Biodegradable polymers, such as polyhydroxyalkanoates (PHAs), have been identified as potential alternatives to traditional plastics, which have been a heavy burden to the environment due to their recalcitrant nature (Kalia et al., 2021). Depending on the carbon atoms per monomer, PHAs can be classified as short-chain-length PHAs (scl-PHAs, C₃–C₅) and medium-chain-length PHAs (mcl-PHAs, C₆–C₁₄) (Li et al., 2016). PHAs are accumulated by many bacteria as a carbon- and energy- storage compound, and the most accumulated PHA is PHB, a polymer composed of 3-hydroxybutyrate, belonging to scl-PHAs (Lee, 1996). Although PHB is the most studied PHA in literature, the bioplastic products derived solely from it are typically brittle and stiff, which are undesirable in many applications (McAdam et al., 2020). Nonetheless, the material properties could be improved by adding 3-hydroxyvalerate units into a PHB biopolymer, obtaining copolymer poly(3-hydroxybutyrate-co-3-hydroxyvalerate), also named PHBV (Li et al., 2016). Due to their inherent biodegradability, excellent biocompatibility and non-toxicity, PHAs have been widely applied in the fields of agriculture, aquaculture, and human health sectors (Kalia et al., 2021).

PHB is produced as an energy- and carbon-storage compound by PNSB under special stress conditions (Monroy and Buitrón, 2020). Production of PHB from CO₂ (carbon source) via photoautotrophy in R. palustris is considered to be sustainable and environmental-friendly. When the electron donor was a poised electrode (photoelectroautotrophy) or a ferrous iron (photoferroautotrophy), it showed the highest PHB electron yield (7.34%) and specific productivity (8.4 × 10⁻¹⁴ mg/cell/h) in R. palustris TIE-1 (Ranaivoarisoa et al., 2019). In the CdS-R. palustris hybrid system, the CdS nanoparticles cotaded on the surface of R. palustris formed by Cd²⁺ bioprecipitation through the cysteine desulfurase can absorb solar energy and release electrons, and the biomass, PHB and carotenoid production were improved by 148%, 147%, and 122%, respectively (Wang et al., 2019). PHB production in R. palustris from lignocellulosic biomass which is considered to be the most economic carbon source in the world has received much attention. R. palustris can utilize p-coumarate, the major lignin breakdown product, to produce PHB (Brown et al., 2020; Alsiyabi et al., 2021). R. palustris can produce a copolymer of PHB called PHBV, which has more ideal thermomechanical properties than PHB alone (Li et al., 2016). Phasin has been shown to control the size and number of granules in the cell, affect PHB accumulation, and promote localization of granules, and heterologous expression of phasin in R. palustris resulted in a significantly higher PHBV titer (0.7 g/L) from p-coumarate (Brown et al., 2021). PHB production from other carbon sources have also been researched, and PHB of different level are obtained with different substrates. Among the different short-chain organic acids utilized, only acetate and propionate can lead to the production of PHB in R. palustris, achieving 17.1% and 11.8% substrate conversion efficiency, respectively (Wu et al., 2012). Moreover, mixtures of acetate, propionate and butyrate as substrates had a significant effect on PHB production, 16.4 mg/g/day (Cardena et al., 2017). When acetate was used as the carbon source, under limiting ammonium concentrations, acetate was consumed through the glyoxylate pathway to produce PHB; however, under nitrogen starvation conditions, acetate was consumed through the TCA cycle and PHB production was increased by 30% (McKinlay et al., 2014). Some agroindustrial residues and energy crops were also investigated for PHB formation, and the highest PHB production, 11.53% TS, was obtained in olive pomace effluent (Corneli et al., 2016). In summary, PHB production in R. palustris from different carbon sources, like CO₂, lignocellulosic biomass, short-chain organic acids, agroindustrial residues and energy crops, has been focused. However, in R. palustris, hydrogen production competes for the reducing equivalents and metabolites with PHB (Wu et al., 2012), genetic engineering to redirect more metabolic flux from hydrogen to PHB is proposed in future study.

Polysaccharide biopolymers are composed of mono saccharides linked together by O-glycosidic bonds (Hangasky et al., 2019). Due to their versatile properties like thickening, crosslinking and adsorption, polysaccharide biopolymers have been widely applied in the petroleum industry (Xia et al., 2020). Moreover, based on its properties of low production cost, nontoxicity and biocompatibility, polysaccharides is considered as green biopolymer for in situ gel formulation for drug delivery (Chowhan and Giri, 2020). R. palustris was reported to produce unipolar polysaccharide for biofilm formation under diverse conditions (Fritts et al., 2017). In the genome of R. palustris CGA009, uppE (RPA2750) and uppC (RPA4833) were identified to be responsible for unipolar polysaccharide production (Fritts et al., 2017). Unipolar polysaccharide production was enhanced in response to three photoheterotrophic conditions, including nutrient limitation, less-preferred nutrients and high salinity; however, cell growth rate was affected (Fritts et al., 2017). Design of the culture conditions of R. palustris is of critical importance to balance the cell growth and unipolar polysaccharide production, which affects the overall process economics. Further genetical modification would improve polysaccharide production.

