Overexpression of the pumpkin (Cucurbita maxima) 16 kDa phloem protein CmPP16 increases tolerance to water deficit

Abstract

The phloem plays an important role in the delivery of nutrients and signals between photosynthetic to heterotrophic tissues. Proteins and RNAs in the phloem translocation stream may have an important role in maintaining the integrity of the sieve tube system, as well as in long-distance signaling. CmPP16 is a pumpkin phloem protein, which has been shown to bind RNA in a non-sequence specific manner, and move it cell-to-cell and conceivably, long-distance. The protein and RNA are found in both companion cell (CC) and sieve elements (SE). However, a more precise function for this protein is not known. In this work we report the overexpression of CmPP16 fused to GFP via transformation of pumpkin (Cucurbita maxima cv. Big Max) plants in the cotyledonary stage by direct inoculation of Agrobacterium tumefaciens and Agrobacterium rhizogenes. Plants overexpressing CmPP16 did not show an obvious phenotype. However, these plants displayed higher photosynthetic capacity during drought than wild-type (WT) pumpkin or transformed with another construct. These results suggest that CmPP16 may be involved in the response to stress through long-distance signaling.

Abbreviations

CmPP16

Cucurbita maxima 16 kDa Phloem Protein

FTSW

fraction of transpirable soil water

WHC

water holding capacity

qRT-PCR

quantitative reverse transcription coupled to polymerase chain reaction

Introduction

The phloem is involved in the delivery of fixed carbon and other nutrients from photosynthetic to heterotrophic tissues; it is, however, becoming increasingly clear that it has a central role in long-distance signaling. Indeed, potential signaling molecules such as phytohormones, lipids, and different kinds of RNAs and proteins circulate in the phloem translocation stream, some of which could regulate development and responses to environmental cues.^1-3 Compared with other plants, the amount of sap that can be obtained and thus analyzed from pumpkin and other cucurbits makes them attractive models to study phloem and xylem long-distance signaling. The phloem proteome of pumpkin and other species has been elucidated, and found to contain proteins potentially involved in the maintenance of the sieve tube system, response to pathogens and oxidative stress, as well as in protein turnover.^4-6 Interestingly, transcription factors as well as cyclins were also indentified. Given the enucleate nature of the sieve element (SE), it is tempting to speculate that some of these proteins may have additional functions, i.e., signaling between distant tissues. In general, the precise role of phloem proteins is not clear, except for those that have a maintenance role and a few others. Phloem Protein 1 and 2 (PP1 and 2) are abundant lectins that help seal wounded sieve tubes,^7,8 although PP2 is capable of binding viroid RNA and has thus been suggested to play a role in signaling.^9,10 The 16 kDa phloem protein from pumpkin (CmPP16) is an abundant protein, which is synthesized in the companion cell (CC) and transported into the sieve element (SE).¹¹ Intriguingly, the CmPP16 mRNA is also found in phloem sap exudates, where it is able to move long-distance to the apical tissue of a heterograft; thus, it is possible that this mRNA exits the phloem and moves to the apex following a post-phloem pathway similar to photoassimilates, or other mRNAs.¹¹ It requires the interaction with the Non-Cell Autonomous Phloem Protein 1 (NCAPP1) in order to traffic through plasmodesmata.¹²

The RNA-binding capacity of CmPP16, as well as the presence of its mRNA in the SE suggest a role in long-distance signaling, presumably through the transport of non-cell autonomous RNAs. However, to date its biological role is far from clear. In order to gain insight on the possible function of this protein, pumpkin plants were transformed through injection of Agrobacterium rhizogenes and A. tumefaciens harboring constructs containing the CmPP16 open reading frame fused to GFP. CmPP16-overexpressing plants did not show an obvious phenotype; however, these carried out photosynthesis under water deficit and thus showed increased drought tolerance compared to wild-type plants and those expressing GUS-GFP. A possible mechanism for such increased tolerance is discussed, and may be extensive to other phloem RNA-binding proteins.

