|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 276, Issue 39, 36543-36549, September 28, 2001
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of
Received for publication, June 1, 2001, and in revised form, July 6, 2001
Fps1p is a glycerol efflux channel from
Saccharomyces cerevisiae. In this atypical major intrinsic
protein neither of the signature NPA motifs of the family, which
are part of the pore, is preserved. To understand the functional
consequences of this feature, we analyzed the pseudo-NPA motifs of
Fps1p by site-directed mutagenesis and assayed the resultant mutant
proteins in vivo. In addition, we took advantage of the
fact that the closest bacterial homolog of Fps1p, Escherichia
coli GlpF, can be functionally expressed in yeast, thus enabling
the analysis in yeast cells of mutations that make this typical major
intrinsic protein more similar to Fps1p. We observed that mutations
made in Fps1p to "restore" the signature NPA motifs did not
substantially affect channel function. In contrast, when GlpF was
mutated to resemble Fps1p, all mutants had reduced activity compared
with wild type. We rationalized these data by constructing models of
one GlpF mutant and of the transmembrane core of Fps1p. Our model
predicts that the pore of Fps1p is more flexible than that of GlpF. We
discuss the fact that this may accommodate the divergent NPA motifs of
Fps1p and that the different pore structures of Fps1p and GlpF may
reflect the physiological roles of the two glycerol facilitators.
Major intrinsic protein
(MIP)1 channels, aquaporins
and glycerol facilitators, occur in all types of organisms ranging from bacteria to humans. MIP channels mediate the gradient-driven diffusion of water, glycerol, and other uncharged compounds and certain ions
across biological membranes. To date, more than 200 MIPs have been
identified comprising an amazing number of isoforms expressed in
different subcellular compartments and tissues, under different
environmental conditions or during different developmental stages. For
example, 10 genes encoding MIP channels are currently known in humans
(1), 35 have been described in the plant Arabidopsis thaliana (2) and the nematode Caenorhabditis elegans
has nine (3). These channel proteins are involved in biological
processes as diverse as urine concentration in the mammalian kidney,
(4) root development in A. thaliana (5, 6), and
osmoregulation in Saccharomyces cerevisiae (7, 8).
Fps1p is a glycerol facilitator of the yeast S. cerevisiae
(7, 8). The central importance of Fps1p in yeast osmoregulation is
illustrated by the sensitivity to a hypo-osmotic shock of yeast cells
in which FPS1 has been deleted. Fps1p has also been shown to
affect signal transduction in yeast osmoregulation in a fashion consistent with its role in controlling the intracellular glycerol content (9, 10). In addition, fps1 Although Fps1p (see Fig. 1) is closely related to bacterial glycerol
channels such as GlpF from E. coli, it is so far unique in
the MIP family for a number of reasons (11). For example, it has long
amino- and carboxyl-terminal hydrophilic extensions, resulting in a
protein of 669 amino acids, compared with 281 amino acids for GlpF.
Only very few fungal MIPs, as well as Drosophila BIB, have
such long extensions. The amino-terminal extension of Fps1p is
important for the regulation of glycerol efflux by external osmolarity
in yeast. Deletion of this domain renders the channel unregulated.
Unregulatable, constitutively open Fps1p causes loss of glycerol from
the cell and hence poor growth in high osmolarity medium (8). As an
additional unique feature, neither of the signature Asn-Pro-Ala (NPA)
motifs of the family in channel-forming loops B and E is fully
preserved; instead they are Asn-Pro-Ser (NPS) and Asn-Leu-Ala (NLA),
respectively (see Fig. 1). In fact, the sequences of only five other
microbial MIPs contain motifs other than NPA. In four of these cases
NPA is preserved in loop B: the presumed glycerol channel from
Enterococcus faecalis (gef 6176) contains an NQA motif in
loop E, the putative aquaporin from Chloribium tepidum (gct
5) has NPV, the Botrytis cinerea putative glycerol channel
contains NPS, and the Saccharomyces kluyveri homolog of
Fps1p has NMA (11, natchaug.labri.u-bordeaux.fr/Genolevures/Genolevures.php3). The only
deviation from NPA in loop B of a microbial MIP occurs in the
Salmonella typhimurium protein (contig 1308), which has an
NLA motif. Unfortunately, these MIPs are incompletely characterized, and thus their transport characteristics cannot be used to aid our
understanding of the structural and/or functional role(s) of these
features. In this study we have therefore chosen to examine the
consequences of the NPS and NLA sequences of Fps1p on channel function.
This is of particular relevance to a general understanding of MIP
channels because their generic NPA motifs have been shown to be
integral to the formation of a continuous solute pore, the so-called
"hourglass" (17), a prediction confirmed in the recently published
structures of GlpF (18) and AQP1 (19).
To understand the functional consequences of their atypical sequence
composition, we have probed loops B and E of Fps1p by site-directed
mutagenesis and in vivo assay. Specifically, we have studied
the consequences of "restoring" NPA motifs in Fps1p. In addition,
we have taken advantage of the fact that the closest bacterial homolog
of Fps1p, E. coli GlpF, which contains two NPA motifs, and
for which a crystal structure has been published (18), can be
functionally expressed in yeast (7). Hence, we have generated mutations
that make this typical MIP resemble Fps1p and tested the function of
these mutant proteins in the same yeast test system. We find that
although Fps1p tolerates NPA in both loops B and E instead of its
unusual pseudo-NPA motifs, GlpF only tolerates NPS in loop B. Changes
in loop E of GlpF abolish protein function as determined in our yeast
test system. We have rationalized these data by generating models of
GlpF containing NPS in loop B and of the transmembrane regions of
Fps1p, using the published structure of GlpF as a starting point (18).
We discuss our results in the context of the pore of Fps1p having a
different structure from that of GlpF. This may accommodate the
divergent NPA motifs Fps1p, and moreover, the different pore structures
of Fps1p and GlpF may reflect the different physiological roles of the
two glycerol facilitators.
Yeast Strains and Growth Conditions--
The S. cerevisiae strains used in this study were isogenic to W303-1A
(20). The deletion mutant fps1 Plasmid Construction--
YEpmyc-FPS1 is a
2µ LEU2 plasmid expressing a fusion protein in which the
c-myc epitope is attached to the carboxyl terminus of Fps1p
(8). Plasmids containing point mutations in the FPS1 gene
were constructed using the megaprimer PCR (23) with
YEpmyc-FPS1 as template and primers as listed in
Table I. Single mutations were introduced
using primers S354A.f or L481P.f. The flanking primers in each case
were M1 and M2. The resultant PCR products were cleaved with
KpnI and ApaI and subcloned into
YEpmyc-FPS1 that had been cleaved with the same
enzymes. All constructs were fully sequenced using the BigDye
Terminator Cycle Sequencing kit (Applied Biosystems). For the
double mutant, two fragments were created; primers M1 and M3 were used
to generate a PCR product using YEpmyc-FPS1L481P
as the template, and YEpmyc-FPS1S354A was digested with NcoI. These fragments were used to transform
S. cerevisiae using the lithium acetate method (24), and the
resulting gap-repaired plasmids (25) were propagated in E. coli DH5 Glycerol Transport Assays--
For glycerol efflux measurements,
the cells were cultured in YNB supplemented with 2% glucose to mid log
phase (typically A600 nm = 0.4-0.5). To
achieve maximal internal glycerol accumulation, the cells expressing
FPS1 were then incubated for 2-3 h in YNB supplemented with
2% glucose and 5% NaCl, whereas cells expressing glpF were
incubated for 5-6 h in the NaCl-containing medium. The cells were
harvested and suspended in YNB supplemented with 2% glucose at
t = 0. The samples (5 ml) were harvested at t = 0, 1, 2, 3, 5, and 10 min, and glycerol efflux was
measured as a function of time as described previously (8). Glycerol concentrations were determined enzymatically using a commercially available kit (Roche Molecular Biochemicals).
To determine glycerol influx, the cells were cultured in YNB
supplemented with 2% glucose to log phase (typically
A600 nm = 0.7-0.8). The cells were harvested
by centrifugation, washed, and suspended in ice-cold MES buffer (10 mM MES, pH 6.0) to a density of 60 mg/ml. Glycerol influx
was then measured as described previously (7, 8) by withdrawing 50-µl
aliquots and collecting cells by filtration. In each case, the dry
weight was determined by harvesting cells on filters, which were then
dried at 80 °C to constant weight.
All transport experiments were performed a minimum of three times, and
the data were expressed as the mean values at each time point ± S.E. A standard two-tailed unpaired t test was used to test
for significant differences between the means, as necessary. In
general, transport data were not related to expression levels to give
specific activities because the standard error associated with membrane
protein expression levels was much higher than that of the transport
assays themselves. Hence it was felt that such adjustment is
misleading, as has previously been reported in studies where levels in
isolated membranes do not reflect those in whole cells (27).
Membrane Preparation and Immunoblots--
The cells were
cultured in YNB supplemented with 2% glucose to late log phase
(typically A600 nm = 0.8), and both the total
membrane and plasma membrane fractions were isolated and visualized as
described previously (8).
Molecular Modeling--
Models of the A70S mutation in GlpF were
generated by extraction of the C Changing NPS and/or NLA to NPA in Fps1p Does Not Affect Glycerol
Transport--
Fps1p differs from the vast majority of its family
members in that it contains NPS and NLA, respectively, in loops B and E (Fig. 1), rather than the signature
sequences of the family: two NPA motifs. We began our analysis by
restoring these amino acid triplets both singly and doubly to NPA. The
mutant genes were expressed from a multi-copy plasmid under the control
of the FPS1 promoter in a yeast strain lacking
FPS1. The localization of the resultant mutant proteins was
examined by cell fractionation and immunoblot. Fig.
2A indicates that wild type
Fps1p and the three mutants were equally localized to the plasma
membrane. Glycerol efflux was assayed by measuring the export of
intracellularly produced glycerol, whereas influx was measured
following the exposure of cells to radiolabeled glycerol. We observed
no statistically significant reduction in glycerol efflux compared with
the wild type (Fig. 2B) for any of our mutant channels. This
was consistent with the growth phenotypes shown in Fig. 2C.
In these tests, yeast cells with a channel that can export glycerol
survive a hypo-osmotic shock, whereas a smaller fraction of yeast cells
survive this treatment and also resume growth more slowly in the
absence of a functional channel (8). Hence we observed that cells with the mutant channels Fps1p S354A, Fps1p L481P, and Fps1p S354A L481P
survived indistinguishably from cells with a wild type channel. Table
II shows that the initial glycerol uptake
rates for the mutant channels were also not significantly different
from the wild type, with rates in the range of 85-116% of the mean
wild type value. This confirmed that changing NPS and/or NLA to NPA in
Fps1p does not substantially affect its glycerol transport properties.
In addition, we note that cells with mutant channels did not show any
growth defect in high osmolarity medium such as 1 M
sorbitol (Fig. 2C) or high salt (not shown), which would otherwise indicate changes in channel regulation.
Changing NPA to NPS and/or NLA in GlpF Affects Glycerol
Transport--
It was clear from the experiments described above that
in our assays neither NPS nor NLA is specifically required for
apparently normal glycerol transport by Fps1p. To begin to understand
this observation, we examined the functional consequences of converting the NPA to NPS and/or NLA in GlpF, which is a more typical member of
the MIP family in terms of sequence. We have shown previously that
GlpF, which is the closest bacterial homolog of Fps1p, can be
functionally expressed in S. cerevisiae. Although cells
expressing the unregulatable GlpF channel initially lose glycerol and
are slightly osmosensitive (8), we observed that after prolonged incubation GlpF-expressing transformants did accumulate glycerol in
hyperosmotic medium (7). This allowed the GlpF constructs to be tested
in exactly the same way as the Fps1p mutants described above.
We therefore constructed NPA to NPS in loop B (A70S), as well as NPA to
NLA in loop E (P204L) and the double NPS/NLA mutant (A70S P204L). These
mutants were analyzed by cell fractionation and immunoblot for
localization (Fig. 3A) and in
both our efflux (Fig. 3B) and influx (Table II) assays. In
contrast to the Fps1p series, we observed that mutations in GlpF
affected protein localization differently in each case. GlpF A70S was
expressed on average at higher levels than wild type, GlpF P204L was
expressed at wild type levels, and the double mutant, GlpF A70S P204L,
was poorly expressed (Fig. 3A). With respect to our
transport assays (Fig. 3B and Table II), GlpF A70S showed
substantial activity, having 75% of the mean uptake rate of wild type
GlpF and approximately wild type export activity. Given the apparent
higher expression level, this may indicate a reduced specific transport
activity per channel. However, transport proficiency of this mutant was supported by growth phenotypes; cells with the mutant channel A70S and
the wild type channel survived a hypo-osmotic shock, indicating that
only these channels were functional (Fig. 3C). For GlpF
P204L, this nonconservative mutation in loop E apparently resulted in a
properly localized but essentially nonfunctional protein. Similarly,
the double mutant was also nonfunctional, although at least some
protein appeared to be localized to the plasma membrane. From these
observations we concluded that the pores of Fps1p and GlpF were not
structurally identical; although Fps1p containing two NPA motifs
appeared to have wild type properties in our assays (Fig. 2), the
reverse was not true for GlpF, because only NPS in loop B could be
tolerated in the GlpF pore (Fig. 3).
Loops B and E of Fps1p Do Not Form a Functional Channel in
GlpF--
The apparent inability of GlpF to tolerate NLA in loop E
indicated that there might be sequences in the B and E loops of Fps1p that are required if the interaction of these NPS- and NLA-containing loops are to form a fully functional channel in GlpF. We therefore precisely aligned the sequences of Fps1p and GlpF using the internet resources, Prosis, TopPred2, TMPred, PhD, and DAS, together with structural information for GlpF (18), and transferred the predicted loops B and E from Fps1p to GlpF both singly and doubly, to make GlpF
even more Fps1p-like. None of these mutants conferred any glycerol
transport activity (Fig. 4 and Table II).
For the E-loop swap, in particular, this is probably because it was not
properly expressed or localized to the plasma membrane. The double loop swap was expressed to wild type levels but was nonfunctional, indicating that GlpF could not tolerate the channel-forming loops of
Fps1p. Again, growth tests confirmed the biochemical analyses of the
channels; only cells with the wild type channel survived a hypo-osmotic
shock, indicating that only this channel confers glycerol export (Fig.
4C). These observations indicate that in addition to loops B
and E, other sequences, probably within the transmembrane domains,
contribute to functional differences between Fps1p and GlpF.
Our mutational studies and analyses in whole yeast cells have
shown that Fps1p can tolerate conventional MIP channel NPA motifs. Changing NPS in loop B and NLA in loop E either singly or doubly into
NPA either did not affect, or only marginally affected, localization and function of the mutant protein. We noted that although the single
Fps1p mutants had somewhat reduced activities compared with wild type,
the double mutant was wild type-like (Table II). In contrast, making
the reverse mutations in the prototypic MIP glycerol facilitator, GlpF,
had pronounced effects. Although NPS in loop B affected but did not
abolish function, the presence of NLA in loop E or the swapping of
loops B and/or E of Fps1p to GlpF abolished function and also affected
localization to the plasma membrane. These observations indicated that
the pores of the two glycerol facilitators, Fps1p and GlpF, are different.
To rationalize our experimental observations, we generated a model of
the transmembrane core of Fps1p based on the aligned regions of the
GlpF structure reported by Fu and colleagues (18). Fig.
5 indicates marked differences between
the pore regions of GlpF and Fps1p. In particular, Fps1p contains
significantly less It is possible that the minor reduction in initial glycerol uptake rate
for Fps1p S354A is due to disruption of the polar environment of the
pore, because alanine is both bulkier and less polar than serine. In
particular potential hydrogen bonding between Ser354 and
Glu377 of transmembrane domain 3 would be abolished. For
Fps1p L481P, the slight reduction in initial glycerol uptake rate may
be due to loop E becoming less flexible in a region crucial for
transport activity. In addition, hydrogen bonding may again be
affected; the introduction of a proline reduces the potential for
hydrogen bonding within the pore, which might be important for
maintaining optimal pore architecture and loop/helix proximities.
Interestingly, the double mutant Fps1p S354A L481P has a wild type-like
initial glycerol uptake rate. In this case the mutant channel contains the MIP family signature NPA/NPA sequence. It appears that in the
absence of the Ser354/Glu377 interaction, the
double mutation causes a realignment of polar groups and hydrogen
bonding within the pore. This compensates for the loss of the serine
and allows efficient tracking of the substrate through the pore. For
instance, it is likely that Asn480 on loop E can come
closer to Asn352 on loop B, allowing the formation of a
favorable hydrogen bond. Hence in the case of Fps1p, NPA/NPA functions
as well as NPS/NLA, suggesting that overall, the ability of a pore to
transport glycerol appears to be a balance between polarity and
stabilizing hydrogen bonds within the pore itself.
In GlpF, we observed that modifications restricting the already limited
pore flexibility resulted in a significant reduction in transport
function. In particular, our model of GlpF A70S (Fig. 6) based on the coordinates of the GlpF
structure (18) may be used to rationalize the apparently somewhat
reduced transport rates incurred by the mutation (taking into account
the apparently increased expression level of this construct). The
introduction of a serine residue at position 70 appears to result in
slightly reduced flexibility of Asn68 because of increased
hydrogen bonding between the side chain oxygen atom of
Ser70 and the side chain nitrogen of Asn68
(4.16 Å apart). However, these effects are likely to be offset by an
increased polar environment within the pore, resulting in only small
effects on transport. In contrast, transport function in GlpF is
significantly compromised by other substitutions, such as the loop E
P204L substitution. This is likely to involve the role of proline as a
"helix breaking" residue, its loss in the mutant leading to a
lengthened helical section in loop E, incorporating the mutant NLA
motif. This alteration most probably significantly reduces the
flexibility of the surrounding amino acids, compared with the situation
when an NPA motif is present.
Analysis of the Pore of the Unusual Major Intrinsic
Protein Channel, Yeast Fps1p*
§,,
,
Cell and Molecular
Biology/Microbiology, Göteborg University, S-40530
Göteborg, Sweden, the ¶ Department of Molecular
Biotechnology, Chalmers University of Technology, S-40530
Göteborg, Sweden, the Department of Biology and Health
Science, University of Luton, Park Square, Luton, Bedfordshire, LU1
3JU, United Kingdom, and the ** Department of Biochemistry
and Biophysics, Göteborg University,
S-40530 Göteborg, Sweden
ABSTRACT
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
mutants both grow
poorly and hyperaccumulate glycerol under anaerobic conditions when
glycerol is produced for redox balancing (8). These observations
establish that Fps1p primarily mediates glycerol diffusion out of the
cell (11) with an energy-independent mechanism (8). Despite the fact
that Fps1p can also mediate measurable glycerol influx into yeast
cells, (7, 8, 12), this is probably not its physiological role because
no growth defects have been found in fps1 mutants cultivated with glycerol as the sole carbon source (8). Furthermore GUP1 and GUP2 have recently been proposed to be
involved in active transport of glycerol in yeasts (13). In contrast,
the physiologically relevant role of GlpF from Escherichia
coli, the closest bacterial homolog of Fps1p with 30.5% identity,
is likely to be import of glycerol for catabolism (14, 15).
EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
::TRP1
was made by replacing a BglII fragment containing the
FPS1 open reading frame with a BamHI fragment
containing the TRP1 gene from YDpW (21). Yeast cells were
routinely grown in medium containing 2% peptone and 1% yeast extract
supplemented with 2% glucose as carbon source. Selection and growth of
transformants carrying a replicating plasmid was performed in yeast
nitrogen base medium (YNB) (22). Plate growth assays were performed by
pregrowing the cells to midlog phase in medium supplemented with 1 M sorbitol. A 10-fold serial dilution of this culture was
made, and 10 µl of each dilution were spotted onto agar plates
supplemented without osmoticum (hypo-osmotic shock). Growth was
monitored after 2-3 days at 30 °C.
. This construct was fully sequenced using the BigDye
Terminator Cycle Sequencing kit (Applied Biosystems).
YEpglpF contains the gene encoding E. coli GlpF
under the control of the yeast PGK1 promoter (7). The gene
was tagged at its 3' end by using megaprimer PCR with primers MluI.f,
RsrII.f, and SexAI.r (Table I) to create an MluI site at
positions 833-838, yielding YEpglpF-MluI. A
synthetic fragment was created by annealing the oligonucleotides myc.f
and myc.r (Table I), which contain the c-myc sequence with
appropriate 5' overhangs, and this was subcloned into
YEpglpF-MluI which had been cleaved with
MluI and SexAI. This procedure yielded
YEpglpFmyc. Plasmids containing point mutations in the
glpF gene were constructed using PCR with
YEpglpFmyc as template and the primer pair A70S.f and
NgoMI.r or P204L.r and NgoMI.f (Table I). Loop swaps were created using
PCR by completely amplifying YEpglpFmyc except for the loop
to be replaced by the corresponding sequence in FPS1. This
corresponding FPS1 sequence was then amplified using
YEpmyc-FPS1 as template with primers containing
5' 30-base extensions with 100% identity to the required
glpF flanking sequence (Table I). These fragments were then
used to transform S. cerevisiae using the lithium acetate
method (24), and the resulting gap-repaired plasmids (25) were
propagated in E. coli DH5. All constructs were fully
sequenced using the BigDye Terminator Cycle Sequencing kit (Applied
Biosystems). All other molecular biological manipulations were
performed using standard techniques, as described in Ref. 26.
Oligonucleotides used for site-directed mutagenesis of FPS1
and glpF
atom coordinates of E. coli GlpF from the Brookhaven Protein Data Bank file, 1FX8, using
Molscript commands in RASMOL version 2.7.2 (28). The mutant model was
subsequently constructed using the MaxSprout algorithm (29), following
substitution of Ala for Ser in the C coordinates file. The
coordinates for Ser70 were then returned to the original
file, substituting the Ala70 coordinates, resulting in a
file that was identical to the original GlpF Brookhaven Protein Data
Bank file, except for the additional oxygen atom of the serine residue.
Hydrogen bonding was predicted using Swiss-PDB Viewer (30). The amino
acid sequences for yeast Fps1p and E. coli GlpF were aligned
using CLUSTALW (31). Models of the Fps1p pore were generated by
extraction of the C atom coordinates of the GlpF structure,
replacement with corresponding amino acid residues of the aligned Fps1p
sequence, and reconstruction of the pore using the MaxSprout algorithm.
The regions of changed secondary structure were corroborated by
analysis of the Fps1p sequence using the -helix indices of Chou and
Fasman (32) and Deleage and Roux (32, 33) and by analysis of the
modeled structure using TMAlpha (34).
RESULTS
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
View larger version (24K):
[in a new window]
Fig. 1.
Schematic diagram of Fps1p. Fps1p
differs from more typical members of the MIP family, such as GlpF, in a
number of ways. For example the signature NPA motifs of the family are
replaced by NPS and NLA in Fps1p, as indicated. The numbers
indicate the six transmembrane domains. The letters A-E
denote the loops of Fps1p, of which B and E are involved in the
formation of the glycerol channel by dipping into the membrane to form
an hourglass. The dashed lines within the amino- and
carboxyl-terminal extensions indicate that they are much longer than
shown (256 and 143 amino acids, respectively). The membrane core
comprises 270 amino acids.
View larger version (36K):
[in a new window]
Fig. 2.
Functional analysis of point mutants of
Fps1p. A, Western blots of plasma membrane extracts.
Representative blots are shown. B, glycerol transport
measurements on whole yeast cells were performed in triplicate for
efflux to determine the transport characteristics of cells with either
Fps1p ( 
), no glycerol channel (
), Fps1p S354A (×), Fps1p L481P
(
), or Fps1p S354A L481P (
). The mean values are presented ± S.E. (n = 3). C, growth characteristics
following hypo-osmotic shock are consistent with the efflux properties
of the mutant proteins. The cells were spotted onto plates in a 10-fold
serial dilution series such that they experienced either no shock
(left panel) or a hypo-osmotic shock (right
panel).
Initial glycerol uptake rates
View larger version (43K):
[in a new window]
Fig. 3.
Functional analysis of point mutants of
GlpF. A, Western blots of plasma membrane extracts.
Representative blots are shown. B, glycerol transport
measurements on whole yeast cells were performed in triplicate for
efflux to determine the transport characteristics of cells with either
GlpF ( ), no glycerol channel (), GlpF A70S (×),GlpF P204L (),
or GlpF A70S P204L (). The mean values are presented ± S.E.
(n = 3). C, growth characteristics following
hypo-osmotic shock are consistent with the efflux properties of the
mutant proteins. The cells were spotted onto plates in a 10-fold serial
dilution series such that they experienced either no shock (left
panel) or a hypo-osmotic shock (right panel).
View larger version (42K):
[in a new window]
Fig. 4.
Functional analysis of loop swap mutants of
GlpF. A, Western blots of plasma membrane extracts.
Representative blots are shown. B, glycerol transport
measurements on whole yeast cells were performed in triplicate for
efflux to determine the transport characteristics of cells with either
GlpF ( ), no channel (), GlpF in which loop B had been exchanged
for that of Fps1p (×), GlpF in which loop E had been exchanged for
that of Fps1p (), or GlpF in which both loops B and loop E had been
exchanged for those of Fps1p (). The mean values are presented ± S.E. (n = 3). C, growth characteristics
following hypo-osmotic shock are consistent with the efflux properties
of the mutant proteins. The cells were spotted onto plates in a 10-fold
serial dilution series such that they experienced either no shock
(left panel) or a hypo-osmotic shock (right
panel).
DISCUSSION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
-chain around the inner part of the pore.
Consequently, our observations can be rationalized in terms of the
comparatively greater flexibility of the Fps1p pore with respect to
GlpF. By comparing the sequences flanking the NPA of loop E of GlpF
with the corresponding NLA in Fps1p, it is clear that the latter also
contribute to this flexibility. Specifically, it appears that the GlpF
sequence PLTGFAMNPARD provides for a strong -helix between the
prolines, whereas the equivalent Fps1p sequence QTGTAAMNLARD has no
such provision. Another consequence of the lower -helical content is
that the region where loops B and E approach each other in Fps1p,
Ser354/Asn480, is more exposed than the
equivalent Ala70/Asn203 in GlpF.
View larger version (45K):
[in a new window]
Fig. 5.
Homology model of the Fps1p pore region
(right) based on the structure of GlpF
(left) displayed in RASMOL. The Fps1p pore
displays markedly less conformation (shown in pink),
resulting in a more flexible structure. The higher degree of
random conformation (shown in white) is most evident around
the inner pore, particularly in the region of the
Ser354/Asn480 loop B/loop E approach (shown in
yellow), which is subsequently more exposed than the
equivalent Ala70/Asn203 in GlpF.
View larger version (46K):
[in a new window]
Fig. 6.
Modeling of the A70S mutation in GlpF,
displayed in RASMOL. The wild type is shown in the left
panel, with Asn68 (center) and Ala 70 (left) rendered. The A70S form is shown in the right
panel and displays the probable hydrogen bond formation
(black line) between Asn68 and Ser70
because of the additional side chain oxygen of Ser70,
limiting the flexibility of Asn68.
In summary, it is clear that the pores of GlpF and Fps1p are different, even though both proteins are members of the "glycerol facilitator" subfamily of the MIP superfamily. The rigid pore of GlpF is sensitive to changes in its signature NPA motifs and cannot accommodate the B and E loops of Fps1p, whereas the pore of Fps1p is more permissive to change. Hence, it is probable that the presence of NPS and NLA, rather than the more typical NPA and NPA, in the pore of Fps1p are tolerated by the more flexible structure.
Why have these two glycerol facilitators evolved different pore designs? A possible explanation may be the fact that Fps1p and GlpF have different physiological roles. Although Fps1p appears to be implicated in the export of glycerol during osmoadaptation (7, 9), GlpF functions in the uptake of glycerol from the growth medium as a first step to catabolize it as a source of carbon and energy (14, 35). However, in addition to their common glycerol transport functions, Fps1p and GlpF each have other unique roles. Fps1p is gated rapidly by changes in medium osmolarity (8) to prevent glycerol leakage and to ensure appropriate intracellular glycerol accumulation. GlpF, on the other hand, interacts with glycerol kinase to ensure rapid glycerol phosphorylation, thereby preventing any glycerol taken up being lost again by diffusion out of the cell (15). We do not yet understand the mechanism that mediates gating in Fps1p, although we have shown previously that certain parts of the amino-terminal extension are needed for channel closing (8). We have also observed that addition of these sequences to GlpF, which is unregulated, is not sufficient for its gating, resulting in a protein with properties that are indistinguishable from wild type GlpF.2 What is clear, however, is that Fps1p responds within seconds to osmotic stress (8), suggesting a conformational change is involved in gating. The flexible pore of Fps1p could be permissive to such a conformational change leading to closure of the pore by a mechanism that remains to be identified. Structural modeling, as employed in this study, may aid us in achieving a better understanding of this mechanism.
A more flexible pore may also have consequences for transport
specificity. Although both proteins appear to transport a similar spectrum of smaller polyols
(18),3 different transport
capacities of Fps1p and GlpF for the polyol-like oxoanions of As(III)
and Sb(III) have been reported. Fps1p has been shown to mediate the
uptake of As(III) and Sb(III) into S. cerevisiae. Through
the regulation of this uptake, the channel has been suggested to have a
role in toxic metalloid resistance in yeast (36). In the same study, no
equivalent uptake could be demonstrated for GlpF, although GlpF has
already been suggested to transport Sb(III), but not As (III), in
E. coli (16), thus indicating a specificity difference
between these close homologs. This hypothesis awaits a more thorough
functional analysis of the transport specificity of Fps1p, which is
currently in progress. Comparison of these data with data emerging on
the recently identified, but as yet uncharacterized, homologs of Fps1p
from different yeast species
(natchaug.labri.u-bordeaux.fr/Genolevures/Genolevures.php3) should shed
light on the intriguing presence of atypical signature motifs in these
unusual MIP family members.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Dr. Cândida Lucas (Universidade do Minho, Braga, Portugal) and Dr. Markus Tamás (Göteborg University) for helpful discussions and Aleksandras Gutmanas (Göteborg University) for assistance with Figs. 5 and 6.
| |
FOOTNOTES |
|---|
* This work was supported by European Commission Contracts BIO4-CT98-0024 (to S. H. and J. R.) and FMRX-CT96-0128 (to S. H.) and by a grant from the Carl Trygger Foundation (to J. R.).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. Tel: 46-31-773-3923; Fax: 46-31-773-2599; E-mail: roslyn.bill@gmm.gu.se.
Special researcher supported by the Swedish Research Council.
Published, JBC Papers in Press, July 9, 2001, DOI 10.1074/jbc.M105045200
2 R. M. Bill and S. Hohmann, unpublished observations.
3 S. Karlgren, R. M. Bill, and S. Hohmann, unpublished observations.
| |
ABBREVIATIONS |
|---|
The abbreviations used are: MIP, major intrinsic protein; contig, group of overlapping clones; PCR, polymerase chain reaction; MES, 4-morpholineethanesulfonic acid.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
| 1. | Borgnia, M., Nielsen, S., Engel, A., and Agre, P. (1999) Annu. Rev. Biochem. 68, 425-458 |
| 2. | Johansson, I., Karlsson, M., Johanson, U., Larsson, C., and Kjellbom, P. (2000) Biochim. Biophys. Acta 1465, 324-342 |
| 3. | Kuwahara, M., Ishibashi, K., Gu, Y., Terada, Y., Kohara, Y., Marumo, F., and Sasaki, S. (1998) Am. J. Physiol. 275, C1459-C1464 |
| 4. | Deen, P. M., and Knoers, N. V. (1998) Curr. Opin. Nephrol. Hypertens. 7, 37-42 |
| 5. | Kaldenhoff, R., Grote, K., Zhu, J. J., and Zimmermann, U. (1998) Plant J. 14, 121-128 |
| 6. | Kjellbom, P., Larsson, C., Johansson, I., Karlsson, M., and Johanson, U. (1999) Trends Plant Sci. 4, 308-314 |
| 7. | Luyten, K., Albertyn, J., Skibbe, W. F., Prior, B. A., Ramos, J., Thevelein, J., and Hohmann, S. (1995) EMBO J. 14, 1360-1371 |
| 8. | Tamás, M. J., Luyten, K., Sutherland, F. C. W., Hernandez, A., Albertyn, J., Valadi, H., Li, H., Prior, B. A., Kilian, S. G., Ramos, J., Gustafsson, L., Thevelein, J. M., and Hohmann, S. (1999) Mol. Microbiol. 31, 1087-1104 |
| 9. | Tamás, M. J., Rep, M., Thevelein, J. M., and Hohmann, S. (2000) FEBS Lett. 472, 159-165 |
| 10. | Tao, W., Deschenes, R. J., and Fassler, J. S. (1999) J. Biol. Chem. 274, 360-367 |
| 11. | Hohmann, S., Bill, R. M., Kayingo, G., and Prior, B. A. (2000) Trends Microbiol 8, 32-36 |
| 12. | Sutherland, F. C., Lages, F., Lucas, C., Luyten, K., Albertyn, J., Hohmann, S., Prior, B. A., and Kilian, S. G. (1997) J. Bacteriol. 179, 7790-7795 |
| 13. | Holst, B., Lunde, C., Lages, F., Oliveira, R., Lucas, C., and Kielland-Brandt, M. C. (2000) Mol. Microbiol. 37, 108-124 |
| 14. | Sweet, G., Gandor, C., Voegle, R., Wittekindt, N., Beuerle, J., Truniger, V., Lin, E. C. C., and Boos, W. (1990) J. Bacteriol. 172, 424-430 |
| 15. | Voegele, R. T., Sweet, G. D., and Boos, W. (1993) J. Bacteriol. 175, 1087-1094 |
| 16. | Sanders, O. I., Rensing, C., Kuroda, M., Mitra, B., and Rosen, B. P. (1997) J. Bacteriol. 179, 3365-3367 |
| 17. | Jung, J. S., Preston, G. M., Smith, B. L., Guggino, W. B., and Agre, P. (1994) J. Biol. Chem. 269, 14648-14654 |
| 18. | Fu, D., Libson, A., Miercke, L. J., Weitzman, C., Nollert, P., Krucinski, J., and Stroud, R. M. (2000) Science 290, 481-486 |
| 19. | Murata, K., Mitsuoka, K., Hirai, T., Walz, T., Agre, P., Heymann, J. B., Engel, A., and Fujiyoshi, Y. (2000) Nature 407, 599-605 |
| 20. | Thomas, B. J., and Rothstein, R. (1989) Cell 56, 619-630 |
| 21. | Berben, G., Dumont, J., Gilliquet, V., Bolle, P. A., and Hilger, F. (1991) Yeast 7, 475-477 |
| 22. | Sherman, F., Fink, G. R., and Hicks, J. B. (1983) Methods in Yeast Genetics , Cold Spring Harbor Laboratory, Cold Spring Harbor, NY |
| 23. | Sarkar, G., and Sommer, S. S. (1990) BioTechniques 8, 404-407 |
| 24. | Ito, H., Funkuda, Y., Murata, K., and Kimura, A. (1983) J. Bacteriol. 153, 163-168 |
| 25. | Muhlrad, D., Hunter, R., and Parker, R. (1992) Yeast 8, 79-82 |
| 26. | Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY |
| 27. | Cain, S. M., Matzke, E. A., and Brooker, R. J. (2000) J. Membr. Biol. 176, 159-168 |
| 28. | Sayle, R. A., and Milner-White, E. J. (1995) Trends Biochem. Sci. 20, 374 |
| 29. | Holm, L., and Sander, C. (1991) J. Mol. Biol. 218, 183-194 |
| 30. | Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, 2714-2723 |
| 31. | Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, 4673-4680 |
| 32. | Chou, P. Y., and Fasman, G. D. (1978) Adv. Enzymol. Relat. Areas Mol. Biol. 47, 45-148 |
| 33. | Deleage, G., and Roux, B. (1987) Protein Eng. 1, 289-294 |
| 34. | Sarakinou, K. S., Antoniw, J. F., Togawa, R. C., and Mullins, J. G. L. (2001) Biochem. Soc. Trans. 29, 36 |
| 35. | Heller, K. B., Lin, E. C., and Wilson, T. H. (1980) J. Bacteriol. 144, 274-278 |
| 36. | Wysocki, R., Chery, C. C., Wawrzycka, D., Van Hulle, M., Cornelis, R., Thevelein, J. M., and Tamás, M. J. (2001) Mol. Microbiol. 40, 1391-1401 |
Copyright © 2001 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:
|
|
 |
|
  Y. Morishita, T. Matsuzaki, M. Hara-chikuma, A. Andoo, M. Shimono, A. Matsuki, K. Kobayashi, M. Ikeda, T. Yamamoto, A. Verkman, et al. Disruption of Aquaporin-11 Produces Polycystic Kidneys following Vacuolization of the Proximal Tubule Mol. Cell. Biol., September 1, 2005; 25(17): 7770 - 7779. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Karlgren, N. Pettersson, B. Nordlander, J. C. Mathai, J. L. Brodsky, M. L. Zeidel, R. M. Bill, and S. Hohmann Conditional Osmotic Stress in Yeast: A SYSTEM TO STUDY TRANSPORT THROUGH AQUAGLYCEROPORINS AND OSMOSTRESS SIGNALING J. Biol. Chem., February 25, 2005; 280(8): 7186 - 7193. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Hedfalk, R. M. Bill, J. G. L. Mullins, S. Karlgren, C. Filipsson, J. Bergstrom, M. J. Tamas, J. Rydstrom, and S. Hohmann A Regulatory Domain in the C-terminal Extension of the Yeast Glycerol Channel Fps1p J. Biol. Chem., April 9, 2004; 279(15): 14954 - 14960. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Tamas, S. Karlgren, R. M. Bill, K. Hedfalk, L. Allegri, M. Ferreira, J. M. Thevelein, J. Rydstrom, J. G. L. Mullins, and S. Hohmann A Short Regulatory Domain Restricts Glycerol Transport through Yeast Fps1p J. Biol. Chem., February 14, 2003; 278(8): 6337 - 6345. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hohmann Osmotic Stress Signaling and Osmoadaptation in Yeasts Microbiol. Mol. Biol. Rev., June 1, 2002; 66(2): 300 - 372. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hansen, J. F. J. Kun, J. E. Schultz, and E. Beitz A Single, Bi-functional Aquaglyceroporin in Blood-stage Plasmodium falciparum Malaria Parasites J. Biol. Chem., February 8, 2002; 277(7): 4874 - 4882. [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 |