Competitive growth experiments with a high-lipid Chlamydomonas reinhardtii mutant strain and its wild-type to predict industrial and ecological risks

Research into microalgal biotechnology has progressed considerably in the last few decades, with successful commercial examples of large-scale production in outdoor raceway ponds (Borowitzka and Moheimani 2013). Attempts are being made to increase productivity through process optimisation and, increasingly, through strain manipulation, which also enables a diversification in the portfolio of compounds that can be produced. One way to improve productivity of microalgal strains for biotechnology is through random mutagenesis, followed by screening for desired characteristics (Guarnieri and Pienkos 2015; Work et al. 2010). Although mutagenesis and screening can be time consuming and costly, it can improve metabolite production several fold. One species where this has been successfully demonstrated is Chlamydomonas reinhardtii (Li et al. 2010).

C. reinhardtii is a popular model organism with a fully sequenced and well-annotated genome (Merchant et al. 2007). Moreover, there are also powerful molecular tools available to modify and characterise its genome (Jinkerson and Jonikas 2015). It is, therefore, commonly used in laboratory scale microalgal biotechnology research and development. Like other microalgal species, C. reinhardtii responds to stress such as nitrogen deprivation (Siaut et al. 2011) with an accumulation of reserve molecules (e.g. lipid, starch). Recently, studies have found that inhibiting starch synthesis in C. reinhardtii, via mutagenesis, can divert the carbon flow towards an increase in intracellular triglycerides (TAGs) (Li et al. 2010; Work et al. 2010). These mutants can accumulate up to 3.5 times more TAGs than the wild-type (WT) strain under nutrient stress conditions (Li et al. 2010). These are promising advances, particularly in the field of microalgal biodiesel research, where process economics remain unfavourable and hence productivity increases through technological advances are keenly sought. However, to scale up production of TAGs through laboratory generated microalgal strains, such as the C. reinhardtii low-starch mutant, there are many challenges that need to be overcome, including appropriate strain cultivation strategies, efficient implementation of nutrient deprivation stress as well as low cost harvesting.

Here two risks associated with large-scale cultivation of microalgal mutant strains are assessed by undertaking a series of competitive growth and productivity experiments. The first is defined as an environmental risk, with the release of genetically altered organisms, which could impact on the biodiversity of the surrounding environment by outcompeting native strains (Snow and Smith 2012). This could further influence the community structure of aquatic ecosystems if the strain generates direct or indirect effects on microbes, grazers or higher trophic levels (invertebrate and vertebrate), ultimately influencing the probability of algae blooms or altered microbial community function (Flynn et al. 2013). Currently, there are no recommended laboratory protocols that can be used to infer or guide the level of risk associated with the escape of specific mutant algal strains into the environment. However, there are environmental risk assessments being generated which can be used for guidance, for example, decision trees asking questions such as: Does the mutant strain survive outside the bioreactor/raceway pond? Does it compete or interact with the WT? (Jeremy Sweet, personal communication).

In addition to the environmental risk, there is also an industrial risk, where the benefit of reduced starch synthesis (Li et al. 2010; Zabawinski et al. 2001) is lost during the cultivation process, and hence the enhanced TAG accumulation during nutrient stress is not realised. This could occur through natural mutation in the genes altered by random mutagenesis, or in the case of a deletion mutation, mating with an invasive WT strain. This could lead to a strain with TAG productivity characteristics similar to the WT dominating the outdoor raceway pond, hence, reducing overall TAG productivity. Moreover, raceway contamination through invasion by a local microalgal species, with lower TAG accumulation levels, is also an industrial risk.

Such risks would benefit from a strategic investigation of competitive dynamics where the fitness of modified or mutant strains are determined against their WT as a baseline. Although conducted in the laboratory, at small scale and in less complex conditions relative to the environment, it can provide key clues to estimate potential risks. If the fitness of the transgenic strain is relatively lower than that of the WT, then it can be argued that any environmental risk analysis could be based solely on the risks of cultivating the WT (Gressel et al. 2013). In this scenario, the industrial risk of reduced productivity, due to WT overtaking the transgenic strain, would need to be estimated.

To further this objective, the competitive outcome between cultivating two C. reinhardtii strains, the CC-124 strain (hereon referred to as WT strain) and the starchless CC-4333 strain (Zabawinski et al. 2001), (hereon referred to as mutant strain) was explored. An initial monoculture assessment of growth was undertaken followed by competing the strains against each other. A response surface competition experiment was developed, varying both the initial proportion of the WT and mutant strains, together with the nitrogen (ammonium chloride, NH₄Cl) concentration in the media. An assessment of both environmental risk during an “escape scenario” and the industrial risk associated with being potentially out competed by the WT strain was made using this design. Subsequently, the industrial risk was further analysed by assessing total biomass accumulation and TAG productivity, again in response to varying the initial proportion of the strains and NH₄Cl concentration.

The fact that the mutant cells, in exponential phase, had a significantly higher μ of 4.08 ± 0.52 day⁻¹ (p = 0.008) (Fig. 1) would suggest that the mutant strain could outcompete the WT strain during an escape scenario presenting an environmental risk. It also hints that there would be a reduced industrial risk from the mutant strain being outcompeted by cells that have reverted to WT lipid accumulation characteristics. In a previous study that compared growth and lipid accumulation in a C. reinhardtii WT and the same mutant strain as this study (Work et al. 2010), the highest cell concentration and growth rate were achieved in the mutant strain, and this was attributed to its smaller cell diameter compared to the WT. This observation was confirmed in our study (Additional file 1: Fig. A2) and previously where it is has been demonstrated that cell size and growth rates are negatively correlated (Schlesinger et al. 1981). It was also observed that there is a slower transition from lag phase to exponential phase with the mutant strain compared to the WT. This difference could be detrimental to the survival of the mutant strain when grown in co-culture with the WT strain. It should be noted that this study was performed in a closed system with uniform resource conditions, without the presence of predators (e.g. zooplankton), symbiotic and antagonistic organisms and without fluctuating environmental conditions. Any of these factors that would favour the mutant or WT strain could create an unbalance and allow one to dominate the other.

