plants Article Cellular and Subcellular Compartmentation of the 2C-Methyl-D-Erythritol 4-Phosphate Pathway in the Madagascar Periwinkle Grégory Guirimand 1,2,† , Anthony Guihur 1,3 , Catalina Perello 4 , Michael Phillips 5 , Samira Mahroug 1,6 , Audrey Oudin 1 , Thomas Dugé de Bernonville 1 , Sébastien Besseau 1 , Arnaud Lanoue 1 , Nathalie Giglioli-Guivarc’h 1 , Nicolas Papon 7 , Benoit St-Pierre 1 , Manuel Rodríguez-Concepcíon 4 , Vincent Burlat 8 and Vincent Courdavault 1, * 1 Biomolécules et Biotechnologies Végétales, EA 2106, Département of Agronomie, productions animale et végétale et agro-alimentaire, Université de Tours, 31 avenue Monge, 37200 Tours, France; gregory.guirimand@univ-tours.fr (G.G.); anthony.guihur@unil.ch (A.G.); samiramahroug@yahoo.fr (S.M.); audrey.oudin@univ-tours.fr (A.O.); thomas.duge@univ-tours.fr (T.D.d.B.); sebastien.besseau@univ-tours.fr (S.B.); lanoue@univ-tours.fr (A.L.); nathalie.guivarch@univ-tours.fr (N.G.-G.); benoit.stpierre@univ-tours.fr (B.S.-P.) 2 Graduate School of Science, Technology & Innovation, Kobe University, Kobe 657-8501, Japan 3 Department of Plant Molecular Biology, Faculty of Biology and Medicine, University of Lausanne, 1007 Lausanne, Switzerland 4 Program of Plant Metabolism and Metabolic Engineering, Centre for Research in Agricultural Genomics (CRAG) CSIC-IRTA-UAB-UB, Campus UAB Bellaterra, 08193 Barcelona, Spain; catalina.perello@gmail.com (C.P.); manuel.rodriguez@cragenomica.es (M.R.-C.) 5 Department of Biology, University of Toronto–Mississauga, Mississauga, 3359 Mississauga Road, ON L5L 1C6, Canada; michaelandrew.phillips@utoronto.ca 6 Department of Environment Sciences, University of Sidi-Bel-Abbes, 22000 Sidi Bel Abbès, Algeria 7 Groupe d’Etude des Interactions Hôte-Pathogène (GEIHP, EA 3142), SFR ICAT 4208, Université d’Angers, UNIV. Brest, F-49333 Angers, France; nicolas.papon@univ-angers.fr 8 Laboratoire de Recherche en Sciences Végétales, Université de Toulouse, CNRS, UPS, 24 chemin de Borde Rouge, Auzeville, BP42617, 31326 Castanet Tolosan, France; burlat@lrsv.ups-tlse.fr * Correspondence: vincent.courdavault@univ-tours.fr † LE STUDIUM RESEARCH FELLOW, Loire Valley Institute for Advanced Studies.  Received: 10 March 2020; Accepted: 1 April 2020; Published: 7 April 2020  Abstract: The Madagascar periwinkle (Catharanthus roseus) synthesizes the highly valuable monoterpene indole alkaloids (MIAs) through a long metabolic route initiated by the 2C-methyl-D-erythritol 4-phosphate (MEP) pathway. In leaves, a complex compartmentation of the MIA biosynthetic pathway occurs at both the cellular and subcellular levels, notably for some gene products of the MEP pathway. To get a complete overview of the pathway organization, we cloned four genes encoding missing enzymes involved in the MEP pathway before conducting a systematic analysis of transcript distribution and protein subcellular localization. RNA in situ hybridization revealed that all MEP pathway genes were coordinately and mainly expressed in internal phloem-associated parenchyma of young leaves, reinforcing the role of this tissue in MIA biosynthesis. At the subcellular level, transient cell transformation and expression of fluorescent protein fusions showed that all MEP pathway enzymes were targeted to plastids. Surprisingly, two isoforms of 1-deoxy-D-xylulose 5-phosphate synthase and 1-deoxy-D-xylulose 5-phosphate reductoisomerase initially exhibited an artifactual aggregated pattern of localization due to high protein accumulation. Immunogold combined with transmission electron microscopy, transient transformations performed with a low amount of transforming DNA and fusion/deletion experiments established that both enzymes were rather diffuse in stroma and stromules of plastids as also observed for the last six enzymes of the pathway. Taken together, these results provide new insights into a Plants 2020, 9, 462; doi:10.3390/plants9040462 www.mdpi.com/journal/plants Plants 2020, 9, 462 2 of 25 potential role of stromules in enhancing MIA precursor exchange with other cell compartments to favor metabolic fluxes towards the MIA biosynthesis. Keywords: 2C-methyl-D-erythritol 4-phosphate pathway; compartmentation; Catharanthus roseus; stromules; GFP imaging; localization artifact; alkaloids 1. Introduction Isoprenoids, also called terpenoids, constitute a large and highly diverse family of natural products, comprising circa 30,000 distinct compounds in plants. Besides being essential components of the primary metabolism acting as photosynthetic pigments, hormones or redox cofactors, isoprenoids also include specialized metabolites involved in pollinator attraction as well as plant defense against pathogens and herbivores [1,2] Among the defensive terpenoids, plants from the Apocynaceae, Nyssaceae, Loganiaceae, and Rubiaceae families especially synthesize monoterpene indole alkaloids (MIAs) for which more than 2500 distinct structures have been described so far [3,4]. For instance, the Madagascar periwinkle (Catharanthus roseus) accumulates more than 130 different MIAs, which limit organ consumption or cause aggressor intoxication upon feeding [5–7]. This cytotoxicity also confers highly valuable pharmaceutical properties to several prominent MIAs, such as the antineoplastic vinblastine and vincristine accumulated in C. roseus. The high economic value of these two compounds has thus prompted many research groups to elucidate MIA biosynthetic pathways in Apocynaceae, making C. roseus a model nonmodel plant [8]. This ultimately led to the characterization of the last steps of vindoline and catharanthine synthesis, the two precursors of vinblastine and vincristine [9–13]. These recent works represent essential advances with a view toward considering the biotechnological production of some interesting MIAs by heterologous hosts [14–16]. MIAs exhibit a complex biosynthetic pathway with almost all MIA being derived from strictosidine, which is considered as the first MIA [17]. While multiple decorations and cyclizations of strictosidine yield the plethora of MIAs, its synthesis always starts with the condensation of the indole precursor tryptamine with the monoterpene secoiridoid precursor secologanin. Tryptamine is a shikimate-derived product generated through a single enzymatic step of tryptophan decarboxylation [18]. Secologanin biosynthesis is a more intricate process relying on a nine-step conversion of geranyl diphosphate (GPP), named monoterpene seco-iridoid (MSI) pathway and involving distinct enzyme isoforms [19,20]. Interestingly, the GPP pool engaged in MIA synthesis exclusively comes from the 2C-methyl-D-erythritol 4-phosphate (MEP) pathway [21]. The MEP pathway produces both isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) through seven enzymatic reactions initiated by the synthesis of 1-deoxy-D-xylulose 5-phosphate (DXP) from pyruvate and glyceraldehyde 3-phosphate (Figure 1). This reaction is catalyzed by DXP synthase (DXS), encoded by a small gene family in higher plants [22–24]. The DXS isogenes have been clustered into two related gene groups: Clade I-DXS including housekeeping genes [25] and Clade II-DXS including genes associated with plant defense and secondary metabolism [23,26]. DXP is then sequentially converted into MEP by DXP reductoisomerase (DXR) and into 4-(cytidine 50 diphospho)-2-C-methyl-D-erythritol (CDP-ME) following the addition of cytidine triphosphate by CDP-ME synthase (CMS). Plants 2020, Plants 9, x462 2020, 9, FOR PEER REVIEW 3 3ofof24 25 Figure 1. The 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway yielding isopentenyl diphosphate Figure 1. The 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway yielding isopentenyl diphosphate (IPP) and its allylic isomer dimethylallyl diphosphate (DMAPP). DXS, 1-deoxy-D-xylulose 5-phosphate (IPP) and its allylic isomer dimethylallyl diphosphate (DMAPP). DXS, 1-deoxy-D-xylulose 5- synthase; DXR, 1-deoxy-D-xylulose 5-phosphate reductoisomerase; CMS, 2-C-methyl-D-erythritol phosphate synthase; DXR, 1-deoxy-D-xylulose0 5-phosphate reductoisomerase; CMS, 2-C-methyl-D- 4-phosphate synthase; CMK, 4-(cytidine 5 diphospho)-2-C-methyl-D-erythritol kinase; MECS, erythritol 4-phosphate synthase; CMK, 4-(cytidine 5′diphospho)-2-C-methyl-D-erythritol kinase; 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; HDS, (E)-4-hydroxy-3-methylbut-2-enyl MECS, 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; HDS, (E)-4-hydroxy-3-methylbut-2- diphosphate synthase; HDR, (E)-4-hydroxy-3-methylbut-2-enyl diphosphate reductase; IDI, isopentenyl enyl diphosphate synthase; HDR, (E)-4-hydroxy-3-methylbut-2-enyl diphosphate reductase; IDI, diphosphate isomerase. isopentenyl diphosphate isomerase. DXS2.1 and DXS2.2 correspond to class II DXS isoforms. cDNA characterized in this work are underlined. This metabolic intermediate is phosphorylated by CDP-ME kinase (CMK) to form CDP- Plants 2020, 9, 462 4 of 25 DXS2.1 and DXS2.2 correspond to class II DXS isoforms. cDNA characterized in this work are underlined. This metabolic intermediate is phosphorylated by CDP-ME kinase (CMK) to form CDP-ME 2-phosphate and cyclized after the loss of the cytidyl group, yielding 2-C-methyl-D-erythritol 2,4-cyclodiphosphate (ME-cPP) in a reaction catalyzed by ME-cPP synthase (MECS). The penultimate step of the MEP pathway relies on the synthesis of (E)-4-hydroxy-3-methylbut-2-enyl diphosphate (HMBPP) catalyzed by HMBPP synthase (HDS) while the final step, catalyzed by HMBPP reductase (HDR), involves the reduction of HMBPP to give a 6:1 mixture of IPP and DMAPP [27]. This unbalanced ratio implicates an additional reaction of IPP isomerization catalyzed by type I IPP isomerase (IDI) to optimize the relative amount of the precursors and to fulfill an efficient biosynthesis of isoprenoids [28]. IPP and DMAPP are then condensed to form GPP that is converted into geraniol to initiate the MSI pathway [29]. In C. roseus, the MIA biosynthetic pathway also exhibits complex compartmentation involving no less than four different cell types and eight distinct subcellular compartments [30,31]. Interestingly, all the MEP pathway genes characterized so far, up to the antepenultimate step of the MSI pathway, display a cellular coexpression restricted to cells of the internal phloem associated parenchyma (IPAP) of leaves [19,29,32–35]. This specific localization thus involves the translocation of the resulting biosynthetic intermediate towards leaf epidermis that houses the subsequent reactions of the pathway. Such a process may represent a regulatory mechanism and a limiting step in the control of the metabolic flux towards MIAs [30]. At the subcellular level, while plastid localization of many MEP pathway enzymes has been described in numerous plants, a plastid targeting has been reported for only HDS and IDI from C. roseus [34–36]. Interestingly, in addition to plastid stroma localization, these studies also pointed out the localization of both enzymes to stromules. These long protrusions budding from plastids are in close proximity with other organelles such as the endoplasmic reticulum (ER), increasing the putative exchange surface and thus may facilitate transportation of molecules from the main plastid body to other organelles including the ER [37,38]. Additionally, IDI is also targeted to mitochondria and peroxisomes, highlighting the triple targeting of this enzyme [34]. Importantly, while many corroborative data have been gathered over the years, this view of the MEP pathway organization in C. roseus remains largely incomplete as less than half of the corresponding genes and proteins has been characterized to date. This prompted us to achieve the cloning and characterization of the missing MEP genes to describe the overall compartmentation of this pathway in C. roseus. We thus first focused on the cloning and the functional characterization of genes encoding CMS, CMK, HDR and possible DXS isoforms. Then, we provided a comprehensive overview of both cellular and subcellular distributions of all MEP pathway gene products except IDI that has been thoroughly characterized using the same sample sets in a previous work [34]. 2. Results 2.1. Cloning and Functional Validation of cDNAs Encoding a Second DXS Isoform, CMS, CMK and HDR Cloning of cDNAs encoding the missing MEP pathway genes (CMS, CMK, HDR) was initiated in the late 2000s before the massive release of C. roseus transcriptomic data. Partial cDNAs encoding each enzyme were thus amplified from a C. roseus cDNA mixture using degenerate primers. The corresponding full-length coding sequences were subsequently isolated from an orientated C. roseus cell cDNA library with specific primers. A similar approach was also used to isolate the second isoform of DXS and phylogenetic studies revealed that both isoforms belong to clade II DXS and were thus named DXS2.1 (initially named DXS) and DXS2.2 (Figure S1; [22,24]). Analysis of the neopublished C. roseus transcriptomic data confirmed the existence of all the cloned coding sequences. The deduced amino acid sequences of C. roseus, CMS, CMK, HDR and DXS2.2, display a high degree of identity with the corresponding A. thaliana orthologs as shown in Table 1. In addition, DXS2.2 exhibits 69% amino acid identity with the previously cloned DXS/DXS2.1. Putative transit peptides (TP) were identified at the N-terminal end of newly characterized enzymes as well as in previously identified Plants 2020, 9, 462 5 of 25 ones (Table 1). Depending on the software used for the predictions, either plastid- or mitochondrion-TP were predicted, even though TargetP predictions were always in favor of a plastid targeting. Table 1. Identity of the C. roseus MEP pathway enzymes excluding transit peptides with the corresponding A. thaliana orthologs and prediction of subcellular localization of the full sequences. Enzyme Identity PSORT Predotar ChloroP MitoProt TargetP 74% ER (membrane) 0.550 P 0.45 Y (0.509) 0.636 cTP (0.375) DXS/DXS2.1 64% Peroxisome 0.102 M 0.01 35 residues 37 residues mTP (0.217) 41% ER (lumen) 0.100 E 0.51 65% P (stroma) 0.570 P 0.91 Y (0.565) 0.327 cTP (0.849) DXS2.2 56% M (matrix) 0.475 M 0.06 62 residues 16 residues mTP (0.301) 34% Cyto 0.450 E 0.08 P (stroma) 0.520 P 0.30 Y (0.509) 0.596 cTP (0.303) DXR 89% Cyto 0.450 M 0.01 83 residues 66 residues mTP (0.049) M (matrix) 0.360 E 0.69 P (stroma) 0.865 P 0.07 0.5259 Y (0.576) cTP (0.932) CMS 76% P (thy mb) 0.515 M 0.20 Not 66 residues mTP (0.083) P (thy space) 0.461 E 0.71 predictable P (stroma) 0.899 P 0.84 Y (0.582) 0.949 cTP (0.895) CMK 75% P (thy mb) 0.595 M 0.01 57 residues 59 residues mTP (0.084) P (thy space) 0.595 E 0.16 Plasma mb 0.700 P 0.97 Y (0.582) 0.5829 cTP (0.955) MECS 85% P (thy mb) 0.547 M 0.03 55 residues 58 residues mTP (0.024) M (inner mb) 0.415 E 0.03 ER (lumen) 0.850 P 0.66 Y (0.526) 0.009 cTP (0.496) HDS 87% Plasma mb 0.790 M 0.02 40 residues 18 residues mTP (0.098) Peroxisome 0.300 E 0.33 M (matrix) 0.850 P 0.80 0.104 Y (0.530) cTP (0.827) HDR 79% Peroxisome 0.322 M 0.02 Not 34 residues mTP (0.129) Nucleus 0.300 E 0.20 predictable DXS/DXS2.1 (CAA09804, [39]); DXS2.2 (ABI35993, this work); DXR (AAF65154, [40]), CMS (ACI16377, this work); CMK (ABI35992, this work); MECS (AAF65155, [40]), HDS (AAO24774, [41]); HDR (ABI30631, this work). For DXS2.1 and DXS2.2, the percent of identity has been established with DXS/CLA1, DXSL2/DXL1, DXS3/DXL2, respectively. For the PSORT and Predotar predictions of subcellular localization, the three most favorable localizations are indicated with the corresponding score. For the ChloroP, MitoProt and TargetP predictions, scores of transit peptide presence and transit peptide length are given. cTP, chloroplast transit peptide; Cyto, cytoplasm; E, elsewhere; ER, endoplasmic reticulum; M, mitochondria; mb, membrane; mTP, mitochondria transit peptide; P, plastid; thy, thylakoid. Functional validations of DXS2.2, CMS, CMK and HDR were achieved by performing complementation assays in mutant E. coli strains, in which a disruption of the corresponding gene results in a lethal phenotype [42,43]. TP-truncated versions of the four proteins were thus expressed in their respective E. coli mutant strains using the pQE-30 plasmid. We found that each TP-truncated protein was able to complement the corresponding deficiency of E. coli mutants, therefore validating the biochemical activity of the cloned C. roseus DXS2.2, CMS, CMK and HDR enzymes (Figure 2A,B). The coding sequences of each enzyme were deposited to Genbank (Table 1). Plants 2020, 9, 462 6 of 25 Plants 2020, 9, x FOR PEER REVIEW 6 of 24 Figure 2. Figure Functional characterization 2. Functional characterization of of DXS2.2, DXS2.2, CMS, CMS, CMK CMK and and HDR HDR byby complementation complementation of E. coli of E. coli mutant strains. DXS, CMS, CMK-deficient E. coli cells (A) or HDR-deficient E. coli cells (B) were mutant strains. DXS, CMS, CMK-deficient E. coli cells (A) or HDR-deficient E. coli cells (B) were transformed with constructs expressing the TP-deleted DXS2.2, CMS, CMK or HDR (HDR2 and HDR3 transformed with constructs expressing the TP-deleted DXS2.2, CMS, CMK or HDR (HDR2 and represent two different deletions of the transit peptide), respectively, or with the original pQE30 vector HDR3 represent two different deletions of the transit peptide), respectively, or with the original as a control. After recovering the transformed cells on plates supplemented with 1 mM mevalonate pQE30 vector as a control. After recovering the transformed cells on plates supplemented with 1 mM (MVA) to rescue the lethal deletion, colonies were replicated on new plates with (+MVA) or without mevalonate (MVA) to rescue the lethal deletion, colonies were replicated on new plates with (+MVA) MVA (−MVA) at 37 ◦ C for DXS, CMS, CMK (A) or on plates containing MVA with Glucose (+Glc) or or without MVA (−MVA) at 37 °C for DXS, CMS, CMK (A) or on plates containing MVA with Glucose Arabinose (+Ara) for HDR. (+Glc) or Arabinose (+Ara) for HDR. 2.2. MEP Pathway Genes Are Highly Coexpressed in IPAP Cells of Young Leaves 2.2. MEP Pathway Genes Are Highly Coexpressed in IPAP Cells of Young Leaves The expression of MEP pathway genes was first analyzed using publically available RNA seq data (TableThe 2). expression Interestingly,of MEP pathway genes the coexpression was first network builtanalyzed using as using DXS2.1 publically availablethe a bait illustrated RNA seq strong data (Table 2). coexpression of Interestingly, all available MEPthe pathway coexpression network genes. built were Six of them usingfound DXS2.1 as athe within bait illustrated first the 20 positions strong coexpression of all available MEP pathway genes. Six of them were found within of the network with the Pearson correlation coefficient >0.95. DXS2.2 was ranked 64 with a PCC > 0.90 the first 20 positions of the network with the Pearson correlation coefficient >0.95. DXS2.2 was ranked and HDR was not found on the data set. We further analyzed the expression pattern of the eight genes 64 with a PCC > 0.90 on fully and HDR developed C.was notplants roseus foundby onRT–qPCR. the data set. We further Transcripts analyzed of each enzymethe were expression pattern detectable of in all the eight genes on fully developed C. roseus plants by RT–qPCR. Transcripts of each tested organs, including young and mature leaves, internodes, flower buds and flowers (Figure 3A–H). enzyme were detectable in all tested organs, including young and mature leaves, internodes, flower buds and flowers (Figure 3A–H). Plants 2020, 9, 462 7 of 25 Table 2. MEP pathway gene RNA seq coexpression pattern. A DXS2.1 coexpression network was drawn using Medicinal Plant Genomics Ressource MPGR RNA seq data [44] showing the clustering of the MEP pathway genes. The codes for the 23 samples and the transcript accession numbers (i.e., cra locus numbers) are from MPGR. See methods for details. RANK 1 2 8 14 16 20 64 NF PCC 1 0,98859 0,96838 0,96244 0,96168 0,955751 0,921181 NF CRA 618 4910 1424 7966 891 4962 381 2720 LOCUS NAME DXS2.1 CMK HDS MECS CMS DXR DXS2.2 HDR CRA_AA 23 66 149 40 26 69 9 CRA_AN 49 65 144 82 33 65 15 CRA_AM 36 81 161 89 30 92 9 CRA_AO 34 37 48 27 10 52 16 CRA_AP 221 113 239 87 39 138 45 CRA_AE 85 70 109 57 23 71 16 CRA_AG 416 227 287 126 46 216 106 CRA_AF 191 129 242 96 45 132 100 CRA_AH 52 32 105 58 10 40 19 CRA_AB 65 44 164 48 13 42 9 CRA_AC 26 26 106 41 10 22 5 CRA_AD 22 25 98 36 12 26 4 NF CRA_AL 26 21 78 32 7 26 5 CRA_AI 38 29 85 42 14 23 15 CRA_AJ 41 33 74 41 15 28 19 CRA_AK 35 34 72 32 5 26 5 CRA_AS 102 103 206 78 31 173 36 CRA_AQ 181 146 310 135 51 254 104 CRA_AR 67 60 151 75 15 123 30 CRA_AT 141 110 234 66 29 243 42 CRA_AU 803 534 1084 382 146 1374 256 CRA_AV 186 148 275 86 41 281 79 CRA_AW 832 573 1154 335 136 1288 529 Plants 2020, 9, x FOR PEER REVIEW 8 of 24 Plants 2020, 9, 462 8 of 25 Figure 3. Expressing pattern of MEP pathway genes in different C. roseus organs. Total RNA was Figure 3.from extracted Expressing the rootpattern of MEP (R), young pathway leaves genes leaves (YL), mature in different C. roseus (ML), first organs. internodes Total (IN), RNAbuds flower was extracted (FB) from the and mature root(F), flower (R),and young leaves to subjected (YL), mature reverse leaves (ML), transcription. first internodes Transcript levels for(IN), MEP flower buds pathway (FB) and mature flower (F), and subjected to reverse transcription. Transcript levels genes and RPS9 were determined by real-time PCR using gene-specific primers. MEP pathway gene for MEP pathway genes and levels expression RPS9 were were determined by real-time normalized using PCRtranscript RPS9. The using gene-specific primers. level of young MEP leaves waspathway gene individually expression levels were normalized using RPS9. The transcript level of young leaves was set to “1” for the eight genes. (A) DXS2.1; (B) DXS2.2; (C) DXR; (D) CMS; (E) CMK; (F) MECS; (G) HDS; individually set HDR. (H) to “1” for the eight genes. (A) DXS2.1; (B) DXS2.2; (C) DXR; (D) CMS; (E) CMK; (F) MECS; (G) HDS; (H) HDR. An overall coordinated expression pattern of the whole set of genes, except DXS2.2, could be An overall observed coordinated with a high expression accumulation pattern in of transcripts of young the whole set of genes, and actively except growing DXS2.2, aerial organscould (young be observed leaves, with buds) flowers a highbutaccumulation of transcripts also to a lesser in young extent in roots and actively and flowers, growing and with lower aerial levelsorgans (young in internode leaves, and flowers mature leaves.buds) Suchbut also to a lesser an expression profileextent in roots also tightly and flowers, correlates with thatand with of the IDIlower levels in gene analyzed internode on the sameand mature samples leaves. [34] and isSuch an expression in agreement profile with the highalso tightly demand ofcorrelates IPP and DMAPPwith thatto of the the fulfill IDI gene analyzed variety on the same samples of MEP pathway-derived [34] and terpenoid is in agreement metabolisms in thesewith the high different demand organs. of IPP and By contrast, the DMAPP to fulfill the variety of MEP pathway-derived terpenoid metabolisms distribution pattern of DXS2.2 transcript distribution was somehow different from DXS2.1 with barely in these different organs. Bytranscripts detectable contrast, theindistribution patternexpression roots and highest of DXS2.2in transcript distribution flower buds, was that suggesting somehow different both isoforms from be could DXS2.1 within barely involved detectable different transcripts physiological in roots and highest expression in flower buds, processes. suggesting that both Subsequently, weisoforms performed could beRNA in situ involved in different hybridization physiological of DXS2.2, processes. CMS, CMK and HDR transcripts Subsequently, on serial we performed sections of young developinginleaves, situ RNA hybridization with DXS2.1, DXR, MECSof DXS2.2, and HDS CMS, CMK (Figure as controls and HDR 4). transcripts on serial sections of young developing leaves, with DXS2.1, DXR, MECS Using the antisense probes, the transcripts of the four genes were specifically detected in the adaxial and HDS as controls (Figure 4). Using the antisense probes, the transcripts of the four genes were specifically Plants 2020, 9, 462 9 of 25 Plants 2020, 9, x FOR PEER REVIEW 9 of 24 part of the vascular region, corresponding to IPAP cells (Figure 4C,G,I,O). Such signal was also observed detected in the adaxial part of the vascular region, corresponding to IPAP cells (Figure 4C,G,I,O). for DXS2.1, DXR, Such signal wasMECS, HDS transcripts also observed for DXS2.1,as previously DXR, established MECS, HDS (Figure transcripts 4A,E,K,M; as previously [32,35,45,46]) established as well as for (Figure IDI transcripts 4A,E,K,M; analyzed [32,35,45,46]) onfor as well as the IDIsame sample transcripts [34]. No analyzed significant on the background same sample [34]. No was observed withbackground significant the corresponding was observed sense withprobes (Figure 4B,D,F,H,J,L,N,P). the corresponding sense probes (FigureThis thus confirms the 4B,D,F,H,J,L,N,P). previous colocalization This thus DXS2.1,colocalization confirms theofprevious DXR, MECSofand HDSDXR, DXS2.1, transcripts MECS andandHDS extends it to the transcripts andherein extends it to the herein identified gene transcripts DXS2.2, CMS, CMK and HDR transcripts. identified gene transcripts DXS2.2, CMS, CMK and HDR transcripts. Taken together, these results thus Taken together, highlight IPAPthese results as the mainthus sitehighlight IPAP as the of coexpression main site of genes of MEP of the coexpression pathway. of genes of the MEP pathway. Figure 4. Coexpression Figure 4. Coexpressionofofthe theMEP MEP pathway genesinin pathway genes internal internal phloem phloem associated associated parenchyma parenchyma (IPAP)(IPAP) of C.ofroseus cellscells young C. roseus leaves. young Paraffin-embedded leaves. Paraffin-embeddedserial serialsections sections of of young young leaves were hybridized leaves were hybridized with digoxigenin-labeled transcripts, with digoxigenin-labeled whichwhich transcripts, were subsequently were subsequentlylocalized withwith localized anti-digoxigenin-alkaline anti-digoxigenin- phosphatase conjugatesconjugates alkaline phosphatase followed followed by nitro by blue tetrazolium nitro chloride blue tetrazolium (NBT)/5-bromo chloride 4-chloro-3-indolyl (NBT)/5-bromo 4-chloro- 3-indolyl(BCIP) phosphate phosphate (BCIP) color color development. development. The antisense The antisense (AS) probes (AS) probes usedused for RNA for RNA labeling labeling andcontrol and control hybridization with sense (S) RNA probes are mentioned in the figure. hybridization with sense (S) RNA probes are mentioned in the figure. (A) DXS2.1 AS; (B) DXS2.1 (A) DXS2.1 AS; (B) S; (C) DXS2.1 S; (C) DXS2.2 AS; (D) DXS2.2 S; (E) DXR AS; (F) DXR S; (G) CMS AS; (H) DXS2.2 AS; (D) DXS2.2 S; (E) DXR AS; (F) DXR S; (G) CMS AS; (H) CMS S; (I) CMK AS; (J) CMK S; (K) CMS S; (I) CMK AS; (J) CMK S; (K) MECS AS; (L) MECS S; (M) HDS AS; (N) HDS S; (O) HDR AS; (P) HDR S. MECS AS; (L) MECS S; (M) HDS AS; (N) HDS S; (O) HDR AS; (P) HDR S. Bars: 100 µm. Bars: 100 µm. 2.3. MEP Pathway Enzymes Are Targeted to Plastids 2.3. MEP Pathway Enzymes Are Targeted to Plastids The subcellular localization of C. roseus MEP pathway enzymes was then analyzed through a The subcellular localization of C. roseus MEP pathway enzymes was then analyzed through a fluorescent protein (FP) imaging procedure recently developed for C. roseus cell cultures [47,48]. The fluorescent protein (FP) imaging procedure recently developed for C. roseus cell cultures [47,48]. The full-length sequence full-length ofofeach sequence eachenzyme enzyme was fusedtotothe was fused theN-terminal N-terminal endend of green of green FP (GFP) FP (GFP) to express to express a a MEPMEP enzyme-GFP fusion maintaining accessibility of the predicted TP. The resulting enzyme-GFP fusion maintaining accessibility of the predicted TP. The resulting proteins wereproteins were transiently expressed transiently expressedininC.C.roseus roseuschlorophyll-containing cells chlorophyll-containing cells forfor imaging. imaging. Depending Depending on enzymes, on enzymes, twotwodifferent patterns different patternsofoffluorescence fluorescence were observed(Figure were observed (Figure 5).5). FromFromCMSCMS to HDR, to HDR, all enzyme all enzyme fusions displayed a GFP fluorescence signal colocalizing with the autofluorescence of chlorophylls, demonstrating that these enzymes were targeted to plastids (Figure 5D1–H4). The signal appeared diffused into the stroma, as well as the stromules as previously reported for HDS and IDI [34,36]. In Plants Plants 2020, 9, 4629, x FOR PEER REVIEW 2020, 10 of 25 10 of 24 fusions displayed a GFP fluorescence signal colocalizing with the autofluorescence of chlorophylls, contrast, the two first demonstrating that steps of the MEP these enzymes werepathway, including targeted to DXS2.1, plastids (Figure DXS2.2The 5D1–H4). andsignal DXR,appeared displayed a diffused into the GFP fluorescence stroma, signal as well as the in punctuated stromules as aggregates previously within reported plastids for5A1–C4). (Figure HDS and IDI [34,36]. In Although diffuse contrast, the two first steps of the MEP pathway, including DXS2.1, DXS2.2 and fluorescence signals were also observed in plastid stroma of a few numbers of cells, the punctuated DXR, displayed a GFP fluorescence signal in punctuated aggregates within plastids (Figure 5A1–C4). aggregates were detected in around 95% of the observed cells for both DXS and DXR. To determine Although diffuse fluorescence signals were also observed in plastid stroma of a few numbers of cells, the punctuated whether this punctuated localization depended on the cell type used for localization, we also performed aggregates were detected in around 95% of the observed cells for both DXS and DXR. To determine biolistic-mediated transient transformations of cells from detached C. roseus leaves. In these cells, while whether this punctuated localization depended on the cell type used for localization, we also HDS-Yellow performed FPbiolistic-mediated (YFP) exhibited atransient diffuse transformations signal in plastids, bothfrom of cells DXSdetached isoformsC.and DXR roseus YFP fusions leaves. In still displayed these cells,awhile punctuated HDS-Yellow fluorescence signal in aplastids FP (YFP) exhibited (Figure diffuse signal in 6A–D). plastids,Such a protein both DXS distribution isoforms and observed DXR in different YFP fusionscell stilltypes was intriguing displayed a punctuatedandfluorescence raised the question of possible signal in plastids subplastidial (Figure 6A–D). Such targeting a of DXS and DXR. protein distribution observed in different cell types was intriguing and raised the question of possible subplastidial targeting of DXS and DXR. FigureFigure 5. Subcellular 5. Subcellular localization localization of the of the MEPMEP pathway pathway enzymes enzymes ininC.C.roseus roseuschlorophyll-containing chlorophyll-containing cells. cells. C. roseus cells (CR6 cell line) were transiently transformed with C. roseus cells (CR6 cell line) were transiently transformed with plasmids expressing plasmids expressing MEPMEPpathway pathway enzymes fused to the N-terminus of GFP as indicated in the first column. The fluorescence enzymes fused to the N-terminus of GFP as indicated in the first column. The fluorescence emitted by GFP-fused proteins (A1–H1) was compared to chlorophyll autofluorescence (A2–H2). Colocalization of the two signals appeared in yellow while merging the two individual (green/red) color images (A3–H3). Cell morphology was observed with differential interference contrast (DIC, A4–H4). Bar: 10 µm. Plants 2020, 9, x FOR PEER REVIEW 11 of 24 emitted by GFP-fused proteins (A1–H1) was compared to chlorophyll autofluorescence (A2–H2). Colocalization of the two signals appeared in yellow while merging the two individual (green/red) color images (A3–H3). Cell morphology was observed with differential interference contrast (DIC, Plants 2020, 9, 462 11 of 25 A4–H4). Bar: 10 µm. Figure 6. Subcellular localization of HDS-YFP, DXS2.1-YFP, DXS2.2-YFP and DXR-YFP fusions in cells Figure 6. Subcellular localization of HDS-YFP, DXS2.1-YFP, DXS2.2-YFP and DXR-YFP fusions in of young C. roseus leaf. Young C. roseus leaves were transiently transformed with plasmids expressing cells of young C. roseus leaf. Young C. roseus leaves were transiently transformed with plasmids HDS (A) and both DXS (B,C) and DXR (D) fused to the N-terminus of YFP as indicated on the top of expressing HDS (A) and both DXS (B,C) and DXR (D) fused to the N-terminus of YFP as indicated on each picture. The fluorescence emitted by YFP-fusion proteins was superimposed to cell morphology the top of each picture. The fluorescence emitted by YFP-fusion proteins was superimposed to cell acquired using differential interference contrast. Arrowhead and star indicate punctuated and diffuse morphology acquired using differential interference contrast. Arrowhead and star indicate localization in plastids, respectively. Bar: 10 µm. punctuated and diffuse localization in plastids, respectively. Bar: 10 µm. 2.4. Punctuated Pattern of Localization of DXS and DXR Requires the Whole Protein Sequence 2.4. Punctuated Pattern of Localization of DXS and DXR Requires the Whole Protein Sequence To gain insight into the punctuated localization of DXS and DXR, we performed additional fusions with To gain insight fluorescent into First, proteins. the punctuated to analyze the localization role of the of DXS part mature andofDXR,thesewe performed proteins, fusionadditional engaging fusions with fluorescent proteins. First, to analyze the role of the mature part only the predicted TP (first 100 residues) of DXS2.1, DXS2.2, DXR or HDS (named tpDXS2.1-GFP, of these proteins, fusion engaging only the predicted TP (first 100 residues) of DXS2.1, DXS2.2, tpDXS2.2-GFP, tpDXR-GFP and tpHDS-GFP) were transiently expressed in C. roseus cells. Compared to DXR or HDS (named tpDXS2.1-GFP, full-length tpDXS2.2-GFP, proteins, deletions of tpDXR-GFP the mature part andoftpHDS-GFP) DXS, DXS2 and wereDXR transiently led to theexpressed loss of the in C. roseus aggregated cells. Compared pattern to full-length of fluorescence proteins, and resulted deletions in diffuse of the patterns mature partwithin of fluorescence of DXS,theDXS2 stromaandandDXR led to stromules the loss of the aggregated pattern of fluorescence and resulted in diffuse (Figure 7A1–C6). By contrast, both tpHDS-GFP, HDS-GFP and HDS-CFP displayed a similar diffuse patterns of fluorescence within the stroma fluorescent signal inandthe stromules stroma and(Figure stromules7A1–C6). By contrast, of plastids both tpHDS-GFP, (Figure 7D1–D6), HDS-GFP clearly distinct fromandthe HDS-CFP displayed a similar diffuse fluorescent signal in the stroma and signal obtained with the DXS2.1-GFP protein (Figure 7E1–E4). In addition, internal fusions with YFP stromules of plastids (Figure were 7D1–D6), also created for clearly bothdistinct DXR and from HDSthe(tpDXR-YFP-DXR signal obtained with andthe DXS2.1-GFP protein tpHDS-YFP-HDS) (Figure(Figure 8A,B).7E1– For E4). In addition, internal fusions with YFP were also created for both DXR and each of these constructs, the punctuated (DXR) and diffuse plastid (HDS) localizations were conserved HDS (tpDXR-YFP- DXR and (Figure tpHDS-YFP-HDS) 8C–J), confirming the(Figure 8A,B). of requirement Forthe each of these mature partconstructs, of DXS2.1,the punctuated DXS2.2 and DXR (DXR) and to their diffuse plastid (HDS) localizations peculiar localization in plastids. were conserved (Figure 8C–J), confirming the requirement of the mature part of DXS2.1, DXS2.2 and DXR to their peculiar localization in plastids. Plants 2020, 9, x FOR PEER REVIEW 12 of 24 Plants 2020, 9, 462 12 of 25 Plants 2020, 9, x FOR PEER REVIEW 12 of 24 FigureFigure 7. Comparison 7. Comparison ofofthe thesubcellular subcellular localization localizationofofDXS2.1, DXS2.2, DXS2.1, DXRDXR DXS2.2, and HDSandexpressed as HDS expressed Figure 7. Comparison full-length-GFP fusionofproteins the subcellular localization or as transit of DXS2.1, peptide-GFP fusion DXS2.2, proteins.DXR and HDS C. roseus cells expressed were as as full-length-GFP fusion proteins or as transit peptide-GFP fusion proteins. C. roseus cells were full-length-GFP fusion with transiently transformed proteins or as plasmids transit apeptide-GFP expressing fusion GFP protein fused proteins. to the C. roseus first 100 residues cells were of the transiently transformed with plasmids expressing a GFP protein fused to the first 100 residues of the enzyme, encompassing transiently transformed the withpredicted plasmidstransit peptide (tp) expressing a GFP of DXS2.1 protein(tpDXS2.1-GFP, fused to the first A1-A3), DXS2.2 of the 100 residues enzyme, encompassing the predicted transit peptide (tp) of DXS2.1 (tpDXS2.1-GFP, A1–A3), DXS2.2 (tpDXS2.2-GFP, B1-B3), DXR (tpDXR-GFP, C1-C3), HDS (tpHDS-GFP, D1-D3), or fused enzyme, encompassing the predicted transit peptide (tp) of DXS2.1 (tpDXS2.1-GFP, A1-A3), DXS2.2 to full-length (tpDXS2.2-GFP, B1–B3), DXR DXS2.1 (DXS2.1-GFP, (tpDXR-GFP, C1–C3), HDS (tpHDS-GFP, D1–D3), or fused to full-length (tpDXS2.2-GFP, B1-B3),A4-A6), DXS2.2 (DXS2.2-GFP, DXR (tpDXR-GFP, C1-C3), B4-B6), DXR (DXR-GFP, HDS (tpHDS-GFP, C4-C6), D1-D3), HDS (HDS- or fused to full-length DXS2.1 (DXS2.1-GFP, A4–A6), DXS2.2 (DXS2.2-GFP, B4–B6), DXR (DXR-GFP, C4–C6), GFP, D4-D6). An additional co-transformation of DXS2.1-YFP (E1) and HDS-CFP (E2) was performed HDS (HDS-GFP, DXS2.1 (DXS2.1-GFP, A4-A6), DXS2.2 (DXS2.2-GFP, B4-B6), DXR (DXR-GFP, C4-C6), HDS (HDS- D4–D6). An additional to highlight co-transformation differences of localization inof plastids DXS2.1-YFP (E1) and as revealed onHDS-CFP the merged (E2) was (E3). image performed Cell to GFP, D4-D6). An additional co-transformation of DXS2.1-YFP (E1) and HDS-CFP (E2) was performed highlight differences morphology wasof observed localization in plastids with as revealed differential on thecontrast interference merged (DIC, image E4). (E3). Cell Bar:morphology 10µm to highlight differences of localization in plastids as revealed on the merged image (E3). Cell was observed with differential interference contrast (DIC, E4). Bar: 10 µm. morphology was observed with differential interference contrast (DIC, E4). Bar: 10µm Figure 8. Subcellular localization of DXR and HDS studied using internal YFP protein fusions. C. roseus cells were transiently transformed with plasmids expressing internal YFP fused to split DXR (A) and HDS (B). The fluorescence emitted by YFP-fused proteins (C,G) was compared to the fluorescence emitted by the plastid-CFP marker (D,H). Colocalization of the two signals appeared in Figure 8. Subcellular localization of DXR and HDS studied using internal YFP protein fusions. C. roseus Figure yellow8.by Subcellular merging thelocalization of DXR two individual and HDS (green/red) colorstudied usingCell images (E,I). internal YFP protein morphology fusions. C. is observed cells were transiently transformed with plasmids expressing internal YFP fused to split DXR (A) and roseus with cells were transiently differential transformed interference contrast withBar: (DIC, F,J). plasmids 10 µm. expressing internal YFP fused to split DXR HDS (B). The fluorescence emitted by YFP-fused proteins (C,G) was compared to the fluorescence (A) and HDS (B). The fluorescence emitted by YFP-fused proteins (C,G) was compared to the emitted by the plastid-CFP marker (D,H). Colocalization of the two signals appeared in yellow by fluorescence emitted by the plastid-CFP marker (D,H). Colocalization of the two signals appeared in merging the two individual (green/red) color images (E,I). Cell morphology is observed with differential yellow by merging the two individual (green/red) color images (E,I). Cell morphology is observed interference contrast (DIC, F,J). Bar: 10 µm. with differential interference contrast (DIC, F,J). Bar: 10 µm. Plants 2020, 9, 462 13 of 25 2.5. Punctuated Localization of DXS and DXR is Likely Caused by Protein Overexpression To identify a potential Plants 2020, plastid 9, x FOR PEER REVIEW subcompartment hosting DXS2.1-, DXS2.2- and DXR-YFP fusions, we 13 of 24 performed cell Totransformations identify a potential plastid subcompartment with distinct markers hosting DXS2.1-, targeted DXS2.2- and DXR-YFP to thylakoids (CAB-CFP;fusions, Figure 9A,B), 2.5.performed Punctuated Localization of DXS and DXR is Likely Causedtargeted by Protein to Overexpression the innerwe membrane cell of transformations plastid envelope with distinct markers (TIC-CFP; Figure 9C,D) thylakoids (CAB-CFP; Figure and plastoglobuli (CCD4-CFP; 9A,B), the inner membrane of plastid envelope (TIC-CFP; Figure 9C,D) and plastoglobuli (CCD4- Figure 9E,F). While thylakoid and inner membrane markers exhibited circular CFP; Figure 9E,F). While thylakoid and inner membrane markers exhibited circular fluorescent fluorescent signals in plastids, the plastoglobuli marker displayed a punctuated signal similar to that of signals in plastids, the plastoglobuli marker displayed a punctuated signal similar to that of DXS2.1, DXS2.1, DXS2.2 DXS2.2 and DXR fusion proteins. However, cotransformations of cells with plasmids and DXR fusion proteins. However, cotransformations of cells with plasmids expressing this marker expressing this marker (CCD4-CFP) or DXS2.1, DXS2.2, DXR, HDS-YFP did not reveal a perfect superimposition of (CCD4-CFP) or DXS2.1, DXS2.2, DXR, HDS-YFP did not reveal a perfect superimposition of the two the two fluorescent signals, preventing us to firmly conclude on a potential targeting of MEP pathway fluorescent signals, preventing(Figure enzyme to plastoglobuli us to9G–Z). firmly conclude on a potential targeting of MEP pathway enzyme to plastoglobuli (Figure 9G–Z). Figure 9. Figure 9. Coexpressions Coexpressions of DXS2.1-YFP,DXS2.2-YFP of DXS2.1-YFP, DXS2.2-YFP andandDXR-YFP DXR-YFPfusions with different fusions plastid with different plastid markers. Chlorophyll-containing C. roseus cells were transiently transformed with plasmids markers. Chlorophyll-containing C. roseus cells were transiently transformed with plasmids expressing expressing the CFP-marker of thylakoid (CAB-CFP, A,B), the inner membrane of chloroplast envelope the CFP-marker of thylakoid (TIC-CFP, (CAB-CFP, (Carotenoid C,D) and plastoglobuli A,B), the inner membrane Cleavage of chloroplast Dioxygenase (CCD) 4-CFP, envelope E,F) or (TIC-CFP, C,D) and plastoglobuli (Carotenoid cotransformed with Cleavage plasmids expressing theDioxygenase CCD4-CFP marker (CCD) 4-CFP, and the E,F) orfusion DXS2.1-YFP cotransformed protein with (G–J). Chlorophyll-free plasmids expressing the CCD4-CFPC. roseus cells were marker andtransiently transformed the DXS2.1-YFP with protein fusion plasmids(G–J). expressing the Chlorophyll-free CCD4-CFP marker and the DXS2.1-YFP (K–N), DXS2.2-YFP (O–R), DXR-YFP (S–V) and HDS-YFP C. roseus cells were transiently transformed with plasmids expressing the CCD4-CFP marker and (W–Z) fusion proteins. Colocalization of the two signals appeared in yellow by merging the two the DXS2.1-YFP (K–N), DXS2.2-YFP (O–R), DXR-YFP (S–V) and HDS-YFP (W–Z) fusion proteins. Colocalization of the two signals appeared in yellow by merging the two individual (green/red) color images (I,M,Q,U,Y). Cell morphology is observed with differential interference contrast (DIC, B,D,F,J,N,R,V,Z). Bar: 10 µm. Plants 2020, 9, x FOR PEER REVIEW 14 of 24 Plants 2020, 9, 462 14 of 25 individual (green/red) color images (I,M,Q,U,Y). Cell morphology is observed with differential interference contrast (DIC, B,D,F,J,N,R,V,Z). Bar: 10 µm. This result prompted us to analyze subcellular localization in native conditions by performing This result prompted anti-DXR immunogold labelingusontocells analyze of C.subcellular localization roseus young leavesinand native conditions by transmission performing electron microscopy anti-DXR immunogold labeling on cells of C. roseus young leaves and transmission electron (TEM) observation. First, we noticed that the immunogold labeling was only barely detectable in microscopy (TEM) observation. First, we noticed that the immunogold labeling was only barely plastidsdetectable of palissadic parenchyma in plastids while of palissadic a pronounced parenchyma signal wassignal while a pronounced observed in plastids was observed of IPAP cells in plastids (Figure of 10A–D), thus confirming our previous results of RNA in situ hybridization (Figure IPAP cells (Figure 10A–D), thus confirming our previous results of RNA in situ hybridization 4). (Figure 4). Figure 10. Immunogold labeling of DXR in the stroma of plastids from C. roseus leaf cells. (A) Morphology of C. roseus young leaves observed on ultrathin cross-sections highlighting epidermis (ep), palissadic parenchyma (pp), spongy parenchyma (sp), xylem (x), phloem (ph) and internal phloem associated parenchyma (* or ipap). (B) Intermediate magnification highlighting plastids from pp and ipap. (C) A very low abundant—but specific—labeling was observed in the stroma of palissadic parenchyma plastids neighboring ipap cells. (D,E) Typical dense labeling was specifically observed in the stroma of ipap plastids devoid of plastoglobuli-like structures but including stromule budding (s). cp, chloroplast; cw, cell wall; sg, starch granule. (ep), palissadic parenchyma (pp), spongy parenchyma (sp), xylem (x), phloem (ph) and internal phloem associated parenchyma (* or ipap). (B) Intermediate magnification highlighting plastids from pp and ipap. (C) A very low abundant—but specific—labeling was observed in the stroma of palissadic parenchyma plastids neighboring ipap cells. (D,E) Typical dense labeling was specifically Plants observed 2020, 9, 462in the stroma of ipap plastids devoid of plastoglobuli-like structures but including stromule 15 of 25 budding (s). cp, chloroplast; cw, cell wall; sg, starch granule. Interestingly, aa homogenous Interestingly, homogenous immunogold labeling was immunogold labeling was observed observed in in the the stroma stroma of of all all observed observed plastids that never displayed plastoglobluli-like structures (Figure 10D,E) plastids that never displayed plastoglobluli-like structures (Figure 10D,E) but could show but could show some some stromule budding stromule budding including including labeling labeling (Figure (Figure 10E). 10E). Such Such anan absence absence of of the the aggregated aggregated localization localization observed through the GFP imaging procedure strongly suggested that it might result procedure observed through the GFP imaging procedure strongly suggested that it might result from the from the of protein expression procedure in cells andinnotably of protein expression cells andthenotably strong thegene overexpression strong caused by caused gene overexpression the use by of the the CaMV35S use of the promoter. To test thisTo CaMV35S promoter. hypothesis, additional transformations test this hypothesis, of the construct additional transformations of theexpressing construct the DXR-YFP fusion were performed in cells of C. roseus young leaves using a expressing the DXR-YFP fusion were performed in cells of C. roseus young leaves using a decreasing decreasing amount of transforming amount plasmid DNA. of transforming Interestingly, plasmid while usingwhile DNA. Interestingly, high amount using highof transforming DNA (800 ng amount of transforming DNAper transformation) still resulted in the observation of a punctuated pattern of localization (800 ng per transformation) still resulted in the observation of a punctuated pattern of localization (Figure 11A), lowering11A), (Figure bombarded lowering DNA (400 and 150 bombarded DNA ng)(400 led to and the150 progressive ng) led to fading of this localization the progressive fading pattern of this up to a complete localization diffuse pattern up todistribution a complete indiffuse plastid distribution stroma (Figure in 11B,C). plastid These stromaresults (Figuretherefore 11B–C).confirm These the localization of DXR observed through immunolabeling and suggest that results therefore confirm the localization of DXR observed through immunolabeling and suggest thatDXR is mainly diffuse in plastid DXR stroma. is mainly Additionally, diffuse these results in plastid stroma. also support Additionally, thesethe idea also results that support an artifactual the idea DXS/DXR that an localizationDXS/DXR artifactual could be caused by gene localization couldoverexpression. be caused by gene overexpression. Figure 11. Transient Transient transformation transformationofofC.C.roseus roseusyoung young leaf leaf using using a decreasing a decreasing amount amount of the of the DXR- DXR-YFP YFP expressing expressing construct. construct. YoungYoung C.leaves C. roseus roseuswere leaves were transiently transiently transformed transformed with amounts with decreasing decreasing of amounts of the the plasmid plasmidDXR expressing expressing fused toDXR the fused to the of N-terminus N-terminus of YFP as YFP as indicated indicated at the bottomatofthe bottom each of picture. each picture. Theemitted The fluorescence fluorescence emitted byproteins by YFP-fusions YFP-fusions proteins was superimposed was superimposed to cell to cell morphology morphology acquired using acquired using differential differential interference interference contrast. contrast. Arrowhead andArrowhead star indicateand star indicate punctuated andpunctuated and diffuse diffuse localization in localization in plastids,(A) plastids, respectively. respectively. 800 ng; (B) (A) 400 800 ng; ng; (C) (B) 150400 ng. ng; Bar:(C) 10150 µm.ng. Bar: 10 µm. 3. Discussion 3. Discussion By cloning By cloning the the missing missing genes genes from from the the MEP MEP pathway pathway in C. roseus, in C. roseus, wewe provide provide aa comprehensive comprehensive overview of the pathway organization in planta, which may potentially impact overview of the pathway organization in planta, which may potentially impact its involvement in its involvement in the the synthesis of synthesis of highly highly valuable valuable MIAs. MIAs. As reported for many As reported for many plants plants synthesizing synthesizing specialized metabolites, we specialized metabolites, we first first established established that that atat least two least two isoforms isoforms ofof clade clade IIII DXS DXS coexist, coexist, namely namely DXS2.1 DXS2.1 and and DXS2.2, DXS2.2, in in the the Madagascar Madagascar periwinkle periwinkle (Figure S1). Interestingly, these two isoforms display slightly distinct expression (Figure S1). Interestingly, these two isoforms display slightly distinct expression profiles in C. profiles in C. roseus roseus organs. DXS2.1 exhibits a coordinated expression with other MEP genes, including organs. DXS2.1 exhibits a coordinated expression with other MEP genes, including IDI, to fulfill IPP IDI, to fulfill IPP and DMAPP and DMAPP to to both both primary primary and and secondary secondary metabolisms (Figure 3; metabolisms (Figure 3; [30,34]). [30,34]). By By contrast, contrast, DXS2.2 DXS2.2 is is mainly expressed in flower buds and much less expressed in roots compared mainly expressed in flower buds and much less expressed in roots compared to other MEP genes. to other MEP genes. This couldcould This potentially be linked potentially to a broader be linked involvement to a broader of this of involvement isoform in the leaf this isoform MIA in the metabolism leaf and the MIA metabolism dedicated and MIA metabolism the dedicated taking place MIA metabolism takingin reproductive organs asorgans place in reproductive described previously as described for a specific previously for isoform of tabersonine 16-hydroxylase [49]. In young leaves, RNA in situ a specific isoform of tabersonine 16-hydroxylase [49]. In young leaves, RNA in situ hybridizationhybridization confirmed that internal phloem associated parenchyma (IPAP) is the main site of expression of MEP pathway genes (Figure 4). Albeit we cannot exclude a lower expression in other leaf tissues, as described with the immunolabeling of HDS and confirmed for DXR (Figure 10; [35]), this result definitively confirms that IPAP is a key tissue in both IPP synthesis and MIA precursor production. The latter thus begins with a DXS-catalyzed reaction up to the antepenultimate step of the MSI pathway catalyzed by 7-deoxyloganic acid hydroxylase and class II cytochrome P450 reductase to form loganic acid [19,50], thus encompassing 16 successive enzymatic steps. While transporters allowing import of loganic acid Plants 2020, 9, 462 16 of 25 into the epidermis have been identified, transporters ensuring its export from IPAP have yet to be discovered [51]. The identification of such transporters remains of importance since transport is a key element in the control of the whole MIA metabolism. At the subcellular level, we showed that all MEP pathway enzymes are targeted to plastids including stroma and stromules, as previously established for HDS and IDI [34,35]. A schematic representation of this distribution is given in Figure 12. While DXS2.1, DXS2.2 and DXR were initially reported to display a punctuated pattern of localization in plastids, fusion/deletion experiments, DXR immunolabeling in planta and cell transformations with decreasing amounts of DNA, confirmed that this result was artifactual and probably resulted from artificially high protein expression levels (Figures 5, 7 and 9–11). These increased amounts of proteins may have impeded the protein folding capacities of DXS and DXR resulting in protein aggregation and formation of the punctuated fluorescence signal. Although partial colocalization with the plastoglobuli markers has been observed (Figure 9), this localization could not be observed with TEM. Additionally, no MEP pathway enzyme has so far been detected in the plastoglobuli proteome, suggesting that early MEP pathway enzymes, while expressed in close proximity to plastoglobuli due to their high density in plastids, are not in fact localized to these structures [52]. Furthermore, such a type of subplastidial distribution has already been described in Arabidopsis thaliana for DXS, which is prone to aggregation and is subject to Clp protease-mediated degradation [53,54]. While localization of DXR in extraplastidial vesicles has been reported [53], these vesicles have not been observed in our experimental conditions that mainly rely on transient cell transformations or IPAP TEM observation. These transient transformations are not compatible with the long analysis times required to observe vesicle formation. Alternatively, the development of vesicles for Arabidopsis enzymes could be slightly different than for the C. roseus counterparts. Based on our results and some previous reports [29,34,55], it appears that plastid stroma and stromules host the first ten steps of the synthesis of the monoterpene precursor of MIAs from DXS up to geraniol synthase. While transporters responsible for export of geraniol into the cytosol remain to be identified, it can be postulated that this subcellular localization particularly involving stromules, may facilitate the release of geraniol in close proximity to the endoplasmic reticulum where geraniol 10-hydroxylase (G10H) catalyzes the subsequent reaction of the MSI pathway [29,30,34,36]. In agreement with this statement, the close proximity between stromules and the endoplasmic reticulum has been already established in C. roseus [36] and Arabidopsis thaliana [56]. These observations were made by taking advantage of the dynamic properties of GFP imaging procedures while proximity between both organelles are more difficult to characterize through TEM analyses since these rely on cell sectioning and differential organelle electron density (Figure 10). Furthermore, exchanges between stromules and the endoplasmic reticulum have been already characterized, thus supporting geraniol channeling in C. roseus [57,58]. In conclusion, besides describing the MEP compartmentation in C. roseus, our work also illustrates how high overexpression of proteins may alter their apparent subcellular localization. This should be taken into consideration in the ever-growing number of studies involving this type of technical approach and aiming at manipulating metabolic fluxes in planta. The latter thus begins with a DXS-catalyzed reaction up to the antepenultimate step of the MSI pathway catalyzed by 7-deoxyloganic acid hydroxylase and class II cytochrome P450 reductase to form loganic acid [19,50], thus encompassing 16 successive enzymatic steps. While transporters allowing import of loganic acid into the epidermis have been identified, transporters ensuring its export Plants from 2020, IPAP have yet to be discovered [51]. The identification of such transporters remains 9, 462 17 of of 25 importance since transport is a key element in the control of the whole MIA metabolism. Figure 12. Schematic representation of the cellular and subcellular organization of the MEP pathway and early stage of the the monoterpene monoterpene seco-iridoid seco-iridoid (MSI) (MSI) pathway pathway inin C. C. roseus roseus leaves. leaves. The enzymes characterized in this study are highlighted in yellow. IDI (type I IPP isomerase) displayed a triple subcellular targeting subcellular targetingininstroma/stromules stroma/stromules ofof plastids, mitochondria plastids, mitochondriaandand peroxisomes as established peroxisomes [34]. as established The proximity between stromules and the endoplasmic reticulum facilitates export of [34]. The proximity between stromules and the endoplasmic reticulum facilitates export of geraniol geraniol in the close in theenvironment of G10HoftoG10H close environment increaseto hydroxylation rate. DXS,rate. increase hydroxylation 1-deoxy-D-xylulose 5-phosphate DXS, 1-deoxy-D-xylulose 5- synthase; DXR, phosphate 1-deoxy-D-xylulose synthase; 5-phosphate DXR, 1-deoxy-D-xylulose reductoisomerase; 5-phosphate CMS, 2-C-methyl-D-erythritol reductoisomerase; CMS, 2-C-methyl-D- 4-phosphate erythritol synthase; synthase; 4-phosphate CMK, 4-(cytidine 50 diphospho)-2-C-methyl-D-erythritol CMK, 4-(cytidine kinase; kinase; 5′diphospho)-2-C-methyl-D-erythritol MECS, 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; HDS, (E)-4-hydroxy-3-methylbut-2-enyl MECS, 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; HDS, (E)-4-hydroxy-3-methylbut-2- diphosphate enyl synthase; diphosphate HDR, (E)-4-hydroxy-3-methylbut-2-enyl synthase; HDR, (E)-4-hydroxy-3-methylbut-2-enyl diphosphate reductase; IDI, diphosphate isopentenyl reductase; IDI, diphosphate diphosphate isopentenyl isomerase. GPS, geranyl diphosphate isomerase. GPS, geranyl synthase. GES, geraniol diphosphate synthase.synthase. CPR, cytochrome GES, geraniol synthase. P450 reductase. CPR, cytochrome G10H, P450 geraniol reductase.10-hydroxylase. G10H, geraniolIPAP, internal phloem 10-hydroxylase. IPAP,associated internal parenchyma. GAP, phloem associated glyceraldehyde 3-phosphate. DXP, 1-deoxy-D-xylulose 5-phosphate. MEP, 2-C-methyl-D-erythritol parenchyma. GAP, glyceraldehyde 3-phosphate. DXP, 1-deoxy-D-xylulose 5-phosphate. MEP, 2-C- 4-phosphate. IPP, isopentenyl methyl-D-erythritol 4-phosphate.diphosphate. IPP, isopentenyl DMAPP, dimethylallyl diphosphate. DMAPP,diphosphate. dimethylallyl GPP, geranyl diphosphate. diphosphate. Created with Biorender.com. GPP, geranyl diphosphate. Created with Biorender.com. 4. Material and Methods 4.1. Plant Material and Cell Culture Conditions Mature C. roseus (L) G. Don, cultivar Pacifica Pink (Apocynaceae) plants grown from seeds (Ball Ducrettet, Thonon, France) in a greenhouse were harvested in summer for microscopy fixation (RNA in situ hybridization experiments) and RNA extraction (cloning experiments and gene expression measurements). Young leaves of C. roseus (Little Bright Eyes cultivar) grown under similar conditions were used for FP imaging. C. roseus cell suspensions used for subcellular localization studies (C20A chlorophyll-free cells and CR6 chlorophyll-containing cells) were propagated in Gamborg B5 medium Plants 2020, 9, 462 18 of 25 (Duchefa) at 24 ◦ C under continuous shaking (100 rpm) for 7 days (C20A cells) or 14 days (CR6 cells) as previously described [36]. 4.2. RNA Extraction and cDNA Synthesis Plant organs were ground in liquid nitrogen. Total RNA was extracted using the NucleoSpin RNA Plant kit (Macherey-Nagel, France) according to the manufacturer’s instructions. First-strand cDNAs was synthesized from 5 µg of total RNA using 500 ng of oligo(dT)18 primers and 15 units of Thermoscript reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. Following retrotranscription, remaining RNA was removed by treatment with E. coli RNase H (Invitrogen) during 20 min at 37 ◦ C. 4.3. Cloning of the cDNA Sequences from the C. Roseus MEP Pathway Enzymes The amino acid sequence of CMS, CMK and HDR orthologues from different plant species were retrieved from GenBank and aligned to identify conserved regions. Degenerated primers were designed to amplify an internal cDNA fragment (Table S1). Amplifications were carried out using retrotranscribed RNA extracted from young leaves using an oriented-cells cDNA library as previously described [59]. The corresponding PCR products were purified and cloned into pGEM-T Easy vector (Promega, Charbonnières-les-bains, France) prior to sequencing. A partial cDNA of the second C. roseus DXS ortholog was obtained by EST sequencing. Analysis of the five partial cDNA sequences allowed designing specific primers to isolate 50 and 30 ends by PCR performed on an oriented C. roseus cDNA library using the M13 reverse and T7 universal primers, respectively (Table S1). After sequencing of the resulting PCR products, specific primers were designed to amplify the full-length open reading frame of the five proteins as described for partial cDNA cloning (Table S1). The sequence of the C. roseus CMS, CMK, HDR and of the second ortholog of DXS were deposited at NCBI under Genbank accession numbers DQ848672 (DXS2.2), FJ177510 (CMS), DQ848672 (CMK) and DQ848676 (HDR). 4.4. Protein Production and E. coli Complementation Based on ChloroP predictions of plastid transit length (Table 1), truncated cDNA of DXS2.2, CMS, CMK and HDR were amplified as described in Table S2, allowing for removal of the predicted transit peptide and to introduce restriction sites at both cDNA ends. Amplifications were performed using Pfu DNA polymerase (Promega) and the corresponding PCR products were cloned into pQE-30 (Qiagen, Courtaboeuf, France). After sequencing, the plasmids expressing the pseudomature form of DXS2.2, CMS, CMK and HDR were used in bacterial complementation assays as described in Table S2 and according to the procedures described by [42,43]. 4.5. Gene Expression Measurements (Real-Time PCR Analysis) The expression of the nine genes of the C. roseus MEP pathway and RPS9 gene (GenBank AJ749993) was analyzed by real-time PCR analysis using the specific forward and reverse primers reported in Table S3. Plant organs including roots, first internodes, young and mature leaves, flower buds and flowers, were ground in liquid nitrogen and total RNA was extracted with the RNAeasy Plant mini kit (Qiagen). Total RNA (1.5 µg) was treated with RQ1 RNase-free DNase (Promega) and used for first-strand cDNA synthesis by priming with oligo d(T17 ) (0.5 µM). Reverse transcription was carried out using superscript III RT (Invitrogen) at 50 ◦ C and according to manufacturer’s instructions. Real-time PCR was run on an ABI Prism 7000 SDS light cycler (Applied Biosystems, Foster City, CA, USA) using the SYBR® Green I technology. Each PCR mixture was performed in a total reaction volume of 25 µl containing 1 µl of a 1/3 dilution RT reaction, 0.2 µM forward and reverse primers and 1 × MESA GREEN qPCR master mix Plus buffer (Eurogentec, Angers, France). The reaction was initiated by a denaturation step (95 ◦ C, 10 min), followed by 40 cycles at 95 ◦ C for 15 s and 60 ◦ C for 1 min. RPS9 was used as a housekeeping control gene to allow normalization in each sample. Each assay was performed in triplicate. Data correspond to average values (n = 3) ±standard deviation (SD). Relative Plants 2020, 9, 462 19 of 25 transcripts levels in each sample are expressed as a ratio of the abundance of the RPS9 transcripts and normalized to the transcript level of young leaves that was set to one for each gene separately. 4.6. Tissue Fixation, Embedding in Paraffin and Microtomy RNase-free conditions were strictly observed for all steps. Glassware was baked for 8 h at 180 ◦ C and nondisposable plastic ware was incubated 10 min in an aqueous 3% H2 O2 solution and rinsed in DEPC-treated water. Leaves from mature C. roseus plants grown in the greenhouse were harvested in late spring/early summer, rapidly fixed in FAA and embedded in Paraplast as previously described [32]. Serial sections (10 µm) were spread on silane-coated slides overnight at 40 ◦ C, and paraffin was removed using xylene (twice for 15 min) before rehydration in an ethanol gradient series up to DEPC-treated water. 4.7. RNA In Situ Hybridization The protocol used for the detection of the MEP pathway transcripts in C. roseus leaves was previously described [32,60]. Full-length cDNAs of DXS2.2, CMS, CMK and HDR amplified with the specific corresponding primers (Table S1) and cloned in pSC-A amp/kan (Agilent Technologies) were used for the synthesis of sense and antisense digoxigenin-labeled RNA probes. For DXS2.1, DXR, MECS and HDS, the previously described plasmids were used for the transcription of riboprobes [32,41]. Briefly, RNA hybridization was performed at room temperature, unless otherwise stated. After rehydration, C. roseus leaf sections were treated with proteinase K (5 µg·mL−1 in 100 mM Tris–HCl and 50 mM EDTA (pH 8.0)) for 30 min at 37 ◦ C, and rinsed twice with TBS150 (10 mM Tris–HCl, 150 mM NaCl (pH 7.5)), followed by blocking of proteinase K with glycine (2 mg·mL−1 in TBS150) for 2 min, and by two rinses in TBS150. After postfixation with 3.7% formaldehyde in PBS during 20 min and two washing steps in TBS150 (5 min), sections were acetylated with acetic anhydride (0.25% in 0.1 M triethanolamine–HCl (pH 8.0)) for 10 min, washed with TBS150, dehydrated in an ethanol series and air-dried. To detect RNA, 120 µL of the hybridization mix (200 ng·mL−1 partially hydrolyzed digoxigenin-labeled RNA transcripts, 40% formamide, 10% dextran sulfate, 1 µg·mL−1 deproteinised yeast total RNA, 0.3 M NaCl, 0.01 M Tris–HCl (pH 6.8), 10 mM sodium phosphate (pH 6.8), 5 mm EDTA and 40 units·mL−1 RNasin ribonuclease inhibitor) were dispersed on the sections before mounting under cover slips to limit evaporation. Hybridization was performed during 16 h at 50 ◦ C in an atmosphere containing 50% formamide. Cover slips were removed from slides through soaking in 2 × SSC at 37 ◦ C (1 × SSC is 0.15 M NaCl and 15 mM sodium citrate). Before detection, slides were treated with RNase A (50 µg·mL−1 in 10 mM Tris–HCl, 0.5 M NaCl, 1 mM EDTA (pH 7.5)) for 30 min at 37 ◦ C followed by washing under gentle agitation in 2 × SSC for 1 h, 1 × SSC for 1 h and in 0.1 × SSC for 1 h at 65 ◦ C. To detect hybridized transcripts, slides were washed in Tween TBS (TTBS) (0.1 M Tris–HCl (pH 8.0), 0.15 M NaCl and 0.3% Triton X-100) for 10 min and blocked with 2% BSA in TTBS for 30 min. Sheep antidigoxigenin Fab fragments–alkaline phosphatase conjugate (Roche) at a 1:200 dilution in a solution of 1% BSA in TTBS was dispensed on the sections and mounted under cover slips. After incubation for 2 h, unbound conjugates were washed two times for 15 min with TTBS and two times for 10 min with AP buffer (0.1 M Tris–HCl (pH 9.5), 0.1 M NaCl and 10 mM MgCl2 ). For color development, slides were immerged in 175 µg·mL−1 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and 350 µg·mL−1 nitro blue tetrazolium chloride in AP buffer for 1–16 h. After development, slides were washed in water, dried and mounted in immersion oil under cover slips. 4.8. Bioinformatic Sequence Analysis The predictions of protein subcellular localization were performed using PSORT [61], Predotar [62], ChloroP 1.1 [63], MitoProt [64] and TargetP 1.1 [65] softwares. Plants 2020, 9, 462 20 of 25 4.9. Coexpression Network Analysis The MPGR RNA seq data [44] were analyzed using Microsoft Excel. DXS2.1 was used as a bait to build a coexpression network with Pearson correlation coefficient for the 23 sample conditions. The network was ranked and filtered for the six other MEP pathway available genes (HDR was not present). A relative red-yellow-green heat map was used to highlight the expression profiles. 4.10. FP-Fusion Constructs for MEP Pathway Enzymes Localization Studies Plasmids expressing GFP and/or YFP fusion proteins of the MEP pathway enzymes including their orientation or truncated variants were constructed using the plasmids and the procedures previously described [36]. Details on primers and on cloning procedures are listed in Table S4 for full-length fusion, in Table S5 for the fusion including the first 100 residues of the DXS2.1, DXS2.2, DXR and HDS and in Table S6 for DXR and HDS internal YFP fusions. 4.11. Organelle Markers For the identification of the subcellular compartments that accumulate the fusion proteins, a set of organelle markers was used in cotransformation experiments with the MEP pathway enzyme constructs. The “plastid”-mcherry (CD3-1000), “plastid”-CFP (CD3-994), “mitochondria”-mcherry (CD3-992), “mitochondria”-CFP (CD3-986), and “peroxisome”-CFP markers, described by [66] were obtained from the ABRC (http://www.arabidopsis.org). The mcherry-GUS cytosolic marker and the mcherry nucleocytosolic marker were described previously [59]. For plastid subcompartment markers, the CAB1 (chlorophyll a/b binding protein 1, At1g29930), TIC40 (translocon at the inner envelope of chloroplasts, At5g16620) and CCD4 (carotenoid cleavage dioxygenase 4, At4g19170) proteins of A. thaliana were used as thylakoid, inner membrane of plastid and plastoglobuli markers, respectively. The plastoglobule localization of CCD4 was previously demonstrated by [67] and [52]. The coding sequence of the three proteins was amplified using specific primers (Table S7) and retrotranscribed from RNA extracted from A. thaliana as a matrix, and subsequently cloned into the pSCA-cassette-CFPi plasmid [59]. 4.12. Cell Transformations and Epifluorescence Microscopy Transient transformation of C. roseus cells by particle bombardment, fluorescence protein imaging and BiFC were performed following the procedures previously described [36,47,48]. Briefly, C. roseus cells plated onto solid culture medium or ditched young leaves were bombarded with gold DNA coated particles (1 µm) and 1100 psi rupture disk at a stopping-screen-to-target distance of 6 cm, using the Bio-Rad PDS1000/He system. Cells were cultivated during 15 h to 48 h and the protein subcellular localization was determined using an Olympus BX-51 epifluorescence microscope equipped with an Olympus DP-71 digital camera. The morphology of the transformed cells was observed with differential interference contrast (DIC). The pattern of localization presented for each protein is representative of circa 100 observed cells. 4.13. Transmission Electron Microscopy Immunogold Labeling Transmission electron microscopy immunogold labeling was performed according to [41] using the anti-DXR antibody described in [68] at 1:5 to 1:10 dilution. The observations were performed using the electron microscopy facilities of the Toulouse Imaging Network. Supplementary Materials: The following are available online at http://www.mdpi.com/2223-7747/9/4/462/s1, Figure S1: A phylogenetic tree was constructed from the amino acid sequence of DXS from the following species. Table S1: Details of primer sequences and cloning procedure used to amplify the DX2.2, CMS, CMK and HDR cDNAs. Table S2: Details of primer sequences and cloning procedures used to express proteins for functional validation. Table S3: The sequence of primers used in qRT–PCR analysis. Table S4: Details of primer sequences and cloning procedure used to express full-length MEP pathway enzyme fused to FP. Table S5: Details of primer sequences and cloning procedure used to fuse the first 100 residues of DXS2.1, DXS2.2, DXR and HDS to GFP. Plants 2020, 9, 462 21 of 25 Table S6: Details of primer sequences and cloning procedures used to create internal YFP fusion for DXR and HDS. Table S7: Details of primer sequences and cloning procedures used to create plastid subcompartment CFP-markers; the CAB1 (chlorophyll a/b binding protein 1, At1g29930), TIC40 (translocon at the inner envelope of chloroplasts, At5g16620), CCD4 (carotenoid cleavage dioxygenase 4, At4g19170) proteins of A. thaliana were used as a marker of thylakoid, inner membrane of plastid and plastoglobules, respectively. The coding sequence of these proteins was amplified using retrotranscribed RNA extracted from A. thaliana as a matrix. Author Contributions: Conceptualization, B.S.-P., V.B. and V.C.; investigation, G.G., A.G., C.P., M.P., S.M., A.O., T.D.d.B., S.B., A.L., N.P., M.R.-C., V.B. and V.C.; writing—original draft preparation, V.C.; writing—review and editing, N.P., N.G.-G., B.S.-P., M.R.-C., V.B. and V.C.; supervision, V.B. and V.C.; funding acquisition, V.C. All authors have read and agreed to the published version of the manuscript. Funding: We acknowledge funding from ARD2020 Biopharmaceutical program of the Région Centre Val de Loire (BioPROPHARM and CatharSIS projects). G.G. acknowledges a research fellowship of Le Studium-Institute for Advanced Studies, Loire Valley, Orléans, France. Conflicts of Interest: The authors declare no conflict of interest. References 1. Bouvier, F.; Rahier, A.; Camara, B. Biogenesis, molecular regulation and function of plant isoprenoids. Prog. Lipid Res. 2005, 44, 357–429. [CrossRef] [PubMed] 2. Gershenzon, J.; Dudareva, N. The function of terpene natural products in the natural world. Nat. Chem. Biol. 2007, 3, 408–414. [CrossRef] 3. Guirimand, G.; Courdavault, V.; St-Pierre, B.; Burlat, V. Biosynthesis and Regulation of Alkaloids. In Plant Developmental Biology-Biotechnological Perspectives: Volume 2; Pua, E.C., Davey, M.R., Eds.; Springer: Berlin/Heidelberg, Germany, 2010; pp. 139–160. ISBN 9783642046704. 4. O’Connor, S.E.; Maresh, J.J. Chemistry and biology of monoterpene indole alkaloid biosynthesis. Nat. Prod. Rep. 2006, 23, 532–547. [CrossRef] [PubMed] 5. Dugé De Bernonville, T.; Carqueijeiro, I.; Lanoue, A.; Lafontaine, F.; Sánchez Bel, P.; Liesecke, F.; Musset, K.; Oudin, A.; Glévarec, G.; Pichon, O.; et al. Folivory elicits a strong defense reaction in Catharanthus roseus: Metabolomic and transcriptomic analyses reveal distinct local and systemic responses. Sci. Rep. 2017, 7, 1–14. [CrossRef] [PubMed] 6. Guirimand, G.; Courdavault, V.; Lanoue, A.; Mahroug, S.; Guihur, A.; Blanc, N.; Giglioli-Guivarc’h, N.; St-Pierre, B.; Burlat, V. Strictosidine activation in Apocynaceae: Towards a “nuclear time bomb”? BMC Plant Biol. 2010, 10, 182. [CrossRef] 7. Luijendijk, T.J.C.; Van Der Meijden, E.D.; Verpoorte, R. Involvement of strictosidine as a defensive chemical in Catharanthus roseus. J. Chem. Ecol. 1996, 22, 1355–1366. [CrossRef] 8. Facchini, P.J.; De Luca, V. Opium poppy and Madagascar periwinkle: Model non-model systems to investigate alkaloid biosynthesis in plants. Plant J. 2008, 54, 763–784. [CrossRef] 9. Kellner, F.; Geu-Flores, F.; Sherden, N.H.; Brown, S.; Foureau, E.; Courdavault, V.; O’Connor, S.E. Discovery of a P450-catalyzed step in vindoline biosynthesis: A link between the aspidosperma and eburnamine alkaloids. Chem. Commun. 2015, 51, 7626–7628. [CrossRef] 10. Qu, Y.; Easson, M.L.A.E.; Froese, J.; Simionescu, R.; Hudlicky, T.; DeLuca, V. Completion of the seven-step pathway from tabersonine to the anticancer drug precursor vindoline and its assembly in yeast. Proc. Natl. Acad. Sci. USA 2015, 112, 6224–6229. [CrossRef] 11. Tatsis, E.C.; Carqueijeiro, I.; Dugé de Bernonville, T.; Franke, J.; Dang, T.-T.T.; Oudin, A.; Lanoue, A.; Lafontaine, F.; Stavrinides, A.K.; Clastre, M.; et al. A three enzyme system to generate the Strychnos alkaloid scaffold from a central biosynthetic intermediate. Nat. Commun. 2017, 8, 316. [CrossRef] 12. Caputi, L.; Franke, J.; Farrow, S.C.; Chung, K.; Payne, R.M.E.; Nguyen, T.-D.; Dang, T.-T.T.; Soares Teto Carqueijeiro, I.; Koudounas, K.; Dugé de Bernonville, T.; et al. Missing enzymes in the biosynthesis of the anticancer drug vinblastine in Madagascar periwinkle. Science 2018, 360, 1235–1239. [CrossRef] [PubMed] 13. Qu, Y.; Easson, M.E.A.M.; Simionescu, R.; Hajicek, J.; Thamm, A.M.K.; Salim, V.; De Luca, V. Solution of the multistep pathway for assembly of corynanthean, strychnos, iboga, and aspidosperma monoterpenoid indole alkaloids from 19E-geissoschizine. Proc. Natl. Acad. Sci. USA 2018, 115, 3180–3185. [CrossRef] [PubMed] Plants 2020, 9, 462 22 of 25 14. Carqueijeiro, I.; Langley, C.; Grzech, D.; Koudounas, K.; Papon, N.; O’Connor, S.E.; Courdavault, V. Beyond the semi-synthetic artemisinin: Metabolic engineering of plant-derived anti-cancer drugs. Curr. Opin. Biotechnol. 2020, 65, 17–24. [CrossRef] [PubMed] 15. Dugé de Bernonville, T.; Papon, N.; Clastre, M.; O’Connor, S.E.; Courdavault, V. Identifying Missing Biosynthesis Enzymes of Plant Natural Products. Trends Pharmacol. Sci. 2020, 41, 142–146. [CrossRef] 16. Courdavault, V.; O’Connor, S.E.; Oudin, A.; Besseau, S.P.N. Towards the Microbial Production of Plant-Derived Anticancer Drugs. Trends Cancer 2020, in press. [CrossRef] 17. St-Pierre, B.; Besseau, S.; Clastre, M.; Courdavault, V.; Courtois, M.; Crèche, J.; Ducos, E.; de Bernonville, T.D.; Dutilleul, C.; Glévarec, G.; et al. Deciphering the Evolution, Cell Biology and Regulation of Monoterpene Indole Alkaloids. Adv. Bot. Res. 2013, 68, 73–109. 18. De Luca, V.; Marineau, C.; Brisson, N. Molecular cloning and analysis of cDNA encoding a plant tryptophan decarboxylase: Comparison with animal dopa decarboxylases. Proc. Natl. Acad. Sci. USA 1989, 86, 2582–2586. [CrossRef] 19. Miettinen, K.; Dong, L.; Navrot, N.; Schneider, T.; Burlat, V.; Pollier, J.; Woittiez, L.; van der Krol, S.; Lugan, R.; Ilc, T.; et al. The seco-iridoid pathway from Catharanthus roseus. Nat. Commun. 2014, 5, 3606. [CrossRef] 20. Dugé de Bernonville, T.; Foureau, E.; Parage, C.; Lanoue, A.; Clastre, M.; Londono, M.A.; Oudin, A.; Houillé, B.; Papon, N.; Besseau, S.; et al. Characterization of a second secologanin synthase isoform producing both secologanin and secoxyloganin allows enhanced de novo assembly of a Catharanthus roseus transcriptome. BMC Genom. 2015, 16, 619. [CrossRef] 21. Contin, A.; van der Heijden, R.; Lefeber, A.W.; Verpoorte, R. The iridoid glucoside secologanin is derived from the novel triose phosphate/pyruvate pathway in a Catharanthus roseus cell culture. FEBS Lett. 1998, 434, 413–416. [CrossRef] 22. Walter, M.H.; Hans, J.; Strack, D. Two distantly related genes encoding 1-deoxy-d-xylulose 5-phosphate synthases: Differential regulation in shoots and apocarotenoid-accumulating mycorrhizal roots. Plant J. 2002, 31, 243–254. [CrossRef] [PubMed] 23. Phillips, M.A.; Walter, M.H.; Ralph, S.G.; Dabrowska, P.; Luck, K.; Urós, E.M.; Boland, W.; Strack, D.; Rodríguez-Concepción, M.; Bohlmann, J.; et al. Functional identification and differential expression of 1-deoxy-D-xylulose 5-phosphate synthase in induced terpenoid resin formation of Norway spruce (Picea abies). Plant Mol. Biol. 2007, 65, 243–257. [CrossRef] [PubMed] 24. Cordoba, E.; Porta, H.; Arroyo, A.; San Román, C.; Medina, L.; Rodríguez-Concepción, M.; León, P. Functional characterization of the three genes encoding 1-deoxy-D-xylulose 5-phosphate synthase in maize. J. Exp. Bot. 2011, 62, 2023–2038. [CrossRef] [PubMed] 25. Kim, B.-R.; Kim, S.-U.; Chang, Y.-J. Differential expression of three 1-deoxy-D: -xylulose-5-phosphate synthase genes in rice. Biotechnol. Lett. 2005, 27, 997–1001. [CrossRef] [PubMed] 26. Okada, A.; Shimizu, T.; Okada, K.; Kuzuyama, T.; Koga, J.; Shibuya, N.; Nojiri, H.; Yamane, H. Elicitor induced activation of the methylerythritol phosphate pathway toward phytoalexins biosynthesis in rice. Plant Mol. Biol. 2007, 65, 177–187. [CrossRef] 27. Rohdich, F.; Zepeck, F.; Adam, P.; Hecht, S.; Kaiser, J.; Laupitz, R.; Gräwert, T.; Amslinger, S.; Eisenreich, W.; Bacher, A.; et al. The deoxyxylulose phosphate pathway of isoprenoid biosynthesis: Studies on the mechanisms of the reactions catalyzed by IspG and IspH protein. Proc. Natl. Acad. Sci. USA 2003, 100, 1586–1591. [CrossRef] 28. Ramos-Valdivia, A.C.; van der Heijden, R.; Verpoorte, R. Isopentenyl diphosphate isomerase: A core enzyme in isoprenoid biosynthesis. A review of its biochemistry and function. Nat. Prod. Rep. 1997, 14, 591–603. [CrossRef] 29. Simkin, A.J.; Miettinen, K.; Claudel, P.; Burlat, V.; Guirimand, G.; Courdavault, V.; Papon, N.; Meyer, S.; Godet, S.; St-Pierre, B.; et al. Characterization of the plastidial geraniol synthase from Madagascar periwinkle which initiates the monoterpenoid branch of the alkaloid pathway in internal phloem associated parenchyma. Phytochemistry 2013, 85, 36–43. [CrossRef] 30. Courdavault, V.; Papon, N.; Clastre, M.; Giglioli-Guivarc’h, N.; St-Pierre, B.; Burlat, V. A look inside an alkaloid multisite plant: The Catharanthus logistics. Curr. Opin. Plant Biol. 2014, 19, 43–50. [CrossRef] 31. Dugé de Bernonville, T.; Clastre, M.; Besseau, S.; Oudin, A.; Burlat, V.; Glévarec, G.; Lanoue, A.; Papon, N.; Giglioli-Guivarc’h, N.; St-Pierre, B.; et al. Phytochemical genomics of the Madagascar periwinkle: Unravelling the last twists of the alkaloid engine. Phytochemistry 2015, 113, 9–23. [CrossRef] Plants 2020, 9, 462 23 of 25 32. Burlat, V.; Oudin, A.; Courtois, M.; Rideau, M.; St-Pierre, B. Co-expression of three MEP pathway genes and geraniol 10-hydroxylase in internal phloem parenchyma of Catharanthus roseus implicates multicellular translocation of intermediates during the biosynthesis of monoterpene indole alkaloids and isoprenoid-derive. Plant J. 2004, 38, 131–141. [CrossRef] [PubMed] 33. Geu-Flores, F.; Sherden, N.H.; Courdavault, V.; Burlat, V.; Glenn, W.S.; Wu, C.; Nims, E.; Cui, Y.; O’Connor, S.E. An alternative route to cyclic terpenes by reductive cyclization in iridoid biosynthesis. Nature 2012, 492, 138–142. [CrossRef] [PubMed] 34. Guirimand, G.; Guihur, A.; Phillips, M.A.; Oudin, A.; Glévarec, G.; Melin, C.; Papon, N.; Clastre, M.; St-Pierre, B.; Rodríguez-Concepción, M.; et al. A single gene encodes isopentenyl diphosphate isomerase isoforms targeted to plastids, mitochondria and peroxisomes in Catharanthus roseus. Plant Mol. Biol. 2012, 79, 443–459. [CrossRef] [PubMed] 35. Oudin, A.; Courtois, M.; Rideau, M.; Clastre, M. The iridoid pathway in Catharanthus roseus alkaloid biosynthesis. Phytochem. Rev. 2007, 6, 259–276. [CrossRef] 36. Guirimand, G.; Burlat, V.; Oudin, A.; Lanoue, A.; St-Pierre, B.; Courdavault, V. Optimization of the transient transformation of Catharanthus roseus cells by particle bombardment and its application to the subcellular localization of hydroxymethylbutenyl 4-diphosphate synthase and geraniol 10-hydroxylase. Plant Cell Rep. 2009, 28, 1215–1234. [CrossRef] [PubMed] 37. Hanson, M.R.; Hines, K.M. Stromules: Probing Formation and Function. Plant Physiol. 2018, 176, 128–137. [CrossRef] [PubMed] 38. Mueller-Schuessele, S.J.; Michaud, M. Plastid Transient and Stable Interactions with Other Cell Compartments. Methods Mol. Biol. 2018, 1829, 87–109. 39. Chahed, K.; Oudin, A.; Giglioli-Guivarc’h, N.; Hamdi, S.; Chénieux, J.C.; Rideau, M.; Clastre, M. 1-Deoxy-d-xylulose 5-phosphate synthase from periwinkle: cDNA identification and induced gene expression in terpenoid indole alkaloid-producing cells. Plant Physiol. Biochem. 2000, 38, 559–566. [CrossRef] 40. Veau, B.; Courtois, M.; Oudin, A.; Chenieux, J.C.; Rideau, M.; Clastre, M. Cloning and expression of cDNAs encoding two enzymes of the MEP pathway in Catharanthus roseus. Biochim. Biophys. Acta 2000, 1517, 159–163. [CrossRef] 41. Oudin, A.; Mahroug, S.; Courdavault, V.; Hervouet, N.; Zelwer, C.; Rodríguez-Concepción, M.; St-Pierre, B.; Burlat, V. Spatial distribution and hormonal regulation of gene products from methyl erythritol phosphate and monoterpene-secoiridoid pathways in Catharanthus roseus. Plant Mol. Biol. 2007, 65, 13–30. [CrossRef] 42. Campos, N.; Rodríguez-Concepción, M.; Sauret-Güeto, S.; Gallego, F.; Lois, L.M.; Boronat, A. Escherichia coli engineered to synthesize isopentenyl diphosphate and dimethylallyl diphosphate from mevalonate: A novel system for the genetic analysis of the 2-C-methyl-d-erythritol 4-phosphate pathway for isoprenoid biosynthesis. Biochem. J. 2001, 353, 59–67. [CrossRef] [PubMed] 43. Sauret-Güeto, S.; Ramos-Valdivia, A.; Ibáñez, E.; Boronat, A.; Rodríguez-Concepción, M. Identification of lethal mutations in Escherichia coli genes encoding enzymes of the methylerythritol phosphate pathway. Biochem. Biophys. Res. Commun. 2003, 307, 408–415. [CrossRef] 44. Medicinal Plant Genomics Ressource MPGR RNA seq Data. Available online: http://medicinalplantgenomics. msu.edu/index.shtml (accessed on 10 March 2020). 45. Courdavault, V.; Thiersault, M.; Courtois, M.; Gantet, P.; Oudin, A.; Doireau, P.; St-Pierre, B.; Giglioli-Guivarc’h, N. CaaX-prenyltransferases are essential for expression of genes involved in the early stages of monoterpenoid biosynthetic pathway in Catharanthus roseus cells. Plant Mol. Biol. 2005, 57, 855–870. [CrossRef] [PubMed] 46. Leménager, D.; Ouelhazi, L.; Mahroug, S.; Veau, B.; St-Pierre, B.; Rideau, M.; Aguirreolea, J.; Burlat, V.; Clastre, M. Purification, molecular cloning, and cell-specific gene expression of the alkaloid-accumulation associated protein CrPS in Catharanthus roseus. J. Exp. Bot. 2005, 56, 1221–1228. [CrossRef] [PubMed] 47. Carqueijeiro, I.; Sepúlveda, L.J.; Mosquera, A.; Payne, R.; Corbin, C.; Papon, N.; de Bernonville, T.D.; Besseau, S.; Lanoue, A.; Glévarec, G.; et al. Vacuole-Targeted Proteins: Ins and Outs of Subcellular Localization Studies. Methods Mol. Biol. 2018, 1789, 33–54. 48. Foureau, E.; Carqueijeiro, I.; Dugé de Bernonville, T.; Melin, C.; Lafontaine, F.; Besseau, S.; Lanoue, A.; Papon, N.; Oudin, A.; Glévarec, G.; et al. Prequels to Synthetic Biology: From Candidate Gene Identification and Validation to Enzyme Subcellular Localization in Plant and Yeast Cells. Methods Enzymol. 2016, 576, 167–206. Plants 2020, 9, 462 24 of 25 49. Besseau, S.; Kellner, F.; Lanoue, A.; Thamm, A.M.K.; Salim, V.; Schneider, B.; Geu-Flores, F.; Höfer, R.; Guirimand, G.; Guihur, A.; et al. A pair of tabersonine 16-hydroxylases initiates the synthesis of vindoline in an organ-dependent manner in Catharanthus roseus. Plant Physiol. 2013, 163, 1792–1803. [CrossRef] 50. Parage, C.; Foureau, E.; Kellner, F.; Burlat, V.; Mahroug, S.; Lanoue, A.; Dugé de Bernonville, T.; Londono, M.A.; Carqueijeiro, I.; Oudin, A.; et al. Class II Cytochrome P450 Reductase Governs the Biosynthesis of Alkaloids. Plant Physiol. 2016, 172, 1563–1577. [CrossRef] 51. Larsen, B.; Fuller, V.L.; Pollier, J.; Van Moerkercke, A.; Schweizer, F.; Payne, R.; Colinas, M.; O’Connor, S.E.; Goossens, A.; Halkier, B.A. Identification of Iridoid Glucoside Transporters in Catharanthus roseus. Plant Cell Physiol. 2017, 58, 1507–1518. [CrossRef] 52. Ytterberg, A.J.; Peltier, J.-B.; van Wijk, K.J. Protein profiling of plastoglobules in chloroplasts and chromoplasts. A surprising site for differential accumulation of metabolic enzymes. Plant Physiol. 2006, 140, 984–997. [CrossRef] 53. Perello, C.; Llamas, E.; Burlat, V.; Ortiz-Alcaide, M.; Phillips, M.A.; Pulido, P.; Rodriguez-Concepcion, M. Differential Subplastidial Localization and Turnover of Enzymes Involved in Isoprenoid Biosynthesis in Chloroplasts. PLoS ONE 2016, 11, e0150539. [CrossRef] [PubMed] 54. Pulido, P.; Llamas, E.; Llorente, B.; Ventura, S.; Wright, L.P.; Rodríguez-Concepción, M. Specific Hsp100 Chaperones Determine the Fate of the First Enzyme of the Plastidial Isoprenoid Pathway for Either Refolding or Degradation by the Stromal Clp Protease in Arabidopsis. PLoS Genet. 2016, 12, e1005824. [CrossRef] [PubMed] 55. Rai, A.; Smita, S.S.; Singh, A.K.; Shanker, K.; Nagegowda, D.A. Heteromeric and homomeric geranyl diphosphate synthases from Catharanthus roseus and their role in monoterpene indole alkaloid biosynthesis. Mol. Plant 2013, 6, 1531–1549. [CrossRef] [PubMed] 56. Schattat, M.; Barton, K.; Baudisch, B.; Klösgen, R.B.; Mathur, J. Plastid stromule branching coincides with contiguous endoplasmic reticulum dynamics. Plant Physiol. 2011, 155, 1667–1677. [CrossRef] [PubMed] 57. Hanson, M.R.; Sattarzadeh, A. Trafficking of proteins through plastid stromules. Plant Cell 2013, 25, 2774–2782. [CrossRef] [PubMed] 58. Mehrshahi, P.; Stefano, G.; Andaloro, J.M.; Brandizzi, F.; Froehlich, J.E.; DellaPenna, D. Transorganellar complementation redefines the biochemical continuity of endoplasmic reticulum and chloroplasts. Proc. Natl. Acad. Sci. USA 2013, 110, 12126–12131. [CrossRef] [PubMed] 59. Guirimand, G.; Guihur, A.; Ginis, O.; Poutrain, P.; Héricourt, F.; Oudin, A.; Lanoue, A.; St-Pierre, B.; Burlat, V.; Courdavault, V. The subcellular organization of strictosidine biosynthesis in Catharanthus roseus epidermis highlights several trans-tonoplast translocations of intermediate metabolites. FEBS J. 2011, 278, 749–763. [CrossRef] 60. Mahroug, S.; Courdavault, V.; Thiersault, M.; St-Pierre, B.; Burlat, V. Epidermis is a pivotal site of at least four secondary metabolic pathways in Catharanthus roseus aerial organs. Planta 2006, 223, 1191–1200. [CrossRef] 61. PSORT, Kenta Nakai; Human Genome Center, Institute for Medical Science, University ot Tokyo: Tokyo, Japan, 2007. 62. Predotar, version 1.3; Unité de Recherches en Génomique Info: Versailles, France, 2010. 63. Emanuelsson, O.; Nielsen, H.; Von Heijne, G. ChloroP, a neural network-based method for predicting chloroplast transit peptides and their cleavage sites. Protein Sci. 1999, 8, 978–984. [CrossRef] 64. Ison, J.; Rapacki, K.; Ménager, H.; Kalaš, M.; Rydza, E.; Chmura, P.; Booth, T. Tools and data services registry: A community effort to document bioinformatics resources. Nucleic Acids Res. 2016, 44, D38–D47. [CrossRef] 65. Emanuelsson, O.; Nielsen, H.; Brunak, S.; Von Heijne, G. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. J. Mol. Biol. 2000, 300, 1005–1016. [CrossRef] [PubMed] 66. Nelson, B.K.; Cai, X.; Nebenführ, A. A multicolored set of in vivo organelle markers for co-localization studies in Arabidopsis and other plants. Plant J. 2007, 51, 1126–1136. [CrossRef] [PubMed] 67. Vidi, P.A.; Kanwischer, M.; Baginsky, S.; Austin, J.R.; Csucs, G.; Dörmann, P.; Kessler, F.; Bréhélin, C. Tocopherol cyclase (VTE1) localization and vitamin E accumulation in chloroplast plastoglobule lipoprotein particles. J. Biol. Chem. 2006, 281, 11225–11234. [CrossRef] [PubMed] Plants 2020, 9, 462 25 of 25 68. Pulido, P.; Toledo-Ortiz, G.; Phillips, M.A.; Wright, L.P.; Rodríguez-Concepción, M. Arabidopsis J-protein J20 delivers the first enzyme of the plastidial isoprenoid pathway to protein quality control. Plant Cell 2013, 25, 4183–4194. [CrossRef] [PubMed] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).