The “beauty in the beast”—the multiple uses of Priestia megaterium in biotechnology

Abstract Over 30 years, the Gram-positive bacterium Priestia megaterium (previously known as Bacillus megaterium) was systematically developed for biotechnological applications ranging from the production of small molecules like vitamin B12, over polymers like polyhydroxybutyrate (PHB) up to the in vivo and in vitro synthesis of multiple proteins and finally whole-cell applications. Here we describe the use of the natural vitamin B12 (cobalamin) producer P. megaterium for the elucidation of the biosynthetic pathway and the subsequent systematic knowledge-based development for production purposes. The formation of PHB, a natural product of P. megaterium and potential petro-plastic substitute, is covered and discussed. Further important biotechnological characteristics of P. megaterium for recombinant protein production including high protein secretion capacity and simple cultivation on value-added carbon sources are outlined. This includes the advanced system with almost 30 commercially available expression vectors for the intracellular and extracellular production of recombinant proteins at the g/L scale. We also revealed a novel P. megaterium transcription-translation system as a complementary and versatile biotechnological tool kit. As an impressive biotechnology application, the formation of various cytochrome P450 is also critically highlighted. Finally, whole cellular applications in plant protection are completing the overall picture of P. megaterium as a versatile giant cell factory. Key points • The use of Priestia megaterium for the biosynthesis of small molecules and recombinant proteins through to whole-cell applications is reviewed. • P. megaterium can act as a promising alternative host in biotechnological production processes.


Introduction
Since its discovery in 1884 (de Bary 1884), Priestia megaterium (formerly known as Bacillus megaterium (Gupta et al. 2020)) provides a powerful cell factory for biotechnology, with numerous patents and applications in industry. The bacterium serves as a model organism for genetic studies and recombinant protein production (Vary 1992;Vary 1994). With its large size of up to 2.5 × 10 μm-"megaterium" literally means "big beast"-it has a significant larger volume compared to that of Escherichia coli (Vary et al. 2007) (Fig. 1).
P. megaterium is a Gram-positive rod, has a low G+C (~38%) genome, and forms endospores. Its size alone has attracted microbiologists for many years to study its physiology and function including cell division, cell wall biosynthesis, and sporulation (Vary 1994). Now within the last decades, recent advances in modern molecular biology have unlocked its potential for biotechnology. With the expansion of synthetic biology, the many advantages of P. megaterium make it an attractive microbial cell factory to rival other model microbes Rebekka Biedendieck and Tobias Knuuti contributed equally to this work.
such as E. coli and Bacillus subtilis (Eisenstein 2016). This review will provide an overview of the diverse potential of P. megaterium as a model Gram-positive organism for biotechnological applications including small molecules like cobalamin (vitamin B 12 ), the polymer polyhydroxybutyrate (PHB), the production of diverse intra-and extracellular recombinant proteins, whole-cell transformations, and its function as a plant growth-promoting bacterium.

Priestia megaterium-history and genome sequencing
In October 2020, Gupta et al. showed that many Bacillus species in addition to the Subtilis and Cereus clade constitute a total of 17 new individual clades based on conserved signature indels (CSIs). They proposed that these clades should be recognized as new genera, with the name Priestia gen. nov. for the Megaterium clade containing the former Bacillus species B. megaterium, B. abyssalis, B. aryabhattai, B. endophyticus, B. filamentosus, B. flexus, and B. koreensis due to two CSIs in the oligoribonuclease NrnB which were uniquely shared by all clade members (Gupta et al. 2020).
The first genome sequences of two P. megaterium strains (DSM319 and QM B1551) were published a decade ago by Eppinger et al. (2011). Up to now, the full genome sequences including corresponding plasmids of around 20 distinct P. megaterium strains are available at the NCBI genome database. Five of these strains lack natural plasmids, while the remaining strains contain up to ten plasmids, consistent with studies already from the early 1980s which found plasmid-less strains to be an exception (Stahl and Esser 1983). The type strain DSM32 (ATCC14581) has been used to conduct basic genetic research. It is also known as the source of the cytochrome P450-BM3 (CYP102A1) (Narhi and Fulco 1986). The plasmid-less DSM319 and its variant MS941, which lacks the gene coding for the major extracellular protease NprM (Wittchen and Meinhardt 1995), are best suited candidates for plasmid-based genetic applications, including the generation of mutants and the recombinant production of proteins . Strain QM B1551 is used in basic research, especially within the context of sporulation genetics (Manetsberger et al. 2018;Riyami et al. 2019). Strain WSH-002 has been used for co-cultivation with Ketogulonicigenium vulgare (Zhang et al. 2010) or Gluconobacter oxydans (Lü et al. 2003) to produce vitamin C (Liu et al. 2011). Strain NCT-2, which was isolated from salinization soil from greenhouses, shows high capacity in bioremediation in salinized soil (Wang et al. 2020). Similarly, strain Q3 was described as an endophytic quinclorac-degrading bacterium for bioremediation purposes , while strains YC4-R4 and TG1-E1 show high salt tolerance as plant growth-promoting rhizobacteria (Vílchez et al. 2018a;Vílchez et al. 2018b). Likewise, strain JX285 acts as plant growth-promoting bacterium, isolated from rhizospheric soil (Huang et al. 2019). Strain SR7 was identified in samples collected from a naturally supercritical carbon dioxide (scCO 2 )-rich environment. SR7 displays resistance to scCO 2 , which is considered to be a promising alternative to classical organic solvents and is already used in in vitro applications (Boock et al. 2019;Freedman et al. 2018). P. megaterium in biotechnological production processes-from the biosynthesis of small molecules through to whole-cell applications These different properties clearly highlight the diversity of P. megaterium and provide the prerequisite for its diverse and Escherichia coli (small cells). P. megaterium and E. coli were individually grown aerobically in rich medium at 37°C, mixed in the middle of their exponential growth phases and examined in a field emission scanning electron microscope (FESEM) Zeiss DSM982 Gemini (magnification 6,500-fold). The picture was taken by Manfred Rohde, Helmholtz Centre for Infection Research, Braunschweig, Germany.