Among the chemicals belonging to isoprenoids, isoprene and squalene are building blocks for biopolymer production. R. palustris is rich in carotenoids, another kind of isoprenoids, which include lycopene, rhodopin, 3,4-didehydro-rhodopin, anhydro-rhodovibrin, rhodovibrin, hydroxy-spirilloxanthin and spirilloxanthin (Chi et al., 2015). The produced carotenoids are connected to the transmembrane proteins through a noncovalent bond, which can stabilize the transmembrane proteins and capture light for the light harvesting complex in the cell membrane (Lang and Hunter, 1994; Takaichi, 2004). The precursors for isoprenoids production, isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), are supplied by the methylerythritol phosphate (MEP) pathway or the mevalonate (MVA) pathway (Li et al., 2019). Genes of the MEP pathway and carotenoid synthetic pathway were identified in the genome of R. palustris (Figure 3). The high carotenoids content indicates that R. palustris has great potential to be developed as a microbial cell factory for isoprenoids production.

Carotenoids, such as lycopene and β-carotene, have great application value as food additives and in the pharmaceutical industry. Research about carotenoids production by R. palustris have been mainly focused on the changes in the culture conditions, such as light sources, light intensity, pH, carbon and nitrogen sources. Among the tested light sources, blue LED showed the highest growth, the highest carotenoid content (1,782 μg/g biomass) and the highest carotenoid productivity (1,800 μg/g/Watt) (Kuo et al., 2012). The light/dark cycle ratio and light intensities also influence the carotenoid yield, and light/dark cycle of 16 h/8 h and light intensity of 150 μmol-photons/m²/s were proven to be the best light parameters in R. palustris (Liu et al., 2019). Moreover, the composition of the carotenoids are different under different light illumination conditions (Muzziotti et al., 2017). Among the various carotenoids in R. palustris, the content of lycopene is the highest, which can reach 79% of the total carotenoids under certain light conditions (Muzziotti et al., 2017). Therefore, lycopene production in R. palustris is very promising. Considering the high total carotenoids production in R. palustris, further engineering at the gene level would enhance the metabolic flux to a specific carotenoid and isoprenoid.

Squalene production in R. palustris has been reported. Due to its antioxidant properties and unique structure, squalene is usually utilized extensively in the pharmaceutical, food, cosmetic and biofuel industries (Kim and Karadeniz, 2012). Squalene derivatives, such as tetramethylsqualene epoxides, botryoxanthins and braunixanthins, could be incorporated to resistant biopolymers by condensation with polyaldehydes (Okada et al., 2000). In R. palustris TIE-1, through deletion of the shc gene (encoding the squalene hopene cyclase), fusion of two consecutive enzymes (CrtE and HpnD) and overexpression of the dxs gene, squalene yield reached 15.8 mg/g (Xu et al., 2016). In another study, the yield of hopanoids, a downstream product of squalene, was more than 36.7 mg/g DCW in wild-type R. palustris TIE-1 under certain culture conditions (Welander et al., 2009), which indicates the high potential of further improvement of squalene production.

Isoprene is a platform chemical for natural rubber (cis-1,4-polyisoprene) production. Studies on the production of isoprene in R. palustris has not been reported; however, our group found that the wild type R. palustris can produce isoprene, which was usually produced in plant. Due to its high carotenoids-producing capacity, high isoprene production is promising in R. palustris. Isoprene is produced from precursor DMAPP catalyzed by isoprene synthase (Li et al., 2018). Except for the utilization of the native MEP pathway to accumulate DMAPP, heterologous expression of the MVA pathway in R. palustris would improve isoprenoid production dramatically. The plasmid harboring the enzymes of the MVA pathway in the vector pBBR1MCS-2 was transformed into another PNSB, R. sphaeroides, resulting in an 8-fold increase in amorphadiene production (Orsi et al., 2019). Transformation of the same plasmid into R. palustris is expected in future research.

