Folic acid instability and bacterial metabolism combine to impact C. elegans development and ageing

Folic acid supplementation is used to prevent folate deficiency, but it is also associated with negative effects on human health. Little is known about how gut bacteria interact with the uptake of synthetic supplements, such as folic acid, and the consequences for host health. Using the simplified C. elegansE. coli host-microbe model system, we examine how folic acid impacts E. coli folate synthesis, and in turn, C. elegans health. We find that folic acid supplements contain a breakdown product that is taken up by the E. coli transporter AbgT, leading to increased bacterial folate levels. We show that this is the main route by which folic acid rescues a C. elegans developmental folate deficiency, but is also the route by which folic acid shortens adult lifespan. Together, this work shows how folic acid instability and bacterial uptake can combine to influence host health. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted December 6, 2017. ; https://doi.org/10.1101/230227 doi: bioRxiv preprint


INTRODUCTION
Folic acid supplementation has been a major public health success, particularly in the prevention neural tube defects (Imbard et al., 2013), but there are numerous health concerns related to oversupplementation (Butterworth and Tamura, 1989;Cole et al., 2007;Kim, 2007;Marean et al., 2011;Milne et al., 1984;Pickell et al., 2011;Portal-Celhay and Blaser, 2012;Selhub et al., 2009). Folic acid is a synthetic oxidized folate. It can be absorbed in the gut in this form (Patanwala et al., 2014) but unlike bacterial and dietary folates, it must be converted into a reduced tetrahydrofolate (THF), before it can be taken up in peripheral cells by reduced folate carriers and used in metabolism (Ducker and Rabinowitz, 2017;Visentin et al., 2014). The routes of folic acid uptake and metabolism are not completely characterized and its interaction with gut microbes has not been well studied (Visentin et al., 2014).
Some bacteria can take up folates directly but many cannot. Instead they either make folate de novo or take up folate precursors such as para-amino benzoic acid (PABA, Figure 1, (Carter et al., 2007;LeBlanc et al., 2013;Magnusdottir et al., 2015). Folic acid and THFs are inherently unstable molecules, comprising of a central PABA moiety linked by a methylene bridge to a pteridine ring to form pteroic acid and by its carboxyl group to a L-glutamic acid residue by a peptide bond (Green and Matthews, 2007). The methylene bridge is prone to disassociation under several parameters, including low pH, to generate pteridine and PABA-glutamic acid (PABA-glu) (De Brouwer et al., 2007;Gazzali et al., 2016;Gregory, 1989;Hanson and Gregory, 2011;Maruyama et al., 1978), where PABA-glu can further dissociate to generate PABA (Der-Petrossian et al., 2007;Thiaville et al., 2016). Thus, the acidic microenvironments of the stomach and upper small intestine may cause folates to degrade to PABA-glu and PABA (Seyoum and Selhub, 1998). Indeed, PABA has been detected as a significant faecal excretory product following folic acid supplementation (Denko Cw Fau -Grundy et al.). Mammalian studies have shown that injection of labelled PABA into the cecum of rats (Rong et al., 1991) and piglets (Asrar and O'Connor, 2005) results in the incorporation of bacterially synthesized folate into host tissues. Human studies have shown that bacterial folate synthesis in both the human small (Camilo et al, 1996) and large intestine (Kim et al., 2004) can be incorporated into host tissues. It is not known, however, if this bacterial folate synthesis is enhanced by folic acid supplementation and if it is, whether there are any consequences for microbiome function and host health.
This study uses a simplified host-microbe model system to examine how folic acid affects bacterial folate synthesis and host health. The model organism, the nematode Caenorhabditis elegans is maintained on lab strains of E. coli for which it depends on for folate as well as for other vitamins (Yilmaz and Walhout, 2014).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted December 6, 2017. ;https://doi.org/10.1101/230227 doi: bioRxiv preprint Like several other Proteobacteria, E. coli is capable of de novo folate synthesis and the salvage of PABA and PABA-glu to make folate (Carter et al., 2007), but cannot take up intact folates (Nickerson and Webb, 1956;Webb, 1955) Figure 1). Here, we find that the major route of folic acid supplementation is an indirect route that this is dependent on the breakdown of folic acid into PABA-glu and subsequent uptake via the E. coli transporter, AbgT. The resultant increase in bacterial folate synthesis is beneficial for C. elegans during development, but has a detrimental impact on C. elegans ageing. This is consistent with our previous work, which has demonstrated that E. coli folate synthesis shortens C. elegans lifespan (Virk et al., 2012) most likely due to a bacterial-dependent chronic toxicity (Virk et al., 2016). Importantly, we also detect PABA-glu and PABA in three folic acid sources, with the highest quantities in a commercial folic acid supplement.
Together, this work highlights an unappreciated bacterial-dependent route via which folic acid can have both positive and negative effects on host health, and thus has wider reaching implications for the importance of the microbiota in determining health outcomes following vitamin supplementation. Folic acid is composed of a central PABA moiety linked by a methylene bridge to a pteridine and to a single glutamic acid residue via a peptide bond. Folic acid is unstable and disassociates to generate PABA-glutamate (PABA-glu), which can further disassociate to PABA. E. coli is able to salvage PABA-glu via transport by the inner membrane protein, AbgT. PABA-glu is then cleaved intracellularly by the heterodimeric carboxypeptidase AbgA/B into PABA, where it can be used as a precursor for folate synthesis to generate the active form of folate, tetrahydrofolate (THF) which is used in the folate cycle. PABA can also diffuse across biological membranes. E. coli is also capable of de novo folate synthesis by generating PABA from chorismate and glutamine through the action of PabA, PabB and PabC. E. coli is unable to import folic acid directly.

Folic acid supplementation rescues C. elegans developmental folate deficiency via E. coli abgT
In order to test whether folic acid interacts with bacterial folate synthesis to supplement host folate in our model system, we used the previously characterized C. elegans folate uptake deletion mutant, gcp-2.1 (ok1004) (Virk et al., 2016). The gcp-2.1 mutant strain lacks GCP-2.1, the C. elegans homologue of the GCPII glutamate carboxypeptidase enzyme that removes glutamates from polyglutamated folates, a prerequisite for folate absorption (Halsted et al., 1998). This strain exhibits a severe developmental folate-deficiency when maintained on E. coli treated with the sulfonamide drug sulfamethoxazole (SMX) (Virk et al., 2016).
This growth defect is rescued with high concentrations of folic acid (Virk et al., 2016). It is not clear, however, whether this supplementation is direct (C. elegans uptake) or indirect (restoration of folate synthesis in E. coli). Here we show that the C. elegans gcp-2.1 mutant grown on the E. coli pabA mutant on defined media (DM) has the same growth defect as these worms grown on wild type E. coli treated with 128 µg/ml SMX ( Figure 2a). Folic acid was found to increase gcp-2.1 mutant body length on the pabA mutant in a dose-dependent manner ( Figure 2b). As E. coli cannot uptake intact folic acid, we examined the dependence of folic acid rescue on the expression of the E. coli abgT gene, which is responsible for the uptake of PABA-glu. A weaker response of the gcp-2.1 mutant to folic acid was observed on the abgTpabA double mutant compared to worms on the pabA single mutant. In contrast, gcp-2.1 mutants maintained on pabA with a plasmid overexpressing abgT (Carter et al., 2007) (pabA (abgT OE)) were more responsive to folic acid (Figure 2b). Analysing the experiment by two-way ANOVA, we find that there is a significant interaction effect of strain type (F=102.67, p<0.0001) and folic acid concentration (F=123.55, p<0.0001) on C. elegans gcp-2.1 body length. Mutation of abgT alone did not influence growth of the gcp-2.1 (Supplementary Figure 1a). Rescue of gcp-2.1 developmental folate deficiency was achieved at a 10-fold lower concentration and independently of abgT expression, with the relatively stable THF, folinic acid (Supplementary Figure 2), suggesting that C. elegans takes up folinic acid directly. These results suggest that the major route of folic acid uptake by the worm is through E. coli via uptake of the folic acid degradation product, PABA-glu by the AbgT transporter.

Folic acid supports E. coli growth via AbgT-dependent uptake of PABA-glu
The abgT-dependence of folic acid to rescue the gcp-2.1 developmental folate-deficiency strongly suggests that PABA-glu is available to E. coli following folic acid supplementation. To test the relative ability of E. coli to take up PABA-glu and folic acid, we added these compounds to pabA, abgTpabA and pabA (abgT OE) E.
coli and assessed growth on DM agar plates after 4 days incubation at 25°C. Under these conditions, we found that all strains containing the pabA mutant showed slower growth than wild type and abgT expression determined the response to folic acid and PABA-glu; 10 µM folic acid rescued growth of the Figure 2. Folic acid supplements C. elegans via an E. coli abgT-dependent route during development a) body length of wild-type and gcp-2.1 mutant C. elegans at L4 stage raised on DM agar plates seeded with WT E. coli (control), pabA mutant or WT E. coli treated with 128 μg/ml SMX b) body length of gcp-2.1 mutant C. elegans at L4 stage raised on DM agar plates seeded with pabA mutant, abgTpabA double mutant or pabA mutant over-expressing abgT with increasing concentrations of folic acid. By two-way ANOVA analyses, we find that there is a significant interaction effect of strain type (F=102.67, p<0.0001) and folic acid concentration (F=123.55, p<0.0001) on C. elegans gcp-2.1 body length. Over-expression is conferred by transformation with a high copy number plasmid, pJ128. pabA and abgTpabA strain are transformed with the empty vector, pUC19. Error bars represent standard deviation of C. elegans body length; n ≥40. pabA mutant, whereas 100 µM was needed to achieve an equivalent rescue in the abgT pabA double mutant ( Figure 3ai). 10 µM folic acid increased the growth of the pabA strain over-expressing AbgT above that of the pabA mutant alone or the wild type strain. Supplementation by PABA-glu had a similar effect but at a 10-fold lower concentration than folic acid ( Figure 3aii). PABA, which can freely diffuse across biological membranes (Tran and Nichols, 1991), rescued bacterial growth was independent of abgT expression and achieved at nanomolar concentrations ( Figure 3aiii). Overall, rescue of bacterial growth by folic acid at low concentrations can be explained by PABA-glu uptake by AbgT while low concentrations of PABA present in folic acid preparations may explain the ability of 100 µM folic acid to rescue E. coli growth and C. elegans gcp-2.1 developmental folate-deficiency independently of abgT ( Figure 2).

Folic acid increases E. coli folate levels in an AbgT-dependent mechanism
It order to verify that E. coli growth following folic acid supplementation is attributable to restored bacterial

Folic acid preparations contain PABA-glu and PABA
Together, the data presented here indicate that the main route of C. elegans folic acid supplementation is indirect via E. coli uptake of PABA-glu and PABA. We used LC-MS/MS to test for the expected presence of these breakdown products in folic acid preparations from Schircks (used in all other experiments in this study), Sigma Aldrich and Boots, a UK retailer. We also tested the impact of the experimental conditions used here on folic acid breakdown by analysing extracts from agar media supplemented with Schircks folic acid and incubated at 25 °C for 4 days. We detected PABA-glu in all three folic acid sources at between 0.3% (Schircks) and 4% (Boots) (Figure 3ci). Under the conditions used for C. elegans experiments PABA-glu increased to 1.18%, suggesting further breakdown. PABA was found at between 0.01% (Schircks) and 0.06% (Boots) (Figure 3cii). [PABA] (µM)

Folic acid shortens C. elegans lifespan via AbgT-dependent uptake of PABA-glu during adulthood
Inhibiting bacterial folate synthesis, without affecting bacterial growth, is known to increase C. elegans lifespan (Virk et al., 2012;Virk et al., 2016). It was therefore hypothesized that factors that increase bacterial folate synthesis, such as folic acid (as shown here), may shorten C. elegans lifespan. Consistent with our previous findings (Virk et al., 2016), we find that C. elegans maintained on any E. coli pabA mutant are long-lived compared to C. elegans fed wild type E. coli (Figure 4, Supplementary Table 1

DISCUSSION
Folic acid supplementation generates a supply of folate in addition to that provided by the diet and gut microbiota. It is not clear, however, exactly how folic acid supplements host folate. Considering the conflicting reports of the safety of folic acid supplementation on host health, it is important to understand all possible routes of uptake. Using the established C. elegans-E. coli host-microbe model in which bacterial folate synthesis has been shown to influence host health (Chaudhari et al., 2016;Han et al., 2017;Virk et al., 2016), this study shows that folic acid can increase bacterial folate synthesis (Figure 3b) in a route that is dependent on its breakdown into PABA-glu ( Figure 3c) and uptake by the bacterial transporter AbgT. We have shown that this route can have beneficial consequences in the case of C. elegans developmental folate-deficiency ( Figure 2) but it can also have a negative impact on long-term host health, by shortening lifespan (Figure 4). Our previous work indicates that lifespan decrease is due to a bacterial folatedependent toxicity. Together, we have uncovered a route by which folic acid interacts with bacterial metabolism to impact host health ( Figure 5).
The detection of significant quantities of PABA-glu and PABA in three folic acid sources in this study, including in a commercial folic acid supplement (Figure 3c) indicates that this route may have implications for human folate supplementation. Indeed, investigations into folic acid supplements have reported significant failings of supplements to meet US (Hoag et al., 1997) or UK (Sculthorpe et al., 2001) standards for dissolution. In light of these issues with folic acid stability, manufacturers have adopted a policy of 'overages' in order to ensure sufficient vitamin is released (Andrews et al. 2017). As the E. coli PABA-glu transporter, AbgT, can influence C. elegans health following folic acid supplementation, depending on both bacterial and host genotype, it is hypothesized that supplement stability and gut microbiota composition may be two previously unexplored variables in understanding the health consequences of folic acid supplementation.
The AbgT protein is a member of a conserved family of over 13, 000 transporters, many of which are encoded for in the genomes of several pathogenic bacteria of the human gut microbiome, including Enterobacter cloacae, N. gonorrhoeae, Salmonella enterica, Shigella boydii and Staphylococcus aureus in addition to E. coli. Interestingly, several diseases are associated with an increased abundance of folatesynthesizing gut bacteria, such as inflammatory bowel disease (IBD (Shin et al., 2015) and small intestinal bacterial overgrowth (SIBO, (Camilo et al., 1996). Human trials are necessary to examine how microbial folate synthesis and microbial compositional changes in the human gut microbiota following folic acid supplementation. We hypothesize that there may be specific groups of patients where effects of folic acid on microbial folate synthesis need to be considered. It may be necessary to design alternative strategies to prevent folate deficiency without increasing bacterial folate synthesis. Figure 5. Schematic of the impact of folic acid supplementation on C. elegans via indirect uptake of breakdown products by E. coli. Folic acid is not taken up well by C. elegans directly. We find that the major uptake of folic acid by C. elegans is dependent on its breakdown into PABA-glu and uptake by the E. coli AbgT transporter. This route increases bacterial folate synthesis in both wild-type and pabA mutant E. coli. Under conditions of folate-deficiency (pabA mutant E. coli), increasing bacterial folate via this route is beneficial for C. elegans development. During C. elegans adulthood, this route has a negative impact on longevity as it promotes a bacterial folate-dependent toxicity.

Culture conditions
Defined media (DM) was prepared as described (Virk et al., 2016), except that 10 nM B12 was added. B12, folic acid and antibiotics are added post-autoclaving for agar plates. DM for liquid culture is filter-sterilised.
0.1 µM PABA added to the liquid DM media used to seed the plates in order to maintain bacterial growth (apart the growth experiments in Figure 3a). 30 µl of 3 ml fresh overnight LB culture is used to inoculate 5ml DM (15 ml Falcon tube). 25 µg/ml kanamycin (50 ug/ml ampicillin if necessary) added to both LB and DM pre-incubation. DM liquid cultures are incubated for 18 hr at 37°C, 220 RPM.

C. elegans strains used
SS104 glp-4(bn2), UF208 (wild-type), and UF209 gcp-2.1(ok1004) (Virk et al., 2016). onto the plate and a glass spreader to scrape off the bacterial lawn. The bacterial suspension was pipetted into a 1.5 ml Eppendorf and the volume was recorded (v). Tubes were vortexed vigorously to obtain a homogenised solution. 100 µl was taken and diluted with 900 µl M9 in a cuvette. A spectrophotometer was used to read bacterial growth at 600 nm. Bacterial growth was calculated by multiplying OD 600 by the volume of the sample (v).

E. coli folate extraction
Bacterial lawns were scraped from plates into micro centrifuge tubes using M9 solution and kept on ice.
Volume (v), multiplied by the OD 600 of the solution (diluted 1:5) gives a measure of the amount of material.
Samples were concentrated in chilled microcentrifuge and pellets were snap frozen in liquid nitrogen.
Samples were vortexed vigorously and left on ice for 15 min before centrifugation in a cooled microcentrifuge for 15 min at full speed. Supernatants were used for analysis.

Folate LC-MS/MS analysis
Folates were detected by multiple reaction monitoring (MRM) analysis using a SCIEX QTRAP 6500 instrument. MRM conditions for folic acid, PABA, PABA-Glu, 5-Me-H 4 PteGlu 3 (5-methylTHF-Glu 3 ) and 5/10-CHO-H 4 PteGlu 3 (formyl THF 3 ) were optimised by infusion of standards into the instrument. The optimised conditions for -Glu 3 folates were applied to other higher folates using MRM transitions described by Lu et al., 2007(Lu et al., 2007. Further confirmation of identity for folates of interest was achieved by performing enhanced product ion scans and comparing the fragment spectra with known standards.
The QTRAP 6500 was operated in ESI+ mode and was interfaced with a Shimadzu Nexera UHPLC system.
Samples were separated using a Thermo PA2 C18 column (2.2 µm, 2.1 x 100 mm) with a gradient of 0.1% formic acid in water (mobile phase A) and acetonitrile (mobile phase B). Samples were maintained at 4⁰C and 2 µL aliquots were injected. The column was maintained at 40⁰C with a flow rate of 200 µL/min, starting at 2% B, held for 2 minutes, with a linear gradient to 100% B at 7 minutes, held for 1 minute, before a 7-minute re-equilibration step at 2% B that was necessary for consistent retention times. The column eluate flow to the MS was controlled via the QTRAP switching valve, allowing analysis between 4 and 8 minutes to minimise instrument contamination. Folates were quantified with reference to external standards and matrix effects were assessed by spiking of standards into extracted samples.

Lifespan analysis
Survival analyses were performed as described (Virk et al., 2012). glp-4(bn2) worms were maintained at 15°C and shifted to 25°C at the L3 stage. At the L4/young adult stage, animals were placed on bacteria under the experimental conditions. All lifespan data is in Table S1.