| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
(Received for publication, August 16, 1994; and in revised form, December 22,
1994) From the
To localize transmembrane segments in the carboxyl-terminal
third of the Neurospora plasma membrane
H
The plasma membrane electrogenic, proton-translocating ATPase
(H Toward this goal,
biochemical, immunological, and genetic approaches have been used to
localize the amino and carboxyl termini of P-type ATPases to the
cytosol (Mandala and Slayman, 1989; Reithmeier and MacLennan, 1981).
This implies an even number of segments. The best characterized
cation-motive porter is the Ca As an approach
toward finding a solution to this problem, we developed an experimental
method based on the idea that transmembrane segments function as
topogenic signals in the topogenesis of polytopic integral membrane
proteins (Lin and Addison, 1994). Using this method, we verified
predicted transmembrane segments in the amino-terminal third of the
H
We used the polymerase chain reaction in constructing fusion
proteins of the Neurospora H
Figure 1:
Constructing fusion proteins. The
single-letter code of the amino acids is used. Numbersabove each correspond to its position in the primary
sequence of the H
Constructs with
putative transmembrane segments were designated pM followed by a number
indicating the position of the segment within the
H
Figure 2:
Fusion proteins membrane topology. Thin
lines with N or C represent the amino or
carboxyl-terminal hydrophilic residues of fusion proteins,
respectively. Rectangular boxes with M
pM7` contains the putative seventh
transmembrane segment of the H
Figure 3:
In vitro synthesis of fusion
proteins. Transcripts were translated with microsomes (nRM) as
indicated. Mixtures were extracted at pH 11.5 and then fractionated on
an alkaline sucrose step gradient into soluble fraction (S)
and membrane pellet (P) as described under ``Experimental
Procedures.'' Oligosaccharide chains were removed by digesting the
total (T) reaction mixture with N-glycosidase F (PNGaseF). Products were precipitated by
trichloroacetic acid and resolved on 10% polyacrylamide gel with SDS;
radioactive proteins were visualized by fluorography. Glycosylated
proteins are indicated by an asterisk.
DNA fragments encoding M5 (amino acid
residues 688-713) and M7 (amino acid residues 755-779) were
also placed in constructs. The residues of M5 are not predicted to span
the membrane by the eight-segment model (Hager et al., 1986).
In the 10-segment model, M5 and M7 are putative initiate-translocation
signals. RNA transcripts of these yielded one product when translated
without microsomes (Fig. 3, lane1). When
translated with microsomes, two products were observed (lane4). The increase in the relative molecular mass of the
slower migrating band, indicated by an asterisk in Fig. 3, is consistent with addition of three N-linked
oligosaccharide side chains to the invertase segment and is
corroborated by results obtained after incubating products with N-glycosidase F, an amidase that cleaves between the
asparagine residues of the glycosylation site and the innermost GlcNAc
of oligosaccharide chains. N-Glycosidase F converted the
slower migrating product to that observed without microsomes (compare lanes5 and 1). This substantiated the
conjecture that the increase in molecular mass was caused by the
addition of oligosaccharide chains to the products. The relative
quantity of pM10 after treatment with N-glycosidase F was
consistently low (lane5). The reason for this is not
apparent. M9 (amino acid residues 827-847), a predicted
initiate-translocation signal, did not initiate transfer of the
carboxyl residues into microsomes (Fig. 3). In the presence of
microsomes, pM9 sedimented with microsomes after alkaline extraction (Fig. 3, compare lanes2 and 4),
suggesting that it is integrated into microsomes, albeit with a very
low efficiency. When M6 (amino acid residues 721-738) and M8
(amino acid residues 807-826) (Fig. 1), as predicted by
the 10-segment model, were engineered into fusion proteins to make pM6
and pM8, the resultant proteins did not associate with microsomes after
alkaline extraction (Fig. 3). The residues 807-826 are
predicted to span the membrane only in the 10-segment model. In this
model, M6 and M8 function as stop-transfer signals. One predicted
stop-transfer signal, M10 (amino acid residues 854-878), in pM10
initiated translocation of the carboxyl-terminal residues into
microsomes (Fig. 3, lane4). The alkaline
extraction method, used for fractionating the products, converts closed
microsomes into opened sheets. Concomitantly, proteins are extracted
from the exposed faces (Fujiki et al., 1982). Peripheral
membrane and lumenal proteins are found in the supernatant fraction,
while integral membrane proteins are found in the membrane pellet.
Taken together, this implies that M5, M7, and M10 served not only as
signals that initiated translocation of the carboxyl-terminal residues,
but also as signals that anchored the proteins in the membrane.
Figure 4:
Digesting products by proteinase K. RNA
transcripts for pM5 (panelA), pM7 (panelB), and pM10 (panelC) were translated
in the Neurospora in vitro system supplemented with microsomes (nRM), as indicated. The resultant mixtures were treated with
100 µg/ml of proteinase K. Triton X-100 (TX-100) was present at 1% during the digestion of products for lane3. Samples for lanes1-3 were trichloroacetic acid-precipitated; others were
immunoprecipitated using polyclonal antiserum (Ab) against the
amino terminus (N) (lane4) or against the
carboxyl terminus (C) (lane5) of the
H
The smaller fragments observed in Fig. 4lane5 for pM10 were generated during
the immunoprecipitation step. pM10 was more susceptible to apparent
proteolytic digestion.
Figure 5:
In vitro synthesis of fusion
proteins with two putative transmembrane segments. RNA transcripts were
translated in a Neurospora in vitro system supplemented with
microsomes (nRM), as indicated. Samples were extracted at pH
11.5 and then fractionated on a sucrose step gradient under alkaline
(pH 11.5) conditions into soluble fraction (S) and membrane
pellet (P). trichloroacetic acid-precipitated products were
resolved on a 10% polyacrylamide gel with SDS; radioactive bands were
visualized by fluorography. See text for more
details.
To further substantiate
the observation that M8 functioned as a stop-transfer signal, we
inserted between M7 and M8 the invertase segment with three consensus N-linked glycosylation sites. pM7InM8 has two invertase
segments. If M8 functions as a stop-transfer signal, pM7InM8 will be
glycosylated three times; if not, it will be glycosylated six times
(see Fig. 2). When synthesized with microsomes, pM7InM8 gave
four products that sedimented with microsomes after alkaline extraction (Fig. 5, lane 4): The topmost product, indicated by an opendiamond in Fig. 5, had an increase in relative
molecular mass of about 15,000 Da, suggesting the addition of six N-linked oligosaccharides chains; the bottommost product is
unmodified protein that sediments with microsomes. A second modified
product, indicated by an asterisk in Fig. 5, had an
increase in relative molecular mass of about 6,000 Da. The third
modified product had an increase of relative molecular mass of 4,000
Da. These products could represent partially glycosylated pM7InM8 that
had both invertase segments in the lumen of microsomes. Alternatively,
this profile of modified proteins would reflect a mixture of
combinations. For example, products with relative increase of molecular
mass of 6,000 and 4,000 Da could represent pM7InM8 that had one
invertase segment in the lumen of microsomes, and the product that had
a relative increase in molecular mass of about 15,000 Da could
represent pM7InM8 with both invertase segments in lumen of microsomes.
To distinguish between these possibilities, the mixture was digested by
proteinase K. Results are depicted in Fig. 6. In the absence of
microsomes, no protected fragment was observed (lane1); in the presence of microsomes, two major and one
minor protected fragments were observed (lane2).
