Oxygen determines the requirement for cobalamin but not riboflavin in the growth of Propionibacterium freudenreichii

Vitamin B₁₂ (hereafter referred as B₁₂), also known as cobalamin, is an essential water-soluble vitamin that plays a critical role in human physiology. It is a unique coordination complex of cobalt, characterized by a corrin-ring structure with two axial ligands. Cobalamins are distinguished from other cobamides by the presence of 5,6-dimethylbenzimidazole (DMBI) as the lower ligand, in contrast to adenine found in pseudocobalamins. Although structurally similar, B₁₂ analogues are not utilized by the human body, highlighting the distinct nutritional value of active B₁₂¹. Cobalamins act as crucial cofactors for enzymes such as cytoplasmic methionine synthase and methylmalonyl-CoA mutase² and are involved in nucleotide synthesis and myelin formation essential for neural health. A deficiency in B₁₂ can cause severe damage, particularly to the brain and nervous systems, leading to symptoms such as weakness, depression, and immune system disorders^3,4. Notably, B₁₂ is produced solely by bacteria and archaea and is not required by plants, making animal-derived food sources the exclusive providers of this vitamin for humans⁵.

In industrial applications, microbial fermentation is the primary method for producing B₁₂⁶, with Thiopseudomonas denitrificans, and Propionibacterium freudenreichii, being the main organisms used^5,6,7. P. freudenreichii, a Gram-positive bacterium from the genus Propionibacterium with a high GC% content (64–70%), belongs to the phylum Actinomycetota^8,9. This bacterium has a long history of use in cheese manufacture⁵. Recent interest has surged in fortifying food with B₁₂ through in situ bacterial fermentation^{10,11,12,13,14}. P. freudenreichii is particularly promising for this application due to its Generally Recognized as Safe (GRAS) status⁵, its efficient cobalamin production with recent reports showing production rates of up to 58.8 mg/L^15,16, and its minimal production of pseudocobalamin^17,18. Additionally, P. freudenreichii produces various functional metabolites, including conjugated linoleic acid (CLA), trehalose, and bifidogenic factors¹⁹.

The synthesis of cobalamin, which is unique to microorganisms, involves over 30 genes⁵. Previous research has elucidated the complete cobalamin synthesis pathway for P. freudenreichii^{20,21,22,23,24,25}. Among the complex steps, the activation and attachment of the lower ligand determines whether the end product is active B₁₂ or pseudovitamin B₁₂. The BluB-CobT2 fusion enzyme is responsible for synthesizing the DMBI base and activating it into α-ribazole phosphate, which is then attached as the lower ligand of cobalamin¹⁸. Fusing the production and activation of DMBI within the same enzymatic complex results in high propensity toward the production of the active form of B₁₂¹⁸. DMBI is derived from reduced flavin mononucleotide (FMN) through the action of the BluB enzyme^18,26,27,28. In P. freudenreichii oxygen plays a critical role in the synthesis of the DMBI ligand. Without adequate oxygen, the DMBI ligand cannot be synthesized, leading to the production of inactive forms of B₁₂, such as pseudocobalamin with adenine as a lower ligand base. Conversely, in the presence of excess oxygen, P. freudenreichii shifts from B₁₂ production to heme synthesis²⁹.

In the cobalamin synthesis pathway of P. freudenreichii, riboflavin is a precursor for the formation of DMBI. Riboflavin, or vitamin B₂, is another crucial B vitamin, essential for human growth, cellular function, and metabolism. It plays a critical role in converting proteins, fats, and carbohydrates into energy^30,31. Deficiency in riboflavin can lead to symptoms such as sore throat, mouth inflammation, and skin conditions³¹. Riboflavin is also vital for microbial metabolism as it serves as the precursor for the synthesis of the coenzymes FMN and FAD, which are essential for various redox reactions³². In cobalamin synthesis by P. freudenreichii, riboflavin is a foundational molecule for the formation of DMBI through the action to the BluB enzyme, ultimately determining the form of B₁₂ produced. The ribA gene, encoding the GTP cyclohydrolase II enzyme, is central to the riboflavin biosynthetic pathway³³. By catalyzing the early steps of riboflavin synthesis, the ribA gene indirectly influences B₁₂ biosynthesis. Given that the key role riboflavin plays in DMBI synthesis and its further influence on B₁₂ synthesis, understanding the relationship between riboflavin synthesis and the ribA gene in P. freudenreichii can provide insights into optimizing B₁₂ production, identifying potential bottlenecks, and highlighting areas for future research.

Bacterial strains and culture conditions

The strains and plasmids used in this study are listed in Table 1. P. freudenreichii strains were routinely cultured in yeast extract sodium lactate (YEL) medium, a standard medium for propionibacteria³⁷. This medium is devoid of vitamin B₁₂ but contains riboflavin provided by yeast extract. P. freudenreichii cultures were incubated routinely at 30 °C under anaerobic conditions, generated using Oxoid™ AnaeroGen™ 2.5 L Sachets (ThermoFisher) within an anerobic jar. These sachets create an atmosphere with < 1% O₂ within 30 min, and these conditions are referred to as anaerobic (pO₂ < 1%) throughout this study. To analyze the effect of riboflavin on growth characteristics, Riboflavin Assay Medium (RAM) (BD™ Difco™, USA) was used. Cysteine or other chemical reductants were not added to the media for P. freudenreichii cultivation. Escherichia coli was grown in Lysogeny Broth (LB) medium. When required, erythromycin (Sigma-Aldrich, USA) was added at 200 µg/mL for E. coli and 10 µg/mL for P. freudenreichii, or ampicillin (Sigma-Aldrich, USA) was added at 50 µg/mL for E. coli.

PCR and other DNA techniques

Platinum SuperFi II DNA Polymerase, restriction enzymes and T4 ligase were obtained from Thermo ScientificTM (Darmstadt, Germany), and were used according to the instructions of the supplier. PCR products were purified with a Qiagen PCR purification kit. Plasmid DNA from E. coli clones was isolated using GenElute Plasmid Miniprep Kit (SIGMA, USA). Genomic DNA was extracted using the MagAttract HMW DNA Kit (Qiagen, Nordic).