In order to decipher which of the predicted scenarios from interpreting the monoculture results above were most likely to take place, co-culture experiments were undertaken where equal initial amounts of WT and mutant strain were competed (50:50), followed by a dominating mutant (25:75) and dominating WT (75:25) experiment. Pure cultures were cultivated as controls and the concentration of nitrogen was also varied (NH₄Cl: 50, 100, 200 and 375 mg L⁻¹). Recent reports have hypothesised that genetic mutations of biotechnological interest (e.g. leading to increased lipid content) are all in the opposite direction of evolution (i.e. optimal growth under natural conditions) and therefore should be outcompeted in growth by the respective WT strains (Flynn et al. 2010; Gressel et al. 2013). The results in this study support this theory. Our results showed that when the WT and mutant were mixed, the overall fitness of the mutant strain is inferior to that of its WT (Fig. 2). Work et al. (2010) suggest several reasons why this would happen: under nitrogen deprivation, starchless mutants metabolise less acetate and have severely attenuated levels of photosynthetic O₂ evolution compared to the WT. This indicates that the mutant strain responds to nitrogen deprivation by decreasing its overall anabolic processes (Work et al. 2010). If the cells in the lag phase of a mixed culture are under similar stress, it could explain why the WT dominates over the mutant in every experimental scenario tested.

Although the competition results showed a clear difference between the WT and mutant, this study was done in a closed system with uniform resource conditions, without the presence of predators (e.g. zooplankton), symbiotic and antagonistic organisms and without fluctuating environmental conditions. Any of these factors that would favour the mutant strain could create an unbalance and allow it to dominate over its WT. For example, in a natural system, differences in grazer palatability can have a dramatic effect on the competitive outcome. Van Donk (1997) performed grazing experiments to determine the role of cell wall structure and nutrient limitation on the digestibility of the C. reinhardtii WT and cell wall-deficient mutant. The study revealed that under nutrient limiting conditions, Daphnia magna clearance rates of C. reinhardtii WT cells were severely debilitated. In contrast, under both non-limiting and limiting nutrient conditions, C. reinhardtii mutant cells were cleared equally (Van Donk 1997). This further supports our prediction that mutant cells remain unlikely to outcompete their WT strain and much less so under nutrient stress conditions necessary to take advantage of the mutation.

Industrially, both the data from this study and Van Donk (1997) suggests that outdoor large-scale cultivation of cell wall-deficient mutant strains has many associated economic risks of failure to produce effective quantities of product. A reversion from the mutant strain to its WT, contamination of the mutant strain culture by its WT or invasion by grazers would result in rapid reduction or disappearance of the mutant. Therefore, while industrial outdoor cultivation of cell wall-deficient mutant strains does not present a significant environmental risk to its WT in an escape scenario, the fragility of the mutant strain could impact on overall industrial productivity. The WT strain would most likely outcompete the mutant in competitive cultures sharing the same resources.

In order to further investigate the level of industrial risk, productivity from a total biomass and total TAG accumulation perspective was quantified during co-culture experiments, over a range of nitrogen concentrations. A RSM was fit to data on the maximum biomass (K, estimated from each of the three replicate time series) and TAG concentrations were calculated, per cell, during the stationary phase to assess biomass and TAG accumulation over time as a function of NH₄Cl and initial WT:mutant cell number ratio. In accordance with the monoculture experiments (Fig. 1), the maximum K was observed when the mutant strain was grown alone at higher NH₄Cl concentrations. There was no significant difference between the K at 200 and 375 mg L⁻¹ NH₄Cl, which corroborated the monoculture experiments. In agreement with the extensive literature on lipid accumulation in C. reinhardtii, our study shows that in stationary phase net growth halts and accumulation of TAGs begins (e.g. Siaut et al. 2011). TAGs accumulated in nitrogen deplete conditions when biomass accumulation stopped after 72 h. To analyse this accumulation, TAGs per million cells were plotted in order to normalise TAG content with biomass production (Fig. 4). Unsurprisingly, TAG accumulation was the highest, per million cells, with the 100% mutant strain at the lowest nitrogen concentration (50 mg L⁻¹ NH₄Cl). However, the industrial risk of reduced productivity was shown in the co-cultures. Due to the success of the WT growth compared to the mutant strain, TAG productivity can be reduced up to 60% (Fig. 4).

The aim of this study was to compare the growth rates of a C. reinhardtii WT and mutant strain in monocultures and in competition (co-cultures) to evaluate the risk of the mutant strain escaping into the environment, as well as the industrial risk of losing the desirable phenotype of the mutant. The response surface design applied suggests relatively little environmental risk if the mutant escapes the cultivation vessel, but a potentially significant industrial productivity loss if a WT equivalent is introduced. Ultimately, the experimental approach is suggested as a predictor of environmental and industrial risk as the experiments were based on laboratory tests only and under specific conditions. A more comprehensive study of environmental risks would require incorporating the complexities associated with ecosystems, including increased biodiversity and changing environmental conditions.