002_AM173 was increased by at least 3 orders of magnitude above the spontaneous escape rate. Deleting A1643 in the contamination screen decreased the escape frequency below the detectable limit when cocultured with either 7002_RLC28 (f) or 7002_RLC29 (g). Red boxes indicate experiments where the escape frequency was higher than the NIH recommended limit of 10⁻⁸ escapees per CFU. Green boxes indicate experiments where the escape frequency was lower than this limit. Escape frequencies reported are the mean of at least three biological replicates (except for (a) where n = 2). Standard errors of the mean frequencies were of the same order of magnitude as the mean frequencies (Supporting Information). Experiments in which no escapees were detected were reported as having a frequency less than the inverse of the number of CFU plated (b,c,f,g).