Production of small molecules: cobalamin (vitamin B 12 ) in P. megaterium P. megaterium is a natural producer of vitamin B 12 (cobalamin) and has played a prominent role in the study of cobalamin biosynthesis and its industrial production. Cobalamin is a key vitamin for higher eukaryotes, which take it from their diet and require it for B 12 -dependent enzymes (Banerjee and Ragsdale 2003). In nature, cobalamin is only produced by certain species of bacteria and archaea. Derived from the tetrapyrrole family, cobalamin contains a central cobalt ion octahedrally coordinated between four pyrrole nitrogens, a lower ligand (DMB, 5,6-dimethylbenzimidazole) and an interchangeable upper ligand (adenosyl or methyl group). Vitamin B 12 is officially named cyanocobalamin, where the upper ligand is replaced by cyanide during downstream processing (cyanide extraction), after microbial fermentation. However, the biologically active forms for cobalamindependent enzymes are either adenosylcobalamin (coenzyme B 12 ) or methylcobalamin (cofactor B 12 ). Microbes typically use cobalamin as a prosthetic group for enzymes in primary and secondary metabolism. P. megaterium possesses a number of cofactor B 12 or coenzyme B 12 -dependent enzymes that aid its survival in the environment. This includes ribonucleotide reductase (NrdJ), methionine synthase (MetH), methylmalonyl CoA mutase (MutAB), and ethanolamine lyase (EutBC). In particular, the coenzyme B 12 -dependent EutBC assimilates ethanolamine as a sole carbon and nitrogen source (Roof and Roth 1989;Wolf and Brey 1986).

P. megaterium as a model to study genetics and biosynthesis of cobalamin
The biosynthesis of cobalamin is complex and requires about 30 enzymes. Therefore, several decades of research has been required to fully decipher the biosynthesis of cobalamin. Originally, P. megaterium was identified as a natural producer of cobalamin by studying its growth on ethanolamine, with auxotroph mutants deficient in cobalamin biosynthesis (Wolf and Brey 1986). Since then, it has provided a suitable model to study the biosynthesis of cobalamin. The biosynthesis of cobalamin in P. megaterium can be summarized in three stages: stage 1, the synthesis of uroporphyrinogen III; stage 2, assembly of the corrin ring; and stage 3, attachment of the upper and lower axial ligands to the central cobalt ion . For P. megaterium DSM319, the majority of its cobalamin genes are organized into the following biosynthetic operons: hem (stage 1), cobI (stage 2), and cobII (stage 3) (Eppinger et al. 2011) (Fig. 2). An exception to this rule includes the genes encoding for cobalt transport (bmd_0328, cobO 1 , cobO 2 ) and a cobyric acid synthetase (cbiP), which are located separately within the genome (Fig. 2). Stage 1 uroporphyrinogen III (uro'gen III) biosynthesis is encoded by the hem operon (hemAXCDBL) in P. megaterium, whose genetic structure is similar to B. subtilis (Hansson et al. 1991). Then, cobalamin is built in two stages. For the first stage, P. megaterium operates the so-called anaerobic pathway to insert cobalt (Raux et al. 1998;Scott 2001) and build the corrin ring to achieve the first oxygen stable intermediate, cobyrinic acid (Moore et al. 2013b). Then the final steps in cobalamin biosynthesis attach the lower and upper axial cobalt ligands. This final stage is found in many prokaryotic lineages, since it also permits salvaging of vitamin B 12 precursors (Maggio-Hall and Escalante-Semerena 1999). Distinctly, P. megaterium assembles the lower ligand (5,6dimethylbenzimidazole) through an aerobic pathway for the final stages of its assembly (Collins et al. 2013). Therefore, P. megaterium has a customized cobalamin pathway to suit its requirement for molecular oxygen at different stages. , native signal peptide, purified from growth medium, crystallization 500 U L −1 (20.6 mg purified enzyme L −1 ) ( M a y e r e t a l . 2019) Bacillus species FJAT PGA 550 U L −1 (30.2 mg purified enzyme L −1 ) Bacillus thermotolerans PGA 220 U L −1 (15.2 mg purified enzyme L −1 ) Hybrid PGAs P xylA opt.
, native signal peptide, purified from growth medium n.d. , coproduction with redox partners AdR and Adx, codon , coproduction with redox partners AdR and Adx, codon optimized Whole-cell transformation hydroxylation of cholesterol, vitamin D 3 and 7-dehydrocholesterol (up to 113.14 mg L −1 48 h −1 ) (Ehrhardt et al. 2016) P. megaterium metabolic engineering of cobalamin production Since cobalamin requires approximately 60 chemical steps for its total synthesis (Battersby 2000), it is essential for biotechnology to make cobalamin through microbial production. P. megaterium provides an excellent host for producing cobalamin Martens et al. 2002;Moore et al. 2013a;Moore et al. 2013b;Moore et al. 2014). For example, while P. megaterium wild-type strains (DSM319, DSM509, and QM B1551) make only low levels (~0.2-1.0 μg L −1 ) of cobalamin in the lab, unpublished industrial strains are believed to reach up to 300 mg L −1 (Martens et al. 2002). Since cobalamin biosynthesis is complex (requires 30 enzymes), there are several bottlenecks that limit its production. This includes biosynthesis of precursors, import of cobalt (Fig. 2), feedback regulation, and rate-limiting enzymes. We will discuss how these individual steps can be optimized in P. megaterium.
