CORE Metadata, citation and similar papers at core.ac.

uk
Provided by Elsevier - Publisher Connector

Biochimica et Biophysica Acta 1496 (2000) 36^51

www.elsevier.com/locate/bba

Review
Class V myosins
a;
Samara L. Reck-Peterson *, D. William Provance Jr. d , Mark S. Mooseker a;b;c
,
John A. Mercer d
a
Cell Biology Department, Yale University School of Medicine, New Haven, CT 06520, USA
b
Pathology Department, Yale University School of Medicine, New Haven, CT 06520, USA
c
Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, CT 06520, USA
d
McLaughlin Research Group, Great Falls, MT 59405, USA

Received 18 August 1999; accepted 30 August 1999

Keywords: Myosin V; Unconventional myosin; mRNA transport; Membrane tra¤c; Vacuolar inheritance

1. Introduction of the mouse and yeast class V myosins provided the

The myosin family of actin-based molecular mo-
tors consists of 15 known classes that are structurally
distinct based on comparisons of the primary struc-
ture of the motor domains of the known myosin
heavy chain genes [1^3]. Information regarding the
function and/or biochemical properties of most of
these myosin classes is sparse relative to the well-
characterized class II and class I myosins; neverthe-
less, the range of proposed functions for these myo-
sins is already remarkably broad [1]. There are a
number of recent reviews that provide an overview
of the rapidly growing myosin gene family [1,4^8].
Among the best characterized and functionally di-
verse of the recently discovered myosin classes are
the class V myosins, the focus of this review (for
other reviews see [9^11]).
Myosin-V was initially characterized as an unusual
calmodulin binding protein from brain with a num-
ber of myosin-like biochemical properties [12^14].
Subsequently, myosin-V heavy chain genes were
cloned from mouse, yeast and chicken, thus de¢ning
the ¢fth class of actin-based motors [15^19]. Studies
* Corresponding author. Fax: +1-203-432-6161;
E-mail: reckpesl@biomed.med.yale.edu
¢rst insights regarding the cellular function of myo-
sin-V. Phenotypes of the mutant dilute mouse and
the temperature-sensitive yeast mutant, myo2-66 led
to the hypothesis that class V myosins may function
in cell polarity and membrane tra¤cking [15,16].
Moreover, mutations in the human ortholog of the
dilute heavy chain gene cause Griscelli syndrome, a
rare recessive disease characterized by pigmentary
dilution and in most, but not all cases immunode¢-
ciency [20^22]. Neurological disorders have also been
reported in Griscelli syndrome patients [22,23]. Since
these ¢rst studies, a great deal has been learned
about the biochemistry, biophysics and cellular func-
tion of the class V myosins. This review will discuss
the emerging evidence that myosin-V is a processive
actin-based motor that has multiple functions in the
cell ranging from mRNA transport, cell polarity and
membrane tra¤cking.
There are currently nine complete myosin-V heavy
chain sequences known (Fig. 1). Analysis of the two
completed eukaryotic genomes of Saccharomyces ce-
revisiae and Caenorhabditis elegans reveals that yeast
have two class V myosins, while C. elegans has a
single class V myosin heavy chain gene. In verte-
brates, there are at least three distinct subclasses of
myosin-V. The three most closely related heavy chain
sequences, that of chicken brain myosin-V, the

0167-4889 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 7 - 4 8 8 9 ( 0 0 ) 0 0 0 0 7 - 0

BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart

 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51 37

myosin in humans, myosin-Vc, has also been identi-

Fig. 1. Phylogenetic analysis of the motor domains of the class
V myosins. Sequence alignments of the motor domains (amino
acids 1^772 for dilute) were performed using the clustal pro-
gram from laser gene. The following accession numbers were
used for this analysis: Mouse MVa dilute (Q99104), Human
MVa (AAD00702), Chicken MVa (Q02440), Rat Myr6
(AAB38840), S. cerevisiae Myo2p (CAA99646), S. cerevisiae
Myo4p (AAC05003.1), S. pombe (CAA22641), Drosophila
(AAC99496), C. elegans hum-2 (AAA97926).
¢ed [28]. Comparison of mRNA and protein expres-
sion data for myosin-Va, b and c, shows that the
three myosin-V isoforms are expressed di¡erentially;
however it remains to be shown whether they have
distinct or overlapping functions within cells
[15,17,26,28]. At least two of these classes have
been shown to be expressed in a single vertebrate
cell [24]. In S. cerevisiae, the two myosin-V genes,
MYO2 and MYO4, appear to have distinct and
non-overlapping functions (see below). While se-
quence alignments show that Dictyostelium MyoJ
and the plant class VIII and XI myosins are closely
related to the class V myosins, only analysis of the
cellular role of these myosins will reveal whether they
share mechanochemical and functional properties
with the class V myosins [1,24,29].

mouse dilute gene and human myosin-V have been

classi¢ed as myosin-Va [24,25]. Rat myr6 [26] repre- 2. Domain structure and function
sents the only full-length myosin-Vb sequence,
although partial murine and human myosin-Vb se- The myosin-V heavy chain consists of three prin-
quences exist in the database [27]. A third class V cipal domains. The amino terminal motor domain

Fig. 2. Domain structure of the class V myosins. To scale bar diagrams representing the domain structure of some representative class
V myosin heavy chain sequences. The most amino terminal domain is the motor domain. The neck domain contains the putative es-
sential light chain (ELC) and calmodulin (CaM) binding sites. The tail domain can be divided into two regions: a region predicted to
from K-helical coiled-coil (C-C) and a carboxyl terminal globular domain (globular tail). Some myosin-V sequences have a PEST site
(*) and all myosin-V sequences have an AF-6 homology domain (t) present in the globular tail domain.

BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart

38 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51

Table 1 domain [31,32]. Most of these mutations are found

Domain homology to dilute myosin-Va and Myo2p in residues that are conserved among all myosins.
Myosin Motor (%) First IQ (%) Globular tail (%) One exception is the dilute P60R mutation. This pro-
(A) Domain homology to dilute myosin-Va line is conserved in all class V myosins and is present
Human Va 98 83 99 in the region of the motor domain that is amino
Chicken Va 94 78 98 terminal to the ATP binding site. This amino termi-
myr6 72 52 65
nal region is highly variable among myosin classes
Drosophila 54 22 28
C. elegans 47 39 24
and its function is unknown. Mice with the P60R
Myo2p 45 44 20 mutation demonstrate severe phenotypes for both
Myo4p 43 35 13 pigmentary dilution and neurological dysfunction
S. pombe 42 22 14 [32]. In vitro motility assays show that this mutation
causes a 7-fold decrease in the rates of actin ¢lament
(B) Domain homology to Myo2p
translocation [33]. We have performed sequence
dilute Va 45 44 20
Human Va 44 44 19 alignments to assess the relative importance of this
Chicken Va 45 39 19 region (amino acids 1^62 in the dilute protein) within
myr6 47 30 16 the class V myosins by comparing the relative rate of
Drosophila 42 35 14 change within the amino terminal region to the rel-
C. elegans 38 22 16
ative rate of change within the conserved core region
Myo4p 71 70 18
S. pombe 50 17 14
of the motor domain (motor domain minus loop 1, 2
(see below) and the variable amino terminal region;
The percent identity of the motor domain (amino acids 1^772
for dilute), ¢rst IQ domain (amino acids 773^796 for dilute),
see [34] for details regarding this type of analysis).
and the globular tail domain (amino acids 1445^1853 for dilute) We found that the extreme amino terminus is not
of all of the class V myosins compared to either dilute myosin- well conserved within the class V myosins. As this
Va (A) or yeast Myo2p (B). Sequence alignments were per- region of the molecule may lie near the neck^tail
formed using the clustal program from laser gene. junction based on the chicken skeletal muscle myosin
II crystal structure [7,35], it is possible that this re-
contains binding sites for both actin and ATP and is gion has evolved to modulate very speci¢c motor^
followed by the neck or regulatory domain that con- neck or motor^cargo interactions. Alternatively, giv-
tains six light chain binding sites. The tail, which en the in vitro motility results, this region may be
constitutes the remaining carboxyl terminal region involved in the regulation of species-speci¢c motor
of the molecule, can be divided into two regions: properties.
an K-helical coiled-coil region that is presumably in- Other regions of the motor domain that could po-
volved in dimerization and a carboxyl terminal glob- tentially encode class-speci¢c functions include two
ular domain (Fig. 2; [30]). The motor domain and regions identi¢ed as surface loops in the crystal struc-
¢rst IQ motif are generally well conserved among the ture of chicken skeletal muscle myosin II that corre-
class V myosins, while the globular tail domain is spond to well-characterized myosin II proteolytic
more divergent. To illustrate this point, we have sites [7,35]. Based on their position in the myosin
compared the percent identity of dilute myosin-Va molecule, these loops may be important for both nu-
and yeast Myo2p to these domains of all other class cleotide and actin binding [7,34,35]. Loop 1 (25^50
V myosins (Table 1). kDa junction) is near the ATP binding site; substi-
tutions within this region in class II myosins have led
2.1. Motor domain to altered ADP release rates (reviewed in [34]). Loop
2 (50^20 kDa junction) has been implicated in actin
All myosins share a core of conserved residues in interaction; substitutions within this domain in class
their motor domains, many of which are known to II myosins have altered the a¤nity of myosin for
participate in nucleotide and actin binding (for re- actin and have altered the actin-activated ATPase
view see [7]). One mutation in MYO2 and seven mu- rates of the myosin (reviewed in [34]). An investiga-
tations in dilute have been mapped to the motor tion of the relative importance of these loop domains

BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart

 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
within the vertebrate myosin-Va sequence, revealed
that loop 1 is relatively conserved, while loop 2 is not
[34]. We have carried out a similar analysis on all of
the full-length class V myosin sequences and have
found that this trend holds true across all of the class
V myosin heavy chain sequences, suggesting that
loop 1, but not loop 2, may have a class-speci¢c
function. Given that loop 2 is thought to contribute
to actin interaction, the lack of conservation of this
region among the class V myosins could indicate that
not all of the class V myosins will be processive mo-
tors (see below) or that this region is not important
for processive movement.
2.2. Neck domain
The neck domain of class V myosins contains six
light chain-binding motifs called IQ motifs [15,17,19].
This motif has the consensus of IQXXXRGXXXR
and is known to be the binding site for calmodulin
and myosin light chains that are members of the EF
hand superfamily (for review see [36]). Unlike the
other class V myosins, the Schizosaccharomyces
pombe myosin-V sequence has only ¢ve IQ motifs.
S. pombe appears to lack the ¢rst IQ motif; however
the amino acid sequence between the end of the mo-
tor domain and the ¢rst bona ¢de IQ motif repre-
sents the length of one IQ motif, raising the possi-
bility that this sequence of S. pombe may bind a
unique light chain. Biochemically puri¢ed chicken
myosin-Va co-puri¢es with 4^5 calmodulin molecules
per heavy chain [30]. In addition, Myosin-Va co-pu-
ri¢es with additional light chains that correspond to
the 17- and 23-kDa essential light chains (ELC) en-
coded by the chicken LC17 and 23 genes, respec-
tively, both of which are also components of chicken
brain non-muscle myosin II [37^40]. Assuming that
the ELC occupies the same position in myosin-V as
it does in myosin II, the ELC would be bound to the
¢rst IQ motif [35,41]. In support of IQ1 having a
speci¢c function in binding the essential light chain,
IQ1 is more conserved between all myosin-V classes
(with the exception of Drosophila myosin-V) than it
is to the other IQ motifs within a given myosin-V
protein. Furthermore, yeast Myo2p has been shown
to bind both calmodulin and Mlc1p, a calmodulin/
EF hand superfamily member that may function as
the essential light chain in yeast [42,43].
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
39
Various roles for calmodulin and other members
of the EF-hand superfamily myosin light chains have
been proposed. Light chain binding may stabilize a
lever arm that is necessary to generate step size and
sliding velocity [35,41,44^46]. Interestingly, yeast
strains lacking all six Myo2p IQ motifs grow nearly
as well as wild-type yeast strains, indicating that the
neck domain is not essential for the function of
Myo2p. On the other hand, yeast over-expressing
Myo2p grow poorly, but in the presence of over-ex-
pressed Mlc1p grow normally [43]. Over-expression
of Myo2p may result in poor growth due to the
depletion of light chains from Myo1p (the yeast class
II myosin), as the phenotype of Myo2p over-expres-
sion resembles that of myo1 mutants [47^49]. Over-
expression of Myo4p also causes a myo1-like pheno-
type, suggesting that Myo4p may also share a light
chain with Myo2p and Myo1p [47]. These results are
consistent with the idea that the primary function of
the light chains may be to provide structural support
to the neck domain of Myo2p. However, as the
length of the neck domain correlates with step size
[44,46], this neckless Myo2p may not be capable of
processive movement along actin ¢laments. Thus, if
yeast Myo2p is a processive motor like myosin-Va
(see below), the viability of the neckless Myo2p raises
the possibility that its essential function may not re-
quire processive movement.
In higher eukaryotes, the light chains appear to
have additional regulatory roles as brain myosin-V
is regulated by calcium in vitro. The actin-activated
ATPase of brain myosin-V increase dramatically in
the presence of micromolar range calcium [30,50]. In
addition, high a¤nity binding of myosin-V to actin
in the presence of ATP requires calcium [50,51]. Cal-
cium-dependent regulation of brain myosin-V may
be mediated by changes in the a¤nity of one or
more calmodulins for the neck domain. As was ¢rst
shown for brush border myosin-I [52^54], addition of
calcium to brain myosin-V results in partial dissoci-
ation of calmodulin from the heavy chain [50]. Thus,
one mode of regulation of myosin-V may be through
altering the £exural rigidity of the neck domain by
either calmodulin light chain dissociation or altera-
tion in light chain binding a¤nity. However, as
noted below, in vitro motility of myosin-Va is inhib-
ited by calcium, leaving the in vivo role of calcium in
myosin-V function unknown [30,55]. The essential
40 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
functions of yeast Myo2p must not be calcium regu-
lated as calmodulin mutants that fail to bind calcium
in vitro are viable [56].
2.3. Tail domain
The tail domain of class V myosins consists of a
region predicted to form K-helical coiled-coil of var-
iable length followed by a globular domain thought
to be involved in cargo binding and/or localization
within the cell. Electron microscopy of myosin-Va
revealed that chicken brain myosin-Va is indeed a
two-headed motor with a globular tail domain [30].
2.3.1. AF-6 homology domain
The only striking homology to the myosin-V tail
domain is to the AF-6/canoe family of proteins. AF-6
was found as an ALL-1 fusion partner (this gene
fusion is found in some patients with acute leukemia)
and by using Ras as bait in a yeast two-hybrid screen
[57,58]. In Drosophila, the AF-6 homolog, canoe, ex-
hibits genetic interactions with both the Notch and
Ras signaling pathways [59,60]. The C. elegans ho-
molog of AF-6 also interacts genetically with the Ras
signaling pathway [61]. AF-6 has a complex domain
structure including a ras-binding domain, a domain
found in kinesin-like proteins, and a PDZ domain, in
addition to the myosin-V-like domain [62]. This com-
plex domain structure suggests that AF-6 may repre-
sent an important link between the cytoskeleton, cell
signaling pathways, and cell^cell interactions. AF-6
has been localized to both the adherens and tight
junctions [63,64]. In support of AF-6 having a role
in adherens and/or tight junction function, the neu-
roectoderm of af-6 knock-out mice has large intra-
cellular gaps, reduced extent of cell^cell junctions,
and loss of cell polarity [65]. These mice die by 10
days post coidum presumably due to placental failure
[65]. However, the role that the myosin-V-like do-
main plays in AF-6 function is not understood.
2.3.2. PEST site
When the chicken myosin-V heavy chain gene was
cloned, a PEST sequence was found within the
coiled-coil containing region of the tail domain
[17,66]. PEST sequences are regions rich in the amino
acids proline, glutamic acid, serine and threonine and
are found in a number of proteins that turnover rap-
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
idly [66]. PEST sequences have been shown to be
important for proteolysis mediated by the 26S pro-
teosome and the calcium-dependent protease caplain
(for review see [67]). Indeed, chicken myosin-Va is
cleaved by calpain and the cleavage site has been
mapped to one amino acid downstream of the
PEST site located in the tail domain [14,50]. All of
the vertebrate myosin-V heavy chain sequences con-
tain PEST sequences in the equivalent region of their
tail domains. It remains to be determined if cleavage
of myosin-V at the PEST site has a physiological role
in uncoupling the motor domain from the putative
cargo-carrying domain. In contrast to the vertebrate
class V myosins, PEST sequences are not present in
the tail domains of the yeast, worm or £y myosin-V
proteins. However, Myo4p does contain a PEST site
located in loop 2 of the motor domain. Chicken my-
osin-Va is also cleaved by calpain in a region pre-
dicted to be in loop 2, although no PEST site is
present in this region of myosin-Va [50]. Thus, it is
possible that both motor domain and tail function
could be regulated by proteolysis.
2.3.3. Tail domain mutations
Characterization of tail mutants from dilute mice,
and myo2 yeast has provided some insight into which
regions of the tail may be important for function
[68,69]. Ten characterized dilute mutations map to
the tail domain of myosin-Va. Three of these muta-
tions (I1510N, M1513K, D1519G) map to within 10
amino acids of each other in a region that is just
amino terminal to the AF-6/canoe homology do-
main. Although these mutations cause a relatively
mild dilute phenotype (some pigment dilution and
no neurological phenotype), the mutated amino acids
are well conserved in all metazoan class V myosins.
Deletion of the last 13 amino acids of the tail domain
causes a relatively strong phenotype-implicating
these amino acids, which are highly conserved from
£y to man, in an important function. Five of these 13
amino acids are invariantly conserved in all class V
myosins except C. elegans and S. pombe. In S. cere-
visiae, mutation of a non-conserved glycine (1248) to
aspartic acid results in defects in cargo binding [69].
2.3.4. Tail domain light chain
Puri¢ed chicken myosin-Va co-puri¢es with an ad-
ditional light chain, the `8 kDa' dynein light chain
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
