Content uploaded by Todd Hennessey
Author content
All content in this area was uploaded by Todd Hennessey on Jan 23, 2024
Content may be subject to copyright.
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS
Vol. 337, No. 2, January 15, pp. 351–359, 1997
Article No. BB969779
RESEARCH REPORT
A Soluble Ecto-ATPase from Tetrahymena t hermophila: Purification and
Similarity to the M embrane-Bound Ecto-ATPase of Smooth Muscle
Thomas M. Smith, Jr., Terence L. Kirley,* and Todd M. Hennessey
1
Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York 14260; and *Department of
Pharmacology and Cell Biophysics, University of Cincinnati College of Medicine, Cincinnati, Ohio 45267
Received October 22, 1996
Key Words: ecto-ATPase; Tetrahymena thermophila;
purinergic reception.
For the first time, a soluble, dedicated E-type ecto-
ATPase has been identified and purified. This fully
soluble ecto-ATPase is released into the growth me-
dia of the single-celled eukaryote, Tetrahymena, at a
constant rate over time (independent of the growth Ecto-ATPases are a ubiquitous set of enzymes that belong
to the class of ATPases known as the E-type ATPases. Al-
phase of the cells) and it has characteristics similar
though the membrane-bound forms of these enzymes have
to those previously described for the membrane-
been found in virtually every type of eukaryotic plasma mem-
bound ecto-enzyme in Tetrahymena. It was purified brane examined, their precise function has been elusive.
by a combination of ion-exchange, size exclusion, and These enzymes have been implicated in possible mechano-
affinity chromatography and nondenaturing gel elec- chemical, virus, cancer cell, transport, ectokinase, cell adhe-
trophoresis. Its molecular weight was determined to sion, nucleotide scavenging, and purinergic functions (1).
be approximately 66,000 Da by denaturing gel elec- Purinergic agonists, such as ATP, are considered to be fast
trophoresis and approximately 69,000 Da by size ex- excitatory neurotransmitters in both the central and periph-
clusion chromatography of the native form. The puri- eral nervous systems (2) and intercellular modulators of
fied soluble enzyme displays the general characteris- many other cell functions. The general importance of pu-
tics of a dedicated E-type ecto-ATPase such as Ca
2/
rinergic reception is seen in the involvement of nucleoside
or Mg
2/
dependence, hydrolysis of ATP and other nu- triphosphates on cardiac, vascular and smooth muscle, excit-
cleoside triphosphates (but not nucleoside diphos- atory and inhibitory effects on neurons, and effects on ion,
phates) and insensitivity to common ATPase inhibi- hormone, exocrine gland, platelet, mast cell, and inflamma-
tors (vanadate, azide, ouabain, N-ethylmaleimide and tory cell secretions (3).
p-chloromercuriphenyl sulfonate). It was further Since Tetrahymena can detect and respond (in a chemore-
shown to be immunologically similar (by polyclonal pellent assay) to nanomolar concentrations of GTP in their
antibodies) to both the membrane-bound ecto- extracellular environment (4), we have been interested in
ATPase of chicken gizzard smooth muscle (66 kDa) nucleoside triphosphates as primitive extracellular signaling
and a 66-kDa protein in Tetrahymena plasma mem- molecules. Our recent finding that Tetrahymena also respond
to external ATP and its analogs (5) makes this a convenientbranes. The ecto-ATPase enzyme activity was also
model system for studying the mechanisms involved in pu-
shown to be present in both the body plasma mem-
rinergic reception. Since ecto-ATPases have rather broad
brane and ciliary plasma membrane fractions but the
substrate specificities, their presence and characteristics
body membrane had slightly higher specific activi- must be known in order to unambiguously study such pu-
ties. We propose that this ecto-ATPase of Tetrahy- rinergic receptor kinetics and specificities.
mena may play a role in inactivating purinergic sig- The E-type class of ecto-ATPases consists of two major
nals, such as in their chemorepulsion responses to types of enzymes: ATP diphosphohydrolases (ATPDases or
external GTP and ATP. It may also play a minor role apyrases) and the dedicated ecto-ATPases (1). Because of
in extracellular nucleotide scavenging.
q1997 Academic
their strong similarities, Dhalla and Zhao (6) developed a
Press
useful set of criteria to distinguish between those enzymes
that hydrolyze ATP or ADP only (ecto-ATPases, ecto-
ADPases) from those enzymes that hydrolyze both ATP and
1
To whom correspondence should be addressed. Fax: (716) 645-
2975. E-mail: THENNES@UBVMS.CC.BUFFALO.EDU.
ADP (ATP diphosphohydrolases). ATP diphosphohydrolases
351
0003-9861/97 $25.00
Copyright q1997 by Academic Press
All rights of reproduction in any form reserved.
352
SMITH, KIRLEY, AND HENNESSEY
fuging at 150gin an IEC Model HN-SII benchtop centrifuge. After
were generally distinguished by three characteristics: (1)
the cells were removed, the medium was again centrifuged at 700g
their ability to hydrolyze both nucleoside triphosphates and
for 5 min to remove any remaining particulate matter. The medium
nucleoside diphosphates at near equal rates, (2) their inhibi-
was placed on ice to lower the temperature to 27C. The chilled me-
tion of activity by incubation with 10–20 m
M
sodium azide,
dium was then centrifuged in a Model L8-55 Beckman ultracentri-
and (3) their lack of stimulation of activity by treatment with
fuge at 105,000gfor 60 min with a 50.2-Ti rotor at 27C. This step
the lectin concanavalin A.
removed any vesicles that may have been present, yielding a truly
The characteristics of the dedicated (‘‘true’’) ecto-ATPases
soluble starting fraction.
may be generally summarized as follows: (1) The enzymes
The high-speed medium supernatant was then chilled in an ice
possess an active nucleotide binding site situated upon the
bath, and to this stirring solution was slowly added solid ammonium
sulfate to a final concentration of 85%. After the last addition of salt,
exterior of the cell; (2) they are glycosylated; (3) they are
the solution was allowed to stir for an additional 30 min to ensure
dependent upon calcium or magnesium for activity, and can
bulk precipitation of the proteins. The suspension was then centri-
use either cation for hydrolysis of the substrate; (4) they are
fuged at 10,000 rpm for 30 min in a Beckman J2-HS centrifuge with
insensitive to specific inhibitors of P-type, F-type, and V-type
a Model JA-14 rotor at 27C.
ATPases; and (5) they are able to hydrolyze both purine and
At the end of the centrifugation, the supernatant was discarded
pyrimidine nucleoside triphosphates as substrates, but not
and the pellets were dissolved in 50 ml of Tris
2
wash buffer (25
nucleoside mono- or diphosphates nor nonnucleoside phos-
m
M
Tris–HCl, pH 7.7). This protein containing solution was then
phates (1, 7).
concentrated to approximately 5 ml in a Nucleopore Model S4370
Although several examples of soluble ATP diphosphohy-
low-pressure stirred cell (Costar, Cambridge, MA) with a 10,000
NMWC regenerated cellulose ultrafiltration disk membrane (Milli-
drolases have been described previously (8 – 10), no detailed
pore Corp., Bedford, MA). This concentrated solution was then di-
identification and description of a soluble dedicated ecto-
luted threefold in Tris wash buffer and added to a 50-ml polypropyl-
ATPase has been reported. Soluble, released ATPases that
ene conical tube (Becton Dickinson and Co., Franklin, NJ) containing
satisfy some of the criteria for E-type ATPase activity have
15 ml of DEAE Sephadex A-50 anion exchanger slurry equilibrated
been described previously in the ciliate, Paramecium (11, 12)
in the same buffer. The protein:Sephadex slurry was then brought
but these were neither identified as ecto-ATPases nor thor-
to 50 ml with Tris wash buffer and allowed to incubate on ice for 30
oughly characterized in the terms of their ecto-ATPase activi-
min with gentle inverting.
ties. Membrane-bound ecto-ATPases have been described in
The unbound proteins, including the ecto-ATPase, were then de-
Tetrahymena (13–16) but no soluble form has been pre-
canted from the matrix and concentrated using the stirred cell in
viously identified as an E-type ATPase in this or any other
preparation for size exclusion chromatography. The concentrated
mixture (approx 2.5 ml) was then adjusted to 8% (v/v) glycerol and
ciliate.
applied to a Sephacryl S-200-HR size exclusion column (fractionation
Ciliates such as Tetrahymena are excellent eukaryotic
range 5000–250,000) equilibrated in 50 m
M
Mops–NaOH, pH 7.1,
model systems for the study of the structure and physiological
100 m
M
NaCl. The column (1.5 1100 cm) was run at a flow rate of
function of ecto-ATPases due to their ease of pure cell culture,
5 ml/cm
2
/h. Two-milliliter fractions were collected and tested for ecto-
high cell density (10
6
cells/ml), behavioral responses to exter-
ATPase activity as described below. The most active fractions were
nal nucleotiside triphosphates as chemorepellents (4), history
pooled and adjusted to 350 m
M
NaCl, 0.5 m
M
MnCl
2
, 0.5 m
M
CaCl
2
of biochemical techniques, and characterizations and amena-
in preparation for lectin affinity chromatography.
bility to modern molecular biological techniques such as ge-
Active fractions were then applied to a 1-ml concanavalin A –Seph-
netic transformations and gene knockout experiments (17).
arose 4B lectin affinity column. The column (0.5 17 cm) was equili-
brated in Con A wash buffer (50 m
M
Mops–NaOH, pH 7.1, 350 m
M
Since Tetrahymena show the same ‘‘avoiding reactions’’ in
NaCl, 0.5 m
M
MnCl
2
, 0.5 m
M
CaCl
2
) with a flow rate of 20 ml/cm
2
/
chemorepellents as the related ciliate, Paramecium (4), be-
h. After all of the protein containing solution had passed through
havioral mutant selections can be done for ‘‘genetic dissec-
the matrix, the column was allowed to wash with Con A wash buffer
tion’’ (18) of the chemosensory transduction pathways for pu-
for 10 bed volumes. The Con A-binding proteins were then eluted
rinergic responses and adaptation.
with 300 m
M
methyl-
a
-
D
-mannopyranoside prepared in Con A wash
buffer. One-milliliter fractions were collected and tested for activity.
MATERIALS AND METHO DS
The most active fractions were pooled and allowed to dialyze over-
night into IEC wash buffer (10 m
M
Tris–HCl, pH 8.2, 2 m
M
MgCl
2
)
Unless otherwise indicated, all chemicals and supplies were pur- with several changes of buffer.
chased from Sigma Chemical Co. (St. Louis, MO). For the subsequent ion-exchange chromatography, the dialyzed
sample was applied to a 1 110-cm column of DEAE Sephadex A-50
Cell Culture
anion exchanger equilibrated in IEC wash buffer. A flow rate of 10
ml/cm
2
/h was maintained throughout binding, washing, and elution.
Tetrahymena thermophila B was a generous gift from N. E. Wil-
liams, University of Iowa. Cells were grown in 2% proteose peptone
(Difco Laboratories, Detroit, MI) and 0.1 m
M
FeCl
3
at 257C with
continuous shaking on an orbital shaker (13). Early-stationary-phase
2
Abbreviations used: Tris, 2-amino-2-(hydroxymethyl)-1,3-pro-
cultures were grown by inoculating 100 ml of medium in 250-ml panediol; EGTA, ethylene glycol bis(
b
-aminoethyl ether) N,N,N*,N*-
Erlenmeyer flasks with 1 ml of stationary-phase stock culture. The tetraacetic acid; SDS–PAGE, sodium dodecyl sulfate–polyacryl-
stock cultures were maintained at 257C in 10-ml test tubes and sub- amide gel electrophoresis; NADPH, nicotinamide adenine dinucleo-
cultured every 3 weeks. For a typical purification, 400 ml of cells in tide phosphate; Mops, 3-[N-morpholino]propanesulfonic acid; EDTA,
medium was used. ethylenediaminetetraacetic acid; CMVs, ciliary membrane vesicles;
BMVs, body membrane vesicles; Con A, concanavalin A; dd water,
Purification of the E-type ATPase
deionized, distilled water; NBT, nitro blue tetrazolium; BCIP, 5-
bromo-4-chloro-3-indolyl phosphate; DMSO, dimethyl sulfoxide;A 96-h conditioned medium was used as the source for the enzyme
purification. Cells were gently removed from the medium by centri- TBS, Tris-buffered saline.
