(Received for publication, December 10, 1996, and in revised form, January 27, 1997)
From the Edward Mallinkrodt Department of Pediatrics and the
Department of Biochemistry and Molecular Biophysics, Washington
University School of Medicine, St. Louis, Missouri 63110
The 39-kDa receptor-associated protein (RAP) is a
molecular chaperone for the low density lipoprotein receptor-related
protein (LRP), a large endocytic receptor that binds multiple ligands. The primary function of RAP has been defined as promotion of the correct folding of LRP, and prevention of premature interaction of
ligands with LRP within the early secretory pathway. Previous examination of the RAP sequence revealed an internal triplication. However, the functional implication of the triplicated repeats was
unknown. In the current study using various RAP and LRP domain constructs, we found that the carboxyl-terminal repeat of RAP possesses
high affinities to each of the three ligand-binding domains on LRP,
whereas the amino-terminal and central repeats of RAP exhibit only low
affinity to the second and the fourth ligand-binding domains of LRP,
respectively. Using truncated soluble minireceptors of LRP, we
identified five independent RAP-binding sites, two on each of the
second and fourth, and one on the third ligand-binding domain of LRP.
By coexpressing soluble LRP minireceptors and RAP repeat constructs, we
found that only the carboxyl-terminal repeat of RAP was able to promote
the folding and subsequent secretion of the soluble LRP minireceptors.
However, when the ability of each RAP repeat to inhibit ligand
interactions with LRP was examined, differential effects were observed
for individual LRP ligands. Most striking, both the amino-terminal and
central repeats, but not the carboxyl-terminal repeat, of RAP inhibited
the interaction of 2-macroglobulin with LRP. These
differential functions of the RAP repeats suggest that the roles of RAP
in the folding of LRP and in the prevention of premature interaction of
ligand with the receptor are independent.
The 39-kDa receptor-associated protein (RAP)1 is an unique receptor antagonist. The target receptors for RAP are cysteine-rich endocytic receptors that belong to the low density lipoprotein (LDL) receptor family (1). The four representative receptors in this family are the LDL receptor (2), the LDL receptor-related protein (LRP, Ref. 3), glycoprotein gp330/megalin (4), and the VLDL receptor (5). Among these receptors, LRP and gp330/megalin are large multifunctional receptors with multiple ligand-binding domains, which bind several structurally and functionally distinct ligands (for reviews, see Refs. 1 and 6). While RAP exhibits high affinities for LRP, gp330/megalin, and the VLDL receptor, it binds only weakly to the LDL receptor (7). Upon binding to these receptors, RAP inhibits the binding and/or endocytosis of all the ligands by the receptors. This unique feature of RAP has allowed its extensive use in biological studies of these endocytic receptors. Recent evidence has suggested that, under normal physiological conditions, RAP is an endoplasmic reticulum (ER) resident protein and functions within the early secretory pathway (8-10). Using LRP as the target protein, it was found that RAP retained within the ER functions as a regulator of LRP activity by transiently interacting with LRP and maintaining LRP in an inactive ligand-binding state. As RAP dissociates from LRP in response to the lower pH within the Golgi, LRP becomes active as it transits to the cell surface (9). The role of RAP in the maturation and trafficking of LRP is further supported by gene-knockout studies (11), which demonstrate that cells lacking RAP exhibit a 75% reduction of functional LRP.
