For trans-Zeatin application experiments, N. benthamiana (accession Nb-1) plants were grown in a greenhouse at 22°C and in 16-h daylengths for 4 weeks prior to infiltration. For transient overexpression experiments, growth conditions were identical, but plants were grown for 4 or 5 weeks before infiltration, depending on the size of the plant. For virus-induced gene silencing (VIGS) experiments, plants were grown in autoclaved soil and in isolation from other plants to prevent the presence of pathogens and pests at 22°C and in 16-h daylengths for 3 weeks prior to silencing. All plants used in the VIGS experiments were photographed prior to infiltration to assess phenotypic differences among AHK3-, AHK4-, and mock β-glucuronidase (GUS)-silenced plants and validate effective silencing with PHYTOENE DESATURASE (PDS)-silenced plants. As previously described (Brunkard et al., 2015a), plants that have effectively silenced PDS exhibit photobleached leaves.

Silencing triggers were cloned as previously described (Brunkard et al., 2015a). Briefly, RNA was isolated from N. benthamiana, Nb-1, with the Spectrum Plant Total RNA kit (Sigma-Aldrich, St. Louis, MO, United States), treating RNA with on-column DNase I digestion (New England Biolabs, Ipswich, MA, United States). Complementary DNA (cDNA) was synthesized from isolated RNA using random hexamers and SuperScript III reverse transcriptase (Fisher Scientific, Waltham, MA, United States). Silencing triggers were amplified with Phusion DNA polymerase (New England Biolabs). Triggers and the TRV2 plasmid pYL156 were digested with XbaI and XhoI (New England Biolabs) and ligated with Promega T4 DNA ligase (Fisher Scientific). Ligations were transformed into XL1-Blue Escherichia coli, were mini-prepped (New England Biolabs, Ipswich, MA, United States), and were Sanger sequenced to confirm insertion sequences. The AHK4 trigger was cloned with oligonucleotides 5'-gat TCT AGA AAC TAT GGA GGA ACG GG-3' and 5'-gat ctc GAG GTT TCA TTA TCA CCG C-3' to silence the two AHK4 homologues, Niben101Scf08855g02013 and Niben101g09260g03002. The AHK3 trigger was cloned with oligonucleotides 5'-cat TCT AGA TGT GAC ACA ACA AGA TTA TGT C-3' and 5'-gat ctC GAG CAA TAG AAG GAC CAA C-3' to silence the two AHK3 homologues, Niben10103911g05013 and Niben101Scf02711g00003.

As with the silencing triggers, AHP5 and AHP6 were amplified with Phusion DNA polymerase (New England Biolabs) from N. benthamiana cDNA that was synthesized as above. AHP5 was cloned using oligonucleotides 5'-aattacaggcctcccgggaccATGAACACCATCGTCGTT-3' and 5'-TTCGCTTCCTGAccc CTAATTTATATCCACTTGAGGAATT-3', while AHP6 was cloned using oligonucleotides 5'-aattacaggcctcccgggaccATGTTGGGGTTGGGTGTG-3' and 5'-TTCGCTTCCTGAcccCTACATTGGATATCTGACTCCTGC-3'. All oligonucleotides contained 15 bp of homology compatible with the SmaI digestion site of binary vectors containing a CaMV 35S promoter, TMV Omega enhancer sequence, and CaMV 35S terminator for transient gene expression. The plasmid was digested with SmaI, and both genes were treated with T5 DNA exonuclease (New England Biolabs) as previously described (Xia et al., 2019). These reactions were used to transform chemically competent DH10B E. coli, were mini-prepped, and were Sanger sequenced to confirm insertion sequences.

Agrobacterium tumefaciens strain, GV3101, was grown overnight in a lysogeny broth medium at 28°C, 250 rpm, with kanamycin, gentamicin, and rifampicin (each at 50 mg ml⁻¹). Cultures were centrifuged at ×700 g for 10 min and then resuspended in an infiltration medium [10 mM MgCl₂, 10 mM 2-(N-morpholino)ethanesulfonic acid (MES), and 200 μM acetosyringone, pH 5.6, adjusted with KOH] to OD_600nm = 1.0. Agrobacteria were then left to induce virulence at room temperature for 2–4 h with gentle shaking prior to infiltration. Immediately before infiltrating, cultures were then further diluted in infiltration media to OD_600nm = 10⁻⁵ for green fluorescent protein (GFP) movement assays or OD_600nm = 0.1 for overexpression. Cultures were left at OD_600nm = 1.0 for VIGS, as previously described (Brunkard et al., 2015a).

Briefly, for VIGS experiments, the first two true leaves of N. benthamiana plants were infiltrated with equal induced inocula of A. tumefaciens carrying the previously described binary vectors: pYL192 (which expresses the TRV1 subgenome) and pYL156 (which expresses the TRV2 subgenome with silencing triggers). A TRV2-GUS trigger was used as a negative control for VIGS, and a TRV2-NbPDS trigger was used as a positive control for VIGS. About 14 days after infiltration of the VIGS inocula, the fourth leaf of each plant was infiltrated with an induced inoculum containing the 35S_PRO:GFP binary vector diluted to OD_600nm = 10⁻⁵. Each plant was left under normal growing conditions (22°C and 16-h daylengths) for 48 h prior to GFP movement assays (described below).

For overexpression experiments, the fourth leaf from 4-to-5-week-old N. benthamiana plants was infiltrated with an induced inoculum containing the 35S_PRO:GFP binary vector diluted to OD_600nm = 10⁻⁵ and the transient overexpression vector diluted to OD_600nm = 0.1, with an empty transient overexpression plasmid used as a negative control. These plants were left under normal growing conditions (22°C and 16-h daylengths) for 72 h prior to GFP movement assays (described below).

A 1 mM stock solution of trans-Zeatin (Cayman Chemical Company) in DMSO was diluted to several concentrations (1.0 nM, 10 nM, and 100 nM) in infiltration media, with the infiltration media containing no trans-Zeatin used as a negative control. The 35S_PRO:GFP binary vector was then diluted to OD_600nm = 10⁻⁵ in these solutions and infiltrated into the fourth leaf from 4-week-old N. benthamiana plants, as above. These plants were left under normal growing conditions (22°C and 16-h daylengths) for 48 h prior to GFP movement assays (described below).

Plasmodesmata movement assays were performed using the fourth leaf from either 4-week-old (for trans-Zeatin treatment experiments) or 4-to-5-week-old (for VIGS and overexpression experiments) N. benthamiana plants. GFP movement assays were conducted as previously described (Brunkard et al., 2015a; Figure 1), observing GFP movement in only the proximal 25% of the leaf. Briefly, leaves were infiltrated with very low inocula of Agrobacterium (OD_600nm < 10⁻⁴) carrying a binary vector to transform cells to express GFP under the CaMV 35S promoter. Only a handful of individual, isolated cells are transformed by the low inocula of Agrobacterium. About 48 h after agroinfiltration, the genetically transformed cells show bright GFP fluorescence and are surrounded by “rings” of cells with lower fluorescence, indicating that GFP has moved into these cells via PD. In this study, we report the greatest distance, in numbers of cells, that GFP has spread.

In all experiments, leaves that were previously infiltrated with the 35S_PRO:GFP vector were infiltrated with water immediately prior to harvesting. For VIGS experiments, leaves were harvested from plants infiltrated with media containing the 35S_PRO:GFP vector 48 h after infiltration. For trans-Zeatin experiments, leaves were similarly harvested from plants infiltrated with media containing trans-Zeatin and the 35S_PRO:GFP vector 48 h after infiltration. For overexpression experiments, leaves were harvested from plants infiltrated with media containing the 35S_PRO:GFP vector 72 h after infiltration. For all conditions, small sections were cut from each infiltrated leaf (~1 × 2 cm), mounted abaxial side up on microscope slides, and imaged (as shown in “Microscopy” below).

For each experiment, at least 5, and as many as 24, plants were assayed for each condition per replicate, and each experiment was replicated three times, resulting in each experiment being conducted in at least 70 plants. GFP movement from 1 to 5 randomly selected transformed cells per plant was observed, depending on how many transformed cells were found in the section used for microscopy. The movement was scored by counting the distance in rings of cells to which GFP had moved from the originally transformed cell (e.g., no movement was scored as zero, since GFP remained only in the transformed cell; movement into one or all cells immediately touching the originally transformed cell but none beyond was scored as one; and so on). The total number of cells containing GFP was also counted for each sample.

GFP was observed in the epidermis of N. benthamiana leaves using a Leica DM6 CS confocal laser scanning microscope, with settings as described by Brunkard et al. (2015a). To ensure no artifacts were introduced during microscopy, identical settings (such as laser strength, gain, emission filters, and aperture) were used in all experiments, and samples were imaged in randomized order to avoid any bias during experimentation. All movement assays were scored by both authors to ensure reproducibility.

Movement assay results are presented as the average movement of GFP and SEM. Differences in GFP movement between two given conditions were compared using unpaired heteroscedastic Student’s t-tests in Excel, with p < 0.05 being considered significantly different.

Given the evidence of enhanced cytokinin signaling and increased PD transport in the ise mutants, we next tested whether exogenous application of cytokinin is sufficient to increase PD transport. To this end, we infiltrated leaves of 4-week-old N. benthamiana plants with infiltration media containing a low inoculum of Agrobacterium (OD_600nm < 10⁻⁴) carrying a 35S_PRO:GFP binary vector and a range of concentrations (1.0, 10, or 100 nM) of the cytokinin trans-Zeatin (Letham and Miller, 1965). Infiltration media with no trans-Zeatin was used as a negative control. About 48 h post-infiltration, we excised sections of infiltrated leaves, imaged GFP foci with confocal microscopy, and statistically analyzed the results to quantitatively assess GFP movement. The final results are expressed as the maximal distance that GFP had spread from the transformed cell into neighboring cells.

We found that higher concentrations of trans-Zeatin correspondingly increased the movement of GFP in leaves (Figure 1). On average, GFP moved 1.36 ± 0.12 cells from the transformed cell in the presence of no exogenous trans-Zeatin. With the addition of 100 nM of trans-Zeatin, GFP moved 1.91 ± 0.14 cells from the transformed cell; when compared with results from the control, we found that this difference is statistically significant (n = 53, p < 0.01). Even with smaller concentrations of trans-Zeatin application, GFP movement also apparently increased, though to proportionately smaller degrees (1.73 ± 0.14 cells with 10 nM of trans-Zeatin, 1.59 ± 0.12 cells with 1.0 nM of trans-Zeatin). Overall, these results demonstrate that cytokinins promote PD movement.

Following these results, we took a genetic approach and tested whether silencing the expression of genes directly involved in cytokinin sensing would result in lowered PD movement. Using VIGS, we silenced AHK3 and AHK4, which encode cytokinin receptors that initiate the cytokinin-AHK-AHP-ARR signal transduction pathway in plant cells. We used a TRV2-GUS trigger as a negative control for silencing and a TRV2-NbPDS trigger as a positive control for silencing. After plants were infiltrated with the VIGS constructs, we allowed 2 weeks for silencing to establish before infiltrating leaves on each plant with the same 35S_PRO:GFP construct used in other experiments. Immediately prior to this, we took photographs of each plant to document phenotypic differences among conditions. We observed no obvious morphological or physiological differences between mock silenced and AHK-silenced plants, so we experimented in the same manner as above for the trans-Zeatin experiments.

In both AHK3‐ and AHK4-silenced plants, we observed a significant decrease in GFP movement relative to the GUS mock treatment (Figure 2). Whereas GFP moved 1.61 ± 0.07 cells from the transformed cell in mock-silenced (TRV-GUS) plants, GFP movement in AHK3-silenced plants was reduced to 1.31 ± 0.07 cells and to 1.35 ± 0.06 cells in AHK4-silenced plants. We found that both results were statistically significant (n = 124, p < 0.01; n = 133, p < 0.01). These findings further bolstered the hypothesis that the regulation of PD transport is intimately linked to the cytokinin signaling pathway.

Given the findings regarding the influence of upstream members of the cytokinin signaling pathway on GFP movement, we asked whether overexpressing proteins downstream of AHKs in the cytokinin signaling pathway would alter PD transport. To test this hypothesis, we cloned AHP5 and AHP6 into binary vectors for transient expression in N. benthamiana leaves. Agrobacteria carrying 35S_PRO:AHP5 or 35S_PRO:AHP6 were then co-infiltrated with 35S_PRO:GFP into N. benthamiana leaves.

When compared to control plants infiltrated with empty vector, we found that leaves overexpressing AHP5 exhibited significantly increased levels of PD transport, whereas leaves overexpressing AHP6 exhibited significantly decreased levels of PD transport (Figure 3). In leaves agroinfiltrated with empty vector, GFP moved 1.58 ± 0.07 cells from the transformed cell vs. 2.13 ± 0.07 for plants overexpressing AHP5 and 1.38 ± 0.06 for plants overexpressing AHP6. The change in GFP movement was significant for both AHP5-overexpressing plants (n = 141, p < 0.001) and AHP6-overexpressing plants (n = 161, p < 0.05) relative to control plants. Together, these results indicate that manipulating the expression of different members of the cytokinin signaling pathway phosphorelay chain directly impacts the rate of PD transport.

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.