Functional characterization of the phosphotransferase system in Parageobacillus thermoglucosidasius

Parageobacillus thermoglucosidasius is a thermophilic bacterium characterized by rapid growth, low nutrient requirements, and amenability to genetic manipulation. These characteristics along with its ability to ferment a broad range of carbohydrates make P. thermoglucosidasius a potential workhorse in whole-cell biocatalysis. The phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS) catalyzes the transport and phosphorylation of carbohydrates and sugar derivatives in bacteria, making it important for their physiological characterization. In this study, the role of PTS elements on the catabolism of PTS and non-PTS substrates was investigated for P. thermoglucosidasius DSM 2542. Knockout of the common enzyme I, part of all PTSs, showed that arbutin, cellobiose, fructose, glucose, glycerol, mannitol, mannose, N-acetylglucosamine, N-acetylmuramic acid, sorbitol, salicin, sucrose, and trehalose were PTS-dependent on translocation and coupled to phosphorylation. The role of each putative PTS was investigated and six PTS-deletion variants could not grow on arbutin, mannitol, N-acetylglucosamine, sorbitol, and trehalose as the main carbon source, or showed diminished growth on N-acetylmuramic acid. We concluded that PTS is a pivotal factor in the sugar metabolism of P. thermoglucosidasius and established six PTS variants important for the translocation of specific carbohydrates. This study lays the groundwork for engineering efforts with P. thermoglucosidasius towards efficient utilization of diverse carbon substrates for whole-cell biocatalysis.

www.nature.com/scientificreports/ to ferment a broad range of monosaccharides, cellobiose, and short-chain oligosaccharides. Its success in bioethanol production, the possibility of genetic manipulation, and recent whole-genome sequencing have highlighted P. thermoglucosidasius as a potential future cell factory for other valuable small molecules. To further increase the understanding and leveraging of P. thermoglucosidasius, it is necessary to characterize the metabolism of the microorganism, including its carbohydrate metabolism and transportation. In prokaryotes, the transport of carbohydrates is mainly catalyzed by the phosphoenolpyruvate(PEP):carbohydrate phosphotransferase system (PTS) 15 . The PTS couples the transport of carbohydrates with subsequent phosphorylation through a four-step phosphoryl transfer system 15,16 . Each PTS consist of two cytoplasmatic proteins, the PTS-general component Enzyme I (EI) that receives the phosphate from PEP, and the histidine-containing phosphocarrier protein (HPr), which is phosphorylated by EI along with a substrate-specific Enzyme II (EII) complex ( Fig. 1) 15,16 . Generally, EI and HPr are common to all PTSs of a cell, meaning that they perform the phosphoryl transfer to all the different EII complexes. Each EII complex is formed by two cytoplasmatic domains; EIIA, which is phosphorylated by HPr, and EIIB, which is phosphorylated by EIIA; and one or two integral membrane domains (EIIC/EIID) that are necessary for substrate translocation (Fig. 1). Furthermore, the three or four EII domains could be either encoded in a single multi-domain protein, or in distinct single-domain proteins 15 . The PTS participates in complex regulatory mechanisms, including both carbon and nitrogen metabolisms. In summary, in low G+C DNA Gram-positives, HPr also works as a sensor of glycolytic intermediates, especially for fructose-1,6-bisphosphate (FBP). High concentrations of FBP increase phosphorylation of HPr on the conserved serine-46 (different to the histidine involved in the PTS phosphorelay) triggering carbon catabolite repression (CCR), generally through the carbon catabolite repressor CcpA 15 . Furthermore, CcpA also regulates the synthesis of branched-chain amino acids which directly stimulates the global regulator CodY, linked to both carbon and nitrogen metabolism 17 . For detailed explanations of the regulatory networks refer to Deutscher et. al (2006) and Sonenshein (2007).
In this study, we investigated the role of the PTSs in the transport of common carbon substrates in P. thermoglucosidasius. By constructing PTS knockouts and measuring the growth of P. thermoglucosidasius on fifteen common carbon substrates, we were able to determine some of the PTSs specificities in P. thermoglucosidasius and show that there is a complex redundancy between different PTS systems. Of the fifteen substrates, thirteen showed a dependence on active PTS-mediated transport. Knockouts of a minimum of one PTS element in each of the fifteen putative PTS gene clusters in P. thermoglucosidasius revealed five PTSs solely responsible for the translocation of arbutin, mannitol, N-acetylglucosamine, sorbitol, and trehalose, respectively, and the main PTS responsible for the translocation of N-acetylmuramic acid. This study establishes the basis for further metabolicand strain engineering of P. thermoglucosidasius for novel biotechnological solutions.