In conclusion, R. palustris has great potential to produce biopolymers and their building blocks. R. palustris can accumulate biopolymers naturally, such as PHB and polysaccharide, under various conditions. Another beneficial property of R. palustris as a microbial cell factory is that industry waste, CO₂ and lignocellulosic biomass can be converted to valuable biopolymers. The light-harvesting system in the cell membrane of R. palustris can capture light as an energy source, contributing to photosynthesis in R. palustris. Bioproduction of these valuable chemicals using many traditional chassis organisms, such as E. coli, Saccharomyces cerevisiae, and Corynebacterium glutamicum, has been researched in recent years, and high production has been realized. However, carbon and energy sources of these chassis organisms limit its sustainable and cost-effective application in industry. Therefore, in terms of economy of the fermentation process, R. palustris is more suitable as a microbial cell factory for valuable chemicals production. However, bioproduction in R. palustris suffers from the low growth rate, and improvement of growth rate is urgently required. Moreover, most research of R. palustris has focused on the optimization of culture conditions to improve chemicals accumulation in R. palustris; but genetic manipulation of R. palustris is seldomly executed due to its complicated genetical engineering process.

R. palustris is widely applied in wastewater treatment and bioremediation as a model organism for its biodetoxification and biodegradation properties (Figure 4). Aromatic compounds in the wastewater, such as 3-chlorobenzoate (3-CBA), can be consumed by R. palustris as the sole carbon source through the innate benzoate-degrading pathway (Harwood et al., 1998). R. palustris WS17 was found to degrade 3-CBA anaerobically in the light; then, R. palustris DCP3 was found to degrade not only 3-CBA but also 2-CBA, 4-CBA and 3,5-CBA (Kamal and Wyndham, 1990; van der Woude et al., 1994; Krooneman et al., 1999). The degradation mechanism of 3-CBA in R. palustris RCB100 was deciphered, which includes three steps, reductive dechlorination of 3-CBA to 3-chlorobenzoyl coenzyme A, dehalogenation of 3-chlorobenzoyl-CoA to benzoyl-CoA, and degradation of benzoyl-CoA to acetyl-CoA (Egland et al., 2001). R. palustris can also grow on cyclohexane carboxylate (CHC), the simplest alicyclic acids, which is degraded to cyclohex-1-enecarboxyl-CoA (CHene-CoA), an intermediate of the benzoate-degrading pathway (Figure 2) (Küver et al., 1995; Hirakawa et al., 2015). Two transcription factors, BadR and BadM, were proven to regulate two operons expression related to the benzoate-degrading pathway (Hirakawa et al., 2015). Additionally, pyridine, a heterocyclic aromatic compound released into the environment as industrial waste, can be degraded by R. palustris JA1 as a sole carbon source for growth (Ramana et al., 2002). Another aromatic compound, phenol, is a toxic compound from the chemical industry and human activity, and it can be converted to 4-hydroxyphenylacetate in R. palustris PL1 (Noh et al., 2002). R. palustris CQV97 was isolated and identified as a microorganism that has the ability to degrade phenanthrene, a polycyclic aromatic compound (Zhao et al., 2011). In addition to aromatic compounds, utilization of taurine, a sulfonate, and degradation of polylactic acid (PLA), have also been reported (Novak et al., 2004; Hajighasemi et al., 2016). The biodegradable PLA is a promising candidate for the replacement of traditional petroleum-based plastics, and depolymerization and recycling utilization of PLA is essential to relieve environment pressure and save resources. The protein RPA1511 from R. palustris was identified to have the hydrolytic activity toward PLA (Hajighasemi et al., 2016).

R. palustris also has the capability of heavy metal tolerance and assimilation. Heavy metals are released in industrial processes, leading to contaminants in soils and water bodies. R. palustris is recognized as a model organism for heavy metal detoxification, especially for arsenic (As), a common soil contaminant. An As-redox transformation system exists in R. palustris that can reduce the highly toxic As⁵⁺ to As³⁺ and oxidize As³⁺ to methylated arsenic (low-toxic) (Batool et al., 2017). The As⁵⁺ resistance of R. palustris is up to 100 mM; and 62.9% of the toxic As⁵⁺ in the medium can be reduced (Batool et al., 2017). In the genome of R. palustris, three ars operons and four arsR genes related to As metabolism were identified (Zhao et al., 2015). Moreover, cobalt (Co) and nickel (Ni), serious pollutants with the development of industry, can also be assimilated by R. palustris. The Co-Ni transporter from R. palustris CGA009 was introduced into Deinococcus radiodurans R1, resulting in nearly 60% Co removal from the contaminated effluents within 90 min and was introduced into Nicotiana tabacum leading to a 5-fold Co acquisition and 2-fold Ni accumulation (Nair et al., 2012; Gogada et al., 2015). In addition, the biodetoxification of other heavy metals, including selenium (Se), uranium (U), cadmium (Cd) and zinc (Zn), have also been reported. After 9 days of cultivation, 99.9% of SeO₃ ²⁻ was reduced to red elemental selenium in R. palustris strain N (Li et al., 2014). U can be reduced to U(IV)-phosphate or U(IV)-carboxylate compounds or accumulate in the cytoplasm and the cell wall in R. palustris (Llorens et al., 2012). Removal of Cd and Zn by 84% and 55%, respectively, were detected in R. palustris TN110 (Sakpirom et al., 2017).

