document

712 NATURE BIOTECHNOLOGY VOL 17 JULY 1999 http://biotech.nature.com

RESEARCH

The dinitroaniline herbicides—including trifluralin, ethalfluralin,oryzalin, and pendimethalin—are used for pre-emergence weedcontrol, particularly in cotton, soybean, wheat, and oilseed crops. Inuse for 25 years, the dinitroanilines selectively inhibit the growth ofmany annual grasses and a few dicotyledonous weeds. This class ofherbicide disrupts meristem development in roots and shoots as aresult of the net depolymerization of cellular microtubules1–3. Theseherbicides bind to purified plant tubulin and inhibit its polymeriza-tion into microtubules4. Although the evolution of resistance to thedinitroanilines is rare, their repeated use has resulted in the selec-tion of resistant (R) biotypes of one of the world’s worst weeds,goosegrass (Eleusine indica)5–7. The R phenotype was found to bethe consequence of a spontaneous mutation in an a-tubulin genecausing Thr239 to be replaced by Ile239 in the R-biotype major a-tubulin8–10.

In a previous experiment, we used the whiskers maize transfor-mation method to overexpress a- and b-tubulins in transgenicmaize calli11. We reported that cell viability was dependent upon thecoexpression of both a- and b-tubulin and that the goosegrass R-biotype a-tubulin could confer an R phenotype9 in the maize calli.In addition, overexpression of transgenic a- and b-tubulin sup-pressed endogenous tubulin synthesis to such an extent that thetransgenic tubulin replaced virtually all endogenous tubulin11.These experiments used a nonregenerable maize suspension cul-ture9,11. In the current study, we ask whether this technology can betransferred to a regenerable plant system without disrupting regen-eration, reproduction, and normal growth. This is particularly rele-vant because the single a- and b-tubulin transgenes altered in this

study encode the major constituents of microtubules.In all eukaryotes, both a- and b-tubulins exist as families of iso-

types resulting from expression of multiple tubulin genes and/orposttranslational modification12. In several cases particular isotypeshave been shown to impart novel characteristics to the micro-tubules into which they become incorporated, provided that theirconcentration reaches a certain threshold level. These novel charac-teristics include the construction of anatomically different classes ofmicrotubules as well as microtubules that differ with respect to coldand drug stability and assembly/disassembly kinetics. These func-tionally distinct microtubules may be essential for normal growthand morphogenesis12.

In this study, we have used tobacco as a test regenerable plantsystem to ascertain whether we could obtain dinitroaniline herbi-cide-resistant transgenic plants overexpressing a mutant a-tubulinand a b-tubulin gene. The data presented here encourage the fur-ther application of this technology to crops in which these herbi-cides are used.

Results and discussionGeneration of stable tobacco transformants harboring a-tubulinand b-tubulin transgenes. DNA cassettes containing either a- andb-tubulin coding sequences or the same coding sequences withintegrated epitope tag sequences were inserted into the vector pJR1and used to transform tobacco using Agrobacteriumtumefaciens–mediated gene transfer. Transformants were selectedon kanamycin, with resistance conferred by expression of the nptIIgene present in the pJR1 vector. The high-activity cauliflower

Dinitroaniline herbicide-resistanttransgenic tobacco plants generated

by co-overexpression of a mutanta-tubulin and a b-tubulin

Richard G. Anthony, Stefanie Reichelt, and Patrick J. Hussey*

School of Biological Sciences, Royal Holloway University of London, Egham, Surrey, UK TW20 OEX. *Corresponding author (e-mail: [email protected]).

