Breakthrough Technologies

Stable Transformation of Ferns Using Spores as Targets:Pteris vittata and Ceratopteris thalictroides1[W][OPEN]

Balasubramaniam Muthukumar, Blake L. Joyce, Mark P. Elless, and C. Neal Stewart Jr.*

University of Tennessee, Knoxville, Tennessee 37996 (B.M., B.L.J., C.N.S.); and Edenspace SystemsCorporation, Manhattan, Kansas 66502 (M.P.E.)

ORCID ID: 0000-0002-8220-2823 (B.M.).

Ferns (Pteridophyta) are very important members of the plant kingdom that lag behind other taxa with regards to our understandingof their genetics, genomics, andmolecular biology. We report here, to our knowledge, the first instance of stable transformation of fernwith recovery of transgenic sporophytes. Spores of the arsenic hyperaccumulating fern Pteris vittata and tetraploid ‘C-fern Express’(Ceratopteris thalictroides) were stably transformed by Agrobacterium tumefacienswith constructs containing the P. vittata actin promoterdriving a GUSPlus reporter gene. Reporter gene expression assays were performed on multiple tissues and growth stages ofgametophytes and sporophytes. Southern-blot analysis confirmed stable transgene integration in recovered sporophytes and alsoconfirmed that no plasmid from A. tumefacienswas present in the sporophyte tissues. We recovered seven independent transformantsof P. vittata and four independent C. thalictroides transgenics. Inheritance analyses using b-glucuronidase (GUS) histochemical stainingrevealed that the GUS transgene was stably expressed in second generation C. thalictroides sporophytic tissues. In an independentexperiment, the gusA gene that was driven by the 23 Cauliflower mosaic virus 35S promoter was bombarded into P. vittata spores usingbiolistics, in which putatively stable transgenic gametophytes were recovered. Transformation procedures required no tissue cultureor selectable marker genes. However, we did attempt to use hygromycin selection, which was ineffective for recovering transgenicferns. This simple stable transformation method should help facilitate functional genomics studies in ferns.

Ferns (Pteridophyta) are nonflowering vascular plantscomprised of 250 genera, the second largest group ofdiversified species in the plant kingdom (Gifford andFoster, 1989). Many interesting traits are inherent invarious fern species, such as arsenic hyperaccumulation(Pteris vittata; Ma et al., 2001), insecticide productionand allelopathy (Pteridium aquilinum; Marrs and Watt,2006), and antimicrobial compound production (Acros-tichum aureum; Lai et al., 2009). Ferns occupy theevolutionary niche between nonvascular land plants,bryophytes, and higher vascular plants such as gym-nosperms and angiosperms. Therefore, extant fern spe-cies still hold a living record of the initial adaptationsrequired for plants to thrive on land. Of these adapta-tions, the most important is tracheids that comprisexylem tissues for water and mineral transport andstructural support. The vascular system allowed pte-ridophytes to grow upright during the sporophytegeneration, leading to greater resource acquisition ca-pacity. Ultimately, sufficient resources allowed for

greater spore production and upright growth, whichfacilitated spore spread. Recent endeavors have inves-tigated lower plant genomics, including the sequencingof the bryophyte Physcomitrella patens (Rensing et al.,2008) and the lycophyte Selaginella moellendorffii (Bankset al., 2011). Both basic and applied biology of ferns lagfar behind that for angiosperms and even bryophytes.

Stable genetic transformation has been accomplishedin a few species outside angiosperms and gymno-sperms, especially among bryophytes (Schaefer et al.,1991; Chiyoda et al., 2008; Ishizaki et al., 2008), butnever in the Pteridophyta. Transient transformationmethods have been developed both in Ceratopterisrichardii ‘C-fern’ and P. vittata, which were limited toheterologous expression in gametophytes (Rutherfordet al., 2004; Indriolo et al., 2010). While transient ex-pression of transgenic constructs does enable research,there is no substitute for stable transformation in func-tional genomics and plant biology. Researchers haveresorted to studying fern gene function using heterol-ogous expression in the angiosperm model plantArabidopsis (Arabidopsis thaliana; Dhankher et al., 2006;Sundaram et al., 2009; Sundaram and Rathinasabapathi,2010), which is far from optimal. Overexpression andknockdown analysis of individual genes in a wide va-riety of fern species would undoubtedly accelerate ourability to learn more about their biology and subse-quently develop novel products from ferns. Further-more, a facile transformation system would acceleratefunctional genomics and systems biology of ferns. Forexample, knockdown analysis of genes involved ininteresting biosynthetic pathways can greatly facilitategene and biochemical discovery.

1 This work was supported by Edenspace Systems Corporationthrough a National Institute of Health award (R42ES014976), the Uni-versity of Tennessee Ivan Racheff Endowment, and the U.S. Depart-ment of Agriculture-National Institute of Food and Agriculture.

* Address correspondence to nealstewart@utk.edu.The author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policydescribed in the Instructions for Authors (www.plantphysiol.org)is: C. Neal Stewart, Jr. (nealstewart@utk.edu).

[W] The online version of this article contains Web-only data.[OPEN] Articles can be viewed online without a subscription.www.plantphysiol.org/cgi/doi/10.1104/pp.113.224675

648 Plant Physiology�, October 2013, Vol. 163, pp. 648–658, www.plantphysiol.org � 2013 American Society of Plant Biologists. All Rights Reserved. www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from

Copyright © 2013 American Society of Plant Biologists. All rights reserved.

P. vittata has the unparalleled ability to accumulatemore arsenic per gram biomass than any other plantspecies and is highly tolerant to arsenic (Gumaeliuset al., 2004). It can thrive in soils containing up to1,500 mg mL–1 arsenic, whereas most plants cannotsurvive 50 mg mL–1 arsenic (Ma et al., 2001). Therefore,P. vittata has been the subject of extensive basic andapplied research for arsenic hyperaccumulation, trans-location, and resistance (Gumaelius et al., 2004) and hasbeen used for arsenic phytoremediation (Shelmerdineet al., 2009). For example, this fern might be of greatutility for the production of a safer rice (Oryza sativa)crop; arsenic can be transported and stored in thegrain, resulting in serious human health ramifications(Srivastava et al., 2012). In a recent greenhouse study,P. vittata has been used to remediate arsenic-contaminatedsoil. Following remediation, rice plants were subse-quently grown, and it was found that arsenic uptakeby rice grains was reduced by 52%, resulting in lessthan 1 mg mL–1 arsenic after two rounds of remediationusing P. vittata phytoremediation (Mandal et al., 2012).Furthermore, this treatment also resulted in increasedrice grain yield by 14% (w/w) compared with control.Ceratopteris is a subtropical-to-tropical fern genus

containing four to six species living in aquatic habitats.The C-fern cultivar was developed as a model fern forteaching (http://www.c-fern.org) and research owingto its small size, short life cycle (120 d), and its ame-nability for in vitro culture (Banks, 1999). The C-fernExpress cultivar was developed by Leslie G. Hickok bycrossing two Japanese varieties of Ceratopteris thalic-troides (L. Hickok, personal communication). cv C-fern

Express, a tetraploid, develops spores in 60 d of cul-ture. Though cv C-fern spores have been shown to be auseful single cell model system and a rapid and effi-cient system for studying RNA interference in ferns(Stout et al., 2003), no stable transformation studieshave been reported. In bryophytes, immature thalli aremost often used as explants (Chiyoda et al., 2008;Ishizaki et al., 2008), whereas in fungi, spores are rou-tinely used for stable transformation studies (Michielseet al., 2005; Utermark and Karlovsky, 2008).