LRP is the largest endocytic receptor identified to date (~600 kDa). It is synthesized as a single polypeptide chain and cleaved in the trans-Golgi into two subunits (12, 13). The 515-kDa extracellular subunit contains 31 copies of complement-type ligand-binding domains arranged in four clusters with 2, 8, 10, and 11 repeats, respectively (1, 3). Also present in this subunit are 22 copies of cysteine-rich epidermal growth factor (EGF) precursor-type repeats which flank the ligand-binding domains. The complement-type repeats in LRP are similar to those in the LDL receptor in which the 40-residue-long cysteine-rich repeats exhibit a highly conserved spacing pattern of six cysteine residues that form three intramolecular disulfide bonds (14). The disulfide bonds are believed to be important for the stability of the ligand-binding sites on the receptor. The complexity of LRP's structure, largely due to the extensive intradomain disulfide bonds, presents a challenging task for proper folding during its biosynthesis. This process may well be assisted by molecular chaperone(s) within the ER. Indeed, using anchor-free, soluble mini-receptors that represent each of the four putative ligand-binding domains of LRP (SLRPs), our most recent studies (15) showed that coexpression of RAP is both necessary and sufficient for the correct folding and subsequent secretion of the SLRPs. Without the coexpression of RAP, SLRPs are misfolded due to the formation of intermolecular disulfide bonds and are retained within the ER with little secretion. It is not known at present whether the role of RAP in the receptor's folding is independent from its function in preventing premature ligand interaction with the receptor.
The HNEL tetrapeptide at the carboxyl terminus of RAP has been shown to
mediate its ER localization and retention (9). In addition to this
ER-retention signal, examination of the RAP sequence also identified an
internal triplication (9). In the present study, using various
molecular and cellular approaches, we have analyzed the function of
each of the three repeats of RAP. We found that while the
carboxyl-terminal repeat of RAP functions similarly to the full-length
RAP in terms of interacting with the receptor and in assisting the
receptor to fold, inhibition of at least one LRP ligand,
2-macroglobulin, can only be achieved with the
amino-terminal or the central repeats of RAP. These differential functions of the RAP repeats suggest that the functions of RAP in
receptor folding and inhibition of ligand interactions are independent.
Construction of
cDNAs for SLRP1, SLRP2, SLRP3, and SLRP4 using polymerase chain
reaction (PCR) have been described previously (15). The same strategies
were used for generating additional soluble minireceptors of LRP with
an HA tag inserted after the signal cleavage site. All oligonucleotides
were synthesized in the Washington University School of Medicine
Protein Chemistry Laboratory. The regions represented in the new SLRPs
are illustrated in Fig. 4, and contain the following amino acid
sequences (3): SLRP2N, 787-994; SLRP2C, 995-1244; SLRP2N-EGF,
826-994; SLRP2C-EGF, 995-1164; SLRP3N, 2462-2712; SLRP3C,
2713-3004; SLRP4N, 3274-3553; and SLRP4C, 3554-3843.
Construction of cDNAs for RAP Repeats
Construction of
RAP repeats for cell-transfection was carried out via two sequential
steps of subcloning. In the first step, the 5 fragment of RAP (signal
peptide + amino acids 1-17) was generated by PCR using pcDNA-RAP
(9) as the template and the following two primers: forward
primer (5
-GATCGGATGATGGCGCCGCGGAGGGTCA-3
) and reverse primer
(5
-GATCCATCATCATCATCATAGCGTAGTCCGGGACGTCGTACGGGTATCCGGACTCGCGTTTCGGGGACGG-3
). The reverse primer contained a sequence encoding an HA epitope and five methionine residues. The HA epitope sequence, YPYDVPDVA, is
derived from influenza hemagglutinin and can be recognized by a
monoclonal anti-HA antibody, 12CA5 (16). The resulting PCR product was
subcloned into pcDNA3 vector, and the resulting plasmid, termed
pcDNA-RAP5
, was used as the base vector for generating various
constructs of RAP. In the second step, the carboxyl-terminal portions
of RAP (amino acids 18-323, 18-110, 91-210, 191-323, or 18-250)
were generated by PCR and subcloned into pcDNA-RAP5
. The resulting
plasmids were used for cell transfection.
The methods for constructing GST/RAP constructs have been described previously (9).
Cell Culture and TransfectionHuman glioblastoma U87 cells were cultured in Earle's minimum essential medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 units/ml penicillin, 100 µg/ml streptomycin, 1 mM sodium pyruvate, and maintained at 37 °C in humidified air containing 5% CO2 (8). For transient transfection, U87 cells were transfected with various plasmids at 40-60% confluence using a calcium phosphate precipitation method (17). For each well of six-well dishes (3.5 cm in diameter), 10 µg of DNA were used in a total volume of 4 ml of medium. Sixteen hours after the start of transfection, cells were washed with medium and cultured continuously for an additional 24 h before use in experiments. The efficiency of transient transfection in these studies was consistently about 20-30% as assessed by immunofluorescent staining of expressed proteins.
