|
J Biol Chem, Vol. 273, Issue 45, 29400-29405, November 6, 1998
Valine 904, Tyrosine 898, and Cysteine 908 in Na,K-ATPase
 Subunits Are Important for Assembly with  Subunits*
Shyang-Guang
Wang and
Robert A.
Farley§¶
From the Departments of Physiology and Biophysics and
§ Biochemistry and Molecular Biology, University of Southern
California School of Medicine, Los Angeles, California 90033
 |
ABSTRACT |
A 26-amino acid sequence in an extracellular loop
of the Na,K-ATPase  subunit between membrane-spanning segments 7 and
8 has been shown to bind to the  subunit of Na,K-ATPase and to promote assembly (Lemas, M. V., Hamrick, M., Takeyasu, K., and
Fambrough, D. M. (1994) J. Biol. Chem. 269, 8255-8259) When this 26-amino acid sequence of the rat Na,K-ATPase
3 subunit was replaced by the corresponding sequence of the rat
gastric H,K-ATPase subunit, the chimeric subunit assembled
preferentially with the rat gastric H,K-ATPase subunit (Wang,
S.-G., Eakle, K. A., Levenson, R., and Farley, R. A. (1997)
Am. J. Physiol. 272, C923-C930). In the present
study, these 26 amino acids (Asn886-Ala911) of
rat Na,K-ATPase 3 were replaced by the corresponding amino acids
Asn908-Ala933 of rat distal colon H,K-ATPase.
Site-directed mutagenesis of the chimeric subunits and Na,K-ATPase
3 showed that Val904, Tyr898, and
Cys908 in the Na,K-ATPase 3 subunit are key residues in
subunit interactions. The V904Q mutation in Na,K-ATPase 3
reduced the Bmax for ouabain binding and the
ATPase activity of 31 complexes by ~95%, and Y898R reduced the
Bmax and ATPase activity by ~60%. The
complementary mutations Q904V and R898Y increased the amount of ouabain
bound by yeast membranes expressing the chimera with the colon
H,K-ATPase sequence. The amount of ouabain bound by complexes assembled
between Na,K-ATPase 3 containing the Y898R,C908G mutations and
gastric H,K-ATPase was less than 10% of wild type Na,K-ATPase 3
expressed with the same subunit. The R898Y,G908C mutations in the
chimeric subunits also increased ouabain binding.
| |
INTRODUCTION |
P-type ATPases are those ion-transporting ATPases that are
phosphorylated transiently by ATP as part of the catalytic mechanism. This class of ion-motive ATPase, including Na,K-ATPase
(NK),1 H,K-ATPase (HK), and
Ca-ATPase, is distributed widely throughout the plant and animal
kingdoms. All of these ion pumps contain a large catalytic subunit
with a molecular mass between 70 and 200 kDa, and the
potassium-transporting ATPases such as HK and NK also require a second,
smaller glycosylated subunit (30-55 kDa) for their enzymatic
functions. The subunit has multiple transmembrane segments and
contains all of the amino acids thus far identified with the enzymatic
functions of ATP hydrolysis and cation transport. The subunit is a
glycoprotein with one transmembrane segment and most of its mass
located on the noncytoplasmic side of the membrane. The role of the subunit in active ion transport is not fully understood.
Lemas et al. (1) have identified 26 amino acids in NK,
predicted to be located in an extracellular loop between transmembrane segments 7 and 8, which mediate interactions between and subunits. These 26 amino acids correspond to
Asn886-Ala911 of the rat NK3 subunit. Wang
et al. (2) examined assembly using a chimeric subunit (NGH26) formed by replacement of
Asn886-Ala911 of rat NK3 with the
corresponding amino acids Gln905-Val930 of rat
gastric HK. When NGH26 was expressed in yeast cells with HK, the
number of ouabain binding sites was the same as for NK3 expressed
with either NK1 or HK. In contrast, only about 10% as many
complexes were formed between NGH26 and NK1. Wang et al.
concluded that some amino acids within the sequence
Gln905-Val930 of rat gastric HK probably
destabilize complexes formed with NK, leading to a reduction
in the number of steady-state pumps. This conclusion is consistent with
the observation that a chimeric subunit with amino acids 1-519
from NK and amino acids 519-1033 from gastric HK did not form
stable complexes with NK1 (3).