The objective of our research was to develop, for thefirst time, a facile stable transformation system forP. vittata and C. thalictroides using spores as the trans-formation targets. This method can be used as an ad-ditional tool to further substantiate and strengthen themolecular mechanism studies in pteridophytes.

RESULTS AND DISCUSSION

Isolation of a Putative Fern Actin Gene and Its Promoter

An important missing component of fern molecularbiology is effective promoters that can be used fortransgene expression. Actin genes are likely candidatesto have constitutive promoters. Thus, we identifiedP. vittata actin genes using degenerate primers designedfrom algae, ferns, and angiosperms, which resulted in a1.4-kb PCR fragment. BlastN and tblastX sequence anal-yses confirmed that the isolated PCR product fromP. vittata was an actin gene. A neighbor-joining treewas constructed based on Blast searches that revealed

Figure 1. In vitro generation of non-transgenic control and transgenicsporophytes of P. vittata (A–H) andC. thalictroides (I–P). Both A. tumefaciens-mediated transformed spores and non-transgenic control spores were germinatedon one-half-strength MS medium.The nontransgenic prothalli (A and I) andtransgenic prothalli (B and J) were trans-ferred to the same medium, where theyformed gametophytes. The nontransgenicgametophytes (C and K) and transgenicgametophytes (D and L) were transferredto one-half-strength MS medium withoutSuc and generated nontransgenic sporo-phytes (E and M) and transgenic sporo-phytes (FandN), whichwere subsequentlytransferred and established in pots (G andO, nontransgenic; H and P, transgenic).Bar = 3 mm.

Plant Physiol. Vol. 163, 2013 649

Stable Transformation of Ferns

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

that the closest neighbors to fern actin were P. patensactin1 and Chlamydomonas reinhardtii actin (SupplementalFig. S1).

Once the actin gene structure was confirmed, weperformed genome walking upstream to isolate theactin promoter. During the first genome walk using aPvuII library, a 720-bp PCR product was obtained.Sequencing results confirmed the isolation of 520-bpputative promoter region located 59 upstream of 200-bpactin gene template. A second genome walk was per-formed using the 520-bp and a new 369-bp upstreamregion that was obtained from an EcoRV library. An889-bp putative promoter sequence was sequence con-firmed and assembled and further used to drive theexpression of the GUSPlus gene. Motif analyses usingPlantCARE and PLACE regulatory element databasesshowed that the actin promoter contained several char-acterized higher plant promoter motifs (SupplementalTable S1).

In Vitro Generation of Transgenic P. vittataand C. thalictroides Sporophytes

In our initial experiments, no antibiotic selection wasused throughout the transformation studies for bothferns. In parallel, nontransgenic control spores werealso germinated and compared toAgrobacterium tumefaciens-transformed spores (Fig. 1). Incubation of suspendedspores in carboxy methyl cellulose for 15 h at room tem-perature before transformation increased the synchroni-zation of growth as well as spore germination by 40% inP. vittata and 20% in C. thalictroides. GUS staining wasused to identify transformants at various life cycle stages.In P. vittata, mature prothalli (more than 90%) began toform 6 weeks after spore transformation in one-half-strength Murashige and Skoog (MS) medium with20% (w/v) Suc in both transgenic and nontransgenic controlsamples (Fig. 1, A and B). After an additional 6 weeks,approximately 90% of prothalli formed gametophytesin the same medium (Fig. 1, C and D). The 3-week-oldgametophytes that were transferred to the same basalmedium without Suc started to produce sporophytesafter 5 weeks from both nontransgenic control andtransgenic gametophytes (Fig. 1, E and F). Approxi-mately 85% of the gametophytes gave rise to sporo-phytes. Efficiency calculations were based on visualestimation for three different transformation experiments.The sporophytes were established in potting media after3 weeks (Fig. 1, G and H). Approximately 60% of thetransferred sporophytes were established in soil. Thetotal time to obtain mature sporophytes from sporeswas 23 to 27 weeks. The efficiency of sporophyte for-mation from spores was approximately 55% to 65%.

In C. thalictroides, 95% of prothalli were formed in3 weeks both in control and transgenic samples (Fig. 1,I and J), which gave rise to 80% efficiency of gameto-phyte production at 4 weeks posttransformation inboth samples (Fig. 1, K and L). Seventy percent ofgametophytes produced sporophytes 3 to 4 weeks

after gametophytes were transferred to Suc-free one-half-strength basal medium (Fig. 1, M and N). About60% of the transferred sporophytes survived to growin potting mix (Fig. 1, O and P). Mature C. thalictroidessporophytes were obtained after 11 to 13 weeks of cul-ture from spores with 50% to 60% efficiency. All thecalculations were based on three independent experi-ments. The established sporophytes started sporulatingafter 8 weeks in pots (Fig. 1, O and P).

Spores from both species germinated normally inone-half-strength MS medium with 20% Suc and con-tinued to grow and produce gametophytes both innontransgenic control and transgenic samples. Previ-ous studies have suggested that Suc is necessary forgermination and early growth development of ferns(Camloh, 1993; Sheffield et al., 2001), yet the influenceof Suc on the later stages of fern development wasunknown. We found that Suc was not necessary forsporophyte development from gametophytes; how-ever, the critical effect of Suc on growth stages was not

Figure 2. GUS histochemical staining of P. vittata T0 prothalli andgametophytes. Six-week-old prothalli (B) and gametophytes (D) de-rived from transgenic spores obtained from A. tumefaciens-mediatedtransformation. A and C are nontransgenic control prothalli and ga-metophytes, respectively. E, Protonemata of P. vittata showing GUSexpression 1 week after bombardment of spores with a 2335S cauli-flower mosaic virus promoter-gusA gene construct. F, GUS expressionin gametophytes derived from spores of P. vittata after bombardment.Bar = 0.25 mm.

650 Plant Physiol. Vol. 163, 2013

Muthukumar et al.

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

studied. As expected, C. thalictroides was able to gen-erate sporophytes much quicker than the slowergrowing P. vittata (Gumaelius et al., 2004), but none-theless, generation of sporophytes was possible in bothof these hermaphroditic ferns (Schedlbauer, 1976;Gumaelius et al., 2004) using same media and similargrowth conditions. Samples that were subjected toA. tumefaciens appeared to lag a few days behind thecontrols, but there was no other phenotypic differencesobserved between nontransgenic control and trans-genic ferns.

Hygromycin as a Potential Selection Agent for Ferns

Hygromycin at 3 and 15 mg L–1 were the lowestconcentrations that severely inhibited growth of P. vittataand C. thalictroides spores, respectively (SupplementalFig. S2). Thus, these concentrations were used to attemptto select transformants for these species in separatetransformation experiments.After cocultivation with A. tumefaciens for 90 h,

P. vittata spores were immediately transferred to one-half-strength MS medium with and without hygromycin.Apparent normal growth was observed in control plates(0 mg L–1 hygromycin and 400 mg L–1 timentin) contain-ing transformed and nontransformed spores. Whereas onthe hygromycin-containing plates, none of the sporesgerminated at 4 weeks. In addition, GUS expressionwas observed in gametophytes obtained from trans-formed control plates. The same result was obtained forC. thalictroides.In another set of experiments, both ferns were cultured

on hygromycin-containing plates following 2 weeks ofincubation on plates without hygromycin after coculti-vation. The spores germinated and formed prothalli, butthen no further growth was observed after transfer tohygromycin-containing plates, wherein the plants sub-sequently died.Thus, we conclude that more experiments are needed

to determine whether hygromycin selection is feasible.Selection regimes can be fine-tuned and/or a promoterstronger than nopaline synthase (nos) is likely required.Because the nos promoter is not a strong constitutivepromoter, even in angiosperms (Christensen and Quail,

1996; Wilkinson et al., 1997; Bhattacharyya et al., 2012),the next likely step would be to replace it with a strongerpromoter in the selectable marker cassette. We noted thatyoung prothalli are sensitive to hygromycin; loweramounts of hygromycin might be required if antibioticselection is desired.

