Expression Levels of the γ -Glutamyl Hydrolase I Gene Predict Vitamin B 9 Content in Potato Tubers

: Biofortiﬁcation of folates in staple crops is an important strategy to help eradicate human folate deﬁciencies. Folate biofortiﬁcation using genetic engineering has shown great success in rice grain, tomato fruit, lettuce, and potato tuber. However, consumers’ skepticism, juridical hurdles, and lack of economic model have prevented the widespread adoption of nutritionally-enhanced genetically-engineered (GE) food crops. Meanwhile, little e ﬀ ort has been made to biofortify food crops with folate by breeding. Previously we reported > 10-fold variation in folate content in potato genotypes. To facilitate breeding for enhanced folate content, we attempted to identify genes that control folate content in potato tuber. For this, we analyzed the expression of folate biosynthesis and salvage genes in low- and high-folate potato genotypes. First, RNA-Seq analysis showed that, amongst all folate biosynthesis and salvage genes analyzed, only one gene, which encodes γ -glutamyl hydrolase 1 (GGH1), was consistently expressed at higher levels in high- compared to low-folate segregants of a Solanum boliviense Dunal accession. Second, quantitative PCR showed that GGH1 transcript levels were higher in high- compared to low-folate segregants for seven out of eight pairs of folate segregants analyzed. These results suggest that GGH1 gene expression is an indicator of folate content in potato tubers.


Introduction
Folates are essential micronutrients in the human diet as they fulfill important cellular functions. The main sources of folate in the human diet are plants, with leafy green vegetables and certain fruits being very good sources. However, staple crops such as potato, rice, or corn currently contain relatively low levels of folates [1,2]. Folate malnutrition is considered to be a global problem, with impoverished and developing regions being some of the most affected areas [3]. In areas of the world that have mandatory folic acid fortification programs such as the United States and Canada, folate intake is still sub optimal [4]. The improvement of staple crops' folate content is an attractive strategy for helping to alleviate health problems related to folate deficiency [2].
Folates are a small family of cofactors involved in one-carbon unit reactions. In mitochondria, they are required for the synthesis of formylated methionyl-tRNAs, the interconversions of serine and glycine, and the catabolism of histidine and purines [5][6][7]. In the cytosol, folates are important cofactors in the synthesis of thymidylate and the remethylation of homocysteine to methionine, the precursor The focus of this study was to investigate the expression of folate metabolism genes as potential folate level determinants in tubers of low-and high-folate potato genotypes. RNA-Seq analysis showed that one gene, GGH1, was expressed at higher levels in high folate compared to low folate genotypes. These results were confirmed on additional high-and low-folate materials by quantitative real-time PCR.

Potato Tuber Material
In a previous study [32], we found high folate concentrations in tubers pooled from four individuals from the wild species Solanum boliviense accession PI 597736. As a follow up fine screening, individual seeds from this accession were planted on 5 December 2011, and plant cuttings were made on 27 February 2012 to produce clonal replicates. Plants were grown under glasshouse in Sturgeon Bay, WI. Tubers from each of four clones named fol 1.6, fol 1.3, fol 1.5 and fol 1.11 were harvested in June 2012. Because of low tuber production, tubers from clonal replicates were pooled as one biological replicate. These four clones were re-propagated from stolon shoots in early May 2012, and then tubers from two plants per clone were harvested in November 2012 and kept separated (= two biological replicates), and re-evaluated for folate. In 2016 and 2017, clonal plants were propagated in vitro in February, and then transferred to soil in a greenhouse. In 2016, tubers were harvested in summer from several plants, pooled and split into two biological replicates. In 2017, tubers were harvested in July from three individual plants and kept separated (= three biological replicates).
True seeds of wild and primitive cultivated species S. boliviense PI 597736, S. tuberosum subsp. andigenum PI 225710 and PI 546023, and S. vernei PI 558149 and PI 500063 were obtained from the U.S. potato genebank (USDA Agricultural Research Service Germplasm Resource Information Network (GRIN), www.ars-grin.gov). Seeds were soaked in 1 g/L GA3 overnight before planting to Metro-mix in May 2014. When plantlets reached approximately 8-cm tall, they were transplanted into 8-cm square individual pots containing Sunshine Mix LA4P in a greenhouse. All-purpose fertilizer 20-20-20 was applied at 200 mg/L once a week until senescence. Plants were watered twice a week until senescence. Vines were killed on 31 October 2014, and tubers were harvested on 10 November. Greenhouse temperature was set at 21 • C day time and 15 • C night time. Supplemental light was provided for 14 h per day from a mixture of 400 Watt high pressure sodium and 1000 Watt metal halide lamps. Tubers were then harvested and processed as described previously [32].
An F 1 population named BRR1 was produced by crossing two high folate lines, fol 1.6 with fol 1.7, the idea being to "purify" high folate. An F 2 population named BRR3 was produced by crossing the high folate fol 1.6 with a low/medium folate clone USW4self#3.

Folate Analysis by Microbiological Assay
Folates were extracted by using a tri-enzyme extraction method as previously published [31,35]. Freeze-dried potato samples (100 mg) were used for all folate extractions. Extracts were flushed with nitrogen and stored at −80 • C until analysis by microbiological assay. Controls containing all reagents but potato samples were used to determine the amount of any residual folates in the reagents. There were no detectable folates in any of the reagents used. Folate concentrations were measured by microbiological assay using Lactobacillus rhamnosus as previously described [32].

RNA Isolation
RNA was extracted using a modified hot phenol method as described previously [45]. One hundred milligrams of freeze dried tuber powder (for qPCR analysis) or 1-2 g fresh tuber tissue (for RNA-Seq analysis) were added to a mixture of 4 mL pre-warmed phenol (pH 4.3) and 4 mL extraction buffer consisting of 100 mM LiCl, 100 mM Tris pH 8.5, 10 mM ethylenediaminetetraacetic acid, 1% sodium dodecyl sulfate, and 15 mM dithiothreitol. Samples were vortexed and incubated at 60 • C for 20 to 30 min. Four mL of chloroform:isoamyl alcohol (24:1) were added to the solution, and the sample was vortexed and centrifuged at 6800× g for 10 min at 4 • C. The aqueous phase was transferred into a new tube containing 4 mL phenol:chloroform:isoamyl alcohol (25:24:1), vortexed, and centrifuged at 6800× g for 10 min at 4 • C. The previous step was repeated twice with phenol:chloroform:isoamyl alcohol (25:24:1) and twice with chloroform:isoamyl alcohol (24:1). RNAs were precipitated with one volume of 4 M LiCl, washed with 70% ethanol, and re-suspended in 50 µL diethylpyrocarbonate-treated water. Genomic DNA was removed by DNase treatment using the DNA-Free kit™ (Ambion, Austin, TX, USA). RNAs were quantified and normalized to 200 ng/µL using a Nanodrop (Thermo Scientific, Wilmington, DE, USA). For qPCR analysis, two RNA isolations were performed on freeze dried material from each individual plant. Tubers from each individual plant (i.e., one plant is from one seed) were bulked, freeze dried, and ground together.

RNA Sequencing
Two biological repetitions of each clone fol 1.3, fol 1.5, fol 1.6, and fol 1.11 that were harvested in Nov. 2012 were used for RNA extraction. RNA samples (duplicate of each of the clones fol 1.3, fol 1.5, fol 1.6 and fol 1.11) were bar coded, pooled, processed together, and sequenced in one Illumina HiSeq2000 lanes (51-cycle v3 Single End). Illumina library preparation was done at the Center for Genome Research and Biocomputing at Oregon State University using TruSeq RNA. Illumina libraries were quantified by qPCR for optimal cluster density. Mapping of the RNA-Seq reads to the DM potato reference genome [46], transcript assembly, and determination of differences in expression levels were performed using JEANS, a modified version of GENE-counter [47], in combination with NBPSeq [48].

Real Time Quantitative Reverse Transcription PCR
One to 2 µg of RNA were converted to cDNA using New England BioLab's M-MuLV reverse transcriptase (New England BioLabs, Ipswich, MA, USA) and oligo-dT18 primer (Thermo Scientific, Wilmington, DE, USA). RNA template (5-10 µL) was mixed with 1 µL oligo-dT18 and nuclease-free water to a final volume of 12 µL. This solution was placed in a 70 • C water bath for 5 min and then cooled on ice. Eight microliters of reverse transcriptase (RT) master mix (composed of 2 µL 10 × M-MuLV buffer, 2 uL 10 mM dNTPs, 0.25 uL M-MuLV reverse transcriptase, and 3.75 uL nuclease-free water) were then added to each sample. RT reactions were carried out on a Bio-Rad C1000 thermocycler (Bio-Rad Laboratories, Hercules, CA, USA). The RT cycle was 25 • C for 5 min, 42 • C for 1 h, 65 • C for 20 min. Samples were then stored in a −20 • C freezer until analysis. cDNA templates were diluted four times prior to use in qPCR reactions.
PCR reactions were run in technical quadruplet in 25 uL total volume (4 uL of diluted cDNA samples, 2.5 uL of 0.1 µM primers, and 12.5 uL Taqman Mastermix II). Comparison of expression was made between individuals from the same segregating population, species, or harvest since the samples selected for this study were not biological replicates. This method was described previously in "example 4" [49]. For instance, there is no justification for comparing GGH1 expression between S. vernei and S. tuberosum subsp. andigenum (i.e., Tbr PI 225710 vs. Vrn PI 500063) because they are different species, so comparisons were only made between a high folate and low folate individual from S. vernei or S. tuberosum subsp. andigenum (i.e., high folate Tbr PI 225710 vs. low folate Tbr PI 546023). Mean and standard deviation for each sample were calculated from technical quadruplates using the [49]. The difference between samples within each population was then calculated to determine the fold change in expression of GGH1 between high folate individuals and low folate individuals.

Folate Content and Profile in S. boliviense PI 597736 Individuals
In a previous study [32], we found high folate concentrations in tubers pooled from four individuals from the wild species S. boliviense accession PI 597736. Subsequently, individual seeds from this accession were planted, and four individual plants, named fol 1.6, fol 1.3, fol 1.5 and fol 1.11, were selected for clonal propagation. Clonal replicates were grown in a greenhouse during four independent seasons (n = 4) and tubers were harvested and evaluated for folate content. There were substantial differences in folate concentrations between clones in all four harvests (Figure 1), indicating that individual clones from the accession PI 597736 are segregating for folate content. Folate concentrations were the highest in clone fol 1.6, followed by, in order, clones fol 1.3, fol 1.5 and fol 1.11 in all four harvests, except in June 2012 where folate concentrations in fol 1.5 were lower than those in fol 1.11. Based on relative folate content, to the exception of 2016, we could clearly group fol 1.6 and fol 1.3 as high folate clones, and fol 1.5 and fol 1.11 as low folate clones. Differences in absolute folate content between harvests could be due to environmental differences (e.g., greenhouse temperature, day length, postharvest storage temperature) and tuber maturity.
PCR reactions were run in technical quadruplet in 25 uL total volume (4 uL of diluted cDNA samples, 2.5 uL of 0.1 μM primers, and 12.5 uL Taqman Mastermix II). Comparison of expression was made between individuals from the same segregating population, species, or harvest since the samples selected for this study were not biological replicates. This method was described previously in "example 4" [49]. For instance, there is no justification for comparing GGH1 expression between S. vernei and S. tuberosum subsp. andigenum (i.e., Tbr PI 225710 vs. Vrn PI 500063) because they are different species, so comparisons were only made between a high folate and low folate individual from S. vernei or S. tuberosum subsp. andigenum (i.e., high folate Tbr PI 225710 vs. low folate Tbr PI 546023). Mean and standard deviation for each sample were calculated from technical quadruplates using the 2 −ΔCt [2 -CtGGH1-CtEF1-α ] method [49]. The difference between samples within each population was then calculated to determine the fold change in expression of GGH1 between high folate individuals and low folate individuals.

Folate Content and Profile in S. Boliviense PI 597736 Individuals
In a previous study [32], we found high folate concentrations in tubers pooled from four individuals from the wild species S. boliviense accession PI 597736. Subsequently, individual seeds from this accession were planted, and four individual plants, named fol 1.6, fol 1.3, fol 1.5 and fol 1.11, were selected for clonal propagation. Clonal replicates were grown in a greenhouse during four independent seasons (n = 4) and tubers were harvested and evaluated for folate content. There were substantial differences in folate concentrations between clones in all four harvests (Figure 1), indicating that individual clones from the accession PI 597736 are segregating for folate content. Folate concentrations were the highest in clone fol 1.6, followed by, in order, clones fol 1.3, fol 1.5 and fol 1.11 in all four harvests, except in June 2012 where folate concentrations in fol 1.5 were lower than those in fol 1.11. Based on relative folate content, to the exception of 2016, we could clearly group fol 1.6 and fol 1.3 as high folate clones, and fol 1.5 and fol 1.11 as low folate clones. Differences in absolute folate content between harvests could be due to environmental differences (e.g., greenhouse temperature, day length, postharvest storage temperature) and tuber maturity.  (n = 2), one biological replication being made of tubers harvested from one plant. Data for 2016 are means ± SE of three technical determinations on each of two biological replicates (n = 2), one biological replication being made of tubers pooled from several plants and then split into two biological replicates. Data for 2017 are means ± SE of three technical determinations on each of three biological replicates (n = 3), one biological replication being made of tubers harvested from one individual plant. Samples that share identical letters were not significantly different as determined by ANOVA and Tukey HSD at a p-value = 0.05.
We then used an HPLC method to detect individual folate derivatives in high and low folate clones. We used samples harvested in November 2012 because of the availability of biological repetitions (as opposed to June 2012) and because tubers from 2016 and 2017 had not yet been grown at the time of the analysis. In all samples, 5-CH 3 -THF constituted the main folate form accumulated in potato tubers with >74% of the total folate pool, 5-CHO-THF comprised around 12% of the total pool followed by 5,10-CH = THF (10%) with very small amounts of THF ( Figure 2).  Then, we determined the glutamylation profile of 5-CH3-THF in high versus low folate clones. Two main glutamylation forms of 5-CH3-THF were present in the clones analyzed: monoglutamate (Glu1) and pentaglutamate (Glu5) (Figure 3). Other forms were Glu2, Glu3, Glu4, and Glu6. In the high folate clones fol 1.3 and fol 1.6, 5-CH3-THF was predominantly found in the Glu1 form, comprising 52 and 47% of the total 5-CH3-THF, respectively, while low-folate clones fol 1.5 and fol 1.11 only had 36 and 26% of this folate in its monoglutamyl form. In low-folate clones fol 1.5 and fol 1.11, the Glu5 form was predominant ( Figure 3) and represented between 44 and 55% of total 5-CH3-THF. In highfolate clones, Glu5 was not predominant and represented between 33 and 36% of total 5-CH3-THF. The amount of Glu3-Glu6 was very similar among all clones. Since 5-CH3-THF is the main folate in potato tubers (Figure 2), the observed increase in the monoglutamyl form in the high-folate clones can be considered as one of the principal contributing factors to the total folate increase. Based on differences in folate profiles and total folate contents between clones, we analyzed gene expression between clones to identify potential key regulatory control genes. Then, we determined the glutamylation profile of 5-CH 3 -THF in high versus low folate clones. Two main glutamylation forms of 5-CH 3 -THF were present in the clones analyzed: monoglutamate (Glu 1 ) and pentaglutamate (Glu 5 ) (Figure 3). Other forms were Glu 2 , Glu 3 , Glu 4 , and Glu 6 . In the high folate clones fol 1.3 and fol 1.6, 5-CH 3 -THF was predominantly found in the Glu 1 form, comprising 52 and 47% of the total 5-CH 3 -THF, respectively, while low-folate clones fol 1.5 and fol 1.11 only had 36 and 26% of this folate in its monoglutamyl form. In low-folate clones fol 1.5 and fol 1.11, the Glu 5 form was predominant ( Figure 3) and represented between 44 and 55% of total 5-CH 3 -THF. In high-folate clones, Glu 5 was not predominant and represented between 33 and 36% of total 5-CH 3 -THF. The amount of Glu 3 -Glu 6 was very similar among all clones. Since 5-CH 3 -THF is the main folate in potato tubers (Figure 2), the observed increase in the monoglutamyl form in the high-folate clones can be considered as one of the principal contributing factors to the total folate increase. Based on differences in folate profiles and total folate contents between clones, we analyzed gene expression between clones to identify potential key regulatory control genes.

Expression of Folate Related Genes in Fol Lines as Determined by RNA-Seq Analysis
We investigated differential gene expression in high versus low folate clones harvested in November 2012. RNA-Seq reads were aligned to the reference DM potato genome. At least 76% of the reads aligned to the potato genome for clones fol 1.3, fol 1.5, and fol 1.6. However, less than 30% of the reads aligned to the reference potato genome in the case of fol 1.11. Further analysis showed that a large proportion of unmapped reads matched with potato virus X sequences, showing that fol 1.11 plants were infected with the virus. Therefore, fol 1.11 samples were removed from further expression analysis. Comparisons of gene expression were made between fol 1.3 (high folate) and fol 1.5 (low folate), and fol 1.6 (high folate) and fol 1.5 (low folate). Using a q cutoff value of 0.05 and a |log2 fold change| cutoff value of 1.5, 464 and 383 genes were differentially expressed between fol 1.6 and fol 1.5, and fol 1.3 and fol 1.5, respectively. Only 21.2% (148) of differentially expressed genes were common between the two comparisons ( Figure S2). Functional enrichment analysis showed that ten and one gene ontology terms of molecular functions were enriched in the comparisons fol 1.6 versus fol 1.5, and fol 1.3 versus fol 1.5, respectively ( Figure S3). The gene ontology term "sulfotransferase activity" was the only common term between the two comparisons.
Next, we examined 14 genes that are known to be involved in folate metabolism (Table S1). Interestingly, only one of these genes, GGH1, showed a log2 fold-change greater than 1.5 between the high and low folate genotypes in both comparisons (GGH3 also had significant fold-change but it was not considered reliable because of the very low count number) (Tables 1 and 2). In the fol 1.6 over fol 1.5 comparison, ADCL also showed a large log2 fold change (i.e., 3.4), but the log2 fold change was only 0.3 in the fol 1.3 over fol 1.5 comparison. Based on these results, we investigated whether GGH1 expression was consistently higher in high folate clones compared to low folate clones.

Expression of Folate Related Genes in Fol Lines as Determined by RNA-Seq Analysis
We investigated differential gene expression in high versus low folate clones harvested in November 2012. RNA-Seq reads were aligned to the reference DM potato genome. At least 76% of the reads aligned to the potato genome for clones fol 1.3, fol 1.5, and fol 1.6. However, less than 30% of the reads aligned to the reference potato genome in the case of fol 1.11. Further analysis showed that a large proportion of unmapped reads matched with potato virus X sequences, showing that fol 1.11 plants were infected with the virus. Therefore, fol 1.11 samples were removed from further expression analysis. Comparisons of gene expression were made between fol 1.3 (high folate) and fol 1.5 (low folate), and fol 1.6 (high folate) and fol 1.5 (low folate). Using a q cutoff value of 0.05 and a |log 2 fold change| cutoff value of 1.5, 464 and 383 genes were differentially expressed between fol 1.6 and fol 1.5, and fol 1.3 and fol 1.5, respectively. Only 21.2% (148) of differentially expressed genes were common between the two comparisons ( Figure S2). Functional enrichment analysis showed that ten and one gene ontology terms of molecular functions were enriched in the comparisons fol 1.6 versus fol 1.5, and fol 1.3 versus fol 1.5, respectively ( Figure S3). The gene ontology term "sulfotransferase activity" was the only common term between the two comparisons.

GGH1 Expression in Various Low and High Folate Germplasm as Determined by Real-Time Quantitative RT-PCR Analysis
To investigate whether the differential GGH1 gene expression observed in low and high folate fol lines by RNA-Seq was a consistent pattern between low and high folate genotypes, GGH1 gene expression was determined in eight additional high and low folate individuals. These individuals were from two segregating populations, BRR1 and BRR3 (see Materials and Methods), and from the species S. tuberosum subsp. angidenum and S. vernei. Individuals from BRR1 and BRR3 populations segregated for folate content ( Figure S5). Overall, folate levels ranged from below 500 ng/g dry weight to more than 2000 ng/g dry weight. High-(above 2000 ng/g dry weight) and low-(below 500 ng/g dry weight) folate individuals were selected from the BRR1 and BRR3 populations for further analysis (Table 3). In addition, high-and low-folate individuals from the wild and primitive cultivated species S. tuberosum subsp. andigenum and S. vernei were selected based on a previous study [35] ( Table 3). The fol lines harvested in November 2012 were also used for further gene expression analysis by qPCR ( Figure 1). Pairwise comparison between high and low folate samples within the fol populations showed significant differences in mean GGH1 expression, with fol 1.6/fol 1.11 and fol 1.6/fol 1.5 showing a 15-fold and 88-fold difference, respectively, and fol 1.3/fol 1.11 and fol 1.3/fol 1.5 showing a 24-fold and 140-fold difference. In 7 out of 8 comparisons GGH1 expression was higher in high folate versus low folate genotypes, with fold change ranging from 2 to 481 (Table 4). Only one pair of genotypes, BRR3 90 and BRR3 56, showed the inverse trend, with a 10-fold higher GGH1 expression in the low folate genotype (BRR3 56) compared to the high folate genotype (BRR3 90). High folate versus low folate genotypes from the species S. vernei showed the greatest difference in GGH1 expression (481-fold difference).

Discussion
Variation in folate content exists in potato tubers [31,32,35], suggesting that breeding potato for increased folate content is possible. However, the genetic control of this variation is unknown. In this study, we showed that the expression of GGH1 was consistently higher in tubers with high-compared to low-folate content, a scenario not found for other known folate biosynthesis and salvage genes. We observed a similar higher GGH1 expression in very small, developmentally-young tubers that had higher folate content compared to larger, more mature tubers [34], which indicates that the correlation between high-folate content and high GGH1 expression is recurrent in potato tubers.
GGHs are vacuolar enzymes that cleave glutamate residues from polyglutamylated folate molecules that are stored in the vacuole [28]. Consistent with a GGH enzymatic activity, high folate genotypes that had higher GGH1 expression had a higher proportion of monoglutamylated than polyglutamylated 5-CH 3 -THF, the predominant folate species in potato tubers. In Arabidopsis, GGH1 cleaves glutamate residues from polyglutamylated folate molecules mainly to di-and triglutamates while GGH2 yields mainly monoglutamates [28]. Therefore, one would expect a higher proportion of di-and triglutamylated rather than monoglutamylated folates in potato tubers that have higher GGH1 expression. However, the amino acid sequence similarity between potato GGHs and either of Arabidopsis GGH1 or GGH2 ranges between 62 and 68% ( Figure S6), making it difficult to identify the closest ortholog. Therefore, enzymatic activities of potato GGHs will need to be tested to determine which glutamylated species these enzymes yield. Based on the glutamate profiles, PGSC0003DMG400007066-encoded GGH, herein named potato GGH1, seems to yield monoglutamates.
We propose two hypotheses to explain the positive correlation between GGH1 expression and folate content in potato tubers. The first one implies that higher GGH1 expression is the cause of higher folate content and is based on the capability of GGHs to cleave the glutamate of p-ABA-glutamate, a product of folate degradation, to free p-ABA that can re-enter the biosynthesis pathway. Under this scenario, higher GGH1 activity increases salvage of p-ABA-glutamate. To effect folate biosynthesis, dihydropterin-6-aldehyde, the corresponding pterin degradation product [18], must be salvaged by NADPH-dependent pterin aldehyde reductase. However, this hypothesis seems unlikely since a three-fold overexpression of GGH in vacuoles caused 40-45% reduction in total folate in Arabidopsis leaves [37]. The second hypothesis implies that higher GGH1 expression is the consequence of higher folate production in potato tubers of high folate segregants. In such scenario, high-folate segregants produce more folate than low-folate segregants because of, for instance, higher folate biosynthesis enzymatic activities (e.g., higher GCHI and/or ADCS activities; note that fol 1.6 had higher ADCL gene expression which could correlate to higher ADCL activity and higher p-ABA production) and/or transport of biosynthesis intermediates (e.g., transport of HMDHP from the cytosol to the mitochondria). In high-folate segregants, folate content reaches a threshold (that is never attained in low-folate segregants) that triggers induction of GGH expression. Elevated GGH expression increases hydrolysis of tetrahydrofolate polyglutamate to tetrahydrofolate monoglutamate, a known inhibitor of the DHPS domain of HPPK/DHPS and potential key regulator of folate biosynthesis [1,50], thereby preventing further accumulation of folate and maintaining folate homeostasis in potato tubers. This inhibition would require transport and accumulation of monoglutamylated folates in the mitochondria where HPPK/DHPS is located.
This second hypothesis suggests that knocking down or out GGH expression (e.g., by silencing or gene editing) in high folate lines could unleash the control of folate homeostasis and lead to further accumulation of folate in potato tubers. This hypothesis is in agreement with a report showing that knocking down GGH activity increased total folate content by 34% in Arabidopsis leaves. Two important consequences should be considered: monoglutamylated folates are more bioavailable but also more prone to chemical and enzymatic degradation. Recent folate engineering work showed that combining the two-genes strategy (i.e., overexpression of GCHI and ADCS) with the overexpression of FPGS, the enzyme that adds glutamates to monoglutamylated folates to yield polyglutamylated folates, increased total folates content and stability during long-term storage in potato tubers [42]. Therefore, a fine balance between monoglutamylated and polyglutamylated folates, and therefore fine-tuning of FPGS and GGH expression, may be a requirement to achieve the most impactful folate biofortification.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4395/9/11/734/s1. Figure S1. Partial alignments of GGH (A) and EF1-α (B) sequences and primers used in qPCR reactions. GGH1, GGH2 and GGH3 transcripts sequences from the DM genotype were aligned in order to design primers specific of GGH1. EF1-α sequences from four potato genotypes were aligned to design primers within highly conserved regions. Figure S2. Venn diagram showing numbers of differentially expressed genes that are common or exclusive in comparisons fol 1.6 versus fol 1.5 and fol 1.3 versus fol 1.5. Figure S3. Functional enrichment analysis in comparisons between fol 1.6 and fol 1.5, and fol 1.3 and fol 1.5. Analyses were done using g:GOSt in g:Profiler. A g:SCS threshold of 0.05 was used. The top panel is a Manhattan plot of enriched terms in the comparison between fol 1.6 and fol 1.5. The middle panel is a Manhattan plot of enriched terms in the comparison between fol 1.3 and fol 1.5. The bottom table provides detailed information such as data source, id and name of the term with corresponding p-value. The light circles in Manhattan plots represent insignificant terms. Figure S4. Phylogenetic tree of UDP-glucose glucosyltransferases from potato and Arabidopsis. Homologs of the At1g05560-encoded protein were searched in the potato genome by using tBLASTn search in Spud DB (http://potato.plantbiology.msu.edu/). The seven top matches were used for phylogenetic analysis in MEGA7. Figure S5. Histogram of number of individuals within folate concentration brackets. A, BRR1 population; B, BRR3 population. Figure S6. Alignment of potato and Arabidopsis GGH proteins. Asterisks indicate conserved residues that are catalytically essential in human GGH or other conserved active site residues that may participate in substrate binding [29]. Alignment was done by CLUSTALW (https://www.genome.jp/tools-bin/clustalw). Shading was done by BOXSHADE (https://embnet.vital-it.ch/software/BOX_form.html). AtGGH1, At1g78660; AtGGH2, At1g78680; AtGGH3, At1g78670; StGGH1, PGSC0003DMG400007066; StGGH2, PGSC0003DMG400021256; StGGH3, PGSC0003DMG400035974. Table S1. Folate metabolism-related genes in Arabidopsis, tomato, and potato. Orthologs of Arabidopsis genes in tomato and potato were retrieved from EnsemblPlants.