(DLC; actual molecular weight is 10 kDa), which
has been shown to bind to a region of the tail do-
main carboxyl terminal to the PEST site [37,70]. The
stoichiometry of this interaction is two DLC's per
myosin-V dimer [37]. The DLC is highly conserved
and has been identi¢ed in a number of organisms
including Chlamydomonas, S. cerevisiae, Drosophila,
Aspergillus, Schistosoma mansoni, and humans [71^
77]. In addition to being a component of the myosin-
Va and dynein holoenzymes, the DLC also interacts
in vitro with NO synthase and IkB, a cytoplasmic
inhibitor of the NFkB transcription factor [78,79].
Biochemical studies of the DLC have shown that it
exists as a dimer when bound to myosin-Va and
dynein [70]. In Aspergillus, the DLC mutant,
nudG8, is a phenocopy of a dynein heavy chain mu-
tant, suggesting that the primary role for the DLC in
this organism may be mediated through its interac-
tion with dynein [76]. However, possible DLC inter-
actions with myosin-V or other proteins in Aspergil-
lus cannot be ruled out. In S. cerevisiae, deletion of
the DLC gene (SLC1) does not result in a myo2-66-
like phenotype [74]. Given the multiple interactions
of the DLC, it has been proposed that the DLC may
have a variety of regulatory roles within the cell
[37,70]. Possible functions for the dynein light chain
in myosin-V function include stabilization of heavy
chain^heavy chain interactions and a role in cargo
binding.
2.3.5. The tail domain and cargo binding
One proposed function of the tail domain is cargo
binding. However, to date, very little is known about
the putative cargo molecules that interact with the
tail domain of myosin-V or the proteins that could
regulate such interactions. Several studies have impli-
cated class V myosins in membrane binding. Myosin-
Va co-puri¢es with synaptic vesicles and co-precipi-
tates with the synaptic vesicle proteins, synaptobre-
vin, synaptophysin, and syntaxin [80,81]. Further-
more, myosin-Va can be chemically cross-linked to
synaptobrevin and synaptophysin, showing that this
interaction may be quite direct [80]. It is unlikely,
though, that myosin-Va associates primarily with
mature synaptic vesicles, as immunogold-labeling of
a synaptic vesicle enriched membrane fraction re-
vealed that the majority of myosin-Va-labeled
vesicles were larger than synaptic vesicles. These my-
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
41
osin-Va vesicles were also labeled by the synaptic
vesicle marker protein, SV2, and could represent en-
dosomes or a synaptic vesicle recycling compartment
[82,83]. Association with these membranes is not
mediated solely by the tail domain of myosin-Va,
as both motor and tail fragments of myosin-Va are
capable of membrane binding [82]. In yeast, some
Myo2p co-puri¢es with vacuolar membranes, and a
point mutation in the tail domain of Myo2p disrupts
this association [69].
2.3.6. The tail domain and localization
In addition to cargo binding, the tail domain has
also been proposed to be important for localization.
In melanocytes, myosin-V staining is quite complex
and has been reported on melanosomes [84,85], the
centrosome [86,87], and other organelles including
the Golgi, mitochondria, and endoplasmic reticulum
[84]. In support of the tail domain playing a role in
localization, GFP-tail domain constructs co-localize
with melanosomes [88] and the centrosome [89]. In
yeast, Myo2p localizes to sites of polarized cell
growth, such as the bud tip of dividing cells [31].
Exogenously expressed globular Myo2p tail domain
is su¤cient for localization to the bud tip and causes
displacement of the endogenous Myo2p from the
bud tip [90]. This exogenously expressed tail domain
causes a phenocopy of the yeast myo2-66 phenotype
[90]. Similarly, over-expression of the tail domain of
myosin-Va in mouse melanocytes causes a dilute phe-
notype [88]. Conversely, strains in which the tail do-
main of Myo2p have been deleted are inviable [69].
Taken together, these results indicate the essential
nature of the class V myosin tail domain. In addi-
tion, a point mutation in the tail domain of Myo2p
also causes mislocalization from sites of polarized
growth [69]. Thus, one function of the tail domain
is to localize class V myosins to di¡erent cellular
locations. However, these studies do not rule out
the possibility that other regions of the molecule
could contribute to localization.
2.3.7. Tail-binding proteins
Myosin-V tail binding proteins that could be in-
volved in cargo binding and/or localization are be-
ginning to be identi¢ed. In yeast, the coiled-coil do-
main of Myo2p associates with Rho3p, one of four
Rho family members in yeast [91]. While the pheno-
42 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
types of rho3 and myo2 mutants suggest that both
proteins are involved in polarized cell growth, the
role of the Rho3p^Myo2p interaction in this process
remains unclear. In vertebrates, the globular tail do-
main of myosin-Vb was recently shown to interact
with the ring ¢nger containing protein, BERP [92].
In vivo, BERP also co-immunoprecipitates with
MVa. Over-expression of the BERP protein lacking
the myosin-V-binding domain results in defects in
NGF-induced neurite outgrowth [92]. However, the
cell biological role of the BERP^myosin-V interac-
tion remains unknown. Additionally, rat brain myo-
sin-Va co-immunoprecipitates with calmodulin-de-
pendent protein kinase II (CaM Kinase II), an
interaction that is mediated by the tail domain of
myosin-V [81]. In vitro, myosin-V is a CaM Kinase
II substrate and can act to stimulate the activity of
CaM Kinase II without a requirement for additional
calmodulin, suggesting that brain myosin-V activates
CaM Kinase II by donating calmodulin molecules
[81]. As is the case with the other myosin-V tail bind-
ing proteins, the in vivo signi¢cance of this interac-
tion remains unknown. Finally, the tail domain of
Myosin-Va has recently been shown to interact
with kinesin and a yeast kinesin related protein,
Smy1p interacts with Myo2p ([93,94] see below).
3. Mechanochemistry
Brain myosin-Va is the only myosin-V to be puri-
¢ed and characterized biochemically [14,30,50,95]. In
addition, studies are beginning to be performed on
baculovirus-expressed tail-less myosin-Va co-ex-
pressed with its light chains [96,97].
Chick brain myosin-Va exhibits marked activation
of its Mg-ATPase in the presence of actin, a diag-
nostic feature of most myosins characterized to date.
Like molluscan muscle myosin-II [98], this activation
requires the presence of calcium [30,50]. At 37³C, in
the presence or absence of calcium, the steady-state
Mg-ATPase rate of myosin-Va is very low ( 6 1
ATP/s per head). In the absence of calcium, addition
of actin has no a¡ect on the Mg-ATPase rate. How-
ever, at calcium concentrations above V3 WM the
Mg-ATPase rate increases dramatically in the pres-
ence of actin, with rates as high as 30^50 ATP/s per
head, making myosin-Va one of the enzymatically
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
fastest myosins characterized [50]. Consistent with
the recent demonstration that myosin-Va is a pro-
cessive motor (see below), maximal activation of the
Mg-ATPase requires very low concentrations of ac-
tin, with a KATPase of only V1 WM (this is V50^100-
fold lower than the KATPase of muscle myosin II). The
Mg-ATPase of baculovirus expressed tail-less myo-
sin-Va is also activated by very low concentrations
of actin, but unlike tissue puri¢ed myosin-V, this
activation does not require calcium [96,97]. Similarly,
separation of the motor from the tail domain by
cleavage with calpain results in a motor that retains
actin activated Mg-ATPase activity, but lacks calci-
um sensitivity [50]. These results suggest that the reg-
ulation of the myosin-V Mg-ATPase by calcium may
require complex, calmodulin-mediated allosteric in-
teractions between the head, neck and tail domains
of the myosin-V molecule.
The low concentrations of actin required to max-
imally activate the Mg-ATPase of myosin-Va indi-
cate that myosin-Va has a high a¤nity for actin.
Indeed, in contrast to other known myosins, myo-
sin-Va exhibits high a¤nity binding to F-actin in
the presence of ATP or ATP-Q-S, as assayed by a
combination of methods including co-sedimentation,
quench of pyrene F-actin £uorescence and electron
microscopy [50,51]. Like activation of the Mg-
ATPase, this high a¤nity binding requires calcium
[51]. In the absence of ATP myosin-V crosslinks ac-
tin ¢laments into bundles [30]. Remarkably, in the
presence of ATP and calcium, myosin-V retains its
ability to crosslink actin ¢laments into bundles [51],
presumably via head^head linkages between two ¢l-
aments since the tail domain does not bind to actin
[50].
The motor properties of myosin-Va have been
studied using in vitro motility assays. Both puri¢ed
brain myosin-Va and baculovirus expressed myosin-
Va move actin at roughly 400 nm/s depending on the
temperature conditions and salt concentrations
[30,55,96]. Chick brain myosin-Va is a barbed-end
directed motor [30,55]. While calcium activates my-
osin-V's ATPase activity, it inhibits myosin-V-based
motility in the in vitro motility assays. This inhibi-
tion is minimized by the addition of exogenous cal-
modulin, but sliding velocities are still reduced [30].
The di¡erence in calcium sensitivity between the
steady-state ATPase of myosin-V in solution and
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
