Cell Motility and the Cytoskeleton 15:210-219 (1990)
Nucleotide Specificity for the Bidirectional
Transport of Membrane-Bounded
Organelles in Isolated Axoplasm
Philip L. Leopold, Robert Snyder, George S. Bloom, and Scott T. Brady
Deparfmenf of Cell Biology and Neuroscience, University of Texas
Southwestern Medical Center, Dallas (P.L.L., G.S.B., S. T.B); Marine Biological
Laboratory, Woods Hole, Massachusetts (P.L.L., R.S., G.S.B., S.T.B.)
Video microscopy of isolated axoplasm from the squid giant axon permits correlated quantitative analyses of membrane-bounded organelle transport both in the
intact axoplasm and along individual microtubules. As a result, the effects of
experimental manipulations on both anterograde and retrograde movements of
membrane-bounded organelles can be evaluated under nearly physiological conditions. Since anterograde and retrograde fast axonal transport are similar but
distinct cellular processes, a systematic biochemical analysis is important for a
further understanding of the molecular mechanisms for each. In this series of
experiments, we employed isolated axoplasm of the squid to define the nucleoside
triphosphate specificity for bidirectional organelle motility in the axon. Perfusion
of axoplasm with 2-20 mM ATP preserved optimal vesicle velocities in both the
anterograde and retrograde directions. Organelle velocities decreased to <50% of
optimal values when the axoplasm was perfused with 10-20 mM UTP, GTP, ITP,
or CTP with simultaneous depletion of endogenous ATP with hexokinase. Under
the same conditions, TTP and ATP-y-S were unable to support significant levels
of transport. None of the NTPs tested had a differential effect on anterograde vs.
retrograde movement of vesicles. Surprisingly, several inconsistencies were revealed when a comparison was made between these results and nucleoside
triphosphate specificities that have been reported for putative organelle motors by
using in vitro assays. These data may be used in conjunction with data from
well-defined in vitro assays to develop models for the molecular mechanisms of
axonal transport.
Key words: organelle motors, nucleoside triphosphates, fast axonal transport, video microscopy
INTRODUCTION
Squid axoplasm has been established as an effective paradigm in the study of organelle transport [Brady
et al., 19821 and, in particular, the ATP dependence of
the process [Brady et al., 1982, 19851. When the axoplasm is perfused with a variety of ATP analogs and
inhibitors of ATP hydrolysis, a number of distinct alterations in organelle motility are observed [Brady et al.,
19851. For example, one ATP analog, adenylylimidodiphosphate (AMP-PNP), causes organelles to “freeze’ ’
along microtubules [Lasek and Brady, 1985; Brady et
al., 19851. This observation led to the discovery of kinesin, a microtubule-stimulated ATPase which has been
0 1990 Wiley-Liss, Inc.
proposed as the organelle motor in the axoplasm [Brady,
1985; Vale et al., 1985a1. This example illustrates the
utility of using observations of organelle motility in an
intact axoplasm to elucidate the molecular mechanisms
of motility. Any analysis of dynamic intracellular processes must be tempered with an appreciation for the
complexity of the cell and, therefore, has limited bio-
Received November 20, 1989; accepted November 29. 1989.
Address reprint requests to Dr. Scott T. Brady, Dept. of Cell Biology
and Anatomy, University of Texas Southwestern Medical Center at
Dallas, 5323 Harry Hines Blvd., Dallas, TX 75235.
Membrane Bounded Organelle Transport
chemical resolution. However, by studying a process
such as organelle transport in an intact cytoplasm, characteristics of organelle motility may be described and
inferences concerning transport motors that are not
readily accessible by other means may be developed.
Given the importance of phosphodiester bond hydrolysis in intracellular motility, characterization of the
nucleoside triphosphate (NTP) specificity of organelle
transport should lead to a further understanding of molecular mechanisms of force production. Often the spectrum of NTP use differs among ATPases, and such information may be used to distinguish functionally related
enzymes. For example, Rozdzial and Haimo [ 1986a1
showed that independent mechanisms account for aggregation and dispersion of pigment granules in melanophores based on the fact that ADP supports only aggregation of granules, while ATP-7-S supports only
dispersion. Similarly, one of the most outstanding differences between cytoplasmic and axonemal dyneins is
found in the rates of hydrolysis of various NTPs [Shpetner et al., 1988; Shimizu, 19871.
The presence of endogenous ATP in the axoplasm
had complicated earlier studies of NTP utilization in axonal transport. Treatments with apyrase [Brady et al.,
19851 and extractions with ATP-free buffer [Adams,
1982; Forman et al., 1983a,b] have been used to deplete
ATP from axoplasm prior to addition of ATP or alternate
substrates. However, these methods have resulted in incomplete reactivation of motility. A procedure has now
been developed for simultaneous elimination of endogenous ATP and introduction of alternative substrates. By
using this procedure, a variety of NTPs were tested for
their ability to support anterograde and retrograde organelle movement in the axoplasm. By defining the biochemical requirements for axonal transport, expectations
regarding the properties and identities of organelle motility motors can be developed. Such studies also provide
a basis for evaluating the utility of in vitro models in
replicating fast axonal transport.
MATERIALS AND METHODS
Axoplasm Extrusion and Video Microscopy
Squid (Loligo pealeii) were provided by the Department of Marine Resources, Marine Biological Laboratory, Woods Hole, MA. A 2-3 cm segment of axon
(= 0.5 mm dia., = 5.0 pl) were dissected, extruded, and
perfused as described in Brady et al. [1985]. Videoenhanced contrast differential interference contrast microscopy [Allen et al., 1981; Allen and Allen, 19831 was
used to examine the squid axoplasm as previously reported [Brady et al., 19851. A video record of experiments was keDt bv using a Sonv VO-5800 video taDe
211
recorder or a Panasonic TQ-2025F optical disc recorder
and optical memory discs (Panasonic TQ-FH224).