Metabolic and Pulse-Chase LabelingMetabolic labeling with [35S]methionine or [35S]cysteine was performed essentially as described before (9). For pulse-chase experiments, cells were pulse-labeled for 1 h and chased with serum-containing medium for 3 h.
Antibodies, Immunoprecipitation, and SDS-Polyacrylamide Gel ElectrophoresisPolyclonal anti-RAP and anti-LRP antibodies have
been described before (9). Monoclonal anti-HA antibody was obtained
from BabCo (12CA5). Immunoprecipitations were carried out essentially as described before (18), except the washing buffer for monoclonal anti-HA antibody contained 0.1% SDS instead of 1% SDS. Preliminary experiments were performed to ensure that the primary antibody used in
each immunoprecipitation was in excess. Protein A-agarose beads were
used to precipitate protein-IgG complexes. The immunoprecipitated material was released from the beads by boiling each sample for 5 min
in Laemmli sample buffer (62.5 mM Tris-HCl, pH 6.8, 2%
(w/v) SDS, 10% (v/v) glycerol) (19). If the immunoprecipitated
material was analyzed under reducing conditions, 5% (v/v)
-mercaptoethanol was included in the Laemmli sample buffer. The
percentage of SDS-polyacrylamide gels is indicated in each figure
legend. Rainbow molecular weight markers (Bio-Rad) were used as the
molecular weight standards.
GST or GST/RAP constructs were electrophoresed on SDS gels and transferred to nitrocellulose membrane. After blocking with tissue culture medium, membranes were incubated with conditioned media harvested from SLRP and RAP cDNAs-cotransfected cells. The bound SLRPs were then detected with anti-HA antibody and enhanced chemiluminescence (ECL, Amersham).
Ligand Blotting, Binding, and Degradation AnalysisFor ligand blotting analysis, purified human LRP was electrophoresed on SDS gel and transferred to nitrocellulose membrane. Strips of membrane were then incubated with GST or GST/RAP constructs, and bound proteins were detected using anti-GST antibody. For binding and degradation analysis, GST/RAP fusion proteins were subjected to thrombin cleavage and cleaved GST was removed with glutathione agarose beads (20). The resulting recombinant proteins of RAP constructs (RAP-(1-323), RAP-(1-110), RAP-(91-210), and RAP-(191-323)) were used for competition assays. Each of analyzed RAP and its constructs was iodinated using the IODOGEN method (21). Binding of 125I-RAP to U87 cells in the absence or the presence of cold competitor has been described previously (8). All the LRP ligands tested have been described before (23-26). Each of these ligands was iodinated and used for cellular degradation analysis by measuring cell-mediated trichloroacetic acid-soluble radioactivity released into the overlying medium, in the absence or the presence of excess unlabeled RAP constructs. Each assay was carried out in triplicate, and the standard deviations were used for error bars.
In our previous studies (9), we
noted a possible internal triplication in the primary structure of RAP.
However, the boundaries and the relationship among the three repeats
were not clear. In the present studies, we have used several DNA
computer analysis programs to examine the RAP sequence. Shown in Fig.
1 is the sequence analysis using the GCG package
(Program Manual for the Wisconsin Package, version 8, Genetics Computer
Group). The programs Compare and Dotplot examine the sequence homology
by identifying both identical and similar residues along the sequence.
