| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
and C8
1
Department of Chemistry and Biochemistry and School of Medicine, University of South Carolina, Columbia, SC 29208
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, C8ß,
and C8
) that are encoded in separate genes. In C8 isolated from
serum, these are arranged as a disulfide-linked C8- dimer that is
noncovalently associated with C8ß. In this study, a recombinant form
of C8- was expressed independently of C8ß in insect cells and
COS-7 cells and was shown to be equivalent to serum-derived C8-
with respect to its ability to combine with C8ß and form functional
C8. Also expressed separately were mutant (mut) forms of C8 and
C8 in which the single interchain disulfide bond was eliminated.
MutC8 exhibited the ability to combine with C8ß and express
hemolytic activity, although at a lower level than human C8. Addition
of purified mutC8 increased this activity, presumably by binding to
mutC8. A possible role for C8 as a retinol binding protein was
also investigated. Absorbance spectroscopy and fluorescence emission
and quenching revealed no specific binding of retinol to mutC8.
Together, these results indicate that 1) the biosynthesis and secretion
of C8- is not dependent on C8ß, which is consistent with in
vivo observations in C8ß-deficient humans; 2) C8 can be
synthesized independently of C8; therefore, protection of C8 from
premature membrane interactions during biosynthetic processing is not a
likely function of C8; 3) C8 enhances but is not required for
expression of C8 activity; and 4) C8 does not bind retinol;
therefore, it cannot function as a retinol transport protein. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(Mr = 64,000), ß
(Mr = 64,000), and
(Mr = 22,000) subunits that are encoded
in separate genes (3, 4, 5). C8 and C8 are linked by a single
disulfide bond to form a C8- dimer that is noncovalently
associated with C8ß. The affinity between C8- and C8ß is
high, as evidenced by their ability to associate when purified and
recombined at low molar ratios (6).
C8 and C8ß fulfill distinct roles in the formation and function of
MAC6 (7). C8 contains the
binding site that mediates interaction of C8- with C8ß, and it
has a binding site for C8, which facilitates intracellular formation
of C8- (8). It also contains a binding site for C9, which
functions to direct the first of multiple C9 molecules into MAC, and it
contains the site recognized by CD59, the membrane-associated protein
that protects homologous cells from lysis by inhibiting formation of a
functional MAC (9, 10). C8ß contains at least two binding sites, one
that mediates interaction with C8 and a second that facilitates C8
incorporation into MAC by binding to the intermediate C5b-7 complex
(11). With the exception of the CD59 binding site, the locations of
these sites within each subunit is unknown.
To date only human C8ß has been expressed independently as a
recombinant protein, but several lines of evidence suggest that
C8- has similar potential (12). Studies using rat hepatocytes
indicate that C8- and C8ß normally associate before secretion;
however, C8- is synthesized at a faster rate and in excess of
C8ß, resulting in the production of free C8- along with intact
C8 (13). Low levels of free C8- have also been detected in normal
human serum (14). In humans with hereditary C8ß deficiency,
significant levels of C8- are present despite the absence of
C8ß (15, 16). Together, these observations suggest that the in vivo
synthesis, secretion, and stability of C8- are not dependent on
C8ß. Whether the in vivo expression of C8 occurs independently of
C8 is not known.
C8 and C8ß are members of the MAC family of structurally related
proteins (2, 17, 18). In contrast, C8 is unrelated and is a member
of the lipocalin family of widely distributed proteins that bind and
transport small lipophilic ligands, e.g., retinol, pheromones,
odorants, etc. (19). Crystal structures of several lipocalins indicate
a common ß-barrel shape with a well-defined binding pocket lined with
nonpolar residues (20, 21, 22). Although an integral part of C8, C8 is
not essential for C8 lytic activity, and its precise role in the
formation and function of MAC is still unknown (23). It has been
suggested that C8 may function to shield C8 from premature
membrane interaction during intracellular processing or while in the
circulation before incorporation of C8 into MAC (7). Others reported
that C8- binds retinol and suggested that C8 may function as a
retinol transport protein analogous to the lipocalins BLG and RBP
(24).