Results

CmPP16-GFP and GUS-GFP transcripts are detected in transgenic pumpkin

7–10 day-old pumpkin plants were transformed as described. An insulin syringe was used to apply the bacterial suspension between both cotyledons and in the hypocotyl (Fig. 1A). Three months after inoculation with A. rhizogenes K599, numerous hairy roots were induced that grew within the stem, close to the wounding site. At this stage, hairy roots were capable of rupturing the stem (Fig. 1B). Hairy roots were not present in plants inoculated with A. tumefaciens C58C1. Instead, these plants only presented wound tumors in several cases, 3 months after inoculation (Fig. 1C). In order to determine whether these plants had been transformed with the pCAMBIA 1304 plasmid, PCRs for GFP were carried out with tissue samples. This vector harbors a GUS-GFP translational fusion under the control of the CaMV 35S promoter. Samples were collected 3 months after inoculation from different tissues from control plants and plants inoculated with A. rhizogenes and A. tumefaciens. The sampling sites selected corresponded to stem next to the wound site, as well as from systemic leaves increasingly removed from the inoculation site, and finally from the shoot apex and hairy roots (in the case of A. rhizogenes). GFP DNA was detected in all cases, suggesting that tissues that arose after inoculation (although not necessarily all) carry the transgene (Fig. 1D and E).

Figure 1.

(A) Inoculation of pumpkin epicotyl with an Agrobacterium suspension. (B) Hairy root tumor growing within the lower stem of an A. rhizogenes-inoculated plant. (C) Aspect of a tumor in a plant inoculated with A. tumefaciens. (D) Agarose gel electrophoresis of PCR products from different tissues from control and plants inoculated with A. rhizogenes harboring a GUS-GFP fusion gene under the control of the CaMV 35S promoter using GFP-specific primers. 1, positive control; 2, leaf from untransformed control plant; 3, hairy root tumor growing within the stem; 4, systemic leaf 5 cm from the apex; 5, root 8 cm from the inoculation site; 6; floral bud 1–2 cm from the apical meristem; 7, stem sample 8 cm from the inoculation site. (E) Agarose gel elctrophoresis of PCR products from different tissues from control and plants inoculated with A. tumefaciens using GFP-specific primers. Lanes 1–7, same as in D, except for 3, which is tumor from stem in the vicinity of the wounding site.

In some cases, plants transformed with A. rhizogenes and A. tumefaciens presented a change in leaf morphology (Fig. S1). Indeed, the 2 leaves that arose later after transformation showed in general a more elongated shape, compared to non-transformed controls. Interestingly, leaves that arose at later stages of development showed a normal phenotype. This was independent of the presence of vector, since this same phenomenon was observed in plants inoculated with either A. rhizogenes or A. tumefaciens harboring no vector (not shown).

CmPP16 is an abundant RNA-binding protein found in pumpkin phloem exudates, although its function is not yet clear.¹¹ The open reading frame of this gene was introduced into pumpkin under control of the CaMV 35S promoter and translationally fused to GFP in its C-end in order to express this gene in tissues aditional to phloem. Plants were inoculated with either A. tumefaciens C58C1 or A. rhizogenes K599 harboring this fusion. Analysis of RNA from transformed plants revealed that CmPP16 RNA levels increased considerably in first true leaves 4 weeks after inoculation (Fig. 2A). CmPP16-GFP mRNA levels were also measured using primers directed to GFP, yielding similar results in terms of increase of CmPP16 RNA levels (Fig. 2B).

Figure 2.

Quantification of CmPP16-GFP transcripts in transformed pumpkin plants by qRT-PCR. Total RNA was obtained from first true leaves 4 weeks after transformation. (A) CmPP16 transcript levels in WT and pumpkin plants transformed with A. tumefaciens and A. rhizogenes without plasmid, and A. tumefaciens and A. rhizogenes harboring either GFP-GUS, or CmPP16-GFP. (B) same as in A, except that GFP transcript levels were measured instead.

CmPP16 overexpression increases drought tolerance in pumpkin

Plants that were shown to express the CmPP16 transgene were further analyzed. Potential phenotypic changes were monitored during a 2-month period. However, plants that harbored the transgene were found to be undistinguishable from wild-type controls.

On the other hand, it was observed that plants that were shown to harbor and express the transgene displayed increased tolerance to water deficit. To monitor this phenotype in a more systematic manner, several parameters were measured, namely, total photosynthesis and CO₂ assimilation, transpiration, and stomatal opening and conductance vs. fraction of transpirable soil water (FTSW). Additionally, an equation was calculated to predict the behavior of FTSW relative to days under drought (Fig. S2).

Table 1.

Effect of CmPP16 overexpression in pumpkin. Plants were analyzed 3 to 4 weeks after inoculation with either A. tumefaciens or A. rhizogenes harboring the pumpkin CmPP16 as described before. Transformation was monitored in the tumor or in tissues distal to the inoculation sites by PCR. Drought tolerance was determined by measurement of photosynthesis and transpiration against soil water content and days without watering as described in the text.

Table 1	A. tumefaciens	A. rhizogenes
Total no. of inoculated plants	36	104
Total no. of tumors (crown gall/hairy root)	3	19
Total no. positive by GFP PCR assay	20	80
No. of positives from crown gall or hairy roots	3	19
No. of positives from leaf	5	35
No. of positives from stem	6	15
No. of positives from cotyledon	6	8
No. of positives from root	0	3
Total no. positive for drought-tolerance phenotype	20	97

Five-week old plants were subjected to drought treatment. While WT plants, as well as those expressing GFP-GUS fell beyond the wilting point rapidly after the start of the treatment, those overexpressing CmPP16 survived 2 more weeks; indeed, these showed much higher photosynthesis rates relative to days after treatment and to FTSW, regardless of the Agrobacterium species used for transformation (Fig. 3 and 4). It was observed that at FTSW lower than 50% almost half of the wild type plants showed minimal stomatal opening and thus transpiration did not occur. Plants overexpressing GFP-GUS continue to transpire reaching negative levels (Fig. S3). In contrast, most CmPP16-overexpressing plants did not transpire nor show stomatal conductance at FTSW lower than 50%, indicating that stomata are effectively closed in these conditions (Fig. S3).

Figure 3.

CmPP16-overexpressing plants show enhanced photosynthetic activity relative to wild type controls and those expressing GFP-GUS.Shown here are plants transformed with A. tumefaciens harboring the CmPP16-GFP fusion construct. Photosynthesis (μmoles of assimilated CO₂ m⁻²s⁻¹) against days under drought treatment. FTSW was measured during this period and also plotted.

Figure 4.

Same as in Figure 3, except that the CmPP16-GFP construct was harbored by A. rhizogenes.

Discussion

Cucurbits, and pumpkin in particular, are excellent models for the study of phloem function, structure and composition, since analytical quantities of phloem sap exudates can be obtained, and several pioneering works have helped elucidate the ultrastructure of phloem.^1-3 Recently, a role for this tissue in long-distance signaling has been emerging. Some of these long-distance signals are proteins and RNA. Indeed, certain proteins have been shown to have a role in intercellular and phloem-signaling. For instance, SHR1 regulates the cell fate of neighboring cells in the root cortex through its intercelluar movement, and, on the other hand KNOTTED1 is also transported through plasmodesmata in the shoot apical meristem where it controls stem cell maintenance.^13,14 FT, which has been demonstrated as an important component of the florigenic signal, is transported through the phloem from mature leaves to the shoot apex,^15-17 while in potato its homolog is transported through the phloem from mature leaves to roots where it induces tuberization in response to photoperiod.¹⁸ There is less information regarding the functional role of RNA long-distance transport through the phloem, although it may also play an important role in plant homeostasis and development. Several microRNAs have been implicated in the response to phosphate starvation and tuberization, wherein these are transported via the phloem from photosynthetic leaves to roots; furthermore, the mRNA for the homeodomain-containing BEL5 transcription factor is transported also through the phloem from leaves to roots where it may also regulate tuberization.¹⁹ Finally, some evidence suggests that long-distance delivery of FT mRNA may also have a role in flower induction^20,21

The cucurbit phloem contains several different RNA-binding proteins; the enucleate nature of the angiosperm sieve tube suggests that these proteins are involved in signaling through the phloem, although a clear functional role has yet to be found.²² One of these proteins, CmRBP50, binds a subset of RNAs, indicating a functional specialization for these proteins.²² CmPP16 is a pumpkin protein with homologues in all land plants; a larger and distant version of this protein is harbored in bryophytes and chlorophytes.²³ It is found in the enucleate sieve element, and is able to bind RNA in a non-sequence specific manner and transport it cell-to-cell through plasmodesmata; however, as in the case of CmRBP50 it may bind only a subset of phloem RNAs in vivo.²² Its function in plants is not known, so it was considered that its overexpression in pumpkin could yield some information in this regard. While plants overexpressing CmPP16 directed by the CaMV35S promoter did not show an obvious phenotype, they displayed an increased tolerance to water deficit. Indeed, these plants maintained high photosynthetic activity longer than wild type plants. It is possible that the role of CmPP16 may be related to additional types of stress, and that this could correlate to its phloem RNA-transport capacity. Plants expressing the GFP-GUS fusion protein did not show significant differences in photosynthetic activity compared with wild type plants.

Plants naturally limit water loss via stomatal closure thus reducing transpiration, which may reach zero during severe water deficit; this is the case of wild type pumpkin plants and plants overexpressing CmPP16; but plants overexpressing GFP-GUS continue transpiration until negative levels, i.e., (Fig. S4). This adaptation leads to the accumulation of gases such as CO₂, which decrease photosynhesis, resulting in production of reactive oxygen species.²⁴ Stomatal conductance shows a pattern similar to transpiration (not shown), in that intracellular CO₂ accumulates; thus, these plants have negative levels of transpiration and stomatal conductance. However, WT plants have stomatal conductance levels close to zero at FTSW of less than 50%; plants overeexpressing CmPP16 carry out photosyntesis even without transpiration for short periods of time, in contrast to WT plants. Similar results have been observed in Amaranthus sp.²⁵

Given, therefore, that CmPP16 induces tolerance to a type of abiotic stress i.e., water deficit, it is possible that this occurs through the phloem-mediated transport of stress-related mRNAs. On a speculative note, such mRNAs, synthesized in the CC of photosynthetic leaves, transported by CmPP16 through the SE, and finally translated in heterotrophic tissues, would be directly involved in the response to the stress. It is also possible that this protein mediates the response to more general responses through this mechanism. This could be tested by introducing either a non-translocatable version of CmPP16, or another one unable to bind RNA; indeed, mutant versions of CmPP16 defective in plasmodesmatal trafficking have been obtained.¹²

Conclusions

The overexpression of the pumpkin endogenous CmPP16 gene led to a well-defined phenotype. Importantly, a possible role for CmPP16 was suggested by drought-tolerance of transformed plants; a more general mechanism of action could be proposed for this protein, via phloem-transport of RNAs that mediate the response to stress. On the other hand, phloem RNA-binding proteins may have similar roles in transporting specific RNAs to heterotrophic tissues (the apices in particular) in response to different types of stress.

Materials and Methods

Vectors

The vectors used in this work were pCAMBIA1304, containing a GFP-GUS translational fusion, the expression of which is directed by the CaMV 35S promoter, and harbors a kanamycin resistance gene. This vector was kindly provided by Dr. Federico Sánchez (IBT UNAM) and has been described elsewhere.²⁶ The other was p*7FWG2 Gateway vector, obtained from Plant Genetic Systems, Ghent, Belgium.²⁷

Agrobacterium strains

Cucumopine-type A. rhizogenes strain K599 (NCPP2659)²⁸ and A. tumefaciens C58C1 strain ²⁹ were used in this work. pCAMBIA 1304 was introduced into both strains by electroporation. A. rhizogenes K599 was grown in Luria-Bertani (LB) medium containing kanamycin (Km) at 50 μg/ml, for 48 hours at 30°C for inoculation. A. tumefaciens C58C1 was maintained on YEB medium containing Km at 50 μg/ml, for 48 hours at 30°C.

Plant material

Approximately 60 seeds of C. maxima cv Big Max (R. H. Shumway´s, Randolph, WI) were surface disinfected for 2 minutes in 70% ethanol with vigorous shaking; subsequently, they were immersed in 2% NaOCl and rinsed 4 times with sterile distilled water for 1 minute. The seeds were placed on petri dishes containing cotton. The dishes were kept in darkness at 23°C until the seeds germinated (8–10 days). The plantlets were then transferred to pots, containing cotton. They were grown up for 8 d at 23°C (Fig. 2); subsequently these were inoculated with Agrobacterium. Eight days post inoculation the plantlets were transferred into pots containing 18 Kg of sterilized soil, until the plants were used. The plants were watered with B&D solution ³⁰ and incubated in a growth chamber with a 12 h light/12 h dark photoperiod at 28°C.

Inoculation with Agrobacterium

A. rhizogenes K599 and A. tumefaciens C58C1 grown on plates were collected and diluted with 1 ml of infiltration medium (5%W/V sucrose, 0.044 mM benzylaminopurine, 50 mL/liter Silwett L-77;³¹ which were used to infect cotyledon axes tissues. For control inoculations, a fresh culture of A. rhizogenes K599 and A. tumefaciens C58C1 without plasmid were grown in LB and YEB medium, respectively, without antibiotic. The bacterial strains was subcultured for 48 h and incubated with acetosyringone 140 μM, final concentration.

C. maxima plants grown as mentioned above were then inoculated with A. rhizogenes strain K599 and A. tumefaciens C58C1 containing the binary vector. The plantlets were inoculated by injection directly into the cotyledonary nodes (Fig. 1) with a syringe containing 50 μL of inoculum (Agrobacterium) into the wound. To transform pumpkin plantlets efficiently, the syringe needle had to cross the vascular bundles. Plants were covered with a dome to maintain humidity at 90%, which is necessary for callus formation and the development of hairy roots. The plants were kept in a growth chamber (12 h light/12 h dark a 28°C) for 15 days; subsequently, the dome was removed. Wild type plants were established as a control transformed with A. rhizogenes K599 and A. tumefaciens C58C1 without plasmid. The plants infected by A. rhizogenes and A. tumefaciens started to show tumors approximately one month after inoculation.

PCR analysis and cloning

Genomic DNA from 0.1 g of stems and detached cotyledons, as well as root tissue was extracted with the DNeasy Plant kit according to the manufacturer's recommendations (Qiagen, Hilden, Germany). The presence of T or R-DNA (from A. tumefaciens and A. rhizogenes) in hairy roots and callus was assessed by amplification of the GFP gene from total DNA. As a positive control for GFP, 1 μl (at 50 ng/μl) of pCAMBIA 1304 plasmid was added directly to the PCR reaction. The primers used were, for GFP (which amplifies the complete open reading frame): GFP Forward 5′-ATGGTGAGCAAGGGCGGAGGAGCTG-3′ and GFP Reverse 5′-TTACTTCTACAGCTCGTCCATGCCGAG-3′. PCRs were carried out with 100 ng of plant DNA (50 ng/μl), 5.0 μl of 10× Taq buffer, 4.0 μl of dNTP (2.5 mM), 1.25 unit of Taq DNA polymerase (5 units/μl, TaKaRa Ex Taq HS, Takara Bio, Japan), and 1.0 μl (12.5 mmol) of each primer in a final volume of 50 μl. For genomic DNA amplification, samples were heated to 94°C for 5 min, followed by 30 cycles at 94°C for 30 s, 62°C for 45 s, and 72°C for 1 min with a T-1000 Professional Basic Gradient Thermocycler (Biometra, Germany). Amplified products were separated by electrophoresis on 1% agarose gels, stained with ethidium bromide and fluorescence visualized under UV light.

The CmPP16 ORF was obtained by RT-PCR from phloem sap exudate RNA using the following primers: forward, CmPP16 FOR 5′-ATGGGGATGGGAATGATGGAGGTC-3′; reverse, CmPP16 REV5′-TTTTCCCATGGGTAACATCCTCCACAT-3′, cloned into pCR8/GW/TOPO (Invitrogen) and introduced into the p*7FWG2 Gateway vector as a fusion with GFP.²⁷ The resulting plasmid was transferred to A. tumefaciens C58C1 and A. rhizogenes K599 by electroporation as described.

Real-time PCR and RT-PCR

GFP ORF was amplified from genomic DNA using a GreenER qPCR Super Mix (Invitrogen) following the manufacturer's recommendations, using the conditions described in the previous section.