Only the protected fragment with a molecular mass of about 40,000 Da
was immunoprecipitated by polyclonal antibodies against
carboxyl-terminal residues of the H
Figure 6:
Digesting pM7InM8 by proteinase K. pM7InM8
was synthesized in the in vitro system supplemented with
microsomes (nRM), as indicated. The resultant mixtures were
digested with 100 µg/ml of proteinase K. For sample in lane3, Triton X-100 (TX-100) was present at
1% during the digestion. Samples of lanes1-3 were trichloroacetic acid-precipitated; others were
immunoprecipitated with polyclonal antiserum (Ab) against the
amino terminus (N) (lane4) or the carboxyl
terminus (C) (lane5) of the
H
The results described herein support the transmembrane
segments as assigned by the 10-segment model (Addison, 1986). A
membrane topological model of the H
Figure 7:
Model of the Neurospora plasma
membrane H
M7 functioned as the strongest of the
sequences that initiated translocation of carboxyl-terminal residues of
the fusion protein into microsomes. The fact that M8 impaired
translocation initiated by M7 suggests that M8 functioned as a
stop-transfer signal. M8 (amino acid residues 807-826) is
predicted to span the membrane only in the 10-segment model. No
protected fragments were observed after digesting membrane integrated
forms of pM5M6, pM7M8, and of pM9M10 by proteinase K that could be
immunoprecipitated by antibodies directed against the amino- or
carboxyl-terminal residues. This is possible only if the proteins were
integrated into microsomes as polytopic integral membrane proteins (see Fig. 2). The notion that M8 is a transmembrane segment can
also be inferred from experimental observations on the
Ca Second, the largest exoplasm
domain in the Ca Of the transmembrane segments predicted as stop-transfer
signals, only M10 initiated translocation. This is not caused simply by
hydrophobicity, for hydropathy plots generated by the algorithm of
Eisenberg et al.(1984) predict M4 as the most hydrophobic of
the presumed stop-transfer signals; pM4 did not associate with
microsomes after alkaline extraction (Lin and Addison, 1995). After
surveying the literature of this field, we found no satisfactory
explanation as to why a predicted stop-transfer signal functions to
initiate translocation. It is clear that as yet we have no insight into
the intrinsic property of a stretch of hydrophobic residues making it
function as a stop-transfer or a start-transfer signal. Any putative
transmembrane segment containing a negatively charged amino acid did
not function when engineered in the fusion proteins. These are M1, M2,
M4, M6, and M9 (Lin and Addison(1995) and this report). M5 contains a
negatively and a positively charged residue seven residues apart. This
segment initiated translocation of the invertase segment into
microsomes. It is possible that the position of these charged residues
allowed for formation of a salt bridge that masked their charges, thus
creating a functional topogenic signal. M10 and M8 each contains a
positively charged residue. M10 initiated translocation of the
carboxyl-terminal residues into microsomes; M8 did not. M8 also
contains two proline residues. Apparently, the combination of charged
and proline residues impaired the ability of M8 to function
individually in the construct (Emr and Silhavy, 1983). Pairing of
transmembrane segments in the H A report on the
topological analysis of the
H A
combination of M5, M6, and M7 from the carboxyl-terminal third of the
H The major
assumption for the basis of the experiments described herein is that
transmembrane segments in fusion proteins mimic their behavior in the
native protein. In view of the similarities of the Neurospora H
Volume 270,
Number 12,
Issue of March 24, 1995 pp. 6942-6948
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
-ATPase (*)
-ATPase, we constructed fusion proteins on the cDNA
level. These contained DNA fragments encoding hydrophilic residues of
the amino and carboxyl termini of the H-ATPase with a
DNA fragment encoding the putative transmembrane segment. To report
translocation into microsomes, a DNA fragment encoding three consensus N-linked glycosylation sites was engineered carboxyl-terminal
to the putative transmembrane segment. Fusion proteins were synthesized
in a Neurospora in vitro translation system supplemented with
homologous microsomes. By the criteria of glycosylation of fusion
proteins by microsomes, sedimentation of products with microsomes after
alkaline extraction, and analysis of protected fragments generated from
proteinase K digestion of integrated products, we localized six
transmembrane segments in the carboxyl-terminal third of the
H-ATPase. These results support a 10-segment model of
the Neurospora H-ATPase.
-ATPase) (
)of Neurospora crassa belongs to a family of cation-motive porters called P-type
ATPases. A prerequisite for understanding this family of remarkable
cation-motive ATPases is knowledge of their membrane topology. This
knowledge will contribute significantly to elucidating their
structure-function relationships. A major impediment to achieving this
goal is the inability to obtain crystals of cation-motive ATPases
suitable for x-ray structural analysis. Accordingly, topological models
of these ATPases are based almost exclusively on hydropathy plots of
their primary sequence. These plots, unfortunately, are ambiguous for
regions spanning the membrane, leading to publication of different
models for the identical protein. For instance, there are three models
of the Neurospora plasma membrane H-ATPase
that have eight- (Hager et al., 1986), 10- (Addison, 1986), or
12-transmembrane segments (Rao et al., 1991). Confusion is
caused primarily by the fact that hydropathy profiles of the
carboxyl-terminal third of this enzyme as well as of other P-type
ATPases are not as clear as those of the amino-terminal third. This has
made it difficult to predict the number of transmembrane segments in
this region. Nevertheless, reliable topological models of these porters
are needed to assist in designing and interpreting mutagenesis studies.
Without structural data, these studies can provide insights into
functionally significant regions of these ATPases. To provide a model
of these ATPases, other experimental approaches will be required to
verify predicted transmembrane segments.
-ATPase of the
sarcoplasmic reticulum. The Ca-ATPase has no large
exoplasmic (trans) domain. A loop of 20 amino acids between
transmembrane segments number 7 and 8 constitutes about 50% of the mass
on the exoplasmic side (Clarke et al., 1990). By site-directed
mutagenesis of the Ca-ATPase cDNA and expression of
the resultant trancripts in an in vivo system,
Ca-binding sites were localized to predicted
transmembrane segments (Clarke et al., 1989). A low resolution
three-dimensional model of the Ca-ATPase constructed
from cryoelectron microscopy and helical image analysis supports a
10-segment model of the Ca-ATPase (Toyoshima et
al., 1993). Another P-type ATPase, the Mg-ATPase
of Salmonella typhimurium, was predicted also to have 10
segments (Smith et al., 1993). In contrast, immunoelectron
microscopy studies of the 
-subunit of the
Na,K-ATPase localized three epitopes
to the exoplasmic side (Mohraz et al., 1994). Since this was
inconsistent with current models, other models were proposed. Image
analysis of three-dimensional crystals of the
Na,K-ATPase from pig kidney and the
Ca-ATPase of sarcoplasmic reticulum suggest that
these enzymes have similar three-dimensional structures (Taylor and
Varga, 1994). The fact that there are various topological models of
this highly conserved protein family of cation-motive ATPases attests
to the absence of reliable methods that can predict and verify if a
stretch of hydrophobic residues span the membrane.-ATPase and identified an integration signal that
consists of two transmembrane segments (Lin and Addison, 1995). In this
report, we use this method to determine the number of transmembrane
segments in the carboxyl-terminal third of the
H-ATPase. The results support a 10-segment model of
the Neurospora H-ATPase (Addison, 1986).
Materials
The 
C-labeled molecular
mass markers were from Life Technologies, Inc. The other reagents were
as described previously (Lin and Addison, 1994; 1995).Plasmid Construction
All constructs contained
portions of the cDNA of the Neurospora H-ATPase cloned between HindIII and SacII sites of pBluescript II KS (Stratagene). Constructs were generated by replacing a DNA
fragment that codes for transmembrane segment number four (M4) and
replacing amino acid residues 322-354 (Lin and Addison, 1995)
with DNA fragments that encode putative transmembrane segments of the
carboxyl-terminal third of the H-ATPase as predicted
by the 10- and 12-segment models (Addison, 1986; Rao et al.,
1991). Constructions were done as described previously (Lin and
Addison, 1994, 1995). pM7InM8 contains an additional invertase segment
(In) inserted between M7 and M8. We replaced the In segment, a BamHI-XbaI fragment, in construct M7 with a BamHI-PstI DNA fragment from M3InM4, which encodes In
and a PstI-XbaI DNA fragment from construct M8, which
encodes M8 and an In carboxyl-terminal to M8. Additional amino acid
residues were introduced by the restriction sites. These were Gly-Ser
between M7 and In and Ser-Ala-Val between M8 and the In segment
amino-terminal to M8. DNA fragments were generated by polymerase chain
reaction (Scharf, 1990) and verified by dideoxyl sequencing.Other Methods
Translation of the RNA transcripts
in the presence of [
S]methionine in the Neurospora in vitro system, fractionation of the products on
an alkaline sucrose step gradient, digestion of products by proteinase
K and by N-glycosidase F, immunoprecipitation of products, and
preparation of samples for SDS-polyacrylamide gel electrophoresis were
as described (Lin and Addison, 1994; 1995).
-ATPase. The
amino and carboxyl termini of the H-ATPase flanked the
fusion proteins (Fig. 1). The primary sequence of the putative
transmembrane segments was from the 10- and 12-segment models (Addison,
1986; Rao et al., 1991). Carboxyl-terminal to the putative
transmembrane segment was placed a segment of yeast invertase with
three consensus N-linked glycosylation sites (Taussig and
Carlson, 1983) to serve as a reporter of translocation into microsomes.
RNA transcripts were generated from DNA constructs using T3 RNA
polymerase and were translated in a Neurospora in vitro translation system supplemented with homologous microsomes
(Addison, 1987). Integration and translocation events were monitored by
sedimentation of products with microsomes after alkaline extraction,
formation of glycosylated products, and protection of products from
protease digestion. Products were identified by immunoprecipitation
using rabbit polyclonal antibodies directed against carboxyl-terminal
residues of the H-ATPase.
-ATPase. Residues marked by an asterisk were introduced by recognition sequences of
restriction endonucleases in the primers of the polymerase chain
reaction. The yeast invertase fragment is represented by the thickbar with three verticallines, putative N-linked glycosylation sites. Proteins with transmembrane
segment are designated pM. The number after M represents the position of the segment in the 10-segment model,
except for M7`, which represents the seventh transmembrane segment of
the 12-segment model.
Translation and Glycosylation of the Fusion Proteins with One
Transmembrane Segment
There are three topological models of the Neurospora H-ATPase that have 8, 10, or 12
segments. The latter is inaccurate in predicting transmembrane segments
in the amino-terminal third of the H-ATPase (Lin and
Addison, 1994). Nevertheless, this model predicts six transmembrane
segments in the carboxyl-terminal third of the
H-ATPase, like the 10-segment model. Discrepancy lies
in the assignment of two of the six segments. To rigorously examine
every stretch of residues capable of serving as a transmembrane
segment, we examined the transmembrane segments as predicted by the
12-segment model by our fusion protein method.-ATPase. The sequential insertion model of polytopic
integral membrane proteins predicts that an odd-numbered transmembrane
segment initiates translocation and an even-numbered segment stops the
translocation process. Therefore, if correctly identified, an
odd-numbered segment engineered into constructs should initiate
translocation of the invertase segment into the lumen of the microsomes
(see Fig. 2).
or M
represent transmembrane
segments. The invertase segment is represented by thickbars with three perpendicularlines,
putative N-linked glycosylation sites. Smallcircles attached to these lines indicate addition of
oligosaccharide chains. Trans and Cis correspond to
exoplasmic (lumenal) and cytosolic sides of the membrane, respectively.
Depending on the construct, the fusion protein may integrate into
microsomes as a type II bitopic integral membrane protein (I)
or a polytopic integral membrane protein with the invertase segment on
the cis side (II) or on both sides (III) of the
membrane.
-ATPase as predicted by
the 12-segment model. This segment includes amino acid residues
660-681 (Rao et al., 1991). These residues were
predicted only by the 12-segment model to span the membrane. RNA
transcripts for pM7` yielded a product with a M
of
34,100 (Fig. 3, lane1). The relative
molecular mass of all fusion proteins was about 4,500 higher than
expected based on the number of residues contained in each. The reason
for this is not apparent. There was no change in the relative molecular
mass of pM7` in the absence or presence of microsomes, nor did it
sediment with microsomes after alkaline extraction (lanes1-4), suggesting that the sequence for M7` has no
topogenic information.
Membrane Topology of the Fusion Proteins
In the
presence of microsomes, pM5, pM7, and pM10 were glycosylated. To
examine the membrane topology, the products were digested by proteinase
K. In the absence of microsomes, no protected fragment was observed (Fig. 4, lane1). In the presence of
microsomes, a protected fragment with a relative molecular mass of
27,000 was observed for all constructs (lane2). When
Triton X-100 was added to dissolve the protective membrane barrier, no
fragment was observed (lane3). The protected
fragment could be immunoprecipitated by polyclonal antibodies directed
against the carboxyl-terminal residues of the H-ATPase (lane5) but not by antibodies directed against the
amino-terminal residues (lane4). This demonstrates
that the fusion proteins had assumed the orientation of a type II
bitopic integral membrane protein in the microsomes (Singer, 1990) with
carboxyl-terminal residues on the exoplasmic (trans) side and the
amino-terminal residues on the cytoplasmic (cis) side of the microsomes
(see Fig. 2).
-ATPase. Precipitates were resolved on a 12.5%
polyacrylamide gel with SDS; radioactive fragments were visualized by
fluorography. Exposure time of data represented in lanes4 and 5 is 3 times longer than that of the data in lanes1-3. Molecular mass markers are
ovalbumin, carbonic anhydrase, and 
-lactoglobulin with molecular
masses of 43,000, 29,000, and 18,400 Da,
respectively.
Translation of Fusion Proteins with Two Putative
Transmembrane Segments
pM7` did not associate with microsomes
after alkaline extraction, suggesting that these residues are not
embedded in the membrane. It is possible, however, that M7` functions
as a topogenic signal with another transmembrane segment. We
demonstrated that certain constructs containing a transmembrane segment
will not associate with microsomes unless combined with another
transmembrane segment (Lin and Addison, 1995). To determine if pM7` had
a similar requirement, we constructed pM7`M5. M5 is a presumed
initiate-transfer signal in the 10-segment model and represents the
eighth transmembrane segment in the 12-segment model. In the latter,
these residues function as a stop-transfer signal. Therefore, pM7`M5
should be integrated into the membrane as a polytopic integral membrane
protein with the termini and the invertase segment on the cis side of
the microsomes (see Fig. 2). When RNA transcripts of this
construct were translated with microsomes, this protein was
glycosylated (Fig. 5, lane4). The topmost
glycosylated product, indicated by an asterisk, had an
increase in relative molecular mass of about 6,000, suggesting the
addition of three N-linked oligosaccharide chains to the
protein. Since pM7`M5 was glycosylated, this suggests that M7` does not
span the membrane as suggested by the eight- and 10-segment models. The
quantity of glycosylated products in pM7`M5 is less than that observed
in pM5 (Fig. 3), suggesting that M7` apparently affected the
decoding of M5 in this construct. When M6, a presumed stop-transfer
signal in the 10-segment model, was combined with M5 to give pM5M6,
this protein was integrated into microsomes as judged by its resistance
to alkaline extraction, but it was not glycosylated (lanes3 and 4). Unlike the combination of M7` and M5,
these segments functioned as predicted based on their positions in the
10-segment model. When M9 and M10 were combined, these also functioned
as predicted based on their positions in the 10-segment model and gave
results similar to those observed with pM5M6 (Fig. 5, lanes1-4). When M8 was engineered downstream of M7,
glycosylation of the resultant product was reduced but not completely
blocked (compare lane4 in Fig. 3with lane4 in Fig. 5). M8 impaired the
translocation process initiated by M7, implying that M8 served as a
stop-transfer signal. Protease K digestion of membrane integrated forms
of pM5M6, pM7M8, and of pM9M10 yielded no protected fragment that could
be immunoprecipitated by polyclonal antisera directed against the
amino- or the carboxyl-terminal residues (data not shown). This
suggests that these proteins were integrated into microsomes with both
termini on the cis side of the microsomes.
-ATPase (lane5). The other fragments were not immunoprecipitated by
antibodies against either the amino- or the carboxyl-terminal residues
of the H-ATPase (lanes4 and 5). This implies that pM7InM8 had two orientations in the
membrane. One product had the orientation of a type II bitopic integral
membrane protein; the other had that of a polytopic integral membrane
protein. This demonstrates that M8 functioned as a weak stop-transfer
signal.
-ATPase. Precipitates were resolved on 12.5%
polyacrylamide with SDS; radioactive bands were visualized by
fluorography. Exposure time of data in lanes 4 and 5 is 5 times that of the data in lanes1-3.
Molecular mass markers are identical to those of Fig. 4.
-ATPase is depicted
in Fig. 7. Numbers within boxes represent the
putative boundaries of the transmembrane segments. The amino acid
residues 688-713 were not predicted to span the membrane in the
eight-segment model (Hager et al., 1986). The fact that pM5
formed a type II integral membrane protein demonstrated that these
residues can serve as a transmembrane segment. pM7` did not interact
with microsomes, nor did M7`, when combined with M5, form an
integration signal. In light of these results, M7` is not a
transmembrane segment.
-ATPase. This model is based on data
presented herein and on previously published data as mentioned in the
text. Rectangularboxes represent transmembrane
segments (M). Boundaries of transmembrane segments are based
on the amino acid sequences used in the fusion proteins. Trans and Cis correspond to exoplasmic and cytoplasmic sides of
the membrane, respectively. Length of lines is not proportional to the
number of amino acids predicted to lie
therein.
-ATPase of sarcoplasmic reticulum. First,
site-directed mutagenesis studies on the Ca-ATPase
identified Glu-908, within M8 of the Ca-ATPase, as a
ligand-binding site (Clarke et al., 1989). This residue is
conserved among the cation-motive ATPases. In the Neurospora H-ATPase, this residue is Glu-805. In the
10-segment model, the amino-terminal boundary of M8 is two residues
carboxyl-terminal to Glu-805. If this residue has a function similar to
that of Glu-908 in the Ca-ATPase, the boundary of M8
must extend beyond what was originally proposed. pM7M8 was engineered
to contain Glu-805 and all native residues of the
H-ATPase between M7 and M8 (see Fig. 1). In
constructing pM7InM8, we used, however, amino acid residues
807-826 for M8. These residues may represent only a portion of
the actual M8. If so, this would explain why M8 in pM7InM8 functioned
as a weak stop-transfer signal. The fusion protein method can not
resolve boundaries of transmembrane segments. Therefore, another
approach will be required before it can be stated with certainty which
amino acids form the boundaries of M8.-ATPase exists between M7 and M8.
Only in the 10-segment model of the Neurospora H-ATPase will a similar loop exist. Taken
together, this supports the conjecture that M8 is a transmembrane
segment.-ATPase will juxtapose
only two pair of segments with opposite charged residues. A simple,
planar model of these segments positioned the charged residues near the
ends of each pair. For example, in the pair M9 and M10, the negatively
charged residue in M9 is positioned near the cis side of the membrane,
while the positively charged residue in M10 is near the trans side.
Clearly, the ion-pairing of segments cannot be the basis of the driving
force to forming the putative paired segments.,K-ATPase (Bamberg and Sachs, 1994)
used a method similar to the one developed in this laboratory to
identify transmembrane segments of this P-type ATPase. The results were
consistent with a 10-segment model. In this ATPase, certain
transmembrane segments did not function as predicted by the sequential
insertion model of polytopic integral membrane proteins. For example,
M5 of the H,K-ATPase, a presumed
initiate-transfer signal, did not initiate translocation of the
carboxyl-terminal residues of the fusion protein into the lumen of
microsomes. Similar observations were reported for other polytopic
integral membrane proteins (Audigier et al., 1987; Ebele et al., 1987; Sengstag et al., 1990). It is assumed,
without any experimental data, that in the topogenesis of P-type
ATPases and perhaps of other porters, transmembrane segments are
decoded individually. Therefore, segments not behaving as predicted by
the sequential insertion model are posited to be integrated into
membrane by another mechanism. If this notion is correct, a particular
polytopic integral membrane protein would use multiple integration
mechanisms and perhaps multiple integration machineries to obtain its
membrane topology. While this explanation may be correct, it should be
kept in mind that it assumes that each transmembrane segment is decoded
individually. This assumption is not valid if transmembrane segments
acted concertedly as they are integrated into the membrane. For
instance, when assayed individually, M9 and M10 did not function as
predicted by the sequential insertion model. When these segments were
combined, pM9M10 was integrated into microsomes as a polytopic integral
membrane protein, suggesting that M9 and M10 acted concertedly in the
integration event. Since transmembrane segments of the Neurospora H-ATPase are separated by a few residues, it is
possible that closely juxtaposed segments are decoded as a unit and are
then integrated into the membrane (Lin and Addison, 1995).,K-ATPase could not integrate
co-translationally into microsomes (Bamberg and Sachs, 1994). This
phenomenon was not observed with segments from the Neurospora H-ATPase. Fusion proteins of the
H-ATPase with various combinations of transmembrane
segments were all integrated co-translationally into microsomes (data
not shown). This apparent discrepancy between the systems may be
explained on the basis of differences in the in vitro systems.
We investigated integration of transmembrane segments from the Neurospora H-ATPase into their cognate
membrane. In the analysis of the
H,K-ATPase, an in vitro system composed of heterologous components was used. Also, we
engineered carboxyl-terminal to the putative transmembrane segment a
segment of invertase, a secretory protein, with the amino and carboxyl
termini of the H-ATPase flanking the fusion proteins.
In constructing the fusion proteins of the
H,K-ATPase, the carboxyl terminus of
the fusion proteins is a portion of the -subunit of the
H,K-ATPase engineered
carboxyl-terminal to the putative transmembrane segments. The function
of the -subunit is unknown. It may participant in the topogenesis
of the -subunit. If so, it is possible the -subunit has
specific interactions with certain transmembrane segments of the
-subunit. This is another potential factor that could explain the
differences between the systems. Clearly, in constructing fusion
proteins, a reporter sequence should be chosen that has no potential
role in the topogenesis of the protein under study. The combined
effects of a heterologous in vitro system and the use of
-subunit in the fusion proteins can probably account for the
apparent differences in the behavior of the transmem-brane segments of
the P-type ATPases observed in the two laboratories.-ATPase to the other P-type ATPases, it can be
anticipated that what is learned about the integration mechanism of the Neurospora H-ATPase will contribute to our
understanding of the integration mechanism of P-type ATPases in
general.
)-ATPase, plasma membrane electrogenic,
proton-translocating ATPase.
©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:
|
|
 |
|
  T. M. Buck and W. R. Skach Differential Stability of Biogenesis Intermediates Reveals a Common Pathway for Aquaporin-1 Topological Maturation J. Biol. Chem., January 7, 2005; 280(1): 261 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Gupta, N. D. DeWitt, K. E. Allen, and C. W. Slayman Evidence for a Salt Bridge between Transmembrane Segments 5 and 6 of the Yeast Plasma-membrane H+-ATPase J. Biol. Chem., December 18, 1998; 273(51): 34328 - 34334. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. V. Moller, G. Ning, A. B. Maunsbach, K. Fujimoto, K. Asai, B. Juul, Y.-J. Lee, A. G. de Gracia, P. Falson, and M. le Maire Probing of the Membrane Topology of Sarcoplasmic Reticulum Ca2+-ATPase with Sequence-specific Antibodies. EVIDENCE FOR PLASTICITY OF THE C-TERMINAL DOMAIN J. Biol. Chem., November 14, 1997; 272(46): 29015 - 29032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Fiedler and G. Scheiner-Bobis Transmembrane Topology of alpha - and beta -Subunits of Na+,K+-ATPase Derived from beta -Galactosidase Fusion Proteins Expressed in Yeast J. Biol. Chem., November 15, 1996; 271(46): 29312 - 29320. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xie, S. A. Langhans-Rajasekaran, D. Bellovino, and T. Morimoto Only the First and the Last Hydrophobic Segments in the COOH-terminal Third of Na,K-ATPase alpha Subunit Initiate and Halt, Respectively, Membrane Translocation of the Newly Synthesized Polypeptide J. Biol. Chem., February 2, 1996; 271(5): 2563 - 2573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Seto-Young, M. J. Hall, S. Na, J. E. Haber, and D. S. Perlin Genetic Probing of the First and Second Transmembrane Helices of the Plasma Membrane H[IMAGE]-ATPase from Saccharomyces cerevisiae J. Biol. Chem., January 5, 1996; 271(1): 581 - 587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bayle, D. Weeks, and G. Sachs The Membrane Topology of the Rat Sarcoplasmic and Endoplasmic Reticulum Calcium ATPases by in Vitro Translation Scanning J. Biol. Chem., October 27, 1995; 270(43): 25678 - 25684. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |
ABSTRACT
