An updated overview on the regulatory circuits of polyhydroxyalkanoates synthesis

Summary Polyhydroxyalkanoates (PHA) are a promising and sustainable alternative to the petroleum‐based synthetic plastics. Regulation of PHA synthesis is receiving considerable importance as engineering the regulatory factors might help developing strains with improved PHA‐producing abilities. PHA synthesis is dedicatedly regulated by a number of regulatory networks. They tightly control the PHA content, granule size and their distribution in cells. Most PHA‐accumulating microorganisms have multiple regulatory networks that impart a combined effect on PHA metabolism. Among them, several factors ranging from global to specific regulators, have been identified and characterized till now. This review is an attempt to categorically summarize the diverse regulatory circuits that operate in some important PHA‐producing microorganisms. However, in several organisms, the detailed mechanisms involved in the regulation of PHA synthesis is not well‐explored and hence further research is needed. The information presented in this review might help researcher to identify the prevailing research gaps in PHA regulation.


Introduction
Polyhydroxyalkanoates (PHA) are natural polymers of hydroxyalkanoates produced by microorganisms, including bacteria and archaea, usually under conditions of limited nutritional supply with excess carbon source (Reddy et al., 2003;Chen and Wu, 2005). They are accumulated by some microorganism as intracellular carbon and energy reserve to combat adverse environmental conditions (Khanna and Srivastava, 2005). PHA are bioplastics and are gaining considerable prominence in both the environmental and medical fields. Synthesis of PHA is a well-regulated process that involves a number of enzymes and regulatory proteins (Sagong et al., 2018). One of the approaches to maximize PHA production and synthesize novel PHA, is to engineer the biosynthetic pathway (Steinb€ uchel, 2001). Another approach is to manipulate the regulatory elements controlling PHA synthesis. To realize the latter approach, a better understanding of the regulatory circuits involved in PHA synthesis is necessary. The present review attempts to provide an overview on the regulation of PHA to the readers. It mainly focuses on the current knowledge on PHA regulation in several model species of bacteria and haloarchaea. Understanding the PHA regulation is a progressive topic, and this review might help to identify the prevailing research gaps. Reviews on the regulation of PHA synthesis is limited. The few already published reviews are focussed on the various types of regulations affecting PHA synthesis (Kessler and Witholt, 2001;Vel azquez-S anchez et al., 2020). This review has been presented to the readers from a different point of view. The present review has provided a vivid description of the various regulation systems affecting PHA synthesis mainly with the help of some model organisms.

An overview on PHA biosynthesis pathways
The type of PHA accumulated is closely related to the microbial species and carbon sources. More than 150 monomeric units of hydroxyalkanoates have been identified as PHA monomers (Steinb€ uchel and L€ utke-Eversloh, 2003). Depending upon the (carbon atom) chain length in the monomers, PHA are classified as short-chain length comprising of C3-C5 monomer (SCL-PHA), medium-chain length comprising of C6-C14 monomer (MCL-PHA) and long-chain length consisting of more than 14 carbons in monomer (LCL-PHA) (Sagong et al., 2018). Most researches are focussed on SCL and MCL-PHA. Synthesis of SCL-PHA from natural and engineered strains is receiving great attention from researchers due to their wide distribution and high production level (Wang et al., 2016). Various Gram negative bacteria genera like Cupriavidus, Burkholderia and Azotobacter, and Gram positive bacteria from genera like Bacillus, Nocardia and Rhodococcus, produce SCL-PHA. Additionally, a few haloarchaeal species including Haloferax mediterranei are SCL-PHA producers. Three monomers of SCL-PHA which have received significant consideration are 3-hydroxybutyrate (3HB), 3-hydroxyvalerate (3HV) and 4-hydroxybutyrate (4HB). MCL-PHA have gained considerable attention as they are more flexible in nature. Most fluorescent Pseudomonas species belonging to rRNA homology group I accumulate MCL-PHA by incorporating 3-hydroxyalkanoate monomers including 3-hydroxyhexanoate, 3-hydroxyoctanoate, 3hydroxydecanoate and 3-hydroxydodecanoate into homo, block-or random copolymers (Fiedler et al., 2000;Chen and Jiang, 2017).

Regulation of PHA synthesis
PHA have a high demanding global market owing to their several potential benefits like biodegradability, biocompatibility and thermoplasticity. However, commercial usage of PHA is still limited. One of the primary reasons is the high production cost. Several fermentation strategies have been employed where low cost substrates including agro-industrial wastes, are utilized to produce PHA at a low cost (Bhattacharyya et al., 2015). However, another possibility to reduce PHA production cost is by constructing high yielding PHA producers through genetic manipulation Ye et al., 2018). PHA metabolism is a bidirectional continuous cycle of synthesis and degradation (Prieto et al., 2016). It is an intricate process controlled by complex regulatory circuits where the carbon flow is suitably balanced between the PHA synthesis and the growth of microorganism. Nutrient limitation with excessive carbon supply is an important criterion to promote PHA production. It can be speculated that such kind of growth conditions affect the PHA metabolic pathway and the overall process is controlled by diverse regulatory mechanisms. Among them, PHA synthesis is directly regulated at the enzymatic level. Determination of the kinetic characteristics (K m , V max ) of the enzymes like b-ketothiolase, 3ketoacyl-CoA reductase, or PHA synthase enzymes have been studied in PHA producers like Zoogloea ramigera (Davis et al., 1987;Ploux et al., 1988), Cupriavidus necator (Steinb€ uchel and Schlegel, 1991) and Thermus thermophilus (Pantazaki et al., 2005). These values provide insights on their substrate specificity as well as clues to further improve their activities to promote PHA formation. Allosteric regulation of b-ketothiolase and PHA synthase has shown to affect PHA accumulation in bacteria like C. necator (Oeding and Schlegel, 1973) Azotobacter beijerinckii (Senior and Dawes, 1973), Pseudomonas putida GPo1 (Ren et al., 2009). Likewise, covalent modification of PHA synthase by phosphorylation also affects its enzymatic activity which may further influence PHA accumulation (Juengert et al., 2018). Therefore, controlling the enzymatic activity of key proteins involved in PHA biosynthetic pathway is one possible method to regulate PHA metabolism. However, this aspect has not been discussed in detail here as it is not the primary focus of this review.
Another mechanism of controlling PHA metabolism involves the regulatory proteins. PHA granules are covered with proteins which are known as PHA granuleassociated proteins (PGAPs) (Maestro and Sanz, 2017). Some of these PGAPs such as PhaR, PhaM, PhaF and PhaQ often serve as regulatory proteins that bind to their own promoters or the promoters of other PHA biosynthetic genes to regulate their transcription to ensure wellorganized PHA granule formation (Lee et al., 2004;P€ otter et al., 2005;Gal an et al., 2011;Pfeiffer et al., 2011). Thus, modification of these regulatory elements can alter the PHA production level. Additionally, some of the global regulatory systems like Gac/Rsm system, PTS and PTS Ntr system, NtrB/NtrC two-component regulatory system and quorum sensing also participate in PHA synthesis in some species. Most species have multiple regulatory systems which are intertwined to give a combined effect on PHA synthesis (Table 1). The following sections attempt to discuss the contribution of these regulation models in different PHA-accumulating bacterial species and haloarchaea.