For transcript levels of the selected genes in phloem sap exudates, quantitative RT-PCR was carried out using the commercial system Express One-Step Syb® GreenER qPCR Super Mix (Invitrogen) according to the manufacturer's instructions. Phloem sap exudates were obtained essentially as described before.³² The following primers were used for actin: CmACT5deg forward: 5′-AAYTTGGAYGAYATGGARAAR-3′ and CmACT3deg reverse: RAANGTYTCRAACATDATYTGNGTCAT. These primers amplify a 200 bp actin fragment, corroborated by sequencing of the cloned fragment. For CmPP16 the following primers were used: CmPP16 forward: 5′-ATGGGGATGGGAATGATGGAGGTCCAT-3′ and CmPP16 reverse: 5′-TAGTTTTCCCATGGGTAACATCCTCC-3′. These amplify the CmPP16 open reading frame (430 bp) without the stop codon. 1 μL of RNA (500 ng/μL) and 50 pM of each primer were used in a 25 μL reaction mix. The Real Time RT-PCR reactions were incubated in a Rotor Gene 3000 apparatus (Corbett Research, Australia). Conditions for amplification were 40 min at 42°C for first strand synthesis, followed by 50 cycles: 40s at 94°C, 45s at 60°C, and 50s at 72°C. RT-PCR reacions were performed by triplicate in each treatment. GFP transcripts were detected in the same manner, using the primers described above. Plasmids were detected using the same methodology. In order to ensure that no dimers were amplified in the reaction, a dissociation curve was generated through progressive heating of the samples (60–95°C), in which only one curve was observed. Assays were carried out for 3 independent samples.

Relative quantification of transcript accumulation was performed according to the method described by Livak and Schmittgen,³³ the 2-ΔCT´ method was used for RNA quantification. This consists of a relative comparison between an endogenous gene and a blank. For this test, 2 repeats were analyzed per treatment. As reported before,³³ it was observed that Ct is exponentially related to copy number, which was important for threshold calculation.

Drought treatment

Two-week seedlings were transferred to 30 cm (height) × 27.5 cm (diameter) pots with a net volume of 18L, 15L of their capacity of which were used. These were filled with loamy sand, containing a mixture of sand/silt/clay in a 60/30/10 percent ratio. The soil was passed through 1-cm screen to remove large aggregates. After that the soil was mixed with 1/3 of their volume of forest litter, resulting in a soil density of 0.95 Kg·L⁻¹. Finally the soil was sterilized at 121°C for 2 hours. The soil had a water content of 30% ± 3% (by volume) at maximum water holding capacity (WHC). A single 2-week old seed was sown per pot.

Plants were maintained for 2 months in greenhouse conditions, under high humidity (70% RH in air) with day/night temperature of 40/15°C and a photoperiod of 12 h light/dark; nutrients were applied with irrigation. The plants were watered until saturation and allowed to drain freely until there was no change in weight. Dry soil weight was determined by baking it at 105°C overnight in an oven. The difference between this weight per liter and wet soil was used to calculate percentage of WHC. The soil was watered until saturation and allowed to dry until constant weight; the weight difference between untreated soil and soil dried at 105°C divided by WHC was calculated to obtain the fraction of transpirable soil water (FTSW). All plants were irrigated daily to maintain 90% WHC. Afterwards two and a half month-old plants were exposed to drought by withdrawing irrigation.

Measurement of Light-saturated photosynthesis and transpiration rate

Ambient CO₂ concentration was 400 mmol per mol of air. Leaf temperature (20°C) and photon flux density incident on the leaf surface (Light-saturated was 2000 μmol m⁻²s¹), CO₂ assimilation rates, stomatal conductance and leaf transpiration were regulated and measured within the chamber of a LI-6400XT infrared gas exchange analyzer (Li-Cor Inc., Lincoln, NE).

Authors’ Contributions

RR-M and BX-C designed the experiments, coordinated the work, carried out plant transformation experiments, and wrote the manuscript; FAR-O, PSH-P and RT-M carried out plant transformation assays, PCR and RT-PCR, and analysis of transformed plants FAR-O and BX-C carried out the drought experiments. All authors read and approved the final version of the manuscript.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Dr. Berenice Calderón-Pérez for critical reading of the manuscript.

Funding

Support for this research was provided by grants from CONACyT No. 156162 to RR-M and CIBIOGEM and CONACyT 105985 to BX-C. FAR-O, PSH-P and RT-M acknowledge doctoral fellowship support from CONACyT.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.