(Received for publication, August 27, 1996, and in revised form, October 14, 1996)
From the Division of Protein Engineering and the
§ Division of Applied Microbiology, Korea Research Institute
of Bioscience and Biotechnology, Korea Institute of Science and
Technology, P. O. Box 115, Yusong, Taejon 305-600, Korea
Z (Glu342 Lys) and
Siiyama (Ser53
Phe) genetic variations of
human
1-antitrypsin (
1-AT) cause a
secretion blockage in the hepatocytes, leading to
1-AT
deficiency in the plasma. Using in vitro folding analysis,
we have shown previously that these mutations interfere with the proper
folding of polypeptides. To understand the fundamental cause for the
secretion defect of the Z and Siiyama variants of
1-AT, we investigated in vivo folding and
stability of these variant
1-AT using the secretion
system of yeast Saccharomyces cerevisiae. Various
thermostable mutations suppressing the folding block of the Z variant
in vitro corrected the secretion defect as well as the
intracellular degradation in the yeast secretion system. Significantly,
the extent of suppression in the secretion defect of Z protein was
proportional to the extent of suppression in the folding defect,
assuring that the in vivo defect associated with the Z
variant is primarily derived from the folding block. In contrast, the
folding and secretion efficiency of Siiyama was not much
improved by the same mutations. In addition, none of the rarely
secreted Siiyama
1-AT carrying the
stabilizing mutations for the wild type and Z variant were active. It
appears that the major defect in Siiyama variant is the
loss of stability in contrast to the kinetic block of folding in the Z
variant.
Human 1-antitrypsin
(
1-AT)1 is a member of the
serine protease inhibitor family, which has a tertiary structure
composed of three
-sheets and several
-helices (1).
1-AT is synthesized as a glycoprotein of 394 amino acids
in the liver and is secreted into the blood. Its major physiological
function is to protect the lungs from excessive elastase activity, and
thus its deficiency causes the lung disease emphysema (2, 3). Some
genetic variations of
1-AT, such as Z
(Glu342
Lys), Siiyama (Ser53
Phe), and Mmalton (Phe52
deleted),
cause
1-AT deficiency in plasma by blocking the secretion of
1-AT (4). The secretion blockage of these
variant
1-AT appears to be due to the aggregation of the
proteins (5, 6, 7). Most of the newly synthesized variant
1-AT accumulate as aggregates and are retained in the
endoplasmic reticulum (ER) of the hepatocytes, which are eventually
degraded. Only a small fraction of the protein that escapes degradation
is deposited as insoluble aggregates within the hepatic ER (8),
resulting in liver diseases.
The structural basis of aggregation of the variant 1-AT
has been suggested to be the loop-sheet polymerization in which a partial opening of the central
-sheet, the A
-sheet, accompanies the insertion of the reactive center loop of another molecule (4). The
loop-sheet polymerization appears to be an important contributing
factor for the secretion blockage of the variant
1-AT.
This implication is based on the finding that the secretion blockage of
the Z or Siiyama was efficiently suppressed in the Xenopus oocyte system either by a thermostable mutation
(Phe51
Leu: F51L) that enhanced the closure of A
-sheet or by mutations that decreased the loop mobility (9). Various
other biochemical and pathological data for the above mutations and
other serpin mutations (10) strongly support that the loop-sheet
polymerization is a structural basis of in vivo aggregation.
However, the mechanism by which each individual mutant
1-AT tends to polymerize and thus cause secretion
blockage remains unclear. One possible mechanism for the polymerization
of the mutant
1-AT is that the kinetic trap during
folding results in the accumulation of intermediates with a high
tendency to polymerize. Another possible mechanism is that the
stability of the already folded structure is lost, resulting in a
conformational change needed for polymerization.
To test these possibilities here, we systematically investigated the
folding, stability, and the secretion of the Z and Siiyama variants of 1-AT, using in vitro folding
assay and yeast secretion system. In order to assess the degree of
individual defects, we examined their suppression by various
thermostable mutations of
1-AT which were identified
recently (11). The results in the present study demonstrate that there
is a direct correlation between the suppression of folding block and
the suppression of secretion block of the Z variant. In contrast, the
defect of Siiyama is much more difficult to rescue than the
Z type defect, which is presumably due to another kind of defect,
i.e. stability loss.
The recombinant yeast strain
expressing the wild type human 1-AT protein has been
described previously (12). Briefly, the recombinant Saccharomyces
cerevisiae Y2805 (MAT
pep::HIS3 prb1-
1.6R can1 his3-200 ura3-52) harbors the
1-AT
expression vector, pYInu-AT, that contains the cDNA of human
1-AT fused in frame with the inulinase signal sequence
downstream of the GAL10 promoter. The yeast expression
vectors for variant
1-AT proteins were constructed by
exchanging the 1.1-kilobase pair BclI/SalI
fragment containing the coding sequence of
1-AT
(starting at 17th codon of mature
1-AT) in pYInu-AT with
the corresponding fragment derived from the pF(BGL)AT plasmids that
were used in the in vitro transcription/translation.
To express the mutant forms of
1-AT protein in vitro, the cDNA fragment
coding for the wild type
1-AT in the plasmid pF(BGL)AT (13) was replaced with the corresponding cDNA fragment encoding the
thermostable mutants derived from the mutagenized plasmid pFEAT30 (11).
The Z and the Siiyama mutation were introduced into the
cDNA encoding the thermostable mutations by
oligonucleotide-directed mutagenesis (14). In vitro
transcription and translation of the cloned gene were carried out and
the translation products were analyzed by transverse urea gradient
(TUG) gel as previously described (13).
The
recombinant yeast cells were grown in methionine-labeling medium
containing 2% galactose, 0.67% yeast nitrogen base without amino
acids, adenine, and amino acids except methionine until the culture
reached the mid-exponential growth phase. After harvesting, the cells
were resuspended and incubated in the same medium containing [35S]methionine (Amersham Corp.). For pulse-chase
analysis, the cells were pelleted and resuspended in the labeling
medium containing 0.2 mg/ml nonradioactive methionine. The cultures
were removed at various times and fractionated into cell pellet and
culture supernatant. The cell pellets were resuspended in buffer A (1% Triton X-100, 150 mM NaCl, 5 mM EDTA, 50 mM Na-HEPES, pH 7.5, 0.5 mg/ml bovine serum albumin, 1 mM phenylmethysulfonyl fluoride, 10 µg/ml each of
aprotinin, leupeptin, and pepstatin, 1 mM dithiothreitol) and disrupted by vortexing with 0.1 ml of glass beads. The obtained cell lysates (intracellular fraction) and the culture supernatants (extracellular fraction) were analyzed by immunoprecipitation with the
polyclonal antibody against human 1-AT
(Sigma) according to the procedure described
previously (15). The immunoprecipitates were analyzed by 10%
SDS-polyacrylamide gel electrophoresis (PAGE). Gels were fixed,
amplified, and fluorographed with EN3HANCE (DuPont). For
endoglycosidase H (endo H) treatment, immunoprecipitates were suspended
in 54 µl of 1% SDS, 10% glycerol, 1% 2-mercaptoethanol and
incubated for 10 min at 95 °C. After cooling, 6 µl of 0.5 M sodium citrate, pH 5.5, and 100 unit endo H were added.
The samples were incubated overnight at 37 °C and analyzed by 10% SDS-PAGE.
Total yeast cell lysates
(intracellular fraction) were obtained by vortexing cell pellets
resuspended in the Laemmli sample buffer (2% SDS, 10% glycerol,
2% 2-mercaptoethanol, 62.5 mM Tris-HCl, pH 6.8, 0.01%
bromphenol blue) together with glass beads. The culture supernatant
(extracellular fraction) was concentrated 20-fold by ultrafiltration
(Amicon, YM30 membrane). Western blot analysis of the samples was
carried with the polyclonal antibody raised against human
1-AT and peroxidase-conjugated anti-rabbit IgG goat
antibody (Sigma). The color development of peroxidase reaction was carried out using 3,3
,5,5
-tetramethylbenzidene stabilized substrate (Promega).
The biological
activity of 1-AT in the culture supernatant was measured
for elastase inhibitory activity according to the method described
previously (12).
To investigate the effects of various
mutations of 1-AT on the folding, the conformational
states of the
1-AT variants translated in
vitro were analyzed by transverse urea gradient gel
electrophoresis (Fig. 1). As observed previously (13),
the folding of the nascent Z polypeptides was blocked at an
intermediate state, but the presence of the F51L mutation (T1)
converted some (40-50%) of the Z-type polypeptides into the native
state. When the Z mutation was combined with the double mutation
(F51L/M374I: T2), or the triple mutation (T59A/T68A/A70G: T3), that
increased the stability of
1-AT more than F51L (T1)
(10), a greater fraction (70-90% versus 40-50%) of the
polypeptides folded into the native form (Fig. 1, middle column). The results suggest that the extent of suppression in the
folding defect is related to the extent of stability increase of the
combining mutations. The effects of the thermostable mutations on the
conformational stability of the native Z protein do not appear to be
additive in all cases, as revealed by the transition midpoint in the
TUG gel (Fig. 1). The stability increase conferred by T2 on urea
denaturation was similar to that of T3 (Fig. 1, left column)
with the wild type background. However, the stability of the native
Z/T2 was much lower than that of Z/T3 (middle column), and
the fraction of the Z polypeptides restored to the native state by T2
was less than that by T3 mutation (70 versus 90%). The
stability of the native Z/T3 was almost the same as that of T3, whereas
the stability of the native Z/T1 and Z/T2 was much lower than that of
T1 and T2, respectively.
The in vitro translation products of 1-AT
carrying Siiyama also could not fold to the native state
(Fig. 1, right column). In contrast to the Z type folding
block, which was suppressed by all three thermostable mutations
(T1-T3), the defect of Siiyama was suppressed only by T3.
About 20% of the Siiyama polypeptides were rescued to the
native state by T3. The stability of the native Siiyama/T3
was much lower than that of T3, and even lower than that of the wild
type.
As a
model system to assess the effects of various mutations on the in
vivo folding and secretion of 1-AT protein, a yeast expression system (12) with high-level secretion of biologically active
human
1-AT into extracellular culture medium was
utilized. To examine the kinetics of secretion of
1-AT
in the recombinant yeast, the cells expressing the wild type human
1-AT were analyzed by pulse-chase and
immunoprecipitation experiments. The immunoprecipitation of the
obtained culture supernatants (Fig. 2A, EC)
showed that significant amounts of the core-glycosylated
1-AT were already secreted outside the yeast cells
within 5-min chase following 2-min pulse labeling with
[35S]methionine. The extensively hyperglycosylated forms
(above 54 kDa) began to show up later, which reflects a longer transit
time for such forms in the secretion pathway due to the additional modification of
1-AT protein. The level of secreted
1-AT increased up to 20 min of chase but after that the
level decreased, probably due to the degradation by the action of yeast
extracellular proteases. Analysis of the soluble intracellular fraction
(Fig. 2A, IC) showed that a major polypeptide of ~46 kDa
remained stable throughout the chase period. The intracellularly
retained
1-AT of ~46 kDa appears to be the
unglycosylated precursor form, since the endo H treatment did not
affect the size at all, while the treatment converted all the
heterogeneously glycosylated forms in the culture supernatant into a
single form of ~46 kDa (Fig. 2B). Most of the unglycosylated form of
1-AT was found in the cytosolic
fraction of yeast and the size of the unglycosylated
1-AT in yeast was slightly larger than that of
1-AT expressed in Escherichia coli (data not
shown). The results imply that some portions of the wild type
1-AT synthesized in yeast were not translocated into the
ER and remained stable in cytoplasm, which is likely due to high-level
expression of the human
1-AT protein in a heterologous host system. The results in Fig. 2, however, show that major portions of the human
1-AT entered into the yeast ER and
transited through the yeast secretion system very rapidly, with a
similar transit time (less than 7 min) as those reported in other yeast
secretory proteins (16).
Effect of Various Thermostable Mutations of
If the folding
defect is the direct cause of the secretion block associated with the Z
variant, the suppression of the folding defect by the thermostable
mutations (Fig. 1) should suppress the secretion blockage of Z variant
observed in vivo. To compare the secretion rate of the
variant forms of 1-AT, the yeast transformants expressing
1-AT were pulse-labeled for 10 min with
[35S]methionine and analyzed by immunoprecipitation.
There was no detectable Z-type
1-AT protein in the
culture supernatant after the 10-min pulse (Fig.
3A, EC, lane Z). In combination with the thermostable mutations, however, some Z proteins were secreted into the
culture media (Fig. 3A, EC). Ten minutes was chosen for the
labeling time to avoid the degradation effect as much as possible. During the pulse, the cells expressing the wild type or the
thermostable mutant proteins alone secreted similar amounts of
1-AT proteins (data not shown). On the combination of Z
mutation, the T1 substitution, which restored only a small part of the
population of the Z polypeptide into the native state in the TUG gel
(Fig. 1), released only a small fraction (10%) of the Z proteins from
the secretion blockage (Fig. 3). With the T3, in which most of the Z
polypeptides could fold into the native form, the secretion yield was
enhanced up to 75% of the wild type protein. The T2 mutation improved
the secretion of Z protein to an the intermediate extent (30%) as observed in the folding analysis. The results clearly show that the
extent of suppression in the secretion was proportional to the extent
of suppression in the folding of Z protein. The suppression of the
secretion defect was also confirmed by Western blot analysis of the
total cell lysates and the culture supernatant (Fig. 3B). There was no significant difference in the relative amounts of the
secreted Z proteins, as measured either by pulse labeling (Fig.
2B, EC) or by Western blot (Fig. 3B, EC). The
results indicate that the secreted Z proteins retain a substantial
stability against degradation comparable with the wild type protein
once they are secreted into the culture medium.
The secretion of Siiyama variant was also blocked in the
yeast secretion system (Fig. 4A, EC), and the
secretion blockage was able to be suppressed slightly by all the three
mutations. Compared with the suppression of the Z-type defect, however,
the overall efficiency of the suppression of the Siiyama
type secretion was much lower. In Western blot (Fig. 4B,
EC), the steady-state level of the secreted Siiyama
carrying the various mutations was 5~20%, while that of Z type was
10~75% (Fig. 3B, EC). The T2 and T3 were more efficient
than T1 in suppressing the secretion defect of Siiyama, but
there was no difference between the suppression by T2 and that by
T3.
To investigate that the secreted protein was correctly folded into the
native form, the inhibitory activity of 1-AT in the culture supernatant was measured by examining the ability to inhibit the activity of elastase (Table I). There is a strong
correlation between the elastase inhibition activity in the culture
supernatant and the amount of
1-AT protein detected in
Western blot of the Z proteins, supporting that the secreted Z proteins
achieve the fully folded native form. The results clearly demonstrate
that the increased secretion efficiency of the Z protein is directly correlated with the improved folding efficiency. In the case of Siiyama protein, however, the secreted proteins did not
show any detectable inhibition activity, implying that the
Siiyama proteins, even if successfully secreted outside
cells, were not in the active native conformation.
|
In general the
transport-impaired polypeptides inside cells are susceptible to
degradation in the secretion pathway (17). We investigated the fate of
the Z and Siiyama 1-AT that accumulated intracellularly in the yeast cells. The kinetics of the ER retention and intracellular degradation of various
1-AT proteins
were examined by the pulse-chase analyses. Fig. 5
displays two populations of the newly synthesized
1-AT
polypeptides with different destiny in the yeast cells: one is the
nontranslocated cytoplasmic
1-AT (a) and the
other is the ER-translocated
1-AT (b). In the
case of the normal human
1-AT (W), most of the
core-glycosylated form of
1-AT (b), which
represents the population entering the secretion pathway, disappeared
from the intracellular fraction within 30-min chase because of the
secretion into extracellular medium. The unglycosylated form,
representing the population that does not enter the secretion pathway,
remained quite stable throughout the chase period. In contrast, the
core glycosylated form of Z
1-AT protein remained
relatively longer in the intracellular fraction than the wild type form
(Fig. 5A), reflecting the longer retention of the Z protein
in the yeast ER as observed in the hepatic ER. The level of the
retained Z
1-AT decreased eventually during the chase.
Since there was almost no detectable Z protien in the extracellular
media (Fig. 3), the disappearance of core glycosylayed Z protein in the
intracellular fraction is caused mainly by degradation rather than by
secretion. The unglycosylated Z form, which presumably resided in
cytoplasm, disappeared more rapidly.
On combination of the Z mutation with the thermostable mutations T1 and
T2 (Fig. 5A), the amount of core-glycosylated form of
1-AT (b) did not decrease as rapidly as that
of the Z type during the chase, reflecting less degradation. The
transit of
1-AT from the ER to the Golgi appears to be
still retarded for the Z/T1, because the intracellular transit of the
glycosylated forms was longer than that of the wild type. The secretion
process appeared to be facilitated by T2 and T3, as the lesser amount of glycosylated forms was detected at 60 min of chase. Interestingly, the amount of core-glycosylated Z/T3
1-AT appears to be
smaller than that of the wild type, implying that Z/T3 appears to
transit the ER more rapidly than the wild type. The unglycosylated
forms (a) of Z
1-AT carrying the thermostable
mutations were also protected from degradation (Fig. 5A) in
proportion to the degree of the suppression in the folding defect by
the stable mutations.
The apparent ER retention of Siiyama 1-AT
was not as obvious during the chase (Fig. 5B), implying
rapid degradation. The ER form of the variant was detected with
Siiyama/T1 and Siiyama/T2 during the chase,
probably due to less degradation compared with the Siiyama
alone. Again the core-glycosylated Siiyama/T3, like the
Z/T3, was not detected. The nonglycosylated cytoplasmic form of
Siiyama was also degraded, though not so much as that of
the Z type. The cytoplasmic form of Siiyama carrying the
thermostable mutations became more resistant to degradation (Fig.
5B).
In eukaryotes, secretory proteins that cannot fold correctly are
generally retained within the ER, and the retained proteins are
ultimately degraded or accumulate as insoluble aggregates (17, 18, 19).
There are two fundamentally distinct reasons why the native
conformation of a mutant protein is not produced: one is a kinetic
block during folding and the other is a drastic loss of stability. Our
present study suggests different causes for the defect of Z and
Siiyama variant 1-AT.
We have shown recently that the Z mutation causes a
kinetic defect that leads to accumulation of a folding intermediate
with a high tendency to aggregate (20). If the folding block of the Z
variant 1-AT is the primary cause of secretion block, it
is to be expected that any mutations suppressing the kinetic
retardation of the Z-type folding will relieve the secretion blockage.
Various thermostable mutations suppressing the folding defect of the Z variant in vitro (Fig. 1) also suppressed the secretion
defect (Fig. 3) as well as the intracellular degradation (Fig. 5) in the yeast secretion system. Significantly, our findings show that the
extent of suppression in the secretion and degradation was proportional
to the extent of suppression in the folding of Z protein. The results
strongly support the notion that the increase in secretion efficiency
of the Z variant
1-AT is directly caused by the improved
folding of the protein, which was conferred by the combined stable
mutations.
The defective step during the folding of Z protein is likely to be the
last stage of folding from a compact intermediate to the native form
(20). How is the kinetic block of the Z type overcome by these stable
equilibrium mutants? We have previously proposed a kinetic partitioning
between the productive folding pathway and a kinetic trap (13). While
the Z mutation induces partitioning into the folding trap, a more
favorable partitioning into the productive pathway would be induced by
the thermostable mutation. These mutations were shown to enhance the
closure of A -sheet, as revealed by the retardation of the mutant
proteins in producing a complex with the peptide mimicking the sequence of reactive center loop (13), and by the delay in converting the native
state into the latent state (11). The results suggest that the
accumulated folding intermediate in the Z mutation is in a state in
which A
-sheet is not completely closed and is likely to be a
precursor of the loop-sheet polymers. Interestingly, the suppression of
the folding and secretion defect of the Z type by T3 was more efficient
than by T2, although T2 and T3 have the same conformational stability
of
G (~3 kcal/mol) (11). It was noticed that urea
dependence of the unfolding rate was much smaller for T3 than for T2
(11), indicating that the unfolding rate of T3 at physiological
condition would be much greater that that of T2. Further biochemical
characterization is required to provide a precise mechanism for the
enhancement of the Z-type secretion by these stable mutation.
The conformational stability of the Z protein in urea, once folded
successfully, appeared to be affected only slightly (20). A reduced
association rate constant of the plasma Z 1-AT with neutrophil elastase was attributed to the altered local conformation of
the reactive center loop rather than the change in stability (21). The
results in Table I also showed that most of the secreted Z proteins,
but not the Siiyama proteins, retain the inhibitory activity, indicating that the primary cause of the Z type retention in
the ER is the folding retardation and not the stability defect.
The
native conformation of Siiyama protein was not observed in
the TUG gel system after 1 h in vitro translation (Fig.
1). Unlike the Z variant, however, the folded form of
Siiyama type was never produced even after a longer period
of folding (data not shown), which suggests that the defect of
Siiyama might be the stability defect. Indeed, the
stability loss of Siiyama 1-AT was so great
that its defect could be suppressed only by T3 (Fig. 1). Even in this
case the stability of the native Siiyama/T3 form was much
lower than that of the wild type, as revealed by the transition
midpoint of the mutant protein shown in the TUG gels. This is quite a
contrast to the result that the stability of the Z/T3
1-AT is as great as that of the T3
1-AT
(Fig. 1, Siiyama/T3 versus Z/T3). The Z mutation
(E342K), being located at the protein surface, does not appear to
induce a substantial loss of stability. In contrast, the conformational
stability of Siiyama (S53F), which is located at the
hydrophobic core, appears to be drastically affected.
The secretion defect of Siiyama could be rescued, though in
a very low yield, by T1 and T2 mutations (Fig. 4), but the folding defect of Siiyama could not be rescued by these mutations
in the in vitro folding (Fig. 1). There is an obvious
distinction between in vivo folding and in vitro
folding. The in vitro folding of the nascent polypeptides is
an equilibrium process between the native and the non-native species,
but the in vivo folding is a vectorial process in that as
soon as the secretion competent form is produced, it is pulled out of
the folding-unfolding equilibrium into the secretory pathway. Thus, if
the major cause for the defect of Siiyama is the loss of
stability and not the kinetic problem, the secretion of the mutant
would be rescued as far as the stability loss is compensated to some
degree, though it is not enough to show the native band on the TUG gel
system. Such compensation may be enough for maintaining the
secretion-competent state. The enhancement in the secretion of
Siiyama by T3 was not greater than that by T2, although the
enhancement of the Z secretion by T3 was much greater than that by T2
(Table I). The results indicate that the folding defect of
Siiyama is intrinsically different from the Z-type folding
defect. The cause of the folding defect of Siiyama does not
appear to be of kinetic origin because incubation of the
Siiyama/T3 at 30 °C prior to electrophoresis to allow
further folding did not increase the production of the native form
(data not shown), whereas the incubation of the Z type or the Z/T1
improved folding efficiency significantly (20). Further evidence for supporting the stability loss of Siiyama came from the
results that none of the successfully secreted Siiyama
1-AT carrying the stable mutations were active (Table
I). It appears that the stability of these secreted
1-AT
was not great enough to confer inhibitory activity.
Our results on Siiyama differ from those of Foreman and
collaborators (9), in that the secretion defect of Siiyama
was suppressed by F51L in Xenopus oocyte more efficiently
than that of the Z type. One possible explanation for this apparent
inconsistency is that the defect of Siiyama is more
sensitive to certain factors like temperature and local protein
concentration. We have observed that at a high temperature such as
37 °C the suppression on the secretion defect of Siiyama
by the thermostable mutation did not occur at all in the yeast system.
This may also be related to the fact that the aggregation of
Siiyama 1-AT is more severe than the Z type
in the plasma (6). It is possible that since the culture temperature of
Xenopus is much lower (~20 °C), the mutational effect
of destabilization might have been reduced. On the other hand, the
kinetic defect, if it existed, would have been magnified at a low
temperature. Our results in the present study and the previous data
from the Xenopus oocyte system (9) support strongly that the
stability loss as opposed to the kinetic defect is the major cause of
the folding and secretion defect of Siiyama.
The ER
retention and subsequent degradation of the transport-impaired
1-AT variants have been studied in various cell systems (22, 23, 24, 25). The amounts of the intracellular level of the Z or
Siiyama protein measured by short pulse labeling of the
recombinant yeast was not much different from that of the wild type
(Figs. 3A and 4A, IC), whereas the steady-state
level of the Z or Siiyama protein in the total
intracellular fraction was lower than that of the wild type (Figs.
3B and 4A, IC), indicating that the intracellular
form of the Z or Siiyama protein is degraded in the yeast
cells. The preferential degradation of Z variant
1-AT in
the yeast exocytic pathway as shown in the present study has been also
observed in another independent study (26). In the present yeast
system, the unglycosylated cytoplasmic form of the variant proteins as
well as the core-glycosylated form retained in the secretion pathway
also underwent degradation (Fig. 5). This is strong evidence that the
intracellularly retained form of these genetic variants is not the
native state, which was also supported by in vitro folding
studies (Fig. 1). Interestingly, the degradation pattern of
Siiyama protein was somewhat different from that of the Z
type. The ER form of the Z type was retained and degraded slowly, while
the ER form of Siiyama degraded much more rapidly (Fig. 5).
On the other hand, the cytoplasmic form of the Z type degraded much
more rapidly than that of Siiyama (Fig. 5). It is quite
possible that the intracellularly accumulated forms of Z and
Siiyama, and consequently the pattern of their degradation,
could be different because of distinct origins of their defects: the
former produced from a trapped folding intermediate and the latter from
a denatured native structure.
The ER serves as a protein folding compartment for secretory proteins
and the role of the ER as a quality control system is now well
established (17, 27). The involvement of chaperones in the ER, such as
calnexin (28), was implicated in the retention and degradation of the
secretion-incompetent 1-AT variant in the ER (29, 30).
However, the detailed molecular mechanism of intracellular degradation
of Z or Siiyama variant has yet to be elucidated. The
degradation of intracellular Z and Siiyama
1-AT does not seem to involve vacuolar proteases, since
the yeast strain employed in the present study has disruption of the
PEP4 gene, which results in quite depressed levels of the
three major proteases, PrA (Pep4p), PrB, and PrC(CPY) (31).
In conclusion, the present study demonstrated that the folding and
stability defect of the Z and Siiyama variant
1-AT is the fundamental cause of secretion block and
intracellular degradation. The results present direct supporting
evidence for the kinetic defect of folding associated with the Z
variant and the stability loss associated with the Siiyama
variant. The defects can be corrected by other mutations that improve
the folding efficiency and stability compensation of the variants,
which appears to be needed for maintaining the secretion-competent
state and protection against proteolysis. A protein folding defect that
is accompanied by aggregation of the polypeptides has been tentatively
identified as a cause of several other human diseases (32, 33). Our
results demonstrate that although the apparent causes for the secretion
block of the
1-AT variants are similar, that is
aggregation, the fundamental causes can be quite different.