Stable GUS Expression in Ferns Transformed byA. tumefaciens and Particle Bombardment

Because spores were used as the transformationtargets, the resulting sporophytes that arise from her-maphrodite gametophytes after fertilization of spermand eggs are referred to as T1 transformants for bothspecies. GUS expression was observed in all life stagesof both fern species: prothalli, gametophytes, andsporophytes. In nontransgenic control prothalli, noGUS expression was observed (Fig. 2A), but expressionwas detected as early as 6 weeks posttransformation intransgenic prothalli of P. vittata (Fig. 2B). When trans-genic P. vittata prothalli further developed into gameto-phytes, GUS expression was observed in most of thegametophyte tissues (Fig. 2D), which was not observedin nontransgenic control gametophytes (Fig. 2C).

Particle bombardment of fresh spores as well as 5-day-old germinating spores yielded fully transformed younggametophytes with GUS expression in rhizoids andprothalli (Fig. 2E). When 15-day-old multicellular pro-tonemata were transformed, the result was apparentchimeric expression of GUS in gametophytes (Fig. 2F).GUS expression driven by the cauliflower mosaicvirus 2335S promoter was observed at an average of12 6 10 out of 100,000 spores, which corresponds toa mean transformation efficiency of 0.012% usingbiolistics.

The choice of targets for transformation plays acrucial role in successful stable transformation byA. tumefaciens (de la Riva et al., 1998; Karami et al.,2009) and by biolistics (Kikkert et al., 2004; Jones andSparks, 2009). When spores were transformed usingbombardment, evenly GUS-stained gametophytes wereobserved, demonstrating that targeting spores fortransformation avoids the potential for chimeras. It also

Figure 3. GUS histochemical staining of transgenic and control T1 P. vittata, T1 C. thalictroides gametophytes, and T2 spo-rophytic fronds. Controls are labeled as B, C, and F. Mature sporophytes were generated from transgenic spores of P. vittata (A).T1 gametophytes (D) and T2 sporophytes (E) of C. thalictroides were generated from transgenic T1 spores obtained from T1sporophytes. Bar = 3 mm.

Plant Physiol. Vol. 163, 2013 651

Stable Transformation of Ferns

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

negates the need for tissue culture. Until this study, onlyfern gametophytes have been targeted for transienttransformation (Rutherford et al., 2004; Indriolo et al.,2010). From our results, it is possible that choosing sporesas transformation targets probably played a crucial rolein successful transformation of these two fern species,which is the same case in fungi (Michielse et al., 2005;Utermark and Karlovsky, 2008), where spores havebeen used as targets. Moreover, A. tumefaciens growthmedium (Utermark and Karlovsky, 2008) coupled withthe phenolic compound acetosyringone, a known in-ducer of virulence genes (Stachel et al., 1986), has beensuccessfully used in experiments for fungal spore trans-formation as well.

Overall, there was no visible difference in the GUSexpression pattern in the first generation sporophytesof both ferns. In sporophytes, no endogenous blue his-tochemical GUS staining was observed in nontransgeniccontrol tissues that were treated with or without carbo-rundum, in contrast to the confirmed transgenic plants(Fig. 3B). In P. vittata sporophytes, blue GUS stainingwas observed only in carborundum-treated maturesporophyte fronds (Fig. 3A). Carborundum treatmentwas not required for effective GUS staining of trans-genic C. thalictroides sporophytes; their fronds GUSstained evenly (Supplemental Fig. S3). One version ofthe GUSPlus marker gene we used contains a ricesignal peptide that facilitates GUS expression inapoplasts. C. thalictroides was transformed with GUSPlusgene with no signal peptide. Nonetheless, a similar pat-tern of GUS expression was observed in fronds of bothspecies.

In P. vittata, GUS expression was visible only aftercarborundum 600-mesh size treatments, possibly be-cause random carborundum perforations helped the

GUS staining solution enter into the otherwise impen-etrable frond tissues. Other mesh sizes, i.e. 300 and1,200, were not effective in GUS assays. The tapesandwich method (Wu et al., 2009), which was used topeel off epidermal layers in an Arabidopsis protoplastisolation method, also yielded effective GUS stainingin P. vittata fronds, but resulted in tearing of fronds,whereas the fronds perforated by carborundum wereintact. Although these fronds were intact after carbo-rundum perforation, GUS staining was observed onlyin those areas where carborundum perforation re-moved the epidermal layers. Carborundum is routinelyused to aid screening of plant susceptibility for patho-gens (Thaler et al., 2004); however, to the best of ourknowledge, this is the first time carborundum was usedto aid in GUS assays.

GUS Expression Analysis in T1 Gametophytesand T2 Sporophytes of C. thalictroides

The gametophytes along with the T2 sporophytesthat developed from T1 spores from two independenttransgenic lines 1 and 7 were assayed using histo-chemical GUS staining. Out of 15 gametophytes screenedfrom each line, 13 and 14 of them were positive in eachline. GUS expression was not observed in nontransgeniccontrol gametophytes (Fig. 3C), whereas in transgenicgametophytes, GUS expression was seen throughoutthe entire gametophyte tissue (Fig. 3D). Seven sporo-phytes from transgenic lines 1 and 7 were also screenedfor GUS expression. In line 1, five sporophytes werepositive for GUS expression, and in line 7, four spo-rophytes were positive for GUS expression. GUS ex-pression was seen in almost all parts of the frond tissue

Table I. Transformation efficiency of P. vittata and C. thalictroides by A. tumefaciens-mediated transformation using R4pGWB501 containingP. vittata actin promoter driving GUSPlus

Rows indicate the results of separate experiments. Transformation efficiency was calculated based on gametophytes that were GUS positives basedon three independent transformation experiments. Prothalli and gametophytes were used to calculate the efficiency because more than 90% ofspores formed prothalli. Sporophytes were not used to calculate efficiency because not all sporophytes were recovered for further growth andanalysis. ANOVA was performed to calculate the average and error rate for transformation.

Fern species Spores TransformationProthalli Analyzed/

GUS Positives

Transformation

Efficiency Using Prothalli

in the Numerator

Gametophytes Analyzed/

GUS Positives

Transformation Efficiency

Using Gametophytes

in the Numerator

no. % %

P. vittata 400 No, control 380 Not applicable 370 Not applicable300 No, control 290 Not applicable 280 Not applicable200 No, control 185 Not applicable 180 Not applicable

50,000 Yes 3/10,000 0.03 17/30,000 0.05640,000 Yes 9/15,000 0.06 8/17,000 0.04750,000 Yes 12/17,000 0.07 4/12,000 0.0333

Mean,0.053 6 0.021 Mean,0.045 6 0.012C. thalictroides 70 No, control 62 Not applicable 55 Not applicable

90 No, control 76 Not applicable 73 Not applicable100 No, control 93 Not applicable 85 Not applicable

60,000 Yes 2/10,000 0.02 9/20,000 0.04670,000 Yes 3/18,000 0.17 13/40,000 0.03390,000 Yes 6/25,000 0.024 7/50,000 0.014

Mean,0.02 6 0.004 Mean,0.031 6 0.016

652 Plant Physiol. Vol. 163, 2013

Muthukumar et al.

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

(Fig. 3E), while in control tissues, no GUS expressionwas observed (Fig. 3F). In GUS staining, potassiumferrocyanide and potassium ferricyanide aids in rapidoxidation of X-Gluc during hydrolysis (Cervera, 2005),which might have prevented GUS gene expressionfrom being observed in all tissues that were analyzed forGUS expression, including the parts of the sporophytes.We expected that all C. thalictroides T1 gametophytes andT2 sporophytes from otherwise-confirmed transgenicswould be GUS positive because this is a hermaphroditespecies, but this was not the case. It is possible thattransfer DNA (T-DNA) integration could have occurredin germinating spores during the 90-h cultivation pe-riod, which could have led to chimeric plants.

Fern Actin Promoter Performance

The 889-bp 59 upstream region of the P. vittata actingene was sufficient to highly express the GUSPlus re-porter gene in both ferns in all life stages, includingprothalli, the sexual gametophytes, and the asexualdiploid sporophytes. Moreover, the actin promoterisolated from P. vittata was able to function efficientlyin another Pteridaceae family member, C. thalictroides,which might indicate regulatory elements in the actinpromoter could be expected to function in other ferntaxa as well.

Transformation Efficiency

For P. vittata, approximately 40,000 to 50,000 sporeswere subjected to three transformation experiments withno antibiotic selection (Table I). The average transfor-mation efficiency, calculated at the prothalli stage, was0.053%. If calculated at the gametophyte stage, it was0.045%. Of the 5,000 sporophytes that were still associ-ated with gametophytes that were randomly screened forGUS, only 40 GUS-positive sporophytes were transferred

to pots. Of these, 29 sporophytes survived and wereestablished in pots. This efficiency likely underrepresentsthe actual transformation efficiency because subsamplingfor GUS expression was performed at each life stage.Moreover, all sporophytes that were formed fromgametophytes were not recovered and analyzed.Hence, for calculating transformation efficiency, sporo-phytes were not considered. All of the 29 established

Figure 5. Southern-blot analysis of T1 sporophytes of P. vittata (A, B,and C) and C. thalictroides (D). Genomic DNA from fronds weredigested with ClaI and probed with a DIG-labeled hpt gene fragment.Lanes 2 to 5, 7 to 10, and 12 to 22: transgenic P. vittata. Lanes 1, 6,and 23: nontransgenic P. vittata. Lanes 11 and 24: ClaI-digested R4pGWB501 actin promoter-GUSPlus plasmid DNA. Lanes 26 to 35: transgenicC. thalictroides. Lane 34: nontransgenic C. thalictroides. DIG-labeledl DNA was used as a marker in all blots. Each plant sample lane rep-resents individual plants arising each from an independent spore. Lanes23 and 24 were exposed longer (8 h) than the rest of the blot (4 h) usingsame X-ray.

Figure 4. Vector map of R4pGWB501 multisite Gateway binary vectorcontaining the P. vittata actin promoter driving the GUSPlus gene withor without an apoplastic signal peptide sequence. There is a uniqueBamHI site in the T-DNA that was used to determine insertion (copy)number in transgenic P. vittata and C. thalictroides sporophytes usingsouthern-blot analysis. A 5.7-kb ClaI fragment is noted; it was utilizedto diagnose whether A. tumefaciens plasmid-derived contaminationwas present using southern-blot analysis from putative transgenic frondsamples.

Plant Physiol. Vol. 163, 2013 653

Stable Transformation of Ferns

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

sporophytes were GUS positive and were found tocontain T-DNA inserts using southern-blot analysis.

For C. thalictroides, approximately 60,000 to 90,000spores were used for three transformation experimentswith no antibiotic selection (Table I). If calculated atthe prothalli stage, the transformation efficiency was0.02%. If calculated at the gametophyte stage, it was0.03%. Of the 6,000 sporophytes that were randomlyselected for GUS assays, only 28 of these were trans-ferred to pots. Of these, 10 of the 15 sporophytes thatsurvived transplanting were GUS stain positive. All ofthe 10 GUS-positive plants were positive for the trans-gene in the southern-blot analysis.

Direct comparison of transformation efficiencies amongtaxa is difficult because each transformation method isinfluenced by various factors and efficiency calculationsare not standardized. However, our fern spore trans-formation efficiency reported here is nearly identical tothat found for fungal (Fusarium oxysporum) spore trans-formation (Mullins et al., 2001), in which there was 0.05%(500 transgenic conidia per 106 spores) efficiency. Ten-fold higher transformation efficiency (0.8%, 800 spo-rangia from 105 spores) has been reported for liverwort(Marchantia polymorpha) transformation based on initialhygromycin-resistant sporangial development fromprothalli (Ishizaki et al., 2008). Similarly, in Arabidopsisplants, 0.5% to 3% transformation efficiency has beenreported using the floral dip transformation method(Bechtold et al., 1993; Clough and Bent, 1998). None-theless, floral dip transformation is highly efficientonly in Arabidopsis and, apparently, only in someecotypes. It is important to demonstrate stable transfor-mation and inheritance, which was shown here in thevarious life stages of two fern species.

We conservatively chose GUS as the visible markerfor these experiments because of its sensitivity; we had

no a priori knowledge about the strength of promoterscontrolling marker gene expression. In the absence ofusing antibiotic selection, the experimental efficiencycould be increased by using a fluorescent marker suchas an orange fluorescent protein gene instead of GUS(Mann et al., 2012). If an orange fluorescence assay couldbe used at the prothalli stage, for example, subjecting afew thousand spores to transformation would lead toselecting just a few transformed ferns, which wouldgreatly reduce the numbers of ferns handled through-out the life cycle.

Transgene Integration

Even though the first generation of sporophytes be-longs to T1 generation, they are largely supported byremaining gametophyte tissues, and hence we thoughtit might be possible that A. tumefaciens could persistin the sporophyte tissues. Southern-blot analysis wasperformed to test for stable integration of T-DNA inthe genomes of both ferns and to estimate T-DNA copynumbers.

Southern-blot hybridization of ClaI-digested genomicDNA from fronds of sporophytes of both ferns usinghygromycin phosphotransferase (hpt) gene probe frompGWB501 vector revealed that a 5.7-kb A. tumefaciensplasmid band (Fig. 4) was not present in any of thetransgenic ferns assayed (Fig. 5; Supplemental Fig. S4A).In all 29 transgenic P. vittata, the hpt probe hybridizedto an approximately 2.0- to 2.1-kb ClaI fragment (Fig. 5,A–C; Supplemental Fig. S4A), whereas in all 10 trans-genic C. thalictroides (Fig. 5D), strong hybridization wasobserved at 2.8 kb. Additionally, in C. thalictroides,there was also a weaker hybridization signal observedat 2.6 kb. In both ferns, similar hybridization signals

Figure 6. Southern-blot analysis of T1sporophytes of P. vittata (A and B) andC. thalictroides (C and D) to determinethe number of T-DNA inserts. GenomicDNA from fronds were digested withBamHI and probed with a DIG-labeledGUSPlus gene fragment. Lane 1: BamHI-digested R4pGWB 501 actin promoter-GUSPlus plasmid DNA. Lanes 2 to 7,9 to 11, and 13 to 22: transgenic P. vittata.Lanes 8 and 12: nontransgenic P. vittata.Lanes 25 to 28 and 31 to 36: transgenicC. thalictroides. Lanes 23 and 29:BamH1-digested plasmid positive con-trol. Lanes 24 and 30: nontransgenicC. thalictroides. DIG-labeled l DNAwas used as a marker in all the gels.Each plant sample lane represents in-dividual plants arising each from anindependent spore.

654 Plant Physiol. Vol. 163, 2013

Muthukumar et al.

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

were observed in most of the individual transgenicferns examined. Moreover, all GUS positive lines werealso hpt positive in the southern blots, which impliesthat all transgenic lines had the full T-DNA insert.In the pGWB501 vector, there is one BamHI site

present within the T-DNA (Fig. 4), and BamHI diges-tion was utilized to estimate number of T-DNA insertsintegrated into the P. vittata and C. thalictroides genomeamong transgenic plants. In P. vittata, most of the trans-genic lines apparently contained more than one insert,and the hybridization patterns were similar amongtransgenic events. Transgenic lines in lanes 2 to 7 and9 and 10 had at least six inserts of the GUSPlus gene(Fig. 6A). Only one line (lane 19) had a single insert(Fig. 6B). Transgenic lines loaded in lanes 14, 18, and22 had at least three inserts (Fig. 6B). Lines loaded inlanes 13, 15, 16 and 17, 21 and 22 had at least twoinserts (Fig. 6B). Lines loaded in lanes 48 to 53 and55 to 57 in Supplemental Figure S4 had only two in-serts. The band sizes ranged from 1.8 to 9.5 kb. Basedon the banding pattern and the copy number estima-tion in the southern-blot analysis, there were approx-imately seven independent transgenic P. vitatta eventsrecovered.C. thalictroides genomic DNA was apparently more

difficult to digest with BamHI compared with P. vittatagenomic DNA. We used 150 units of BamHI to digest15 mg DNA of P. vittata DNA, but C. thalictroides re-quired 400 units of enzyme to digest most of the genomicDNA. Despite this problem, both ClaI and BamHI en-zymes effectively digested genomic DNA of both ferns(Supplemental Fig. S5). In C. thalictroides, four lines (lanes25 to 28) were found to contain only a single T-DNAinsert (Fig. 6C). Transgenic lines loaded in lanes 31 and 32had an estimated four copies, and lines in lanes 33 and 34had at least six copies of the T-DNA. Lines in lanes35 and 36 were also found to have at least six copiesbut in different apparent locations compared with linesloaded in lanes 31 and 32 (Fig. 6D). The band sizesranged from 1.8 to 8.5 kb. Based on this analysis, weestimate there were up to 10 independent transgenicevents analyzed.Southern-blot analysis confirmed that there was no

A. tumefaciens-derived plasmid contamination and full-length integration of the T-DNA in transgenic plants.The presence of the 2.0-kb hpt gene insert in both fernsand 2.8 and 2.6 kb in C. thalictroides depended on thenearest ClaI sites near the left border in the genomewhere it was integrated. It may be possible that a ClaIsite might exist proximal to 2 kb at the LB-plant ge-nome junction. Multiple inserts that might not be ran-dom have been observed previously in transgenicangiosperms (Gelvin and Kim, 2007), fungi (Michielseet al., 2005), and bryophytes (Ishizaki et al., 2008). Similarto our system, in which no selection pressure was used toselect transgenic lines, it was found that in Arabidopsis,T-DNA was integrated at multiple locations in the ge-nome without any discrimination of DNA sequence(Kim et al., 2007). Southern blots for both the hpt geneand GUS gene substantiated the integration of the entire

T-DNA in the fern genome. However, in eight lines ofP. vittata and two lines of C. thalictroides, southern-blotanalysis using the GUS probe showed many small-sizedbands in both blots, which might be the result of indis-criminate restriction enzyme digestion occurring in bothferns, i.e. BamHI. Although, in A. tumefaciens-mediatedtransformation, truncated transgene integration is lesslikely to occur than in biolistics, it has been observed inhigher plants, especially in Pinus strobus and in wheat(Triticum aestivum; Levee et al., 1999; Hensel et al., 2012).In the remaining C. thalictroides lines and in P. vittata, itcan be concluded that T-DNAs are mostly integratedintact. Multiple T-DNA insertion events per fern in bothspecies were observed during this study. Because bothferns are homosporous and hermaphroditic in nature,multiple T-DNA insertions might have occurred in theinitial haploid stages, as T-DNA insertion can occurmultiple times during the transformation process.Multiple T-DNA insertions in a single line have beenobserved in Arabidopsis (Alonso et al., 2003; O’Malleyand Ecker, 2010), and multiple T-DNA insertions werealso observed even within a single locus in Arabidopsis(Valentine et al., 2012). Based on our results, multipleT-DNA insertions might have occurred within a narrowregion of the fern genome. Analyses of more transgenicfern lines will help determine the pattern of T-DNAinsertion and will also determine whether more singleinsertion events can be obtained using this spore trans-formation method.

CONCLUSION

We have demonstrated that two fern species can bestably transformed by targeting spores using eitherA. tumefaciens or biolistics. We also discovered andvalidated a fern actin promoter, which is active in twofern species at each life stage. These tools should en-able expanded functional genomics studies in ferns,which will help to provide answers to many questionscurrently unknown in fern biology, including thoseregarding sex determination. Furthermore, because cvC-fern Express is emerging as a fern model, the abilityto stably transform it will enhance its usefulness as amodel. Practical research in P. vittata should also beenabled, especially with regards to arsenic hyper-accumulation (Gumaelius et al., 2004), to further developits capabilities for phytoremediation and phytosensingapplications. We are encouraged that ferns are appar-ently amenable to A. tumefaciens-mediated transforma-tion and that spores could be universal targets for ferntransformation.

MATERIALS AND METHODS

Actin Promoter Isolation and Vector Construction

Because there were no available fern promoters for transformation studies,we sought likely candidates in Pteris vittata. The P. vittata actin gene was isolatedfrom genomic DNA using plant-conserved degenerate primers: forward primer,59-ATGGCNGAYGGNGARGA-39 and reverse primer (Bhattacharya et al.,

Plant Physiol. Vol. 163, 2013 655

Stable Transformation of Ferns

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

1993), 59-GAAGCAYTTGCGRTGSACRAT-39. The 59 upstream region of theputative actin gene was isolated by two-step genome-walking procedure us-ing the Genomewalker kit (Clontech). The putative actin promoter was used toregulate the expression of the GUSPlus reporter gene. Two GUSPlus variants,one with and one without a rice (Oryza sativa) signal peptide sequence, wereused to determine if one was more effective than the other in fern transfor-mation. pCAMBIA1305.2 (GenBank accession no. AF354046) features GUSPlusthat contains a castor bean (Ricinus communis) intron and a rice apoplastic signalpeptide. This version was used to transform P. vittata. For Ceratopteris thalictroidestransformation studies, the same P. vittata actin promoter was used to controlGUSPlus gene taken from pCAMBIA 1305.1 (GenBank accession no.AF354045), which lacks the rice apoplastic signal peptide. The actin pro-moter and GUSPlus genes were subcloned into pDONRP4P1R and PCR8/GW/TOPO (Life Technologies), respectively, using the manufacturer’s in-structions for primer design and cloning. The promoter and the GUSPlusfragments were cloned into the Gateway binary vector R4pGWB 501, whichalso harbors hpt for plant hygromycin selection (Nakagawa et al., 2008; Fig. 4),using a multisite Gateway three fragment vector construction kit (Life Tech-nologies). After sequence confirmation, the binary vector (12.6 kb with T-DNAof 6.0 kb) was transformed into Agrobacterium tumefaciens strain EHA105 (Hoodet al., 1993) using a freeze-thaw method (Höfgen and Willmitzer, 1988). TheP. vittata actin promoter and actin gene sequences were deposited into GenBank(KC463697).

A. tumefaciens-Mediated Transformation of Spores

P. vittata was acquired from Edenspace Systems Corporation, and sporeswere collected from mature sporophyte fronds. C. thalictroides ‘C-fern Express’spores were obtained from Carolina Biological Supply Company. One hun-dred milligrams of spores were suspended in 2 mL of sterile deionized water,cleaned using a 60-mm nylon mesh followed by surface sterilization using2.5% (v/v) commercial bleach containing a few drops of Tween 20 for 5 min,and then washed five times in equal volumes of sterile water. After each wash,the spores were pelleted using a tabletop centrifuge at 10,000ɡ for 3 min.Spores were suspended in 0.5 mL 1.5% (w/v) carboxy methyl cellulose (lowviscosity, Fisher Scientific) as described in Sheffield et al. (2001). The sporeconcentration was calculated based on hemocytometer counts. Based on thehemocytometer counts, the spore volume was adjusted to 300 spores 50 uL–1.To increase the efficiency of spore germination, the suspended spores wereincubated at room temperature for 15 h before transformation or germination.

A 2-mL A. tumefaciens inoculant culture was started in induction mediumwith the required antibiotic (Utermark and Karlovsky, 2008) but withoutacetosyringone. After 8 h, 250 mL from the 2-mL culture was inoculated into25 mL of the same medium. The culture was grown to 0.8 optical density for15 h at 28°C, and then 10 mL of the culture was centrifuged at 5,000ɡ for15 min and resuspended in 20 mL of the same A. tumefaciens growth medium.Two hundred micromolar acetosyringone was then added to induce virulencegene expression. The culture was further incubated for 24 h by shaking thesuspension at 60 rpm at room temperature (25°C) on a rotary shaker. Then,0.5 mL of induced culture was mixed with 0.5 mL of sterilized spore sus-pension. This mixture was incubated for 15 min, and the entire suspensionwas plated as 50-mL (300 spores per plate) aliquots on top of a hydrophilicpolyvinylidene difluoride membrane (Fisher Scientific) in 100-mm cocultivationagar plates containing A. tumefaciens growth medium with 200 mM acetosyr-ingone and 2 g L–1 gellan gum for 72 h. Membranes were then transferred to one-half-strength MS plus 20 g Suc plates containing 400 mg L–1 timentin to kill theresidual A. tumefaciens. The spores were subcultured every 2 weeks in the samemedium. The transformation experiment was repeated five times, and efficiencywas calculated from three independent experiments. After transformation, wesimply allowed the ferns to progress naturally through their life cycle. Sporesgerminated into gametophytes, which were transferred to one-half-strength MS(0 g Suc) plates containing 200 mg L–1 timentin for sporophyte development. Allfern cultures were grown at 22°C with 65 mmol m–2 s–1 light intensity usingfluorescent lights with 16-h/8-h light/dark cycle. Mature sporophytes wereeventually transferred to 4-L pots containing PRO-MIX potting media. Sporeswere collected and analyzed for transgene inheritance analyses. The maturesporophytes were grown in growth chambers at 24°C with 105 mmol m–2 s–1

irradiance from fluorescent lights under a 16-h/8-h light/dark cycle. We usednontransgenic ferns from the same accessions as controls, and they were treatedidentically (except transformation).T1 spores were germinated in pots and allowedto grow until they formed gametophytes and sporophytes. Tissues were collectedand analyzed for GUS expression.

Hygromycin Sensitivity Assay

To determine whether hygromycin could be used as a selection agent, weperformed a toxicity analysis by plating spores on one-half-strength MS me-dium containing 20 g Suc and 400 mg L–1 timentin supplemented with hygro-mycin (0, 3, 5, 10, 15, and 25 mg L–1 ). After 4 weeks, hygromycin toxicitywas determined based upon live/dead status of each fern. The experimentwas repeated twice.

Particle Bombardment of P. vittata Spores

P. vittata spores were surface sterilized using 70% (v/v) ethanol and 10%(v/v) bleach. Surface sterilized spores were then plated onto polyvinylidenedifluoride membranes on one-half-strength MS with 20 g Suc medium in 1-mLaliquots for biolistic transformation. Fresh spores, 5-day-old germinatingspores or protonemata, and 15-day-old multicellular protenemata were usedfor transformation. The pMDC32 vector containing a dual 35S promoterdriving the gusA was used for transformation (Curtis and Grossniklaus, 2003).Biolistic transformation was carried out using conditions described for Glycinemax that used 0.6-mm gold particles (Hazel et al., 1998). Nicotiana tabacum‘Xanthi’ leaves grown from in vitro cultures were used as a positive control.P. vittata gametophytes and N. tabacum leaves were GUS stained after 3 weeksafter the gametophytes were fully developed.

Marker Gene Analysis

GUS histochemical staining was performed using a standard protocol(Cervera, 2005) with modifications to help infiltrate the P. vittata sporophytefronds, which were recalcitrant to standard staining. Fronds were perforatedwith carborundum 600 (Fisher Scientific) using cotton swab by painting onboth adaxial and abaxial surfaces, which was subsequently removed bywashing three times with 0.1 M Tris-HCl buffer, pH 7.0. Fern tissues wereincubated in GUS staining solution, destained using 70% (v/v) ethanol toremove chlorophyll, rinsed with 50% (v/v) glycerol, and viewed under a lightstereoscope.

Southern-Blot Analysis

Genomic DNA from each fern species was isolated using a cetyl-trimethylammonium bromide method (Dong et al., 2005). Between 3 to 5 g ofyoung transgenic and nontransgenic control sporophyte fronds were collectedand frozen in liquid nitrogen. Tissues were homogenized in 500 mL extractionbuffer containing 2% (w/v) cetyltrimethylammonium bromide, 100 mM

Tris-HCl, pH 8.0, 1.4 M NaCl, 20 mM EDTA, and 0.2% (v/v) 2-mercaptoetanoland incubated at 65°C for 20 min with 50 mg mL–1 ribonuclease A. Then, anequal volume of chloroform:isoamyl alcohol (24:1, v/v) was added and mixedvigorously. The mixture was centrifuged immediately at 6,000ɡ for 20 min.The aqueous phase was transferred to a fresh tube, and 0.7 mL–1 of isopropylalcohol was added and incubated at –20°C for at least 2 h. The DNA waspelleted by centrifuging the mixture at 6,000ɡ for 20 min. The DNA pellet waswashed twice with 70% (v/v) ethanol and resuspended in 400 mL steriledistilled water. DNA was quantified using fluorometry as well as visual es-timation using agarose gel electrophoresis.

For each sample, 15 mg genomic DNA was digested with BamHI or ClaIrestriction endonucleases (Fig. 4) and electrophoresed in 0.8% (w/v) agarosegels for 22 h. All gels were loaded with 3 to 4 mL of digoxigenin (DIG)-labeledl DNA marker III (HindIII-EcoRI digested, Roche Applied Science) as sizemarker and 100 picogram mL–1 ClaI- or BamHI-digested R4pGWB501 plasmidDNA containing P. vittata actin promoter and GUSPlus, both of which can bedetected in x-ray film along with experimental samples. DNA was transferredto a nylon membrane (GE Healthcare) by capillary blotting (Sambrook et al.,1989) and cross linked by UV irradiation.

Membranes were hybridized with a 570-bp GUSPlus fragment gene or a580-bp hpt fragment as probes in 35 mL DIG EasyHyb solution at 42°C (forGUS probe, 40°C) for 2 h. The GUSPlus primers were forward, 59-CAAG-CACCGAGGGCCTGAGC-39 and reverse, 59-CTCAGTCGCCGCCTCGTTGG-39.The hpt primers were forward, 59-GCGCTTCTGCGGGCGATTTG and re-verse, 59-GTCCGAATGGGCCGAACCCG. Each probe was DIG labeled(Roche Applied Science), and hybridizations were conducted at 42°C for thehpt gene and 40°C for the gus gene in a rotating hybridization oven. The hy-bridization temperature was based on the melting temperature of the probe

656 Plant Physiol. Vol. 163, 2013

Muthukumar et al.

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

and the formula based on DIG hybridization buffer as specified in the man-ufacturer’s protocol (Roche DIG application manual for filter hybridization).The membranes were then washed twice in 23 SSC (Sambrook et al.,1989) and0.1% (w/v) SDS at room temperature for 5 min each and twice in 0.13 SSCand 0.1% (w/v) SDS at 60°C for 15 min each (Neuhaus-Url and Neuhaus,1993; Dietzgen et al., 1999). Detection of the hybridized probe DNA wascarried out as described in the manufacturer’s protocol. Ten-fold diluted CDP-Star chemiluminescent reagent (Roche Applied Science) was used (incubationtime 10 min), and signals were visualized on x-ray film after 6- to 18-h exposure.The GUSplus probe was used for BamHI-digested blots, and the hpt probe wasused for ClaI-digested blots.

Sequence data from this article can be found in the GenBank data librariesunder accession number KC463697.

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure S1. The phylogenetic relationship of the P. vittataactin gene with selected actin genes from other plant species.

Supplemental Figure S2. Hygromycin kill curve analysis of P. vittata andC. thalictroides, which were scored based on gametophytes formed after3 to 4 weeks of culture on one-half-strength MS medium.

Supplemental Figure S3. GUS histochemical staining of transgenic andcontrol T1 C. thalictroides sporophytic fronds.

Supplemental Figure S4. Southern-blot analysis of T1 sporophytes ofP. vittata to screen for A. tumefaciens plasmid contamination and for T-DNAinsert number from a subsample of 10 of 29 T1 P. vittata sporophytes.

Supplemental Figure S5. Agarose gel electrophoresis of ClaI- and BamHI-digested samples that were used for P. vittata and C. thalictroides southern-blot analysis.

Supplemental Table S1. Putative regulatory motifs of P. vittata actin pro-moter.

ACKNOWLEDGMENTS

We thank Jonathan Willis for helping with DIG southern-blot experiments,Jason Burris for assistance in transformation experiments, Kellie Burris, JenniferHinds, Reggie Millwood, Yijia Zhang, and Mike Blaylock for their encour-agement and help, Mitra Mazarei and Reza Hajimorad for their assistancein carborundum perforation experiments, and Les Hickok for sharing in-formation about cv C-fern, which was useful in this study.

Received July 10, 2013; accepted August 6, 2013; published August 9, 2013.

LITERATURE CITED

Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, StevensonDK, Zimmerman J, Barajas P, Cheuk R, et al (2003) Genome-wide inser-tional mutagenesis of Arabidopsis thaliana. Science 301: 653–657

Banks JA (1999) Gametophyte development in ferns. Annu Rev PlantPhysiol Plant Mol Biol 50: 163–186

Banks JA, Nishiyama T, Hasebe M, Bowman JL, Gribskov M, dePamphilis C,Albert VA, Aono N, Aoyama T, Ambrose BA, et al (2011) The Selaginellagenome identifies genetic changes associated with the evolution of vascularplants. Science 332: 960–963

Bechtold N, Ellis J, Pelletier G (1993) In planta Agrobacterium mediatedgene transfer by infiltration of adult Arabidopsis thaliana plants. Life Sci316: 1194–1199

Bhattacharya DS, Stickel D, Sogin ML (1993) Isolation and molecular phylo-genetic analysis of actin coding regions from the prymnesiophyte alga,Emilianiahuxleyi, using reverse transcriptase and PCR methods. Mol Biol Evol10: 689–703

Bhattacharyya J, Chowdhury AH, Ray S, Jha JK, Das S, Gayen S, ChakrabortyA, Mitra J, Maiti MK, Basu A, et al (2012) Native polyubiquitin promoter ofrice provides increased constitutive expression in stable transgenic riceplants. Plant Cell Rep 31: 271–279

Camloh M (1993) Spore germination and early gametophyte development ofPlatycerium bifurcatum. Am Fern J 83: 79–85

Cervera M (2005) Histochemical and fluorometric assays for uidA (GUS)gene detection. In L Pena, ed, Transgenic Plants: Methods and Protocols,Vol 286. Humana Press, Totowa, New Jersey, pp 203–213

Christensen AH, Quail PH (1996) Ubiquitin promoter-based vectors forhigh-level expression of selectable and/or screenable marker genes inmonocotyledonous plants. Transgenic Res 5: 213–218

Chiyoda S, Ishizaki K, Kataoka H, Yamato KT, Kohchi T (2008) Direct trans-formation of the liverwort Marchantia polymorpha L. by particle bom-bardment using immature thalli developing from spores. Plant Cell Rep27: 1467–1473

Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16: 735–743

Curtis MD, Grossniklaus U (2003) A gateway cloning vector set for high-throughput functional analysis of genes in planta. Plant Physiol 133:462–469

de la Riva GA, Gonzalez-Cabrera J, Vázquez-Padrón R, Ayra-Pardo C(1998) Agrobacterium tumefaciens: a natural tool for plant transformation.Electron J Biotechnol 1: 18–133

Dhankher OP, Rosen BP, McKinney EC, Meagher RB (2006) Hyper-accumulation of arsenic in the shoots of Arabidopsis silenced for arsenatereductase (ACR2). Proc Natl Acad Sci USA 103: 5413–5418

Dietzgen RG, Abedina M, Higgins CM, Karunatratne S, Vickers J (1999)Nonradioactive detection of transgenes in plant genomic Southern blots.Biochemica (Indianap, Ind) 1: 19–20

Dong R, Formentin E, Losseso C, Carimi F, Benedetti P, Terzi M, SchiavoFL (2005) Molecular cloning and characterization of a phytochelatinsynthase gene, PvPCS1, from Pteris vittata L. J Ind Microbiol Biotechnol32: 527–533

Gelvin SB, Kim S-I (2007) Effect of chromatin upon Agrobacterium T-DNA in-tegration and transgene expression. Biochim Biophys Acta 1769: 410–421

Gifford E, Foster A (1989) Morphology and Evolution of Vascular Plants.Freeman and Company, New York, pp 626

Gumaelius L, Lahner B, Salt DE, Banks JA (2004) Arsenic hyperaccumulationin gametophytes of Pteris vittata. A new model system for analysis of arsenichyperaccumulation. Plant Physiol 136: 3198–3208

Hazel CB, Klein TM, Anis M, Wilde HD, Parrott WA (1998) Growthcharacteristics and transformability of soybean embryogenic cultures.Plant Cell Rep 17: 765–772

Hensel G, Oleszczuk S, Daghma DES, Zimny J, Melzer M, Kumlehn J (2012)Analysis of T-DNA integration and generative segregation in transgenicwinter triticale (3 Triticosecale Wittmack). BMC Plant Biol 12: 171

Höfgen R, Willmitzer L (1988) Storage of competent cells for Agrobacteriumtransformation. Nucleic Acids Res 16: 9877

Hood EE, Gelvin SB, Melchers LS, Hoekema A (1993) New Agrobacteriumhelper plasmids for gene transfer to plants. Transgenic Res 2: 208–218

Indriolo E, Na GN, Ellis D, Salt DE, Banks JA (2010) A vacuolar arsenitetransporter necessary for arsenic tolerance in the arsenic hyperaccumulatingfern Pteris vittata is missing in flowering plants. Plant Cell 22: 2045–2057

Ishizaki K, Chiyoda S, Yamato KT, Kohchi T (2008) Agrobacterium-mediatedtransformation of the haploid liverwort Marchantia polymorpha L., anemerging model for plant biology. Plant Cell Physiol 49: 1084–1091

Jones HD, Sparks CA (2009) Stable transformation of plants. In DJ Somers,P Langridge, JP Gustafson, eds, Methods in Molecular Biology, Vol 513.Humana Press, New York, pp 111–130

Karami O, Esna-Ashari M, Kurdistani GK, Aghavaisi B (2009) Agrobacterium-mediated genetic transformation of plants: the role of host. Biol Plant 53:201–212

Kikkert JR, Vidal JR, Reisch BI (2004) Stable transformation of plantcell by particle bombardment/biolistics. In L Pena, ed, TransgenicPlants: Methods and Protocols. Humana Press, Totowa, New Jersey, pp61–78

Kim S-I, Veena , Gelvin SB (2007) Genome-wide analysis of AgrobacteriumT-DNA integration sites in the Arabidopsis genome generated undernon-selective conditions. Plant J 51: 779–791

Lai Hy, Lim YY, Tan SP (2009) Antimicrobial tyrosinase inhibiting activities ofleaf extracts from medicinal ferns. Biosci Biotechnol Biochem 73: 1362–1366

Levee V, Garin E, Klimaszewska K, Seguin A (1999) Stable genetic transfor-mation of white pine (Pinus strobus L.) after cocultivation of embryogenictissues with Agrobacterium tumefaciens. Mol Breed 5: 429–440

Ma LQ, Komar KMM, Tu C, Zhang W, Cai Y, Kennelley ED (2001) A fernthat hyperaccumulates arsenic. Nature 409: 579

Plant Physiol. Vol. 163, 2013 657

Stable Transformation of Ferns

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.

Mandal A, Purakayastha TJ, Patra AK, Sanyal SK (2012) Phyto-remediation of arsenic contaminated soil by Pteris vittata L. II. Effect onarsenic uptake and rice yield. Int J Phytoremediation 14: 621–628

Mann DGJ, Abercrombie LL, Rudis MR, Millwood RJ, Dunlap JR,Stewart CN Jr (2012) Very bright orange fluorescent plants: endoplas-mic reticulum targeting of orange fluorescent proteins as visual re-porters in transgenic plants. BMC Biotechnol 12: 17

Marrs RH, Watt AS (2006) Biological flora of the British Isles 245: Pteridiumaquilinum (L.) Kuhn. J Ecol 94: 1272–1321

Michielse CB, Hooykaas PJ, van den Hondel CA, Ram AF (2005) Agrobacterium-mediated transformation as a tool for functional genomics in fungi. CurrGenet 48: 1–17

Mullins ED, Chen X, Romaine P, Raina R, Geiser DM, Kang S (2001)Agrobacterium tumefaciens-mediated transformation of Fusarium oxy-sporum: an efficient tool for insertional mutagenesis and gene transfer.Phytopathology 91: 173–180

Nakagawa T, Nakamura S, Tanaka K, Kawamukai M, Suzuki T, Nakamura K,Kimura T, Ishiguro S (2008) Development of R4 gateway binary vectors(R4pGWB) enabling high-throughput promoter swapping for plant re-search. Biosci Biotechnol Biochem 72: 624–629

Neuhaus-Url G, Neuhaus G (1993) The use of nonradioactive digoxigeninchemiluminescent technology for plant genomic Southern blot hybridi-zation: A comparison with radioactivity. Transgenic Res 2: 115–120

O’Malley RC, Ecker JR (2010) Linking genotype to phenotype using theArabidopsis unimutant collection. Plant J 61: 928–940

Rensing SA, Lang D, Zimmer AD, Terry A, Salamov A, Shapiro H,Nishiyama T, Perroud P-F, Lindquist EA, Kamisugi Y, et al (2008) ThePhyscomitrella genome reveals evolutionary insights into the conquest ofland by plants. Science 319: 64–69

Rutherford G, Tanurdzic M, Hasebe M, Banks J (2004) A systemic genesilencing method suitable for high throughput, reverse genetic analysesof gene function in fern gametophytes. BMC Plant Biol 4: 6.5

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A LaboratoryManual. Cold Spring Harbor Lab Press, Plainview, New York, pp 7.6–7.9

Schaefer D, Zryd JP, Knight CD, Cove DJ (1991) Stable transformation ofthe moss Physcomitrella patens. Mol Gen Genet 226: 418–424

Schedlbauer MD (1976) Fern gametophyte development: controls of di-morphism in Ceratopteris thalictroides. Am J Bot 63: 1080–1087

Sheffield E, Douglas GE, Hearne SJ, Huxam S, Wynn JM (2001) En-hancement of fern spore germination and gametophyte growth in arti-ficial media. Am Fern J 91: 179–186

Shelmerdine PA, Black CR, McGrath SP, Young SD (2009) Modellingphytoremediation by the hyperaccumulating fern, Pteris vittata, ofsoils historically contaminated with arsenic. Environ Pollut 157:1589–1596

Srivastava S, Suprasanna P, D’Souza SF (2012) Mechanisms of arsenictolerance and detoxification in plants and their application in transgenictechnology: a critical appraisal. Int J Phytoremediation 14: 506–517

Stachel SE, Nester EW, Zambryski PC (1986) A plant cell factor inducesAgrobacterium tumefaciens vir gene expression. Proc Natl Acad Sci USA83: 379–383

Stout SC, Clark GB, Archer-Evans S, Roux SJ (2003) Rapid and efficient sup-pression of gene expression in a single-cell model system, Ceratopteris ri-chardii. Plant Physiol 131: 1165–1168

Sundaram S, Rathinasabapathi B (2010) Transgenic expression of fernPteris vittata glutaredoxin PvGrx5 in Arabidopsis thaliana increases planttolerance to high temperature stress and reduces oxidative damage toproteins. Planta 231: 361–369

Sundaram S, Wu S, Ma LQ, Rathinasabapathi B (2009) Expression of aPteris vittata glutaredoxin PvGRX5 in transgenic Arabidopsis thaliana in-creases plant arsenic tolerance and decreases arsenic accumulation in theleaves. Plant Cell Environ 32: 851–858

Thaler JS, Owen B, Higgins VJ (2004) The role of the jasmonate response inplant susceptibility to diverse pathogens with a range of lifestyles. PlantPhysiol 135: 530–538

Utermark J, Karlovsky P (2008) Genetic transformation of filamentous fungi byAgrobacterium tumefaciens. Nat Protocols 10:

Valentine ME, Wolyniak MJ, Rutter MT (2012) Extensive phenotypicvariation among allelic T-DNA inserts in Arabidopsis thaliana. PLoSONE 7: 9e44981

Wilkinson JE, Twell D, Lindsey K (1997) Activity of CaMV 35S and nospromoters in pollen: implications for field release of antisense transgenicplants. J Exp Bot 48: 265–275

Wu FH, Shen SC, Lee LY, Lee SH, Chan MT, Lin CS (2009) Tape-ArabidopsisSandwich - a simpler Arabidopsis protoplast isolation method. PlantMethods 5: 16

658 Plant Physiol. Vol. 163, 2013

Muthukumar et al.

www.plantphysiol.orgon May 12, 2020 - Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved.