Purple phototrophic bacteria are outcompeted by aerobic heterotrophs in the presence of oxygen

Highlights

• Photoheterotrophic growth was suppressed in the presence of oxygen.

• PPB were outcompeted by aerobic heterotrophs under aerobic conditions.

• The out-competition is explained by lower aerobic growth rates of PPB.

• PPB are resilient to oxygen due to their chemoheterotrophic capabilities.

Abstract

There is an ongoing debate around the effect of microaerobic/aerobic conditions on the wastewater treatment performance and stability of enriched purple phototrophic bacteria (PPB) cultures. It is well known that oxygen-induced oxidative conditions inhibit the synthesis of light harvesting complexes, required for photoheterotrophy. However, in applied research, several publications have reported efficient wastewater treatment at high dissolved oxygen (DO) levels. This study evaluated the impact of different DO concentrations (0-0.25 mg·L⁻¹, 0-0.5 mg·L⁻¹ and 0-4.5 mg·L⁻¹) on the COD, nitrogen and phosphorus removal performances, the biomass yields, and the final microbial communities of PPB-enriched cultures, treating real wastewaters (domestic and poultry processing wastewater). The results show that the presence of oxygen suppressed photoheterotrophic growth, which led to a complete pigment and colour loss in a matter of 20-30 h after starting the batch. Under aerobic conditions, chemoheterotrophy was the dominant catabolic pathway, with wastewater treatment performances similar to those achieved in common aerobic reactors, rather than those corresponding to phototrophic systems (i.e. considerable total COD decrease (45-57% aerobically vs. ± 10% anaerobically). This includes faster consumption of COD and nutrients, lower nutrient removal efficiencies (50-58% vs. 72-99% for NH₄⁺-N), lower COD:N:P substrate ratios (100:4.5-5.0:0.4-0.8 vs. 100:6.7-12:0.9-1.2), and lower apparent biomass yields (0.15-0.31 vs. 0.8-1.2 g COD_biomass·g COD_removed⁻¹)). The suppression of photoheterotrophy inevitably resulted in a reduction of the relative PPB abundances in all the aerated tests (below 20% at the end of the tests), as PPB lost their main competitive advantage against competing aerobic heterotrophic microbes. This was explained by the lower aerobic PPB growth rates (2.4 d⁻¹ at 35 °C) when compared to common growth rates for aerobic heterotrophs (6.0 d⁻¹ at 20 °C). Therefore, PPB effectively outcompete other microbes under illuminated-anaerobic conditions, but not under aerobic or even micro-aerobic conditions, as shown by continuously aerated tests controlled at undetectable DO levels. While their aerobic heterotrophic capabilities provide some resilience, at non-sterile conditions PPB cannot dominate when growing chemoheterotrophically, and will be outcompeted.

Introduction

Purple phototrophic bacteria (PPB) have been proposed as an interesting mediator to transform wastewater constituents into potentially valuable products, such as hydrogen, polyhydroxyalkanoates, carotenoids, and the biomass itself (e.g. single-cell protein (SCP) or fertilizer) (Capson-Tojo et al., 2020; Winkler and Straka, 2019). PPB perform anoxygenic photosynthesis (as opposed to oxygenic photosynthesis) to generate energy, which enables the assimilation of pollutants from waste streams via photoheterotrophic growth (using organic compounds as carbon source), at biomass yields around 1.0 g COD·g COD_removed⁻¹ (Hülsen et al., 2018b, 2018a, 2014; Winkler and Straka, 2019). PPB can be applied for simultaneous secondary and tertiary treatment, partitioning soluble organics, nitrogen (N) and phosphorus (P) into suspended, aggregated (flocculant or granular) or attached (biofilm) biomass. A recent review of 177 studies reported median COD, N and P removal efficiencies under illuminated-anaerobic conditions (in various configurations), of 76%, 53% and 58% respectively, with optimal COD:N:P uptake ratios around 100:6-10:1-2 (Capson-Tojo et al., 2020; Lu et al., 2019b; Puyol et al., 2017). Therefore, PPB-based processes represent a promising novel alternative for resource recovery from waste streams.

During anoxygenic photosynthesis, light energy is collected via light harvesting complexes (LHC), containing various carotenoids and bacteriochlorophylls (BChl). The harvested light energy is further transformed into chemical energy in reaction centres. BChls absorb wavelengths in the near infrared (NIR) range (780-890 nm for BChl a, and 970-1040 nm for BChl b (with common maximum absorption peaks at 800 nm, 850 nm or 1010 nm), while carotenoids absorb mostly fractions of visible (VIS) light (e.g., 500 nm for lycopene or 550-555 nm for spirilloxanthin) (Canniffe and Hunter, 2014; Niedzwiedzki et al., 2015; Okubo et al., 2006; Saer and Blankenship, 2017; Wang et al., 2012)). Besides VIS-light harvesting (with further transfer of excited electrons to BChls and to reaction centres for intracellular energy generation), carotenoids also fulfil photoprotective functions (Frank and Polívka, 2009; Hartigan et al., 2002; Saer and Blankenship, 2017). These carotenoids give phototrophically grown PPB cultures their very characteristic “purple” colour.