When RAP was compared with itself using a window of 30 residues and a
stringency of 15 residues, we found significant homology between
regions within 1-100 and 201-300, as well as between regions 101-200
and 201-300 (Fig. 1A). Only scattered homology was found
between regions of 1-100 and 101-200. Thus, we assigned the RAP
sequence into three repeats approximately equal in sizes (repeat 1 = 1-100; repeat 2 = 101-200; repeat 3 = 201-323). Using
these boundaries, we compared the sequence homology among the three
repeats. The lower homology between repeat 1 and repeat 2 prompted us
to compare each of the two repeats separately, using the Gap program of
the GCG package. We found that both repeat 1 and repeat 2 have high homology with repeat 3 (46.4% and 45.5% similarity, respectively), whereas repeat 1 and repeat 2 have relatively low homology (38.9% similarity). Thus, we speculate that if the three repeats of RAP were
derived from the same ancestral sequence, repeat 1 and repeat 2 may
have been derived from repeat 3 separately during evolution. Two
regions between repeat 1 and 3 were further identified for high
homology, which are represented by the two elongated homology lines in
Fig. 1A. Only one (albeit longer) region was found between repeat 2 and repeat 3 that shares high homology. These homologous regions are graphically illustrated in Fig. 1B.
Differential Interactions of RAP Repeats with LRP
To analyze
whether each of the three triplicated repeats interact similarly with
LRP, we performed ligand blotting analysis using GST/RAP constructs
representing each of the three repeats (9). In addition to the
full-length RAP (GST/RAP-(1-323)), three GST/RAP repeat constructs
were made, which slightly overlapped at the assigned boundaries
(i.e. GST/RAP-(1-110), GST/RAP-(91-210), and
GST/RAP-(191-323), see Fig. 2A). When these
GST fusion proteins were tested for interaction with LRP via ligand
blotting (Fig. 2B), we found that repeat 3 possessed similar
affinity for LRP as full-length RAP. Although they have much lower
affinities, repeat 1 and repeat 2 of RAP also appear to be capable of
interacting with LRP. Due to low specific radioactivity following
iodination of repeats 1 and 2, we were unable to conclusively compare
the affinities of RAP repeats with LRP on the cell surface directly. However, to examine the relationships of these interactions at the cell
surface, we performed competition analysis for the binding of
full-length RAP to U87 cells (8, 9) by each of the RAP repeat
constructs. To eliminate potential steric hindrance effects of GST on
ligand inhibition, we removed the GST portion from each of the fusion
constructs, and thereafter repurified the resulting RAP fragments.
Shown in Fig. 2C are the binding of either
125I-RAP-(1-323) or 125I-RAP-(191-323) to U87
cells in the absence or the presence of excess unlabeled competitors
(100-fold excess). As seen in the figure, in addition to the
full-length RAP, only repeat 3, but not repeats 1 and 2, demonstrated
inhibition of the binding of 125I-RAP-(1-323). For binding
of 125I-RAP-(191-323), both unlabeled full-length RAP and
repeat 3 maximally inhibited, while repeat 1 and repeat 2 inhibited
binding slightly. These results suggest that, among the three repeats
of RAP, the third repeat possesses the highest affinity for LRP,
consistent with the results obtained from the ligand blotting
analysis.
Differential Interactions of RAP Repeats with Ligand-binding Domains of LRP
In our previous studies, we have shown that RAP
binds to the second, third, and fourth ligand-binding domains of LRP
(15). To analyze to which of the ligand-binding domains each of the RAP
repeat binds, we performed experiments in which potential interactions
between a given RAP repeat and a given ligand-binding domain of LRP
were examined. Ligand-binding domains of LRP were represented by SLRPs,
which upon coexpression with RAP, are secreted into the media (15)
(also see illustrations in Fig. 4 for schematic representations of
SLRPs). GST/RAP constructs, separated via SDS-polyacrylamide gel
electrophoresis, were transferred onto nitrocellulose and incubated
with conditioned media containing SLRP2 (Fig.
3A), SLRP3 (Fig. 3B), or SLRP4
(Fig. 3C). The SLRPs that interacted with GST/RAP constructs
were then detected using anti-HA antibody (15). As seen in the figures,
no interaction of the negative control, GST, with any SLRP was
detected. Both full-length RAP (GST/RAP-(1-323)) and repeat 3 of RAP
(GST/RAP-(191-323)) interacted with each of the three SLRPs,
suggesting that repeat 3 of RAP contains the binding determinants for
interacting with each of the three ligand-binding domains on LRP. Weak
interaction of repeat 1 of RAP was detected only with SLRP2, whereas
weak interaction of repeat 2 of RAP was seen with only SLRP4. These
weaker interactions were enhanced when the amounts of GST/RAP fusion
protein were increased by 10-folds (as is the case presented in Fig.
3). These results suggest differential interactions of RAP repeats with
ligand-binding domains on LRP. Two additional RAP repeat constructs
were tested in these experiments, one with the ER-retention signal HNEL
deleted (GST/RAP-(1-319)), and one containing repeats 1 and 2, and
part of repeat 3 of RAP (GST/RAP-(1-250)). As shown in Fig. 3,
GST/RAP-(1-319) interacted similarly to full-length RAP with LRP,
indicating that the HNEL signal is not important for RAP interaction
with LRP. The fact that GST/RAP-(1-250) does not interact with SLRP3
is consistent with the notion that neither repeat 1, nor repeat 2 of
RAP interacts with the third ligand-binding domain of LRP, and suggests
residue 250-323 are important for interacting with SLRP3.
Our previous cell surface saturation binding analyses have suggested that each LRP molecule contains 5-7 binding sites for RAP (22). Thus, to further define the binding sites of RAP on the subdomains of LRP, we prepared constructs representing approximately each half of the three ligand-binding domains, with or without the flanking EGF precursor type repeats (see "Materials and Methods"). These regions are graphically illustrated in Fig. 4. Using these SLRPs and the coexpression of RAP, we have analyzed the interaction with each of the repeats within RAP, as well as the full-length RAP. The results are summarized in Table I. As seen in the table, the amino-terminal (SLRP2N) and carboxyl-terminal (SLRP2C) halves of the second ligand-binding domain of LRP (SLRP2) interacted equally well with both full-length RAP and the repeat 3 of RAP, suggesting that at least two independent RAP-binding sites exist within the second ligand-binding domain of LRP. Removal of the EGF precursor repeats from either the amino-terminal half (SLRP2N-EGF) or the carboxyl-terminal half (SLRP2C-EGF) had no effect on their interactions with RAP, suggesting that the EGF precursor repeats are not required for RAP's interactions with LRP. When the third ligand-binding domain of LRP was divided in half, only the carboxyl-terminal half (SLRP3C) exhibited interaction, albeit weak, with RAP, consistent with the notion that this repeat contains only one RAP-binding site likely localized near the middle of this repeat. Although the carboxyl-terminal half (SLRP4C) exhibited somewhat weaker interaction, both the amino-terminal (SLRP4N) and carboxyl-terminal (SLRP4C) halves of the fourth ligand-binding domain of LRP reacted with RAP, suggesting that this ligand-binding domain of LRP likely contains two RAP-binding sites. Taken together, we have thus identified at least five independent RAP-binding sites on LRP. The exact number of RAP-binding sites on LRP and their precise localization will require further investigation.
|
Also shown in Table I are results of interactions between different SLRPs and the three RAP repeat constructs. It appears that the binding sites for repeats 1 and 2 of RAP are both localized on the amino-terminal halves of the second and fourth repeats of LRP, respectively.
Role of RAP Repeats in the Folding and the Secretion of SLRPsTo examine the ability of RAP repeats to promote the
folding process of LRP, we constructed cDNAs corresponding to the
full-length RAP and each of its three repeats of RAP. An HA epitope and
five methionine residues were included in the constructs to monitor the
expression of these proteins after cell transfection. When these
cDNAs were transfected into U87 cells (8), we found abundant expression for each of them following metabolic labeling with [35S]methionine and immunoprecipitation with anti-HA
antibody (Fig. 5A). The expression levels of
RAP and its individual repeats after transfection were approximately
50-fold higher when compared with the endogenous RAP (data not shown,
see Ref. 15). In our previous studies, we developed a system by which
the folding of LRP can be evaluated using SLRPs (15). Correct folding
and the subsequent secretion of each of the three ligand-binding
domains of LRP (SLRP2, SLRP3, and SLRP4) require the coexpression of
RAP. Shown in Fig. 5B is an experiment in which SLRP2 is
expressed in U87 cells without or with coexpression of full-length RAP
(RAP-(1-323)). The secretion level of SLRP2 was monitored by analyzing
radiolabeled SLRP2 in the media and cell lysates after pulse labeling
with [35S]cysteine for 1 h and chase for 3 h.
As seen in the figure, without coexpression of RAP-(1-323), little
secretion (~8% of total radioactivity) of SLRP2 was seen, whereas
when RAP-(1-323) was coexpressed, about 50% of total
35S-SLRP2 was secreted. Using the same assay, we analyzed
the ability of each RAP repeat construct in the secretion of SLRPs.
Shown in Fig. 5C are results from a representative
experiment of the four performed. The percent of SLRP secretion after
pulse-chase labeling was plotted against individual constructs. As seen
in the figure, repeat 3 of RAP (RAP-(191-323)) functioned at least as
well (SLRP4) if not better (SLRP2 and SLRP3) than the full-length RAP
in its ability to assist the folding and secretion of each of the three
SLRPs. Both repeat 1 and 2 of RAP had little effect on the secretion of
SLRPs, possibly due to their lower affinity for the SLRPs. Also
examined in these experiments was RAP-(1-250), which functioned
similarly to the full-length RAP in the secretion of SLRP2 and SLRP4,
but not SLRP3, consistent with its ability to bind LRP (see Fig.
3).
Role of RAP Repeat Constructs in the Inhibition of Ligand Interactions with LRP
RAP has been used extensively in the study
of LRP ligands due to its ability to antagonize ligand interactions
with the receptor (1). To examine the ability of each of the RAP
repeats to inhibit LRP ligands, we analyzed the interaction of several
ligands with LRP in the absence or the presence of either the
full-length RAP or each of the repeat constructs of RAP. To exclude any
possible steric hindrance by GST, we proteolytically removed it from
each of the fusion proteins and repurified the RAP constructs. Since the primary binding sites on the cell surface of some LRP ligands are
not on LRP (e.g. tissue factor pathway inhibitor, or TFPI; see Ref. 23), but subsequent internalization and degradation is, we
performed ligand degradation assays using U87 cells. Each of the
ligands examined was radioiodinated and incubated with U87 cells for
4 h in the absence or the presence of RAP constructs. Shown in
Fig. 6 are summaries for four LRP ligands: t-PA (24), 2M* (25), antithrombin III-thrombin complex (26), and
TFPI (23). As seen in the figure, full-length RAP inhibited the
degradation of each of the ligands with LRP. Repeat 3 of RAP inhibited
the degradation of 125I-t-PA, 125I-antithrombin
III-thrombin complex, and 125I-TFPI, but not
125I-
2M*. Interestingly, while both repeat 1 and repeat 2 of RAP were generally inefficient in inhibiting other
ligands, these repeats were very effective in inhibiting the
interaction of
2M* with LRP. Thus, LRP ligands were
differentially inhibited by RAP repeat constructs, consistent with the
hypothesis that LRP ligands bind to different sites on the
receptor.
Despite being widely used as an antagonist to LRP on the cell surface, RAP has recently been defined as a specialized ER chaperone and functions during LRP's folding and subsequent trafficking along the early compartments of the secretory pathway (9, 10, 15). Previously it had been noted that RAP contained a triplicated repeat within its sequence (9). However, the functional aspects of these repeats have not been defined. In the current report, we have re-examined the sequence of RAP using several sequence analysis programs. Although no clear boundaries can be defined among the three repeats, we found that the homologous regions are retained within each boundary if RAP is divided into three approximately equally sized regions (i.e. repeat 1, amino acids 1-100; repeat 2, amino acids 101-200; and repeat 3, amino acids 201-323).
Several studies have suggested that there are multiple RAP-binding sites on each LRP molecule. However, the exact number of RAP-binding sites on LRP is unknown. Using cell surface binding analyses, we have previously shown that there were approximately 5-7 times more RAP-binding sites on hepatocytes when compared with t-PA binding sites (i.e. functional LRP molecules, see ref. 22). However, other reported binding studies of RAP to purified LRP concluded that two RAP-binding sites are present on each LRP molecule (27). Using the soluble LRP minireceptors, we have recently reported that each of the three putative ligand-binding domains of LRP is capable of interacting with RAP (15). In the current report, we have further analyzed the RAP-binding sites on LRP using domain and subdomain constructs of LRP. We found that there were at least two RAP-binding sites on the second and fourth, and one on the third ligand-binding domains of LRP, yielding a total of at least five RAP-binding sites on the receptor. It should be pointed out that, given the fact that each half of the second and fourth ligand-binding domains of LRP is capable of binding RAP, these repeats may contain more than two RAP binding sites. The exact number of RAP-binding sites, as well as the regions where RAP binds, will require further investigation.
The physiological function of RAP has become more clear during the past 2 years. The fact that RAP is localized primarily within the ER with little or no secretion suggests an intracellular role for this protein. It appears that RAP associates with LRP during or immediately after the biosynthesis of the receptor (9). Association of RAP with LRP would ensure that LRP remains in an inactive ligand-binding state during its trafficking along the early secretory pathway. This appears to be important for receptor trafficking as LRP ligands are secreted proteins and often expressed in the same cells as the receptor. By associating with the receptor, RAP prevents premature interaction of ligands with the receptor (9, 10). This function of RAP resembles that of the invariant chain in preventing premature binding of peptides with the MHC class II molecules during their trafficking along the secretory pathway (28). Interestingly, using the soluble minireceptor system, our recent studies also suggest that RAP is required for the folding process of LRP by preventing the formation of intermolecular disulfide bonds (15). It is not clear at present whether the function of RAP in the folding of the receptor and in preventing premature ligand interaction with the receptor are related. For example, if a ligand interacts with LRP before the receptor completes its folding, the folding process might be impaired. In this case, preventing ligand interaction with the receptor ensures the proper folding of LRP. On the other hand, it is possible that RAP is independently involved in the folding, but remains associated with the receptor after the folding. Thereafter, premature ligand interaction with the receptor may be prevented. The precise role of RAP as a chaperone for LRP, as well as the mechanisms involved, await further definition.
The differential functions of the RAP repeats may also explain their
roles and appearance during evolution. Since repeat 3 of RAP is capable
of assisting the folding of each of the ligand-binding domains of LRP
and of inhibiting most of the ligand interactions with the receptor, it
is likely that this repeat is essential for RAP's function and may be
the ancestral region for the whole molecule; consistent with the
indications from sequence alignment. However, the fact that interaction
of some of the LRP ligands (e.g.
2-macroglobulin, shown in this study) is not inhibited by repeat 3 of RAP suggests the need for the first two repeats, which
do inhibit these ligand interactions with the receptor. It is
interesting to note that LRP is present in an organism as primitive as
the nematode Caenorhabditis elegans (29). Examination of the
C. elegans gene bank also identifies a 290-amino acid
protein sequence (gene accession no. Z75527[GenBank]) that shares high sequence
homology with human RAP, particularly in the portion where repeat 2 and
repeat 3 of human RAP share most homology. However, no ER-retention
signal is present at the carboxyl terminus of this sequence. Whether
this protein is the nematode equivalent of RAP requires functional
studies. It will be interesting in future studies to examine the
appearance of LRP ligands during evolution. For example,
2-macroglobulin has been described in the horseshoe crab
(30). If most LRP ligands are absent in the nematode, the primary
function of RAP may be to aid in receptor folding. The role of RAP in
inhibiting ligand interaction with the receptor may have evolved only
after LRP expression became high in tissues (e.g. liver and
brain) of higher organisms and with the appearance of the diverse array
of LRP ligands.
We thank Dr. Joachim Herz for providing cDNA for human LRP, Dr. Mark Heiny for assistance in the analysis of RAP sequences, and Dr. Alan Schwartz for general support.