Click on author name to view affiliation information
Poison ivy [Toxicodendron radicans (L.) Kuntz] is a widely recognized native plant species because of its production of urushiol, which is responsible for delayed contact dermatitis symptoms in humans. Poison ivy is predicted to become both more prevalent and more noxious in response to projected patterns of climate change. Future studies on poison ivy chemical ecology will require reverse genetics to investigate urushiol metabolism. A prerequisite for reverse genetic procedures is the introduction and expression of recombinant DNA into poison ivy tissues. Poison ivy leaves and cotyledons were marginally susceptible to vacuum- and syringe-agroinfiltration and expression of two firefly luciferase (LUC)–based reporter genes. The efficacy of agroinfiltration and transient LUC expression was dependent on leaf age and plant growth environmental conditions, with young leaves grown in magenta boxes showing highest transient LUC expression levels. Agroinfiltrated leaves showed an Agrobacterium-dependent accumulation of brown–colored pigments. Biolistic transformation of a LUC reporter gene did not show brown pigment accumulation and readily displayed transient LUC bioluminescence in both leaves and cotyledon tissues. These studies establish best practices for introducing and transiently expressing recombinant DNA into poison ivy leaf and cotyledon tissues, on which future reverse genetic procedures can be developed.
Poison ivy [Toxicodendron radicans (L.) Kuntz] is a native North American species best known for its capacity to induce allergenic dermatitis symptoms after contact with human skin. The natural product responsible for the dreaded skin rashes are alk(en)yl-catechol congeners generically called urushiol, which is found in all members of the Toxicodendron genus (Gross et al., 1975; Hill et al., 1934; Kurtz and Dawson, 1971; Majima, 1922; Markiewitz and Dawson, 1965; Symes and Dawson, 1953, 1954). Urushiol chemical ecology is largely enigmatic because humans appear to be uniquely sensitive to the allergenic effects of urushiol, whereas wild and domesticated animals appear to be largely unaffected on contact with poison ivy. In fact, deer (Martin et al., 1951) and goats (Popay and Field, 1996) eat poison ivy foliage with no apparent ill effects. There are few studies focused on poison ivy physiology and/or ecology. Two studies are noteworthy as they establish that poison ivy responds to elevated atmospheric CO2 levels by growing faster, accumulating more biomass, and shifting urushiol congener levels toward more allergenic forms of urushiol (Mohan et al., 2006; Ziska et al., 2007). Given the current pattern of increasing anthropogenic CO2 emissions, poison ivy is poised to become more prevalent and more noxious, yet most aspects of urushiol metabolism and ecology are largely unexplored. For example, Dewick (1997) postulated that urushiol is derived from fatty acid metabolism beginning with a C16 fatty acid–CoA starter molecule that is extended by a presumed polyketide synthase activity into a tetraketide, that is subsequently modified to form alk(en)yl-catechol (urushiol) congeners. With that said, none of the proposed urushiol metabolites, enzyme activities, enzymes, or genes have been empirically validated to date. DeWick’s hypothesized formation of a fatty acid tetraketide that is cyclized into an alkylphenol by a polyketide synthase activity is likely because a number of plant polyketide synthases use fatty acid–CoA starting molecules in the production of alkylphenols (Abe et al., 2004; Kim et al., 2010, 2013; Matsuzawa et al., 2010; Taura et al., 2016, 2009). The capacity to investigate poison ivy metabolism and chemical ecology was recently advanced with the publication of the poison ivy (T. radicans) leaf and root transcriptome (Weisberg et al., 2017).
The capacity to express recombinant DNA molecules in plants enables detailed investigations of the role of both endogenous and foreign genes. Recombinant genes in plants can be expressed either as transient unintegrated transgenes or as transgenes that are stably integrated into the plant nuclear or plastid genomes. Understanding the molecular basis for urushiol biosynthesis and urushiol chemical ecology will require a variety of molecular genetic methods, all of which require the introduction and expression of recombinant DNA in poison ivy plant cells and tissues. The principle methods of introducing recombinant DNA constructs into plant cells are biolistic and Agrobacterium-based transformation procedures.
Agrobacterium tumefaciens was initially investigated as a soil-borne phytopathogen responsible for crown gall tumor formation. The etiology of pathology requires segments of bacterial tumor-inducing (Ti) plasmid DNA transfer, stable integration, and expression of these bacterial DNA sequences within the plant cell nucleus (Drummond et al., 1977). This inherent plant gene transformation capability of A. tumefaciens was reengineered to replace the tumor formation regions on the Ti plasmid (i.e., disarmed binary Ti plasmids) with recombinant “genes of interest.” The potential utility of Agrobacterium-mediated transient transformation was first demonstrated by expressing high levels of transient GUS reporter gene activity in bean, tobacco, and poplar leaves by simply infiltrating A. tumefaciens harboring a GUS reporter gene on a binary T-DNA plasmid directly into the leaf mesophyll interstitial space (Kapila et al., 1997). The general applicability of the technique is shown by transient reporter gene expression in the leaves of diverse plant species such as switchgrass (VanderGheynst et al., 2008), habanero pepper (Arcos-Ortega et al., 2010), cowpea epicotyl (Bakshi et al., 2011), rice (Andrieu et al., 2012), soybean (King et al., 2015), and persimmon (Mo et al., 2015), to name a few.
Biolistic transformation is often used when Agrobacterium-mediated transformation proves problematic. Biolistic transformation is the acceleration of nucleic acid–coated microparticles to high velocity (using a so-called gene gun) resulting in entry into the plant cell cytoplasm, where the nucleic acid subsequently uncoats from the microparticle, and is either transiently expressed or stably integrates into the plant nuclear genome (Klein et al., 1987). Alternatively, if combined with appropriate plastid DNA sequences the transgene can integrate into, and be expressed from, the plastid genome (Daniell et al., 1990). Biolistic transformation is the method of choice for stable transformation of major cereal crops (Klein et al., 1988; Wang et al., 1988) that are otherwise recalcitrant to Agrobacterium-mediated stable transformation.
Development of reverse genetic methods in poison ivy will enable molecular genetic investigations of poison ivy ecophysiology and chemical ecology, and provide foundational molecular genetic approaches in other members of the Toxicodendron genus (e.g., poison oak and poison sumac). As an initial step toward the implementation of such molecular genetic studies, the present report describes the introduction and transient expression of recombinant firefly LUC reporter gene constructs in poison ivy cells and tissues through agroinfiltration and biolistic transformation.
Poison ivy (T. radicans subsp. radicans) drupes were sourced from the RoaCo-1 liana located in Catawba VA (Benhase and Jelesko, 2013). Drupes were mechanically and chemically scarified (Benhase and Jelesko, 2013). Axenic seedlings were germinated on petri plates of 0.5X Murashige & Skoog (MS) basal salts media (Plant Natural, Bozeman, MT), in the dark for four nights. Germinating seedlings were then transferred to either sterile 0.5X MS media in magenta boxes or pots containing nonsterile Sunshine Mix 1 (Plant Natural). The plants were grown at 28 °C under 16 h light/day cycle. Poison ivy leaves were typically agroinfiltrated with a syringe at the three to four true leaf stage. Nicotiana benthamiana plants were directly germinated and grown in pots containing Sunshine Mix, under the same environmental conditions as the poison ivy seedlings. Nicotiana benthamiana plants were syringe agroinfiltrated 4 weeks postgermination.