Received 8 February 1999; accepted 30 April 1999

Dinitroaniline herbicides are used for the selective control of weeds in arable crops. Dinitroanilineherbicide resistance in the invasive weed goosegrass was previously shown to stem from a spontaneousmutation in an a-tubulin gene. We transformed and regenerated tobacco plants with an a/b-tubulin dou-ble gene construct containing the mutant a-tubulin gene and showed that expression of this constructconfers a stably inherited dinitroaniline-resistant phenotype in tobacco. In all transformed lines, thetransgene a- and b-tubulins increased the cytoplasmic pool of tubulin approximately 1.5-fold whilerepressing endogenous a- and b-tubulin synthesis by up to 45% in some tissues. Transgene a- and b-tubulin were overexpressed in every plant tissue analyzed and comprised approximately 66% of the totaltubulin in these tissues. Immunolocalization studies revealed that transgene a- and b-tubulins wereincorporated into all four microtubule arrays, indicating that they are functional. The majority of the a/b-tubulin pools are encoded by the transgenes, which implies that the mutant a-tubulin and the b-tubulincan perform the majority, if not all, of the roles of microtubules in both juvenile and adult tobacco plants.

Keywords: dinitroaniline herbicide resistance, a- and b-tubulin, microtubules, transgenic tobacco

© 1999 Nature America Inc. • http://biotech.nature.com©

199

9 N

atu

re A

mer

ica

Inc.

• h

ttp

://b

iote

ch.n

atu

re.c

om

NATURE BIOTECHNOLOGY VOL 17 JULY 1999 http://biotech.nature.com 713

RESEARCH

mosaic virus 35S/maize alcohol dehydrogenase intron 1 hybrid pro-moter (35Si) directed expression of both the a- and b-tubulin cod-ing sequences. The a-tubulin coding sequence used was eitherEiStua1 or EiRtua1, from the biotype of goosegrass that was, respec-tively, sensitive (S) or resistant (R) to dinitroaniline herbicide.EiStua1 and EiRtua1 differ only at amino acid residue 239, wherethreonine is replaced by isoleucine in EiRtua1 (ref. 9). The b-tubu-lin coding sequence used was Zmtub2 (ref. 13).

Production of the mutant a-tubulin, EiRtua1, confers dini-troaniline herbicide resistance on transgenic tobacco. Gross mor-phology and seed germination efficiency were used as key indicatorsof phenotypic similarities between nontransformed and trans-formed plants. Figure 1 shows a composite picture of control non-transformed plants and plants containing a single copy of constructR-tag (R1Nt plants). No morphological dissimilarities could be dis-tinguished between the plants (Fig. 1). Germination efficiencies of1,633 seeds from the control and R1Nt plants were also very similar(96.6% for the control compared with 95.8% for the R1Nt seeds).

Seeds from nontransformed tobacco and T2 lines containingeither a single copy or two copies of construct S, S-tag, R, or R-tagwere germinated and grown on solid media containing 0.4 mg/Lpendimethalin. This dose of dinitroaniline herbicide distinguishednontransformed tobacco seedlings, as well as T2 seedlings contain-

ing either construct S or S-tag, from T2 seedlings containing con-struct R or R-tag. The T2 lines containing one copy of construct S-tag (S1Nt) and one and two copies of construct R-tag (R1Nt, R2Nt,respectively) were used for further analyses (Fig. 2).

To evaluate the response of nontransformed tobacco and T2lines S1Nt, R1Nt, and R2Nt to different herbicides, we measured rootlength of seedlings germinated on media containing various con-centrations of herbicide (0–1.0 mg/L). All seedlings exhibiteddose-dependent inhibition of root or shoot growth (Fig. 3).Growth of the nontransformed tobacco seedlings and the S1Nt

seedlings was almost completely inhibited on 0.4 mg/L dinitroani-line herbicide (pendimethalin and oryzalin), whereas growth ofeither R1Nt or R2Nt was unaffected at this concentration. In con-trast, growth of the R1Nt and R2Nt seedlings was significantly inhib-ited (up to 90%) at herbicide concentrations of 1.0 mg/L. Similarresults were obtained using a phosphorothioamidate herbicide,amiprophosmethyl, which has a mode of action similar to that ofthe dinitroanilines and which likely interacts with the same targetsite14. All seedlings were equally sensitive to pronamide, anotherantimicrotubule herbicide similar in size to the dinitroanilines andphosphorothioamidates, but with a different chemistry and modeof action.

Effect and usage of overexpressed a-tubulin. Expression oftransgenic a-tubulins in S1Nt, R1Nt, and R2Nt transformedseedlings, nontransformed control seedlings, and in the differenttissues and organs of R1Nt plants was assessed by one-dimensionalgel electrophoresis and western blot analysis using a pan-reactivemonoclonal antibody against a-tubulin (Fig. 4). In seedlings of allthree transformed lines, transgenic a-tubulin increased the overallpool of tubulin by approximately 1.5-fold. This increase was cou-pled with suppression of endogenous a-tubulin synthesis. Thedecrease in endogenous tubulin synthesis was as much as 45%, ascompared with the tubulin content of nontransformed tobaccoseedlings (Fig. 4A). Similar results were obtained when the levels ofb-tubulin were analyzed (data not shown). These data demonstratethat the contribution of the transgenic a- and b-tubulin to thepool of tubulin is as much as about 66%. A similar value for this

Figure 1. Phenotype of transgenic tobacco. Gross morphology of thecontrol (C) nontransformed tobacco plants compared with R1Nt plants.Whole plants (10 weeks), buds, and flowers (13 weeks) are shown.

Figure 2. Dinitroaniline herbicide–resistant transgenic tobacco.Growth response of (A), the control nontransformed tobacco (B) S1Nt

(C) R1Nt, and (D) R2Nt seedlings to 0.4 mg/L of the dinitroanilineherbicide, pendimethalin.


199

9 N

atu

re A

mer

ica

Inc.

• h

ttp

://b

iote

ch.n

atu

re.c

om


RESEARCH

contribution was found in all of the different tissues of floweringR1Nt plants analyzed (Fig. 4B). Overproduction of transgenic a-and b-tubulin and corresponding suppression of endogenous a-and b-tubulin synthesis was previously observed in similarly trans-formed maize calli9,11. The results shown here demonstrate thatsuppression of endogenous a- and b-tubulin synthesis that accom-panies overexpression of both a- and b-tubulin can occur in func-tionally differentiated cell types and not just in a randomly divid-ing population of dedifferentiated maize cells.

To determine whether the transgenic a- and b-tubulins werefunctional, immunolocalization studies were carried out using anti-bodies against an antihemagglutinin epitope tag to detect the trans-genic a-tubulin in seedling root cells of R1Nt. The transgenic a-tubulin was shown to incorporate into interphase cortical, pre-prophase band, spindle, and phragmoplast microtubule arrays (Fig.5). Similar results were obtained using the anti–c-myc epitope tagantibodies to detect the transgenic b-tubulin (data not shown).

We have shown that a dinitroani-line-resistance phenotype can beconferred on transgenic tobaccoexpressing the mutant a-tubulincoding sequence, EiRtua1. This a-tubulin is expressed at high levels inR1Nt and R2Nt seedlings and in alltissues of R1Nt plants examined.Little difference was observedbetween the level of expression oftransgenic a-tubulin in R1Nt andR2Nt. In transgenic maize calli, thereis a correlation with transgenic a-and b-tubulin gene expression andcopy number, but this relationship isnot proportional and in this latterrespect is similar to our observa-tions for tobacco11. The overexpres-sion of transgenic a- and b-tubulinhas resulted in the suppression ofendogenous tubulin synthesis in alltobacco tissues. These results sug-gest that synthesis of transgenictubulin has a direct effect on synthe-sis of endogenous tubulin.Alternatively, synthesis of trans-genic tubulin may proceed at ahigher rate because it has a moreefficient promoter or is translated

more efficiently, as previously proposed for transgenic a- and b-tubulin gene expression in maize calli11. Evidence of posttranscrip-tional regulation of tubulin synthesis has been provided from stud-ies using animal cells15,16; however, it remains to be establishedwhether such controls occur in plants.

