Bacillus subtilis as potential producer for polyhydroxyalkanoates

Polyhydroxyalkanoates (PHAs) are biodegradable polymers produced by microbes to overcome environmental stress. Commercial production of PHAs is limited by the high cost of production compared to conventional plastics. Another hindrance is the brittle nature and low strength of polyhydroxybutyrate (PHB), the most widely studied PHA. The needs are to produce PHAs, which have better elastomeric properties suitable for biomedical applications, preferably from inexpensive renewable sources to reduce cost. Certain unique properties of Bacillus subtilis such as lack of the toxic lipo-polysaccharides, expression of self-lysing genes on completion of PHA biosynthetic process – for easy and timely recovery, usage of biowastes as feed enable it to compete as potential candidate for commercial production of PHA.


Background
The natural bioploymers Polyhydroxyalkanote(s) (PHAs) are natural biopolymers. Many prokaryotic organisms accumulate PHAs as reserve material when carbon (C) source is available in excess in the environment and there is a limitation of nutrients essential for growth. It serves as a food source, which is mobilized by PHA depolymerase under stressed environmental conditions [1]. Although PHAs may generally account upto 90% of the dry cell weight (DCW) of the microbes [2], however their production on industrial scale is still very costly in comparison to petrochemicalbased plastics [3,4]. The other basic drawbacks which hinder their exploitation on industrial scale are the highly crystalline nature and very low strength of poly(3hydroxybutyrate) (PHB), the most well studied PHA. In contrast to homopolymers of 3HB (PHBs), copolymers of small chain length C3-C5 hydroxy acids (scl-HA) and medium chain length C6-C14 mcl-HA of PHA are more ductile, easier to mold and tough [5]. These copolymers have better film forming and mechanical properties quite similar to low-density polyethylene. These features improve their strength and processability [6,7]. Certain microbes can even produce a natural-synthetic hybrid block copolymer of polyhydroxyoctanoate-diethylene glycol, which results in significant changes in their physiochemical and material characteristics [8,9]. Efforts are thus needed towards improving product quality [10] and efficiency of the recovery process, which will result in optimization of yields. In order to reduce the cost of production, efforts are being made to search [11,12] or develop (genetically engineered) strains [13] capable of producing PHAs from inexpensive renewable sources [5,14,15], or even develop transgenic plants for this purpose [16]. The purpose of this review is primarily to consolidate the features of Bacillus subtilis as a potential candidate for producing biopolymers -PHAs.
A perusal of the capacities of different microbes to produce PHAs reveals that certain Gram-negative bacterial species belonging to Alcaligenes, Ralstonia and Pseudomonas lead this group. Alcaligenes and Ralstonia are versatile organisms with well established abilities to utilize pure substrates, agricultural wastes, oily wastes, dairy products and carbon dioxide (CO 2 ) for PHA production [5,[18][19][20]. Pseudomonas can normally synthesize mcl-PHA on various aliphatic alkenes or fatty acids, agricultural and oily wastes [20,21]. Pseudomonas sp. can however, simultaneously produce scl-mcl PHAs [22][23][24][25][26]. Among the Gram-negative bacteria, certain archaeal strains of Haloferax, Halobacterium, Haloarcula and Haloquadratum have been reported for their abilities to synthesize PHA from inexpensive C sources as feed material [27,28]. Although Gram-positive bacteria have not been widely studied, a few genera reported to produce PHB and certain copoly-mers include Bacillus, Clostridium, Corynebacterium, Nocardia, Rhodococcus, Streptomyces and Staphylococcus [29].
Bacillus as PHA producer B. subtilis is generally regarded as safe (GRAS) organism by Food and Drug Administration (FDA) [30,31] and thus offers additional benefits. B. subtilis has been accorded the designation of industrial workhorse for being among the most widely used microbes for large scale production of recombinant proteins, amino acids and fine chemicals [32,33]. It will not be inappropriate to call them as "cell factory" [34] since B. subtilis is already known for production of valuable metabolites, bioremediation and generation of bioenergy [35][36][37] [39], B. thuringiensis [37], B. mycoides RLJ B-017 [40] and Bacillus sp. JMa5 [41]. Bacillus sp. INT005 and B. cereus UW85 could produce PHA with a wide range of compositions varying from PHB, P(3HBco-3HV: copolymer of butyrate and valerate), P(3HB-co-3HHx: copolymer of butyrate and hexanoate), P(3HB-co-4HB-co-3HHx: copolymer of 3-hydroxy-, 4-hydroxybutyrates and hexanoate) to P(3HB-co-6HHx-co-3HHx: terpolymer of butyrate, 6-hydroxy-and 3-hydroxy hexanoates) depending upon the substrate [42,43]. Various Bacillus spp. have been shown by different researchers to synthesize copolymers when co-fed with various substrates. Using B. cereus UW85, the production of terpolymer of 3HB, 3HV and 6HHx was recorded with εcaprolactone as sole C source in mineral salts medium without any glucose. However, addition of glucose along with ε-caprolactone seemed to suppress copolymer synthesis and the result was the production of PHB [42]. Bacillus sp. INT005 could accumulate PHB when glucose was used alone as C substrate in the medium. However, addition of various C sources along with very low glucose concentration resulted in copolymers of 3HB and 3HHx on octanoate and decanoate, copolymers of 3HB-4HB-4HHx on 4-hydroxybutanoate and 3HB-3HHx-6HHx on supplementation with ε-caprolactone [44]. Recent studies have produced still more interesting information. B. cereus SPV when grown on structurally unrelated C sources such as fructose, sucrose and gluconate resulted in the incorporation of 4HB with the first two substrates and 4HB and 3HV with gluconate in the medium [45]. Although limitation of nitrogen (N), phosphorous (P) and oxygen in the culture conditions are known to influence PHB production, however, potassium limiting media led to the production of a copolymer containing 3HB and 3HV monomers in contrast to only PHB under sulphur, P or N limitation [43]. B. megaterium yielded PHB on cane molasses [46], B. cereus CFR06 on starch [47] and other Bacillus spp. could also produce PHB from industrial food waste water, soya waste and malt waste from beer brewery plant and pea-shell slurry [48]. Bacillus sp. 256 utilize an unrefined natural substrate -mahua (Madhuca sp.) flower as C source (containing 57% w/w as sugars) to produce copolymers (90:10 mol% P(HB-co-HV)) [49].
There are a few reports which have shown B. subtilis to be natural producers [37][38][39]50]. These results need to be viewed more cautiously since B. subtilis is a very heterogeneous group. Recent studies have shown that B. subtilis can be subdivided into two subspecies namely B. subtilis subsp.subtilis and B. subtilis subsp.spizizenii [51,52], such that only one of them may possess members which are PHA producers, where as those belonging to B. subtilis subsp. subtilis (B. subtilis 168 belongs to this group) may be naturally non-producers types. This results in a very interesting situation with respect to B. subtilis, which can thus be "labeled" as non-producers just like Escherichia coli which has been widely used for expression of many genes including those of PHA biosynthesis [3,5,[53][54][55][56].
B. subtilis has been recently used as a host for over expression of phaCAB genes from P. aeruginosa and R. eutropha. Expression of phaC1 from P. aeruginosa and phaAB genes from R. eutropha in B. subtilis resulted in the production of copolymers P(HD-co-HDD, hydroxydecanoate-cohydroxydodecanoate) and P(HB-co-HD-co-HDD) from malt waste [57]. Since B. subtilis is not a human pathogen, further supporting its usage as a host for expression of foreign genes [58]. B. megaterium phaPQRBC genes were cloned into B. subtilis 1A304(Φ105MU331). This report of stable plasmid integration into the B. subtilis chromosome paved the way for large scale fermentation process for production of PHA which eliminated the use of antibiotics, a desirable feature for reducing production cost [58]. Recombinant B. subtilis could utilize malt waste as a C source, further raising the hopes for producing PHA at cheaper rates. This study showed that phaPQ of B. megaterium were essential for PHA production along with phaRBC, although they could not observe any sequence homology in NCBI database. It has been previously shown that phaP in R. eutrophus might play a role in determining the shape and size of the inclusion bodies [59].
Although B. subtilis has the potential to compete with E. coli, the most commonly used host for heterologous expression of genes [34] and it may even be superior in certain aspects, however, the time pressure limits the change of host organism in later stages of development of process optimization [60]. Westers et al., [60] listed four drawbacks in using Bacillus as host organism -a) lack of suitable expression vectors, b) plasmid instability, c) pres-ence of protease and d) presence of malfolded proteins. Recent efforts have circumvented certain issues such as the generation of suitable expression vectors [61] and stable plasmids particularly for PHA production [58]. B. subtilis WB800 strain lacking eight extracellular proteases enabled the production of homologous proteins, including those which were otherwise susceptible to rapid degradation [34]. Certain other features which further support B. subtilis as potential PHA producer have been discussed here.

Genomic status of PHA biosynthesis in Bacillus
In the post genomic era, 22 Bacillus strains have been sequenced and 32 projects are in progress [62]. Among all Bacillus spp., B. subtilis is one of the most heterogeneous representative [52,63,64]. The concerted efforts of European Union funded consortia [65][66][67] lead B. subtilis [68] to be the first Gram-positive, soil microorganism to be completely sequenced and opened the era of functional analysis of Gram-positive bacteria. The major resources made available as a results of these efforts in B. subtilis include datasets on transcriptome, proteome, secretome and metabolome. These also include a collection of more than 3000 mutants, vectors, tools and techniques for rapid production of heterologous proteins [35,69,70].
Screening of metabolic (KEGG) and genomic (NCBI) databases for the presence of enzymes [71] involved in PHA biosynthesis reveals that genes for PhaA and PhaB are observed to be present in almost all the sequenced genomes of Bacillus except certain strains of B. megaterium [29,72], B. thuringiensis, B. subtilis and Bacillus spp. (Table  1). However, quite a bit of variation is recorded in the case of phaC gene and phaR. PhaC is frequently observed largely in the members of B. cereus group, however, it's conserved domain is absent or partially present in most other Bacillus spp. The presence of conserved domain of PhaR was largely partial or absent. In general B. subtilis lacks genes related to PHA biosynthesis [1] providing opportunity to circumvent the need to eliminate or reduce the background effect caused due to homologous genes of the host during heterologous gene expression. The concomitant presence of PHA biosynthesis and depolymerization system has proved beneficial for efficient production of PHA [35]. Incidentally, B. subtilis does contain phaZ encoding for PHA depolymerase.

Genome scale reconstruction
The basic quest in biotechnology is to construct microbial strains capable of accomplishing the rapidly expanding array of desired biotransformations [34]. The initial steps towards enhancing product efficiency and recovery process are to look for the presence of pathways that divert fluxes towards undesirable products or compete for the utilization of precursors and cofactors [73]. Genome scale reconstruction of B. subtilis was converted into an in silico  B. amyloliquefaciens FZB42 F n/a n/a n/a n/a n/a B. subtilis subsp. subtilis str. JH642 n/a n/a n/a n/a n/a n/a B. subtilis subsp. subtilis str. SMY F F F n/a n/a n/a B. subtilis subsp. subtilis str. 168 n/a F e F e n/a n/a n/a model to trace the metabolic pathways which can be verified through experimental works in B. subtilis [73,74]. This can be executed with relatively high confidence with the availability of databases on transcriptomes, proteomes and metabolomes [73,74]. Biolog's Phenotype Microar-ray™ technology enables phenotype analysis [74] of B. subtilis for a wide range of C, N, P and sulphur sources. A wide variety of C sources and their corresponding transporters present in B. subtilis reflected on its high adaptability to environmental conditions. The interesting part was its inability to utilize glycine probably because of production of H 2 O 2 , a toxic byproduct of glycine oxidase activity. The absence of glycoxylate shunt in B. subtilis justifies the observation that no growth is observed on acetate as sole C source although NADH and ATP could be produced by the metabolic network [74]. The phenotypes showing the utilization of various substrates are indicative of the presence of corresponding pathways. In the work of Oh et al., [74] 350 intracellular metabolites were identified but only 160 were present in the model, which implies that a large portion of B. subtilis metabolism is yet to be elucidated.

Synthetic genomics: a reductionist's approach
The urge to exploit microorganisms for maximum benefits to human beings has emerged in creation of cells with predictable behaviour. Genome reduction process for building collection of knockout mutants has been initiated for a wide range of bacteria such as Gram-positives -B. subtilis (Firmicutes), Corynebacterium glutamicum (Actinobacteria), and Gram-negatives -E. coli K-12, Haemophilus influenzae, P. aeruginosa, P. aeruginosa strain PA14 and Acinetobacter baylyi ADP1 (Gamma Proteobacteria) [73]. Efforts in this direction have been made with E. coli and B. subtilis as model organisms (   a: with respect to B. subtilis 168 genome deletion mutants of E. coli [76]. The 16 mutants of E. coli K-12 strain MG1655 lacked 2.4-29.7% of the parental chromosome. These deletions resulted in aberrant cell shape, slow growth rate and modified nucleoid organization. Mutants of E. coli MG1655 and W3110 with the removal of mobile DNA, virulence genes and genes for amino acids, lipopolysaccharides, transporters etc. exhibited high electroporation efficiency [77] and higher threonine secretion [75] compared to wild type strain. B. subtilis the Gram-positive counter-part of the industrially important E. coli has only 271 essential genes out of a 4.2 Mb genome [78]. The optimization of Bacillus as cell factory by deleting prophages, prophage-like regions and polyketide synthase operon of B. subtilis amounted to 7.7% reduction in genome size equivalent to elimination of 332 protein encoding genes [32]. Subsequent efforts by Ara et al., 2007 lead to DNA deletion to the extent of 24.7% without any apparent benefits [79]. However, Morimoto et al., 2008 [61] have demonstrated that genome reduction may create bacterial cells with practical utility to industry -enhanced extracellular cellulase and protease and improved efficiency to utilize a wide range of C sources.
B. subtilis Δ6 has a few major advantages compared to conventional B. subtilis strains. B. subtilis Δ6 lacks the BsuM restriction-modification system. It confers advantages to Δ6, since BsuM restriction reduces transformation efficiency of B. subtilis with recombinant plasmids and is also responsible for structural plasmid instability. It consequently affects the application potential of plasmids for high-level protein production [32]. This system can be assisted by highly stable expression vectors developed for the rapid purification of recombinant protein in B. subtilis [80]. However, this cell factory engineering can be used to re-insert genes of importance into the chromosome of B. subtilis.

Horizontal Gene Transfer (HGT)
HGT plays an important role in the diversity and evolution of B. subtilis. Phage integration accounts for 16% of the HGT regions in B. subtilis 168 [81]. Extensive HGT in Bacillus sp. increases the chance of "easier" and extensive manipulation. HGT from A+T rich islands can be accomplished by transposes. B. subtilis has comparatively more stable chromosomal structure compared to other Bacillus spp. particularly those of B. cereus group. B. subtilis genome shows 10 transposons and related rearrangements compared to 112 such events in B. halodurans C-125 chromosome [82]. In B. subtilis the ability to survive in a wide range of environmental niches can be because of large number of genes (586 equivalent to 10% of the ORFs) related to SpoA, regulatory protein important for initiating sporulation, genes encoding proteins involving antibiotics and cell wall synthesis [63].

Manipulating carbon catabolite repression
In view of the fact that feedstock being a significant component of manufacturing costs of PHA, maximizing the C yield and reduced byproduct formation becomes a priority. The metabolic bottleneck for maximum yield of all products is either stoichiometrically constrained or limited by energy. Sauer et al., [83] suggested that to improve yields, it will be better to look for microorganisms and processes that may provide the necessary energetic efficiency [83]. They recommended either to use alternative C substrates such as glycerol or sucrose (instead of glucose) which have reduced or non-existing stoichiometric limitations or go in for metabolic engineering [83]. For maximum efficiency and regulation of metabolic processes in B. subtilis, an alternative C catabolite repression was suggested since it doesn't possess cAMP [84]. A CcpA defective mutant of B. subtilis was found to grow at a slower rate in minimal medium with glucose and ammonia as C and N sources, in comparison to the wild type cells [84]. CcpA represses the expression of citZ gene encoding citrate synthase. This reaction involves condensation (utilization) of acetyl-CoA with oxaloacetate [84]. Manipulation of CcpA can enable larger availability of acetyl-CoA for PHA biosynthesis. Further, B. subtilis sporulates in response to depletion of nutrients [64] and glucose [79]. Incidentally, there is a conflict between PHA production and spore formation, since both get induced under similar environmental conditions. A non-spore forming strain (wild or mutant) of Bacillus holds promise of better performance [2]. So a knockout will enable it to grow vegetative and make sporulation genes redundant. In fact, this mechanism lead to increased cellulase activity in B. subtilis 168 [79]. Under laboratory conditions, it will be worth the effort to knockout sporulations genes and make room for re-introducing relevant genes, if necessary. Gram-positive bacteria can produce copolymers of PHA from simple, inexpensive and structurally un-related C sources. In contrast, Gram-negative bacteria depend upon expensive and structurally related substrates for producing copolymers of PHA [29].

Novel recovery system for PHA
Since PHA accumulates intracellularly, cell disruption is indispensable to recover it. On an industrial scale, the PHA recovery process involves the usage of a large amount of chemical reagents and/or enzymes [2]. An efficient recovery system implies reduced cost of production [85]. An optimum time to recover PHAs is immediately after the C source gets exhausted and before the commencement of degradation by depolymerase enzyme [86]. A novel self disruption cell system has been developed in B. megaterium. In this system, a gene cassette carrying the cell lysis system (holin and endolysin of B. amyloliquefaciens phage) [87] was inserted into the E. coli -B. subtilis shuttle vector pX. In this expression system, xylR-xylA' target genes are induced by xylose but inhibited by glucose, which acts as an anti-inducer [86]. It synchronizes the processes of spontaneous cell lysis and substrate exhaustion, which thus results in the release of accumulated PHAs [86]. The efficiency of this regulatory process can be enhanced by manipulating the YoeB, a cell wall-associated protein, which gets induced in response to antibiotics stress. The expression of yoeB in B. subtilis is under a xylose-inducible promoter [88]. yoeB mutants display an increased rate of autolysis in response to nutrient depletion and various cell envelope stress conditions [88]. The process of autolysis in B. subtilis 168 can be aided by mutating yoeB gene and in the process making it independent of xylose regulation.

Lacks Lipopolysaccharides
Gram-negative bacteria such as Ralstonia, A. latus and recombinant E. coli are among those which have been exploited for industrial scale PHB production. The outer cell membrane of most Gram-negative bacteria including E. coli contains lipo-polysaccharides (LPS), which are endotoxins [89] and are pryogenic in human beings [60]. The purification of byproducts including PHAs is complicated due to the presence of endotoxins [60]. Since LPS induces a strong immunogenic reaction, this feature is undesirable for biomedical applications of the PHAs [7,90]. A review on the toxic nature of LPS reveals that cyanobacterial (Gram-positive bacteria) LPS are less toxic compared to those produced by members of Enterobacteraceae [89]. B. subtilis offers the advantage of lacking LPS and excreting proteins at a high rate into the medium [61]. It thus stands a better chance as PHA producer for biomedical applications [90].

Potential for producing hydrogen
The use of CO 2 as a potential inexpensive renewable C source can help in reducing PHA production cost [5]. Synechococcus sp. MA19 was observed to produce 55% w/w PHA from CO 2 [91]. A. eutrophus was shown to accumulate PHA at the rate of 1.55 g/L/h, which was higher than that recorded with cyanobacteria or photosynthetic bacteria [92]. The strategy being proposed is that if hydrogen (H 2 ) production becomes cheap then R. eutropha can produce PHB from CO 2 and oxidation of (H 2 ), under dark fermentative conditions [93].

Conclusion
B. subtilis as an organism for industrial scale production of several fine chemicals is already established but not for PHA production. A comparison of all the PHA producers (Additional file 1, Table S1) reveals that the most versatile PHA producers are Bacillus spp. Their abilities to produce PHAs range from homopolymers to copolymers fromsimple sugars to complex industrial wastes. It reflects on how Bacillus can exploit its natural abilities to produce hydrolytic enzymes for easy metabolism of biowastes to be used as inexpensive C source. Bacillus can be easily grown to very high cell density of 70 g/L [41] and does not have a major codon bias [29,58]. A recovery of around 70 to 80% of bacterial dry matter as PHA production is potentially sufficient for establishing an economically feasible process [35]. In addition, the robust stress tolerant B. subtilis lacking the toxic LPS, and permitting heterologous expression of self-lysing genes concomitant with completion of PHA biosynthetic process (for efficient recovery), enable it to be a strong contender in the future as an industrial PHA producer.