353
SOLUBLE ECTO-ATPase OF Tetrahymena
After allowing 10 bed volumes of IEC wash buffer to flow through sodium acetate– acetic acid (pH 4.5– 5.5), Mops– NaOH (pH 6.0– 7.5),
Tris–HCl (pH 8.0– 9.5). The reaction was initiated by the additionthe column, proteins were eluted by a gradient of 0–200 m
M
NaCl
in IEC wash buffer. Active fractions were pooled and concentrated of 2 m
M
ATP and allowed to continue for 5 min. The protein concen-
tration was adjusted so that the activity was linear with respect toin a Millipore Ultrafree-Cl Polysulfone microconcentrator (10,000
NMWL) for final purification by nondenaturing gel electrophoresis. the time of the assay. Liberated phosphate was detected by the
method of Lanzetta et al. (21). One unit of activity is defined as 1
m
mol phosphate released per hour.
Gel Electrophoresis
Denaturing (SDS) discontinuous gel electrophoresis was conducted
Size Determination of the Native Ecto-ATPase
as described by Ausubel (19). Electrophoresed proteins were detected
either by the use of an ISS Pro-Blue Coomassie staining kit (Inte- To determine the native molecular weight of the ecto-ATPase, the
grated Separation Systems, Natick, MA) or by silver staining (20). elution volumes for blue dextran (M
r
2,000,000), yellow dextran (M
r
Nondenaturing discontinuous gel electrophoresis (19) was used to 20,000), cytochrome c (M
r
12,400), and vitamin B
12
(M
r
1355) molecu-
separate the native forms of this enzyme. After electrophoresis, the lar weight standards were determined in the Sephacryl S-200-HR
gel was stained for ecto-ATPase activity by one of two methods de- size exclusion column and a standard curve was prepared. The elu-
pending upon the desired final product. tion volume for the ecto-ATPase was then determined and the native
molecular weight was deduced from the standard curve.
Detection of Ecto-ATPase Activity in Native Gels
Body Membrane Vesicles Preparation
For analysis of purified enzyme by SDS– PAGE, the gel was
stained for ecto-ATPase activity by a modification of the phosphate Body membrane vesicles from T. thermophila were prepared as
detection method of Lanzetta et al. (21). The electrophoresed gel was described by Smith and Hennessey (23) with the following modifica-
washed two times for 5 min each in 200 ml deionized, distilled (dd) tions: A 100-ml culture of 96-h T. thermophila (approx 490,000 cells/
water at 257C. The gel was then cut in half and each half was placed ml) was used as the source of starting material and the cells were
into 200 ml buffer consisting of 50 m
M
Mops–NaOH, pH 7.1, 0.1 m
M
washed three times in modified Dryl’s solution (2 m
M
Mops–NaOH,
sodium ortho-vanadate, 10 m
M
sodium azide, 2 m
M
ATP, and either pH 7.1, 2 m
M
sodium citrate, 1.5 m
M
calcium chloride) prior to incu-
3m
M
CaCl
2
or 5 m
M
EGTA and 5 m
M
EDTA. After incubation for bation in STEN buffer (500 m
M
sucrose, 20 m
M
Tris–HCl, 2 m
M
20 min at 257C, the incubation buffer was removed, the gels were EDTA, 6 m
M
NaCl, pH 7.5). All other steps were conducted as de-
each washed for 1 min with 200 ml dd water, and 50 ml phosphate scribed previously.
detection mixture (21) was then added to each gel. After a few min-
utes, a green phosphate precipitate formed in the 3 m
M
CaCl
2
con-
taining gel half, showing the location of the ecto-ATPase enzyme,
Ciliary Membrane Vesicles Preparation
which was then excised with a clean razor blade and prepared for Ciliary membrane vesicles from T. thermophila were prepared ex-
SDS–PAGE. The gel slice was quickly rinsed in water and allowed actly as described by Adoutte et al. (24). Both the ciliary and plasma
to soak for 2 min in 1 ml of SDS sample buffer (19). The gel was membrane vesicles were washed once in 10 m
M
Tris–HCl, pH 8.0,
then transferred to 1 ml of fresh SDS sample buffer and allowed to pelleted at 105,000gfor 60 min using a 50.2-Ti rotor at 27Cina
incubate for 90 min at 307C. The slice was then loaded onto an SDS – Model L8-55 Beckman ultracentrifuge, and suspended in 200
m
lof
polyacrylamide gel and subjected to both Coomassie and silver stain- fresh buffer. This step served to remove any residual polymers or
ing after electrophoresis. sucrose from the vesicle preparation.
For analysis of the enzymatic characteristics, the method of Picher
et al. (22) was used after electrophoresis to avoid exposing the en-
zyme to the denaturing 1
M
HCl present in the Lanzetta et al. (21)
Polyclonal Antibody Production
method. After detection by calcium phosphate precipitation, the ecto-
Polyclonal antibodies to the chicken ecto-ATPase were generously
ATPase-containing region was excised from the native gel using a
provided by T. Kirley, Department of Pharmacology and Cell Bio-
clean razor blade. The gel slice was finely minced, crushed, and sus-
physics, University of Cincinnati, and were produced by Cocalico
pended in 200
m
lof25m
M
Mops–NaOH, pH 7.1, 0.1% Triton X-100.
Biologics (Reamstown, PA). Briefly, immunoaffinity-purified ecto-
The slurry was then transferred to a microfuge tube containing a
ATPase was obtained as described previously (25) and the antigen
small piece of dialysis membrane covering the top of the tube. The
was prepared by electroblotting SDS– polyacrylamide gels to nitro-
tube was inverted into the dialysis buffer and allowed to dialyze for
cellulose membrane. Membranes were stained with 0.1% ponceau
24hat377C in the same buffer. After dialysis, the acrylamide parti-
S (26) and the 66-kDa protein was excised. The protein-containing
cles were removed by centrifugation and the resulting supernatant
nitrocellulose strip was destained in water, dissolved in DMSO, and
was used for ecto-ATPase enzyme assays.
the denatured antigen used for antibody production.
In Vitro Ecto-ATPase Assay
Western Blotting
ATPase assays were conducted as follows: Reactions were con-
ducted in 12 175-mm disposable glass test tubes (VWR Scientific) Western blotting was performed as described by Harlow and Lane
(26). Briefly, SDS–polyacrylamide gels were electroblotted to Im-in a water bath at 257C. The standard final reaction mixture (250
m
l) contained purified enzyme in 25 m
M
Mops–NaOH, pH 7.1, 3.5 mobilon PVDF transfer membrane (Millipore Corp., Bedford, MA)
as described by Matsudaira (27) and allowed to block overnight inm
M
potassium chloride, 25 m
M
sodium chloride, 1% Triton X-100,
10 m
M
sodium azide, 1 m
ML
-(/)-tartaric acid, 0.1 m
M
sodium ortho- Blotto (5% (w/v) nonfat dry milk dissolved in Tris-buffered saline
(TBS)). After washing the membrane in TBS, primary antibody wasvanadate, and either 1 m
M
CaCl
2
and/or 1 m
M
MgCl
2
for cation
dependent ATPase activity, or 10
m
M
EGTA for cation independent diluted 1:200 into Blotto and allowed to react for 2 h. After washing
the membrane in TBS, goat anti-rabbit secondary antibody (alkalineATPase activity. For free cation concentration, the computer pro-
gram Calcium version 1.1 was used. For inhibitor studies, the respec- phosphatase conjugate) was diluted 1:4000 into Blotto and allowed
to react for 1 h. The membrane was developed using Sigma Fasttive chemicals were allowed to preincubate with the purified enzyme
for 15 min at 257C. For pH studies, three buffer systems were used: BCIP/NBT substrate tablets.
354
SMITH, KIRLEY, AND HENNESSEY
The characteristics of the soluble enzyme are typical of an
E-type ATPase and are similar to that of the cell surface ecto-
ATPase. Much like the membrane-associated ecto-ATPase,
the purified soluble ecto-ATPase displays a relatively high
K
m
for ATP. Lineweaver– Burke analysis of the enzyme yields
the expected straight line from which a K
m
for ATP of 0.24
m
M
was calculated.
The soluble ecto-ATPase has a requirement for divalent
cations, and a high affinity for calcium and magnesium
(Fig. 3). The K
d
for free calcium was determined to be 15.9
m
M
, while the K
d
for free magnesium was determined to be
10.2
m
M
.
The soluble, purified ecto-ATPase is highly insensitive to
many of the inhibitors of the classic ATPase inhibitors (Table
II). Of special note is the relatively strong inhibition observed
after incubation with the lectin concanavalin A.
The nucleotide specificity of the purified enzyme is very
broad, yet specific for nucleotide triphosphates (Table III).
FIG. 1. Soluble ecto-ATPase activity is released from intact cells
Enzyme was reacted with 2 m
M
substrate concentration in
in a time-dependent manner. A 100-ml culture of Tetrahymena ther-
the standard ecto-ATPase assay. The ability to hydrolyze the
mophila cells was grown as described under Materials and Methods.
The culture reached stationary phase by 70 h, as indicated by total
adenosine mono- and diphosphates as well as nonnucleotide
cell protein (see inset). At the indicated times, cells were removed
phosphates was nearly absent.
by low-speed centrifugation and the remaining medium was centri-
The purified ecto-ATPase possesses a narrow pH range for
fuged at 105,000gfor 60 min. The high-speed supernatant was tested
optimum activity (Fig. 4). The enzyme was tested in a pH
for released, soluble ecto-ATPase as described under Materials and
range from 4.5 to 9.5 using three different buffers systems.
Methods (nÅ3).
The ecto-ATPase of Tetrahymena is immunologically simi-
lar to the previously described membrane-associated chicken
gizzard smooth muscle ecto-ATPase. The highly enriched pH
RESULTS 8.2 anion exchange fraction of Tetrahymena was probed with
an affinity-purified polyclonal antibody to the chicken ecto-
An enzyme with the characteristics of an E-type ATPase ATPase. The fraction displayed a cross-reacting band at 66
(i.e., glycosylation, specificity for nucleoside triphosphates, kDa (Fig. 5), the same molecular weight for the native gel-
cation dependency, insensitivity to classical ATPase inhibi- purified enzyme. This protein is identical in size to the
tors) is released from Tetrahymena into the surrounding cul- chicken gizzard membrane ecto-ATPase (28).
ture fluid as a function of time (Fig. 1). This enzyme does not Purified body membrane and ciliary membrane vesicles
appear to be selectively secreted during any particular phase were also obtained from T. thermophila. When these purified
of growth (see Fig. 1, inset); rather, the activity is found to membrane fractions were examined for ecto-ATPase activity
be released throughout all phases. It is soluble by the defini- in an in vitro ecto-ATPase assay, the body plasma membrane
tion that it remains in the supernatant after a 1-h spin at vesicles had a somewhat higher specific activity. The specific
100,000g. activity of the body plasma membrane was 19.6 {0.8 Units/
The ecto-ATPase was highly enriched from the sur- mg, while the specific activity of the ciliary membrane vesi-
rounding culture medium by ammonium sulfate precipita- cles was 14.9 {0.3 Units/mg (nÅ3).
tion, reverse anion-exchange chromatography at pH 7.7, size Western blot analysis of the purified ciliary membrane
exclusion chromatography, lectin affinity chromatography, fraction using a 1:200 dilution of immune serum displayed
and anion exchange chromatography at pH 8.2. The enzyme the 66-kDa protein reacting (Fig. 6). Analysis of purified
could be further purified to homogeneity by a final step of BMVs from Tetrahymena also revealed a 66-kDa protein rec-
nondenaturing gel electrophoresis (Table I). The enzyme was ognized (data not shown).
purified over 650-fold compared to the conditioned cell me-
dium with a final yield of 4.3%.
To determine the molecular weight of the enzyme, a 7% D ISCUSSION
nondenaturing gel was prepared as described and a sample
of a highly enriched fraction loaded and run under nondena- A soluble, dedicated ecto-ATPase that fits all of the criteria
for ecto-ATPase activity has never been described in any sys-turing conditions. After excision of the ecto-ATPase and sub-
sequent SDS – PAGE, a protein with a relative mobility of tem. For the first time, we have described an ATPase that is
released into the extracellular medium (without any cellularapproximately 66 kDa was observed (Figs. 2A and 2B). Silver
staining did not reveal any additional proteins present (data manipulations or special treatments) that possesses the main
hallmarks of a dedicated (‘‘true’’) ecto-ATPase.not shown).
The ecto-ATPase of T. thermophila appears to have a na- We first noticed an ecto-ATPase-like activity present in the
conditioned media of Tetrahymena. We examined thetive molecular weight of approximately 65 –70 kDa as deter-
mined by size exclusion chromatography. The V
e
/V
o
ratio of amounts of enzyme released over time and found that the
amount increased with time in culture, irrespective of thethe ecto-ATPase was determined to be 1.23, giving a native
molecular weight of approximately 69 kDa. phase of growth. The enzyme did not appear to be the only
355
SOLUBLE ECTO-ATPase OF Tetrahymena
TABLE I
Enzyme Activity Table for the Purification of the Soluble Ecto-ATPase
Volume Protein conc. Total protein Specific activity Total units Fold Recovery
Fraction (ml) (mg/ml) (mg) (units/mg) (
m
moles/hr) purification (%)
T. thermophila B cell
media (96 h) 350 0.086 30.31 183 5570 1 100
85% ammonium sulfate cut 14 2.00 28.01 179 5039 .97 90
DEAE flow through 38 0.213 8.12 466 3786 2.5 67
Size exclusion cut
34 2 0.054 0.108 608 65 3.3 1.1
35
a
2 0.072 0.144 1768 254 9.6 4.5
36
a
2 0.088 0.177 2235 395 12.1 7.1
37
a
2 0.086 0.172 2326 400 12.6 7.1
38
a
2 0.077 0.154 2959 455 16.1 8.1
39
a
2 0.068 0.137 2831 387 15.4 6.9
40 2 0.057 0.114 1405 160 7.6 2.8
41 2 0.053 0.107 367 39 1.9 0.7
Concanavalin A cut
1 1 nd nd — 3 — 0.05
2
a
1 0.017 0.017 17301 294 94 5.2
3
a
1 0.030 0.030 7883 236 42.9 4.2
4
a
1 0.023 0.023 6011 138 32.7 2.4
5
a
1 0.017 0.017 4995 84 27.2 1.5
6 1 0.012 0.012 2907 34 15.8 0.62
DEAE, pH 8.2, IEC (0–200
m
M
NaCl) 1 0.015 0.015 18734 281 102.3 5.0
Native PAGE 0.2 0.009 0.001 120689 238 659.5 4.3
Note. The 96-h cultures of Tetrahymena were grown as described previously and the conditioned medium was collected. After removing
the cells and centrifuging the supernatant at 105,000gfor 60 min, the ecto-ATPase was purified as described under Materials and Methods.
The native gel-purified fraction was enriched over 650-fold from the starting material extract and had a specific activity of 120,689 units/
mg.
a
Indicates fractions were pooled for subsequent steps.
protein secreted, however, and the general amount of protein was not performed, the molecular weight is consistent with
our purified fraction. Another soluble ATPase activity wassecreted also increased as a function of time in culture (data
not shown) yielding little change in the specific activity. previously described in Tetrahymena (14) but it was de-
scribed as a cytoplasmic enzyme and it was not evaluatedTherefore, 96-h cell medium was chosen as a convenient
source with which to purify the enzyme. This enzyme activity for ecto-ATPase characteristics. Other membrane bound ecto-
ATPase-like activities have also been described in Tetrahy-was also released into a buffered solution of washed cells but
the yield was less because the cells could not be grown in mena by other investigators (15, 16).
Three separate lines of evidence point to the 66-kDa pro-this nonnutritive solution (data not shown).
We tried to obtain a soluble form of the enzyme by incuba- tein of T. thermophila being the genuine soluble ecto-ATPase.
They are as follows:tion of either intact cells or cell homogenates with proteases
(papain, trypsin). This did not cause release of soluble ecto- 1. Analysis of an enriched fraction by native gel electro-
ATPase activity or any decrease in enzymatic activity (data phoresis, ecto-ATPase staining, and SDS – PAGE, displayed
not shown). However, since there are many different types a protein of 66 kDa.
of proteases with different specificities, we do not rule out 2. Probing the pH 8.2 ion exchange fraction with a poly-
the possibility that a specific protease may elicit release of clonal antibody to the chicken gizzard ecto-ATPase displays a
this ecto-ATPase. Lin and Russel (29) also saw no significant cross-reacting protein at 66 kDa, the same molecular weight
release of ecto-ATPase activity when rat hepatocytes were observed for the chicken enzyme (28).
similarly treated with trypsin, chymotrypsin, and papain. 3. Size exclusion chromatography indicates that the native
Previously, Dentler (13) used a Triton X-114 procedure to form of the enzyme is approximately 69 kDa, similar to that
enrich for an E-type ATPase activity from the ciliary mem- observed after SDS – PAGE of the denatured form.
brane of Tetrahymena. By using sucrose gradient fraction-
ation, he was later able to tentatively identify the ATPase The characteristics of the purified soluble ecto-ATPase are
similar to, but not identical to, the membrane-associated (de-activity as being associated with one of two proteins of 66 or
88 kDa, each of which is externally biotinylated, stained with tergent extracted) form of the enzyme. Although they both
share relatively high K
m
’s for ATP, a similar insensitivity toCon A, and is, therefore, exposed to the ciliary surface (30).
Although definitive identification of the ecto-ATPase enzyme the classic inhibitors of the V-, P-, and F-type ATPase inhibi-
356
SMITH, KIRLEY, AND HENNESSEY
FIG. 3. The purified soluble ecto-ATPase can use either magnesium
or calcium for activity, although magnesium yielded higher activities
at all concentrations tested. A sample of the native gel-purified solu-
ble ecto-ATPase was tested in the in vitro ecto-ATPase assay for free
cation dependency using 2 m
M
ATP as substrate (see Materials and
Methods). The free cation concentration was determined with the
aid of a computer program. The enzyme has a high affinity for cal-
cium (open circles) and magnesium (closed circles), showing an esti-
mated K
d
for free calcium of 15.9
m
M
and a K
d
for free magnesium
of 10.2
m
M
.
with concanavalin A. The criteria set forth by Dhalla and
FIG. 2. The soluble ecto-ATPase was purified from the conditioned
Zhao (6) stated that ecto-ATPases may be distinguished from
cell medium and displayed a relative mobility of 66 kDa. (A) SDS–
other E-type ATPase by their inability to hydrolyze ADP,
PAGE indicates that many proteins are present in conditioned cell
medium when either 2
m
g (lane 3) or 20
m
g (lane 2) of conditioned cell
their lack of inhibition by sodium azide, and the stimulation
medium was analyzed on a 7.5% denaturing gel and silver stained as
of their activity by concanavalin A. The enzyme under study
described under Materials and Methods. The gels were purposely
overdeveloped to enable visualization of the minor, diffusely staining
66-kD protein (see lane 2). (B) A sample from the final anion ex-
TABLE II
change fraction was further purified on a 7% nondenaturing gel and
stained for ecto-ATPase by the method of Lanzetta et al. (21). The
The Purified Soluble Ecto-ATPase Is Resistant to Many
ecto-ATPase containing region was excised, soaked in SDS sample
Inhibitors of the Classic ATPases
buffer as described, and analyzed on a 12% SDS –polyacrylamide gel.
When this gel was stained with Coomassie blue, a prominent protein
Compound Concentration Relative activity (%)
was seen at 66 kDa (lane 1). No other proteins were detected when
the gel was destained and later silver stained (not shown).
None — 100
L
-(/)-Tartaric acid 10 m
M
100
Dithiothreitol 10 m
M
100
tors, and a high affinity for divalent cations, the relative rate
N-Ethylmaleimide 5 m
M
100
of GTP hydrolysis was lower for the purified enzyme than for
Sodium azide 10 m
M
100
the membrane-associated form of the enzyme. This could be
p-CMBS 0.5 m
M
100
due to the presence of residual or contaminating NTPases
ortho-Vanadate 0.1 m
M
100
present on the cell surface of intact Tetrahymena (35). This
ortho-Vanadate 0.2 m
M
93.5 {3.0
is supported by the fact that the ecto-ATPase represents only
Triton X-100 1% 100
Molybdic acid 10
m
M
100
about 30% of the total ATP hydrolyzing activity present on
Concanavalin A 25
m
M
43.0 {0.6
the cellular surface (31) and yet nearly 100% of the total
of the native gel electrophoresis-purified fraction. Therefore,
Note. Many compounds that are effective at inhibiting the activity
differences between these two assays may be due to other
of other classes of ATPase were tested in the in vitro ecto-ATPase
contaminating ATPase activities in the membrane-bound
assay. Activity was measured as
m
mol phosphate/h as described un-
fractions. In vivo ecto-ATPase assays with intact cells (5)
der Materials and Methods.
L
-(/)-Tartaric acid, dithiothreitol, N-
showed characteristics that were similar enough to those of
ethylmaleimide, sodium azide, molybdic acid, and Triton X-100 were
the extracted membrane-bound form and purified soluble
all ineffective. At high concentrations, vanadate inhibited about 7%
form to conclude that the same enzyme was being assayed
of the activity. Incubation with the lectin concanavalin A led to a
under all three conditions.
decrease of nearly 60% of the activity. Inhibitors are expressed as
percentage of control activity (nÅ3).
The ecto-ATPase activity is highly sensitive to incubation
357
SOLUBLE ECTO-ATPase OF Tetrahymena
TABLE III
The Purified Soluble Ecto-ATPase Possesses a Broad
Substrate Range for Nucleoside Triphosphates
Substrate Relative activity (%)
Adenosine 5*-triphosphate 100
Inosine 5*-triphosphate 95.7 {1.2
Uridine 5*-triphosphate 89.8 {3.5
Cytosine 5*-triphosphate 70.6 {0.8
Guanosine 5*-triphosphate 23.7 {0.6
Adenosine 5*-diphosphate 2.1 {1.1
Adenosine 5*-monophosphate 0.0
Guanosine 5*-monophosphate 0.0
Adenosine 2*- and 3*-monophosphate 0.0
p-Nitrophenyl phosphate 0.0
Note. Native gel purified ecto-ATPase was tested in the in vitro
ecto-ATPase assay using a concentration of 2 m
M
for each substrate
tested (see Materials and Methods). The enzyme was shown to hy-
drolyze ITP nearly as well as ATP, followed by UTP, CTP, and less
effectively for GTP. The ability to hydrolyze ADP was nearly absent.
Other nucleotide monophosphates were not hydrolyzed. Activities
are expressed as percentage of control activity (nÅ3).
FIG. 5. A soluble protein at 66 kDa from conditioned medium cross-
reacts with polyclonal antibodies directed against the chicken gizzard
here fits the first two criteria, but fails to satisfy the third.
smooth muscle ecto-ATPase. An aliquot of the highly enriched pH
Unlike the smooth muscle ecto-ATPase of chicken described
8.2 anion exchange fraction from Tetrahymena was electrophoresed
by Stout and Kirley (28), which shows immunological and
and transferred to PVDF membrane as described under Materials
structural similarity to the Tetrahymena enzyme and may
and Methods. The membrane was then cut in half for further analy-
sis. (A) One side of the membrane was Coomassie stained to examine
be stimulated nearly 1900% by concanavalin A treatment,
the distribution of proteins present. (B) The other side was incubated
the ecto-ATPase of Tetrahymena is actually inhibited nearly
with an affinity-purified polyclonal antibody directed against the
60% by a 15-min preincubation with the lectin. This inhibi-
ecto-ATPase of chicken. This revealed a minor cross-reacting protein
tion by Con A is similar to that observed for the rabbit skele-
at 66 kDa. Other proteins in the fraction were not labeled. (C) Molec-
tal muscle ecto-ATPase (after digitonin solubilization) which
ular weight standards were also Coomassie stained.
was inhibited nearly 55% by concanavalin A treatment (32)
and frog muscle ecto-ATPase, which is also highly inhibited
by concanavalin A (33).
To study the distribution of the enzyme on the Tetrahy-
mena cell surface membrane, we used a procedure for the
isolation of pure ciliary membrane vesicles (CMVs, 24) from
Tetrahymena and a two-phase aqueous partitioning proce-
dure to separate plasma membranes from other contaminat-
ing membranes, successfully isolating body membranes vesi-
cles (BMVs, 23). This procedure yielded a body membrane
fraction that had an E-type ATPase detectable both with and
without membrane permeabilizing detergent. Detergent was
included in the in vitro assay because the presence of it did
not interfere with the enzyme activity and it also allowed
access of substrate into any inside-out membrane vesicles
that may have been present. Approximately 85% of the total
E-type ATPase of the body plasma membrane vesicles could
be detected in the absence of detergent (data not shown).
The body membrane vesicle fraction had a specific activity
FIG. 4. The purified soluble ecto-ATPase showed a sharp pH opti-
statistically higher than that of the purified ciliary mem-
mum near pH 7.5. Native gel-purified enzyme was tested in the in
brane fraction (tÅ9.52, Põ0.975, df Å1). If this accurately
vitro ATPase assay at different pH values, as described under Mate-
reflects the situation in vivo, the body may contain propor-
rials and Methods. The buffers used were sodium acetate–acetic acid
tionally more ecto-ATPase activity. It may be possible, how-
(pH 4.5–5.5), Mops– NaOH (pH 6.0 – 7.5), and Tris – HCl (pH 8.0 –
ever, that the method of preparation of CMVs reduced the
9.5). The enzyme displayed a narrow pH optimum, with maximal
activity at pH 7.5.
specific activity somewhat, as ciliary membrane preparation
358
SMITH, KIRLEY, AND HENNESSEY
Recently, we have found that the behavioral sensitivity of
Tetrahymena to extracellular nucleoside triphosphates (as
chemorepellents) is dependent upon the activities of the ex-
tracellular phosphohydrolyzing enzymes (5). While both Tet-
rahymena (4) and Paramecium (4, 34) show significant chem-
orepellent responses in micromolar to nanomolar concentra-
tions of external GTP, their behavioral responses to ATP
required millimolar concentrations. However, when external
ATP hydrolysis was inhibited either by removing divalent
cations or by using nonhydrolyzable analogs (
b
–
g
-methylene
ATP), Tetrahymena responded well to ATP at 50
m
M
and
methylene ATP at 50 n
M
(5). This suggests that these cells
respond to lower concentrations of GTP in the divalent ion-
containing solutions commonly used because the active ecto-
ATPase in vivo prefers hydrolysis of ATP over GTP almost
4:1. Therefore, the ecto-ATPase helps to set the selectivity of
the purinergic responses of the cell by preferential hydrolysis
of ATP. It also removes the purinergic signal, in a manner
similar to how acetylcholinesterase inactivates acetylcholine
signaling molecules.
The ecto-ATPase of Tetrahymena is not necessary for ex-
tracellular nucleotide scavenging because there are other
nonspecific external phosphatases that can hydrolyze
nTPs, nDPs, and nMPs (35, 36). Since the ecto-ATPase re-
quires relatively high concentrations of Ca
2/
or Mg
2/
for
FIG. 6. Antibodies which recognize the soluble form of the enzyme
activity, it is relatively inefficient at hydrolyzing extracel-
also react with the membrane-associated form of the ecto-ATPase.
lular nucleoside triphosphates in dilute solutions (such as
Ciliary membrane proteins were obtained by Triton X-114 phase
some pond waters). The fact that extracellular ATP and
partitioning of isolated whole cilia (13). The membrane proteins were
GTP would not only be in short supply in a pond but also
run in duplicate on a 7.5% SDS gel and electroblotted, as described
actively avoided (4, 5) further strengthens the hypothesis
under Materials and Methods. (A) One side of the membrane was
that this ecto-ATPase is more involved in the behavioral
Coomassie stained to show the molecular weight markers (lane 1)
responsiveness to purinergic signals than in extracellular
and to examine the distribution of ciliary proteins present (lane 2).
(B) The other side was used for Western blot analysis. The other
nucleotide scavenging.
strip of PVDF membrane was probed with antibody against the
chicken ecto-ATPase. A cross-reacting protein at 66 kDa was ob-
ACKNO WLEDGMENT S
served, the same molecular weight as that for the soluble form of
the enzyme. This work was supported by Grants NIH R01 AR38576, NIH K04
AR01841 and American Heart Association Grant 96013960 to T.L.K.
and NSF MCB 9410756 to T.M.H.
involves a vigorous vortexing for 2–3 min to strip the mem-
branes from the axonemes. Therefore, the small but signifi-
cant difference between the two membrane domains may not REFERENCES
necessarily reflect the situation in vivo.
1. Plesner, L. (1995) Int. Rev. Cytol. 158, 141 – 214.
Polyclonal antibodies obtained against the chicken ecto-
2. Seguela, P., Haghighi, A., Soghomonian, J. J., and Cooper, E.
ATPase (25) cross-reacted with the soluble form of the en-
(1996) J. Neurosci. 16, 448–455.
zyme of Tetrahymena and were used to examine the localiza-
3. Harden, T. K., Boyer, J. L., and Nicholas, R. A. (1995) Annu. Rev.
tion of the enzyme in purified membrane fractions. When
Pharmacol. Toxicol. 35, 541–579.
purified body and ciliary membrane vesicles were probed
4. Francis, J. T., and Hennessey, T. M. (1995) J. Euk. microbiol.
with these antibodies, which clearly recognize the soluble 66-
42, 78–83.
kDa ecto-ATPase, a 66-kDa cross-reacting protein was ob-
served in both fractions. Hence, based upon the specific activi-
5. Smith, T. M., Kim, M., Kirley, T., and Hennessey, T. M. (1996)
in Ecto-ATPases: Recent Progress on Structure and Function
ties of the purified membranes and the presence of the immu-
(Plesner, L., Kirley, T., and Knowles, A., Eds.), Plenum Press,
noreactive protein in both the body and ciliary membrane
New York (in press).
fractions, we can conclude that the enzyme is not selectively
6. Dhalla, N. S., and Zhao, D. (1988) Prog. Biophys. Mol. Biol. 52,
localized to only one membrane domain.
1–37.
Ecto-ATPases have eluded purification for decades due to
7. Banerjee, R. K. (1981) Mol. Cell. Biochem. 37, 91–99.
their low abundance, sensitivity to many common detergents,
and copurification with other proteins. The fact that a soluble
8. Bhargava, T., Datta, S., Ramachandran, V., Ramakrishnan, R.,
form of the enzyme can be purified in the absence of deter-
Roy, R. K., Sankaran, K., and Subrahmanyam, V. B. K. (1995)
Curr. Sci. 68, 293–300.
gents allows for an easier and more effective purification.
This should aid in the study and possible elucidation of the
9. Champagne, D. E., Smartt, C. T., Ribeiro, J. M. C., and James,
A. A. (1995) Proc. Natl. Acad. Sci. USA 92, 694 – 698.
endogenous function of the enzyme.
359
SOLUBLE ECTO-ATPase OF Tetrahymena
10. Komoszynski, M. A. (1996) Comp. Biochem. Physiol. B 113, 581 – 23. Smith, T. M., and Hennessey, T. M. (1993) Anal. Biochem. 210,
299–308.591.
24. Adoutte, A., Ramanathan, R., Lewis, R. M., Dute, R. R., Ling, K-
11. Riddle, L. M., Rauh, J. J., and Nelson, D. L. (1982) Biochim. Bio-
Y., Kung, C., and Nelson, D. L. (1980) J. Cell Biol. 84, 717– 738.
phys. Acta 688, 525–540.
25. Stout, J. G., and Kirley, T. L. (1996) Biochemistry 35, 8289–
12. Levin, A. E., Travis, S. M., DeVito, L. D., Park, K. A., and Nelson, 8298.
D. L. (1989) J. Biol. Chem. 264, 4544 – 4550.
26. Harlow, E., and Lane, D. (1988) Antibodies: A Laboratory Man-
13. Dentler, W. L. (1988) J. Cell Biol. 107, 2679–2688. ual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
14. Chua, B., Elson, C., and Shrago, E. (1977) J. Biol. Chem. 252, 27. Matsudaira, P. (1987) J. Biol. Chem. 262, 10035 – 10038.
7548–7554. 28. Stout, J. G., and Kirley, T. L. (1994) J. Biochem. Biophys. Meth-
15. Burnasheva, S. A., Fain, F. S., Kireeva, I. Y., and Poglazov, B. F. ods 29, 61–75.
(1989) Biokhimiya 54, 788–795. 29. Lin, S.-H., and Russell, W. E. (1988) J. Biol. Chem. 263, 12253–
16. Doughty, M. J. (1978) Comp. Biochem. Physiol. B 60, 339–345. 12258.
17. Gaertig, J., and Gorovsky, M. A. (1995). Methods Cell Biol. 47, 30. Dentler, W. L. (1992) J. Protozool. 39, 368 – 378.
559–569. 31. Smith, T. M., and Hennessey, T. M. (1995) Mol. Biol. Cell. 6,
18. Kung, C. (1971) Genetics 69, 29 – 45. 180a.
19. Ausubel, F. M. (1987) Current Protocols in Molecular Biology, 32. Kirley, T. L. (1988) J. Biol. Chem. 263, 12682 – 12689.
Vol. I. Wiley, New York. 33. Still, J. S., Manery, J. F., and Dryden, E. E. (1986) Biochem. Cell
20. DeMoreno, M. R., Smith, J. F., and Smith, R. V. (1985) Anal. Biol. 64, 1272 – 1280.
Biochem. 151, 466–470. 34. Clark, K. D., Hennessey, T. M., and Nelson, D. L. (1992) Proc.
Natl. Acad. Sci. USA 90, 3782–3786.
21. Lanzetta, P. A., Alvarez, L. J., Reinach, P. S., and Candia, O. A.
(1979) Anal. Biochem. 100, 95–97. 35. Kiy, C., Vosskuhler, L., and Tiedtke, A. (1993) Exp. Cell Res.
205, 286–292.
22. Picher, M., Cote, Y. P., Beliveau, R., Potier, M., and Beaudoin,
A. R. (1993) J. Biol. Chem. 268, 4699 –4703. 36. Tiedtke, A., and Rasmussen, L. (1989) J. Protozool. 36, 378 – 382.
