Rare sugars and their derivatives (Supplementary Fig. 1) with respective purities of 100% were prepared by the Rare Sugar Production Station, International Institute of Rare Sugar Research and Education, Kagawa University, Japan using methods described previously2,3,49. Common sugars and other reagents described in the Methods section were purchased from Fujifilm-Wako (Tokyo, Japan) unless noted otherwise.
Effects of d-tagatose on diseases in pot and field trials
d-Tagatose was tested against diseases caused by various oomycete, ascomycete, or basidiomycete pathogens (Table 1) to determine the effective dose (percentage in weight per volume, w/v) that reduced disease severity to less than 50% of that caused by the mock treatment (untreated).
For pot trials, a control value was calculated based on disease severity after a spray of d-tagatose solution at 0.5–10% (w/v) in distilled water containing New Gramin sticker (3 ml 10 liter−1) (Mitsui Chemicals Agro Inc., Tokyo, Japan) compared to severity without the d-tagatose spray by methods described previously43,44,50 (Experimental conditions for the respective host-pathogen combinations for pot trials in Table 1 are summarized in Supplementary Table 1).
Typical disease symptoms and less-severe symptoms after other treatments for cucumber downy mildew are shown as examples in Fig. 1a. A solution (5 ml per pot) of 1 or 5% (w/v) d-tagatose, d-allulose, or d-allose was sprayed on the first true leaf of cucumber seedlings (cv. Sagami–Hanshiro; 1- to 1.2-leaf stage) 5 days before inoculation with 5 ml (2 × 104 sporangia ml−1) per pot of downy mildew pathogen (Pseudoperonospora cubensis, FS-1 strain [stock isolate of National Federation of Agricultural Cooperative Associations]). Inoculated plants were incubated at 20 °C in a moist chamber under 100% humidity for 24 h. A solution (5 ml) of 0.01% (w/v) probenazole [PBZ], 0.005% (w/v) acibenzolar-S-methyl [ASM], or 0.03% (w/v) metalaxyl was also sprayed on the first true leaf of cucumber seedlings (cv. Sagami–Hanshiro) 5 days before inoculation with downy mildew pathogen (5 ml, 2 × 104 sporangia ml−1) as described above. These plants and untreated plants without any sugar or agrochemical treatment were incubated for 7 days in a greenhouse at 20 °C. The timing of the treatments with sugars and agrochemicals also varied from 1 to 7 days before inoculation and from 12 to 72 h after inoculation (5 ml, 1 × 104 sporangia ml−1). Symptoms were then assessed after 7 days under the conditions described above. Severity was calculated from a rating 0–3: 0, no symptoms; 0.1, diseased area 3%; 0.3, 10% diseased area; 0.8, 25% diseased area; 1.5, 50% diseased area; 2, 70% diseased area; 3, 95% or more diseased area (see also Supplementary Table 1) as described previously43,44,50. The relative disease severity was calculated against that caused for the untreated plants based on the average (±SD) (N = 3 independent replications), then the data from the respective curative or preventive experiment were analyzed using a Tukey–Kramer multiple comparison test (p < 0.05) in the program JMP 12 (SAS Institute, Cary, NC, USA) (Fig. 1b).
In field trials, the control efficacy of 1 or 5% (w/v) d-tagatose against five downy mildew–host combinations (grapevine cv. Kyoho caused by Plasmopara viticola, cucumber cv. Ancor 8 by Pseudoperonospora cubensis, Chinese cabbage cv. Kigokoro75 by Hyaloperonospora parasitica, onion cv. Gifuki by Peronospora destructor, or spinach cv. Jiromaru by Peronospora farinosa f. sp. spinaciae) was reexamined and compared with that on plants treated or untreated with an appropriate reference fungicide: cyazofamid FL (CZF; Ranman, Ishihara Sangyo Kaisha, Osaka, Japan), chlorothalonil FL (CTN; Daconil, SDS Biotech K. K., Tokyo, Japan), or mancozeb WP (MCZ; Dithane, Corteva Agriscience, Wilmington, DE, USA) based on the average of disease severity (±SD) (N = 50–163 leaves per replication, three independent replications). Since effective concentrations for the respective fungicides differ against the specific downy mildews, CZF was used at 94 parts per million (ppm) against grapevine and spinach downy mildew, CTN at 400 ppm against Chinese cabbage and cucumber downy mildew, and MCZ at 1875 ppm against onion downy mildew. All isolates of the pathogens, except those from natural infections, were obtained from standard stocks of the Agrochemical Research Center, Mitsui Chemicals Agro, Inc. and maintained in the laboratory as appropriate.
Inoculation of the pathogens and treatment of d-tagatose or respective agrochemicals were summarized in Supplementary Table 2. Disease severity was calculated from the rating based on the Fungicide Evaluation Manual of Japan Plant Protection Association for field trials51 (also see Supplementary Table 2) as described previously44,50. Briefly, plants grown until a specific Biologische Bundesanstalt, Bundessortenamt und CHemische Industrie (BBCH) stage52, were sprayed with 200 to 300 liters d-tagatose solution per 10 a with a sprayer. As one of examples, sporangia of Plasmopara viticola (1 × 103–104 sporangia ml−1) were sprayed 2 days after the d-tagatose or agrochemical treatment, and the treatment of d-tagatose or agrochemical was repeated two more times with about one week interval. The disease severity of the tested plants was scored in accordance with the following criteria 7 days after the final treatment (Supplementary Table 2). The disease severities shown in Fig. 1c–g were calculated based on a disease rating specific for each disease and based on percentage diseased leaf area or number of lesions (e.g., grapevine downy mildew: 0, no symptoms; 1, area <10%; 2, area 11 to 30%; 3, area 31–50%; and 4, area ≥ 51%) (Supplementary Table 2). Respective disease ratings were used to calculate disease severity as 100[(n1 + 2n2 + 3n3 + 4n4)/4 N]; N = total number tested (50–163 leaves per replication, with three replications to calculate the average), n1–n4 = number of leaves rated, respectively as disease rating 1–4 for grapevine, cucumber, onion and spinach, and 100(n1 + 2n2 + 3n3)/3 N; N = total numbers tested (160 leaves per replication, with three replications to calculate the average), n1–n3 = number of leaves for each disease rating 1–3 for Chinese cabbage. The tests were repeated three times for each treatment, and the average disease severity ±SD was calculated and is shown in Fig. 1. All data in Fig. 1 were analyzed using a Tukey–Kramer multiple comparisons (p < 0.05) in the program JMP 12. Experimental conditions for other host-pathogen combinations for field trials in Table 1 are also summarized in Supplementary Table 2.
For testing d-tagatose effects on downy mildew of Arabidopsis (Fig. 2a), seedlings of A. thaliana ecotype Colombia-0 (Col-0), grown in soil in plastic pots (7.5 cm diameter) at 22 °C with 16 h light (100–150 µmol m−2 s−1)/8 h dark for 3 weeks, were sprayed with or without 10 ml of 1% (w/v) aqueous d-tagatose per pot. After 1 day in the same conditions, the plants were sprayed with 1 × 105 conidiospores ml−1 of Hyaloperonospora arabidopsidis isolate Noco2, and then incubated at 18 °C and 90–100% relative humidity with 10 h light (150–200 µmol m−2 s−1)/14 h dark. Plants were photographed at 0, 10, or 20 days post inoculation (dpi) (Fig. 2a, b).
For the tests of d-tagatose effects on downy mildew of Arabidopsis described in Fig. 2c–h, seedlings of A. thaliana ecotype Col-0 were grown on a filter paper (Whatman 3MM Chr) (GE Healthcare, Chicago, IL, USA) on Murashige–Skoog medium with 1% (w/v) sucrose and 0.8% (w/v) agar at 22 °C with 16 h light (50–200 µmol m−2 s−1)/8 h dark for 10 days then transferred to Murashige–Skoog medium (2 ml) containing d-tagatose (up to 100 mM) or 50 mM of deoxygenated d-tagatose at the C-6 position (6-deoxy-d-tagatose, 6dT) (Supplementary Fig. 1). Seedlings on the paper filter were kept for 1 day in the same conditions, and then the seedlings were sprayed with 1 × 105 conidia ml−1 of Hyaloperonospora arabidopsidis isolate Noco2 and incubated at 18 °C and 90–100% relative humidity with 10 h of light (150–200 µmol m−2 s−1)/14 h dark for 7 days.
