Proc. Nati. Acad. Sci. USAVol. 81, pp. 1440-1444, March 1984Cell Biology

Tubulins from different higher plant species are immunologicallynonidentical and bind colchicine differentially

(microtubules/cultured plant cells/tubulin antibodies/immunoautoradiography/antimicrotubule drugs)

LOUIS C. MOREJOHN*, THOMAS E. BUREAUt, LEE P. TocCHI, AND DONALD E. FOSKETtDepartment of Developmental and Cell Biology, University of California, Irvine, CA 92717

Communicated by Keith R. Porter, November 22, 1983

ABSTRACT We have initiated immunological and drug-binding studies on the tubulins from different higher plantspecies. Antibodies were raised against electrophoretically sep-arated rose (Rosa sp.) tubulin a- and ,&subunits and charac-terized by immunoblot autoradiographic assays. Each IgGpreparation bound to its antigen and cross-reacted differen-tially with the respective tubulin subunits from an alga, seaurchin, rabbit, and cow. Antigenic determinants were sharedmore among the 13-subunits than among the a-subunits fromthese organisms. Tubulins were isolated from cultured cells ofcarrot (Daucus carota) and hibiscus (Hibiscus rosa-senensis).Immunoautoradiography and quantitation of cross-reactivityon blots showed nonidentity among homologous subunits fromrose, carrot, hibiscus, and alga tubulins, with more antigenicdifferences among a-subunits than among 13subunits. Com-parative colchicine-binding assays showed that rose and hibis-cus tubulins bound 33% and 65%, respectively, of the colchi-cine bound by carrot tubulin and that higher plant tubulinsbound much less colchicine than bovine brain tubulin underidentical conditions.

Higher plant microtubules participate in a variety of cellularfunctions including chromosome migration during mitosis,cell plate formation in cytokinesis, organelle transport, andorientation of cellulose microfibril deposition in developingcell walls (1). The main structural component of microtu-bules is tubulin, a conserved heterodimeric protein having amolecular weight of 100,000 (2-4) and composed of one a-subunit and one 3-subunit (Mr, -50,000 each). Antibodiesraised against animal tubulin have been shown to bind to tu-bulin-like components of plant cell extracts (5, 6) and to dec-orate microtubules by indirect immunofluorescence (7, 8) orimmunogold staining (9). This cross-reactivity of animal tu-bulin antibodies has suggested that plant and animal tubulinsare immunologically similar. However, quantitative immu-nological differences have been shown to exist not only be-tween the tubulins from different vertebrate species (10) butalso between the tubulins in different organelles from a givenspecies (11). Comparative peptide mapping studies on the a-and 83-subunits of tubulins from taxonomically distant spe-cies have revealed dramatic structural differences betweenthe a-subunits from plants and animals (12-15). Further-more, the resistance of plant microtubules in cells to severalwell-characterized antimicrotubule agents such as colchicinehas implied that plant and animal tubulins have differentpharmacological properties (1, 16).Because higher plant tubulin was not isolated until recent-

ly, little is known of its immunological and biochemicalproperties. As a consequence, we do not know the degree towhich higher plant tubulins conform to the model for tubulinstructure that has arisen from studies of mammalian braintubulin. Recently, we developed a method for the isolation

of tubulin from cultured rose cells (13). We report here that(i) pure IgG preparations from antisera raised to rose tubulinsubunits cross-react to varying degrees with tubulin subunitpolypeptides from an alga, sea urchin, rabbit, and cow, (ii)tubulins isolated from different higher plants are immunolog-ically nonidentical, and (iii) the amount of colchicine boundby each plant tubulin not only is different but also is muchlower than that bound by bovine brain tubulin.

MATERIALS AND METHODSPlant Cell Suspension Cultures. Suspension cultures of

rose (Rosa sp. cv. Paul's scarlet) were derived from stemexplants (17) and grown as described (13, 18). Suspensioncultures of carrot (Daucus carota) and hibiscus (Hisbiscusrosa-senensis) were initiated from root callus (from D. Ra-din, University of California, Irvine) and leaf callus (from F.Hoffman, University of California, Irvine), respectively.Carrot and hibiscus suspension cultures were grown in medi-um containing Murashige and Skoog salt formulation (19),2,4-dichlorophenoxyacetic acid (1 mg/liter), and sucrose (30g/liter). Carrot medium was supplemented with thiamine(0.4 mg/liter)/i-inositol (0.1 mg/liter) and adjusted to pH5.7; hibiscus medium contained benzyl aminopurine (2 mg/liter)/naphthalene acetic acid (2 mg/liter) and was adjustedto pH 5.6. All suspension cultures used in these studies weretransferred to fresh medium at 14-day intervals and weregrown in the dark at 22-24°C on gyratory shakers. Only cellsin exponential growth phase (days 4-7) were used in tubulinisolation.

Purification of Tubulins. Tubulins from cultured higherplant cells were isolated as described (13) with the followingmodifications. Cells chilled to 4°C were washed with coldisolation buffer containing 50 mM Pipes KOH, pH 6.9/1 mMEGTA/0.5 mM MgCl2/1 mM dithiothreitol/5 ,uM leupeptinhemisulfate (Sigma)/5 ,uM pepstatin A (Sigma)/10 mMdiethyldithiocarbamic acid (Sigma). Cells were homogenizedin isolation buffer containing 2 mM GTP (lithium salt; Cal-biochem), and the additions of DNase I and RNase wereeliminated. After filtration and centrifugation, the 48,200 x gsupernatant was gently swirled in a flask containing DEAE-Sephadex A50 at a ratio of 0.15 ml of ion-exchange resin perml of supernatant for 30 min at 4°C in the dark. Protein-con-taining fractions were eluted with step gradients of KCl asdescribed (13), and fractions A and B were collected by(NH4)2SO4 precipitation at 85% of saturation; fraction C wascollected at 50% of saturation. Precipitates were dissolved in1 ml of isolation buffer containing 1 M sucrose/0.02%NaN3/0.1 mM GTP (sucrose isolation buffer) and dialyzedversus 1 liter of sucrose isolation buffer for 10 hr at 4°C.

Abbreviations: a-IgG, a-subunit IgG; ,¢IgG, /3-subunit IgG.*Present address: Department of Genetics and Cell Biology, Univer-sity of Minnesota, St. Paul, MN 55108.

tPresent address: Department of Botany, University of Texas, Aus-tin, TX 78712.tTo whom reprint requests should be addressed.

1440

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

17,

202

1

Proc. NatL Acad Sci USA 81 (1984) 1441

Dialyzates were clarified by centrifugation at 48,200 x g for1 hr, and the supernatants were frozen in dry ice-chilledmethanol and stored at -80TC.Bovine brain tubulin was purified by DEAE-Sephadex

A50 chromatography according to the batch method of Wei-senberg et al. (20) and Weisenberg and Timasheff (21) asmodified by Lee et al. (22), except that the MgC12 precipita-tion step was eliminated. Tubulin (NH4)2SO4 precipitateswere dialyzed (peptidase inhibitors were omitted), clarifiedby centrifugation, and stored as described for plant tubulins.Tubulin was .91% pure (23).

Rabbit brain tubulin was purified from freshly dissectedcerebra of New Zealand White rabbits by two cyclic roundsof microtubule polymerization and depolymerization usingthe method of Sloboda et al. (24). Microtubule pellets werestored in 4 M glycerol at -80'C.Alga (Chlamydomonas reinhardtii) flagellar tubulin and

sea urchin (Strongylocentrotus purpuratus) egg and spermflagellar outer doublet tubulins were gifts from the labora-tory of L. Wilson (University of California, Santa Barbara).NaDodSO4/Polyacrylamide Gel Electrophoresis. All pro-

teins in these studies were separated on NaDodSO4/7.5%polyacrylamide slab gels having a thickness of 0.8 mm basedon the method of Studier (25), with modifications as de-scribed (13). Gels were stained overnight with 0.025% Coo-massie brilliant blue R-250 in 5% methanol/7.5% acetic acid.

Protein Determinations. Protein dye-binding (26) was usedto determine protein concentrations and bovine serum albu-min (fraction V, Sigma) was used as a standard. However,this spectrophotometric assay (Bio-Rad) yields absorbancevalues of bovine serum albumin solutions that are -2.1 timeshigher than those obtained from several other gravimetrical-ly determined proteins in solution (Bio-Rad Instruction Man-ual). For this reason, we determined the extent to which theuse of bovine serum albumin standards in dye-binding assaysunderestimates tubulin in solution. Bovine brain tubulin wasdialyzed extensively against distilled water and lyophilized.Gravimetrically determined amounts of lyophilized tubulinwere run on gels along with quantities of tubulin as estimatedby the dye-binding method. Stained gels containing both se-ries of tubulins were scanned with a densitometer, and stan-dard curves constructed from areas of a- and /3-subunitpeaks showed linearity for loadings of 0.5-3 ,ug of lyophi-lized tubulin per lane. Comparison of standard curves fromeach tubulin series showed that the dye-binding method un-derestimates tubulin by a factor of 2.04 when bovine serumalbumin is used as a standard protein. Thus, tubulin concen-trations measured by the Bio-Rad assay were adjusted up-ward by this amount.

Preparation of Antibodies. Polyclonal antibodies to the in-dividual subunits of rose tubulin were prepared in rabbits.Tubulin was run on 9.25% polyacrylamide slab gels (13, 25);gels were stained in 0.2% Coomassie blue in 50% methanolfor 30 min and destained in 5% methanol for 15 min. Electro-phoretically separated a- and,3-subunit bands were excised,and polypeptides were electroeluted into dialysis tubing andprecipitated with acetone. Each subunit precipitate (100 ug)was emulsified in phosphate-buffered saline and completeFreund's adjuvant and injected subcutaneously at the shoul-ders and haunches of a different female New Zealand Whiterabbit using :12.5 Mg per site. Rabbits were injected withthe same antigen preparations at the same sites 2 weeks lat-er. Four intradermal booster injections of -10 ,ug per sitewere given at 2-week intervals using subunits in diced poly-acrylamide gel bands emulsified in phosphate-buffered sa-line and incomplete Freund's adjuvant. Rabbits were bled at2-week intervals, titers of sera were tested by immunoblot-ting (27), and antisera with the highest titers were pooled.The immunoglobulin fraction of each antiserum was isolatedby (NH4)2SO4 precipitation and purified by DEAE-cellulose

(Cellex D, Bio-Rad) chromatography according to the batchmethod of Hudson and Hay (28). Preimmune serum takenfrom each rabbit prior to antigen injection was used for con-trol studies.Immunoautoradiography. Tubulin samples from different

organisms were run on NaDodSO4/7.5% polyacrylamideslab gels and transferred to nitrocellulose filters by electro-blotting (27) at 0C for 2.5 hr. Electroblotted gels werestained overnight with Coomassie blue to verify that tubulinsubunits were completely transferred. Filters were closed insealable plastic bags along with a solution of 1% bovine se-rum albumin/10 mM Tris HCl, pH 7.4/5 mM EDTA/0.15 MNaCl/0.02% NaN3 (solution A) and agitated for 30 min atroom temperature. IgG was then added to desired final dilu-tions and filters were rotated slowly at 370C for 16 hr. Preim-mune serum for each antigen was used at the same dilutionsas IgG. Filters were washed with three changes of 25 ml of10 mM Tris HCl, pH 7.4/5 mM EDTA/0.15 M NaCl (solu-tion B) at room temperature. Five milliliters of solution Acontaining 4 x 106 cpm of 125I-labeled Staphylococcus aure-us protein A (7.99 gCi/,ug; 1 Ci = 37 GBq; New EnglandNuclear) was added to bags and filters were rotated at 370Cfor 2 hr. Filters were removed from the bags and washedwith two changes of solution B (200 ml) for 2 hr at roomtemperature. Filters were dried between sheets of chroma-tography paper, covered with SaranWrap, and exposed toKodak X-Omat AR x-ray film on DuPont Cronex LightningPlus intensifying screens at -80'C for 2-4 hr.

Colchicine-Binding Assays. The colchicine-binding reac-tions of rose, hibiscus, carrot, and bovine brain tubulins(4.6-4.9 ,uM) were assayed by the gel filtration method (20)as modified by Detrich et al. (29). The molecular weight oftubulin was assumed to be 100,000 (2-4). [3H]Colchicine wasobtained from New England Nuclear and had a specific ac-tivity of 37.2 Ci/mmol.

RESULTSRose Tubulin IgGs and Their Reactivities with Diverse Tu-

bulins. Tubulin was isolated from cultured cells of rose (13),and antisera were raised in rabbits against the electrophoreti-cally separated a- and /3-subunits. IgGs from each antiserumwere purified by DEAE-cellulose chromatography and theirbinding specificities were examined on nitrocellulose blotsby immunoautoradiography. Each IgG preparation bound toits respective subunit antigen on blots without cross-reactionwith the other subunit (Fig. 1). Neither the a- nor the p-sub-unit preimmune serum reacted with tubulin subunits, indi-cating a minimum of nonspecific binding.When each IgG was used at a dilution of 1:100, the autora-

diographic band intensity produced by the a-subunit IgG (a-IgG) was nearly twice that produced by the ,-subunit IgG (,/-IgG) (lanes 2 and 4), indicating a lower titer of the 83-IgG.When p-IgG was used at 1:50 dilutions for probing immuno-blots, similar autoradiographic signals were obtained asshown by densitometric traces of autoradiographic films.Therefore, the,B-IgG was used at this dilution for subsequentexperiments.The cross-reactivity of antibodies raised against animal tu-

bulin with tubulin-like factors in plant extracts (5, 6) and withmicrotubules in plant cells (7-9) has suggested similar immu-nological properties of diverse tubulins. To examine thisphenomenon more directly, we studied the specificities andextents of cross-reactivity of rose tubulin IgGs with the tubu-lin subunits from selected higher and lower plant and animalspecies. When tubulins from rose cells, alga (C. reinhardtii)flagella, sea urchin (S. purpuratus) flagellar outer doubletand egg, rabbit brain, and cow brain were run together on aNaDodSO4/polyacrylamide slab gel, a comparison of theelectrophoretic mobilities of the subunits revealed more

Cell Biology: Morejohn et aL

Dow

nloa

ded

by g

uest

on

June

17,

202

1

1442 Cell Biology: Morejohn et al.

B C2 3 4

rose carrot hibiscusSSB AB

S A B C S A B C S A B C

front-

FIG. 1. Binding of IgGs to the subunits of rose tubulin. (A)NaDodSO4/7.5% polyacrylamide gels containing separated a- and/3-subunits of rose tubulin (2 Ag) were electroblotted to strips of ni-trocellulose paper and probed with either preimmune serum or sub-unit IgG followed by '25I-labeled protein A. (B) Autoradiograph offilters probed with 1:100 dilutions of a-subunit preimmune serum(lane 1) and a-IgG (lane 2). (C) Autoradiograph of filters probed with1:100 dilutions of /3-subunit preimmune serum (lane 3) and /-IgG(lane 4). Autoradiographs were exposed for 2 hr. Apparent molecu-lar weights x lo- of protein standards and a- and /3-subunit posi-tions are indicated on the left.

variability among the a-subunits than among the /3-subunits(Fig. 2A). Furthermore, the rose and alga tubulin subunitbands were not as well separated as those of the animals.Very similar results were reported by Little et al. (12) withcarboxymethylated tubulins from plants and animals.

Tubulins were blotted from gels to nitrocellulose filtersand probed with rose subunit IgGs. Autoradiographs showedthat a-IgG cross-reacted most with the alga a-subunit, only

2 3 4 5 6A _ _

-ES _Mi--.-d *

V*WN

B___vq

FIG. 3. Isolation of rose tubulin and tubulin-like proteins fromcultured carrot and hibiscus cells. Extracts of cultured cells ofthree higher plants (rose, carrot, and hibiscus) were prepared, andsupernatants were fractionated by DEAE-Sephadex A50 chroma-tography into three protein-containing fractions (A, B, and C) andrun on a NaDodSO4/7.5% polyacrylamide gel. S, unfractionated su-pernatants; A, unbound proteins of fraction A; B, weakly boundproteins (0.4 M KCl step) of fraction B; and C, tightly bound pro-teins (0.4-0.8 M KCI step) of fraction C. S, A, and B lanes eachcontain 20 Ag of protein; C lanes contain 2 tug of protein. Apparentmolecular weights x l0'- and a- and 3-subunit positions are indicat-ed on the right.

slightly with those of sea urchin egg and flagellum, and not atall with those of rabbit brain or cow brain (Fig. 2B). Evenafter prolonged exposure of autoradiographs, no reactivitywas observed with the mammalian a-subunits, which indi-cates that the low level of cross-reactivity with sea urchin a-subunits was not due to nonspecific binding. When /3-IgGwas used to reprobe the same blot, it cross-reacted stronglywith the alga /B-subunit and at low and similar levels with allthe animal B-subunits (Fig. 2C).IgG Reactivities with Different Plant Tubulins. A modified

version of the DEAE-chromatographic procedure to isolateplant tubulin (13) was used to isolate tubulins from culturedcells of carrot (D. carota) and hibiscus (H. rosa-senensis).Analysis of the supernatants and step-gradient fractions byNaDodSO4/7.5% polyacrylamide gel electrophoresis showedthat proteins eluting in the 0.4-0.8 M KCl step (fraction C) ofboth carrot and hibiscus were highly enriched for two tubu-lin-like polypeptides of equal concentration that comigratedwith the a- and /subunits of rose tubulin (Fig. 3). Heavilyloaded gels revealed a few high molecular weight and lowmolecular weight polypeptides copurifying with the putativetubulins. Identification of carrot and hibiscus tubulins wasaccomplished by assembly of microtubules after incubation

A

FIG. 2. Reactivity of rose tubulin IgGs with tubulins from phylo-genetically distant species. (A) NaDodSO4/7.5% polyacrylamide gelcontaining a- and /3-subunits of tubulins (3 Mg) from various species.Stained a- and /-subunit bands of different tubulins displayed char-acteristic widths and thicknesses in preliminary gels. Multiplicationof the area of densitometer tracings of each tubulin band by its widthprovided an estimation of the amount of each sample to be loaded toyield equal amounts of tubulin in each lane. (B) Autoradiograph ofblot probed with a 1:100 dilution of a-IgG. (C) Autoradiograph of thesame blot shown in B after reprobing with a 1:50 dilution of /3-IgGantibody. Autoradiographs were exposed for 4 hr. Lanes: 1, rosecell tubulin; 2, alga flagellar tubulin; 3, sea urchin sperm flagellarouter doublet tubulin; 4, sea urchin egg tubulin; 5, rabbit brain tubu-lin; 6, bovine brain tubulin. Positions of a- and /-subunits are desig-nated on the left.

B

C

a---

_ I

ij

2 3 4

FIG. 4. Cross-reactivities with-w Wcarrot, hibiscus, and alga tubu-lins. (A) NaDodSO4/7.5% poly-acrylamide gel containing 2 uig ofrose (lane 1), carrot (lane 2), hibis-cus (lane 3), and alga (Chlamydo-

_ _mp monas) flagellar tubulins (lane 4).(B) Autoradiograph blot probedwith 1:100 dilution of a-IgG. (C)Autoradiograph of the same blotshown in B reprobed with 1:50 di-lution of /-IgG. AutoradiographsB and C were exposed for 3.25and 3.5 hr, respectively. Positions

it. Am ~of a- and ,B subunits are shown on*brtxh';-' ax the left.

A

68-

m43-.-

35-

-116

-68

Gom --a 3--&-43-43-35

_ ,,_-front

Proc. NatL Acad Sd USA 81 (1984)

.t 1.i. i .,.11.1'. -kcct -10,

-3 --w lo

Dow

nloa

ded

by g

uest

on

June

17,

202

1

Proc. NatL Acad ScL USA 81 (1984) 1443

Table 1. Quantitation of rose tubulin IgG reactivitiesa-Subunit /3-Subunit

Source of cpm % of cpm % oftubulin bound control bound control

Rose 15,855 100 17,936 100Carrot 13,222 83 9,326 52Hibiscus 2,237 14 10,526 59Alga 6,013 38 7,168 40

Determination of bound radioactivity was made by placing the ni-trocellulose filter used in Fig. 4C over the exposed autoradiographicfilm on a light box, carefully excising areas corresponding to a- and/-subunit bands, and assaying the filter pieces in a Beckman 5500Gamma Radiation Counter. Values were corrected for backgroundradioactivity on the filter by using pieces of equal size from areasthat contained no protein. Data are expressed as % of control (IgGbinding to rose subunit = 100%).

with 40 ,uM taxol at 240C (13, 30). Abundant microtubuleswere formed as revealed by increased turbidity at 400 nm(31) and electron microscopy of negatively stained samples(data not shown). Quantitative densitometry of fraction Ctubulins run on NaDodSO4/7.5% polyacrylamide gels gavepurities of 85% for rose tubulin and 90% for both carrot andhibiscus tubulins.The cross-reactivities of rose tubulin IgGs with carrot, hi-

biscus, and alga tubulins were examined on blots by immu-noautoradiography (27). Fig. 4 shows that the a-IgG andIgG reacted with the corresponding tubulin polypeptidesfrom each plant. The amount of cross-reactivity of a-IgG ap-peared to be more variable between the different speciesthan that of the /3IgG. Furthermore, staining of the blottedgels showed this was not the result of differential electropho-retic transfer of the proteins to the nitrocellulose filters.

Quantitative comparisons of the cross-reactivities of therose IgGs with plant tubulin subunits were made by cuttingthe radioactive bands from probed nitrocellulose blots andcounting gamma radiation. Table 1 shows that the amount ofradioactivity bound to the a-subunits was, in fact, more vari-able than that bound to the /3subunit. Surprisingly, morecross-reaction was seen with the a-subunit of the alga thanwith that of hibiscus, a higher plant.

Plant Tubulmi Colchicine Binding. It has not been shownpreviously whether the resistance of plant microtubules to

20 0.40

10 -0.2 -CC

05-P 0.1

'' O10 20 30 40 60 70 80

Fraction

FIG. 5. Hibiscus tubulin colchicine binding. Tubulin (4.9 uM)was incubated with 0.29 ,uM [3H]colchicine in sucrose isolation buff-er at 24°C for 2 hr. Colchicine-tubulin complex was separated fromunbound coichicine by gel filtration. The sample (1 ml) was appliedto a column containing Sephadex G25-fine (1.6 cm x 11 cm) equili-brated with isolation buffer. Fractions of 0.5 ml were collected at aflow rate of 0.5 ml/min. Fractions 1-30 and every 10th fractionthereafter were analyzed for protein and radioactivity. Aliquots of100 ,ul were used for protein assays (26), and the remaining 400-,ulsamples were added to 10 ml of Hydrofluor (National Diagnostics,Somerville, NJ) and assayed on a Beckman LS-230 liquid scintilla-tion spectrophotometer. *, Corrected values of radioactivity; o,corrected values for protein.

Table 2. Colchicine binding of different tubulinsTubulin mol of colchicine % of total

Source of concentration, bound per mol colchicinetubulin AuM of tubulin boundRose 4.6 3.3 x 1O-4 0.45Hibiscus 4.9 6.5 x 10-4 0.94Carrot 4.9 1.0 x 10-3 1.48Brain 4.9 1.1 x 10-2 15.20Gel filtration assays of colchicine binding of tubulins from rose,

hibiscus, carrot, and bovine brain were carried out as described inthe legend to Fig. 5. Determination of radioactivity bound to tubulinfor each assay was made with a curve integration program on anApple II computer and is expressed in mol of colchicine bound permol of tubulin.

colchicine is due to low binding affinities of plant tubulin forcolchicine or to any of several other modes of resistance(16). To address this problem, we compared the colchicine-binding reactions of rose, hibiscus, and carrot tubulins withthat of bovine brain tubulin by the gel filtration method (29).Animal tubulins bind colchicine optimally at 370C (32, 33),but because plant materials are usually treated with colchi-cine at ambient laboratory temperatures and because rose,hibiscus, and carrot suspension cultures are grown at roomtemperature (22-25°C), we chose to examine the colchicine-binding reactions at 24°C.

Preliminary binding assays of hibiscus tubulin indicatedthat binding equilibrium was achieved within 2 hr at 240C.Thus, all assays were carried out at these standard condi-tions. The elution profiles of protein and radioactivity fromthe gel filtration column showed that tubulin-colchicinecomplex was formed by each plant preparation. As an exam-ple, the hibiscus colchicine-binding experiment is shown inFig. 5.Table 2 summarizes the results of the gel filtration assays

and shows that each plant tubulin bound a different amountof colchicine. Rose and hibiscus tubulins bound 33% and65%, respectively, of the amount of colchicine bound by car-rot tubulin. Furthermore, the plant tubulins bound verysmall amounts of colchicine compared with that bound bybrain tubulin, with carrot tubulin binding only 9% of thatbound by brain tubulin, even though the brain tubulin bind-ing reaction had not yet come to equilibrium under these ex-perimental conditions (32).

DISCUSSIONWe have reported herein that antibodies raised in rabbitsagainst the electrophoretically separated a- and 3subunitsof cultured rose cell tubulin bind to their respective antigenson blots and show no cross-reaction between subunits.While the amino acid sequence of plant tubulin has not beenreported, sequence analyses of animal tubulins (2-4) haveshown that the a- and /-subunits share several regions ofhomology and that -41% of their sequences are identical. Ithas been postulated that the sequence homology of a- and ,8-subunits indicates they arose from a common ancestral se-quence early in eukaryotic evolution (34). Nevertheless, sev-eral studies have reported that monoclonal or polyclonalantibodies against either the a- or 83-subunit do not cross-react with the other subunit (35-40) but do cross-react withcorresponding subunits of tubulins from diverse species (37-40). While it is not clear whether the antigenic determinantsof each subunit are derived from primary amino acid se-quence, secondary structure, or post-translational modifica-tion, the above observations indicate that regions of homolo-gy between a- and /3subunits are so highly conserved thatthey do not elicit immune responses readily.

Cross-reaction of each rose tubulin IgG with tubulin sub-units from diverse sources (Fig. 2) shows immunological

Cell Biology: Morejohn et aL

Dow

nloa

ded

by g

uest

on

June

17,

202

1

1444 Cell Biology: Morejohn et al.

similarities between these polypeptides, yet the extent ofcross-reaction decreased rapidly with increased phylogenet-ic distance. Most cross-reactivity was seen with alga tubulin;only a very small amount of cross-reactivity with animal a-and 3-subunits and a total lack of cross-reactivity with mam-malian a-subunits were observed. These results suggest thatthe immune responses elicited in rabbits by rose tubulin sub-units were predominantly directed against immunogenic de-terminants not found on animal tubulin subunits. Further-more, the lack of rose a-IgG cross-reaction with mammaliana-subunits, its highly variable cross-reaction with plant a-subunits, and higher titer than f-IgG suggest that a-subunitsof diverse organisms have more antigenic differences thanthe /3-subunits.We have improved the anion-exchange chromatography

method for the isolation of cultured rose cell tubulin (13)such that increased tubulin purity (-85%) is obtained. Fur-thermore, the use of this method for the purification of tubu-lins from cultured cells of carrot and hibiscus indicates thatanion-exchange chromatography is widely applicable for theisolation of tubulin from cultured higher plant cells. Purityfor both carrot and hibiscus tubulins was -90%.

Microtubules in plant cells are 100-1000 times more resist-ant to colchicine treatment than are microtubules in animalcells (16), and attempts to isolate plant tubulin by means ofcolchicine binding have met with failure (1, 16). We haveshown that the colchicine-binding capacities of rose, hibis-cus, and carrot tubulins are different from each other and aremuch lower than that of brain tubulin under identical condi-tions, even when the temperature for the brain tubulin col-chicine-binding reaction is less than optimum. Apparentlythe tubulins from higher plants have low colchicine-bindingaffinities, and it is this property that is largely responsible forthe resistance of microtubules in cells to disruption by col-chicine. It will be of great interest to characterize the colchi-cine-binding reaction of plant tubulin in greater detail, in or-der to compare binding kinetics, decay rates, and affinityconstants for each different tubulin.Both the drug-binding and immunological properties of tu-

bulin are a function of its structure at some level. Thus, dif-ferences in antibody or colchicine binding between the vari-ous species examined here reflect interesting structural dif-ferences between tubulins. Although the full extent andsignificance of the differences are not revealed by thesestudies, it is clear that there are subtle structural differencesamong the tubulins of the plants examined, and even greaterdifferences between them and mammalian brain tubulins.Further studies in which the amino acid sequences of severalplant tubulins are compared with each other and with thoseof mammalian brain tubulin could be expected to show re-gions of similar structure that would be central to the com-mon functions of all microtubules as well as the dissimilarregions responsible for the observed divergence.

We wish to thank Drs. G. Gutman and G. Granger for their helpfulsuggestions on the antibody preparations, and Robert Yamamotoand Richard Cyr for technical advice and assistance. We are gratefulto Dr. L. Wilson for the generous gifts of sea urchin and Chlamydo-monas tubulin samples. We also extend our appreciation to Dr. H.Koopowitz for the use of the Apple II Computer and to HoangNguyen for help with the curve integration program. This work wassupported by grants from the Monsanto Co. and from the NationalScience Foundation (PCM 80-16035).

1. Gunning, B. E. S. & Hardham, A. R. (1982) Annu. Rev. PlantPhysiol. 33, 651-698.

2. Valenzuela, P., Quiroga, M., Zoldivar, J., Rutter, W. J.,

Kirschner, M. W. & Cleveland, D. W. (1981) Nature (Lon-don) 289, 650-665.

3. Ponstingl, H., Krauhs, E., Little, M. & Kempf, T. (1981) Proc.Natl. Acad. Sci. USA 78, 2757-2761.

4. Krauhs, E., Little, M., Kempf, T., Hofer-Warbinek, R., Ade,W. & Ponstingl, H. (1981) Proc. Natl. Acad. Sci. USA 78,4156-4160.

5. Koehn, J. A. & Olsen, R. W. (1978) Biochem. Biophys. Res.Commun. 80, 391-397.

6. Rikin, A., Atsmon, D. & Gitler, C. (1982) Planta 154, 402-406.7. Lloyd, C. W., Slabas, A. R., Powell, A. J., MacDonald, G. &

Badley, R. A. (1979) Nature (London) 279, 239-241.8. Wick, S. M., Seagull, R. W., Osborn, M., Weber, K. & Gun-

ning, B. E. S. (1981) J. Cell Biol. 89, 685-690.9. DeMey, J., Lambert, A. M., Bajer, A. S., Moeremans, M. &

DeBrabander, M. (1982) Proc. Natl. Acad. Sci. USA 79, 1898-1902.

10. Morgan, J. L., Holladay, C. R. & Spooner, B. S. (1978) Proc.Natl. Acad. Sci. USA 75, 1414-1417.

11. Kowit, J. D. & Fulton, C. (1974) J. Biol. Chem. 249, 3638-3646.

12. Little, M., Luduefia, R. F., Langford, G. M., Asnes, C. F. &Farrell, K. (1981) J. Mol. Biol. 149, 95-107.

13. Morejohn, L. C. & Fosket, D. E. (1982) Nature (London) 297,426-428.

14. Little, M., Luduefia, R. F., Keenan, R. & Asnes, C. F. (1982)J. Mol. Evol. 19, 80-86.

15. Little, M., Luduefia, R. F., Morejohn, L. C., Asnes, C. F. &Hoffman, E. (1984) Origins Life 13, in press.

16. Hart, J. W. & Sabnis, D. D. (1976) Curr. Adv. Plant Sci. 26,1095-1104.

17. Nickell, L. G. & Tulecke, W. (1959) Bot. Gaz. 120, 245-250.18. Fosket, D. E., Morejohn, L. C. & Westerling, K. E. (1981) in

Metabolism and Molecular Activities of Cytokinins, eds.Guern, J. & Pdaud-Lenoel, C. (Springer, Berlin), pp. 193-211.

19. Murashige, T. & Skoog, F. (1962) Physiol. Plant. 15, 473-497.20. Weisenberg, R. C., Borisy, G. G. & Taylor, E. W. (1968) Bio-

chemistry 7, 4466-4479.21. Weisenberg, R. C. & Timasheff, S. N. (1970) Biochemistry 9,

4110-4116.22. Lee, J. C., Frigon, R. P. & Timasheff, S. N. (1973) J. Biol.

Chem. 248, 7253-7262.23. Lee, J. C., Tweedy, N. & Timasheff, S. N. (1978) Biochemis-

try 17, 2783-2790.24. Sloboda, R. D., Dentler, W. L. & Rosenbaum, J. L. (1976)

Biochemistry 15, 4497-4505.25. Studier, F. W. (1973) J. Mol. Biol. 79, 237-248.26. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254.27. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl.

Acad. Sci. USA 76, 4350-4354.28. Hudson, L. & Hay, F. C. (1976) Practical Immunology (Black-

well, Oxford), pp. 162-164.29. Detrich, H. W., III, Williams, R. C., Jr., & Wilson, L. (1982)

Biochemistry 21, 2392-2400.30. Schiff, P. B., Fant, J. & Horwitz, S. B. (1979) Nature (Lon-

don) 277, 665-667.31. Gaskin, R., Cantor, C. R. & Shelanski, M. L. (1974) J. Mol.

Biol. 89, 737-758.32. Lambeir, A. & Engelborghs, Y. (1981) J. Biol. Chem. 256,

3279-3282.33. Pfeffer, T. A., Asnes, C. F. & Wilson, L. (1976) J. Cell Biol.

69, 599-607.34. Luduefia, R. F. & Woodward, D. 0. (1973) Proc. Natl. Acad.

Sci. USA 70, 3594-3598.35. Silflow, C. D. & Rosenbaum, J. L. (1981) Cell 24, 81-88.36. Gozes, I. & Barnstable, C. J. (1982) Proc. Natl. Acad. Sci.

USA 79, 2579-2583.37. Piperno, G. & Luck, D. J. L. (1977) J. Biol. Chem. 252, 383-

391.38. Cleveland, D. W., Lopata, M. A., Sherline, P. & Kirschner,

M. W. (1981) Cell 25, 537-546.39. Chang, M. T., Kilmartin, J. V. & Dove, W. F. (1981) J. Cell

Biol. 91, 333a (abstr.).40. Asai, D. J., Brokaw, C. J., Thompson, W. C. & Wilson, L.

(1982) Cell Motil. 2, 599-614.

Proc. NatL Acad Sd USA 81 (1984)

Dow

nloa

ded

by g

uest

on

June

17,

202

1