PCR amplification and cloning of the AtTCTP2 ORF into the pCR8/GW/TOPO vector (Invitrogen, Carlsbad CA) was carried out previously (Toscano-Morales et al., submitted). Briefly, the AtTCTP2 ORF was recombined into the pB7FWG2 Gateway binary vector (Plant Genetic Systems, Ghent, Belgium). The resulting construct was then introduced by electroporation into A. rhizogenes strain K599.

Transformation of tobacco leaf explants was carried out by the puncture method described by Hinojosa-Moya et al. (2013). Basically, cultures containing A. rhizogenes K599 harboring the recombinant plasmids were grown on selective liquid media, concentrated, and resuspended in fresh media. Young tobacco leaves (2–3 cm diameter) were sterilized using ethanol 70% (1 min) followed by soaked in 10% sodium hypochlorite for 30 min and several washes with sterile water. Then, these leaves were inoculated with the previously prepared bacterial suspension employing a sterile insulin syringe. After transformation the explants were placed on MS medium with no hormone supplementation [1.0 MS salts, 2% sucrose, and 0.4% agar (Gelrite)] and finally transferred to controlled environment growth chamber under long-day conditions (16 h light/8 h dark) during 30 days. The regenerated plantlets were transferred to sterile-soil, watered and kept in transparent plastic bags to maintain their turgor and low CO₂ concentrations. One week later, plastic bags were removed and plants kept under greenhouse conditions until reaching maturity.

F1 seeds were harvested and sowed on soil, where these were subject to herbicide (Ammonium glufosinate-FINALE, Bayer, Germany) selection to isolate the transgenic offspring (segregation analysis were not performed). Total DNA was extracted from all candidate plants (including wild type) and each one of these was used as template for transgene PCR detection, employing GFP primers (forward: 5′-ATGGTGAGCAAGGGCGAGGAGCTG-3′; reverse: 5′-CCTTGTACAGCTCGTCCATGC-3′). PCR positive plants were then isolated in individual pots and grown 1–2 weeks under greenhouse conditions.

Grafting assays were conducted using a combination of the grafting methods described by Palauqui et al. (1997) and Imlau et al. (1999). Essentially, the stock is decapitated 20–25 cm above the soil, and the outermost cortex of the stock stem is cut longitudinally to produce a cortex flap; after this the terminal apex of the scion (carrying 2–3 leaves of 0.5–1 cm) is excised, chamfered and attached to the stock in the gap between the flap and the stem, to finally be wrapped with parafilm. To avoid desiccation and turgor loss, grafted plants were maintained under transparent plastic bags for 10 days under controlled conditions in a growth chamber. Then, plastic bags were removed and plants were kept for another 5–7 days under these conditions. Finally, grafted plants were transferred to a greenhouse; 2–3 weeks later, newly developed sink leaves near the scion apex (<1 cm) and fully expanded source leaves in the stock (>2 cm) were harvested, frozen with liquid nitrogen and stored at −80°C for further use.

Total RNA was extracted from each sample, with an average weight of 15–20 mg per sample, using the RNAeasy Kit (QIAgen, Hilden, Germany). Total RNA concentrations from samples were normalized, and used as templates to perform one-step RT-PCR reactions for each sample employing a commercial RT-PCR system (KAPA FAST Universal One-Step q-RT PCR Kit) in order to detect AtTCTP2-GFP and 18S (as control) mRNAs using specific primers [GFP (see above, transgenic plant selection) and 18S (direct: 5′-GCCCGGGTAATCTTTGAAATTTCAT-3′; reverse: 5′-GTGTGTACAAAGGGCAGGGACGTA-3′)]. The conditions of the one step RT-PCR were as follows: (1) 42°C–10 min, (2) 95°C–5 min, and (3) 95°C–5 s/62°C–30 s/72°C–15 s (40 cycles). Finally, amplicons were visualized on a 1.2% agarose gel. The amplicons obtained are 720 bp for AtTCTP2-GFP and 152 bp for 18S.

GFP, AtTCTP2, BAR, NtTCTP, and 18S RNA levels were determined as follows: total RNA was extracted for each graft sample (stock or scion) and used for one-step RT-PCR (100 ng in a 10 μL reaction). A commercial system was used according to the manufacturer′s recommendations (KAPA SYBR FAST Universal One-Step qRT-PCR Kit). Specific primers for GFP, 18S (shown previously in this section), AtTCTP2 (direct: 5′-ATGTTGGTCTACCAGGATATTCTTACA-3′; reverse: 5′-GCACTTGATCTCTTTCAAGCCGTAGGC-3′), and NtTCTP (direct: 5′-GGAAGTGGGTTGTTCAGGGAGCTGTTGATG-3′; reverse: 5′-TGTTCTTAAAAACTTCTTCCTGCTCTGCGCCT-3′) were used. The Real Time qRT-PCR reactions were incubated in a Rotor Gene 3000 apparatus (Corbett Research, Australia) using the following PCR conditions: 5 min at 42°C for reverse transcription followed by 3 min at 95°C with 40 cycles of denaturation (95°C for 3 s), annealing (60°C for 20 s) and extension (72°C for 3 s). To verify that no additional products were amplified in the reaction, a dissociation curve was generated through progressive sample heating (60–95°C). The Ct value for each product was determined by triplicate in each treatment. 18S rRNA mRNA was used to normalize gene expression. Relative quantification for transcript accumulation was performed according to the comparative ΔΔ CT method (Livak and Schmittgen, 2001) relating fold changes in GFP, AtTCTP2, between transgenic samples and its WT counterpart (either stocks or scions).