In addition to heavy metals, R. palustris has high tolerance and degradation effects toward pesticides and herbicides, which are widely applied in agriculture and forestry. The residual pesticides and herbicides in the agro-ecosystem have caused significant environmental and human health concerns in recent years (Katsuda, 1999; Ray and Fry, 2006). Among them, pyrethroid pesticides have been extensively utilized for 30 years. R. palustris JSC-3b has the ability to degrade pyrethroid effectively; and the gene est3385, encoding a pyrethroid degradation ester, was identified in strain PSB-S (Zhang et al., 2014; Luo et al., 2018). Cyhalofop-butyl herbicides remaining in the soil can also cause serious environmental pollution and risk. The cyhalofop-butyl in soybean processing wastewater can be completely degraded after 5 days by R. palustris (Wu et al., 2019). Other toxic compounds detected in various environmental matrixes, such as hexabromocyclododecane and tributyl phosphate, which can cause serious human health problems, can also be degraded by R. palustris (Berne et al., 2005; Chang et al., 2020).

Food waste generated from restaurants and schools has caused many management problems in recent years. R. palustris CGA009 was reported to remove a wide variety of organic compounds, ammonia, nitrate and starch from food waste, and BOD and COD were reduced by 70% and 33%, respectively (Mekjinda and Ritchie, 2015).

An obvious characteristic of R. palustris is the possibility of simultaneous wastewater treatment and valuable chemicals bioproduction. When soybean processing wastewater was adopted as a carbon source, the toxic cyhalofop-butyl in the wastewater was degraded and the production of single cell protein, carotenoid and bacteriochlorophyll were enhanced (Wu et al., 2019). A maximum of 80 ml H₂/g COD/day hydrogen was concurrently produced with efficient removal of BOD, COD, ammonia, nitrate, starch in the food waste (Mekjinda and Ritchie, 2015).

These outstanding functions make R. palustris a potential organism to be applied in wastewater treatment and bioremediation. However, due to the cell washout and unstable biodegradation, these traditional use of suspended R. palustris is inefficient. Immobilization microorganism technology, which loads bacteria into a solid carrier to enhance the abundance of bacteria in the wastewater, is attempted in recent years. The immobilized R. palustris with alginate, polyvinyl alcohol or agar beads showed 30%∼40% higher ammonia removal rate than that by free R. palustris (Zhan and Liu, 2013). R. palustris P1 was immobilized to glass pumice with a high adsorption capacity of 4.02 × 10⁸ cells g⁻¹, and the maximum NH₄ ⁺-N and NO₂ ⁻-N removal rates were 134.82 ± 0.67% and 93.68 ± 0.14% higher than those of free R. palustris P1, respectively (Xiao et al., 2019). The immobilized technology showed properties like high removal stability, easy to use in continuous reactors, high cell densities, and protecting the bacteria from predation by plankton, contributing to its highly effective aquaculture wastewater treatment.

R. palustris is traditionally utilized as a biofertilizer in agriculture, and as food additives and water purification microbes in aquaculture and livestock breeding. In agriculture, due to the detoxification and degradation properties of R. palustris mentioned above, it can be utilized to clean up the pollutants in the soil and improve the soil quality. An alternative method is to clone the specific genes from R. palustris into plants, which process endows the plants with the ability to tolerate and degrade heavy metals and pesticides in the soil. R. palustris also has the capability to promote plant growth due to its function of nitrogen fixation (Oda et al., 2005) and the production of two phytohormones, indole-3-acetic acid and ALA (Koh and Song, 2007). Plant Vigna mungo treated with R. palustris CS2 showed a 17% increase in shoot length and a 21.7% increase in root length (Batool et al., 2017). Additionally, R. palustris has the ability to induce resistance against plant disease, such as tobacco mosaic virus (TMV), one of the most destructive plant viruses. The Rhp-PSP protein with inhibitory activities against TMV in vivo and in vitro was isolated and purified from R. palustris JSC-3b (Su et al., 2015). An R. palustris GJ22 suspension was sprayed on tobacco, which induced tobacco resistance against TMV and enhanced the immune response under subsequent TMV infection (Su et al., 2017). The growth and germination of the tobacco were also promoted at the same time (Su et al., 2017). These outstanding properties, detoxification and degradation functions, plant growth promoting effects and anti-virus abilities, make R. palustris a potential candidate as a biofertilizer in agriculture.

R. palustris is also widely applied in aquaculture and livestock breeding in terms of its probiotic properties as feed additives and its wastewater treatment ability. Supplementing the drinking water of broiler chickens with R. palustris resulted in the growth promotion and the improvement of meat quality (Xu et al., 2014). The growth performance and immune response of tilapia (Oreochromis niloticus) were also improved by using R. palustris as a water additive (Zhou et al., 2010). In addition, the probiotic properties of R. palustris were further demonstrated in a model animal, rats (Fang et al., 2012). Tissue damage and bacteria translocation were not detected in rats, and no R. palustris remained in the feces of the rats 3 days after the termination of intake, indicating the biological safety of R. palustris as a probiotic bacterium (Fang et al., 2012). However, the molecular mechanism of its probiotic properties is not very clear. In 2019, an extracellular polysaccharide RPEPS-30 from R. palustris was found to enhance the host immune system and improve the growth of beneficial microbes in the gut (Zhang et al., 2019). As we mentioned above, R. palustris is suitable for wastewater treatment, which is generated from the aquaculture and livestock breeding industries. R. palustris WKU-KDNS3 was isolated from an animal waste lagoon, and its capability to eliminate skatole, a major contributor to the malodor emission from feces, was proven (Sharma et al., 2015). In another study, the isolated R. palustris from eutrophicated ponds can remove the odorous organic acids and phosphate swine wastewater (Kim et al., 2004). At present, research about the applications of R. palustris in the aquaculture and livestock industries have focused on the isolation of strains and demonstrations of their activity. Further analysis of the molecular basis of its probiotic properties is necessary, and subsequent genetic manipulation of R. palustris would improve its probiotic properties.

Considering the environmental pollution caused by the consumption of large amounts of nonrenewable fossil fuels, it is urgent to find alternative clean energy sources. Numerous studies have shown that R. palustris can produce biofuels, such as hydrogen, methane, ammonia and butanol (Table 3).

Ammonia is a clean fuel with features like high energy density, no carbon emission, low storage cost, difficult to explode (Bélanger-Chabot et al., 2015; Jiao and Xu, 2019). Ammonia is particularly considered as a strong option for long-haul shipping (Phillips et al., 2010; Zhao et al., 2021). Ammonia is chemically produced from hydrogen and nitrogen by the Haber-Bosch process that is conducted at a high temperature and high pressure in industry (Kandemir et al., 2013). In nature, ammonia can be produced by biological nitrogen fixation catalyzed by nitrogenase (Figure 5, Eqs 1–3), which is highly sensitive to oxygen (Oelze and Klein, 1996; Vitousek et al., 1997). In R. palustris, three different nitrogenase gene clusters, anf, vnf and nif, encoding three nitrogenase isozymes, were identified by genomic analysis, and further functional analysis demonstrated their function (Larimer et al., 2003; Oda et al., 2005). Different from algae and cyanobacteria, in which the inhibitory effect of oxygen on nitrogenase limits its wide application, R. palustris is an anoxygenic photosynthetic bacteria and the inhibitory effect of oxygen is absent (McKinlay and Harwood, 2010b). Hydrogen and methane production catalyzed by nitrogenase in R. palustris have also been reported. Actually, researchers have focused more on hydrogen and methane production than ammonia production by R. palustris.

Hydrogen has been recognized as promising alternative due to properties like high conversion efficiency, renewability, clean burning nature and high mass-based energy density (Patel et al., 2012; Hu et al., 2018). Biological hydrogen production has significant potential in the hydrogen economy (Christopher and Dimitrios, 2012). Photo-fermentative hydrogen production by photosynthetic bacteria can be using energy directly from sunlight and reduced organic compounds (Hallenbeck and Ghosh, 2009). In R. palustris, hydrogen is produced as an obligate product during the nitrogen reduction process catalyzed by nitrogenase (McKinlay and Harwood, 2010b; Heiniger et al., 2012) (Figure 5, Eqs 1–3), even in the condition of nitrogen deficiency (Figure 5, Eq. 4) (McKinlay and Harwood, 2010b). In this process, electrons are obtained from organic or inorganic compounds and then transported to ferredoxin through the photosynthetic electron transport chain, which is utilized for hydrogen production via nitrogenase in combination with generated ATP (McKinlay and Harwood, 2010b). In R. palustris, the wide type nitrogenase consumed 75% reductant (ferredoxin) for ammonia production and the hydrogen production is at a low level (Rey et al., 2007). Moreover, the nitrogenase activity is inhibited by ammonia at the transcriptional and posttranslational level and is also inhibited by ADP, which competes with ATP for binding to the enzyme (Dixon and Kahn, 2004; Heiniger et al., 2012; Zheng and Harwood, 2019). Moreover, in R. palustris, PHB is synthesized as an energy- and carbon-storage compound, which competes with the metabolites and reducing equivalents for hydrogen production (Vincenzini et al., 1997). Some genetic engineering strategies were explored to improve hydrogen production efficiency, such as improvement of ammonium tolerance of the nitrogenase, elimination of PHB synthesis and disruption of the Calvin cycle. Regulatory gene nifA mutants, which lead to high nitrogenase expression even in the presence of ammonium, were screened, and up to five times hydrogen production (332 μmol/mg protein) relative to the wild type was detected (Rey et al., 2007). Disruption of the PHB synthesis gene phbC was performed in R. palustris WP3-5, and a 1.7-fold increase in hydrogen content (70%) was obtained (Yang and Lee, 2011). Hydrogen production can be improved by 4.6-fold to 145 μmol/mg protein in a ΔcbbLS ΔcbbM ΔcbbP mutant strain, in which electrons entering the Calvin cycle were blocked and redirected to hydrogen production (Zheng and Harwood, 2019). An obvious increase of hydrogen production has been obtained after genetic engineering strategies were applied.

Photo-fermentation of hydrogen suffers from the low light conversion efficiency and high energy demand by nitrogenase, leading to the low hydrogen yield. Most research on the improvement of hydrogen production in R. palustris has mainly focused on optimization of basic cultivation parameters, including the substrate, light intensity, pH, temperature and immobilization of the cells. Organic acids, such as acetate, propionate, malate and lactate, can be utilized as substrates to produce hydrogen by R. palustris WP3-5 (Wu et al., 2012). Four different carbon sources were evaluated for hydrogen production by R. palustris DSM127, and the highest hydrogen yield, 1.69, 2.38, 2.57, and 4.92 mol H₂/mol substrate, were obtained using acetate (3 g/L), malate (2 g/L), lactate (2.5 g/L), and butyrate (2 g/L), respectively (Hu et al., 2018). However, when butyrate (2 g/L) was utilized as the carbon source, hydrogen production started after 400 h, longer than other carbon sources (Hu et al., 2018). Mixtures of different substrates were evaluated for the distribution of hydrogen and PHB production, and mixtures of propionate and acetate showed the highest hydrogen production, 391 ml/g/day (Cardena et al., 2017). The possibility of using various industrial and agricultural wastewater for hydrogen production has been widely investigated. Photofermentation of crude glycerol, a major side product of the biodiesel industry, to hydrogen was also studied in R. palustris, and the conversion rate was improved through adjusting the cultivation conditions (Sabourin-Provost and Hallenbeck, 2009; Ghosh et al., 2012; Pott et al., 2013). Agroindustrial residues and energy crops were also investigated for hydrogen production in R. palustris. For instance, a R. palustris CGA676 mutant cultured with wheat bran and maize effluent produced 648.6 and 320.3 ml/L hydrogen, respectively (Corneli et al., 2016). Light systems are also essential for hydrogen production. Two light systems, the fluorescent and incandescent, were investigated for cell growth and hydrogen production, and the incandescent system was proven to be more effective (Hu et al., 2018). Immobilization of R. palustris through biofilm formation and latex coatings formation can concentrate its biological reactivity and stabilize the microbes, which is beneficial for hydrogen production and light conversion efficiency (Liao et al., 2010; Tian et al., 2010). R. palustris CQK 01 was immobilized on the surface of baked glass beads to form a biofilm, and the bioreactor showed high hydrogen production, 38.9 ml/L/h and 0.2 mol/mol glucose (Tian et al., 2010). Latex coatings of R. palustris CGA009 were constructed in both dry and wet conditions, and high-level hydrogen production was detected (Gosse et al., 2012; Piskorska et al., 2013). In summary, hydrogen production by R. palustris has been widely studied, and abundant genetic engineering strategies and optimization of culture conditions have been performed.

Moreover, the intracellular redox balance can also influence hydrogen production in R. palustris. During photoheterotrophic conditions, R. palustris will produce a large amount of reducing equivalents, a part of which are consumed by nitrogenase for hydrogen production for intracellular redox balance (McKinlay and Harwood, 2011). However, the native Calvin cycle can also absorb a large section of the reducing equivalents (McKinlay and Harwood, 2010a; McKinlay and Harwood, 2011). The CbbRRS system in the Calvin cycle can respond to the redox signal and regulate the carbon flux (Romagnoli and Tabita, 2007; Laguna et al., 2010). Therefore, the Calvin cycle competes for reducing equivalents with hydrogen production. When the Calvin cycle was blocked in R. palustris, the hydrogen produced was enhanced on all substrates (McKinlay and Harwood, 2011). Hence, for hydrogen production, attention should be paid to the oxidation state of substrates and the Calvin cycle flux to avoid redox imbalance.

The efficiency of hydrogen production can be further improved via complementation of photo-fermentation and dark-fermentation. In this sequential two-stage dark- and photo-fermentation process, dark hydrogen production was conducted using acidogenic bacteria like Clostridium butyricum (Su et al., 2009a), C. pasteurianum (Chen et al., 2008), Caldicellulosiruptor saccharolyticus (Özgür et al., 2010), and E. coli (Sangani et al., 2019), and photo hydrogen production was conducted using photosynthetic bacteria, like R. palustris. A sequential dark- and photo-fermentation system including C. pasteurianum and R. palustris has been constructed for hydrogen production using sucrose as a feedstock (Chen et al., 2008). The overall hydrogen yield increased from the maximum of 3.8 mol H₂/mol sucrose in dark fermentation to 10.02 mol H₂/mol sucrose by sequential dark and photofermentation (Chen et al., 2008). In the sequential dark- and photo-fermentation system, photo-fermentation can utilize the resulting effluent from dark fermentation, mainly consisting of butyric and acetic acid. It is reported that the sequential dark- and photo-fermentation system could achieve a theoretically maximum hydrogen yield (Su et al., 2009b); therefore, is has been considered as an efficient system to increase hydrogen production yield.

Methane is a commonly used energy source, and nearly half of hydrogen production in industry is from methane. A purified Mo-dependent nitrogenase from Azotobacter vinelandii with two mutations, V70A and H195Q, in NifD near the FeMo cofactor, was proven to be capable of reducing carbon dioxide to methane in vitro (Yang et al., 2012). Then, homologous mutations, V75A and H201Q, in NifD from R. palustris, were produced and methane production from carbon dioxide was also detected in vivo (Figure 5, Eq. 5) (Fixen et al., 2016; Zheng and Harwood, 2019). Further optimization indicated that utilization of nongrowing cells which are not carrying out the competing biosynthesis can enhance the yield of methane by 9-fold to 900 nmol/mg total protein using fumarate as the substrate (Zheng and Harwood, 2019). The analogous mutation of the V-dependent nitrogenase (VnfD ^V57AH180Q) can also catalyze the synthesis of methane in R. palustris (Zheng et al., 2018). The wild type of another Fe-only nitrogenase in R. palustris also has the ability to synthesize methane, ammonia and hydrogen simultaneously, but the proportion of methane is relatively low (Zheng et al., 2018). With in-depth study of the mechanisms of nitrogenase, improving the ability of nitrogenase to synthesize hydrogen and methane through genetic engineering and synthetic biology methods are the direction of further research.

In addition, the production of butanol by R. palustris has also been reported. Compared to ethanol, butanol is an attractive renewable biofuel with a higher energy content, lower volatility, higher viscosity and higher intersolubility with traditional fuels (Jin et al., 2011). The adhE2 gene from Clostridium acetobutylicum ATCC 824 encodes alcohol-aldehyde dehydrogenase, which catalyzes the synthesis of n-butanol from n-butyrate, was expressed in R. palustris (Doud et al., 2017). A selectivity (butanol production from butyrate) up to 40%, close to the theoretical maximum selectivity 45%, was achieved in the engineered R. palustris (Doud et al., 2017). However, a very low n-butanol production rate, 0.03 g L⁻¹ d⁻¹, was obtained as a result of the slow growth of R. palustris, and the redox flux is speculated to be the limiting factor (Doud et al., 2017). Further engineering to improve its growth rate and metabolic rate is proposed to improve butanol and other biofuels production.

Biofuel production, mostly hydrogen, in R. palustris has been widely researched in the fields of diverse substrates and genetical modification to improve the metabolic flux. However, the low production yield still limits its application in industry. Combination of biofuel production with industrial and agricultural wastewater treatment or biodegradation might be a promising strategy to reduce cost.

In recent years, some novel application fields were reported for R. palustris, utilizing its electron exchange ability with solid-phase materials, such as MCF, MES and photocatalytic synthesis (Figure 6). Construction of MFC has attracted much attention in recent years due to its electricity generation ability. In the MFC system, microbes can oxidize organic substrates in the anode and release electrons and protons, which are transferred to the cathode and react with oxygen to generate water, while generating electricity at the same time (Figure 6A) (Slate et al., 2019). Photo-microbial fuel cells (PMFC), which can convert light energy to electricity due to the photosynthetic activity of the photosynthetic microorganisms, are also being developed. R. palustris has been utilized as a model organism in PMFC. The power density of R. palustris DX-1 achieved 2,720 mW/m² under light with acetate as the electron donor (Xing et al., 2008). Carbon sources from wastewater and contaminated sites can also be utilized as electron donors in the PMFC system. R. palustris RP2 can produce a current (power density 720 μW/m²) using petroleum hydrocarbons from oil contaminated sites as the energy source (Venkidusamy and Megharaj, 2016). In another study, R. palustris G11 could accumulate polyphosphate and PHB when a sufficient carbon source was provided, and the stored polyphosphate and PHB were further used for electricity production in a carbon source-insufficient condition (Lai et al., 2017). Power production under carbon source-insufficient conditions indicates the simultaneous application of PMFC for energy production and wastewater treatment, in which the carbon source is not enriched. Recently, a novel hybrid hydrogen-photosynthetic microbial fuel cell was constructed, which can produce hydrogen and a maximum power density of 2.39 mW/m² in R. palustris ATCC 17007 at the same time (Pankan et al., 2020). The performance of PMFC may be affected by the electrode material, reactor, membrane, substrate and operating conditions (Aghababaie et al., 2015). Application of PMFC is mostly limited by its low power generation and expensive electrode materials, and research should be focused on improving the rate of electron transfer to/from the electrode and the exploration of advanced materials as the electrode.

Microorganisms applied in MCF can release electrons to the solid-phase material and generate electricity. However, microorganisms can also absorb electrons from the cathode for MES, a bioelectrochemical process (Rabaey et al., 2011). In an MES system, desired chemicals are produced from an inorganic substrate and the absorbed electrons at the cathode-microbe interface (Figure 6B) (Karthikeyan et al., 2019). Photoautotrophic iron-oxidizing R. palustris TIE-1 can accept electrons from a solid-phase electrode, with carbon dioxide as the sole carbon source and electron acceptor (Bose et al., 2014). In R. palustris TIE-1, the pio operon is responsible for electron uptake from the electrode, and electron uptake stimulates ruBisCo form I expression, indicating the enhancement of carbon fixation during MES (Jiao and Newman, 2007; Bose et al., 2014). The major obstacle for using R. palustris TIE-1 for MES is the low electron uptake efficiency from cathode, which is reflected by the low maximum current density values (Ranaivoarisoa et al., 2019). To further enhance the electron uptake ability by R. palustris TIE-1, several iron-based redox mediators coated cathodes were tested and the immobilized Prussian blue (PB)-coated cathode was proven to be the best (Rengasamy et al., 2018). In another study, to enhance the electron uptake rates of R. palustris TIE-1, a photobioelectrochemical system with two uncoupled reactors, an electrochemical system and a photobioreactor, was developed, which had an electron uptake rate 56-fold higher than the single system (Doud and Angenent, 2014). In summary, the application of R. palustris in MES has been reported in recent years, but electron transfer efficiency is still the key factor limiting its efficiency. Actually, even after a decade of development, using MES for chemical synthesis cannot compete with the traditional fossil-fuel-derived production (Prevoteau et al., 2020). However, MES is still promising for chemical production from carbon dioxide, and the performance of MES can be improved.

R. palustris can also absorb electrons from the nanoparticles (NPs) coated on its cell surface. An inorganic-biological hybrid system, a CdS-R. palustris hybrid system, was constructed, in which CdS NPs coated on the cell surface can absorb solar energy, release electrons, then participate in various biosynthetic pathways (Figure 6C) (Wang et al., 2019). In this semiartificial photosynthetic platform, solar energy is directly utilized for the chemical production. CdS NPs on the surface of R. palustris are formed by Cd²⁺ bioprecipitation through the cysteine desulfurase (Bai et al., 2009). In the CdS-R. palustris hybrid system, carbon dioxide reduction was enhanced and the production of solid biomass, carotenoids and PHB were increased to 148%, 122%, and 147%, respectively (Wang et al., 2019). The synthesized CdS NPs can also locate to the cytoplasm, absorb solar energy, and enhance the nitrogen fixation catalyzed by nitrogenase (Sakpirom et al., 2019). Research about the nanoparticles-photosynthetic hybrid system is still in the initial stage. Further research on different nanomaterials and the genetic modification of R. palustris are expected to improve its catalytic efficiency.