| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Division of Tumor Biology, The Netherlands Cancer Institute, Amsterdam, The Netherlands; and
Department of Molecular Immunology, German Cancer Research Center, Heidelberg, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Peptides derived from cellular or viral proteins, predominantly generated by the proteasome, are transported by TAP from the cytoplasm into the endoplasmic reticulum (ER). In the ER lumen, a subset of these peptides is loaded onto MHC class I molecules, which are then sorted as stable complexes to the cell surface for T cell surveillance (for review, see Refs. 6, 7). The ER-resident chaperone tapasin forms a bridge between TAP and class I molecules, and thereby may enhance the rate and/or efficiency of peptide loading (8, 9). TAP can therefore be considered a scaffold for assembly of MHC class I molecules.
TAP has the capacity to transport peptides with a wide range of length and sequence, although the optimal substrate size is between 8 and 15 amino acids (10, 11). The pore should have considerable flexibility, because peptides with very long side chains can be translocated by TAP (12). Allelic differences in rat TAP are reflected in an altered peptide transport profile (13, 14, 15). Also, a clear preference for hydrophobic C-terminal peptides has been demonstrated for mouse TAP (14). These observations indicate that TAP may restrict the repertoire of antigenic peptides ultimately presented by class I molecules.
Peptide translocation has been divided into two steps: ATP-independent binding of peptides at the cytoplasmic side of membranes and ATP-dependent transport (16). A viral inhibitor of the initial peptide-binding step has been identified in the HSV ICP47 protein, which functions as a cytoplasmic high-affinity competitor (17, 18). In contrast, the human cytomegalovirus-encoded glycoprotein US6 prevents the actual translocation event by binding to TAP at the lumenal side of the ER (19, 20). The existence of various viral proteins, which abrogate the activity of TAP, underscores its important role in Ag presentation.
It is apparent that TAP is involved in numerous protein-protein interactions inside the ER, as well as in peptide binding and transport powered by ATP hydrolysis. To get a better understanding of the different features of TAP, we designed a series of deletion constructs to determine its topology and to elucidate the structure-function relationship. Based on the hydrophobicity profile of the TAP subunits, we cloned a set of C-terminal deletions fused to a reporter cassette. This cassette consists of a vesicular stomatitis virus (VSV) epitope for detection purposes and two N-linked glycosylation consensus sites, which allows the determination of the orientation of the C terminus with respect to the ER membrane. Our results are incorporated into a model of TAP in which a putative translocation pore consists of the N-terminal 6 and 5 TMs of TAP1 and TAP2, respectively. Various segments within the pore-forming sequence seem to determine ER retention. Downstream of that pore a peptide-binding domain can be distinguished, which is mainly cytoplasmic, but is linked to the membrane via a TM doublet in each TAP subunit. Finally, a C-terminal ATP-binding domain concludes the peptide transporter subunits.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
COS-7 cells were cultured in DMEM supplemented with 7.5% FCS and transfected using the DEAE-dextran method. Abs used are P5D4-peroxidase (Boehringer Mannheim, Mannheim, Germany), anti-TAP2 polyclonal raised against GST fusion protein containing aa 434–703 of human TAP2 and monoclonal TAP2.70 (21) and anti-calreticulin (anti-CRT; Affinity Bioreagents, Golden, CO).
DNA constructs
pMT2IiVSV contains two consensus N-linked
glycosylation sequences derived from the human invariant chain (Ii).
