OOZO-7519/88$3.00+0.

00
Pergumon Presspic
0 1988 Ausrmhn SocietyforParasirologv.

THE ROLE OF THE CYTOSKELETAL PROTEIN, TUBULIN, IN THE

MODE OF ACTION AND MECHANISM OF DRUG RESISTANCE TO
BENZIMIDAZOLES

E. LACEY

CSIRO Division of Animal Health, McMaster Laboratory, Private Bag No.1, Glebe. NSW 2037, Australia

(Received 30 May 1988; accepted 22 June 1988)

CONTENTS
INTRODUCTION 886
MICROTUBULES 886
Introduction 886
Structure of microtubules 886
Endogenous co-factors 888
Non-tubulin protein regulation 890
Microtubules as an integral component of the cell 890
Isolation of tubulin 892

METHODOLOGY FOR INVESTIGATION OF MKROTUBULE INHlBlTORY ACTlVlTY 894

Polymerization/depolymerization 894
Ligand binding 895
GTPase activity 895
Alkylation 895
Bis-alkylationiaberrant isoforms 895
Growth toxicity 896

BIOCHEMICAL PHARMACOLOGY OF ~MMALIAN BRAIN TUBULIN 896

Non-selective microtubule inhibitors 896
Selective microtubule inhibitors 898
Vinblastine binding site 898
Tax01 binding site 899
Colchicine binding site 901
Colchicine 901
Podophyllotoxin 903
Benzimidazotes 903
Other CLC-site inhibitors 907

PHARMACOLOGY OF BENZlMIDAZOLE ACTION IN HELMINTHS 908

General in viva aspects of benzimidazole pharmacology 908
In vitro benzimidazole pharmacology 910
Biochemical pharmacology of benzimidazoles 911

BENZIMIDAZOLE-TUBULIN INTERACTION IN HELMINTH RESISTANCE AND SPECIES

SPECIFICITY 915

BENZIMIDAZOLE-RESISTANCE: GENETIC MUTANTS, MODELS AND PARASITES 918

THE NATURE OF MICROTUBULE INHIBITOR INTERACTION AT THE COLCHICINE

BINDING SITE 919
General binding site hypothesis 922
Location of the colchicine site 926
Functions of the colchicine site 926

CONCLUSION 927

ACKNOWLEDGEMENTS 928

REFERENCES 928

885
886 E.LACbY

INTRODUCTION single most important limitation to their continued

IT is now 27 years since the introduction of thiabenda- use, drug resistance.
zole (TBZ) as a safe broad-spectrum anthelmintic
(Brown, Matzuk, Ilves, Peterson, Harris, Sarett, Eger- MICROTUBULES
ton, Yakstis, Campbell & Cuckler, 196 1). This more Introduction
than any other drug heralded the commencement of The existence of microtubules in eukaryotic cells
our modern concepts of parasite chemotherapy, has been sporadically recorded over the last century,
establishing many of the toxicological and efficacy however their intensive study in cell biology is relat-
criteria for subsequent developments. The intensive ively recent. The term ‘microtubules’ was only coined
investigations which followed in the 1960s and 1970s in 1961 (see Dustin, 1984). Since this time our
led to the development of a series of benzimidazoles knowledge of the role of the self-association equilib-
(BZ) and BZ prodrugs as anthelmintics (Fig. 1). Each rium between the soluble sub-unit tubulin and the
drug represented an improvement in efficacy and insoluble polymeric microtubule has undergone
spectrum of activity which firmly established the explosive growth, although our knowledge of the
current predominance of BZs within our chemothera- involvement of microtubules in many cellular events is
peutic arsenal. still evolving. While the role of microtubules within
These developments proceeded almost entirely by the mitotic spindle is fundamental and ubiquitous to
classical in vivo parasitological screening and our all cell division in eukaryotes, it is perhaps the subtler,
understanding of how BZs exerted their activity metabolic effects of these cytoskeletal proteins that
remained largely unexplored. By the late-1970s the control by direct or indirect means the primary
biochemical pharmacology of BZs consisted largely functions essential to cell homeostasis.
of apparently unrelated in vitro biochemical studies The key to much of our knowledge of tubulin relies
for individual members of this class. Considering on its interaction with the alkaloid, CLC. Proof of
these studies and the concept of mechanism of action, CLC binding was a primary criterion for the presence
the question is raised as to how a closely related class of tubulin and the involvement of tubulin within
of compounds can exert such mechanistic diversity cellular events. Although the scope of methods for
and yet display such similarities in toxicity and identifying tubulin has broadened with the discovery
efficacy. of other microtubule inhibitors (such as vinblastine
While our understanding of how BZs act at the bio- (VBL), PDT and taxol) and immunological and
chemical level and the implications of that action is genetic approaches, CLC is still commonly used in
still far from complete, the concept of the primary site assessment of potential microtubule involvement.
of BZ interaction within the microtubule matrix in Indeed, no other soluble protein has been so heavily
eukaryotic cells is now firmly established. This hypo- dependent on the use of pharmacological probes in
thesis, originating with the studies of Borgers & De uncovering its biochemistry.
Nollin (1975) stimulated great interest in both the in Several excellent reviews have been published as
vitro and in viva actions of BZs on the tubulin- monographs (Dustin, 1978,1984; Roberts & Hyams,
microtubule equilibrium from the mid-1970s to the 1979) and as conference proceedings (De Brabander
present day. &r De May, 1980; Soifer, 1975, 1986) covering
This review initially focuses on the more general general aspects of microtubule biochemistry. More
aspects of microtubule biochemistry and pharmacol- specific reviews on microtubule inhibition (Wilson,
ogy in eukaryotic organisms. Much of the literature 1975; Schiff & Horwitz, 1981; Mareel & De Mets,
reviewed refers to non-helminth species. This broad 1984) microtubule assembly (Timasheff & Grisham,
approach is essential if we are to understand the 1980; Purich & Kristofferson, 1984) microtubule
significance of a microtubule-based mode of action mutants (Oakley, 1985) and molecular biology
for BZs, the origin of their species specificity and their (Cleveland & Sullivan, 1985) also serve to provide
observed pharmacology in helminths. The early excellent background.
observations that BZs bind to the same site on tubulin
as many well recognized inhibitors of microtubule Structure of microtubules
activity such as colchicine (CLC), podophyllotoxin The microtubule is, as its name suggests, a hollow
(PDT) and their respective analogues, further justify tubular organelle, approximate lumen diameter
this approach. Consequently, this review attempts to 15 nm and outer diameter 25 nm, of variable and
draw together the anthelmintic pharmacology and transient length. By electron microscopy the microtu-
mode of action of BZs within the context of the more bule is comprised of a series of 13 protofilaments.
general role of microtubule inhibitors culminating in a This arrangement, while characteristic of mammalian
general hypothesis of the nature of the CLC/BZ/PDT microtubules in vivo, is dependent on species and in
binding site (CLC-site). It is hoped by this approach to vitro assembly conditions (see Dustin, 1984). In
the mode of action of BZs that the review may also nematodes, protofilament number has been charac-
serve as a basis for future anthehnintic development terized for the free-living species Cuenorhabditis
and address the potential of BZs to overcome the elegans by Chalfie & Thompson (1979) and the
 Role of tubuiin in benzimidazoles mode of action 887

(a) Benzlmldazoles

Drug Abbrev. -R Drug Abbrev -R

CMbl?PllZlM nac H
flebendarole tlez

Parbendirnln JBZ
Nocodazo!~ ND2
Oxrbendatcle U%Z

Rlbcndazole RBZ Flubendazole FL%2 -C C6H,-4-F

ti
Ryrobrndazole RN2 Fenbendatole FE2 -S-&H,

DxFendazole OF2 -$-C&

Csciobcnd&?ole CIBZ tl
-clso,-C&,X,-4-F

Thlabcndazote TN2 H Canbendatole CBZ.

(b) Pro-drugs

Benoay 1 Febantel

Thiophanate Netoblmln

FIG. 1. Structures of (a) benzimidazoles and (b) benzimidazole pro-drugs developed for experimental or commercial use.
888 E. LACEV

parasites Ascaridia gulli and Tric~ostron~~Lu~colu- Both a - and fi-tubulin are heterogeneous and can
briformis by Davis & Gull (1983). In these studies an be partially resolved by PAGE with the isolation of
11-protofilament array was most commonly various a ,, a,. . . and p,, /!I2 , proteins (Little &
observed. C. elegans also contains a 15unit array Luduena, 1985). The extent of variability is again
unique to six touch receptor neurons, while T. colu- dependent on the source of tubulin. Under isoelectro-
briformis possesses an additional 12-unit array with focusing, the extent of microheterogeneity is more
similar cellular distribution to the 11 protofilament apparent (Wolff, Denoulet Bi Jeantet, 1982). The
arrays. Microtubule protofilament stability (11 vs 15) precise nature and roles of isotubulins have not yet
in C. eZeg&t.scan be correlated to differential drug, been fully elucidated, however several groups have
temperature and fixation effects (Chalfie & Thomp- demonstrated that many of tubulin’s in v&o functions
son, 1979), however, the functional significance of are to varying degrees iso-form specific (Gundersen,
these differences has yet to be confirmed in other Kalnoski & Bulinski, 1984). Whether this implies
species. unique or multiple functional roles for isotubulins
The microtubule is in dynamic equilibrium, under- within the eukaryotic cell is equivocal. The recent
going a ‘treadmilling’ addition and subtraction of work of Lewis, Gu & Cowan (1987) and Lopata &
soluble tubulin units at opposite ends of the develop- Cleveland (1987) suggests that while isotubulin
ing tubule. The state of growth is thus delicately poised specific functions exist, communal association in the
between assembly and disassembly by changes to the microtubule matrix occurs.
rates of subunit addition or loss. In vitro the energy Limited proteolysis of both a- and ~-tubulins by
change required to perturb this equilibrium can be trypsin, chymotrypsin and other proteases reveals a
supplied by temperature variation between 0” (dis- consistent sensitivity to cleavage within narrow
assembly) and 37°C (assembly). Essential to this structural domains. For a -tubulin, trypsin cleavage
process is the concentration of tubulin itself since a results in the formation of a 34-38,000 mol. wt NH?-
minimum critical concentration must be present for terminus peptide with a 14,000-16,000 residue.
assembly to occur. The critical concentration is not an Chymotrypsin preferentially cleaves /3-tubulin to
invariable factor and depends on the type of tubulin yield a 32,000-34,000 NH,-terminus peptide
and in vitro solution variables (buffer type, pH, ionic together with a 19,000-20,000 peptide (Brown &
strength, co-factors and purity). In a more general Erickson, 1983; Serrano, Avila & Maccioni, 1984;
sense critical concentration can be altered by any Serrano 8: Avila, 1985; Sackett & Wolff, 1986).
factor capable of binding or stabilizing either the Estimates of the size of these peptides vary between
microtubule or tubulin subunits. On disassembly, the laboratories, probably due to differences in electro-
microtubule dissociates to dimeric tubulin with phoretic conditions (Clayton et al., 1980). Both a-
variable concentrations of tubulin oligomers. Oligo- and @-tubulin are cleaved by proteases at a second site
meric tubulin is essential to the in vitro initiation of to yield a 1000-2000 mol. wt COOH-terminus
microtubule assembly and is thought to function as a peptide. Interestingly, when this fragment is speci-
‘seed’ for growth. Removal of oligomers renders fically cleaved by subtilisin from both Q- and ,5-
normal thermodynamic control of the equilibrium tubulin, the dimer, named tubulin-S, possesses a
inoperable in the absence of alternative microtubule critical concentration for polyme~zation of
stabilizing agents, such as glycerol, taxol, dimethyl- 0.1 mg ml-’ (compared with 2 mg ml-’ for tubulin)
sulphoxide (DMSO) and glutamate. indicating the important regulatory role of this
The dimer is comprised of closely related proteins, 2000 mol. wt COOH-terminus in modulation of the
a- and /I-tubulin, of approximate molecular weight tubulin-microtubule equilibrium (Bhattacharyya,
50,000 and sequence lengths of 450 and 445 amino Sackett & Wolff, 1985). These results support the
acids, respectively, for pig brain tubulin. Structural hypothesis that both tubulins contain two major
homology between the proteins is normally >40%; subunits joined by a relatively exposed sequence
both are highly acidic (p1 4-6) with 40% of the final accessible to proteases together with a polar COOH-
30 COOH-terminus residues acidic (Ponstingl, terminus (Fig. 2). Limited proteolysis represents a
Krauhs, Little, Kempf, Hofer-Warbinek & Ade, major strategy for the localization of binding domains
1982). Although tubulins isolated from most species of ligands such as co-factors, microtubule inhibitors,
fall within these broad parameters, considerable size, endogenous binding proteins and antibodies (Serrano
homology and acidity heterogeneities exist between et al., 1984; Bhattacharyya et al., 1985; Serrano,
tissues, isolates and species. Separation of cx- and /3- Wandosell & Avila, 1986; Sackett, Bhattacharyya &
tubulin (and their definition) is achieved by polyacryl- Wolff, 1986). The structural relationship between
amide gel electrophoresis (PAGE) after reduction microtubules and tubulin is schematically presented
and derivatization, although this last step is no longer in Fig. 2.
considered essential (Clayton, Quinlan, Roobot,
Pogson & Gull, 1980). By electrophoresis, a - and /?- Endogeaous co-~ct0r.s
tubulin mobility is more characteristic of 55,000 and Modification of the tubulin-microtubule equilib-
53,000 mol. wt. proteins, respectively, than the calcu- rium in vivo is clearly not dependent on temperature
lated 50,000. fluxes but is regulated by a range of endogenous
 Role of tubulin in benzimidaroles mode of action 889

I
, 4nm ,
-1 --I I

FIG. 2. The structure of microtubules. Enlargement of a portion of a typical immunofluorescence pattern observed with
tubulin antibodies of the cellular microtubular matrix shows the relationship between the soluble dimers and oligmers and the
polymeric microtubule. PAGE and IEF demonstrate the ~croheterogene~~ of tubulin, while protease treatment with C
(chymot~in), T (trypsin) and S (sub&sin) reveals the major structural domains within each monomer as proposed by
Sackett & Wolff (1986).
890 E.LACEY

co-factors which control the status of the tubulin oped from the sequence homologies of other GTP/
dimer and thus the required critical concentration for Mg?+ dependent proteins (Steinlicht ef al., 1987).
polymerization. Other cations such as Zn2+, Co2+, Mn2+, NH:, Na+,
Central to initial isolation procedures for native Kf and Li+ and various anions alter the in vitro
tubulin was the discovery of a requirement for added dynamics and stability of microtubule formation
nucleotide, specifically GTP, for both stability and (Gaskin & Kress, 1977; Buttlaire, Czuba, Stevens,
self-assembly. The dimer contains two GTP binding Lee & Himes, 1980; Himes, Lee, Eagle, Haskins,
sites denoted as the E-site (exchangeable) and N-site Babler & Ellermeier, 1982; Bhattacharyya & Wolff,
(non-exchangeable), reflecting their relative affinity 1976; Lee & Timasheff, 1977; Arakawa & Timasheff,
for exogenous GTP. The E-site is rendered non- 1984; Hamel & Lin, 1981a; Croom, Correia &
exchangeable when incorporated into the microtu- Williams, 1986). However, such activity does not
bule on polymerization. The requirement for GTP appear to be relevant to endogenous modulation.
appears essential for all tubulins regardless of species
and is probably an evolutionary stable binding Non-tub&n protein regulation
domain within tubulin. GTP can be substituted by Extraction of microtubule matrix from cells reveals
other GTP analogues and other nucleotides (ATP, an abundance of proteins which interact with either
UTP and CTP) which are of lower affinity (Penning- tubulin or microtubules. The majority of these are
roth & Kirschner, 1977, 1978). E-site bound GTP is microtubule-associated proteins (MAPS) which act to
incorporated into the growing microtubule and is stabilize microtubule formation while also binding to
subsequently hydrolysed from GTP to GDP (GTPase tubulin to form the soluble oligmeric component.
activity). GTPase activity is extensively modulated by Characterization of MAP interactions by in vivo and
microtubule inhibitors (Lin & Hamel, 1981). The in vitro biochemical and immunological techniques
precise role of GTPase activity in microtubule supports the hypothesis that MAPS are functionally
dynamics is under investigation (Pantaloni & Carlier, specific to microtubules. Several inhibitory proteins
1986; Caplow, 1986), however. the effectiveness of have also been identified (Table 1). Within this group,
non-hydrolysable GTP analogues in polymerization some proteins demonstrate direct inhibition of [‘HI
studies suggests GTPase activity is non-essential to CLC binding (Lockwood, 1978; Sherline, Schiavone
the polymerization process (Penningroth & Kirsch- & Brocato, 1979; Gupta & Gupta, 1984) while other
ner, 1978). In vitro, the role of GTP in polymerization proteins have been shown to interfere with drug
can be over-ridden by the use of non-selective micro- induced changes, probably indirectly, by inducing
tubule stabilization reagents such as glycerol and conformational changes (Fellous, Luduena, Prasad,
sucrose (Shelanski, Gaskin & Cantor, 1973). Jordan, Anderson, Ohayon & Smith, 1985).
Localization of the E-site has been attempted by MAPS and tau proteins appear to bind to the
photo-affinity labelling of GTP analogues with COOH-terminus of P-tubulin between residues 392-
varying degrees of success (Nath, Eagle & Himes, 445, particularly the heptapeptide region 434-440.
1986). The recent use of direct photo-affinity labelling Weak interaction to tau proteins but not MAPS occurs
with GTP suggests the site is located on /I-tubulin in the parallel a sequences (Littauer, Giveon, Thier-
(Hesse, Marnta & Isenberg, 1985; Nath et al., 1986). auf, Ginsburg & Ponstingl, 1986). Although well
Both (x - and /3-tubulin contain glycine-rich clusters at characterized in mammalian systems, little is known
residues 140-146 consistent with the phosphate- of the nature of MAPS and related proteins in lower
binding regions of other nucleotide binding proteins eukaryotes. In general, these proteins should be
(Ponstingl et al., 1982). A more general model for considered as examples of types of in vivo regulating
both GTP sites encompasses the involvement of four actions within the microtubule matrix since it is
‘loops’ between B-pleated sheets and a-helices unlikely that they are the sole modulators in view of
involving amino acid residues from -60 to 300 our relatively limited knowledge of species diversity
(Sternlicht, Yaffe & Farr, 1987). (Roobol, Pogson & Gull, 1980).
Tubulin contains single high affinity sites for both Figure 3 depicts the sites and relationships of these
Ca’+ and Mg?+. In vitro these ions show opposing proteins to the tubulin-microtubule equilibrium.
actions; Mg I+ is required for assembly while Ca2+,
either free or complexed with the 14,000 mol. wt Ca’+ Microtubules as an integral component of the cell
binding protein, calmodulin, inhibits assembly and To fully appreciate the mechanisms of microtubule
induces disassembly. inhibition, it is necessary to consider the nature of the
Localization of the Ca?+ high (and low) affinity tubulin-microtubule equilibrium not just in bio-
site(s) on the COOH-terminus of both a- and p- chemical isolation but also as part of the eukaryotic
tubulin has been achieved by limited proteolysis by cell and the whole organism. As will become apparent,
subtilisin, trypsin and chymotrypsin (Serrano, Valen- it is the cellular implications of the drug-tubulin
cia, Caballero & Avila, 1986). Biochemical data complex that are the most poorly understood and yet
suggest that the Mg2+ site is proximal to either of the most extensively studied aspect of microtubule in-
GTP sites (Correia, Baty & Williams, 1987) which is hibitors.
consistent with the GTP-site model for tubulin devel- A number of discrete functions are ascribed to
 TABLE~-PROTEINSASSOCIATED WITH THE EQUILIBRIA
TUBULIN-MICROTUBULE

Source/ Molecular
Protein distribution weight Action References

MAPla,b,c neuronal* - 350,000 stabilize microtubule and oligomers see Vallee, Bloom & Luca, 1986
MAPZa,b neuronal -280,000 stabilize microtubule and oligomers see Valiee, Bloom & Luca, 1986
MAP3a,b neuronal 180,000 stabilize microtubule and oligomers Matus & Riederer, 1986
MAP4 neuronal and non-neuronal 215,000 stabilize microtubule and oligomers Olmsted, Asnes, Parysek, Lyon &
Kidder, 1986
MAP5 neuronal/post-natal - 320,000 stabilize microtubule and oligomers Matus & Riederer, 1986
Tau (several) primarily neuronal 55,000-68,000 stabilize microtubule and oligomers see Binder, Frankfurter & Rebhun,
1986
non-neuronal 42,000, cross react with tau antibodies, Drubin, Kobayashi & Kirschner,
125,000,200,000 function unknown 1986
TAP (Tubutin CHO cells? 68,000 binds tubulin, acts as an elongation Lockwood. 1978
Assembly Protein) factor
Chartins neuronal/non-neuronal 64,000-80,000 uncharacterized Magendantz & Solomon, 1985
CSFs (Cold Stabil- neuronal 35,000,64,000, induce cold ~crotubuie stability, Margolis & Rauch, 1981; Piroliet,
izing Factors) 67,000 microtubule inhibitor insensitive Job, Fischer & Margolis, 1983
STOP neuronal - 145,000 similar to CSFs Margolis, Job, Pabion & Rauch,
1986
250 K protein neuronal 250,000 inhibits CLC binding and polymer- Sherline et al., 1979
ization
Tubulin binding neuronal <4000,10,000- inhibits CLC binding and polymer- Lockwood, 1978
factor-1,2,3 15,000,25,000- ization
40,000
Pl,P2 CHO cells - 70,000 modulates CLC binding Gupta & Gupta, 1984
33 K protein neuronal 33,000 inhibits polyme~zation, induces Kotani, Muro~shi, Nishida & Sakai,
partial depolymerization 1984
Substance P neuronal -1000 inhibits polymerization binds to Maccioni, Cann & Stewart, 1986
COOH-terminus

*For most of these proteins the tissue distribution has not been extensively characterized outside of neuronal sources, little
information available.
tCh.inese hamster ovary cells.
892 E. LACEY

nicrotubule

FIG. 3. Interactions of associated proteins with the tubulin-microtubule equilibria. Modified from Lockwood (1979).

microtubules at the celluiar Level: (i) formation of the logical probes since it is from inhibitors like CLC,
mitotic spindle in cell division; (ii) maintenance of cell VBL and PDT and non-inhibitors such as lumicolchi-
shape; (iii) ceti motility; (iv) cellular secretion; (v) tine, colchiceine and isocolchicine that much of our
nutrient absorption; and (vi) intracellular transport. knowledge of microtubule involvement in cellular
These functions are well documented by Dustin function is derived. The use of these approaches is
(1984) and references cited therein. Intracellular based on two premises: (i) that the drugs act only by
association of either tubulin or microtubules include microtubule inhibition; and (ii) that in vitro microtu-
most organelles: mitochondria, Golgi apparatus, ribo- bule activity reflects the nature and extent of the drug-
somes, lysosomes. cell membranes and, of course, the tubulin interaction affecting particular cellular
nucleus (Dustin, 1984). It is via these interactions and functions in vivo.
the probable direct association of tubuiin and micro- It is apparent that only the first premise is applied in
tubules with physiological mechanisms of hormone, cases where both inhibitors and non-i~ibitors are
neurotransmitter, nutrient, enzyme and receptor active (Mizel & Wilson, 1972; Loike & Horwitz,
action that the cellular responses of microtubule 1976), that is, the mechanism is considered to be non-
inhibition are manifested (Mizel & Wilson, 1972; microtubule dependent since the first premise is
Loike & Horwitz, 1976; Azhar, Hwang & Reaven. violated. Consideration of the second premise would
1983; Crie, Ord, Wakeland & Wildenthal, 1983; Do bring into question whether altered specificity of
Khac, Tanfin & Harbon, 1983; Faulkner, Henderson microtubules within the mechanism had occurred.
& Peaker, 1984; McKay, Aronstam & Schneider, Since the roles of tubulin can be considered as
1985; Durrieu, Bernier-Valentin & Rousset. 1987; either isotype specific or communal, extrapolation of
Kar~off-Schweizer & Knull, 1957). drug action from a brain tubulin pool as in pofymer-
The disruption of the tubulin-microtubule equilib- ization or [jH] CLC binding studies may not neces-
rium can be seen as leading to a cascade of direct and sarily be indicative of individual isotubuiin drug
indirect biochemical/physiological changes resulting sensitivities. Thus, in situations where atypical activity
in the loss of cellular homeostasis. Conditions of profiles for inactive in vitro inhibitors arise, the
cellular ‘disequilibrium’. if maintained, are lethal. conclusion of a mechanism as microtubule indepen-
Lethal effects are most evident in actively dividing or dent may not be valid.
growing cells. While it is postulated that such cells
require greater utilization of tub&n and are therefore
more sensitive to microtubule inhibitors, it is also
probable that it is the disruption to the timing of Methods for the purification of native tub&n from
critical physiological events (such as cell division) both neuronal and non-neuronal sources have been
which is lethal. A study of the cellular role of micro- reviewed (Farrell, 1982; Vallee, 1986). Strategies for
tubules is heavily dependent on the use of pharmaco- the isolation of tubulin rely on three basic charac-
 Role of tubulin in benzimidazoles mode of action 893

teristics: (1) temperature-dependent cycles of general applicability to differing tubulin sources

polymerization/depolymerization; (2) interaction (Farrell, 1982).
with basic ligands; and (3) specific ligand affinities. Isolation of tubulin from non-neuronal but particu-
In general, the use of temperature-cy~l~g is limited larly non-mammalian sources where tubulin normally
to rich tubulin sources (such as brain) where the criti- represents less than 1% of total soluble protein
cal concentration of tubulin is obtained in the crude (compared with brain at 20%) has met with consider-
supernatant or can be obtained by concentrating the able difficulties. Very few non-mammalian tubulin
protein. Many permutations of the solution conditions purification protocols have been applicable to other
have been reported since initial publication by organisms, with most methods yielding only partially
Shelanski et al. (1973). Some of the more radical purified preparations. Recently, a general two-step
modifications reported utilize glutamate (Hamel & protocol for isolating plant tubulin by use of DEAE-
Lin, 1981b), DMSO (Himes, Burton & Gaito, 1977) Sephadex AS0 and ammonium sulphate precipitation
and, more recently, taxol (Collins & Vallee, 1987). was reported (Morejohn & Fosket, 1982). To date
The use of ion-exchange chromatography with purified tubulin has been isolated from rose, carrot
basic Iigands is surprisingly the most underdeveloped and hibiscus by this method (Morejohn, Bureau,
of the isolation techniques. Little manipulation of Tocchi & Fosket, 1984).
diethyl aminoethyl (DEAE) linked gels has been In parasitic helminths partial purification of tubulin
attempted since its introduction (Weisenburg, Borisy from Hymenolepis diminutu has been achieved by a
&Taylor, 1968) with tubulin absorbed either by batch DE-52 cellulose batch technique (Watts, 1980,
or column techniques and eluted by linear or step- 198 1). The alternative column approach with DEAE
wise gradients of high ionic strength buffer. Tubulin is sephadex has been used for partial purification of
eluted at 0.5-0.8 M-NaCI or KCI. In a recent study. A. galfi tubulin (Ireland, Clayton, Gutteridge, Pogson
Lacey & Snowdon (submitted for pubIication) & Gull, 1982), a method analogous to that reported
demonstrated that by choosing less hydrophobic for the fungus Asperg~~~l~ n~d~Zans (see Davidse &
amino Iigands such as polylysine, aminoethyl, lysine Flach, 1977). This step was incorporated in a three-
and arginine, tubulin was purified without exposure to step total purification of A. gaili tub&n (Dawson,
high salt concentrations, a factor known to induce Gutteridge & Gull, 1983) in which tubulin was eluted
partial denaturation of tubulin (Croom et al., 1986). from DEAE-sephadex with salt (0.4 M), concentrated
Isolation of tubulin by ligand affinity is based on the by ultrafiltration, centrifuged to remove actin fila-
specificity of tubulin’s interaction with a range of small ments and polymerized using DMSO. Complete
molecular weight compounds, for example, CLC purification was achieved by temperature-dependent
affinity gels (Hinman, Morgan, Seeds & Cann, 1973) cycling with DMSO-assisted polyme~zation steps.
VBL precipitation (Bryan, 1971) and proteins such as Modification of this technique has been reported to
lactoperoxidase (Rousset & Wolff, 1980). tubulin yield pure tubulin from Acuris swim. Fusciolu hepa-
antibodies (Ikeda & Steiner, 1976) and tubulin itself ticu and C. eleguns (see Gull, Dawson, Davis &
(Lockwood, 1978). Recent studies by Kocha, Fukuda, Byard, 1987). However, this strategy is not without
Isobe & Okuyama (1986) suggest that CLC affinity limitations since it combines both the disadvantages of
columns may represent a combination of specific polymerization and DEAE techniques with DMSO-
affinity and non-specific hydrophobic interactions, induced denaturation (Lacey & Prichard, 1986).
since CLC can be replaced with a variety of hydro- While of little consequence for polymerization studies
phobic ligands with similar purification being since inhibition is sub-stoichiomet~c, use of tub&n
achieved. purified by this method for drug binding studies and
Other adjuncts to these methods for initial and/or isotubuhn characterization should be viewed with
final purification steps include (NH&SO, precipita- caution.
tion, phosphocellulose chromatography and Mg2+ In this laboratory, purification has been achieved by
precipitation. combining polar amino acid ligands (arginine and
Several limitations to the above methods should be polylysine) and hydroxyappatite chromatography.
noted. Using temperature cycling, polymerization and The isolation conditions for F. heputicu are shown in
isolation of tubulin can be achieved in the presence of Fig. 4 (Gill & Lacey, unpublished results). By combin-
variable quantities of MAPS (6-20%), however, only ing arginine and hydroxyappatite cohunns, greater
tubulin above the critical concentration is removed at than 90% pure tubulin was obtained from F. he~utica
each cycle while sizeable losses of cold-stable micro- and &ionie&z expansa. For nematodes, low tubulin
tubules occur during depolymerization on ice recovery at 2.5% NH,SO, on arginine columns
(Margolis & Rauch, 3981). The use of DEAE occurred and polylysine was substituted for arginine.
chromatography yields essentially pure tubulin but Tubulin isolated from Dirojiluria immitus, Ostertagiu
requires the use of DMSO for subsequent polymeriza- circumcinctu , T. colubriformis and Ijl. contortus
tion studies. Further, the salt concentrations required adults by this procedure contains impurities at
to elute tubulin impair [3H] ligand stoichiometric molecular weights of 63,000 and 85,000 which can
studies (Croom ef al., 1986). The major limitations of be removed by varying the initial buffer conditions
the specific ligand techniques relate to their lack of (use of high ionic strength buffers to prepare crude
894

(A) AMNINE-SEPHAROSE 4B (25 ml)

110
tool-l r20

@JM Mes ’ 0.4M Mes Eluant

FIG. 4. Purification of F, ~e~uf~~ff tubulin by (A) arginine ~~r~rna~o~ra~~~ (2.5 ml column) and s~bse~uent~~ (B)
~~droxyappa~te (t&m-gel HA, 5 rni cokmn). The &ted protein for each 4 mI fraction is presented as a histogram with the
presence of tuhuiin monitored by f3H]MBZ binding (- - -). Elution of fraction 11.from A with phosphate buffers at pH 6.5
(0.05-0.05 M) yielded >90% pure tubulin in fraction IO of (R), tubulin represents 0.1% of the total 100,000 g crude
supernatant applied to the column.

s~p~rnatant). Isofation by this technique requires less

than 5 h with recovery of X5% of the to&I tubuhn Tubuhn ip2v&o in the presence of CTP, Mg‘+ and a
pool without exposure to denaturing conditions, and Ca*+ chelater (normally EGTA) at pH 6.4-7.0
has been achieved using less than 10 mg of crude undergoes polymerization to form microtubules on
supernatant. warming to 37°C. The transition can be monitored by
spectrophotometry (350-400 nm), viscosity or cen-
METHODOLOGY FOR INVESTIGATION OF trifugation followed by protein estimation. Depending
M~CROTUBUL~ INHIBITORY ACTIVITY on the method of isolation (presence or absence of
Investigation of microtubule inhibitors utihzes a MA&f, poI~e~zation can be carried out in typicat
range of biochemical techniques to identify activity tubuhn buffers (0.025-0.1 M, MES, PIPES or PO:-)
and site of action. A review of these approaches or induced by glycerol, DMSO, I M glutamate or
provides a means of collating published results and taxol. Depolymerization is assessed by cooling to 4 “C.
thus classifying drug binding sites. Many variations of Drugs are added either in buffer or DMSO (maximum
several of these techniques have been published. concn. lo/s, unless DMSO induction is used. A typical
 Role of tubulin in benzimidazoles mode of action 895

inhibition profile for a spectrophotometric assay is &and binding

shown in Fig. 5. On warming, a lag time of 2-3 min is Methods for PHI-labelled ligand binding to crude
ubserved followed by a linear increase in absorbance or purified tub&in have been reported (Table 2).
(turb~dity~ which plateaus by about 15 min,Increitsing Charcoal extraction of unbound isotope is the most
inhibitor ~on&entra~ons progressively increase lag convenient of the techniques, however it is timited to
time and reduce both the rate of turbidity change and 13H] CLC in higher eukaryotes and f”H] BZCs in
the height of the plateau, helminths, fungi and some other lower eukaryotes.
A graph of per cent inhibition of either rate or For other ligands, charcoal treatment strips bound
extent of polymerization vs inhibitor concentration [‘H] ligand from tubulin either partially ([“HI PDT) or
enables interpolation of an 15(, value (that is, the completely (?H] VBL) (Cortese, Bhattacharyya &
inhibitor concentration required to inhibit rate or Wolff, 1977; Lacey Edgar & Culvenor, 1987).
extent by 50%)). Concentration-response plots are Desorption of bound ligand during boundifree
generally linear over l&85% inh~bjt~on~ Above 85% separation is reported with other techniques and is
inhibition turbidity changes are due to aggregation independent of the potency of the hgand as a micro-
independent of inhibitor concentration. tubule inhibitor. Xn such cases equilibrium get filtra-
Polymerization is very sensitive to all microtubute tion techniques are favoured.
inhibitors since inhibition requires only a small pro- Non-radioactive binding assays based on fluores-
portion of tubulin to be bound (sub-stoichiometric); cence enhancement (Garland, 1978), circular dichro-
however, it cannot distinguish the tightness of binding. ism (Yeh, Chrzanowska & Brossi, 1988) and
The technique is useful for determining whether spectrophotometry (Head, Lee, Field &r Lee, 1985)
structurat modifications have a favourable or un- have also been reported. Ligand binding requires only
favourable effect on activity but is poor for differen- small quantities of tubulin (0.5-50 ~cg), is indepen-
tiating potent inhibitors, since achieving the required dent of the purity of tub&in and is highly repro-
level of drug binding for inhibition tends to group ducible.
these compounds in an I,,, range of OS-5 HUM(de-
pending on technique). Both spectrophotometric and GTFa.w activity
viscosity detection methods require relatively large In 1.0 M glutamate, polymerization of tubulin is
quantities of tubulin (0.5-1.5 mgper sample) and thus coupled to [%I] GTP hydrolysis. In the presence of
are normally restricted to brain tubulin. For studies microtubule inhibitors the rate of hydrolysis to [3H]
where only limited quantities of tubulin are available GDP is altered (Lin & Hamel, 1981). To date all
(such as parasites) the ~e~t~~ga~on/prote~n assay is microtubule inhibitors tested affect this reaction by
preferred (Dawson, Gutteridge & Gulf, 1984). either enhancement (CLC, nocodazole (NDZ)) or
inhibition (VBL, maytansine (MTS) and PDT) (Table
3). The mechanism by which inhibitors act on GTPase
is unknown.

Alkylation
Microtubule inhibitors inhibit the alkylation of
tubulin by ]‘“CJ iodoacetamide (Roach, Bane &
Luduena, 2985). The ability to protect -SH moieties
is characteristic of many inhibitors but not structurally
related non-inhibitors. To date, the technique has onfy
been used for a limited number of compounds and
little information on general applicability can be
inferred.

Bis-al~ylation/aberrant isoforrns
When tubulin is alkylated with N;N’-ethylene
bis(iodoacetam~de), the formation of an abnormal fi-
isotubulin, p*, derived by a specific cross-link
between spatially proximal -SH moieties is observed
Timefm)
on PAGE. fn the presence of CLC or other com-
pounds which inhibit [3H]CLC binding the formation
Fto. 5. Inhibition of sheep brain tubulin polymerization by of j?* is inhibited while being enhanced by ligands
phomopsin A. 7,141of either DMSO (control) or various
inhibitor concentrations was added to the tubulin solution.
binding to the VBL-site (Roach & Luduena, 1984).
After mixing, the cuvettes were warmed to 37°C and the When alkylation is undertaken in the absence of
change in turbidity (absorbance) monitored. With increasing GTP, a second specific cross-link is formed with
drug concentration, a progressive increase in the onset time differing /3-tubuhn electrophoretic mobility, ,f3”~
and reduced rate of poiymerizat~on and extent (change in. Formation of @*is inhibited in the presence of VBL
absurbance at the plateau) are observed. and VBL-site hgands while being enhanced by the
896 E. LACEY

TABLE ~-COMPILATION OF LIGAND BINDING TECHNIQUES

Technique Ligand Comments References

1. Gel filtration
- column (standard) CLC Widely used in initial Wilson, 1970
studies
- column (Hummel & Dryer)* NDZ, GTP Not widely used Head et al., 1985
- Mini-column (Penefsky) GTP, VBL Useful for BZ-brain tubulin Zeeberg & Caplow, 1979
interaction
- Batch (Hirose & Kano) ND2 Useful for weak inter- Head et al., 1985
actions
2. DEAE paper Most ligands Applicable to all l&and Borisy, 1972
interactions
3. Charcoal absorption CLC, BZCs in Simple, highly reproducible Sherline, Bodwin & Kipnis,
lower eukaryotes only narrow ligand applica- 1974; Lacey & Prichard.
bility 1986
4. Fluorescence CLC, other colchi- Limited to tropolone CLC- Ray et al., 198 1;Bane el
noids site ligands al., 1984
Bis ANS Decrease at 335 nm emis- Prasad et al., 1986
sion, increase at 490 mn
DAPI Bonne et al., 1985
5. Circular dichroism CLC Specific to CLC analogues Detrich ef al., 198 1; Yeh et
al., 1988
6. Dialysis MBZ Not widely used Laclette et al., 1980

*Original method citation; details in Head el al. (1985); Zeeberg & Caplow (1979).
tlacey, unpublished observations.

presence of CLC-site analogues (Little & Luduena, with tubulin compared with other proteins. Non-
1985). selective inhibitors elicit many of the cellular effects
characteristic of microtubule inhibition such as meta-
Growth toxicity phase arrest, inhibition of cell growth, in vitro poly-
Microtubule inhibitors block the division of cells in merization and in some cases [“HI hgand binding, but
in vitro culture (Dustin, 1984). In itself the inhibition also exhibit activity against many other proteins at
of growth of organisms is in no way specific to micro- comparable concentrations. Such compounds include
tubule inhibitors but determination of dose-response alkylating agents (Roach & Luduena, 1984) acylating
profiles to derive LD,,, and ED,,, data is an important agents (Lee, Houston & Himes, 1976; Mellado, Slebe
aspect of structure-activity and resistance studies & Maccioni, 1982) thiols (Banerjee, Jordan &
(Quinn & Beisler, 1981; Lacey & Watson, 1985b; Luduena, 1985; Potchoo, Braguer, Peyrot, Chauvet-
Sheir-Neiss, Lai & Morris, 1978). In conjunction with Monges, Sari & Crevat, 1986) heavy metal ions,
techniques known to be microtubule-dependent, for organometals (Vogel, Margolis & Mottet, 1985;
example, mitotic index (see Dustin, 1984) specific Peyrot, Briand, Momburg & Sari, 1986) aldehydes or
morphological changes (Oakley, 1985) and immuno- ketones (McKinnon, Davidson, De Jersey, Shanley &
fluorescence of microtubule matrix (Rubino, Fiori, Ward, 1987; Boekelheide, 1987; Luduena, Roach.
Lubinu, Monaco & Cappuccinelli. 1983) tentative Trcka, Mallevais & MacRae, 1987) and quinones
identification of tubulin-dependent mechanisms can (O’Brien, White, Jacobs, Boder & Wilson. 1984; Epe,
be made. The importance of growth toxicity and Hegler & Metzler, 1987). Typically, such ligands are
immunofluorescence in assigning a microtubule- well recognized substrates for nucleophilic substitu-
related mechanism for a new inhibitor is elegantly tion by thiol, imino, amino and hydroxyl moieties of
described in the case of 2,4-dichlorobenzylthiocyan- amino acids such as cysteine, lysine, tryptophan and
ate (DCBT) where other in vitro techniques were serine residues which are involved in tubulin polymer-
initially unsuccessful (Abraham, Dion, Duanmu, ization
Gottesman & Hamel, 1986). A second less well-defined group of non-selective
microtubular inhibitors is constituted by basic com-
BIOCHEMICAL PHARMACOLOGY OF pounds. The affinity of amines for tubulin is well
MAMMALIAN BRAIN TUBULIN established (Lee, Tweedy & Timasheff, 1978), how-
Non-selective microtubule inhibitors ever it is not known whether they interact at defined
Microtubule inhibitors can be broadly subdivided sites analogous to those of MAPS within acidic
based on selectivity or non-selectivity of interaction residues. Ifr vitro, polyamines such as polylysine and
 lNHlBITORS~*
TABLE 3-ACTIVITY OF MlCROTUBULb

Dispia~ement
B* B” GTPase
I,,,(poly)§ PHI VBL [%j CLC [“HI GTP Akylation formation formation activity

VBL-siteligunds
Vinblastine 0.5.5-1.7 I E I
Vincristine 1.4 I E I
Maytansine 2.7 I E I
Phomopsin 0.56 - E -

CLC-site lignnds
Colchicine 10-16 I I K E
MTPT 4.8 I - E
~ombretastafin 11 I - E
~odophylloto~n 1.7 (A) 1 I I
Stegnacin 3.5 (A) I - E
NSC-350102 5-10 (A) I I I
Nocodazole 2-5 I I E
NSC-181928 ilU - I E
NSC-215914 >I0 - I - E
NSC-251635 na I - E
TN-16 - I I
Non-specificligands
XX-362499 5-10 - N - - - E
MPMAP <lO - N I N -
DAPI N N - - - - I
Bis-ANS 3-15 - N - - - -

iOnly inhibitors tested in a minimum of four techniques are documented. Since methods vary between publications only general
observations are recorded.
$Code: E, enhancement; 1: in~bition; N, no effect; -, not recorded, na, not available.
§Where possible J,,, values are taken from polymerization studies using @.I M-MES or PIPES buffer, CTP with addition of
inhibitor folowed by warming to 37 “C with spectrophotometric detection. I,, values obtained by other procedures are noted as ‘A
(atypical).
X98 E. LACEY

other basic macromolecules mimic the action of in many proteins and the lack of comparative studies,
MAPS in reducing the critical concentration for insufficient data exist to define the location of these
polymerization and stabilizing microtubule growth binding sites at present.
(Lee et at., 1978) as well as inducing rapid dis- While these interactions are useful biochemical
assembly of ~crotubules (Iwata, Matsui, Hino & probes in the characte~zation of tubulin (Roach &
Nakano, 1982) and tubulin aggregation. This latter Luduena, 1984) and toxicologically relevant to the
effect is also observed with aminoglycoside anti- study ot microtubules (many of these compounds are
biotics, neomycin, viomycin, streptomycin, gentamicin industrial chemicals), they are of little pharmaco-
and kanamycin at mM concentrations (Akiyama, logical importance since activity is due to interaction
Tanaka, Tanaka & Nonomura, 1978). with other proteins in general, not specifically tubulin.
While no definitive free ligand studies have been
reported, the recent chromatographic studies of Selective microtubule inhibitors
Kocha et ai. (1986) suggest the existence of hydro- Three pha~acolo~cally relevant selective sites
phobic binding domains on tubulin. They demon- have been defined by various methodologies and are
strate that by the use of CLC, its inactive derivative denoted according to the prototype inhibitor as the
lumicolchicine and a variety of unrelated hydro- CLC-, VBL- and taxol-sites. The CLC- and VBL-sites
phobic ligands (naphthyl, diphenylmethyl, trimeth- are present on the dimeric tubulin while taxol binds
oxyphenyl, phenyl), tubulin can be adsorbed onto the predominantly to a microtubule site with only vestigial
gel matrix. With the exception of CLC, these ligands interaction with the dimer. Despite extensive charac-
do not demonstrate activity at < 100 p M in polymer- terization of the CLC- and VBL-sites, their location
ization or [“H] ligand displacement studies. The within tubulin is enigmatic.
existence of potential hydrophobic domains is also
suggested by the interaction of polyaromatic ligands ~~b~ffsti~e binding site
such as 4’,6-di~idino-2-phenylindole (DAPI, VBL (Fig. 7), a clinically used anticancer agent
Fig. 6) and bis-(8anihno naphthalene-l-sulphonate) originally isolated from periwinkle ( vincu rosa L.), is
(bis ANS) and related compounds with tub&n in the prototype of the class of indoles known as Vinca
fluorescence studies (Bonne, Heusele, Simon & alkaloids incorporating both natural and semi-
Pataloni, 1985; Horowitz, Prasad & Luduena, 1984; synthetic derivatives (Zavala, Guenard & Potier,
Prasad, Luduena & Horowitz, 1986), and may go 1978) and is distinct from other indoles such as
some way to account for the range of structurally resperpine, yohimbine, melatonin and harmine (Tan
unrelated compounds such as the 5,6diphenylpyri- & Lagnado, 1975). VBL and other Vinca alkaloids
dazin-3-ones (NSC 362446, Fig. 6) (Batra, Powers, inhibit tubulin polymerization at 0.5-S HUMin vitro,
Hess & Hamel, 19&6), gossypol (Medrano & Andreu, inducing the formation of spiral aggregates and
1986), daunomycin (Na & Timasheff, 1977), acti- paracrystal formation at high concentration (Zavala et
nomycin D (Rajagopalan & Gumani, 1986), stypol- al., 1978). [‘H] VBL binding is rapid, reaching
dione (O’Brien, et al., 19X4), rifampicin (Rajagopalan equilibrium within 15 min, and reversible with bound
& Gurnani, 19851, oestrogens (Hartley-ASP, Deinum ligand easily dissociated during separation of bound
& Wallin, 1985) and progesterones (Loizzi, 1985) and free drug under non-equilibrium conditions.
known to possess weak microtubule inhibitory Under equilibrium conditions, a 2 : 1 stoichiometry of
activity distinct from established CLC and VBL sites. VBL binding to tubulin has been established (see
The cellular response of microtubule inhibition in Luduena, Anderson, Prasad, Jordan, Ferrigni, Roach,
such cases probably reflects the essential role of the Horowitz, Murphy & Fellous, 1986). The association
microtubule matrix in cell homeostasis as opposed to constants (Kc,}for VBL binding are highly variable
ligand specificity. Zfr vitro, their effects are generally ranging from 0.16 to 45 PM-‘, depending on metho-
characterized by high concentrations of compound dology and source of tubulin (Luduena et al., 1986).
(normally >50 ,uM), multiple stoichiometry, isotope The activity of VBL in various microtubule inhibitor
labelling dispersed in both cz- and /3-tubulin and the sensitive techniques is presented in Table 3.
absence of defined structure-activity relationships Maytansine (MTS, Fig. 7), a 19-member macrolide
(SAR). Within the non-selective classes, sufficient isolated from genera of Maytenus and Putterlickia is
departures from these generalizations occur to imply the most extensively studied of the mayt~sinoid
the existence of unique binding domains for Bis ANS, microtubule inhibitors (Kupchan, Sneden, Branfman,
DAPI. DCBT, 2(4-methyl-l-piperazinylmethyl)- Howie, Rebhun, Mclvor, Wang & Schnaitman, 1978).
acryiophenone (MPMAP) and NSC 362449 (Abra- MTS is a competitive inhibitor of [“H] VBL and
ham et al., 1986; Batra et al., 1986; Luduena et al., vincristine binding and demonstrates similar reversi-
1987; Bonne et al., 1985; Prasad et al., 1986). It is bility of binding and dependency on ionic strength
probable that local steric, pl and hydrophobic ligand and temperature (Mandelbaum-Shavit, Wolpert-De
binding sites have distinct optima, which lead to Filippes & Johns, 1976).
segregation of non-selective inhibitors based on their A third structurally distinct class of inhibitors, the
physico-chemical properties. In view of the distribu- cyclic hexapeptide phomopsinoids isolated from the
tion of nucleophilic, acidic and hydrophobic domains causative agent of lupinosis, ~ho~ops~ ~e~tostro~i-
 Role of tubulin in benzimidazoles mode of action x99

flPl?RP

CH2SCzN

-bls ANS

0CB-l

NH2

NSC 362449 DRPI

FIG. 6. Structures of non-selective microtubule inhibitors, with specific binding domains on tubulin.

formis, also inhibits [“HI VBL binding (Lacey et al., Tax& binding site
1987). Phomopsin A (Fig. 7) and B and the semi- Tax01 (Fig. 12) is a natural product isolated from
synthetic derivatives, phomopsin~ine and octa- the yew tree, Taxus br~vi~~~i~(see Horwitz, Parness,
hydrophomopsin, are the most potent microtubule Schiff & Manfredi, 1982). Taxol’s action as a microtu-
inhibitors thus far reported (Zso polymerization bule inhibitor is unique in that it is achieved by
<t /AM). inducing tubulin to polymerize, increasing both the
Recently, a further inhibitor, rhizoxin (Fig. 7), a rate and yield of microtubules in both polymerization
macrocyclic ketone isolated from the fungus Rhizo- competent (with MAPS) and incompetent tubuhn and
pus chinensis, a pathogen causing rice seedling blight, effectively reducing critical concentration for assem-
was reported to inhibit polymerization by binding to bly by greater than 20-fold (Horwitz et al., 1982).
the VBL-site (Takahashi, Iwasaki, Kobayashi, Okuda, This action renders possible the isolation of microtu-
Murai & Sato, 1987). bules from sources previously obtainable only after
The VBL-site is known to induce changes to the lengthy isolation processes. The recent report of
CLC-site (Lacey, ef al., 1987), GTP E-site (Huang, conditions by which taxol-induced microtubules
Lin & Hamel, 1985), Bis ANS site and the MAP and could be dissociated extends the potential usefulness
tau binding regions (Luduena et al., 1986). Despite of this approach (Collins & Vallee, 1987). [“HI Taxol
the importance of this site to tubulin activity, evidence binds almost exclusively to microtubules with stoichi-
of its existence outside mammals and echinoderms is ometry equivalent to 0.6-0.8 moles per mole tubulin
limited (Luduena et al., 1986). No report of ligands dimer. The binding site is thought to be distinct from
active at the VBL-site on hehninth tubulin has been those of VBL, CLC, Ca”+ and GTP (Parness &
pubIished. Horwitz, 1981). Due to the complex nature of the
900 E. Lacsv

“3C

R = CH, Vlnblastxne

R = CHO Vincristine Maytanslne

fl c~cH2cH3

Cl
CH-C-N- CH-F--N& C-L-
0 0

c=o
“T
CH
CHIN; ;;-NH$~CH3
0
C”2

Phomopsin R

ne

Rhizoxin

FIG. 7. Structures of vinbiastine binding site ligands.

 Role of tubulin in benzimidazoles mode of action 901

interactions of taxol in binding to and altering dif- using 2-methoxy-5(2’,3’,4’-trimethoxyphenyl) tro-

ferent components of the tubulin-microtubule equi- pone (MTPT, Fig. 8) an analogue lacking the
librium, the precise location of the domain on dimeric aliphatic B ring (Bane, Puett, MacDonald & Williams,
tubulin and the microtubule is unknown. 1984) support the role of the methoxytropone group
in fluorescence (Bhattacharyya & Wolff, 1974). This
Colchicine binding site phenomenon is a significant aspect of the interaction
Colchicine. CLC is a tricyclic acetylated alkaloid of CLC-site ligands and its mechanism will be con-
originally isolated from meadow saffron (Colchicum sidered later. Complex formation is also associated
automnale), the ring systems of which are designated with changes in the circular dichroism spectrum
A, B and C as shown in Fig. 8. CLC binds to mammal- (Detrich, Williams, MacDonald, Wilson & Puett,
ian brain tubulin at a single high affinity site with 198 l), accessibility of altered protease sites (Serrano
maximal stoichiometry of 1: 1. This maximum is only et al., 1984), reactivity of cysteinyl residues (Roach et
observed under equilibrium binding conditions with al., 1985) GTPase activity (Lin & Hamel, 1981)
non-equilibrium stoichiometry varying between 0.5 antibody recognition (Morgan & Spooner, 1983) and
and 0.8 moles CLC per mole tubulin (Table 4). electron paramagnetic resonance (EPR) spectrum
Binding is slow with equilibrium time >4 h, highly (Deinum & Lincoln, 1986) which, together, with the
temperature dependent (optimum 37°C) with virtu- fluorescence studies, support the occurrence of con-
ally no binding observed at 4°C. The complexation of formational changes both to CLC and tubulin on
CLC with tubulin results in a ‘tight’ interaction which interaction. The kinetics of complex formation
is not readily dissociated and is often referred to as involve a minimum two-step CLC-induced tubulin
‘pseudo irreversible’. The complex dissociates readily conformational change: Tubulin + CLC * (Tubulin-
on tubulin denaturation by solvent extraction, heat or CLC) + (Tubulin-CLC)*
detergent and thus is non-covalent, The CLC-binding The initial reaction is rapid leading to the formation
site on tubulin is not stable and decays with first order of a reversible complex which undergoes a slow CLC-
kinetics. Stabilization of binding site decay can be induced conformational change in tubulin to give the
achieved by the addition of VBL (or VBL-site final complex. This final step is thought responsible for
ligands), GTP, CLC-site ligands and CLC itself the observed pseudo-irreversibility of CLC binding
(Wilson, 197.5). The mechanism of this decay has not (Garland, 1978). Subsequent studies (see Bane et al.,
been elucidated. Formation of the CLC-tubulin 1984) have not departed from this basic model. Exten-
complex is associated with enhancement of intrinsic sive SAR have been reported for CLC analogues
CLC fluorescence intensity (>15OX) at a broad (Shiau, De &Harmon, 1976; Rosner, Capraro, Jacob-
emission maximum between 430 and 440 nm (Bhat- son, Atwell, Brossi, Iorio, Williams, Sik & Chignell,
tacharyya & Wolff, 1974; Garland, 1978). This 1981; Quinn & Beisleer, 1981; Brossi, Yeh, Chrza-
phenomenon is specific to the complex as the inactive nowska, Wolff, Hamel, Lin, Quin, Suffness & Silver-
isomer, isocolchicine, shows only slight enhancement ton, 1988) which demonstrate the scope of structural
(3% of that for CLC). Comparative studies of fluores- limitations within this domain. Modification of the C
cence induced by changes in solvent viscosity and ring by hydrolysis to colchiceine or photodecomposi-
immobilization-dependent fluorescence effects (using tion to p- and y-lumicolchicine abolishes activity
desacetyl CLC coupled to proteins) suggest that (Wilson, 1975; Rosner et al., 1981). Reversing the
CLC-tubulin fluorescence may be due to immobiliza- positions of the methoxy and keto groups as occurs in
tion of CLC within the binding site (Bhattacharyya & isocolchicine, removal of the methoxy group (colchi-
Wolff, 1984). Fluorescence enhancement is reported tide) and hydrogenation of a double bond in the C ring
also for other methoxytropone analogues of CLC (pre-colchicine) also results in loss of activity (see
(Ray, Bhattacharyya & Biswas, 198 1). Data produced Brossi et al., 1988). Replacement of the methoxy

Tax01

-
0\ /
FIG. 8. Structureof taxol.
 Temp.
Ligand Technique Stoichiometry (“C) K, (FM-‘) Comments References

Colchicine * 0.5-0.8 37 l-l.25 Slow, irreversible Hams et al., 1978

MTFT Fluorescence 23-37 0.29-0.48 Rapid, reversible See Engelborghs &
Fitzgerald, 1987
~odophylloto~n /‘H]/DEAE paper -0.8 37 1.8 Rapid, reversible Cortese et al., 1977
binds at 4°C
MB2 [“‘C]/dialysis 0.73 37 0.28 Rapid, reversible Laclette et al., 1980
Additional low affinity
sites of observed
NDZ Equilibrium dialysis 0.45 25 - Binds at 4” C Hoebeke et al., 1976
Gel filtration 2.0 25 0.4 High % DMSO used Head et al., 1985

*CLC binding has been extensively characterized by most of the methods cited in Table 2, reviewed in Hams et al. (1978). The data
presented for CLC represent the range of reported data in that article.
 Role of tubulin in benzimidazoles mode of action 903

group by a methylthio moiety gives a slight improve- it was by no means essential. As such the postulated
ment in potency, however, increasing size beyond ‘trimethoxy phenyl’ binding (Andreu & Timasheff,
methylthio (thiocolchicine) reduces activity (Shiau et 1982) is perhaps much less sensitive to structural
al., 1976). Ammonolysis of CLC to yield the amino- variation than previously thought. The results support
tropone, colchiceinamide, maintains activity (Zweig the hypothesis that the role of the B ring in inhibition
& Chignell, 1973). Inversion of the (-) optical is to achieve optimal stereochemical alignment of the
activity of the 7-acetamido group of CLC to (+) in the A-C rings which for some compounds such as MTPT
B ring abolishes activity. This is thought to be due to a is unnecessary but which for others such as the
change in the conformation of the phenyltropone (A- 4-phenylbenzoates is essential. Based on competitive
C) rings from a predominantly S to an R configuration drug binding sites, EPR and other techniques, the
(Yeh et al., 1988). Removal of the acetamido group CLC-site is not proximal to either the GTP E-site,
to give the unsubstituted B ring maintains micro- Mg- 2+> Ca-2+, MAP-sites or VBL-sites, however a
tubule inhibitory activity as does the introduction of a complex series of allosteric interactions appear
double bond in this ring (Rijsner et al., 1981). to exist between some of these sites and the CLC-
Hydrolysis of the acetamide group to the free amino site.
(desacetylcolchicine) reduces activity by lo- to 20- Podophyllotoxin. Podophyllotoxin (PDT, Fig. 9), a
fold (Riisner et al., 1981), however, substitution with naturally occurring lignan isolated from the May
other amides, carbamates or ureas results in variable Apple (Podophyllurn peltutum), inhibits polymeriza-
effects. For substituted acetamides both inhibition of tion by binding to the CLC-site. Although a competi-
[‘HI CLC binding and anti-tumour activity is tive inhibitor of CLC binding, the [3H] PDT-tubulin
improved (Quinn & Beisler, 1981; Riisner et al., interaction is quite different in that it is freely revers-
198 1). No size limitation to replacement of the amide ible, equilibrates rapidly (<30 min) and occurs at 4°C
is apparent since desacetylcolchicine conjugated (Table 4, Cortese et al., 1977). While sharing many
to fluorescein, S-(4,6-dichlorotriazin-2yl) amino- properties in common with CLC-induced changes to
fluorescein (DTAF), a haem-nonapeptide or an azi- tubulin some differences are noticeable (Table 3).
donitrophenylaminohexamide photo-affinity ligand Several SAR studies of PDT analogues have been
yields active inhibitors (Clark & Garland, 1978; published. Isomerization of the tram lactone to cis
Zimmermann, Doenges & Moll, 1982; Salmon & (picropodophyllin, Fig. 9) or removal of the carbonyl
Wadsworth, 1986; Williams, Mumford, Williams, to give the furan (anhydropodophyllol, Fig. 9) both
Floyd, Aivaliotis, Martinez, Robinson & Barnes, result in 1 O-fold loss of activity, while opening the ring
1985). Modifications to the A ring have received little to give either the cis or tram acids abolishes activity
attention. Hydrolysis of the methoxy groups leads to a altogether (Kelleher, 1977). Modification of either the
reduction of activity while the presence of larger cyclohexane or aromatic rings results in only slight
water-soluble sugars (colchicoside) in the 4-position changes in activity with KI values for the displacement
leads to loss of activity (Riisner et al., 198 1). The of [“HI CLC ranging from 0.12 to 1.2 compared with
SAR of CLC analogues can be correlated to a hydro- PDT (KI = 0.51 ,LLM). With the exception of the
phobicity model (optimum log P = 1.17) with most lactone and the presence of bulky polar moieties, high
of the structurally specific effects leading to a loss of tolerance for structural change is observed for PDT
activity residing within the C ring (Quinn & Beisler, analogues (Kelleher, 1977). Steganacin (Fig. 9),
1981; RGsner et al., 1981; Brossi et al., 1988). In an another lignan related to PDT isolated from Stegunot-
elegant study, Fitzgerald (1976) demonstrated that uenia uruliuciu, also inhibits polymerization by bind-
the tropolone can be substantially modified without ing to the CLC-site (Zavala, Guknard, Robin &
loss of activity and that the B ring is not essential for Brown, 1980). A series of substituted benzylbenzox-
activity using allo-CLC (C ring methylbenzoate, oles based on the partial structure of PDT also act as
Fig. 8) and MTPT, a CLC analogue which lacks the B CLC-site inhibitors. Although the compounds are
ring. Subsequent studies support the lack of import- significantly less active than PDT, this represents the
ance of the B ring in inhibition but have shown that its first versatile synthetic approach to testing structural
presence confers the ‘pseudo’ irreversible character analogues of the aromatic rings of PDT. Extensive
on CLC binding (Bane et al., 1984). A more recent data for this series have been reported (Batra, Jurd &
study (Lacey, Burden &Watson, unpublished results) Hamel, 1985). The structures of the more potent
extended the Fitzgerald study by synthesizing a series analogues from this series (NSC 350102 and 32 1567)
of A-C derivatives of allo CLC. Interestingly, the are shown in Fig. 9.
methyl_4(substituted phenyl) benzoates (p-MTPB, Benzimiduzoles. The involvement of BZs in micro-
Fig. 8) were only poor microtubule inhibitors tubule inhibition was first indicated by studies of the
whereas the atypical allo-CLC 3-substituted phenyl- mode of action of the prodrug benomyl and its active
benzoates (m-MTPB, Fig. 8) exhibited potent activ- principle carbenazim (MBC) in fungi (Clemons &
ity. By the synthesis of various mono-, di- and Sisler, 1971). This was confirmed by cytological
tri-methoxyphenyl substituents of both series it was observation and analogy to other known microtubule
apparent that although the 2,3,4-trimethoxy arrange- inhibitors (Hammerschlag & Sisler, 1973; Davidse,
ment was optimal in both 3- and 4-substituted series, 197 3). Similar cytological studies both in mammalian
904 E. LACEY

Olle

CLC F11 lo-CLC

MTPT m-VTPB p-MTBP

C-Ring partial structures

p-Lumicolchicine

Colchiceine

0\
/ -
0
Isocolchicine
’ Otle
0

FIG. 9. Structures of colchicine and related compounds.

 Role of tubulin in benzimidazoles mode of action 905

cell lines and in rats with MBC, benomyl and PPOlO, obtained. For this series, these parameters were highly
an MBC prodrug (Styles & Garner, 1974), and in the colinear and thus not distinguishable as shown in eqs
nematode A. mum (Borgers & De Nollin, 1975) (1) and (2).t
confirmed the antimitotic and microtubule inhibitory
pIC,,, = 1.06x - 0.24~~~ + 0.29 Zalkv,+ 4.28
actions. However, in early studies MBC in com- (1)
n = 24 r = 0.94 s = 0.18 ’
parison to CLC and VBL only weakly inhibited
polymerization of tubulin in crude mammalian brain pIC,,, = 0.84MR - 0.15 MRz + 0.31 I,,,,, + 4.17
supernatant (Hoebeke & Van Nijen, 197.5). The n =24 r = 0.90 s = 0.23 (2)
initial biochemical evidence for potent microtubule
inhibition by BZs was reported for NDZ using where the negative logarithm of the IC,,, for polymer-
purified rat brain tubulin. An Is,, for polymerization ization (pIC,,,) was correlated to hydrophobicity, JC
of 0.63 p M was found making NDZ more potent than [eqn. (l)] and MR [eqn. (2)] of the 5-substituents and
both CLC and PDT, Under equilibrium dialysis an indicator variable for the presence (I = 1) or
NDZ binding to tubulin was rapid with a temperature absence (I = 0) of a minimum -CH,CH2-R or
optimum of approximately 25°C. Stoichiometry -OCH>-R structure within the 5-substituent. n, r and
approaches 1: 1 and binding is competitively in- s represent the number of inhibitors, the correlation
hibited by CLC (Tables 3 and 4, Hoebeke, Van Nijen coefficient and standard deviation of the regression
& de Brabander, 1976). In vitro inhibition of bovine equation, respectively.
brain tubulin polymerization by commercially avail- Optimum inhibition (IC,,, = 2-4 PM) for this series
able BZCs, TBZ and cambendazole (CBZ) demon- occurred over an approximate IogP range 1.8-4.2.
strated a range of I,,, values from 2 to 550 PM From this analysis a fundamental difference was
calculated from reduction in the rate of polymeriza- identified between the interaction of the aliphatic and
tion (Table 5) (Friedman & Platzer, 1978; Ireland, aromatic substituents which required the addition of a
Gull, Gutteridge & Pogson, 1979; Lacey & Watson, positive indicator variable (I) to the QSAR as
1985a). Preliminary SAR data confirmed both described above. In comparison with a CLC SAR
superior activity of the 2-carbamoyl group over study (Quinn & Beisler, 1981), the optimum hydro-
thiazole and the importance of a 5 (or 6)*-substituent phobicity for 5-substituted BZs was 5- to lOOO-fold
for potent (< 10 p M) inhibition (Friedman & Platzer, higher. Subsequent data however suggested that sub-
1978). The BZCs, oxibendazole (OBZ) and fenbend- stituent size and not hydrophobicity was the determi-
azole (FBZ) were competitive inhibitors of [3H] CLC nant for activity since for 69 5-substituted BZCs an
binding while the thiazole CBZ was non-competitive overall regression to MR (r= 0.875) was superior to
(Friedman & Platzer, 1978). A subsequent report of that observed for ?G (r = 0.759). Within this regres-
MBZ inhibition of bovine brain tubulin together with sion parabolic dependence on substituent size was not
equilibrium dialysis data confirmed previous studies observed suggesting the absence of a major size
with its bioisostere NDZ. The binding profile did,
however, suggest the presence of another class of tQuantitative SAR studies attempt to correlate the
lower affinity binding site(s) as saturation at high con- physico-chemical parameters of the inhibitors to biological
centration could not be demonstrated (Laclette, activity. Four parameters JC, MR. u and I are commonly
Guerra & Zetina, 1980). Detailed investigation of used. n is defined as the hydrophobicity of a substituent, x,
NDZ-tubulin interaction by both equilibrium and and is calculated by
non-equilibrium methodologies indicated a 2 :1 Jr, = logP,_, - logP,_,
stoichiometry occurring via a minimum two-step
mechanism, analogous to that of CLC, with an initial where for any x, n is obtained as the difference between the
oil/water partition coefficient of a compound containing x
fast component followed by a slow phase (Head et
and the parent derivative where x = H.
al., 1985). Thus, MR (molar refractivity) is a parameter derived from the
NDZ + Tubulin + NDZ * Tubulin + NDZ . Tubulin*. Lorentz-Lorentz equation:
fast slow n'-I
MR= - +MW
Unlike CLC, however, this interaction is freely revers- nz+2 d
ible. Despite the numbers of BZs synthesized over the where MW is the molecular weight of the compound; d,
past 30 years, only one comprehensive SAR of density of the atoms/molecules; and n, the index of refrac-
microtubule inhibition has been reported (Lacey & tion. MR is colinear with substituent size and is thought to
Watson, 1985a). By examining the effect of variation represent the role of London dispersion forces in binding.
of the 5-substituent, correlations to both hydropho- u, Hammett constant, reflects the electronic nature of the
bicity and molar refractivity (MR, size) were substituent. Positive u values indicate electron withdrawing
substituents where negative values are electron donating.
I, indicator variable, is arbitrarily defined for the presence
*BZs are tautomeric when substituted on the aromatic of a specific structural subset (I = 1); if absent, I = 0.
ring. The two indistinguishable tautomers formed are With the exception of I values, all physico-chemical
normally noted as 5 (or 6). In this review this is contracted to parameter values for most commonly used substituents are
5- for convenience. tabulated in Hansch & Leo (1979).
906 E. LACEY

TABLE S-ACTIVITY OF BEN~~M~DA~~LES

AS INHIBITORSOF MAMMALIANAND PARASITETIJBULIN
RELATEDTECHNIQUES*t

Mammalian Parasite

lC,,, ED,,, LD,,,

Drug (p?!y) (Ll??O) $6) % Inh. (MBZ) (eggs) (larvae)

Carbenazim 71 2 to 7 NI 5.0 7.6 0.094

Parbendazole 3.1 36 0.30 0.33 0.014
Oxibendazole 2.2 onIt 37 0.30
Albendazole 6.9 nt 13 0.21 OS on;34
Rycobendazole NI 100 nt nt 2.3 NI 100 1.6
Ciclobendazole 5.5 nt 37 1.4 8.5 0.24

Mebendazole 6.1 0.31 6.3 88 0.19 4.1 0.048

Nocodazole 2.1 0.07 4.5 90 0.12 19 0.038
Flubendazole 3.5 40 0.17
Fenbendazole 5.4 on:7 18 0.18 NIn: 0 on;19
Oxfendazole NI 100 11.3 - 33 1.3 NI 100 0.57
Luxabendazole nt 0.30 nt nt
Thiabendazole 549 nt nt 5.5 0.33 co.04
Cambendazole 64.2 nt nt 0.54 nt nt

*All data are quoted in PM concentrations.

tr,,, (poly): concentration required to inhibit polymerization by 50%. Data derived from Lacey &
Watson (1985a) with the exception of ciclobendazole (Lacey, thesis cited above), thiabendazole and
cambendazole (Friedman & Platzer, 1978).
LD,,, (L12 10): Concentration required to inhibit the growth of L 12 10 leukaemia cells by 50%. Data
from Lacey & Watson (1985b).
IC,,, (CLC), % inh.: Concentration of inhibitor required to inhibit [“HI CLC (1 ,u M) binding to sheep
brain tubulin by 50% and the % inhibition observed at saturation concentration (200 PM). Note,
most BZs are only soluble up to lo-20 p M. Data from Lacey, unpublished results.
IC,,, (MBZ): Inhibition of 13H] MBZ (PM) binding by pre-incubation of inhibitor followed by addi-
tion of [“H] MBZ to H. contorfus L, crude tubulin extracts. Data from Lacey, Gill & Murray, unpu-
blished results.
ED,,, (eggs), LD,,, (larvae): Inhibition of egg and larval development of H. contortus. Eggs were iso-
lated from faeces and incubated on agar containing increasing inhibitor concentrations. After 7 days
the numbers of eggs, L,/L2 and L, larvae were counted and the data computed to derive the ED,,,
and LD,,, values (Lacey, Redwin & Waller, unpublished results). This technique results in substan-
tial differences in the ED,,, data for egg hatch where water is used (compare with Lacey ef al., 1987).
nt: Not tested.

restriction in the binding domain of the Ssubstituent however, further increases to propyl, butyl and pentyl
region in tubulin. Several stetic and electronegativity carbamate progressively reduce activity. For all
factors were noted as important for potent inhibitory 5-substituents where variation of the carbamate was
activity for alkyl substituents (I,,,,,) and for substi- undertaken, the isopropyl carbamate was consistently
tuted phenoxy derivatives (u2.__ Zx’.ha,lde
and (T:..,,~) in potent (Lacey, 1982, thesis cited above). Alkylation of
eqn. (3) (Lacey E., unpublished Ph.D. thesis, Univer- the l-position also reduces activity lo- to 20-fold.
sity of Sydney, 1982). Replacement of the -NH- of the BZ with -S- or-O-to
give the benzothiazoles and benzoxazoles, respect-
pIC,,, = 0.28MK + 0.35 I,,,,, + 2.35 u2.+ + ively, abolishes inhibitory activity (Lacey, 1982, thesis
- 1.15 a*‘-‘“h+4.511. (3)
0.27 ~~~~~~~~~ cited above). Mono-substitution in the 4 (or 7)
n=69 r = 0.875 s = 0.216 position produces weak activity, being approximately
20-50X less active than the 5-isomer (for example,
The effect of structural modification of the parent 6-phenoxy I,,, = 3.3,~~; 4-phenoxy I,,, = 62,~~).
molecule, MBC, on microtubule inhibitory activity is However, the introduction of a 4 [or 7) substituent
illustrated in Fig. 10. The presence of a carbamate into an active, but not inactive 5-substituted BZC
group (NHCOOCH,R) in the 2-position is essential results in a slight improvement in activity (4-Me ABZ,
for potent activity. Irrespective of the 5-substituent, I,,, = 5.1 PM, ABZ, I,,, = 6.9 PM). With the excep-
replacement with an amide (NHCOCH,R) reduces tion of the dimethyl analogue, 5,6-disubstitution
activity lo-fold, while hydrolysis to the amine reduced or abolished activity. To date no examples of
abolishes activity. Increasing the size of the carbamate 4,5- or 4,7-disubstituted BZCs have been examined.
from methyl to ethyl improves activity 1.2- to 2-fold, Of the approximately 500 BZ analogues tested in this
 Role of tubulin in benzimidazoles mode of action 907

OH

OCH3 OCH,

bCH3
Steganac In
PO1

Lac tone partial strut tures

Picropodophyllln

Podophyllic acid

R = 0CH3 NSC 350102

POT-cyc lit ether
R = OCH,CH, NSC 321567

FIG. 10. Structures of podophyllotoxin and related compounds

laboratory, most potent microtubule inhibitory activ- Other CLC-site inhibitors. Griseofulvin and iso-
ity was found to reside in the 5-substituted BZCs. propylphenyl carbamate (IPC) (Fig. 11) were both
Within this series, maximum inhibition of tubulin recognized as microtubule inhibitors and were
polymerization was observed between 0.7 and 4 ,UM thought to act by binding at the CLC-site (Fig. 12).
for both aliphatic and aromatic substituents. As However, demonstration of this activity in biochemi-
described in eqn. (3), this activity was dependent on cal studies with mammalian tubulin has been unsuc-
the size of the substitutent, however for aromatic sub- cessful (Wehland, Herzog & Weber, 1977; Goss,
stituents such as the phenylthio (FBZ) and phenoxy, Bloodgood, Brower, Pickett-Heaps & McIntosh,
substitution of the ring modulated activity, as de- 1975). This was despite cytological and in vitro
scribed by the Z.i.ha,lde,(T~._~~
and u:,,,~ terms in eqn. (3). growth inhibition studies in lower eukaryotes which
Irrespective of the size of the substituent, activity was confirmed their activity and specificity for tubulin.
heavily dependent on the nature of the moiety Surprisingly, the simplicity of IPC structure has not
adjacent to the BZ. For either branched substituents led to a more detailed analysis of the SAR for phenyl
(isopropyl, t-butyl) or polar moieties (-CHOH-, SO, carbamates. Several more complex carbamates have
SO,) activity was either reduced or abolished. This been identified such as NSC 215914, tubulozole and
and other observations su gest that both the geometry NSC 181928 (Fig. 11) which also inhibit tubulin
and polarity of this 2-3 1 region adjacent to the BZ polymerization. The stereochemistry of tubulozole
play a major role in the interaction with tubulin. It is represents a unique probe for microtubule-dependent
thought that these limitations are important in defin- mechanisms in that the cis-isomer is an active in-
ing structural variability compatible with the binding hibitor (I,, = 0.6~~) while the tram-isomer is essen-
of both the BZC and .5-substituents to their respective tially inactive (Van Ginckel, de Brabander, Vanherck
domains, in essence acting as a bridge. The inhibitory & Heeres, 1984). This relationship offers consider-
potency of BZs against polymerization of mammalian able advantage over other active/inactive probes such
tubulin was highly correlated with the LD,,, for L 12 10 as CLC/lumicolchicine and CLC/isocolchicine, since
growth supporting the notion that tubulin is the sole the tubulozoles are stable, water-soluble and possess
site of action in inhibiting cell growth (Lacey & nearly identical chemical and kinetic behaviour. The
Watson, 1985b). deazadihydropteridine carbamate, NSC 18 1928,
008 E. 1 ACk.:Y

chemically closely related to methotrexate, is one specificity which constitute the general framework for
of an extensive series of compounds which show consideration of the modes of action of BZs. These
antitumour activity (Bowdon, Waud, Wheeler, Hain, principles are:
Dansby & Temple, 1987). These compounds are (1) The potency of BZs (mg per kg dose) based on
potent inhibitors of in vitro polymerization and in vivo efficacy studies demonstrates a rank order
inhibit (“HI CLC binding. Detailed SAR studies have efficacy of TBZ < CBZ < PBZ < OBZ < MBZ <
been reported. Within the deazadihydropteridine FBZ - ABZ - OFZ - RBZ.SO over a 1O-fold range.
nucleus, modifications such as aromatization, -NH, For gastrointestinal nematodes. typical dose ranges
derivatization or inclusion of -NH-bioisosteres such arc 14 to 5 mg kg-’ in sheep, 66- 100 to 4.5 mg kg-’ in
as -S-, -0-, -CH,-, either reduce or abolish activity. cattle, 88-100 to 6 mg kg-’ in horses and SO to
3-Position activity appears subject to non-specific 4 mg kg’ in pigs (Marriner & Armour, 1 Y86).
features such as hydrophobicity and size analogous to (2) The doses required to achieve efficacy against
the S-substituents of BZs, however this is yet to be nematodes are lower than those used for cestode and
confirmed. A number of other inhibitors have been trematode control. In the latter classes control is
reported to act at the CLC-site. These include com- normally only obtained at higher doses or on multiple
bretastatin (Pettit, Singh, Niven, Hamel & Schmidt, treatment (Van den Bossche et 01.. 1982).
19X7), TN-16 (Roach & Luduena, 1985) and rote- (3) Extra-intestinal parasites, particularly intra-
none (Brinkley, Barham, Barranco & Fuller, 1974). vascular and interstitial dwelling parasites, are less
sensitive than gastrointestinal parasites (Van den
PHARMACOLOGY OF BENZIMIDAZOLE ACTION Bossche et al., 1982).
IN HELMINTHS (4) Activity against developing stages is superior to
The study of the pharmacology of BZs (or any other that against arrested or adult stages in comparable
drug class) is comprised of extension from studies at habitats. Two examples are (a) gastrointestinal para-
the biochemical level to the in vitro level and finally to site egg development (in vivo and in vitro), hatching
in vivo studies. At each stage. the complexities of the and larval development are inhibited at doses which
drug’s interactions and dynamics within the system are sub-efficacious against adults in vivo (Kirsch &
are vastly increased. These complexities are crucial to Schleich. 1982). and (b) interstitial filarial nematodes.
understanding how BZs act in vivo but when con- such as Onchocerca gibsoni, display microfilarial
sidering the basic mode of action are best viewed as control in the absence of adult efficacy (Forsyth.
modulating factors (Lacey, 1985). Mitchell & Copeman. 1984). Examples of developing
stages being relatively insensitive under conditions of
General in vivo aspects of benzimidazole adult control are known, for example F. hepatica
pharmacoloD (Boray. 1986). however such cases are generally due
The in vivo efficacy of BZs has been examined in to different habits of developing and adult stages.
most clinically significant helminthoses in a diverse Important host-related pharmacological principles
range of animals (for recent reviews see Van den affecting in vivo efficacy are:
Bossche, Rochette & Horig, 1982; Van den Bossche. (1) Host toxicity is low at efficacious anthelmintic
Thienpont & Janssens, 1985; Campbell & Rew, doses, however transient adverse reactions at higher
1986). Further review of such an extensive data base or chronic doses are observed. Acute LD,,, host
serves little purpose, but from these reviews it is values are generally greater than 5-20 X normal dose
possible to establish a number of general principles of levels (Seiler, 1975).
the pharmacology of individual BZs and their species (2) Efficacious BZ doses are host-dependent due

FIG. 11. Structural prerequisites for benzimidazole carbamate for optimal microtubule inhibitory activity.
 Role of tubulin in benzimidazoles mode of action 909

Cl+0

c”3~ Cl
0
o-
c N”~ocHK”3)2
Griseofulvin IPC

Combretastatin Ro tenone

TN-16

Tubulozole
H E
NHC-OCH2CH3

-
\ /
u-

NSC 181928

NSC215914
0
CH,O-tN
0
NHkOCH,

NSC 251635

FIG. 12. Structures of other CLC-site bindingligands.

910 E. LACEY

to differing dynamics and clearance rates of drugs in Douvres, 1986) L, and L, (Rew et al., 1986) and L,
vh,o (Van den Bossche et al., 1985). and young adults (Jenkins, 1982; Rapson, Jenkins &
(3) BZs are extensively metabolized in host species Topley, 1985) in many parasitic nematode species,
usually to less potent anthelmintics. The pathways and Several in vitro techniques for detecting larvicidal
rates of metabolism for each BZ are similar for all activity (egg to L3) have been reported (Ibarra &
hosts. First pass metabolism by soluble liver enzymes Jenkins, 1984; Wailer & Lacey, 1985). Inhibition of
(reductases and oxidases) constitutes a rapid pathway development and poputation growth of non-parasitic
of deactivation for several BZs, notably MBZ, PBZ species (for example, C. elegans) by BZs has also been
and ABZ (see Van den Bossche et al., 1985; Hen- reported (Platzer, Eby & Friedman, 1977; Spence,
nessy, 1985). Malone, Novak & Woods, 1982). Techniques which
(4) Despite earlier reports of poor systemic involve the isolation of obhgatory parasitic stages for
absorption of BZs after oral ad~nistration in viva testing BZs are reported but their use is minimal. The
(most notably MBZ), subsequent data suggest >50% recent in vitro assessment of motility of adult and
absorption (Van den Bossche et al., 1982). Reduced non-developing stages (Folz, Pax, Thomas, Bennett,
bioavailability was initially assumed in the absence of Lee & Conder. 1987) suggests that this situation may
biliary excretion data. Within the BZ class, biliary change.
excretion is favoured by MBZ, FBZ and OFZ while Other in vitro BZ sensitive techniques based on
predominantly urinary excretion is observed for TBZ, cytological examination of either cultured cells or
CBZ, PBZ, OBZ and ABZ (see Hennessy, 1985). tissue sections have also been reported recently
(5) BZs are hydrophobic and water insoluble and (Rubino et al., 1983; Howells & Delves, 1985). With
therefore bioavailability and pharmacodynamics and the exception of these more recent techniques, many
thus efficacy can be altered by formulation and assays are not BZ-specific but are generally applicable
presentation (see Hennessy, 1985). to anthelmintic screening (Jenkins, 1982).
In summary, there are three fundamental aspects of The most widely used technique for determining
BZ pharmacology in vita: the role of host pharma- BZ activity is the ovicidal egg hatch assay. Nematode
codynamics, host-parasite interactions and the bio- parasite eggs isolated from faeces are incubated with
chemical pha~acology of BZs which must be BZ for 24-48 h, after which time the proportions of
addressed to account for the efficacy variations larvae and eggs or larvae plus embryonated eggs and
observed against helminths. unembryonated eggs are calculated to derive an ED,,,
value (the dose required to reduce egg hatch by 50”/0).
In vitro benzimidazole pharmacology Initially reported by Egerton (1969) in comparative
In vitro techniques for detection of anthelmintic studies of ovicidal activity of TBZ in several parasite
action by removing host pharmacodynami~s as a species, the technique found prominence in detection
component of drug modulation are inte~ediate to of BZ-resistance in nematodes. To date, data from
our understanding of mode of action. In vitro tech- egg-hatch assays have been correlated to in viva BZ
niques have been reviewed elsewhere (Jenkins, 1982). resistance in H. contortus (Le Jambre, 1976),
In utilizing in vitro data three basic assumptions are 71 colubriformis (Coles & Simpkin, 1977). 0. cir-
made: (i) the host plays no role in the intrinsic activity rumcincta (Martin, Anderson, Jarrett, Brown &
of the drug; (ii) the site of action is a vital component of Ford, 1982) and N. spathiger (Obendorf, Parsons &
the viability of the stage of parasite life cycle used in Nicholls, 1986). Egg development has also been
the in viiro technique, and (iii) dynamics of drug shown to be inhibited by BZs in non-nematode
availability comparable to in viva situations are species (Fetterer, 1986).
achieved. Recently, Lacey, Brady, Prichard CycWatson (i 987)
The third assumption is independent of the assay reported a correlation between inhibition of hatching
since drug uptake, solubility and stability which of H. contortus eggs and inhibition of mammalian
determine dynamics can be deduced by other means. brain tubulin polymerization by BZCs. For commer-
Many failures of the first two assumptions have been cial BZCs, PBZ, OBZ, ABZ, MBZ, FLBZ, FBZ and
reported or can be rationalized in retrospect, for OFZ and major metabolites of MBZ, FBZ, OFZ and
example, the discovery of levamisole (Raeymaekers, ABZ, inhibitors of mammalian tubulin polymer-
Allewijn, Vandenberk, Demoen, Van Offenwert & ization were also potent inhibitors of egg hatch, while
Janssen, 1966) and lack of pro-drug activity (Jenkins, non-inhibitors failed to prevent hatching. This led to
1982). Unfortunately, misinterpretation of the limita- the postulate that the primary mode of action of BZs
tions imposed by these assumptions has led to a in the developing egg involved inhibition of
general scepticism of in vitro approaches in parasi- mi~rotubule-dependent processes. Subsequent
tology. studies (Lacey, Redwin & Wailer, unpublished)
BZ activity has been reported in many in vitro examining a range of BZCs with differing hydro-
screens; however, in very few cases have extensive phobicities have confirmed this correlation, demon-
SAR data been presented. BZs inhibit in vitro strating however that ovicidal activity can only be
development, motility and survival of various transi- achieved over a relatively narrow window of log C’
tions between larval states L, and L, (Rew, Urban & values from 1 to 3.4. Compounds with IogP values
 Role of tubulin in benzimidazoles mode of action 911

below or above this range were not ovicidal but did to 20-fold more potent inhibitors of larval develop-
show larvicidal activity (L, to L, development). The ment (L, to LJ than egg hatch. Comparison of egg
requirement for optimal physico-chemical properties hatch and larval development inhibitory activity of
of drugs for in vitro activity is well established commercial BZs is presented in Table 5.
(Hansch & Leo, 1979). Graphical presentation of Most in vitro techniques for BZ action involve
negative logarithm of the ED,,, values (pED,,,) vs determination of toxicity associated with a transition
hydrophobicity results in a parabolic curve, the between distinct developmental stages. Dose-
positive slope demonstrating increasing activity with dependent BZ toxicity to non-developing stages of
increasing fat solubility and relates to the summation nematodes has not been conclusively proven. In fact,
of hydrophobic properties required for drug absorp- available data support the contention that BZs are not
tion (Fig. 13). Inhibition of hatching is therefore lethal in vitro to adult Nippostrongylus braziliensis
related not only to mechanism of action but also the (see Coles & McNeillie, 1977), F. hepatica (Rew,
ability to partition through the shell into the egg personal communication) and A. suum (see Kohler &
lumen. For hydrophilic (low logf’) derivatives, higher Bachmann, 1981). Studies in H. contortus and
drug concentrations are required to achieve an T. colubriformis adults and L, larvae failed to show
adequate concentration at the site of action. Where toxicity on in vitro incubation with BZs (Lacey,
this concentration is in excess of the drug’s solubility, unpublished results). Such negative reports are prob-
no inhibition is observed. Highly hydrophobic ably more common than has been reported in the
compounds (high log P) partition into the shell but published literature.
final equilibration into the egg lumen is not favoured Failure to demonstrate toxicity to non-developing
by virtue of the relative hydrophobicity of the stages in vitro does not imply failure of action, since it
environs of the shell and lumen (Fig. 13). This gives is possible to detect perturbations in parasite motility
rise to the negative slope of the pED,,,vslogP plot (Folz et al., 1987) nutrient uptake, enzyme secretion
since more drug is required to saturate the shell in or glycolytic enzyme activities (see later). It is appar-
order to achieve the required internal concentration ent that overt toxicity is associated exclusively with
for inhibiting egg development. developmental phenomena implying that disruption
These limitations are dependent on the shell and are of other biochemical actions within the parasite is not
removed in developing larvae where inhibitors out- lethal in non-developing stages. In Fig. 14, collation of
side the ovicidal window can be utilized to derive the in vitro and in vivo effects of BZ action on
LD,,, values. Good correlations exist for the relation- helminths are presented as a composite of the phar-
ship between LD,,, and ability of BZs to inhibit macological responses within the life-cycle of a
binding of [‘HI MBZ binding to parasite tubulin ‘typical’ nematode.
(Lacey, unpublished results). In general BZs are lo-
Biochemical pharmacology of benzimidazoles
Initial studies of the mode of action of BZs focused
on their role in carbohydrate metabolism. In com-
parative studies of BZ-resistant (R) and susceptible
70- (S) H. contortus, differential inhibition of the anaer-
obic enzyme fumarate reductase by TBZ was
observed (Prichard, 1973). Inhibition of fumarate
60- reductase by other BZs (CBZ, FBZ, OFZ and MBZ)
/-----I
with BZ-S and -R isolates supported this hypothesis
(Malkin & Camacho, 1972; Romanowski, Rhoads,
Colglazier & Kates, 1975; Rahman & Bryant, 1977).
Fumarate reductase satisfied many of the criteria for
selectivity of BZs: a selective target in anaerobic
OVICIDAL parasites with reduced sensitivity in BZ-R isolates.
40
H’fDROPHlLlC’ WINDOW HYDROPWOBK
However, it failed to address other observations such
as the lack of SAR compatible with in vivo efficacy at
therapeutic concentrations, inhibition by other un-
related anthelmintics (disophenol and levamisole)
(Prichard, 1973), lower control (uninhibited) activi-
ties in BZ-R compared with BZ-S, highly isolate-
dependent control activities and inhibition
independent of resistance status (Bryant & Bennet,
FIG. 13. Relationship between egg-hatch activity and
1983). These observations are inconsistent with a pri-
hvdroohobicitv flog P). The eauilibria between drug
concentrations in‘thi solution (OF\, the egg shell (0s) and mary target as they suggest significant modulation
the egg lumen (Ot) are illustrated to demonstrate the independent of BZ-resistance status.
pharmacological limitations of the egg-hatch for hydrophilic BZs have also been shown to inhibit glucose uptake
and hydrophobic derivatives. both in vitro and in vivo in many helminth species, for
912 E. LACEY

iefhat

FIG. 14. Colage of the in vir*o effects of ben2imid~o~es on helminths. Heavy arrows are used to indicate the structural
changes on BZ exposure: metaphase arrest in ovaries, atypical in U&W egg development, intestinal degradation and inhibited
egg and larval development. Impaired functions are denoted by X.

example, A. suum, Trichinella spiralis, Schistosoma demonstrated differential BZ-S and -R activity
mansoni, iWoniezia expansa and H. d~~~~~uta,in some occurring, however at lower 32. concentration ranges
cases associated with a compensatory depletion of (< 10 FM) compatible with it2 V&Odrug levels. wypo-
glycogen stores. However, glucose uptake was not thesis of a primary site of action by this mechanism is
inh~~i~e~ in other species such as F. he~at~ca, T. coiu- again considered dubious in view of the diverse range
br~orm~, N. dab& and H. tontortus (see Behm & of inhibitors known to inhibit glucose uptake. These
Bryant, I485). me-iuvestjgation of H. co~torr~ include amoscanate, d~amfenetide, dit~iaz~ine,
glucose uptake demonstrated a two-component niridazole and others (Rew & Fetterer, 1986; Bennett
mechanism, whereby at low concentrations of glucose & Thompson, 1986; Van den Bossche, 1985).
an active component was predominant which was BZ- Other less well-characterized effects of BZs on
sensitive while at higher concentrations of glucose a energy metabolism include uncoupling oxidative
passive BZ insensitive mechanism occurred. There- phosph~rylatiou (Van den Bossche, 1972), depres-
fore it is probable that failure to observe inhibitory sion of ATP levels (Rahman, Cornish, Chcvis &
activity may be due to either the absence of an active Bryant, 1977) and alteration of metabolic pathways
component or a saturation of glucose uptake sites {Rahman & Bryant. 1977; Rew, Smith & Co&lazier,
such that passive transport represents the primary 1982; Sangster Br Prichard. 1985). Many miscellane-
uptake mechanism (Rew, P&hard RLLacey, in pre- ous inhibitory activities on transmembrane proton
paration). Like fnmarate reductase, glucose uptake discharge (McCracken, Stillwell & Hudson, 1982,
 Role of tubulin in benzimidazoles mode of action 913

Abstract in Molecular and Biochemical Parasitology was not without justification since the concept of a
(Supplement), p. 726) synthesis or breakdown of ubiquitous protein as a primary site of action is hardly
serotonin (Metzger & Diiwel, 1974, Abstract in an attractive hypothesis for a class of drugs with a wide
Proceedings Third International Congress of Parasi- therapeutic index.
tology III: 1444-l 44.5) membrane-bound mono- The initial characterization of BZs as inhibitors of
amine oxidase activity (Moreno & Barrett, 1979) and [“HI CLC binding to A. suum embryonic tubulin
Na+ uptake (Beames, Merz & Donahue, 1976) have demonstrated a 2.50- and 400-fold greater inhibition
been noted while higher levels of glutathione-s- constant for MBZ and FBZ compared with mammal-
transferase were observed in BZ-R H. contortus by ian tubulin, providing the first indications of selecti-
Kawalek, Rew & Heavner (1984). Whether these vity (Friedman & Platzer, 1980). However, a
activities are mechanistically dependent on either subsequent report of direct [jH] MBZ binding to
fumarate reductase or glucose uptake is unknown. partially purified intestinal A. suum tubulin failed to
BZs have also been shown to inhibit the in vitro support this study and the hypothesis was advanced
secretion of acetylcholinesterase (AChE) in N. brasi- that selectivity was achieved by differential drug
Ziensis with resultant accumulation within the parasite kinetics in the host and parasite (Kohler & Bachmann,
(Watts, Rapson, Atkins & Lee, 1982). Studies in 1981).
T. colubriformis support these results showing differ- These conflicting hypotheses were addressed in a
ential activity in BZ-S and -R isolates (Sangster, study by Dawson et al. (1984) which compared the
Prichard & Lacey, 1985). Comparison of this action ability of BZs and CLC to inhibit DMSO-induced
with inhibition of neurotransmitter release induced by polymerization of A. galli tubulin. For FBZ, PBZ,
CLC led to the suggestion of microtubule involvement MBZ and CLC, the I,,, values for the inhibition of
(Watts et al., 1982). In vivo inhibition of AChE A. galli polymerization were between 4 and 6 ELM,
release, an essential component of a ‘chemical hold- similar to those in parallel studies with mammalian
fast’ mechanism for the parasite within its habitat, is brain tubulin (3.5-9~ M). OFZ and TBZ were signifi-
thought to lead to expulsion. The generality of this cantly more selective for A. galli tubulin than mam-
inhibition among other unrelated anthelmintics malian tubulin (33- and 120-fold, respectively). On
suggests that AChE levels act as a criterion for para- first glance this provides support for selectivity of
site viability (Rapson, Chilwan & Jenkins, 1986). action against parasite tubulin for only two analogues,
The concept of parasite expulsion without lethal with no selectivity for the other BZs or CLC. How-
action is addressed by Rahman et al. (1977). After in ever, on more thorough analysis a different situation
vivo administration of MBZ to sheep infected with emerges. Since only a small proportion of tubulin
H. contortus, M. expansa and F. hepatica, parasites dimers needs to bind to the drug to inhibit polymeriza-
were recovered viable up to 24 h post-dosing despite tion, any change in selectivity of binding would
severe changes in the biochemical levels of total primarily be detected by those drugs unable to bind
nucleotides and glucose while other metabolite levels mammalian tubulin effectively, in this case TBZ and
were unaffected. These and other observations OFZ. For the other ligands, binding (detected by
(Kirsch & Schleich, 1982; Sangster & Prichard, 1985) either inhibition of polymerization or displacement of
support the more general hypothesis that BZs induce [jH] CLC) to mammalian tubulin is easily demon-
multiple biochemical changes, only some of which are strated, therefore a selective interaction will not be
associated with an immediate lethal effect. This leads detected. As stated previously, polymerization is an
to the conclusion that parasite expulsion from the extremely sensitive indirect technique for detecting
prediliction site is due to inability to maintain homeos- ligand binding by virtue of the sub-stoichiometric
tatis and not a biocidal action. The existence of non- levels of binding required to inhibit microtubule
lethal action is well recognized in BZ treatment of assembly. However, polymerization assays cannot
fungal cells which are not growing (Clemons & Sisler, distinguish tightness of binding outside the extremes
1971). of whether sufficient or insufficient ligand is bound to
Coincident to examination of the role of BZs in inhibit polymerization. To exemplify this, VBL, PDT
energy metabolism, Borgers & De Nollin (1975) and MBZ are all more potent inhibitors of mammal-
postulated the involvement of BZs (particularly ian tubulin polymerization than CLC, yet it is only
MBZ) in the induction of microtubule disintegration CLC that forms a tight pseudo-irreversible complex
in intestinal cells of A. suum. This was supported by with tubulin. In summary, selectivity of BZ action in
other studies in both nematodes and cestodes polymerization assays can only be detected with poor
(Borgers, De Nollin, Verheyen, De Brabander & or inactive mammalian tubulin inhibitors and active
Thienpont, 1975; Verheyen, Borgers, Vanparijs & derivatives are irrelevant in this respect. The study by
Thienpont, 1976). While this work stimulated Dawson et al. (1984) in fact confirms the hypothesis
research into mammalian and non-parasitic lower of Friedman & Platzer (1980) that selectivity for BZs
eukaryotes such as fungi and slime moulds, bio- is achieved by a change in the BZ-tubulin interaction.
chemical and genetic studies in parasites received Further support for the interaction of BZs with
little attention until the concept of BZs as microtubule parasite tubulin was obtained by comparison of the
inhibitors was firmly established. In reflection, this binding of [‘HI PBZ, [jH] OBZ and [jH] CLC in
914 E. LACEY

displacement studies using BZ-S and -R T. colubri- literature. Tub&in binding sites have been identified
formis isolates (Sangster et al., 198.5). For both [“H] on the mitochondrial membrane suggesting the
BZs significantly lower association constants (K,) and existence of tubulin-mitochondria interactions which
maximum amounts bound of [“k-r]BZ (B,,,) were could represent a regulatory function (Bernier-
observed in the R-isolate (Table 6). These observa- Valentin & Rousset, 1982). SimiIar observations have
tions were consistent with other ultrast~ctural and been extended to plasma membr~es (Bemier-
AChE studies which supported the involvement of Valentin, Aunis & Rousset, 1983). Many soluble
microtubules in BZ-resistance. glycolytic enzymes such as glyceraldehyde-3-phos-
As discussed above, BZs are known to inhibit a phate dehydrogenase (Durrieu et al., 1987), aldolase,
wide variety of apparently unrelated mechanisms. Of pyruvate kinase and lactate dehydrogenase (muscle
these mechanisms, fumarate teductase, glucose type) can be demonstrated to interact with tubulin
uptake and microtubule inhibition each satisfy many (Karkhoff-Schweizer & Knuil, 1987). Detection of
of the criteria considered relevant for a putative site of these interactions by electrophoresis appears to
action. This gives rise to the question of whether these suggest that association is specific and not an
mechanisms are directly or indirectly related. Based experimental artefact due to acid-base protein asso-
on the inhibitor profiles of both fumarate reductase ciation since other enzymes with similar pI values such
and glucose uptake it is apparent that these systems as glucose phosphate isomerase, triose phospha-
are not specific to BZs. In this respect they are tisomerase, phosphoglyceromutase and enolase do
perhaps more precisely considered as sensitive sites not interact with tubulin (Karkhoff-Schweizer &
for assessing bioche~cal homeostasis within the Knull, 1987), while CLC in~bition of glucose uptake
parasite and thus are susceptible to a variety of and altered glycolytic pathways in mammary glands
indirect effects associated with chemical or environ- (Faulkner et al., 1984) are consistent with observa-
mental insult. The differential behaviour of BZ-S and tions in helminths.
-R isolates observed in both fumarate reductase and Recent studies with the microtubule inhibitors,
glucose-uptake could be secondary, since BZ selec- CLC and PDT, and non-inhibitors, colchiceine, /?-
tion enhances the viability of the parasite in the lumi-CLC and iso-CLC, demonstrated that both CLC
presence of the drug, differences in sensitive meta- and PDT were inhibitors of both glucose uptake (Rew,
bolic enzymes would reflect the observation that the P&hard & Lacey, in preparation) and fumarate
extent of chemical insult is less in BZ-R than BZ-S reductase (Prichard, Lacey & Darwish, unpublished
isolates. results). The coincidence of three structurally distinct
Alternatively, arguments for microtubule depen- microtubule inhibitor classes acting on these mechan-
dence of these mechanisms can be drawn from the isms supports the hypothesis of microtubule depen-

TABLE ~-ASSOCIATION CONSTANTS OF CLC-SITE LIGANDS TO PARASITE TUBULCN

Ligand Species Purity K, (PM-‘) Reference

cLc* A. suum :
early embryonic crude 0.3-0.5s Friedman, Platzer &
Carroll, 1980
late embryonic crude 0.15-0.22 Friedman et at., 1980
A. suum:
intestinal enriched 0.059 Kohler & Bachmann,
1981
H. diminuta enriched 0.076 Watts, 198 1
ABZ H. contort& crude 0.46-6.3 Lacey & Prichard, 1986;
Lacey et al., 1987
OBZ T. colubriformis crude 0.84 Sanaster et al., 1985
H. contortus crude 1.8-3.7 La&y & Prichard, 1986;
Lacev et al.. 1987
PBZ T. colubriformis crude 1.56 Sangster et al., 1985
H. contort& crude 0.7-4.3 Lacey & Prichard, 1986;
Lacey et al., 1987
MBZ H. contortus crude 0.7-1.1 Lacey et al., 1987
T. colubrijormis crude 3.6-7.7 Lacev, unpublished
FBZ H. contortuxt crude 1.1-4.5 Lace; & &chard, 1986;
Lacey et al., 19X7
OFZ H. ~~n~orius crude 0.17-0.22 Lacey et al., 1987

*CLC binding is observed also in F. hepaticn (Fetterer, 1986) and T. colubriformis

(Sangster et al., 1985), however the values are not recorded.
tK, values are protein dependent, lower K,vahres are associated with higher protein
concentrations.
 Role of tub&in in ben~imid~oles mode of action 915

dence, however the situation is complicated by the tubulin for CLC is lo-50 times lower than that
activity of non-inhibitors in both techniques, Valida- observed for mammalian tubulin (Table 3). Inter-
tion of the proposed dependence between these estingly k3H]NBC has only low levels of stable binding
mechanisms and microtubules is required. Despite in parasite extracts while no binding is detected for
this it is apparent that sufficient data exist to support a I’Hj TBZ by the charcoal extraction technique (Lacey,
general concept of a primary microtubule action unpubtished results}. Partial pur~~cation of H. con&~-
leading to a series of biochemical effects which either tus tubulin by polylysine chromatography indicated
directly or indirectly elicit physiological changes as that >85% of crude [“H] MBZ charcoal stable binding
schematically presented in Fig. 15, was located in the tubulin fraction. In this study [3H]
BENZIMIDAZOLE-TUBULIN INTERACTION IN BZ binding was observed to be dependent not only on
the isolate’s resistance status but also on BZ structure
HELMINTH RESISTANCE AND SPECIES with resistance factors (ratio of BZ-S and BZ-R
SPECIFICITY binding) varying from 6.9 for MBZ to 69 for OBZ.
The concept of BZ specificity for parasite tubulin Subsequent studies with other ii contotirrs isolates
derived from both [“H] BZ binding and polymeriza- supported the concepts that BZ resistance was depen-
tion studies supports the hypothesis that BZ selec- dent on the extent of charcoal stable binding and that
tivity is achieved at the level of the BZ-tubulin the extent of change in charcoal stable binding was
complex. While 13HjBZ binding can be quantitated by dependent on the structure of the BZ (Lacey, Snow-
several methods (Head et al., 1985; Sangster el al., don, Eagleson & Smith, 1987). BZ resistance may
1985, Laclette et al., 1980), the observation of also be detected by this technique in other parasites,
charcoal stable binding of (‘“C] MBC to fungal tubulin such as Q. ckcltnrctncta , T. c~~~~br~~rrnis(see Lacey,
represents a peculiarity previously only reported for in press) and N. spdziger (see Lacey, Obendorf,
I’H) CLC interactions. examination of charcoal Nichols & Eagleson, unpublished results).
extraction for the j3H] labelled BZCs, MBZ, PBZ, Analogous to the CLC-tubulin complex, the BZC-
OBZ, ABZ, FBZ and OFZ bound to BZ-S and -R parasite tubulin charcoal stable binding appears to be
ti. contorfus demonstrated that this property was also a tight, pseudo-irreversible interaction, in that once
characteristic of the BZC-tubulin complex (Lacey & bound the label is not readily dissociated by the
Prichard, 1986). Association constants for BZs and subsequent addition of excess unlabelled ligand. The
CLC binding to various helminth tubulin extracts are complex is, however, non-covalent since quantitative
presented in Table 6. In general, the affinity of parasite dissociation on denaturation of tubulin by either

k AChE

Fro, 15 Proposed mechanistic relationship between tubulin and other sites of BZ action.
916 E. LACEY

boiling, detergent or solvent extraction occurs (Lacey, coal extraction indicates that this technique may also
unpublished results). be able to determine the selective toxicity of BZs as
The simplicity and robust nature of the charcoal anthelmintics. This hypothesis was recently examined
extraction technique formed the basis of the first by comparing 13H]MBZ binding and in viva efficacy
routine technique for biochemical detection of resis- for a selection of helminths (Lacey, in press). These
tance in parasitic nematodes (Lacey & Snowdon, data are reproduced in Table 7. For gastrointestinal
1988). Resistance can be detected at any stage of the parasites and host species the quantity of j3H] MBZ
life cycle, however the technique was standardized for bound is reflected in the observed in viva toxicity with
L, larvae as the most convenient isolation and storage the exception of the intravascular dwelling parasite
stage. [“HI MBZ binding to crude 10.000 g super- D. immitus and the interstitial parasite 0. gibsoni. In
natants was quantitated after 30 min incubation at these parasites, only poor efficacy is achieved against
37°C followed by a 5 min charcoal extraction and adult parasites at doses which reflect the intrinsic
centri~gation. For H. cun~~~us, 0. ~~r~lirnci~c~~ sensitivity of the binding site. Since in viva BZ inhibi-
and T. coi~br~formis all [“H] MBZ binding sites were tion of microfilaria, embryotoxicity in the adult female
saturated at concentrations >0.3 _uM with no signifi- (Awadzi, Schulz-Key, Howells, Haddock & Giiles,
cant kinetic differences observed between the species 1982; Forsyth et al., 1984) and in vitro inhibition of
(K, values, 8.6, 7.4 and 7.7~~-‘; B,,, values 101, D. immitis excised ovary cells (Howells & Delves,
105 and lOOpmolesmg_‘, respectively). In 24 isolates 1985) can be demonstrated, it is probable that the lack
tested the extent of ]‘H] MBZ binding correlated with of adult efficacy is due to other factors.
known in viva and/or in vi&o resistance status of the Since BZs exhibit lethal action both it2 vitro and in
isolates. To date over 80 isolates have been tested by t@voat developing stages of the life cycle and removal
this technique with no cases of false positives or of non-developing stages from the host is non-lethal,
negatives noted. efficacy may be dependent on the mechanism and
Correlation of resistance status to a structural dynamics of parasite elimination from the host. In the
change in the binding site which is sensitive to char- gastrointestinal tract, removal from the predeliction

TABLE 7-CORRELATION OF [SH] MBZ CHARCOALSTABLEBINDING AND in viva

TOXIC11.Y’

[3HI MBZ boundt Dose

Organism (pmoles/mg crude protein) (mg kg-‘)

T. colubr~orm~ (BZ-S) 71 c12.5

H. contortus (BZ-S) 63 c12.5
0. circumcin& ’ 65 c12.5
N. spathiger 25 15
H. contortus (BZ-R) 10 >25
Intravascular
D.immitus 96 Not effective
Interstitial
0. gibsoni 50 Not effective
Trematode
F. hepatica 10 100

Cestodes
T. pisiformis 10 200 (m)t;
T. h~datige~a 5 160-320 (m)
E. granulosus (cyst) 10 100-500
(highly variable)

MtUlUUdS
Ovis sp. (brain) <I >1280
Rattus sp. @rain) <1 >320

*Data derived from Lacey, in press.

t13Hf MBZ binding was obtained from plots of 13H] MBZ bound over a range of
concentrations O-2.5 PM. The figures quoted are maximum binding observed at
saturation over the range of l-2.5 PM, with the exception of H. conforms (BZ-S)
and N. spathiger where binding at the maximum concentration is quoted since
saturation is not achieved over this range.
$m, Multiple treatments.
 Role of tubulin in benzimidazoles mode of action 917

site is not a hindrance given high fluid clearance rates. ible interactions. The presence of additional revers-
However, in the absence of effective clearance, the ible [‘HI MBZ binding over [“HI MBZ charcoal stable
elimination kinetics of the parasite are balanced binding has been characterized for F. hepatica by gel
against the kinetics of the BZ-tubulin interaction and filtration (Lacey, unpublished).
recovery can occur depending on the extent of The role of duration of exposure in both host
biochemical damage (Zintz & Frank, 1982). The role elimination dynamics and efficacy against insensitive
of time of exposure (either increasing the dose or or resistant species can be summarized by considering
multiple dosing) is well established in anthelmintic the equilibria:
efficacy (see Hennessy, 1985). 41
(1) BZ + Tubulin i;_, BZ-Tubulin Type I
Time dependence is a crucial component of BZ
action for any parasite in which the host removal
mechanism takes a significantly longer time than the k
(2) BZ +Tubulin -2
i;_,
drug’s in vivo pharmacological half-life. Empirically BZ-Tubulin Type 2
this limitation is overcome by increasing drug action
by higher doses and prolonged or repeated treatment,
thus achieving increased efficacy even in insensitive where the reversible (Type 1) complex dissociates at a
parasites such as E. granulosus cysts (Heath, Christie rate (k_,) as the BZ concentration falls by in vivo
& Chevis, 1975). While such improvements are excretion, while for charcoal stable pseudo-
considered as dose-related, it should be noted that irreversible binding (Type 2) dissociation occurs over
once equilibrium of the drug between host, parasite a longer time frame (k_; + k-,). Under in vivo condi-
and more specifically site of action is achieved, it is the tions Type 1 binding exists only when sufficient levels
duration of exposure not dose that is relevant, pro- of BZ are maintained while Type 2 binding will
vided sufficient receptor sites are filled. This was decrease at a rate independent of host pharmaco-
elegantly demonstrated in sub-divided dose studies dynamics. For both complexes, efficacy is dependent
where efficacy against BZ-resistant parasites was on whether k_, (or k_,) is greater than the rate of
increased compared with a single dose (Prichard, parasite expulsion.
Hennessy & Steel, 1978). Why the extent of BZ (or CLC) charcoal stable
While other aspects of BZ pharmacodynamics such binding varies with species is unknown. However, in
as metabolism, absorption, pharmacokinetics and view of the extent of microheterogenicity of tubulin it
excretion extensively modulate the action of BZs, is probable that the tubulin pools within each species
specificity of the BZ-tubulin complex and host are composed of multiple isotypes with structural
dynamics of parasite elimination remain the most changes conferring altered stability to BZ-isotubulin
important components of in vivo efficacy. complexes. Such a hypothesis would suggest that for
Since specificity is characterized by charcoal stable any species or any individual the overall BZ-tubulin
binding it is important to consider how efficacy is interaction is a composite of the isotubulin affinities.
achieved in those species where only low levels of Thus,
charcoal stable (‘H]BZ binding are observed. In the
BZ-Tubulin = C BZ-Tubulin, + BZ-Tubulinz +
extreme case of mammalian tubulin, BZs inhibit both
BZ-Tubulin, . . . BZ-Tubulin,,
polymerization and [‘H]CLC binding and can be
demonstrated to form a reversible complex with where each complex possesses a different affinity
tubulin (Tables 3 and 4). In vitro, toxicity to BZs can which could be due to either a change at the binding
be demonstrated under static conditions where drug is site or an allosteric effect due to changes which alter
in continuous contact with cells (Styles & Gamer, the conformation of the binding site (Lacey, in press).
1974; Lacey & Watson, 1985b) but not in dynamic in The relative affinity of other microtubule inhibitors
vivo situations (Table 8). This implies the existence of for parasite tubulin can be measured by pre-
a biological reversible equilibrium between the extra- incubation of inhibitors with H. contortus L, tubulin
cellular and intracellular drug pools which are in turn using the resistance assay technique for [3H] MBZ
at rapid reversible equilibrium with tubulin. When (Lacey & Snowdon, 1988; Lacey, Gill & Murray,
drug concentrations decrease by either metabolic or unpublished results). The IC,,, values for a selection of
excretory processes, complex dissociation occurs these compounds are presented in Tables 6 and 8.
reducing the extent of BZ inhibition. Therefore it is Within the BZ class the relationship between TBZ,
likely that BZ binding in all species is composed of a CBZ and the newer BZCs follows that expected from
minimum of two components, charcoal stable and in vivo efficacy studies. BZCs show subtle differences
reversible charcoal non-stable fractions. Thus in vivo in activity, but all compounds are potent inhibitors at
use of divided doses or prolonged administration is a concentrations well below those observed in viva. For
compromise between the dynamic conditions of drug structurally unrelated inhibitors an interesting pattern
elimination in the host and static conditions of drug of activities is observed. CLC is approximately 600-
availability seen in most in vitro techniques where fold less active, an observation consistent with CLC
eggs and larvae are constantly exposed to drug levels activity in lower eukaryotes while the only other
necessary to sustain microtubule inhibition by revers- colchicinoid derivative which possesses activity is the
918 E. Lncsv

TABLE S-INNIBITION OF PHI MB2 BINDING

TO extensiveiy reviewed elsewhere (Uialler & Prichard.
Fi. CO~~~~~ZlS L, CRIJDE T”B”LtN EXTRACT BY CLC-SITE 1986; Waller, 1986).
LIGANDS”? In parasitic nematodes several attempts to under-
stand the nature of this inheritability have been made
Drug IC,,, % Inhibition$
by quantitative genetic approach. Using BZ-R and
C&hi&e 100 90 (1OOO~~) BZ-S H c5~~or~~ isolate crosses, Le Jambre, Royal
Allocolchicine - NI (2000 p M) & Martin (1979) concluded that BZ-resistance was
pn-MTPB - NI (2000 PM) due to a single dominant trait within a polygenic back-
MTPT - 47 (2000 PM) ground based on egg-hatch assay results. in vivo
a.1
efficacy studies with CBZ-resistant H. contortus
Podophylloto~n 93 (IOO~M)
Steganacin - 25 (loo FM) cusses, however, suggested a recessive gene also with
NSC 350102 - NI (100 PM) polygenic involvement (Herlich, Rew & Cotglazier,
NSC 321567 - NI(l00~~) 1981). Recent work in T. c~~~br~orf~~ also found
unacceptable departure from the expected population
CbloroIPC 1630 54 (2000 ,UM)
- ratios for a single gene effect (Martin, 1987). These
Tubulozole C NI (lool?pM)
Tubulozole T - NI (1000 ,UM) genetic studies in parasites can now be viewed in the
NSC-215915 - 19 (1OOPM) light of [“HI BZ binding studies which demonstrate the
NSC-225635 - NI(l00~~) role of tub&in in BZ-resistance. ~~dividuai worm
WC-181928 2.9 90 (20 ,fkMf analysis by quantitation of I’H] OBZ binding by BZ-S
90 (IO FiM)
and BZ-R isolates of H. ~o~~o~~l~~demonstrated that
MB2 0.2
while the BZ-R isolate was composed of a single
*Other inhibitors tested which were inactive included population, the BZ-S isolate was heterogeneous con-
coumarin, rotenone, DAPI, Griseofulvin, chlorpromazine, taining a bimodal distribution with six local maxima or
promazine, melatonin, vinblastine, maytansine and inflections within the population (Fig. 16, from Lacey,
phomopsin A. Eagleson & Smith, in preparation). The BZ-S popula-
*Data derived from Lacey, Gill & Murray, unpublisb~ tion binding overlapped that of BZ-R population
results. (approximately 2% of BZ-S) and extended to 1OX the
$Percenrage inhibition observed at the highest concentra- BZ-I? binding mean. This suggested that susceptible
tion tested, in brackets. isolates were not homogeneous and that in BZ-S X
BZ-R crosses the emergence of apparent polygenicity
was unavoidable. The existence of microheterogenity
AC ring analogue, MTPT. Both the microtubule of [“H] BZ binding is not unexpected since [‘HI MBZ
specific isomer tubulozole C and its inactive analogue binding in a number of different BZ-S isolates of
tubulazole T were inactive as were several other Ei. confortus and also T. colubr~or~~s have been
carbamate analogues, confirming that the presence of shown to differ (Lacey & Snowdon, 19SS). The
the carbamate moiety per se is not a determinant for origins of variation of [‘Hj BZ binding to individual
activity (as suggested by Gupta, 1984). The deazadi- parasites in unselected BZ populatii~ns is unkn~~wn
hydropteridines (for example, NSC 18 1928) showed but is consistent with the mjcroh~terogeni~ity of
comparable inhibitory activity to BZs. Subsequent tubulin within eukaryotes. The existence of multiple
testing of a series of dihydropteridines as inhibitors of isotypes of both a- and ~-tubulin can be traced
both egg hatch and larval development honeyed this directly to the genome of tubulin and does not neces-
activity as potentially anthelmintic (Lacey, Gill, sarily involve post-translationat modification (Lee,
Temple, Combes & Rener, unpublished results). Field, George & Head, 1986). With the exception of
Interestingly, no activity was observed for the VBL yeast and Tm&zymenn, both a - and P-tubulin genes
binding site ligands or several other classes of CLC from all species thus far examined are multigene
binding site ligands. Of the other inhibitors PDT, but families (Cleveland Rr Sullivan, 1985). In summary.
not its partial analogues NSC 350102 or NSC B&resistance is probably a polygenic phenomenon
32 I567, showed moderate activity. a property recog- related to the tub&n genome.
nized in the medicinal uses of ~od~~~~~~l~~ species Intensive interest in the functionsof microtubules in
(Hartwell & Schrecker, 1958). eukaryotic cells led in the late-1970s to development
of tubulin mutants in many model organisms (Table
BENZIMIDAZOLE-RESISTANCE: GENETIC 9). For parasitology these fundamental studies prv-
MUTANTS, MODELS AND PARASITES vide an important perspective of the potential resist-
Resistance to BZs was first reported in the early- ance mechanisms to BZ chemotherapy.
I%% foltowing the release of TBZ (Conway, 1965; Resistance in these models to BZs and other CLC-
Drudge, Szanto, Wyant & Eiam, 1964). Since this site ligands arises by structural changes which involve
time, selection of BZ-resistance in the field has both tubulin and non-tubulin proteins (Table 9). the
emerged as a widespread, albeit sporadic, problem in latter including proteins specific to the microtubule
the control of gastrointestinal nematodes in grazing matrix and proteins independent of it such as perme-
animals. These aspects of resistance have been ability mutants (Ling & Thompson. 1974; Welker &
 Role of tubulin in ben~imid~oles mode of action 919

occurrence of conformational changes to the CLC-

site in tubulin mutants lends itself to the existence of
both resistant and sensitive mutants (Cabral, Brady &
Schibler. 1986). Observation of differential sensitivity
between CLC-site hgands in some mutants provides a
helpful insight into the nature of lig~d-tubulin inter-
actions within the CLC-site. For instance, [‘“Cl MBC
binding to the benomyl resistant mutant ben 13 and
the hypersensitive mutant ben 14 in A. nidulans
reflects in vitro growth toxicity to benomyl (Davidse
8r Flach, 1977) but not to TBZ (Sheir-Neiss et al.,

ys\:;
1978). The resistance factors of ben 13 and ben 17 to
benomyl and TBZ indicate that they possess a differ-
““i/ T; ent resistance mechanism to the majority of the ben
mutants (Sheir-Neiss et al., 1978), while the inverse
relationship of highly benomyl-resistant yet relatively
TBZ-sensitive mutants for ben 1 and ben 16 suggests
a completely different mechanism.
In CHO cells, selection of mutati~)ns with NDZ and
FIG. 16. The population frequency of f”H]OBZ binding to PDT provides evidence supporting a closer bio-
adult female H. con&~rfus.T,, T, , . . denote IocaI maxima or chemical interaction of NDZ, PDT, NSC 18 1928 and
inflections within the distributions. TN-16 with tubulin compared to that of CLC and
indicates a lack of correlation between TBZ and the
Williams, 1983). Mutations occur in both a- and /I- other BZs (Gupta, 1986).
tubulin, often associated with altered apparent molec- Comparison of the results derived from these
ular weight and isoelectric points (Sheir-Neiss et a/., models to parasitic nematodes suggests that only a
1978). Of the BZ-resistant mutants characterized to small proportion of resistant mutants are relevant to
date, the majority of mutations are located on /?- the mechanism of resistance to BZ in viva where
tubulin with only a few a-tubulin mutants reported. resistance is specifically due to a reduction in [“H] HZ
This has led many researchers to conclude that the binding. Within laboratory models the potential exists
CLC-site is located either on /3-tubulin or at the a -/? to define both the extent and nature of mutations
interface (Oakley, 1985). The latter hypothesis is a which confer BZ resistance and to develop a frame-
compromise between the tubuhn mutation data and work for characterizing resistance mechanisms. At
affinity labelhng and electrophoretic studies which present 13H]BZ binding in helminths can only identify
suggest an a-tubuiin CLC binding site. However, an resistance and not distinguish whether the same
alternative explanation for the preponderance of /3- mechanisms are involved in different isolates or
tubulin BZ mutants which cannot be discounted may species.
reflect that lethal mutations tend to occur in a-
tubulin.
THE NATURE OF MICROTUBULE INHIBITOR
[‘HI Ligand binding has been correlated to in vitro
toxicity to A. ~~dliIa~s mutants, ben 13 and ben 14 I~E~CTION AT THE COLCHI~INE BINDING
(Davidse & Fiach, 1977), and coicemid CHO cell SITE
mutants (Keates, Sarangi & Ling, 1981). However, The pharmacology of BZs at the biochemical, in
phenotypic resistance is not always associated with vitro and in vivo levels is consistent with a primary
alterations in [“H] l&and binding (Cabral, Sobel & site of action on tubuhn. Thus it is appropriate to
Gottesman, 1980; Gupta & Gupta. 1984). For the consider the nature of the binding site as the focus not
majority of mutants, [“H] ligand binding is uncharac- only of BZs but also of the other classes of compounds
terized. which occupy this site. Despite extensive biochemical
In many mutants the occurrence of aberrant resist- characterization of many CLC-site ligands, no broad
ance profiles reflecting multiple ligand site changes consensus has emerged which is compatible as a
(CLC, VBL and taxol), temperature sensitivity and general hypothesis for ligand interaction. Several
abnormal microtubule morphology suggest that many models for specific CLC analogue interactions with
resistant mutations involve altered microtubule varying degrees of complexity are relevant to an
dynamics (Oakley & Morris, 1981; Oakley, 1985). overah model, however these lack general application
These phenomena probably result from structural (Margulis, 1974; Andreu & Timasheff, 1982; Rossi,
changes in tubulin which alter the normal conforma- Link & Lee, 1984; Bane et al., 1984). In developing a
tional changes associated with ligand binding and model for the CLC-site it is necessary to address not
polymerization. In such mutants it is the functional only the structural heterogeneity of ligands but also
role of the site as an inhibitory control mechanism that provide a framework for consideration of other
is being overridden, not ligand binding per se. The factors such as species-dependent ligand interaction
 TABLE 9-CLC-sl-re LIGAND RESISTANCE STUDIES IN LABORATORY MODEL ORGANISMS*

organism Selection
criteria Method Characterization Comments RefereRCeS

Yeast
Saccharomyces benomyl spontaneous sensitivity to 173 R mutants isolated, Thomas, Neff & Botstein, 1985
cerevisiae temp. of those tested all map
morphology, to tub2 @-tubulin in gene)
molecular 15/173 dominant, 50/173
genetics recessive
(1 mutant) 1 l/173 ts for growth
One mutant contains amino
acid change at 241 Arg
for His on fi-tubulin

SchizosaccharomycesTBZ spontaneous sensitivity to mutants map to three loci- Yamamota, 1980

pombe TBZ temp. ben, 1,2,3
nitrosoguanidine cross-R to MBC
ben 1 mutants dominant,
ben 2, 3 recessive, ts
spontaneous sensitivity to nda 2 mutants BZ-SS and cs Umenesono, Toda, Hayashi & Yanagida,
TBZ, MBZ, 1983
NDZ nda 3 mutants both R and SS

and cs

Fungi
A. nidulans benomyl EMS sensitivity to R in 3 unlinked loci ben A, Sheir-Neiss ef ai., 1978
2D
electrophoresis B, C
peptide mapping 18/26 ben A mutants show
Co-
polymerization altered /&tubulin
with brain
tubulin 4/28 mutants show differences
in RF between TBZ and
benoymi
TBZ RF range 1.5-16, benomyl
range 0.24-40
3/26 ben A mutants ts

Slime moulds
Dicfyosreiium benomyl nitrosoguanidinc sensitivity to 18 mutant loci-6/l 8 possibly Welker & Williams, 1983
discoideum coumarin TBZ, MBC tubulin retated
acriflavin benomyl,
temp., 12/18 permeability or uptake
morphology mutants
S/61 TBZ R mutants are ts
5/6 1 altered spore morphology
Role of tubulin in benzimidazoles mode of action 921
922 E. LACEY

and the fundamental role of the site. As part of this but to a conformation with only partial carbonyl
review a hypothesis for the generalized CLC-site character.
ligand interactions is advanced, derived from an over- One further feature of a u complex hypothesis
view of the current literature (Lacey, Burden & which must be accounted for within the tropolones is
Watson, in preparation). the enhancement of CLC fluorescence on binding.
Bhattachar~a & Wolff (1984) have proposed that
General binding site hypothesis this phenomenon is linked to the immobilization of
CLC-site ligands bear two identifiable chemical CLC within the site, however a second interpretation
similarities: (i) a moiety sensitive to nucleophilic sub- which can be derived from the u complex hypothesis
stitution, and (ii) a high degree of structural restraint is that fluorescence is due to extended conjugation
which distinguishes potent from weak activity within caused by electron delocalization. To examine this
each class. possibility the fluorescence spectrum of CLC in the
For CLC and its active analogues, the nucleophilic presence of hydroxide and thiols was compared to
site is the methoxytropone system. This becomes that of colchiceine in base. In all cases, fluorescence
more obvious when the system is considered chemi- enhancement (20- 1 OO-fold) occurred compared with
cally as a vinylogous ester (Cook & Loudon, 1951). CLC or unionized colchiceine. This suggests that the
Nucleophilic displacement by water (hydrolysis), existence of an ionized tropolone as postulated by this
amines and thiols are well documented (Zweig & hypothesis will enhance fluorescence (Lacey &
Chignell, 1973; Shiau et al., 1976). For BZCs and Moore, unpublished results).
other carbamates displacement of the methoxyl group While the formation of a o complex is the final
(hydrolysis and transcarbamoylation) of the carbam- stabilized stage of nucleophilic ligation, intermediates
ates occurs under similar conditions (Lacey, 1982, where a 6+ . . d- polarization exists between the
thesis cited above). Trans-esterification, hydrolysis ligand and cysteine moiety will occur. For many less
and displacement of the lactones (podophyllotoxin) active ligands (for example TBZ), these intermediates
or esters (allo-CLC) are also fundamental nucleo- will constitute the maximal extent of nucleophilic
philic chemical reactions. Nucleophilic sites are also reaction, while for other ligands where no polarization
present in ~~-unsaturated ketones (gr~seo~lvin) and exists binding will depend on the absence of unfavour-
more obvious reagents such as alkylating and aceylat- pble steric interactions with the thiol together with
ing agents which interact at the CLC-site. acceptable interactions with adjacent amino acids.
This common chemical feature implies reactivity However, the existence of 6- character at the nucleo-
with a nucleophile such as SH, -NH, and -OH philic site should be incompatible with activity as
moieties. The observation that selective derivatiza- observed for colchiceine, 2-aminobenzimidazole and
tion of cysteine (Kuriyama & Sakai, 1974; Ikeda & the podophyllic acids.
Steiner, 197X) but not lysine (Mellado et af., 1982; The general shape of this binding domain is relat-
Szasz, Burns & Stemlicht, 1982) inhibits CLC-site ively planar as reflected by the observation that all
binding implies that cysteine may be involved in ligands capable of either forming a (T complex or
ligation as part of complexation. Possible involvement achieving partial stabilization are themselves planar
of serine may be postulated, however its lower nucleo- or able to adopt relatively low energy planar confor-
philicity would make this unlikely for all but esters and mations while non-inhibitors such as picropodophyl-
lactones. lin are unable to adopt such conformations (Zavala et
For CLC, PDT and BZCs, the postulated scheme al., 1980).
for nucleophilic substitution would involve reversible Acceptance of this hypothesis implies that kinetic
formation of a tetrahedral ‘u complex’ which would models for CLC and tropone binding based on
ultimately lead to covalent-linkage by elimination of fluorescence (Garland, 1978; Bane et al., 1984;
the alcohol. Since covalency is not observed for these Engelborghs & Fitzgerald, 1987) are only.monitoring
ligands, it is reasonable to assume that within the site, this aspect of the ligand binding in the C rmg.
nucleophilic attack is held to a stabilized cr complex Within existing CLC-site models, the hypothesis of
(Fig. 17). For CLC and other tropolones the o a bi~nction~ A-C ring model has been established.
complex is resonance-stabilized by electron delocal- The trimethoxyphenyl (TMP) A ring is considered a
ization. Of the resonance structures possible, the common domain for CLC, PDT and steganacin. To
enolate form is a dominant contributor to electron achieve similar orientation of PDT and CLC to a
density since the more electronegative (a-) atom nucleophilic site requires the methylenedioxybenzene
(oxygen) bears the negative charge. Consequently, the moiety to occupy the TMP CLC-site. This observa-
9-keto group would possess appreciably less double tion is consistent with the structural prerequisites of
bond character. This provides an interesting extension the benzylbenzodioxoles where only a 4-methoxy-
to the hypothesis relating to CD spectral changes in phenyl confers activity (Batra et al., 198.5) in what
bound CLC proposed by Detrich er al. (198 1) in that would constitute the PDT equivalent of the TMP
on CLC binding to tubulin the tropolone ‘puckering’ ligand in this series. The occurrence of a 3,4-disubsti-
undergoes a conformational change which leads to tuted phenyl binding to the A ring site is also the
not only a more planar conformation (as postulated) second most favourable substitution within the
 tb> Podophyllotoxln

Pl@“CDCH3
R t M-R
‘*me t HS-R HS-R

j -CH30H
i j -CH30H

NN-
I
R q CH&H
//.-\
$!c-- P R Ni&SR

FOG.17. The proposed mechanism of nucleophilic attack by a cysteinyf residue on colcbicine, podophyllotoxin and benzimidazole carbamates.
For colchicine the major resonance contributors are noted.

.__. ._I,,
_.“_. _._._--.. .I
. -.._-_ .-
.- .., ._ - - -
_. _. . - - _I
_, _ _ ., ,_ _. -
924 E.LACEY

m-MTPB series (Lacey, Burden & Watson, unpub- hydrophobic and/or electrostatic interactions. This
lished results). complexation is driven by the relative repulsion of
This binding domain for the methylenedioxy ring ligands from a hydrophilic environment but re-
would place the TMP ring of PDT and the steganacins strained by an absolute requirement for compatible
in a region corresponding to the B ring substituent for geometry between AC and the AC site. Incompati-
CLC, B ’ . Whether the binding of the A ring is due to bility may lead to either an AB ’ or possibly B ‘C
an optimal hydrophobic or size effect is unknown as preferred fit or complex dissociation. Subsequent
insufficent data exist to distinguish these properties, conformational optimization will involve formation of
For the A-C region an optimal length between the a 6+ d- interaction between the C ligand and
nucleophilic centre of CLC and its analogues and the available polar amino acids. Where 6+. d- is due to
4-methoxy group of the TMP ring is approximately 9- cysteine and relevant geometric and energetic con-
10 A (Lacey & Burden, unpublished results). siderations are favourable, the u complex forms. The
Binding in the B ’ region is relatively structurally conversion of a sp? carbon (trigonal planar geometry)
non-specific compared to the A ring (Quinn & to sp” (tetrahedral) induces a major conformational
Beisler, 1981) with activity again dependent on the change in the ligand. Further interaction probably by
lipophilicity and/or size of the substituent. Super- hydrogen (H)-bonding to the protein is considered
imposing the structure of the BZCs and dihydropteri- necessary to stabilize the 6- character created on
dines on CLC with interaction to tubulin through a either the carbonyl oxygen of carbamate or PDT or
single attachment to a common nucleophilic site the tropolone 9keto group (Fig. 17). Due to the
cannot accommodate the 5-substituent of BZ or the distance of cysteine from the b- oxygen, a second
3-substituent of the deazadihydropteridines within polar amino acid, such as tryptophan, arginine,
the structural limitations of an A-C model. Since the histidine, serine or lysine must be adjacent to the
unsubstituted BZ and deazadihydropteridine carba- cysteine to stabilize the increased electronegativity by
mates themselves are greater than 10 A, optimal fit of acting as an H-bonding donor.
these derivatives and most other aromatic carbamates The pseudo-irreversible binding kinetics of CLC
will result in the substituents’ binding to an extended differ from those of all other CLC analogues not
A region, A ’ which exists either above or below the possessing an amido moiety in the 7-position (deme-
A-C plane (for convenience only the above plane coline, desacetamido colchicine and MTPT) and
option will be considered). other structurally unrelated inhibitors which show
The large size of CLC B ring substituted analogues reversible kinetics (Table 4). Since the interaction of
(Clark & Garland, 1978; Zimmerman et al., 1982; brotro-CLC with tubulin followed by acid treatment
Salmon & Wadsworth, 1986; Williams et al., 1985) leads to formation of cysteinyl acetic acid
and their structural diversity are consistent with the (CH,COOH is derived from BrCHCONH- of the
structure of the 5-substituted BZs and 3-substituted 7-substituent by alkylation and hydrolysis) (Schmitt &
deazadihydropterines and suggest that both A ’ and Kram, 14,7X),a further d+ . b- stabilization between
B ’ regions are structurally relatively non-specific a cysteine and the acetamide carbonyl could be postu-
within the geometric restraints imposed by the A-C lated to occur within the B ’ domain for CLC. This
binding region. interaction may represent the relatively slow equilib-
For this general binding site hypothesis, the B ring is rium which leads to CLC pseudo-irreversibility and
fused within the A-C planar domain consistent with the requirement for higher positive enthalpy for CLC
the maintenance of inhibitory activity on introduction compared to other tropolones leading to the charac-
of unsaturation within the B ring (Rosncr ef al., 198 I), teristic temperature dependency of free CLC-tubulin
to a general site described according to the modified interaction.
CLC template as AA ’ B ’ C as shown in Fig. 18. The In general it would appear that sensitivity to colchi-
dimensions can be approximated based on the size of noids is a relatively recent development in the evolu-
various inhibitors. For both A ’ and B ’ no upper size tion of the CLC-site with high affinity [“HI CLC
restriction has been observed. binding restricted to higher eukaryotes (Hains,
Using this hypothesis CLC-ligands can be divided Dickerson, Wilson & Owellen. 1978). The majority of
into four major classes AC (MTPT, MTBT, MBC, lower eukaryotes such as yeast, fungi and slime moulds
TBZandgriseofulvin), AA ’ C (substituted BZCs, sub- are relatively insensitive to CLC (up to mM concentra-
stituted deazadihydropteridines, tubulozole and other tions) with little or no detectable [“HI CLC binding.
aromatic carbamates), AB ‘C (CLC, PDT and stega- For helminths. [“HI CLC binding can be detected
nacin) and AB ’ (substituted benzylbenzodioxoles, (Table 6) but in [‘HI MBZ displacement studies, CLC
combretastatin and possibly TN- 16). For mammalian is two-three orders of magnitude less active, while
tubulin the extremely weak activity of so-called ‘single many of the other colchinoids are completely inactive
site’ analogues such as methyltropolone and trimetho- (Table 8). Podophyllotoxin has been less extensively
xybenzenes suggests that a minimum occupation of two studied. however, comparing helminth and mammal-
binding domains is required for potent activity. ian data. its activity is reduced but not to the same
It is probable that binding to the CLC-site requires extent as CLC (Tables 3 and 8). This suggests that
an initial fit to all regions simultaneously by weak changes to the A and B ’ domains have occurred.
 Role of tubulin in benzimidazoles mode of action 925

(i) [ii)
i

FIG. 18. A generalized view of the shape of the proposed CLC-site based on the ABC rin8 system of colchicine. A, A’, B’
and C reoresent the structural domains for hgand interaction. This model was derived from structural overlap of MB’& PDT,
CLC and isopicrostegone by assumption of ccommon nucleophilic site for binding. Regions B ’ and A ’ have been extended in
accordance with the size of other substituents known to bind in these regions. a and b correspond to the steric binderance
region of BZs noted in Fig, 10 and the region of the nucleophilic site, respectively, while c represents the region of this
alkylation for CLC described by Schmitt & Kram (1978). Inserts (i) and (ii) are representations of the postulated cysteiny1
arrangement within the binding site.

BZs are potent inhibitors of all eukaryotic tubulins data derived from ‘single site’ analogues, for example,
thus far studied. Within this broad specificity, how- trimethoxybenzenes, benzylalcohols, benzaldehydes,
ever, optimum specificity due to a change in the BZ- methoxytropone and others, have been discounted,
tub&in interaction occurs in lower eukaryotes. since none of these derivatives are microtubule in-
Comparison of the relative activity of BZs towards hibitors within two-three orders of magnitude of their
helminths and mammalian tubuiin supports the view parent structure {Fitzgerald, 1976; Andreu &
that some structural refinement of the A’ domain Timasheff, 1982). JR fact most are less active: than
has occurred, since compounds such as OFZ, RBZ many ‘non-inhibitors’ such as isocolchicine and
and other more polar BZs are significantly less active colchiceine. Further, these derivatives lack the geo-
against mammalian tubulin, but that its broad struc- metric specificities inherent in their parent analogues
tural specificity has remained unchanged. From [“H) which are essential for their placement within the
MBZ displacement studies, it is apparent that struc- postulated binding domains, In the absence of ‘singLIe
tural changes in all binding domains have occurred. site’ ligands active at realistic concentrations
The extent of these changes are more dramatic in the (<~~O,UM) with supporting SAR data, structural
B ’ region since both AB ’ C and AB ’ ligands are inferences drawn from these derivatives are out of
largely devoid of activity. proportion to their importance as CLC-site specific
Considering a general CLC-site mode of action, ligands and as such should be considered with caution.
926 E.LACEY

Location of the colchicine site that 50% inhibition of [“HI CLC binding requires the
Evidence for the location of the CLC-site on a a- alkylation of a minimum of three cysteinyl residues
tubulin has emerged in several alkylation and photo- (Kuriyama & Sakai, 1974; Ikeda & Steiner, 1978).
activation studies with substituted B ring analogues While such a hypothesis accounts for many of the
(Schmitt & Atlas, 1976; Williams et al., 1985). observations in the literature there are several critical
Ligation has been shown to occur at approximately 2 experiments necessary to substantiate this model.
and 12 A, respectively, from the B ring itself. These Firstly, this location suggests that by appropriate sub-
consistent observations of a -tubulin association stitution of ligands it should be possible to specifically
suggested that the entire B ’ binding domain is located cross-link a- and p-tubulin. Secondly, by altering the
on a -tubulin (Schmitt & Atlas, 1976; Williams et al ., position of photo-affinity label on various CLC-site
1985) and that the region is proximal to a cysteinyl ligands it should be possible to selectively label either
residue (Schmitt & Kram, 1978). a - or B-tubulin and finally it should also be feasible to
CLC binding to tubulin prior to and after trypsino- develop a specific cross-link between the cysteinyl
lysis has localized the CLC-site within a 5000 mol. wt residues on a -tubulin.
fragment of a-tubulin (Serrano et al., 1984). Based
on the trypsin cleavage pattern, it localizes this Function of the colchicine site
sequence between Arg 339 (Sackett & Wolff, 1986) In contrast to our knowledge of the role of MAPS
and possibly the Arg 373 residue. At present insuffi- and other proteins in control of the tubulin-
cient analogue displacement binding data exist to microtubule equilibrium, the role of the CLC-site in
characterize whether binding is due to either indivi- endogenous regulation has received little attention. In
dual or multiple A-, B ’ or C-site binding. Inter- a series of papers Lockwood (1978, 1979) identified
estingly, a cysteine is located adjacent to a potential the existence of several proteins and peptides which
H-bonding donor, tryptophan at 346, with a highly inhibited [“HI CLC binding. Other studies (Sherline et
conserved 8 amino acid sequence within this 5000 al., 1979; Koehn & Olsen, 1980) have reported the
mol. wt region. The involvement of tryptophan as isolation of other proteins both soluble and particul-
H-bonding donor is compatible with tubulin fluores- ate with similar actions (Table 1). Although several
cence studies and tryptophan selective alkylation with competitively inhibit [“HI CLC binding, the nature of
Koshland’s reagent supporting tryptophan interaction their interaction is uncharacterized. Given the relative
in ligand binding (Maccioni & Seeds, 1982). Further sizes of the proteins (> 15,000 and 250,000) specific
this tryptophan is thought to be proximal to the a--P allosteric modulation of the CLC-site by surface
interface (Prasad et al., 1986). binding to the dimer may occur. Although no ligands
The observation that CLC-site ligands inhibit of comparative size to CLC-site ligands (< 1000) have
formation of the /3*-tubulin cross-link between been identified, the activity by a haem nonapeptide
cysteinyl residues at 239 and 354, which in the tertiary conjugated to desacetyl-CLC (Zimmerman et al.,
structure are only 9 A apart, has provided evidence for 1982) suggests that competitive endogenous peptides
P-tubulin involvement (Little & Luduena, 1985). This (<4000) described by Lockwood (1978) are com-
distance is compatible with either an AC or AA’ patible with the CLC-site structure.
arrangement of cysteines (see Fig. 18). An alternative hypothesis to the endogenous ligand
Constructing a model for the CLC-site based on concept is that CLC-site ligands are intercalated into
available data reveals a site with three or four cysteinyl an essential regulatory domain, the status of which is
residues in close proximity. A /3-tubulin cysteine, normally controlled by allosteric interactions. In this
probably residue 239, is associated with the A ring hypothesis, the binding domain would exist in equilib-
domain while the 354 residue lies either in the B’ rium between an active or inactive conformation,
domain 9 A away or at the nucleophilic site in the C perhaps achieved by variation in the spatial arrange-
domain. An as yet unidentified a-tubulin cysteine ments of key amino acids such as cysteine within the
occupies the B ’ -tubulin. Alternatively, a fourth domain.
cysteine at residue 347 on a-tubulin binds the C ring By allosteric modulation by either other proteins,
as a u complex as shown in Fig. 18,Assignment of the C GTP or other mechanisms, the equilibrium may be
domain as either 354 cysteine in P-tubulin or the 347 directed towards conformational changes which
position in a -tubulin is based on their conservation in maintain the site in an active or inactive state for self-
species sensitive to BZCs where the 239 position of p- assembly. CLC-site ligands may act by binding into
tubulin is serine (Little & Luduena, 1985; Burland, the active state site inhibiting the subsequent changes
Paul, Oetliker & Dove, 1988). The a -tubulin cysteinyl required for polymerization. The concept of the CLC-
arrangement is based on the likelihood that the u site as a regulatory domain as opposed to an endogen-
complex is the most stable domain and thus less ous ligand site would account for its evolutionary
sensitive to destruction by trypsinolysis. Each of the conservation as critical to the self-assembly process
identified cysteines is enclosed in a relatively con- and also account for the parallels that exist between
served sequence not associated with other recognized [‘H] CLC binding and self-assembly status such as
binding sites. The hypothesis of a cysteinyl rich CLC- temperature dependent time-decay, effects of chern-
site is consistent with thiol labelling studies that suggest cal modification, stability to pH and ionic strength.
 Role of tubulin in benzimidazoles mode of action 927

This hypothesis would enable a more satisfactory mammalian and parasite data it is apparent that the
explanation of the diversity of CLC resistant mutants ability to bind tubulin is subject to varying degrees of
since these mutants need not specifically alter the selectivity within the class (for example, OFZ and
structure of the CLC-site but just the regulatory RBZ) which suggest a broadening of the SARs for
function of the site as suggested in many of the hyper- interaction within the 5substituent binding domain
stabilized microtubule mutants (Oakley, 1985). and may indicate structural changes in the A ’ region
Further, the existence of inactive conformations could of tubulins of these species. However, it is the nature
account for failure to observe molar stoichiometry of binding kinetics and the existence of a pseudo-
between CLC-site ligands and tubulin under non- irreversible equilibrium in helminths which is respon-
equilibrium conditions. sible for the observed species specificity of BZs as a
class. The extent of binding as defined by the charcoal
CONCLUSION extraction technique parallels with the extent of drug
The tubulin-microtubule equilibrium is essential to resistance in H. contortus, T. colubriformis, 0. cir-
all eukaryotes. The status of this equilibrium is cumcincta and N. spathiger and also the dose-
controlled by the interaction of both activating (GTP, efficacy spectrum of BZs in other parasites. In vivo
Mg2+) and inhibitory (Ca2+) co-factors as well as a this relationship is modulated to a minor extent by
diverse range of associated proteins. Central to host pharmacodynamics, however these factors do
endogenous regulation is a number of structural not significantly alter overall pharmacology. The most
domains, the conformations of which are extensively important factor responsible for extensive modulation
modulated during the assembly of tubulin. While of tubulin binding and efficacy relates to the mechan-
elucidation of the structures of binding domains is still ism of parasite expulsion. Since the action of BZs is
some way off, the structural diversity of inhibitors non-lethal in non-developing stages such as the adult
which interact at these sites and their resulting phar- or arrested larvae, efficacy is dependent on the
macological activity are consistent with an essential relationship between the kinetics of the BZ-tubulin
role for these domains within tubulin. interaction and parasite expulsion. Where the rate of
The relationship between inhibitor binding and expulsion is slow compared with BZ-tubulin binding,
activity at the biochemical, in vitro and in vivo levels efficacy will be low and erratic despite both bio-
is most obvious in developing cells where the utiliza- chemical and parasitological observations that BZs
tion of tubulin within the mitotic spindle is dramati- are exerting an effect. There are two aspects to this
cally affected with lethal consequences. In problem. Firstly, for helminths with high levels of BZ
non-developing cells, the relationship between bind- charcoal stable binding (D. immitis and 0. gibsoni) it
ing and physiological action is less distinct, in that is necessary to alter drug availability such that the
while extensive biochemical changes are observed kinetics of binding and expulsion occur during com-
they are not associated with lethal effects over the parable time frames. Secondly, for parasites with low
same time scale. The diverse actions of inhibitors can charcoal stable binding, it is necessary to re-design
in some cases be directly correlated to tubulin binding BZs with higher levels of pseudo-irreversible binding.
but in many cases the relationship is obscured by This approach is already being undertaken to develop
indirect interactions occurring by routes such as BZs with activity against BZ-resistance isolates
inhibition of hormone, neurotransmitter and enzyme (Lacey, unpublished data).
activity. The involvement of tubulin in many of these The factors that control the nature of BZ binding
processes is, in most cases, poorly understood. kinetics are unknown. It is probable that BZ binding
Within the structural domain for CLC-site ligand and thus efficacy is not due to a single equilibrium but
interactions a general hypothesis is presented to to the sum of the constituent isotubulin affinities. The
account for ligand heterogeneity. This hypothesis, diversity of data derived from resistance studies in
based on the tricyclic ring structure of CLC as a laboratory model organisms suggests that consider-
template, defines a common site for nucleophilic able flexibility within tubulin exists to modulate BZ
reactivity with a cysteinyl residue for the C ring with binding by mutation. Particularly interesting is the
regions of broad structural specificity in the A ’ and association of BZ-R and temperature sensitivity
B ’ regions with a more structurally specific A region. suggesting that species specificity to BZ and other MT
In essence the most active CLC-site ligands can be inhibitors is perhaps pre-conditioned by environ-
considered as geometrically constrained ‘soft’ alkyla- mental temperature.
tion reagents. The required geometries of the in- The affinity of CLC-site ligands for helminth
hibitors with interaction by either AC, AA ’ B, AB ’ C tubulin suggests that changes in ligand specificity for
or AB ’ binding are different though not mutually helminth tubulin is not limited to BZs. The potent
exclusive. From the available literature it is probable activity of the deazadihydropteridines suggests that
that this site contains a minimum of three cysteines other classes of microtubule inhibitors can be devel-
derived from both GL- and P-tubulins. oped as anthelmintics and these may provide alterna-
In parasites, the pharmacological action of BZs is tive approaches to the development of new generation
consistent with a primary mode of action against the BZs.
tubulin-microtubule equilibrium. When comparing In a broader sense, the characterization of the
928 E. LACEY

biochemical pharmacology of mammalian tubulin at polymerization and mitosis. Molecular Pharmacology 27:
other sites such as the VBL, taxol, associated protein 94-102.
and non-selective domains represents an exciting and BATRA J. K., POWERS L. J., HESS F. D. & HAMEL E. 1986.
largely unexplored arena for further development Derivatives of 5,6-diphenylpyridazine-3-one: synthetic
within the hehninth parasites and other lower eukary- antimitotic agents which interact with plant and mammal-
ian tubulin at a new drug-binding site. Cancer Research
otes. In this respect it is curious that while over 16 BZs
46: 1889-1893.
and related compounds which exploit one aspect of the BEAMES C. G., MERZ J. M. & DONAHUE M. J. lY76. Effect
biochemical pharmacology of tubulin in helminths of anthelmintics on the short circuit current of the
have been developed commercially, little development intestine of Ascaris suum. In: Biochemistry of Parasites
in other metazoan and protozoan parasites has been and Host-Parasite Relationships [Edited by BOSSCHE H.
reported. VAN DEN), pp. 581-587. North Holland, Amsterdam.
BEHM C. A. Rr BRYANT C. 1985. The modes of action of
Ackno~~edgen~nts-I would like to grateFully acknowledge some modern anth~imintics. In: Res~tance in nematodes
the contributions of my colleagues Drs Jenny Gill, Warwick to Anthelmintic Drugs (Edited by ANDERSON N. &
Grant, Nicholas Sangster, MS Phillipa Smith and Mr Greg WALLER P. J.), pp. 57-67. CSIRO Division of Animal
Russell for their appraisal of the review. Thanks are also Health, Glebe, NSW.
extended to MS Sally Pope and Mr Peter Burden for assisting BENNETT J. L. & THOMPSON D. P. 1986. Mode of action of
with the figures, MS Antoinette D’Cruz for her tireless antitrematodal agents. In: Chemotherapy of Parasitic
efforts in the production of the manuscript and MS Jill Diseases (Edited by CAMPBELL W. C. & REW R. S.),
Franklin for collation and checking of the references. I pp. 427-443. Plenum Press, New York.
would also like to extend my thanks to the Australian Wool BERNIER-VALE~TIN F. & Rouss~r B. 1982. Interaction of
Corporation under whose auspices many of the published tubulin with rat liver mitochondria. Journal of Biological
and unpublished studies derived from my laboratory were Chemistry 257: 7092-7099.
funded. BERNIER-VALENTIN F., AUNIS D. & Roussm B. 1983.
Evidence for tubulin-binding sites on cellular membranes:
plasma membranes, mitochondrial membranes, and
REFERENCES secretory granule membranes. Journal of CelLBiology 97:
209-2 16.
BHATTACHARYYA 3. &WOLFF J. 1974.Promotionof~uores-
ABRAHAM I.: DION R. L., D~ANMU C., GOTTESMAN M. M. & cence upon binding of colchicine to tubuiin. Proceedings
HAMEL E. 1986. 2,4-Dichlorobenzyl thiocyanate, an of the National Academy of Sciences of the United States of
antimitotic agent that alters microtubule morphology. America 71: 2627-2631.
Proceedings of the National Academy of Sciences of the BAATTACHARYYA B. & WOLFF J. 1976. Anion-induced
UnitedStates ofAmerica 83: 6X39-6843. increases in the rate of colchicine binding to tubulin.
AKIYAMA T., TANAKA K., TANAKA N. Sr Nowo~c~n Y. Biochemistry 15: 2283-2288.
197X. Interaction of viomycin and aminoglycoside anti- BHATT~CHARYYA B. & W~LFP J. 1984. Immobilization-
biotics with tubulin and microtubules. Journal of Anti- dependent fluorescence of colchicine. Journal of Bio-
biotics 31: 1306-1309. logical Chemistry 259: 11,836 11,843.
ANIJREU J. M. & TIMASHEFF S. N. 1982. Interaction of BHATTACHARYYAB., SACKETT D. L.&Z WOOLF J. 19X5.
tubulin with single ring analogues of colchicine. Bio- Tubulin, hybrid dimers, and tubulin S. Journal of Rio-
chemistry 21: 534-543. logical Chemistry 260: 10,208-10.216.
ARAKAWA T. & TIMASHEFF S. N. 1984. The mechanism of BINDER L. J., FRANKFURTER A. & REBHUN L. I. 1986.
action of Na glutamate, lysine HCI, and piperazine- Differential localization of MAP-2 and tau in mammalian
N.N’-bis (2-ethanesulfonic acid) in the stabilization of neurons in situ. In: Lyynamic Aspects of I~~icrotubu~~
tubulin and microtubule formation. Journal of 3~ulogic~ri Biology. Annals of the New York Academy of Sciences,
Chernisrrv 259: 4979-4986. Vol. 466 (Edited by SOIFER D.), pp. 145-166. New York
Awanzr K., SCHULZ-KEY H., HOWELLS R. E., HADDOCK Academy of Sciences, New York.
D. R. W. & GILLES H. M. 1982. The chemotherapy of BO~KELHEIDE K. 1987.2,5-Hexanedione alters microtubule
onchocerciasis-VIII. Levamisole and its combination assembly-II. Enhanced polymerization of crosslinked
with the benzimidazoles. Annals of Tropical Medicine and tubulin. Toxicology and Applied Pharmacology 88: 383-
Purasitology 76: 459-473. . . 396.
AZHAR S., Hwn~c S. F. & REAVEN E. P. 1983. Effect of BINGE D., HEUS$LE C., SIMON C. Sr PAKTALONI D. 1985.
antimicrotubule agents on terminal ~ycosyltransferases 4 ‘,6-Diamidino-2-phenylindoie, a fluorescent probe for
and other enzymes associated with rat liver subcellular tubulin and microtubules. Journalof Biological Chemistry
fractions. BiochemicalJournal 212: 721-731. 260: 2819-2825.
BANE S., PUETT D., MACDONALD T. L. & WILLIAMS R. C. DORAY J. C. 19X6. Trematode infections of domestic
1984. Binding to tubulin of the colchicine analoaue animals. In: Chemotherapy of Parasitic Diseases (Edited
2-methoxy-5-G ‘, 3 ‘, 4’-trimethoxyphenyl) tropone. by CAMPBELL W. C. & REW R. S.), pp. 401-425. Plenum
Journal of Bioloaical Chemistrv 259: 7391-7398. Press, New York.
BANERJEE A., Jo& M. A. & <~DUENA R, F. 1985.Inter- BORGERS M. &DE NOLLIN S. 1975. Ultrastructuralchanges
action of reduced glutathione with bovine brain tubulin. in Ascar~.~su~~mintestine after mebendazole treatment in
Biochemical and Riophysical Research Communications yivo. Journal ofParasitology 60: 1 IO- 122.
128: 506-512. BOROERS M., DE NOLLIN S., VERHEYEN A., DE BRABANDER
BATRA J. K., JURD L. & HAMEL E. 1985. Structure-function M. & THIENPONT D. 1975. Effects of new anthelmintics
studies with derivatives of 6-benzyl- 1,3benzodiozole. a on the microtubular system of parasites. In: Microtubules
new class of synthetic compounds which inhibit tubulin and Microtubule Inhibitors (Edited by BORGERS M. & DE
 Role of tubulin in benzimidazoles mode of action 929

BRABANDER M.), pp. 4Y7-SO8. North Holland/Elsevier. Microtuhute Biology. Annat.s of the New York Academy of
Amsterdam. Sciencees, Vol. 466 (Edited by SOIFER D.), pp. 745-756.
BORISY G. G. lY’72. A rapid method for quantitative deter- The New York Academy of Sciences, New York.
mination of microtubule protein using DEAE-cellulose CAMPBELLW. C. & RF.W R. S. (Eds) 1986. Chemofherapyo~
filters. Analy~icalBiochemist~ SO: 373-X-S. Parasitic 1Xseases. Plenum Press. New York.
B~SSCHE H. VAN DEV 1072. Biochemical effects of the CAPLOW M. I Y86. Location of the guanosine triphosphate
anthelmintic drug mebendazole. In: fompwufive Bio- (GTP) hydrolysis site in microtubules. In: Dynamic
chemistry ofPurusites (Edited by ROSSCHF. H. VAN IIEN). Aspects of Microruhule Biology. Annuls ofrhe New York
pp. 139-l 57. Academic Press, New York. Acuden?y of S?iences, Vol. 466 (Edited by SOIFER D.),
ROSKHE H. VAN DEN. ROCH~TTE F. & HORIG C. 1982. $tr.S! O-S I H. The New York Academy of Sciences, New
Mehendazole and related anthelmintics. Rd~~roxxs in
Pharmucofogyand Chemothempy 19: 67- 128, CHALFIE M. & THOMSON J. N. 1979. Organisation of
BOSSCHE H. VANDEN 1985. Pharmacology of anthelmintics. neuronal microtubules in the nematode Cutnorhabditi.s
In: Ch~motherap~I of~u,s~roint~.stinufHeiminrhs (Edited efegans. Journal ofCell Rioiogy 82: 278-280.
byBosscnE H.VAN DEN,T~~~ENPoHTD. &J~usst~s P. G.). CLARK J. 1. & GAIU.AXD D. 1978. Fluorescein colchicine.
D,D.125-I 8 1.Springer, Berlin. Synthesis, purification and biological activity. Journat of
&ix~~ H.VAN D&‘?HIENPONT I). &JANSSENS P. G. (Eds) CPtlBiofogy 76: 619-627.
1985. Chemorherupv of C&troinrestinai Heimirzrhs. CLAYTON L., QUINLAN R. A., ROOBOL A., Pocso~ C. I. &
Springer, Berlin. ’. ” GULL K. 1980. A comparison of tubulins from mammal-
BO&ON B. J., WAUD W. R., WHEELER G. P., HAIN R., ian brain and fh)jsurum polvcephatum using SDS-
DANSBY L. & TEMPLE C. L. 19X7. Comnarison of polyacrylamide gel electroph&&and peptine mapping.
I .2_dihydropyrido [3,4-hl pyrazines ’ (I -deaza- F.!XS Letters 115: 30 l-305.
7,X-dihydropteridines) with several other inhibitors of CLEMONS G. P. bi SISLER H. D. IY71. Localization of the
mitosis. Chnc~?erRe.search 47: 162 I- 1626. site of action of a fungitoxic benzomyl derivative. Pesricide
BRINKLEY 3. R., BARHAM S. S., BARRA~CV S. C. & FUI.LBR ~jo~hen1i.s~~and Pfz~~iofogy 1: 32-43.
G. M. 1974. Rotenone inhibition of spindle microtubule CI.EVELAUD D. W. & SLYLLIVXN K. F. 1YX.5.~[~lecular
assembly in mammalian cells. ~xpe~rn~n~at Celf Research biology and genetics of tuhulin. ~4nnuaf Review of Bio-
85:41-46. chemisr? 54: 33 l--365.
HK~SSI A., YEH H. J. C., CHRZANOWSKA M., WOLFF J., COLES G. C. & MCNEIUIE R. M. lY77. The response of
HAMEI. E.. LIN C. M.. QLII~ F.. SUPFNESS M. & SILVER- nematodes in v~vo and in vitro to some anthelmintics.
TON J. 19X8. Colchicine and its analogues: recent findings. Journul of tfrtmintholog) 51: 323-326.
Medicinal Research Reviews 8: 77-94. COLES G. C. & SIMPKIN K. G. 1077. Resistance of nema-
BROWN H. D., MATZUK A. R., ILVES I. R., PETERSON 1,. H., tode eggs to the ovicidal activity of henzimidazoles.
HARRIS S. A.. SARETT L. Ii., ECERTON J. R., YAKSTIS J. J.. Research in Veterinary Science 22: 386-387.
CAMPBELL W. C. & CUCKLER A. C. 1961. Antiparasitic COI.LINS C. A. & VAUEE R.B. 19X7. Temperature-
drugs--IV. Z-(4 ‘-Thiazolvl~-b~nzimidazole. a new anthel- dependent reversible assembly of taxol-treat~d.microtu-
min&. Joitrnbl of the &erican Chernjeal Society 83: bules. Journal of <‘elfBiofogv 105: 2847-2854.
1764-I 765. CONWAY D. P. iY64. Vari&ce in the effectiveness of
BROWN H. R. Rr ERICKSON H. P. 1983. Assembly of proteo- thiabendazole against Huemonchus contornts in sheep.
lytically cleaved tubulin. An,hire.s of Riochemi.sfq~ und Americun Journat of C’ererinarvResearch 25: 105-l 06.
Riophysics 220: 46-5 I. COOKJ. W. &LouooGJ. D. I95 i.Thetropolones. Quaner(v
BRYAN J. I Y7 I. Vinblastine and microtubules-1. Induction Reviews (ChemicalSocierv) 5: 99-l 29.
and isolation of crystals from sea urchin oocytes. Experi- CORTESE F.~. BHAT.TACHARYY’A’B. & WOLFF J. lY77. Podo-
mental Ceil Research 66: 129- 136. phyllotoxin as a probe for the coichicine binding site of
BRYANT C. 81 BENNETT E.-M. 1Y83. Observations on the tubulin. Jofrt-nrri of Rioioxicaf C‘hemism? 252: 1 134-
fumarate reduetase system in ~aen~on~t?us conrorfus and 1140.
their relevance to anthelmintic resistance and to strain Goss R. A.. BI.~OD~OOD R. A.. RROWER D. I.., PICKETT-
variations of energy metabolism. .~~ol~c~~a~ and Bio- HEAPS J. D. 6i MCINTOSH J. R. 1975. Studies on the
chemical Parasirotogv 7: 281-2Y2. mechanism of action of isopropyl N-phenyl carnamate.
BLIRLANO T. C.. SCHEDI. T., GL:LL K. & DOVE W. F. 1984. Experimental C’etlResearch 92: 3Y4-3YX.
Genetic analysis of resistance to benzimidazoles in CORREIA J. J., RATY I.. T. & WILLIAMS R. C. 10X7. Mg’+
Physarum: differential expression of /!I-tubulin genes. dependence of guanine nucleotide binding to tubulin.
Gmetics 108: 123-141. Joumul of Biological Chemis@ 262: 17,278-l 7,284.
BURLAND T. G., PAUL E. C.A., OETLIKER M. RL DOVE CRIE J. S.. ORD J. M., WAKELA~D J. R. & WIUIENTHAL K.
W. F. 198X. A gene encoding the major beta tubulin of the IYX3. Inhibition of cardiac proteolysis by colchicine.
mitotic spindle in Ph~~arum poiycephatum ptu.snzodia. 3ios~~en~i~ai.torcrnaf 2 10: 63-7 1.
.~olecl~far and Celil~lar Biology 8: 12 75 1281. CROOM H. B., CORREIA J. J. & WILLIAHS R. C. lY86. The
BUITLAIRE D. H.. CZUBA B.A., STEVENS T.H.. LEE effects of elevated pH and high salt concentrations on
Y. C. & HIMES R. H. 1980. Manganous ion binding to tubulin. Archives of Biochemivql and Riophysirs 249:
tubulin. Journal of Biological C’hemistuy 255: 2164- 397-406.
2168. DAVIDSE L. C. 1973. Antimitotic activity of methyl
CABRAL F. R., SOBEL M. E. & GOTTESMAN M. M. lY80. benzimidazol-2-yl carbamate (MBC) in Aspergittus
CHO mutants resistant to colchicine, colcemid of griseo- niduluns. Pesticide Niochemisq and Physiology 3: 3 17-
fulvin have an altered /?-tuhulin. Cell 20: 29-36. 325.
CABRAL F. R., BRADY R. C. & S~HIBLER M. J. 1986. A DAVIDSE L. C. & FLACH W. 1977. Differential binding of
mechanism of cellular resistance to drugs that interfere methyl henzimidazole-2-yl carbamate to fungal tubulin as
with microtubu~e assemblv. In: f~vnarn~~ Asoecrs ot_( a mechanism of resistance to this antimitotic agent in
930 E. LACEY

mutant strains of Aspergillus nidulans. Journal of Cell FELLOUSA., LUDUENA R. F., PRASAD V., JORDANM. A.,
Biology 72: 174-193. ANDERSONW., OHAYONR. &SMITH P. T. 1985. Effect of
DAVIS C. & GULL K. 1983. Protofilament number in tau and MAP, on the interaction of maytansine with
microtubules in cells of two parasitic nematodes. Journal tubulin: inhibitory effect of maytansine on vinblastine-
ofParasitology 69: 1094-1099. induced aggregation of tubulin. Cancer Research 45:
DAWSON P. J., GUTTERIDGEW. E. &GULL K. 1983. Purifi- 5004-5010.
cation and characterisation of tubulin from the parasitic FETTERER R. H. 1986. The effect of albendazole and tricla-
nematode, Ascaridia galli. Molecular and Biochemical bendazole on colchicine binding in the liver fluke Fusciolu
Parasitology 7: 267-277. hepatica. Journal of Veterinary Pharmacology and Thera-
DAWSON P. .I., GUTTERIDGE W. E. & GULL K. 1984. A peutics 9: 49-54.
comparison of the interaction of anthelmintic benzimida- FITZGERALD T. J. 1976. Molecular features of colchicine
zoles with tubulin isolated from mammalian tissue and the associated with antimitotic activity and inhibition of
parasitic nematode Ascaridia galli. Biochemical Pharma- tubulin polymerization. Biochemical Pharmacology 25:
cology 3: 1069-1074. 1383-1387.
DE BRABANDERM. & DE MAY J. 1980. Microtubules and FOLZ S. D., PAX R. A., THOMAS E. M., BENNETT J. L., LEE
Microtubule Inhibitors 1980. Elsevier/North Holland, B. L. & CONDER G. A. 1987. Development and valida-
Amsterdam. tion of an in vitro Trichostrongylus colubriformis motility
DEINUM J. & LINCOLN P. 1986. Binding of tubulin of an assay. International Journal for Parasitology 1 I: 144 l-
allocolchicine spin probe: interaction with the essential 1444.
SH groups and other active sites. Biochimicu et Bio- FORSYTH K. P., MITCHELL G. F. & COPEMAN D. B. 1984.
physica Acta 870: 226-233. Onchocercu gibsoni: increase of circulating egg antigen
D~T&CH H. W., WILLIAMS R. C., MACDONALD T. L., with chemotherapy in bovines. Experimental Parasitology
WILSON L. & PUETT D. 1981. Changes in the circular 58: 41-55.
dichroic spectrum of colchicine associayed with its binding FRIEDMAN P. A. & PLATZER E. G. 1978. Interaction of
to tubulin. Biochemistry 20: 5999-6005. anthelmintic benzimidazoles and benzimidazole deriva-
Do KHAC L., TANFIN Z. & HARBON S. 1983. Differential tives with bovine brain tubulin. Biochimica et Biophysicu _
role of microtubules in the control of prostaglandin E, and Acta 544: 605-614.
B-adrenerpic stimulation of cvclic AMP accumulation in FRIEDMAN P. A. & PLATZER E. G. 1980. Interaction of
‘the rat miometrium. Biochkmical Pharmacology 32: anthelmintic benzimidazoles with Ascaris suum embry-
2535-2541. onic tubulin. Biochimica et Biophysics Acta 630: 271-
DRUBIN D., KOBAYASHI S. & KIRSCHNER M. 1986. Associa- 278.
tion of tau protein with microtubules in living cells. In: FRIEDMAN P. A., PLATZER E. G. & CARROLL E. J. 1980.
Dynamic Aspects of Microtubule Biology. Annals of the Tubulin characterization during embryogenesis of Ascaris
New York Academy of Sciences, Vol. 466 (Edited by suum. Developmental Biology 76: 47-57.
SOIFER D.). pp. 257-268. New York Academy of GARLAND D. L. 1978. Kinetics and mechanism of colchicine
Sciences, New York. binding to tubulin: evidence for ligand-induced conforma-
DRUDGE J. H., SZANTO J., WYANT Z. N. & ELAM G. 1964. tional change. Biochemistry 17: 4266-4272.
Field studies on parasite control in sheep: comparison of GASKIN F. & KRESS Y. 1977. Zinc ion-induced assembly of
thiabendazole, ruelene, and phenothiazine. American tubulin. Journal of Biological Chemistry 252: 6918-
Journal of Veterinary Research 25: 1512-l 518. 6924.
DURRIEU C., BERNIER-VALENTIN F. & ROUSSET B. 1987. GINCKEL R., VAN, DE BRABANDER M., VANHERCK W. &
Binding of glyceraldehyde 3-phosphate dehydrogenase to HEERES J. 1984. The effects of tubulazole, a new synthetic
microtubules. Molecular and Cellular Biochemistry 14: microtubule inhibitor on experimental neoplasms. Euro-
55-65. pean Journal of Cancer and Clinical Oncology 20: 99-
DUSTIN P. 1978. Microtubules. Springer, Berlin. 104.
DUSTIN P. 1984. Microtubules. Springer, Berlin. GULL K., DAWSON P. J., DAVIS C. & BYARD E. H. 1987.
EGERTON J. R. 1969. The ovicidal and larvicidal effect of Microtubules as target organelles for benzimidazoles
thiabendazole on various helminth species. Texas Reports anthelmintic chemotherapy. Biochemical Society Trans-
ofBiologyand Medicine 21: Supplement 2, 561-580. actions 15: 59-60.
ENCELBORCHS Y. & FITZGERALDT. J. 1987. A fluorescence GUNDERSEN G. G., KALNOSKI M.H. & BULINSKI J. C.
stopped flow study of the competition and displacement 1984. Distinct populations of microtubules: tyrosinated
kinetics of podophyllotoxin aid the colchiccne analog and non-tyrosinated alpha tubulin are distributed differ-
2-methoxy-5-(2’,3’,4’-trimethoxyphenyl) tropone on ently in vivo. Cell 38: 779-789.
tubulin. Journal of Biological Chemistry 262: 5204- GUPTA R. S. & GUPTA R. 1984. Mutants of Chinese hamster
5209. ovary cells affected in two different microtubule-
EPE B., HECLER J. & METZLER M. 1987. Site-specific associated proteins. Journal ofBiological Chemistry 259:
covalent binding of stilbene-type and steroidal estrogens 1882-1890.
to tubulin following metabolic activation in vitro. Car- GUPTA R. S. 1986. Cross-resistance of nocodazole-resistant
cinogenesis 8: 1271-1275. mutants of CHO cells toward other microtubule inhibi-
FARRELL K. W. 1982. Isolation of tubulin from non- tors: similar mode of action of benzimidazole carbamate
neuronal sources. In: Methods of Enzymology, Vol. 85 derivatives and NSC 181928 and TN-l 6. Molecular
(Edited by FREDERIKSON D. W. & CUNNINGHAM L. W.), Pharmacolonv 30: 142-148.
pp. 385-393. Academic Press, New York. HAINS F. O., DICKERSONR. M., WILSON L. & OWELLEN
FAULKNER A., HENDERSON A. J. & PEAKER M. 1984. The R. J. 1978. Differences in the binding urouerties of vinca
I. L

effects of colchicine and vincristine on the concentrations alkaloids and colchicine to tubulin by varying protein
of glucose and related metabolites in goat’s milk. Bio- sources and methodology. Biochemical Pharmacology
chimica et Biophysics Acta 802: 335-339. 27: 71-76.
 Role of tubulin in benzimidazoles mode of action 931

HAMEL E. & LIN C. M. 198 1. Stabilization of the colchicine- 1982. Taxol: a new probe for studying the structure and
binding activity of tub&n by organic acids. Biochimica et function of microtubules. Cold Spring Harbour Sympos-
BiophysicaActa 675:226-231. ium on Quantitative Biology 46: 219-226.
HAMEL E. & LIN C. M. 1981b. Glutamate-induced poly- HOWELLS R. E. &DELVES C. J. 1985. A simple method for
merization of tubulin: characteristics of the reaction and the identification of compounds which inhibit tubulin
application to the large-scale purification of tubulin. polymerization in filarial worms. Annals of Tropical
Archives ofBiochemistry and Biophysics 209: 29-40. Medicine and Parasitology 79: 507-512.
HAMMERSCHLAGR. S. & SISLER H. D. 1973. Benomyl and HUANG A. B., LIN C. M. & HAMEL E. 1985. Maytansine
methyl-2-benzimidazolecarbamate (MBC): biochemical inhibits nucleotide binding at the exchangeable site on
cytological and chemical aspects of toxicity in Ustilugo tubulin. Biochemical and Biophysical Research Com-
maydis and Saccharomyces cerevesiae. Pesticide Bio- munications 128: 1239-1246.
chemistry and Physiology 3: 42-54. IBARRA 0. F. & JENKINS D. C. 1984. The relevance of in
HANSCH C. & LEO A. 1979. Substituent Constants for vitro anthelmintic screening tests employing the free-
Correlation Analysis in Chemistry and Biology. Wiley living stages of trichostrongylid nematodes. Journal of
Interscience, New York. Helminthology 58: 107-112.
HARTLEY-ASP B., DEINUM J. & WALLIN M. 1985. Diethyl- IKEDA Y. & STEINER M. 1976. Isolation of platelet microtu-
stilbestrol induces metaphase arrest and inhibits microtu- bule protein by an immunosorptive method. Journal of
bule assembly. Mutation Research 143: 231-235. Biological Chemistry 25 1: 6135-6141.
HARTWELL J. L. & SCHRECKERA. W. 1958. The chemistry IKEDA Y. & STEINER M. 1978. Sulfhydryls of platelet
of podophyllum. Fortschritte der Chemie organischer tubulin: their role in polymerisation and colchicine
Nuturstoffe 15: 83-166. binding. Biochemistry 17: 3454-3459.
HEAD J., LEE L. L.-Y., FIELD D. J. & LEE J. C. 1985. IRELAND C. M., GULL K., GUTTERIDGE W. E. & POGSON
Equilibrium and rapid kinetic studies on nocadazole- C. I. 1979. The interaction of benzimidazole carbamates
tubulin interactions. Journal of Biological Chemistry 266: with mammalian microtubule protein. Biochemical
11,060-11,066. Pharmacology 28: 2680-2682.
HEATH D. D., CHRISTIE M. J. &CHEWS R. A. F. 1975. The IRELAND C. M., CLAYTON L., GUTTERIDGE W. E., POGSON
lethal effect of mebendazole on secondary Echinococcus C. I. & GULL K. 1982. Identification and drug binding
granulosus, cysticerci of Taenia pisiformis and tetra- capabilities of tubulin in the nematode Ascaridia galli.
thyridia of Mesocestoides corti. Parasitology 70: 273- Molecular and Biochemical Parasitology 6: 45-53.
285. IWATA M., MATSUI T., HINO A. & NAKANO E. 1982.
HENNESSY D. R. 1985. Manipulation of panthelmintic Polyamine-induced disassembly of reconstituted microtu-
pharmacokinetics. In: Resistance in Nematodes to Anthel- bules in vitro. Cell Structure and Function 7: 155-164.
mintic Drugs (Edited by ANDERSON N. & WALLER P. J.), JENKINS D. C. 1982. In vitro screening tests for anthel-
pp. 79-81. CSIRO Division of Animal Health, Glebe, mintics. In: Animal Models in Parasitology (Edited by
NSW. OWEN D. G.), pp. 173-186. MacMillan Press, London.
HERLICH H., REW R. S. & COLGLAZIER M. L. 1981. Inherit- KARKHOFF-SCHWEIZER R. & KNULL H. R. 1987. Demon-
ance of cambendazole resistance in Haemonchus con- stration of tubulin-glycolytic enzyme interactions using a
tortus. American Journal of Veterinary Research 42: novel electrophoretic approach. Biochemical and Bio-
1342-1344. physical Research Communications 146: 827-831.
HESSE J., MARUTA H. & ISENBERG G. 1985. Monoclonal KAWALEKJ. C.,REWR. S. &HEAVNERJ. 1984,Glutathione-
antibodies localize the exchangeable GTP-binding site in S-transferase, a possible drug-metabolising enzyme, in
B and not a-tubulin. FEBSLetters 179: 91-95. Haemonchus contortus: comparative activity of a
HIMES R. H., BURTON P. R. & GAITO J. M. 1977. Dimethyl cambendazole-resistant and a susceptible strain. Znter-
sulphoxide-induced self-assembly of tubulin lacking asso- nationalJournalforParasitology 14: 173-175.
ciated proteins. Journal of Biological Chemistry 252: KEATES R. A. B., SARANCI F. & LING V. 1981. Structural
6222-6228. and functional alterations in microtubule protein from
HIMES R. H., LEE Y. C., EAGLE G. R., HASKINS K. M., Chinese hamster ovary cell mutants. Proceedings of the
BABLER S. D. & ELLERMEIER J. 1982. The relationship National Academy of Sciences of the United States of
between cobalt binding to tubulin and the stimulation of America 78: 5638-5642.
assembly. Journal of Biological Chemistry 257: 5839- KELLEHER J. K. 1977,Tubulinbindingaffinitiesofpodophyl-
5845. lotoxin and colchicine analogues. Molecular Pharmu-
HINMAN N. D., MORGAN J. L., SEEDS N. W. & CANN J. R. cology 13: 232-241.
1973. Isolation of brain tubulin by affinity chromato- KIRSCH R. & SCHLEICH H. 1982. Morphological changes in
graphy. Biochemical and Biophysical Research Com- trichostrongylid eggs after treatment with fenbendazole.
munications 52: 752-758. Veterinary Parasitology 11: 375-380.
HOEBEKE J. & NUEN G. VAN 1975. Quantitative turbidi- KOCHA T., FUKUDA T., ISOBE T. & OKLJYAMA T. 1986.
metric assay for potency evaluation of colchicine-like Hydrophobic chromatography of tubulin on immobilised
drugs. Life Sciences 17: 591-596. colchicine columns with various species. Chemical and
HOEBEKE J., NUEN G. VAN&DE BRABANDERM. 1976.Inter- Pharmaceutical Bulletin 34: 3791-3802.
action of oncodazole (R17934), a new antitumoral drug, KOEHN J. A. &OLSEN R. W. 1980. Endogenous particulate
with rat brain tubulin. Biochemical and Biophysical inhibitor of microtubule assembly in developing mammal-
Research Communications 69: 319-324. ian brain. Archives of Biochemistry and Biophysics 201:
HOROWITZ P., PRASAD V. & LUDUENA R. F. 1984. Bis (1,8- 207-215.
anilinonaphthalenesulphonate), a novel and potent in- KSHLER P. & BACHMANN R. 1981. Intestinal tubulin as
hibitor of microtubule assembly. Journal of Biological possible target for chemotherapeutic action of mebenda-
Chemistry 259: 14,647-14,650. zole in parasitic nematodes. Molecular and Biochemical
HORWITZ S. B., PARNESS J., SCHIFF P. B. & MANFREDI J. J. Parasitology 4: 325-336.
932 E. LACEY

KOTANI S., MUROFUSHI H., NISHIDA E. & S.~KAI H. 19X4. of the self-assembly of tubulin by diethylpyrocarbo~te
33 K protein-an inhibitory factor of tubulin polymerisa- and photooxidation. Biochemical and Biophysical
tion in porcine brain. Journal of Biochemistry (Tokyo) 94: Research Communications IO: 50-57.
95%969. LE JAMBRE I,. F. 1976. Egg hatch as an in vitro assay of
KUPCHAN S. M., SNEDEN A. T., BRANFMANA. R., HOWIE thiabendazole resistance in nematodes. Veterinar?jParasi-
G. A., REBHUN L. I., MCIVOR W. W., WANG R. W. & tology 2: 385-391.
SCHNAITMAN T. C. 1978. Structural reauirements for LE JAMBRE L. F., ROYAL W. M. & MARTIV P. J. 1970. The
antileukemic activity among the naturally’occurring and inheritance of thiabendazole resistance in Huemonchus
semisynthetic maytansinoids. Journaf of i~ed~cjnaf contortus. Parasitology 78: 107- 119.
Chemistry 21:?1-36. LEWIS S. A., Gu W. & COWAN N. J. lY87. Free inter-
KURIYAMAR. & SAKAI H. 1974. Role of tubulin -SH groups mingling of m~maIian 8-tubulin isotvpes among
in polymerization to microtubules. Journal ofBiochemis- fun&onalIy distinct microtubules. Cell 49: 539-548. -
try 76: 651-654. LIN C. M. & HAMEL E. 198 1. Effects of inhibitors of tubulin
LACEY E. 198.5. The biochemistry of anthelmintic resist- polymerization on GTP hydrolysis. Journul of Biological
ance. In: Resistance in Nematodes to Anthelrnintic Drugs Chemistry 256: 9242-9245.
(Edited by ANDERSON N. & WALLER P. .I.), pp. 69-78. LING V. & THOMPSON L. H. 1974. Reduced permeability in
CSIRO Division of Animal Health, Glebe, NSW. CHO cells as cells as a mechanism of resistance. to
LACEY E. & WATSON T.R. 1985a. Structure-activity colchicine. Journal of Cell Physiology 83: 103- 1 16.
relationships of benzimidazole carbamates as inhibitors of 1.1~~ V.. AUDIN J. E., CHASE A. & SARANCI F. lY?Y.
mammalian tubutin in vitro. Biochemical Pharmacoio~ Mutants of Chinese hamster ovary (CHO) cells with
34: 1073-1077. altered colcemid-binding affinity. Cell 18: 423-430.
LACEY E. & WATSON T. R. 1985b. Activity of benzimida- LITTAUER U. Z., GIVEON D.. THIERALX M., GIXZBURG I. &
zole carbamates against L1210 mouse leukaemia celts: PONSTIXGL H. 1986. Common and distinct tubulin bind-
correlation with in vitro tubulin polymerization assay. ing sites for microtubule-associated proteins. Proceedings
Biochemical Pharmacology 34: 3603-3605. ofthr National Acudenty ofSciences of the United States of
LACEY E. & PRICMARDR. K 1986. Interactions of benzi- Amoricu 83: 7162-7166.
midazoles (BZ)
\ I
with tubulin from BZ-sensitive and BZ- LITI.I.E M. & LUDUENA R. F. 1985. Structural differences
resistant isolates of Haemonchus contortus. Molecular between brain PI- and P2-tubulins: implications for
and Biochemical Parasitology 19: 17 I -18 1. microtubule assembly and colchicine binding. EMU0
LACEY E., BRADY R. L., PRICHARDR. K. Sr WATSON T. R. Journal 4: S I-56.
1987. Comparison of inhibition of poIyme~sation of LOCKWOOD A. H. 1978. The biological regulation of micro-
mammalian tubulin and helminth ovicidal activity by tubule assembly and function. In: Cefi Reprodz~~rfon
benzimidazole carbamates. Veterinary Parasitology 23: (Edited byDIRKSE4 E. R., PRESCOTTD. M. &Fox C. F.).
105-I 19. pp. 27 I-283. Academic Press, New York.
LACEY E., EDGAR J. A. & CULVENOR C. C. J. 1987. Inter- LOCKWOOD A. H. 1979. Molecules in mammalian brain that
action of phomopsin A and related compounds with interact with the colchicine site on tubulin. Pr0ceeding.s of
purified sheep brain tubulin. Biochemical Pharmacology .__ the National Academy of Sciences of the United States of
‘36: 2133-23’38. America 76: I 184-I 188.
LACEYE., SNOWDONK. L., EAGLESONG. K. & SMITHE. F. L~IK~ J. D. SL HOROWITZS. B. IY76. Effects of podophyllo-
I 987. Further investigations on the primary mechanism of toxin and VP-16-213 on microtubule assembly in vitro
benzimidazole resistance in Haernon~h~~.~contortus. and nucleoside transport in HeLa cells. Bjo~hern~st~ 15:
Internationui Joi~rnaf~or Parasitofo~~ 17: 142 1-I 42Y. 5435-5443.
LACEY E. & SNOWDON K. L. 1988. A routine diagnostic LOII,ZI R. F. 1985. Progesterone withdrawal stimulates
assay for the detection of benzimidazole in parasitic mammary gland tubulin polymerization in pregnant rats.
nematodes using tritiated benzimidazole carbamates. Endocrinology 116: 2543-2547.
Veterinary Parasitology 21: 309-324. LOPATA M. A. & CLEVELAVD D. W. 1Y87. In viva microtu-
LACEYE. (in press). Comparative biochemistry of parasites bules are copolymers of available P-tubulin isotypes:
and its role in drug resistance-an investigation of species localization of each of six vertebrate B-tubulin isotvnes
differences in tubulin. In: Parasitic Helminths, a Bio- using polyclonal antibodies elicited by synthetic peptide
chemical Perspective (Edited by BRYANT C.. BEHW antigens. Journal ofCell Biofogy 105: 17lf7-1720.
C. A. &BENNETT E. M.). Chapman and Hall. LUD&NA R.F.. A&ERS~N WY’H., PRASAD V.. JORDAN
LACLETTEJ. P., GUERRAG. & ZETINA C. 1980. Inhibition of M. A., FERRIGNI K. C., ROACH M. C.. HOROWITZ P. M.,
tubulin polymerization by mebendazole. Biochemical and MURPHY D. B. & FELL~US A. 1986. Interactions of
Biooh\*sicai Research Communications 92: 417-423. vinblastine and maytansine with tuhulin. In: Dynamir
LEE J: C. & TIMASHEFF S. N. 1977. In vitro reconstitution Aspect.s of Microtubufe Biology. Annals of the New York
of calf brain mirotubules: effects of solution variables. Aiademy” of Sciences, Vol. 4??~ (Edited by SO~FER D.).
Biochemistry 16: 1754-I 764. pp. 718-732. The New York Academy of Sciences, New
LEE J. C., TWEEDY N. & TIMASHEFF S. N. 1978. In vitro York.
reconstitution of calf brain microtubules: effects of LU~UENA R. I?., ROACH M. C., TRCKA P. P., MALLEVAIS
macromolecules. Biochemistry 17: 2783-2790. M. L. & MACRAE T. 1987. The effect of 2-(4-methyl-l-
LEE J. C.. FIELD D. J.. GEORGE H. J. & HEAD J. 1986. pip~razinylmethyl) acryiophenone dihydrochloride on
Biochemical and chemicai properties of tub&n sub- the atkvlation of tubulin. Archives of B~u~henzist~ and
species. In: Dynamic Aspects of l~li~rotlfbufe Biolo~. ~jophy~i~s 255: 453-459.
Annals of the New York Academy of Sciences, Vol. 466, MACCIOFII R. B. & SEEDS N. W. 1982. Involvement of
pp. 11 I-1 28. The New York Academy of Sciences. New tryptophan residues in colchicine binding and the
York. assembly of tubulin. Biochemical and Biophysical
LEE Y. C., HOUSTONI.. L. & HIMES R. I-I. 1976. Inhibition Research Communications 108: 896-903.
 Role of tubulin in benzimid~oles mode of action 933

MACCIONI R. B. & SEEDS N. W. 1982. Involvement of MIZEL S. B. & WILSON L. 1972. Nucleoside transport in
tryptophan residues in colchicine binding and the assembly mammalian cells. Inhibition by colchicine. Biochemistry
of tub&n. Biochemical and Biophysical Research Com- 11: 2573-2578.
munications 108: 896-903. MOREJOHN L.C. & FOSKET D. E. 1982. Higher plant
MAFENDAWZ M. & SOLOMON F. 1985. Analyzing the tubulin identified bv self-assembly into microtubules in
components of microtubules: antibodies against chartins, vitro. Nature (London) 297: 426-428.
associated proteins from cultured cells. Proceedings ofthe MOREJOHN L.C., BUREAU T. E.. TOCCHI L. P. & FOSKF.T
National Academy of Sciences of the United States of D. E. 1984. Tubulins from different higher plant species
America 82: 6581-6585. are immunologically non-identical and bind colchicine
MALKIN M. F. & CAMACHO R. M. 1972. The effect of differentially. Proceedings of the National Academy of
thiabendazole on fumarate reductase from thiabendazole- Sciences ofthe United States ofAmerica 81: 1440-l 444.
sensitive and resistant Haemonchus contortus. Journal of MORENO M. S. & BARRETTJ. 1979. Monoamine oxidase in
Purasitology 58: 845-846. adult Hymenoiepis diminuta (Cestoda). Parasitology 78:
MANDELBAU~-SHA~IT F., WOLPERT-Do FILIPPES M. K. & 1-5.
JOHNS D. G. 1976. Binding of maytansine to rat tubulin. MORGAN J. L. & SPOONER B. S. 19133. Immunoio~cal
Biochemical und Biophysical Research Communications detection of microtubule poison-induced conformational
72: 47-54. changes in tubulin. Journal of Biological Chemistry 258:
MAREEL M. M. & DE METS M. 1984. Effect of microtubule 13,127-13,133.
inhibitors on invasion and on related activities of tumor NA C. & TIMASHEFF S. N. 1977. Physical-chemical study of
cells. International Review of Cytology 90: 125-167. daunomycin-tubulin interactions. Archives of Bio-
MARGOLIS R. L. & RAUCH C. T. 198 1. Characterization of Chemist~a~B~oph~ics 182: 147-154.
rat brain crude extract microtubule assembly: correlation NATH J. P., EAGLE G. R. & HIMES R. H. 1986. Studies on
of cold stability with the phosphorylation state of a the exchangeable nucleotide binding site of tubulin. In:
microtubule-associated 64K protein. Biochemistry 20: Dynamic A>pects of Microtubule Biology. Annals of the
4451-4456. New York Academv of Sciences, Vol. 466 (Edited bv
MARGOLIS R. L.., JOB D., PABION M. & RAUCH C. T. 19X6. SOWER D.), pp.4i2-295. New York Academy df
Sliding of stop proteins on microtubules: a model system Sciences, New York.
for diffusion-dependent microtubule motility. In: OAKLEY B. R. & MORRIS N. R. 198 1.A /?-tubulin mutation
Dynamic Aspects-of Microtab~~e Biology. Annals of the in Aspergi&s nid~~ans that blocks microtubule function
New York Academy of Sciences, Vol. 466 edited by without blocking assembly. Cell 24: X37-845.
SOIFER D.), pp. 308-3i1. The New York A‘cademy df OAKLEY B. R. & MORRIS N. R. 1981. A f3-tub&n mutation
Sciences, New York. in Aspergillus nidulans that blocks microtubule function
MARGULIS T. N. 1974. Structure of the mitotic spindle without blockinn assemblv. Cell 24: 837-845.
inhibitor colcemid. N-Desacetyl-N-methylcolchicine. OBENDORF D. I.., FARSONS j. & NICHOLLS J. 1986. An egg
Journal ofthe American ChemicalSociety 96: 899-902. development test for the evaluation of benzimidazole
MARRINER S. & ARMOUR J. 1986. Nematode infections of resistance in Nematodirus sputhiger. Australian Veteri-
domestic animals: gastrointestinal infections. In: Chemo- na~Jo~rna~ 63: 382-383.
therapy of Parasitic Diseases {Edited by CAMPBELL O’BRIEN E. T., WHITE S., JACOBS R. S., BODER G. B. &
W. C. & REW R. S.), pp. 287-305. Plenum Press, New WILSON L. 1984. Pharmacological properties of a marine
York. natural product, stypoldione, obtained from the brown
MARTIN P. J., ANDERSON N., JARRETT R. G., BROWN algae Stypodium zonale. Hydrobiologia 116- 117: 14 l-
T. H. & FORD G. E. 1982. Effects of a preventive and 145.
suppressive control scheme on the development of OLMSTED J. B., ASNES C. F., PARYSEKL. M., LYON H. D. &
thiabendazole-resistance in Ostertagia spp. Australian KIDDER G. M. 1986. Distribution of MAP-4 in cells and
Veterinary Journal 58: 1X5-190. in adult and developing mose tissues. In: DynamicAspects
MARTIN R. J., ANUERSOW N., JARRETT R. G., BROWN T. H. of .~i~rotubu~e Bioiogy. Annals ofthe New York Academy
& FORD G. E. 1982. Effects of a preventive and suppres- ofSciences, Vol. 466 (Edited by SOIFER D.), pp. 292-305.
sive control scheme on the development of thiabendazole- New York Academy of Sciences. New York.
resistance in Ostertagia spp. Australian Veterinary PANTALONI D. & CARLIER M.-F. 1986. Involvement of
Journal 58: 185-190. guanosine triphosphate (GTP) hydrolysis in the mechan-
MATUS A. & RIEDERER B. 1986. Microtubule-associated ism of tubulin polymerization: regulation of microtubule
proteins in the developing brain. In: Dynamic Aspects of dynamics at steady state by a (_TP cap. In: Dynumic
~li~rotabule ~joZo~. Annals of New York Academy of Asoecfs o~M~?rotubuie Biofo~. Annals of the New York
Sciences, Vol. 466-(Edited by &FER D.), pp. 167-i7$. Aiademi of Sciences. Vol. 4186 (Edited by SOWER D.),
New York Academy of Sciences, New York. p&6-509. The New York Academy of Sciences, New
McKAYD.B.,ARoNsT~MR.S.&SCHNF.IDERA.S. 1985.lnter-
actions of microtubule-active agents with nicotinic acety- PARNESS J. & HORWITZ S. B. 198 1. Taxol binds to polymer-
cholinereceptors. MolecularPharmacokogy28: 10-16. ized tubulin in vitro. Journal of CellBiology 91: 479-487.
MCKINNON G., DAVIDSONM., DE JERSEY J., SHANLEYD. & PENNINGROTHS. M. & KIRSCHNER M. W. 1977. Nucleotide
WARD L. 1987. Effects of acetaldehyde on polymeriza- binding and nhosphorylation in microtubuie assembly in
tion of microtubule proteins. Bruin Research 416: W-99. vitro. Journal of~~o~e~~~arBiology 115: 643-673. _
MEDRANO F. J. & ANDREL’ J. M. 19X6. Binding of gossypol PENNINGROTHS. M. & KIRSCHNER M. W. 1978. Nucfeotide
to purified tubulin and inhibition of its assembly into specificity in microtubule assembly in vitro. Biochemisrry
microtubules. European Journal of Biochemistry 158: 17: 734-740.
63-69. PETTIT G. R., SINGH S. B., NWEN M. L., HAMEL E. &
MELLADO W., SLEBE J. C. & MACCIONI R. B. 1982. Tuhulin SCHMIDT J. M. 1987. Isolation, structure, and synthesis of
carbamoylation. Biochemical Journal 203: 675-68 1. combretastatins A-l and B-l, potent new inhibitors of
934 E. LACEY

mi~rotubule assembly, derived from Combretum cafium. trongyh colubrifornzis; in vitro culture of parasitic stages
.Iournal
ofNatural Products 50: 119- 13 1. and their use for the evaluation of anthelmintics. Research
PEYROT V., BRIAND C., MOMBURG R. &SARI J. C. 1986. bl in Veterinary Science 39: 90-94.
vitro mechanism study of microtubule assembly inhibi- RAPSON E. B., CHILWAN A. S. & JENKINS D. C. 1986.
tion by cis-dichlorodiamine-platinum (II). Biochemical Acetylcholinesterase secretion-a parameter for the inter-
Pharmacology 35: 371-384. pretation of in vitro anthelmintic screens. Parasitology
PIROLLET F., JOB D., FISCHER E. H. & MARGOLIS R. L. 92: 425-430.
1983. Purification and characterisation of sheep brain RAY K., BHATTACHARYYAB. & BISWAS B. B. 1981. Role of
cold-stable microtubules. Proceedings of the National B-ring of colchicine in its binding to tubulin. Journal of
Academy of Sciences of the United States ofAme~ca SO: Biolog~cai Chern~~~ 256: 6241-6244.
1560-1564. Rew R. S., SMITH C. & COLCLAZIER M. L. 1982. Glucose
PLATZER E. G., EBY J. E. &FRIEDMAN P. A. 1977. Growth metabolism of Haemonchus contortus adults: effects of
inhibition of Caenorhabditis elegans with benzimidazoles. thiabendazole on susceptible versus resistant strain.
Journal of Nematology 9: 280. Journal OfParasitoioav 68: 845-850.
PONSTINGL H., KRAUHS E., LITTLE M., KBMPF T., HOFER- REW R. S. “& FETTERE; R. H. 1986. Mode of action of
WARBIN~K R. & ADE W. 1982. Amino-acid sequence of antinematodal drugs. In: Chemotherapy of Parasitic
a- and /3-tubulins from pig brain: heterogeneity and Diseases (Edited by CAMPBELL W. C. & REW R. S.).
regional similarity to muscle proteins. Cold Spring pp. 321-337. Plenum Press, New York.
Harbour Symposium on Quantitative Bioloa 46: 191- REW R. S., URBAN J. F. & DO~VRES F. W. 1986. Screen for
198. anthelmintics using larvae of Ascaris SIIUM. Americurr
POTCHOO Y., BRACUER I)., PEYROT V., CHAUVET-MONGES Jour?~alof Vete~na~~Research 47: 869-873.
A. M., SARI J. C. & CRETAN A. 1986. In vitro inhibition ROACH M. C. B: LUDUEN~ R. F. 1984. Different effects of
of microtubule assembly by disulfiram. International tubulin ligands on the intrachain cross-linking of @,-
Journal of Clinical Pharmacology, Therapy and Toxico- tubulin. Journal of Biological Chemistry 259: 12,063-
logy 24: 499-504. 12,071.
PRASA~ A. R. S., LUDUENA R. F. &HOROWITZ P. M. 1986. ROACH M. C., BANE S. & L.UDUENAR. F. 198.5. The effect
Detection of energy transfer between tryptophan residues of colchicine analogues on the reaction of tubulin with
in the tubulin molecule and bound bis (X-anilinonaphtha- iodo [‘“Cl acetamide and iv, N’-ethylenebis (iodoaceta-
lene-1 -sulphonate), an inhibitor of mi&otubule assembly, mide). Journal of Biological Chemistry 260: 3015-3023.
that binds to a flexible region of tub&in. Biochernist~ 25: ROACH M. C. & LUDUENA R. F. 1985. The effect of Tn-16
3536-3540. on the alkylation of tubutin. Biochemical and 3~oph_~ici~l
PRICHARD R. K. 1973. The fumarate reductase reaction of Research Communications 129: 200-205.
Haemonchus contortus and the mode of action of some ROBERTSK. & HYAMS J. S. 1979. Microtubules. Academic
anthelmintics. International Journal for Parasitology 3: Press New York.
409-417. ROMAN~WSK~ R. D., RHOADS M. L., COLCLAZIER M. I,. &
PRICHARD R. K., HENNESSY D. R. & STEEL J. W. 1978. KATES K. C. 1975. Effects of cambendazole, thiabenda-
Prolonged administration: a new concept for increasing zole and levamisole on fnmarate reductase in cambenda-
the spectrum and effectiveness of anthelmintics. Veferi- zole-resistant and -sensitive strains of Haemonchus
nary Parasitology 4: 309-315. contortus. Journal of Parasitology 6 1: 777-778.
PURICH D. L. & KRXSTOFFERSOND. 1984. Microtubule ROOBOL A., POGSON C. I. & GULL K. 1980. Identification
assembly: a review of progress, principles and perspec- and characterization of microtubule proteins from
tives. In: Advances in Protein internists, Vol. 36 (Edited myxamoebae of Physaru~z polycephalum 3iocheF~icul
by ANFINSEN C. B., EDSALL J. T. & RICHARDS F. M.), Journai 189: 305-3 12.
pp. 133-212. Academic Press, New York. RGSNERM., CAPRAROH.-G., JACOBSONA. E., ATWELL L...
OUINN F. R. & BEISLER J. A. 1981. Quantitative structure- RROSSI A., lORIO M. A., WILL.IAMS T. H., SIK A. H. &
activity relationships of colchicines against P388 leukemia CHICNELL C. F. 1981. Biological effects of modified
in mice. Journal of Medicinal Chemistry 24: 25 l-256. colchicines. Improved preparation of 2-demethylcolchi-
RAEYMAEKERSA. H. M., ALLEWIJN F. T. N., VANDBNBERK tine, 3-demethylcolchicine. and (+)-colchicine and
J., DEMOEN P. J. A., OFFENWERT T. T. T. VAN& JANSSEN reassignment of the position of the double bond in
P. A. J. 1966. Novel broad-spectrum anthelmintics. dehydro-7-deacetamidocolchicines. Journal of Medicinal
Tetramisole and related derivati& of &arylimidazo[2,1- Chemistry 24: 257-261.
blthiazole. Journal ofMedical ~hern~t~ 9: 545-551. ROSSI M., LINK J. & LEE J. C. 19X4. The structure of
RA&AF~ M. S. & BRYANT C. 1977. Studies of regulatory 2-methoxy-5-(2’,3’.4’-trimethoxyphenyl) tropone, an
metabolism in Monieza expansa: effects of cambendazole effective analog of colchicine. Archives of Biochemistv?,
and mebendazole International Journal for Parasitology and Biophysics 231: 470-476.
7: 403-409. RO~SSET B. & WOLFF J. 1980. Purification of brain tubulin
RAHMAN M. S., CORNISH R. A., CHEVIS R. A. F. & BRYANT by affinity chromatography on immobilized lactoperoxi-
C. 1977. Metabolic changes in some helminths from dase. Journal of Biological Chemistr?, 255: 11,677-
sheep treated with mebendazoie. New Zealand Veterinary 11.681.
Journal 25: 79-83. RUBINOS., FIORI P. L., LI;BINU G., MONACO G. YLCAPP~(:-
RAJAGOPALAN R. & G~RNANI S. 1985. Effect of rifampicin CINELI.I P. 1983. The cytoskeleton of hydatid cyst
on the biological activity of tubulin. ~ioche~zical Pharn~a- cultured cells and its sensitivity to inhibitors. European
rology 34: 3415-3419. Jo~{rnai of Cell Biology 30: 182-l 90.
RAJAGOPALANR. & GURNANI S. 1986. Inhibition of microtu- SACKETT D.L., BHATTACHARYYAB. & WOLFF J. 1986.
bule assembly by actinomycin D, an anti-tumour drug. Promotion of tubulin assembly by carboxyterminal charge
Chemical-Biological Interactions 60: 201-206. reduction. In: Dynamic Aspects of Microtubule Bio1og.y.
RAPSON E. B., JENKINS D. C. & TOPLEY P. 1985. Trichos- Annals of the New York Academy of Science, Vol. 466
 Role of tubulin in benzin tidazoles mode of action 935

(Edited by SO~FERD.), pp. 460-466. New York Academy antileukemic activity. Journal of Pharmaceutical Sciences
of Sciences, New York. 64: 646-648.
SACKETTD. L. &WOLFF J. 1986. Proteolysis of tubulin and SOIFER D. (Ed.) 1975. Annals of the New YorkAcademyof
the substructure of the tubulin dimer. jolournul of Bio- Science, Vol. 253. The New York Academy of Sciences,
logical C~e~~.~~~ 261: 9~7~-9~76. New York.
SAI.MON E. D. & WADSWORTH P. 1986. Fluorescence SOIFER D. (Ed.) 1986. Dynamics Aspects of ~icrot~bu~e
studies of tubulin and microtubule dynamics in living cells. Biology. Annals of the New York Academy of Sciences,
In: Applications of Fluorescence in the Biomedical Vol. 466. The New York Academy of Sciences, New
Sciences (Edited by TAYLOR D. L., WAGGONER A. S., York.
MURPHY R. F., LANNI F. & BIRCE R. R.).I DD. IL 377-403. SPENCE A.M., MALONE K. M. B., NOVAK M.M.A. &
Alan R. Liss Inc., New York. Wooos R. A. 1982. The effects of mebendazole on the
SANCSTER N. C. & PRICHARD R. K. 1985. The contribution growth and development of Caenorhabditis elegans.
of a partial tricarboxylic acid cycle to volatile end- Canadian Journal @Zoology 60: 26 16-2623.
products in thiabendazole-resistant and -susceptible Tti- STERNLICHTH., YAFFE M. B. & FARR G. W. 1987. A model
chostrongy~~ ~olzib~formis. ~o~er~~ar and Biochemists of the nucleotide-binding site in tubulin. FEBS Letters
Para.~ito~o~ 14: 261-274. 214: 226-235.
SANCSTERN. C., PRICHARDR. K. & LACEY E. 1985.Tubulin STYLES J. A. & GARNER R. 1974. Benzimidazoiecarbamate
and benzimidazole-resistance in Trichostrongylus cola- methyl ester-evaluation of its effects in vivo and in vitro.
hriformis (Nematoda). Journal of Parasitology 11: 64.5- Mutation Research 26: 177-187.
651. SZASZ J., BURNS R. & STERNLICHT H. 1982. Effects of
SCHEDL T., OWENS J., DOVE W. F. & BURLAND T. G. 1984. reductive methylation on microtubule assembly. Journal
Genetics of the tubulin gene families of Physarum. of Biological Chemistry 257: 3697-3704.
Genetics 108: 143-164. TAKAHASHI M., IWASAKI S., KOBAYASHI H., OKUDA S.,
ScatfF P. B. & HORWITZ S. B. 1981. Tubulin: a target for MURAL T. & SATOY. 1987. Rhizoxin binding to tubulin at
chemotherapeutic agents. In: ~o~e~~~u~ Actions und the maytansine binding site. Biochimica et Biophysics
Targets for Cancer Chemotherapeutic Agents (Edited by Acta 926: 215-223.
SARTORELLIA. C.. LAZO J. S. & BERTINO J. R.), pp. 483- TAN L. P. & LA~NADO .I. R. 1975. Effects of indole
507. Academic Press, New York. alkaloids and related compounds on the properties of
SCHMITT H. & ATLAS D. 1976. Specific affinity labelling of brain microtubular protein. Biochemical Societv Trans-
tubulin with bromocolchicine. Journal of Molecular actions 3: 121-124.’
Biology 12: 743-758. TIMASHEFF S. N. & GRISHAM L. M. 1980. In vitro assembly
SCHMITT H. & KRAM R. 1978. Binding of antimitotic drugs of cytoplasmic microtubules. Annual Review of Bio-
around cysteine residues of tubuhn. Experimental Cell chemistry 49: 565-591.
Research 115: 408-4 11. THOMAS J. H., NEFF N. F. & BOTSTEIN D. 1985. Isolation
SEILER J. P. 1975. Toxicology and genetic effects of and characterization of mutations in the P-tubulin gene of
benzimida~ole compounds. ~u~~l~on Research 32: 15 I- Sn~charomy~es cerevisiae. Genetics 112: 7 15-734.
168. UMENESONO K., TODA T., HAYASHI S. & YANACXDA M.
SERRANO I~., AVILA J. & MACCIOF~I R. B. 1984. Limited 1983. Two cell division cycle genes NDA 2 and NDA 3 of
proteolysis of tub&n and the localization of the binding the fission yeast Schizosaccharomyces pombe control
site for colchicine. Journal of Biological Chemistry 259: microtubular organization and sensitivity to anti-mitotic
6607-6611. benzimidazole compound. Journal of Molecular Biology
SERRANO L. & AVILA J. 1985. The interaction between 168: 271-284.
subunits in the tubulin diner. Biochemicul Journul 230: VALLEE R. B. 1986. Reversible assembly purification of
551-556. microtubules without assembly-promoting agents and
SERRANOL., VALENCIA A., CABALLERO R. &AVILA J. 1986. further purification of tubulin, microtubule-associated
Localization of the high affinity c~cium-binding site on proteins and MAP fragments. In: .~et/zods in Enzym-
tub&n molecule. ~oz4rnu~ of BioIogical Chemistry 261: ology, Vol. 134 (Edited by VALLEE R. B.), pp. 89-104.
7076-7081. Academic Press, New York.
SERRANO L., WANDOSELL. F. & AVILA 3. 1986. Location of VALLEE R. B., BLOOM G. S. & LUCA F. C. 1986. Differen-
the regions recognised by five commercial antibodies on tial structure and distribution of the high molecular
the tubulin molecule. Analytical Biochemistry 159: 253- weight/brain microtubule-associated proteins, MAP-l
259. and MAP-2. In: Dynamic Aspects of Microtubule Biology.
SHEIR-NEISS G., LAI M. H. &MORRIS N. R. 1978. Identifi- Annals of the New York Academy of Sciences, Vol. 466
cation of a gene for ,!I-tubulin in Aspergillus nidulans. Cell [Edited by SOIFER D.), nn. 134-144. The New York
15: 639-647. Academy of Sciences, NewYork.
SHELAKSKI M. L., GASKIN F. & CANTOR C. R. 1973. VERHEYEN A., BORGERS M., VAHPARIJS 0. & THIENPONT D.
Microtubule assembly in the absence of added nucleo- 1976. The effects of mebendazo:e on the u~trastructure of
tides. Proceedings of the National Academy of Sciences of cestodes. In: B~o~hern~r~ of Parasites and Host-Parasite
the United States ofilmen’ca 70: 763-768. Relationships (Edited by BOSSCHEH. VANDEN), pp. 60%
SHERLINE P., BODWIN C. K. & KIPNIS D. M. 1974. A new 618. North-Holland Publishing, Amsterdam.
colchicine binding assay for tubulin. Analytical Bio- \ ‘OGEL D. G., MARGOLIS R. L. & MOTTET N. K. 1985. The
chemistry 62: 400-407. effects of methyl mercury binding to microtubules. Tooxi-
SHERLINE P., SCHIAVONEK. bi BROCATO S. 1979. Endogen- cologyand Applied Pharmacologv 80: 473-486.
ous inhibitor of colchicine-tubulin binding in rat brain. WALLER P. J. & LACEY E. 1985. Nematode growth regula-
Science 205: 593-595. tors. In: Resistance in Nematodes to Anthelmintic Drugs
SHlAU G. T.. DE K. K. & HARMON R. E. 1976. Alkylthio- (Edited by ANDERSON N. & WALLER P. J.), pp. 137-147.
cotchicines and ~-deacetyl-alkylthiocolchicines and their CSIRO Division of Animal Health, Glebe, NSW.
936 E. LM3.Y

WALLER P. J. 1986. Anthelmintic resistance in nematode WILSON L. 1975. Microtubuies as drug receptors: nharma-
parasites of sheep. In: Agricultural Zoology Reviews. cological properties of microtubule protein. In: Atrnafs of
Vol. 1 (Edited by RUSSELL G. E.), pp. 333-373. Inter- the New York Academv ofSciences. Vol. 253 fEdited bv
cept, Newcastle upon Tyne. SOIFER D.), pp. 2121231. New York Academy of
WALLER P. J. & PRICHARD R. K. 1986. Drug resistance in Sciences, New York.
nematodes. In: Chemotherapy of Parasitic Diseases WOLFF A., DENOULET P. & JEANTET C. 1982. High level of
(Edited by CAMPBELL W. C. & REW R. S.), pp. 339-362. tubulin microheterogeneity in the mouse brain. Neuro-
Plenum Press, New York. science Letters 31: 323-328.
WATTS S. D. M. 1980. The preparation of a fraction with YAMAMOTO M. 1980. Genetic analysis of resistant mutants
tubutin-like properties from the rat tapeworm ~ymenuie- to antimitot~c benzimid~ole compounds in Schizosuc-
pis diminuta. Biochen~ica~
Society Transactions 8: 7 l-72. ~harorn~,~es uombe. ~~oie~u~~r and General Genetics
WATTS S. D. M. 1981. Colchicine binding in the rat tape- 180: 231-234.
worm, Hymenolepis diminuta. Biochimica et Biophysics YEH H. J. C.. CHRZANOWSKA M. & BRCSSI A. 1988. The
Acta 667: 59-69. importance of the phenyl-tropolone ‘as’ configuration in
WATTS S. D. M., RAPSON E. B., ATKINS A. M. &LEE D. L. coichicine’s binding to tub&. /XBS Letters 229: 82-86.
1982. Inhibition of acetylcholinesterase secretion from ZAVALA F.. GU~NARD D. & POTIER P. 1978. Interaction of
Nippostrongylus brasiliensis by benzimidazole anthelmin- vinblastine analogues with tubulin. Experientia 34: 1497-
tics. Biochemical Pharmacology 31: 3035-3040. 1499.
WEHLAND J., HERZOC W. & WEBER K. 1977. Interaction of ZAVALA F., GU$NARD I)., ROBIN J.-P. & BROWN E. 1980.
eriseofulvin with microtubuIes, microtubule protein and Structure-~t~tubu~in activity relationships in steganacin
Fubulin. Jortrnal of ~~o~e~i6~ar Biology 111: 329-342. congeners and analogues. rnhibition ot tubuhn polymer-
WEISENBURG R. C., BORISY G. G. & TAYLOR E. W. ization in vitro by (~)-isodeoxypodophyllotoxin. Jozrvnaf
1968. The colchicine-binding protein of mammalian ofMedicinal Chemistry 3: 546-546.
brain and its relation to microtubules. Biochemistry 7: ZEEBERC 8. & CAPI.OW M. 1979. Determination of free and
4466-4478. bound microtubular protein and guanine nucleotide
WELKER D. L. & WILLIAMS K. L. 1983. Genetic loci asso- under equilibrium conditions. Biochemistrv 18: 3880-
ciated with altered resistance to microtubule inhibitors 3886. .
and with spore shape in Dictyostelium discuideum. ZIMMERMAN H.-P.. DOENGES K. H. & MOLL E. 1982.
Journal of General Microbiofonv 29: 2X7-22 16. Deacetyicolchicine conjugated to a heme-nonapeptide act
WILLIAMS R. F., MUMFORD C. L, WILLIAMS G. A., FLOYD with both soluble tub&in and microtubules polymerized
L. J., AIVALIOTIS M. J.. ,MARTINEZ R. A., ROBINSON in vitro EL~ropeanJo~~rnufof Celf Biofo~~ 26: 3 1 O-3 14.
A. K. & BARNES L. D. 1985. A photoaffin~ty derivative of Z~NTZ K. & FRANK W. 1982. ~~ltrastructurai modifications
colchicine: 6 ‘-(4 ‘-azido-2 ‘-nitrophenylamino) hexanoyl- in Heterakis spumosa after treatment with febantel and
deacetyicolchicine. Journal of Biologica/ Chemistry 260: mehendazole. Veterinary Parasitology 10: 47-56.
13,794-13,802. ZWEIC M. H. & CHIGNELL C. F. 1973. Interaction of some
WILSON L. 1970. Properties of colchicine binding protein colchicine analogs, vinblastine and podophyllotoxin with
from chick embryo protein. Interactions with vinca alka- rat brain microtubule protein. Riochemical Pharmaco/og~
loids and podophyllotoxin. Biochemistry 9: 4999-5007. 22:2141-2150.