Unlike gastric HK, other HK subunits do not appear to
discriminate between different subunits. For example, functional pumps are assembled between a distal colon HK subunit and either NK1 or HK subunits. Codina et al. (4) showed that
86Rb uptake into Xenopus oocytes increased when
colon HK was expressed with either NK1 or gastric HK, and
Cougnon et al. (5) observed an increase in the
86Rb uptake rate of oocytes injected with cRNA for colon
HK and an amphibian HK. A human HK subunit (ATP1AL1) with 86%
amino acid sequence identity to rat colon HK was cloned by Modyanov et al. (6), who observed that any of several different subunits could be coimmunoprecipitated with ATP1AL1 when expressed in
Xenopus oocytes. Expression of rabbit gastric HK with
ATP1AL1 in Xenopus oocytes resulted in a 3-fold increase in
86Rb uptake compared with uninjected cells (7).
To identify amino acids that are involved in interactions between the
and subunits, a new chimeric subunit (NCH26) was made by
replacing the 26-amino acid sequence
Asn886-Ala911 of the rat NK3 subunit with
the corresponding sequence Asn908-Ala933 of
rat distal colon HK subunit. A series of mutations was introduced into NK3 or the chimeric NGH26 and NCH26 subunits, and the presence of functional complexes was measured by ouabain binding or ATPase activity after expression of the polypeptides in yeast with
either NK1 or gastric HK. Stability of the complexes was estimated from the ability of each complex to bind ouabain at
elevated temperatures. As a result of these measurements,
Val904, Tyr898, and Cys908 in
NK3 have been identified as important amino acids for assembly of
subunits and subunits.
| |
MATERIALS AND METHODS |
Construction of the Chimeric Subunit NCH26--
The plasmid
pRD-CHK containing the cDNA of the rat colon HK subunit was a
gift of Dr. Gary Shull (University of Cincinnati). A 102-base pair
fragment of pRD-CHK, encoding positions 2711-2812, was amplified by
polymerase chain reaction, and ClaI and HpaI restriction sites were introduced at the same time. For construction of
plasmid pNCH26m, the polymerase chain reaction fragment was digested
with ClaI and HpaI and was ligated into the
corresponding region of clone pNGH26m (2) whose
ClaI-HpaI fragment (79 base pairs) had been
removed. Three mutations (N886D,2
A911V, and F912N), resulting from the
introduction of ClaI and HpaI sites, were
corrected by polymerase chain reaction. The resultant plasmid pNCH26
encodes the rat NK3 subunit with the region
(Asn886-Ala911) replaced by the 26 amino acids
(Asn908-Ala933) of the rat colon HK
subunit. The AflII-BglII fragment (1,486 base
pairs) of the yeast expression plasmid YEpNGH26 (2) was replaced by the
corresponding AflII-BglII fragment of pNCH26. The
final plasmid (YEpNCH26) was analyzed by restriction digestion and by
DNA sequencing.
Site-directed Mutagenesis--
Mutations in the cDNA were
made using the polymerase chain reaction, as described previously (2).
Pfu DNA polymerase (Stratagene) was used to perform the
site-directed mutagenesis, and the resultant mutants were screened by
restriction enzymes and were confirmed by DNA sequencing using the
Sequenase version 2.0 (U. S. Biochemical Corp.).
Expression of NCH26 or Mutants with Subunits in Yeast
Cells--
The yeast strain 30-4 (MAT , trp1, ura3, Vn2,
GAL+) obtained from R. Hitzeman (Genentech, South San Francisco)
was transformed with different combinations of the chimeric subunit
or mutant expression plasmid and one of the subunit expression
plasmids pG1T-R1 and pG1T-HK (9) by the method of Elble (8).
After identification of transformants containing and subunits
on selective medium, frozen glycerol stocks from four different clones were made and were stored at  80 °C. Cultures for experiments described in this report were started from these glycerol stocks. A
membrane fraction of transformed yeast cells was prepared as described
previously (9). Membranes were extracted with 0.1% (w/v) SDS as
described previously (10).
[3H]Ouabain Binding--
Ouabain binding to yeast
membranes was done as described previously (2). Experiments shown in
Figs. 5 and 7 were done using approximately 20 nM
[3H]ouabain. To determine the number of steady-state pump
complexes (Bmax) and the ouabain dissociation
constant (Kd), binding data were fit by a
self-competition model (11) within the ouabain concentration range
0-1,000 nM. For determination of the effects of heat on
stability, 3 mg of yeast microsomal membrane protein was
dissolved in 400 µl of 25 mM imidazole-HCl, 1 mM EDTA (sodium-free), pH 7.4, and was heated at different
temperatures (40-50 °C) for 90 s. Membranes were placed in ice
for 15-30 min, and the amount of ouabain bound at 37 °C was
measured. The amount of ouabain bound by membranes without heating was
used as 100%.
SDS-Polyacrylamide Gel Electrophoresis and Immunoblots--
100
µg of yeast microsomal membrane protein was separated on 10%
SDS-polyacrylamide gels and then was transferred to Immobilon-P membranes (Millipore). The blots were first incubated with monoclonal antibody 5 (D. Fambrough, Johns Hopkins University), then incubated with the alkaline phosphatase-conjugated goat anti-mouse IgG
(Calbiochem). The subunits were visualized with
5-bromo-4-chloro-3-indolyl phosphate (Sigma) and nitro blue tetrazolium
(Sigma). The density of each band was determined by scanning with a
Bio-Rad scanner (Imaging Densitometer, model GS-670). The average of
the expression levels of NCH26 + NK1 and NCH26 + HK with three
different clones was determined and compared with the NK3 + NK1
and NK3 + HK controls.
ATPase Activity--
NK activity was determined in triplicate by
measuring ouabain-inhibitable phosphate release, as described
previously (12).
| |
RESULTS |
Expression Levels of NCH26 with NK1 or HK--
The sequence
Asn886-Ala911 of rat NK3 was replaced by
the corresponding sequence Asn908-Ala933 of
rat colon HK, and this chimeric subunit NCH26 was expressed in
yeast cells with either NK1 and gastric HK. Fig.
1A shows an immunoblot of
membranes prepared from three different clones expressing either NK3
or NCH26 with either NK1 or HK. Three clones expressing only
NK3 are also shown. Because the antibody 5 recognizes the same
epitope on both NK3 and NCH26, the relative abundance of each subunit can be compared directly. The abundance of NCH26 expressed with
NK1 is 82 ± 19% of NK3 expressed with NK1, and the
amount of NCH26 expressed with HK is 87 ± 11% of NK3
expressed with HK (Fig. 1B). When expressed in the
absence of a subunit, NK3 is present at only 10 ± 10% of
NK3 levels found with NK1. In the absence of the subunit, the
subunit is degraded rapidly, and the higher steady-state abundance
of subunits expressed with either NK1 or HK reflects
stabilization of the subunit by association with a subunit
(13). These results show that the steady-state level of NCH26 expressed
in yeast with NK1 or HK is not significantly different from
NK3 (p > 0.05).
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 1.
Expression levels of NK3 and NCH26
expressed in yeast with NK1 or HK. Panel A, yeast
membranes containing 100 µg of protein were separated by
SDS-polyacrylamide gel electrophoresis and transferred to a
polyvinylidene difluoride membrane. The blot was probed sequentially
with monoclonal antibody 5 and alkaline phosphatase-conjugated goat
anti-mouse IgG. Three different clones were used to determine the
expression level of subunit of each complex. Panel
B, relative expression levels were measured by densitometry. The
averages of three clones of NK3 expressed with NK1 or HK are
the 100% controls, and the expression levels of NCH26 expressed with
NK1 or HK were normalized to the controls. Bars show
means ± S.D.
|
|
Ouabain Binding by NCH26--
Functional NCH26 + complexes
were quantitated by ouabain binding, and the
Bmax and Kd values for NK3
or NCH26 expressed with different subunits were determined for
three different clones each. The results presented in Fig.
2 demonstrate that functional complexes
are formed equally well in yeast between NK3 and either NK or
HK. Although the NCH26 polypeptide is present at the same level as
NK3, fewer functional complexes are formed between NCH26 and either
NK1 or HK than with NK3. Yeast membranes containing NCH26 + NK1 form about 40% of the number of functional pumps as NK3 + NK1, and the number of functional NCH26 + HK complexes is about
20% of NK3 + HK. The Kd for ouabain binding
by NCH26 + NK1 is 50 ± 13 nM, and for NCH26 + HK it is 206 ± 95 nM. These values are 8- and
27-fold higher than the Kd values of NK3 + NK1
(6 ± 2 nM) and NK3 + gHK (7 ± 3 nM), respectively.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
Ouabain binding by NK3 + and NCH26 + subunits. The ouabain binding capacity
(Bmax) of membranes from yeast expressing the
indicated + subunits was determined as described under
"Materials and Methods." Bars represent mean values ± S.D. of triplicate measurements from three different clones.