The extent of hyphal growth of H. arabidopsidis isolate Noco2 7–10 days after inoculation of Arabidopsis seedlings with or without d-tagatose treatment was examined microscopically after trypan blue (0.01% w/v in lactophenol) staining with chloral hydrate destaining36. Typical hyphal growth of H. arabidopsidis isolate Noco2 in Arabidopsis seedlings after treatment with or without 1% (w/v) d-tagatose treatment were imaged at 10 days after inoculation (Fig. 2b). Hyphal growth on 111–138 cotyledons for each treatment 7 days after inoculation was rated based on percentage of leaf area with hyphae: +++ for 75–100% of entire leaf; ++ for 25% to 74%; + for <24%; and – for no hyphae. Typical hyphal growth for each rating, classified as hyphal growth efficiency (%), is shown in Fig. 2c as means ± SD of three independent replications.
Conidiospores produced by H. arabidopsidis isolate Noco2 on 20 (Fig. 2d) or 30 (Fig. 2h) Arabidopsis seedlings with four leaves including cotyledons were collected in water by vortexing at 7 days after inoculation for each treatment described above with or without d-tagatose treatment (up to 100 mM) or 50 mM of 6dT, then counted with a hemocytometer. Each treatment was done six times, and the mean number of conidia per 20 seedlings (±SD) per treatment (Fig. 2d) or per 10 seedlings (±SD) per treatment (Fig. 2h) of six independent replications was calculated.
Conidiophores and oospores produced by H. arabidopsidis isolate Noco2 on or in each cotyledon (N = 37–49) were observed microscopically and counted for each treatment at 7 days after the inoculation described above with d-tagatose treatment up to 100 mM or without. The tests were repeated three times for each treatment, and means (±SD) per cotyledon calculated (Fig. 2e, f).
Typical conidiospores, conidiophores and oospores observed without d-tagatose treatment are shown in Fig. 2g. All data in Fig. 2c–h were analyzed using a Tukey–Kramer multiple comparison test (p < 0.05) in the program JMP 12.
In vitro effects of d-tagatose on Hyaloperonospora arabidopsidis isolate Noco2
To test direct effects on germination and germ tube elongation 6 h after germination, we collected fresh conidiospores of H. arabidopsidis isolate Noco2 and placed 10 µl of 2 × 104 conidiospores ml−1 in each well of a 96-well microtiter plate (Falcon no. 3075; NJ, USA) that held 100 µl Gamborg B5 medium salt mixture solution (Nihon Pharmaceutical, Tokyo, Japan) and 10 µg/ml meropenem with or without d-tagatose or d-mannitol up to 400 mM in each well. The plate was then covered with parafilm and incubated for 48 h (total 54 h) at 12 °C in the dark. Ten microliters of 2 × 104 conidia ml−1 was also placed on Gamborg B5 medium salt mixture solution (Nihon Pharmaceutical) with or without 50 mM d-tagatose or 6-deoxy-d-tagatose in microtiter plates and incubated the same way.
Typical germ tubes in each test were photographed (Fig. 2i), and their lengths were measured 48 h after the sugar was added (at 6 h after germination; 54 h total incubation); mean length ± SD (n = 60–93 independent germ tubes) for each treatment is shown in Fig. 2j, and the mean percentage germination ± SD (n = 85–263 for each of 4–6 independent replications) at 24 h is shown in Fig. 2k. Means among treatments were compared for significant differences using a Tukey–Kramer multiple comparison test (p < 0.05) in JMP 12.
RT-qPCR, microarray, and RNA-seq analyses of regulation of gene expression by d-tagatose
For RT-qPCR analyses of PR-protein and defense-related gene expression in cucumber, seeds of cucumber (Cucumis sativus L. cv. Sagami–hanshiro) were germinated in the dark on a filter paper (Whatman 3MM Chr) moistened with water. After 2–3 days, seedlings were planted in plastic pots (7.5 cm diameter) and grown in a growth chamber at 22 °C for 2 weeks in natural light in July (Kagawa, Japan). Immediately after plants were sprayed with or without 10 ml of 1% (w/v) aqueous d-tagatose per pot (7.5 cm diameter) and air-dried, they were sprayed with or without 1 × 105 conidia ml−1 of Pseudoperonospora cubensis per pot, and all leaves from the plant were sampled at 0, 24, and 48 h after inoculation. The leaves were ground to a fine powder in liquid nitrogen with a mortar and pestle. Total RNA was extracted using Tri Reagent (Sigma, St. Louis, MO, USA). The cDNAs for cucumber genes were prepared with a PrimeScript RT Master mix (Takara, Shiga, Japan), and expression of several PR-protein genes was quantified using RT-qPCR with SYBR Premix Ex Taq II (Takara) and a Thermal Cycler Dice TP800 (Takara). Cycling conditions were as follows: 95 °C for 30 s, followed by 40 cycles of 95 °C for 30 s and 60 °C for 30 s. The transcript level was normalized by comparison with actin (AB010922), and data were analyzed as described previously14,17. Means with SD values (2 to 3 independent replications) were analyzed for significant differences among treatments using a Tukey–Kramer multiple comparison test (p < 0.05) using JMP 12 (Fig. 3a). For selecting target defense-related genes examined above, the same cDNAs from cucumber leaves treated with or without d-tagatose for 24 h under the above conditions were amplified using a Clontech PCR Select cDNA Subtraction Kit (Takara), and 288 random clones that were expected to be upregulated in d-tagatose-treated leaves were sequenced (Supplementary Table 3). Clones with annotated sequences for putative defense-related genes among those 288 clones were selected, and a partial region of each sequence was used for this RT-qPCR analysis as the putative PR-protein and defense-related genes encoding peroxidase (POX)24, lipoxygenase (LOX)25, pore-forming toxin-like protein (Hrf)26, 4-coumarate-CoA ligase (4CL)27,28, and caffeoyl-CoA O-methyltransferase (CCoAMT)27,28. Primers used for these analyses are listed in Supplementary Table 4.
For comparisons of rice PR-protein gene expression after various sugar treatments based on microarray analyses, an Agilent Rice Oligo Microarray (44k, custom-made; Agilent Technologies, Redwood City, CA, USA) was used with the methods described previously16. Briefly, seedlings of two-leaf-stage rice plants were cultured in Kimura B broth containing 0.5 mM of d-tagatose, d-allulose14, d-allose16,17, d-glucose14,16,17, or no-sugar for 2 days. Total RNA was extracted using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). The RNAs (400 ng) were labeled with Cy3 or Cy5 using a Low RNA Input Linear Amplification/Labeling Kit (Agilent Technologies) after hybridization and washing according to the instructions. Hybridized microarrays were then scanned with an Agilent Microarray Scanner (Agilent Technologies), and signal intensity of each spot in the array was delineated, measured and normalized using Feature Extraction Software (version 9.1; Agilent Technologies). The microarray analyses were done three times. Scatter plot analyses, done using the Subio platform 1.22.5473 (Subio, Amami, Japan), are shown in Supplementary Fig. 3. Data extraction processes were performed using three independent replications for each treatment according to the instructions, and relative expression of the respective PR-protein genes was calculated with SD values in relation to data from the no-sugar treatment. The relative expression values were compared using a Tukey–Kramer multiple comparison test (p < 0.05) using JMP 12 in Fig. 3b. All microarray data were deposited as data files in the Gene Expression Omnibus Database with accession GSE19595 for d-allulose14, GSE15479 for d-allose16,17 and d-glucose14,16,17, and GSE136313 for d-tagatose.
For RNA-seq analyses of d-tagatose effects on downy mildew-inoculated Arabidopsis (Fig. 3c, Supplementary Fig. 4), seedlings of A. thaliana ecotype Col-0 grown in soil in plastic pots (7.5 cm diameter) at 22 °C with 16 h light (100–150 µmol m−2 s−1)/8 h dark for 10 days were sprayed either with or without 10 ml of 1% (w/v) aqueous d-tagatose per two plastic pots (7.5 cm diameter) and kept for a day in the same conditions. After the sugar treatment, Arabidopsis plants were sprayed with 1 × 106 conidia ml−1 of H. arabidopsidis isolate Noco2, then incubated at 18 °C and 90–100% relative humidity with 10 h light (150–200 µmol m−2 s−1)/14 h dark for another 6 days. Twenty to thirty biological replicates of plants with conidiospores and conidiophores that were inoculated and treated with or without d-tagatose were ground to a fine powder with a mortar and pestle in liquid nitrogen. Total RNA was extracted using Tri Reagent (Sigma), then purified using an RNeasy Plant Mini Kit (Qiagen). The quality and quantity of the extracted RNA was determined using the Nanodrop Spectrophotometer (Thermofisher) and Agilent 2100 Bioanalyzer (Agilent Technologies). The A260/A280 values and the 28 S/18 S ribosomal RNA ratios of each total RNA sample were 2.06 and 1.7 (d-tagatose-treated sample), and 2.13 and 1.6 (mock-treated sample), respectively. The purified total RNA was sent to Hokkaido System Science (Sapporo, Japan) for RNA-seq. The library for RNA-seq analyses was prepared with 5 µg total RNA using the TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, CA, USA) and the manufacturer’s instructions. The libraries were sequenced using the HiSeq 2000 sequencing platform and TruSeq SBS Kit v3 reagents with 101 cycles. Base-calling and data filtering were performed using CASAVA ver.1.8.1 software (Illumina), ultimately yielding approximately mapped reads or 122 million (d-tagatose-treated sample) or 84 million (mock sample) bases. The mean quality scores for the reads were 35.8 (d-tagatose-treated sample; 91.81% had ≥Q30 bases) and 35.57 (mock sample; 91.18% had ≥Q30 bases). The adapter sequences were trimmed by Cutadapt (v1.1)53. After preprocessing, the reads were mapped to the A. thaliana genome (TAIR10.17) and H. arabidopsidis Emoy2 genome (HyaAraEmoy2_2.0) from the Ensembl database (http://ensemblgenomes.org) using TOPHat (v2.0.2)/Bowtie54. Expression levels of each gene in both d-tagatose-treated and mocked samples were quantified by the number of fragments (paired-end reads) mapped to the coding region of each gene with a value of FPKM (fragments per kilobase of exon per million fragments mapped) using Cuffdiff (v2.0.2) program54. Differential expression between d-tagatose-treated and mocked samples was calculated as a Log2 ratio of respective FPKM values. Statistical significance (q-value) of the comparison using FPKM values was also calculated using Cuffdiff (v2.0.2) program54. All data, including raw sequence files for samples, were deposited in the Gene Expression Omnibus Database as accession GSE136568.
Selected genes for the differential expression analyses were PR-protein and defense-related genes jasmonic acid (JA) and/or ethylene (ET)-signaling genes (JA/ET [PR-proteins]: PR-3/CHIB [At3g12500], PR-4 [At3g04720], PDF1.2a [At5g44420], PDF1.2b [At2g26020]; JA markers: LOX2 [At3g45140], VSP2 [At5g24770], JAZ1 [At1g19180], JAZ5 [At1g17380], JAZ10 [At5g13220], TPS4 [At1g61120], CLH1 [At1g19670]; ET markers: ERF1 [At3g23240], ORA59 [At1g06160], EIN2 [At5g03280], EIN3 [At3g20770]), salicylic acid (SA)-signaling genes (SA-related PR-proteins: PR-1 [At2g14610], PR-2 [At3g57260], PR-5 [At1g75040]; SA markers: EDS1 [At3g48090], ICS1 [At1g74710], TGA2 [At5g06950], TGA5/OBF5 [At5g06960], TGA6 [At3g12250]); MAMPs response genes (chitin/peptidoglycan/flagellin receptor genes: CERK1 [At3g21630], CEBiP/LYM2 [At2g17120], LYM1 [At1g21880], LYM3 [At1g77630], FLS2 [At5g46330], BAK1 [At4g32430]), and cell wall modification-related genes: cellulose synthase/polygalacturonase inhibitor (PGIP)/pectin methyl esterase (IRX1/CESA8/LEW2 [At4g18780], IRX3/CESA7 [At5g17420], IRX5/CESA4 [At5g44030], PGIP1 [At5g06860], PGIP2 [At5g06870], ATPMEPCRA [At1g11580]), respectively. Differential expression of these genes between d-tagatose-treated and mocked samples was summarized in Supplementary Fig. 4a, and the values of log2 ratio shown in Supplementary Fig. 4a were visualized using graduated color bars (Fig. 3c). All genes up- or downregulated more than 10 times from either A. thaliana or H. arabidopsidis isolate Noco2 were also listed with the annotation in Supplementary Fig. 4b.
Detection of monosaccharide kinase activity in crude enzyme extracts from germinated conidia of Hyaloperonospora arabidopsidis isolate Noco2
Monosaccharide kinase activity of crude enzyme extracts from germinated conidia of H. arabidopsidis isolate Noco2 was determined. Germinated conidiospores (2 × 106 conidiospores) that had been grown for 24 h in Gamborg B5 medium salt mixture solution (Nihon Pharmaceutical) were ground with a mortar, pestle and zirconia beads (0.5 mm diameter) in 500 μl of extraction buffer (50 mM Tris-HCl buffer, pH 7.5, containing EDTA-free Complete protease inhibitor [Sigma]), then centrifuged at 13,000 rpm for 5 min at 4 °C. The supernatant was desalted using 50 mM Tris-HCl buffer (pH 7.5) and an Amicon Ultra-4 (10,000 MWCD; Millipore, Billerica, MA, USA), and the supernatant (=crude enzyme extracts) was mixed with 100 mM of d-tagatose in the reaction mixture (50 mM Tris-HCl pH 7.5, 10 mM MgCl2 and 6 mM ATP). The mixture was incubated for 24 h at 25 °C, and the reaction products were labelled using the p-aminobenzoic acid ethyl ester (ABEE) system (J-Chemical, Tokyo, Japan) and detected using HPLC as detail in the section on characterization of enzymes for d-tagatose phosphorylation using HPLC.
cDNA cloning and heterologous expression of various genes encoding sugar kinases
A cDNA library of H. arabidopsidis isolate Noco2 was constructed using total RNA (1 µg) extracted as described for the RNA-seq analyses and a SMARTer PCR cDNA Synthesis Kit (Takara) according to the instructions. cDNA fragments of the coding region from fructokinase (LC500344), glucokinase (LC500564), xylulose kinase (LC500562), or ribokinase (LC500561) of H. arabidopsidis isolate Noco2 were subcloned in frame into the pGEX5X-1 vector (GE Healthcare) and expressed in SoluBL21 Competent E. coli cells (Genlantis, San Diego, CA, USA) according to the manufacturer’s instructions. Primers used for cloning are listed in Supplementary Table 4. The recombinant proteins of each coding region, generated by the heterologous expression system with 0.4–0.6 mM Isopropyl β-d-1-thiogalactopyranoside (IPTG) for 3 h at 37 °C, were purified using a GSTrap HP column (GE Healthcare) and the manufacturer’s instructions. Images of the recombinant proteins separated using SDS-PAGE with a standard protocol16 are shown in Fig. 4b and Supplementary Fig. 5, and the prepared recombinant proteins were used for initial examination of their d-tagatose phosphorylation activities (Fig. 4c, Supplementary Fig. 5).
For enzymatic characterizations, the recombinant protein of fructokinase (LC500344) was generated using a mass culture system with 10 liter LB broth containing 100 µg ml−1 carbenicillin at 37 °C and 200 rpm for 6 h (preculture), and 25 °C and 200 rpm for 1 day with 0.6 mM IPTG in a Jar Fermenter TS-M15L (Takasugi MFG, Tokyo, Japan) and purified using a GSTrap HP column (GE Healthcare) as per the instructions and dialyzed against 5 mM Tris-HCl buffer (pH 7.5) for 4 h. The prepared recombinant protein was used to determine phosphorylation activity for a kinetics analysis (Fig. 4d, Supplementary Fig. 6a–d, j, k), optimal temperature (Supplementary Fig. 6e), thermal stability (Supplementary Fig. 6f), optimal pH (Supplementary Fig. 6g), pH stability (Supplementary Fig. 6h), and optimal cofactor (metal ion) (Supplementary Fig 6i).
Characterization of enzymes for d-tagatose phosphorylation using HPLC
The enzymatic reaction mixture (1 ml containing 100 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 6 mM ATP, 25 mM d-tagatose or d-fructose, and 100 ng target enzymes) was incubated for 24 h at 25 °C. For detecting kinase products using HPLC, ABEE labelling was performed using the method of Yasuno et al.55 with modifications16. Briefly, 10 µl of the reaction mixture or sugar standard (25 mM d-fructose, d-tagatose, d-fructose 6-phosphate, or d-tagatose 6-phosphate) was added to 40 μl of ABEE reagent solution with a borane–pyridine complex (from the kit) and heated at 80 °C. Chloroform and distilled water (200 μl each) were added, the mixture centrifuged at 3000 rpm for 5 min. The upper aqueous layer was used for HPLC analyses (Prominence; Shimadzu, Kyoto, Japan) using an Xbridge C18 column (4.6 mm ID × 250 mm; Waters, Milford, MA, USA) with a 50-min separation at a flow rate of 1.0 ml min–1 at 30 °C with a running solvent system of 0.2 mM of potassium borate buffer (pH 8.9)/acetonitrile (93/7), followed by a 20-min wash with 0.02% trifluoroacetic acid/acetonitrile (50/50) and equilibration for 15 min with the running solvent. The peaks were monitored with a fluorescence detector (RF-10A XL or RF-20A XL, Shimadzu) with emission of 360 nm and excitation of 305 nm or UV detector (UV-VIS Detector SPD-10AV, Shimadzu) at an absorbance of 305 nm.
Inhibition of monosaccharide phosphorylation activity of fructokinase by d-tagatose
Activity of fructokinase (10 µg) was determined spectrophotometrically at 340 nm at 25 °C for 5 min by coupling production of ADP to oxidation of NADH via pyruvate kinase (15 U ml−1) (Oriental Yeast, Tokyo, Japan) and lactate dehydrogenase (10 U ml−1) (Oriental Yeast) reactions as described by Miller and Raines56 with 0.5–2.5 mM d-fructose and 10–250 mM d-tagatose in 1 ml reaction mixture containing 80 mM Tris-HCl (pH 7.5), NADH (0.3 mM), phosphoenolpyruvate (0.8 mM), ATP (4 mM), and MgCl2 (8 mM). The same data (mean ± SD of four replications of the reaction) were used for calculations for the kinetic analyses including Dixon plot (Fig. 4d), Lineweaver–Burk plot (Supplementary Fig. 6j), and bar graph (Supplementary Fig. 6k).
Characterizations of enzymes in d-fructose 6-phosphate metabolic pathway and inhibition of activity by d-tagatose 6-phosphate
The cDNA fragment (described above) of the coding region from phosphomannose isomerase (LC500563) of H. arabidopsidis isolate Noco2 was subcloned in frame into the pColdI vector (Takara), and expressed in E. coli SHuffle Express Competent cells (New England BioLabs, MA, USA) according to the manufacturer’s instructions. Primers used for cloning are listed in Supplementary Table 4. The recombinant proteins of the coding region generated by the heterologous expression system were obtained using a mass culture system with 10 liters LB broth containing 1% (w/v) d-glucose and 100 µg ml−1 carbenicillin at 30 °C and 200 rpm for 6 h (preculture), 15 °C and 200 rpm for 1 h (cold shock induction), and 15 °C and 200 rpm for 3 days in Jar Fermenter TS-M15L and purified using a HisTrap HP column (GE Healthcare) as per the manufacturer’s instructions, then dialyzed against 10 mM Tris-HCl buffer (pH 7.4). The recombinant proteins, separated by SDS-PAGE using a standard protocol16, are shown in Fig. 5b.
A reaction mixture (100 µl) containing 100 mM Tris-HCl (pH 7.4), 10 mM MgCl2, 10 µg recombinant protein of phosphomannose isomerase (LC500563), and 2.5 mM d-fructose 6-phosphate (F6P) was incubated at 30 °C for 6 h (Fig. 5c), or reaction mixtures (100 µl) containing 100 mM Tris-HCl (pH 7.4), 10 mM MgCl2, 5 µg recombinant protein with 2.5 mM of F6P and d-tagatose 6-phosphate (T6P) (from 0 to 37.5 mM) were incubated at 30 °C for 1 h. Reaction products were labelled and purified for ABEE as described above, and the ABEE-labelled sugars (10 µl) were analyzed using an HPLC system (Prominence) with an Xbridge C18 column (4.6 mm ID × 250 mm) and 20-min separation at a flow rate of 1.0 ml min–1 at 30 °C with solvent A of the Glyscope solvent set (J-Chemical), followed by a 5-min wash with solvent B from the solvent set and equilibration for 10 min with the running solvent. The peaks were monitored with the fluorescence detector (RF-20A XL) with emission of 360 nm and excitation of 305 nm. Respective data in Fig. 5d are means ± SD of three replications calculated from the peak area on the HPLC spectrum with a calibration curve using standards with known concentrations, and they were compared for significant differences using a Tukey–Kramer multiple comparison test (p < 0.05) in JMP 12.
Statistics and reproducibility
All data in this work were collected in multiple and respective data are means ± SD of replications and statistical analyses shown in figures were performed using a Tukey–Kramer multiple comparison test (p < 0.05) in JMP 12 program (SAS Institute) unless noted otherwise as described above.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.