Three LUC-containing plasmids were used in this study. Plasmid pJGJ204 contains an Arabidopsis RBCS1B–LUC gene fusion comprised the RBCS1B promoter-exon 3, resulting in an in-frame RBCS1B–LUC chimeric fusion protein (Jelesko et al., 1999).The firefly LUC was polymerase chain reaction (PCR) amplified from pJGJ102 (Jelesko et al., 1999) using oligonucleotide primers (oJGJ234 5′-GGGGAGAAGTTTGTACAAAAAAGCAGGCTATGGAAGACGCCAAAAACATA-3′; and oJGJ235 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATTTTACAATTTGGACTTTC-3′) that incorporated 5′-attB1 and 3′-attB2 sequences. The resulting LUC PCR fragment was BP subcloned into pDONR221 using BP Clonase II (Thermofisher, Waltham, MA) to yield plasmid pJGJ404. The LUC gene from pJGJ404 was subcloned into pMCD32 (Curtis and Grossniklaus, 2003) using LR Clonase II (Invitrogen, Carlsbad, CA) to yield plasmid pJGJ410 expressing the LUC gene from a double CaMV35S promoter. Plasmid pJGJ411 containing a LUC–INT gene was similarly subcloned, except using pLUK07 (Mankin et al., 1997) as the template in the initial PCR reaction. Plasmid p19 (Voinnet et al., 2003) was used for enhancing transient expression levels of the transgenes in plant cells. Agrobacterium tumefaciens strain GV3101 containing either p19 (Voinnet et al., 2003), pJGJ204, pJGJ410, or pJGJ411 were grown overnight (16 h) in 5 mL Luria-Bertani medium supplemented with 50 μg·mL−1 kanamycin and 50 μg·mL−1 gentamycin. The culture optical density (OD) at 600 nm (OD600) was estimated from a 1:10 culture dilution before initial centrifugation (15 min at 4 °C, 3000 × g). Cultures were resuspended in MMA buffer [10 mm 2-N-Morpholino ethanesulfonic acid (MES), 10 mm MgCl2, and 20 μm acetosyringone] by vortexing for either final OD600 of 4.0 or 0.4. The Agrobacterium strains containing a LUC–containing plasmid (pJGJ204, pJGJ410, or pJGJ411) were mixed with GV3101/p19 at equivalent concentration and incubated at 22 °C for 1 h before agroinfiltration (Kapila et al., 1997). Leaf tissues were either vacuum-agroinfiltrated or syringe-agroinfiltrated using a 3 mL Luer Lok syringe (Becton Dickinson, Franklin Lakes, NJ), on the leaf underside (Kapila et al., 1997). Plants were returned to a 28 °C growth chamber for 24 h before assay of LUC activity.
Plasmid pJGJ411 was isolated from Escherichia coli OmniMax strain first by large scale alkaline lysis (Ausubel et al., 2006), then purification using a QIAprep Miniprep kit (Qiagen, Valencia, CA). Purified pJGJ411 plasmid DNA was coated on (0.6 μm) gold particles, accelerated using a 1100 psi rupture disk (Analytical Scientific Instruments US Inc., Richmond, CA) in a Bio-Rad (Richmond, CA) Biolistic PDS-1000 He Particle Delivery System. Poison ivy cotyledon/first true leaf-stage seedlings in magenta boxes were subjected to particle bombardment at a distance of ≈13 cm. After bombardment, plants were returned to a 28 °C growth chamber for 24 h before assay of LUC activity.
Before LUC imaging, transformed poison ivy tissues were placed onto water agar plates and then sprayed with luciferin and subjected to in vivo LUC imaging, as previously described (Jelesko et al., 1999). Qualitative photon emission imaging was obtained from “slice” images. Quantification of vacuum-agroinfiltration and biolistic-transformed poison ivy cotyledon and first true leaf in vivo–LUC activity (photon emission) was estimated using “gravity” images. A fixed area defined by the area that encompassed either the largest cotyledon (6358 pixels2) or largest first true leaf (30,458 pixels2) was used to measure the number of emitted photons for each subsequent cotyledon, leaf, and background. The background heat photon emission was estimated for each specified tissue by randomly measuring the apparent gravity photon emission in 10 regions corresponding to the absence of poison ivy tissue in the reflected light images, and the upper 95% confidence interval value was calculated. The upper 95% confidence interval value established the upper boundary of background heat photon emission levels corresponding to cotyledon and leaves, respectively. The respective upper background photon emission value was subtracted from the value of the respective tissue’s highest photon emission level, and the resulting difference was equally divided into five bins, thereby providing a rough distribution of photon emission levels for each transformed tissue.