|
|
The number of ouabain-binding complexes formed by NCH26 and HK is
only 20-40% of the number formed by either NK3 + HK or NGH26 + HK (2). This result might be explained if residues within the
sequence Asn908-Ala933 of rat colon HK are
less suitable for assembly with gastric HK than corresponding
residues of rat NK3 or rat gastric HK. A comparison of the amino
acid sequences Asn886-Ala911 of rat NK3,
Gln905-Val930 of rat gastric HK, and
Asn908-Ala933 of rat colon HK shows that
the three sequences have identical residues or conservative
substitutions in all positions except at amino acids 898, 908, and 909. In both NK3 and rat gastric HK, a tyrosine or a phenylalanine is
located at position 898, and a cysteine is located at position 908. In
rat colon HK these residues are arginine and glycine, respectively.
In position 909, the amino acids are different for all three sequences.
To see whether the amino acids in positions 898 and 908 are important for the assembly with HK, mutations R898Y and/or G908C were
introduced into NCH26, and the amount of ouabain bound by the mutants
expressed in yeast with different subunits was measured.
Ouabain Binding by NCH26 Mutants--
The maximum amount of
ouabain bound (Bmax) by the R898Y, G908C, and
R898Y,G908C mutants of NCH26 expressed in yeast cells with either
NK1 or gastric HK was used to indicate the number of functional
complexes (Fig. 3). The mutation
R898Y increases the Bmax of NCH26 + NK1
complexes to the same value as NK3 + NK1. When expressed with
HK, the Bmax of the R898Y mutant increases 2.7 times higher than that of NCH26 + HK but is still less than that
of NK3 + HK. The mutation G908C does not affect the number of
NCH26 + NK1 complexes. When expressed with HK, however, the mutation G908C reduces the Bmax of NCH26 from
22% to 3% of NK3 + HK. When the two mutations R898Y and G908C
are made in NCH26, the Bmax for ouabain binding
is the same as NK3 + HK (p > 0.05), and is 1.8 times higher than that of NCH26/R898Y + HK.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 3.
Ouabain binding by NCH26 and related mutants
expressed in yeast with NK1 or gHK. Left panel,
NCH26 and related mutants expressed with NK1. Right
panel, NCH26 and related mutants expressed with gastric HK. The
maximum amount of ouabain bound by each combination
(Bmax) is the average of the values of three
different clones, and the Bmax of each clone was determined
in triplicate. Bars show means ± S.D. The values of
NK3 + NK1 and NK3 + HK are also shown.
|
|
Ouabain Binding by NGH26 Mutants--
Wang et al.
studied the chimeric NGH26 subunit and concluded that amino acids
of gastric HK between Gln905 and Val930
interact more stably with the extracellular domain of HK than NK
(2). Charged amino acids have been implicated in the assembly of some
membrane proteins (14, 15), and within the 26 residues that were
exchanged during chimera formation, the charged amino acids
Lys902 and Glu905 are conserved among all of
the NK subunits and also in the colon HK subunits. In the gastric
HK subunits, these amino acids are leucine and glutamine,
respectively. Because NGH26 forms fewer functional pumps with NK1
than with HK, charged residues in positions 902 and 905 may be
important for interactions between subunits and NK1. To test
this possibility, the mutations L902K and Q905E were introduced
separately or together into NGH26, and the subunits were expressed
in yeast with either NK1 or HK. Only the Q905E mutation reduced
the ouabain binding capacity of yeast membranes containing the mutant
subunits assembled with HK (p < 0.05).
Introduction of the double mutation (NGH26-KE), however, restored the
binding capacity and reduced the Kd for ouabain
binding by NGH26 from 72.7 to 23 nM (Table
I).
View this table:
[in this window]
[in a new window]
|
Table I
Ouabain binding affinity and binding capacity of subunits expressed
with HK
Ouabain binding to yeast membranes containing either NK3 + HK or NGH26 + HK was done as described under "Materials
and Methods." Mutations made in the NGH26 chimeric subunit are
indicated. The ouabain binding affinity (dissociation constant
(Kd, in nM) and maximum amount of ouabain bound
(Bmax, in pmol/mg protein) were determined from a
fit to the data by a single-site self-competition model (11). Values
are shown as means ± S.D. (n = 3).
|
|
Site-directed Mutagenesis of NGH26-KE--
Although the mutations
L902K and Q905E make the amino acid sequence of NGH26 more nearly like
that of NK3, the number of pumps is not increased above the NGH26 + NK1 level. This may be a result of the presence in NGH26 and
NGH26-KE of amino acids whose side chains are sterically or
electrostatically incompatible with assembly with NK1or because of
the absence in these subunits of amino acids whose side chains are
important for specific interactions with NK1. There are 10 amino
acid differences between the NK3 sequence and NGH26-KE (Fig.
4). Each of these amino acids was changed
in NGH26-KE, either individually or in pairs, to those amino acids in
NK3, and ouabain binding was used to identify amino acids that are
important for assembly of NK3 with NK1.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 4.
Site-directed mutagenesis of NGH26-KE.
Bold letters indicate the 10 amino acid differences between
the NK3 (upper sequence) and NGH26-KE (lower
sequence) in the region that was replaced during chimera
construction. The 26 amino acids that were replaced in NK3 and their
replacements in NCH26 are indicated by double underlines.
The arrows indicate the L902K and Q905E mutations in
NGH26-KE. The mutations Q886N, Q889E, E895Q, double mutations
F898Y,G899E, and Y903V,Q904V, and Y906F, Y909H, and V911A were
introduced into NGH26-KE.
|
|
Ouabain Binding by NGH26-KE and Related Mutants--
Yeast
membranes containing the subunit mutants derived from NGH26-KE
together with either NK1 or gastric HK were equilibrated with 20 nM [3H]ouabain, and specific binding was
measured as before. As shown in Fig. 5,
when the double mutation Y903V,Q904V is made in NGH26-KE and this
mutant subunit is expressed with NK1, yeast membranes bind 10 times more ouabain than when NGH26-KE is expressed with NK1. In
addition, the double mutation Y903V,Q904V in NGH26-KE increases the
amount of ouabain bound by when expressed in yeast with HK. The
amount of ouabain bound by the rest of the mutants when expressed with
either NK1 or HK is not significantly different from that of
NGH26-KE + NK or NGH26-KE + HK (p > 0.05),
respectively.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 5.
Ouabain binding by NGH26-KE mutants expressed
in yeast with NK1 or gastric HK. Left panel, NGH26-KE
mutants expressed with NK1. Right panel, NGH26-KE mutants
expressed with gastric HK. Yeast membranes, containing either
NGH26-KE (-KE), or the mutants, were incubated with 20 nM [3H]ouabain at 37 °C for 1 h, and
the amount of ouabain bound was determined as described under
"Materials and Methods." The binding assay was performed in
duplicate with four different clones, and the bars show
means ± S.D. within each group.
|
|
Ouabain Binding by NK3 Mutants--
The mutations Y903V,Q904V
in the chimera NGH26-KE and R898Y,G908C in the chimera NCH26 lead to
large increases in the amount of ouabain bound by yeast membranes
containing these chimeric subunits and either NK1 or HK.
These increases could be due to changes in the affinity of the mutants
for ouabain and/or to increased stability of the complexes. If
amino acids in positions 898, 903, 904, and/or 908 participate in
interactions between and subunits, then the reverse mutations
in NK should reduce the amount of ouabain bound by yeast membranes
expressing each complex. Thus, the mutations V904Q, Y898R,
C908G, and Y898R,C908G were made in NK3 and the ouabain binding
affinity and capacity of each complex were measured after
isolation of yeast membranes. Non-polar amino acid side chains are
conserved among NK subunits in position 904 but not in position 903, and so the effect of amino acid substitutions in position 903 was not
tested. The left panel of Fig.
6 shows the Bmax
values of the NK3 mutants expressed with NK1, and the right
panel shows the Bmax values for NK3 mutants expressed with HK. The Bmax of
NK3/V904Q + NK1 is only 5% of the Bmax
for NK3 + NK1, and the Bmax of
NK3/V904Q + HK is 18% of NK3 + HK, confirming that
Val904 is important for the functional assembly of NK
with both subunits. The Bmax of NK3/Y898R + NK1 is 42% of NK3 + NK1, and the
Bmax of NK3/Y898R + HK is not
significantly different from that of NK3 + HK. For the mutation
C908G in NK3, no significant increase or decrease in the number of
complexes was seen when assembled with either subunit. Even
though the individual mutation Y898R or C908G has no significant effect
on the number of NK3 + HK complexes, the double mutation
Y898R,C908G was associated with a reduction in the
Bmax for ouabain binding to about 10% of NK3 + HK levels. Table II shows that for
the V904Q mutation, no change in ouabain affinity was observed when the
mutant NK3 subunit was expressed with either NK1 or HK.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 6.
Ouabain binding by NK3 mutants expressed
with NK1 or HK. Left panel, NK3 mutants
expressed with NK1. Right panel, NK3 mutants expressed
with gastric HK. The maximum amount of ouabain bound by each
complex is shown as the average of the values of three different
clones, and the Bmax of each clone was
determined in triplicate. Bars show means ± S.D. The
values of NK3 + NK and NK3 + HK were taken from Fig.
2.
|
|
View this table:
[in this window]
[in a new window]
|
Table II
Ouabain binding affinities of NK3 and NK3-related mutants
expressed with NK1 or HK
The Kd values (nM) were determined as
described under "Materials and Methods." Values are shown as
means ± S.D. (n = 4).
|
|
ATPase Activity of NK3 Mutants Expressed with NK1--
Yeast
membranes containing the different + NK1 complexes were
extracted with SDS (16) to see whether reduction in the number of pump
complexes caused by mutation of residues in NK3 is accompanied by a
similar reduction in ATPase activity. When expressed with NK1, the
ATPase activities of the mutants V904Q, Y898R, C908G, and the double
mutant Y898R,C908G were reduced to the same extent as the
Bmax values (Table
III). As shown previously (9), the + HK complexes are unstable in SDS, and the influence of the mutations
on the ATPase activity of these complexes could not be determined.
View this table:
[in this window]
[in a new window]
|
Table III
Comparison of the effects of mutations in NK3 on ouabain binding
capacity and on ouabain-inhibitable ATPase activity in yeast membranes
The data for the Bmax values are from the
experiments summarized in Fig. 7. Values for Bmax
and ATPase activity are expressed as a percentage of the values for
NK3, 3.8 pmol of ouabain bound/mg and 1.6 µmol of ATP
hydrolyzed/mg/h, respectively. Values shown are means ± S.D.
|
|
Thermal Stability--
The stability of the NK3 mutants
expressed in yeast with NK1 or HK was investigated by heating the
membranes at different temperatures (40-50 °C) for 90 s and
then measuring the amount of ouabain bound at 37 °C. Fig.
7 (upper panel) shows that all of the mutants are as stable as NK3 + NK1 when expressed with NK1. In contrast, when expressed with HK, both NK3 and the mutants denature at much lower temperatures (lower panel).
Compared with NK3 + HK, the + HK complexes containing the
mutations Y898R and C908G are significantly less stable. The mutant
containing the double mutation Y898R,C908G is extremely unstable when
expressed with HK. Ouabain binding by complexes of this mutant
expressed with HK was not detected after the membranes were heated
at 45 °C for 90 s (data not shown). In contrast to Y898R and
C908G, the mutation V904Q in NK3 has no effect on the thermal
stability of the functional + HK complexes. Glutamine is
conserved in all gastric HK subunits in the position corresponding
to Val904 in NK.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 7.
Heat inactivation of ouabain binding.
Upper panel, NK3 and related mutants expressed with
NK1. Lower panel, NK3 and related mutants expressed
with gastric HK. 3 mg of yeast membranes containing different
complexes were heated at different temperatures (40-50 °C) for
90 s, and then the amount of ouabain binding was determined in
quadruplicate at 37 °C by incubation with 20 nM
[3H]ouabain, 4 mM MgCl2, 4 mM Pi, pH 7.5. Values shown are expressed as
percent of [3H]ouabain bound by the unheated membranes.
Standard deviations are indicated by error bars.