Immunolocalization studies confirmed that the transgenic a-and b-tubulins are functional. The observation that transgenictubulin is more abundant than the endogenous tubulin in every tis-sue of the R1Nt plant examined suggests that these single transgenictubulin isotypes can perform the majority, if not all, of roles ofmicrotubules in the tobacco plant. These data suggest that dini-troaniline herbicide resistance can be transferred to regenerablecrop plants for the beneficial control of weeds by herbicide applica-tion after planting. However, because of the complex evolution oftubulin isotype families in eukaryotes, we cannot rule out the possi-bility that suppression of endogenous tubulin synthesis may reduceexpression of isotypes involved in conferring subtle variation to

Figure 4. Expression of transgenic a-tubulin and effects on endogenous a-tubulin synthesis. Western blots were probed with a pan-reactivemonoclonal antibody against a-tubulin. Lanes contain 50 mg of total protein extracts (A) from control nontransformed tobacco seedlings (C),S1Nt, R1Nt, and R2Nt seedlings, and (B) from control and various tissues of an R1Nt plant. The densitometric analyses below each blot show theabundance of endogenous and transgene a-tubulin relative to the control value of 100%. The slower migrating bands on the western blotscorrespond to the transgenic a-tubulin/hemagglutinin epitope tag (a/HA) compared with the endogenous a-tubulin.

A B

Figure 3. Effect of increasing concentrations of herbicide on seedling growth. Growth responses ofnontransformed seedlings (q), S1Nt (X), R1Nt (G), and R2Nt (n) seedlings to the herbicides pendimethalin, oryzalin,amiprophosmethyl, and pronamide.


199

9 N

atu

re A

mer

ica

Inc.

• h

ttp

://b

iote

ch.n

atu

re.c

om

NATURE BIOTECHNOLOGY VOL 17 JULY 1999 http://biotech.nature.com 715

microtubules at certain developmental stages under certain envi-ronmental conditions. Such subtle variation in microtubule prop-erties in relation to the conditions for growth and developmentwill be established only after several generations of field trials.

Experimental protocolVector construction. All cloning techniques were standard17. DNA cassettescontaining a- and b-tubulin coding sequences, with or without epitope tagsequences, fused to the 35Si promoter were inserted into pJR1 (ref. 11) (Fig.6). All vector constructions were checked by restriction analysis and DNAsequencing before being introduced into tobacco.

Tobacco transformation. Each vector construct was introduced intocompetent A. tumefaciens strain LBA4404. Each A. tumefaciens clone wasthen used to inoculate sterile leaf discs of Nicotiana tabaccum var Samsun,as described previously18. T1 seeds were collected and germinated on 1/2MS(Murashige and Skoog)) medium supplied with 200 mg/ml kanamycin.Genomic DNA was extracted and purified using the hexadecyltrimethy-lammonium bromide procedure19. Transformation of kanamycin-resistantplants was verified by PCR of the transgenic a-tubulin cassette, using theforward primer 5¢-CTACTCCAAAAATGTCAAAGAT-3¢ (35S promoterregion) and the reverse primer 5¢-GGCGTAGTCGGGCACGTCGTA-3¢(hemagglutinin-epitope tag region).

T1 populations harboring construct S, S-tag, R, or R-tag segregated forkanamycin resistance: Independent lines showed either a 3:1 (kanr: kans)segregation ratio, indicating a single-copy insertion, or a 15:1, ratio indicat-ing the insertion of two copies. The T1 lines were selfed to generate the T2lines, which were uniformly resistant to kanamycin. Seeds of all T2 linesharboring construct R or R-tag germinated and grew normally on 0.4 mg/Lpendimethalin. A control line containing a single copy of construct S-tag(designated S1Nt) was analyzed in parallel with two lines containing con-struct R-tag, one of which contained a single copy and the other, two copies(designated R1Nt and R2Nt, respectively).

