Key Findings

This section shows our findings investigating B12 production in E. coli, for which the ability has been reported to produce the active coenzyme form of B12, adenosylcobalamin (AdoCbl), utilizing the substrates Cbi and DMB [1]. The three forms of B12 that can be found in bacteria are AdoCbl, methylcobalamin (MeCbl) and hydroxocobalamin (OHCbl), converted interchangeably inside the cells according to the microorganism’s needs [2]. The two coenzyme forms, AdoCbl and MeCbl, are converted to OHCbl under UV-A light exposure in aqueous solution [3], which is more stable and kept for storage. In the following, we mostly use the broader term B12, as we detect a different form of it than the one that is originally produced in cells.

Aim

We first started testing B12 production in E. coli BL21(DE3) cells by exogenously adding cobinamide (Cbi) and 5,6-dimethylbenzimidazole (DMB) for verification of the reports from literature.

Experimental setup

E. coli BL21(DE3) cells from a glycerol stock were grown overnight in 10 mL of LB medium, washed with DPBS, resuspended in 10 mL of M9 medium and then used to inoculate the experimental setups, precisely 200 mL of M9 medium containing the respective substrates in 500 mL shaking flasks, to a starting OD600 of 0.001. We tested cultures with different concentrations of Cbi in combination with 10 µM DMB. For controls, we supplemented cultures with no substrate and 1 nM, 100 nM and 500 nM adenosylcobalamin (AdoCbl). These setups were then cultivated at 37°C, shaking at 200 rpm, for 12 hours with OD600 measurements every hour (Figure 1). After the last measurement, the whole cultures were centrifuged in falcon tubes, supernatants were discarded and cell pellets stored at -20°C for measurement of B12 concentrations.

Results

There appears to be no difference in growth between conditions within the first 6 hours (Figure 2). An improved growth was observed for supplementation with 100 nM or 500 nM AdoCbl, reaching an OD600 of about 0.5 or 0.8, respectively, after 12 hours, while most other conditions ranged around an OD600 of 0.3. Slightly less growth was observed in the culture supplemented with 500 nM Cbi and 10 µM DMB, resulting in a final OD600 of 0.2.

Conclusion

The starting OD600 of 0.001 was suboptimal as measurable changes in optical density only occur after 4 hours of cultivation. This can be improved by starting production cultures at a higher OD600 of 0.1, as notable shifts in optical density started to occur in the range of this value (Figure 2). The results of this experiment suggest that supplementation of cells with AdoCbl enhances growth of E. coli BL21(DE3), which matches the scientific consensus on this topic [2]. However, combined addition of Cbi and DMB to the cultures appear to marginally inhibit cell growth. This contradicts our expectation that E. coli cells with Cbi and DMB would grow similarly to those supplemented with AdoCbl, as they should have been able to produce AdoCbl themselves. Ultimately, the samples kept for detection of B12 concentrations in the respective conditions were discarded, as we had not established a reliable detection method at that point and the experiment was deemed to be repeated with improved experimental setup.

The first experiment gave us insight on what to take care of when cultivating E. coli BL21(DE3) for B12 production, so we could adjust the experimental setup according to what we had learned. Therefore, the experiment was repeated with an improved setting for more conclusive results. Here, the production temperature was decreased to 30°C upon substrate addition, as this is supposed to improve protein expression.

In the following experiment we exposed all culture samples to light to convert all B12 forms to OHCbl. This was a needed step for measuring the B12 concentration using Liquid Chromatography - Mass Spectrometry (LC-MS).

Aim

We tested B12 production with E. coli BL21(DE3) from Cbi and DMB in an improved experimental setting. Measurements were conducted for cell growth and for B12 concentrations in the respective cultures. Furthermore, we determined the most ideal concentration of Cbi for optimized B12 yield out of a range of substrate concentrations derived from literature [4].

Experimental setup

E. coli BL21(DE3) cells from a glycerol stock were grown overnight in 10 mL of LB medium, washed with DPBS, resuspended in 10 mL of M9 medium and then used to inoculate 50 mL cultures in 250 mL shaking flasks to a starting OD600 of 0.1. The cultures were grown to OD600~0.6, then the respective substrates were added. We tested cultures with different concentrations of Cbi in combination with 10 µM DMB. The cultures without substrates grew at 37°C, 200 rpm. Upon substrate addition, temperature for all cultures was reduced to 30°C. First OD600 was measured after 5 hours and then every 12 hours. Samples for LC-MS were taken after 5, 29 and 53 hours (Figure 3). For detection of B12 concentrations by LC-MS, we took 1 mL of cell culture, centrifuged it, discarded the supernatant and stored the cell pellet at -80°C. Later on, samples were purified and exposed to white ambient light. We then sent the samples to the Hannibal Lab at the University Medical Center Freiburg, which kindly performed the LC-MS analysis for us.
After receiving the LC-MS results, to estimate the g/g DCW value for the sample with the highest concentration, we measured the dry cell weight of a 1 mL sample from an independent E. coli BL21(DE3) culture with similar OD600.

Results

All cultures grew similarly, except for the one with only Cbi added, which grew a bit less (Figure 4). Comparing the three conditions with combined supplementation of Cbi and DMB, there seems to be a slight tendency towards better growth for higher Cbi concentrations.

LC-MS analysis shows the highest OHCbl content of 1.71 µg/g dry cell weight (DCW) for supplementation with 500 nM Cbi and 10 µM DMB (Figure 5). Supplementation with 100 nM Cbi and 10 µM DMB resulted in fivefold lower OHCbl contents. All the other samples measured have relatively negligible amounts of OHCbl lower than 0.3 µg/g DCW. The only exception was the negative control, which showed 25% less OHCbl compared to the 100 nM Cbi and 10 µM DMB condition after 53 hours.

Conclusion

We show that by supplementation with 500 nM Cbi and 10 µM DMB E.coli BL21(DE3) can produce B12 comparable to literature data, which range between 0.65 and 528 µg/g DCW [5,6]. Given that the condition with 100 nM Cbi and 10 µM DMB showed a fivefold lower OHCbl concentration than the supplementation with 500 nM Cbi and 10 µM DMB, it is reasonable to assume that the yield of B12 production in E. coli BL21(DE3) is proportionally correlated to the Cbi substrate concentration when DMB is available in the appropriate amount. The relatively high OHCbl concentrations detected in control samples where no substrate was added most likely results from sample carryover during the measurement, as these samples were measured directly after the 500 nM Cbi + 10 µM DMB sample from the respective earlier time points. Detection and quantification of B12 from cell pellets using LC-MS apparently worked very well, making it a suitable method for our further experiments.

Now that we showed B12 production in E. coli BL21(DE3) supplemented with Cbi and DMB, the next step is the introduction of the bluB gene into these cells to make them produce DMB by themselves (Design). This ideally makes exogenous supplementation with DMB unnecessary, therefore enhancing bioproduction efficiency by recombinant protein expression.

Aim

Cultures of E. coli BL21(DE3) carrying piG_01a were induced with different concentrations of doxycycline (DOX) to test bluB expression from this construct. We measured cell growth and investigated production of the BluB enzyme.

Experimental setup

E. coli BL21(DE3)[piG_01a] cells from a glycerol stock were grown overnight in 10 mL of LB medium, washed with DPBS, resuspended in 10 mL of M9 medium and then used to inoculate 50 mL cultures in 250 mL shaking flasks to a starting OD600 of 0.04. The cultures were grown until OD600~0.8, then we added the respective amounts of DOX to the conditions (Figure 6). The cultures grew at 37°C, shaking at 200 rpm. Samples for OD600 measurement and Western Blot were taken every 4 hours. The cultivation lasted 58 hours from induction onwards.

Results

Induction with 0.1 to 10 ng/mL DOX showed no significant difference in growth compared to the uninduced setup, although it is worth mentioning that the culture with 10 nM DOX grew the best throughout the whole experiment. Following induction, the cultures induced with 50 and 100 ng/mL DOX did not grow for the most part. The optical density of both cultures appeared to increase after 40 hours.

The Western Blot results show that E. coli BL21(DE3)[piG_1a] cells are able to produce BluB (Figure 8). We could detect that BluB protein levels increased with rising concentrations of DOX at different time points after induction. However, the non-induced sample (0 ng/mL DOX) also shows bluB expression with similar intensity as for 0.1 ng/mL and 1 ng/mL DOX. To exclude the possibility that the samples were accidentally induced, we tested an independent overnight culture of E. coli BL21(DE3)[piG_1a] taken directly from the glycerol stock. Unfortunately, even in this culture we could detect BluB. The BluB protein level was comparable to induced samples 16 hours after induction with 0.1 ng/mL DOX (Figure 8C).

BluB has successfully been expressed in E. coli BL21(DE3) with our designed construct [piG_1a]. However, the promoter appears to be leaky. Even without induction with DOX, we could see bluB expression in an overnight culture of E.coli BL21(DE3)[piG_1a]. We therefore had a closer look on the design of the promoter sequence and how it possibly needed to be changed (Parts). Considering the observed growth inhibition for 50 and 100 ng/mL DOX, the DOX concentration that is most appropriate to proceed with appears to be 10 ng/mL.

Aim

Having characterized the functionality of piG_01a to our purpose, in this experiment we cultivate E. coli BL21(DE3)[piG_01a], induced and supplemented with Cbi to confirm its hypothesized ability to produce B12 independent of exogenous DMB. Cultures were analyzed for growth behavior and B12 concentrations.

Experimental setup

E. coli BL21(DE3)[piG_01a] cells from a glycerol stock were grown overnight in 10 mL of LB medium, washed with DPBS, resuspended in 10 mL of M9 medium and then used to inoculate 50 mL cultures in 250 mL shaking flasks to a starting OD600 of 0.1. The cultures were grown until OD600~0.9, then the respective amounts of Cbi and DOX were added to the conditions (Figure 9). Initially, cultivation was carried out at 37°C, 200 rpm, before substrate and inducer were added. Afterwards, the temperature for all cultures was reduced to 30°C. OD600 was measured after 0, 26 and 50 hours, LC-MS samples were taken after 26 and 50 hours. The cultivation lasted 50 hours, starting from substrate addition / induction. For detection of B12 concentrations by LC-MS, we took 1 mL of cell culture, centrifuged it, discarded the supernatant and stored the cell pellet at -80°C. Samples were purified and exposed to ambient light deliberately before sending them to external LC-MS analysis, blinded to prevent bias. We additionally collected cell pellets of the whole 50 mL cultures as well as the corresponding supernatants after 50 hours for LC-MS analysis. We sent the samples to the Hannibal Lab at the University Medical Center Freiburg, which kindly performed the LC-MS analysis for us.

To estimate the g/g DCW value for the sample with the highest concentration, we measured the cell weight of a 1 mL sample of an independent E. coli BL21(DE3) culture with similar optical density.

Results

The growth assay did not show a significant difference of induced cultures treated with different concentrations of Cbi. The setup supplemented with 500 nM Cbi, without DOX induction seems to grow slightly better than the other setups (Figure 10).

OHCbl concentrations were measured in cell pellets and supernatants of E.coli BL21(DE3)[piG_1a] samples. The supernatant of all samples did not contain any detectable OHCbl. Very low amounts of OHCbl were measured in cell pellets of non-treated (-) and only DOX treated controls. However, non-induced samples treated with 100 nM Cbi showed some amounts of OHCbl. DOX-induction and treatment with increasing concentrations of Cbi resulted in an upward trend in the amount of detectable OHCbl (Figure 11A). The sample induced with 500 nM Cbi shows the highest OHCbl yield out of all conditions, about fivefold higher than the sample induced with 100 nM Cbi, which aligns with the results in Figure 5.

In order to investigate the possible total amount of OHCbl produced in the culture, we processed the whole 50 mL culture at the final time point, 50 hours after induction. A significant amount of OHCbl was measured in the induced sample treated with 500 nM Cbi, correlating to an estimate of about 13 µg/g DCW (Figure 11).

High values of OHCbl were measured in the supernatant of non-treated (-) and only DOX treated endpoint samples (50h). Since these results could be due to measurement errors the samples were re-measured for validation but with additional blanks between each sample. Significantly less OHCbl was detected in the supernatant (Figure 11B) compared to the initial measurement (Figure 11A). Furthermore, the measurement confirmed that supplementation with 500 nM Cbi yielded five times more OHCbl than supplementation with 100 nM Cbi.

Conclusion

By overexpressing BluB, E. coli BL21(DE3)[piG_01a] can produce sufficient amounts of DMB endogenously, leaving DMB supplementation unnecessary regarding B12 production. The fact that nearly the same concentration of OHCbl was measured in both non-induced and induced samples, supplemented with 100 nM Cbi, is another indication for a leaky tet-promoter in piG_01a.

We successfully confirmed that bluB expression with piG_01a provides sufficient amounts of DMB for B12 production in E. coli BL21(DE3) from Cbi. However, the promoter of this construct is leaky. To optimize this part, we slightly changed the promoter region to create piG_01b (Parts). This plasmid was now introduced into the different expression strain E. coli MG1655, a standard K-12 strain. We decided to do this, as we received feedback from multiple sources that BL21(DE3) might be additionally stressed by our backbone because of its inheritance of the T7 promoter as the T7-RNA-polymerase is natively expressed in BL21(DE3).

Aim

After cloning of piG_01b and transformation into E. coli MG1655, we induced with different concentrations of DOX to test, if this construct is less leaky than piG_01a in E. coli BL21(DE3) cells.

Experimental setup

E. coli MG1655[piG_01b] cells from glycerol stock were used to inoculate a 250 mL pre-culture in M9 medium and incubated at 37°C, 200 rpm. At OD600~0.9, the pre-culture was distributed into four 50 mL flasks and the respective amounts of DOX were added (Figure 12). The different conditions were cultivated at 37°C, shaking at 200 rpm. We took samples for OD600 measurement and Western Blot every 4 hours for 36 hours.

Results

For the growth assay we could not find any difference between non-induced and DOX induced samples (Figure 13).

The Western Blot of samples taken 24 hours after induction showed a significant signal for the non-induced setup (Figure 14). Expression of bluB appears similar for 10 and 50 ng/mL DOX, while the culture induced with 100 ng/mL DOX exhibits the brightest signal.

Conclusion

The promoter of piG_01b seems to be as leaky as the one of piG_01a. In this experiment we also see that cell growth is not inhibited for 50 and 100 ng/mL DOX different to what we have seen before (Figure 7). It is likely that the growth inhibition does not occur here, because the strain we are using now works more efficiently. Considering the results of this experiment, we will use 100 ng/mL DOX for induction in further experiments.

Aim

The main goal of this experiment was the comparison between optimized B12 production setups for E. coli MG1655[piG_01b] and E. coli MG1655 supplemented with DMB. piG_07, a version of piG_01b with mutated bluB gene, and pGGAselect, the empty backbone, were used as controls to prove the functionality of piG_01b. Additionally, we had a first look at the functionality of piG_10a, although here it is only tested briefly. The construct is later characterized more in-depth. Composite Results

Experimental setup

Name	Short description	BioBrick
piG_01b	TetR_bluB	BBa_K4604015
piG_07	tetR_mut_bluB	BBa_K4604020
piG_10a	tetR_bluB_riboK12_mazF	BBa_K4604021

E. coli MG1655[piG_01b], E. coli MG1655[piG_07], E. coli MG1655[pGGAselect] and E. coli MG1655[piG_10a] from glycerol stocks were used to inoculate respective 100 mL pre-cultures in M9 medium. These pre-cultures then grew to sufficient optical densities to inoculate 50 mL cultures to an OD600 of 0.3. Substrates and inducer were added at OD600~0.6. We tested induced and non-induced setups for piG_01b, piG_07, pGGAselect and piG_10a with Cbi and with DMB. The cultures grew at 37°C, shaking at 200 rpm. Optical density was measured every 6 hours, samples for Western Blot were taken every 12 hours, LC-MS samples were taken 0 and 24 hours after induction and substrate supplementation. For detection of B12 concentrations with LC-MS, we took 1 mL of cell culture, centrifuged it, discarded the supernatant and stored the cell pellet at -80°C. For determining the dry-cell-weight (DCW), sample tubes were weighed beforehand. Once all samples were collected, they were heated together at 80°C overnight for drying. Sample tubes containing the dry cells were weighed again to determine DCW. These samples were then purified and exposed to light deliberately before sending for external LC-MS analysis, blinded to prevent bias. The experiment was repeated a total of two times.

Results

The growth assay did not show a clear result for the different tested conditions. The cultures of E. coli MG1655[pGGAselect] supplemented with Cbi and DMB and induced E. coli MG1655[piG_10a] supplemented with Cbi grew to a final OD600 of about 2 (Figure 17). All other setups reached an OD600 of 1.4, on average with similar growth behavior.

We next examined the expression of bluB using Western Blot. The induced cultures of E. coli MG1655[piG_01b] and E. coli MG1655[piG_10a] show clear bands for BluB after 12 and 24 hours in the Western Blot images (Figure 18). There are slight bands indicating bluB expression for non-induced E. coli MG1655[piG_01b] cultures after 12 hours (Figure 18A). There was no bluB expression in cells carrying piG_07 or pGGAselect.

LC-MS measurement shows high OHCbl contents in cells of the induced and non-induced E. coli MG1655[piG_01b] cultures supplemented with Cbi, as well as in the cultures of E. coli MG1655[pGGAselect] treated with Cbi and DMB (Figure 19A). Induced E. coli MG1655[piG_01b] supplemented with Cbi, the setup that was intended to produce the most B12, yielded about 5 µg/g DCW. Uninduced cells in otherwise same conditions display OHCbl contents that are roughly 1 µg/g DCW higher than induced ones. In addition, significant OHCbl contents were observed in induced E. coli MG1655[piG_10a] cultures supplemented with Cbi, yet they were about 30% as high as the contents measured in samples of induced E. coli MG1655[piG_01b] with Cbi (Figure 19A).

Cbi was detected in relatively similar amounts in all setups supplemented with 500 nM Cbi (Figure 19B). In supernatants, low OHCbl concentrations were observed across all cultures (Figure 20). Here we see high Cbi concentrations in media that were supplemented with Cbi.

B12 detection in the experimental setups using the ethanolamine utilization assay shows significant growth for the induced and non-induced setups of E. coli MG1655[piG_01b] setups supplemented with Cbi (Figure 21). Cultures of E. coli MG1655[pGGAselect] treated with Cbi and DMB display the highest growth out of all setups. E. coli carrying piG_10a grew similarly to the setups with piG_01b. All of the setups that grew reached optical densities higher than that of the control setup containing E. coli MG1655[pGGAselect] supplemented with 200 nM AdoCbl.

Conclusion

The induced cultures of E. coli MG1655[piG_01b] express bluB (Figure 18), which produces DMB, leading to B12 synthesis when cells are supplied with Cbi. E. coli cells, which do not carry the bluB gene, also produce B12 when treated with Cbi and DMB. Yet, this experiment shows about double the B12 content in induced E. coli MG1655[piG_01b] supplemented with Cbi than in E. coli MG1655[pGGAselect] with Cbi and DMB (Figure 20). Light signals for BluB were detected in the Western Blot for non-induced setups with E. coli MG1655[piG_01b] (Figure 18). This matches the previous finding that the tet-promoter in piG_01b is leaky. The observed leakiness appears to facilitate sufficient bluB expression for high B12 production, as OHCbl content in non-induced E. coli MG1655[piG_01b] with Cbi was slightly higher than that detected in induced E. coli MG1655[piG_01b] with Cbi (Figure 20).

E. coli MG1655[piG_10a] displayed strikingly low B12 production, given that the piG_10a is a first prototype version of our final system. Still, it is one without an established toxin-antitoxin (TA) system, therefore all this experiment tells us for now is that the B12 production part in piG_10a works. This plasmid surely needs further optimisation.

Considering the observed growth behavior, productive cells carrying piG_10a or pGGAselect seem to grow better than those carrying piG_01b, which grew similar to the unproductive setups. This might indicate improved cell growth for cells producing B12. Yet, productive E. coli MG1655[piG_01b] cells produced significantly higher amounts of B12 compared to E. coli MG1655[pGGAselect] and E. coli MG1655[piG_10a]. Perhaps at this production rate, metabolic burden of B12 overproduction comes into play.

The large deviation observed in the LC-MS measurement for OHCbl and Cbi in nearly every setup might be related to heating samples at 80°C overnight, to measure dry cell weight. This likely caused partial degradation of the respective molecules.

Cbi concentrations in the media of setups supplemented with Cbi decrease over time (Figure 20C), while those detected inside cells increase. This proves that E. coli MG1655 cells take up the precursor to then produce B12. The growth observed for the setups in the ethanolamine utilization assay matched the qualitative findings from the LC-MS measurement. All setups that showed significant growth showed notable OHCbl contents. Yet, the relative growth between setups in ethanolamine utilization assay do not match the amounts of B12 detected with LC-MS. Quantities of B12 measured during this run of LC-MS do not appear reliable due to the heat treatment of samples mentioned above.

Key Findings

This part is dedicated to the establishment of B12 production in cyanobacteria, primarily in the model organism S. elongatus PCC 7942. First, we wanted to validate reports on the ability of S. elongatus PCC 7942 being able to produce B12 when supplemented only with the precursor 5,6-dimethylbenzimidazole (DMB). If this turned out to be successful, we would move on to introducing the two genes ssuE and bluB, derived from S. elongatus PCC 7002 and M. smegmatis, into the organism that supposedly are necessary to produce DMB and therefore B12 endogenously inside the cells [7]. Initially, we did not find any B12 produced in the cyanobacteria strains we tested. We tested S. elongatus PCC 7942 from University of Tübingen and AG Wilde at University of Freiburg, as well as S. sp. PCC 6803 from University of Tübingen. Neither did we see any pseudocobalamin (pseudoCbl) production, except for one sample of S. sp. PCC 6803 supplemented with DMB. This did not match the scientific consensus we came across, reporting that S. elongatus PCC 7942 is able to produce pseudoCbl, while apparently at least some enzymes necessary for the B12 synthesis pathway have been found in S. sp. PCC 6803 [8,9]. The only explanation we could come up with is that we did not manage to extract either of the two substances from the cells, yet it would not explain the high pseudoCbl concentration found for that one S. sp. PCC 6803 sample. The last experiment we did was one more try to find a hint that would indicate at least pseudoCbl production in the strains we worked with. There, pseudoCbl was detected, significantly more than in the samples supplemented with DMB. This would indicate that DMB is incorporated into the pseudoCbl molecule. Following this train of thought, S. sp. PCC 6803 might be lacking at least one of the genes catalyzing the final conversion steps to B12. Of course, this is only a brief discovery that needs validation.

Aim

We tested B12 production in S. elongatus PCC 7942 cultures supplemented with DMB. Measurements were conducted to investigate growth behavior and B12 concentrations in respective cultures.

Experimental setup

S. elongatus PCC 7942 cells from a BG-11 liquid culture were used to inoculate a 100 mL culture in a CellDEG HD100 cultivator to a starting OD750 of 0.5. The culture then grew at 30°C, shaking at 200 rpm with ~115 µE light intensity. When the culture reached OD750~6.9, cells were washed and resuspended in fresh iron deficient medium. The cells were then distributed into 10 mL setups in CellDEG HD10 cultivators to a starting OD750 of 3.72. The setups contained different concentrations of DMB, three replicates per setup. The cultures grew at 30°C, 200 rpm. Light intensity started at ~115 µE and was later increased to ~250 µE and then ~575 µE, when the value for (OD750)*75 equaled the respective light intensity, according to advice from the CellDEG company.
For detection of B12 concentrations by LC-MS, we took 1 mL of cell culture, centrifuged it, discarded the supernatant and stored the cell pellet at -80°C. Later on, samples were purified and exposed to light deliberately to convert the light-sensitive B12-forms adenosylcobalamin (AdoCbl) and methylcobalamin (MeCbl) to hydroxocobalamin (OHCbl). We then sent them for external LC-MS analysis, blinded to prevent bias.

Results

All setups grew similarly, only the growth of cultures supplemented with 50 µM DMB appeared to stagnate after 84 hours (Figure 23).

LC-MS analysis shows low OHCbl concentrations across all setups with values that are in or close to the range of 0-25 pM (Figure 24). There is no significant difference in the concentrations of OHCbl measured in cells and supernatants.

Conclusion

OHCbl concentrations measured with LC-MS are close to the range of 0-25 pM OHCbl, which can be considered “zero or trace amounts”, according to Luciana Hannibal, leader of the lab that conducted the LC-MS measurement. Considering these results, we conclude that either the specific strain we used did not produce OHCbl at all or we did not manage to extract OHCbl from the cells successfully.

Aim

In this experiment, we tested B12 production in S. elongatus PCC 7942 from 2 different cryostocks as well as in S. sp. PCC 6803, all supplemented with DMB. The cultures were analyzed for growth behavior and concentrations of OHCbl and pseudocobalamin (pseudoCbl).

Experimental setup

S. elongatus PCC 7942 and S. sp. PCC 6803 cells from liquid BG-11 cultures were used to inoculate 40 mL cultures in BG-11(-Fe). The according setups contained 10 µM DMB from the start. OD750 was measured every 24 hours until 72 hours, then once more after 48 hours. The setups were cultivated at 30°C, shaking at 120 rpm. Used strains: S. sp. PCC 6803 and S. elongatus PCC 7942 from University of Tübingen, S. elongatus PCC 7942 from AG Wilde at University of Freiburg. For detection of B12 concentrations by LC-MS, we took a 2 mL sample of cell culture, centrifuged it, discarded the supernatant and stored the cell pellet at -80°C. The eppis for each sample were weighed out beforehand. Later on, we heated all samples together at 80°C overnight for drying and then weighed out the eppis with dry cells afterwards to later correlate concentrations from LC-MS measurements to cell mass. These samples were then purified and exposed to light deliberately. We then sent them for external LC-MS analysis, blinded to prevent bias.

Results

S. sp. PCC 6803 grew to a higher OD750 than both variants of S. elongatus PCC 7942 (Figure 26). For S. sp. PCC 6803 and the better growing S. elongatus PCC 7942 variant (AG Wilde), cultures supplemented with 10 µM DMB grew better than those without DMB. For the variant of S. elongatus PCC 7942 growing comparatively slowly (Tübingen), the culture without DMB grew better than the one with DMB added.

The only setup showing considerable amounts of pseudoCbl is the culture of S. sp. PCC 6803 supplemented with 10 µM DMB (Figure 27). Measured OHCbl contents in all setups range from 0.06 to 0.13 µg/g DCW.

Conclusion

None of the setups show OHCbl contents that would be indicative of B12 production in these cells with measured amounts of only up to 0.13 µg/g DCW. Additionally, we only observed significant pseudoCbl contents for S. sp. PCC 6803 supplemented with 10 µM DMB, containing about 3.3 µg/g DCW. This can be considered to be significantly more than the trace amounts detected in all other setups, which was below 0.5 µg/g DCW. Both S. elongatus PCC 7942 and S. sp. PCC 6803 are supposed to produce pseudoCbl natively, therefore we would expect samples to contain similar amounts of pseudoCbl, certainly those respective to each strain. One explanation could be that the extraction prior to LC-MS did not work out equally well.

Aim

We tested B12 production in S. sp. PCC 6803 cultures supplemented with DMB. LC-MS measurements were conducted to investigate B12 concentrations in respective cultures.

Experimental setup

S. sp. PCC 6803 cells from a BG-11 liquid culture were used to inoculate 10 mL cultures in CellDEG HD10 cultivator to a starting OD750 of 0.5. The cultures grew at 30°C, shaking at 120 rpm with ~115 µE light intensity. Cells were cultivated with and without DMB in FOM and FOM(-Fe). For detection of B12 concentrations by LC-MS, we took the whole 10 mL cell cultures, centrifuged them, discarded the supernatant and stored the cell pellet at -80°C. Later on, samples were purified and exposed to light deliberately to convert the light-sensitive B12 forms AdoCbl and MeCbl to OHCbl. We then sent them for external LC-MS analysis, blinded to prevent bias.

Results

LC-MS shows no notable amount of OHCbl in any of the samples. For pseudoCbl contents, we see about 13 fold higher amounts detected for the setups without DMB than for those supplemented with DMB (Figure 29).

Conclusion

We see a strong decrease in pseudoCbl for the samples that were supplemented with DMB, yet we did not see any B12 production. This implies to us that pseudoCbl is remodeled, but the whole conversion to B12 is lacking. We did employ the iron-deficient medium, because literature suggested that pgam3 is crucial for the final conversion step to AdoCbl and present in S. elongatus PCC 7942, but only expressed under iron-deficient conditions [7]. As S. elongatus PCC 7942 did not deliver any conclusive results, this time we tested a similar setup for S. sp. PCC 6803. Here we demonstrate, that S. sp. PCC 6803 is able to produce pseudoCbl. Moreover, it seems likely that S. sp. PCC 6803 is capable of remodeling the pseudoCbl molecule when supplemented with DMB, although not capable of of fully converting to B12.

Cultivation of the cyanobacteria strains we worked with takes a decent amount of time, with the average time for one cell division taking 6-12 hours [10]. Therefore, we wanted introduce ssuE and bluB into S. elongatus PCC 7942 in parallel, on a shuttle vector designed by iGEM Marburg 2019 [11]. The newly created plasmid we named piG_CBM. Multiple rounds of conjugation and transformation by electroporation, as well as one attempt for natural transformation with multiple different setups were done, all unsuccessful. We consulted with Wilde, Hess and Selim groups, all doing research specialized in cyanobacteria at University of Freiburg, Faculty for Biology, many times throughout this process and adjusted our experiments accordingly. To see the respective protocols, have a look at the Labbook section between the end of July and start of September. The only time we got a somewhat positive result was when we tested transformation by electroporation for S. elongatus PCC 7942 and S. sp. PCC 6803 with piG_CBM and an unrelated plasmid in parallel to see if the strain or the plasmid we were working with might be the cause for the experienced complications. Here we saw a faint colony for S. sp. PCC 6803 electroporated with piG_CBM DNA, which appeared about 3 days after plating the cells out and disappeared 2 days later. After all, it turned out that the S. sp. PCC 6803 culture used was contaminated.

We did acquire S. elongatus PCC 7942 cells from a different cryo-stock afterwards, yet in the following experiments, attempts to transfer the necessary genes for B12 production into the new strain were unsuccessful as well. This leads us to the conclusion that either we had a fundamental error across all iterations of conjugation and transformation by electroporation or the used shuttle vector does not work as expected. At the start of September, after we reached out to the aforementioned labs in Freiburg working with cyanobacteria for some more advice, Jun.-Prof. Khaled Selim suggested trying to clone with the low copy backbone pVZ322 [12], in case the plasmid we used until then was the problem. We started with this, but it did not work out smoothly and it became clear that time would not suffice to transfer the construct containing ssuE and bluB into S. elongatus PCC 7942 to then cultivate these bacteria and analyze samples with LC-MS.

We do not know what exactly went wrong and it would take us more time to investigate, time we did not have enough of within the scope of this project. We hope our documented process can help you with your microalgal journey. Maybe you, reading this, will be able to establish de novo B12 synthesis in cyanobacteria.