the motility of surface-adsorbed molecules remains
unexplained. It is possible that adsorption to the
motility chamber surface in the absence of calcium
induces a structural change that mimics the e¡ects of
calcium on the molecule in solution. Alternatively,
the inhibitory e¡ects of calcium on motility might
be caused by enhanced adsorption of the neck do-
main to the surface resulting from calcium-induced
calmodulin dissociation. This would restrain the lev-
er arm and result in impaired motility. The protective
e¡ects of exogenous calmodulin addition are consis-
tent with this idea.
Several puri¢ed organelle populations also exhibit
myosin-V motor activity in vitro. Chick brain vesicle
populations enriched for both myosin-Va and synap-
tic vesicle markers can support actin movements at
speeds equivalent to puri¢ed myosin-Va [82]. Inter-
estingly, vesicle associated myosin-Va must be acti-
vated with low levels of detergent to support motil-
ity, suggesting that the myosin-Va on these vesicles
may have been puri¢ed in an inactive state [82]. This
raises the possibility that myosin-Va motor activity
may be regulated spatially in the cell. Squid giant
axon endoplasmic reticulum has also been shown
to support myosin-V based motility [99].
The Mg-ATPase and actin binding studies indicate
that myosin-Va has a high a¤nity for actin in the
presence of ATP; a property unlike other myosins,
but consistent with properties of a processive motor.
Processive motors can undergo multiple cycles of
ATP hydrolysis coupled with movement along a ¢l-
ament without di¡using away from it. As low den-
sities of myosin-V have been proposed to function in
the directed movement of organelles and RNA, this
is an important feature of myosin-V to verify. Recent
studies using in vitro motility assays have demon-
strated that myosin-Va is the ¢rst example of an
actin-based processive motor. In vitro motility assays
revealed that the sliding velocity of actin ¢laments
did not decrease as myosin-Va surface density de-
creased from 1000 to 2.7 molecules per mm2 [46].
Furthermore, at the lowest myosin-Va densities, ac-
tin ¢laments were tethered to the surface at a single
contact site and such ¢laments swiveled about that
point as the ¢lament moved forward through it [46].
The Sellers laboratory presented similar ¢ndings for
baculovirus expressed myosin-V at the 1998 ASCB
meeting [96]. To verify that such ¢lament movements
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
43
are due to single myosin-V molecules two quantita-
tive assays ¢rst developed for analysis of kinesin
processivity were used [100^102]. Determination of
¢lament landing rates, and the ratio of ¢laments
moving a distance greater than their length as a func-
tion of motor density, gave results consistent with
processivity [46].
The biophysical properties of single myosin-V mol-
ecules were also examined using an optical trapping
assay (see [103] for a discussion of this method).
These experiments demonstrated that single myosin-
Va molecules can take multiple steps before stalling
at a resistive trap force of V3 pN, a force somewhat
less than that measured for kinesin [46,104]. More-
over, the dwell time between successive steps in-
creased at 2 mM ATP but not 1 WM ATP (where
ATP binding would be rate limiting), indicating that,
as seen for RNA polymerase, mechanical load can
a¡ect the catalytic cycle [46,105]. Additionally, as
¢rst shown for kinesin [104,106^108], myosin-V can
step backward under load, indicating that load may
a¡ect the probability and direction of the steps [46].
The myosin-Va step size was measured to be 30^38
nm versus the 4^17 nm range reported for myosin II
[46,109]. This large step size is plausible given the
length (V23 nm) and neck angle (greater than
100³) between the two myosin-V heads [30]. As 30^
38 nm approximates the pseudo helix (36 nm) repeat
of the actin ¢lament, a single myosin-Va molecule
could move along a ¢lament without having to spiral
around the ¢lament to ¢nd its next binding site (see
[102] for discussion).
4. Function
A number of functions have been proposed for
class V myosins. The analysis of mutants in yeast
and mouse has provided most of the information
regarding function to date. The emerging evidence
suggests that a single class V myosin may have a
number of di¡erent functions within a single cell.
For instance, in yeast, Myo2p appears to function
in vacuolar inheritance, polarized cell growth and
cell polarity. The mouse dilute protein also appears
to have multiple functions including melanosome
and SER membrane tra¤cking. What remains gen-
erally unclear in the case of Myo2p and dilute is
44 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
which proposed cell biological functions are direct
versus indirect consequences of the absence of func-
tional myosin-V protein.
4.1. mRNA transport
Currently one of the most well-de¢ned functions
for a class V myosin is the role of yeast Myo4p in
mRNA transport. After cell division in yeast, only
mother cells are able to switch mating types. The HO
endonuclease induces mating type switching in moth-
er cells by causing a double stranded break at the
MAT locus (for review see [110]). Daughter cells
are kept from switching mating types because a tran-
scriptional repressor of the HO endonuclease, Ash1p,
is speci¢cally localized to daughter cells [111,112].
Five genes were identi¢ed as being needed for moth-
er-cell-speci¢c HO expression. All ¢ve of these genes,
termed SHE genes, are also required to promote
daughter-cell-speci¢c Ash1p nuclear localization
[112,113]. Cloning of the SHE1 gene revealed that
SHE1 is identical to MYO4, a non-essential yeast
class V myosin gene whose function was previously
unknown [47,112,113].
The discovery that ASH1 mRNA, along with
Ash1 protein, is asymmetrically distributed to daugh-
ter cells was the ¢rst example of mRNA localization
contributing to cell fate decisions in a single-celled
organism [114,115]. Both ¢lamentous actin and
Myo4p are required for ASH1 mRNA localization
[114,115]. In further support of a direct involvement
of Myo4p in ASH1 mRNA localization, GFP-la-
beled ASH1 mRNA particles move in a directed
manner at 400 nm/s, speeds that are similar to those
reported for puri¢ed chicken brain myosin-V
[30,46,116]. However, approximately 10-fold slower
rates were recently reported by Beach et al. [117]
using a similar localization method. Importantly, in
myo4 knock-out cells, ASH1 mRNA remains in
mother cells and no long-range directed movements
can be detected [116,117]. The ¢nding that ASH1
mRNA co-immunoprecipitates with Myo4p further
demonstrated a direct involvement of Myo4p in
ASH1 mRNA transport [118]. Thus, it seems likely
that Myo4p functions to directly move an ASH1
mRNA containing particle from mother cells to the
bud tip. The other She proteins appear to play roles
in both Myo4p/ASH1 particle assembly (She2p,
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
She3p) as well as roles in transport to or tethering
in the daughter cell (She4p/Dim1p, She5p/Bni1p;
[114^119]. Additionally, Bud6p/Aip3p, a protein re-
quired for organization of the actin cytoskeleton, is
required for tethering of ASH1 mRNA in the bud tip
of daughter cells [117,120].
A number of interesting questions remain, such as:
How is Myo4p linked to ASH1 mRNA? Is ASH1
mRNA transported as a polysome particle? How is
translation of ASH1 regulated? For instance, it was
recently shown that translation of Ash1 protein is
necessary for anchoring ASH1 mRNA to the daugh-
ter cell cortex [121]. Finally, is this role in RNA
transport speci¢c to Myo4p- or might other class V
myosins be involved in RNA transport?
4.2. Membrane tra¤cking
Many of the phenotypes associated with loss of
function of a class V myosin involve a membrane
tra¤cking defect. dilute mice demonstrate defects in
both melanosome tra¤cking in melanocytes and
smooth endoplasmic reticulum (SER) tra¤cking in
neurons. Biochemical evidence suggests that myo-
sin-Va may have other membrane tra¤cking roles
as well. In yeast, Myo2p is essential for vacuolar
inheritance and may also function in other mem-
brane tra¤cking pathways, such as exocytosis.
4.2.1. dilute melanosome tra¤cking
As the name implies, dilute mice have a lightened
coat color compared to their littermates. Normally,
melanocytes synthesize pigment in melanosomes and
deliver them to keratinocytes via their dendritic pro-
cesses. dilute melanocytes have morphologically nor-
mal dendritic processes and show no defects in me-
lanosome morphology [122,123]. However, in dilute
melanocytes, melanosomes are clustered around the
nucleus, while wild-type melanosomes are found
evenly distributed throughout the cytoplasm
[122,123]. In further support of myosin-Va having a
direct role in melanosome tra¤cking, myosin-Va is
concentrated in regions of the cell that contain me-
lanosomes in primary melanocyte cultures [122]. In
melanoma cell lines, immuno£uoresence and electron
microscopy have shown that myosin-Va partially co-
localizes with melanosomes [84,85]. These ¢ndings
show that myosin-Va is needed to either directly
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
transport melanosomes to the cell periphery or tether
melanosomes in the cell periphery.
Recent experiments visualizing melanosome move-
ments in live cells support the tethering model [88].
In both wild-type and dilute melanocytes, melano-
somes move rapidly and bidirectionally in a micro-
tubule-dependent manner between the cell center and
the cell periphery. While melanosomes in both wild-
type and dilute melanocytes exhibit microtubule-in-
dependent, presumably actin-dependent, movements;
much less microtubule-independent movement is ob-
served in the dilute melanocytes [88]. Long-range
melanosome transport from the cell center to the
periphery via the actin cytoskeleton is unlikely, as
actin ¢laments are concentrated at the cell periphery
and are not organized linearly [88]. However, actin is
clearly involved in melanosome distribution as its
disruption in melanoma cells leads to a phenocopy
of the dilute mutant phenotype [124]. A reasonable
model based on these studies is that the primary
mode of transport to the cell periphery is microtu-
bule-based, while movement or retention in the actin-
rich dendritic processes is actin and myosin-Va-de-
pendent [10,88].
Melanosome movement in Xenopus and ¢sh mela-
nophores has long been shown to be dependent on
microtubules and their associated motors ([125,126];
see [127] for review). Recently, both actin and myo-
sin-V-dependent motility has been demonstrated in
these systems as well [128^130]. In ¢sh melano-
phores, melanosomes can be stimulated to move
from a perinuclear region to a uniform distribution
in the cytoplasm. This dispersion of melanosomes is
actin and myosin-V-dependent [128^130]. In the
Xenopus melanophore system, puri¢ed melanosomes
move in vitro along actin ¢laments. Myosin-V is
found on these melanosomes [128]. Interestingly, re-
cent work has shown that myosin-V association with
Xenopus melanosomes is cell cycle regulated [130].
Melanosomes treated with interphase extracts are as-
sociated with myosin-V and support movement in
motility assays, while melanosomes treated with mi-
totic extracts are not associated with myosin-V and
movement in motility assays is severely inhibited
[130]. As myosin-V is more heavily phosphorylated
in mitotic extracts versus interphase extracts, this cell
cycle regulation is most likely regulated by phosphor-
ylation [130].
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
45
The existence of alternative splicing in the globular
tail region suggests that myosin-Va has di¡erent car-
goes in di¡erent cell types [131,132], providing a
mechanism for the semidominant suppression of
the dilute coat-color phenotype ^ but not the dilute
neurological phenotype ^ by a mutation at the dilute
suppressor (dsu) locus [133]. Molecular cloning of the
dsu, ashen [134], and leaden [135] loci should provide
important insights into pathways of myosin-V func-
tion, since the coat-color phenotypes of both ashen
and leaden, which are indistinguishable from that of
dilute, are both suppressed by dsu [133].
4.2.2. Griscelli syndrome in humans
Griscelli syndrome in humans is characterized by
pigmentary dilution in which melanosomes are not
present in the dendrites of melanocytes, but rather
are clumped around the nucleus [21,136]. Many Gris-
celli syndrome patients also have immunode¢ciency
and/or a neurological phenotype [21^23,136]. The
similarity between the dilute phenotype and that of
Griscelli syndrome patients made human myosin-Va
a good candidate gene for Griscelli syndrome. In-
deed, linkage analysis placed the Griscelli syndrome
gene in the vicinity of the myosin-Va gene and two
out of three Griscelli syndrome patients examined
bear mutations in the myosin-Va gene [20]. One pa-
tient was homozygous for a point mutation in the
coiled-coil region of the tail domain in an amino
acid that is not conserved among all class V myosins,
but is conserved among the myosin-Va genes. The
second patient was homozygous for a missense mu-
tation that would result in a truncation of myosin-Va
at the end of the motor domain. The third patient
had no mutations in the coding sequence of myosin-
Va. The cell biological basis for the immunological
and neurological Griscelli syndrome phenotypes is
not known. In fact, as not all patients present the
neurological and immunode¢ciency phenotypes, it is
unclear whether these phenotypes are linked to the
myosin-Va gene. So far, no immune system pheno-
type has been demonstrated for the dilute mouse.
Some Griscelli syndrome patients have been reported
to have defects in natural killer (NK) cell function
[136], but similar assays using NK cells from a
null allele of the dilute mouse showed that dilute
NK cells killed target cells as well as wild-type cells
[137].
46 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
4.2.3. dilute and SER tra¤cking
In addition to pigmentation abnormalities, dilute
mice also su¡er from presumed neurological dysfunc-
tions characterized by seizures that eventually result
in death of the dilute-lethal mouse at an age of ap-
proximately 3 weeks [138]. The molecular basis for
these neurological defects is not well understood.
Myosin-V may have a role in ¢lopodial extension
as chromophore-assisted laser inactivation (CALI)
of myosin-V in growth cones, results in defects in
¢lopodial extension [139]. However, neurite out-
growth appears to be normal in dilute-lethal mice
and the cytoskeleton of dilute-lethal growth cones
also appears to be normal [140]. One explanation
for these seemingly contradictory results is that an-
other protein or pathway can perform myosin-V's
function in ¢lopodial extension in the dilute mouse.
Alternatively, CALI could inactivate a myosin-V tail-
binding partner that is involved in ¢lopodial exten-
sion, as the antibody used in the CALI experiments
was directed to the tail domain of myosin-Va.
The ¢nding that the dendritic spines of dilute Pur-
kinje cells lack smooth endoplasmic reticulum (SER),
led to the hypothesis that the neurological disorders
observed in dilute mice could be a result of lack of
SER transport to, or tethering in, the dendritic spines
[141,142]. The SER in the dendritic spines of Pur-
kinje cells sequesters cytosolic calcium and is there-
fore important for maintaining calcium homeostasis
after synaptic transmission. Mutations in the opt
gene, an IP3- gated calcium channel found on the
SER membrane, cause neurological phenotypes rem-
iniscent of dilute mice, thus supporting the idea that
the dilute neurological disorders are due to aberrant
calcium homeostasis in dendritic spines [143]. Fur-
ther support for a role of myosin-V in ER transport
comes from studies using squid axoplasm. As men-
tioned above, myosin-V is present on squid axoplasm
ER-derived vesicles and can support actin-based mo-
tility of these ER-derived vesicles [99].
4.2.4. Myo2p and vacuolar inheritance
During cell division, cellular organelles as well as
DNA must be equally distributed between the two
dividing cells. In yeast, the vacuole (or lysosome) is
partitioned between mother and daughter cells early
in the cell cycle via a tubular^vesicular segregation
structure [144]. Several components of the actin cy-
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
toskeleton are required for vacuolar segregation in-
cluding actin, pro¢lin, and Myo2p [145]. Several lines
of evidence suggest that Myo2p's role in vacuolar
inheritance could be to directly transport vacuoles
from mother cells to daughter cells. First, vacuoles
appear to align along actin ¢laments during segrega-
tion and some Myo2p co-localizes with these vacu-
oles [145]. Secondly, some Myo2p co-fractionates
with vacuoles and a point mutation in the tail do-
main of Myo2p disrupts this vacuole association [69].
It is also possible that Myo2p's role in vacuolar in-
heritance is to retain vacuoles in the daughter cell ^ a
mechanism analogous to the tethering model of my-
osin-Va function in melanosome tra¤cking [88].
However, as vacuolar inheritance is not an essential
process, while Myo2p is an essential gene, Myo2p
must have other cellular functions distinct from va-
cuolar inheritance.
4.2.5. Myo2p and vesicle tra¤cking
One of the most notable phenotypes of myo2-66
yeast cells is the accumulation of 50^100 nm vesicles
within the cytoplasm [16,146]. This phenotype led to
the hypothesis that Myo2p may function in vesicular
tra¤cking. However, the identity of the contents of
these vesicles remains elusive. Most markers for se-
cretion show no tra¤cking defects in myo2-66 cells
[146^148]. Yet, total secretion, as measured by cell
surface growth, is partially defective in myo2-66 cells
(T. Karpova, S. Reck-Peterson, B. Elkin, M. Moose-
ker, P. Novick and J. Cooper, manuscript submit-
ted). Perhaps the most dramatic e¡ect of myo2 mu-
tations is on the polarity of secretion, which, as
measured by total cell surface growth, is completely
depolarized in myo2 mutants (Karpova et al., manu-
script submitted).
A wealth of indirect evidence suggests that Myo2p
may transport secretory vesicles to sites of polarized
secretion. For instance, myo2 is synthetically lethal
with a number of genes whose protein products func-
tion in the post-Golgi secretory pathway in yeast
[146]. Epistasis experiments also place Myo2p func-
tion upstream of genes that function in secretory
vesicle docking and fusion [146]. Secretory vesicles
depend on Myo2p for localization, as Sec4p, a rab
GTPase that functions in exocytosis and is found on
secretory vesicles, is rapidly depolarized in myo2 mu-
tants [90,147,149]. Furthermore, secretory vesicles
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
marked by Sec4p are also rapidly depolarized, along
with Myo2p, in tropomyosin mutants, in which actin
cables are disrupted [147]. The speed and timing of
the Myo2p and Sec4p delocalization and relocaliza-
tion in these tropomyosin mutants led Pruyne et al.
[147] to conclude that Myo2p may function to di-
rectly transport Sec4p-containing vesicles to polar-
ized regions of the cell.
If Myo2p moves secretory vesicles along actin ¢l-
aments to sites of exocytosis, some aspects of Myo2p
localization should be actin-dependent. Indeed, in
dividing cells, actin ¢laments are required to main-
tain Myo2p localization at the bud tip [147]; Karpo-
va et al., manuscript submitted). However, some as-
pects of Myo2p localization appear to be actin and
motor independent. Some Myo2p can localize in la-
trunculin- (an actin depolymerizing drug) treated
cells that have been released from G0 [150]. Addi-
tionally, the tail domain of Myo2p alone is capable
of localization, showing that a tail binding site exists
at the bud tip [90]. Thus, actin ¢laments appear to be
required to maintain, but may not be required to
establish bud tip localization.
While all of the available data are consistent with
the hypothesis that Myo2p may indeed be an actin-
based motor that moves secretory vesicles to polar-
ized regions of the cell, direct proof is lacking. It
remains a possibility that Myo2p's key function is
to establish the polarized site needed for a number
of cellular processes, including secretion. Myo2p
could also function to tether or trap secretory
vesicles at polarized regions of the cell, again analo-
gous to the presumed role of Myosin-Va in melano-
some transport. The answers to several questions
must be addressed before it can be determined if
Myo2p is indeed the motor responsible for secretory
vesicle movement. For instance, does Myo2p reside
on secretory vesicles? Do GFP-labeled secretory
vesicles move in myo2 mutants?
4.3. Class V myosins and the microtubule cytoskeleton
A number of lines of evidence suggest that class V
myosins have close ties to the microtubule cytoskel-
eton. The myo2-66 mutation is suppressed by over-
expression of SMY1, a non-essential kinesin-related
protein [151]. However, Smy1p can function as a
myo2-66 suppressor in the absence of microtubules
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
47
and Smy1p bearing a mutation that should destroy
its ability to function as a microtubule motor still
suppresses myo2-66 [152]. Thus, the nature of the
Myo2p^Smy1p interaction appears to be microtubule
independent. Myo2p and Smy1p have also been
shown to physically interact [94]. This interaction is
mediated by the globular tail domain of Myo2p and
is necessary for Smy1p localization and for suppres-
sion of myo2-66 by SMY1 (K. Beningo, S. Lillie and
S. Brown, unpublished data). Thus, the cellular func-
tion of Smy1p and the Smy1p^Myo2p interaction
remains elusive [151].
In vertebrates, a number of interactions between
myosin-V and the microtubule cytoskeleton have
been demonstrated. In neurons, myosin-Va co-local-
izes with both the microtubule and actin cytoskeleton
[140]. Myosin-Va localization in the centrosome in
both interphase and dividing cells has been reported
for a number of di¡erent cell types [86,87]. Recently
the AF-6/canoe domain of the myosin-Va tail has
been shown to physically interact with kinesin [93].
The two proteins co-immunoprecipitate from mouse
brain extracts and partially co-localize in melano-
cytes [93]. As already mentioned, myosin-V also
shares a light chain with dynein [37]. What is the
functional signi¢cance of these myosin-V/microtu-
bule cytoskeleton interactions? One obvious possibil-
ity is that myosin-V is a passenger on vesicles trans-
ported by the microtubule cytoskeleton until vesicles
reach the actin rich cortex, where myosin-V could
then function to transport or tether vesicles. Indeed,
ER vesicles derived from squid axoplasm support
both microtubule and actin-based movement [9].
5. Future perspectives
There are a number of unanswered questions in
the myosin-V ¢eld and many exciting experiments
on the horizon. Some of these include: mechano-
chemical properties and motor domain structure;
functions of multiple myosins V within a single
cell; and identi¢cation of myosin-V cargo.
5.1. Mechanochemical properties and motor domain
structure
A major limitation to the detailed biochemical,
48 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
biophysical and structural analysis of myosin-V (as
well as most other unconventional myosins) is that
expression levels are too low to allow for tissue pu-
ri¢cation of su¤cient quantities for such studies. Sev-
eral laboratories have now been able to produce
functional myosin-V motor and neck domain with
its associated light chains using the baculovirus-ex-
pression system [96,97], although as noted above,
these preparations thus far lack calcium sensitivity.
This approach should provide su¤cient protein for
both high resolution structural studies as well as de-
termination of the kinetic parameters of the myosin-
V catalytic cycle. Such studies, thus far only done for
class I and II myosins [153^155] are essential in order
to understand the molecular bases of myosin-V pro-
cessivity, high a¤nity actin binding and large step
size. The recent ¢nding that myosin-Va is the ¢rst
example of a processive actin-based motor demon-
strates that di¡erent myosin classes have evolved to
have very di¡erent mechanochemical properties.
How well will these properties be conserved within
the class V myosins? Will all class V myosins be
processive motors?
5.2. Functions of multiple myosins V within a single
cell
As vertebrates express three di¡erent class V
myosins, one area of future research will be to
determine the function of each. Are the functions
of myosin-Va, b and c overlapping? Do shared
mechanisms for regulation and cargo association ex-
ist? In yeast, the data for MYO2 and MYO4 suggest
that two class V myosins can perform very di¡erent
functions.
5.3. Identi¢cation of myosin-V cargo
Perhaps the most important line of investigation to
understand the function of class V myosins is to
identify their cargo and the molecular basis of car-
go^motor interaction. Biochemical puri¢cation of
cargo-associated myosin-V combined with genetic
approaches should yield this information. Given
that a single myosin-V may perform a number of
functions in the cell, it will be particularly interesting
to determine how binding to di¡erent cargo mole-
cules is regulated. For instance, in yeast, Myo2p as-
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
sociates with the vacuole. But, as vacuolar inheri-
tance is not essential Myo2p must bind other cargo
molecules as well. What are these? And, how is bind-
ing to each regulated?
6. Note added in proof
While this manuscript was in proof several impor-
tant myosin V papers were published. We refer the
reader to these references: P.C. Bridgman, Myosin
Va movements in normal and dilute-lethal axons pro-
vide support for a dual ¢lament motor complex, J.
Cell Biol. 146 (1999) 1045^1060; K.M. Trybus, E.
Krementsova, Y. Freyzon, Kinetic characterization
of a monomeric unconventional myosin V construct,
J. Biol. Chem. 274 (1999) 27448^27456; E.M. De La
Cruz, A.L. Wells, S.S. Rosenfeld, E.M. Ostap, H.L.
Sweeney, The kinetic mechanism of myosin V, Proc.
Natl. Acad. Sci. USA 96 (1999) 13726^13731; D.
Schott, J. Ho, D. Pruyne, A. Bretscher, The
COOH-terminal domain of Myo2p, a yeast myosin
V, has a direct role in secretory vesicle targeting,
J. Cell Biol. 147 (1999) 791^808; K.E. Miller, M.P.
Sheetz, Characterization of myosin V binding
to brain vesicles, J. Biol. Chem. 275 (2000) 2598^
2606.
Acknowledgements
We wish to thank Valerie Mermall and Rania
Zaarour for critical comments on the manuscript.
Work discussed from the authors laboratories was
supported by the American Heart Association and
Grant 9874907 from the National Science Founda-
tion (J.A.M.) and NIH DK 25387 (M.S.M.).
References
[1] V. Mermall, P.L. Post, M.S. Mooseker, Science 279 (1998)
527^533.
[2] F.J. Probst, R.A. Fridell, Y. Raphael, T.L. Saunders, A.
Wang, Y. Liang, R.J. Morell, J.W. Touchman, R.H. Lyons,
K. Noben-Trauth, T.B. Friedman, S.A. Camper, Science 280
(1998) 1444^1447.
[3] A. Wang, Y. Liang, R.A. Fridell, F.J. Probst, E.R. Wilcox,
J.W. Touchman, C.C. Morton, R.J. Morell, K. Noben-
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
Trauth, S.A. Camper, T.B. Friedman, Science 280 (1998)
1447^1451.
[4] J.P. Baker, M.A. Titus, Curr. Opin. Cell Biol. 10 (1998) 80^
86.
[5] M.A. Hoyt, A.A. Hyman, M. Bahler, Proc. Natl. Acad. Sci.
USA 94 (1997) 12747^12748.
[6] T. Hasson, M.S. Mooseker, Curr. Opin. Neurobiol. 7 (1997)
615^623.
[7] M.J.T. Cope, J. Whisstock, I. Rayment, J. Kendrick-Jones,
Structure 4 (1996) 969^987.
[8] M.S. Mooseker, R.E. Cheney, Annu. Rev. Cell Dev. Biol. 11
(1995) 633^675.
[9] G.M. Langford, B.J. Molyneaux, Brain Res. Rev. 28 (1998)
1^8.
[10] M.A. Titus, Curr. Biol. 7 (1997) R301^R304.
[11] R.E. Larson, Braz. J. Med. Biol. Res. 29 (1996) 309^318.
[12] R.E. Larson, D.E. Pitta, J.A. Ferro, Braz. J. Med. Biol. Res.
21 (1988) 213^217.
[13] R.E. Larson, F.S. Espindola, E.M. Esprea¢co, J. Neuro-
chem. 54 (1990) 1288^1294.
[14] F.S. Espindola, E.M. Esprea¢co, M.V. Coelho, A.R. Mar-
tins, F.R. Costa, M.S. Mooseker, R.E. Larson, J. Cell Biol.
118 (1992) 359^368.
[15] J.A. Mercer, P.K. Seperack, M.C. Strobel, N.G. Copeland,
N.A. Jenkins, Nature 349 (1991) 709^713.
[16] G.C. Johnston, J.A. Prendergast, R.A. Singer, J. Cell Biol.
113 (1991) 539^551.
[17] E.M. Esprea¢co, R.E. Cheney, M. Matteoli, A.A. Nasci-
mento, P.V. De Camilli, R.E. Larson, M.S. Mooseker,
J. Cell Biol. 119 (1992) 1541^1557.
[18] G. Sanders, B. Lichte, H.E. Meyer, M.W. Kilimann, FEBS
Lett. 311 (1992) 295^298.
[19] R.E. Cheney, M.S. Mooseker, Curr. Opin. Cell Biol. 4
(1992) 27^35.
[20] E. Pastural, F.J. Barrat, R. Dufourcq-Lagelouse, S. Certain,
O. Sanal, N. Jabado, R. Seger, C. Griscelli, A. Fischer, G. de
Saint Basile, Nat. Genet. 16 (1997) 289^292.
[21] C. Griscelli, A. Durandy, D. Guy-Grand, F. Daguillard, C.
Herzog, M. Prunieras, Am. J. Med. 65 (1978) 691^702.
[22] H. Hurvitz, R. Gillis, S. Klaus, A. Klar, F. Gross-Kiesel-
stein, E. Okon, Eur. J. Pediatr. 152 (1993) 402^405.
[23] A. Haraldsson, C.M. Weemaes, J.A. Bakkeren, R. Happle,
Eur. J. Pediatr. 150 (1991) 419^422.
[24] W.M. Bement, T. Hasson, J.A. Wirth, R.E. Cheney, M.S.
Mooseker, Proc. Natl. Acad. Sci. USA 91 (1994) 6549^
6553.
[25] T. Hasson, J.F. Skowron, D.J. Gilbert, K.B. Avraham, W.L.
Perry, W.M. Bement, B.L. Anderson, E.H. Sherr, Z.Y.
Chen, L.A. Greene, D.C. Ward, D.P. Corey, M.S. Moose-
ker, N.G. Copeland, N.A. Jenkins, Genomics 36 (1996) 431^
439.
[26] L.P. Zhao, J.S. Koslovsky, J. Reinhard, M. Bahler, A.E.
Witt, D.W. Provance Jr., J.A. Mercer, Proc. Natl. Acad.
Sci. USA 93 (1996) 10826^10831.
[27] W.M. Huang, L. Reed-Fourquet, E. Wu, J.Y. Wu, Proc.
Natl. Acad. Sci. USA 87 (1990) 8491^8495.
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
49
[28] O.C. Rodriguez, R.E. Cheney, Mol. Viol. Cell 9, (1998) 20a
(abstract).
[29] M.D. Peterson, A.S. Urioste, M.A. Titus, J. Muscle Res.
Cell Motil. 17 (1996) 411^424.
[30] R.E. Cheney, M.K. O'Shea, J.E. Heuser, M.V. Coelho, J.S.
Wolenski, E.M. Esprea¢co, P. Forscher, R.E. Larson, M.S.
Mooseker, Cell 75 (1993) 13^23.
[31] S.H. Lillie, S.S. Brown, J. Cell Biol. 125 (1994) 825^842.
[32] J.D. Huang, M.J. Cope, V. Mermall, M.C. Strobel, J. Ken-
drick-Jones, L.B. Russell, M.S. Mooseker, N.G. Copeland,
N.A. Jenkins, Genetics 148 (1998) 1951^1961.
[33] V. Mermall, J.D. Huang, N.G. Copeland, N.A. Jenkins,
M.S. Mooseker, Mol. Biol. Cell 7 (1996) 372a (abstract).
[34] H.V. Goodson, H.M. Warrick, J.A. Spudich, J. Mol. Biol.
287 (1999) 173^185.
[35] I. Rayment, W.R. Rypniewski, K. Schmidt-Base, R. Smith,
D.R. Tomchick, M.M. Benning, D.A. Winkelmann, G. We-
senberg, H.M. Holden, Science 261 (1993) 50^58.
[36] A.R. Rhoads, F. Friedberg, FASEB J. 11 (1997) 331^340.
[37] F.S. Espindola, R.E. Cheney, S.M. King, D.M. Suter, M.S.
Mooseker, Mol. Biol. Cell 7 (1996) 372 (abstract).
[38] Y. Nabeshima, Y. Nabeshima, Y. Nonomura, Y. Fujii-Kur-
iyama, J. Biol. Chem. 262 (1987) 10608^10612.
[39] M. Kawashima, Y. Nabeshima, T. Obinata, Y. Fujii-Kur-
iyama, J. Biol. Chem. 262 (1987) 14408^14414.
[40] Y. Nabeshima, Y. Nabeshima, M. Kawashima, S. Naka-
mura, Y. Nonomura, Y. Fujii-Kuriyama, J. Mol. Biol. 204
(1988) 497^505.
[41] R. Dominguez, Y. Freyzon, K.M. Trybus, C. Cohen, Cell 94
(1998) 559^571.
[42] S.E. Brockerho¡, R.C. Stevens, T.N. Davis, J. Cell Biol. 124
(1994) 315^323.
[43] R.C. Stevens, T.N. Davis, J. Cell Biol. 142 (1998) 711^722.
[44] T.Q. Uyeda, P.D. Abramson, J.A. Spudich, Proc. Natl.
Acad. Sci. USA 93 (1996) 4459^4464.
[45] J. Howard, J.A. Spudich, Proc. Natl. Acad. Sci. USA 93
(1996) 4462^4464.
[46] A.D. Mehta, R.S. Rock, M. Rief, S.A. Spudich, M.S.
Mooseker, R.E. Cheney, Nature 400 (1999) 590^593.
[47] B.K. Haarer, A. Petzold, S.H. Lillie, S.S. Brown, J. Cell Sci.
107 (1994) 1055^1064.
[48] F.Z. Watts, G. Shiels, E. Orr, EMBO J. 6 (1987) 3499^3505.
[49] E. Bi, P. Maddox, D.J. Lew, E.D. Salmon, J.N. McMillan,
E. Yeh, J.R. Pringle, J. Cell Biol. 142 (1998) 1301^1312.
[50] A.A.C. Nascimento, R.E. Cheney, S.B.F. Tauhata, R.E.
Larson, M.S. Mooseker, J. Biol. Chem. 271 (1996) 17561^
17569.
[51] S.B.F. Tauhata, D. Vital dos Santos, R.E. Larson, M.S.
Mooseker, Mol. Cell Biol. 9 (1998) 19a (abstract).
[52] M.S. Mooseker, J.S. Wolenski, T.R. Coleman, S.M. Hay-
den, R.C. Cheney, E. Esprea¢co, M.B. Heintzelman, M.D.
Peterson, in: M.S. Mooseker, J.S. Morrow (Eds.), Current
Topics in Membranes: Ordering the Membrane Cytoskleton
Trilayer, Vol. 38, Academic, Orlando, FL, 1991, pp. 31^55.
[53] K. Collins, J.R. Sellers, P. Matsudaira, J. Cell Biol. 110
(1990) 1137^1147.
50 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51
[54] J.S. Wolenski, S.M. Hayden, P. Forscher, M.S. Mooseker,
J. Cell Biol. 122 (1993) 613^621.
[55] J.S. Wolenski, R.E. Cheney, M.S. Mooseker, P. Forscher,
J. Cell Sci. 108 (1995) 1489^1496.
[56] J.R. Geiser, D. van Tuinen, S.E. Brockerho¡, M.M. Ne¡,
T.N. Davis, Cell 65 (1991) 949^959.
[57] R. Prasad, Y. Gu, H. Alder, T. Nakamura, O. Canaani, H.
Saito, K. Huebner, R.P. Gale, P.C. Nowell, K. Kuriyama et
al., Cancer Res. 53 (1993) 5624^5628.
[58] L. Van Aelst, M.A. White, M.H. Wigler, Cold Spring Harb.
Symp. Quant. Biol. 59 (1994) 181^186.
[59] H. Miyamoto, I. Nihonmatsu, S. Kondo, R. Ueda, S. Toga-
shi, K. Hirata, Y. Ikegami, D. Yamamoto, Genes Dev. 9
(1995) 612^625.
[60] T. Matsuo, K. Takahashi, S. Kondo, K. Kaibuchi, D. Ya-
mamoto, Development 124 (1997) 2671^2680.
[61] Y. Watari, K. Kariya, M. Shibatohge, Y. Liao, C.D. Hu, M.
Goshima, M. Tamada, A. Kikuchi, T. Kataoka, Gene 224
(1998) 53^58.
[62] C.P. Ponting, Trends Biochem. Sci. 20 (1995) 265^266.
[63] K. Mandai, H. Nakanishi, A. Satoh, H. Obaishi, M. Wada,
H. Nishioka, M. Itoh, A. Mizoguchi, T. Aoki, T. Fujimoto,
Y. Matsuda, S. Tsukita, Y. Takai, J. Cell Biol. 139 (1997)
517^528.
[64] T. Yamamoto, N. Harada, K. Kano, S. Taya, E. Canaani,
Y. Matsuura, A. Mizoguchi, C. Ide, K. Kaibuchi, J. Cell
Biol. 139 (1997) 785^795.
[65] A.B. Zhadanov, D.W. Provance, Jr., C.A. Speer, D. Goss,
J.A. Blixt, J.D. Co¤n, C.M. Reichert, J.A. Mercer, Curr.
Biol., in press.
[66] S. Rogers, R. Wells, M. Rechsteiner, Science 234 (1986) 364^
368.
[67] M. Rechsteiner, S.W. Rogers, Trends Biochem. Sci. 21
(1996) 267^271.
[68] J.D. Huang, V. Mermall, M.C. Strobel, L.B. Russell, M.S.
Mooseker, N.G. Copeland, N.A. Jenkins, Genetics 148
(1998) 1963^1972.
[69] N.L. Catlett, L.S. Weisman, Proc. Natl. Acad. Sci. USA 95
(1998) 14799^14804.
[70] S.E. Benashski, A. Harrison, R.S. Patel-King, S.M. King,
J. Biol. Chem. 272 (1997) 20929^20935.
[71] S.M. King, R.S. Patel-King, J. Biol. Chem. 270 (1995)
11445^11452.
[72] R. Phillis, D. Statton, P. Caruccio, R.K. Murphey, Develop-
ment 122 (1996) 2955^2963.
[73] T. Dick, K. Ray, H.K. Salz, W. Chia, Mol. Cell Biol. 16
(1996) 1966^1977.
[74] T. Dick, U. Surana, W. Chia, Mol. Gen. Genet. 251 (1996)
38^43.
[75] K.F. Ho¡mann, M. Strand, J. Biol. Chem. 271 (1996)
26117^26123.
[76] S.M. Beckwith, C.H. Roghi, B. Liu, N. Ronald Morris,
J. Cell Biol. 143 (1998) 1239^1247.
[77] G.J. Pazour, C.G. Wilkerson, G.B. Witman, J. Cell Biol. 141
(1998) 979^992.
[78] S.R. Ja¡rey, S.H. Snyder, Science 274 (1996) 774^777.
BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart
[79] P. Crepieux, H. Kwon, N. Leclerc, W. Spencer, S. Richard,
R. Lin, J. Hiscott, Mol. Cell. Biol. 17 (1997) 7375^7385.
[80] R. Prekeris, D.M. Terrian, J. Cell Biol. 137 (1997) 1589^
1601.
[81] M.C. Costa, F. Mani, W. Santoro Jr., E.M. Esprea¢co,
R.E. Larson, J. Biol. Chem. 274 (1999) 15811^15819.
[82] L.L. Evans, A.J. Lee, P.C. Bridgman, M.S. Mooseker,
J. Cell Sci. 111 (1998) 2055^2066.
[83] O. Cremona, P. De Camilli, Curr. Opin. Neurobiol. 7
(1997) 323^330.
[84] A.A. Nascimento, R.G. Amaral, J.C. Bizario, R.E. Larson,
E.M. Esprea¢co, Mol. Biol. Cell 8 (1997) 1971^1988.
[85] X. Wu, B. Bowers, Q. Wei, B. Kocher, J.A. Hammer 3rd.,
J. Cell Sci. 110 (1997) 847^859.
[86] E.M. Esprea¢co, D.E. Coling, V. Tsakraklides, K. Krogh,
J.S. Wolenski, G. Kalinec, B. Kachar, Proc. Natl. Acad.
Sci. USA 95 (1998) 8636^8641.
[87] X. Wu, B. Kocher, Q. Wei, J.A. Hammer 3rd., Cell Motil.
Cytoskel. 40 (1998) 286^303.
[88] X. Wu, B. Bowers, K. Rao, Q. Wei, J.A.R. Hammer, J.
Cell Biol. 143 (1998) 1899^1918.
[89] V. Tsakraklides, K. Krogh, L. Wang, J.C. Bizario, R.E.
Larson, E.M. Esprea¢co, J.S. Wolenski, J. Cell Sci. 112
(1999) 2853^2865.
[90] S.L. Reck-Peterson, P.J. Novick, M.S. Mooseker, Mol.
Biol. Cell 10 (1999) 1001^1017.
[91] N.G. Robinson, L. Guo, J. Imai, E.A. Toh, Y. Matsui, F.
Tamanoi, Mol. Cell Biol. 19 (1999) 3580^3587.
[92] A.E. El-Husseini, S.R. Vincent, J. Biol. Chem. 274 (1999)
19771^19777.
[93] J.D. Huang, S.T. Brady, B.W. Richards, D. Stenolen, J.H.
Resau, N.G. Copeland, N.A. Jenkins, Nature 397 (1999)
267^270.
[94] K.A. Beningo, S.H. Lillie, S.S. Brown, Mol. Biol. Cell 8
(1997) 372a (abstract).
[95] R.E. Cheney, Methods Enzymol. 298 (1998) 3^18.
[96] F. Wang, E.V. Harvey, J.A. Hammer III, J.R. Sellers, Mol.
Cell. Biol. 9 (1998) 19a (107).
[97] E.M. De La Cruz, A.L. Wells, E.M. Ostap, H.L. Sweeney,
Mol. Cell. Biol., 9 (1998) 19a (108).
[98] J.M. Chalovich, P.D. Chantler, A.G. Szent-Gyorgyi, E. Ei-
senberg, J. Biol. Chem. 259 (1984) 2617^2621.
[99] J.S. Tabb, B.J. Molyneaux, D.L. Cohen, S.A. Kuznetsov,
G.M. Langford, J. Cell Sci. 111 (1998) 3221^3434.
[100] J. Howard, A.J. Hudspeth, R.D. Vale, Nature 342 (1989)
154^158.
[101] W.O. Hancock, J. Howard, J. Cell Biol. 140 (1998) 1395^
1405.
[102] J. Howard, Nature 389 (1997) 561^567.
[103] A.D. Mehta, J.T. Finer, J.A. Spudich, Methods Enzymol.
298 (1998) 436^459.
[104] K. Svoboda, S.M. Block, Cell 77 (1994) 773^784.
[105] M.D. Wang, M.J. Schnitzer, H. Yin, R. Landick, J. Gelles,
S.M. Block, Science 282 (1998) 902^907.
[106] E. Meyhofer, J. Howard, Proc. Natl. Acad. Sci. USA 92
(1995) 574^578.
 S.L. Reck-Peterson et al. / Biochimica et Biophysica Acta 1496 (2000) 36^51 51

[107] C.M. Coppin, D.W. Pierce, L. Hsu, R.D. Vale, Proc. Natl. Messiaen, Biochem. Biophys. Res. Commun. 252 (1998)
Acad. Sci. USA 94 (1997) 8539^8544. 329^333.
[108] H. Kojima, E. Muto, H. Higuchi, T. Yanagida, Biophys. J. [133] K.J. Moore, D.A. Swing, E.M. Rinchik, M.L. Mucenski,
73 (1997) 2012^2022. A.M. Buchberg, N.G. Copeland, N.A. Jenkins, Genetics
[109] A.D. Mehta, J.A. Spudich, Adv. Struct. Biol. 5 (1999) 229^ 119 (1988) 933^941.
270. [134] P. Lane, J. Womack, J. Hered. 70 (1979) 133^135.
[110] J.E. Haber, Annu. Rev. Genet. 32 (1998) 561^599. [135] J.M. Murray, Am. Nat. 67 (1933) 278^283.
[111] A. Sil, I. Herskowitz, Cell 84 (1996) 711^722. [136] C. Klein, N. Philippe, F. Le Deist, S. Fraitag, C. Prost, A.
[112] N. Bobola, R.P. Jansen, T.H. Shin, K. Nasmyth, Cell 84 Durandy, A. Fischer, C. Griscelli, J. Pediatr. 125 (1994)
(1996) 699^709. 886^895.
[113] R.P. Jansen, C. Dowzer, C. Michaelis, M. Galova, K. Nas- [137] E.K. Haddad, X. Wu, P.A. Henkart, J.A. Hammer III,
myth, Cell 84 (1996) 687^697. Mol. Cell. Biol. 9 (1998) 20a (abstract).
[114] R.M. Long, R.H. Singer, X. Meng, I. Gonzalez, K. Nas- [138] A.G. Searle, Heredity 6 (1952) 395^401.
myth, R.P. Jansen, Science 277 (1997) 383^387. [139] F.S. Wang, J.S. Wolenski, R.E. Cheney, M.S. Mooseker,
[115] P.A. Takizawa, A. Sil, J.R. Swedlow, I. Herskowitz, R.D. D.G. Jay, Science 273 (1996) 660^663.
Vale, Nature 389 (1997) 90^93. [140] L.L. Evans, J. Hammer, P.C. Bridgman, J. Cell Sci. 110
[116] E. Bertrand, P. Chartrand, M. Schaefer, S.M. Shenoy, R.H. (1997) 439^449.
Singer, R.M. Long, Mol. Cell 2 (1998) 437^445. [141] K. Dekker-Ohno, S. Hayasaka, Y. Takagishi, S. Oda, N.
[117] D.L. Beach, E.D. Salmon, K. Bloom, Curr. Biol. 9 (1999) Wakasugi, K. Mikoshiba, M. Inouye, H. Yamamura, Brain
569^578. Res. 714 (1996) 226^230.
[118] S. Munchow, C. Sauter, R.P. Jansen, J. Cell Sci. 112 (1999) [142] Y. Takagishi, S. Oda, S. Hayasaka, K. Dekker-Ohno, T.
1511^1518. Shikata, M. Inouye, H. Yamamura, Neurosci. Lett. 215
[119] B. Wendland, J.M. McCa¡ery, Q. Xiao, S.D. Emr, J. Cell (1996) 169^172.
Biol. 135 (1996) 1485^1500. [143] V.A. Street, M.M. Bosma, V.P. Demas, M.R. Regan, D.D.
[120] D.C. Amberg, J.E. Zahner, J.W. Mulholland, J.R. Pringle, Lin, L.C. Robinson, W.S. Agnew, B.L. Tempel, J. Neuro-
D. Botstein, Mol. Biol. Cell 8 (1997) 729^753. sci. 17 (1997) 635^645.
[121] I. Gonzalez, S.B. Buonomo, K. Nasmyth, U. von Ahsen, [144] L.S. Weisman, W. Wickner, Science 241 (1988) 589^591.
Curr. Biol. 9 (1999) 337^340. [145] K.L. Hill, N.L. Catlett, L.S. Weisman, J. Cell Biol. 135
[122] D.W. Provance Jr., M. Wei, V. Ipe, J.A. Mercer, Proc. (1996) 1535^1549.
Natl. Acad. Sci. USA 93 (1996) 14554^14558. [146] B. Govindan, R. Bowser, P. Novick, J. Cell Biol. 128 (1995)
[123] Q. Wei, X. Wu, J.A. Hammer III, J. Muscle Res. Cell 1055^1068.
Motil. 18 (1997) 517^527. [147] D.W. Pruyne, D.H. Schott, A. Bretscher, J. Cell Biol. 143
[124] Y.I. Koyama, T. Takeuchi, Anat. Rec. 196 (1980) 449^459. (1998) 1931^1945.
[125] V.I. Rodionov, F.K. Gyoeva, V.I. Gelfand, Proc. Natl. [148] H. Liu, A. Bretscher, J. Cell Biol. 118 (1992) 285^299.
Acad. Sci. USA 88 (1991) 4956^4960. [149] C. Walch-Solimena, R.N. Collins, P.J. Novick, J. Cell Biol.
[126] M.C. Tuma, A. Zill, N. Le Bot, I. Vernos, V. Gelfand, 137 (1997) 1495^1509.
J. Cell Biol. 143 (1998) 1547^1558. [150] K.R. Ayscough, J. Stryker, N. Pokala, M. Sanders, P.
[127] L.H. Haimo, C.D. Thaler, BioEssays 16 (1994) 727^732. Crews, D.G. Drubin, J. Cell Biol. 137 (1997) 399^416.
[128] S.L. Rogers, V.I. Gelfand, Curr. Biol. 8 (1998) 161^164. [151] S.H. Lillie, S.S. Brown, Nature 356 (1992) 358^361.
[129] V.I. Rodionov, A.J. Hope, T.M. Svitkina, G.G. Borisy, [152] S.H. Lillie, S.S. Brown, J. Cell Biol. 140 (1998) 873^883.
Curr. Biol. 8 (1998) 165^168. [153] S.B. Marston, E.W. Taylor, J. Mol. Biol. 139 (1980) 573^
[130] S.L. Rogers, R.L. Karcher, J.T. Roland, A.A. Minin, W. 600.
Ste¡en, V.I. Gelfand, J. Cell Biol. 146 (1999) 1265^1276. [154] E.M. Ostap, T.D. Pollard, J. Cell Biol. 132 (1996) 1053^
[131] P.K. Seperack, J.A. Mercer, M.C. Strobel, N.G. Copeland, 1060.
N.A. Jenkins, EMBO J. 14 (1995) 2326^2332. [155] J.D. Jontes, R.A. Milligan, T.D. Pollard, E.M. Ostap, Proc.
[132] J. Lambert, J.M. Naeyaert, T. Callens, A. De Paepe, L. Natl. Acad. Sci. USA 94 (1997) 14332^14337.

BBAMCR 14597 6-3-00 Cyaan Magenta Geel Zwart