Poison ivy leaves were subjected to syringe-agroinfiltration using A. tumefaciens strain GV3101 harboring plasmid pJGJ411 containing a double CaMV promoter driving the expression of firefly LUC gene containing an artificial intron (LUC–INT). The intron in the LUC–INT gene abolished low expression levels in Agrobacterium cells relative to the continuous LUC open reading frame in pJGJ410 (Supplemental Fig. 1), and thus all LUC activity was derived from transient LUC expression in poison ivy cells. Poison ivy leaves from potting soil–grown plants subjected to syringe-agroinfiltration generally did not display LUC activity greater than background heat photon emission levels in qualitative pseudocolor superimposed slice photon emission images (Fig. 1A). On very few occasions weak photon emission more than background levels of syringe-agroinfiltrated leaves was observed. In contrast, syringe-Aagroinfiltrated leaves on poison ivy plants grown axenically in magenta boxes displayed small patches of consistent photon emission over regions that were syringe-agroinfiltrated (Fig. 1A). Thus, magenta box–grown poison ivy leaves were more susceptible to syringe-agroinfiltration–mediated transient DNA transformation. However, not all leaves from Magenta box grown poison ivy plants were equally susceptible to syringe-agroinfiltration. Older leaves of magenta box–grown plants were less susceptible to syringe-agroinfiltration transient LUC–INT expression compared with younger leaves (Fig. 1B). Younger leaves more readily took up the Agrobacterium-infiltration solution delivered from the blunt syringe. Consequently, young poison ivy leaves in magenta box–grown plants consistently showed higher transient LUC expression levels than older leaves from the same plants, whereas changing the bacterial concentration 10-fold had little effect on transient LUC expression levels (Supplemental Fig. 4). With that said, even the best poison ivy transient LUC expression levels were dramatically lower than N. benthamiana leaves syringe-agroinfiltrated with pJGJ411 (Fig. 1C). The poison ivy leaves did not take up as much of the Agrobacterium solution as the N. benthamiana leaves with each syringe-agroinfiltration attempt. However, these differences in relative infiltration volumes were far less than the difference in magnitude of photon emission between poison ivy and N. benthamiana (Fig. 1A and C).
Citation: HortScience horts 53, 2; 10.21273/HORTSCI12421-17
Syringe-agroinfiltrated poison ivy leaves displayed brown discoloration within the regions that were effectively infiltrated with bacteria (Fig. 1D). This brown discoloration was dependent on the presence of the Agrobacterium, because syringe infiltration with just the MMA media lacking A. tumefaciens did not acquire the brown pigmentation. When plants are infiltrated with bacteria, many plant species induce the synthesis and accumulation of phytoalexin compounds as part of a plant defense response (Hammerschmidt, 1999). Thus, the overall low transient LUC expression levels may have been a consequence of reduced Agrobacterium vigor in response to a biotic stress response, however this possibility was not further investigated in this study.