The supply of precursors such as uroporphyrinogen III (uro'gen III) for the main tetrapyrrole scaffold and Sadenosyl-L-methionine (SAM) for methylation represents a major bottleneck for cobalamin biosynthesis in P. megaterium. For example, glutamyl-tRNA reductase (HemA) is regulated at both the transcriptional and posttranslation level (Schobert and Jahn 2002), through a negative feedback mechanism in heme biosynthesis. Overexpression of a proteolysis-resistant hemA mutant in P. megaterium increases cobalamin levels 11-fold to 2.8 μg L −1 . Furthermore, to increase uro'gen III supply directly, chromosomal overexpression of the uro'gen biosynthesis operon (hemAXCDBL operon) increases cobalamin levels up to 8.5 μg L −1 . Therefore, the supply of 5aminolevulinic acid (5-ALA) and uroporphyrinogen III is a major limiting factor in cobalamin biosynthesis.
Cobalt is essential for cobalamin biosynthesis (Martens et al. 2002). Crucially, like any transition metal, regulation is required to avoid toxicity. P. megaterium has a range of unique regulatory features to control cobalt levels. In P. megaterium DSM319, the addition of cobalt (1-10 μM) alone increases cobalamin levels up to 13 μg L −1 (Moore et al. 2014). However, cobalt homeostasis and its incorporation into cobalamin biosynthesis is poorly understood. At the enzyme level, the cobaltochelatase CbiX L inserts cobalt into the tetrapyrrole macrocycle and may play a role in cobalt h om e os t as i s ( Fig . 2 ) . O v er pro du ci ng C biX L i n P. megaterium DSM509 in the presence of cobalt increases cobalamin levels by 6-fold . Interestingly, CbiX L has an extended C-terminal domain that harbors a 4Fe-4S cluster and polyhistidine-rich motif (Leech et al. 2003). While the C-terminal extension is not essential for its chelatase activity (Leech et al. 2003), it may regulate or sense cobalt levels. For example, cobalt can substitute iron in Fe-S clusters (Ranquet et al. 2007) and polyhistidine motifs coordinate transition metals. For cobalt transport, P. megaterium has two potential cobalt transporters from the cbiO ATPase family or a single-component dual cobalt and nickel transporter (bmd_0328) (Komeda et al. 1997). Interestingly, bmd_0328 is part of two gene operons, containing an uncharacterized gene with another shorter polyhistidine motif (HXXXHH) (bmd_0329). Both genes are co-localized with an upstream cobalamin riboswitch (Fig. 2), suggesting their role in cobalamin biosynthesis. B 12 riboswitches are cisregulatory RNA elements that provide tight negative feedback control when cobalamin levels are high by sequestering either the Shine-Dalgarno site or by forming a transcription attenuator. While the role of bmd_0329 is unknown, overexpression of bmd_0328 in P. megaterium DSM319 leads to growth sensitivity in the presence of cobalt, suggesting increased cobalt import and toxicity (Moore 2011).
Unlike most prokaryotic metabolic pathways, there are no known specific transcription factors to regulate gene expression of cobalamin biosynthesis. Instead, global signals such as molecular oxygen repress cobalamin biosynthetic genes in Salmonella typhimurium (Escalante-Semerena and Roth 1987). Intriguingly, overproduction of the global anaerobic respiratory regulator FNR (fumarate and nitrate reductase regulator) in P. megaterium DSM509 increased cobalamin synthesis by 4-fold , suggesting that cobalamin biosynthesis is globally regulated by oxygen. Instead of transcription factors, cobalamin biosynthesis is regulated by cobalamin riboswitches. The P. megaterium DSM319 genome contains eight cobalamin riboswitches. This includes genes encoding B 12 -independent enzymes (metE, nrdEF), complete pathways (cobI and cobII operons), cobyric acid synthetase (cbiP), and cobalt homeostasis genes (bmd_0328-bmd_0329, bmd_0512 (see above)) (Fig. 2). The cobI operon is regulated by a cobalamin riboswitch and transcription terminator and is highly sensitive (nM levels) to cobalamin (Moore et al. 2014). This is not surprising since prokaryotic cells only require trace levels of cobalamin for unrestricted growth. Instead scavenging cobalamin from the environment (Nahvi et al. 2004) is also supported by an ABC transporter btuF (located within the cobII operon) and an uncharacterized transporter (bmd_0512), both co-localized with a cobalamin riboswitch (Fig. 2).
The cobalamin riboswitches represent the major bottlenec k in eng in eering cobalamin biosynthe sis i n P. megaterium (Fig. 2). To bypass this metabolic feedback, the entire cobI operon was placed under the control of a constitutive promoter on a multi-copy plasmid. Remarkably, expression in P. megaterium DSM319 in the presence of 10 μM cobalt led to major increases in cobalamin levels to 220 μg L −1 , a 27.5-fold increase over the control strain (Moore et al. 2014).
Production of biopolymers using P. megaterium: polyhydroxybutyrate (PHB) Polyhydroxyalkanoates (PHAs) are naturally occurring biopolymers synthesized by many microorganisms in response to environmental stress. They are considered to have promising potential to substitute traditional petrol-based plastics, as these so-called bioplastics show similar chemical and physical properties as conventional plastic (Chen 2009;Lu et al. 2009). PHAs were described in 1926 by the French scientist Lemoigne, who observed that P. megaterium accumulated polyhydroxybutyrate (PHB), a specific form of PHA, in the cells as distinct granules (Lemoigne 1926). Inside the cells, PHB acts as a storage device for carbon and energy and can be used again when conditions change. The hydrophobic granules are surrounded by a phospholipid monolayer in which a number of specific proteins are embedded, thereby associating with the granules (Jendrossek 2009). For the biological synthesis of PHAs, a variety of different C-sources, even crude waste material like glycerol derived from biofuel production, can be used (de Jesus et al. 2016;Naranjo et al. 2013;Solaiman et al. 2006). The key step in this process, the enzymatic polymerization of hydroxyacyl-coenzyme A (CoA) to PHA and CoA, is catalyzed by a PHA synthase. In 2001, McCool and Cannon identified the genetic organization of the five involved P. megaterium genes in two divergent orientated operons consisting of phaRBC and phaQP (McCool 2001). The phaC and phaR genes encode the two subunits of the PHA synthase (McCool 2001;Tsuge et al. 2015). Within the heterodimer, PhaC is the catalytic subunit localized with the granules, while PhaR is needed for polymerization (McCool 2001). PhaR from P. megaterium should not be confused with PhaR from other organisms like Ralstonia eutropha, where the name designates a transcriptional regulator of PHB synthesis (Lee et al. 2004). For the protein PhaB, a NADPH-dependent acetoacetyl coenzyme A reductase function was proposed. PhaB is involved in the supply of (R)-3HB-CoA monomer for the polymerization of PHB (Tsuge et al. 2015). PhaP as a phasin is localized with the granules (McCool and Cannon 1999). These non-enzymatic proteins are commonly found in PHA producers and have been shown to influence the PHA granule morphology and size (Jendrossek 2009). In P. megaterium the phaQ gene codes for a transcriptional regulator that negatively regulates the expression of phaP and phaQ. It interacts directly with PHB like PhaR in R. eutropha, although it has evolved independently (Lee et al. 2004). To ensure the abundant occurrence of phasin PhaP, but secure the required low level of the regulator PhaQ, Lee et al. speculate that the phaQ mRNA, as part of the phaQP transcript, is systematically degraded (Lee et al. 2004).
Optimizing PHB production in P. megaterium To date, there are hardly any attempts described to develop P. megaterium toward an increased production of PHB through genetic engineering. In contrast much effort has been placed on optimizing cultivation conditions of environmental isolates of P. megaterium to increase PHB production with mainly molasses as carbon source, resulting in almost 70% PHB of cell dry weight (Gouda et al. 2001;Rodríguez-Contreras et al. 2013). For the well-known and genome sequenced strain DSM319, Godard et al. (2020) observed that the PHB content increased 5-fold to almost 30 % of cell dry weight under high salt conditions (Godard et al. 2020). Further, the production of functionalized PHB granules provides an exciting new application for P. megaterium. One recent study showed that the mammalian cytochrome P450 CYP11A1 could be immobilized and purified with PHB granules produced in P. megaterium, thereby circumventing the problem of low stability of recombinantly produced cytochromes. Here, CYP11A1 was readily localized in the phospholipid monolayer of the PHB granule in its native form verified by denaturing PAGE (Stenger et al. 2018). Another study showed that the IgG binding domain of Protein A from Staphylococcus aureus (ZZ domain) could be produced, purified, and presented on PHB granules when fused to PhaC in P. megaterium. The isolated functionalized PHB beads were capable of purifying IgG from human serum, thereby proving their functionality (Grage et al. 2017).
Production of recombinant proteins using P. megaterium: intra-und extracellular formation at g/L scale The Gram-negative E. coli represents a well-established and heavily used host for the production and purification of recombinant proteins. However, E. coli has some major drawbacks including the presence of endotoxins (LPS) or limitations in the secretion of proteins into the growth medium, which permits easier downstream processing (Lakowitz et al. 2018;Terpe 2006). In contrast, Gram-positive bacteria lack an outer membrane, thereby omitting endotoxins and making protein secretion much more efficient. Many Grampositives have shown promising potential for recombinant protein production, for example, multiple members of the genus Bacillus (Terpe 2006).
In general, P. megaterium is renowned for its high stability of recombinant plasmids, even in the absence of selective antibiotics (Radha and Gunasekaran 2008). The production of recombinant proteins using P. megaterium is typically performed by plasmid-based gene expression. The majority of these plasmids are based on the oriU/repU system derived from the pBC16 plasmid, originally found in Bacillus cereus (Bernhard et al. 1978;, or on the compatible ori100/repM100 system from a plasmid found in P. megaterium QM B1551 (Eppinger et al. 2011;Gamer et al. 2009;Stevenson et al. 1998). In addition, the temperature-sensitive ori ts /E194ts system from plasmid pE194 is suited for genomic integration experiments . All plasmids are designed as shuttle vectors enabling all cloning in E. coli and subsequent transfer to P. megaterium via protoplast transformation ), a technique even suited for new environmental P. megaterium isolates (Boock et al. 2019). Standard antibiotics such as tetracycline, kanamycin, chloramphenicol, erythromycin, or spectinomycin can be used as selection markers for P. megaterium (Fig. 3). Importantly, all plasmids allow stable replication and production of recombinant proteins in B. subtilis or Bacillus licheniformis (Lakowitz et al. 2017;Larsen and Bjerga 2018).
Promoter systems for recombinant protein production in P. megaterium A controlled high-level production of recombinant proteins in P. megaterium is based on the native xylose-inducible promoter/repressor system. The promoter P xylA is induced in the presence of xylose through a de-repression mechanism based on the inactivation of the repressor XylR via xylose binding. The corresponding gene xylR is encoded on the same expression plasmid ). The P xylA -based expression system has undergone several systematic optimization steps resulting in intracellular recombinant protein production rates of up to 1 g L −1 (Stammen et al. 2010a) and more than 500 mg L −1 extracellularly (Korneli et al. 2013). This has resulted in the widely used P. megaterium recombinant protein production system, which comprised almost 30 plasmids and four strains and is available for commercial use from MoBiTec (Göttingen, Germany).
Moving on from the xylose-inducible system, a number of alternative promoter systems have recently been studied for protein production, expanding the P. megaterium plasmid toolbox (Fig. 3). Among these promoters are homologous sugar-inducible promoters (sucrose, arabinose, galactose, lactose) which were identified from transcriptome analyses. The employment of these systems for recombinant protein production resulted in up to 80 % yield compared to the optimized P xylA -based system (Biedendieck et al. 2007b;Hartz et al. 2019). Also a heterologous starch-inducible promoter (P a m y L ) from Bacillus amyloliquefaciens was found less effective compared to P xylA -based system (Radha and Gunasekaran 2008). Furthermore, the isopropyl-β-D-thiogalactopyranosid (IPTG)-inducible hyper-spank promoter (P Hysp ) yielded approximately 60 % of recombinant proteins compared to the P xylA -based system (Boock et al. 2019). As an alternative to these bacterial systems, phage-derived RNA polymerase (RNAP) systems have been successfully applied for protein production in multiple bacteria. For P. megaterium the genes coding for the RNAPs from the bacteriophage T7 (Gamer et al. 2009) and the E. coli phage K1E (Stammen et al. 2010b) residing on a separate plasmid were expressed under the control of P xylA . Upon xylose-based gene induction, the phage RNAPs are produced and specifically recognize the corresponding phage promoters driving the target gene expression localized on the second plasmid. The phage promoter-dependent gene expression resulted in up to 10 times more recombinant protein compared to the P xylAbased system (Gamer et al. 2009;Stammen et al. 2010b).
Beside inducible promoters, a number of constitutive and growth phase-dependent homologous promoters were tested for recombinant protein production. The promoters of the pyruvate dehydrogenase operon (pdhABCD) and of genes involved in glycolysis and gluconeogenesis (fba, fbp, gap, pgc, pgi, and pgk) from strain DSM319 yielded up to 75% of recombinant proteins compared to optimized P xylA -based system (Moore et al. 2018). In addition, some growth phase-dependent promoters were identified with slightly increased protein yields as the P xylA -based optimized system (Hartz et al. 2021). Fig. 3 Schematic summary of Priestia megaterium plasmids used for the production, secretion, and purification of recombinant proteins. All plasmids are constructed as shuttle plasmids for cloning in E. coli (yellow elements) and replication (dark blue, different compatibility classes), selection (blue), and production of recombinant proteins in P. megaterium. Suitable promoters (black arrow) are the native (P xylA ) and the optimized (P xylA opt. ) xylose-inducible promoter, the lactose inducible (P lac ), sucrose (P suc , P sacB ), arabinose (P ara ), galactosidase (P gal ), IPTG (P Hysp ), and starch (P amyL ) promoter, the T7-RNA-polymerasedependent promoter which is based on a two-plasmid system, and several constitutive (P const ) and growth phase-dependent (P growth ) promoters.
Genes encoding recombinant proteins can be fused to coding sequences of different signal peptides (purple) of the lipase A (SP LipA ), the unknown secreted proteins YocH (SP YocH ) and YngK (SP YngK ), the natural protease NprM (SP NprM ), and the serine protease VPR (SP Vpr ). In addition original signal peptides of the foreign recombinant protein can be used (SP native ). For purification of intra-or extracellular recombinant proteins, a fusion to N-or C-terminal His 6 or StrepII tag is possible (orange). Nterminal tags can be removed of using tobacco each virus (TEV) or factor X a protease cleavage (light blue). Black stars indicate stable places for integration of additional genetic elements as tRNAs or genes for coexpression.
Secretion of recombinant proteins with P. megaterium As mentioned above, Bacillus excels as a good secretion host for proteins. Ninety percent of all extracellular Bacillus proteins are secreted by the secretion(SEC)-dependent pathway (Tjalsma et al. 2004;Tjalsma et al. 2000) guided by an Nterminally fused signal peptide (SP). The nascent and unfolded polypeptide is directly secreted, prior to spontaneous folding. Subsequently, the SP gets cleaved off outside of the cell, the protein is folded and released into the growth medium (Freudl 2018). For B. subtilis and others, it has been demonstrated that the combination of a specific SP with a certain recombinant protein determines the efficiency of the overall secretion process which cannot be predicted (Brockmeier et al. 2006;Freudl 2018;Hemmerich et al. 2016;Mathiesen et al. 2008). The secretion of proteins using the SECdependent pathway provides an excellent route if a recombinant protein is known to form insoluble inclusion bodies intracellularly (Freudl 2018). For B. subtilis 173 SEC-dependent SPs were described (Brockmeier et al. 2006). A similar number of SEC-dependent SPs was identified for P. megaterium DSM319 (www.megabac.tu-bs.de; Hiller et al. 2004). The efficiency of seven P. megaterium SPs on secretion was evaluated using the heterologous Thermobifida fusca hydrolase, resulting in highly variable levels of secreted protein (Stammen et al. 2010a) (Fig. 3). A later study found similar effects when looking at the secretion of the endoglucanase EGI1 and the multimodular cellulase Cel9AT with five (for EGI1) and four (for Cel9AT) different SPs, respectively, tested (Kalbarczyk et al. 2018). Furthermore, some studies show secretion of recombinant proteins using their original SP including the dextransucrase DsrS from Leuconostoc mesenteroides (Malten et al. 2005a), penicillin G acylases from different Bacillus species (Mayer et al. 2019;Yang et al. 2006), and a keratinase from B. licheniformis (Radha and Gunasekaran 2007;Radha and Gunasekaran 2008) (Tab. 1). For most secretion experiments, the P. megaterium strain MS941, a DSM319 variant lacking the gene coding for the extracellular neutral metalloprotease NprM, was used as this strain reveals a reduction of 98.5% of extracellular protease activity (Wittchen and Meinhardt 1995).
Further adaptations of the recombinant plasmid system in P. megaterium Subsequently, to enable one-step protein purification, various tag-based affinity chromatography methods can be used (Terpe 2003). Thus, the P. megaterium plasmid systems were designed to create a plethora of N-and/or C-terminal fusion with His 6 -or StrepII-affinity tags in combination with protease cleavage sites for tag removal (Biedendieck et al. 2007c).
When combined with extracellular protein production, affinity tags provide rapid tools for protein purification, especially along with continuous cultivation processes (Gädke et al. 2017a;Gädke et al. 2017b). Moreover, a codon plus system employing the co-expression of genes for tRNAs with rare codons often found in heterologous target genes showed a general positive effect on recombinant protein production by P. megaterium (Finger et al. 2015) (Fig. 3). For genomic modifications, a plasmid system based on a temperature sensitive promoter was established allowing gene integration and deletion . Just recently, the group of Hannemann developed a CRISPR-Cas9 system for the genome editing of P. megaterium with an efficiency up to 100% (Hartz et al. 2021).
Production of biotechnological important proteins: multiple cytochrome P450 suitable for whole-cell transformations P450 enzymes are frequently employed in metabolic reactions to catalyze challenging chemical reactions (Pochapsky 2020). They use a broad range of structurally diverse substrates and form their products with high stereo-and regio-selectivity. Humans carry 57 different P450 (CYP) enzymes involved in the metabolism of steroid hormones, other sterols, vitamin D 3 , eicosanoids, fatty acids, and retinoic acid (Luo and Liu 2020;Rendic and Guengerich 2021;Sarparast et al. 2020). They catalyze almost exclusively monooxygenase reactions through the activation of molecular oxygen using a single electron. The major type of reaction is the hydroxylation of difficult to activate C-H bonds. For this purpose, they contain a heme group as single-electron transfer agent. Consequently, catalyzed reactions require electron donors like NADH or NADPH. Flavin or iron-sulfur proteins transfer electrons from NADH or NADPH to the P450-bound heme (Chiliza et al. 2020;Li et al. 2020).
Since most bacterial and fungal P450s are cytosolic and soluble, these variants are better suited for biotechnology applications compared to their membrane-bound plant and mammalian homologs (Distefano et al. 2021;Finnigan et al. 2020;Iizaka et al. 2021;Toplak et al. 2021;Zhang et al. 2021b). In P. megaterium a number of different P450 enzymes were found. Seven cytosolic P450s (CYP) from different P. megaterium strains have been described. CYP106A2 from strain ATCC13368 (Berg et al. 1976;Schmitz et al. 2018) and the CYP109E1 from strain DSM319 (Jóźwik et al. 2016) rely on a FAD-dependent ferredoxin reductase and a corresponding ferredoxin as electron donor and transfer proteins. The best studied CYP102A1 from strain DSM32, also known as cytochrome P450-BM3, consists of an Nterminal P450 domain followed by a flavodoxin and a reductase domain. Therefore, this in vitro system is self-sufficient and channels electrons directly from NADPH, which accelerates monooxygenase rate in comparison to other P450s (Cook et al. 2016;Miura and Fulco 1974;Whitehouse et al. 2012). The remaining four P. megaterium P450s all require an external redox partner. These include CYP106A1 from strain DSM32 (He et al. 1989;Lee et al. 2015) and from strain DSM319 (Brill et al. 2014), CYP109A2 from strain DSM319 (Abdulmughni et al. 2017a), and CYP107DY1 from strain QM B1551. Interestingly, CYP107DY1 is the first plasmid encoded P450 found in Bacillus species. Since no CYP107 homologs can be found in the genomes of P. megaterium strains, CYP107DY1 may have been obtained by horizontal gene transfer. This observation suggests a possible role of P450s in the adaptation and even evolution of bacteria (Milhim et al. 2016).
Whole-cell systems for recombinant production of P450s in P. megaterium To study P450s, this is typically performed using E. coli recombinant expression and studied in vitro with corresponding electron transfer proteins and cofactors. However, in vitro substrate conversion rates are often limited, which may constitute a problem for large-scale industrial applications. The use of whole-cell systems can overcome these in vitro limitations, although the import of the substrates and the export of the product might also be limited (Bernhardt and Urlacher 2014). Whole-cell systems have been biotechnologically employed for CYP106A1 from P. megaterium strain DSM319, CYP106A2 from P. megaterium strain ATCC13368 (Bleif et al. 2010;Bleif et al. 2012), and CYP109A2 (Abdulmughni et al. 2017a) and CYP109E1 (Abdulmughni et al. 2017b) from strain DSM319, all using P. megaterium as a production host (Tab. 1). In addition, E. coli-based whole-cell systems using CYP107DY1 from P. megaterium strain QM B 1551 (Milhim et al. 2016) and CYP102A1 from strain DSM32 (Chu et al. 2016) have been reported. Finally, membrane-bound mammalian P450 CYP11A1 was recombinantly produced in P. megaterium (Stenger et al. 2018).
Production of recombinant proteins: the P. megaterium cell-free transcription-translation system For future progress with P. megaterium, a novel cell-free transcription-transcription tool was recently developed to study fundamental molecular biology and accelerate the testing of gene expression systems (Moore et al. 2018). Within synthetic biology, there has been a renewed interest in cell-free transcription-translation systems (Tinafar et al. 2019;Cole et al. 2020). Cell-free systems require a cell extract, energy solution, and plasmid DNA to synthesize recombinant proteins. These reactions can be performed in the microscale range (nL to mL range) either in test tube reactions, microtiter (96, 384, 1536 well format) plates, or microfluidics (Laohakunakorn et al. 2020). While E. coli remains the dominant cell-free system, a range of new cell-free systems has recently been developed from other major prokaryotic expression systems, including B. subtilis, Streptomyces spp., Clostridium autoethanogenum, Pseudomonas putida, and Vibrio natriegens (Gregorio et al. 2019;Cole et al. 2020). In terms of protein yield, although many systems are still in development, cell-free protein production is often monitored with GFP as a standard. For this, P. megaterium (134 ng μL −1 GFP) compares favorably to other popular Grampositive hosts such as B. subtilis (21.6 ng μL −1 ) and Streptomyces lividans 66 (~100-400 ng μL −1 ) (Cole et al. 2020).
For P. megaterium DSM319, an optimized cell-free protocol (also active in DSM509) was recently developed (Moore et al. 2018). One key advantage of cell-free systems is the study of the biological numbers that underpin protein synthesis in combination with computational prediction models (Moore et al. 2018). A key finding of the P. megaterium cell-free system was potential rate-limiting steps in protein synthesis. For example, under the conditions studied, maximal translation rates were an order of magnitude slower than that of E. coli cell-free (Garamella et al. 2016).
In addition, the kinetics of the xylose-inducible promoter system (see above) were characterized in detail. For example, the dissociation constant for XylR and operator binding was determined at 12.9-14.2 nM with a Hill coefficient of 1.74-1.8. Finally, as an example of the power and speed of cell-free systems, using a liquid handling robot, up to 500 P. megaterium plasmids with varying promoter and RBS regions were rapidly screened for activity within 24 h. This permits the rapid characterization of DNA plasmid designs for forward engineering in P. megaterium cells (Moore et al. 2018). In summary, this recent development provides a highyield and rapid cell-free tool for the study and engineering of P. megaterium for future metabolic engineering and synthetic biology applications.
Whole-cell applications: P. megaterium as a plant growth-promoting bacterium Recently, within the last 20 years, scientists have discovered that the plant microbiome is essential for the survival of plants in changing environmental conditions including invasion by pathogenic microorganisms and insects (Berger and Gutjahr 2021;Genre et al. 2020;Haskett et al. 2020;Ortiz and Sansinenea 2021;Prsic and Ongena 2020;Vishwakarma et al. 2020;Zhang et al. 2021a). In conclusion, the health and growth of cultural plants can be influenced by the composition of its root and leave microbiome (Ray et al. 2020). In recent years, the beneficial effect of P. megaterium on plant growth has become a growing matter of interest. It has been described for a number of different plants including the model organism Arabidopsis thaliana (López-Bucio et al. 2007;Ortíz-Castro et al. 2008), the commercially important plants tomato (Solanum lycopersicum) (Ibort et al. 2017;Porcel et al. 2014), tea (Camellia sinensis) (Chakraborty et al. 2006), maize (Zea mays) (Al-Enazy et al. 2017;Marulanda et al. 2010), mustard (Brassica juncea L.) (Kang et al. 2014;Rajkumar and Freitas 2008), rice (Oryza sativa L.) (Feng et al. 2017), bean (Phaseolus vulgaris) (Korir et al. 2017;Ortíz-Castro et al. 2008), soybean (Glycine max) (Zhou et al. 2017), and oilseed rape (Brassica napus) (Hu et al. 2013). Three different mechanisms of plant growth promotion by P. megaterium have been described.
The first role can be summarized as biofertilizer. Phosphorus is essential for plant growth; however, its bioavailable form is often present in very low amounts (Liu 2021). The transformation of phosphorus in minerals and organic sources to their bioavailable forms occurs through secretion of organic acid in combination with acid phosphatases and phytases (Kang et al. 2014;Martínez-Viveros et al. 2010). P. megaterium secrets organic acids providing the main basis of phosphate biofertilization using secreted acid phosphatase and phytases (Hu et al. 2013). In addition, P. megaterium can provide reduced nitrogen to plants (Ding et al. 2005;Liu et al. 2006;Singh et al. 2020). Currently, a number of different P. megaterium fertilizer preparations often in combination with other bacteria are commercially available by different manufacturers for large-scale agricultural applications. These combinations including P. megaterium are also subjects of different patents. Wang et al. claim the release of potassium and phosphorus, the fixation of nitrogen, the inhibition of harmful bacteria in the soil, and the prevention of different diseases using a mixture of P. megaterium, further Bacillus strains and Gram-negative bacteria (Wang et al. 2009). Recently a fertilizer and its preparation was patented for the hydrolysis of phosphorus containing only P. megaterium but combined with organic matter and ammonium sulfate (Jianzhong 2019).
Secondly, plant growth-promoting bacterium can also be responsible for changes in environmental concentrations of phytohormones and other regulators of plant growth. Interestingly, the production of the auxin indole acetic acid (IAA) by P. megaterium resulted in a plant growthpromoting effect, as reported repeatedly for different plants (Chakraborty et al. 2006;Feng et al. 2017). Furthermore, the cultivation of A. thaliana with P. megaterium lead to higher concentrations of the isoprenoid plant hormone abscisic acids (ABA) in plant leaves, thus improving drought stress tolerance (Zhou et al. 2016). It was postulated that secretion of the polyamine spermidine by P. megaterium was responsible for the upregulation of the expression of ABA-associated genes and subsequent production of ABA in the plant. Ortíz-Castro et al. reported that cytokinin signaling plays a central role in the plant growth-promoting effect by P. megaterium on A. thaliana under defined lab conditions (Ortíz-Castro et al. 2008). In addition to the classical phytohormones, the role of so-called volatile compounds as plant growth-promoting substances is recently gaining more interest (Ryu et al. 2003;Sharifi and Ryu 2018). For P. megaterium the positive effect of the volatile compound 2-pentylfuran on the growth of A. thaliana has been observed (Zou et al. 2010). The mechanism by which this compound promotes plant growth is still unknown. Finally, acetoin produced by a P. megaterium strain promoted the growth of A. thaliana (Ryu et al. 2003).
Thirdly, P. megaterium can act as a biopesticide or biocontrol agent. The multiple anti-pathogenic mechanisms of P. megaterium are divers. An antifungal activity against the tea pathogen Fomes lamaoensis, the cause of brown root rot, is possibly related to the production of iron-chelating siderophores. Additionally, the enhanced secretion of the plant defense-related peroxidase, phenylalanine ammonia lyase, chitinase, and β-1,3-glucanase by P. megaterium was observed (Chakraborty et al. 2006). These enzymes are postulated to act directly against the fungal cell wall, thereby protecting the plant. The secretion of chitinase, β-1,3glucanase, and protease by P. megaterium also mediated a protective effect against the fungus Rhizoctonia solani, the causative agent of "damping-off", a destructive disease of plant seedlings, and against root rot in tomato (Lycopersicon esculentum Mill) (Solanki et al. 2012). The treatment of R. solani-caused diseases with P. megaterium strain ATCC55000 was patented in 1995 where its function as a biological control agent was described but lacks explanation. Moreover, the additional role of strain ATCC55000 in stimulating growth and yield in soybeans was part of the invention (Liu and Sinclair 1995). Another antifungal property of P. megaterium mediated by an unidentified volatile compound was found against the aflatoxin-producing Aspergillus flavus found on rice grains (Mannaa et al. 2017). A mixture of the three rhizobacterial bacteria P. megaterium, (Peri)Bacillus simplex, and Sinorhizobium fredii coated on soybean seeds revealed a clear protective effect against the "soybean cyst nematode" (Heterodera glycines) (Zhou et al. 2017).
Finally, coming back to the P. megaterium P450s, CYP102A1 (P450-BM3) most likely plays an important role in the regulation quorum sensing by soil bacteria through the inactivation of acyl homoserine lactones (AHLs). These molecules are known signaling molecules in the communication of Gram-negative bacteria (Chowdhary et al. 2007). Interestingly, the ability of P. megaterium to degrade AHLs suggests a link of their plant-protective, quorum-quenching activity to the quorum sensing of plant pathogenic bacteria (Dong et al. 2001). In summary, P. megaterium revealed a whole variety of molecular strategies of plant growthpromoting effects.

Conclusion and perspectives
P. megaterium is a fast-growing giant cell factory, with past and current industrial applications, and a promising alternative to standard model organisms (Eisenstein 2016). In 1994, Patricia S. Vary wrote "Prime time for Bacillus megaterium" (Vary 1994) which was followed 13 years later by the update "Bacillus megaterium -from simple soil bacterium to industrial protein production host" (Vary et al. 2007). Now, 14 years later, we have a new name for our well-known bacterium, while there are more than 20 fully sequenced genomes, a commercialized recombinant plasmid toolkit for high-yield recombinant protein production to the g per L scale, and a first cell-free transcription-translation in vitro system. In addition, our knowledge of whole-cell systems for the production of important P450 enzymes and also as a plant growthpromoting bacterium, the use in the production of bioplastics, and a prominent role in the understanding of B 12 biosynthesis increased significantly.
So, what is needed next to develop P. megaterium into a competitive biotechnological production host? To support rapid and more straightforward biotechnological research, a complete genome-level single-gene knock-out library will provide an accurate picture of all non-essential genes, as is available for many model microbes. This is desirable to identify genes that may limit/benefit questions relating to recombinant protein production, metabolic capacity, or selfregulatory processes. Although the transformation of P. megaterium protoplasts is completely sufficient for the introduction of single plasmids, to test entire gene banks in this organism, a better transformation system is also needed. Since almost all genes necessary for the formation of natural competence are present in P. megaterium, an easier integration of DNA should be possible, which will serve as another important developmental feature. In addition, due to its natural size, P. megaterium represents a perfect tool for cell biological studies. Combining time-lapse microscopic studies with corresponding bioinformatic tools and modeling approaches is of interest (Münch et al. 2015).
Funding Open access funding enabled and organized by Projekt DEAL. This work was funded by the German Research Foundation (DFG), within the priority program SPP1617, "Phenotypic heterogeneity and sociobiology of bacterial populations" (TK).

Compliance with ethical standards
Ethics approval This article does not contain any studies with human participants or animals performed by any of the authors.

Conflict of interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.