PHA regulation in Cupriavidus necator
Cupriavidus necator is conveniently handled and cultured on cheap feedstock to produce PHB (Raberg et al., 2018). The designation of C. necator has underwent series of changes. Initially, known as Hydrogenomonas eutropha, it has been renamed as Alcaligenes eutrophus, Ralstonia eutropha, Wautersia eutropha and finally as Cupriavidus necator. PHB synthesis in this organism is controlled by at least three regulatory factors, PhaR, PhaM and PTS system (Fig. 1).
Phasins are low-molecular weight amphiphilic proteins, predominantly present on the surface of PHA granules (Steinb€ uchel et al., 1995;Maestro and Sanz, 2017). The hydrophobic domain of PhaP phasin effectively binds to the hydrophobic surface of PHA inclusion body, forming a network-like layer on PHA granules (McCool and Cannon, 1999). The hydrophilic domain of PhaP remains exposed to the cytoplasm (Maestro and Sanz, 2017). The PhaP expression is regulated by a DNA-binding protein named PhaR. Co-occurrence of phaR gene and phaP gene has been revealed in many bacteria and haloarchaea (Maehara et al., 2001;P€ otter et al., 2002;Segura et al., 2003;Chou et al., 2009;Kadowaki et al., 2011;Cai et al., 2012;Nishihata et al., 2018). PhaR homologs are well-distributed among SCL-PHAproducing microorganisms and the phaR gene is located near the other PHA-synthesis genes, including phaC, phaA, phaB and phaZ in the genome (Maehara et al., 2002). The phaR gene in C. necator is located adjacent to its phaCAB gene cluster (P€ otter et al., 2002). C. necator has seven phasin proteins (PhaP1-PhaP7), among which PhaP1 is the major phasin affecting PHB synthesis and granule size (P€ otter et al., 2005;Pfeiffer and Jendrossek, 2012). Deletion of only phaP1 impaired PHB production whereas the other six phaP genes had no significant effect (York et al., 2001;P€ otter et al., 2005;Pfeiffer and Jendrossek, 2012). PhaR represses the expression of phaP1 by binding to its promoter, under PHB non-accumulating conditions in this strain (P€ otter et al., 2002;York et al., 2002). Additionally, it binds to an intragenic region of phaP3 and represses its transcription (P€ otter et al., 2005). Such a regulation of phaP genes by PhaR is necessary at the onset of PHB DrpoN mutant accumulated more PHA during nitrogen limitation compared with nitrogen excess. Rehm (2004, 2005) RpoS Global regulator RpoS might serve as a negative regulator of phaC1 promoter.
Deletion of rpoS gene increased PHA degradation. (2008) PsrA Transcriptional regulator PsrA is involved in the fatty acid -PHA metabolic network.

Raiger-Iustman and Ruiz
Deletion of psrA gene reduced the PHA production and also changed the monomer composition.  synthesis to avoid the interference by PhaP in the initiation process of PHB synthesis (York et al., 2002). Deletion of phaR in C. necator reduced PHB yield. Intriguingly, the DphaRDphaP1 double mutant produced further reduced PHB yield compared with DphaP1 mutant, suggesting PhaR also contributes in promoting PHB synthesis in a PhaP-independent manner in C. necator (York et al., 2002). Possibly, PhaR represses additional PHA-synthesis proteins such as PHA synthase whose inaccurate expression might have hindered PHB production or led to premature PHB utilization (York et al., 2002;Quelas et al., 2016). Once the cells start PHB synthesis, PhaR has the ability to sense the presence of PHB granules and binds to the native PhaP-free PHB granules, thus releasing the inhibitory effect on phaP1 and phaP3 transcription (P€ otter et al., 2002, 2005). However, the amount of PhaP3 synthesized is lower than PhaP1 which may be due to stronger repression of phaP3 transcription than phaP1 by PhaR (P€ otter et al., 2005). As cells synthesize PhaP1 and small amounts of PhaP3, they gradually cover the PHB granule surface due to their higher affinity towards it and thus PhaR gets gradually detached. During the process of granule formation and polymer chain elongation, the granule surfaces are capable to accommodate both PhaR and PhaPs and thus keeping the cytosolic concentration of free PhaR low enough to allow continuous phaP1 and phaP3 transcription. It was observed that DphaP1 mutant synthesized naked PHB granules with their exposed hydrophobic surfaces coalescing to form only one large PHB granule. On the other hand, DphaR mutant synthesized large amounts of PhaPs that covered the PHB granule and stabilized them, forming smaller sized PHB granules than the wild-type strain. Thus, it suggests that PhaR-mediated controlled expression of phaP proteins is critical for PHB granule morphogenesis (P€ otter et al., 2002). Finally, at the end of PHB synthesis process, the PHB granules attain a critical volume and the surface area becomes limited (Quelas et al., 2016). Thus, the continuously expressed PhaP1 and PhaP3 proteins displace PhaR, thereby, increasing the cytosolic concentration of free PhaR. These unbound PhaR binds to the phaP1 and phaP3 binding sites, again repressing their transcription (P€ otter et al., 2005). Simultaneously, to inhibit further PhaR expression, the excess unbound PhaR also binds to the promoter region of its own gene and represses its expression. Thus, PhaR is an autoregulated transcriptional repressor which regulates PhaP1, PhaP3 and PhaR expressions and thereby controls PHB biosynthesis in C. necator (Fig. 1).
Another PHA granule-associated protein detected in C. necator is PhaM (Fig. 1) (Pfeiffer et al., 2011). It is a 32 kDa DNA-binding protein possessing phasin-like properties and shows non-significant similarity (15%) A. Regulation of PHA synthesis by PhaR. In the absence of PHB chain, free PhaR binds to the promoter of phaR and phaP and inhibits their transcription. In the presence of nascent PHB chain, PhaR binds to PHB granules and the negative effect of PhaR on phaR and phaP transcription is released and their transcription continues. As PhaP increases, it displaces PhaR from PHB granules. As free PhaR increases, it again binds to the phaR and phaP promoter, inhibiting their transcription. B. Association of PHB granule with nucleoid via PhaM in C. necator. C. Regulation of PHB synthesis by PTS system. PTS system regulates PHB synthesis by modulating PHA mobilization system, possibly by transferring the phosphoryl group or influencing ppGpp level via SpoT1. Inactivated PHB degradation may improve PHA accumulation. The red and dashed red arrows indicate repression and proposed repression effects, respectively. with PhaR protein (Wahl et al., 2012). The predicted structure of PhaM reveals a carboxy terminal similar to histone proteins, which explains the DNA binding ability of PhaM (Wahl et al., 2012). Its amino terminal contains two potential transmembrane domains which might be involved in the binding with PHB granules. The four lysine residues present in the carboxy terminus PAKKA motifs in PhaM is responsible for attaching PHB granules to the nucleoid region (Bresan and Jendrossek, 2017). Under PHA-accumulating conditions, PHB granules synthesized by wild type are attached to the nucleoid region (Wahl et al., 2012). The association of PHB granules with the nucleoid region is probably related to PhaM. PhaM and PhaC1 are constitutively expressed and PhaM directly interacts with PhaC1 and forms the complex of PhaM-PhaC1 even in the absence of PHB. The interaction between PhaM and PhaC1 is specific as PhaM is unable to interact with PhaC of Aeromonas caviae in vitro although both belong to class I synthases (Ushimaru and Tsuge, 2016). The PhaM-PhaC1 initiation complex bound to nucleoid serve as a scaffold for PHB granules Jendrossek, 2013, 2014). During cell division and prior to PHB formation, this initiation complex is colocalized at least two discrete nucleoid-associated foci. As PHB granules formation begin, they attach to the complex following which they distribute among two daughter cells. Thus, each daughter cell receives at least one formed PHB granule. At the later stages of growth, the PhaM and PhaC1 proteins detach from the granules (Bresan and Jendrossek, 2017). The constitutive overexpression of PhaM increased the number of small PHB granules which completely covered the nucleoid surface (Wahl et al., 2012). Deletion of phaM generated fewer PHB granules with a larger diameter compared with the wildtype strain. This is probably because PhaM acts as a physiological activator of PhaC1 as one molecule of PhaM activated 10-11 molecules of PhaC1 (Pfeiffer and Jendrossek, 2014). Moreover, the distribution of PHB granules among the daughter cells in the DphaM mutant was hampered as only one of the daughter cells contained PHB granules (Wahl et al., 2012). Thus, PhaM ensures proper segregation of PHB granules into daughter cells during cell division by facilitating the interaction of PHB granules with the nucleoid region. In addition, PhaM also interacts with PhaP5 and forms a PhaM-PhaP5 complex which are attached to the nucleoid region (Pfeiffer and Jendrossek, 2013). In the wild-type strain, phaP5 expression is very low and its deletion did not affect the granule number and localization (Wahl et al., 2012). Notably, when phaP5 was overexpressed, the number of PHB granules increased but their diameter decreased (Wahl et al., 2012). Moreover, the PHB granules were detached from the nucleoid region and were localized near both the cell poles in the form of clusters. It is possible that binding of the overexpressed PhaP5 to PhaM, restricts the binding of PhaM to DNA and/or to PhaC1, leading to detachment of PHB granules from nucleoid region. Alternatively, it is also possible that, the overexpressed PhaP5 displaces the PhaM molecules from the PHB granule surface leading to altered subcellular localization of the granules. Therefore, subcellular localization of PHB granules is determined by integrated expressions of PhaM, PhaP5 and PhaC1 in C. necator.
Another point of PHA regulation in C. necator is exerted by the PTS system (Fig. 1). The sugar PTS system (sugar phosphotransferase system) is a multicomponent system that participates in uptake and concomitant phosphorylation of carbohydrates (Erni, 2013). It's paralog system is PTS Ntr (nitrogen-related phosphotransferase system). Phosphoenolpyruvate (PEP) serves as a common phosphoryl donor in both the systems. Sugar PTS system is carbohydrate-specific where phosphoryl group from PEP is relayed to enzyme I and then to a histidine protein (HPr). Enzyme I and HPr are carbohydrate non-specific. The phosphoryl group is transferred from HPr to the carbohydrate-specific enzyme IIA (Kraube et al., 2009). Enzyme IIB accepts the phosphoryl group from IIA and transfers to the sugar attached to another enzyme IIC. PTS Ntr system is nitrogen-related where the phosphoryl group from PEP is sequentially transferred to nitrogen-related enzyme I (EI Ntr ), nitrogenrelated protein (NPr), and then to nitrogen regulatory enzyme II (EIIA Ntr ) (Wang et al., 2005). C. necator possesses an incomplete PEP-PTS system that is composed of enzyme I, HPr and EIIA Ntr , EIIA Man encoded by ptsI, ptsH, ptsN and ptsM, respectively (Kaddor and Steinb€ uchel, 2011) (Fig. 1). It lacks EIIB and EIIC components and thus cannot phosphorylate sugars. Intriguingly, the C. necator HPr shows 33% amino acid similarity with the NPr of E. coli and it strongly phosphorylates enzyme IIA Ntr rather than EIIA Man (Rabus et al., 1999;Kraube et al., 2009). Interestingly, IIA Ntr plays a regulatory role in PHB metabolism (Kraube et al., 2009). Inactivation of the ptsN gene resulted in a higher PHB accumulation in the mutant strain (Kaddor and Steinb€ uchel, 2011). Contrarily, deletion of ptsI or ptsH led to a reduced PHB accumulation when excess carbon source was present. After carbon source was exhausted, the PHB content rapidly reduced in the ptsI or ptsH mutants. Possibly, the PtsI and PtsH proteins regulate the PHBmobilizing system in this strain (Pries et al., 1991). In the presence of carbon source, these proteins transfer the phosphoryl group to the PHB mobilizing enzyme system and inactivates the latter. Perhaps, in their absence, the PHB-mobilizing system cannot be inactivated, leading to a faster degradation of the accumulated PHB. Alternatively, it is also possible that these proteins regulate PHB mobilization at the transcription level. In 2014, it has been further suggested that the PTS system affects the stringent response in C. necator which influences the PHB accumulation (Karstens et al., 2014). The stress response during nitrogen limitation is achieved through synthesis of ppGpp nucleotide. SpoT1 and SpoT2 proteins determine the level of ppGpp in cells in response to nitrogen availability. SpoT1 is a bifunctional (p)ppGpp synthase/hydrolase enzyme. However, SpoT2 possesses only ppGpp synthase activity. DspoT1DspoT2 double mutant showed strongly impaired PHB accumulation and a more active PHB mobilization system compared with the wild-type and DspoT2 mutant (Juengert et al., 2017). On the contrary, simulated DspoT1 mutant constructed by complementing spoT2 in DspoT1DspoT2 showed higher PHB accumulation and reduced PHB mobilization. Further evidences suggest that the absence of ppGpp synthesis in DspoT1DspoT2 triggers PHB mobilization by PhaZa1 PHB depolymerase. On the other hand, a constant high level of ppGpp in the simulated DspoT1 mutant due to the absence of ppGpp hydrolase activity possibly inhibits PhaZa1-mediated PHB mobilization. Thus, there exists a connection between stress response and PHB accumulation in C. necator. Notably, the non-phosphorylated EIIA Ntr resulted from ptsI or ptsH deletion interacts with SpoT1 (Karstens et al., 2014). Thus, it is possible that this interaction changes the ppGpp level which affects the PHB content of the mutants.
PHA regulation in Pseudomonas putida KT2440 (or KT2442) P. putida KT2440 and its spontaneous rifampicin resistant mutant P. putida KT2442, are natural MCL-PHA producers and possesses almost identical expression profile (Follonier et al., 2011). In P. putida KT2442/ KT2440, PhaF, PhaD, Crc, PsrA, RpoN, RpoS, stringent response, PTS system and GacS/GacA system plays regulatory role in PHA synthesis (Fig. 2). The particular effect of the different regulatory factors on PHA synthesis in various other strains of P. putida such as P. putida GPo1, P. putida U and P. putida CA-3 have been summarized in Table 1.
The PHA gene cluster of P. putida KT2440 comprises of two operons, phaC1ZC2D and phaIF (Prieto et al., 2016). The latter operon is located downstream of the former one. The transcription of phaC1ZC2D and phaIF are driven by the promoters of phaC1 and phaI, respectively (De Eugenio et al., 2010). The phaC1 and phaC2 genes encode two type II PHA synthases with the preference for 3-hydroxyacyl-CoA with length of C6-C14, phaZ encodes an intracellular PHA depolymerase, phaD encodes a TetR-like transcriptional regulator and phaI and phaF encode PHA granule-associated phasin proteins (De Eugenio et al., 2010). PhaF is involved in maintaining the molecular architecture of PHA granules (Gal an et al., 2011;Obeso et al., 2015). According to the structural characterization study from Maestro et al., (2013), PhaF belongs to the intrinsically disordered protein family. PhaF has an interesting structural property as it possesses three separate motifs (Gal an et al., 2011;Maestro et al., 2013;Tarazona et al., 2019). The amino terminal region containing the long and uninterrupted amphipathic alpha-helical structure is stabilized by the interaction with PHA granules. The carboxy terminus containing histone-like DNA-binding domain is natively unfolded in the absence of DNA but acquires a super helical structure upon its non-specific binding to DNA. It has been demonstrated that PhaF segregates the PHA granules among the two daughter cells during cell division process by binding to both PHA granules and nucleoid region (Gal an et al., 2011;Tarazona et al., 2020). The central core of PhaF contains a leucine zipper motif, which is involved in the oligomerization of PhaF protein. Such kind of coiled-coil sequence is also predicted in the primary structure of PhaI, which indicated that PhaF and PhaI are likely to form a heterodimer or heterotetramers to stabilize the PHA granules (Maestro et al., 2013). P. putida KT2442 uses octanoic acid as the preferred carbon source for PHA synthesis since the expression of phaC1 and phaI was higher when octanoic acid was fed than glucose, citric acid or gluconate (Gal an et al., 2011). PhaF plays a role in regulating the expression of phaC1 and phaI. In the DphaF mutant, the transcription level of phaC1 was reduced by 3.5-fold and the highest transcriptional level of phaI was differed from the mid-exponential phase to the stationary phase compared with the wild type. Moreover, the PHA granules in DphaF mutant failed to distribute among the daughter cells, which reduced the PHA content during continuous fermentation where PHA formation and cell division take place simultaneously (Gal an et al., 2011). Hence, PhaF maintains the heterogeneity of the cell population with respect to PHA synthesis and granule distribution. Recently, it has been revealed that PhaF stabilizes PHA granules and functions in properly segregating them during cell division by interacting with the PhaI located on the granule surface (Tarazona et al., 2020). Additionally, it is also capable of interacting with PhaD and modulates the regulatory activity of PhaD on transcription of pha operon in P. putida KT2442. PhaD is not associated with PHA granules rather randomly localized throughout the cytoplasm. PhaD interacts with the phaI and phaC1 promoter regions to activate the transcription of the phaIF and phaC1ZC2D operon, respectively. (Hoffmann and Rehm, 2004;De Eugenio et al., 2010). The upregulation in the transcription levels of the PhaD-regulated PHA genes is more pronounced when octanoate is used compared with glucose. Thus, PhaD serves as a carbon-dependent activator of the pha gene cluster in P. putida KT2442 (De Eugenio et al., 2010. It is likely that PhaF may act as a coactivator of PhaD as it has been found that presence of PhaF affected the unilateral binding of PhaD to phaI promoter (Fig. 2) (Tarazona et al., 2020). However, exact interactive role of PhaF with PhaD and phaI is yet to be determined.
There are evidences that phaIF and phaF transcription in P. putida KT2440 are differentially regulated. The phaF expression is induced by nitrogen limitation in this strain (Hoffmann and Rehm, 2005). RpoN, RNA polymerase sigma factor 54 (r 54 ), is a global regulator that affects the transcription of nitrogen-regulated genes. Interestingly in the DrpoN mutant, phaF was expressed irrespective of nitrogen status in the culture. Thus, RpoN acts as a negative regulator of phaF expression in P. putida KT2440, under nitrogen excess conditions  . 2). In contrast, the co-transcription of phaI and phaF was only induced during nitrogen limitation in both the wild-type and DrpoN mutant. Thus, the phaIF expression is RpoN-independent. P. putida KT2440 possesses a typical PTS Ntr system composed of Enzyme I Ntr , histidine protein (NPr) and enzyme IIA Ntr , encoded by ptsP, ptsO and ptsN genes, respectively (Fig. 2). PHA accumulation was significantly impaired in the case of ptsP or ptsO deletion (Vel azquez et al., 2007). However, PHA synthesis increased for ptsN deletion. Probably, deletion of ptsN gene creates a factual situation of excess carbon source with respect to other limiting nutrients. This channelizes the available carbon source to PHA synthesis. On the contrary, the PHA synthesis machinery of DptsP or DptsO might sense a shortage of carbon source which directs the carbon source to other metabolic pathways. Intriguingly, the non-phosphorylated EIIA Ntr inhibits the activity of pyruvate dehydrogenase complex (Pfl€ uger-Grau et al., 2011). But when its encoding gene was deleted, the activity of pyruvate dehydrogenase was improved. It is possible that more acetyl-CoA is channelled towards PHA synthesis in DptsN. Thus, the role of EIIA Ntr in carbon metabolism may exert a metabolic regulation in PHA synthesis in P. putida KT2440. Stringent response is mediated by ppGpp synthetase RelA, and ppGpp synthetase/hydrolase SpoT, encoded by relA and spoT genes, respectively. Using oleic acid as the carbon source, the wild-type P. putida KT2440 accumulated high amounts of PHA during nitrogen limitation (Mozejko-Ciesielska et al., 2017). However, the production significantly reduced under non-limiting conditions. Interestingly, the DrelA/spoT mutant produced similar levels of PHA in both nitrogen limiting and nonlimiting conditions. Thus, deletion of relA and spoT genes eliminated the need for nitrogen limitation for PHA synthesis. It was also noted that the expression of phaIF genes was significantly higher in DrelA/spoT, indicating that the phaIF operon was negatively regulated by stringent response.
In the case of nutrient balance, Crc (Catabolite repression control) protein represses phaC1 expression and inhibits PHA synthesis in P. putida KT2442 (La Rosa et al., 2014). In nutritionally balanced medium like LB, the phaC1 expression was strongly inhibited by Crc during the exponential phase. Even addition of excess octanoic acid did not induce phaC1 expression. However, inactivation of crc gene increased the phaC1 expression by five folds and led to PHA accumulation in small amounts. In the stationary phase, Crc protein is counteracted by two sRNAs, CrcZ and CrcY and its inhibitory effect on phaC1 was diminished. The strength of Crc-mediated repression is determined by the level of the two sRNAs, CrcZ and CrcY. Notably, a two-component system, CbrA-CbrB, involved in the amino acid uptake and assimilation, influences the transcription of CrcZ and CrcY sRNAs. In the imbalanced medium, CrcZ and CrcY sRNAs completely antagonized Crc protein but in balanced medium, the Crc protein was partially antagonized. Thus, the phaC1 expression and PHA accumulation was unaffected by Crc protein in imbalanced medium. However, in the balanced medium, deletion of the crc gene increased PHA production by 57% during the stationary phase.
The sigma factor RpoS is a global regulator that regulates stress response. RpoS is reported to be involved in PHA metabolism in P. putida KT2440. Deletion of rpoS gene in this strain increased PHA degradation rate due to higher expression of phaZ gene during the stationary phase (Raiger-Iustman and Ruiz, 2008). Moreover, depletion of the carbon source in this phase lowered the rate of PHA synthesis, resulting in a lower PHA content in the DrpoS. Hence, it has been suggested that RpoS negatively regulated phaC1 promoter, possibly in an indirect manner. P. putida KT2442/KT2440 also possesses the GacS/GacA system that is involved in the regulation of its PHA synthesis (Prieto et al., 2016). Disruption of the GacS sensor kinase led to a reduction in PHA synthesis. The transcription rate of the entire pha cluster was reduced. However, the underlying mechanism involved in regulating PHA synthesis by GacS/GacA system is not yet determined. In addition, a transcriptional repressor PsrA negatively regulated the genes involved in fatty acid metabolism pathway in this strain (Fonseca et al., 2014). Deletion of psrA genes upregulated the fabB gene expression and suggested that fatty acid metabolism was more active in DpsrA. Thus, less number of intermediates were channelled towards PHA biosynthesis leading to a reduced production level.

Complex regulatory circuit in Azotobacter vinelandii
Azotobacter vinelandii is a free living, nitrogen fixing bacteria capable of accumulating PHB. This species has a complex regulatory circuit which involves PhbR, PEP-PTS system and Gac/Rsm system (Fig. 3). The genome of A. vinelandii possesses a biosynthetic phbBAC operon, the regulatory gene phbR encoding the transcriptional regulator of phbBAC (PhbR), the phbP gene and phbF encoding the phasin regulator PhbF (known in other bacteria as PhaR) (Segura et al., 2000(Segura et al., , 2003Peralta-Gil et al., 2002). Inactivation of the phbR gene reduced PHB accumulation by 70%, and the activity of the acetoacetyl-CoA reductase, b-ketothiolase and PHB synthase enzymes were reduced by 30%, compared with the wildtype strain (Peralta-Gil et al., 2002). Thus, PhbR activates the transcription of the phbBAC operon. Notably, the transcription of phbB gene is driven by two overlapping promoters, pB1 and pB2. In the DphbR mutant, the level of transcript corresponding to the pB1 promoter was significantly reduced and that corresponding to pB2 promoter increased. This indicates that PhbR activates the transcription of phbB driven by pB1 promoter. Possibly, in the absence of PhbR, RNA polymerase more efficiently initiates transcription from pB2 promoter. Similar to P. putida KT2440, A. vinelandii possesses PTS Ntr system (Enzyme I Ntr , histidine protein NPr and enzyme IIA Ntr ) which is encoded by ptsP, ptsO and ptsN genes. Inactivation of the ptsP or ptsO gene reduced PHB accumulation, whereas the ptsN inactivation doubled PHB synthesis (Segura and Esp ın, 1998;Noguez et al., 2008). In the case of ptsP or ptsO inactivation, the terminal phosphoryl acceptor enzyme IIA Ntr is not phosphorylated (Noguez et al., 2008). Further inactivation of ptsN in these mutants removes the non-phosphorylated EIIA Ntr , which completely or partially restores PHB accumulation. Intriguingly, these pts mutations affect the transcription levels of the phbR and phbBAC genes. The phbR mRNA levels and subsequent transcription of phbBAC genes in the DptsP mutant and DptsO mutant are reduced, whereas they increase in the DptsN mutant (Noguez et al., 2008). It suggests that the non-phosphorylated EIIA Ntr represses PHB synthesis by downregulating phbR and phbBAC expression in A. vinelandii. It is noteworthy that the phbR expression is regulated by the sigma factor RpoS (Hernandez-Eligio et al., 2011). RpoS serves as an activator in the PHB synthesis process as its inactivation decreased the phbR and phbB transcript levels and subsequently, reduced PHB accumulation (Fig. 3). A deeper insight into the regulatory role of PTS system reveal that non-phosphorylated EIIA Ntr induces proteolytic degradation of RpoS by ClpAP complex (composed of hexameric ATPase/protein-unfoldase, ClpA and the tetradecameric proteolytic component, ClpP) in A. vinelandii (Fig. 3) (Muriel-Mill an et al., 2017). RpoS degradation lowers the expression level of phbR, the transcriptional activator of phbBAC, thereby, diminishing the transcription of the biosynthetic operon. The phbR expression is also posttranscriptionally regulated by a small RNA named ArrF (Azotobacter regulatory RNA involving Fe) (Muriel-Mill an et al., 2014) (Fig. 3). ArrF links iron concentration in medium and PHA synthesis in A. vinelandii. In the condition of iron excess, ferric uptake repressor (Fur) protein, a regulator of iron acquisition, complexes with Fe 2+ and represses the transcription of arrF gene. During iron limitation, this repression is released and arrF is transcribed (Jung and Kwon, 2008). Notably, the phbR expression is upregulated and PHB synthesis increased during iron-limiting conditions (Muriel-Mill an et al., 2014). The arrF inactivation reduced the phbR and phbB transcripts levels, and PHB accumulation, indicating ArrF sRNA is a positive regulator of phbR expression during iron limitation. GacS/GacA system composed of the sensor kinase (GacS) and the response regulator (GacA) is a twocomponent system present in several Gram negative bacteria including A. vinelandii. In the presence of appropriate extracellular signal, GacS protein autophosphorylates (Heeb and Haas, 2001). Eventually, GacA is phosphorylated and then acts as a transcriptional activator of several small RNA (sRNA), also known as Rsm (repressor of secondary metabolites) sRNA. These sRNAs further complex with RsmA protein which is a mRNA binding negative translational regulator (Manzo et al., 2011). The rsm sRNA-RsmA complex posttranscriptionally regulates the translation of the target mRNAs and controls protein synthesis. In A. vinelandii, Gac/Rsm system regulates the expression of pha genes by controlling the post-transcriptional expression of PhbR (Fig. 3). The phosphorylated GacA activates the transcription of one rsmY, and seven rsmZ1-rsmZ7 genes (Manzo et al., 2011). These sRNAs bind to the RsmA protein and counteracts its repressor effect. Deletion of gacS gene resulted in reduced PHB accumulation, indicating that the GacS/GacA cascade plays a positive role in the PHB synthesis (Castañeda et al., 2000). While rsmA deletion increased the PHB production by 25%, indicating that RsmA play a negative role in PHB synthesis (Hernandez-Eligio et al., 2012). Notably, the transcript levels of phbR and phbB was reduced in DgacA mutant, whereas they increased in DrsmA mutant. This suggests that GacS/GacA system counteracts the RsmA repressor and promotes phbR translation which further activates transcription of phbBAC. On contrary, deletion of gacA reduces the rsm sRNA transcription level, which is insufficient to counteract the repressor effect of RsmA protein. Therefore, binding of the free RsmA protein to 5' leader region of phbR transcripts leads to degradation of the latter and hence reduces PhbR protein expression, and subsequent PHB production (Hernandez-Eligio et al., 2012).
Another possible regulatory protein involved in PHA metabolism of A. vinelandii is CydR (Fig. 3). It is an oxygen-dependent global regulatory protein. This FNRlike transcription factor necessary for the regulation of cytochrome bd expression in A. vinelandii (Wu et al., 1997). Interestingly, a putative CydR binding site is present upstream of phbB in this strain (Setubal et al., 2009). The DcydR mutant showed increased b-ketothiolase and acetoacetyl-CoA reductase activity, compared with the wild-type strain (Wu et al., 2001). PHB was accumulated in the DcydR mutant throughout the exponential phase, whereas wild-type accumulated PHB only during the stationary phase. Thus, deletion of cydR imparts a positive impact on the expression of PHA genes and PHA synthesis in A. vinelandii.
The phaP gene located upstream of the phaEC encodes the major phasin protein in H. mediterranei. Knockout of the phaP gene significantly reduced PHBV accumulation and affected the number and size of PHA granules in H. mediterranei (Cai et al., 2012). However, the H. mediterranei PhaP shares no sequence similarity with bacterial phasins and its homologs are only distributed among haloarchaea. Probably, the PhaP protein of H. mediterranei represents a novel haloarchaeal type phasin family. The phaP gene in this strain is cotranscribed with the phaR gene. The promoter region of phaRP operon consists of four tandemly repeated sequences (Cai et al., 2015). Most likely, this negativecis element acts as the binding site for the PhaR protein.
The haloarchaeal PhaR homologs constitute of an AbrBlike domain in the C-terminus, which is critical for the repressor function of PhaR and supposedly acts as the DNA-binding motif. The PhaR protein acts as a negative transcriptional regulator of phaRP operon by specifically interacting with its promoter. Importantly, in haloarchaea, PhaR regulates the transcription of phaR and phaP by binding to their common promoter. However, in bacteria, phaP and phaR possesses independent promoters and PhaR interacts to the respective promoters to regulate their expression (Maehara et al., 2002;P€ otter et al., 2002). Strikingly, PhaR promotes PHBV synthesis in H. mediterranei in a PhaP-independent pathway as phaR complementation in DphaRP mutant completely restored PHBV accumulation (Cai et al., 2015). Such kind of phenomenon is similar to C. necator (York et al., 2002). The DphaRP cells produced only one or two medium-sized PHA granules, compared with the multiple moderate-sized PHA granules in the wild-type strain (Cai et al., 2015). Complementation of only phaR in DphaRP led to the synthesis of one or two large-sized PHA granules. On the other hand, only phaP complementation produced several irregularly shaped small-or mediumsized PHA granules. A similar granule morphology was also observed when phaP was overexpressed using a strong promoter in DphaRP. Thus, it was suggested that PhaR controlled the phaP expression that might be necessary for synthesis of PHA granules with regular morphology in H. mediterranei.
The PEP/pyruvate interconversion in H. mediterranei is mediated by pyk and pps genes (Chen et al., 2019). The pyk gene encoding pyruvate kinase catalyses the conversion of PEP to pyruvate whereas the reverse reaction is catalysed by PEP synthetase, encoded by pps gene. Deletion of the pps gene led to a 35.9% increase in PHBV production most likely by channelling more pyruvate towards PHBV synthesis. Intriguingly, H. mediterranei genome possesses a novel protein, PPS-like, that shows high homology with PPS protein. However, this protein is not involved in the PEP/pyruvate interconversion, instead, deletion of the pps-like gene led to a 70.46% increase in PHBV accumulation. It has been speculated that PPS-like protein evolved as a regulator protein from the PPS protein (Chen et al., 2020). Deletion of pps-like gene promoted the expression of the PHA monomer supplying pathway and upregulated the expression of the phaEC, phaR and phaP genes in H. mediterranei. It activates the transcription of the three cryptic genes, especially phaC1 (Chen et al., 2020). Moreover, both PhaC1 and PhaC3 forms functional PHA synthase with PhaE and leads to PHBV synthesis in the Dpps-likeDphaC mutant of H. mediterranei. Intriguingly, PPS-like protein effectively interacts with the promoter region of the phaC1 gene and forms protein-DNA complex. This indicates that PPS-like protein acts as a repressor of the phaC1 expression and hence, its gene deletion leads to an upregulation of the phaC1 expression. Such type of regulatory function of PPS-like in PHA synthesis is reported for the first time and thus requires further characterization.

Regulation in other bacteria
PhaQ, a transcriptional regulator in Bacillus megaterium. PhaQ is a transcriptional regulatory protein identified in Bacillus megaterium (Lee et al., 2004). PhaQ differs from the PhaR protein of C. necator in amino acid sequence, N-terminal portion, cis-acting element sequence, as well as size (Maehara et al., 2002;P€ otter et al., 2002). It has high amino acid sequence similarity with hypothetical PhaQs from Bacillus anthracis and Bacillus cereus (Ivanova et al., 2003;Lee et al., 2004). Possibly, PhaQs constitute a separate class of transcriptional regulator of PHB synthesis in Bacillus species. PhaQ in B. megaterium negatively autoregulates its own expression by interacting with its own promoter region and interfering the binding of RNA polymerase with its promoter (Lee et al., 2004). The phaQ gene is located upstream of the phaP gene and they are co-transcribed (McCool and Cannon, 1999;Lee et al., 2004). The genome of this strain consists of another pha gene cluster, phaRBC, transcribed as a tricistronic operon where phaR and phaC encode heterodimeric PHB synthase subunits. Generally, PhaP proteins are abundant and the concentration of transcriptional regulatory proteins like PhaR or PhaQ is low in PHAproducing bacteria (Lee et al., 2004). In B. megaterium, the intergenic region of phaQ-phaP shows no existence of a promoter. Thus, the differential expression of PhaQ and PhaP is possibly because the cell selectively degrades the phaQ transcript in the phaQP cotranscript by its post-transcriptional machinery, leaving the phaP transcript intact (Lee et al., 2004). The expression level of PhaP protein was low in B. megaterium with phaQ overexpressed. This indicates that PhaQ negatively regulates the expression of phaP in B. megaterium. PhaQ interacts with DNA and artificial PHB granules in vitro. It is also a PHB-responsive repressor as it could sense the presence of PHB granules in vivo. Perhaps, PHB may act as an inducer for phaP expression in the PhaQ-mediated regulation.
Novel regulatory proteins in Synechocystis sp.. Synechocystis sp. PCC 6803 is a PHA-accumulating cyanobacterium. During nitrogen depletion, this strain undergoes chlorosis in which the cells first accumulate glycogen and later synthesize PHB by metabolizing the intracellular glycogen pool (Koch et al., 2019). Two novel proteins, Slr0058 and PirC, are known to regulate PHB synthesis in this strain (Koch et al., 2020b;Orthwein et al., 2021). Additonally, response regulator Rre37 and SigE also regulates PHB synthesis in this strain. Slr0058 shows structural similarity with regulatory phasin PhaF from Pseudomonas spp. but lacks the DNA binding domain (Koch et al., 2020b). The Dslr0058 mutant synthesizes PHB but strikingly, no visible PHB granules were detected during the vegetative growth; whereas, it had abnormally increased number and improperly aggregated PHB granules during chlorosis. Possibly, Slr0058 is involved in initiation of PHB granule formation. During vegetative growth, Slr0058 aggregates in foci and then disappears during chlorosis. In the absence of Slr0058, the PHB granules aggregates in an uncontrolled manner. Thus, Slr0058 is a novel regulatory protein involved in the controlled PHB synthesis and granule formation in Synechocystis sp. PCC 6803.
The other protein PirC negatively regulates the activity of 2,3-phosphoglycerate-independent phosphoglycerate mutase (PGAM) catalysing the conversion of 3phosphoglycerate to 2-phosphoglycerate in lower glycolysis leading to acetyl-CoA production (Orthwein et al., 2021). Under sufficient nitrogen concentration, PirC protein complexes with a signal processor protein, PII. However, during nitrogen limitation, PirC is released from the complex and it inhibits the PGAM activity. Thus, the carbon flux is re-directed towards glycogen synthesis instead of acetyl-CoA formation. Interestingly, deletion of pirC led to increased glycogen catabolism, leading to acetyl-CoA production and hence, over-accumulation of PHB under nitrogen limitation in this strain. Further expression of the phaA and phaB genes from C. necator, under the control of a strong promoter, in the pirCdeleted mutant of Synechocystis sp. PCC 6803 significantly improved PHB production (Koch et al., 2020a). During nitrogen/phosphorus limitation and in the presence of acetate as carbon source, this DpirC mutant accumulated up to 81% (wt) PHB compared with only 32%(wt) by wild type under alternating light/dark regime. Thus, it is evident that cultivating regulatory mutants in optimized culture conditions is a promising method to maximize PHA production.
SigE, a group 2 RNA polymerase sigma factor, is reported to regulate PHB biosynthesis in Synechocystis sp. PCC 6803 (Osanai et al., 2013). Overexpression of sigE increased PHB production under nitrogen depletion. Moreover, the enzymes involved in glycogen catabolism and oxidative pentose phosphate pathway were upregulated. Thus, engineering of sigE modified the metabolic pathway from glycogen to PHB biosynthesis during nitrogen depletion. Another regulator that regulates carbon storage distribution in this strain is Rre37. This OmpR-type response regulator is induced by nitrogen depletion. Overexpression of rre37 decreased the glycogen level as it accelerated the catabolism of glycogen and enhanced PHB accumulation (Osanai et al., 2014). Overexpression of sigE and rre37 further enhanced PHB accumulation. Rre37 preferentially activated phaA and phaB, whereas SigE activated phaC and phaE expressions.
PHB synthesis regulation in Herbaspirillum seropedicae SmR1. NtrB/NtrC system is a two-component sensoractivator regulatory system comprising of a signal transduction protein (P II and/or P z ), effector protein, sensor (NtrB) and regulatory (NtrC) protein. It is the key regulatory system for nitrogen metabolism in bacteria. During nitrogen limitation, the nitrogen sensing protein, uridylyl transferase/uridylyl-removing enzyme uridylylates P II or P z protein (Persuhn et al., 2000). The uridylylated P II or P z protein enables autophosphorylated NtrB to transfer a phosphoryl group to NtrC. NtrC acts as a transcriptional regulator that controls the expression of nitrogen-dependent genes. In the presence of excess nitrogen, P II or P z protein is non-uridylylated that inhibits the kinase activity of NtrB. Moreover, it stimulates the phosphatase activity of NtrB that results in nonphosphorylation of NtrC, thereby inactivating the regulatory function of NtrC (Herv as et al., 2010). Herbaspirillum seropedicae SmR1 is a PHAaccumulating nitrogen-fixing diazotrophic bacterium belonging to b-Proteobacteria. This strain possesses the components of the NTR system. Deletion of the ntrC gene increased the PHB accumulation, irrespective of the nitrogen status (Sacomboio et al., 2017). A deeper insight shows that ntrC deletion upregulated the expression of zwf gene, which encodes G6PDH (glucose-6-phosphate dehydrogenase) In the DntrC mutant, the G6PDH activity and NADPH/NADP + ratio increased by 2.8-fold and 2.1-fold, respectively. The increased NADPH favours PHB biosynthesis as it is a cofactor for PhaB. Moreover, it also creates a higher pool of reducing power and PHB generation acts as a redox sink. Thus, up-regulation of zwf gene is the key factor leading to higher PHB production in the DntrC mutant (Sacomboio et al., 2017).
There exists a cyclic dependency between PHB synthesis and redox-responsive transcriptional regulator Fnr, in H. seropedicae SmR1. Transcriptional activation of phaC genes by Fnr is required for proper PHB synthesis, whereas optimal PHB production is important for maintaining the Fnr activity (Batista et al., 2018). Deletion of the three Fnr encoding genes reduced PHB accumulation and the transcription levels of the three phaC genes (phaC1, phaC2 and phaC3). During oxygen limitation, the absence of PHB in the DphaC1 mutant disturbed the redox balance of the cell as the mutant was unable to maintain the NAD(P)H/NAD(P) + ratios, resulting in increased sensitivity of the cells to oxidative stress. Notably, the redox balance of cell is Fnr-dependent because synthesis of cytochrome c proteins involved in electron transport chain is regulated by Fnr. Thus, redox balance and PHB biosynthesis is tightly controlled in H. seropedicae SmR1.
Herbaspirillum seropedicae SmR1 also has a special type of PHB synthesis regulation known as responsive backup circuit (RBC). The genome of this strain consists of three genes encoding putative phasins, phaP1, phaP2 and phaP3 (Alves et al., 2016). PhaP1 is the major phasin followed by PhaP2 in wild type, while PhaP3 is very less abundant and detectable only in the DphaP1 mutant. The expression of both the phaP1 and phaP2 genes are regulated by the PhaR regulator (Kadowaki et al., 2011). PhaR possesses 83% sequence identity with C. necator PhaR and is also capable of binding to DNA and PHB granules. It represses the expression strength of its own promoter and phaP1 and phaP2 promoters. In the wild-type strain, the transcription strength of phaP2 promoter was 8-fold lower compared with the phaP1 promoter (Alves et al., 2016). Strikingly, the expression of phaP2 increased by 6-fold in the DphaP1 mutant. The expression of phaP2 is not fully activated in wild type but it serves as a backup phasin gene to guarantee proper phasin expression and allows PHB synthesis to some extent in the DphaP1 mutant. Such a regulation type, known as RBC, increases robustness of the microorganisms under unfavourable conditions. Riboregulation in Sinorhizobium meliloti. Small RNAs are known to regulate complex metabolic processes by modulating the expression of multiple target genes in response to various stimuli (Beisel and Storz, 2010). Sinorhizobium meliloti produces PHB under unbalanced growth conditions. The fate of the carbon source in such unbalanced conditions is strictly regulated by MmgR sRNA in this strain. MmgR (Makes more granules Regulator) is a negative regulator of PHB accumulation (Lagares et al., 2017). During growth, depletion of nitrogen source acts as a stimulus for inducing mmgR expression. Once MmgR sRNA is activated, it finely tunes the PHB synthesis. It limits the PHA accumulation to an optimum level, despite of the presence of excess carbon source in the medium. Clearly, PHB accumulation was enhanced and granules were irregularly shaped in the DmmgR mutant. This was accompanied by overexpression of the PhaP1 and PhaP2 proteins. Thus, MmgR sRNA mediated PHB accumulation in S. meliloti involves negative post-transcriptional regulation of PhaP expression.
PHA synthesis regulation exerted by Quorum sensing and other global regulators. Quorum sensing (QS) is a global regulatory mechanism of gene expression in response to extracellular concentration of certain signal molecules called autoinducers (Bassler, 2002). The concentration of the autoinducers increase with bacterial cell density, and hence QS system allows bacteria to monitor their population density by altering various gene expressions (Venturi, 2006). Gram negative bacteria use acylated homoserine lactone (AHL) as autoinducer signal molecules. In Pseudomonas aeruginosa, QS system via LasI/LasR participates in regulating PHA biosynthesis (Xu et al., 2011). LasI, an AHL synthase, catalyses the production of N-3-oxo-dodecanoyl homoserine lactone (3-O-C 12 -HSL) as a signal molecule which binds to the transcriptional regulator, LasR (Steindler et al., 2009). The 3-O-C 12 -HSL interacts with the N-terminal domain of the LasR protein and induces structural changes that allows the protein to bind to the DNA sequences of target genes and regulate their transcription (Kiratisin et al., 2002). Deletion of lasI or lasR or both in P. aeruginosa reduced the PHA content (Xu et al., 2011). Moreover, the mutants showed significantly reduced levels of phaC1 expression. Possibly, LasR mediated QS system downregulates phaC1 expression that reduces intracellular PHA biosynthesis in P. aeruginosa.
Pseudomonas chlororaphis PA23 has three distinct QS systems, PhzI/PhzR, CsaI/CsaR and AurI/AurR, among which PhzI/PhzR is known to affect PHA biosynthesis as its deletion led to a reduced PHA production (Nandi et al., 2016;Mohanan et al., 2019). This strain also possesses another global regulator ANR which is an oxygen-sensitive transcriptional factor (Sawers, 1991). ANR acts by binding to the conserved anr box present in the promoter of its target genes (Mohanan et al., 2019). In P. chlororaphis PA23, ANR and PhzI/ PhzR are subjected to cross-regulation (Nandi et al., 2016). ANR positively regulates phzI and phzR, whereas PhzR negatively regulates anr transcription. Intriguingly, both the QS-deficient and Danr mutant showed similar levels of reduced PHA accumulation. However, the pha gene expression profiles in the two mutants were different. The downregulation of the pha genes including phaC1, phaC2, phaZ, phaD, phaF and phaI were more pronounced in QS-deficient strain than Danr mutant. Notably, the promoter region of the anr gene possesses a phz box which suggests that it is recognized by the PhzR, activated by AHL. Thus, it is possible that the effect of ANR on PHA synthesis is mediated via QS system (Mohanan et al., 2019). Moreover, as ANR protein is a global regulator affecting multiple genes, it possibly has an overlapping effect on PHB production. It is revealed that anr deletion in P. seudomonas extremaustralis decreases the oxidative stress resistance and disturbs the redox balance maintenance under limited oxygen supply (Tribelli et al., 2013). Furthermore, recent studies show that ANR-controlled genes include those related to central carbon metabolism and regeneration of NADPH in P. putida (Tribelli et al., 2019). Taken together, all these factors probably have an added influence resulting in a reduced PHB accumulation in the Danr mutant of P. chlororaphis PA23 (Mohanan et al., 2019).
In Burkholderia thailandensis, QS is a complex regulatory circuit. The scmR gene is a global regulator for synthesis of various secondary metabolites and also represents a key component of the QS system in this strain. Possibly, ScmR controls gene expression by regulating synthesis of AHL signalling molecule, or by directly binding to target genes, or by indirectly modulating gene expression via intermediate regulators (Le Guillouzer et al., 2020). B. thailandensis possesses three LuxI/LuxR-type QS systems, BtaI1/BtaR1 (QS-1), BtaI2/ BtaR2 (QS-2) and BtaI3/BtaR3 (QS-3). Among them, BtaI1, BtaI2 and BtaI3 are LuxI-type synthases, and BtaR1, BtaR2 and BtaR3 are the LuxR-type transcriptional regulators (Le Guillouzer et al., 2017). In the DbtaR1 and DbtaR3 mutants, the transcription level of scmR gene was low, indicating that it is activated by QS-1 and QS-3 (Martinez et al., 2020). Interestingly, an interdependence between the QS-1 and QS-3 systems exists in this strain where BtaR1 possibly modulates QS-3 system and vice versa (Le Guillouzer et al., 2017). scmR deletion reduced phaC expression and led to 50% decrease in PHA production (Martinez et al., 2020). Complementation of the scmR gene restored the wildtype PHA production. Hence, ScmR positively affects PHA accumulation in B. thailandensis. Notably, the DbtaI3 or DbtaR3 mutant showed decreased PHA accumulation which might be due to the reduced expression of scmR gene. Thus, QS-3 positively controls the PHA synthesis probably by modulating the expression of scmR gene. Hence, QS-dependent regulation of the scmR gene contributes to the regulation of PHA synthesis in B. thailandensis (Martinez et al., 2020).

Conclusions
Bioplastics are an alternative solution to the increasing plastic pollution caused by over usage of synthetic plastics. PHA are promising bioplastics having potential application in various sectors including biomedical, and packaging industries. The main objectives of PHA research is to facilitate synthesis of PHA with novel characteristics and also with characteristics similar to the plastics so as to offer a greener alternative to the existing products. Moreover, to promote commercial PHA production for widespread applications. To achieve these, strain improvement is a basic requirement. Designing microbial strains having improved PHA-producing capabilities would reduce production cost of bioplastics. Most studies have showed that engineered strains for PHA will increase PHA biosynthesis and their competitiveness. However, genetic modification of organisms can have unpredictable consequences. Transferring gene from one microorganism to another, gene deletion or rearrangement of gene sequences can result into genetic instability or affect other physiological functions of microorganisms that can create obstacle for industrial applications. A single gene may be regulated by several regulatory factors and may control different physiological traits. Therefore, knowledge of regulatory mechanisms is important and might help to design improved strains with better genetic stability. Regulation is an integral part of PHA synthesis. Identifying the key players of PHA regulation and a deeper characterization of the regulatory pathway is necessary. In this review, different regulatory mechanisms involved in PHA synthesis from typical bacteria and haloarchaea have been discussed. Types of regulation in PHA metabolism is quite diverse, ranging from global regulators to specific regulators. Most organisms possessed multiple regulation models which affected the expression of pha gene cluster either positively or negatively. Notably, these regulators controlled PHA metabolism in different organisms at different levels. This review provided a simplified idea on how the regulatory models operated in some specific PHAsaccumulating bacterial and haloarchaeal species. Such information on PHA regulation is crucial as it would direct the PHA researchers to develop improved PHA-producing microbial strains with the aid of metabolic engineering.
However, not all regulators involved in PHA synthesis have been identified till now. Hence, continued effort and further investigation is needed to elucidate the unknown regulatory circuits present in PHA metabolism. Moreover, the already characterized regulation models have been hardly applied for PHA production improvement. Thus, it is necessary to implement these regulation models in research work to achieve engineered strains with improved PHA production ability.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.