Heterologous 1,3-Propanediol Production Using Different Recombinant Clostridium beijerinckii DSM 6423 Strains

1,3-propanediol (1,3-PDO) is a valuable basic chemical, especially in the polymer industry to produce polytrimethylene terephthalate. Unfortunately, the production of 1,3-PDO mainly depends on petroleum products as precursors. Furthermore, the chemical routes have significant disadvantages, such as environmental issues. An alternative is the biobased fermentation of 1,3-PDO from cheap glycerol. Clostridium beijerinckii DSM 6423 was originally reported to produce 1,3-PDO. However, this could not be confirmed, and a genome analysis revealed the loss of an essential gene. Thus, 1,3-PDO production was genetically reinstalled. Genes for 1,3-PDO production from Clostridium pasteurianum DSM 525 and Clostridium beijerinckii DSM 15410 (formerly Clostridium diolis) were introduced into C. beijerinckii DSM 6423 to enable 1,3-PDO production from glycerol. 1,3-PDO production by recombinant C. beijerinckii strains were investigated under different growth conditions. 1,3-PDO production was only observed for C. beijerinckii [pMTL83251_Ppta-ack_1,3-PDO.diolis], which harbors the genes of C. beijerinckii DSM 15410. By buffering the growth medium, production could be increased by 74%. Furthermore, the effect of four different promoters was analyzed. The use of the constitutive thlA promoter from Clostridium acetobutylicum led to a 167% increase in 1,3-PDO production compared to the initial recombinant approach.


Introduction
The demand for 1,3-propanediol (1,3-PDO) has been increasing in the last few years and will continue to rise in the coming years. With a "Compound Annual Growth Rate" of 14.2% as of 2018, an optimistic market expectation for 1,3-PDO is US$ 1,442.77 million in 2027 [1]. 1,3-PDO is an important basic chemical and is widely used as an organic solvent in the food, cosmetics, and pharmaceutical industries. However, the main application of 1,3-PDO is in the polymer industry as a raw material or intermediate, especially as a key monomer in the production of polytrimethylene terephthalate [2,3]. In the past, ethylene oxide hydroformylation and acrolein hydration-hydrogeneration were the two processes mainly used for the chemical synthesis of 1,3-PDO. Both processes depend on petroleum products as precursors for 1,3-PDO production [4][5][6]. However, those processes have many disadvantages, such as high investment, technical difficulties, substrate toxicity, and environmental issues [2,7]. Therefore, a biobased synthesis of 1,3-PDO is desirable.
A biosynthetic 1,3-PDO production approach was developed by DuPont and Genencor in the early 2000s using recombinant Escherichia coli cells and glucose as substrate [8]. The process was commercialized by DuPont Tate & Lyle in Loudon (USA) in 2006, with a production capacity of 63,500 tons per year [9][10][11]. Another possibility is the use of glycerol or crude glycerol as a substrate for the microorganisms [12,13]. About 10% crude glycerol is generated as the main by-product from the biodiesel production. With the expanding biodiesel market, the availability of crude glycerol has increased over the last few years. As coenzyme B12-dependent glycerol dehydratase (DhaB, DhaC, DhaE) hydroxypropionaldehyde (3-HPA), followed by the conversion of 3-HPA via a N linked 1,3-PDO dehydrogenase (DhaT) to 1,3-PDO [27]. The glycerol reductive pa of C. butyricum and C. beijerinckii consists of a coenzyme B12-independent gl dehydratase (DhaB1, DhaB2) for the dehydration of glycerol to 3-HPA, also follow the conversion of 3-HPA to 1,3-PDO via a NADH-linked 1,3-PDO dehydrogenase ( [28,29]. However, glycerol is not only reduced to 1,3-PDO but also oxidized and us biomass and by products e.g., acetate, butyrate, ethanol, and butanol [30]. An overv the glycerol consumption pathway is given in Figure 1.
In the past, most publications focused on the conversion of glycerol to 1,3-PDO wild-type clostridial strains. Clostridia are difficult to manipulate genetically, due limited availability of genetic tools as well as the occurrence of native restr modification systems, resulting in major obstacles in obtaining recombinant s [31,32]. Therefore, development of new genetic tools and methods to overcome restr modification systems are essential for the construction of an optimized production for the waste valorization from glycerol to 1,3-PDO in the future. One of thes developments is the newly published transformation protocol for C. beijerinckii DSM [31], which was used in this study to create recombinant 1,3-PDO production strain The two different 1,3-PDO production pathways form C. pasteurianum DSM 52 C. beijerinckii DSM 15410 were expressed in C. beijerinckii DSM 6423 in a plasmidmanner. Furthermore, the effect of different promoters on 1,3-PDO production examined in this study. The results presented here demonstrate the heterol production of 1,3-PDO with genetically engineered C. beijerinckii DSM 6423 strains f first time.  In the past, most publications focused on the conversion of glycerol to 1,3-PDO with wild-type clostridial strains. Clostridia are difficult to manipulate genetically, due to the limited availability of genetic tools as well as the occurrence of native restriction-modification systems, resulting in major obstacles in obtaining recombinant strains [31,32]. Therefore, development of new genetic tools and methods to overcome restriction-modification systems are essential for the construction of an optimized production strain for the waste valorization from glycerol to 1,3-PDO in the future. One of these new developments is the newly published transformation protocol for C. beijerinckii DSM 6423 [31], which was used in this study to create recombinant 1,3-PDO production strains.
The two different 1,3-PDO production pathways form C. pasteurianum DSM 525 and C. beijerinckii DSM 15410 were expressed in C. beijerinckii DSM 6423 in a plasmid-based manner. Furthermore, the effect of different promoters on 1,3-PDO production was examined in this study. The results presented here demonstrate the heterologous production of 1,3-PDO with genetically engineered C. beijerinckii DSM 6423 strains for the first time.
Prior to the transformation of C. beijerinckii, the newly constructed plasmids were transformed into electrocompetent E. coli SCS110 cells. Those E. coli cells are dcm and dam deficient, leading to unmethylated plasmid DNA after replication and isolation of plasmids. Transformation of C. beijerinckii was performed as described by Diallo et al., 2020 [31].