For this, a PCR using as template pRc/CMV
(22) yielded
a fragment containing aa 90–126 of Ii, flanked by a 5' SalI
site and a 3' XhoI site, which was cloned into the
SalI site of pMT2SM-tag. The consensus glycosylation sites
are at aa 113/115 and 119/121.
hTAP1A and hTAP2E cDNAs (23) were cloned into pBluescriptII KS(+) as XbaI fragment (into XbaI site resulting in pFM366.1) and Eco47III-NotI fragment (using HincII and NotI sites yielding pFM368.1), respectively. Standard PCR was performed with these templates to generate TAP deletions. The 5' primer T3 was combined with truncation-specific 3' primers containing a SalI restriction site, encoding Val and Asp, to clone into the SalI site of pMT2IiVSV, either as SalI fragments (TAP1) or XhoI/SalI fragments (TAP2). The C-terminal amino acid of TAP is used to name the truncation constructs. Similarly, a full-length VSV-tagged TAP1 derivative was cloned into pMT2IiVSV using as 3' primer gacacgtcgacttctggagcatctgcaggagc to generate pTAP1-FLVSV; pTAP2-IS-E243: 5' primer: atccgtcgacagccATGgagactaagacaggggagctgaac with 3' primer Tap2-V473 and pTAP1-IS-D297: 5' primer: atccgtcgacagccATGgattctctgagtgagaatctgagc with 3' primer Tap1-K423; and PCR fragments cloned as SalI fragments into SalI site pMT2IiVSV. Restriction sites are underlined and the ATG start codon is indicated with capital letters.
A PstI fragment coding for aa 20–73 (gene-internal
PstI site) of human Ii was generated with 5' primer
gcgctgcagaccATGcctggggccccggagagcaag by PCR on pRc/CMV
and cloned into the PstI site of pMT2IiVSV to generate
pPstIiVSV. A PCR using 5' primer
ggcctcgaggcgctgccccgcatattctccctg plus 3' primer
TAP1-H257 on pFM366.1 yielded a XhoI/SalI
fragment which was cloned into the SalI site of pPstIiVSV to
generate pIi-TAP1-H257.
The BamHI-HindIII fragment (blunted with Klenow/dNTPs) of pCMUIV-CD8 (24) was cloned into the PstI site (blunted with T4 DNA polymerase/dNTPs) of pMT2IiVSV to generate pCD8BH. A PCR with 5' primer caggtcgacggctgggagggctgtgggggctgc plus 3' TAP2-R210-primer on template pFM368.1 yielded a SalI fragment, which replaced the SalI fragment of pCD8BH to generate pCD8-TAP2-R210.
Immunoprecipitation
COS cells were transfected and after 2 days, 2 µM lactacystin was added 4 h before lysis in buffer with 1% digitonin. Lysates were precleared twice with normal rabbit serum and divided into a normal rabbit serum control and an anti-TAP2 polyclonal serum for immunoprecipitation using protein A-Sepharose. Proteins were analyzed by 12% SDS-PAGE.
Extract preparation/endoglycosidase H/N-glycosidase F
Cells were harvested in PBS, and cell pellets were lysed in 50 mM Tris (pH 7.5), 1 mM EDTA, 150 mM NaCl, and 1% Nonidet P-40 (TEN/NP-40) on ice for 30 min. Glycanase (Boehringer Mannheim) treatments were given for 3 h at 37°C in the lysis buffer supplemented with 0.2% SDS and 1 mM PMSF. N-glycosidase F (10 mU/µl) and endoglycosidase H were given (50 µU/µl).
Membrane preparation
Cells were harvested in 1 ml of PBS and disrupted by the European Molecular Biology Laboratory (Heidelberg, Germany) cell cracker. After removal of debris by centrifugation for 5 min at 2500 rpm, membranes were pelleted for 30 min at 14000 rpm (Sigma 1K15 table centrifuge; Sigma, St. Louis, MO). Membranes were resuspended in 500 µl of 100 mM Na2CO3 (pH 11.5) and incubated on ice for 30 min. Extracted membranes were pelleted as before. Before loading, the carbonate supernatants were diluted twice in TEN/NP-40. The remaining membrane sample was taken up in 1 ml of TEN/NP-40.
Immunofluorescence
COS cells were fixed with 4% formaldehyde in PBS, 24 h after transfection, and permeabilized with 0.2% Triton X-100. After blocking in 10% FCS, coverslips were incubated with P5D4 and anti-CRT Abs. Secondary Abs were goat anti-mouse IgG-FITC and goat anti-rabbit IgG-Texas Red (Life Technologies, Rockville, MD). Images were obtained using a 600 MRC confocal microscope (Bio-Rad, Hercules, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine the membrane topology of human TAP, we generated by PCR technology a set of truncation mutants with a C-terminal reporter cassette. This cassette has a dual purpose: 1) detection of the proteins by virtue of a VSV-derived Ab epitope and 2) the introduction of two N-linked glycosylation consensus sites. The topology of the C terminus is deduced from the glycosylation status of the reporter, because the addition of any carbohydrates requires accessibility to ER lumenal enzymes.
The Kyte-Doolittle hydrophobicity plots of human TAP1 and TAP2 suggest
a number of potential membrane-spanning stretches (1, 2).
These regions are depicted as boxes and were used as guides for the
introduction of C-terminal truncations (Fig. 1
). PCR products of the chosen deletions
were cloned in a mammalian expression vector which contains the
reporter cassette. COS cells were transiently transfected, and the
expressed proteins were analyzed by SDS-PAGE and Western blotting using
the VSV epitope. An aliquot of the samples was treated with glycanase
to remove the N-linked sugars. A change in mobility upon
digestion is indicative of glycosylation of the reporter cassette and
its presence in the lumen of the ER. Absence of glycosylation, in
contrast, indicates a cytoplasmic localization of the C terminus. In
general, we observed that short expression times (24 h) and low
expression levels resulted in a more uniform and consistent
glycosylation of the various deletion constructs. The results for TAP1
and TAP2 are shown in Figs. 2 and
3, respectively, and the interpretation
is depicted in the model presented in Fig. 4.
|
|
|
TAP1.
The shortest construct, 1T38 (containing the sequence up to the Thr
residue at position 38), shows a mixture of glycosylated and
nonglycosylated products with a preference for the glycosylated form
(Fig. 2
, compare lanes 1 and 2). To establish
membrane integration, the membranes were extracted with sodium
carbonate (pH 11.5), which completely removes the soluble ER resident
protein disulfide isomerase from the membranes (data not shown). Fig. 5 shows that the polypeptide T38 is
inserted into the membrane. Thus, the glycosylation of 1T38 is not the
result of complete translocation into the ER lumen. Note that in this
experiment 1T38 is almost completely glycosylated. Apparently,
construct 1T38 functions as a type 2 signal/anchor sequence. This
conclusion is further validated by construct 1R56, which shows
extensive glycosylation. Construct 1R64 is also mainly glycosylated,
indicating the absence of a closely neighboring TM (open box in Fig. 1), as suggested by the hydrophobicity plot. In contrast, construct
1E85 shows a predominance of nonglycosylated products, an effect seen
even more pronounced with 1K97 (truncated within the next hydrophobic
segment, see Fig. 1). Apparently, the sequence between E85 and K97
contributes to proper membrane integration of TM2. We conclude that one
TM is passed between 1R56/1R64 and 1E85/1K97. The glycosylation
patterns of the next four constructs (1E115, 1K156, 1T212, and 1H257)
show an alternating pattern in accordance with the presence of four
additional TMs (lanes 11–18). The
construct 1H257 shows some glycosylated products due to the unstable
insertion of the C-terminal TM into the membrane, a unique feature for
this construct that will be described elsewhere (40). The closely
spaced constructs 1M320, 1M329, and 1K344 are in a region where one or
two potential TMs are located. The only partial glycosylation patterns
of both 1M320 and, to a greater extent, 1M329 suggest the possibility
that a pair of TMs is positioned within this part of TAP1, as
previously proposed by Gileadi and Higgins (25). The first
TM may not be efficiently inserted into the membrane in the absence of
the second. Further support for this hypothesis is presented below.