Perfusions
Axoplasms were perfused with four volumes (=20
p1) of buffer Xi2 (half-strength buffer X from Brady et
al. [1985]): 175 mM potassium aspartate, 65 mM taurine, 35 mM betaine, 25 mM glycine, 10 mM HEPES
(N-2-Hydroxyethylpiperazine N'-2-ethanesulfonic acid),
6.3 mM MgCl,, 5 mM potassium EGTA (ethyleneglycol
bis-(2-aminoethyl ether)N, "-tetraacetic acid, potassium salt), 1.5 mM CaCl,, 0.5 mM glucose, pH adjusted
to 7.2 with KOH. Varying concentrations of ATP (vanadate free) or other NTPs were included in the buffer Xi2
perfusions. The pH of buffers was adjusted to 7.0 following NTP additions. Buffer X/2 rather than fullstrength buffer X was used to facilitate slight extraction
of the axoplasm so that individual microtubules and associated organelles could be readily observed at the periphery [Brady et al., 19851. Reagents were obtained
from Sigma Chemical Co. (St. Louis, MO) unless otherwise stated.
ATP Depletion Buffer
In experiments which required the hydrolysis of the
endogenous ATP, the perfusion buffer was supplemented with 40 unitsipl hexokinase (Sigma, # H-5500)
which transfers a phosphate group from ATP to a hexose
sugar. The glucose in buffer Xi2 was substituted with 5
mM 2-deoxyglucose, which is equally efficient as a substrate for hexokinase [Sols et al., 19581 but which cannot
produce ATP through glycolysis; 5 mM 2-deoxyglucose
was diluted by the axoplasm to 4 mM upon addition of
buffer to the perfusion chamber. Therefore, the depletion
buffer was capable of depleting 4 mM ATP in the chamber. Loligo axoplasm contains 0.1-1 .0 mM ATP [Caldwell, 19601 which is diluted to <0.2 mM ATP during
perfusion and is well within the depletion capacity of the
buffer. ATP is the phosphate donor of choice for hexokinase and the enzyme has very limited ability to utilize
other NTPs [Darrow and Colowick, 19621. As a result,
endogenous ATP was preferentially degraded while
other NTPs were perfused with the ATP depletion
buffer.
Dinitrophenol was added to block ATP production
by mitochondria and has previously been found to be an
effective inhibitor of bidirectional organelle transport in
axoplasm [Brady et al., 19821. The addition of the adenylate kinase inhibitor, diadenosine pentaphosphate
[Lienhard and Secemski, 19731, reduced the likelihood
that ADP would be directly converted to ATP by
transphosphorylation. Creatine phosphate was not likely
to contribute significantly to the ATP pool given the
observation that this compound, when injected into ATP-
212
Leopold et al.
depleted Loligo axons, was unable to restore ATP-dependent functions [Caldwell et al., 19601. In sum, the
ATP depletion system was designed to deplete endogenous ATP and to block ATP production in the axoplasm
while leaving exogenous NTPs other than ATP intact.
ATP depletion of axoplasms was accompanied by a
simultaneous perfusion with 2, 10, or 20 mM NTP or 20
mM adenosine-5’-0-(3-thiotriphosphate)(ATP-y-S, tetralithium salt, Boehringer Mannheim, Indianapolis,
IN). Depletion of ATP in the absence of any NTP causes
a rapid cessation of vesicle motility [Brady et al., 1982,
19851 that appears to be only partially reversible (data
not shown).
Velocity Determinations
Speed estimations were made with a Photonics Microscopy, Inc., video manipulator (C2117) in the speed
estimation mode. Following calibration, movement of
cursors was matched to the average velocity of organelle
movement in the anterograde or retrograde direction. Alternatively, movement of cursors was matched to the
velocity of individual organelles along isolated microtubules at the periphery of the perfused axoplasm. Anterograde and retrograde organelle velocities were recorded
in unperlused axoplasm to providc a basis for comparison following perfusion. For each condition, velocity
measurements were made in at least two axoplasm preparations and in a number of different fields.
RESULTS
Measurements of vesicle motility were made both
in the interior of the axoplasm and at the periphery. Each
type of data provided useful information, but imposed
characteristic limitations on interpretation. To compare
anterograde and retrograde motility, data were taken
from the interior of the axoplasm where the cytoskeletal
elements retain their in vivo polarized organization
[Lasek, 19841. Velocities in the interior reflected population averages due to the limits of resolution and to the
fact that organelles passed in and out of the plane of
focus. Observations in the interior of the axoplasm offered the ability to observe motility prior to perfusion.
Therefore, observations of unperfused axoplasm served
as the standard with which experimental perfusions could
be compared. At the periphery, individual microtubules
became extracted from the mass of the axoplasm [Allen
et al., 1985; Schnapp et al., 19851. Due to the fact that
long sections of individual microtubules (up to -20 p m
in length) were plainly visible, the velocities of individual organelles were easily estimated on these structures.
However, the polarity of the extracted microtubules
could no longer be determined, so the direction of vesicle
translocation could not be assigned. While extracted mi-
2.00
__
-----
20 mM
10 mM
2 mM
\
\
0.00
’
0
,‘.---,---
5
,
10
15
20
Time After PerfuSicfi (mid
25
30
Fig. 1. Time course: Organelle transport along peripheral microtubules in the presence of GTP. Extruded axoplasm was perfused with
experimental NTPs in the presence of ATP depletion buffer. Organelle
velocities were measured at various times along isolated peripheral
microtubules. The results shown here are for GTP at 2 mM, 10 mM,
and 20 mM. Velocities changed rapidly in the first 5-10 rnin following
perfusion and became stable for at least 20 min. Each curve represents
data averaged from a minimum of three experiments. Slight increases
in velocity at 10 min are artifacts of the curve-drawing program.
crotubules were not a suitable model for addressing questions about differences between anterograde and retrograde movement, these studies provided confirmation of
the relative organelle velocities determined in the interior. Moreover, they permitted evaluation of membranebounded organelle movement without the complication
of the intact axoplasmic structure, providing a bridge
between a true in vitro reconstitution and the extruded
axoplasm system.
During perfusion of alternate NTPs, endogenous
ATP was removed by including the ATP depletion
buffer. When the ATP depletion buffer system was not
supplemented with NTPs, all movement in the axoplasm
stopped within 10 min. Boiling hexokinase for 10 min
prior to perfusion alleviated the inhibition of motion with
movement continuing after 15 min.
The effect of each experimental buffer on organelle
motility was observed over the course of a 30 min incubation. Figure 1 shows that organelle velocities on peripheral microtubules changed rapidly following perfusion with the experimental buffer containing GTP before
reaching a plateau between 10 and 30 min after treatment. The plateau velocity in Figure 1 depended on the
concentration of GTP included in the perfusion buffer.
Similar results were obtained with other NTPs. The delay before the system became stable may reflect an equilibration period for NTPs, axoplasmic ATPases, and depletion of endogenous ATP. Axoplasms which were
observed for longer periods of time were found to have
slowly decreasing vesicle velocities, presumably due to
the lack of an NTP regenerating system. Therefore, in
Membrane Bounded Organelle Transport
213
TABLE I. Anterograde and Retrograde Organelle Velocities in the Axoplasmic Interior
Anterograde
Energy
source
Control‘
ATP~
ATP
+ ADBe
UTP
GTP
ITP
CTP
TTP
ATP-?I-S
Retrograde
(mM)
N“
Mean t SEb
( pmisec)
N
Mean 2 SE
(pm/sec)
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
20
117
16
6
4
3
6
6
12
8
13
12
13
12
3
7
4
2
7
9
6
13
6
4
1.85 rfr 0.03
1.79 t 0.10
1.70 t 0. I4
1.69 t 0.10
0
1.36 t 0.08
1.49 t 0.08
0.24 t 0.07
0.78 t 0.06
0.40 t 0.03
0.09 t 0.06
0.46 t 0.04
0.38 t 0.03
0.05 2 0.05
0.32 2 0.15
0.33 t 0.04
0
0.53 t 0.08
0.26 t 0.04
0.17 t 0.08
0.20 t 0.04
0.10 t 0.05
0
I18
16
6
4
3
6
6
11
7
14
8
11
5
3
6
7
2
6
7
3
13
3
4
1.34 t 0.03
1.09 t 0.0s
1.18 t 0.08
1.32 t 0.06
0
1.13 t 0.09
1.20 t 0.16
0.19 t 0.09
0.62 t 0.06
0.42 t 0.04
0.04 t 0.04
0.30 t 0.03
0.25 t 0.07
0
0.31 t 0.04
0.30 t 0.03
0
0.41 t 0.06
0.22 t 0.05
0
0.16 t 0.03
0
0
Conc.
aEach observation reflectc a population of organelles in the axoplasmic interior; for cases in which the mean velocity = 0 Fmlsec, N = number
of axoplasm preparations examined; in all other cases, observations were made on at least two axoplasm preparations.
bVelocities are reported as the mean velocity t standard error.
“Unperfused axoplasm.
dValues for ATP experiments were significantly different from values for all alternate NTPs by the Students T test ( P < 0.05).
‘ADB = ATP depletion buffer; alternate NTPs were always perfused in the presence of the ATP depletion buffer.
Anterograde and retrograde organelle velocities were recorded prior to perfusion (Table I, Control). These values confirm the differences in velocity of anterograde
and retrograde transport reported previously [see, for example, Allen et al., 19821. In order to determine the
efficacy of each NTP, anterograde and retrograde organelle velocities were compared to the control values.
Perfusion with 2 mM ATP yielded transport velocities which were nearly identical to the control values
measured in the unperfused axoplasm. Addition of the
ATP depletion buffer halted motility within 10 min in the
presence of 2 mM ATP. ATP at 10 or 20 mM supported
transport at velocities comparable to the control values.
In the presence of the depletion buffer, 10 and 20 mM
ATP perfusions also supported organelle transport but at
slightly lower velocities than for ATP alone. This decreased velocity is likely to be due to the generation of
ADP during ATP depletion (see Discussion).
NTPs other than ATP were unable to support a
Velocities in the Interior of the Axoplasm
significant amount of motility in either direction when
Observations made in the interior of the axoplasm present at 2 mM. At 10 and 20 mM, UTP, GTP, ITP,
allowed a comparison of anterograde and retrograde or- and CTP were capable of producing organelle velocities
ganelle velocities in the presence of each NTP (Table I). at approximately 15-40% of corresponding control val-
the data which follow, mean velocities reflect data taken
between 10 and 30 min following perfusion. The curves
in Figure 1 represent averages of data from either three or
four axoplasms for each condition. In order to minimize
sampling error in the construction of Figure 1 , data from
each of the experiments for a given condition were averaged over 5 min intervals and plotted. Each of the
experimental curves was averaged to give the time
course for each condition. Motility ceased between 10
and 30 min following perfusion in most experiments with
2 mM alternate NTPs. Since motility never restarted after stopping, data following the point of cessation was
taken as 0 p d s e c for purposes of weighting experiments
equally. The value on the plot at T = 0 was selected as the
mean value of anterograde and retrograde transport in
unperfused axoplasm. Smooth curves were drawn by using the cubic spline method by the Slidewrite Plus program (Advanced Graphics Inc., Sunnyvale, CA).
214
Leopold et al.
ues. Organelle velocities at 10 and 20 mM were comparable for a given NTP, suggesting that utilization of
NTPs was maximal at these concentrations. Concentrations of TTP as high as 20 mM were ineffective in supporting motility showing movement at < 12% of control
values and substantial reductions in the number of moving organelles. 20 mM ATP-y-S was unable to sustain
motility in either direction. No conditions were found
which exhibited differential effects on anterograde and
retrograde transport.
Under conditions which support the most rapid vesicle translocations (e.g., Control or ATP), anterograde
velocity is greater than retrograde velocity. However,
under conditions which reduce translocation velocities
by >50% in both directions (e.g., alternate NTPs), no
statistical distinction between the two directions could be
found.
TABLE 11. Velocities of Individual Vesicles on Peripheral
Microtubules
Energy
source
Conc .
(mM)
N"
Mean rt SEb
(Fmisec)
ATF"
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
2
10
20
20
9
7
3
16
6
11
12
11
19
4
6
11
14
19
21
12
7
7
4
4
6
4
2.21 rt 0.02
2.12 rt 0.14
2.05 0.12
0.12 rt 0.04
1.17 C 0.09
1.66 rt 0.13
0.14 C 0.05
0.72 rt 0.05
0.46 C 0.02
0
0.35 rt 0.05
0.36 i- 0.03
0.10 rt 0.04
0.33 ir 0.03
0.41 rt 0.02
0.08 rt 0.05
0.44 rt 0.08
0.29 rt 0.05
0
0.18 rt 0.07
0.09 rt 0.04
0
ATP
+ ADBd
UTP
GTP
ITP
CTP
Velocities on Extracted Microtubules
A small fraction of axoplasmic microtubules and
organelles was extracted from the mass of axoplasm during perfusion. Since the axoplasm had to be perfused
before these structures were liberated, there were no data
from unperfused axoplasms. Table I1 summarizes the
data. When ATP was present in buffer Xi2 at a concentration of 2 mM, organelle velocities on extracted microtubules were comparable to previously reported values from extruded axoplasms perfused with ATP [Brady
et al., 19851, but were higher than values from intact
squid axons [Allen et al., 19821 or unperfused axoplasm;
10 and 20 mM ATP supported rapid organelle translocations at comparable velocities. When the axoplasm
was perfused with 2 mM ATP and the ATP depletion
system, minimal movement remained on extracted microtubules after 10 min; 10 mM ATP was able to overcome partially the effects of the inhibitors although full
velocity (as compared to 10 mM ATP without the depletion buffer) was not obtained (see Discussion); 20 mM
ATP in the presence of the depletion system demonstrated that high concentrations of ATP brought organelle velocities closer to levels with ATP alone.
NTPs tested at 2 mM were not able to rescue the
transport on extracted microtubules from ATP depletion;
velocities were <7% of ATP values. When the concentrations of the other NTPs were raised to 10 mM, continuing motility was evident with several NTPs. As in the
interior, UTP, GTP, ITP, and CTP were able to support
organelle velocities on extracted microtubules at velocities 15-35% of the value for the same concentration of
ATP. TTP failed to produce significant motility at a concentration of 10 mM (<lo% of the ATP value). The
results for the alternate NTPs at 20 mM (Fig. 2) were
very similar to the results at 10 mM. UTP, GTP, ITP,
and CTP supported motility at 14-22% of 20 mM ATP
TTP
ATP-7-S
*
"Each observation reflects the velocity of a single vesicle moving
along a peripheral microtubule; for cases in which the mean velocity
= 0 pm/sec, N = number of axoplasm preparations examined; in all
other cases, observations were made on at least two axoplasm preparations.
bVelocities are reported as the mean velocity rt standard error.
'Values for ATP experiments were statistically different from results
with alternate NTPs by the Student's T test ( P < 0.05).
dADB = ATP depletion buffer; alternate NTPs were perfused in the
presence of the ATP depletion buffer.
values while perfusion with TTP failed to produce significant motility (<5% of ATP) even at 20 mM. Similarly, the ATP analog, ATP-y-S, was unable to support
motility at a concentration of 20 mM. A comparison of
the data obtained in the interior and on extracted microtubules is shown in Figure 2 and served to verify the
comparison of NTP utilization.
The putative organelle motors, kinesin and cytoplasmic dynein, have been shown to hydrolyze ATP in
conjunction with the divalent cation, Mg2+ [Brady et
al., 1985; Kuznetsov and Gelfand, 1986; Cohn et al.,
1987, 1989; Shpetner et al., 1988; Wagner et al., 19891.
In some cases, the presence of ATP in excess of the
concentration of Mg2' has been shown to inhibit the
Mg2+-ATPase activity and force production of these
molecules [Cohn et al., 1987, 19891. Accordingly, a
separate set of experiments was performed to confirm
that Mg2 from endogenous supplies and from the perfusion buffer was not a limiting factor in the availability
of hydrolytic substrate for the molecular motors. Perfusions containing 20 mM NTP and 20 mM Mg2' pro+
Membrane Bounded Organelle Transport
2.50
-
2.00
2 mM
c]
2.00
T
6.3 nM
ANiEROGRADE
(D
20 nM Ma 2+
\
2
g
EXTRACTED MTS
.oo
2+
1.50
RETROGRADE
1.50
215
1
.-
E
x
Y
0.50
1.00
.-
&d
-u
hcn
0.00
2.50
-8
T
5
10 mM
2.00
5
5
1.50
>
1.00
2W
0.50
-E
6
0.50
0.00
A RMT
ATP
>
0.00
2.50
20 mM
?
2.00
1.50
1
A RMT
UTP
A RMT
TTP
Fig. 3 . Demonstration that organelle motility following perfusion was
not due to insufficient MgZf levels. In order to eliminate the possibility that high concentrations of NTPs were inhibiting the MgNTPase
of the organelle motor(s), a set of perfusions was conducted with 20
mM Mg2+ and observations were made of anterograde (A) and retrograde (R) motility as well as of motility on extracted microtubules
(MT). The presence of 20 mM MgZi. did not change the response of
the axoplasm to the endogenous substrate (ATP) or to alternate substrates which were either capablc (UTP) ui iiicapable (TTP) of supporting motility.
.oo
1
the substrate specificity of the ATPase(s) acting as a
motor for fast axonal transport. Axonal transport has
long been recognized as an ATP-dependent process [Sa0.00
<
ATP ATP UTP GTP ITP CTP P
T
I
ATP-$-S
bri and Ochs, 1972; Adams, 19821 and, predictably,
+AD8
ATP is most efficiently utilized by the transport apparaFig. 2. Comparison of organelle velocities in the axoplasmic interior tus (Tables I and 11). In the interior, 2, 10, and 20 mM
and on extracted microtubules. Anterograde and retrograde organelle ATP supported vesicle velocities which were very simipopulation velocities were essentially unchanged before and after per- lar to transport in the unperfused axoplasm. Under the
fusion with 2, 10, and 20 mM ATP. Changes in retrograde motility same conditions, vesicle velocities in the absence of the
mirrored changes in anterograde motility following perfusion with
cytoskeletal matrix, as exemplified by movement on exalternate NTPs. Organelle velocities on extracted microtubules responded to changes in buffer conditions in a qualitatively identical tracted microtubules, were always greater than corremanner. At high velocities (e.g., where anterograde velocities were > sponding velocities at the interior of the axoplasm (Fig.
1.5 pm/sec), velocities on extracted microtubules exceeded both an- 2), suggesting that the matrix was capable of slowing
terograde and retrograde velocities. Under conditions supporting re- movement in both directions. These observations produced vesicle velocities, mean values of anterograde and retrograde
vide quantitative support for suggestions that cytoskelemotility were similar to values for motility on extracted microtubules.
tal interactions may be at least partially responsible for
the in vivo differences between the rates of anterograde
duced results which were consistent with the results us- (smaller, faster) and retrograde (larger, slower) oring the standard 6.3 mM MgZf in buffer X/2 (Fig. 3 ) . ganelles through steric hindrance or biochemical mechanisms.
Previous reports indicate that organelle motility
DISCUSSION
might be sensitive to a temporary absence of ATP resultIn order to characterize the spectrum of NTP utili- ing in decreased vesicle velocities upon reactivation [Adzation during organelle translocations in fast axonal ams, 1982; Forman et al., 1984; Brady et al., 19851.
transport, we have perfused extruded squid axoplasm Significantly, the extruded axoplasm model replicated in
with a variety of these compounds. The results address vivo organelle velocities ( z1.8 pmisec anterograde,
0.50
216
Leopold et al.
=1.1 pm/sec retrograde) after perfusion with levels of
ATP equal to or slightly greater than physiological concentrations (1-2 mM, Table I and data not shown). The
simultaneous addition of NTP and depletion of endogenous ATP may account for this result.
In the presence of the ATP depletion buffer, perfusion of 2 mM ATP was unable to support motility. This
result verified that endogenous ATP and ATP added as a
contaminant of commercial preparations of alternate
NTPs could be eliminated by the depletion system. The
depletion buffer was swamped by the addition of 10 or 20
mM ATP although with some reduction in velocity as
compared to perfusion with the same concentration of
ATP in the absence of the depletion system. The reduction
in organelle velocity was most likely due to the production
of ADP after hydrolysis of ATP by the depletion buffer.
ADP greater than 5 mM has been found to reduce organelle velocities in the axoplasm when included in the
perfusion buffer (Brady and Leopold, unpublished results). In addition, Cohn et al. [1989] reported that ADP
competitively inhibits in vitro motility driven by kinesin,
a putative axoplasmic organelle motor.
Since endogenous ATP had been removed from the
axoplasm and the major avenues of ATP production in
thc axoplasm wcre blocked, extended periods of vesicle
movement in axoplasms containing alternate NTPs were
due to the presence of those NTPs in the perfusion
buffer. In all cases, the concentration of alternate NTP
required to support motility was several orders of magnitude higher than both physiological concentrations and
the minimum concentration of ATP required to support
motility. This result presumably arose from a lower kinetic efficiency for these hydrolytic substrates as compared to the primary substrate of the motor(s), ATP; 10
or 20 mM UTP, GTP, ITP, and CTP were able to support
organelle velocities at 14-40% of the rates produced by
ATP. In most cases, differences among the effects of
these four NTPs were statistically indistinguishable although under some conditions, UTP appeared to be
slightly more effective as a substrate than the other
NTPs. Axoplasms in which vesicle velocities fell below
0.2 pm/sec also exhibited large reductions in the number
of moving organelles. Therefore, motility in these axoplasms was not considered to be representative of the
effect of the experimental condition. Quantitation of the
number of moving vesicles was impossible due to limits
of resolution. TTP was unable to support a significant
amount of motility, even at a concentration of 20 mM.
These results agree with the observations of Forman et
al. 219841 that high levels of GTP, ITP, UTP, and CTP
were capable of partially reactivating organelle transport
after detergent permeabilization.
The presence of ATP in excess over Mg2+ has
been shown to inhibit the ATPase and force-producing
activities of one putative organelle motor, kinesin [Cohn
et al., 1987, 19891. At equimolar concentrations of
Mg2+ and ATP, kinesin ATPase activity is slightly depressed [Cohn et al., 19891 while motility is unaffected
[Cohn et al., 19871. The possibility that a deficit of
Mg2 ' was responsible for inhibition of the organelle
motors at high NTP concentrations was ruled out by a set
of perfusions at equimolar NTP and Mg2+. The fact that
elevated Mg2 concentrations failed to alter the relationships between organelle velocities with various substrates demonstrated that the Mg2 concentration was
not responsible for the range of effects on motility (Fig.
3 ) . In addition, the similarity in maximal velocities for 2
mM and 20 mM ATP indicates that NTP concentration
did not adversely affect motility (Fig. 2).
Each of the alternate NTPs tested yielded anterograde and retrograde velocities which were statistically
indistinguishable from each other by the Student's T test
( P < 0.05). This result contrasted with the clear distinction between anterograde and retrograde organelle velocities in unperfused axoplasm. When ATP was perfused
into the axoplasm, rapid organelle translocations and a
distinction between anterograde and retrograde velocities
were maintained. Presumably, the differential effect of
the rriatrix on anterograde and retrograde vesicles (as
discussed above) is less influential at the lower vesicle
velocities supported by other NTPs.
Analogs of ATP are also useful as probes of ATPase activity. In some cases, analogs may functionally
replace ATP in cellular processes. ATP-y-S has been
used to distinguish the two directions of transport of
pigment granules in melanophores [Rozdzial and Haimo,
1986a,b]. In fish-scale melanophores, ATP-y-S was able
to reactivate dispersion but not aggregation of pigment
granules after depletion of ATP, suggesting that dispersion may employ an ATPase which hydrolyzes the thiophosphate analog. ATP-y-S can be hydrolyzed very
slowly by myosin [Bagshaw et al., 19721, but cannot be
hydrolyzed by axonemal dynein [Penningroth et al.,
19821. In the axoplasm, 20 mM ATP-y-S was unable to
support translocation, indicating that the organelle motor
is not able to utilize this analog. No distinction between
anterograde and retrograde transport was noted. The lack
of organelle movements in either anterograde or retrograde directions with ATP-y-S or ADP (Brady and
Leopold, unpublished results) contrasts with the movement of pigment granules in melanophores and implies
that pigment granule movements are not homologous to
fast axonal transport.
Anterograde and retrograde motility had comparable responses to the alternate NTPs. Therefore, a model
for axonal transport should include either a single bidirectional ATPase or separate anterograde and retrograde
ATPases with very similar NTP requirements. One cur+
+
Membrane Bounded Organelle Transport
rent model for organelle transport involves two putative
organelle motors, kinesin [Vale et al., 1985a; Brady,
19851 and cytoplasmic dynein [Paschal et al., 19871.
Based on unidirectional gliding of microtubules caused
by each of these proteins, kinesin has been proposed as
an anterograde motor [Vale et al., 1985b1 while cytoplasmic dynein (also called MAP 1 C) has been suggested as a retrograde motor [Paschal and Vallee, 19871.
In vitro microtubule gliding assays using kinesin
and cytoplasmic dynein demonstrate several qualitative
differences between each of the proteins and organelle
transport. Kinesin promotes gliding at velocities which
are approximately 0.5 pm/sec [Vale et al., 1985a; Porter
et al., 1987; Cohn et al., 19891 while cytoplasmic dynein
moves microtubules at approximately 1.25 prnisec [Paschal et al., 19871. In addition, the motors differ in their
NTP specificity for gliding. Kinesin was reported to induce microtubule gliding in the presence of 1-5 mM
GTP or ITP [Porter et al., 19871. Gliding velocities with
GTP were equivalent to those produced by ATP while
ITP velocities were half of that value. More recently, a
detailed report of kinesin motility in the presence of 10
mM NTPs has been published [Cohn et al., 19891. Kinesin-induced microtubule gliding was supported by
UTP, ITP, and CTP at similar ratios to ATP as those
reported here. GTP exhibited the most effective production of motility at 78% of the rate of 10 mM ATP. In
contrast to the results in the axoplasm, TTP was able to
produce motility at rates comparable to UTP, ITP, and
CTP (38% vs. 38%, 27%, and 31% respectively). On the
other hand, 5 mM alternate NTPs could not be used by
cytoplasmic dynein to produce microtubule gliding [Paschal and Vallee, 19871. Interestingly, both GTP and ITP
supported bidirectional organelle transport in the axoplasm, while TTP was ineffective as a substrate in either
direction (Table I).
Although the velocities of organelles in axoplasm
were reduced with UTP, GTP, ITP, and CTP as compared to the control, no significant difference could be
observed between anterograde and retrograde transport
in utilization of specific NTPs. This result was surprising
in light of the substantive differences between gliding of
microtubules mediated by kinesin and by cytoplasmic
dynein in vitro in the presence of various NTPs. Several
possibilities exist to explain this dicrepancy . The simplest interpretation of the axoplasm data suggests a
model in which the same motor or very similar motors
conduct transport in both directions. Such a model has a
precedent in the giant amoeba, Reticulomyxa. Euteneuer
et al. [1988] have shown that this organism contains a
microtubule-based motor which is related to dynein.
There is evidence that this dynein-like motor produces
bidirectional organelle movement in vitro in a phosphorylation-regulated fashion.
217
Several types of observations are consistent with
the possibility of a bidirectional motor being involved in
fast axonal transport, including biochemical, pharmacological, and immunochemical studies. Recently, Brady
et al. [1989] showed that perfusion of a monoclonal antibody to the heavy chain of kinesin into isolated axoplasm inhibited organelle motility in both directions with
similar kinetics and stoichiometries. This observation
suggested that kinesin was associated with vesicles moving in both directions and perhaps may play a role in both
anterograde and retrograde fast axonal transport. A collection of pharmacologic data in addition to the NTP data
indicates that a wide variety of treatments also fail to
distinguish clearly between the two directions. These include vanadate [Forman et al., 1983al (Brady and
Leopold, unpublished observations), N-ethylmaleimide
[Pfister et a]., 19891, pyrophosphate and ADP (Brady
and Leopold, unpublished observations), and AMP-PNP
[Lasek and Brady, 1985; Brady et al., 19851. In each of
these cases, treatments that would be expected to distinguish between kinesin and cytoplasmic dynein affect
both directions of transport at similar concentrations with
a similar time course.
Reports that one direction of transport may be selectively arrested [Forman et al., 1983b; Edmonds and
Koenig, 1987; Smith, 19881 do not clearly demonstrate
separate anterograde and retrograde motors due to several ambiguities. Either a subset of the anterograde particles is observed [Forman et al., 1983b; Smith, 19881 or
inhibition of one direction has been accompanied by a
drastic reduction in the velocity of transport in the other
direction [Edmonds and Koenig, 19871. Similarly, two
recent studies using in vitro assay systems report that
ultraviolet irradiation in the presence of vanadate affects
retrograde organelle movement more than anterograde
movement [Schroer et al., 1989; Schnapp and Reese,
19891. This treatment has been shown to cause degradation of axonemal dynein [Lee-Eiford et al., 19861 and
cytoplasmic dynein [Paschal et al., 19871 but not kinesin
[Lye et al., 19891. However, the interpretation of these
reports is complicated by variability between experiments and inefficiency of organelle translocation by the
systems employed.
Other models may also be invoked to explain the
available evidence on the molecular mechanisms underlying bidirectional organelle movements. These include
two distinct motors interacting synergistically on the surface of vesicles and two biochemically distinct motors
that share many pharmacological properties. Clearly, additional studies will be necessary using a variety of assay
systems to establish the molecular basis of bidirectional
transport in the neuron and other cell types. Although
video microscopic studies of isolated axoplasm do not
provide a high degree of biochemical resolution, corre-
218
Leopold et al.
lation of in vitro studies to cellular models on a continuing basis is important in order to develop an accurate
picture of cellular biochemistry. This study provides a
basis for such a correlation between ATPase activity and
in vitro motility data from potential organelle motors and
the transport of membrane-bounded organelles in cytoplasm, providing insights into the molecular mechanisms
of fast axonal transport in situ.
ACKNOWLEDGMENTS
The authors would like to thank Dr. Raymond J.
Lasek for the use of his microscope at the MBL and Mark
Wagner for valuable comments on the manuscript. This
work was supported by National Institutes of Health
(NIH) grants NS23868 (S.T. Brady and G.S. Bloom),
and NS23320 (S.T. Brady), National Science Foundation Biological Instrumentation Program grant DMB8701164 (G.S. Bloom and S.T. Brady), and Welch
Foundation grant 1-1077 (G.S. Bloom and S.T. Brady).
REFERENCES
Adams, R.J. (1982): Organelle movement in axons depends on ATP.
Nature 297:327-329.
Allen, R.D., and Allen, N.S. (1983): Video-enhanced microscopy
with a computer frame memory. J. Microsc. (Part I ) 129:3-17.
Allen, R.D., Allen, N.S., and Travis, J.L. (1981): Video-enhanced
contrast, differential interference contrast (AVEC-DIC) microscopy. A new method capable of analyzing microtubule
related motility in the reticulopodial network of Allogromia
laticollaris. Cell Motil. 1:291-302.
Allen, R.D., Metuzals, J . , Tasaki, I., Brady , S.T., and Gilbert, S.P.
(1982): Fast axonal transport in squid giant axon. Science 218:
1127-1 128.
Allen, R.D., Weiss, D.G., Hayden, J.H., Brown, D.T., Fujiwake,
H., and Sinipson, M. (1985): Gliding movement of and bidirectional transport along single native microtubules from squid
axoplasm: Evidence for an active role of microtubules in cytoplasmic transport. J. Cell Biol. 100:1736-1752.
Bagshaw, C.R., Eccleston, J.F., Trentham, D.R., Yates, D.W., and
Goody, R.S. (1972): The magnesium-dependent adenosine
triphosphatase of myosin. Cold Spring Harbor Symp. Quant.
Biol. 37: 127- 135.
Brady, S . T. (1985): A novel brain ATPase with properties expected
for the fast axonal transport motor. Nature 317:73-75.
Brady, S.T., Lasek, R.J., and Allen, R.D. (1982): Fast axonal transport in extruded axoplasm from squid. Science 218:
I 129-1 13 1.
Brady, S.T., Lasek, R.J., and Allen, R.D. (1985): Video microscopy
of fast axonal transport in extruded axoplasm: A new model for
studying molecular mechanisms. Cell Motil. 5:X 1-101.
Brady, S.T., Pfister, K.K., and Bloom, G.S. (1989): A monoclonal
antibody against the heavy chain of bovine brain kinesin inhibits both anterograde and retrograde axonal transport in isolated
squid axoplasm. Proc. Natl. Acad. Sci. USA (in press).
Caldwell, P.C. (1960): The phosphorus metabolism of squid axons
and its relationship to the active transport of sodium. J. Physiol. (Lond.) 1.52545-560.
Caldwell, P.C., Hodgkin. A.L., Keynes, R.D., and Shaw, T.I.
(1960): The effect of injecting ‘energy-rich’ phosphate compounds on the active transport of ions in the giant axons of
Loligo. J. Physiol. (Lond.) 152:561-590.
Cohn S.A., Ingold, A.L., and Scholey, J.M. (1987): Correlation
between the ATPase and microtubule translocating activities of
sea urchin egg kinesin. Nature 328:160-163.
Cohn S.A., Ingold, A.L., and Scholey, J.M. (1989): Quantitative
analysis of sea urchin egg kinesin-driven microtubule motility.
J. Biol. Chem. 264:4290-4297.
Darrow, R.A., and Colowick, S.P. (1962): Hexokinase from baker’s
yeast. In Colowick, S.P., and Kaplan, N.O. (eds.): “Methods
in Enzymology.” New York: Academic Press, Vol. 5, pp.
226-235.
Edmonds, B., and Koenig, E. (1987): Powering of hulk transport
(varicosities) and differential sensitivities of directional transport in growing axons. Brain Res. 406:288-293.
Euteneuer, U., Koonce, M.P., Pfister, K.K., and Schliwa, M. (1988):
An ATPase with properties expected for the organelle motor of
the giant ameoba, Reticulomyxu. Nature 322: 176-178.
Forman, D.S., Brown, K.J., and Livengood, D.R. (1983a): Fast axonal transport in permeabilized lobster giant axons is inhibited
by vanadate. J. Neurosci. 3:1279-1288.
Forman, D.S., Brown, K.J., and Promersberger, M.E. (198%): Selective inhibition of retrograde axonal transport by erythro9-[3-(2-hydroxynonyl)]adenine. Brain Res. 272: 194-197.
Forman, D., Brown, K., Promersberger, M., and Adelman, M.
(1984): Nucleotide specificity for reactivation of organelle
movements in pernieabilized axons. Cell Motil. 4: 121-128.
Kuznetsov, S.A., and Gelfand, V.I. (1986): Bovine brain kinesin is a
microtubule-activated ATPase. Proc. Natl. Acad. Sci. USA
83:8530-8534.
Lasek, R.J. (1984)The structure of the axoplasm. Curr. Top. Membr.
Transport 22:39-53.
Lasek, R.J., and Brady, S.B. (1985): Attachment of transported vesicles to microtubules in axoplasm is facilitated by AMP-PNP.
Nature 316:645-647.
Lee-Eiford, A., Ow, R.A., and Gibbons, I.R. (1986): Specific cleavage of dynein heavy chains by ultraviolet radiation in the presence of ATP and vanadate. J. Biol. Chem. 261:2337-2342.
Lienhard, G.E., and Secemski, 1.1. (1973): P’,P-Di(adenosine-S’)
pentaphosphate, a potent multisubstrate inhibitor of adenylate
kinase. J. Biol. Chem. 248:1121-1123.
Lye, R.J., Pfarr, C.M., and Porter, M.E. (1989): Cytoplasmic dynein
and microtubule translocators. In Warner, F.D., and Mclntosh,
J.R. (eds.): “Cell Movement, Volume 2: Kinesin, Dynein, and
Microtubule Dynamics.” New York: Alan R. Liss, Inc., pp.
141-154.
Paschal, B.M., Shpetner, H.S., and Vallee, R.B. (1987) MAP IC is
a microtubule-activated ATPase that translocates microtubules
in vitro and has dynein-like properties. J. Cell Bid. lOS(3):
1273-1282.
Paschal, B.M., and Vallee, R.B. (1987): Retrograde transport by the
microtubule-associated protein MAP IC. Nature 330:181-183.
Penningroth, S.M., Cheung, A,, Olehnik, K., and Koslosky, R.
(1982): Mechanochemical coupling in the relaxation of rigorwave sea urchin sperm flagella. J . Cell Biol. 92:733-741.
Pfister, K.K., Wagner, M.C., Bloom, G.S., and Brady, S.T. (1989):
Modification of the microtubule-binding and ATPase activities
of kinesin by N-ethylmaleimide (NEM) suggests a role for sulfhydryls in fast axonal transport. Biochemistry 28:9006-90 12.
Porter, M.E., Scholey, J.M., Stemple, D.L., Vigers, G.P.A., Vale,
R.D., Sheetz, M.P., and Mclntosh, J.R. (1987): Characterization of the microtubule movement produced by sea urchin egg
kinesin. J . Biol. Chem. 262(6):2794-2802.
Membrane Bounded Organelle Transport
Rozdzial, M.M., and Haimo, L.T. (1986a): Reactivated melanophore
motility: Differential regulation and nucleotidereyuirements of
bidirectional pigment granule transport. J . Cell Biol. 103:
2755-2764.
Rozdzial, M.M., and Haimo, L.T. (l986b): Bidirectional pigment
granule movements of melanophores are regulated by protein
phosphorylation and dephosphorylation. Cell 47: 1061-1070.
Sabri, M.I., and Ochs, S. (1972): Relation of ATP and creatine phosphate to fast axoplasmic transport in mammalian nerve. J. Neurochem. l9:282 1-2828.
Schnapp, B.J., and Reese, T.S. (1989): Dynein is the motor for retrograde axonal transport of organelles. Proc. Natl. Acad. Sci
USA 86:1548-1552.
Schnapp, B.J., Vale, R.D., Sheetz, M.P., and Reeses, T.S. (1985):
Single microtubules from squid axoplasm support bidirectional
movement of organelles. Cell 40:455-462.
Schroer, T.A., Steuer, E.R., and Sheetz, M.P. (1989): Cytoplasmic
dynein is a minus end-directed motor for membranous organelles. Cell 56937-946.
Shimuzu, T. (1987): The substrate specificity of dynein from Termhymenu cilia. J. Biochem. 102:1159-1 165.
219
Shpetner, H.S., Paschal, B.M., and Vallee, R.B. (1988): Characterization of the microtubule-activated ATPase of brain cytoplasmic dynein (MAP 1C). J. Cell Biol. 107:1001-1009.
Smith, R.S. (1988): Studies on the mechanism of the reversal of rapid
organelle transport in myelinated axons of Xenopus luevis. Cell
Motil. Cytoskeleton 10:296-308.
Sols, A,, de la Fuente, G., Villar-Palasi, C.. and Asensio, C. (1958):
Substrate specificity and some other properties of baker’s yeast
hexokinase. Biochim. Biophys. Acta 30:92-101.
Vale, R.D., Reese, T.S., and Sheetz, M.P. (1985a): Identification of
a novel force-generating protein, kinesin, involved in microtubule-based motility. Cell 42:39-50.
Vale, R.D., Schnapp, B.J., Mitchison. T., Steuer, E., Reese, T.S.,
and Sheetz, M.P. (1985b): Different axoplasmic proteins generate movement in opposite directions along microtubules in
v i m . Cell 43623-632.
Wagner, M.C., Pfister, K.K., Bloom, G.S., and Brady, S.T. (1989):
Copurification of kinesin polypeptides with microtubule-stimulated Mg-ATPase activity and kinetic analysis of enzymatic
processes. Cell Motil. Cytoskeleton 12: 195-215.