Growth response of transgenic plants to herbicides. Tobacco seeds weresurface-sterilized, and 52 seeds were positioned on gridded plates contain-ing 1/2MS gelled medium incorporating herbicide at concentrations from 0to 1.0 mg/L. The tobacco seedlings were grown for four weeks. Seedlingswere removed from the agar, and the lengths of the roots were measured.

Western blot analysis. Tobacco total protein extracts were prepared,fractionated on a 7.5% one-dimensional SDS–PAGE gel, transferred tonitrocellulose membrane, and immunostained as described20. Each lane

was loaded with 50 mg of protein as quantified by the Bio-Rad (London,UK) protein assay. The primary antibody used was anti–a- tubulin(Amersham, Little Shalfont, UK) at a 100-fold dilution in Tris BufferedSaline containing 0.1% (wt/vol) Tween-20. This antibody is well character-ized and cross-reacts with a-tubulins from a wide range of distantly relatedspecies. The secondary antibody used was antimouse peroxidase-conjugat-ed immunoglobulin at a dilution of 1000-fold (Sigma, Poole, UK).

Immunofluorescence of tobacco root cells. Two-week-old tobacco rootswere fixed, digested, and squashed onto poly-L-lysine–coated coverslips asdescribed21. Root cells were stained with the antibody 12CA5 (ref. 22), spe-cific for the hemagglutinin (HA) epitope, diluted 100-fold in phosphate-buffered saline (PBS). The secondary antibody used was fluorescein isoth-iocyanate (FITC)–conjugated rabbit anti-mouse IgG again diluted 100-foldin PBS. Samples were viewed using a Leica TCS4D confocal laser scanningmicroscope (Leica, Wetzlar, Germany).

AcknowledgmentsThis work was funded by the Biotechnological and Biological Sciences ResearchCouncil. We would like to thank J. Payne (Zeneca Agrochemicals) for assis-tance with the tobacco transformation technique.

1. Sumida, S. & Ueda, M. Effect of O-ethyl O-(3-methyl-6-nitrophenyl) N-sec-butyl-phosphoroamidate (S-2846), an experimental herbicide, on mitosis inAllium cepa. Plant Cell Pysiol. 17, 1351–1354 (1976).

2. Appleby, A.P. & Valverde, B.E. Behaviour of dinitroaniline herbicides in plants.Weed Technol. 3, 198–206 (1989).

3. Anthony, R.G. & Hussey, P.J. Dinitroaniline herbicide resistance and the micro-tubule cytoskeleton. Trends Plant Sci.4, 112–116 (1999).

4. Morejohn, L.C., Bureau, T.E., Mole-Bajer, J., Bajer, A.S. & Fosket, D.E.Oryzalin, a dinitroaniline herbicide, binds to plant tubulin and inhibits micro-tubule polymerization in vitro. Planta 172, 252–264 (1987).

5. Mudge, L.C., Gossett, B.J. & Murphy, T.R. Resistance of goosegrass (Eleusineindica) to dinitroaniline herbicides. Weed Sci. 32, 591–594 (1984).

6. Vaughn, K.C., Marks, M.D. & Weeks, D.P. A dinitroaniline resistant mutant ofEleusine indica exhibits cross-resistance and supersensitivity to antimicro-tubule herbicides and drugs. Plant Physiol. 83, 956–964 (1987).

7. Waldin, T.R., Ellis, J.R. & Hussey, P.J. Tubulin-isotype analysis of two grassspecies-resistant to dinitroaniline herbicides. Planta 188, 258–264 (1992).

8. Cronin, K.E., Hussey, P.J., Ray, J.A. & Waldin, T.R. Herbicide resistant plants.World Intellectual Property Organization No. WO 93/24637 (1993).

9. Anthony, R.G., Waldin, T.R., Ray, J.A., Bright, S.W.J. & Hussey, P.J. Herbicideresistance caused by spontaneous mutation of the cytoskeletal protein tubu-lin. Nature 393, 260–263 (1998).

RESEARCH

Figure 5. Immunolocalization of the transgenic a-tubulin. Root tipcells from R1Nt seedlings were stained with the antihemagglutininepitope tag, 12CA5. (A) Interphase cortical; (B) preprophase band; (C)spindle; and (D) phragmoplast microtubule arrays are identified.

Figure 6. Construction of plasmid vectors for tobaccotransformation. DNA cassettes containing a/b-tubulin codingsequences fused to the hybrid promoter containing the cauliflowermosaic virus (CaMV) 35S promoter and the maize alcoholdehydrogenase intron 1 (adh1)9. These DNA cassettes wereinserted into plasmid pJR1, which contains the left and rightborders (LB and RB, respectively) for integration into the tobaccogenome and the nptII gene for conferring kanamycin resistance(kanr) on the tobacco transformants. EiStua1 and EiRtua1, sensitiveand resistant a-tubulin coding sequences, respectively; Zmtub2, b-tubulin coding sequence; nos 3¢ term, nopaline synthase 3¢terminus.

A B

C D


199

9 N

atu

re A

mer

ica

Inc.

• h

ttp

://b

iote

ch.n

atu

re.c

om


RESEARCH

10. Yamamoto, E., Zeng, L. & Baird, W.V. a-Tubulin missense mutations correlatewith antimicrotubule drug resistance in Eleusine indica. Plant Cell 10, 297–308(1998).

11. Anthony, R.G. & Hussey, P.J. Suppression of endogenous a- and b-tubulin syn-thesis in transgenic maize calli overexpressing a- and b-tubulins. Plant J. 16,297–304 (1998).

12. Raff, E.C. in Microtubules (eds Hyams, J. & Lloyd, C.) 85–109 (Wiley-Liss, NewYork 1994).

13. Hussey P.J. et al. The b-tubulin gene family in Zea mays: two differentiallyexpressed tubulin genes. Plant Mol. Biol. 15, 957–972 (1990).

14. Ellis, J.R., Taylor, R. & Hussey, P.J. Molecular modelling indicates that twochemically distinct classes of anti-mitotic herbicide bind to the same receptorsite(s). Plant Physiol. 105, 15–18 (1994).

15. Cleveland, D.W. & Theodorakis, N.G. in Microtubules (eds Hyams, J. & Lloyd,C.) 47–58 (Wiley-Liss, New York 1994).

16. Gonzalez-Garay, M.L. & Cabral, F. a-Tubulin limits its own synthesis: evidence

for a mechanism involving translational repression. J. Cell Biol. 135,1525–1534 (1996).

17. Ausubel, F.M. et al. Current protocols in molecular biology. (Greene PublishingAssociates, New York; 1992).

18. Horsch, R.S. et al. A simple and general method for transferring genes intoplants. Science 227, 1229–1231 (1985).

19. Doyle, J.J. & Doyle, J.L. Isolation of plant DNA from fresh tissue. Focus 12,13–15 (1990).

20. Hussey, P.J. & Gull, K. Multiple isotypes of a and b-tubulin in the plantPhaseolus vulgaris. FEBS Lett. 181, 131–118 (1985).

21. Smertenko, A., Blume, Y., Viklicky, V., Opatrny Z. & Draber, P. Post-translation-al modifications and multiple tubulin isoforms in Nicotiana tabacum L cells.Planta 201, 349–358 (1997).

22. Field, J. et al. Purification of a RAS-responsive adenylyl cyclase complex fromSaccharomyces cerevisiae by use of an epitope addition. Meth. Mol. Cell. Biol.8, 2159–2165 (1988).


199

9 N

atu

re A

mer

ica

Inc.

• h

ttp

://b

iote

ch.n

atu

re.c

om

document

Documents