Insertional inactivation of the bluB/cobT2 and ribA genes

The protocol for inactivating the bluB and ribA genes in P. freudenreichii was adapted from previous research with some adjustments³⁸. First, a suicide vector was constructed by inserting an erythromycin resistance gene (ermE, GenBank: M11200.1) into a pBluescript SK + plasmid. The ermE gene was amplified by PCR with the primers ermfor and ermrev (Table 2) from genomic DNA of Saccharopolyspora erythraea DSM 40517 (DSMZ, Germany, Acc. No. PDBV01000001.1). The PCR product was then digested with EcoRV (Thermo Scientific™, Germany) and ligated with EcoRV-cut pBluescript SK +. The ligation mixture was transformed into competent E. coli DH5α and transformants were selected on LB agar (Condalab, Spain) (1.5%) containing erythromycin (200 µg/mL). Plasmid DNA was isolated from selected clones, and its identity was confirmed first by restriction digestion and agarose gel-electrophoresis, and finally by Nanopore sequencing (See below). This resulting plasmid was named pBsErmE.

For the inactivation of bluB/cobT2 two integration constructs were prepared. Primer pairs bluBcobtfor1/bluBcobtrev1 and bluBcobtfor2/bluBcobtrev2 were used to PCR amplify 550-bp and 545-bp internal bluB/cobT2 DNA fragments (Table 2), which were designed to be homologous to nucleotides 77–625 and 1151–1695 of the bluB/cobT2 coding region, respectively. These PCR products were then cloned as a EcoRI-XbaI fragment into pBsErmE, resulting in the plasmids pBsErmEbluB and pBsErmEcobT. Similarly, 607-bp ribA DNA fragment was amplified with the primers ribAfor and ribArev, which were designed to be homologous to nucleotides 63–669 of the ribA coding sequence. This PCR product was cloned as an XbaI-EcoRI fragment into pBsErmE resulting in pBsErmEribA for the inactivation of ribA.

P. freudenreichii electrocompetent cells were prepared as described previously with some modifications³⁸. Colonies of the P. freudenreichii DSM 4902 from YEL agar plate were used to inoculate YEL medium supplemented with 0.5 M sucrose and 2% glycine and incubated at 30 °C until the OD₆₀₀ reached 0.1. The cells were then harvested by centrifugation (4000×g, 15 min, 4 ℃) and washed three times with ice-cold 0.5 M sucrose. After washing, the cells were resuspended in an electroporation buffer (0.5 M sucrose, 10% glycerol, and 1 mM potassium acetate, adjusted to pH 5.5). For the electroporation step, 90 µL of the electrocompetent cells were mixed with 3–4 µg of pBsErmEbluB, pBsErmEcobT, or pBsErmEribA plasmid DNA in cooled electroporation cuvettes. The electroporation of P. freudenreichii DSM 4902 was performed with Gene Pulser XcellTM (Bio-Rad Laboratories, Richmond, CA, United States) at 20 kV/cm, 200-Ω resistance, and 25-µF capacitance followed by mixing the cells with 900 µL of YEL medium containing 0.5 M sucrose, 20 mM MgCl₂, and 2 mM CaCl₂ and incubation at 30 ℃ for 24 h. To select the mutants harboring inserted plasmids, the transformed cells were plated on YEL plates containing 10 µg/mL erythromycin and for pBsErmEbluB and pBsErmEcobT also 1 mg/L of vitamin B₁₂ was added. The plates were cultivated for 7 days at 30 ℃ under anaerobic (pO₂ < 1%) conditions. Despite several attempts, transformant colonies were obtained only with pBsErmEbluB and pBsErmEribA plasmids, but not with pBsErmEcobT. The mutant strains were further confirmed by colony PCR and Illumina genome sequencing (Sect.2.9).

Vitamin B₁₂ analysis

The content of B₁₂ in cell pellets was determined in its cyanocobalamin form and analysed using previously reported extraction and ultra-high-performance liquid chromatographic (UHPLC) methods^11,12. P. freudenreichii DSM 4902 strain, P.f ∆bluB mutant, and P.f ∆ribA mutant were grown at 30 ℃ in 10 mL of medium for 4 days in triplicates. Cells were harvested by centrifugation, and both the pellet and the supernatant were stored at −20 °C. The cell pellet suspended in 10 mL of the extraction buffer (8.3 mM sodium hydroxide and 20.7 mM acetic acid, adjusted to pH 4.5) and supplemented with 1% sodium cyanide was subjected to boiling water extraction (for 30 min) followed by centrifugation upon cooling to obtain 25 mL of the final extract. The extract was then analyzed with a Waters Acquity UPLC system (Milford, MA, US) equipped with a photodiode array detector (PDA; 210 ~ 600 nm) using an Acquity HSS T3 C18 column (2.1 × 100 mm, 1.8 μm). The mobile phase was a gradient flow of MilliQ water and acetonitrile, both containing 0.025% TFA. The chromatogram was obtained by recording the absorbance at 361 nm. Six cyanocobalamin standards (0.015 ~ 0.075 ng/µL) were injected (10 µL in duplicate) to create a calibration curve for each sample set. Cellular B₁₂ yield (µg/g fresh cell biomass) for each strain is reported as the average of three biological replicates.

Furthermore, the identity of the cobamide in the extracts (as cyanocobalamin) was confirmed with UPLC-MS/MS using a high resolution quadrupole time-of-flight mass spectrometer (Q-TOF, Synapt G2-Si; Waters, MA, USA) following the instrumental settings reported previously¹².

Measurement of riboflavin content

Riboflavin was extracted from the supernatants of the cultures mentioned above (See Vitamin B₁₂ analysis) using the European standard method (EN 14152:2014) with some modifications and analysed with a UHPLC method¹¹. Briefly, the culture supernatant (1mL) was mixed with 15 mL of 0.1 M hydrochloric acid and extracted in a boiling water bath for 60 min. After cooling to the room temperature, the pH was adjusted to 4.5 with 2.5 M sodium acetate and the extract was incubated for 24 h at 37℃ with Taka-Diastase (50 mg; Pfaltz and Bauer, CT, USA) and β-amylase (5 mg; Sigma-Aldrich). The analysis was performed with a Waters UPLC system equipped with a fluorescence detector using a Waters Acquity BEH C18 column (2.1 mm × 100 mm; 1.7 μm). The mobile phase contained 20 mM ammonium acetate dissolved in 30% aqueous methanol and was eluted at a constant flow of 0.2 mL/min. The excitation and emission wavelengths were set at 432 nm and 520 nm respectively. An external calibration curve obtained by injecting six riboflavin standards (0.01 ~ 1 ng/µL) in duplicate was used for the quantitation.

Analysis of acids

The supernatants from the strains and control media were diluted with MilliQ water and filtered (0.45 μm; Pall, USA) to HPLC vials. Succinic acid, lactic acid acetic acid and propionic acid were quantified using the HPLC method reported previously^11,39. The analysis was performed using an HPLC system equipped with a pump (Waters 515), an autosampler, a UV detector (Waters 717) and a refractive index detector (HP 1047 A, HP, USA) on a Hi-Plex H column (Agilent, CA, USA; 300 mm × 6.5 mm) as a pre-column and HiPlex H (Agilent, CA, USA; 50 mm × 7.7 mm) as a guard column. 10 mM H₂SO₄ was used as a mobile phase with the flow rate set at 0.5 mL/min for 30 min with the column temperature maintained at 40 °C.

Impact of vitamin B₁₂ on the metabolism of P. freudenreichii: an isothermal calorimetric analysis

The heat flow measurements were conducted following the protocol established by⁴⁰ with some modifications. Both P. freudenreichii DSM 4902 and P.f ∆bluB (with 10 µg/mL erythromycin addition) inoculated from 3-day liquid cultures were cultured in YEL media at 30 °C overnight. The cultures were diluted in fresh YEL media to achieve OD₆₀₀ 0.05, preparing them for subsequent analyses under two distinct environments: a micro-oxygen condition, which maintained an anaerobic atmosphere without explicitly removing oxygen, and a fully anaerobic setting, wherein oxygen was actively excluded by normalising the media in an anaerobic chamber (Whitley A85 anaerobic workstation, Don Whitley Scientific, UK) over night. Additionally, both strains were grown in the presence and absence of B₁₂ supplementation (1 µg/mL).

300 µL of the inoculated cultures were then transferred into plastic inserts (calWells, Symcel, Sweden) in four biological replicates, housed within titanium vials of a 48-well calPlate (Symcel, Sweden). The calPlate was incubated at 30 °C for 72 h within the calScreener (Symcel, Sweden), where heat flow signals were continuously recorded. For concurrent optical density measurements, the cultures (200 µL) were loaded to a 96-well microtiter plate (82.1581001, SARSTEDT, Germany), with the anaerobic culture prepared anaerobically (Whitley A85 anaerobic workstation, Don Whitley Scientific, UK) in one half of the plate and sealed with a clear Adhesive Sealing Sheet (Thermo-Fischer Scientific), and the micro-oxygen cultures were inoculated outside the anaerobic chamber. This plate was then incubated in the Varioskan Flash (Thermo Fisher Scientific, USA) at 30 °C for 72 h, during which OD₆₀₀ readings were taken every two h.

For determination of growth-limiting levels of vitamin B₁₂, both the DSM 4902 strain and its mutant derivative were prepared in the same way as above, but the P.f ∆bluB inoculum was supplemented with a gradient of vitamin B₁₂ (cyanocobalamin) concentrations, specifically at 1.00 µg/mL, 0.50 µg/mL, 0.10 µg/mL, 0.05 µg/mL, 0.01 µg/mL, and 0.005 µg/mL to examine the dose-dependent effects of B₁₂ on the metabolic activity of P.f ∆bluB compared to unsupplemented wild type, under fully anaerobic conditions.

Impact of riboflavin on the growth of P. freudenreichii

The growth curves were determined by following optical density at 600 nm wavelength in cultures using a Varioskan LUX Multimode Microplate Reader (Thermo Fisher Scientific, USA). Both P. freudenreichii DSM 4902 and P.f ∆ribA (with 10 µg/mL erythromycin addition) were cultured in 10 mL of YEL media under anaerobic (pO2 < 1%) conditions at 30 °C for 3 days. The cultures were then centrifuged; washed with RAM medium and resuspended in fresh RAM medium to an OD₆₀₀ value of 0.03.

For the subsequent growth curve measurements, 200 µL of the cultures were prepared in 6 replicates in a 96 well microplate (Thermo Fisher Scientific, USA) under micro-oxygen condition (10% O₂; maintained by N₂ gas flushing). P.f ∆ribA strains were inoculated with 10 µg/mL erythromycin. To determine the growth-limiting levels of riboflavin, both the DSM 4902 strain and P.f ∆ribA inoculum were supplemented with a gradient of riboflavin concentrations: 0.00 µg/mL, 0.01 µg/mL, 0.05 µg/mL, 0.10 µg/mL, 0.50 µg/mL, 1.00 µg/mL. This was done to examine the impact of different riboflavin doses on growth of P.f ∆ribA compared to supplemented wild type.

The inoculated cultures were then transferred into plastic inserts Nunc™ Microwell 96-Well Microplate with Nunclon™ Supra Surface (Thermo Fisher Scientific, USA). The microplate was incubated in the Varioskan at 30 °C, 10% O₂ for 72 h, during which OD₆₀₀ readings were taken every two h.

Sequencing and bioinformatics

For plasmid sequencing, E. coli strain carrying pBsErmE was cultivated overnight at 37 °C with shaking at 225 rpm (Forma Orbital Shaker, Thermo-Fischer Scientific). The cells were harvested via centrifugation at 8000 × g for 1 min (Accu Spin 17, Thermo-Fischer Scientific, Germany). Plasmid DNA was extracted from the pelleted cells using a GenElute Plasmid Miniprep Kit (Sigma-Aldrich, USA) according to the manufacturer’s instructions. Libraries were prepared using Native Barcoding Kit 96 V14 from Oxford Nanopore, UK (SQK-NBD114.96) and barcoding protocol NBE_9171_v114_revO_15sep2022. Libraries were loaded and sequenced on a PromethION 2 Solo (PRO-SEQ002) for 72 h using standard settings. Sequencing data were base-called and assembled using the workflow (https://github.com/yanhui09/ONTrapid) with some modifications. Flye assembler was used for assembly with the parameters: –nano-corr, –plasmid, –min-overlap 5000. The plasmid was annotated with pLannotate⁴¹.

For the P.f ∆bluB and P.f ∆ribA mutants DNA sequencing was performed at CeGaT (Germany). Briefly, 100 ng of genomic DNA was used for library preparation with Illumina DNA Prep kit (Illumina) and sequenced as 2 × 100 bp reads on NovaSeq 6000 sequencer (Illumina). The reads were demultiplexed using Illumina bcl2fastq v2.20 and the adapters were trimmed with Skewer v.0.2.2. The quality of the reads was assessed by FastQC v.0.11.5-cegat. The reads were then assembled with SPAdes genome assembler v3.14.1⁴². The resulting contigs were searched for the vector insertion sites using Megablast and annotated with PROKKA⁴³. Additionally, the trimmed reads were mapped to the reference genome LT618787 through alignment via Burrows-Wheeler transformation tool BWA v.0.7.17-r1188⁴⁴ to search for deletion site and exclude the possibility of off-target mutations.

Riboswitch Scanner^45,46 was used for detecting riboswitches in the genome sequence of P. freudenreichii DSM 4902³⁶.

Insertional inactivations of bluB/cobT2 and ribA genes

An integrative plasmid suitable for gene inactivation in P. freudenreichii was constructed by cloning ermE from S. erythraea DSM 40517 into the E. coli vector pBluescript SK+ (Additional Fig. 1). The ermE has previously been successfully used as a selection marker in P. freudenreichii⁴⁷. In this study, homologous regions in the integration constructs ranged from 545 to 607-bp. Since the E. coli vector containing the internal fragments of P. freudenreichii genes cannot persist in P. freudenreichii, erythromycin-resistant colonies are expected to arise exclusively through single crossover recombination events with the corresponding chromosomal allele. Despite repeated attempts, no erythromycin-resistant transformants were obtained using the construct with the shortest homologous region (545 bp) located within the cobT region of the bluB/cobT2 gene. Although a shorter homologous region (520 bp) was previously sufficient for insertional inactivation of the slpB in P. freudenreichii³⁸the absence of transformants in the current study with a 545 bp homologous region may reflect lower recombination efficiency compared to other constructs with slightly longer homologous regions. Another possible explanation is that mutations in the cobT region might be lethal to the strain, even when the vitamin B₁₂ is supplemented. In contrast, with the other two integration constructs, erythromycin-resistant transformants were obtained following electroporation. Subsequent genomic DNA sequencing and bioinformatics analyses confirmed that integration events occurred at predicted genomic locations (Additional Fig. 2).

Sequence analyses of the genomic arrangement in the P. freudenreichii bluB/cobT mutant, named as P.f ∆bluB, revealed that integration created two truncated copies of the bluB/cobT2 gene. The first copy lacks 1251 bp from the 3’ end of the coding region, while the second copy lacks the first 93 nucleotides of the coding region at the 5’ end, in addition to the upstream control region of the gene. Between these copies, 4388 bp of vector sequence is inserted. In the P. freudenreichii ribA mutant, named as P.f ∆ribA, genomic sequencing revealed two truncated copies of ribA separated by a 4.4 kb vector sequence. The first copy of ribA lacks 786 bp from the 3’ end of the coding region, and the second copy lacks the first 62 nucleotides of the coding region at the 5’ end, as well as the upstream control region of the gene (Additional Fig. 2).

Phenotypic characterization of the P.f ∆bluB and P.f ∆ribA strains

To assess the colony forming ability of the mutants in the absence of B₁₂ and riboflavin, serially diluted cultures were spotted on YEL (riboflavin-containing) or RAM (riboflavin deficient) agar plates. Media were either unsupplemented or supplemented with B₁₂ or riboflavin as appropriate. Plates were incubated at 30 °C for six days under either anaerobic (pO₂ < 1%; AnaeroGen system) or in a CO₂ incubator (~ 20% O₂, 5% CO₂) (Fig. 1).

As expected, under anaerobic (pO₂ < 1%) conditions (Fig. 1A and B), the wild-type formed visible colonies at all dilutions (10⁻³ to 10⁻¹⁰) on both media. In contrast, P.f ∆bluB strain showed reduced colony size on YEL agar, exhibiting a partial growth defect that could be restored by B₁₂ supplementation (Fig. 1A). This indicates that P. freudenreichii possesses a transport system capable of importing B₁₂ from the growth media, although this transporter remains to be genetically identified. Conversely, P.f ∆ribA exhibited minimal growth on riboflavin-deficient RAM agar, but its growth was fully restored with riboflavin supplementation (Fig. 1B), suggesting the presence of an as-yet-uncharacterized riboflavin transport system.

Under aerobic (~ 20% O₂) conditions, all strains formed colonies of roughly equal size across all dilutions (10⁻¹ to 10⁻⁸) on YEL agar, regardless of B₁₂ supplementation (Fig. 1C). On RAM agar, growth was restricted to the lowest dilutions (10⁻¹−10⁻²) and P.f ∆ribA again grew only when riboflavin was supplemented (Fig. 1C). Interestingly, P.f ∆bluB exihibited wild-type-like growth on both media, regardless of B₁₂ supplementation (Fig. 1C). These results indicate that riboflavin is essential under both anaerobic (pO₂ < 1%) and aerobic (~ 20% oxygen) conditions, whereas the growth-promoting effect of B₁₂ is limited to low-oxygen conditions.

The phenotype of the ribA disruption mutant is consistent with the predicted function of ribDEAH operon (locus tags PFR_JS15-1_2050, PFR_JS15-1_2049, PFR_JS15-1_2048, PFR_JS15-1_2047), which likely encodes the riboflavin biosynthesis machinery. An FMN riboswitch located approximately 100 nucleotides upstream may regulate this operon. Another FMN riboswitch was identified upstream of PFR_JS15-1_1083, encoding a putative ABC transporter, whose possible involvement in riboflavin uptake warrants further investigation⁴⁸.

The partial growth defect of the P.f ∆bluB strain under anaerobic (pO₂ < 1%) conditions confirms that that B₁₂ biosynthesis is important, but not strictly essential, for growth in anaerobic (pO₂ < 1%) conditions. In contrast, the P.f ∆bluB was indistinguishable from the wild type under aerobic conditions (~ 20% oxygen), suggesting that B₁₂ is dispensable in higher oxygen concentrations. This is consistent with recent reports showing that P. freudenreichii can grow well under aerobic bioreactor conditions while producing minimal or no B₁₂²⁹.

These observations reflect the hybrid nature of B₁₂ biosynthesis in P. freudenreichii, which uses the anaerobic (cobalt-early) pathway for the core corrin ring and cobalt insertion^5,25 but relies on the oxygen-dependent enzyme BluB to synthesize DMBI, the lower ligand of cobalamin²⁸. Therefore, B₁₂ biosynthesis in P. freudenreichii depends on oxygen only at the final stage, creating a physiological constraint: under oxygen-limited conditions, demand for B₁₂ increases due to metabolic requirements (e.g., for methylmalonyl-CoA mutase activity), yet the ability to synthesize B₁₂ is reduced due to impaired BluB function. This explains the importance of exogenous B₁₂ under anaerobic conditions and the paradox of low B₁₂ production despite robust growth in oxygen-rich environments.

B₁₂ and organic acids production by wild type, P.f ∆ribA and P.f ∆bluB strains

To further investigate the impact of impaired B₁₂ biosynthesis in P. freudenreichii, we cultivated the wild type, P.f ∆ribA, and P.f ∆bluB strains in YEL medium supplemented with CoCl₂. Cultures were incubated statically under atmospheric conditions for seven days. While initially oxygenated, the medium likely became progressively oxygen-depleted over time, simulating oxygen-limited conditions. Given the lack of active O₂ control, we avoid the term “microaerobic”. YEL medium, supported the growth of P.f ∆ribA owing to riboflavin present in the yeast extract. After seven days, the P.f ∆bluB strain reached a cell-density ~ 50% of that of the wild type. B₁₂ concentrations reached 10.91 µg/g (wild type) and 13.5 µg/g (P.f ∆ribA) but remained below the detection limit in the P.f ∆bluB strain (Fig. 2A).

Organic acid profiling showed that lactate was fully consumed by wild type and P.f ∆ribA cultures, while P.f ∆bluB consumed only 50%. Propionate and acetate production by P.f ∆bluB was also ~ 50% lower than by the other strains (Fig. 2B). These results confirm that B₁₂ deficiency impairs both growth and metabolic output under oxygen-limited conditions. Alternatively, trace amounts of B₁₂ below the detection threshold of our assay may suffice to support limited growth.

Prior data²⁹ further confirm that P. freudenreichii can grow well under aerobic conditions without producing substantial B₁₂. This implies that the dependence on B₁₂ is condition-specific, linked to the requirement of certain metabolic reactions (e.g., methylmalonyl-CoA mutase) that are more active or essential under low-oxygen conditions.

Isothermal calorimetric analysis of wild type and P.f ∆blub strains

To gain insight into the role of B₁₂ on the metabolic activity, we conducted isothermal calorimetric analyses comparing the wild type and the P.f ∆bluB mutant strains. Heat flow and OD₆₀₀ were monitored over 72 h under both aerobic and anaerobic conditions, with or without B₁₂ supplementation (Fig. 3A).

When B₁₂ was supplemented, P.f ∆bluB reached OD₆₀₀ and accumulated heat values similar to wild type. Oxygen also enhanced total heat output by ~ 60% in the wild type and ~ 50% in P.f ∆bluB. However, oxygen delayed the peak heat flow by 2–4 h, suggesting a metabolic shift in response to O₂.

To identify the minimal B₁₂ concentration needed to restore P.f ∆bluB growth, we conducted a dose-response experiment under strictly anaerobic conditions. At ≥ 0.1 µg/mL B₁₂, P.f ∆bluB achieved growth comparable to wild type, with peak heat production reached within 25 h (Fig. 3B). Lower concentrations (≤ 0.01 µg/mL) delayed and diminished heat output. These data indicate that at least 0.1 µg/mL of B₁₂ is required to support robust anaerobic growth.

Compared to the vitamin B₁₂ production yields of other bacteria (e.g., Bacillus megaterium, Acetobacter pasteurianus)^49,50P. freudenreichii requires relatively high intracellular concentrations to maintain full metabolic activity, supporting the idea that this species evolved efficient B₁₂ biosynthesis to meet high cellular demand.

Impact of riboflavin concentration on growth of wild-type and P. f ∆riba

To determine the minimal riboflavin concentration required for growth, we cultured wild-type and P. f ∆ribA strains in RAM medium under 10% O₂. While RAM is not optimized for P. freudenreichii, its lack of riboflavin makes it suitable for this purpose.

Wild-type growth was unaffected by varying riboflavin concentrations. P.f ∆ribA grew comparably to wild type when supplemented with ≥ 0.05 µg/mL riboflavin. Lower concentrations (0.01 and 0.00 µg/mL) led to reduced growth rates and final OD₆₀₀ (Fig. 4). These results confirm that riboflavin is essential under low-O₂ conditions and that supplementation at ≥ 0.05 µg/mL is sufficient to restore growth in P.f ∆ribA.

Effect of B₁₂ supplementation on wild type, P.f ∆ribA and P.f ∆bluB strains

Based on isothermal calorimetric results, we next evaluated whether supplementing cultures with 0.1 µg/mL B₁₂ would normalize growth, organic acid production, and intracellular B₁₂ levels across all strains under B₁₂-producing conditions (YEL supplemented with CoCl₂ and under a microaerobic atmosphere)^29,34,35. After seven days, all strains reached comparable final cell densities and acid production levels (Fig. 5). Lactate was exhausted in all cases, and B₁₂ concentrations were similar across strains. These results confirm that 0.1 µg/mL B₁₂ is sufficient to fully restore growth and metabolism in P.f ∆bluB under these conditions. 

Importantly, P.f ∆ribA showed no defect in B₁₂ synthesis or metabolic output when supplemented with riboflavin, confirming that endogenous riboflavin biosynthesis does not affect B₁₂ production when exogenous riboflavin is available.

In this study, we successfully constructed two mutant derivatives of P. freudenreichii DSM 4902, each lacking the ability to synthesize either vitamin B₁₂ or B₂, through insertional inactivation of a key biosynthesis gene in each respective pathway. The genetic modifications in the bluB and ribA mutants was confirmed via genome sequencing and bioinformatics analysis. Phenotypic characterization showed that P. freudenreichii can import both vitamins from its environment. We found that riboflavin (B₂) is essential for optimal growth under both aerobic and anaerobic (pO2 < 1%) conditions, while B₁₂ becomes particularly important under oxygen-limited conditions.

Our results demonstrated impaired growth and organic acid production in the absence of B₁₂, particularly under low oxygen availability. Supplementation experiments revealed that 0.10 µg/mL of B₁₂ and 0.05 µg/mL are sufficient to restore normal growth in the respective mutants. Notably, the presence of exogenous B₂ eliminated the dependency on de novo riboflavin synthesis for B₁₂ production, confirming that riboflavin does not limit cobalamin biosynthesis when provided externally.

Importantly, P. freudenreichii follows the anaerobic (cobalt-early) pathway for corrin ring and cobalt insertion steps in B₁₂ biosynthesis, which are oxygen-independent. However, it requires oxygen to complete the biosynthesis via the enzyme BluB, which catalyzes the formation of the lower ligand base DMBI. This dependency creates a metabolic bottleneck: under anaerobic conditions, B₁₂-dependent metabolic reactions become more important, yet BluB activity is impaired due to the lack of oxygen. In contrast, under aerobic conditions, the need for B₁₂ diminishes despite favorable conditions for its synthesis. These findings help explain the observed disconnect between oxygen availability and B₁₂ production.

In summary, this study refines our understanding of the interplay between oxygen, vitamin biosynthesis, and metabolic requirements in P. freudenreichii. It highlights how oxygen availability shapes both the need for and the capacity to synthesize B₁₂ and suggests that optimizing cobalamin production in this organism will require tight control of oxygen to balance enzymatic activity and metabolic demand. Future studies employing transcriptomic and metabolic profiling under defined oxygen gradients will be valuable for elucidating the regulatory mechanisms governing this balance.

1. Butler, P. A. & Kräutler, B. Biological organometallic chemistry of B₁₂ in Bioorganometallic Chemistry (ed Simonneaux, G.) 1–55 (Springer, (2006).

2. Nielsen, M. J., Rasmussen, M. R., Andersen, C. B. F., Nexø, E. & Moestrup, S. K. Vitamin B₁₂ transport from food to the body’s cells—a sophisticated, multistep pathway. Nat. Rev. Gastroenterol. Hepatol.9, 345–354. https://doi.org/10.1038/nrgastro.2012.76 (2012).

   Article  CAS  Google Scholar 

3. Reynolds, E. Vitamin B₁₂, folic acid, and the nervous system. Lancet Neurol.5, 949–960. https://doi.org/10.1016/S1474-4422(06)70598-1 (2006).

   Article  CAS  Google Scholar 

4. O’Leary, F. & Samman, S. Vitamin B₁₂ in health and disease. Nutrients2, 299–316. https://doi.org/10.3390/nu2030299 (2010).

   Article  CAS  Google Scholar 

5. Martens, J. H., Barg, H., Warren, M. & Jahn, D. Microbial production of vitamin B₁₂. Appl. Microbiol. Biotechnol. 58, 275–285. https://doi.org/10.1007/s00253-001-0902-7 (2002).

   Article  CAS  Google Scholar 

6. Fang, H., Kang, J. & Zhang, D. Microbial production of vitamin B₁₂: A review and future perspectives. Microb. Cell. Fact.16, 15. https://doi.org/10.1186/s12934-017-0631-y (2017).

   Article  CAS  Google Scholar 

7. Rudra, B. & S Gupta, R. Phylogenomic And comparative genomic analyses of species of the family pseudomonadaceae: proposals for the genera halopseudomonas gen. Nov. And atopomonas gen. Nov., merger of the genus oblitimonas with the genus thiopseudomonas, And transfer of some misclassified species of the genus Pseudomonas into other genera. Int. J. Syst. Evol. Microbiol. 71, 005011. https://doi.org/10.1099/ijsem.0.005011 (2021).

   Article  CAS  Google Scholar 

8. Scholz, C. F. P. & Kilian, M. The natural history of cutaneous propionibacteria, and reclassification of selected species within the genus Propionibacterium to the proposed Nov.l genera Acidipropionibacterium gen. Nov., Cutibacterium gen. Nov. and Pseudopropionibacterium gen. Nov.. Int. J. Syst. Evol. Microbiol.66, 4422–4432. https://doi.org/10.1099/ijsem.0.001367 (2016).

   Article  CAS  Google Scholar 

9. Oren, A. & Garrity, G. M. Valid publication of the names of forty-two phyla of prokaryotes. Int. J. Syst. Evol. Microbiol.71, 005056. https://doi.org/10.1099/ijsem.0.005056 (2021).

   Article  Google Scholar 

10. Capozzi, V., Russo, P., Dueñas, M. T., López, P. & Spano, G. Lactic acid bacteria producing B-group vitamins: A great potential for functional cereals products. Appl. Microbiol. Biotechnol.96, 1383–1394. https://doi.org/10.1007/s00253-012-4440-2 (2012).

    Article  CAS  Google Scholar 

11. Chamlagain, B. et al. Effect of the lower ligand precursors on vitamin B12 production by food-grade propionibacteria. LWT72, 117–124. https://doi.org/10.1016/j.lwt.2016.04.023 (2016).

    Article  CAS  Google Scholar 

12. Chamlagain, B., Edelmann, M., Kariluoto, S., Ollilainen, V. & Piironen, V. Ultra-high performance liquid chromatographic and mass spectrometric analysis of active vitamin B12 in cells of Propionibacterium and fermented cereal matrices. Food Chem.166, 630–638. https://doi.org/10.1016/j.foodchem.2014.06.068 (2015).

    Article  CAS  Google Scholar 

13. Gu, Q., Zhang, C., Song, D., Li, P. & Zhu, X. Enhancing vitamin B₁₂ content in soy-yogurt by Lactobacillus reuteri. Int. J. Food Microbiol.206, 56–59. https://doi.org/10.1016/j.ijfoodmicro.2015.04.033 (2015).

    Article  CAS  Google Scholar 

14. Molina, V., Médici, M., Font de Valdez, G. & Taranto, M. P. Soybean-based functional food with vitamin B₁₂-producing lactic acid bacteria. J. Funct. Foods. 4, 831–836. https://doi.org/10.1016/j.jff.2012.05.011 (2012).

    Article  CAS  Google Scholar 

15. Wang, P., Zhang, Z., Jiao, Y., Liu, S. & Wang, Y. Improved propionic acid and 5,6-dimethylbenzimidazole control strategy for vitamin B12 fermentation by Propionibacterium freudenreichii. J. Biotechnol.193, 123–129. https://doi.org/10.1016/j.jbiotec.2014.11.019 (2015).

    Article  CAS  Google Scholar 

16. Calvillo, Á., Pellicer, T., Carnicer, M. & Planas, A. Bioprocess strategies for vitamin B₁₂ production by microbial fermentation and its market applications. Bioengineering9, 365. https://doi.org/10.3390/bioengineering9080365 (2022).

    Article  CAS  Google Scholar 

17. Friedmann, H. C. Partial purification and properties of a single displacement trans-N-glycosidase. J. Biol. Chem.240, 413–418. https://doi.org/10.1016/S0021-9258(18)97664-7 (1965).

    Article  CAS  Google Scholar 

18. Deptula, P. et al. Blub/CobT2 fusion enzyme activity reveals mechanisms responsible for production of active form of vitamin B₁₂ by Propionibacterium freudenreichii. Microb. Cell. Fact.14, 186. https://doi.org/10.1186/s12934-015-0363-9 (2015).

    Article  CAS  Google Scholar 

19. Poonam, P., Pophaly, S. D., Tomar, S. K., De, S. & Singh, R. Multifaceted attributes of dairy propionibacteria: A review. World J. Microbiol. Biotechnol.28, 3081–3095. https://doi.org/10.1007/s11274-012-1117-z (2012).

    Article  CAS  Google Scholar 

20. Kiatpapan, P. & Murooka, Y. Construction of an expression vector for propionibacteria and its use in production of 5-aminolevulinic acid by Propionibacterium freudenreichii. Appl. Microbiol. Biotechnol.56, 144–149. https://doi.org/10.1007/s002530100603 (2001).

    Article  CAS  Google Scholar 

21. Falentin, H. et al. The complete genome of Propionibacterium freudenreichii CIRM-BIA1^T, a hardy actinobacterium with food and probiotic applications. PLoS One5, e11748. https://doi.org/10.1371/journal.pone.0011748 (2010).

    Article  ADS  CAS  Google Scholar 

22. Hashimoto, Y., Yamashita, M. & Murooka, Y. The Propionibacterium freudenreichii HemYHBXRL gene cluster, which encodes enzymes and a regulator involved in the biosynthetic pathway from glutamate to protoheme. Appl. Microbiol. Biotechnol.47, 385–392. https://doi.org/10.1007/s002530050945 (1997).

    Article  CAS  Google Scholar 

23. Piao, Y. et al. Production of vitamin B₁₂ in genetically engineered Propionibacterium freudenreichii. J. Biosci. Bioeng.98, 167–173. https://doi.org/10.1016/S1389-1723(04)00261-0 (2004).

    Article  CAS  Google Scholar 

24. Sattler, I. et al. Cloning, sequencing, and expression of the uroporphyrinogen III methyltransferase CobA gene of Propionibacterium freudenreichii (shermanii). J. Bacteriol.177, 1564–1569. https://doi.org/10.1128/jb.177.6.1564-1569.1995 (1995).

    Article  CAS  Google Scholar 

25. Roessner, C. A., Huang, K. X., Warren, M. J., Raux, E. & Scott, A. I. Isolation and characterization of 14 additional genes specifying the anaerobic biosynthesis of cobalamin (vitamin B₁₂) in Propionibacterium freudenreichii (P. shermanii). Microbiology 148, 1845–1853. https://doi.org/10.1099/00221287-148-6-1845 (2002).

    Article  CAS  Google Scholar 

26. Collins, H. F. et al. Bacillus megaterium has both a functional BluB protein required for DMB synthesis and a related flavoprotein that forms a stable radical species. PLoS One8, e55708. https://doi.org/10.1371/journal.pone.0055708 (2013).

    Article  ADS  CAS  Google Scholar 

27. Gray, M. J. & Escalante-Semerena, J. C. Single-enzyme conversion of FMNH₂ to 5,6-dimethylbenzimidazole, the lower ligand of B₁₂. Proc. Natl. Acad. Sci. U. S. A.104, 2921–2926. https://doi.org/10.1073/pnas.0609270104 (2007).

    Article  ADS  CAS  Google Scholar 

28. Taga, M. E., Larsen, N. A., Howard-Jones, A. R., Walsh, C. T. & Walker, G. C. BluB cannibalizes flavin to form the lower ligand of vitamin B₁₂. Nature 446, 449–453. https://doi.org/10.1038/nature05611 (2007).

    Article  ADS  CAS  Google Scholar 

29. Loivamaa, I. et al. Aerobic adaptation and metabolic dynamics of Propionibacterium freudenreichii DSM 20271: insights from comparative transcriptomics and surfaceome analysis. mSystems 9, e00615–00624. https://doi.org/10.1128/msystems.00615-24 (2024).

    Article  CAS  Google Scholar 

30. Powers, H. J. Riboflavin (vitamin B-2) and health. Am. J. Clin. Nutr.77(6), 1352–1360. https://doi.org/10.1093/ajcn/77.6.1352 (2003).

    Article  CAS  Google Scholar 

31. Thakur, K., Tomar, S. K., Singh, A. K., Mandal, S. & Arora, S. Riboflavin and health: A review of recent human research. Crit. Rev. Food Sci. Nutr.57, 3650–3660. https://doi.org/10.1080/10408398.2016.1145104 (2017).

    Article  CAS  Google Scholar 

32. Hiltunen, H. M. & Thesis, P. D. Functional analysis of RIBA, the introductory enzyme for riboflavin biosynthesis. Humboldt-Universität zu Berlin (2016).

33. Lin, J. W., Chao, Y. F. & Weng, S. F. Riboflavin synthesis genes ribE, ribB, ribH, RibA reside in the Lux Operon of Photobacterium leiognathi. Biochem. Biophys. Res. Commun. 284, 587–595. https://doi.org/10.1006/bbrc.2001.5013 (2001).

    Article  CAS  Google Scholar 

34. Dank, A. et al. Propionibacterium freudenreichii thrives in microaerobic conditions by complete oxidation of lactate to CO₂. Environ. Microbiol.23, 3116–3129. https://doi.org/10.1111/1462-2920.15532 (2021).

    Article  CAS  Google Scholar 

35. Dank, A., Biel, G., Abee, T. & Smid, E. J. Microaerobic metabolism of lactate and propionate enhances vitamin B₁₂ production in Propionibacterium freudenreichii. Microb. Cell. Fact.21, 225. https://doi.org/10.1186/s12934-022-01945-8 (2022).

    Article  CAS  Google Scholar 

36. Deptula, P. et al. De novo assembly of genomes from long sequence reads reveals uncharted territories of Propionibacterium freudenreichii. BMC Genomics18, 790. https://doi.org/10.1186/s12864-017-4165-9 (2017).

    Article  CAS  Google Scholar 

37. Malik, A. C., Reinbold, G. W. & Vedamuthu, E. R. An evaluation of the taxonomy of Propionibacterium. Can. J. Microbiol.14, 1185–1191. https://doi.org/10.1139/m68-199 (1968).

    Article  CAS  Google Scholar 

38. do Carmo, F. L. R. et al. Propionibacterium freudenreichii surface protein SlpB is involved in adhesion to intestinal HT-29 cells. Front. Microbiol.8, 1033. https://doi.org/10.3389/fmicb.2017.01033 (2017).

    Article  Google Scholar 

39. Hugenschmidt, S., Schwenninger, S. M., Gnehm, N. & Lacroix, C. Screening of a natural biodiversity of lactic and propionic acid bacteria for folate and vitamin B12 production in supplemented Whey permeate. Int. Dairy. J. 20, 852–857. https://doi.org/10.1016/j.idairyj.2010.05.005 (2010).

    Article  CAS  Google Scholar 

40. Beilharz, K. et al. Protocol to assess metabolic activity of Pseudomonas aeruginosa by measuring heat flow using isothermal calorimetry. STAR. Protoc. 4, 102269. https://doi.org/10.1016/j.xpro.2023.102269 (2023).

    Article  CAS  Google Scholar 

41. McGuffie, M. J. & Barrick, J. E. PLannotate: Engineered plasmid annotation. Nucleic Acids Res.49, W516–W522. https://doi.org/10.1093/nar/gkab374 (2021).

    Article  CAS  Google Scholar 

42. Prjibelski, A., Antipov, D., Meleshko, D., Lapidus, A. & Korobeynikov, A. Using spades de Novo assembler. Curr. Protoc. Bioinf. 70, e102. https://doi.org/10.1002/cpbi.102 (2020).

    Article  CAS  Google Scholar 

43. Seemann, T. Prokka: Rapid prokaryotic genome annotation. Bioinformatics30, 2068–2069. https://doi.org/10.1093/bioinformatics/btu153 (2014).

    Article  CAS  Google Scholar 

44. Li, H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. Preprint At. https://doi.org/10.48550/arXiv.1303.3997 (2013).

    Article  Google Scholar 

45. Mukherjee, S. & Sengupta, S. Riboswitch scanner: An efficient pHMM-based web-server to detect riboswitches in genomic sequences. Bioinformatics32, 776–778. https://doi.org/10.1093/bioinformatics/btv640 (2016).

    Article  CAS  Google Scholar 

46. Singh, P., Bandyopadhyay, P., Bhattacharya, S., Krishnamachari, A. & Sengupta, S. Riboswitch detection using profile hidden Markov models. BMC Bioinformatics10, 325. https://doi.org/10.1186/1471-2105-10-325 (2009).

    Article  CAS  Google Scholar 

47. Jore, J. P. M., van Luijk, N., Luiten, R. G. M., van der Werf, M. J. & Pouwels, P. H. Efficient transformation system for Propionibacterium freudenreichii based on a novel vector. Appl. Environ. Microbiol.67, 499–503. https://doi.org/10.1128/AEM.67.2.499-503.2001 (2001).

    Article  ADS  CAS  Google Scholar 

48. Gutiérrez-Preciado, A. et al. Extensive identification of bacterial riboflavin transporters and their distribution across bacterial species. PLoS One. 10, e0126124. https://doi.org/10.1371/journal.pone.0126124 (2015).

    Article  CAS  Google Scholar 

49. Mohammed, Y., Lee, B., Kang, Z. & Du, G. Development of a two-step cultivation strategy for the production of vitamin B12 by Bacillus megaterium. Microb. Cell. Fact.13, 102. https://doi.org/10.1186/s12934-014-0102-7 (2014).

    Article  CAS  Google Scholar 

50. Bernhardt, C., Zhu, X., Schütz, D., Fischer, M. & Bisping, B. Cobalamin is produced by Acetobacter pasteurianus dsm 3509. Appl. Microbiol. Biotechnol.103, 3875–3885. https://doi.org/10.1007/s00253-019-09704-3 (2019).

Download references

Authors and Affiliations

1. Department of Food and Nutrition, University of Helsinki, P.O. Box 66, 00014, Helsinki, Finland

   Ruoxi Zhang, Bhawani Chamlagain, Minnamari Edelmann, Kirsi Savijoki, Vieno Piironen & Pekka Varmanen

2. Department of Food Science, University of Copenhagen, Frederiksberg C, 1958, Denmark

   Yuandong Sha & Paulina Deptula

Contributions

R.Z., Y.S., P.D. and P.V. wrote the main manuscript text. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Ruoxi Zhang or Pekka Varmanen.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions