Research Report

Isolation of Cells Specialized in Anticancer AlkaloidMetabolism by Fluorescence-Activated Cell Sorting1[OPEN]

Inês Carqueijeiro2,3, Ana Luísa Guimarães2,4, Sara Bettencourt, Teresa Martínez-Cortés, Joana G. Guedes,Rui Gardner, Telma Lopes, Cláudia Andrade, Cláudia Bispo, Nuno Pimpão Martins, Paula Andrade,Patrícia Valentão, Inês M. Valente, José A. Rodrigues, Patrícia Duarte, and Mariana Sottomayor*

CIBIO/InBIO-Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto,4485-661 Vairão, Portugal (S.B., T.M.-C., J.G.G., M.S.); Instituto de Investigação e Inovação em Saúde, Institutode Biologia Molecular e Celular, Universidade do Porto, 4200-135 Porto, Portugal (I.C., A.L.G., P.D.);Departamento de Biologia, Faculdade de Ciências da Universidade do Porto, 4169-007 Porto, Portugal (I.C.,M.S.); Instituto Gulbenkian de Ciência, 2780-156 Oeiras, Portugal (R.G., T.L., C.A., C.B., N.P.M.);REQUIMTE/Laboratório de Farmacognosia, Departamento de Química, Faculdade de Farmácia, Universidadedo Porto, 4050-313 Porto, Portugal (P.A., P.V.); and REQUIMTE/LAQV, Departamento de Química eBioquímica, Faculdade de Ciências, Universidade do Porto, 4169-007 Porto, Portugal (I.M.V., J.A.R.)

ORCID IDs: 0000-0001-9576-5556 (I.C.); 0000-0002-5162-6359 (S.B.); 0000-0002-7270-5136 (R.G.); 0000-0002-8104-0957 (N.P.M.);0000-0002-0740-4396 (P.V.); 0000-0003-1438-1353 (I.M.V.); 0000-0001-5893-3362 (P.D.); 0000-0003-2411-710X (M.S.).

Plant specialized metabolism often presents a complex cell-specific compartmentation essential to accomplish the biosynthesis ofvaluable plant natural products. Hence, the disclosure and potential manipulation of such pathways may depend on thecapacity to isolate and characterize specific cell types. Catharanthus roseus is the source of several medicinal terpenoid indolealkaloids, including the low-level anticancer vinblastine and vincristine, for which the late biosynthetic steps occur in specializedmesophyll cells called idioblasts. Here, the optical, fluorescence, and alkaloid-accumulating properties of C. roseus leaf idioblastsare characterized, and a methodology for the isolation of idioblast protoplasts by fluorescence-activated cell sorting isestablished, taking advantage of the distinctive autofluorescence of these cells. This achievement represents a crucial step forthe development of differential omic strategies leading to the identification of candidate genes putatively involved in thebiosynthesis, pathway regulation, and transmembrane transport leading to the anticancer alkaloids from C. roseus.

Plants depict a uniquemetabolic plasticity includinga wealth of natural products that help them to dealwith a wide range of environmental threats. Many ofthose specialized metabolites have valuable applica-tions as medicines, colorants, flavors, fragrances, etc.,but their exploitation is often hindered by their lowlevels and limited availability of plant material. Theintrinsic complexity of specialized pathways has beena major challenge to their full characterization, butit can now deliver potent tools, if combined with therecent development of omic technologies. Therefore,the association of differential omic screenings to theregulation of specialized pathways by environmental/developmental cues and to their organ/tissue-specificorganization is enabling the fast disclosure of importantparts of long time searched pathways (Miettinen et al.,2014; Stavrinides et al., 2015). However, there is a level ofcomplexity that is particularly challenging concerningthe application of these approaches—the expression ofdifferent parts of the pathway in specific cell typeswithin a tissue or organ. The low relative abundance ofthose cell types within the respective plant organs ham-pers the retrieval of candidate genes, but if isolation ofsuch cells is achieved, it constitutes a powerful tool. Twomajor techniques may be explored to this end: laser mi-crodissection and fluorescence-activated cell sorting(FACS). The use of laser microdissection for plant

microsampling has increased significantly over the pastdecade (Fang and Schneider, 2014). FACS, however, inspite of being a method possibly more cost- and time-effective for obtaining sampleswith quantity and qualityguaranteeing successful cellular “omics,” has beenmostly applied to the sorting of protoplasts from Ara-bidopsis (Arabidopsis thaliana) root and shoot apical cellsspecifically labeled with GFP (Birnbaum et al., 2005;Galbraith, 2010; Carter et al., 2013).

A paradigmatic example of cell-specific compart-mentation involving autofluorescence is the terpenoidindole alkaloid (TIA) pathway in the medicinal plantCatharanthus roseus (Supplemental Fig. S1), whoseleaves accumulate in low levels the anticancer TIAsvinblastine (VLB) and vincristine (De Luca et al., 2014).The TIA pathway presents multicellular compart-mentation in C. roseus leaves, with early and middlesteps being expressed in internal phloem-associatedparenchyma and the epidermis, whereas late stepsare expressed in laticifers and idioblasts, character-ized by a conspicuous fluorescence (St-Pierre et al.,1999; Courdavault et al., 2014). A diversity of C. roseustranscriptomic projects (MPGR, Phytometasyn, Cathacyc)has enabled a recent burst of research in the TIApathway, leading to the full characterization of itsiridoid/terpenoid precursor part, of late biosyntheticsteps, and of the first TIA transporter (Yu and DeLuca,

Plant Physiology�, August 2016, Vol. 171, pp. 2371–2378, www.plantphysiol.org � 2016 American Society of Plant Biologists. All Rights Reserved. 2371 www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from

Copyright © 2016 American Society of Plant Biologists. All rights reserved.

2013; Miettinen et al., 2014; Stavrinides et al., 2015;Kellner et al., 2015). However, the last bottleneck partof the pathway leading to the anticancer alkaloids re-mains elusive, as well as most TIA transmembranetransporters and transcriptional regulators affectingVLB/vincristine levels.

This study describes the distinctive properties of C.roseus leaf idioblasts and the optimization of FACSconditions allowing the successful separation of idi-oblast protoplasts from those of common mesophyllcells. The amount/quality of cells sorted and thebuffer composition makes the sorting methodologydeveloped here compatible with the implementationof subsequent transcriptomic, metabolomic, andproteomic studies, crucial to fully unravel the TIApathway.

RESULTS AND DISCUSSION

Idioblasts and Laticifers from C. roseus Leaves AreDifferentially Stained by Alkaloid Indicators and Depict aCharacteristic Autofluorescence

Infiltration of leaves with both the Dragendorff rea-gent and the Mayer reagent, widely used to identifyalkaloids, characteristically labeled a particular set ofcells with distinct scattered patterns in the adaxial andabaxial sides of the leaves, previously identified as id-ioblasts (Fig. 1; Yoder and Mahlberg, 1976; Mersey andCutler, 1986; St-Pierre et al., 1999). Labeling was alsoassociated with small veins, likely corresponding toreactionwith alkaloids accumulated in the unbranched,nonarticulated laticifers described byYoder andMahlberg(1976). Focusing of different planes of the leaf surfaceenabled to conclude that the epidermal cells did notreact with the reagents and that the two differentabaxial/adaxial patterns of idioblasts likely corre-sponded to the spatial organization of such cells in thetwo respective underlying tissues, the spongy and thepalisade parenchymas, as could also be concluded fromthe shape of the cells (Fig. 1, B, D, F, and H). This dif-ferential accumulation of alkaloids in laticifers and id-ioblasts of C. roseus leaves was previously reported byYoder and Mahlberg (1976) and is also supported bythe work of Mersey and Cutler (1986), who detected a

Figure 1. Bright-field optical microscopy images of C. roseus whole leaves infiltrated with alkaloid indicators reveal a positivereaction in idioblasts (stained isolated mesophyll cells) and laticifers (stained threads). A to D, Infiltration with the Dragendorffreagent reveals alkaloid-accumulating cells stained in brown. E to H, Infiltration with the Mayer reagent reveals alkaloid-accumulating cells stained in dark gray. A, B, E, and F, Images of the adaxial face. C, D, G, and H, Images of the abaxial face.Focusing of leaf images was performed based on the detection of stained cells, which appeared only below the epidermis celllayer. Bars = 200 mm (A, C, E, and G) and 50 mm (B, D, F, and H).

1 This work was supported by the following: (1) FEDER fundsthrough the Operational Competitiveness Programme COMPETEand National Funds through Fundação para a Ciência e a Tecnologia(FCT) under the projects FCOMP-01-0124-FEDER-037277 (Pest-C/SAU/LA0002/2013) , FCOMP-01-0124-FEDER-019664(PTDC/BIA-BCM/119718/2010), and FCOMP-01-0124-FEDER-028125 (PTDC/BBB-BIO/2231/2012); (2) FEDER funds and FCT/MEC national funds, under the Partnership Agreement PT2020,within the project UID/QUI/50006/2013-POCI/01/0145/FERDER/007265; (3) FCT scholarships cosupported by FCT and theEuropean Social Fund, SFRH/BD/4190/2007 (I.C.), SFRH/BD/97590/2013 (J.G.G.), and SFRH/BPD/111181/2015 (I.M.V.); and (4)a Scientific Mecenate Grant from Grupo Jerónimo Martins SA.

2 These authors contributed equally to the article.3 Present address: EA2106 Biomolécules et Biotechnologies

Végétales, Département de Biologie et Physiologie Végétales, UFRPharmacologie, Université François-Rabelais de Tours, Parc deGrandmont, 37200 Tours, France.

4 Present address: Berlin Institute for Medical Systems Biology, MaxDelbrück Center for Molecular Medicine in the Helmholtz Association,Robert-Rössle-Strasse 10, H. 89, Rm. 2.12, 13125 Berlin, Germany.

* Address correspondence to mssottom@fc.up.pt.The author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Mariana Sottomayor (mssottom@fc.up.pt).

M.S. conceived the research, supervised most experiments, anddesigned and performed experiments; I.C., A.L.G., S.B., T.M.-C.,J.G.G., T.L., C.A., C.B., N.P.M., I.M.V., and P.D. designed and/orperformed different sets of experiments; P.A. and P.V. providedtechnical assistance; R.G. and J.A.R. supervised specific experi-ments; A.L.G. prepared most of the figures; M.S. wrote the manu-script with contributions of all the authors.

[OPEN] Articles can be viewed without a subscription.www.plantphysiol.org/cgi/doi/10.1104/pp.16.01028

2372 Plant Physiol. Vol. 171, 2016

Carqueijeiro et al.

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

higher level of TIAs in cell fractions enriched in idio-blast protoplasts. Therefore, our results confirm theimportance of idioblasts as alkaloid accumulation tar-gets, with their accumulation capacity likely influenc-ing the whole metabolic flux of the pathway. As such,idioblasts are certainly home of key alkaloid trans-porters and possibly also of some key biosyntheticsteps, as already shown by St-Pierre et al. (1999).Under the fluorescence microscope, the exact same

pattern of cells reactingwith alkaloid indicators depicteda distinctive vacuolar autofluorescence characterized byblue and green emissions upon UV excitation (Fig. 2).The strong autofluorescence of idioblasts observed inthis study was also previously reported, albeit withsomewhat different emission profiles (Mersey andCutler,1986; St-Pierre et al., 1999). Considering the fluo-rescence excitation and emission spectra of the main

C. roseus TIAs determined by Renaudin (1985), only theTIA serpentinemay be partially contributing to the blueautofluorescence observed in this study, with the bulkof the autofluorescence signal, namely in the greenspectrum, being due to unknown compounds.

Overall, these results indicated the presence inC. roseusleaves of cells specifically accumulating alkaloids anddepicting a distinctive autofluorescence that makes thempotentially amenable to FACS approaches.

FACS Allows the Isolation of a Highly EnrichedPopulation of C. roseus Leaf Idioblasts Suitable forDifferential Omic Screenings

To make possible the sorting of the fluorescent idio-blasts, the cell wall of C. roseus leaves was digested with

Figure 2. Epifluorescence micros-copy images of C. roseus wholeleaves. A, Image of the adaxial face.B, Image of the abaxial face. 1, Redchannel revealing autofluorescenceof chlorophyll in chloroplasts. 2,Blue channel revealing the presenceof an autofluorescence signal in id-ioblasts. 3, Green channel revealingthe presence of an autofluorescencesignal in idioblasts. 4, Merged im-ages of the three channels. Bars =100 mm.

Figure 3. A, Bright-field image (A1) of C. roseus leaf protoplasts and the corresponding fluorescence image showing the viableprotoplasts that incorporated and modified fluorescein diacetate to generate green fluorescence (A2). B, Bright-field image of pro-toplasts treated with the Dragendorff reagent showing an idioblast with the vacuole content stained in brown. C, Bright-field image ofprotoplasts treatedwith theMayer reagent showing an idioblastwith the vacuole content stained in dark gray (manyprotoplasts burst inthe presence of this reagent). D, Bright-field images of protoplasts focused at themidplane (D1) and above themidplane (D2) showingan idioblast with a distinctive bright halo. E, Bright-field and epifluorescencemicroscopy images of leaf protoplasts with cells: focusedat the midplane (E1), focused above the midplane showing that idioblasts display a distinctive bright halo (E2), observed with the bluechannel revealing idioblast protoplasts (E3), observed with the green channel revealing idioblast protoplasts (E4), and represented bythe merged image of E1, E3, and E4 (E5). Bars = 10 mm (A, B, and D) and 20 mm (C and E).

Plant Physiol. Vol. 171, 2016 2373

Isolation of Mesophyll Specialized Cells by FACS

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

an optimized protocol that enabled the production of ahigh yield of pure and stable protoplasts (Fig. 3A).Treatment with the Mayer and Dragendorff reagentsindicated that idioblast protoplasts retained their dis-tinctive capacity to accumulate alkaloids (Fig. 3, B andC). Furthermore, it was possible to observe that those

reagents stained specifically the vacuole, where alka-loids are known to accumulate. Interestingly, under thebright-field microscope and when focused above theirmidplane, idioblasts presented a distinctive bright halo,likely resulting from a higher refraction index of thevacuole content (Fig. 3, D and E). This further indicated

Figure 4. Autofluorescence emissionspectra of C. roseus mesophyll cells. A,Confocal (A1) and bright-field (A2) im-ages of the adaxial face of a wholemounted leaf focused at the level of thepalisade parenchyma. B, Confocal (B1)and bright-field (B2) images of the ab-axial face of a whole mounted leaffocused at the level of the spongyparenchyma. C, Confocal (C1) andbright-field (C2) images of mesophyllprotoplasts. D, Emission spectra of thecells indicated in A, B, and C upon ex-citation with a 405-nm laser. Bars =50 mm (A and B) and 20 mm (C).

Figure 5. Dot plots obtained during sorting of C. roseus leaf protoplasts. A, Dot plot of red versus green emissions with indicationof the two populations selected as common mesophyll cells (I) and idioblasts (II), based on bright field and fluorescence mi-croscopy of the sorted cells (Fig. 6). Events corresponding to debriswere excluded from the dot plot. B, Dot plots of FSC versus SSCof the two populations highlighted in A. Crosses indicate medians of populations in both FSC and SSC.

2374 Plant Physiol. Vol. 171, 2016

Carqueijeiro et al.

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

a distinct chemical composition of idioblasts, at thesame time it provided a way to identify them under thebright-field microscope. Upon UV excitation, idioblastprotoplasts showed intense blue and green fluores-cence, indicating they also retained the original fluo-rescence properties (Fig. 3E). In fact, the emissionspectra of these cells were highly similar to the emissionspectra of the leaf idioblasts, and markedly differentfrom other mesophyll leaf cells or from other proto-plasts (Fig. 4), clearly confirming the identity of thefluorescent, refractive protoplasts as idioblasts. Overall,it was clear that idioblast protoplasts retained theiroriginal distinctive chemical composition and identity,and therefore constituted highly interesting targets forcell sorting and differential characterization.

To sort C. roseus protoplasts relying on their intrin-sic fluorescence properties, the photomultiplier tubedetector voltages of the sorter were set to maximizedifferences between the minimum and maximum flu-orescence for the blue, green, and red channels. PBSworked well as sheath buffer for sorting of the C. roseusprotoplasts, but it lowered the stability of the sorted cellpopulations in comparison with the use of protoplastsuspension buffer as sheath. Therefore, PBS should beused only when further manipulation of cells is notneeded, such as for metabolomic or transcriptomicstudies.

Analysis of the sorted events using all the differentcombinations of fluorescence signals in two- and three-dimensional dot plots revealed that the combination of

Figure 6. Epifluorescencemicroscopy images of a sourceC. roseus leaf protoplast suspension (A) and of populations ofmesophyllcells (B and C) and idioblasts (D and E) isolated by FACS. 1, Red channel revealing autofluorescence of chlorophyll in chloro-plasts. 2, Blue channel revealing the presence of an autofluorescence signal in idioblasts. 3, Green channel revealing the presenceof an autofluorescence signal in idioblasts. 4, Merged images of the three channels. 5, Merge of the blue and green channelimages with the corresponding bright-field image. Bars = 40 mm (A, B, and D) and 5 mm (C and E).

Plant Physiol. Vol. 171, 2016 2375

Isolation of Mesophyll Specialized Cells by FACS

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

the red and green signals depicted in Figure 5Awas theone giving a better definition of the different cell pop-ulations. Sorted cells from all over the spectrum ofevents detected were collected in mannitol-MES (MM)buffer and observed under the fluorescence micro-scope. This enabled to identify populations I and II astargets (Fig. 5A): population I was constituted exclu-sively of cells with a high number of chloroplasts, withlow blue/green autofluorescence and low refraction ofthe vacuolar content, clearly enabling their identifica-tion as a pure population of common mesophyll cells(Fig. 6, B andC); population II was constituted of 90% ofcells depicting an intense blue/green autofluorescenceand a high refraction of the vacuolar content, enablingtheir identification as idioblasts (Fig. 6, D and E). Whencomparing the forward (FSC) and side (SSC) scatterproperties of the two populations (Fig. 5B), it was ob-served that they overlap partially, with population II(idioblasts) having an average higher FSC and lowerSSC, but with a high heterogeneity of values. Since FSCincreases with particle refraction index and size, andSSC increases with granularity or higher amount ofcellular components, these results match microscopeobservations of idioblasts, which show a higher refrac-tion index and often a big cell size, although with a highheterogeneity in size (Figs. 3, D and E, and 6A). Onaverage it was possible to sort and recover;50,000 cellsof population II (idioblasts) per operation day, and10-fold more of common mesophyll cells.

To test the feasibility of using the sorted cells fortranscriptomic studies, the sorted cells were collecteddirectly to RNA extraction buffer and their RNA wasisolated, with about 50,000 cells yielding 200 ng of RNA

for the two sorted populations, as well as for the wholeprotoplast mother suspension, all with high quality(RNA quality indicator $ 8). qPCR analysis of thegenes involved in the late steps of TIA biosynthe-sis, desacetoxyvindoline 4-hydroxylase (D4H) anddeacetylvindoline 4-O-acetyltransferase (DAT), fur-ther confirmed the suitability of the sorted cells fortranscriptomic studies, and showed a ;3-fold increasein the transcript levels of those genes in idioblasts, inline with previous in situ hybridization studies of thesetwo genes (Fig. 7; St-Pierre et al., 1999).

To demonstrate that the sorted cells may also be usedfor metabolomic studies, namely in what concerns al-kaloid profile, methanolic extracts obtained fromroughly 200,000 sorted common mesophyll cells and;25,000 sorted idioblasts were analyzed by liquidchromatography-tandem mass spectrometry. Figure 8shows the easy detection of the leaf-abundant VLBmonomeric precursor vindoline (Carqueijeiro et al.,2013) in sorted common mesophyll cells and idioblasts.The results confirm that idioblasts are impressive TIAaccumulation targets, showing levels of vindoline;3,000-fold higher than common mesophyll cells(315.22 ng per idioblast versus 0.105 ng per mesophyllcell).

In conclusion, the mesophyll of C. roseus leaves pre-sents scattered idioblasts that specifically accumulatehigh levels of alkaloids and depict a distinctive con-spicuous autofluorescence, which allows the isolationof idioblast protoplasts retaining their original identityby means of an optimized FACS method. The proce-dure enables the isolation of a high number of cells,50,000 idioblast cells per operation day (from ;3.5 g ofleaves), easily reaching 250,000 upon 5 d of work,

Figure 7. Expression of D4H and DAT in sorted common mesophylland idioblast protoplasts. Transcript levels were determined by qPCRand normalized using CrRPSL24 (P , 0.05 for both genes, Student’s ttest).

Figure 8. Base peak chromatogram ofm/z 457.23 and respective massspectrometry fragment pattern. A, Vindoline standard (0.5 mg mL21). B,Sorted idioblast protoplasts (115 cells mL21). C, Sorted common me-sophyll protoplasts (800 cells mL21).

2376 Plant Physiol. Vol. 171, 2016

Carqueijeiro et al.

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

highly exceeding the total 2,500 to 5,000 idioblast cellsisolated by Murata and De Luca (2005) using laser mi-crodissection. The idioblast cell fraction obtained in thisstudy is highly enriched, and is now being used for theimplementation of differential transcriptomic and meta-bolomic studies to uncover the differentiation status andfunctions of idioblasts, as well as novel candidate genespotentially involved in key events of the biosynthesis,transport, and regulation of the medicinal TIA pathway.

MATERIALS AND METHODS

Plant Material

Catharanthus roseus (L.) G. Don ‘Little Bright Eye’ plants were grown at 25°C,in a growth chamber, under a 16/8 h photoperiod, using white fluorescent lightwith a maximum intensity of 70 mmol m22 s21. Seeds were acquired from B&TWorld Seeds. Plants used for protoplast isolation were 6 to 8 months old.

Alkaloid Indicators

Dragendorff and Mayer reagents were prepared as described by Bruneton(2009). Both reagents were diluted 1:10 with water just before use. For detectionof alkaloids in leaves, the reagents were infiltrated through the abaxial face of awhole leaf using a syringe without needle. For alkaloid detection in proto-plasts, a small drop of diluted reagent was added to an excess of protoplastsuspension.

Isolation and Characterization of C. roseus Leaf Protoplasts

The isolation of leaf protoplasts was performed as described by Carqueijeiroet al. (2013). Protoplasts were diluted in MM buffer (0.4 M mannitol in 20 mMMES, pH 5.6–5.8) to 53 105 to 23 106 cells mL21, and observed using an opticalmicroscope (Olympus) with a coupled Olympus DP 25 digital camera and CellB software. Protoplasts were also observed using a Leica DMI6000 microscopebody equipped with a Hamamatsu Flash Orca 4.0 LT sCMOS camera with 103Plan Fluor 0.3 NA and 203 Plan Apochromat 0.7 NA objectives controlledthrough the Leica LAS X software (Leica Microsystems). For acquisition of blueand green autofluorescence, the excitation band-pass filter was 350/50 nm,associated to an emission filter of 455/50 nm (blue) and of 525/36 nm (green).For acquisition of red autofluorescence, the excitation was 490/20 nm andemission was 705/72 nm.

Viability of protoplastswas assayed asdescribedbyCarqueijeiro et al. (2013).Autofluorescence characterization of protoplasts was performed in a Leica TCSSP5 II confocal microscope (Leica Microsystems) controlled through the LeicaLAS 2.6 software, using the diode 405-nm laser to excite and recording aspectral series from 420 to 600 nm with a detection bandwidth of 5 nm.

FACS of C. roseus Protoplasts

Leaf protoplasts suspended in MM buffer were run in a MoFlo (BeckmanCoulter), at 4°C, immediately after isolation. The protoplast suspension wascarefully homogenized immediately prior to use and whenever needed duringthe sorting. Protoplasts were sorted using either protoplast isolationmedium orPBS as sheath, at a constant pressure of 200 kPa (;30 psi), using a 100-mmnozzle and a drop-drive frequency of ;40 kHz (;40,000 drops/s). The totalevent rate was kept below 8,000 events/s. Autofluorescencewas detected usinga water-cooled I90-C Coherent argon laser operating at multiline UV wave-lengths (330–360 nm) with 50-mW output as excitation source, and two dif-ferent emission band-pass filters, 455/30 nm and 530/40 nm for blue and greenfluorescence detection, respectively. Scatter measurements were carried out at488 nm using a Coherent Sapphire 488-200 CDRH laser, with 140-mW output,which was also used for detection of chlorophyll autofluorescence togetherwith a 670/40-nm band-pass emission filter. Sorted cells were collected inmicrocentrifuge tubes containing a buffer adapted to subsequent processing.For imaging, 200 cells were sorted onto a slide containing 10 mL of protoplastisolation medium and immediately observed in the Leica High ContentScreening microscope as described above.

RNA Extraction and cDNA Synthesis

RNAwas isolated using the RNeasy PlantMini kit (Qiagen) according to themanufacturer’s instructions with the modifications described by Birnbaumet al. (2005). RNA quantity and quality assessment was performed in EXPE-RION automated electrophoresis system (Bio-Rad). For the synthesis of cDNA,150 ng of the isolated RNA was treated with DNaseI (Thermo Scientific), fol-lowed by enzyme inactivation by heating at 65°C for 10 min in the presence of20 mM EDTA. First-strand cDNA was synthesized from total RNA using theiScript cDNA synthesis kit (Bio-Rad) and oligo(dT) according to the manufac-turer’s instructions.

qPCR

The sequences of the primers used are given in Supplemental Table S1. ThecDNA samples were analyzed with the iCycler iQ system (Bio-Rad). The qPCRreaction mixes contained 10 mL of Bio-Rad 13 iQ SYBR Green Supermix, 0.5 mM

each primer, and 1 mL of cDNA for a 20-mL end volume reaction. The qPCRprogram started with a 3-min initial denaturation step at 95°C, followed by40 cycles of amplification (95°C for 10 s, 54°C for 15 s, and 72°C for 30 s) withcontinuous monitoring of the SYBR Green fluorescence. The reaction endedwith a melting curve step from 55°C to 95°C at 0.5°C per second. Data analysiswas carried out with the Bio-Rad Optical System Software 5.0.

HPLC-Electrospray Ionization-Tandem MassSpectrometry Analysis

Methanolic extracts of lyophilized sorted cells were separated on an AccelaHPLC (Thermo Fischer Scientific) using a Phenomenex Gemini C18 column(1503 4.6 mm; 3-mmparticle size). The mobile phase consisted of methanol (A)and 25 mM ammonium acetate pH 10.0 (B) mixed to form the following gra-dient: 40% B at 0min, 38% B at 6min, 38% B at 35min, 25% B at 40min, 10% B at45 min, 40% B at 50 min, and 40% B at 55 min. Elution was performed with aflow rate of 0.5 mL min21, and the injection volume was 20 mL. Mass spec-trometry analysis was performed on an LTQ Orbitrap XL hybrid mass spec-trometer (Thermo Fischer Scientific) controlled by LTQ Tune Plus 2.5.5 andXcalibur 2.1.0. The electrospray ionization source settings were as follows:source voltage at 3.1 kV, capillary temperature at 275°C with a sheath gas flowrate of 40 and an auxiliary gas flow rate of 10 (arbitrary unit as provided by thesoftware settings), capillary voltage at 48 V, and tube lens voltage at 115 V. Themass spectrometry data-handling software Xcalibur QualBrowser (ThermoFischer Scientific) was used to search for predicted metabolites by their exactm/z value. All peaks were checked for m/z value and fragmentation products.The vindoline standard used was from Sigma-Aldrich.

Accession Numbers

Accession numbers in the database Cathacyc (http://www.cathacyc.org)are as follows: RPSL24, Caros004968.2; D4H, Caros006133.1; and DAT,Caros016117.1.

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. General scheme of the C. roseus monoterpenoidindole alkaloid pathway leading to the anticancer alkaloids vinblastineand vincristine.

Supplemental Table S1. List of primers used in this work.

ACKNOWLEDGMENTS

We would like to thank Pedro Lima (IGC) for help with fluorescence andconfocal microscopy, Paula Sampaio (IBMC) for help with the preparation offigures, Filipe Borges (IGC) for assistance during RNA extraction, ManuelJoaquim Bastos Marques (Department of Physics and Astronomy, FCUP) forhelp with the interpretation of bright-field images, and Silvia Maia from Centrode Materiais da Universidade do Porto for technical assistance with massspectrometry.

Received May 31, 2016; accepted June 20, 2016; published June 29, 2016.

Plant Physiol. Vol. 171, 2016 2377

Isolation of Mesophyll Specialized Cells by FACS

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

LITERATURE CITED

Birnbaum K, Jung JW, Wang JY, Lambert GM, Hirst JA, Galbraith DW,Benfey PN (2005) Cell type-specific expression profiling in plants viacell sorting of protoplasts from fluorescent reporter lines. Nat Methods2: 615–619

Bruneton J (2009) Pharmacognosie: Phytochimie, Plantes Médicinales, Ed4e, Revue et Augmentée. Lavoisier, Paris

Carqueijeiro I, Noronha H, Duarte P, Gerós H, Sottomayor M (2013)Vacuolar transport of the medicinal alkaloids from Catharanthus roseus ismediated by a proton-driven antiport. Plant Physiol 162: 1486–1496

Carter AD, Bonyadi R, Gifford ML (2013) The use of fluorescence-activated cell sorting in studying plant development and environmen-tal responses. Int J Dev Biol 57: 545–552

Courdavault V, Papon N, Clastre M, Giglioli-Guivarc’h N, St-Pierre B,Burlat V (2014) A look inside an alkaloid multisite plant: the Cathar-anthus logistics. Curr Opin Plant Biol 19: 43–50

De Luca V, Salim V, Thamm A, Masada SA, Yu F (2014) Making iri-doids/secoiridoids and monoterpenoid indole alkaloids: progress onpathway elucidation. Curr Opin Plant Biol 19: 35–42

Fang J, Schneider B (2014) Laser microdissection: a sample preparationtechnique for plant micrometabolic profiling. Phytochem Anal 25: 307–313

Galbraith D (2010) Flow cytometry and fluorescence-activated cell sortingin plants: the past, present, and future. Biomedica 30: 61–69

Kellner F, Geu-Flores F, Sherden NH, Brown S, Foureau E, CourdavaultV, O’Connor SE (2015) Discovery of a P450-catalyzed step in vindoline

biosynthesis: a link between the aspidosperma and eburnamine alka-loids. Chem Commun (Camb) 51: 7626–7628

Mersey BG, Cutler AJ (1986) Differential distribution of specific indolealkaloids in leaves of Catharanthus roseus. Can J Bot 64: 1039–1045

Miettinen K, Dong L, Navrot N, Schneider T, Burlat V, Pollier J, WoittiezL, van der Krol S, Lugan R, Ilc T, et al (2014) The seco-iridoid pathwayfrom Catharanthus roseus. Nat Commun 5: 3606

Murata J, De Luca V (2005) Localization of tabersonine 16-hydroxylase and16-OH tabersonine-16-O-methyltransferase to leaf epidermal cells de-fines them as a major site of precursor biosynthesis in the vindolinepathway in Catharanthus roseus. Plant J 44: 581–594

Renaudin JP (1985) Extraction and fluorimetric detection after high-performance liquid-chromatography of indole alkaloids from cultured-cells of Catharanthus roseus. Physiol Veg 23: 381–388

St-Pierre B, Vazquez-Flota FA, De Luca V (1999) Multicellular compart-mentation of Catharanthus roseus alkaloid biosynthesis predicts inter-cellular translocation of a pathway intermediate. Plant Cell 11: 887–900

Stavrinides A, Tatsis EC, Foureau E, Caputi L, Kellner F, Courdavault V,O’Connor SE (2015) Unlocking the diversity of alkaloids in Catharanthusroseus: nuclear localization suggests metabolic channeling in secondarymetabolism. Chem Biol 22: 336–341

Yoder LR, Mahlberg PG (1976) Reactions of alkaloid and histochemicalindicators in laticifers and specialized parenchyma cells of Catharanthusroseus (Apocynaceae). Am J Bot 63: 1167–1173

Yu F, De Luca V (2013) ATP-binding cassette transporter controls leafsurface secretion of anticancer drug components in Catharanthus roseus.Proc Natl Acad Sci USA 110: 15830–15835

2378 Plant Physiol. Vol. 171, 2016

Carqueijeiro et al.

www.plantphysiol.orgon February 10, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.