(Received for publication, February 7, 1997, and in revised form, March 18, 1997)
From the Department of Biological Sciences, Columbia University, New York, New York 10027
C173 and W125 are pet mutants of Saccharomyces cerevisiae, partially deficient in cytochrome oxidase but with elevated concentrations of cytochrome c. Assays of electron transport chain enzymes indicate that the mutations exert different effects on the terminal respiratory pathway, including an inefficient transfer of electrons between the bc1 and the cytochrome oxidase complexes. A cloned gene capable of restoring respiration in C173/U1 and W125 is identical to reading frame YGR112w of yeast chromosome VII (GenBank Z72897[GenBank]). The encoded protein is homologous to the product of the mammalian SURF-1 gene. In view of the homology, the yeast gene has been designated SHY1 (Surf Homolog of Yeast). An antibody against the carboxyl-terminal half of Shy1p has been used to localize the protein in the inner mitochondrial membrane. Deletion of part of SHY1 produces a phenotype similar to that of G91 mutants. Disruption of SHY1 at a BamHI site, located approximately 2/3 of the way into the gene, has no obvious phenotypic consequence. This evidence, together with the ability of a carboxyl-terminal coding sequence starting from the BamHI site to complement a shy1 mutant, suggests that the Shy1p contains two domains that can be separately expressed to form a functional protein.
Respiratory defective pet mutants1 of Saccharomyces cerevisiae have been useful in understanding some of the processes underlying the biogenesis of mitochondria. In earlier studies, pet mutants have been grouped into different phenotypic classes based on their spectral properties and their respiratory and ATPase activities (1, 2). A substantial number of complementation groups consist of mutants displaying defects in single enzymes of the respiratory chain (e.g. cytochrome oxidase, ubiquinol-cytochrome c reductase) or ATPase. Such strains have been exploited in different laboratories to identify and isolate genes coding for the protein constituents of the respiratory complexes (3, 4) and for ancillary factors acting at different stages of their assembly (5-7). Another commonly encountered phenotype is characterized by the pleiotropic absence of oligomycin-sensitive ATPase and the respiratory chain complexes, whose synthesis depends in part on the expression of the mitochondrial gene products. This phenotype is frequently elicited by mutations in constituents of the mitochondrial translational machinery (8).
In addition, numerous mutants exhibit the presence of all the respiratory chain components but at levels below those found in wild type yeast. This class of mutants has received little attention because the gross biochemical phenotypes do not provide obvious clues about the primary lesions responsible for the respiratory defect. To enlarge on current information about the contribution of the nuclear genome toward the maintenance of respiratory competent mitochondria, we have begun studies of pet mutants with partial pleiotropic phenotypes. In this communication we report on mutants from complementation group G91 of our collection of pet strains (9). Enzyme assays have revealed that the depression of mitochondrial respiration in this group of mutants cannot be explained in any simple way by the activities of different segments of the electron transport chain but rather appears to be related to inefficient transfer of electrons in the span between the bc1 and cytochrome oxidase complexes. The mutations responsible for the respiratory defect have been localized to the yeast homolog of the SURF-1 gene, previously shown to be in a highly conserved gene cluster of several mammalian (10) and chicken genomes (11). We present evidence that the product of the yeast SHY1 gene is a membrane constituent of mitochondria and therefore is likely to function directly in some aspect of mitochondrial organization and function.
The genotypes and sources of the strains of S. cerevisiae used in this study are listed in Table I. The media used to grow yeast have been described elsewhere (8).
|
SHY1 was cloned by
transformation of the pet mutant C173/U1 ( ura3-1
shy1-2) with a yeast genomic plasmid library by the method of
Beggs (13). The library used for the transformation was constructed
from partial Sau3A fragments of nuclear DNA (averaging 7-15
kb) cloned into the BamHI site of the shuttle vector YEp24 (14). This library was kindly provided by Dr. Marian Carlson, Department of Genetics and Development, Columbia University.
Approximately 1 × 108 cells were transformed with 10 µg of plasmid DNA. The transformation mixtures were plated on minimal
glycerol medium to select clones complemented for the uracil auxotrophy
and respiratory defect. The transformation yielded two
uracil-independent and respiratory competent clones. The transformed
phenotypes of both clones were verified by segregation tests to be
plasmid-dependent.
Wild type and mutant yeast were grown to stationary phase in YPGal (2% galactose, 1% yeast extract, and 2% peptone), and mitochondria were prepared by the procedure of Faye et al. (15), except that Glusulase was replaced by Zymolyase 20,000 (ICN Biomedicals, Inc.) during preparation of spheroplasts. ATPase activity was assayed at 37 °C by the colorimetric determination of inorganic phosphate released from ATP (16). NADH oxidase was measured polarographically with a Clark electrode. The reactions were done in 20 mM potassium phosphate, pH 7.5, containing 2 mM NADH. NADH-cytochrome c reductase, succinate-cytochrome c reductase, and cytochrome oxidase were all assayed at 25 °C by previously described procedures (2).
Construction of W303Two
different mutant alleles of SHY1 were introduced into a
respiratory competent strain of yeast. The first was a simple disruption of the SHY1 coding sequence at the internal
BamHI site, with a fragment containing the yeast
HIS3 gene. This allele was made by cloning the 630-base pair
EcoRI-SphI fragment of pG91/ST5 in YEp352 (17).
The resultant plasmid (pG91/ST6) was linearized with BamHI
and ligated to HIS3 on a 1.7-kb BamHI fragment to
yield shy1::HIS3. In the second construct, the
sequence extending from the SphI to the ClaI site
was removed and substituted by a 1-kb fragment with the yeast
URA3 gene. This deletion allele
(shy1::URA3) was made by linearizing pG91/ST12
at the unique ClaI site 215 nucleotides downstream of
SHY1. The ClaI site was blunted with Klenow
polymerase, ligated to an eight nucleotide long SphI linker, and recircularized. After digestion of this plasmid with
SphI, the linear plasmid, deleted for more than two-thirds
of SHY1, was ligated to a 1-kb fragment with the yeast
URA3 gene. Both the shy1::HIS3 and
shy1::URA3 alleles were recovered as linear fragments and used to disrupt the corresponding wild type genes by the
one-step gene replacement method (18) in the respiratory competent
diploid strain W303. Haploid progeny with the mutant alleles were
recovered following sporulation of diploid transformants.
To obtain antibodies
against Shy1p, the sequence coding for the carboxyl-terminal half of
the protein was expressed as a fusion protein in Escherichia
coli. The 1.5-kb BamHI-HindIII fragment containing the region of SHY1 coding for 166 carboxyl-terminal residues of the protein was ligated to pATH22 (19).
The resultant plasmid expressed a 55-kDa hybrid protein consisting of
the amino-terminal half of anthranylate synthetase component I fused
in-frame to the Shy1p sequence. The fusion protein constituted most of
the insoluble protein fraction of the E. coli cells
harboring the pATH22 construct. This fraction was dissolved in a 10 mM Tris-HCl, 1 mM EDTA, pH 7.5 buffer,
containing 2% SDS, 5 mM -mercaptoethanol, and 20 µg/ml phenylmethylsulfonyl fluoride and was further purified on a
Bio-Gel A0.5 column developed with a buffer containing 10 mM Tris-HCl, 0.1 mM EDTA, 5 mM
-mercaptoethanol. Fractions enriched for the fusion protein were
pooled, concentrated by acetone precipitation, and used to raise
antibodies in rabbits.
Standard procedures were used for the preparation and ligation of DNA fragments and for transformation and recovery of plasmid DNA from E. coli (20). The preparation of yeast nuclear DNA and the conditions for the Southern hybridizations were as described by Myers et al. (8). DNA probes were labeled by random priming (21). DNA was sequenced by the method of Maxam and Gilbert (22). Proteins were separated by polyacrylamide gel electrophoresis in the buffer system of Laemmli (23), and Western blots were treated with antibodies against the Shy1p fusion protein followed by a second reaction with 125I-protein A (24). Protein concentrations were determined by the method of Lowry et al. (25).
C173
is one of two independent pet isolates previously assigned
to complementation group G91 of a collection of respiratory deficient
mutants of yeast (9). The respiratory defect of the mutants is
complemented by ° testers (cytoplasmic petite mutants lacking
mitochondrial DNA), indicating that each has a recessive mutation in a
nuclear gene. The G91 mutants are genetically stable and, unlike some
nuclear pet mutants (8), do not acquire deletions in
mitochondrial DNA at any appreciable frequency. In vivo
labeling of the mitochondrial gene products indicated that
mitochondrial protein synthesis is not affected in the mutants (data
not shown).
The cytochrome composition of mutant mitochondria was determined from
the visible spectrum of extracts obtained under conditions known to
quantitatively solubilize all the respiratory components of the
organelle. A comparison of the spectra obtained from various mutants
showed a reduction in cytochromes a and
a3 and an increase in the -absorption bands
at 550 nm corresponding to c-type cytochromes (Fig.
1). To distinguish between cytochromes c and
c1, mitochondria were first treated with high
salt to remove cytochrome c and were then extracted with
deoxycholate to solubilize the remaining cytochromes. Spectra of the
salt and detergent extracts (Fig. 2) show that the
increase in absorption at 550 nm in the shy1 null mutant is due to cytochrome c. Its concentration, based on the
difference spectra, was estimated to be 0.57 nmol/mg mitochondrial
protein in wild type and 0.9 nmol/mg protein in the mutant. The spectra of the cytochromes remaining after the salt wash confirm the lower concentration of a-type cytochromes but do not show any
appreciable differences in cytochromes b and
c1.
The mutants were also assayed for electron transport activities representing different spans of the respiratory chain. These, as well as mitochondrial ATPase activity, were measured in two independent G91 mutants and in two different mutant constructs, only one of which expressed a respiratory defective phenotype (see below). These assays indicated mutant mitochondria to have reduced cytochrome oxidase activity, consistent with their spectral properties (Table II). No significant differences were found in the oligomycin-sensitive ATPase. All the mutants, however, had approximately two times higher NADH-cytochrome c reductase and succinate c reductase activities (Table II).
|
The presence in W125, C173/U1, and W303SHY1 of higher than normal
NADH- and succinate-cytochrome c reductase and of lower but
nonetheless respectable cytochrome oxidase activity was not consistent
with their growth phenotype. All three mutants fail to show any
appreciable growth on glycerol. To further assess their respiratory
potential, mutant and wild type mitochondria were assayed for NADH and
succinate oxidase. As expected, these assays revealed both activities
to be lower in the mutant mitochondria. The decrease, however, is
quantitatively greater than predicted from the individual rates of
cytochrome c reduction and oxidation by the terminal
oxidase. The 70% decrease of NADH oxidase measured in W125
mitochondria is difficult to reconcile with the 2-fold enhancement in
NADH-cytochrome c reductase and a reduction of only 23% in
the cytochrome oxidase rate. A similar trend is evident in the results
obtained with C173/U1 and W303
SHY1/U, although the discrepancies
between the reductions in NADH oxidase and cytochrome oxidase were not
as large.
The different respiratory activities of the mutant mitochondria were
also difficult to relate to their growth properties on non-fermentable
substrates. For example, even though W303SHY1/U and W125 retain 20 and 30% of the wild type NADH oxidase activity, respectively, neither
mutant shows any appreciable growth on glycerol or ethanol.
To clone
the wild type gene mutated in G91 mutants, C173/U1 was transformed with
a yeast genomic library. Two respiratory competent clones obtained from
the transformation were analyzed for their plasmid content. Plasmids
isolated from the transformants were amplified in E. coli
and their nuclear DNA inserts characterized. The restriction maps of
the two plasmids indicated both to have the same inserts. This plasmid
was designated as pG91/T1 (Fig. 3).
The gene responsible for restoring respiration in G91 mutants was identified by transferring different regions of the pG91/T1 insert to YEp352 and transforming C173/U1 with the new plasmids (Fig. 3). These experiments indicated that the 2.3-kb SphI fragment internal to the pG91/T1 insert restored respiration (pG91/ST4), whereas the smaller plasmid pG91/ST3 lacking the 400 base pairs of DNA from the BamHI to the SphI site did not, suggesting the BamHI site to be internal to the gene. The gene was identified by sequencing the span between the two SphI sites. This region has a reading frame matching open reading frame YGR112w on chromosome VII (GenBank accession number Z72897[GenBank]). The calculated molecular weight of the primary product encoded by this reading frame is 45,062. Even though the insert of pG91/ST4 lacks the sequence coding for the amino-terminal 104 residues of the encoded protein, this plasmid restores respiration as efficiently as pG91/T1, containing the entire gene (see also below).
The predicted sequence of the YGR112w product is homologous to the
proteins encoded by the human and mouse SURF-1 genes (10, 11) (Fig. 4). The sequence similarity is highest in the
central portion of the protein. This region of the yeast protein
beginning with residue 81 and ending with residue 191 shares 50 identities (44%) with the mouse homolog. A shorter conserved region
occurs near the carboxyl terminus where there are 12 identities in a span of 23 residues. The yeast and mammalian proteins have two potential membrane spanning domains close to their amino and carboxyl termini. The sequence of the amino-terminal 20-30 residues has a
preponderance of basic residues conforming to the composition generally
found in cleavable mitochondrial import signals (26). Because of its
homology to SURF-1, the yeast gene and proteins will
henceforth be referred to as SHY1 (Surf
Homolog of Yeast) and Shy1p, respectively.
Properties of Shy1 Null Mutants
Restoration of respiratory
activity in G91 mutants by SHY1 could be mediated either by
complementation or extragenic suppression. To distinguish between these
mechanisms, SHY1 was disrupted/deleted in a respiratory
competent strain of yeast. Two different mutant alleles were made. The
first was constructed by insertion of the HIS3 gene at the
BamHI site in the SHY1 reading frame.
Introduction of shy1::HIS3 into the chromosomal
DNA of haploid yeast did not affect the transformant's ability to grow
on glycerol. The second mutant construct was made by replacing the
sequence from the internal SphI site to the ClaI
site in the 3-flanking sequence with the URA3 gene. This
allele lacks the sequence of SHY1 coding for the carboxyl-terminal 285 residues starting from codon 110 (Fig.
5). Substitution of the
shy1::URA3 mutation for the normal chromosomal copy of the gene resulted in a respiratory deficient phenotype.
The shy1::URA3 allele was ascertained to be
linked to the mutations in C173/U1 and W125, indicating that
SHY1 confers respiration by complementation rather than by
extragenic suppression. For these allelism tests, SHY1 was
transferred to the integrative plasmids YIp352 and YIp351 (17) to yield
pG91/ST16 and pG91/ST17, respectively. Following digestion of pG91/ST16
and pG91/ST17 at the unique NruI site of SHY1,
the linear plasmids were used to transform C173/U1 and W125. Most of
the Ura+ clones obtained from the transformation of the
mutants were respiratory competent suggesting that the integration had
occurred at the locus of the mutations. This was verified by crosses of
a respiratory competent clone from each transformation to the wild type
haploid strain W303-1A or W303-1B and to the corresponding mutants with the
shy1::URA3 allele. Diploid cells issued
from each cross were sporulated and asci dissected. The meiotic spore
progeny from the back crosses to the wild type W303 strain were
respiratory competent (total of 18 complete tetrads). Analysis of the
tetrads obtained from the crosses to the shy1 mutants
confirmed a 2:2 segregation of the respiratory competent phenotype
(total of 16 complete tetrads). The respiratory competent progeny were
all Ura+. Transformation of W125 with pG91/ST17 linearized
at the NruI site of SHY1 also yielded respiratory
competent Leu+ clones. One such clone was crossed to
W303-1B and also to W303
SHY1/U. The meiotic segregation results
obtained with the diploid cells were identical to those discussed
above, except that the respiratory competent phenotype of the spores
obtained from the cross to W303
SHY1/U cosegregated with the
LEU2 marker, as expected.
The respiratory defect resulting from
mutations in SHY1 suggested that the encoded protein was
likely to be a mitochondrial constituent. This was confirmed with an
antibody prepared against a fusion protein consisting of the
amino-terminal half of E. coli anthranylate synthetase
component I and the carboxyl-terminal 166 residues of Shy1p. The
antibody detected a protein of approximately 45 kDa in the
mitochondrial but not in the post-mitochondrial supernatant fraction of
wild type yeast (Fig. 6). This size is consistent with
the molecular weight of Shy1p predicted by the gene sequence. The
identity of this protein as the SHY1 product is supported by
its higher abundance in mitochondria of a transformant harboring the
gene on a multicopy plasmid (Fig. 6) and its absence in a strain with a
disrupted copy of the SHY1 (not shown).
The solubility of Shy1p was determined by fractionation of sonically
disrupted mitochondria into soluble matrix proteins and membrane
vesicles after disruption of wild type mitochondria by sonic
irradiation. Western blot analysis of the two fractions indicate that
Shy1p is enriched in the membrane fraction. Extraction of Shy1p is
effected by concentrations of deoxycholate that are generally used to
solubilize hydrophobic proteins, providing additional support for its
membrane association (Fig. 6). To localize Shy1p more precisely, the
outer and inner mitochondrial membranes were separated by isopycnic
centrifugation. The gradient fractions were probed with an antibody
against subunit 5 of cytochrome oxidase (an inner membrane marker) as
well as antibodies against total outer membrane proteins (the serum
used detects chiefly porin), and with the antibody against Shy1p. The
procedure used to separate the two membranes yields outer membrane
relatively free of inner membrane; fractions enriched in inner
membrane, however, contain significant amounts of the outer membrane.
The results of the Western analysis indicate that Shy1p fractionates
with the inner membrane. This is evidenced by the cosedimentation of
Shy1p with the cytochrome oxidase subunit 5 marker and its absence in
the less dense fractions of the gradient that are enriched in the outer
membrane marker (Fig. 7). Localization of Shy1p in the
inner membrane was confirmed by the difference in its susceptibility to
proteinase K in mitoplasts and mitochondria (Fig. 8).
The results of this experiment indicate that Shy1p is protected against
proteolysis in intact mitochondria. Disruption of the outer membrane by
exposure of mitochondria to hypotonic conditions causes the protein to become sensitive to proteinase K. Subunit 5 of cytochrome oxidase, being almost entirely embedded in the inner membrane (29), is resistant
to proteolysis in both mitochondria and mitoplasts. Sco1p, an inner
membrane protein facing the intermembrane space (30), behaves
analogously to Shy1p.
Shy1p does not appear to be a component of a larger complex. Sedimentation in sucrose gradients calibrated with size standards ranging from 60 to 500 kDa indicates Shy1p to have a size consistent with the monomer molecular weight (data not shown).
Conferral of Respiration by Partially Deleted Copies of SHY1The complete SHY1 reading frame and the regions
starting from the SphI or BamHI sites of the gene
were inserted in single copy either at the URA3 locus of
C173/U3 and at the URA3 or LEU2 loci of W125. The
same constructs were also integrated at the LEU2 locus of
W303SHY1/U, a respiratory defective strain with a partial deletion
of SHY1. Transformants harboring the different constructs
were tested for growth on the non-fermentable substrate glycerol as the
carbon source. These experiments indicate that the entire gene is
required to complement the respiratory defect of W303
SHY1/U. This,
however, is not true of the point mutants. A construct containing the
sequence coding for the carboxyl-terminal 286 codons (pG91/ST21)
restored respiration in both C173/U1 and in W125 (Fig.
9). Truncation of an additional 120 codons, leaving only
the sequence coding for the carboxyl-terminal 166 residues (pG91/ST18),
complements the respiration defect of W125 but not C173/U1. The same
results were obtained when the different constructs were introduced
into the mutants on the multicopy plasmids, pG91/ST3 and pG91/ST4 (see
also Fig. 3).
The differential effect of the partially deleted SHY1 genes on the growth phenotypes of C173/U1 and W125 could indicate that the two strains have mutations in different genes. This was excluded by the tight linkage of the two mutations. Diploid cells issued from a cross of C173/U1 to the respiratory competent transformant W125/T1 were selected on minimal glucose and were sporulated. Meiotic progeny from 13 complete tetrads were cured of the pG91/T1 plasmid on glucose medium containing 5-fluoroorotic acid. All the spores from the 13 tetrads tested became respiratory defective after loss of the plasmid.
The ability of pG91/ST3 and pG91/ST4 (and their integrative
counterparts) to complement W125 implies that the partial
SHY1 sequences present in these plasmids express proteins
which, in conjunction with the mutant Shy1p, are capable of performing
the function of the normal protein. The requirement for the mutant protein produced by W125 is evident from the failure of either plasmid
to complement W303SHY1/U. The presence of truncated forms of Shy1p
in W125/ST3 and W125/ST4 was confirmed by Western analysis of the
proteins in mitochondria and the post-mitochondrial supernatant fractions of the two transformants. The antibody to the fusion protein
detected novel lower molecular weight proteins in the mitochondrial but
not post-mitochondrial supernatant fractions of the transformants (Fig.
10). The 30- and 21-kDa proteins observed in W125/ST4
and W125/ST3, respectively, approximate the sizes expected of the
truncated products based on the sequences retained in the two plasmids.
The first in-frame methionine codon in pG91/ST4 corresponds to residue
124 of the protein. The predicted molecular weight of the product
translated from this codon is 31,600. In pG91/ST3, the first methionine
and 10 additional codons are contributed by the sequence upstream of
the BamHI site in the multiple cloning region of YEp352. The
molecular weight of the product translated from this frame is 20,350, a
value in close agreement with the size estimated by SDS.
Since both truncations retain only one of the two putative transmembrane domains, it was of interest to determine whether the shorter polypeptides behave as intrinsic membrane proteins. This was tested by sonic disruption of mitochondria and sequential extraction of the resultant submitochondrial membranes with salt and with detergent. The results of this experiment show that the 31- and 20-kDa peptides have properties similar to that of Shy1p (Fig. 10). Shy1p and the two truncated proteins are absent in the soluble fractions released by sonic treatment of mitochondria and by extraction of the membranes with high salt. Like native Shy1p, solubilization of the shorter proteins requires disruption of the phospholipid bilayer with detergent.
It is of interest that the mitochondria of the mutant W125 contain a protein of approximately 30 kDa (marked by the asterisk in Fig. 10) which is not present in mitochondria of the parental W303-1A strain. The solubility properties of this protein are similar to Shy1p and the truncated derivatives. This novel protein could be a prematurely terminated translation product of the mutant gene. The absence of the protein in W125 may indicate stabilization of the protein by the polypeptides expressed from pG91/ST3 and pG91/ST4.
SURF-1 has been identified in mammalian (10), chicken (11), Drosophila melanogaster (31), and S. cerevisiae (GenBank, Z7297) genomes. Human, mouse, and chicken SURF-1 are part of a conserved locus that includes several other genes (10, 11). The absence of this organizational feature in the fly and yeast genomes casts some doubt on its functional significance (27). Even though the SURF-1 gene appears to be widely, if not universally, distributed in eucaryotic organisms, nothing is known about its function. In our studies, the yeast SURF-1 homolog was identified because of its ability to restore respiration in mutants previously placed in complementation group G91 of a pet mutant collection (9). Several lines of evidence indicate that the yeast homolog of SURF-1 (SHY1) confers respiratory competence to G91 mutants by complementation rather than suppression. The biochemical properties of a shy1 null mutant construct mimic those of point mutants. The mutant alleles in two independently isolated members of complementation group G91 were found to be genetically linked to the shy1 null mutation.
The partial pleiotropic phenotype of shy1 mutants is representative of a substantial number of pet complementation groups. Mutants of this phenotypic class have not been studied in any systematic way for the principal reason that their biochemical properties are not conducive to any easy interpretation of the primary lesion. To learn more about the basis of the respiratory defect in shy1 mutants, we have compared the partial activities associated with different spans of the electron transport chain to the net flux through this pathway with either NADH or succinate as electron donors. The enzyme assays revealed consistently higher succinate- and NADH-cytochrome c reductase and lower cytochrome oxidase activity in the mutant mitochondria. The 2- to 3-fold increase in reductase activity may be related to the higher content of cytochrome c in the mutant mitochondria. The presence of more endogenous cytochrome c in the membrane probably allows for a more rapid equilibration/reduction of the free cytochrome c used as the electron acceptor in the assay. The reduction in cytochrome oxidase correlates with a lower mitochondrial concentration of cytochromes a/a3, although the reason for this is not clear. Significantly, the loss of both succinate and NADH oxidase is greater than can be accounted for by the reduction in cytochrome oxidase. This discrepancy suggests that overall electron transport in shy1 mutants is also impaired, as a result of less efficient transfer of electrons from the bc1 complex to cytochrome oxidase. The interaction of these complexes could be affected by their organization in the membrane or by the efficiency with which cytochrome c mediates electron transfer between the two complexes. The absence of detectable growth on non-fermentable substrates, even when the oxidation of succinate or NADH is 20-30% of wild type, points to still other effects of the mutations on mitochondrial function.
Shy1p has been localized to mitochondria and is stably associated with the inner mitochondrial membrane. Release of Shy1p from the inner membrane requires detergent in the concentration range normally used to solubilize intrinsic membrane proteins. The primary sequence suggests the presence of two hydrophobic sequences of sufficient length to act as membrane anchoring domains. They are located near the amino and carboxyl termini of the protein. The membrane anchoring property of the carboxyl-terminal sequence is supported by the membrane association of the partial proteins expressed from pG91/ST3 and ST4, both of which lack the amino-terminal region of Shy1p.
Notwithstanding its relatively small size, Shy1p appears to be composed
of at least two separate functional domains which, when expressed
separately, are capable of providing the function of Shy1p. The two
domains can be dissected at the BamHI site as evidenced by
the absence of a phenotype in W303SHY1/H, a strain in which
SHY1 was disrupted at the BamHI site. The ability
of the sequence 3
of the BamHI site to act as a separate
functional domain is also supported by complementation of W125 by
pG91/ST3 and pG91/ST18 whose inserts start at the BamHI site
and detection of the corresponding truncated protein in the
transformant W125/ST3. Restoration of respiratory growth in C173/U1 is
observed with pG91/ST4 and pG91/ST19 both of whose inserts start from
the SphI site. This mutant is not complemented by pG91/ST3
or pG91/ST18 suggesting that the mutation is located between the
BamHI and SphI sites. The mutation was also
localized to this region by cloning the gene from C173/U1 and
swapping the SphI-BamHI fragment between the wild
type and mutant genes. Only plasmids with the wild type
SphI-BamHI fragment complemented the mutant. The
failure of the polypeptides expressed from the truncated
genes to complement W303
SHY1/U containing only the
amino-terminal coding region up to the SphI site suggests
that the amino-terminal domain must include at least part of the
sequence encoded by the region between the SphI and
BamHI sites.
The restoration of respiratory function in C173/U1 and W125 by the amino-terminal truncated polypeptides suggests that their orientation in the membrane is the same as the native protein. In the absence of information about the transport route for Shy1p, how this is accomplished can only be speculated. In the most general sense, it is possible that by virtue of their interaction (perhaps through contact points in the central hydrophilic region), the two separate domains insert into the membrane in a manner analogous to that of the native protein. The simplest way to visualize this is if membrane insertion of the carboxyl-terminal transmembrane segment occurs from the intermembrane space. The sensitivity of Shy1p in mitoplasts to proteinase K digestion suggests that the hydrophilic central region faces the intermembrane space, an orientation consistent with insertion from that compartment. Even though the partial proteins with the carboxyl-terminal domain (whether expressed from the two partially deleted genes or from the gene disrupted at the BamHI site) lack the putative amino-terminal import signal they may fortuitously contain sequences capable of targeting them to mitochondria. This would not be unusual in view of the rather loose sequence requirements of targeting signals (32).
We thank Dr. Liza Pon, Department of Anatomy and Cell Biology, Columbia University, for her generous gift of the outer membrane antibody.