Organization and control of the ascorbate biosynthesis ......Jun 25, 2020 · 114 the cytosol while L-galactono-1,4-lactone crosses the outer membrane of the mitochondria to 115 be

Short title: Ascorbate biosynthesis in plants 1

2

Author for Contact Details: 3

Miguel A. Botella, Instituto de Hortofruticultura Subtropical y Mediterranea "La Mayora" 4

(IHSM-UMA-CSIC), Departamento de Biología Molecular y Bioquímica, Facultad de Ciencias, 5

Universidad de Malaga, Campus de Teatinos s/n, E-29071 Malaga, Spain. 6

Nicholas Smirnoff, Biosciences, College of Life and Environmental Sciences, 7

University of Exeter, Exeter, EX4 4QD, UK. 8

9

Article title: Organization and control of the ascorbate biosynthesis pathway in plants 10

11

Authors: Mario Fenech1, Vítor Amorim-Silva1, Alicia Esteban del Valle1, Dominique Arnaud2, 12

Araceli G. Castillo3, Nicholas Smirnoff2*, Miguel A. Botella1*. 13

14

Mario Fenech, [email protected], ORCID 0000-0001-9879-3456. 15

Vítor Amorim-Silva, [email protected], ORCID 0000-0002-3978-7205. 16

Alicia Esteban del Valle, [email protected], ORCID 0000-0003-3039-1172. 17

Dominique Arnaud, [email protected], ORCID 0000-0003-4898-1656. 18

Araceli G. Castillo, [email protected], ORCID 0000-0003-3990-5475. 19

Nicholas Smirnoff, [email protected], ORCID 0000-0001-5630-5602. 20

Miguel A. Botella, [email protected], ORCID 0000-0002-8867-1831. 21

22

Author affiliations: 23 1 Instituto de Hortofruticultura Subtropical y Mediterranea "La Mayora" (IHSM-UMA-CSIC), 24

Departamento de Biología Molecular y Bioquímica, Facultad de Ciencias, Universidad de 25

Malaga, Campus de Teatinos s/n, E-29071 Malaga, Spain. 26 2 Biosciences, College of Life and Environmental Sciences, University of Exeter, Exeter, EX4 27

4QD, UK 28 3 Instituto de Hortofruticultura Subtropical y Mediterranea "La Mayora" (IHSM-UMA-CSIC), 29

Departamento de Genética, Facultad de Ciencias, Universidad de Malaga, Campus de 30

Teatinos s/n, E-29071 Malaga, Spain. 31

* Author for Contact. 32

33

One sentence summary: 34

Metabolic engineering, genetic analysis and functional mutant complementation identify GDP-35

L-galactose phosphorylase as the main control point in ascorbate biosynthesis in green 36

tissues. 37

38

Author contributions: 39

M.F., V.A.S., M.A.B. and N.S. conceived the project and designed the experiments; M.F. 40

performed the experiments and analyzed the data; N.S. performed the MCA simulation; 41

V.A.S., M.A.B., and N.S. supervised the experiments; D.A. generated vtc2/VTC2-GFP 42

transgenic lines and tested mutant complementation by AO assay; A.G.C. performed yeast 43

two-hybrid assay with technical assistance by M.F.; A.E.V. provided technical assistance to 44

M.F.; M.F., M.A.B. and N.S. wrote the article; V.A.S., M.A.B. and N.S. supervised and 45

completed the writing; M.A.B. agrees to serve as the author responsible for contact and 46

ensures communication. 47

48

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted June 25, 2020. ; https://doi.org/10.1101/2020.06.23.167247doi: bioRxiv preprint

https://doi.org/10.1101/2020.06.23.167247

Funding information: 49

This work was supported by a grant from the Spanish Ministerio de Educación, Cultura y 50

Deporte para la formación del Profesorado Universitario (FPU014/01974), The Ministerio de 51

Economía, Industria y Competitividad (cofinanced by the European Regional Development 52

Fund; grant no. BIO2016-81957-REDT, BIO2017-82609-R) and by the Spanish "Ministerio de 53

Economia, Industria y Competitividad/FEDER" (AGL2016-75819-C2-1-R). NS was supported 54

by the Biotechnology and Biological Sciences Research Council (BBSRC) grants 55

BB/G021678/1 and BB/N001311/1. 56

57

Present addresses: 58

Dominique Arnaud current address: Laboratory of Molecular Plant Physiology and Functional 59

Genomics and Proteomics of Plants, CEITEC-Central European Institute of Technology, 60

Masaryk University, Kamenice 5, CZ-625 00, Brno, Czech Republic. 61

62

Email address of Author for Contact 63

Miguel A. Botella, [email protected] 64

Nicholas Smirnoff, [email protected] 65

66

ABSTRACT 67

68

The enzymatic steps involved in L-ascorbate biosynthesis in photosynthetic organisms (the 69

Smirnoff-Wheeler, SW pathway) has been well established and here we comprehensively 70

analyze the subcellular localization, potential physical interactions of SW pathway enzymes 71

and assess their role in control of ascorbate synthesis. Transient expression of GFP-fusions 72

in Nicotiana benthamiana and Arabidopsis (Arabidopsis thaliana) mutants complemented with 73

genomic constructs showed that while GME is cytosolic, VTC1, VTC2, VTC4, and L-GalDH 74

have cytosolic and nuclear localization. While transgenic lines GME-GFP, VTC4-GFP and L-75

GalDH-GFP driven by their endogenous promoters accumulated the fusion proteins, the 76

functional VTC2-GFP protein is detected at low level using immunoblot in a complemented 77

vtc2 null mutant. This low amount of VTC2 protein and the extensive analyses using multiple 78

combinations of SW enzymes in N. benthamiana supported the role of VTC2 as the main 79

control point of the pathway on ascorbate biosynthesis. Interaction analysis of SW enzymes 80

using yeast two hybrid did not detect the formation of heterodimers, although VTC1, GME and 81

VTC4 formed homodimers. Further coimmunoprecipitation (CoIP) analysis indicted that 82

consecutive SW enzymes, as well as the first and last enzymes (VTC1 and L-GalDH), 83

associate thereby adding a new layer of complexity to ascorbate biosynthesis. Finally, 84

metabolic control analysis incorporating known kinetic characteristics, showed that previously 85

reported feedback repression at the VTC2 step confers a high flux control coefficient and 86

rationalizes why manipulation of other enzymes has little effect on ascorbate concentration. 87

88

INTRODUCTION 89 90

L-Ascorbate (ascorbate, vitamin C) is the most abundant soluble antioxidant in plants 91

and functions as a cofactor in many enzymatic reactions (Fenech et al., 2019; Smirnoff, 2018). 92

Some groups of animals, among which humans are included, lost the ability to biosynthesize 93

ascorbate (Smirnoff, 2018). Therefore, we need to incorporate this essential nutrient through 94

the diet, with fruits and vegetables being the major source. In plants, ascorbate is synthesized 95

from the hexose phosphate pool via D-mannose/L-galactose through the Smirnoff-Wheeler 96


https://doi.org/10.1101/2020.06.23.167247

(SW) pathway (Wheeler et al., 1998; Fig. 1). The pathway is supported by abundant 97

biochemical and genetic studies (Bulley and Laing, 2016; Wheeler et al., 2015). D-Glucose 6-98

P is sequentially transformed into D-fructose 6-P, D-mannose 6-P and D-mannose 1-P by 99

phosphoglucose isomerase (PGI), phosphomannose isomerase (PMI) and 100

phosphomannomutase (PMM). GDP-D-mannose pyrophosphorylase (VTC1; GMP) then 101

converts D-mannose 1-P into GDP-D-mannose, which is subsequentially epimerized by GDP-102

D-mannose epimerase (GME) into GDP-L-galactose. GDP-D-mannose and GDP-L-galactose 103

are additionally used in synthesis of cell wall polysaccharides and protein glycosylation 104

thereby interconnecting the SW pathway with cell wall biosynthesis (Conklin et al., 1999; 105

Fenech et al., 2019; Reiter and Vanzin, 2001). The next step is catalyzed by GDP-L-galactose 106

phosphorylase (GGP), producing L-galactose 1-P. This step is proposed to be the main control 107

point of the pathway (Dowdle et al., 2007; Laing et al., 2015, 2007; Linster et al., 2007; 108

Yoshimura et al., 2014). In Arabidopsis thaliana (Arabidopsis), GGP activity is encoded by two 109

paralogues, VTC2 and VTC5. Next, L-galactose 1-phosphate phosphatase (VTC4, GPP) and 110

other unidentified phosphatases (Conklin et al., 2006; Torabinejad et al., 2009), transform L-111

galactose 1-phosphate into L-galactose. L-Galactose is then oxidized to L-galactono-1,4-112

lactone by a L-galactose dehydrogenase (L-GalDH). All these steps are believed to occur in 113

the cytosol while L-galactono-1,4-lactone crosses the outer membrane of the mitochondria to 114

be oxidized by L-galactono-1,4-lactone dehydrogenase (L-GalLDH) which is an integral 115

component of mitochondrial electron transport Complex 1 on the inner membrane (Pineau et 116

al., 2008; Schimmeyer et al., 2016). 117

Ascorbate concentration is dependent on the tissue (Franceschi and Tarlyn, 2002; 118

Lorence et al., 2004; Müller-Moulé, 2008; Zhang et al., 2011) and the ascorbate pool adjusts 119

to changes in environmental conditions, particularly light (Bartoli, 2006; Page et al., 2012; 120

Plumb et al., 2018). Surprisingly, despite being the most abundant soluble antioxidant, a good 121

understanding of the mechanisms that control ascorbate concentration remains limited. While 122

mutant analyses of genes involved in biosynthetic pathway have corroborated the involvement 123

of the various enzymes (Table 1), they provide little information on how the pathway is 124

controlled. This is further complicated because the ascorbate pool is determined by the 125

balance between synthesis, oxidation, recycling and breakdown. The oxidation product of 126

ascorbate is the monodehydroascorbate radical. This is either directly reduced back to 127

ascorbate or can give rise to dehydroascorbate (DHA) in its hydrated bicyclic hemiketal form, 128

which is reduced back to ascorbate through the Halliwell-Foyer-Asada cycle (Asada, 1999). 129

In addition, a proportion of ascorbate and DHA can be degraded to various products such as 130

oxalate, threonate, and tartrate (Debolt et al., 2007; DeBolt et al., 2006; Dewhirst et al., 2017; 131

Green and Fry, 2005; Pallanca and Smirnoff, 2000; Terai et al., 2020; Truffault et al., 2017). 132

These recycling and breakdown reactions are evident under severe oxidative stress, which 133

results in decreased cellular ascorbate (Terai et al., 2020; Waszczak et al., 2018). 134

Multiple lines of evidence point to VTC2 as the critical controlling step of ascorbate 135

biosynthesis. The transcription of VTC2 gene (Dowdle et al., 2007; Müller-Moulé, 2008; Urzica 136

et al., 2012) and the activity of the encoded enzyme (Dowdle et al., 2007)are highly responsive 137

to environmental factors. Furthermore, VTC2 translation is subject to feedback repression via 138

a conserved upstream open reading frame (uORF) in the 5’-UTR (Laing et al., 2015). As a 139

consequence, removing the uORF results in increased ascorbate in Arabidopsis (Laing et al., 140

2015), tomato (Li et al., 2018) and lettuce (Zhang et al., 2018). In addition, VTC2 is the only 141

gene of the pathway whose overexpression consistently increases the ascorbate content 142

(Bulley et al., 2012; Yoshimura et al., 2014) and QTL analysis shows that GGP paralogs 143

located within regions of the genome associated to fruits containing high ascorbate content 144


https://doi.org/10.1101/2020.06.23.167247

(Mellidou et al., 2012). There have been reports that overexpression of other genes of the 145

pathway such as VTC1 or GME also increase ascorbate concentration in Arabidopsis (Zhou 146

et al., 2012 [1.3-fold]; Li et al., 2016 [1.5-fold]), rice (Zhang et al., 2015 [1.4-fold]), tobacco 147

(Wang et al., 2011 [2-4-fold]), tomato (Cronje et al., 2012 [1.7-fold]; Zhang et al., 2011 [1.4-148

fold]) and kiwifruit (Bulley et al., 2009 [1.2-fold]). However, overexpression of these genes 149

does not always have an effect on ascorbate concentration (Sawake et al., 2015; Yoshimura 150

et al., 2014). In addition, various studies have reported ascorbate feedback inhibition of the 151

biosynthetic enzymes PMI (Maruta et al., 2008), GME (Wolucka and Van Montagu, 2003) and 152

L-GalDH (Mieda et al., 2004). This feedback control is further supported by the decreased 153

incorporation of labelled sugars into ascorbate in vivo after feeding ascorbate (Pallanca and 154

Smirnoff, 2000; Wolucka and Van Montagu, 2003). 155

In order to increase our understanding of the organization of the ascorbate 156

biosynthesis pathway, we have performed a systematic over-expression of single and multiple 157

combinations of all the enzymes of the pathway from VTC1 downstream in Nicotiana 158

benthamiana. This allowed the systematic investigation of their effects on the ascorbate pool 159

and supported the formation of a multiprotein complex. In addition, complementation of the 160

Arabidopsis mutants with the respective GFP-tagged enzymes confirmed the functionality of 161

the constructs and provided information on their subcellular localizations. Finally, we 162

constructed a kinetic model based on the known properties of the pathway whose predictions 163

support that activity of VTC2 is the only significant controlling step and which also explains 164

other observed features of the pathway. In addition to this information, this work provides a 165

set of tools that will allow a better understanding of ascorbate biosynthesis at the molecular 166

and cellular level. 167

168

RESULTS 169

170

Generation and analyses of GFP-tagged ascorbate biosynthesis Arabidopsis 171

complementation lines reveal major differences in protein expression along the 172

pathway 173

174

In order to study the function of SW proteins we aimed to generate Arabidopsis lines 175

transformed with genomic regions (including their promoters) of ascorbate biosynthesis genes 176

that resulted in proteins fused to GFP at their C-terminus. To investigate the functionality of 177

these constructs, they were introduced into Arabidopsis mutants and the complementation of 178

their phenotypes was analyzed. Arabidopsis mutants in SW genes typically show lower 179

ascorbate concentration and/or lethality in early stages of development (Table 1). Selfing of 180

Arabidopsis heterozygous mutants for GME (SALK_150208, gme-3, hereafter gme; Supp. Fig. 181

1A) and L-GalDH (SALK_056664, Supp. Fig. 1B) did not produce homozygous mutants (Supp. 182

Fig. 2). Since exogenous supplementation of ascorbate prevented growth arrest of several 183

SW mutants (Dowdle et al., 2007; Pineau et al., 2008) we germinated seeds from 184

heterozygous gme-3 and gldh plants on solid media without and supplemented with 0.5 mM 185

ascorbate (Fig. 2A). As a control we included seeds from a heterozygous gldh mutant plant, 186

whose growth arrest is prevented by exogenous ascorbate (Pineau et al., 2008). The analysis 187

of heterozygous gme progeny identified 40 wild-type (WT) seedlings, 46 heterozygous 188

seedlings and no homozygous for the T-DNA, confirming the lethality of a loss-of-function 189

allele of GME (Qi et al., 2017; Voxeur et al., 2011) (Fig. 2A; Supp. Fig. 2A). Similarly, no 190

homozygous gme seedlings were identified when the growth medium was supplemented with 191

0.5 mM ascorbate in contrast to gldh (Pineau et al., 2008; Supp. Fig. 2B), confirming that GME, 192


https://doi.org/10.1101/2020.06.23.167247

similarly to VTC1 (Lukowitz et al., 2001), has additional roles to that of ascorbate biosynthesis 193

(Supp. Fig. 2A; Gilbert et al., 2009; Qi et al., 2017; Voxeur et al., 2011). On the other hand, 194

diagnostic PCR of heterozygous lgaldh progeny identified 23 WT seedlings and 69 195

heterozygous T-DNA seedlings while no homozygous mutants were found in the medium not 196

supplemented with ascorbate. Remarkably, a number of seedlings showed growth arrest and 197

yellowing similarly to gldh progeny (Fig. 2A, magnified square). In contrast, we identified 25 198

WT, 41 heterozygous T-DNA and 25 homozygous T-DNA seedlings by diagnostic PCR in 199

ascorbate containing medium (Fig. 2A, Supp. Fig. 2B). Therefore, and in contrast to gme 200

mutants, Igaldh mutants could be rescued by supplementation with exogenous ascorbate (Fig. 201

2A, Supp. Fig. 2B). This supports that the expression product of L-GalDH is responsible for 202

the L-galactose dehydrogenase activity present in Arabidopsis. Additionally, the finding that, 203

like gldh, Igaldh is an ascorbate auxotroph also indicates that the activity of this enzyme is 204

essential for ascorbate biosynthesis. 205

In order to investigate the functionality of a GME protein fused with a GFP at the C-206

terminus, we transformed WT plants with a GMEp:GME-GFP construct. Of the several 207

transformant that showed high expression of GME-GFP and a single insertion (Supp. Fig. 3A) 208

line 6 (L6) was selected to introduce the transgene into the gme mutant background using 209

reciprocal crosses. Using heterozygous gme plants as male parent, we did not obtain F1 210

gme/GMEp:GME-GFP (hereafter named gme/GME-GFP L6) descendants, supporting the 211

role of GME in pollen development and pollen tube elongation (Qi et al., 2017). However, when 212

line L6 was used as a male parent we obtained viable gme/GME-GFP F1 seedlings that 213

allowed the identification in the F2 of homozygous gme/GME-GFP L6 seedlings (Supp. Fig. 214

3B). Surprisingly, while the GME-GFP fusion protein complement the sterility defects, it 215

contained 55% of the ascorbate content of WT plants (Fig. 2B) which, considering the lower 216

ascorbate content of knock-down gme Arabidopsis mutants (Qi et al., 2017), indicates that it 217

complements only partially the ascorbate content. Similarly, a L-GalDHp:L-GalDH-GFP 218

construct was transformed into WT plants and selected line 1 (L1) that showed high 219

expression and a single insertion (Supp. Fig. 3C). Then, reciprocal crosses between 220

heterozygous lgaldh plants and L-GalDHp:L-GalDH-GFP L1 were performed. In this occasion, 221

we did obtain F1 plants when lgaldh plants were used as male or female parent. Homozygous 222

lgaldh/GalDH-GFP L1 seedlings were identified in F2 progeny using diagnostic PCR in 223

medium lacking ascorbate (Supp. Fig. 3D), indicating that the L-GalDH-GFP protein driven by 224

the L-GalDH promoter restores the enzyme activity lost in lgaldh mutant. In addition, ascorbate 225

content of lgaldh/L-GalDH-GFP L1 plants contained WT levels of ascorbate (Fig. 2B). 226

In contrast to gme and lgaldh, loss of function vtc2 and vtc4 homozygous mutants are 227

viable and show WT growth at standard conditions despite presenting a 20% and 55% of WT 228

ascorbate content, respectively (Fig. 2B; Conklin et al., 2000; Dowdle et al., 2007; Torabinejad 229

et al., 2009). In the case of vtc2 mutant, GDP-L-galactose phosphorylase activity is also 230

provided by the paralogous VTC5 gene. Consistently, vtc2vtc5 double mutants, like lgaldh and 231

gldh mutants (Fig. 2A; Supp. Fig. 2B, C), show growth arrest that can be complemented with 232

exogenous ascorbate (Dowdle et al., 2007; Lim et al., 2016). However, considering the 233

phenotypes described (Table 1) the viability of vtc4 must rely on other enzymes with L-234

galactose 1-phosphate phosphatase activity (Nourbakhsh et al., 2015; Torabinejad et al., 235

2009). In order to test the functionality of GFP-tagged VTC2 and VTC4, we relied on restoring 236

WT ascorbate content by introducing VTC2p:VTC2-GFP and VTC4p:VTC4-GFP constructs 237

into their respective mutants. T-DNA knock-out mutant vtc2-4 (SAIL_769_H05; hereafter 238

named vtc2; Supp. Fig. 1C; Lim et al., 2016) was directly transformed and lines with single 239

insertion were selected (Supp. Fig. 3E). Among these lines, line 13 (L13) was selected 240


https://doi.org/10.1101/2020.06.23.167247

because it showed the highest accumulation of VTC2-GFP (Supp. Fig. 3F). Also, 241

VTC4p:VTC4-GFP was introduced into vtc4-4 mutant (Supp. Fig 2D) following the same 242

strategy described for GME and L-GalDH (Supp. Fig. 3G). A single insertion VTC4p: VTC4-243

GFP line 5 was selected to cross with vtc4 homozygous mutant and the homozygous 244

vtc4/VTC4-GFP line was isolated by diagnostic PCR (Supp. Fig. 3H). Ascorbate concentration 245

of vtc2/VTC2-GFP and vtc4/VTC4-GFP lines was restored to WT levels validating the 246

functionally of the GFP-tagged enzyme (Fig. 2B). 247

We analyzed the protein levels of the different lines generated in a single blot in order 248

to compare the amount of proteins accumulated in each line. Immunoblot analysis indicated 249

that GME-GFP and VTC4-GFP were easily detectable while L-GalDH-GFP showed reduced 250

levels of proteins (Fig. 3C). Interestingly, despite the complementation of ascorbate levels in 251

L13, VTC2-GFP protein was not detected (Fig. 2C, upper panel) and was detected only after 252

overexposure of the blot using a high sensitivity ECL substrate for the detection of femtogram 253

amounts of protein. This low amount of VTC2 protein cannot be explained by a low expression 254

of VTC2 at the transcriptomic level because, based on public available transcriptomic 255

databases, VTC2 and GME show the highest expression of all SW genes in vegetative tissue 256

(Supp. Fig. 4). 257

258

Combinatorial expression of ascorbate biosynthesis genes in N. benthamiana leaves 259

identifies VTC2 as the main control point of ascorbate biosynthesis 260

261

We have shown that that SW proteins are functional when a tag is added to their C-262

terminus and that a small amount of VTC2-GFP protein is required to increase the content of 263

vtc2 ascorbate to WT levels. In order to investigate the effect of the activity of ascorbate 264

biosynthesis genes on ascorbate concentration we cloned the coding sequences (CDS) of 265

Arabidopsis VTC1, GME, VTC2, VTC4, L-GalDH and L-GalLDH with C-terminal GFP and HA, 266

driven by the CaMV35S promoter (Fig. 3A) and transiently expressed them in N. benthamiana 267

leaves. Immunoblot analysis of the SW GFP tagged proteins showed that the molecular 268

masses of the different fusion proteins were consistent with their predicted sizes (Fig. 3B). 269

Moreover, all the fusion proteins showed little degradation of the proteins at all time points 270

tested (Supp. Fig. 5). Three days after agroinfiltration, VTC1-GFP, GME-GFP, VTC2-GFP and 271

VTC4-GFP proteins showed the highest expression while L-GalDH-GFP and L-GalLDH-GFP 272

were also well expressed (Fig. 3B). 273

Then, we performed a systematic analysis of the effect of transient overexpression of 274

individual or multiple combinations of these SW Arabidopsis genes on the ascorbate 275

concentration in N. benthamiana leaves. Based on our previous data (Fig. 3B), we analyzed 276

the ascorbate content 3 days after agroinfiltration. All genes were expressed either individually 277

or in pairs (Fig. 3C-G; Supp. Table 1), and due to the large number of possible combinations, 278

three genes were expressed individually and two genes out of the three were co-expressed in 279

order to analyze the ascorbate content in the same experiment (Fig. 3C-G). This experimental 280

design validated the reproducibility of our experimental system since individual genes have 281

been analyzed three times with their proper controls (except for VTC1, with only two). As 282

shown in Fig. 3C, the transient expression of individual VTC1-GFP and GME-GFP genes did 283

not increase ascorbate content relative to the naïve (not agroinfiltrated) or control (infiltrated 284

with GFP) N. benthamiana leaves. However, expression of VTC2-GFP increased ascorbate 285

concentration between 2 and 3-fold. The expression of individual VTC1, GME and L-GalDH 286

genes or any combination of these three genes without VTC2-GFP did not significantly 287

increase ascorbate concentration in any of the multiple assays (Fig. 3D-G). It also 288


https://doi.org/10.1101/2020.06.23.167247

demonstrates that higher levels of VTC1 (Fig. 3C), VTC4 (Fig. 3G) and L-GalDH (Fig. 3F) do 289

not contribute to increase the ascorbate content, either individually or combined. The 290

combination of VTC2 with GME caused a significant increase in ascorbate relative to VTC2 291

(Fig. 3E), although the effect was not consistent in all experiments (Fig. 3E, H). An increased 292

ascorbate content by combining GME and VTC2 has been previously reported using transient 293

overexpression of kiwifruit GME and VTC2 homologs in tobacco leaves (Bulley et al., 2009). 294

Finally, we tested the effect of the simultaneous transient overexpression of all SW 295

cytosolic Arabidopsis genes on the ascorbate concentration in N. benthamiana leaves. The 296

expression of all cytosolic components was confirmed by immunoblot analysis (Fig. 3H) and 297

increased the ascorbate content to similar levels as produced by expressing VTC2 alone (Fig. 298

3H). Expressing all the cytosolic enzymes together with the final mitochondrial enzyme L-299

GalLDH also did not increase ascorbate concentration (Fig. 3H). Overall, these results point 300

to VTC2 as the main control point in the biosynthesis of ascorbate. 301

302

Subcellular localization of ascorbate biosynthesis enzymes using GFP-tagged proteins 303

304

The biosynthesis of ascorbate up to the L-galactose dehydrogenase step is proposed 305

to take place in the cytosol with the exception of the last step that occurs at the intermembrane 306

space of the mitochondria (Dunkley et al., 2006; Heazlewood et al., 2004; Østergaard et al., 307

1997; Pineau et al., 2008; Schimmeyer et al., 2016). In silico analysis using the software 308

Compartments (https://compartments.jensenlab.org/Search) predicted a cytosolic subcellular 309

localization for GME, VTC4 and L-GalDH, a dual cytosolic-nuclear localization of VTC1 and 310

VTC2 and a mitochondrial localization of L-GalLDH (Supp. Fig. 6). Confocal microscopy 311

imaging of C-terminal GFP-tagged proteins transiently expressed in N. benthamiana revealed 312

a cytosolic subcellular localization for VTC1, GME, VTC2, VTC4 and L-GalDH while L-GalLDH-313

GFP appeared in structures that resembled mitochondria (Fig. 4). Interestingly, VTC1-GFP, 314

VTC2-GFP, VTC4-GFP and L-GalDH-GFP also showed a nuclear localization in addition to 315

their cytoplasmic localization, whereas we observed a GME-GFP signal surrounding the nuclei 316

(Fig. 4). Evidence for a dual cytosolic-nuclear localization has been provided for VTC1 and 317

VTC2 using stable transgenic Arabidopsis lines (Müller-Moulé, 2008; Wang et al., 2013), 318

suggesting a potential nuclear function of these proteins. Remarkably, similarly to VTC1, 319

nuclear localization signals are not predicted for VTC4 or L-GalDH and, therefore, the 320

localization reported here was unexpected (Supp. Table 2). 321

Because overexpression can result in ectopic localization of proteins, we used the 322

Arabidopsis transgenic lines that carry functional biosynthetic ascorbate proteins C-terminally 323

fused to GFP and driven by their own promoters to analyze their in vivo subcellular localization 324

as well as their tissue distribution. Cotyledons, cotyledon-hypocotyl junction, hypocotyl and 325

root tip of four-day-old Arabidopsis seedlings were analyzed. GME-GFP, VTC4-GFP and L-326

GalDH-GFP lines showed a strong fluorescence signal in the cytoplasm (Fig. 5A). In addition, 327

VTC4-GFP and L-GalDH-GFP were present within nuclei-like structures. In roots, these two 328

proteins were present within nucleus as a dot that resembles nucleolus, but they also 329

displayed a ring-like pattern in nucleus whereas GME-GFP showed an even distribution 330

across cytosol (Fig. 5A). We did not detect fluorescence signal for VTC2-GFP above the 331

background level (Fig. 5B) consistent with the weak signal previously observed in immunoblot 332

analysis (Fig. 2H). Except for VTC2, these data are consistent with the over-expression data 333

of SW-GFP constructs in N. benthamiana. 334

335


https://doi.org/10.1101/2020.06.23.167247

Protein-protein interaction assays support a physical association of proteins involved 336

in ascorbate biosynthesis 337

338

Enzymes involved in metabolic pathways can associate in order to increase the 339

efficiency of consecutive enzymatic transformations (Smirnoff, 2019; Sweetlove and Fernie, 340

2018). Since VTC1, GME, VTC2, VTC4 and L-GalDH proteins showed cytoplasmic 341

localization (Fig. 4, Fig. 5) we investigated their interaction in vivo using yeast two-hybrid (Y2H) 342

(Fig. 6A; Supp. Fig. 7). For this, all ascorbate biosynthesis CDS (including the mitochondria-343

localized L-GalLDH as a negative control) were cloned into both binding domain (BD, pGBKT7) 344

and activation domain (AD, pGADT7) yeast vector and tested all 36 possible interactions 345

resulting from the combination of all the enzymes of the pathway in both directions (fused to 346

AD or to BD; Supp. Fig. 7). As positive controls, we included SNF1/SNF2 (Fields and Song, 347

1989) and p53/SV40-Large AgT (Iwabuchi et al., 1993; Li and Fields, 1993) interactions, 348

whereas Laminin C/SV40-Large AgT was used as a negative control (Bartel et al., 1993; Ye 349

and Worman, 1995). This analysis did not provide evidence of direct interactions between the 350

enzymes of the pathway, although VTC1, GME and VTC4 showed self-interaction, showing 351

that these proteins were expressed in yeast (Fig. 6A). A dimerization interface is predicted for 352

GME and VTC4, while a trimerization domain is predicted for VTC1 (Fig. 6B), consistent with 353

our yeast two-hybrid interaction results. 354

Despite the apparent lack of direct interactions of ascorbate biosynthesis enzymes in 355

yeast, we investigated whether these proteins could associate in vivo. For this, we co-356

expressed biosynthesis enzymes of the pathway tagged with C-terminal GFP and HA in N. 357

benthamiana and performed coimmunoprecipitation (CoIP) assays. We used GFP-Trap beads 358

to inmunoprecipitate GFP tagged proteins and α-HA antibodies to detect co-359

immunoprecipitated HA-fused proteins using immunoblot. We first tested biosynthesis 360

proteins that are consecutive in the pathway (Fig. 6C-F), i.e., VTC1/GME, GME/VTC2, 361

VTC2/VTC4, VTC4/L-GalDH. We co-expressed VTC1-GFP and GME-HA to test the 362

interaction, and GFP together with GME-HA as a negative control (Fig. 6C) and checked by 363

immunoblot if protein expression was suitable for immunoprecipitation prior to the assay 364

(Supp. Fig. 8). After immunoprecipitation of VTC1-GFP and free GFP, we detected a specific 365

association between GME-HA with VTC1-GFP but not with free GFP (Fig. 6C). Following the 366

same strategy, we found associations between all consecutive enzymes of the pathway (Fig. 367

6D-F). In addition, we also performed CoIP between the first and last cytosolic enzymes of the 368

pathway, i.e., VTC1 and L-GalDH and found a specific association between VTC1-GFP and 369

L-GalDH-HA (Fig. 6G). 370

371

A kinetic model of ascorbate biosynthesis indicates GDP-L-galactose phosphorylase 372

as the main control point 373

374

To further understand the results of over-expression, a kinetic model of ascorbate 375

biosynthesis was constructed using Complex Pathway Simulator (COPASI; Hoops et al., 376

2006). Reactions were described by reversible Michaelis-Menten kinetics, except for L-Gal 1-377

P phosphatase/VTC4 and L-GalLDH which were described by irreversible Michaelis-Menten 378

kinetics (Fig. 7A). Published Km values were used except for GGP, which was set to 0.1 mM 379

in the absence of published data (Gatzek et al., 2002; Hoeberichts et al., 2008; W. A. Laing et 380

al., 2004; Linster et al., 2007; Maruta et al., 2008; Østergaard et al., 1997; Wolucka and Van 381

Montagu, 2003). The relative activity of enzymes has been reported in only one case (Dowdle 382

et al., 2007) but these are not guaranteed to reflect the maximum rate of each enzyme. Low 383


https://doi.org/10.1101/2020.06.23.167247

expression of GGP compared to other pathway enzymes is suggested from very low 384

expression of the GFP fusion which nevertheless rescues ascorbate concentration in the 385

mutant background (Fig. 3D, H). However, conservatively, the relative activity of each enzyme 386

(V) was set to the same value (10 mM s-1) but was varied during simulations. Evidence from 387

radiolabeling suggests that ascorbate synthesis is subject to feedback inhibition by ascorbate 388

(Conklin et al., 1997; Laing et al., 2015; Pallanca and Smirnoff, 2000; Wolucka and Van 389

Montagu, 2003). The proposed uORF-mediated feedback repression of GGP translation 390

(Laing et al., 2015) was modelled by including non-competitive inhibition by ascorbate. 391

Competitive inhibition of PMI and L-GalDH by ascorbate (Maruta et al., 2008; Mieda et al., 392

2004) were included, although in the case of L-GalDH, the inhibition could be an assay artefact 393

(William A Laing et al., 2004). Once synthesized, ascorbate and its oxidation product 394

dehydroascorbate (DHA) can be broken down to a wide range of products (Dewhirst et al., 395

2020; Dewhirst and Fry, 2018; Green and Fry, 2005). The oxidation of ascorbate to 396

monodehydroascorbate (MDHA) and DHA was included in simplified form including only 397

ascorbate and DHA. The breakdown rate of DHA was described by irreversible 1st order 398

kinetics (Conklin et al., 1997; Pallanca and Smirnoff, 2000). It is assumed in the model that all 399

breakdown products are recycled to D-Glc 6-P. This is unlikely to be correct but did not affect 400

model predictions. 401

The starting parameters used in the model are shown in Supplementary Table 3. The 402

model carried out metabolic control analysis (MCA) at steady state. The effect of changing the 403

strength of feedback repression of GGP by ascorbate on the flux control coefficients for each 404

step are shown in Figure 7B (and data in supplemental Table 4). At the highest Ki value 405

(relatively weak feedback), all steps up to GGP contributed equally to flux, indicated by similar 406

flux control coefficients. However, biosynthesis steps beyond GGP had very low flux control 407

coefficients and changes in activity of these enzymes would be predicted to have a very small 408

effect on pathway flux. The turnover step also had a significant effect on flux. Decreasing Ki 409

by ten-fold or more resulted in all flux control being shared between GGP and turnover, the 410

pre-GGP steps having very low flux control coefficients. Therefore, for the biosynthesis steps, 411

the model predicts that only GGP controls the flux through the pathway if feedback repression 412

is sufficiently strong. Correspondingly, concentration control coefficients (i.e. the change in the 413

concentration of each pathway intermediate caused by a change in enzyme activity) for each 414

step were calculated. The results for ascorbate are shown in Figure 7C and for all 415

intermediates in Supplementary Table 5. At the highest Ki value, all steps up to GGP had 416

similar concentration control coefficients indicating change in activity at each step would exert 417

a similar effect on ascorbate concentration. Increasing feedback strength decreased the 418

concentration control coefficients of all steps except GGP predicting that only this step controls 419

ascorbate concentration. The outcome could be affected by the starting conditions (relative 420

enzyme activities and kinetic constants) but testing the model with varying Km values or with 421

and without competitive feedback inhibition of PMI and L-GalDH had little effect on ascorbate 422

concentration (Supplementary Table 6). To simulate metabolic engineering experiments, the 423

enzyme activity of each step was increased over a large range and the predicted change in 424

ascorbate concentration at various feedback repression strengths was determined (Fig. 7D 425

and supplementary Table 7). As expected from the MCA, only varying GGP activity had a 426

significant effect on ascorbate over a very wide range of activities and this effect was strongly 427

dependent on the extent of feedback repression. In comparison, it was necessary to decrease 428

the activity of the other enzymes to very low values before ascorbate concentration was 429

affected and this behavior was particularly marked for the post-GGP steps. The kinetic model 430

therefore reflects the effect of transient expression of the pathway enzymes in N. 431


https://doi.org/10.1101/2020.06.23.167247

benthamiana: only increasing GGP activity increases ascorbate concentration. Simulation of 432

mannose 6-P addition caused a small transient increase in ascorbate accumulation, while L-433

galactose addition had a large effect (Fig. 7E). This differential response reflects the operation 434

of feedback repression at the GGP step. 435

436

DISCUSSION 437

Here, we report the effect of systematic overexpression, individually and in 438

combination, of all genes involved in the SW pathway have on the endogenous ascorbate 439

content using N. benthamiana leaves. The GFP-tagged enzymes were functional since 440

translational fusion of GFP at the C-termini of all enzymes increased ascorbate concentration 441

in their respective Arabidopsis mutants. Further, immunoblots of the expressed enzymes 442

showed that they were all well expressed in N. benthamiana (Supp. Fig. 8). Under these 443

experimental conditions, only VTC2/GGP over-expression was able to increase ascorbate 444

concentration, which confirms previous studies (Ali et al., 2019; Bulley et al., 2012, 2009; Laing 445

et al., 2015; Li et al., 2018; Yoshimura et al., 2014; Zhang et al., 2018). Interestingly, a vtc2 446

mutant containing a VTC2-GFP construct driven by its own promoter and including the 5’-UTR 447

showed an extremely low GFP fluorescence compared to other SW GFP-tagged proteins. This 448

low expression of VTC2-GFP was also reflected in the immunoblot comparing the expression 449

of all the components of the pathway at the protein level (Fig. 2E), likely reflecting the 450

translational repression of VTC2 via the conserved uORF in the 5’-UTR (Laing et al., 2015). 451

This is further confirmed by the finding that the overexpression in N. benthamiana of VTC2 452

without the uORF showed a similar amount of VTC2 proteins to other proteins of the pathway 453

(Supp. Fig. 8). The expected low level of endogenous VTC2 protein in N. benthamiana, which 454

also contains the uORF (Laing et al., 2015), may explain the marked increase of ascorbate 455

after the overexpression of Arabidopsis VTC2 lacking the uORF. 456

Based on the data presented here and other reports, we constructed a kinetic model 457

based on the known properties of the pathway enzymes that includes the repression of GGP 458

by ascorbate. However, it does not explicitly distinguish between transcriptional, translational, 459

post-translational and metabolite feedback on enzyme activity as the mechanism of 460

repression. This model was used for metabolic control analysis, an approach which 461

determines the distribution of the control of flux and concentration of pathway intermediates 462

in response to small changes in activity of each enzyme in the pathway by calculating flux 463

control and concentration control coefficients for each step (Fell, 1992). This analysis showed 464

that the only step in the biosynthesis pathway that controls pathway flux and ascorbate pool 465

size is GGP as long as feedback repression is included. The other critical step is ascorbate 466

breakdown while the flux control coefficients of all the other steps were negligible. Decreasing 467

the strength of feedback repression by ascorbate on GGP eventually decreased its flux and 468

concentration control coefficients, resulting in control being more evenly distributed across the 469

steps up to GGP. The model predicts that the steps post GGP exert little control and that in 470

the presence of feedback repression, GGP is the only critical step. Consequently, in virtual 471

metabolic engineering experiments, in which each enzyme was changed over a large range, 472

only GGP affected ascorbate concentration. 473

The model prediction is consistent with the transient expression results shown in this 474

report and with the relatively small effect of over-expressing enzymes other than GGP. In a 475

previous study of transient expression in Nicotiana tabacum, GME alone did not affect 476

ascorbate, in accordance with the model, but enhanced the effect of expressing GGP (Bulley 477

et al., 2009). However, in our hands this increase in ascorbate was not always reproducible 478

suggesting that other factors need to be considered such as different carbohydrate status 479


https://doi.org/10.1101/2020.06.23.167247

caused by growth conditions. Stable over-expression of GMP and GME has been reported to 480

increase ascorbate in tomato and N. tabacum (Badejo et al., 2008, 2007; Li et al., 2019) and 481

the model predicts that these enzymes could be effective if endogenous activity is low. 482

Similarly, increasing the expression of L-GalDH in N. tabacum had no effect on ascorbate 483

while antisense suppression in Arabidopsis only had a small effect in high light (Gatzek et al., 484

2002). 485

The model outcomes are robust over a range of starting conditions and are minimally affected 486

by the Km values tested nor by the presence/absence of competitive inhibition of PMI and L-487

GalDH. The model also explains another feature of the pathway, which is that VTC2 and 488

VTC5, the paralogues of GGP in Arabidopsis, do not compensate for each other when 489

knocked out (Dowdle et al., 2007). The control features of the pathway appear to emerge from 490

its architecture in which a strongly repressible GGP step is followed by the irreversible L-Gal 491

1-P phosphatase and L-galactonolactone dehydrogenase steps. 492

In a previous study comparing the activity of SW enzymes where ascorbate pool size differs 493

under low and high light intensity, only GGP showed increased activity in high light in 494

Arabidopsis (Dowdle et al., 2007). Transcript levels suggest that only VTC2 and VTC5 (and 495

sometimes GME) are responsive to environmental conditions (Li et al., 2013), and that there 496

is not a coordinated induction of SW genes under high light and low temperature, conditions 497

that increase ascorbate concentration (Laing et al., 2017). The control of ascorbate 498

biosynthesis therefore differs from pathways in which there is coordinated induction of multiple 499

pathway enzymes by transcription factors, for example anthocyanin synthesis (Zhang et al., 500

2014). 501

Importantly, the model shows that ascorbate pool size responds to GGP over a wide 502

range and thereby explains how a combination of transcription and feedback repression 503

controls the amount/activity of GGP and provide a mechanism to adjust ascorbate pool size 504

to prevailing conditions. Light is the best studied factor and ascorbate adjusts to the prevailing 505

light intensity over several days (Page et al., 2012). The mechanism likely involves a changed 506

balance between increased transcription of GGP via a possible photosynthesis-derived signal 507

and light response promoter elements (Gao et al., 2011; Yabuta et al., 2007) and ascorbate 508

repression of translation via the uORF. In addition to the transient expression assays showing 509

the role of the uORF in translational repression of GGP (Laing et al., 2015), CRISPR-Cas9 510

editing of the uORF in Arabidopsis, lettuce and tomato increases ascorbate, confirming its role 511

in repression of ascorbate synthesis in diverse species (Li et al., 2018; Zhang et al., 2018). 512

Faster control could also occur if ascorbate status more directly affects GGP activity via direct 513

inhibition or post-translational modification. These latter possibilities have not been 514

investigated but could operate since ascorbate addition relatively rapidly decreases 515

incorporation of 14C-glucose or mannose into ascorbate (Pallanca and Smirnoff, 2000; 516

Wolucka and Van Montagu, 2003). It is not known if ascorbate or a proxy of ascorbate 517

concentration exerts repression and, also, the mechanism by which the uORF controls GGP 518

translation is not known. 519

In addition to investigating the effect of enzyme overexpression on ascorbate 520

accumulation, we used tagged enzymes to investigate subcellular localization and the 521

formation of enzymatic complexes. With the exception of L-GalLDH, the cytosolic localization 522

of the SW enzymes is consistent with the in-silico prediction (Supp. Fig. 6). In this study, we 523

have shown that all SW proteins are present in the cytosol (Fig. 4; Fig. 5). Interestingly, in 524

addition to VTC1 and VTC2 that were previously reported to localize in the nucleus (Müller-525

Moulé, 2008; Wang et al., 2013) VTC4 and L-GalDH also showed nuclear localization (Fig. 4; 526

Fig. 5) despite the lack of nuclear localization signals (Supp. Table 2). Whether or not this 527


https://doi.org/10.1101/2020.06.23.167247

nuclear localization is functionally important remains to be determined but several metabolic 528

enzymes moonlighting in the nucleus have been previously reported (Boukouris et al., 2016). 529

Using CoIP in N. benthamiana we have shown that enzymes catalyzing consecutive steps in 530

the pathway associate (Fig. 6C-F) as well as proteins catalyzing distant enzymatic steps such 531

as VTC1 and L-GalDH (Fig. 6G). However, no direct interactions were found using yeast two-532

hybrid analysis. The lack of direct interaction but the positive in vivo associations may indicate 533

that the enzymes might attach to scaffold proteins so that the interaction would not be 534

detectable in the yeast two-hybrid assay. Although this deserves further investigation it is 535

tempting to speculate that some of the ascorbate biosynthesis enzymes might form a 536

functional enzymatic complex (metabolon) that could potentially channel pathway 537

intermediates to increase flux and perhaps aid partitioning of GDP-D-mannose and GDP-L-538

galactose between ascorbate and polysaccharide synthesis. Such channeling has been 539

shown for a number of pathways (Amorim-Silva et al., 2019; Smirnoff, 2019; Sweetlove and 540

Fernie, 2018; Zhang et al., 2017). 541

In conclusion, metabolic engineering experiments using the last six enzymes of the 542

SW pathway show that only GGP has significant control over the pathway, and we have 543

developed a kinetic model that provides a rationale for this result. It is clear that the balance 544

of VTC2/VTC5 amount is controlled by a combination of transcription and translation 545

repression, largely, but not entirely mediated by the uORF mechanism. In addition, we have 546

produced a set of complemented mutants from GME to L-GalDH tagged with GFP that 547

provides a resource for further investigation of the SW pathway. 548

549

Materials and Methods 550

551

Plant material 552

Arabidopsis thaliana (L.) Heynh wild-type (WT) and transgenic lines generated were in 553

Col-0 ecotype. Arabidopsis mutants vtc2-4 (SAIL_769_H05; Lim et al., 2016), vtc4-4 554

(SALK_077222; Torabinejad et al., 2009) and gldh (SALK_060087; Pineau et al., 2008) used 555

in this study have been previously described. However, to the best of our knowledge, this is 556

the first time that SALK_150208 (here named gme-3) and SALK_056664 (here named lgaldh-557

1), obtained from the The European Arabidopsis Stock Centre (NASC: 558

http://arabidopsis.info/), were described (M&M Supp. Table 1). Diagnostic PCR was 559

performed to identify the presence of T-DNA insertion in individuals plants using the allele-560

specific primers listed in M&M Supp. Table 2A. The generation of stable transgenic lines either 561

in WT or in the respective mutant backgrounds is described in the “Generation of transgenic 562

lines” section and lines are listed in (M&M Supp. Table 1). 563

564

Standard growth conditions 565

Arabidopsis seeds were surface sterilized using chlorine gas by pouring 3 mL of 37% 566

HCl into 100 mL of commercial bleach and airtight sealed for 4 hours. Then, seeds were air-567

cleared for at least 4 hours in a laminar flow cabinet and stored at 4ºC for 3-days stratification. 568

Seeds were sowed on half-strength Murashige-Skoog (MS) agar-solidified (0.6% [w/v] agar) 569

medium supplemented with 1,5% [w/v] sucrose under sterile conditions and grown with a long-570

day photoperiod (16-h light/8-h darkness cycle, 22±1 ᵒC, 150±50 µmol photons m-2 s-1) for 7-571

10 days. Seedlings were then collected for analysis, transferred to liquid media or to soil (4:1 572

(v/v) soil:vermiculite). 573

574

Plasmid Constructs 575


https://doi.org/10.1101/2020.06.23.167247

Two different types of DNA fragments were used as templates (genomic and coding 576

DNA sequence) to generate the constructs used in this work. First, genomic fragments were 577

amplified starting from at least 2 kb upstream the +1 ATG unless otherwise specified (Supp. 578

Fig. 1) until the stop codon (not included) using genomic Col-0 DNA as template. Second, 579

coding DNA sequences (CDS) were amplified, starting at +1 ATG until the stop codon (not 580

included), using as template Col-0 Arabidopsis cDNA that was generated using total RNA. In 581

the case of VTC2, due to a failure in amplifying the entire VTC2p:VTC2 product, about 1.4 kb 582

(according to Gao et al., 2011) of the VTC2 promoter including the 5'UTR was amplified (Supp. 583

Fig. 1; M&M Supp. Table 3). The primers used to generate these above-mentioned DNA 584

fragments are detailed in M&M Supp. Table 2B and 2C. A proofreading DNA polymerase was 585

used for all these PCR amplifications (TAKARA PrimeSTAR Max DNA Polymerase, CAT. 586

#R045A). 587

VTC2 promoter was digested with HindIII and directly cloned into the pGWB504 vector. 588

All other DNA fragments (excluding VTC2 promoter) were cloned into pENTR/D-TOPO using 589

the pENTR Directional TOPO cloning kit (Invitrogen). Next, on the one hand, DNA genomic 590

fragments (which includes promoter and lacks the stop codon) were recombined from 591

pENTR/D-TOPO into pGWB4 by LR reaction (Invitrogen) to generate SWp:SW-GFP 592

constructs for GME, VTC4 and L-GalDH. On the other hand, all CDS fragments (without 593

promoter, 5’-UTR and stop codon) were sub-cloned into pGWB5 and pGWB14 to generate 594

CaMV35S:SWCDS-GFP and CaMV35S:SWCDS-HA, respectively. In the case of VTC2, the CDS 595

was also recombined to VTC2p-pGWB504 by LR reaction (Invitrogen) thus obtaining the 596

VTC2p:VTC2CDS-GFP construct Likewise, CDS (pENTR/D-TOPO) were sub-cloned into both 597

pGADT7-GW (activation domain, AD) and pGBKT7-GW (binding domain, BD) to generate 598

SWCDS-AD and SWCDS-BD, respectively, for yeast two-hybrid assay (see Yeast two-hybrid 599

assay section). The expression vectors generated using the destination vectors pGWB5 and 600

pGWB14 contain a kanamycin and a hygromycin resistance gene for bacteria and plant 601

selection, respectively. The yeast two-hybrid vectors pGADT7 and pGBKT7 contain an 602

ampicillin and kanamycin resistance gene, respectively for selection in bacteria. A summary 603

of generated constructs is showed in M&M Supp. Table 3. All constructs were analyzed by 604

colony PCR (using the primers indicated in M&M Supp. Table 2D), enzymatic DNA restriction 605

and sequencing (using the primers indicated in M&M Supp. Table 2E). 606

The plasmids used from the pGWB vector series were provided by Tsuyoshi 607

Nakagawa (Department of Molecular and Functional Genomics, Shimane University; 608

Nakagawa et al., 2007). The pGADT7(GW) and pGBKT7(GW) destination vectors were 609

provided by Salomé Prat (Nacional de Biotecnología-Consejo Superior de Investigaciones 610

Científicas). 611

612

Generation of Arabidopsis transgenic lines 613

The constructs generated using pGWB4 plasmid for native expression of the SW 614

genes (pGWB504 for VTC2) were transformed into Agrobacterium tumefaciens 615

GV3101::pMP90 by electroporation and confirmed by colony PCR using the primers indicated 616

in M&M Supp. Table 2D. Arabidopsis WT (Col-0) plants were then transformed using these A. 617

tumefaciens strains by floral dipping (Clough and Bent, 1998). T2 plants were analyzed for 618

hygromycin resistance ratio resistant:sensitive (R:S) and lines with single insertion (T2 619

HygR:HygS=3, χ² (α=0.05), n>300) were selected. Selected T2 plants were grown in long day 620

photoperiod and a sample of rosette leaf was taken to perform immunoblot analysis (Fig. 2; 621

Supp. Fig. 3). Seeds from different T2 lines with a high expression were selected and 622

homozygous T3 and T4 plants were used in this work. 623


https://doi.org/10.1101/2020.06.23.167247

For generation of the complementation lines, selected T2 plants (with single insertion) were 624

crossed with their respective mutants in both directions as male and female. In order to obtain 625

homozygous plants for both the T-DNA insertion and the SWp:SW-GFP construct (HygR), we 626

selected hygromycin resistant plants and performed diagnostic PCR to identify T-DNA 627

insertion using the primers in M&M Supp. Table 2A. Homozygous F3 plants for T-DNA and 628

SWp:SW-GFP insertion were used in this work. 629

630

Ascorbate complementation assay 631

WT, gme (SALK_150208), lgaldh (SALK_056664) and gldh (SALK_060087) seeds 632

were surface sterilized and stratified for 3 days at 4 ºC. Seeds were then sowed in agar-633

solidified (0.5% [w/v] agar) modified Scholl medium with or without 0,5 mM ascorbate as 634

described elsewhere (Lim et al., 2016) and grown with a long-day photoperiod. After 10 days, 635

plants were individually collected and PCR-diagnosed to identify homozygous mutants. 636

637

Ascorbate measurements 638

High Performance Liquid Chromatography (HPLC) for combinatorial expression in Nicotiana 639

benthamiana 640

Ascorbate was extracted from approximately 250 mg fresh tissue using 2 mL of a 2% 641

(w/v) metaphosphoric acid, 2 mM EDTA ice-cold buffer (modified from Davey et al., 2006). 642

Samples were centrifuged at 18000 g for 20 min at 4ᵒC, and supernatants were filtered (0.45 643

µm) before injection into an HPLC (Agilent Technologies 1200 series; Rx-C18, 4.6x100 mm 644

3.5 µm column (Agilent Technologies 861967-902), diode array detector (Agilent 645

Technologies G1315D). Samples were measured at 254 nm using filtered (0.45 µm) 0.1 M 646

NaH2PO4, 0.2 mM EDTA, pH 3.1 (orthophosphoric acid) as mobile phase (Harapanhalli et al., 647

1993) at 4 ᵒC (column at 20 ᵒC; 5 µL sample injection, 0.7 mL/min flow, 200 bar). Standards 648

were prepared using ascorbic acid dissolved in extraction buffer, with concentrations ranging 649

from 0 mM to 2.5 mM. 650

651

Ascorbate oxidase (AO) activity for Arabidopsis transgenic lines 652

Ascorbate and dehydroascorbate were extracted from approximately 80 mg fresh 653

tissue using ice cold 1 mL of a 3% (w/v) metaphosphoric acid plus 1 mM EDTA. Samples were 654

centrifuged at 16000 g for 10 min at 4ᵒC. Absorbance measurements were carried out at 265 655

nm using 20 µL sample/standard aliquots in a 96 well UV-transparent plate (Greiner UV-Star® 656

96-Well Microplate) mixed with 100 µL of either phosphate buffer (0.2 M KH2PO4, pH 7 to 657

measure ascorbate) or Tris(2-carboxyethyl)phosphine (TCEP)-phosphate buffer (0.2 M 658

KH2PO4, pH7, 2 mM TCEP, to measure ascorbate and dehydroascorbate). Then, 5 µL 40 659

U/mL AO was added to each well, mixed and incubated at room temperature for 20 minutes, 660

when the absorbance was again measured. The difference in the absorbance was interpolated 661

in the standard curve, using standards ranging from 0 mM to 1 mM in extraction buffer. TCEP 662

and AO stocks were prepared in phosphate buffer. 663

664

Immunoblot 665

Proteins were separated by size through a sodium dodecyl sulphate-polyacrylamide 666

gel electrophoresis (SDS-PAGE) and transferred onto a polyvinylidene difluoride (PVDF) 667

membrane (Immobilon-P, Millipore, 0.45 µm pore size; CAT. #IPVH00010) previously 668

methanol-equilibrated. Next, the PVDF membrane was blocked with 5% (w/v) skimmed milk 669

followed by an incubation with a primary antibody synthesized in mouse against GFP (Santa 670

Cruz Biotechnology, CAT. #sc-9996; 1:600) or HA (Sigma-Aldrich, CAT. #H3663; 1:3000) 671


https://doi.org/10.1101/2020.06.23.167247

epitopes. Then, the membrane was subsequentially incubated with a horseradish peroxidase 672

(HRP)-conjugated α-mouse whole IgG antibody produced in rabbit (Sigma-Aldrich, CAT. 673

#A9044; 1:80000). Protein-HRP chemiluminescence was detected using Chemidoc XRS1 674

System (Bio-Rad) after incubation with Clarity ECL Western Blotting Substrate or SuperSignal 675

West Femto Maximum Sensitivity Substrate according to the manufacturer’s instructions. 676

677

Transient expression in Nicotiana benthamiana leaves 678

Agrobacterium tumefaciens GV3101::pMP90 was transformed by electroporation with 679

the constructs generated using pGWB5 and pGWB14 positive colonies were confirmed by 680

colony PCR. Two leaves per plant (leaves 3 and 4 from apex) 4-week-old N. benthamiana 681

were infiltrated in the abaxial side of using 1 mL syringe (without needle). Agrobacterium 682

cultures were grown overnight in liquid Luria-Bertani medium containing rifampicin (50 μg/mL), 683

gentamycin (25 μg/mL), and the construct-specific antibiotic kanamycin (50 μg/mL). Cells were 684

then harvested by centrifugation (for 15 min at 3000 g in 50-mL falcon tubes) at room 685

temperature, pellets were resuspended in agroinfiltration solution (10 mM MES, pH 5.6, 10 686

mM MgCl2, and 1 mM acetosyringone), and kept for 2-3 h in dark conditions at room 687

temperature. In order to avoid overexpression-associated gene silencing, Agrobacterium 688

strain containing a p19 expression vector was coinfiltrated. For single gene expression, the 689

gene of interest and p19 were mixed in such a way that the optical densities (600 nm, OD600) 690

were 0.8 and 0.2, respectively. For multiple co-infiltration experiments, the genes of interest 691

were diluted in the same proportion so the final OD600 was 1 (0,8 for the combination of genes 692

of interest and 0,2 for p19). Agrobacterium strains expressing either free GFP or TTL3-HA (an 693

ascorbate biosynthesis non-related protein (Amorim-Silva et al., 2019)) were used to 694

compensate OD600 in co-infiltration experiments. 695

To achieve optimal protein expression, immunoblot time course analysis was addressed 2, 3 696

and 4 days after infiltration (Fig. 3B). Protein loads were adjusted so comparable amounts of 697

GFP-tagged protein were detected two days after infiltration for each construct using the Java-698

based image-processing program FIJI (Schindelin et al., 2012; Schneider et al., 2012). 699

700

Confocal laser scanning microscopy 701

All confocal images of N. benthamiana (Fig. 4) were obtained using a Leica TCS SP5 702

II confocal microscope equipped with a 488-nm argon laser for GFP excitation and a PMT for 703

its detection using the HCX PL APO CS 40x1,25 OIL objective with additional 2x digital zoom. 704

Arabidopsis imaging was carried out using a Leica TCS SP8 equipped with HC PL APO CS2 705

40x/1,30 OIL objective and a solid state 488-nm laser used to excite GFP and a high-sensitivity 706

SP hybrid detector (Fig. 5A) and a Zeiss LSM 880 (Fig. 5B) with a 488-nm argon laser and 707

PMT detectors for GFP and transmitted light using the C-Apochromat 40x/1.2 W Korr M27 708

objective. All image processing was performed using FIJI (Schindelin et al., 2012; Schneider 709

et al., 2012). 710

711

Yeast two-hybrid assay 712

The Gal4-based yeast two-hybrid system (Clontech Laboratories) was used for testing 713

the interaction. Saccharomyces cerevisiae Meyen ex E. C. Hansen strain PJ69-4A was 714

transformed with the constructs described in the plasmid constructs section (SW genes clones 715

into pGADT7 and pGBKT7) as described elsewhere (Gietz and Schiestl, 1995). Constructs 716

transformed are explained in the plasmid constructs section (pGADT7 and pGBKT7). 717

Transformants were grown on plasmid-selective medium (synthetic defined (SD)/ -Trp -Leu) 718

and grown at 28 ᵒC for 4 days. Two independent colonies were taken per transformation event 719


https://doi.org/10.1101/2020.06.23.167247

and resuspended in 200 µL of sterile water. Ten-fold serial dilutions were made and 5 µL of 720

each dilution were spotted onto three different interaction-selective medium: SD -Trp (minus 721

tryptophan) -Leu (minus leucin) -His (minus histidine) +2 mM 3-AT (3-amino-1,2,4-triazole); 722

SD -Trp -Leu -Ade; SD -Trp -Leu -Ade +3-AT. Full SD medium was also used to check strain 723

survival. Plates were grown at 28 ᵒC and pictures were taken after 3, 5 and 9 days (only 5 724

days is shown). Concentrations are available at Rodríguez-Negrete et al. (2014). 725

726

Bioinformatic analyses 727

Protein domains predictions were performed using Protein Basic Local Alignment 728

Search from the National Center for Biotechnology Information (Altschul et al., 1997, BLASTp; 729

https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Proteins) followed by functional annotation of 730

protein domain using the Conserved Domain Database (Marchler-Bauer et al., 2017, 2015, 731

2011; Marchler-Bauer and Bryant, 2004; 732

https://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml). Protein subcellular localization was 733

predicted using COMPARTMENTS (Binder et al., 2014; 734

https://compartments.jensenlab.org/Search). Predicted subcellular localization signals were 735

performed using LOCALIZER (Sperschneider et al., 2017; http://localizer.csiro.au/). Metabolic 736

control Analysis was carried out using Complex Pathway Simulator (COPASI, Hoops et al., 737

2006; http://copasi.org/). All the details needed for the simulations are collected in Supp. 738

Tables 3-7). VTC2 mRNA expression datasets were extracted from Transcriptome Variation 739

Analysis database (Klepikova et al., 2016, TRAVA; http://travadb.org/) and eFP-seq Browser 740

(Sullivan et al., 2019; https://bar.utoronto.ca/eFP-Seq_Browser/). 741

742

Coimmunoprecipitation assay 743

Four-week-old N. benthamiana plants were used for transient expression assays as 744

described above. Leaves were ground to fine powder in liquid nitrogen. ~0.5 g of ground 745

leaves per sample was used, and total proteins were extracted with 1 mL of non-denaturing 746

extraction buffer (50 mM Tris-HCl pH7.5, 150 mM NaCl, 1% (v/v) Nonidet P-40, 10 mM EDTA, 747

1mM Na2MoO4, 1 mM NaF, 10 mM DTT, 0.5 mM PMSF**, 1% (v/v) protease inhibitor (Sigma 748

P9599)) and incubated for 30 min at 4 ᵒC using an end-over-end rocker. Protein extracts were 749

centrifuged at 20000 g for 20 min at 4 ᵒC and then filtered by gravity using Poly-Prep 750

chromatography columns (731-1550, Bio-Rad). Supernatants were filtered by gravity through 751

Poly-Prep chromatography columns (731-1550, Bio-Rad), and 100 mL was reserved for 752

immunoblot analysis as input. The remaining supernatants were used for immunoprecipitation 753

of GFP-fused proteins agarose using GFP-Trap coupled to agarose beads (Chromotek) and 754

following the manufacturer’s instructions. Total (input), immunoprecipitated (IP), and CoIP 755

samples were resuspended with 2x concentrated Laemmli sample buffer and heated at 95°C 756

for 5 min protein denaturation. Finally, samples were separated in a 10% SDS-PAGE gel and 757

analyzed as described above. 758

759

Statistical analysis 760

All analyses were performed using SigmaPlot v11 for Windows considering n>3 and α=0.05. 761

762

Accession Numbers 763

Sequence data were sourced from the The Arabidopsis Information Resource (TAIR) 764

(https://www.arabidopsis.org/) under the following accession numbers: AT2G39770 for VTC1, 765

AT5G28840 for GME, AT4G26850 for VTC2, AT3G02870 for VTC4, AT4G33670 for L-GalDH, 766

and AT3G47930 for L-GalLDH. 767


https://doi.org/10.1101/2020.06.23.167247

Supplemental material 768

769

Supplementary Figure 1. Gene structure, primers disposition and mutants’ T-DNA location. 770

771

Supplementary Figure 2. Diagnostic PCR of the gme, lgaldh and gldh seedlings analyzed in 772

the complementation assay described in Figure 2A. Whereas ascorbate does not recover the 773

lethality of gme mutation (A), it does recover lgaldh (C). As a control, we used gldh (B), 774

previously published to be recovered by ascorbate (Pineau et al., 2008). 775

776

Supplementary Figure 3. Generation of Arabidopsis transgenic lines containing ascorbate 777

biosynthesis enzymes fused to GFP. A) Immunoblot (α-GFP) analysis of fourteen GME-GFP 778

T2 lines identifies several lines expressing different amounts of GME-GFP protein 779

(background WT). After the identification of lines harboring single insertions (T2 HygR:HygS=3, 780

χ² (α=0.05)), we selected line L6 (arrowhead) to introduce the GMEp:GME-GFP construct into 781

the gme mutant (SALK_150208) background using reciprocal crosses. B) Diagnostic PCR of 782

F2 generation from the cross indicated identified plants harboring GME-GFP in homozygous 783

mutant background (arrowheads). Ascorbate content of this line is shown in Figure 2. C) 784

Immunoblot (α-GFP) analysis of VTC2-GFP of two lines among which ascorbate content 785

ranges. Four days old seedlings were grown under long day conditions in solid half-strength 786

MS agar plates. D) Immunoblot (α-GFP) analysis of L-GalDH-GFP T2 generation showing the 787

six lines expressing L-GalDH-GFP protein (background WT). After the identification of lines 788

harboring single insertions (T2 HygR:HygS=3, χ² (α=0.05)), we selected line L1 to introduce the 789

L-GalDHp:L-GalDH-GFP construct into the lgaldh mutant (SALK_056664) background using 790

reciprocal crosses. H) Diagnostic PCR of F2 generation from the cross indicated identified 791

plants harboring L-GalDH-GFP in homozygous mutant background (arrowheads). Ascorbate 792

content of this line is showed in Figure 2. E) Ascorbate content of the T3 generation of lines 793

generated consisting of vtc2 mutants harboring VTC2p:VTC2-GFP. Three fully expanded 794

leaves were collected on five-week-old T3 plants grown on soil in a growth chamber under 795

short-day conditions (10 h light at 22°C/14 h dark at 19°C), at 60% humidity and 200 µmol m–796 2 s–1 light intensity. Different letters denote statistically significant differences for ascorbate 797

content (One-Way ANOVA, α=0,05). G) Immunoblot (α-GFP) analysis of VTC4-GFP T2 798

generation showing the six lines expressing VTC4-GFP protein (background WT). After the 799

identification of lines harboring single insertions (T2 HygR:HygS=3, χ² (α=0.05)), we selected 800

line L5 to introduce the VTC4p:VTC4-GFP construct into the vtc4 mutant (SALK_077222) 801

background using reciprocal crosses. H) Diagnostic PCR of F2 generation from the cross 802

indicated identified plants harboring VTC4-GFP in homozygous mutant background 803

(arrowheads). Ascorbate content of this line is showed in Figure 2. 804

805

Supplementary Figure 4. VTC2 mRNA levels are similar to those of VTC1 and GME, although 806

VTC2 protein is hardly detectable. Data were taken from two different public RNA sequencing 807

(RNA-seq) data sets: Transcriptome Variation Analysis (TRAVA, travadb.org; Klepikova et al., 808

2016) and the eFP-Seq Browser (https://bar.utoronto.ca/eFP-Seq_Browser/; Sullivan et al., 809

2019) corresponding to different adult plants’ vegetative and green tissues, respectively. 810

811

Supplementary Figure 5. Complete blots corresponding to Figure 2B (red rectangle). Blots A, 812

B and C have the same time of exposure using an ultra-sensitive enhanced chemiluminescent 813

substrate for low femtogram protein level detection. Blot C and D are the same membrane 814


https://doi.org/10.1101/2020.06.23.167247

with different exposure times. Protein loads were calculated using FIJI from a first immunoblot 815

(not shown) and diluted so the expression of GFP-tagged proteins in the first timepoint is 816

similar among constructs. 817

818

Supplementary Figure 6. Predicted subcellular localization of ascorbate biosynthesis proteins 819

in the plant cell according to COMPARTMENTS (https://compartments.jensenlab.org/Search; 820

Binder et al., 2014). 821

822

Supplementary Figure 7. Yeast two-hybrid whole screening. Red rectangles correspond to the 823

spotting presented in Figure 6A. As positive controls, we used SNF1/SNF4 (Fields and Song, 824

1989) and p53/SV40-Large AgT (AgT; Iwabuchi et al., 1993; Li and Fields, 1993) interactions. 825

As a negative control, we included LamC/AgT cotransformation (Bartel et al., 1993; Ye and 826

Worman, 1995). His: histidine, 3-AT: 3-amino-1,2,4-triazole, Ade: adenine. 827

828

Supplementary Figure 8. Immunoblot (α-GFP) of N. benthamiana leaves crude extracts 829

following agroinfiltration of above-indicated combinations of proteins (also see Appendix 1). 830

Two leaves of 2 or 3 independent plants were infiltrated, using GFP or TTL3-HA (an HA-831

tagged protein non-related to ascorbate synthesis) to rise optical density up to 1. 832

833

Supplementary Table 1. Infiltration of Nicotiana benthamiana corresponding to Figure 3C-H. * 834 Data transformed with log10 x to accomplish normality and homoscedasticity required for 835 ANOVA. 836 837 Supplementary Table 2. Prediction of transit peptides and nuclear localization signals within 838 VTC genes. Values in the brackets are (probability | residues involved). Results obtained from 839 LOCALIZER (http://localizer.csiro.au/; Sperschneider et al., 2017) 840 841 Supplementary Table 3. Initial conditions for the COPASI kinetic model of ascorbate 842 biosynthesis and turnover. 843 844 Supplementary Table 4. Flux control coefficients for ascorbate biosynthesis and turnover with 845 various strengths of non-competitive feedback inhibition at the GDP-L-Gal phosphorylase step 846 (Ki GGP). Values were calculated in COPASI (Hoops et al., 2006) using conditions indicated 847 in Supplementary Table 3. 848 849 Supplementary Table 5. Concentration control coefficients at each step for pathway 850 intermediates with various strengths of non-competitive feedback inhibition at the GDP-L-Gal 851 phosphorylase step (Ki GGP). Values were calculated in COPASI (Hoops et al., 2006) using 852 conditions indicated in Supplementary Table 3. 853 854 Supplementary Table 6. Flux (A) and concentration (B) control coefficients at each step with 855 competitive feedback inhibition of PMI and L-GalDH relaxed to 100 mM (compare 856 Supplementary Tables 4 and 5 for published Ki values). Values were calculated in COPASI 857 (Hoops et al., 2006) using conditions indicated in Supplementary Table 3. 858 859 Supplementary Table 7. The effect of varying relative enzyme activity (V or Vf) on the 860 concentration of pathway intermediates with various strengths of non-competitive feedback 861 inhibition at the GDP-L-Gal phosphorylase step (Ki GGP). Values were calculated in COPASI 862 (Hoops et al., 2006) using conditions indicated in Supplementary Table 3. 863 864


https://doi.org/10.1101/2020.06.23.167247

Material and Methods Supplementary Table 1. Arabidopsis mutants and lines used in this 865 work. 866 867 Material and Methods Supplementary Table 2. Arabidopsis stable lines used in this work. 868 869 Material and Methods Supplementary Table 3. A) Sizes of constructs generated in this work 870

using the vectors listed in B) and the resultant construct. 871

872

Acknowledgements 873

We acknowledge Glenda Gillaspy (Department of Biochemistry, Virginia Tech) for kindly 874

providing us the vtc4-4 mutant used in this work. 875

876

Tables 877

878

Table 1. Phenotypes associated to Arabidopsis mutants affected in the genes involved in 879

ascorbate biosynthesis available in the literature. Phenotype analysis shows that VTC2 is the 880

first enzyme fully dedicated to ascorbate synthesis since growth arrest occurring in the 881

vtc2/vtc5 double mutant and downstream mutants is rescued by ascorbate supplementation. 882

Although a reduced function (knock-down mutation) of genes upstream of GGP (VTC2, VTC5) 883

also results in lower ascorbate concentration, the lethality occurring in knock-out mutants 884

cannot be prevented by exogenous ascorbate supplementation. 885

886 Enzyme Mutant Allele Phenotype Reference

GMP cyt1-1 Knock-out Embryo lethal Lukowitz et al., 2001

vtc1-1 Knock-down 30% of WT ascorbate Conklin et al., 2000, 1996

GME gme-1 Knock-out

Male gametophyte lethal. Growth defect, rescued by boron but not by ascorbate supplementation.

Qi et al., 2017

gme-2 Knock-down 30% of WT ascorbate

GGP

vtc2-4 Single knock-out 20% of WT ascorbate Lim et al., 2016

vtc5-2 Single knock-out 80% of WT ascorbate Dowdle et al., 2007; Lim et al., 2016 vtc2/vtc5 Double knock-out

Growth arrest, rescued by ascorbate supplementation

GPP vtc4-4 Knock-out 65% of WT ascorbate Torabinejad et al., 2009

L-GalDH lgaldh Not reported 30% of WT ascorbate in antisense suppression lines

Gatzek et al., 2002

L-GalLDH gldh Knock-out Growth arrest, rescued by ascorbate supplementation

Pineau et al., 2008

887

888

Figure legends 889

890

Figure 1. The ascorbate biosynthesis via GDP-mannose and L-galactose (the Smirnoff-891

Wheeler [SW] pathway) in Arabidopsis thaliana. Yellow area refers to the cytosolic enzymes 892

used in this work. mOM: mitochondrion outer membrane, mIMS: mitochondrion 893

intermembrane space, mIM: mitochondrion inner membrane. In the left-hand side of the 894

arrows, the genes encoding each enzyme are displayed while in the right-hand side the 895

encoded proteins/enzymatic activities are shown. Glc=glucose, Fru=fructose, Man=mannose, 896

Gal=galactose, GalL=galactono-1,4-lactone, Asc= ascorbic acid. PGI=phosphoglucose 897

isomerase, PMI=phosphomannose isomerase, PMM=phosphomannomutase, GMP=GDP-898

Mannose pyrophosphorylase, GME=GDP-mannose 3’,5’ epimerase, GGP= GDP-L-galactose 899

phosphorylase, GPP= L-galactose 1-phosphate phosphatase, L-GalDH= L-galactose 900


https://doi.org/10.1101/2020.06.23.167247

dehydrogenase, L-GalLDH= L-galactono-1,4-lactone dehydrogenase. CytRED=cytochrome c 901

reduced, CytOX=cytochrome c oxidized. 902

903

Figure 2. Characterization of C-terminal GFP fusions of ascorbate biosynthesis enzymes. A) 904

Ascorbate complementation assay rescues growth arrest from gldh and lgaldh mutants but 905

not gme lethality. In small dashed circles, homozygous mutants show arrested growth in media 906

lacking ascorbate, but not in media supplemented with ascorbate. B) Complementation of 907

ascorbate concentration by expression of GFP fusion proteins in ascorbate deficient 908

Arabidopsis mutants. Different letters denote statistically significant differences for ascorbate 909

concentration (One-Way ANOVA, α=0,05). Ascorbate was extracted from three independent 910

samples composed of three 6-weeks old fully expanded rosettes grown in short day regime at 911

150 µmol photons m-2 s-1, one and a half hours after lights turned on. C) Immunoblot (α-GFP) 912

of complemented lines show that little amount of protein VTC2 compared other components 913

restores ascorbate content. Red arrowhead indicates VTC2-GFP. A representative sample for 914

each genotype is shown out of the three tested by immunoblot (not shown). 915

916

Figure 3. The effect of transient expression of C-terminal GFP and HA fusions of ascorbate 917

biosynthesis enzymes on ascorbate concentration in Nicotiana benthamiana leaves. A) 918

Strategy followed to clone and to overexpress in N. benthamiana leaves ascorbate 919

biosynthesis genes from Arabidopsis translationally fused to GFP and HA at the protein C-920

terminus. B) Immunoblots (α-GFP) of fusion protein accumulation 2, 3 and 4 days after 921

agroinfiltration. Expression was driven by the 35S promoter. Protein amount was measured 922

and loaded to obtain similar intensities in the blot at 2 days after infiltration. C-H) Leaf 923

ascorbate concentration 3 days after agroinfiltration with different combinations of C-terminal 924

GFP ad HA fusions of ascorbate biosynthesis enzymes. Constructs used in each infiltration 925

are shown in Supplementary Table 1. Two leaves of at least three N. benthamiana plants were 926

infiltrated and samples were collected at 3 days after infiltration. 927

928

Figure 4. Subcellular localization of free GFP and C-terminal GFP fusions of ascorbate 929

biosynthesis enzymes transiently expressed in Nicotiana benthamiana leaves under the 930

control of the 35S promoter 3 days after agroinfiltration. GFP was visualized by laser scanning 931

confocal microscopy. Scale bar = 30 µm. 932

933

Figure 5. Subcellular localization of ascorbate biosynthesis enzymes fused to GFP at the C-934

terminus and expressed under the control of their native promoters in Arabidopsis. A) The 935

expression of GFP-tagged enzymes is ubiquitous in four days-old seedlings and their 936

subcellular localization is compatible with cytosol and nucleus except for GME, which only 937

locates in cytosol. B) GFP signal of a functional VTC2-GFP (line 13) protein was not 938

detectable. GFP was visualized by laser scanning confocal microscopy. Scale bar = 30 µm. 939

940

Figure 6. Interaction of ascorbate biosynthesis enzymes assessed by yeast two-hybrid 941

analysis and coimmunoprecipitation (CoIP). (A) Yeast two-hybrid assay detected no direct 942

interaction between the enzymes, but VTC1, GME and VTC4, which contain predicted 943

dimerization/trimerization domains (B), form dimers/trimers. B) Protein scheme for each of the 944

enzymes assayed in yeast two-hybrid showing length (in amino acids), annotated domains 945

and predicted dimer/trimerization interfaces. (C-G) CoIP assay reported association in vivo in 946

Nicotiana between consecutive steps as well as between the first (VTC1) and the last (L-947

GalDH) steps occurring in cytosol. GFP-tagged enzyme from Nicotiana crude protein extracts 948


https://doi.org/10.1101/2020.06.23.167247

containing two overexpressed consecutive enzymes, GFP and HA-tagged, was pulled-down 949

using α-GFP agarose beads and coimmunoprecipitated protein was detected using α-HA 950

antibody. 951

952

Figure 7. A kinetic model of ascorbate biosynthesis and turnover. A) Scheme of the pathway 953

modelled in COPASI showing intermediates catalytic steps (arrows) and feedback (red lines). 954

The initial model parameters are shown in Supplementary Table 3. B) Metabolic control 955

analysis at steady state was used to calculate flux control coefficients for each enzyme with 956

various strengths of non-competitive inhibition exerted by ascorbate (indicated by Ki values) 957

on GGP (step 6). C) Metabolic control analysis at steady state was used to calculate 958

concentration control coefficients for ascorbate with various strengths of non-competitive 959

inhibition exerted by ascorbate (indicated by Ki values) on GGP (step 6). D) The effect of 960

varying enzyme activity on ascorbate concentration with various strengths of non-competitive 961

inhibition exerted by ascorbate (indicated by Ki values) on GGP (step 6). E) Time course of 962

the change in ascorbate concentration in response to the addition of D-mannose 6-phosphate 963

(D-Man 6-P), L-galactose (L-Gal) and L-galactono-1,4-lactone (L-GalL). 1: PGI, 2: PMI, 3: PMM, 964

4: GMP, 5: GME, 6: GGP, 7: GPP, 8: L-GalDH, 9: L-GalLDH, 10: Turnover, 11: Oxidation, 12: 965

Reduction). 966

967

968

Literature cited 969

970

Ali, B., Pantha, S., Acharya, R., Ueda, Y., Wu, L.B., Ashrafuzzaman, M., Ishizaki, T., 971 Wissuwa, M., Bulley, S., Frei, M., 2019. Enhanced ascorbate level improves multi-972 stress tolerance in a widely grown indica rice variety without compromising its 973 agronomic characteristics. J. Plant Physiol. https://doi.org/10.1016/j.jplph.2019.152998 974

Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 975 1997. Gapped BLAST and PSI-BLAST: A new generation of protein database search 976 programs. Nucleic Acids Res. https://doi.org/10.1093/nar/25.17.3389 977

Amorim-Silva, V., García-Moreno, Á., Castillo, A.G., Lakhssassi, N., Del Valle, A.E., Pérez-978 Sancho, J., Li, Y., Posé, D., Pérez-Rodriguez, J., Lin, J., Valpuesta, V., Borsani, O., 979 Zipfel, C., Macho, A.P., Botella, M.A., 2019. TTL proteins scaffold brassinosteroid 980 signaling components at the plasma membrane to optimize signal transduction in 981 arabidopsis. Plant Cell. https://doi.org/10.1105/tpc.19.00150 982

Asada, K., 1999. THE WATER-WATER CYCLE IN CHLOROPLASTS: Scavenging of Active 983 Oxygens and Dissipation of Excess Photons. Annu. Rev. Plant Physiol. Plant Mol. Biol. 984 50, 601–639. https://doi.org/10.1146/annurev.arplant.50.1.601 985

Badejo, A.A., Jeong, S.T., Goto-Yamamoto, N., Esaka, M., 2007. Cloning and expression of 986 GDP-d-mannose pyrophosphorylase gene and ascorbic acid content of acerola 987 (Malpighia glabra L.) fruit at ripening stages. Plant Physiol. Biochem. 988 https://doi.org/10.1016/j.plaphy.2007.07.003 989

Badejo, A.A., Tanaka, N., Esaka, M., 2008. Analysis of GDP-D-Mannose Pyrophosphorylase 990 Gene Promoter from Acerola (Malpighia glabra) and Increase in Ascorbate Content of 991 Transgenic Tobacco Expressing the Acerola Gene. Plant Cell Physiol. 49, 126–132. 992 https://doi.org/10.1093/pcp/pcm164 993

Bartel, P., Chien, C.T., Sternglanz, R., Fields, S., 1993. Elimination of false positives that 994 arise in using the two-hybrid system. Biotechniques 14, 920–924. 995

Bartoli, C.G., 2006. Inter-relationships between light and respiration in the control of ascorbic 996 acid synthesis and accumulation in Arabidopsis thaliana leaves. J. Exp. Bot. 57, 1621–997 1631. https://doi.org/10.1093/jxb/erl005 998

Binder, J.X., Pletscher-Frankild, S., Tsafou, K., Stolte, C., O’Donoghue, S.I., Schneider, R., 999


https://doi.org/10.1101/2020.06.23.167247

Jensen, L.J., 2014. COMPARTMENTS: unification and visualization of protein 1000 subcellular localization evidence. Database 2014, bau012–bau012. 1001 https://doi.org/10.1093/database/bau012 1002

Boukouris, A.E., Zervopoulos, S.D., Michelakis, E.D., 2016. Metabolic Enzymes 1003 Moonlighting in the Nucleus: Metabolic Regulation of Gene Transcription. Trends 1004 Biochem. Sci. 41, 712–730. https://doi.org/10.1016/j.tibs.2016.05.013 1005

Bulley, S., Laing, W., 2016. The regulation of ascorbate biosynthesis. Curr. Opin. Plant Biol. 1006 33, 15–22. https://doi.org/10.1016/j.pbi.2016.04.010 1007

Bulley, S., Wright, M., Rommens, C., Yan, H., Rassam, M., Lin-Wang, K., Andre, C., 1008 Brewster, D., Karunairetnam, S., Allan, A.C., Laing, W.A., 2012. Enhancing ascorbate 1009 in fruits and tubers through over-expression of the L-galactose pathway gene GDP-L-1010 galactose phosphorylase. Plant Biotechnol. J. 10, 390–397. 1011 https://doi.org/10.1111/j.1467-7652.2011.00668.x 1012

Bulley, S.M., Rassam, M., Hoser, D., Otto, W., Schünemann, N., Wright, M., MacRae, E., 1013 Gleave, A., Laing, W., 2009. Gene expression studies in kiwifruit and gene over-1014 expression in Arabidopsis indicates that GDP-L-galactose guanyltransferase is a major 1015 control point of vitamin C biosynthesis. J. Exp. Bot. 60, 765–778. 1016 https://doi.org/10.1093/jxb/ern327 1017

Clough, S.J., Bent, A.F., 1998. Floral dip: a simplified method forAgrobacterium-mediated 1018 transformation ofArabidopsis thaliana. Plant J. 16, 735–743. 1019 https://doi.org/10.1046/j.1365-313x.1998.00343.x 1020

Conklin, P.L., Gatzek, S., Wheeler, G.L., Dowdle, J., Raymond, M.J., Rolinski, S., Isupov, 1021 M., Littlechild, J.A., Smirnoff, N., 2006. Arabidopsis thaliana VTC4 encodes L-1022 galactose-1-P phosphatase, a plant ascorbic acid biosynthetic enzyme. J. Biol. Chem. 1023 281, 15662–15670. https://doi.org/10.1074/jbc.M601409200 1024

Conklin, P.L., Norris, S.R., Wheeler, G.L., Williams, E.H., Smirnoff, N., Last, R.L., 1999. 1025 Genetic evidence for the role of GDP-mannose in plant ascorbic acid (vitamin C) 1026 biosynthesis. Proc. Natl. Acad. Sci. 96, 4198–4203. 1027 https://doi.org/10.1073/pnas.96.7.4198 1028

Conklin, P.L., Pallanca, J.E., Last, R.L., Smirnoff, N., 1997. L-ascorbic acid metabolism in 1029 the ascorbate-deficient arabidopsis mutant vtc1. Plant Physiol. 115, 1277–85. 1030 https://doi.org/10.1016/s0168-9452(03)00277-2 1031

Conklin, P.L., Saracco, S.A., Norris, S.R., Last, R.L., 2000. Identification of ascorbic acid-1032 deficient Arabidopsis thaliana mutants. Genetics 154, 847–56. 1033

Conklin, P.L., Willlams, E.H., Last, R.L., 1996. Environmental stress sensitivity of an 1034 ascorbic acid-deficient Arabidopsis mutant. Proc. Natl. Acad. Sci. 93, 9970–9974. 1035

Cronje, C., George, G.M., Fernie, A.R., Bekker, J., Kossmann, J., Bauer, R., 2012. 1036 Manipulation of L-ascorbic acid biosynthesis pathways in Solanum lycopersicum: 1037 elevated GDP-mannose pyrophosphorylase activity enhances L-ascorbate levels in red 1038 fruit. Planta 235, 553–564. https://doi.org/10.1007/s00425-011-1525-6 1039

Davey, M.W., Kenis, K., Keulemans, J., 2006. Genetic control of fruit vitamin C contents. 1040 Plant Physiol. https://doi.org/10.1104/pp.106.083279 1041

DeBolt, S., Cook, D.R., Ford, C.M., 2006. L-Tartaric acid synthesis from vitamin C in higher 1042 plants. Proc. Natl. Acad. Sci. 103, 5608–5613. 1043 https://doi.org/10.1073/pnas.0510864103 1044

Debolt, S., Melino, V., Ford, C.M., 2007. Ascorbate as a Biosynthetic Precursor in Plants. 1045 Ann. Bot. 99, 3–8. https://doi.org/10.1093/aob/mcl236 1046

Dewhirst, R.A., Clarkson, G.J.J., Rothwell, S.D., Fry, S.C., 2017. Novel insights into 1047 ascorbate retention and degradation during the washing and post-harvest storage of 1048 spinach and other salad leaves. Food Chem. 1049 https://doi.org/10.1016/j.foodchem.2017.04.082 1050

Dewhirst, R.A., Fry, S.C., 2018. The oxidation of dehydroascorbic acid and 2,3-1051 diketogulonate by distinct reactive oxygen species. Biochem. J. 1052 https://doi.org/10.1042/BCJ20180688 1053

Dewhirst, R.A., Murray, L., Mackay, C.L., Sadler, I.H., Fry, S.C., 2020. Characterisation of 1054


https://doi.org/10.1101/2020.06.23.167247

the non-oxidative degradation pathway of dehydroascorbic acid in slightly acidic 1055 aqueous solution. Arch. Biochem. Biophys. 681, 108240. 1056 https://doi.org/10.1016/j.abb.2019.108240 1057

Dowdle, J., Ishikawa, T., Gatzek, S., Rolinski, S., Smirnoff, N., 2007. Two genes in 1058 Arabidopsis thaliana encoding GDP-L-galactose phosphorylase are required for 1059 ascorbate biosynthesis and seedling viability. Plant J. https://doi.org/10.1111/j.1365-1060 313X.2007.03266.x 1061

Dunkley, T.P.J., Hester, S., Shadforth, I.P., Runions, J., Weimar, T., Hanton, S.L., Griffin, 1062 J.L., Bessant, C., Brandizzi, F., Hawes, C., Watson, R.B., Dupree, P., Lilley, K.S., 2006. 1063 Mapping the Arabidopsis organelle proteome. Proc. Natl. Acad. Sci. 103, 6518–6523. 1064 https://doi.org/10.1073/pnas.0506958103 1065

Fell, D.A., 1992. Metabolic control analysis: a survey of its theoretical and experimental 1066 development. Biochem. J. 286, 313–330. https://doi.org/10.1042/bj2860313 1067

Fenech, M., Amaya, I., Valpuesta, V., Botella, M.A., 2019. Vitamin C content in fruits: 1068 Biosynthesis and regulation. Front. Plant Sci. 9, 1–21. 1069 https://doi.org/10.3389/fpls.2018.02006 1070

Fields, S., Song, O., 1989. A novel genetic system to detect protein–protein interactions. 1071 Nature 340, 245–246. https://doi.org/10.1038/340245a0 1072

Franceschi, V.R., Tarlyn, N.M., 2002. L-Ascorbic Acid Is Accumulated in Source Leaf 1073 Phloem and Transported to Sink Tissues in Plants. Plant Physiol. 130, 649–656. 1074 https://doi.org/10.1104/pp.007062 1075

Gao, Y., Badejo, A.A., Shibata, H., Sawa, Y., Maruta, T., Shigeoka, S., Page, M., Smirnoff, 1076 N., Ishikawa, T., 2011. Expression Analysis of the VTC2 and VTC5 Genes Encoding 1077 GDP-L-Galactose Phosphorylase, an Enzyme Involved in Ascorbate Biosynthesis, in 1078 Arabidopsis thaliana. Biosci. Biotechnol. Biochem. 75, 1783–1788. 1079 https://doi.org/10.1271/bbb.110320 1080

Gatzek, S., Wheeler, G.L., Smirnoff, N., 2002. Antisense suppression of L-galactose 1081 dehydrogenase in Arabidopsis thaliana provides evidence for its role in ascorbate 1082 synthesis and reveals light modulated L-galactose synthesis. Plant J. 1083 https://doi.org/10.1046/j.1365-313X.2002.01315.x 1084

Gietz, R.D., Schiestl, R.H., 1995. Transforming yeast with DNA. Methods Mol. Cell. Biol. 5, 1085 255–269. 1086

Gilbert, L., Alhagdow, M., Nunes-Nesi, A., Quemener, B., Guillon, F., Bouchet, B., Faurobert, 1087 M., Gouble, B., Page, D., Garcia, V., Petit, J., Stevens, R., Causse, M., Fernie, A.R., 1088 Lahaye, M., Rothan, C., Baldet, P., 2009. GDP-d-mannose 3,5-epimerase (GME) plays 1089 a key role at the intersection of ascorbate and non-cellulosic cell-wall biosynthesis in 1090 tomato. Plant J. 60, 499–508. https://doi.org/10.1111/j.1365-313X.2009.03972.x 1091

Green, M.A., Fry, S.C., 2005. Vitamin C degradation in plant cells via enzymatic hydrolysis 1092 of 4-O-oxalyl-l-threonate. Nature 433, 83–87. https://doi.org/10.1038/nature03172 1093

Harapanhalli, R.S., Howell, R.W., Rao, D. V., 1993. Testicular and plasma ascorbic acid 1094 levels in mice following dietary intake: a high-performance liquid chromatographic 1095 analysis. J. Chromatogr. B Biomed. Sci. Appl. https://doi.org/10.1016/0378-1096 4347(93)80314-T 1097

Heazlewood, J.L., Tonti-Filippini, J.S., Gout, A.M., Day, D.A., Whelan, J., Millar, A.H., 2004. 1098 Experimental Analysis of the Arabidopsis Mitochondrial Proteome Highlights Signaling 1099 and Regulatory Components, Provides Assessment of Targeting Prediction Programs, 1100 and Indicates Plant-Specific Mitochondrial Proteins. Plant Cell 16, 241–256. 1101 https://doi.org/10.1105/tpc.016055 1102

Hoeberichts, F.A., Vaeck, E., Kiddle, G., Coppens, E., Van De Cotte, B., Adamantidis, A., 1103 Ormenese, S., Foyer, C.H., Zabeau, M., Inzé, D., Périlleux, C., Van Breusegem, F., 1104 Vuylsteke, M., 2008. A temperature-sensitive mutation in the Arabidopsis thaliana 1105 phosphomannomutase gene disrupts protein glycosylation and triggers cell death. J. 1106 Biol. Chem. 283, 5708–5718. https://doi.org/10.1074/jbc.M704991200 1107

Hoops, S., Gauges, R., Lee, C., Pahle, J., Simus, N., Singhal, M., Xu, L., Mendes, P., 1108 Kummer, U., 2006. COPASI - A COmplex PAthway SImulator. Bioinformatics. 1109


https://doi.org/10.1101/2020.06.23.167247

https://doi.org/10.1093/bioinformatics/btl485 1110 Iwabuchi, K., Li, B., Bartel, P., Fields, S., 1993. Use of the two-hybrid system to identify the 1111

domain of p53 involved in oligomerization. Oncogene 8, 1693–6. 1112 Klepikova, A. V., Kasianov, A.S., Gerasimov, E.S., Logacheva, M.D., Penin, A.A., 2016. A 1113

high resolution map of the Arabidopsis thaliana developmental transcriptome based on 1114 RNA-seq profiling. Plant J. 88, 1058–1070. https://doi.org/10.1111/tpj.13312 1115

Laing, W., Norling, C., Brewster, D., Wright, M., Bulley, S., 2017. Ascorbate concentration in 1116 <em>Arabidopsis thaliana</em> and expression of ascorbate related genes 1117 using RNAseq in response to light and the diurnal cycle. bioRxiv 138008. 1118 https://doi.org/10.1101/138008 1119

Laing, W. A., Bulley, S., Wright, M., Cooney, J., Jensen, D., Barraclough, D., MacRae, E., 1120 2004. A highly specific L-galactose-1-phosphate phosphatase on the path to ascorbate 1121 biosynthesis. Proc. Natl. Acad. Sci. 101, 16976–16981. 1122 https://doi.org/10.1073/pnas.0407453101 1123

Laing, William A, Frearson, N., Bulley, S., MacRae, E., 2004. Kiwifruit L-galactose 1124 dehydrogenase: molecular, biochemical and physiological aspects of the enzyme. 1125 Funct. Plant Biol. 31, 1015–1025. https://doi.org/10.1071/FP04090 1126

Laing, W.A., Martínez-Sánchez, M., Wright, M.A., Bulley, S.M., Brewster, D., Dare, A.P., 1127 Rassam, M., Wang, D., Storey, R., Macknight, R.C., Hellens, R.P., 2015. An Upstream 1128 Open Reading Frame Is Essential for Feedback Regulation of Ascorbate Biosynthesis 1129 in Arabidopsis. Plant Cell. https://doi.org/10.1105/tpc.114.133777 1130

Laing, W.A., Wright, M.A., Cooney, J., Bulley, S.M., 2007. The missing step of the L-1131 galactose pathway of ascorbate biosynthesis in plants, an L-galactose 1132 guanyltransferase, increases leaf ascorbate content. Proc. Natl. Acad. Sci. 104, 9534–1133 9539. 1134

Li, B., Fields, S., 1993. Identification of mutations in p53 that affect its binding to SV40 large 1135 T antigen by using the yeast two‐hybrid system. FASEB J. 1136 https://doi.org/10.1096/fasebj.7.10.8344494 1137

Li, J., Liang, D., Li, M., Ma, F., 2013. Light and abiotic stresses regulate the expression of 1138 GDP-L-galactose phosphorylase and levels of ascorbic acid in two kiwifruit genotypes 1139 via light-responsive and stress-inducible cis-elements in their promoters. Planta 238, 1140 535–547. https://doi.org/10.1007/s00425-013-1915-z 1141

Li, S., Wang, J., Yu, Y., Wang, F., Dong, J., Huang, R., 2016. D27E mutation of VTC1 1142 impairs the interaction with CSN5B and enhances ascorbic acid biosynthesis and 1143 seedling growth in Arabidopsis. Plant Mol. Biol. 92, 473–482. 1144 https://doi.org/10.1007/s11103-016-0525-0 1145

Li, T., Yang, X., Yu, Y., Si, X., Zhai, X., Zhang, H., Dong, W., Gao, C., Xu, C., 2018. 1146 Domestication of wild tomato is accelerated by genome editing. Nat. Biotechnol. 1147 https://doi.org/10.1038/nbt.4273 1148

Li, X., Ye, J., Munir, S., Yang, T., Chen, W., Liu, G., Zheng, W., Zhang, Y., 2019. 1149 Biosynthetic Gene Pyramiding Leads to Ascorbate Accumulation with Enhanced 1150 Oxidative Stress Tolerance in Tomato. Int. J. Mol. Sci. 20, 1558. 1151 https://doi.org/10.3390/ijms20071558 1152

Lim, B., Smirnoff, N., Cobbett, C.S., Golz, J.F., 2016. Ascorbate-Deficient vtc2 Mutants in 1153 Arabidopsis Do Not Exhibit Decreased Growth. Front. Plant Sci. 7, 1–9. 1154 https://doi.org/10.3389/fpls.2016.01025 1155

Linster, C.L., Gomez, T.A., Christensen, K.C., Adler, L.N., Young, B.D., Brenner, C., Clarke, 1156 S.G., 2007. Arabidopsis VTC2 encodes a GDP-L-galactose phosphorylase, the last 1157 unknown enzyme in the smirnoff-wheeler pathway to ascorbic acid in plants. J. Biol. 1158 Chem. https://doi.org/10.1074/jbc.M702094200 1159

Lorence, A., Chevone, B.I., Mendes, P., Nessler, C.L., 2004. myo-Inositol Oxygenase Offers 1160 a Possible Entry Point into Plant Ascorbate Biosynthesis. Plant Physiol. 134, 1200–1161 1205. https://doi.org/10.1104/pp.103.033936 1162

Lukowitz, W., Nickle, T.C., Meinke, D.W., Last, R.L., Conklin, P.L., Somerville, C.R., 2001. 1163 Arabidopsis cyt1 mutants are deficient in a mannose-1-phosphate guanylyltransferase 1164


https://doi.org/10.1101/2020.06.23.167247

and point to a requirement of N-linked glycosylation for cellulose biosynthesis. Proc. 1165 Natl. Acad. Sci. 98, 2262–2267. https://doi.org/10.1073/pnas.051625798 1166

Marchler-Bauer, A., Bo, Y., Han, L., He, J., Lanczycki, C.J., Lu, S., Chitsaz, F., Derbyshire, 1167 M.K., Geer, R.C., Gonzales, N.R., Gwadz, M., Hurwitz, D.I., Lu, F., Marchler, G.H., 1168 Song, J.S., Thanki, N., Wang, Z., Yamashita, R.A., Zhang, D., Zheng, C., Geer, L.Y., 1169 Bryant, S.H., 2017. CDD/SPARCLE: functional classification of proteins via subfamily 1170 domain architectures. Nucleic Acids Res. 45, D200–D203. 1171 https://doi.org/10.1093/nar/gkw1129 1172

Marchler-Bauer, A., Bryant, S.H., 2004. CD-Search: protein domain annotations on the fly. 1173 Nucleic Acids Res. 32, W327–W331. https://doi.org/10.1093/nar/gkh454 1174

Marchler-Bauer, A., Derbyshire, M.K., Gonzales, N.R., Lu, S., Chitsaz, F., Geer, L.Y., Geer, 1175 R.C., He, J., Gwadz, M., Hurwitz, D.I., Lanczycki, C.J., Lu, F., Marchler, G.H., Song, 1176 J.S., Thanki, N., Wang, Z., Yamashita, R.A., Zhang, D., Zheng, C., Bryant, S.H., 2015. 1177 CDD: NCBI’s conserved domain database. Nucleic Acids Res. 1178 https://doi.org/10.1093/nar/gku1221 1179

Marchler-Bauer, A., Lu, S., Anderson, J.B., Chitsaz, F., Derbyshire, M.K., DeWeese-Scott, 1180 C., Fong, J.H., Geer, L.Y., Geer, R.C., Gonzales, N.R., Gwadz, M., Hurwitz, D.I., 1181 Jackson, J.D., Ke, Z., Lanczycki, C.J., Lu, F., Marchler, G.H., Mullokandov, M., 1182 Omelchenko, M. V., Robertson, C.L., Song, J.S., Thanki, N., Yamashita, R.A., Zhang, 1183 D., Zhang, N., Zheng, C., Bryant, S.H., 2011. CDD: a Conserved Domain Database for 1184 the functional annotation of proteins. Nucleic Acids Res. 39, D225–D229. 1185 https://doi.org/10.1093/nar/gkq1189 1186

Maruta, T., Yonemitsu, M., Yabuta, Y., Tamoi, M., Ishikawa, T., Shigeoka, S., 2008. 1187 Arabidopsis phosphomannose isomerase 1, but not phosphomannose isomerase 2, is 1188 essential for ascorbic acid biosynthesis. J. Biol. Chem. 283, 28842–28851. 1189 https://doi.org/10.1074/jbc.M805538200 1190

Mellidou, I., Chagne, D., Laing, W.A., Keulemans, J., Davey, M.W., 2012. Allelic Variation in 1191 Paralogs of GDP-l-Galactose Phosphorylase Is a Major Determinant of Vitamin C 1192 Concentrations in Apple Fruit. Plant Physiol. 160, 1613–1629. 1193 https://doi.org/10.1104/pp.112.203786 1194

Mieda, T., Yabuta, Y., Rapolu, M., Motoki, T., Takeda, T., Yoshimura, K., Ishikawa, T., 1195 Shigeoka, S., 2004. Feedback inhibition of spinach L-galactose dehydrogenase by L-1196 ascorbate. Plant Cell Physiol. 45, 1271–1279. https://doi.org/10.1093/pcp/pch152 1197

Müller-Moulé, P., 2008. An expression analysis of the ascorbate biosynthesis enzyme VTC2. 1198 Plant Mol. Biol. 68, 31–41. https://doi.org/10.1007/s11103-008-9350-4 1199

Nakagawa, T., Kurose, T., Hino, T., Tanaka, K., Kawamukai, M., Niwa, Y., Toyooka, K., 1200 Matsuoka, K., Jinbo, T., Kimura, T., 2007. Development of series of gateway binary 1201 vectors, pGWBs, for realizing efficient construction of fusion genes for plant 1202 transformation. J. Biosci. Bioeng. https://doi.org/10.1263/jbb.104.34 1203

Nourbakhsh, A., Collakova, E., Gillaspy, G.E., 2015. Characterization of the inositol 1204 monophosphatase gene family in Arabidopsis. Front. Plant Sci. 5, 1–14. 1205 https://doi.org/10.3389/fpls.2014.00725 1206

Østergaard, J., Persiau, G., Davey, M.W., Bauw, G., Van Montagu, M., 1997. Isolation of a 1207 cDNA coding for L-galactono-gamma-lactone dehydrogenase, an enzyme involved in 1208 the biosynthesis of ascorbic acid in plants. Purification, characterization, cDNA cloning, 1209 and expression in yeast. J. Biol. Chem. 272, 30009–16. 1210

Page, M., Sultana, N., Paszkiewicz, K., Florance, H., Smirnoff, N., 2012. The influence of 1211 ascorbate on anthocyanin accumulation during high light acclimation in Arabidopsis 1212 thaliana: further evidence for redox control of anthocyanin synthesis. Plant. Cell 1213 Environ. 35, 388–404. https://doi.org/10.1111/j.1365-3040.2011.02369.x 1214

Pallanca, J.E., Smirnoff, N., 2000. The control of ascorbic acid synthesis and turnover in pea 1215 seedlings. J. Exp. Bot. 51, 669–674. https://doi.org/10.1093/jexbot/51.345.669 1216

Pineau, B., Layoune, O., Danon, A., De Paepe, R., 2008. L-galactono-1,4-lactone 1217 dehydrogenase is required for the accumulation of plant respiratory complex I. J. Biol. 1218 Chem. 283, 32500–32505. https://doi.org/10.1074/jbc.M805320200 1219


https://doi.org/10.1101/2020.06.23.167247

Plumb, W., Townsend, A.J., Rasool, B., Alomrani, S., Razak, N., Karpinska, B., Ruban, A. 1220 V., Foyer, C.H., 2018. Ascorbate-mediated regulation of growth, photoprotection, and 1221 photoinhibition in Arabidopsis thaliana. J. Exp. Bot. https://doi.org/10.1093/jxb/ery170 1222

Qi, T., Liu, Z., Fan, M., Chen, Y., Tian, H., Wu, D., Gao, H., Ren, C., Song, S., Xie, D., 2017. 1223 GDP-D-mannose epimerase regulates male gametophyte development, plant growth 1224 and leaf senescence in Arabidopsis. Sci. Rep. 7, 10309. 1225 https://doi.org/10.1038/s41598-017-10765-5 1226

Reiter, W.D., Vanzin, G.F., 2001. Molecular genetics of nucleotide sugar interconversion 1227 pathways in plants. Plant Mol. Biol. 47, 95–113. 1228

Sawake, S., Tajima, N., Mortimer, J.C., Lao, J., Ishikawa, T., Yu, X., Yamanashi, Y., 1229 Yoshimi, Y., Kawai-Yamada, M., Dupree, P., Tsumuraya, Y., Kotake, T., 2015. 1230 KONJAC1 and 2 Are Key Factors for GDP-Mannose Generation and Affect l-Ascorbic 1231 Acid and Glucomannan Biosynthesis in Arabidopsis. Plant Cell. 1232 https://doi.org/10.1105/tpc.15.00379 1233

Schimmeyer, J., Bock, R., Meyer, E.H., 2016. L-Galactono-1,4-lactone dehydrogenase is an 1234 assembly factor of the membrane arm of mitochondrial complex I in Arabidopsis. Plant 1235 Mol. Biol. 90, 117–126. https://doi.org/10.1007/s11103-015-0400-4 1236

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., 1237 Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J.-Y., White, D.J., 1238 Hartenstein, V., Eliceiri, K., Tomancak, P., Cardona, A., 2012. Fiji: an open-source 1239 platform for biological-image analysis. Nat. Methods 9, 676–682. 1240 https://doi.org/10.1038/nmeth.2019 1241

Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Image to ImageJ: 25 years of 1242 image analysis. Nat. Methods 9, 671–675. https://doi.org/10.1038/nmeth.2089 1243

Smirnoff, N., 2019. Engineering of metabolic pathways using synthetic enzyme complexes. 1244 Plant Physiol. https://doi.org/10.1104/pp.18.01280 1245

Smirnoff, N., 2018. Ascorbic acid metabolism and functions: A comparison of plants and 1246 mammals. Free Radic. Biol. Med. 0–1. 1247 https://doi.org/10.1016/j.freeradbiomed.2018.03.033 1248

Sperschneider, J., Catanzariti, A.M., Deboer, K., Petre, B., Gardiner, D.M., Singh, K.B., 1249 Dodds, P.N., Taylor, J.M., 2017. LOCALIZER: Subcellular localization prediction of both 1250 plant and effector proteins in the plant cell. Sci. Rep. https://doi.org/10.1038/srep44598 1251

Sullivan, A., Purohit, P.K., Freese, N.H., Pasha, A., Esteban, E., Waese, J., Wu, A., Chen, 1252 M., Chin, C.Y., Song, R., Watharkar, S.R., Chan, A.P., Krishnakumar, V., Vaughn, 1253 M.W., Town, C., Loraine, A.E., Provart, N.J., 2019. An ‘<scp>eFP</scp> ‐Seq Browser’ 1254 for visualizing and exploring <scp>RNA</scp> sequencing data. Plant J. 100, 641–654. 1255 https://doi.org/10.1111/tpj.14468 1256

Sweetlove, L.J., Fernie, A.R., 2018. The role of dynamic enzyme assemblies and substrate 1257 channelling in metabolic regulation. Nat. Commun. 9, 2136. 1258 https://doi.org/10.1038/s41467-018-04543-8 1259

Terai, Y., Ueno, H., Ogawa, T., Sawa, Y., Miyagi, A., Kawai-Yamada, M., Ishikawa, T., 1260 Maruta, T., 2020. Dehydroascorbate Reductases and Glutathione Set a Threshold for 1261 High-Light-Induced Ascorbate Accumulation. Plant Physiol. 1262 https://doi.org/10.1104/pp.19.01556 1263

Torabinejad, J., Donahue, J.L., Gunesekera, B.N., Allen-Daniels, M.J., Gillaspy, G.E., 2009. 1264 VTC4 Is a Bifunctional Enzyme That Affects Myoinositol and Ascorbate Biosynthesis in 1265 Plants. Plant Physiol. 150, 951–961. https://doi.org/10.1104/pp.108.135129 1266

Truffault, V., Fry, S.C., Stevens, R.G., Edinburgh, T., Wall, C., Building, D.R., Gautier, H., 1267 Edinburgh, T., Wall, C., Building, D.R., 2017. Ascorbate degradation in tomato leads to 1268 accumulation of oxalate , threonate and oxalyl threonate. Plant J. 89, 996–1008. 1269 https://doi.org/10.1111/tpj.13439 1270

Urzica, E.I., Adler, L.N., Page, M.D., Linster, C.L., Arbing, M.A., Casero, D., Pellegrini, M., 1271 Merchant, S.S., Clarke, S.G., 2012. Impact of oxidative stress on ascorbate 1272 biosynthesis in Chlamydomonas via regulation of the VTC2 gene encoding a GDP-L-1273 galactose phosphorylase. J. Biol. Chem. 287, 14234–14245. 1274


https://doi.org/10.1101/2020.06.23.167247

https://doi.org/10.1074/jbc.M112.341982 1275 Voxeur, A., Gilbert, L., Rihouey, C., Driouich, A., Rothan, C., Baldet, P., Lerouge, P., 2011. 1276

Silencing of the GDP-D-mannose 3,5-epimerase affects the structure and cross-linking 1277 of the pectic polysaccharide rhamnogalacturonan II and plant growth in tomato. J. Biol. 1278 Chem. 286, 8014–8020. https://doi.org/10.1074/jbc.M110.198614 1279

Wang, J., Yu, Y., Zhang, Z., Quan, R., Zhang, H., Ma, L., Deng, X.W., Huang, R., 2013. 1280 Arabidopsis CSN5B Interacts with VTC1 and Modulates Ascorbic Acid Synthesis. Plant 1281 Cell. https://doi.org/10.1105/tpc.112.106880 1282

Wang, H. Sen, Yu, C., Zhu, Z.J., Yu, X.C., 2011. Overexpression in tobacco of a tomato 1283 GMPase gene improves tolerance to both low and high temperature stress by 1284 enhancing antioxidation capacity. Plant Cell Rep. 30, 1029–1040. 1285 https://doi.org/10.1007/s00299-011-1009-y 1286

Waszczak, C., Carmody, M., Kangasjärvi, J., 2018. Reactive Oxygen Species in Plant 1287 Signaling. Annu. Rev. Plant Biol. 69, 209–236. https://doi.org/10.1007/978-3-642-1288 00390-5 1289

Wheeler, G., Ishikawa, T., Pornsaksit, V., Smirnoff, N., 2015. Evolution of alternative 1290 biosynthetic pathways for vitamin C following plastid acquisition in photosynthetic 1291 eukaryotes. Elife 2015, 1–14. https://doi.org/10.7554/eLife.06369 1292

Wheeler, G.L., Jones, M.A., Smirnoff, N., 1998. The biosynthetic pathway of vitamin C in 1293 higher plants. Nature 393, 365–369. https://doi.org/10.1038/30728 1294

Wolucka, B.A., Van Montagu, M., 2003. GDP-Mannose 3′,5′-Epimerase Forms GDP-L-1295 gulose, a Putative Intermediate for the de Novo Biosynthesis of Vitamin C in Plants. J. 1296 Biol. Chem. 278, 47483–47490. https://doi.org/10.1074/jbc.M309135200 1297

Yabuta, Y., Mieda, T., Rapolu, M., Nakamura, A., Motoki, T., Maruta, T., Yoshimura, K., 1298 Ishikawa, T., Shigeoka, S., 2007. Light regulation of ascorbate biosynthesis is 1299 dependent on the photosynthetic electron transport chain but independent of sugars in 1300 Arabidopsis. J. Exp. Bot. 58, 2661–2671. https://doi.org/10.1093/jxb/erm124 1301

Ye, Q., Worman, H.J., 1995. Protein-protein interactions between human nuclear lamins 1302 expressed in yeast. Exp. Cell Res. https://doi.org/10.1006/excr.1995.1230 1303

Yoshimura, K., Nakane, T., Kume, S., Shiomi, Y., Maruta, T., Ishikawa, T., Shigeoka, S., 1304 2014. Transient expression analysis revealed the importance of VTC2 expression level 1305 in light/dark regulation of ascorbate biosynthesis in Arabidopsis 78, 60–6. 1306 https://doi.org/10.1080/09168451.2014.877831 1307

Zhang, C., Liu, J., Zhang, Y., Cai, X., Gong, P., Zhang, J., Wang, T., Li, H., Ye, Z., 2011. 1308 Overexpression of SlGMEs leads to ascorbate accumulation with enhanced oxidative 1309 stress, cold, and salt tolerance in tomato. Plant Cell Rep. 30, 389–398. 1310 https://doi.org/10.1007/s00299-010-0939-0 1311

Zhang, G.Y., Liu, R.R., Zhang, C.Q., Tang, K.X., Sun, M.F., Yan, G.H., Liu, Q.Q., 2015. 1312 Manipulation of the rice L-galactose pathway: Evaluation of the effects of transgene 1313 overexpression on ascorbate accumulation and abiotic stress tolerance. PLoS One. 1314 https://doi.org/10.1371/journal.pone.0125870 1315

Zhang, H., Si, X., Ji, X., Fan, R., Liu, J., Chen, K., Wang, D., Gao, C., 2018. Genome editing 1316 of upstream open reading frames enables translational control in plants. Nat. 1317 Biotechnol. 36, 894–898. https://doi.org/10.1038/nbt.4202 1318

Zhang, Y., Beard, K.F.M., Swart, C., Bergmann, S., Krahnert, I., Nikoloski, Z., Graf, A., 1319 George Ratcliffe, R., Sweetlove, L.J., Fernie, A.R., Obata, T., 2017. Protein-protein 1320 interactions and metabolite channelling in the plant tricarboxylic acid cycle. Nat. 1321 Commun. 8. https://doi.org/10.1038/ncomms15212 1322

Zhang, Y., Butelli, E., Martin, C., 2014. Engineering anthocyanin biosynthesis in plants. Curr. 1323 Opin. Plant Biol. https://doi.org/10.1016/j.pbi.2014.05.011 1324

Zhou, Y., Tao, Q.C., Wang, Z.N., Fan, R., Li, Y., Sun, X.F., Tang, K.X., 2012. Engineering 1325 ascorbic acid biosynthetic pathway in Arabidopsis leaves by single and double gene 1326 transformation. Biol. Plant. 56, 451–457. 1327

1328

1329


https://doi.org/10.1101/2020.06.23.167247

1330

1331

1332

1333

1334

1335

1336

1337

1338

1339

1340

1341

1342

1343

1344

1345

1346 1347

Figure 1. The ascorbate biosynthesis via GDP-mannose and L-galactose (the Smirnoff-Wheeler [SW] pathway) in Arabidopsis thaliana. Yellow area 1348 refers to the cytosolic enzymes used in this work. mOM: mitochondrion outer membrane, mIMS: mitochondrion intermembrane space, mIM: 1349 mitochondrion inner membrane. In the left-hand side of the arrows, the genes encoding each enzyme are displayed while in the right-hand side the 1350 encoded proteins/enzymatic activities are shown. Glc=glucose, Fru=fructose, Man=mannose, Gal=galactose, GalL=galactono-1,4-lactone, Asc= 1351 ascorbic acid. PGI=phosphoglucose isomerase, PMI=phosphomannose isomerase, PMM=phosphomannomutase, GMP=GDP-Mannose 1352 pyrophosphorylase, GME=GDP-mannose 3’,5’ epimerase, GGP= GDP-L-galactose phosphorylase, GPP= L-galactose 1-phosphate phosphatase, L-1353 GalDH= L-galactose dehydrogenase, L-GalLDH= L-galactono-1,4-lactone dehydrogenase. CytRED=cytochrome c reduced, CytOX=cytochrome c 1354 oxidized. 1355


https://doi.org/10.1101/2020.06.23.167247

1356

1357

1358

1359

1360

1361

1362

1363

1364 1365

Figure 2. Characterization of C-terminal GFP fusions of ascorbate biosynthesis enzymes. A) Ascorbate complementation assay rescues growth arrest 1366 from gldh and lgaldh mutants but not gme lethality. In small dashed circles, homozygous mutants show arrested growth in media lacking ascorbate, 1367 but not in media supplemented with ascorbate. B) Complementation of ascorbate concentration by expression of GFP fusion proteins in ascorbate 1368 deficient Arabidopsis mutants. Different letters denote statistically significant differences for ascorbate concentration (One-Way ANOVA, α=0,05). 1369 Ascorbate was extracted from three independent samples composed of three 6-weeks old fully expanded rosettes grown in short day regime at 150 1370 µmol photons m-2 s-1, one and a half hours after lights turned on. C) Immunoblot (α-GFP) of complemented lines show that little amount of protein 1371 VTC2 compared other components restores ascorbate content. Red arrowhead indicates VTC2-GFP. A representative sample for each genotype is 1372 shown out of the three tested by immunoblot (not shown). 1373

1374


https://doi.org/10.1101/2020.06.23.167247

1375

1376

1377

1378

1379 1380

Figure 3. The effect of transient expression of C-terminal GFP and HA fusions of ascorbate biosynthesis 1381

enzymes on ascorbate concentration in Nicotiana benthamiana leaves. A) Strategy followed to clone and to 1382

overexpress in N. benthamiana leaves ascorbate biosynthesis genes from Arabidopsis translationally fused 1383

to GFP and HA at the protein C-terminus. B) Immunoblots (α-GFP) of fusion protein accumulation 2, 3 and 4 1384

days after agroinfiltration. Expression was driven by the 35S promoter. Protein amount was measured and 1385

loaded to obtain similar intensities in the blot at 2 days after infiltration. C-H) Leaf ascorbate concentration 3 1386

days after agroinfiltration with different combinations of C-terminal GFP ad HA fusions of ascorbate 1387

biosynthesis enzymes. Constructs used in each infiltration are shown in Supplementary Table 1. Two leaves 1388

of at least three N. benthamiana plants were infiltrated and samples were collected at 3 days after infiltration. 1389

1390


https://doi.org/10.1101/2020.06.23.167247

1391

. 1392

1393

1394

1395

1396

1397

1398

1399

1400

1401

1402

1403

1404

1405

1406

1407

1408

1409

1410

1411

1412

1413

1414

1415

1416

1417

1418

1419

1420

1421

1422

1423

1424

1425

1426

1427

1428

1429

1430

1431

1432

1433

1434

Figure 4. Subcellular localization of free GFP and

C-terminal GFP fusions of ascorbate biosynthesis

enzymes transiently expressed in Nicotiana

benthamiana leaves under the control of the 35S

promoter 3 days after agroinfiltration. GFP was

visualized by laser scanning confocal microscopy.

Scale bar = 30 µm.


https://doi.org/10.1101/2020.06.23.167247

1435

1436 1437

Figure 5. Subcellular localization of ascorbate biosynthesis enzymes fused to GFP at the C-terminus and expressed under the control of their native 1438 promoters in Arabidopsis. A) The expression of GFP-tagged enzymes is ubiquitous in four days-old seedlings and their subcellular localization is 1439 compatible with cytosol and nucleus except for GME, which only locates in cytosol. B) GFP signal of a functional VTC2-GFP (line 13) protein was not 1440 detectable. GFP was visualized by laser scanning confocal microscopy. Scale bar = 30 µm. 1441

1442


https://doi.org/10.1101/2020.06.23.167247

1443

1444 1445

Figure 6. Interaction of ascorbate biosynthesis enzymes assessed by yeast two-hybrid analysis and coimmunoprecipitation (CoIP). (A) Yeast two-1446 hybrid assay detected no direct interaction between the enzymes, but VTC1, GME and VTC4, which contain predicted dimerization/trimerization 1447 domains (B), form dimers/trimers. B) Protein scheme for each of the enzymes assayed in yeast two-hybrid showing length (in amino acids), annotated 1448 domains and predicted dimer/trimerization interfaces. (C-G) CoIP assay reported association in vivo in Nicotiana between consecutive steps as well 1449 as between the first (VTC1) and the last (L-GalDH) steps occurring in cytosol. GFP-tagged enzyme from Nicotiana crude protein extracts containing 1450 two overexpressed consecutive enzymes, GFP and HA-tagged, was pulled-down using α-GFP agarose beads and coimmunoprecipitated protein was 1451 detected using α-HA antibody. 1452

1453

Figure 6. 1454


https://doi.org/10.1101/2020.06.23.167247

1455

1456

1457

1458

1459

1460

1461 1462 Figure 7. A kinetic model of ascorbate biosynthesis and turnover. A) Scheme of the pathway modelled in COPASI showing intermediates catalytic 1463 steps (arrows) and feedback (red lines). The initial model parameters are shown in Supplementary Table 3. B) Metabolic control analysis at steady 1464 state was used to calculate flux control coefficients for each enzyme with various strengths of non-competitive inhibition exerted by ascorbate 1465 (indicated by Ki values) on GGP (step 6). C) Metabolic control analysis at steady state was used to calculate concentration control coefficients for 1466 ascorbate with various strengths of non-competitive inhibition exerted by ascorbate (indicated by Ki values) on GGP (step 6). D) The effect of varying 1467 enzyme activity on ascorbate concentration with various strengths of non-competitive inhibition exerted by ascorbate (indicated by Ki values) on GGP 1468 (step 6). E) Time course of the change in ascorbate concentration in response to the addition of D-mannose 6-phosphate (D-Man 6-P), L-galactose (L-1469 Gal) and L-galactono-1,4-lactone (L-GalL). 1: PGI, 2: PMI, 3: PMM, 4: GMP, 5: GME, 6: GGP, 7: GPP, 8: L-GalDH, 9: L-GalLDH, 10: Turnover, 11: 1470 Oxidation, 12: Reduction). 1471

1472


https://doi.org/10.1101/2020.06.23.167247

1473

1474

1475

1476

1477

1478 1479 Supplementary Figure 1. Gene structure, primers disposition and mutants’ T-DNA location. 1480

1481


https://doi.org/10.1101/2020.06.23.167247

1482 1483 Supplementary Figure 2. Diagnostic PCR of the gme, lgaldh and gldh seedlings analyzed in the complementation assay described in Figure 2A. 1484 Whereas ascorbate does not recover the lethality of gme mutation (A), it does recover lgaldh (C). As a control, we used gldh (B), previously published 1485 to be recovered by ascorbate (Pineau et al., 2008). 1486

1487


https://doi.org/10.1101/2020.06.23.167247

1488

1489

1490

1491

1492

1493

1494 1495 Supplementary Figure 3. Generation of Arabidopsis transgenic lines containing ascorbate biosynthesis enzymes fused to GFP. A) Immunoblot (α-1496 GFP) analysis of fourteen GME-GFP T2 lines identifies several lines expressing different amounts of GME-GFP protein (background WT). After the 1497 identification of lines harboring single insertions (T2 HygR:HygS=3, χ² (α=0.05)), we selected line L6 (arrowhead) to introduce the GMEp:GME-GFP 1498 construct into the gme mutant (SALK_150208) background using reciprocal crosses. B) Diagnostic PCR of F2 generation from the cross indicated 1499 identified plants harboring GME-GFP in homozygous mutant background (arrowheads). Ascorbate content of this line is shown in Figure 2. C) 1500 Immunoblot (α-GFP) analysis of VTC2-GFP of two lines among which ascorbate content ranges. Four days old seedlings were grown under long day 1501 conditions in solid half-strength MS agar plates. D) Immunoblot (α-GFP) analysis of L-GalDH-GFP T2 generation showing the six lines expressing L-1502 GalDH-GFP protein (background WT). After the identification of lines harboring single insertions (T2 HygR:HygS=3, χ² (α=0.05)), we selected line L1 1503 to introduce the L-GalDHp:L-GalDH-GFP construct into the lgaldh mutant (SALK_056664) background using reciprocal crosses. H) Diagnostic PCR 1504 of F2 generation from the cross indicated identified plants harboring L-GalDH-GFP in homozygous mutant background (arrowheads). Ascorbate 1505 content of this line is showed in Figure 2. E) Ascorbate content of the T3 generation of lines generated consisting of vtc2 mutants harboring 1506 VTC2p:VTC2-GFP. Three fully expanded leaves were collected on five-week-old T3 plants grown on soil in a growth chamber under short-day 1507 conditions (10 h light at 22°C/14 h dark at 19°C), at 60% humidity and 200 µmol m–2 s–1 light intensity. Different letters denote statistically significant 1508 differences for ascorbate content (One-Way ANOVA, α=0,05). G) Immunoblot (α-GFP) analysis of VTC4-GFP T2 generation showing the six lines 1509 expressing VTC4-GFP protein (background WT). After the identification of lines harboring single insertions (T2 HygR:HygS=3, χ² (α=0.05)), we selected 1510 line L5 to introduce the VTC4p:VTC4-GFP construct into the vtc4 mutant (SALK_077222) background using reciprocal crosses. H) Diagnostic PCR 1511 of F2 generation from the cross indicated identified plants harboring VTC4-GFP in homozygous mutant background (arrowheads). Ascorbate content 1512 of this line is showed in Figure 2. 1513

1514


https://doi.org/10.1101/2020.06.23.167247

1515

1516

1517

1518

1519

1520

1521

1522

1523

1524

1525 1526 Supplementary Figure 4. VTC2 mRNA levels are similar to those of VTC1 and GME, although VTC2 protein is hardly detectable. Data were taken 1527 from two different public RNA sequencing (RNA-seq) data sets: Transcriptome Variation Analysis (TRAVA, travadb.org; Klepikova et al., 2016) and 1528 the eFP-Seq Browser (https://bar.utoronto.ca/eFP-Seq_Browser/; Sullivan et al., 2019) corresponding to different adult plants’ vegetative and green 1529 tissues, respectively. 1530

1531

1532


https://doi.org/10.1101/2020.06.23.167247

1533

1534

1535

1536

1537

1538

1539

1540

1541

1542

1543

1544

1545

1546 1547 Supplementary Figure 5. Complete blots corresponding to Figure 2B (red rectangle). Blots A, B and C have the same time of exposure using an ultra-1548 sensitive enhanced chemiluminescent substrate for low femtogram protein level detection. Blot C and D are the same membrane with different 1549 exposure times. Protein loads were calculated using FIJI from a first immunoblot (not shown) and diluted so the expression of GFP-tagged proteins 1550 in the first timepoint is similar among constructs. 1551

1552


https://doi.org/10.1101/2020.06.23.167247

1553

1554

1555

1556

1557

1558

1559

1560

1561

1562

1563

1564

1565

1566

1567

1568 1569 Supplementary Figure 6. Predicted subcellular localization of ascorbate biosynthesis proteins in the plant cell according to COMPARTMENTS 1570 (https://compartments.jensenlab.org/Search; Binder et al., 2014). 1571

1572


https://doi.org/10.1101/2020.06.23.167247

1573 1574 Supplementary Figure 7. Yeast two-hybrid whole screening. Red rectangles correspond to the spotting presented in Figure 6A. As positive controls, 1575 we used SNF1/SNF4 (Fields and Song, 1989) and p53/SV40-Large AgT (AgT; Iwabuchi et al., 1993; Li and Fields, 1993) interactions. As a negative 1576 control, we included LamC/AgT cotransformation (Bartel et al., 1993; Ye and Worman, 1995). His: histidine, 3-AT: 3-amino-1,2,4-triazole, Ade: 1577 adenine. 1578

1579


https://doi.org/10.1101/2020.06.23.167247

1580 1581 Supplementary Figure 8. Immunoblot (α-GFP) of N. benthamiana leaves crude extracts following agroinfiltration of above-indicated combinations of 1582 proteins (also see Appendix 1). Two leaves of 2 or 3 independent plants were infiltrated, using GFP or TTL3-HA (an HA-tagged protein non-related 1583 to ascorbate synthesis) to rise optical density up to 1. 1584

1585


https://doi.org/10.1101/2020.06.23.167247

Documents

Organization and control of the ascorbate biosynthesis ......Jun 25, 2020 · 114 the cytosol while L-galactono-1,4-lactone crosses the outer membrane of the mitochondria to 115 be