Received: 21 July 2017 | Revised: 11 August 2017 | Accepted: 24 August 2017 DOI: 10.1002/pld3.12 ORIGINAL RESEARCH Two Arabidopsis late pollen transcripts are detected in cytoplasmic granules Marıa R. Scarpin1 | Lorena Sigaut2 | Silvio G. Temprana3 | Graciela L. Boccaccio3 | Lıa I. Pietrasanta2,4 | Jorge P. Muschietti1,5 1 tica y Biologıa Instituto de Ingenierıa Gene ctor N. Torres” (INGEBIMolecular “Dr. He CONICET), Buenos Aires, Argentina Abstract Many of mRNAs synthesized during pollen development are translated after germi- 2 Instituto de Fısica de Buenos Aires (IFIBACONICET), Departamento de Fısica, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina 3 n Instituto Leloir, IIBBAFundacio CONICET, Facultad de Ciencias Exactas y Naturales, Departamento de Fisiologıa y Biologıa Molecular y Celular, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina 4 Centro de Microscopıas Avanzadas, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina 5 Departamento de Biodiversidad y Biologıa Experimental, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina nation, and we hypothesize that they are stored in cytoplasmic granules. We analyzed the cellular localization of the SKS14 and AT59 Arabidopsis mRNAs, which are orthologues of the tobacco NTP303 and tomato LAT59 pollen mRNAs, respectively, by artificially labeling the transcripts with a MS2-GFP chimera. A MATLAB-automated image analysis helped to identify the presence of cytoplasmic SKS14 and AT59 mRNA granules in mature pollen grains. These mRNA granules partially colocalized with VCS and DCP1, two processing body (PB) proteins. Finally, we found a temporal correlation between SKS14 protein accumulation and the disappearance of SKS14 mRNA granules during pollen germination. These results contribute to unveil a mechanism for translational regulation in Arabidopsis thaliana pollen. KEYWORDS MATLAB, MS2, pollen, processing body, translational regulation Correspondence Jorge Muschietti, Instituto de Ingenierıa tica y Biologıa Molecular “Dr. He ctor Gene N. Torres” (INGEBI-CONICET), Buenos Aires, Argentina. Email: prometeo@dna.uba.ar Funding information JPM, GLB, LS, and LIP are investigators of the National Research Council (CONICET) from Argentina. MRS and SGT are recipients of postdoctoral fellowships from the CONICET. This work was supported by grants from ANPCyT (PICT 2014-0423 and PICT 2015-0078) and from the Universidad de Buenos Aires to JPM. ---------------------------------------------------------------------------------------------------------------------------------------------------------------------This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2017 The Authors. Plant Direct published by American Society of Plant Biologists, Society for Experimental Biology and John Wiley & Sons Ltd. Plant Direct. 2017;1–10. wileyonlinelibrary.com/journal/pld3 | 1 2 | SCARPIN ET AL. Current addresses: Marıa R. Scarpin: Plant Gene Expression Center and Department of Plant and Microbial Biology, Albany, CA, USA. Silvio G. Temprana: Molecular and Cell Biology Department, University of California, Berkeley, CA, USA 1 | INTRODUCTION or undergo a rearrangement where translation initiation factors are recruited, thus allowing mRNAs to reenter polysomes (Decker & Par- Arabidopsis pollen development involves two stages: an early stage ker, 2012). that includes microspores and bicellular pollen followed by a late As in yeast and animals, it has been shown that Arabidopsis stage of tricellular and mature pollen. These two stages differ in their thaliana PBs include decapping factors and coactivators, such as transcriptional profiles: early genes expressed during the microspore decapping protein one (DCP1), decapping protein two (DCP2), stage decrease their abundance before pollen maturation. In turn, decapping protein five (DCP5), and varicose (VCS) (Xu & Chua, late genes are expressed after asymmetric mitosis and accumulate 2009; Xu, Yang, Niu, & Chua, 2006). The knockout of these genes during pollen maturation, bringing about a stable pool of mRNAs affects the growth of vascular and epidermal cells, stomata, and that govern pollen germination and pollen tube early growth. root hairs, suggesting that decapping and/or PBs have a fundamen- Thereby, at anthesis, all proteins or mRNAs required for germination tal role during plant development (Maldonado-Bonilla, 2014; Xu , and pollen tube growth would be present (Boavida, Becker, & Feijo et al., 2006). 2005). Consistent with this, in many species, pollen germination is Here, we focus in the Arabidopsis mature pollen SKS14 and independent of transcription but dependent on translation (Twell, AT59 mRNAs (Loraine, McCormick, Estrada, Patel, & Qin, 2013) 1994). which are putative orthologues of the tobacco NTP303 and tomato Pollen tube growth is a process that occurs in an explosive LAT59 mRNAs, respectively. We found SKS14 and AT59 mRNAs in way, and tip extension requires a rapid increase in pectin amount. cytoplasmic granules that colocalize with the PB markers VCS and Some of the late pollen genes encode proteins homologous to DCP1. Finally, we show that SKS14 protein accumulates during pol- enzymes linked to pectin metabolism, including polygalacturonases len germination while the number of SKS14 mRNA granules (Brown & Crouch, 1990; Niogret, Dubald, Mandaron, & Mache, decreases. These observations are compatible with the notion that 1991; Rogers & Lonsdale, 1992), pectin methylesterases (Mu, the SKS14 mRNA is released from PBs to allow translation in a con- Stains, & Kao, 1994; Wakeley, Rogers, Rozycka, Greenland, & trolled manner. Hussey, 1998), and pectate lyases (Rogers, Harvey, & Lonsdale, 1992; Wing et al., 1989). The tomato late gene LAT59 codifies a 2 | MATERIAL AND METHODS protein related to the pectate lyase family, potentially involved in cell wall degradation by pectin cleavage. The translation of LAT59 mRNA is highly regulated and occurs since final stages of pollen development (Curie & McCormick, 1997). In turn, the tomato LAT52 is a pollen gene that codifies a cysteine-rich extracellular protein involved in pollen hydration and pollen germination (Muschietti, Dircks, Vancanneyt, & McCormick, 1994). LAT52 transcript levels gradually increase during pollen development reaching its maximum at pollen maturity (Twell, Klein, Fromm, & McCormick, 1989). Another example is the tobacco NTP303 gene, which is transcribed through pollen development from the early binucleate stages (Weterings et al., 1992) and translated once germination occurs (Wittink et al., 2000). NTP303 has homology with ascorbate oxidases, and according to its time of expression, it would be linked to pollination or fertilization (Schrauwen et al., 1999). 2.1 | Plant material and growth conditions Sterilized seeds from Arabidopsis thaliana (ecotype Columbia-0) wildtype, single mutants, and transgenic plants were plated on 0.5X Murashige and Skoog (1962) medium with 1% sucrose and selective agent (50 mg/l kanamycin) if necessary and cold stratified 4 days in dark at 4°C. Seeds were germinated and grown under continuous light at 22°C for 7 days. Seedlings were then transferred to soil or peat, mixed with vermiculite and perlite (2:1:1), and grown in a chamber at 22°C under long-day (16/8 hr light/dark) photoperiod and 60% relative humidity. 2.2 | Plasmid constructs MS2 system is based on the strong association between bacterio- Processing body (PB) are highly conserved cytoplasmic organelles phage MS2 capsid protein (MCP) and six repeat loops of a 19- involved in translation inhibition, mRNA degradation, and storage. nucleotide fragment containing bacteriophage’s replicase start codon PB formation includes translationally repressed messenger ribonucle- (Bertrand et al., 1998). The pMS2-GFP and pSL-MS2-12X plasmids oproteins (mRNPs) that aggregate into larger structures through pro- were donated by Robert Singer (Addgene plasmid #27121 and tein–protein interactions. mRNPs localized in PBs can be degraded #27119, respectively) (Bertrand et al., 1998; Fusco et al., 2003). SCARPIN | ET AL. 3 MS2 control vector was generated through an LR recombination deviations from the average fluorescence of the corresponding box system (Invitrogen) using the binary vector pK7WG2D. After PCR were marked with a yellow point. The grid was moved 20 pixels to amplification from the pMS2-GFP plasmid, the GFP-MCP fragment the right and down, and a new round of analysis was carried out. was inserted into the pZD05 vector under the control of LAT52 pro- Those groups of pixels that followed similar parameters were marked moter, obtaining the pLAT52::GFP-MCP. Then, the pLAT52::GFP-MCP with a light blue circle (Fig. S2D). Pixels detected with both analysis fragment was inserted in the pENTR1a entry vector and then recom- (yellow point and light blue circle) were processed in a new round of bined in the binary vector pK7WG2D. All plasmids were confirmed the script to confirm the cytoplasmic granules (Fig. S2E). The vali- by sequencing. dated cytoplasmic granules were marked with a red point and a SKS14 and AT59 50 UTRs and coding regions were cloned by PCR green circle, indicating the center and the approximate area of the from Arabidopsis mature pollen cDNA in the pSL-MS2-12X plasmid. granule, respectively (Fig. S2F). Determination of the number of The SKS14-MS2-12X and AT59-MS2-12X fragments were PCR- cytoplasmic granules was carried out by taking one image of the cloned into pZD05. pLAT52::SKS14-MS2-12X and pLAT52::AT59- central plane of each pollen grain; by this approach, we obtained an MS2-12X fragments were PCR-cloned and inserted into the binary approximated value of the number of cytoplasmic granules for each vector pK7WG2D-pLAT52::GFP-MCP obtaining pLAT52::GFP-MCP/ pollen grain analyzed. All the analysis was carried out in the pres- pLAT52::SKS14-MS2-12X and pLAT52::GFP-MCP/pLAT52::AT59-MS2- ence of puromycin 50 lg/ml. 12X. All plasmids were confirmed by sequencing. pLAT52::RFP-VCS and pLAT52::RFP-DCP1 vectors were generated using the binary vector pK7WG2D. The VCS or DCP1 fragments were 2.4.2 | PBs detection amplified from the pda09249 DNA stock (ABRC) or Arabidopsis A similar MATLAB script used for the mRNA granules was applied mature pollen cDNA, respectively. They were inserted in the pENTR1a for PBs detection except that positive PB marker protein foci were entry vector and then recombined in pK7WG2D. The pLAT52-RFP defined between 5 and 80 pixels. fragment was amplified from the pZD05 vector, digested, and cloned in pK7WG2D containing VCS or DCP1. All plasmids were confirmed by sequencing. 2.4.3 | Colocalization analysis pSKS14::SKS14-RFP vector was generated using binary vector To address statistical significance of the proximity between the pGWB408. The SKS14 coding region and the RFP fragment were mRNAs granules and PB marker protein foci (i.e., whether or not amplified from Arabidopsis cDNA and from the pZD05 vector, they may colocalize by chance given the specific layout of mRNAs respectively, and cloned in the pENTR1a vector. The 35S promoter and PBs within an image), a shuffling analysis was performed using a of the pGWB408 vector was replaced by a 1091-bp fragment corre- custom-made MATLAB routine. After confirming the cytoplasmic sponding to the SKS14 promoter that was cloned by PCR from Ara- granules and the PB marker protein foci, both were modeled as cir- bidopsis cDNA. All plasmids were confirmed by sequencing. cles in the 2D plane using the center of mass and area obtained EHA105 Agrobacterium strain was used for transformation. Plant from previously described analysis. The random distribution of the transformation was carried on by floral dip method (Clough & Bent, distance between the mRNA granule and the nearest PB foci was 1998). obtained by randomly repositioning the mRNA granule 10,000 times while keeping the PB’s foci layout fixed. From that random distribu- 2.3 | Microscopy analysis Mature pollen grains of twenty-day-old plants were collected and stored in Tris–EDTA buffer in the presence or absence of puromycin (100 lg/ml) for 1 hr at 22°C. Images were taken with a fluorescence microscope Olympus BX41 or in a confocal microscope Olympus tion, we proceeded to obtain the p-value for the experimentally obtained distance. The sum of the radii of the mRNA granule and the nearest PB marker protein foci was used to determine whether they were colocalizing, proximal, or distant. All the analyses were carried out in the presence of puromycin 50 lg/ml. IX81 FV1000 (Laser 488 nm, filter BP 505-525, or Laser 543 nm, filter BP 560-620) and analyzed by MATLAB scripts. 2.5 | Statistical analysis 2.4 | MATLAB scripts Results are expressed as scattered plots. Points represent data dis- 2.4.1 | mRNA granules detection pollen grain (each n corresponds to the media of the number of In order to analyze the cytoplasmic granules by MATLAB (Fig. S2), determination of fluorescence intensity (each n corresponds to initially a mask was applied to eliminate vegetative nucleus and facil- one vegetative nucleus fluorescence intensity measurement); prob- itate the cytoplasm visualization. Then, the image was split into 49 ability values of <.1 were considered statistically significant. Statis- boxes and the average fluorescence of each box was determined tical analysis of the data and further processing were performed (Fig. S2C). Those groups of pixels that were between 5 and 30 pixels using GraphPad Prism version 5.00 for Windows (GraphPad Soft- in size and had an average fluorescence higher than three standard ware). persion for n = 7-10 quantifications of cytoplasmic granules per cytoplasmic granules of 10-20 pollen grains) and n = 10-25 for 4 | SCARPIN Paralleling the fate of maternal mRNA in animal oocytes, we SKS14-MS2 LB KanR MCP-bs SKS14 LAT52 LAT52 GFP MCP NLS RB hypothesize that SKS14 and AT59 mRNAs are stored in cytoplasmic granules during pollen development. To analyze the subcellular local- AT59-MS2 ization of these transcripts in vivo, we artificially labeled the mRNA LB KanR MCP-bs AT59 LAT52 LAT52 GFP MCP NLS RB MS2 LB KanR LAT52 ET AL. using the MS2-MCP system, which is based on the strong binding of the bacteriophage MS2 coat protein (MCP) to specific RNA loops GFP MCP NLS RB DIC MS2 system (a) CONTROL F I G U R E 1 mRNA detection by the MS2 system. The two depicted constructs were inserted in the same vector: GFP fused to the MS2 coat protein (MCP) with a nuclear localization signal (NLS), and the SKS14 or AT59 transcripts fused to MCP-binding site (MCPbs). A control construct termed MS2 encodes the GFP-MCP chimera and no target mRNA. The three constructs SKS14-MS2, AT59-MS2, and MS2 were under the control of the pollen-specific promoter LAT52. LB and RB, left and right borders of the translocation cassette, respectively (b) 3.1 | SKS14 and AT59 mRNAs form cytoplasmic granules in mature pollen SKS14-MS2 3 | RESULTS (c) LAT59 protein is present at final stages of pollen development and its amount increases after pollen germination (Curie & McCormick, 1997) while NTP303 is synthesized only once pollen germinates (Wittink et al., 2000). We sought to investigate the post-transcriptional regulation of the Arabidopsis LAT59 and NTP303 orthologues. (d) Arabidopsis AT59 (At1 g14420) is the LAT59 ortholog (Kulikauskas & McCormick, 1997). Among the putative NTP303 orthologues, we focused on the SKU5 similar (SKS) family that has 19 members in SKS13, and SKS14 are expressed in pollen (Honys & Twell, 2003). We then considered the length and free energies of their 50 UTRs regions and compared these parameters to those of NTP303 50 UTR. AT59-MS2 Arabidopsis (Sedbrook et al., 2002). Among them, SKS11, SKS12, (e) Furthermore, SKS14 is the pollen SKS gene that shows the highest increment in expression levels as pollen development proceeds, showing a maximum in mature pollen (Honys & Twell, 2003). This makes SKS14 a good candidate to test our hypothesis and therefore was chosen for this study. (f) *** 1.5 #of granules per PG F I G U R E 2 Arabidopsis SKS14 and AT59 mRNAs form cytoplasmic granules in mature pollen grains. (a) Confocal images of a representative pollen grain from the control MS2 line. A mask was applied to eliminate vegetative nucleus and facilitate cytoplasm visualization. (b-e) Representative images of two independent SKS14-MS2 (b and c) and two AT59-MS2 lines (d and e). In the left panels, white arrowheads show examples of cytoplasmic granules identified by the MATLAB script while empty arrowheads show cytoplasmic aggregates that were not detected by the MATLAB script. Right panels, DIC images. Size bar, 5 lm. (f) Quantification of cytoplasmic granules. Each point corresponds to the mean value of an independent sample that included 20 pollen grains. The media and standard error for each transgenic line are shown. Statistical significance (Mann–Whitney test) relative to the control line MS2 3-1 is indicated (***p < .001 and **p < .01) 1.0 ** 0.5 ** ** 0.0 M S 2 3- 1 M S 2 5S 2 K S 14 3- S K 2 S 14 6A 3 T 59 1- 1 A T 59 2- 4 SCARPIN | ET AL. T A B L E 1 Quantification of cytoplasmic granules per pollen grain in WT pollen. Mean values from 20 pollen grains in independent samples are shown, and the media and standard error for each transgenic line are indicated. N: number of pollen grains analyzed. ND: Not determined (Mann–Whitney test) 5 The number of mRNA granules per pollen grain is somehow lower than expected considering that the strong LAT52 promoter was used. Given that a whole scan of the entire pollen grain along Z-axis rapidly quenched the fluorescent GFP signal, a maximum of five confocal sections of each pollen grain were analyzed, with a final size of 2.5 um Line Media Standard error N p-value p-value summary MS2 3-1 0.014 0.010 10 ND ND len grain, and thus, we speculate that the total number of RNA gran- SKS14 3-2 0.413 0.156 10 .0003 *** ules would be 10 times larger than the experimental value. In addition, SKS14 6-3 0.133 0.042 10 .0053 ** the presence submicroscopic RNA granules of less than the resolution AT59 1-1 0.087 0.021 10 .0029 ** of the confocal microscope (~150 nm) should also be considered. A AT59 2-4 0.093 0.031 9 .0092 ** second factor that may be introducing a systematic error is that gran- thick. This represents about 10% of the volume of an Arabidopsis pol- ules close to the perinuclear region were not included, due to the strong GFP nuclear signal and because most of the potential positives termed MCP-binding site (Lampasona & Czaplinski, 2016). The GFP- close to the perinuclear region were extensions of the vegetative cell MCP protein was fused to a nuclear localization signal (NLS) so that nuclei indentations. Thus, a significant proportion of mRNA granules only the GFP-MCP bound to the target mRNAs is found in the cyto- was lost in our analysis, but at the same time, we largely reduced the plasm. Fig. S1 shows that the free GFP-MCP protein is largely accu- chances of including false positives. mulated in the vegetative nuclei identified by DAPI staining. We To test that the differences observed in the number of granules obtained independent Arabidopsis transgenic lines containing the were not due to dissimilar expression levels of the GFP-MCP chi- GFP-MCP construct together with either the SKS14 or the AT59 mera in the different Arabidopsis lines, we measured the fluores- mRNA fused to the MCP-binding site, termed SKS14-MS2 and AT59- cence intensity in the vegetative nuclei, where nonbound GFP-MCP MS2, respectively (Fig. 1). All constructs are under the control of the accumulates due to the NLS included in the construct. Although tomato pollen-specific promoter LAT52 (Twell, Yamaguchi, Wing, GFP-MCP intensity varies between lines, we found that there is no Ushiba, & McCormick, 1991). For comparison, we generated control correlation between the nuclear fluorescence intensity and the num- transgenic plants termed MS2 lines carrying only the GFP-MCP con- ber of cytoplasmic granules, strongly supporting that these granules struct (Fig. 1). With these tools, we found that both SKS14 and are specific and are not due to the overexpression of heterologous AT59 messengers form granules in the cytoplasm of mature pollen GFP-MCP protein (Fig. S3). grains (Fig. 2; Table 1). Representative images of pollen grains showing cytoplasmic granules of SKS14-MS2 mRNA and AT59-MS2 mRNA are depicted in Fig. 2b-e. A representative control MS2 pollen grain is shown in 3.2 | SKS14 and AT59 mRNA cytoplasmic granules colocalize with PB proteins After demonstrating that the SKS14 and AT59 mRNAs accumulate in Fig. 2a. To perform an automated unbiased analysis of these images, we cytoplasmic granules in mature pollen, we sought to determine implemented a MATLAB script that detects granules in the cyto- whether these granules contain PBs marker proteins. We first inves- plasm, applying a mask to the nucleus and measuring pixel size and tigated the localization of VCS and DCP1 in mature pollen by fluorescence intensity (see Material and Methods and Fig. S2). expressing RFP-VCS and RFP-DCP1 under the LAT52 promoter. Fig. 2f shows that the number of cytoplasmic granules of fluorescent Fig. 3 shows that RFP-VCS and RFP-DCP1 appeared as cytoplasmic protein in the SKS14-MS2 and AT59-MS2 lines was 10-40 times foci. As expected, the size of these bodies increased upon incubation higher than in the control MS2 line (Table 1). with puromycin, a drug that inhibits translation elongation-releasing Puromycin RFP-VCS F I G U R E 3 The PB proteins VCS and DCP1 form cytoplasmic foci in mature pollen. Representative images of mature pollen grains in the absence (-) or presence (+) of puromycin 50 lg/ml. DIC images are shown in the right panels. Size bar, 5 lm RFP-DCP1 – + 6 | SCARPIN SKS14 VCS Overlay DIC Zoom 10X ET AL. Colocalization (a) (b) (c) F I G U R E 4 Colocalization of SKS14 mRNA with RFP-VCS. (a) Confocal image of a representative mature pollen grain showing high colocalization between SKS14 mRNA and a RFP-VCS body. (b) Confocal image of a representative mature pollen grain showing a SKS14 mRNA cytoplasmic granule contiguous to a RFP-VCS body. (c) Confocal image of a representative cytoplasmic granule with no relationship with any RFP-VCS body. In the left panels, white arrowheads show examples of cytoplasmic granules identified by MATLAB while empty arrowheads show cytoplasmic aggregates not detected by MATLAB. The insets in the merged (“Overlay”) column are enlarged on the 10X panels. In the “Colocalization” column, the blue point and black circle indicate the localization and size of the VCS body, respectively, and the red point and circle correspond to the SKS14 mRNA granule. DIC images are shown. Size bar, 5 lm mRNAs from ribosomes, thus allowing mRNA recruitment to PBs To address the extent of random colocalization in these images, (Thomas, Loschi, Desbats, & Boccaccio, 2011). The effect was stron- we performed a shuffling analysis (for description of the script, see ger for the RFP-VCS construct where PBs increased both in size and “Materials and Methods”). Fig. S4 shows examples of shuffling analy- number upon exposure to puromycin (Fig. 3). sis for each one of the colocalization groups. At an alpha value of Next, we generated four different Arabidopsis lines by crossing 0.05, the experimental distance observed for colocalized and proxi- the SKS14-MS2 or AT59-MS2 with either the RFP-VCS or RFP-DCP1 mal granule pairs, but not for distant granules, was significantly dif- lines. Automated PBs and mRNA granule detection was performed by ferent than the values expected by chance. a MATLAB script as above (for details see “Materials and Methods”). Then, we quantified the number of mRNA granules that colocal- To investigate colocalization, the two structures were modeled as cir- ized or were nearby PBs (Table 2). We found that around 20% of cles using the area and center obtained with the MATLAB algorithm, the SKS14 mRNA bodies colocalized with either VCS or DCP1, while as detailed in Materials and Methods. Colocalization was considered to occur when the distance between the centers of the structures was equal or lower than the sum of their radii. Proximity was defined when the distance was larger but less than twice the sum of their radii. Finally, no relationship was assumed if the mRNA granule and the PB T A B L E 2 SKS14 and AT59 mRNA cytoplasmic granules colocalize with PB proteins. Percentages of granules showing colocalization or proximity for each mRNA and PB protein pair are indicated. N: number of pollen grains analyzed were separated by more than twice the sum of their radii (Fig. S4). The mRNA/protein position and size of the mRNA granules and PBs were determined by SKS14-VCS specific MATLAB scripts (for details see “Materials and Methods”), merged the results and determined whether the mRNA granules colo- SKS14-DCP1 calized, were proximal, or showed no relationship to the PBs protein markers. Fig. 4 a, b, and c shows representative SKS14-MS2 RFP-VCS SKS14-MS2 RFP-DCP1 (Fig. S5), AT59-MS2 RFP-VCS (Fig. S6), and AT59-MS2 RFP-DCP1 (Fig. S7), with similar results. AT59-DCP1 Proximal (%) N 19.6 6.5 46 21.4 16.7 42 23.8 0 21 10 30 20 AT59-VCS pollen grains depicting the three types of spatial colocalization relationships. We performed the same study for three additional lines, Colocalization (%) 13 9.5 63 14.3 11.4 35 11.4 11.4 35 0 34 8.8 SCARPIN | ET AL. Protein DIC (a) 7 approximately 10% of them were found proximal. Similarly, 10% of the AT59 mRNA granules colocalized with the VCS or DCP1 bodies Early Stage and a similar fraction was found nearby them (Table 2). These results suggest that both SKS14 and AT59 mRNAs showed partial localization with PBs identified by VCS or DCP1. 3.3 | SKS14-RFP protein is expressed at early stages during pollen germination Late Stage (b) Next, we asked when the SKS14 mRNA is translated. We generated two independent Arabidopsis lines (S2 and S13) that express both pLAT52::GFP-MCP/pLAT52::SKS14-MS2 and pSKS14::SKS14RFP simultaneously. Fig. 5 shows that the SKS14-RFP protein started to be detected during pollen germination, being localized to (c) the region where the pollen tube will emerge (Fig. 5a-b). Later, SKS14-RFP is accumulated at the margins of pollen tubes as long as they grow (Fig. 5c-d), suggesting that there is an increase in the Pollen tube total amount of SKS14-RFP. These observations agree with previous results that show that the tobacco NTP303 is present at the cell wall and at callose plugs of growing pollen tubes (Wittink et al., 2000). In general, mRNA translation correlates with silencing foci remodeling or dissolution (reviewed in Thomas et al., 2011; Pascual, Maschi, Luchelli, & Boccaccio, 2014). We wonder whether SKS14 expression involves changes in the number of SKS14 mRNA gran- (d) Pollen tube tip ules. We analyzed in the double-transgenic lines (S2 and S13) the presence of SKS14 mRNA granules at the early (ES) and late (LS) stages of pollen germination (Fig. 5a-b). We found a statistically significant decrease in the number of granules at the late stage of pollen germination when compared to the corresponding early stage in the S13 line (Fig. 5e). We also found for both S2 and S13 lines that 0.6 # of granules per PG (e) *** the number of fluorescent mRNA granules at the early stage, but ** not at the late stage, was significantly higher in comparison with the 0.4 control MS2 line (ES-MS2/ES-S2 pair p = .019 (*p < .1) and LSMS2/LS-S2 0.2 pair p = .25; ES-MS2/ES-S13 pair p = .0002 (***p < .001) and LS-MS2/LS-S13 pair p = .48). These results suggest that during pollen germination, when the SKS14-RFP protein is being synthetized, there is a simultaneous decrease in the amount of MS2 LS S E LS E S 0.0 S13 F I G U R E 5 SKS14-RFP protein is expressed from early stages of germination and localizes at the margins of pollen tubes. Representative images of S2 line pollen grains expressing SKS14-RFP protein (left panels). Right panels, DIC images. Size bars, 5 lm for early and late stages (a and b) and 15 lm for the pollen tube (c). (d) A 5X magnification shows localization at the tip (Size bar, 5 lm). (e) Quantification of SKS14 mRNA cytoplasmic granules per pollen grain compared to a MS2 control (MS2 3-1). ES and LS, early and late stages of pollen germination, respectively. Each point represents the mean value of independent samples including 9-10 pollen grains. Lines link data from paired samples (same experiment). Statistically different values are indicated (***p < .001 and **p < .01). The p-value for the ES-MS2/LS-MS2 pair was 0.99, and for the LS-MS2/LS-S13 pair, 0.48 (two-way ANOVA randomized block, Bonferroni post-test) SKS14 mRNA granules, suggesting that the SKS14 mRNA granules release their content to allow a controlled production of the SKS14 protein. 4 | DISCUSSION Both in plants and animals, gamete development depends on the translation of stored mRNAs. The post-transcriptional control of genes expressed during the spermatogenesis in fly and mouse is well known (Lasko, 2012; Nguyen-Chi & Morello, 2011). In plants, the translational inhibitor cycloheximide, but not the transcriptional inhibitor actinomycin D, inhibits early pollen tube growth, suggesting that translation of preexisting mRNAs is required (Hao, 8 | SCARPIN ET AL. Li, Hu, & Lin, 2005). Studies on the regulation of the tobacco in close contact. We speculate that pollen grains may contain several NTP303 and tomato LAT59 genes demonstrated that they are types of PBs, and only a fraction of them would contain VCS and both transcribed during pollen development while their proteins DCP1. Several animal cell types display a variety of PB containing are mostly or exclusively synthesized after germination, respec- subsets of PB components, and in addition, it has been proposed tively (Curie & McCormick, 1997; Hulzink et al., 2002). Antisense that mRNAs are differentially located on PBs depending on their NTP303 plants are male sterile due to the arrest of pollen tubes translational requirements (Weil et al., 2012). In mammalian neurons, within the style (de Groot et al., 2004). Fig. 5 shows a negative only 50% of dendritic DCP1a bodies contain Hedls, the VCS ortho- correlation between the presence of SKS14 mRNA granules and log, and vice versa, 50% of Hedls-positive dendritic puncta contain RFP-SKS14 expression, which is compatible with a role for the DCP1a (Luchelli, Thomas, & Boccaccio, 2015). Likewise, in untreated granules in mRNA storage. We found that upon initiation of pol- mammalian cells, only 18% of the beta-actin and Cerulean-mini-dys- len germination, the related SKS14 protein localizes where the trophin mRNAs colocalized with PB using Hedls as a protein marker pollen tube emerges, and later in the pollen tube margins. (Aizer et al., 2014). Further analysis including super-resolution micro- Tobacco NTP303 is related to ascorbate oxidases and localizes at scopy and the study of additional PB proteins and RNA regulation the plasma membrane (Wittink et al., 2000), and we propose that pathways will shed light on the dynamic relationship between pollen SKS14 could have a similar role in Arabidopsis pollen. mRNAs and PBs and will allow a more complete understanding of Here, we visualized the mRNAs of the Arabidopsis SKS14 and their translational regulation. AT59 mRNAs in mature pollen and during germination. We identi- In several examples, translation is repressed by proteins that bind fied SKS14 and AT59 mRNA in cytoplasmic granules related to PBs to the UTRs and mRNA silencing is necessary for the proper localiza- in mature pollen and propose that these bodies would function in tion and function of the encoded proteins (Chartrand et al., 2002; mRNA storage while waiting for being translated. In addition, the Jambor, Brunel, & Ephrussi, 2011). Several common elements in the presence of PB components in these RNA granules may be linked 50 UTR of pollen-expressed genes have been identified (Hulzink et al., to mRNA degradation. However, we favor the notion of mRNA 2003). Some of these consensus sequences are present in the storage given that SKS-RFP accumulates latter during pollen tube NTP303 50 UTR and affect translation efficiency (Hulzink et al., growth (Fig. 5c). 2002), thus opening the possibility that these pollen-specific 50 UTR It is generally accepted that mRNAs translationally inactive are sequences play a regulatory role during development and germina- stored in stress granules (SGs), processing body (PB), or related orga- tion. While the repression mechanism remains poorly understood, it nelles (Decker & Parker, 2012; Thomas et al., 2014). We propose has been proposed that translation of the NTP303 mRNA is acti- that an active exchange of mRNAs between PBs and polysomal vated by factors that bind to the 50 UTR after pollen germination mRNPs finely tune gene expression during the late stages of pollen (Hulzink et al., 2002). In turn, binding of regulatory factors to the development and during germination. Supporting this, in a pioneer 50 UTR of the LAT59 mRNA would inhibit translation at early stages work, the NTP303 mRNA was found in polysomal and in a fraction and the release of the repressors upon pollen maturation would of particles resistant to EDTA/puromycin treatment referred to as allow LAT59 protein synthesis (Curie & McCormick, 1997). Whether EPPs (Honys, Combe, Twell, & Capková, 2000). These particles these 50 UTRs sequences mediate the targeting of these transcripts have been proposed as long-term storage complexes. EPPs include to PBs remain to be investigated. In this regard, it has been recently a set of mRNAs that are stored and translationally repressed at reported that CGG repeats in the 50 UTR of the Fragile X Mental early stages of pollen development. Some of the stored mRNAs are Retardation 1 (FMR1) RNA mediate RNA localization into cytoplas- massively translated at late stages of pollen development and/or mic granules (Rovozzo et al., 2016). transported to the pollen tube where they are translated. Rele- To securely define which pollen mRNAs are stored or translated, vantly, most of the translationally inactive NTP303 mRNA was pre- a robust and tight post-transcriptional regulation is expected to sent both in the polysomal and EPP fractions. Whether the EPPs occur. The presence of cytoplasmic granules that contain the SKS14 and the granules identified here represent the same entity remains and AT59 mRNAs in mature pollen suggests a mechanism for trans- open. Against this possibility, VCS and DCP1 are absent from lational regulation. Our results suggest that granules related to PBs tobacco EPPs (Honys et al., 2009). store mRNAs to postpone their translation until necessary or that We also described the presence of processing body in Arabidop- they regulate mRNA levels. An appealing speculation is whether the sis mature pollen. As in other cell types and organisms, RFP-VCS observations here reported for SKS14 and AT59 mRNAs can be and RFP-DCP1 localized to discrete cytoplasmic bodies that become extrapolated to other late pollen genes. larger and more abundant upon treatment with puromycin, suggesting that pollen PBs recruit transcripts that are released from active polysomes. Accumulation of translationally repressed mRNAs in PBs AUTHOR CONTRIBUTIONS has been previously reported in yeast, Drosophila, and mammals MRS and JPM conceived the experiments; MRS, LS, SGT, GLB, LIP, (Decker & Parker, 2012; Thomas et al., 2011). However, when ana- and JPM designed the research; MRS performed the experiments; lyzing the colocalization of SKS14 and AT59 mRNAs with VCS or LS, SGT, and LIP provided technical assistance; MRS, LS, SGT, GLB, DCP1, we found partial overlapping with about 30% of the granules LIP, and JPM analyzed and interpreted the data; MRS and JPM SCARPIN ET AL. wrote the manuscript with contributions of all the authors; GLB complemented the writing. ACKNOWLEDGMENTS We thank Dr. Federico Fuentes for the excellent confocal microscopy technical assistance. REFERENCES Aizer, A., Kalo, A., Kafri, P., Shraga, A., Ben-Yishay, R., Jacob, A., . . . ShavTal, Y. (2014). Quantifying mRNA targeting to P bodies in living human cells reveals a dual role in mRNA decay and storage. Journal of Cell Science, 127, 4443–4456. https://doi.org/10.1242/jcs.152975 Bertrand, E., Chartrand, P., Schaefer, M., Shenoy, S. M., Singer, R. H., & Long, R. M. (1998). Localization of ASH1 mRNA particles in living yeast. Molecular Cell, 2(4), 437–445. https://doi.org/10.1016/S10972765(00)80143-4  , J. A. (2005). The making of gametes Boavida, L. C., Becker, J. D., & Feijo in higher plants. International Journal of Developmental Biology, 49(5– 6), 595–614. https://doi.org/10.1387/ijdb.052019lb Brown, S. M., & Crouch, M. L. (1990). Characterization of a gene family abundantly expressed in Oenothera organensis pollen that shows sequence similarity to polygalacturonase. The Plant Cell, 2(3), 263– 274. https://doi.org/10.1105/tpc.2.3.263 Chartrand, P., Meng, X. H., Huttelmaier, S., Donato, D., & Singer, R. H. (2002). Asymmetric sorting of ash1p in yeast results from inhibition of translation by localization elements in the mRNA. Molecular Cell, 10(6), 1319–1330. Clough, S., & Bent, A. (1998). Floral dip: A simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal, 16(6), 735–743. Curie, C., & McCormick, S. (1997). A strong inhibitor of gene expression in the 50 untranslated region of the pollen-specific LAT59 gene of tomato. The Plant Cell, 9(11), 2025–2036. Decker, C. J., & Parker, R. (2012). P-bodies and stress granules : Possible roles in the control of translation and mRNA degradation. Cold Spring Harbor Perspectives in Biology, 4(9), 1–16. https://doi.org/10.1101/csh perspect.a012286 Fusco, D., Accornero, N., Lavoie, B., Shenoy, S. M., Blanchard, J. M., Singer, R. H., & Bertrand, E. (2003). Single mRNA molecules demonstrate probabilistic movement in living mammalian cells. Current Biology, 13(2), 161–167. https://doi.org/10.1016/S0960-9822(02)014367 de Groot, P., Weterings, K., de Been, M., Wittink, F., Hulzink, R., Custers, J., . . . Wullems, G. (2004). Silencing of the pollen-specific gene NTP303 and its family members in tobacco affects in vivo pollen tube growth and results in male sterile plants. Plant Molecular Biology, 55(5), 715–726. https://doi.org/10.1007/s11103-0041964-6 Hao, H., Li, Y., Hu, Y., & Lin, J. (2005). Inhibition of RNA and protein synthesis in pollen tube development of Pinus bungeana by actinomycin D and cycloheximide. New Phytologist, 165(3), 721–729. https://doi. org/10.1111/j.1469-8137.2004.01290.x Honys, D., Combe, J. P., Twell, D., & Capkova, V. (2000). The translationally repressed pollen-specific ntp303 mRNA is stored in non-polysomal mRNPs during pollen maturation. Sexual Plant Reproduction, 13 (3), 135–144. https://doi.org/10.1007/s004970000047 Honys, D., Ren, D., Fecikova, J., Jedelsky, P. L., Nebesa, J., & Dobrev, P. (2009). Cytoskeleton-associated large RNP complexes in tobacco male gametophyte (EPPs) are associated with ribosomes and are involved in protein synthesis, processing, and localization. Journal of Proteome Research, 8, 2015–2031. | 9 Honys, D., & Twell, D. (2003). Comparative analysis of the arabidopsis pollen transcriptome 1. Plant Physiology, 132(2), 640–652. https://doi. org/10.1104/pp.103.020925.fitness Hulzink, R. J. M., de Groot, P. F. M., Croes, A. F., Quaedvlieg, W., Twell, D., Wullems, G. J., & van Herpen, M. M. A. (2002). The 5 -untranslated region of the ntp303 gene strongly enhances translation during pollen tube growth, but not during pollen maturation. Plant Physiology, 129(1), 342–353. https://doi.org/10.1104/pp.001701.342 Hulzink, R. J. M., Weerdesteyn, H., Croes, A. F., Gerats, T., van Herpen, M. M. A., & van Helden, J. (2003). In silico identification of putative regulatory sequence elements in the 5 -untranslated region of genes that are expressed during male gametogenesis. Plant Physiology, 132 (1), 75–83. https://doi.org/10.1104/pp.102.014894 Jambor, H., Brunel, C., & Ephrussi, A. (2011). Dimerization of oskar 30 UTRs promotes hitchhiking for RNA localization in the Drosophila oocyte. RNA, 17(12), 2049–2057. https://doi.org/10.1261/rna. 2686411 Kulikauskas, R., & McCormick, S. (1997). Identification of the tobacco and Arabidopsis homologues of the pollen-expressed LAT59 gene of tomato. Plant Molecular Biology, 34(5), 809–814. Lampasona, A. A., & Czaplinski, K. (2016). RNA voyeurism: A coming of age story, Methods, 98, (pp. 10–17). London: Elsevier Inc. https://doi. org/10.1016/j.ymeth.2015.11.024 Lasko, P. (2012). mRNA localization and translational control in Drosophila oogenesis. Cold Spring Harbor Perspectives in Biology, 4(10), 1–15. Loraine, A. E., McCormick, S., Estrada, A., Patel, K., & Qin, P. (2013). RNA-seq of arabidopsis pollen uncovers novel transcription and alternative splicing. Plant Physiology, 162(2), 1092–1109. https://doi.org/ 10.1104/pp.112.211441 Luchelli, L., Thomas, M. G., & Boccaccio, G. L. (2015). Synaptic control of mRNA translation by reversible assembly of XRN1 bodies. Journal of Cell Science, 128(8), 1542–1554. https://doi.org/10.1242/jcs.163295 Maldonado-Bonilla, L. D. (2014). Composition and function of P bodies in Arabidopsis thaliana. Frontiers In Plant Science, 5, 201. https://doi. org/10.3389/fpls.2014.00201 Mu, J. H., Stains, J. P., & Kao, T. (1994). Characterization of a pollenexpressed gene encoding a putative pectin esterase of Petunia inflata. Plant Molecular Biology, 25(3), 539–544. https://doi.org/10. 1007/bf00043881 Muschietti, J., Dircks, L., Vancanneyt, G., & McCormick, S. (1994). LAT52 protein is essential for tomato pollen development: Pollen expressing antisense LAT52 RNA hydrates and germinates abnormally and cannot achieve fertilization. Plant Journal, 6(3), 321–328. https://doi.org/ 10.1046/j.1365-313X.1994.06030321.x Nguyen-Chi, M., & Morello, D. (2011). RNA-binding proteins, RNA granules, and gametes: Is unity strength? Reproduction, 142(6), 803–817. https://doi.org/10.1530/REP-11-0257 Niogret, M. F., Dubald, M., Mandaron, P., & Mache, R. (1991). Characterization of pollen polygalacturonase encoded by several cDNA clones in maize. Plant Molecular Biology, 17(6), 1155–1164. https://doi.org/ 10.1007/bf00028732 Rogers, H. J., Harvey, A., & Lonsdale, D. M. (1992). Isolation and characterization of a tobacco gene with homology to pectate lyase which is specifically expressed during microsporogenesis. Plant Molecular Biology, 20(3), 493–502. https://doi.org/10.1007/bf00040608 Rogers, H. J., & Lonsdale, D. M. (1992). Genetic manipulation of male sterility for the production of hybrid seed. Plant Growth Regulation, 11(1), 21–26. Rovozzo, R., Korza, G., Baker, M. W., Li, M., Bhattacharyya, A., Barbarese, E., & Carson, J. H. (2016). CGG repeats in the 50 UTR of FMR1 RNA regulate translation of other RNAs localized in the same RNA granules. PLoS ONE, 11(12), 1–23. https://doi.org/10.1371/journal.pone. 0168204 Schrauwen, J. A. M., Derksen, J., Knuiman, B., & Wullems, G. J. (1999). ment, E. The location of the pollen-specific NTP303 protein. In C. Cle 10 | Pacini & J. C. Audran (Eds.), Anther and pollen (pp. 113–118). Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-642-599859_10 Sedbrook, J. C., Carroll, K. L., Hung, K. F., Masson, P. H., & Somerville, C. R. (2002). The arabidopsis SKU5 gene encodes an extracellular glycosyl phosphatidylinositol – anchored glycoprotein involved in directional root growth. The Plant Cell, 14(7), 1635–1648. https://doi.org/ 10.1105/tpc.002360.1996 Thomas, M. G., Loschi, M., Desbats, M. A., & Boccaccio, G. L. (2011). RNA granules: The good, the bad and the ugly. Cellular Signalling, 23 (2), 324–334. https://doi.org/10.1016/j.cellsig.2010.08.011 Thomas, M. G., Pascual, M. L., Maschi, D., Luchelli, L., & Boccaccio, G. L. (2014). Synaptic control of local translation: The plot thickens with new characters. Cellular and Molecular Life Sciences, 71(12), 2219– 2239. https://doi.org/10.1007/s00018-013-1506-y Twell, D. (1994). The diversity and regulation of gene expression in the pathway of male gametophyte development. In R. Scott & A. Stead (Eds.), Molecular and cellular aspects of plant reproduction (Society for Experimental Biology Seminar Series)(pp. 83–136). Cambridge: Cambridge University Press. https://doi.org/10.1017/ cbo9780511752339.007 Twell, D., Klein, T. M., Fromm, M. E., & McCormick, S. (1989). Transient expression of chimeric genes delivered into pollen by microprojectile bombardment. Plant Physiology, 91(4), 1270–1274. Twell, D., Yamaguchi, J., Wing, R. A., Ushiba, J., & McCormick, S. (1991). Promoter analysis of genes that are coordinately expressed during pollen development reveals pollen-specific enhancer sequences and shared regulatory elements. Genes and Development, 5(3), 496–507. https://doi.org/10.1101/gad.5.3.496 Wakeley, P. R., Rogers, H. J., Rozycka, M., Greenland, A. J., & Hussey, P. J. (1998). A maize pectin methylesterase-like gene, ZmC5, specifically expressed in pollen. Plant Molecular Biology, 37(1), 187–192. https://d oi.org/10.1023/a:1005954621558 Weil, T. T., Parton, R. M., Herpers, B., Soetaert, J., Veenendaal, T., Xanthakis, D., . . . Davis, I. (2012). Drosophila patterning is established by differential association of mRNAs with P bodies. Nature Cell Biology, 14(12), 1305–1313. https://doi.org/10.1038/ncb2627 Weterings, K., Reijnen, W., van Aarssen, R., Kortstee, A., Spijkers, J., van Herpen, M., . . . Wullems, G. (1992). Characterization of a pollen- SCARPIN ET AL. specific cDNA clone from Nicotiana tabacum expressed during microgametogenesis and germination. Plant Molecular Biology, 18(6), 1101–1111. https://doi.org/10.1007/BF00047713 Wing, R. A., Yamaguchi, J., Larabell, S. K., Ursin, V. M., & McCormick, S. (1989). Molecular and genetic characterization of two pollenexpressed genes that have sequence similarity to pectate lyases of the plant pathogen Erwinia. Plant Molecular Biology. Springer, 14(1), 17–28.  Twell, D., Schrauwen, Wittink, F. R. A., Knuiman, B., Derksen, J., Vera, C.,  a, V. (2000). The pollen-specific gene Ntp303 J. A. M., . . . Capkov encodes a 69-kDa glycoprotein associated with the vegetative membranes and the cell wall. Sexual Plant Reproduction, 12(5), 276–284. https://doi.org/10.1007/s004970050195 Xu, J., & Chua, N.-H. (2009). Arabidopsis decapping 5 is required for mRNA decapping, P-body formation, and translational repression during postembryonic development. The Plant Cell, 21(10), 3270–3279. https://doi.org/10.1105/tpc.109.070078 Xu, J., Yang, J.-Y., Niu, Q.-W., & Chua, N.-H. (2006). Arabidopsis DCP2, DCP1, and VARICOSE form a decapping complex required for postembryonic development. The Plant Cell, 18(12), 3386–3398. https://doi.org/10.1105/tpc.106.047605 SUPPORTING INFORMATION Additional Supporting Information may be found online in the supporting information tab for this article. How to cite this article: Scarpin MR, Sigaut L, Temprana SG, Boccaccio GL, Pietrasanta LI, Muschietti JP. Two Arabidopsis late pollen transcripts are detected in cytoplasmic granules. Plant Direct. 2017;00:1–10. https://doi.org/10.1002/pld3.12