Western blot assay was performed as described by Hinojosa-Moya et al. (2013). In resume, total protein was extracted by grinding plant tissue in liquid nitrogen, homogenized in protein extraction buffer, and centrifuged at 13,000 g for 2 min. Supernatants were recovered and dissolved in SDS-PAGE sample buffer. Protein concentration was determined using a spectrophotometer (NanoDrop™ 1000; Thermo Scientific, Waltham, MA) and all samples homogenized to the same concentration followed by 12% SDS-PAGE. Total proteins were transferred for 1 h at 100 V to polyvinylidene difluoride (PVDF) membranes (Whatman), blocked for 2 h in blocking solution (PBS 1X, 5% non-fat milk, and 0.1% Tween 20) followed by rinsing with PBS 1X, and incubated overnight at 4°C with the polyclonal GFP antibody (diluted 1:2000 in 1X PBS, 5% non-fat milk, and 0.1% Tween 20). Membranes were washed (1X PBS, 0.1% Tween 20) 5 times and incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology, CA) at 1:5000 in PBS 1X containing 2.5% skim milk during 2 h. Finally, after several washes, the signal was detected with HRP color development reagent (Amersham Prime Western Blotting Kit Reaction-ECL™-GE Healthcare™) and revealed in an Amersham Hyperfilm (ECL™-GE Healthcare™) film.

Samples from stocks and scions were analyzed on a Leica confocal laser scanning microscope (model TC-SP5/MO-TANDEM) using a krypton/argon with a laser excitation fluorescence/emission of 488/525 nm for green fluorescence and 580 nm/665 nm emission. All images were recorded and analyzed with Leica Las AF software, followed by processing using Photoshop 8.0 software (Adobe) as described (Xoconostle-Cázares et al., 1999).

Parafilm-junction was removed from grafted plants after 30 days, after which adventitious root appearance was visualized using a Sony Cybershot DSC-W300 digital camera (13.6 megapixels) harboring a 3× (35–105 mm) Carl Zeiss Vario-Tessar lens f/2.8–5.5.

Transgenic tobacco plants harboring the overexpression construct 35S:AtTCTP2-GFP-TNOS were obtained using the regeneration protocol described by Hinojosa-Moya et al. (2013). Taking advantage of this, the progeny of these regenerated plants was selected for herbicide resistance and tested for presence of trangene by PCR. Transgenic and WT plants of similar size and developmental stage were used to perform reciprocal heterografting experiments as well as WT homografts. Total RNA was extracted from young and mature leaves in all grafting experiments; these were used as templates to perform the RT-PCR detection of the AtTCTP2-GFP transcript to test for long-distance movement. Signal corresponding to GFP was detected in three out of six WT scions grafted onto transgenic stocks (Table 1), indicating long-distance movement of this mRNA (Figure 1A). Interestingly, in the case of the reciprocal heterograft, this signal was identified in one of seven grafts (Table 1, Figure 1B) suggesting directionality of the movement of this mRNA, likely from stock to scion (AtTCTP2-GFP/WT). The control homografts (WT/WT) did not show any transgenic signal (Figure 1C); 18S RNA was detected in all the samples at similar levels.

The movement of AtTCTP2-GFP mRNA from transgenic stocks to WT scions, and from transgenic scions to WT rootstocks was determined by quantitative RT-PCR. The results indicate that the percentage of transcript that moved across the graft union, obtained as the ratio of AtTCTP2-GFP mRNA present in the non-transgenic scion, and that in the transgenic stock, is between 7 and 9% (Figure 2). Interestingly, AtTCTP-GFP mRNA was also detected moving rootward from transgenic scion to non-transgenic WT stock, albeit at lower efficiencies (between 1 and 1.5%).

To discard that only AtTCTP2-GFP transcript is capable of moving long-distance, the presence of npt2 mRNA was analyzed in WT/AtTCTP2-GFP and AtTCTP2-GFP/WT heterografted plants, as well as in WT controls. Quantitative RT-PCR was carried out for this end. As shown in Figure 2, no npt2 mRNA was detected in non-transgenic scions or stocks grafted to AtTCTP2-GFP expressing plants. Thus, only AtTCTP2-GFP mRNA is transported long-distance.

The previous results could not discriminate whether the AtTCTP2-GFP transcript, the protein, or both, were able to move across the graft union. In order to answer this question, the presence of the fusion protein was determined by Western blot analysis of total proteins from scions and stocks of transgenic plants. Independent scions and stocks were used in which AtTCTP2 mRNA movement was observed. In all cases, the fusion protein was detected in both WT scion grafted onto a transgenic stock; in the case of WT stock onto which a transgenic scion had been grafted, one out of two AtTCTP2 could be detected in the former (Figure 3). These results indicate that AtTCTP2-GFP moves long distance as both mRNA and protein, and that such movement is not always in the direction from source to sink.

To confirm the long-distance movement of AtTCTP2 protein, as well as the localization pattern of the fluorescent fusion protein, confocal microscopy was used to detect GFP-associated fluorescence and thus AtTCTP2. Tissue samples were collected from the same leaves used for RT-PCR (marked in red in Figure 4G), and prepared for confocal analysis. GFP fluorescence was detected in mesophyll, stomata and nuclei in leaf tissue from transgenic stocks or scions (Figures 4A,D), which is the same accumulation pattern of AtTCTP2 in Arabidopsis (Toscano-Morales et al., submitted). GFP signal was also identified in all WT scions and in four out of seven stocks (Figures 4B,E; Table 1) showing a highly similar pattern of localization. No signal was detected in WT/WT controls either in stock or scion (Figures 4C,F). These observations suggest the AtTCTP2-GFP fusion protein moves long distance in both directions, but preferentially from stock to scion (Table 1). Interestingly, AtTCTP2-GFP protein was detected in all WT scions, even in those samples negative by RT-PCR (Table 1). This also suggests that the fluorescence signal detected could not only be attributed to the translation of this mRNA once it moved long-distance but to the actual movement of AtTCTP2 as protein. Overall, both RT-PCR and fluorescence detection of AtTCTP2-GFP indicate its capacity for long-distance movement as protein and/or as transcript.

Adventitious roots are normally found in some plant species and have a role in adaptation to stress or nutrient deficiency (Drew et al., 1979), as well as for vegetative propagation of some tree species (Naiman and Décamps, 1997). The emergence of aerial roots in the adventitious region in the graft interface was observed in several of the grafting experiments performed (Figure 5). Indeed, such adventitious roots were observed in the sites adjacent to the graft union in all AtTCTP2/WT grafts (Figure 5A), and in some of the WT/AtTCTP2 grafts (Figure 5B) but not in the control (Figure 5C). Interestingly, the appearance of these adventitious roots was specifically associated to the grafts where GFP-associated fluorescence (and hence AtTCTP2) was detected in WT scion or stock, and thus, in which long-distance transport of the protein occurred (Table 1); interestingly, in some cases fluorescence was observed, even though AtTCTP2 mRNA was detected. This suggests that the emergence of these aerial roots is linked to the movement of the protein rather than to movement of the transcript. No GFP-associated fluorescence was observed in equivalent tissues from WT homografts.

Adventitious aerial roots were analyzed by final point PCR and RT-qPCR to determine the presence of the transgene and its expression levels. All aerial roots were transgenic (harboring the 35S::AtTCTP2-GFP construct) and positive for the AtTCTP2-GFP transcript (data not shown). Moreover, the AtTCTP2 mRNA detection is in agreement with previous work showing that the expression pattern of the promoter region of AtTCTP2 is active in the root with an atypical pattern in the lateral root primordia, similar to the PUCHI gene (Toscano-Morales et al., submitted; Karim et al., 2009).

In addition, confocal fluorescence laser microscopy was performed to detect AtTCTP2-GFP protein signal and determine its localization as well. The AtTCTP2-GFP fluorescent signal was specifically found in the nuclei of the mid-root region in all aerial roots that emerged (Figure 6A) which is consistent with the localization pattern found in the 35S::AtTCTP2-GFP transgenic primary roots (Figure 6B; Toscano-Morales et al., submitted) and in contrast with the absence of nuclear signal in the primary mid-root region of WT stock controls (Figure 6C). Interestingly, no nuclear localization signal is predicted for AtTCTP2, or several plant TCTPs tested, using a nuclear predictor server (http://www.sbc.su.se/~maccallr/nucpred/). In a similar manner, the same localization pattern was observed in lateral root primordia both in aerial and transgenic primary root (Figures 6D,E,G,H), unlike the WT stock control (Figures 6F,I). Remarkably, the protein accumulation pattern (in cells at the base of the lateral root primordia) observed in all aerial roots of WT/35S::AtTCTP2-GFP scions was reminiscent of the transgenic primary roots, suggesting an important role of AtTCTP2 and its nuclear localization in midroot growth and lateral root development; this also suggests that AtTCTP2 is capable of moving into the scion and into nuclei of aerial roots, which appear to be induced by AtTCTP2 itself. It must be noted that in these roots fluorescence is not restricted to the nucleus. Instead, signal can also be observed in the cell periphery, and, at lower levels, in cytoplasm; this is much higher than the background autofluorescence of WT roots.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.