Results
Genome analysis and variant design. To assess the capacity of P. thermoglucosidasius DSM 2542 to metabolize carbon substrates through PTS and associated elements thereof, we searched on the available genome sequence (GenBank Accession No. CP012712) 9 for genes encoding putative PTSs. The genomic analy- Figure 1. Schematic representation of the general bacterial phosphotransferase system (PTS). The two general cytoplasmatic components, which are used for all putative PTS systems are indicated in pink. These are the PTSgeneral component Enzyme I (EI) and the histidine-containing phosphocarrier protein (HPr). The area marked in tan is a representation of an Enzyme II (EII) complex. EII complexes confer the carbohydrate specificity and are specific for each PTS system. Each EII complex is formed either by distinct proteins or by a single multidomain protein and consists of two hydrophilic domains (EIIA and EIIB), and one or two transmembrane domains (EIIC and EIID). www.nature.com/scientificreports/ sis, performed as indicated in Bioinformatic analysis in the Materials and Methods section, revealed that P. thermoglucosidasius contains 15 genomic regions with at least one such PTS element (Fig. 2), scattered throughout the bacterial genome. A minimum of one PTS-associated gene from each gene cluster was knocked out by a scarless-genome edition method based on allelic replacement 6 , yielding 15 deletion strains. In addition, the gene encoding the EI (ptsI; AOT13_08105) ( Fig. 1) was also knocked out, thereby yielding a total of 16 deletion variants of P. thermoglucosidasius ( Fig. 2 and Supplementary Fig. 16).

Knockout of ptsI.
After the mutants were constructed, we focused on quantitative physiology experiments towards identifying the associated growth phenotypes. To this end, 15 carbon sources were selected for their potential as substrates for P. thermoglucosidasius. Besides the carbon sources typically used for bacterial growth experiments (e.g., glucose, fructose, mannose, and xylose), other substrates were likewise included (N-acetylglucosamine, N-acetylmuramic acid, glycerol, mannitol, sorbitol, cellobiose, maltose, sucrose, trehalose, arbutin, and salicin). To determine which of the 15 substrates were transported by the PTS in P. thermoglucosidasius, the WT strain along with the common enzyme knockout strain, ∆ptsI (strain GTS17 in Table 1), were grown in minimal media supplemented with the 15 carbon substrates, respectively. Comparing the growth profiles of these two strains revealed that the WT strain grew in all conditions (Fig. 3A), while the ∆ptsI deletion mutant had abolished growth with 12 out of the 15 feedstocks (Fig. 3B). The ∆ptsI strain had no significant growth when the medium was supplemented with arbutin, cellobiose, fructose, glucose, glycerol, mannitol, mannose, N-acetylglucosamine, sorbitol, salicin, sucrose, or trehalose. These results indicate PTS-dependent processing of the 12 molecules. Although we cannot exclude that other functional elements could be involved in the transport of the substrates tested herein (e.g., hexose permeases or facilitators, which could promote growth on glucose or fructose) [18][19][20] . In addition to the 12 carbon substrates on which growth was abolished, N-acetylmuramic acid yielded diminished growth indicating that this substrate is mainly, but not exclusively, metabolized through PTS. Interestingly, the ∆ptsI strain was still able to grow when the media was supplemented with maltose or xylose.
These results indicate that the transport of these sugars is either independent of phosphorylation processes catalyzed by PTSs or, as indicated above, could be mediated by another transport mechanism, such as a permease.

Discussion
P. thermoglucosidasius-related species hold potential to produce valuable compounds from a broad variety of carbon sources. Rapid growth, availability of genetic tools, and, most importantly, its thermophilic nature, make this strain a prominent workhorse for large-scale industrial production. In light of the features, in this study, we systematically investigated the role of the PTSs of P. thermoglucosidasius in the transport of 15 common molecules used as carbon feedstock by knocking out both common and specific PTS components. The elimination of some components of the PTS systems affected carbon substrate utilization on the host strain. The extent of this impact is variable, and carbon substrate-dependent. Knockout of the shared EI enzyme resulted in abolished growth for all tested carbohydrates except maltose, xylose, and N-acetylmuramic acid. Although many carbohydrates depend on PTS-mediated translocation with subsequent phosphorylation, other transporters are also known for transporting carbohydrates in bacteria. Here, the ATP-binding cassette (ABC) and the major facilitator superfamily (MFS) transporters also play a major role in the sugar uptake in prokaryotes [21][22][23] . The uptake of maltose has previously been described to occur through an ABC transporter in other bacterial species [24][25][26][27] . However, previous works on Bacillus have had contradictory results regarding maltose transport. Reizer et al. 27 showed that inactivation of the putative EIIBC (malP) in B. subtilis resulted in a seven-fold increase of the doubling time in a minimal medium supplemented with maltose. Similarly, Schönert et al. 28 have shown a lack of [ 14 C]maltose uptake in cell suspensions of B. subtilis ΔmalP after growth in LB supplemented with 1% maltose. Contrary, both in B. subtilis and in B. licheniformis, another major industrial Gram-positive bacteria, it has been suggested that maltose is transported by a proton symport mechanism, which does not occur via the PTS but is regulated by the PTS, and further metabolized through a maltose phosphorylase 29,30 . Furthermore, complementation of an E. coli strain deficient for maltose transport genes, with the MFS transporter MalA from Geobacillus stearothermophilus suppressed the growth defects on maltose 31 . A homologous genomic region encoding the same putative genes, including an MFS transporter homologue to MalA, is also found in P. thermoglucosidasius (AOT13_18465). In addition, a homologue of the maltose/maltodextrin transport system permease protein MalG from E. coli, part of the MalEFGK ABC transporter complex 32 , is also found in P. thermoglucosidasius (AOT13_07020). The results presented in our work, the presence of a MalA homologue, and the presence of a putative maltodextrin phosphorylase in the genome of P. thermoglucosidasius DSM 2542 (AOT13_18705), suggest a PTS-independent pathway. www.nature.com/scientificreports/ Xylose is transported across the cell membrane either mediated by a PTS-independent mechanism or by a PTS-dependent but phosphorylation-independent mechanism. Facilitated diffusion of xylose catalyzed by the enzyme II complex (EII) of the PTS specific to mannose, has previously been reported in three species of www.nature.com/scientificreports/ lactobacilli (L. pentosus, L. plantarum, and L. casei) 33 . The transport was demonstrated to be independent of phosphorylation, which could explain why the growth of ∆ptsI supplemented with xylose remains unaffected. Genome analysis of six Geobacillus strains showed that xylose transport and metabolism are encoded in a gene cluster containing ABC transporters 34 . It has also been shown that xyl genes in thermophilic Bacillus sp. are clustered encoding xylO (ATP-binding protein), xylP (xylose permease), xylA (xylose isomerase), xylB (xylulose kinase) 35 , and although P. thermoglucosidasius DSM 2542 has an operon with both homologues to xylA and xylB (AOT13_11570 and AOT13_11575 respectively), it lacks homologues of xylO and xylP. Since strain ∆10525-10530 is not able to grow when supplemented with xylose, the EII complex encoded in that gene cluster probably catalyzes the facilitated diffusion of xylose with a similar mechanism as reported for lactobacilli, which would further be metabolized through the operon xylAB. In addition to the described mechanisms, xylose has been shown to be transported by AraE (part of the MFS) in Bacillus subtilis 23 . However, no homologue of AraE is found in P. thermoglucosidasius to our knowledge. In both Gram-negative and -positive bacteria, N-acetylmuramic acid is mainly transported across the membrane by the PTS MurP and subsequently phosphorylated yielding the 6-phospho sugar utilized for peptidoglycan formation, or as carbon source through the N-acetylglucosamine-6P degradation pathway 36,37 . This seems to be the case also for P. thermoglucosidasius DSM 2542, which contains a putative operon encoding homologues to the regulator MurR, the etherase MurQ and the transporter MurP. Moreover, this is supported by the diminished growth of the strain ∆11075 on this carbon source.
Still remain to be determined which mechanism allows the partial metabolism of N-acetylmuramic acid observed on the strains ∆ptsI and ∆11075. In bacteria, the first PTS-independent N-acetylmuramic acid transporter has been identified in the periodontal pathogen Tannerella forsythia 38 . This organism contains an operon www.nature.com/scientificreports/ encoding a specific N-acetylmuramic acid transporter, a sugar kinase, and a MurQ etherase. No homologues of T. forsythia transporter and kinase are found in the genome of P. thermoglucosidasius, however, some of its many uncharacterized transporters and kinases could have unspecific activity towards N-acetylmuramic acid. For all carbon sources tested, except for the three discussed above, growth was inhibited by the deletion of ptsI. While this is not generally surprising, we could have expected PTS-independent growth on glycerol. This triol is known to be transported both in Gram-negative and Gram-positive bacteria by energy-independent diffusion mediated by GlpF, a conserved glycerol uptake facilitator 39,40 . Although P. thermoglucosidasius DSM 2542 encodes the corresponding homologue (AOT13_09870), the mutant ∆ptsI had impaired growth when the medium was supplemented with glycerol. Previous studies also report inhibited growth on glycerol by mutants of Gram-positive and Gram-negative bacteria defective in one of the common enzymes of the PTS [41][42][43][44][45][46] . The mechanisms involved in this regulation are different in Gram-negative compared to Gram-positive bacteria but both are mediated by the glycerol kinase responsible for the formation of glycerol-3-phosphate trapping the substrate in the cell upon uptake 41 . For Gram-positive bacteria, the glycerol kinase was found to be phosphorylated by PEP and the common enzymes of the PTS; EI and HPr, causing an increase in glycerol kinase activity 41,43,46,47 . In fact, the His-232 of glycerol kinase from Enterococcus casseliflavus has been identified as the site of PEP-dependent PTS-catalyzed phosphorylation 47 . Given that glycerol kinase from P. thermoglucosidasius DSM 2542 (AOT13_09875) shares 64% homology to the enzyme from E. casselliflavus, and it contains the highly conserved histidine-232 residue, we suggest the same regulatory mechanism.
The findings of this study could have important implications for the future scalability and industrial applications of P. thermoglucosidasius as a cell factory or whole-cell biocatalysis. By identifying the specific PTS systems responsible for the transport and phosphorylation of various carbon substrates, this study lays the groundwork for future engineering efforts aimed at enhancing the strain's ability to efficiently utilize diverse carbon sources. www.nature.com/scientificreports/ Such efforts could potentially lead to the development of a highly versatile whole-cell biocatalyst capable of converting a wide range of substrates into valuable products. Overall, this study provides valuable insights into the metabolic capabilities of P. thermoglucosidasius.

Materials and methods
Bacterial strains and plasmids. The strains and plasmids used in this study are listed in Table 1 and Supplemental table 1   Construction of recombinant strains. Construction of the mutant strains was performed by two-step allelic exchange through homologous recombination, exploiting the native machinery of P. thermoglucosidasius DSM 2542 51 . DNA fragments containing flanking regions of the targeted genes designed to include only the start and stop codons of the knockout-target genes were obtained by PCR using P. thermoglucosidasius DSM 2542 chromosomal DNA and the corresponding oligonucleotide pairs. The PCR products were purified using a Nucle-oSpin Gel and PCR kit (Macherey-Nagel, Germany) and cloned into the backbone of pMTL61110 (obtained by PCR using primers 23 and 24d) by USER cloning (New England Biolabs, US). Chemically competents E. coli DH5α were transformed by heat shock with 3 μl of the USER reactions and after 1h recovery cells were plated on LB with kanamycin. The resulting plasmids were purified using a NucleoSpin Plasmid kit (Macherey-Nagel, Germany), and its sequence verified (Eurofin Genomics). P. thermoglucosidasius DSM 2542 electrocompetents 6 were transformed with each of these plasmids using a single electric pulse in a Bio-Rad GenePulser Xcell (10 μF, 600 Ω, 25 kV/cm) and recovered in 1mL of pre-warmed SPY supplemented with 1% glycerol at 52 °C with agitation (200 rpm) for 3 hours. Selection of kanamycin-resistant colonies was done on TSA plates with 12.5 μg ml −1 kanamycin overnight at 52 °C. A kanamycin-resistant colony from each transformation was incubated overnight at 62 °C on SPY supplemented with kanamycin to force the first recombination and later plated on TSA plates with 12.5 μg ml −1 kanamycin. After confirming the right integration site by colonyPCR, the selected colonies were grown 3 days in 5mL of SPY without kanamycin at 60 °C with agitation (200 rpm), doing subcultures in fresh media each morning and evening. Cells were plated on TSA and incubated at 60 °C overnight. The next day the plates were replicaplated on TSA plus kanamycin. Antibiotic-sensitive clones were isolated and, among them, one for each gene was selected (GTS17-GTS32 strains) in which a second recombination event led to the excision of the plasmid and deletion of the targeted gene. DNA sequencing reactions of the appropriate PCR products (see Supplementary Fig. 16 N-acetyl-muramic acid; 20mM cellobiose, maltose, sucrose, or trehalose; 8mM arbutin or salicin). All carbohydrates/glucosides were purchased from Sigma-Aldrich (USA) or Carbosynth (Compton, Berkshire, UK). Cultures were incubated in 96-well plates sealed with Titer-Tops® (Sigma-Aldrich, USA) at 60 °C with agitation (567 cpm) in an Epoch2 microplate spectrophotometer (BioTek, Agilent, USA), and bacterial growth was monitored every 10 minutes for 24 h measuring OD 600 nm . At least three independent biological replicates for each growth curve were obtained. Results were expressed as means ± standard deviations.

Data availability
All data generated or analyzed during this study are included in this published article (and its Supplementary Information files).