Finally, the constructs 1K423 and 1Q453 show no evidence for
glycosylation and thus indicate the absence of further TMs in TAP1 and
confirm the cytoplasmic orientation of the C terminus.
|
, lanes 1 and 2) is
nonglycosylated and membrane integrated as concluded from carbonate
extraction studies (Fig. 5). Trypsin digestion experiments show that
the C terminus is in the cytoplasm (data not shown). Therefore, TM1 of
TAP2 itself is a type 1 signal/anchor sequence directing the N terminus
into the lumen of the ER in contrast to TAP1. The next construct, 2R80,
is predominantly glycosylated, as concluded from the shift in m.w. upon
glycanase treatment (compare lanes 3 and 4).
Thus, the sequence between residues 46 and 80 contains a TM region, as
also suggested by the hydrophobicity profile. Three other TMs are
apparent from the glycosylation status of the three constructs shown in
lanes 4-10: 2E128 and 2R210 are mostly
nonglycosylated, whereas 2R175 is mainly glycosylated. The other TAP2
constructs show no or little glycosylation (lanes
11–20) with the exception of 2L292, which has a low, but
consistent level of added sugars. Therefore, we consider it possible
that, like for TAP1, this is an indication for a TM doublet (see also
below). In conclusion, the TAP2 subunit contains at least five, and
possibly seven, TMs with the N terminus being located in the ER lumen
and the C terminus in the cytoplasm.
|
The results presented in Figs. 2 and 3 suggest the possibility of
a pair of TMs with a very small lumenal loop downstream of TM6 of TAP1
and TM5 of TAP2 (see Fig. 4). To obtain independent evidence for the
existence of such TMs, we constructed N-terminal deletions of TAP1 and
TAP2, where the first six or five TMs are deleted, respectively. After
transient expression in COS cells, we investigated whether these
proteins become stably inserted into the membrane using sodium
carbonate extractions. Proteins TAP1-IS-D297 and TAP2-IS-E243 (Fig. 5B) are predominantly retained in the membrane fraction,
although a minor fraction of TAP2-IS-E243 is still found in the
cytosol. Possibly, the absence of a natural signal sequence reduces the
efficient targeting to and insertion in the ER membrane. Because the
protein TAP1-IS-D297 is membrane inserted and nonglycosylated, we
conclude that both its N and C terminus are cytoplasmic and thus the
protein contains TM7 and TM8 of TAP1 as depicted in Fig. 4. Similarly,
we conclude that TAP2-IS-E243 contains TM6 and TM7 of TAP2.
Dimerization of TAP1 and TAP2
TAP1 and TAP2 require heterodimerization for peptide binding and
translocation (10, 16). In COS cells, we expressed
full-length TAP2 along with TAP1 truncations or full-length TAP1
(around 71 kDa) containing a VSV epitope to define interacting domains.
Lysates (see Fig. 6A for
expression controls) were immunoprecipitated with TAP2-specific
polyclonal Abs and probed for the presence of TAP1 derivatives with
anti-VSV mAbs (Fig. 6B). Full-length TAP1 is found in
association with TAP2, as expected (lane 1).
Constructs 1K423 and 1K344 associate equally efficient with TAP2, as
well as a smaller C-terminal side product that still contains the VSV
tag and whose N terminus is estimated to be around TM5 based on m.w.
Construct 1H257 (containing TM1–6) is found to be relatively weakly
associated with TAP2, although its expression level is higher. Further
deletions, such as 1T212, do not show specific coimmunoprecipitation,
which points to the importance of the region encompassing TM6 of TAP1.
Thus, TAP1-H257 heterodimerizes with TAP2, possibly forming a
transmembrane pore domain. This does not exclude the existence of
additional dimerization domains between TM6 and the nucleotide-binding
domain of TAP. Indeed, TAP1-IS-D297, which starts downstream of TM6
(see Fig. 1), can be coisolated with TAP2, showing the existence of two
nonoverlapping dimerization domains within TAP1.
|
TAP molecules are situated in the ER membrane (26, 27). The sequence of both TAP1 and TAP2 does not contain any
known ER retention signal. Therefore, we analyzed by immunofluorescence
whether extensive C-terminal truncation results in a different
intracellular distribution or even plasma membrane staining. All our
constructs showed a typical reticular fluorescence pattern, indicating
ER localization. The results for the smallest constructs are shown in
Fig. 7. COS cells were transfected with
either the TAP1 construct 1T38 (Fig. 7A) or the TAP2
construct 2R45 (Fig. 7C), each truncated after the first TM.
The cells were fixed and stained with Abs against the VSV epitope and
against CRT serving as an endogenous luminal ER marker. For both TAP
derivatives, the staining showed a characteristic ER reticular staining
that closely matched the fluorescence of CRT. This suggests that any ER
retention signal is located in either the short cytoplasmic (in case of
TAP1) or the lumenal (TAP2) N-terminal tail or within the TMs
themselves. To investigate whether the N terminus is the sole
determinant for ER retention in the pore-forming domain of TAP1, we
exchanged the N terminus including TM1 of TAP1-H257 for the TM of
invariant (Ii) chain (a type II molecule) with an N-terminal deletion
of 20 amino acids of its cytoplasmic tail to ensure plasma membrane
expression (28). Likewise, the N terminus including TM1 of
TAP2-R210 was exchanged for the TM of the cell surface molecule CD8 (a
type I molecule) along with its extracellular segment
(24). As control, we tested the behavior of the Ii chain
TM coupled to the reporter cassette and detected cell surface
expression by immunofluorescence and FACS analysis (data not shown),
which excludes that the reporter cassette either contains a cryptic ER
retention sequence or is retained due to malfolding. The
immunofluorescence of chimeras Ii-TAP1-H257 (Fig. 7B) or
CD8-TAP2-R210 (Fig. 7D) is similar to that of the single TM
TAP constructs 1T38 and 2R45. Therefore, removal of the N-terminal
signal sequence/anchor residues of either TAP1 or TAP2 does not result
in cell surface expression. This suggests that, throughout the domain
that constitutes the putative pore domain, TAP1 and TAP2 contain
multiple redundant ER retention signals, including the first TM. The
presence of ER retention signals in the C-terminal nucleotide-binding
domains was not detected; fusion of CD8 with the ABC domains of TAP1 or
TAP2 did not prevent cell surface expression of CD8 (data not
shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), suggesting that
sequences downstream of T38 may be involved in properly orienting TM1.
Nevertheless, evaluation of the results leads to a model where the two
TAP subunits show a high degree of structural similarity. A correct
insertion of the various truncation constructs is further supported by
the dimerization and other subunit interactions within the pore-forming
domains (40).
The glycosylation patterns of the TAP1 and TAP2 deletion constructs as
presented in Figs. 2 and 3 were obtained using short expression times.
Under these conditions, we found the most uniform and reproducible
results. For both TAP1 and TAP2, we find that the first hydrophobic
segment can function as a signal anchor sequence, albeit with different
orientations. By contrast, the first TM of P-glycoprotein is not stably
inserted into the membrane, but requires the cooperative action of the
second TM (31). Because truncation constructs did not
conclusively prove the existence of the closely spaced TM7 and TM8 of
TAP1, as well as TM6 and TM7 of TAP2, we employed an alternative
approach. The proper integration of the internal start constructs
1IS-D297 and 2IS-E243 shows that internal TMs can have the potential to
function as signal/anchor. Furthermore, these pairs of TMs seem to rely
on each other for membrane integration, a feature also found for
multispanning integral membrane proteins like P-glycoprotein
(31) and Sec61p (29). Gileadi and Higgins
(25) also proposed the existence of a TM doublet at the
same position within the TAP1 molecule, although the orientation in the
membrane of Escherichia coli is opposite. A natural
glycosylation consensus sequence is present within TAP1 at residues
279/281. In our model, these residues are within a cytoplasmic loop
consistent with a lack of glycosylation of full-length TAP1
(27).
The translocation pore of the ABC transporter is generally considered to be composed of 12 TMs (5). Our coprecipitation experiments show the association of TAP1-H257, a TAP1 derivative containing the six N-terminal TMs, with full-length TAP2. Further C-terminal truncations of TAP1 do not result in a complex, which is stable in lysis buffer containing 1% digitonin. We conclude that TM6 of TAP1 is essential for dimer interactions. By comparing the overall structure of TAP1 with TAP2, we suggest that TM1–5 of TAP2 are functionally equivalent to TM1–6 of TAP1. Further independent evidence that TM5 of TAP2 and TM6 of TAP1 mark the C-terminal ends of the pore-forming domains has been obtained by studying subunit interactions using alternative approaches (40).
Genetic analysis and photocrosslinking with peptide ligands have
implicated sequences downstream of TM6 of TAP1 and TM5 of TAP2 in
mediating substrate specificity and substrate binding by the
transporter. Allelic forms of rat TAP2 show marked differences in
transport specificity with regard to the C-terminal amino acid of
peptide substrates (13, 14, 32). The functional and
structural equivalence of rat and human TAP is large, since
heterologous subunits can form an active complex (33). The
functional analysis of chimeric rTAP2a and
rTAP2u revealed that three independent clusters
of polymorphic amino acids (TAP2 residues 217/218; 262, 265, 268; 374,
380) are involved in substrate specificity (15, 34). It is
intriguing that the residues 217/218 follow TM5, the border of the
putative pore domain, by only eight amino acids. Furthermore, a single
point mutation in human TAP2 at residue 374 modifies transport
specificity (34). Cross-linking studies using
photoreactive peptides delineated hTAP1 sequences encompassing residues
375–487 and hTAP2 residues 301–433, respectively, to be involved in
the ATP-independent binding of peptides to the transporter (21, 35). Therefore, we propose discrete peptide-binding domains in
the TAP subunits starting with the cytoplasmic loop C-terminal of TM6
in TAP1 and TM5 in TAP2 and reaching beyond the hydrophobic stretches
in the center of the molecules (pairs of open boxes in Fig. 1) that are
apparently not membrane integrated. Importantly, heterodimerization of
TAP1 and TAP2 is a prerequisite for the formation of a specific
peptide-binding unit (10, 16, 21). The substrate-binding
segments are positioned between the putative pore domains, which also
dimerize, and the nucleotide-binding domains. We have not been able to
show stable interaction between the two nucleotide-binding domains.
Based on the above considerations, we propose a three-domain structure
of TAP1 and TAP2 as shown schematically in Fig. 4. In a previous study,
the membrane topology of human TAP1 was determined by using C-terminal
truncations linked to ß-lactamase as reporter protein and expressing
the fusion constructs in E. coli (25). With the
exception of the cytoplasmic orientation of both the N terminus and the
nucleotide-binding domain, the topological details of that study are,
however, at complete variance with the results shown here. The simplest
explanation for this discrepancy is that E. coli membranes
do not allow for a proper integration of the TMs of TAP1. Because
peptide binding to TAP is energy independent (16), it
seems very unlikely that all structures involved in substrate
recognition are located inside the ER as proposed by Gileadi and
Higgins (25). The similar organization of TAP1 and TAP2
proteins in the ER membrane of mammalian cells as characterized in this
study is fully consistent with an ATP-independent binding step
involving exposed cytoplasmic loops. This binding would precede the
actual substrate translocation through a hydrophobic or amphipathic
pore crossing the lipid bilayer.
We show that the pore of both TAP1 and TAP2 contain multiple ER retention signals. Two types of ER retention/retrieval signals have been defined for transmembrane proteins. In a coatamer-dependent manner, a cytoplasmic di-lysine or di-arginine signal at either the C or N terminus, respectively, results in retrieval from the Golgi to the ER (24, 36, 37). Furthermore, transmembrane segments have been found to retain proteins in the ER (38, 39). We demonstrate here that the first TMs of TAP1 and TAP2, type II and type I, respectively, including their short N-terminal sequences, are sufficient for ER localization. Because these sequences contain no known motifs for retention and have opposite orientations, it is most likely that the TMs themselves determine ER retention. A C-terminal, cytoplasmic retention signal in the ABC domains is absent. The specific amino acid motifs and the mechanism underlying a TM-mediated retention remain to be established.
The topology model of the TAP subunits presented in Fig. 4 provides a
framework for further studies. The model indicates a high degree of
similarity in the secondary structure of TAP1 and TAP2. In comparison,
TAP2 only lacks one N-terminal TM, resulting in an opposite orientation
of the N terminus. Three homologous domains are found with discrete
functions. The model is an important step toward understanding the
dynamics of peptide translocation, the role of TAP as a scaffold in
peptide loading of class I molecules, and the molecular basis for TAP
inhibition by viral proteins like ICP47 and US6.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jacques Neefjes, Division of Tumor Biology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands. E-mail address:
3 Abbreviations used in this paper: CFTR, cystic fibrosis transmembrane conductance regulator; TM, transmembrane; ER, endoplasmic reticulum; VSV, vesicular stomatitis virus; CRT, calreticulin; NP-40, Nonidet NP-40.
Received for publication July 26, 1999. Accepted for publication October 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Keusekotten, R. M. Leonhardt, S. Ehses, and M. R. Knittler Biogenesis of Functional Antigenic Peptide Transporter TAP Requires Assembly of Pre-existing TAP1 with Newly Synthesized TAP2 J. Biol. Chem., June 30, 2006; 281(26): 17545 - 17551. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Schrodt, J. Koch, and R. Tampe Membrane Topology of the Transporter Associated with Antigen Processing (TAP1) within an Assembled Functional Peptide-loading Complex J. Biol. Chem., March 10, 2006; 281(10): 6455 - 6462. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Halenius, F. Momburg, H. Reinhard, D. Bauer, M. Lobigs, and H. Hengel Physical and Functional Interactions of the Cytomegalovirus US6 Glycoprotein with the Transporter Associated with Antigen Processing J. Biol. Chem., March 3, 2006; 281(9): 5383 - 5390. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Koch, R. Guntrum, S. Heintke, C. Kyritsis, and R. Tampe Functional Dissection of the Transmembrane Domains of the Transporter Associated with Antigen Processing (TAP) J. Biol. Chem., March 12, 2004; 279(11): 10142 - 10147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Ritz, I. Drexler, D. Sutter, R. Abele, C. Huber, and B. Seliger Impaired Transporter Associated with Antigen Processing (TAP) Function Attributable to a Single Amino Acid Alteration in the Peptide TAP Subunit TAP1 J. Immunol., January 15, 2003; 170(2): 941 - 946. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Raghuraman, P. E. Lapinski, and M. Raghavan Tapasin Interacts with the Membrane-spanning Domains of Both TAP Subunits and Enhances the Structural Stability of TAP1{middle dot}TAP2 Complexes J. Biol. Chem., October 25, 2002; 277(44): 41786 - 41794. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Lankat-Buttgereit and R. Tampe The Transporter Associated With Antigen Processing: Function and Implications in Human Diseases Physiol Rev, January 1, 2002; 82(1): 187 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Velarde, R. C. Ford, M. F. Rosenberg, and S. J. Powis Three-dimensional Structure of Transporter Associated with Antigen Processing (TAP) Obtained by Single Particle Image Analysis J. Biol. Chem., November 30, 2001; 276(49): 46054 - 46063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Ritz, F. Momburg, H.-P. Pircher, D. Strand, C. Huber, and B. Seliger Identification of sequences in the human peptide transporter subunit TAP1 required for transporter associated with antigen processing (TAP) function Int. Immunol., January 1, 2001; 13(1): 31 - 41. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