Vacuum-agroinfiltration was investigated as an alternative method for transient heterologous plant gene expression in poison ivy. An Arabidopsis chimeric AtRBCS1B–LUC chimeric gene fusion (pJGJ204 comprised the AtRBCS1B promoter, exons I, intron I, exon II, intron II, and partial exon III fused in-frame to the firefly LUC open reading frame), was used to vacuum-agroinfiltrate excised poison ivy leaves and cotyledons from plants grown in magenta boxes. Vacuum-agroinfiltrated leaves were placed on 0.5 × MS media plates, sprayed with luciferin at 24 h postinfiltration, and then imaged for three sequential 1 h-imaging sessions beginning at ≈48 h post agroinfiltration. The AtRBCS1B–LUC chimeric gene produces a low rate of photon emission in transgenic Arabidopsis plants. Superimposed “slice” images with complete subtraction of background photons displayed a few blue spots over leaves/cotyledons (Supplemental Fig. 2A), whereas superimposed “slice” imaging that retained some background photons displayed higher qualitative photon accumulation over leaves vacuum-agroinfiltrated with the plasmid containing the AtRBCS1B–LUC chimeric gene, relative to vacuum-agroinfiltrations with the vector control plasmid pSLK7292 (Supplemental Fig. 2B) indicating low but demonstrable AtRBCS1B–LUC recombinant protein expression over background heat photons. Photon emission levels were significantly higher in pJGJ204 (AtRBCS1B–LUC) vacuum-agroinfiltrated leaves and cotyledons compared with pSLK7292 vacuum-agroinfiltrated leaves and cotyledons (P value < 0.05), during the first two of three sequential 1 h-imaging sessions quantified using “gravity” imaging (Fig. 2). There was a consistent trend of gradually lower total photon emission levels over the course of three hours suggesting that luciferin substrate levels were a limiting factor during the imaging session. The vacuum-agroinfiltrated leaves and cotyledons also showed browning of infiltrated tissues similar to the syringe-agroinfiltrated leaves. This was particularly apparent in cases where the cotyledon or leaf segments were not fully vacuum-agroinfiltrated, and showed a lighter green color (see arrows in Supplemental Fig. 2C) typical of uninfiltrated leaves. These results demonstrate that a heterologous Arabidopsis genomic RBCS1B promoter and intron containing RBCS1B–LUC gene fusion was expressed in poison ivy leaves and cotyledons.
Citation: HortScience horts 53, 2; 10.21273/HORTSCI12421-17
The induction of an apparent biotic stress response (i.e., accumulation of brown pigmentation) in agroinfiltrated poison ivy leaf tissue might limit the efficiency of agroinfiltration transient DNA transformation. To avoid such a possibility, biolistic transformation of isolated pJGJ411 (LUC–INT) plasmid DNA directly into poison ivy leaves and cotyledons was investigated. Using superimposed slice photon imaging, punctuate foci of photon emission were observed on both cotyledons and leaves (Supplemental Fig. 3). The distribution of LUC expression was not uniform, both across tissue type and between replicated tissues on the same targeting field. Likewise, quantification of gravity images demonstrated that the amount of photon emission from any given leaf or cotyledon was variable, ranging from background heat photons to quite intense photon emission (Fig. 3). This variability is likely a consequence of the uneven and unpredictable spreading of plasmid-coated gold particles during particle acceleration. Unlike all poison ivy tissues transformed by agroinfiltration, the biolistic transformed poison ivy cotyledons and leaf tissues did not produce demonstrable brown pigmentation.
Citation: HortScience horts 53, 2; 10.21273/HORTSCI12421-17
These studies lay the foundation for future transgenic reverse genetic approaches to investigate various poison ivy physiologies by demonstrating that exogenous DNA can be effectively introduced and expressed in poison ivy cotyledon and leaf tissues. Because the CaMV35S promoter is weakly active in A. tumefaciens (Supplemental Fig. 1; Mankin et al., 1997) it was essential to use intron-containing firefly LUC gene constructs (CaMV35S-LUC-INT and AtRBCS1B-LUC) to validate poison ivy–specific transient LUC activity. This was particularly germane in the case of the AtRBCS1B–LUC construct used in the vacuum-agroinfiltrated leaves and cotyledons because this reporter gene construct is expressed at relatively low levels in plant cells (Jelesko et al., 1999, 2004). The two introns in the AtRBCS1B–LUC gene ensured that all LUC activity was due to expression in transiently transformed poison ivy cells and not within the A. tumefaciens strain GV3101.
The age of poison ivy leaves was an important factor for the relative susceptibility of syringe agroinfiltration transient transformation. Leaves from potting soil–grown poison ivy plants were mostly recalcitrant to syringe-agroinfiltration transient transformation. On the other hand, young poison ivy leaves from plants grown in magenta boxes showed consistent syringe-agroinfiltration transient LUC–INT expression. Moreover, overall plant age was not the critical determinant, but rather the relative age of target leaves. As shown in Fig. 1B, the fifth emerged leaf (young leaf) showed greater syringe-agroinfiltration transient LUC expression levels than that of the second emerged leaf (older leaf) on the same plant. The qualitative levels of LUC–INT expression levels in poison ivy leaves were dramatically lower than the levels observed in N. benthamiana leaves syringe-agroinfiltrated with the same reporter gene construct. Poison ivy cotyledons and leaves were also susceptible to vacuum-agroinfiltration transient transformation using a chimeric AtRBCS1B–LUC transgene. Leaves infiltrated with A. tumefaciens strain GV3101 displayed discoloration consistent with an inducible plant biotic stress response. Both syringe- and vacuum-agroinfiltrated leaves demonstrated the accumulation of uncharacterized brown pigments (Fig. 1D; Supplemental Fig. 2C). This pigment accumulation was not observed for leaves infiltrated with MMA buffer lacking Agrobacteria (Fig. 1D). The relative LUC expression levels in the brightest poison ivy leaves was dramatically less that that observed in control N. benthamiana leaves syringe agroinfiltrated with the same construct. This difference could have been due to a variety of parameters including the quantity of Agrobacteria penetrating into the apoplastic space of leaves, and/or a biotic stress response that inhibited bacterial vigor or T-DNA transfer into the plant cells. Nevertheless, poison ivy leaves and cotyledons showed significant transient LUC–INT and AtRBCS1B–LUC expression levels over controls indicating the feasibility of using agroinfiltration transient poison ivy transformation.
As expected, poison ivy cotyledons and leaves were readily transiently transformed using a biolistic method. Biolistically transformed poison ivy leaves and cotyledons did not produce brown pigmented regions and thus averted a poison ivy biotic stress response against bacterial pathogen associated molecular patterns. As is typical of biolistic transformation of plant tissues, the transient LUC–INT expression levels were highly variable among both leaf and cotyledon tissues. This is in large part to previously recognized inconsistencies in both plasmid DNA coating of gold micro particles (Sanford et al., 1993) and heterogenous particle spread on the macro carrier disk, resulting in heterogeneous distribution during particle acceleration toward the target tissue.
These results demonstrate the feasibility of poison ivy leaf and cotyledon transient transformation with recombinant reporter gene constructs using either agroinfiltration or biolistic methods. Both poison ivy DNA transformation methods are important technical advancements enabling the introduction of recombinant DNA constructs designed for a variety of reverse genetic (e.g., RNA interference or viral induced gene silencing) and genome-editing methods (e.g., zinc finger, TALEN, or CRISPR-CAS9) to enable molecular genetic investigations of poison ivy ecophysiology and metabolism.
Contributor Notes
This work was supported by funds from the Virginia Agricultural Experiment Station and the USDA National Institute of Food and Agriculture, U.S. Department of Agriculture (Washington, DC).
C.C.D. performed the syringe-agroinfiltration, biolistic transformation, and luciferase-imaging experiment data analysis, and writing and editing of the manuscript. A.J.W. performed the vacuum-agroinfiltration, luciferase imaging, and editing of manuscript. J.G.J. conceived and supervised the experiments, performed data analyses, and wrote the manuscript.
Corresponding author. E-mail: jelesko@vt.edu.