Other than phototrophic growth, PPB can also grow without light, via fermentation (anaerobic), or chemoheterotrophic respiration (aerobically, or performing denitrification) (Capson-Tojo et al., 2020). In the presence of oxygen, energy for cellular metabolism is mostly derived from aerobic respiratory chains and virtually no BChl or carotenoids are synthesized (Zeilstra-Ryalls et al., 1998). This is because the oxidative conditions resulting from the prolonged presence of oxygen inhibit the expression of most of the genes coding for the light-harvesting antenna (responsible for energy capture) and reaction centre complexes (the main components of photosystems and responsible for energy transformation via electrical charge separation) (Bauer et al., 2003; Gregor and Klug, 1999; Sganga and Bauer, 1992; Zhu et al., 1986). Using Rhodobacter capsulatus as model organism, it has been found that, in the presence of oxygen, a regulator known as CrtJ responds to oxidising growth conditions by oxidising a molecular HS-HS bond to an S-S bond, which stimulates the DNA binding activity of the repressor (Bauer et al., 2003). This process is responsible for the repression of photosynthesis gene expression, with a redox pair potential of -270 mV (Madigan et al., 2011). It has been hypothesized that repressing gene expression is a protection mechanism of the cells to avoid photooxidative damage (Zhu et al., 1986). This suppression results in the inhibition of photoheterotrophy and the shift towards chemoheterotrophy, which results in reduced biomass yields and COD:N:P substrate ratios (Izu et al., 2001; Jiao et al., 2003; Siefert et al., 1978; Yue et al., 2015). Both quantitative studies and theoretical analyses have confirmed this phenomenon under micro-aerobic conditions, with oxidative phosphorylation, rather than cyclic photophosphorylation, being the predominant ATP-production route in PPB under these conditions (Lu et al., 2011a, 2011b). This is in agreement with studies showing that PPB require an oxidation reduction potential (ORP) below -200 mV for efficient photoheterotrophic growth (Ormerod, 1983; Siefert et al., 1978), which was further supported by modelling PPB addition to activated sludge reactors (Huang et al., 2001). The latter concluded that PPB do not synthesize pigments under aerobic conditions (Huang et al., 2001). Some PPB can survive naturally in media where oxidative conditions are prevalent (such as marine sediments), but even in those situations, it is widely accepted that PPB live in niches where reducing conditions (low in oxygen) are available (Hiraishi et al., 2020; Madigan et al., 2011; Vethanayagam, 1991).

Due to the aforementioned reasons, most research applying PPB for resource recovery has been carried out under anaerobic, illuminated conditions. Nevertheless, a number of articles have suggested that photoheterotrophic growth occurred under illuminated, microaerobic (0.5-1.0 mg DO·L⁻¹; dissolved oxygen), and even aerobic conditions (2.0-8.0 mg DO·L⁻¹). This was suggested by either directly claiming that photophosphorylation occurred in the presence of oxygen (Meng et al., 2017; Peng et al., 2018), by reporting optimal PPB growth conditions under simultaneous light and oxygen supply (Lu et al., 2019a; Yang et al., 2018c; Zhou et al., 2016, 2015), or by reporting effective carotenoid and BChl synthesis under aerobic conditions (Lu et al., 2019a; Meng et al., 2020, 2017; Yang et al., 2018b; Zhi et al., 2020). Conversely, previous studies from the same research group reported oxidative phosphorylation as main ATP generation process in the presence of oxygen, which inherently points towards aerobic chemoheterotrophy, rather than photoheterotrophy (Lu et al., 2013, 2011a, 2011b; Wang et al., 2016). This was further supported by the generally lower biomass yields reported under aerobic conditions (0.19-0.51 g COD·g COD_removed⁻¹ (Lu et al., 2011b; Meng et al., 2017; Wu et al., 2012)), which are far from those achieved photoheterotrophically (up to 1.0 g COD·g COD_removed⁻¹ (Hülsen et al., 2018b, 2018a, 2014; Winkler and Straka, 2019)). Low yields point towards the consumption of COD for ATP generation (e.g. via oxidative phosphorylation). In addition, several studies have reported faster COD removal rates aerobically when compared to their anaerobically grown PPB (both illuminated) (Lu et al., 2019a, 2013, 2011b; Meng et al., 2020, 2017; Wang et al., 2016; Yang et al., 2018b; Zhao and Zhang, 2014). This further points towards chemoheterotrophic growth under aerobic conditions (not necessarily of PPB in mixed cultures), as growth rates for common aerobic heterotrophs are higher when compared to those of PPB growing photoheterotrophically (maximum specific growth rate for heterotrophic biomass of 6.0 d⁻¹ at 20 °C (uptake rate of 8.95 d⁻¹ assuming a yield of 0.67 g COD·g COD⁻¹) (Henze et al., 1987) vs. photoheterotrophic acetic acid specific uptake rate by PPB of 2.4 g COD·g COD⁻¹·d⁻¹ (20 °C) (Puyol et al., 2017)). The aerobic uptakes rates via chemoheterotrophy for PPB have been recently determined using different pure cultures (i.e. Rhodopseudomonas palustris, Rhodospirillum rubrum, Rhodobacter sphaeroides and Rhodobacter capsulatus), obtaining values of 3.2-8.0 d⁻¹ at 28°C with fructose as substrate (growth rates of 1.6-4.4 d⁻¹), which are also higher than those achieved photoheterotrophically (Alloul et al., 2021). Considering the information presented above, it is unclear how anaerobic photoheterotrophy can be performed under aerobic conditions and whether anaerobic photoheterotrophy and aerobic chemoheterotrophy take place simultaneously or are mutually exclusive.

It is also unclear how the competition of PPB with aerobic heterotrophs would play out under aerobic, non-sterile conditions, treating real wastewater. Under anaerobic, NIR-illuminated conditions, the faster growth rates compared to common anaerobic VFA-degrading microbes provide a selective advantage for PPB (particularly in the case of acetate consumption photoheterotrophically compared to methanogenesis). This is not the case under aerobic conditions. In fact, PPB have been reported to grow slower via respiration than common aerobic heterotrophs (Izu et al., 2001) and PPB are therefore likely to be outcompeted over time (Hülsen et al., 2019; Izu et al., 2001; Siefert et al., 1978). This is indicated by the very low abundances of PPB in activated sludge systems (Siefert et al., 1978; Zhang et al., 2003). It is noteworthy that most studies reporting positive effects of (high) DO levels have never considered this competition, as tests have mostly been carried out using pure/isolated cultures fed with artificial/sterile media (Liu et al., 2016, 2015; Lu et al., 2019a, 2013, 2011b; Wang et al., 2016; Wu et al., 2019, 2012; Zhou et al., 2014).

This is exacerbated by the general absence of microbial data in most studies claiming a positive effect of oxygen on the PPB treatment performance (particularly at the end in batch studies; t_end). Among the studies reporting the microbial communities, batch tests (with durations of 3-4 days) have shown that aerobic heterotrophs (e.g. Pseudomonas or Hyphomonas) are present at considerable proportions at t_end under illuminated-aerated conditions (Yang et al., 2018c, 2018a; Zhou et al., 2015), and that the PPB diversity is reduced (Zhao and Zhang, 2014). In addition, the presence of strict anaerobes at the end of batches suggests that anaerobic regions might have existed in the reactors despite aeration (Peng et al., 2018; Yang et al., 2018c). None of the aforementioned studies reported high proportions of bacteria commonly found in PPB-mediated reactors (e.g. Rhodobacter or Rhodopseudomonas (Capson-Tojo et al., 2020)), which suggests that they are not dominant or cannot compete when oxygen is supplied. Interestingly, none of these studies reported the presence of pigment-producing aerobic anoxygenic phototrophs (i.e. Erythrobacter, Roseobacter, Roseovarius, Rubriomonas or Roseivivax; Oren (2011)) when oxygen was provided.

In addition to obvious scientific reasons, this lack of agreement on the effect of oxygen on the performance of PPB mediated treatment systems, and its importance for resource recovery, is clearly relevant for practical application. On one hand, PPBs might be outcompeted in a matter of days. On the other, a non-detrimental effect of DO (and oxidative conditions) could expand the application spectrum of PPB (e.g. for in-situ aquaculture resource recovery or for recovery form high DO streams). In addition, this is highly relevant when considering the growth of PPB in open pond systems, with large surfaces for oxygen transfer, as an excessive oxygen concentration (e.g. due to diffusion or excessive mixing using paddle-wheels) might lead to oxidative environments if the organic matter concentration in the liquid cannot maintain reducing conditions.

This article aims to shed light on the effect of oxygen on the treatment performance of PPB-based systems, using enriched cultures and non-sterile, real wastewaters. Particular attention was paid to the competition between PPB and common aerobic heterotrophs. A number of batch tests at different controlled DO ranges were carried out to determine the oxygen uptake rates (OURs) and the maximum DO concentrations that would allow efficient photoheterotrophic growth. A flat plate reactor was also used to study the effect of air supply at constant flow rates, to determine the lowest air flows at which photoheterotrophic growth is suppressed. The COD and nutrient recoveries, the absorption spectra (for carotenoids and BChls assessment) and the microbial populations were studied to elucidate the prevalent metabolic routes in each condition.