In this report, we describe the expression and functional
characterization of a recombinant form of human C8- and provide
direct evidence that it can be synthesized independently of C8ß. In
addition, recombinant forms of C8 and C8 were expressed
separately, indicating no interdependence between these subunits for
synthesis and secretion. Characterization of recombinant C8 and
C8 confirmed that C8 enhances but is not required for C8
hemolytic activity, while spectroscopic studies established that C8
is not a retinol binding protein.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human C8 was purified from plasma fraction III provided by Miles
(Berkeley, CA) (25). C8- and C8ß were purified by gel
filtration in high ionic strength buffer (26). Radioiodinations were
performed using the Iodogen procedure. Molar concentrations of C8,
C8-, and C8ß were determined based on published
2801% values and established m.w. (23). Rabbit
antiserum against human C8- was produced by standard procedures.
A purified mAb specific for C8 was provided by Dr. Peter J. Sims of
the Blood Center of Southeastern Wisconsin. Bovine BLG (variant B) was
purchased from Sigma (St. Louis, MO) and was determined to be free of
retinol by the absence of absorbance at 330 nm. The molar concentration
was calculated using 278 = 17,600.
Baculovirus expression constructs
To prepare wtC8, a previously described cDNA clone A1 was
modified as follows (3). The full-length C8 cDNA insert in pAT153
PvuII-8 was excised with HindIII/SalI
and cloned into pBluescript II (Stratagene, La Jolla, CA). The 5'
untranslated sequence was truncated by unidirectional exonuclease
deletions of a SpeI/HindIII digest of this
plasmid. After ligation and subcloning, a full-length
SpeI/SalI fragment was released, modified by
addition of HindIII linkers, and cloned into the
HindIII site of pBluescript II. This plasmid was also used
as a template to generate mutC8 containing a
Cys164
Gly164 substitution. This was
accomplished by PCR-mediated site-directed mutagenesis using
pBluescript II universal primers and overlapping primers encoding the
desired codon substitution. The resulting full-length PCR product was
digested with HindIII and subcloned into pBluescript II, and
a 5' KpnI/BamHI fragment was exchanged with the
corresponding fragment in the wtC8 construct.
To prepare wtC8, a previously described full-length C8 cDNA was
excised with PstI and subcloned into pBluescript II (5). The
5' untranslated sequence was truncated by exonuclease deletions of a
SpeI/BamHI digest of this plasmid. After
religation and subcloning, a full-length
XbaI/HindIII fragment was released, modified by
addition of HindIII linkers, and cloned into pBluescript II.
This plasmid was also used as a template for PCR site-directed
mutagenesis to create mutC8 containing
Cys40Gly40. This product was likewise cloned
into the HindIII site of pBluescript II.
Recombinant baculovirus constructs were prepared using the MaxBac Baculovirus Expression System and protocols provided by the supplier (Invitrogen, San Diego, CA). For each construct, full-length HindIII fragments were isolated, partially filled in, and ligated to the NheI site of pBlueBac transfer plasmid. DNA was purified by CsCl gradient centrifugation and cotransfected with linear wild-type AcMNPV DNA into Spodoptera frugiperda (Sf9) cells using the cationic liposome method. Transfected cells were incubated at 27°C in supplemented Grace’s insect medium containing 10% FBS. Recombinant virus was plaque purified, and high titer stocks were prepared by amplification in Sf9 cells cultured in EX-CELL 400 medium (JRH Biosciences, Lenexa, KS).
COS cell expression constructs
The wtC8 and wtC8 constructs were prepared by excising
full-length HindIII fragments from their respective
pBluescript II plasmids and ligating after partial fill-in to the
XbaI site of the expression vector pcDNA3 (Invitrogen). The
orientation and integrity of the inserts were confirmed by PCR and
sequencing.
Expression in insect cells
High titer recombinant baculovirus was used to infect
Trichoplusia ni (High Five) cells (Invitrogen). Monolayer
cultures were grown in EX-CELL 400 medium at 27°C for 72 h.
Control medium was prepared using wild-type baculovirus. After
harvesting, media were centrifuged at 1000 x g and
concentrated twofold by ultrafiltration. For immunoblotting, medium
samples were precipitated with 10% TCA and electrophoresed on SDS-PAGE
gels. Proteins were transferred to nitrocellulose using standard
procedures, probed with rabbit anti-human C8-, and visualized
using goat anti-rabbit IgG horseradish peroxidase. Because of
problems with quantitation by ELISAs, concentrations of rC8- and
mutC8 in expression medium were estimated from immunoblots using
human C8- as a standard.
Expression in COS-7 cells
Purified plasmid DNA was used to transfect COS-7 cells by the
DEAE-dextran method. Following transfection, cells were grown in 5%
CO2 at 37°C in DMEM (Life Technologies, Grand
Island, NY) supplemented with 10% FBS. After 24 h, medium was
removed, and cells were washed with HBSS and incubated for an
additional 48 h in Opti-MEM I reduced serum medium (Life
Technologies). Control medium was prepared by culturing nontransfected
cells under identical conditions. Harvested medium was centrifuged at
2300 x g and concentrated by ultrafiltration.
Immunoblots were prepared as described above using goat anti-rabbit
IgG alkaline phosphatase as the secondary Ab and the Immun-Lite II
Chemiluminescent Assay Kit (Bio-Rad, Richmond, CA) to visualize immune
complexes. Concentrations of rC8- and mutC8 were estimated by
ELISA using a mAb specific for C8 and human C8- as the
standard.
Purification of recombinant proteins
MutC8 was partially purified using a human C8ß-agarose
affinity column. Concentrated medium from insect cells was applied at
25°C to resin equilibrated in 50 mM sodium phosphate, pH 7.0. Bound
protein was eluted in the same buffer containing 0.5 M NaCl and
dialyzed back into starting buffer. Because trace C8ß contaminants
are introduced from the affinity resin in the above step, the sample
was subsequently incubated with rabbit anti-human C8ß IgG
affinity resin in the same buffer. The final concentration was
determined by ELISA using a mAb specific for C8 with human C8-
as the standard. The resulting mutC8 was about 90% pure as
determined by SDS-PAGE, with yields of about 25 µg/100 ml of
medium.
MutC8 was harvested from insect cell expression medium by batch
adsorption. Equal volumes of medium and 10 mM sodium phosphate/20 mM
NaCl, pH 6.0, were incubated with carboxymethyl-Sephadex C-50 at
25°C. Bound protein was eluted with 10 mM sodium phosphate/600 mM
NaCl, pH 6.0, dialyzed into 10 mM sodium phosphate/20 mM NaCl, pH 6.0,
and applied to a carboxymethyl-Sepharose column. MutC8 was eluted
with a linear gradient using 1.5 M NaCl in the limit buffer. The
product was about 95% pure as determined by SDS-PAGE, with yields of
about 300 µg/100 ml of expression medium. Amino acid analysis
confirmed the identity of the product, while sequencing indicated a
blocked N-terminus, consistent with C8 isolated from human serum
(5). The 2801% in 10 mM sodium phosphate/20 mM
NaCl, pH 6.0, was determined to be 17.4 by quantitative amino acid
analysis.
Sucrose density gradients
rC8- or mutC8 in expression medium or control medium
supplemented with human C8- was incubated at varying molar
excesses over radioiodinated C8ß for 30 min in 5 mM imidazole, 72.2
mM NaCl, 0.5 mM MgCl2, 0.15 mM CaCl2, and
1 mg/ml BSA, pH 7.4, at 25°C. Binding to C8ß was analyzed on linear
5 to 10% (w/v) sucrose density gradients prepared in the same
buffer (9).
Hemolytic assays
rC8- or mutC8 in expression medium was mixed with human
C8ß in 5 mM imidazole, 72.2 mM NaCl, 0.5 mM MgCl2,
0.15 mM CaCl2, 2.5% D-glucose, 0.05% gelatin,
and 1 mg/ml BSA, pH 7.4. Purified human C8- in the same buffer
was used as a standard. Samples were incubated at 25°C for 30 min and
then subjected to hemolytic assays using sheep EAC1–7 as described
previously (27). In assays using purified mutC8 and mutC8,
samples were treated similarly, except that mutC8 was preincubated
for 10 min at 25°C with an excess of mutC8.
Spectroscopic analysis of retinol binding
All trans-retinol was purchased from Sigma.
Stock solutions were prepared in absolute ethanol and stored under
N2. The molar concentration in ethanol was determined using
325 = 52,800 (28). All proteins were prepared in 50 mM
sodium phosphate/150 mM NaCl, pH 7.4. Retinol was introduced by
stepwise addition of a stock solution such that the final ethanol
concentration was maintained at <1.5% (v/v). Samples were incubated
for 5 min at 25°C before spectroscopic analysis. To detect
fluorescence emission of bound retinol, excitation and emission
wavelengths were 330 and 470 nm, respectively. To measure tryptophan
fluorescence quenching, excitation and emission wavelengths were 285
and 340 nm, respectively.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
The results in Figure 1
show that a
rC8- dimer similar in size to human C8- can be produced by
coinfecting insect cells or cotransfecting COS-7 cells with wtC8 and
wtC8 expression constructs. Upon reduction, rC8- from either
cell type yields C8 and C8 chains similar in size to those in
human C8-. This similarity is most evident for recombinant
products from COS-7 cells, presumably because these cells add
complex-type N-linked carbohydrate that is normally found on
C8 (3). In the case of insect cells, a doublet is observed for
rC8-. This may be due to heterogeneity in high mannose-type
N-linked carbohydrate that is typical of recombinant
glycoproteins produced in insect cells, or it may reflect variation in
internal cross-linking of disulfide bonds in C8. The loss of
heterogeneity after reduction favors the latter explanation. Also noted
is the low expression level in control cells that are
infected/transfected separately with wtC8 or wtC8. This suggests
intracellular instability due to a free Cys in each protein, the
exception being wtC8 in insect cells which is secreted at a high
level as a monomer and as a cross-linked dimer.
|
- to combine with purified human C8ß is
demonstrated in Figure 2. The low protein
concentration used in these particular experiments (
100 ng/ml) and
the fact that only eightfold (insect cells) or twofold (COS-7 cells)
excesses of rC8- are required to achieve complete combination
indicate a high affinity for C8ß. Whether the higher level required
with insect cell rC8- is significant is questionable because of
uncertainty in quantitating recombinant product in the medium. Also
noted is the broader sedimentation profile and shoulder in gradients
using COS-7 cell medium (bottom) compared with
insect cell medium (top). This is attributed
to differences in radiolabeled C8ß preparations and variability
between gradients. Importantly, it is not indicative of incomplete
combination, since rC8- at two- and fourfold excesses yields the
same profile, which is essentially identical with that obtained with a
twofold excess of C8-.
|
- when combined with C8ß is shown in Figure 3. With rC8- from insect cells, the
same hemolytic activity is observed at two- and eightfold excesses over
C8ß. Although the results in Figure 2
indicate that rC8- is not
fully combined at a twofold excess, similar activities at both levels
suggest that the same amount of C8 is being formed. This apparent
contradiction is probably due to the ability of EAC1–7 to bind
transiently formed C8 and shift the equilibrium toward further C8
formation. Importantly, the activity at either excess is comparable to
that observed with human C8-. With rC8- from COS-7 cells,
the results indicate that the activity is equivalent to human
C8-.
|
and mutC8
The interdependency of C8 and C8 for synthesis and secretion
is examined in Figure 4. Mutant forms of
each were prepared in which the single Cys involved in interchain
cross-linking was eliminated. When expressed separately in insect or
COS-7 cells, monomer forms of mutC8 and mutC8 are independently
secreted at relatively high levels compared with wtC8 and wtC8
(Fig. 1). Also noted is a small amount of mutC8 dimer in the medium
from insect cells, which is eliminated upon reduction. This dimer
probably results from a disulfide exchange involving the single
internal disulfide bond in mutC8. The results in Figure 5 show that mutC8 is capable of
combining with C8ß. Only low molar excesses of mutC8 from insect
cells (eightfold) or COS-7 cells (twofold) are required for complete
combination; thus, the affinity of mutC8 for C8ß is comparable to
those of rC8- and human C8-.
|
|
is compared with C8- in Figure 6. With mutC8 from insect cells, the
activity is similar at four- and eightfold excesses over C8ß. With
mutC8 from COS-7 cells, the activity is identical at two- and
fourfold excesses. In both cases, however, mutC8 is only about 10 to
15% as active as human C8- when compared at 50% lysis.
Increasing mutC8 to higher excesses produced no further increase in
activity (data not shown).
|
and C8 retain binding affinity for each other
after cleavage of the interchain disulfide bond in C8- (8). To
investigate whether mutC8 has a similar affinity for C8 and if
the presence of C8 influences the activity of C8, as suggested by
the results shown in Figure 6, mutC8 and mutC8 were purified and
used in comparative hemolytic assays. As shown in Figure 7, the activity of mutC8 in the
presence of C8ß is progressively increased by preincubating with
increasing amounts of purified mutC8. At the highest excess
(300-fold), the activity is nearly identical with that obtained with
human C8-. This suggests that mutC8 is binding mutC8 to
produce a complex that is functionally equivalent to C8-.
|
- and mutC8
To determine whether retinol is a ligand for C8, several
spectroscopic methods were used to detect binding to mutC8. BLG, a
well-characterized retinol binding protein, and human C8- were
used as controls. C8- was included because it reportedly binds
retinol at high ionic strengths (24).
It is well established that the OD330 of retinol bound
to proteins such as BLG or RBP in aqueous solution is comparable to
that of free retinol in nonpolar solvents because the ligand is in a
stable, hydrophobic environment in both cases (29). In contrast, the
relative absorbance of unbound retinol is greatly reduced due to
inherent instability and breakdown in an aqueous environment. This
behavior is apparent in the control experiment in Figure 8 (top), in
which the absorbance spectrum of BLG after addition of an equimolar
amount of retinol is compared with retinol in buffer. The broad peak at
330 nm is typical of a BLG-retinol complex (29, 30). In experiments not
shown, less ligand produced a lower OD330, while higher
amounts produced no net increase, indicating that saturation of the
retinol binding site is complete at an approximately 1:1 ratio of added
ligand to protein. Because they are at the same concentration, one
would predict the OD330 to be similar to the BLG-retinol
complex if an equivalent amount of retinol binds to C8- and
mutC8. However, addition of an equimolar amount to either protein
yields an OD330 similar to that of the buffer control,
which corresponds to unbound and unprotected retinol. Addition of
higher amounts of ligand produced no net increase in
OD330.
|
- and mutC8. Results
from the control experiment shown in Figure 8
(middle) show that as retinol is added to BLG,
fluorescence increases steadily until saturation of the binding site is
achieved. This behavior is in agreement with results obtained by others
(31, 32). Also noted is that saturation occurs at <1:1 molar
equivalent of added retinol. This has been observed for both BLG (33)
and RBP (32) and probably reflects variation in the quality of the
protein preparations. Importantly, no increase in fluorescence is
observed relative to the buffer control when as much as a twofold molar
excess of retinol is added to C8- or mutC8.
Binding of retinol to BLG or RBP results in significant quenching
of tryptophan fluorescence (32, 33), which suggests that one or more
tryptophan residues is in close proximity to the binding site. This is
consistent with crystallographic data for these proteins (20, 21).
Comparative sequence alignments indicate that the two tryptophans in
BLG and two of the four in RBP closely align with the two in C8. For
BLG, RBP, and C8, these residues are predicted to reside within
segments that form the core ß-barrel structure that is characteristic
of the lipocalins (19). Furthermore, molecular modelling predicts that
C8 and BLG have similar overall folding patterns with a binding
pocket that could accommodate retinol (24). In light of this, BLG,
C8-, and mutC8 were titrated with retinol, and fluorescence
quenching was measured. As shown in Figure 8
(bottom), fluorescence of BLG is quenched
about 38% at 1 molar equivalent of added retinol, which agrees well
with results reported by others (33). In contrast, addition of an
equimolar amount of retinol quenches the fluorescence of C8- and
mutC8 by only about 6 and 14%, respectively. When considered in
relation to above results, this small amount of quenching is probably
caused by nonspecific binding.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- can occur
independently of C8ß. This conclusion is consistent with the finding
of free C8- in C8ß-deficient and normal human sera.
Furthermore, the data show that rC8- produced in either
nonmammalian or mammalian cells is capable of associating with human
C8ß and forming hemolytically active C8. The high affinity for C8ß
and similar activity indicate that rC8- is functionally
equivalent to C8- derived from human serum. This result and
earlier studies which determined that functional recombinant C8ß can
be produced independently of C8- suggest the feasibility of using
recombinant subunits in deciphering structure-function relationships in
C8 (12).
The ability to synthesize a recombinant form of C8 is
significant with respect to understanding the function of C8.
Previous studies showed that after limited cleavage of disulfide bonds
in C8-, C8 has the capacity to noncovalently bind C8 (8).
However, the scope of these studies was restricted by the inability to
select for cleavage of interchain vs internal disulfide bonds and the
fact that only trace amounts of C8 and C8 could be produced.
Nevertheless, they established that C8 isolated in this manner binds
to C8, and it was concluded that C8 contains a specific binding
site for C8. This and the fact that C8 is a lipocalin led to the
proposal that C8 may function to shield or otherwise protect
hydrophobic regions of C8 from premature interaction with membranes
during biosynthetic processing (7). Because the present results show
that mutC8 can be synthesized independently of C8, this
possibility can now be excluded from consideration as a likely function
for C8.
The fact that mutC8 associates with C8ß and forms lytically
active C8 is in agreement with results obtained with C8 and C8
isolated from C8- (8, 9). Those studies found that C8 contains
a binding site for C8ß and a site for C9, and that C8 · C8ß
is hemolytically active toward EAC1–7. The binding affinity of
C8 · C8ß for EAC1–7 was considerably less than that of intact
C8, but increased to comparable levels in the presence of added C8
(23). Similar behavior was observed with mutC8 and mutC8 in the
present study. When combined with C8ß, mutC8 is hemolytically
active but at a lower level than C8- (Fig. 6). Upon adding
mutC8, the activity increases to a level approaching that obtained
with C8- (Fig. 7). This result and the earlier binding studies
suggest that the interaction of C8 with C8 increases the affinity
of C8 for EAC1–7, although the exact mechanism by which this occurs is
unknown. These results also provide further support for the conclusion
that C8 contains binding sites for C8ß and C9 and that C8 is
not required for lytic activity. In addition, they demonstrate that
C8 and C8 can be produced separately as recombinant proteins that
retain functionality.
A role for C8 in complement-mediated cell lysis remains to be
identified. It is not required for the synthesis or function of C8 as
shown in the present study, nor is it essential for the formation and
lytic activity of MAC. Because it is a lipocalin, it may bind an as yet
unidentified small ligand. One group suggested that this ligand may be
retinol because of the predicted similarity in folding to BLG and the
observed binding of radiolabeled retinol to C8- (24). However,
the significance of the latter result is tempered by several concerns.
One is that binding was performed in 2 M NaCl, which reportedly was
necessary to provide access to the putative ligand binding site on
C8 by disrupting electrostatic interactions in C8-. Secondly,
analysis of the published data indicates that the amount of ligand
bound was extremely small under these conditions and corresponded to a
molar ratio of only about 1:6000 (retinol:C8-). Direct binding to
C8 was not measured due to an inability to preserve internal
disulfide bonds upon cleavage of C8-. Considering the high ionic
strength used, the hydrophobic nature of retinol, and the small amount
of ligand bound, it is questionable whether the observed binding is
meaningful.
In the present study, the availability of a recombinant form of
C8 enabled its retinol binding potential to be examined directly and
by methods commonly used with well-characterized retinol binding
proteins. Under conditions approximating physiologic ionic strength and
where specific binding to BLG is readily detectable, retinol binding to
C8- and mutC8 was not observed. In experiments not shown,
C8- and mutC8 were also incubated with high molar excesses
(50-fold) of retinol in buffer containing either 0.15 or 2 M NaCl.
After removing free retinol by dialysis, absorbance profiles revealed
only small amounts of bound material with the spectral characteristics
of retro-retinol breakdown products (29). Therefore, we conclude that
neither C8- nor C8 specifically bind retinol and that C8 is
not a retinol transport protein. The small amount of binding to
C8- observed by others at an extreme ionic strength is probably
due to nonspecific partitioning into hydrophobic sites on the
protein.
In conclusion, the ability to independently produce functional,
recombinant C8 and C8ß will facilitate efforts to localize and
fine-map the multiple sites of interaction on these subunits. It will
also be possible to more precisely define the role of each subunit in
the formation, function, and regulation of MAC and to gain a general
understanding of the basis for the highly specific binding interactions
that occur between MAC proteins. With respect to C8, the
availability of recombinant protein will enable the use of more focused
approaches to identifying a possible biologic function for this
subunit, which to date remains elusive. In particular, crystallization
of C8 is now feasible to consider. Results from such an effort could
confirm the prediction of a lipocalin-like structure for C8 as well
as provide insight into the identity of its natural ligand.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Present address: U.S. Army Medical Research Institute of Chemical Defense, Aberdeen Proving Ground, MD 21010.
3 Present address: Arthritis and Immunology, Oklahoma Medical Research Foundation, 825 NE 13th St., Oklahoma City, OK 73104.
4 Present address: Laboratory of Pathology, National Cancer Institute, Building 10, Room 2N106, Bethesda, MD 20892.
5 Address correspondence and reprint requests to Dr. James M. Sodetz, Department of Chemistry and Biochemistry, University of South Carolina, Columbia, SC 29208. E-mail address:
6 Abbreviations used in this paper: MAC, the membrane attack complex of complement; BLG, ß-lactoglobulin; RBP, serum retinol binding protein; wtC8 and wtC8, wild-type recombinant human C8 and C8, respectively; mutC8 and mutC8, the respective mutant forms of recombinant human C8 and C8 containing a single CysGly substitution; rC8-, recombinant human C8-; EAC1–7, sheep E carrying human C1-C7.
Received for publication January 15, 1998. Accepted for publication February 26, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
subunit and C9. Biochemistry 26:3565.[Medline]
subunit of the eighth component of human complement: identification and characterization of a distinct binding site for the chain. Biochemistry 24:4603.[Medline]
subunit of C8. Biochemistry 24:4598.[Medline]
- and ß subunits. J. Immunol. 139:3021.[Abstract]
-. Pediatr. Res. 27:234.[Medline]
and further examination of its linkage to C8 ß. Hum. Genet. 95:513.[Medline]
chain reveals it has no direct role in cytolysis. Biochemistry 23:832.[Medline]
, a member of the lipocalin protein family. Mol. Immunol. 28:123.[Medline]
is not the target of homologous restriction factors. J. Immunol. 144:3087.[Abstract]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), C8ß
plus twofold C8
), and C8ß (
). Bottom,
Results obtained with medium from the corresponding transfected COS-7
cells: C8ß plus twofold rC8
), and twofold C8
).
Bottom, Results using medium from transfected COS-7 cells:
C8ß plus twofold rC8