Strain Verification
Newly constructed C. beijerinckii strains and strains used for growth experiments were verified by 16S rRNA gene sequencing, and plasmids were retransformed in E. coli XL1-Blue MRF' for further analysis. Therefore, genomic DNA was isolated using the "MasterPure TM Gram Positive DNA Purification Kit" (Lucigen Corp., Middleton, WI, USA). Genomic DNA served as a template for amplification of the 16S rRNA gene (primers: 16S-27F and 1492r) using the "ReproFast proofreading polymerase" (Genaxxon Bioscience GmbH, Ulm, Germany). The amplified 16S rRNA gene was sequenced by Microsynth AG (Balgrach, Switzerland), and the sequence was blasted using NCBI blastn with the RNA/ITS database. For plasmid verification, E. coli XL1-Blue MRF was transformed with 3 µL of genomic DNA. After growth, plasmids were isolated and checked via analytic digestion.

Growth Conditions of Batch Experiments
For batch growth in bottles, the different C. beijerinckii strains were inoculated in a glycerol medium. The medium was prepared as described above. The pH of the glycerol medium was adjusted to 7.5 with KOH. After preparing the medium, 50 mL aliquots were filled in bottles and closed airtight. The gas phase was exchanged with N 2 :CO 2 (80: 20), and bottles were autoclaved. Before inoculation, the medium was supplemented with 40 mM xylose and 100 mM glycerol. For recombinant C. beijerinckii strains, clarithromycin (5 µg/mL) was added to the medium. To analyze the influence of vitamin B 12 supplementation, 5 mg/L sterile vitamin B 12 was added to the autoclaved glycerol medium of C. beijerinckii [pMTL83251_P thlA _dhaBCET. Cpas]. OD 600 and pH were monitored during the growth experiments. 2-mL samples for the analysis of substrate consumption and product concentration were taken. The samples were withdrawn with syringes and frozen until analysis took place.

Analytical Methods
The 2-mL samples withdrawn during growth experiments were thawed and subsequently centrifuged (18,000× g; 30 min; 4 • C). Acetate, butyrate, butanol, ethanol, acetoin, and isopropanol concentrations were determined using a "Clarus 600 gas chromatograph" (Perkin Elmer Inc., Waltham, MA, USA) equipped with a flame ionization detector heated to 300 • C and a flowrate of synthetic air of 450 mL min -1 . H 2 was used as the carrier gas (45 mL min -1 ). Prior to analysis, 480 µL supernatant was acidified with 20 µL 2 M HCl. 1 µL of acidified supernatant was injected onto an "Elite-FFAP" column (30 m × 0.32 mm; Perkin Elmer Inc., Waltham, MA, USA) with the injector heated to 225 • C. For analysis, the following temperature profile was used: 40 • C for 3 min, 40 • C to 250 • C by 40 • C min -1 , 250 • C for 1 min. Xylose and 1,3-PDO concentrations were measured using an "Agilent 1260 Infinity Series HPLC" system (Agilent Technologies, Santa Clara, CA, USA) equipped with a refractive index detector and a diode array detector. 20 µL of supernatant were injected into a "CS organic acid" precolumn (40 × 8 mm) followed by a "CS organic acid" column (300 × 8 mm; CS-Chromatographie Service GmbH, Langerwehe, Germany). The column was heated to 40 • C, and a mobile phase consisting of 5 mM H 2 SO 4 with a flow rate of 0.6 mL min -1 was used. Glycerol concentration was determined using the "Glycerol Assay Kit MAK117" (Sigma-Aldrich ® , St. Louis, MO, USA) following the manufacturer's manual. For glycerol standards, 1 mM, 0.8 mM, 0.6 mM, 0.4 mM, 0.2 mM, and 0.1 mM were used. Prior to analysis, the supernatant of the growth experiment samples was diluted (1:200 or 1:100) to fit the standard range used for the assay.

Recombinant 1,3-PDO Production
Different batch growth experiments were carried out to evaluate if the constructed recombinant C. beijerinckii strains produced 1,3-PDO and under what conditions the highest 1,3-PDO concentration could be achieved. Since C. beijerinckii DSM 6423 could not grow with glycerol as the sole carbon source, xylose was also added to the glycerol medium. The first batch growth experiment was performed with unbuffered glycerol medium for cultivation of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis], C. beijerinckii [pMTL83251], and the wild-type C. beijerinckii ( Figure 3). Compared to the control strains C. beijerinckii [pMTL83251] and the C. beijerinckii wild-type, the recombinant strain C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] grew faster and reached the stationary phase after 24 h with an OD 600 of 0.73. However, the three analyzed C. beijerinckii strains reached a final OD 600 of about 1 ( Figure 3A). The drastic pH drop during the first 24 h of cultivation was only detected in C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] ( Figure 3B). Unexpectedly, 1,3-PDO could only be detected in the growth medium of the recombinant strain C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis]. After almost 250 h of incubation, 18.9 mM 1,3-PDO was measured in case of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] ( Figure 3C). The glycerol concentration in the case of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] decreased during the experiment by 33.3 mM. In contrast, the glycerol concentration of the wild-type strain and C. beijerinckii [pMTL83251] decreased by 7 mM and 13 mM, respectively ( Figure 3D). Xylose was not consumed completely by any of the strains tested. The main product of the analyzed C. beijerinckii strains was butyrate (wild-type: 20.9 mM; C. beijerinckii [pMTL83251]: 21.3 mM; C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis]: 20.4 mM) ( Figure 3E). Only traces of ethanol were detected for all tested C. beijerinckii strains (0.2 mM). The lowest acetate concentration (2.2 mM) was measured in case of the wild-type strain. However, this strain also produced the most butanol, with 4.5 mM during the growth experiment. C. beijerinckii [pMTL83251] produced 3.6 mM acetate and 2.9 mM butanol throughout the cultivation. The highest acetate concentration was measured in the growth medium of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] (6.4 mM). However, this strain produced only traces of butanol (0.2 mM). concentration (2.2 mM) was measured in case of the wild-type strain. However, this stra also produced the most butanol, with 4.5 mM during the growth experiment. C. beijerinc [pMTL83251] produced 3.6 mM acetate and 2.9 mM butanol throughout the cultivatio The highest acetate concentration was measured in the growth medium of C. beijerinc [pMTL83251_Ppta-ack_1,3-PDO.diolis] (6.4 mM). However, this strain produced only tra of butanol (0.2 mM).

Influence of Buffered Glycerol Medium on 1,3-PDO Production
Due to the drastic drop in pH values in the first growth experiment ( Figure 3B), t effect of MOPS-buffered medium on the production of 1,3-PDO was examined. T following batch growth experiment was executed with the two recombinant strains beijerinckii [pMTL83251_PthlA_dhaBCET.Cpas] and C. beijerinckii [pMTL83251_Ppta-ack_1 PDO.diolis]. As controls, the wild-type strain as well as C. beijerinckii [pMTL83251] w used. The results of this batch growth experiment are shown in Figure 4. Again, beijerinckii [pMTL83251_Ppta-ack_1, 3

Influence of Buffered Glycerol Medium on 1,3-PDO Production
Due to the drastic drop in pH values in the first growth experiment (Figure 3B), the effect of MOPS-buffered medium on the production of 1,3-PDO was examined. The following batch growth experiment was executed with the two recombinant strains C. beijerinckii [pMTL83251_P thlA _dhaBCET.Cpas] and C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis]. As controls, the wild-type strain as well as C. beijerinckii [pMTL83251] were used. The results of this batch growth experiment are shown in Figure 4. Again, C. beijerinckii [pMTL 83251_P pta-ack _1,3-PDO.diolis] reached the stationary phase after 24 h of incubation (OD 600 : 0.85). In contrast, the OD 600 of C. beijerinckii [pMTL83251_P thlA _dhaBCET.Cpas] as well as the wild-type strain and C. beijerinckii [pMTL83251] decreased after 24 h and 48 h, respectively. Afterwards, the OD 600 of all three strains rose again. All tested strains reached a similar final OD 600 after 243 h ( Figure 4A). Compared to the first growth experiment, the pH dropped in the case of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] during the first 48 h of incubation only to a pH value of 5.47. Afterwards, the pH value decreased to a final value of 5.39 ( Figure 4B). Again, production of 1,3-PDO was only detected in the case of C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis]. The amount of 1,3-PDO was increased to 32.8 mM 1,3-PDO for the recombinant C. beijerinckii strain harbouring the 1,3-PDO production genes of C. beijerinckii DSM15410 ( Figure 4C). Minor concentrations of 1,3-PDO production could be measured for the wild-type strain, C. beijerinckii  Figure 4D). As in the first growth experiment, butyrate was the main product. The highest butyrate production was detected in the culture broth of C. beijerinckii [pMTL83251] with 40.1 mM, followed by C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis] and C. beijerinckii [pMTL83251_P thlA _ dhaBCET.Cpas], which produced 39.6 mM and 38.1 mM butyrate, respectively. The wildtype strain produced 33.3 mM butyrate. The highest acetate production was also detected in the culture broth of the strain C. beijerinckii [pMTL83251] at 3.1 mM. The strains C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis], C. beijerinckii [pMTL83251_P thlA _dha BCET.Cpas], as well as the wild-type strain, produced 2.2 mM, 3.0 mM, and 1,3 mM acetate. Similar to the first growth experiment, the highest butanol concentration was obtained by the wild-type strain (4.1 mM butanol). In the case of C. beijerinckii [pMTL83251] and C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis], 1.5 mM and 0.5 mM butanol were produced. No butanol could be detected in the culture medium of C. beijerinckii [pMTL83251_ P thlA _dhaBCET.Cpas]. Only traces of ethanol were detected by all tested C. beijerinckii strains (0.4 mM). An overview of the synthesized products is shown in Figure 4E. The use of MOPS-buffered glycerol medium resulted in a 74% increase in 1,3-PDO production using C. beijerinckii [pMTL83251_P pta-ack _1,3-PDO.diolis].

Effect of Vitamin B12 Supplementation
Since the used glycerol dehydratase of C. pasteurianum is coenzyme B12-dependent, the effect of supplementation of vitamin B12 in the MOPS-buffered growth medium of C. beijerinckii [pMTL83251_PthlA_dhaBCET.Cpas] was studied. Therefore, the glycerol medium of C. beijerinckii [pMTL83251_PthlA_dhaBCET.Cpas] was supplemented with vitamin B12. The results are shown in Figure 5. During this growth experiment, the recombinant strain C. beijerinckii [pMTL83251_Ppta-ack_1,3-PDO.diolis] grew slower than during previously described growth experiments. Cells reached the stationary phase after

Discussion
C. beijerinckii DSM 6423 (=C. beijerinckii NRRL B-593) is described as a natural 1,3-PDO producer [22,37,45]. However, the strain obtained from the DSMZ was unable to do so due to the loss of the 1,3-PDO dehydrogenase gene. In contrast to the published literature, the C. beijerinckii DSM 6423 wild-type strain only produced traces of 1,3-PDO. To reconstitute the 1,3-PDO production of the wild-type strain, a newly published transformation protocol of C. beijerinckii was employed to establish recombinant 1,3-PDO production in C. beijerinckii DSM 6423. Therefore, the heterologous gene expression of two different 1,3-PDO gene clusters, the effect of buffered growth media on the recombinant strains, and the use of different promoters for the gene expression were examined. The data presented clearly show that higher concentrations of 1,3-PDO could only be produced with recombinant C. beijerinckii DSM 6423 strains. In the published study of Gungormusler et al. the wild-type produced up to 131 mM 1,3-PDO (10 g/L) [37]. Genome analysis of C. beijerinckii DSM 6423 showed that the 1,3-PDO gene cluster is not complete. The dhaB1 and dhaB2 genes encoding a supposedly coenzyme B 12 -independent glycerol dehydratase could be identified. However, a dhaT gene encoding a 1,3-PDO dehydrogenase could not be found in the genome of C. beijerinckii DSM 6423. It is possible that another gene with a similar function can also convert 3-HPA to 1,3-PDO, resulting in low 1,3-PDO production. For example, alcohol dehydrogenases could also take over the same function due to their broad substrate specificity [46].
Higher concentrations of 1,3-PDO were only achieved using recombinant C. beijerinckii strains carrying the 1,3-PDO gene cluster of C. beijerinckii DSM 15410. The plasmid-based expression of the 1,3-PDO genes (dhaBCE and dhaT) of C. pasteurianum did not lead to an increase in 1,3-PDO production compared to the wild-type strain and C. beijerinckii [pMTL83251]. As mentioned before, the glycerol dehydratase complex of C. pasteurianum is coenzyme B 12 -dependent [27]. However, very little is known about the synthesis of coenzyme B 12 in C. beijerinckii DSM 6423. In many cases, supplementation of vitamin B 12 was necessary to produce 1,3-PDO when a coenzyme B 12 -dependent glycerol dehydratase was used [47]. In this study, even with supplementation of vitamin B 12 (2.5 mg/L), C. beijerinckii [pMTL83251_P thlA _dhaBCET.Cpas] was unable to produce 1,3-PDO. The reason could be that C. beijerinckii DSM 6423 does not harbour the necessary genes for the conversion of vitamin B 12 to coenzyme B 12 . Furthermore, Fokum et al. recently showed that the addition of vitamin B 12 in high concentrations (7.5-10 mg/L) to the culture medium has a negative effect on the 1,3-PDO production of C. beijerinckii CCIC 22954 [48]. For industrial approaches, it is also more desirable to use a coenzyme B 12 -independent production strain, in order to keep cultivation costs low due to expensive vitamin B 12 supplementation.
A further increase in 1,3-PDO production was accomplished by buffering the glycerol medium and preventing a fast pH decrease. The activities of enzymes as well as cofactors within the cell depend strongly on the pH. Therefore, a more stable pH throughout the bacterial cultivation is desired [49]. A positive effect of a controlled pH throughout the fermentation of glycerol was already described before [37]. Experiments with mixed cultures showed that the highest 1,3-PDO production yields could be measured at pH 7 and 8 [50]. Furthermore, most studies on the production of 1,3-PDO from glycerol are performed in fermenters with a controlled pH. Thus, it would be interesting to examine our recombinant production strains in a fermentation experiment with a steady pH.
The biggest improvements were made by exchanging the pta-ack promoter. Unfortunately, the direct activity test as described before for C. saccharoperbutylacetonicum and Eubacterium limosum [35,36] could not be applied for C. beijerinckii DSM 6423. In preliminary tests, C. beijerinckii carrying pMTL83251_P bgaL _FAST did not show differences in fluorescence compared to the wild-type strain. Thus, growth experiments were performed to examine the effects of different promoters. Although the lactose-inducible bgaL promoter of C. perfringens did not lead to the highest 1,3-PDO production, the inducible promoter system is almost tight in C. beijerinckii DSM6423 as indicated by the low 1,3-PDO concentration of uninduced cultures. Based on the presented results, the bgaL promoter can be described as the second inducible promoter for gene expression in C. beijerinckii DSM 6423. Another inducible system is based on the xylB promoter from Clostridium difficile strain 630, in which gene expression is induced by the addition of xylose [31,51]. The xlyB promoter was not examined in this study, since xylose was used as substrate in the growth experiments. Due to the close relationship of C. saccharoperbutylacetonicum and C. beijerinckii, both strains belong to the second clade of solvent-forming clostridia [52], which is why the bld promoter was also tested. Furthermore, the bld promoter showed the highest activity in C. saccharoperbutylacetonicum compared to promoters thlA, bgaL, and pta-ack [36]. Thus, we assumed high activity of the bld promoter in C. beijerinckii. However, the highest 1,3-PDO concentration and yield were detected in the case of C. beijerinckii [pMTL83251_P thlA _1,3-PDO.diolis] harboring the thlA promoter. This promoter also showed high activity in C. saccharoperbutylacetonicum [36].
The use of recombinant clostridial strains to produce 1,3-PDO was already reported before. In 2005, Gonzalez-Pajuelo et al. published an article using genetically modified C. acetobutylicum strains [53]. C. acetobutylicum was chosen as a host strain, as no genetic tools were available for C. butyricum at that time. The recombinant C. acetobutylicum strain was engineered for the heterologous production of 1,3-PDO using the genes from C. butyricum. Since the described fermentation approaches were fed-batch experiments, a comparison to this study is difficult. The modified C. acetobutylicum strain produced more 1,3-PDO compared to the tested strains of the present study due to the different cultivation method. However, all the recombinant C. beijerinckii strains presented here showed higher yields when grown in buffered glycerol medium. In comparison with other wild-type clostridial strains, the obtained concentrations are lower [7], but the yield is higher in the case of C. beijerinckii [pMTL83251_P thlA _1,3-PDO.diolis]. As mentioned before, a comparison of batch growth cultures and fermentation experiments under controlled conditions (pH) are hardly possible. However, continuous culture experiments with the newly constructed strains might even show higher 1,3-PDO production, thus proving the way for a cost-competitive bioprocess, when compared to chemical synthesis routes. Funding: This work was supported by the Bioeconomy International submission platform within the project "Sustainable bio-based 1,3-propanediol production from C5/C3 sources by metabolically engineered clostridia (SupperC)"(Förderkennzeichen: 031B0814).

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.