Filled triangles, NK3; open circles, Y898R;
filled squares, C908G; filled circles,
Y898R,C908G; open squares, V904Q
|
|
| |
DISCUSSION |
It has been shown previously that both and subunits are
required by NK and gastric HK to catalyze active ion transport (17-19). NK subunits assemble equally well with NK1 and gastric HK (9), but gastric HK does not form functional pumps with NK
(3). In contrast to gastric HK, coexpression of colon HK in
Xenopus oocytes with either gastric HK or NK (4) or with toad urinary bladder HK (5) leads to functional ion pumps. This
result indicates that interactions between subunits and colon HK
or gastric HK are mediated by amino acids that are different in the
two subunits. Wang et al. (2) have shown that
substitution of amino acids Gln905-Val930 from
rat gastric HK for the corresponding sequence
Asn886-Ala911 of rat NK3 reduces the
assembly of this chimeric subunit (NGH26) with NK1 compared with
assembly with HK. In the experiments reported here, a chimeric subunit (NCH26) was formed by replacing the sequence
Asn886-Ala911 of rat NK3 with the
corresponding sequence Asn908-Ala933 of the
rat colon HK subunit to examine the structural basis for the
selectivity of subunits for assembly with different subunits.
The formation of functional pumps was compared when NK3, NCH26, or
NGH26 was expressed in yeast with either NK1 or gastric HK, and
amino acids in the subunits that are important for interactions
between and subunits were identified after site-directed mutagenesis.
The steady-state abundance of the NCH26 polypeptides expressed in yeast
with NK1 or HK is the same as that of NK3 (Fig. 1A). About 40% of the NCH26 + NK1 complexes are
functional, as determined by the ability to bind ouabain, and
about 20% of the NCH26 + HK complexes are functional (Fig. 2). This
result is consistent with the observation of Codina et al.
(4) that the colon HK is capable of functional assembly with either
NK1 or gastric HK. The result also suggests that some NCH26 + complexes are inactive in the yeast membranes.
Codina et al. (4) reported that high concentrations of
ouabain inhibited colon HK expressed in Xenopus oocytes
with either NK1 or HK (IC50 = 400-600
µM in the presence of 1 mM external KCl). In
the absence of KCl, the Kd for ouabain binding to
NK3 expressed in yeast with either NK1 or HK is 6-8
nM (Tables 1 and 2). The Kd for ouabain
binding to NCH26 + NK1 is 50 nM, and for NCH26 + gastric
HK the Kd is greater than 200 nM. It
is likely, therefore, that the low affinity of colon HK for ouabain
is due at least in part to amino acids
Asn908-Ala933. These amino acids are found in
a loop predicted from hydropathy analysis to be located on the
non-cytoplasmic side of the cell membrane, between transmembrane
segments 7 and 8. The loop between these transmembrane segments of
NK has been implicated in ouabain binding by Schultheis et
al. (20), who observed that mutations in Arg880 led to
a reduction in the affinity of the sodium pump for ouabain.
The chimeric subunit NGH26 contains amino acids
Gln905-Val930 from rat gastric HK
substituted for Asn886-Ala911 of rat NK3.
When this chimera was expressed in yeast, about 10 times as many pumps
were assembled with gastric HK as with NK1 (2). The difference in
the number of pumps assembled in yeast from NK, from gastric HK,
or from the chimeric polypeptides NCH26 and NGH26 and either NK1 or
HK probably reflects differences in assembly or in the stability of
the different complexes. Consequently, mutations were made in
the chimeras NGH26 and NCH26 and also in NK3 to identify amino acid
side chains that might mediate subunit interactions.
Introduction of charged amino acids in positions 902 and 905 of NGH26
(L902K and Q905E) increased the affinity of the pump for ouabain but
did not increase the number of pumps assembled with HK (Table I).
Because the single mutations alone did not influence the dissociation
constant of ouabain binding of NGH26 + HK, it is likely that neither
Leu902 nor Gln905 interacts directly with
ouabain. The lower Kd of the double mutant is
probably caused by an induced tertiary structure in the ouabain binding
site of the double mutant similar to that of the NK3 subunit.
Because neither the single mutations (L902K and Q905E) nor the double
mutation (L902K,Q905E) increase the Bmax of
NGH26 expressed with either NK1 or HK, these amino acids also are
probably not located in the interaction interface.
The chimera containing the L902K and Q905E mutations (NGH26-KE) was
used as a template for additional mutations that changed amino acids in
the chimera to those found in NK3. Most of the mutations in NGH26-KE
did not affect ouabain binding by the chimera; however, the mutations
Y903V and Q904V led to a significant increase in the amount of ouabain
bound when the mutant/chimeric subunit was expressed with NK1
(Fig. 5). This result suggests that one or both of these small
hydrophobic amino acids may be important for complex formation with
NK1. The valine at position 903 is not conserved among NK
subunits. Polar amino acids including threonine, glutamine, and
glutamic acid are also found at this position in some isoforms or
species. A non-polar amino acid such as valine, leucine, or isoleucine
is conserved at position 904 in all NK subunits, and a glutamine is
found at this position in all gastric HK subunits. Because gastric
HK subunits do not assemble with NK subunits, the mutation V904Q
was made in NK3 to test whether the glutamine in position 904 of
gastric HK could be the reason that gastric HK does not assemble
with NK1. The V904Q mutation in NK3 caused a reduction in both
the number of functional complexes (Fig. 6) and the ATPase
activity (Table III) to only 5% of NK3 + NK1, without affecting
the affinity of the mutant for ouabain (Table II). This result shows
that the presence of glutamine at position 904 in gastric HK
subunits is sufficient to prevent assembly of HK subunits with
NK1. It also suggests that valine or another small hydrophobic amino
acid in position 904 in NK subunits is important for assembly with NK1.
In addition to Val904, Tyr898 also appears to
be important for assembly of subunits with NK1. The mutation
R898Y in NCH26 caused a 2-fold increase in the number of functional + NK1 complexes (Fig. 3), and the reverse mutation Y898R in NK3
caused a 50% reduction in the number of functional pumps when
assembled with NK1 (Fig. 6). Arginine is conserved in colon HK
subunits at the position corresponding to amino acid 898 of NK3, and
the positive charge may limit assembly of colon HK subunits with NK subunits. Interestingly, the Y898R mutation did not lead to a
decrease in the number of pumps formed with gastric HK (Fig. 6).
The thermal stability of pumps containing the V904Q or Y898R mutation
in NK3 is comparable to that of nonmutated NK3 (Fig. 7). This
result indicates that the reduced number of functional pumps containing
these mutations is probably not the consequence of unstable
complexes. The limiting factor may be the initial assembly of the two
polypeptides, such that the V904Q or Y898R mutation in NK3 prevents
the majority of the two subunits from forming functional complexes.
Beggah et al. (21) reported that mutations to hydrophobic
amino acids near the carboxyl terminus of NK3 interfere with
complex formation in Xenopus oocytes (21). In particular,
the double mutation V269N,F271N abolished the cellular accumulation of
subunits, which is an indication of complex formation. The
finding here that Val904 and Tyr898 in NK3
are important for assembly with NK1 is intriguing in this context.
Perhaps the valine-aromatic amino acid pair on each subunit interacts
with one another to provide a stable contact between the subunits.
The double mutation Y898R,C908G in NK3 caused a reduction in the
number of functional pumps assembled with HK by 90%, despite the
absence of an effect of either mutation alone (Fig. 6). This effect of
the double mutation on the assembly of NK3 with HK is consistent
with the observation that the reciprocal mutations R898Y,G908C in NCH26
led to a 4-fold increase in the amount of ouabain bound when when this
chimeric/mutant subunit was expressed in yeast with HK. The
thermal stability profile (Fig. 7) demonstrates that NK3 with the
double mutation Y898R,C908G is extremely unstable when assembled with
HK. Thus, the small number of pumps assembled either from NK3
containing either these mutations or from NCH26, and HK, may be the
consequence of instability in + HK complexes caused by arginine
and glycine at these positions. The double mutations Y898R,C908G in
NK3 also caused a 13-fold reduction in the ouabain affinity of pumps
assembled with NK1 (Table II) with little effect on complex
stability (Fig. 7). Because neither Y898R nor C908G alone affected
ouabain binding, it is likely that neither Tyr898 nor
Cys908 is located within the ouabain binding site, and the
double mutation reduces ouabain affinity by indirectly affecting the
ouabain binding site.
| |
FOOTNOTES |
*
This work was supported by National Institutes of Health
Grant GM-28673.The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
¶
To whom correspondence should be addressed: Dept. of
Physiology and Biophysics, University of Southern California School of Medicine, 1333 San Pablo St., MMR 250, Los Angeles, CA 90033. Tel.:
213-342-1240; Fax: 213-342-2283; E-mail: rfarley{at}hsc.usc.edu.
The abbreviations used are:
NK, Na,K-ATPase; HK, H,K-ATPase; NGH26, rat Na,K-ATPase 3 subunit with amino acids
Asn886-Ala911 replaced by amino acids
Gln905-Val930 of rat gastric H,K-ATPase subunit; NCH26, rat Na,K-ATPase |
Top
Abstract
Introduction