|
|
|
|||||||
|
|||||||||
(Received for publication, June 15, 1995; and in revised form, August 8, 1995) From the
Pharmacologically active compounds with an imidazoline and/or
guanidinium moiety are recognized with high affinity by a family of
membrane-bound proteins collectively known as imidazoline binding sites
or imidazoline/guanidinium receptive sites. Two such receptive sites
may correspond to imidazoline binding domains identified on the A and B
isoforms of monoamine oxidase (MAO), but the detection of monoamine
oxidase isoforms in multiple tissues contrasts with the restricted
expression of imidazoline-binding proteins. To address these issues,
we determined the relationship between monoamine oxidase isoforms and
subtypes of imidazoline-binding proteins in human tissues known to
express one or both isoforms of MAO.
2-(3-Azido-4-[ However,
the imidazoline binding domain was not detected in platelet membrane
preparations containing amounts of MAO-B equivalent to those in the
photolabeled liver membranes indicating that recognition of this domain
is tissue-restricted. Restricted access to the imidazoline binding
domain on platelet MAO-B was not altered by membrane washing with 500
mM KCl or by solubilization and partial purification of the
enzyme suggesting that there are distinct subpopulations of MAO.
Identification of a binding domain on MAO that recognizes this class of
pharmacologically active compounds suggests a novel mechanism for
regulation of substrate oxidation/selectivity or that the enzyme may
subserve an as yet undefined function.
Imidazoline/guanidinium receptive sites or imidazoline binding
sites are defined as the nonadrenergic receptor binding sites for a
group of structurally related compounds containing imidazoline or
guanidinium moieties. Although many of these compounds interact with
adrenergic receptors, they also produce ill-defined effects on ion
transport, insulin secretion, and blood pressure regulation that are
mediated by interactions with multiple, pharmacologically distinct
imidazoline-binding proteins (1, 2, and references therein). ( The imidazoline-binding protein in liver is predominantly localized
to mitochondrial membranes and cannot be separated from the
mitochondrial enzyme monoamine oxidase (MAO, ( If indeed MAO and subtypes of
imidazoline-binding proteins are identical, then one must also explain
the apparent discrepancy that exists in the tissue localization of the
two entities. Whereas MAO is widely distributed, members of the family
of imidazoline-binding proteins exhibit a more restricted expression.
Definition of the relationship between MAO and imidazoline-binding
proteins is also complicated by differences in the stoichiometry of the
two entities in tissues where both proteins are apparently expressed
suggesting that either there are subpopulations of monoamine oxidase
that do not bind imidazoline/guanidinium ligands or that the
imidazoline binding domain is not accessible in all tissues. In
addition, it is difficult to explain the functional effects of these
compounds based solely upon their interaction with monoamine oxidases
suggesting that there are imidazoline-binding proteins that are not
monoamine oxidases. To address these issues, we utilized a
photoaffinity probe ([
Figure 1:
Monoamine oxidase isoforms and
imidazoline/guanidinium receptive sites in human placental and liver
mitochondrial membranes. Monoamine oxidase isoforms and
imidazoline/guanidinium receptive sites in human placental and human
liver mitochondrial membranes were identified by immunoblotting or
photoaffinity labeling, respectively. In each experiment, membranes
were solubilized and electrophoresed under denaturing conditions on 8%
SDS-polyacrylamide gels. Autoradiographs were obtained by exposing the
dried gels at -70 °C for 5-7 days. The migration of
midrange molecular weight standards is indicated by the numbers to the left of the immunoblot or autoradiographs (M
Peptides with apparent molecular weights similar to
that of the MAO isoforms were also covalently labeled in the two human
tissues with the photoaffinity adduct,
[
Figure 2:
Immunoprecipitation of
photoaffinity-labeled human placental and human liver mitochondrial
membranes. Human liver mitochondrial (400 µg of membrane protein)
and human placental (250 µg of membrane protein) membranes were
photolabeled with [
Figure 3:
Relationship between binding sites for
imidazoline or guanidinium ligands and monoamine oxidase inhibitors.
Placental membranes (500 µg of membrane protein) (A) and
liver mitochondrial membranes (100 µg of membrane protein) (B) were photolabeled with [
The
ligand recognition properties of the [
Figure 4:
Relationship between
imidazoline/guanidinium receptive sites and MAO-B in platelet and liver
mitochondrial membranes. A, immunoblot of MAO-B in platelet
and liver mitochondrial membranes. Platelet membranes (300 µg of
membrane protein) and liver mitochondrial membranes (50 µg of
membrane protein) were solubilized, electrophoresed on a 10%
SDS-polyacrylamide gel, and transferred to membranes. The blot was
probed with the anti-MAO-B monoclonal antibody MAO-B-1C2 (1:1000), and
immunoreactive proteins were identified as described under
``Experimental Procedures.'' The arrow indicates the
migration of the immunoreactive protein with an apparent molecular
weight of
A similar lack
of availability of the imidazoline binding domain on MAO-B in platelet
membranes was observed using the imidazoline ligand
[ As an
initial approach to determine the cause of the limited access to the
imidazoline binding domain on platelet MAO-B, experiments were
performed to address the possibility that this domain is masked in the
platelet membrane environment. In the first series of experiments,
liver mitochondria and platelet membranes were mixed and incubated for
10 min at 24 °C prior to [
Figure 5:
Accessibility of the imidazoline binding
domain in liver mitochondria and platelet membranes. A,
autoradiograph of [
In a second
approach to address this issue, solubilized membrane proteins from
liver and platelet were fractionated by chromatofocusing. Proteins were
eluted with a pH gradient from pH 7 to 4, and the peak of MAO-B
immunoreactivity was detected in the pH 5.4-5.8 fractions for
both the liver and platelet preparations (Fig. 6, A and B). This pH range is similar to the previously determined pI
of
Figure 6:
Partial purification and photolabeling of
liver and platelet MAO-B. Solubilized extract from liver membranes (3.5
mg of protein in 2.5 ml) (A) or platelet membranes (100 mg in
100 ml) (B) was fractionated using a chromatofocusing matrix
(bed volumes of 5 ml and 17 ml, respectively). Absorbed proteins were
eluted with a pH gradient of pH 7-4, and protein was monitored by
absorbance at 280 nm (
Although many of the cellular effects of
imidazoline/guanidinium compounds are mediated by known
neurotransmitter receptor systems (i.e. adrenergic), some of
their effects apparently involve interaction with a family of
imidazoline binding sites of unknown identity. The present study
indicates that two members of this protein family identified in human
placenta and liver are identical with the A and B isoforms of the
enzyme monoamine oxidase, respectively. The identities of additional
imidazoline binding proteins that differ in their ligand recognition
properties and subcellular distribution have yet to be
determined(8, 26, 27, 28, 29) . Members of the family of imidazoline binding proteins are subtyped
as I MAO isoforms are
identified in a wide variety of tissues by enzyme activity,
immunoreactivity, and the use of radiolabeled enzyme
inhibitors(31) . In contrast, the tissue distribution of
imidazoline binding sites is more restricted. Such a discrepancy exists
in human platelet membranes, which express both immunoreactive and
functional MAO-B that is poorly recognized by the photoaffinity adduct
[ Although analysis of cDNA clones
encoding monoamine oxidase B isolated from various tissues do not
indicate additional sequence diversity(18, 32) , the
enzyme genes are complex and consist of multiple exons. Alternative
splicing could result in tissue-specific expression of MAO subtypes
containing the imidazoline binding domain. Thus, in tissues where
detection of the imidazoline binding site is limited, such as
platelets, the MAO population may consist predominantly of enzyme
subtypes lacking this domain. MAO also undergoes several
post-translational modifications including the covalent attachment of
the flavin cofactor and the formation of several disulfide bonds which
are required for enzyme activity. Additionally, there is a consensus
site for N-linked glycosylation at amino acids 181 and 145 of
MAO-A and -B, respectively, but the proteins appear not to be
glycosylated(18, 33) . Possibilities 3 and 4 are of
note for several reasons, particularly the large gain in detectable
imidazoline binding sites observed during two chromatographic steps
used for the purification of rabbit kidney imidazoline-binding
protein(16) . Such an observation may be due to the separation
of the imidazoline-binding protein from an associated protein or a
small, endogenous organic ligand. The latter possibility is in line
with the demonstration of endogenous substances such as
clonidine-displacing substance or agmatine which are postulated to be
endogenous ligands for members of the family of imidazoline-binding
proteins(3, 4, 5, 6, 7) .
This point also parallels several reports indicating the existence of
endogenous substances that regulate MAO activity (34, 35, 36, 37) . However, the
imidazoline binding domain on platelet MAO-B remains inaccessible
following attempts to remove associated substances by high salt washes,
removal of the enzyme from its membrane environment by detergent
solubilization, and partial purification of the enzyme by
chromatofocusing. These data suggest that the differential recognition
of the imidazoline binding domain on MAO-B in liver and platelet is due
to structural differences in the enzyme itself. Imidazoline/guanidinium compounds are not apparent substrates for
MAO and do not compete for binding of radiolabeled inhibitors to the
enzyme (21) . Similarly, MAO inhibitors that bind to the enzyme
active site do not inhibit labeling of the enzyme in human liver and
placenta by the photoaffinity adduct [ Although there is clearly an imidazoline binding
domain on the enzyme, the consequences of occupation of this site on
enzyme activity or substrate selectivity are not clear. Relatively high
concentrations of imidazoline/guanidiniums are required to observe an
effect on enzyme activity(9, 20) . Enzyme activity is
noncompetitively inhibited by certain imidazolines at ligand
concentrations 100-1000-fold higher than their K
Volume 270,
Number 46,
Issue of November 17, 1995 pp. 27961-27968
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
RELATIONSHIP TO SUBTYPES OF IMIDAZOLINE-BINDING PROTEINS AND
TISSUE-SPECIFIC INTERACTION OF IMIDAZOLINE LIGANDS WITH MONOAMINE
OXIDASE B (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
I]iodophenoxy)methylimidazoline
([
I]AZIPI), a photoaffinity adduct that
selectively labels imidazoline-binding proteins, photolabeled an M
= 59,000 peptide in liver and an M
= 63,000 peptide in placenta,
consistent with the M
of the MAO isoforms
identified by immunoblots in these tissues. The photolabeled species in
liver was immunoprecipitated with MAO-B selective antibodies, whereas
the photolabeled species in placenta was immunoprecipitated by MAO-A
selective antibodies consistent with the isoform of MAO predominantly
expressed in these tissues. The imidazoline/guanidinium ligands
interact with the enzyme at a site distinct from the substrate
recognition domain, and the immunoprecipitated peptides in liver and
placenta display distinct ligand recognition properties consistent with
those reported for subtypes of imidazoline binding sites.
)The imidazoline/guanidinium receptive sites also recognize
endogenous substances that mimic some of these
effects(3, 4, 5, 6, 7) . 
)EC 1.4.3.4)
during purification, suggesting a potential relationship between these
two entities(8) . Partial amino acid sequencing of a purified
rabbit kidney imidazoline binding protein revealed high sequence
similarity to monoamine oxidases of other species whose sequences are
known, and heterologous expression of human MAO-A or MAO-B indicates
that both isoforms recognize the imidazoline,
[
H]idazoxan, a ligand commonly used to identify
imidazoline binding sites in various tissues(9) . However, the
affinity of the enzymes for [H]idazoxan following
expression in Saccharomyces cerevisiae is 10-50-fold
lower than that expected for the imidazoline binding site (9) ,
and some members of the imidazoline binding protein family do not
recognize this radioligand when evaluated in their natural environment
within the cell(10) .I]AZIPI) to label the
imidazoline-binding proteins in different tissues and monoclonal
antibodies that selectively recognize the MAO isoforms to
immunoprecipitate the labeled membrane proteins. We report that the
[
I]AZIPI-labeled species in human liver
mitochondria and placenta are identical with the MAO isoforms expressed
in the two tissues. The imidazoline/guanidinium ligands interact with
MAO isoforms at a site distinct from the substrate binding domain and
may represent a novel site for MAO regulation. The interaction of such
ligands at the imidazoline binding domain of the B isoform of MAO was
not observed in platelets suggesting that availability of this site is
tissue-selective. The differential photolabeling of liver and platelet
MAO-B was maintained following washing of the membranes in high salt
buffer or solubilization and partial purification of the enzymes from
the two tissues.
Materials
Renaissance Western blot
Chemiluminescence Reagent kit, [
I]sodium
iodide, and [
H]pargyline (36 Ci/mmol) were
purchased from DuPont NEN, and [H]idazoxan (41
Ci/mmol) was purchased from Amersham. Idazoxan, clonidine, guanabenz,
clorgyline, deprenyl, and pargyline were provided by Research
Biochemicals Int. (Natick, MA), and cirazoline was a gift from
Synthelabo (Paris, France). Ascites fluids containing monoclonal
antibodies to human MAO-A and -B (11, 12) were kindly
provided by Dr. Richard Denney (University of Texas Medical Branch,
Galveston, TX). Horseradish peroxidase-conjugated secondary antibodies
were purchased from Chemicon (Temecula, CA). GammaBind
G-Sepharose
brand protein G-Sepharose beads and
Polybuffer exchanger 94 resin were purchased from Pharmacia Biotech
Inc. Polyvinylidene difluoride membranes were purchased from Gelman
Sciences (Ann Arbor, MI). Kodak XAR 5 film was purchased from
Chesapeake X-Ray (Florence, SC). Acrylamide, bisacrylamide, SDS, Tween
20, and prestained midrange molecular weight protein standards were
purchased from Bio-Rad. Ecoscint A scintillation fluid was purchased
from National Diagnostics (Atlanta, GA). Centricon-10 centrifugal
concentrators were purchased from Amicon Inc. (Beverly, MA).
Membrane Preparation
Full-term human placenta were
collected within 2 h of delivery and washed in 100 mM Tris-HCl, pH 7.4, at 4 °C. Tissue samples were homogenized in
lysis buffer (5 mM Tris-HCl, pH 7.5, 5 mM EDTA, 5
mM EGTA, and 0.1 mM phenylmethylsulfonyl fluoride
(PMSF), 4 °C) and filtered through two layers of gauze mesh. A low
speed pellet was generated by centrifugation at 1000 
g for 10 min at 4 °C. The resulting supernatant was centrifuged
at 35,000 g for 10 min at 4 °C, and this high
speed pellet was washed twice in membrane buffer (50 mM Tris,
pH 7.5, 0.6 mM EDTA, 5 mM MgCl, 0.1
mM PMSF), rapidly frozen, and stored at -70 °C for
up to 2 months. A fraction of human liver enriched in mitochondria was
obtained as described previously(13) . Outdated human platelets
were collected by centrifugation at 800 g for 30 min
at 4 °C in the presence of 2 mM EDTA and 10 µM indomethacin, resuspended, and frozen in 5 mM Tris-HCl,
pH 7.4 containing 2 mM EDTA, 10 µM indomethacin,
and 75 µg/ml PMSF. Platelets were lysed by freeze-thawing twice in
lysis buffer and homogenized using a glass Dounce. Membranes were
pelleted by centrifugation at 35,000 g for 10 min at 4
°C, washed, and resuspended in membrane buffer.Photoaffinity Labeling
The cirazoline derivative,
2-(3-amino-4-iodophenoxy)methylimidazoline was synthesized, iodinated,
and converted to the photolabile azide
([I]-AZIPI) for use as a photoaffinity adduct
as described previously(2, 10) . Membranes were
incubated in reduced light with 1-2 nM
[
I]AZIPI for 30 min at 24 °C, chilled on
ice, and diluted 10-fold with 4 °C membrane buffer containing 2
mM dithiothreitol immediately prior to photolyzing at 4 °C
for 5 min in a Ray-O-Vac photolysis chamber (320 nm). Membranes were
pelleted in a microcentrifuge, solubilized in loading buffer (100
mM Tris-HCl, pH 6.8, containing 1% SDS, 50% glycerol, 25%
-mercaptoethanol, and bromphenol blue) at 100 °C for 5 min and
subjected to SDS-polyacrylamide gel electrophoresis(14) .
Competing ligands were preincubated with the membranes for 5-10
min at room temperature. Membrane protein was measured by the method of
Lowry et al.(15) .Immunoblots
Membrane proteins were solubilized in
loading buffer as above and subjected to SDS-polyacrylamide gel
electrophoresis. Proteins were transferred to polyvinylidene difluoride
membranes with a semidry electroblotter (Integrated Separation Systems,
Hyde Park, MA) for 1 h at 200 mA. The blots were blocked with 5% nonfat
dried milk in wash buffer (20 mM Tris-HCl, pH 7.6, 165 mM NaCl, and 0.2% Tween 20) for 1 h, washed twice, and incubated for
1 h at 24 °C with primary antibodies diluted in wash buffer. After
washing, blots were incubated with goat anti-mouse IgG horseradish
peroxidase-conjugated secondary antibody diluted 1:10,000 in wash
buffer for 20 min at 24 °C. The horseradish peroxidase reaction was
visualized by Renaissance Western blot Chemiluminescence
Reagent (DuPont NEN) and exposure to Kodak XAR-5 film.
Immunoprecipitation
Membranes were photolabeled
and resuspended in 100 µl of solubilization buffer (membrane buffer
with 1% Triton X-100 and 0.05% SDS). The soluble extract was precleared
with 10 µl (packed volume) of Sepharose G beads and incubated with
the indicated ascites fluid (1:100 final dilution) containing
antibodies to MAO-A or MAO-B for 12-16 h at 4 °C. The
antibody-antigen complexes were precipitated by the addition of 10
µl (packed volume) of Sepharose G beads and brief centrifugation in
a microcentrifuge. The resulting pellet was washed in 500 µl of
membrane buffer, solubilized in loading buffer at 100 °C for 10
min, and subjected to SDS-polyacrylamide gel electrophoresis.Radioligand Binding
Membranes were diluted in
membrane buffer and preincubated with competing ligands for 10 min at
24 °C. [H]Pargyline (100 nM) or
[H]idazoxan (3-150 nM) was added,
and the samples were incubated with shaking at 24 °C for 30 min.
[H]Idazoxan binding was performed in the presence
of 10 µM rauwolscine, and nonspecific binding was
determined in the presence of 10 µM cirazoline. In
[H]pargyline binding studies, nonspecific binding
was determined in the presence of 100 µM pargyline.
Membranes were collected by vacuum filtration on glass fiber filters
(No. 32, Schleicher & Schuell) and washed in 4 3 ml of 100
mM Tris, pH 7.4, 4 °C. Radioactivity retained on the
filters was determined by liquid scintillation spectrometry with
approximately 50% efficiency.Solubilization and Partial Purification of MAO-B by
Chromatofocusing
Membranes were solubilized and the enzyme was
partially purified using a chromatofocusing matrix as described
previously(16) . Briefly, human liver (9 mg of protein) and
platelet membranes (300 mg of protein) were resuspended in
solubilization buffer containing MOPS (25 mM), pH 7, EGTA (2
mM), PMSF (1 mM), and 1% digitonin at a
detergent:protein ratio of 3:1. The mixture was incubated at 4 °C
for 30 min, and the soluble fraction was collected following
centrifugation at 100,000 g for 60 min at 4 °C.
The chromatofocusing matrix (Polybuffer Exchanger 94) was equilibrated
with 0.1% digitonin in MOPS (25 mM), pH 7, EGTA (2
mM), PMSF (1 mM), and solubilized material was
applied at 0.4 ml/min. Proteins were eluted with Polybuffer 74 (diluted
1:8), pH 4.0, containing digitonin (0.1%), EGTA (2 mM), and
PMSF (0.1 mM). Fractions were collected and protein was
monitored by optical density at 280 nm. The pH of fractions containing
MAO-B was adjusted to 7.4 prior to concentration by Centricon centrifugal concentrators and subsequent photolabeling. To remove
free [
I]AZIPI from the solubilized
preparations, the samples were desalted on Sephadex G-25 gel filtration
columns prior to photolysis. The eluate was photolyzed in the presence
of 2 mM dithiothreitol and concentrated by lyophilization
prior to addition of loading buffer and gel electrophoresis.
Identification of MAO Isoforms and Photoaffinity
Labeling of Imidazoline-binding Proteins in Human Liver Mitochondria
and Human Placenta
As a first step to define the relationship
between MAO and imidazoline binding sites, we characterized the
proteins in human liver and placenta, tissues that express both the MAO
isoforms and imidazoline-binding proteins. As described previously,
human placenta expressed predominantly MAO-A(11) , and human
liver mitochondria expressed high levels of the MAO-B isoform (17) (Fig. 1A). Immunoblots using antibodies
selective for human MAO-A (MAO-A-4D3 ascites)(11) , or MAO-B
(MAO-B-1C2 ascites) (12) , identified a single protein with
apparent molecular weight of 63,000 in human placental membranes
and a peptide with apparent molecular weight of
59,000 in human
liver mitochondrial membranes, respectively. (
)The apparent
molecular weights of the MAO isoforms were similar to previously
reported estimates based on gel mobility and the calculated molecular
weights of 59,700 and 58,800 determined from the nucleotide sequences (18) .
10). The arrows indicate the migration of proteins with apparent molecular weights
of
63,000 and
59,000. A, identification of MAO-A and
MAO-B by immunoblotting. Nitrocellulose transfers of polyacrylamide
gels of human and rat liver mitochondria (200 µg of membrane
protein) and human placenta (350 µg of membrane protein) were
immunoblotted with MAO-A-4D3 (1:200) or MAO-B-1C2 ascites (1:1000) as
described under ``Experimental Procedures.'' Immunoreactive
proteins were identified by a chemiluminescent reaction with
horseradish peroxidase-conjugated secondary antibodies and subsequent
exposure to film. Similar results were obtained in three experiments. B, autoradiograph of photoaffinity-labeled human placental and
human liver mitochondrial membranes. Human placental membranes (400
µg of membrane protein) and human liver mitochondrial membranes (50
µg of membrane protein) were photolabeled with
[
I]AZIPI (1.4 nM) in the presence or
absence of 10 µM cirazoline. Similar results were obtained
in three experiments.
I]AZIPI, that selectively labels
imidazoline-binding proteins (Fig. 1B). Photolabeling
of the M
= 59,000 and
63,000
peptides was blocked by cirazoline, an imidazoline that exhibits high
affinity for members of the family of imidazoline-binding proteins.
Thus, the apparent molecular weight of the imidazoline-binding protein
in human placental and human liver mitochondrial membranes corresponds
to that of the MAO-A or -B isoform predominantly expressed in these
tissues.
Immunoprecipitation of
[
To determine whether the
ligand binding subunit of imidazoline-binding proteins in human liver
mitochondrial and placental membranes was identical to MAO, we
attempted to immunoprecipitate the
[I]AZIPI-labeled Proteins from Liver
Mitochondrial and Placental Membranes
I]AZIPI-labeled species with monoclonal
antibodies selective for the MAO-A and -B isoforms (MAO-A-4D3 and
MAO-B-1C2 ascites fluids)(19) . The species labeled by
[
I]AZIPI in placenta (M
= 63,000) or liver mitochondria (M
= 59,000) were immunoprecipitated by MAO-A-4D3 and
MAO-B-1C2 ascites, respectively (Fig. 2A). The
selectivity of immunoprecipitation of the labeled species was tested by
immunoprecipitating aliquots of
[
I]AZIPI-labeled membranes from the tissues
with both ascites fluids (Fig. 2B). The photolabeled
species in the placental membranes was not immunoprecipitated with the
MAO-B-1C2 ascites, nor was the photolabeled species in liver
mitochondrial membranes immunoprecipitated by the MAO-A-4D3 ascites (Fig. 2B). The efficiency of immunoprecipitation for
monoamine oxidase A or B was
95% and 50%, respectively, and
paralleled the relative distribution of photolabeled peptides in the
supernatant and immunoprecipitation pellet. (
)Thus, the
[I]AZIPI-labeled imidazoline-binding proteins
were selectively immunoprecipitated by antibodies recognizing the
isoform of MAO predominantly expressed in each tissue. The apparent
molecular weight of the immunoprecipitated photolabeled peptide was
directly related to that of the MAO isoform present in the tissue,
suggesting that the photolabeled species are indeed MAO and not a
co-immunoprecipitated protein. The ligand binding subunits of
imidazoline-binding proteins in human placental and in human liver
mitochondrial membranes were therefore identified as the isoforms of
the mitochondrial enzyme, MAO.
I]AZIPI (0.9 nM).
In A, the MAO isoform present in the membranes was
immunoprecipitated with MAO-A-4D3 ascites (placenta) or MAO-B-1C2
ascites (liver mitochondria) both at 1:100 final dilution. These
results are representative of 2-3 experiments. In separate
experiments (B), both MAO-A-4D3 ascites and MAO-B-1C2 ascites
(1:100) were used to immunoprecipitate the MAO isoforms present in each
tissue.
Identification of Multiple Ligand Binding Domains on
MAO
Various imidazoline/guanidinium ligands interact poorly with
the MAO active
site(20, 21, 22, 23) . To determine
whether [I]AZIPI photoincorporates at the
enzyme active site, membranes were photolabeled with
[
I]AZIPI in the presence of the
mechanism-based, irreversible MAO inhibitors pargyline, clorgyline, and
deprenyl (Fig. 3, A and B). Saturating
concentrations of the MAO inhibitors did not eliminate photoaffinity
labeling of the imidazoline-binding proteins in placental or liver
mitochondrial membranes. Similarly, imidazoline or guanidinium ligands
did not compete for binding of pargyline, an inhibitor of both MAO-A
and -B (Fig. 3C). [
H]Pargyline
binding in placental membranes was inhibited by the MAO-A selective
ligand, clorgyline, but not by the MAO-B selective ligand, deprenyl. In
liver mitochondrial membranes, [H]pargyline
binding was inhibited by deprenyl, but not by clorgyline as expected
for a tissue expressing predominantly MAO-B (Fig. 3C).
However, the imidazoline ligand cirazoline (1 µM), which
effectively blocks [I]AZIPI photoincorporation,
did not inhibit [
H]pargyline binding in any of
the tissues examined (Fig. 3C). These results indicate
that [I]AZIPI incorporates into a site on MAO
distinct from the binding site of the acetylenic MAO inhibitors.
I]AZIPI (2
nM) in the presence of buffer, 1 µM cirazoline, 1
µM idazoxan, 10 µM amiloride, 10 µM clonidine, 10 µM rauwolscine, 100 µM
pargyline, 0.1 µM clorgyline, or 1 µM deprenyl. Following electrophoresis on an 8% polyacrylamide gel
under denaturing conditions, autoradiographs were obtained by exposing
the dried gels at -70 °C for 1 (A) or 7 days (B). The migration of midrange molecular weight standards are
indicated by the numbers to the left of the
autoradiographs (M
10).
The arrows indicate the migration of the photolabeled proteins
with apparent molecular weights of
63,000 (A) and
59,000 (B). C, identification of MAO-A in
placental membranes and MAO-B in liver mitochondrial membranes using
the MAO inhibitor [
H]pargyline. Binding of
[H]pargyline (100 nM) was determined in
100 µg of liver mitochondrial membrane protein or 400 µg of
placental membrane protein in the presence of buffer (TB), 100
µM pargyline (parg), 0.1 µM clorgyline (clor), 1 µM deprenyl (depr), or 1 µM cirazoline (ciraz) in
duplicate as described under ``Experimental Procedures.''
Total bound radioactivity for liver and placenta was 4,900 cpm and
15,600 cpm, respectively, and was measured with approximately 50%
efficiency. The data are expressed as percent total binding (TB) and are representative of the results of two
experiments.
I]AZIPI
photoincorporation sites on MAO-A and -B were examined to determine the
relationship between these sites and imidazoline-binding proteins
identified in other tissues (Fig. 3, A and B).
Photolabeling in both liver mitochondrial and placental membranes was
equally sensitive to inhibition by cirazoline; however, the recognition
of idazoxan and amiloride varied in the two tissues. Photoaffinity
labeling of the imidazoline-binding protein in human placenta was more
sensitive to competition by the guanidinium ligand, amiloride, while
the imidazoline compound, idazoxan, competed more effectively in liver (Fig. 3). These results indicate that the pharmacological
heterogeneity of imidazoline-binding proteins may reflect the presence
of different MAO isoforms.
Photolabeling of MAO-B from Human Liver Mitochondria and
Human Platelets
Although the preceding results indicate the
identity of MAO and subtypes of imidazoline-binding proteins, there is
an apparent discrepancy between the tissue distribution of the two
entities. The quantitative relationship between MAO and
imidazoline-binding proteins was thus further investigated in two
tissues that express high levels of MAO-B, liver mitochondria and
platelets. The relative amounts of MAO-B present in liver
mitochondrial, and platelet membranes were determined by immunoblotting
using MAO-B-selective monoclonal antibodies (MAO-B-1C2 ascites).
Similar levels of MAO-B immunoreactivity were identified in 50 µg
of membrane protein from liver mitochondrial membranes and 300 µg
of membrane protein from platelet membranes (Fig. 4A).
Photolabeling was therefore compared in aliquots of liver mitochondrial
and platelet membranes containing similar amounts of MAO-B.
[I]AZIPI photoincorporated into a M
= 59,000 peptide in liver
mitochondrial membranes but not in platelet membranes (Fig. 4B). The relative amounts of MAO-B in the two
tissues were also determined using the MAO radioligand,
[
H]pargyline (Fig. 4C). Binding
of this mechanism-based inhibitor indicated the presence of active
enzyme in both platelet and liver mitochondrial membranes suggesting
that the site of [I]AZIPI photoincorporation on
MAO-B is differentially accessible in these tissues.
59,000 in both platelet and liver mitochondrial
membranes. B, autoradiograph of
[
I]AZIPI-labeled species in liver mitochondrial
and platelet membranes. Platelet membranes (300 µg of total
protein) and liver mitochondrial membranes (50 µg of total protein)
were photolabeled with [
I]AZIPI (1.4
nM) in the presence or absence of 10 µM cirazoline. Following electrophoresis on a 10% SDS-polyacrylamide
gel, autoradiographs were obtained by exposing the dried gels at
-70 °C for 11 days. A photolabeled species in platelet
membranes was detected with longer exposure. The arrow indicates the migration of the photolabeled protein with M
of 59,000. In A and B,
the migration of midrange molecular weight standards are indicated by
the numbers to the left of the immunoblot or
autoradiograph (M
10). C, [
H]pargyline binding in liver
mitochondrial and platelet membranes.
[H]Pargyline (100 nM) was incubated with
aliquots of liver mitochondrial (100 µg of membrane protein) or
platelet membranes (600 µg of membrane protein) in the presence of
buffer (total binding) or 100 µM pargyline (nonspecific
binding). Specific binding = total binding minus nonspecific
binding. Nonspecific binding represented 4% and 6% of total binding for
liver and platelets, respectively. Bound radioactivity was measured
with approximately 50% efficiency. Data represent the average ±
range for two experiments. D,
[H]idazoxan binding in liver mitochondrial and
platelet membranes. Increasing concentrations of
[H]idazoxan (3-150 nM) were
incubated with aliquots of liver mitochondrial (, 150 µg of
membrane protein) or platelet (
, 500 µg of membrane protein)
membranes in the presence of 10 µM rauwolscine.
Nonspecific binding was determined in the presence of 10
µM cirazoline. At radioligand concentrations near the K, specific binding represented 90% of total binding in liver
and 32% of total binding in platelets. Data are representative of two
experiments performed in duplicate.
H]idazoxan (Fig. 4D). The
relative amounts of MAO-B in liver mitochondria and platelet membranes
were determined by both immunoblotting and
[H]pargyline binding, and membrane aliquots
containing similar amounts of MAO-B were incubated with increasing
concentrations of [H]idazoxan. The
-adrenergic receptor antagonist rauwolscine (10
µM) was included in the incubation buffer to prevent
ligand interaction with -adrenergic receptors.
[H]Idazoxan exhibited similar affinities for the
imidazoline binding sites in these two preparations (K
10-20 nM); however, the binding capacity of
the liver membranes was greater than that of the platelet membranes.
These data indicate that the lack of photoaffinity labeling of MAO-B in
platelet membranes is not due to a difference in affinity for
imidazoline ligands, but rather to a lower binding capacity.
I]AZIPI labeling
to determine whether the imidazoline binding domain on platelet MAO-B
was masked by a diffusable substance. The presence of platelet
membranes did not inhibit the photolabeling of liver mitochondria (Fig. 5A). Next, we removed peripheral membrane
proteins and small molecular weight substances present in the crude
membrane preparations by washing the membranes in buffer containing 500
mM KCl, a cation that allosterically increases the
dissociation rate of imidazoline/guanidinium ligands from imidazoline
binding sites(24, 25) . However, the availability of
the imidazoline binding domain did not increase following high salt
washes (data not shown). The differential photolabeling of liver and
platelet MAO-B was also maintained after detergent solubilization of
the enzymes from the membrane (Fig. 5B).
I]AZIPI-labeled species in
liver mitochondria, platelet membranes, or a mixture of both. Liver
membranes (50 µg of protein), platelet membranes (300 µg of
protein), or a mixture (50 µg of liver mitochondria protein and 300
µg of platelet membrane protein) were incubated at 24 °C for 10
min prior to photolabeling with [
I]AZIPI (0.7
nM) in the presence or absence of cirazoline (10
µM). Following electrophoresis on a 10% SDS-polyacrylamide
gel, autoradiographs were obtained by exposing the dried gels at
-70 °C for 10 days. The arrow indicates the
migration of the photolabeled protein with M
of
59,000. B, immunoblot and autoradiograph of
[
I]AZIPI-labeled MAO-B in detergent-solubilized
preparations of liver and platelet membranes. Approximately 35% of
membrane proteins and
40% of MAO-B detectable by immunoblotting
were solubilized by extraction of platelet and liver membranes with 1%
digitonin. Left, aliquots of solubilized liver (25 µg of
protein) and platelet (100 µg of protein) membranes were
electrophoresed on a 10% SDS-polyacrylamide gel and transferred to
membranes. The blot was probed with the anti-MAO-B monoclonal antibody
MAO-B-1C2 (1:1000), and immunoreactive proteins were identified as
described under ``Experimental Procedures.'' Right,
aliquots of solubilized liver (25 µg of protein) and platelet (125
µg of protein) membranes containing similar amounts of MAO-B
immunoreactivity, as indicated in the left panel, were
photolabeled with [
I]AZIPI (1.4 nM) in
the presence or absence of 10 µM cirazoline. Following
electrophoresis on a 10% SDS-polyacrylamide gel, autoradiographs were
obtained by exposing the dried gels at -70 °C for 8 days. The
migration of midrange molecular weight standards are indicated by the numbers to the left of the autoradiographs or
immunoblot (M
10). The arrows indicate the migration of photolabeled or
immunoreactive species with M
of
59,000.
5.5 for the imidazoline-binding protein purified from rabbit
kidney(2, 16) . Fractions from liver (pH
5.4-5.5) and platelet (pH 5.6) were concentrated, and aliquots
containing similar amounts of MAO-B were photolabeled (Fig. 6C). The imidazoline binding domain on platelet
MAO-B remained inaccessible after fractionation of the solubilized
material using the chromatofocusing matrix.
-). MAO-B content was
determined by immunoblotting (insets). Aliquots of collected
fractions from liver (50 µl) and platelet (100 µl) were
electrophoresed on 10% SDS-polyacrylamide gels and transferred to
membranes. The blots were probed with the anti-MAO-B monoclonal
antibody MAO-B-1C2 (1:1000), and immunoreactive proteins were
identified as described under ``Experimental Procedures.''
The fraction numbers are indicated under each lane of the immunoblots.
The arrows indicate the migration of the immunoreactive
protein with an apparent molecular weight of 59,000 in both liver
mitochondria and platelet fractions. Fraction size: 1 ml (liver) and 4
ml (platelet). C, immunoblot and autoradiograph of
[
I]AZIPI-labeled fractions eluted from the
chromatofocusing matrix. Aliquots of the eluted fractions enriched in
MAO-B (fraction 23 from platelet and fractions 23-25 from liver)
were concentrated, and similar amounts of enzyme (left panel)
were photoaffinity-labeled using [
I]AZIPI (right panel). Blots were probed with the anti-MAO-B
monoclonal antibody MAO-B-1C2 (1:1000), and immunoreactive proteins
were identified as described under ``Experimental
Procedures.'' The arrow indicates the migration of the
immunoreactive protein with an apparent molecular weight of
59,000
in both liver and platelet fractions (left panel).
Autoradiographs of the photolabeled samples were obtained by exposing
the dried gels at -70 °C for 8 days. In the right
panel, the arrow indicates the migration of a
photolabeled species with M
of 59,000. The
migration of midrange molecular weight standards are indicated by the numbers to the left of the autoradiograph or
immunoblot (M
10).
and I based on their ability to recognize
various imidazoline or guanidinium ligands. Differences in ligand
recognition properties of the I and I sites
include selective recognition of the imidazoline clonidine by I sites(30) . The relative insensitivity of
[I]AZIPI photoincorporation in placenta or
liver mitochondria to competition by clonidine indicates that both of
the imidazoline-binding proteins identified in these tissues belong to
the I
subgroup of imidazoline-binding proteins which are
localized to the outer mitochondrial membrane. The I subtype can be further subclassified based upon differential
recognition of the guanidinium compound amiloride (30) . The
imidazoline binding domain on MAO-A in placental membranes exemplifies
the amiloride-sensitive I subtype while photolabeling of
the imidazoline binding domain on MAO-B in liver mitochondrial
membranes is relatively insensitive to amiloride. Similarly,
imidazoline binding sites identified in human placenta and liver by
radioligand binding with [H]idazoxan
differentially recognize amiloride(8, 28) . Thus, two
pharmacologically defined I subtypes in these tissues are
identical with the A and B isoforms of MAO.I]AZIPI. There are several possible
explanations for the observation that the imidazoline binding domain is
not equally detected in all tissues expressing MAO: 1) the existence of
additional isoforms of monoamine oxidase, generated perhaps by
alternative splicing, that differ in the enzyme domain that recognizes
imidazoline/guanidinium ligands; 2) cell-specific post-translational
modification of the enzyme such that the binding domain for
imidazoline/guanidinium ligands is selectively masked; 3) the existence
of tissue-specific protein(s) that allosterically influence
accessibility to the imidazoline binding domain; or 4) occupation of
the imidazoline binding domain by an endogenous substance that is
present in selected tissues.
I]AZIPI
and do not inhibit [
H]idazoxan binding to the
heterologously expressed enzymes(9) . Thus the binding of these
compounds likely involves a site on the enzymes distinct from the
substrate recognition site. In addition, in rat liver and rabbit
cerebral cortex, various MAO inhibitors exhibit K
values in the micromolar range for imidazoline binding sites
identified with
[H]idazoxan(20, 22) . However,
in rat brain membranes, the MAO-A isoform selective inhibitor
clorgyline exhibits picomolar affinity for a subpopulation of
imidazoline-binding proteins and the interaction of clorgyline with
these sites is irreversible, as is its ability to inhibit substrate
oxidation(23) . Thus, the relationship between these two enzyme
domains and the actual structure of the functional enzyme complex
remain unclear. determined in radioligand binding studies corresponding to a
concentration that is 10-100-fold higher than the estimated
concentration required to saturate available imidazoline binding
sites(9, 20) . It is thus difficult to correlate all
of the various cellular effects of imidazoline or guanidinium compounds
with alterations in MAO-induced neurotransmitter metabolism, suggesting
the involvement of other imidazoline-binding proteins. Alternatively,
perhaps MAO is a multifunctional enzyme that possesses as yet unknown
actions initiated by occupation of the imidazoline binding domain.
Demonstration of a binding site on MAO for this class of
pharmacologically active compounds that is detected in a
cell-type-specific manner is of particular significance given the
putative role of the enzyme in the etiology and/or therapeutic
management of various neurodegenerative diseases(31) .
) imidazoline binding
sites (IBS), imidazoline/guanidinium receptive sites, or I
receptors.)I]AZIPI,
2-(3-azido-4[
I]iodophenoxy)methylimidazoline;
PMSF, phenylmethylsulfonyl fluoride; MOPS,
3-(N-morpholino)propanesulfonic acid.
))90% of the
photolabeled species. MAO-B was immunoprecipitated from liver
mitochondrial membranes with lower efficiency (
45% in
immunoprecipitation pellet), and, correspondingly,
50% of the
photolabeled species was present in the immunoprecipitation pellet.
We express our appreciation to Drs. V. Bakthavachalam
and J. L. Neumeyer of Research Biochemicals for providing
2-(3-aminophenoxy)methylimidazoline used as a precursor to AZIPI. We
thank Dr. Richard M. Denney of the University of Texas Medical Branch,
Galveston, TX for the generous gift of MAO monoclonal antibodies, Dr.
Singh (Division of Pediatrics, Medical University of South Carolina)
for the human liver mitochondria, and Dr. P. Halushka (Dept. of
Pharmacology, Medical University of South Carolina) for human platelet
concentrate. We also thank Mark Dole for technical assistance.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Remaury, R. Raddatz, C. Ordener, S. Savic, J. C. Shih, K. Chen, I. Seif, E. De Maeyer, S. M. Lanier, and A. Parini Analysis of the Pharmacological and Molecular Heterogeneity of I2-Imidazoline-Binding Proteins using Monoamine Oxidase-Deficient Mouse Models Mol. Pharmacol., November 1, 2000; 58(5): 1085 - 1090. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. Raddatz, S. L. Savic, V. Bakthavachalam, J. Lesnick, J. R. Jasper, C. R. McGrath, A. Parini, and S. M. Lanier Imidazoline-Binding Domains on Monoamine Oxidase B and Subpopulations of Enzyme J. Pharmacol. Exp. Ther., March 1, 2000; 292(3): 1135 - 1145. [Abstract] [Full Text]
|
|
|
|
 |
|
 P. G. Guyenet Is the hypotensive effect of clonidine and related drugs due to imidazoline binding sites? Am J Physiol Regulatory Integrative Comp Physiol, November 1, 1997; 273(5): R1580 - R1584. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Raddatz, A. Parini, and S. M. Lanier Localization of the Imidazoline Binding Domain on Monoamine Oxidase B Mol. Pharmacol., October 1, 1997; 52(4): 549 - 553. [Abstract] [Full Text]
|
|
|
|
|
|
C. Gargalidis-Moudanos, A. Remaury, N. Pizzinat, and A. Parini Predominant Expression of Monoamine Oxidase B Isoform in Rabbit Renal Proximal Tubule: Regulation By I2 Imidazoline Ligands in Intact Cells Mol. Pharmacol., April 1, 1997; 51(4): 637 - 643. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |