| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institute of Immunology, Vienna International Research Cooperation Center at Novartis Forschungsinstitut and
Institute of Immunology, University of Vienna, Vienna, Austria; and
Novartis Research Institute, Vienna, Austria
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
In this study, we report the molecular cloning of the CDw92 molecule by using mAbs VIM15 and VIM15b and a retroviral vector-based expression cloning approach, which has been recently established in our laboratory. Furthermore, we show that the CDw92 protein is specifically regulated upon maturation and differentiation of dendritic cells (DCs),4 and we provide evidence for a role of CDw92 in regulation of expression of IL-10 by DCs.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Both CDw92 mAbs VIM15 (isotype IgG1) and VIM15b (isotype IgG2b)
were established at our institute by immunization of BALB/c mice with
cells of the myeloid cell line MV4-11 (1). The CD4 mAb
VIT4, the CD8 mAb VIT8, the CD14 mAb VIM13, the CD15 mAb VIM16, the
anti-MHC class II mAb VID1, as well as the negative control mAb
VIAP (to calf intestine alkaline phosphatase) were also generated by
us. The CD14 mAb MEM-18 was purchased from An der Grub/Scandic (Vienna,
Austria). The hybridomas producing the CD16 mAb 3G8 and the CD19 mAb
BU-12, respectively, were obtained from the American Type Culture
Collection (Manassas, VA). The negative control mAb AFP-01 directed to
human 
-fetoprotein was a kind gift of V. Horejsi (Institute of
Molecular Genetics, Academy of Sciences of the Czech Republic, Prague,
Czech Republic).
For the functional assays, the mAbs were used at a concentration of 10 µg/ml. The working dilutions of the mAb preparations contained less than 10 pg/ml LPS. No biological effects of these very low LPS concentrations were detectable in our assay systems.
Cell lines and cell preparations
Cell lines, including human myeloid KG1a cells, Sp2/6 mouse myeloma cells, P815 mouse mastocytoma cells, EL-4 mouse lymphoma cells, CPII mouse mast cells, and BW5147 mouse thymoma cells were cultured in RPMI 1640 medium supplemented with 10% FCS in a fully humidified atmosphere of 5% CO2 at 37°C. Phoenix packaging cells were maintained in DMEM containing 10% FCS. Endothelial cells were cultured in DMEM containing 20% FCS, 1% basic fibroblast growth factor, and 10 µg/ml heparin.
PBMCs were isolated from heparinized whole blood of normal healthy
donors by standard density gradient centrifugation using Ficoll-Paque
(Pharmacia Biotech, Uppsala, Sweden). To examine CDw92 surface
expression on activated PBLs, PBMCs were seeded in 96-well cell culture
plates (105 cells/well) and cultured for 2 days
with and without ionomycin (1 µM; Sigma-Aldrich, St. Louis, MO) or
PHA (5 µg/ml; Abbott Laboratories, Vienna, Austria) in RPMI 1640
medium supplemented with 5% human AB serum (PAA Laboratories, Munich,
Germany). Monocytes were separated from freshly isolated PBMC
preparations by magnetic sorting using biotinylated CD14 mAb VIM13 in
conjunction with the MACS magnetic bead technology of Miltenyi Biotec
(Bergisch Gladbach, Germany). Monocyte-derived DCs (Mo-DCs) were
generated by culturing purified blood monocytes for 7 days in RPMI 1640
medium supplemented with 10% FCS and a combination of recombinant
human GM-CSF and recombinant human IL-4 (50 ng/ml and 100 U/ml,
respectively; both kindly provided by the Novartis Research Institute,
Vienna, Austria). For some experiments, immature Mo-DCs as well as
peripheral blood monocytes were further treated for 1 day with IFN-
(100 U/ml; kindly provided by Dr. G. Adolf, Boehringer
Ingelheim, Vienna, Austria), IFN-
(300 U/ml; Boehringer
Ingelheim), LPS (1 µg/ml, Escherichia coli serotype
0127-B8; Sigma-Aldrich), ionomycin (1 µM; Sigma-Aldrich),
or calcium ionophore (100 ng/ml; Sigma-Aldrich). For IL-10 (10 ng/ml;
R&D Systems, Minneapolis, MN) stimulation, ionomycin- or calcium
ionophore-treated Mo-DCs and peripheral blood monocytes were cultured
for 1 day in the presence of this IL.
Immunofluorescence analysis
The direct and indirect immunofluorescence techniques were used to determine mAb binding. For indirect immunofluorescence, either Oregon Green-labeled goat (Molecular Probes, Eugene, OR) or FITC-labeled sheep (An der Grub/Scandic) F(ab')2 anti-mouse IgG and IgM (H plus L) Abs were used as secondary reagents. The cell staining was assessed by flow cytometry and fluorescence microscopy.
Retroviral cDNA library construction
Total RNA was extracted from human myeloid KG1a cells using the
guanidinium thiocyanate method as described by Chomczynski and Sacchi
(3). Poly(A)+ mRNA was isolated
using an Oligotex mRNA minikit (Qiagen, Chatsworth, CA) according to
the manufacturer’s instructions. Double-stranded cDNA was synthesized
from 5 µg of poly(A)+ mRNA by using a Pharmacia
Biotech cDNA synthesis kit. 5'-Phosphorylated adapters were ligated to
the double-stranded cDNA. After removal of excess adapters by
chromatography, the cDNA library was ligated into the retroviral
expression vector pBabeMN-lacZ, kindly provided by G. Nolan (Stanford
University, Stanford, CA). Then, the cDNA library was used to transform
E. coli DH5, which were plated on Luria Bertani agar
plates containing 100 µg/ml ampicillin (Sigma-Aldrich). Bacteria were
harvested from the plates by scraping in Luria Bertani medium, and the
plasmid cDNA was extracted by using a Maxi Plasmid kit (Qiagen). This
retroviral library was characterized by 
2 x
106 independent bacterial clones and cDNA inserts
ranging from 0.8 to 4 kb.
Transfection and infection procedures
One day before transfection, 3 x 106 cells of the ecotropic retroviral packaging cell line Phoenix (kindly provided by G. Nolan) were seeded in 6-cm tissue culture dishes (Nunc, Roskilde, Denmark). Cells were grown overnight and were transfected with Superfect transfection reagent (Qiagen) using 5 µg of library plasmid cDNA per dish according to the manufacturer’s instructions. After 2 days, the culture supernatant containing secreted viruses representing the library was filtered and polybrene (Sigma-Aldrich) was added to a concentration of 10 µg/ml. Five hundred microliters of this viral supernatant was used to infect 5 x 105 mouse target cells (including Sp2/6 mouse myeloma, P815 mouse mastocytoma, EL-4 mouse lymphoma, and BW5147 mouse thymoma cells) in 500 µl of RPMI 1640 medium and 10% FCS in 6-well culture dishes (Nunc). After cultivation for 6 h at 32°C in a 5% CO2 atmosphere, 2 ml of fresh RPMI 1640 medium containing 10% FCS was added and the cells were grown overnight at 37°C. Afterward, the infected cells were washed and grown in fresh medium for another 2 days at 37°C.
Screening procedure
Infected mouse target cells (2 x 107) were washed twice with PBS/1% BSA and were incubated with the CDw92 mAbs VIM15 and VIM15b (each 10 µg/ml) for 20 min on ice. Then, the cells were incubated with MACS goat anti-mouse IgG microbeads (Miltenyi Biotec) according to the manufacturer’s instructions. After washing with PBS/1% BSA, cells were resuspended in 500 µl of MACS-sorting buffer (PBS/0.5% BSA/2 mM EDTA) and were loaded onto MACS RS+ selection columns (Miltenyi Biotec) for positive selection of CDw92-transduced cells. Selected cells were maintained in RPMI 1640 medium and 10% FCS and were expanded to a cell number of 2 x 107 and again subjected to MACS sorting. After four cycles of sorting, 60–90% of the cells (depending on the mouse cell line used) reacted with the CDw92 mAbs VIM15 and VIM15b. One positive single-cell clone was isolated by limiting dilution and was used to recover the CDw92 cDNA.
RT-PCR recovery of the CDw92 cDNA
The CDw92 single-cell clone was expanded and total RNA was extracted. The CDw92 cDNA insert was rescued using RT-PCR (Superscript II RT; Life Technologies, Rockville, MD) and the Advantage-GC polymerase system (Clontech Laboratories, Palo Alto, CA) and primers flanking the multiple cloning site of the retroviral vector pBabeMN-lacZ. The PCR was run for 32 cycles (1 min at 94°C, 1 min at 58°C, and 6 min at 68°C) on a DNA Thermal Cycler (PerkinElmer/Cetus, Norwalk, CT). The amplified 3-kb PCR fragment was gel purified and subcloned into pBabeMN-lacZ. The resulting vector (termed CDw92A6) was transfected into Phoenix packaging cells using the DEAE-Dextran transfection method described before (4), with some modifications. Phoenix packaging cells (1.5 x 106) were suspended in 4 ml of DMEM containing 1% NuSerum (Genome Therapeutics, Waltham, MA), 200 µg/ml DEAE-Dextran, 100 µM chloroquine diphosphate, and 4 µg of CDw92A6 plasmid DNA. After 2 h of incubation at 37°C, the Phoenix cells were washed with DMEM containing 10% FCS and were transferred into 6-well culture plates (Nunc). After overnight cultivation at 37°C, the medium was exchanged and 1 ml of fresh DMEM containing 10% FCS was added. The cells were then cultured for 2 days and the culture supernatant was used to infect mouse target cells. Reaction of both CDw92 mAbs with these cells was examined at day 3 postinfection by immunofluorescence analysis.
Sequence analysis
Sequencing of the cDNA insert of the CDw92A6 plasmid was performed by using an automatic sequencing facility based on a 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA). The putative transmembrane domains (TMDs) of CDw92 and the sequence topography were estimated using TopPred II 1.3 (5). This program was used with the settings for eukaryotic proteins and the Kyte-Doolittle (6) hydrophobicity scale with an upper cutoff of 1.5 for certain TMDs and a lower cutoff of 1.1 for putative TMDs. Also, the Goldman, Engelman, and Steitz (7) hydrophobicity scale with the default setting was applied. In addition, we analyzed the sequence topology by the PHDTopography system (8) (http://www.EMBL-Heidelberg.DE/predictprotein).
Surface labeling of cells and immunoprecipitation analysis
Cells were surface biotinylated using 0.5 mg/ml sulfo-NHS-LC-biotin (Pierce, Rockford, IL) in PBS (pH 8.5). After 50 min of incubation at 4°C, the reaction was stopped by adding 20 mM Tris-HCL (pH 8.2). Cells (1 x 107/ml) were lysed for 40 min at 4°C in lysis buffer (20 mM Tris-HCL (pH 8.2), 140 mM NaCl) containing 1% NP40 (Pierce) and 0.2% deoxycholic acid (Sigma-Aldrich) as detergents and a mixture of protease inhibitors (1 mM PMSF, 10 µg/ml aprotinin, and 10 µg/ml leupeptin). The cell lysates were centrifuged at 12,000 rpm for 20 min at 4°C and afterward were precleared by using an irrelevant mAb coupled to CNBr-Sepharose for 1 h. Then, the lysates were subjected to immunoprecipitation using the Solid Phase Immunoisolation Technique as described (9). Immunoprecipitates were separated by SDS-PAGE using 10% gels under nonreducing conditions followed by transfer to Immobilon P polyvinylidene difluoride membranes (Millipore, Bedford, MA). Membranes were blocked with 5% BSA in TBS (pH 7.9), and biotinylated proteins were visualized on the blot using a streptavidin-peroxidase conjugate (Amersham, Aylesbury, U.K.) and the chemiluminescence detection system from Roche Molecular Biochemicals (Mannheim, Germany).
Determination of cytokine production
Mo-DCs (1 x 106/ml) were cultured
with or without LPS (1 µg/ml; Sigma-Aldrich) in the presence or
absence of the CDw92 mAbs VIM15 or VIM15b, the anti-MHC-class II
mAb VID1, or the control mAb VIAP (to calf intestine alkaline
phosphatase) in 24-well plates (Costar, Cambridge, MA). After 48
h, the supernatants were harvested and cytokines were measured by a
sandwich ELISA using matched-pair Abs. Capture as well as detection Abs
for human IL-1
were obtained from Genzyme (Cambridge, MA), for IL-10
and IL-12-p40 from R&D Systems, and for TNF- from BD PharMingen (San
Diego, CA). As standards, we used human recombinant cytokines purchased
from R&D Systems. Assays were performed in duplicates according to the
recommendations of the manufacturers. The lower limit of detection was
10 pg/ml for IL-1 and 20 pg/ml for IL-10, IL-12-p40, and
TNF-.
Measurement of choline incorporation
Cells (1 x 106) were incubated in 1 ml of RPMI 1640 medium/10% FCS containing 0.5 µCi of 14C-choline chloride (Amersham) for 10, 30, or 60 min at 37°C. Afterward, the cells were washed twice with PBS. The samples were divided and one-tenth of the cells was used to determine the protein content in a Bradford assay (Pierce) for normalization. The other part of the cell preparation (nine-tenths) was pelleted and resuspended in 500 µl of PBS. After the addition of 5 ml of scintillation mixture, the samples were counted in a beta counter for 2 min.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The primary cellular reactivities of the CDw92 mAbs VIM15 and
VIM15b were described before (1). However, binding
analyses to individual subsets of lymphocytes were not performed in
detail. Thus, we examined the distribution of CDw92 on human peripheral
blood leukocyte subsets by double-staining immunofluorescence analysis
using markers to individual leukocyte subsets. As can be seen in Fig. 1
, CDw92 was most strongly expressed by
CD14+ monocytes. Granulocytes characterized by
CD15 were stained more weakly. To assess the distribution of VIM15
within PBL subsets, we used VIM15-PE in combination with FITC-labeled
markers for B cells (CD19), NK cells (CD16), and T cell subsets (CD4
and CD8), and gated PBLs according to their forward and side scatter
characteristics. Almost all B lymphocytes (97 ± 2%) were VIM15
positive. T lymphocytes were discriminated by VIM15 into a positive and
negative population; 80% of the CD4+ cells and
60 ± 9% of the CD8+ T lymphocytes were
VIM15 positive. Among the CD16+ cells, only a
small population (23 ± 9%) gave a weak score.
|
Isolation of the CDw92 cDNA
To study the structure and function of the CDw92 molecule, we isolated the CDw92 cDNA by using a retroviral cloning system that had been recently established in our laboratory. This system is based on the retroviral vector pBabeMN-lacZ and the packaging cell line Phoenix, both of which were developed by G. Nolan and colleagues (10). We constructed a pBabeMN-lacZ retroviral cDNA library from myeloid KG1a cells that strongly expressed the CDw92 molecule. We transiently transfected the library into virus-producing Phoenix cells and used the resulting supernatants (which contain virions representing the library) to infect different mouse target cell lines. Those transduced cells, which bound the CDw92 mAb VIM15, were sorted by using the MACS magnetic beads technology. Upon four MACS-sorting cycles, 60–90% (depending on the mouse cell line used) of cells scored positively with VIM15 mAb.
The MACS-sorted population was subjected to limiting dilution to
separate one single-cell clone that was strongly stained by the CDw92
mAbs. This cell clone was expanded and the RNA was isolated to amplify
the CDw92 cDNA insert by RT-PCR using primers flanking the multiple
cloning site of the retroviral vector. As can be seen in Fig. 2, a 3-kb DNA fragment was obtained. To
analyze whether this fragment represents the CDw92 cDNA, we subcloned
it back into pBabeMN-lacZ. The resulting plasmid was transfected into
Phoenix cells to produce retroviral supernatants for transduction of
BW5147 mouse cells. The reactivities of the CDw92 mAbs VIM15 and VIM15b
with the established transductant were analyzed by immunofluorescence
staining and flow cytometric analysis. A strong staining of both mAbs
was observed with the transduced cells but not with untransduced BW5147
mouse cells used as a control (Fig. 3).
This specific binding of both CDw92 mAbs to the transduced cells
indicated that the established plasmid, termed CDw92A6, contained the
CDw92 cDNA as insert.
|
|
To further prove that plasmid CDw92A6 encodes the CDw92 protein,
we compared by SDS-PAGE the recombinant protein encoded by CDw92A6 with
the native CDw92 protein on KG1a cells. Lysates of CDw92A6-transduced
and -untransduced BW5147 mouse cells as well as of human myeloid KG1a
cells were subjected to immunoprecipitation using CDw92 mAb VIM15b.
Precipitates were resolved under nonreducing SDS-PAGE conditions.
VIM15b immunoprecipitates derived from lysates of CDw92A6-transduced
BW5147 cells displayed a polypeptide band of 70 kDa that migrated as
the native CDw92 protein precipitated from myeloid KG1a cells. This
70-kDa band was not seen in lysates of untransduced BW5147 cells and in
immunoprecipitates obtained by using a control mAb (Fig. 4). Thus, this experiment confirmed that
we had cloned the cDNA encoding the CDw92 protein.
|
Sequencing of the cDNA insert of plasmid CDw92A6 revealed a
length of 2679 bp with an open reading frame of 1959 bp coding for 652
amino acid residues. This sequence has been deposited at the
EMBL/GenBank nucleotide sequence database (accession no. AJ272365). The
predicted molecular mass of the CDw92 polypeptide chain is 72.98 kDa.
This computational calculation in comparison with the molecular mass of
the native CDw92 molecule of 70 kDa determined by SDS-PAGE shows that
CDw92 is not or is only weakly glycosylated on one of the three
potential N-glycosylation sites found in the
sequence(Fig. 5).
|
and Table I).
Based on this prediction, 16 different membrane topologies are possible
containing either all four (Fig. 6, structure 1), some (Fig. 6, structures 2–15), or none (Fig. 6, structure 16) of the putative TMDs. By applying the
"positive-inside" rule, i.e., the observation that positively
charged amino acids are more abundant in cytoplasmic segments
(11, 12, 13), the best topology is predicted for structure 11,
which has the largest difference (Fig. 6; KR Diff = 20) of lysines
(Fig. 6, K) and arginines (Fig. 6, R) in cytoplasmic over exoplasmic
segments. Structure 11 contains 10 TMDs (TMDs 1–7, 9, 11, and 12) and
cytoplasmic positions for both the N and C termini (Fig. 6). When we
used PHDTopography (8) as second algorithm, a similar
10-transmembrane topology was predicted. According to this model, TMDs
1–4, 6–8, and 11 and 12 and an additional region between amino acids
178 and 195 are membrane integrated. However, the KR difference of this
structure is 12 (data not shown). In any case, each algorithm used
assigned a cytoplasmic position for both termini. This is of particular
interest because the C-terminal segment contains, at position 587–593,
the amino acid stretch SIYEMV, which closely resembles the consensus
sequence V/I/xYxxL/V of the immunoreceptor tyrosine-based inhibitory
motif (ITIM).
|
|
-chain or the rat mast-cell function-associated
Ag (MAFA). Occasionally, serines are also found at the critical
position three amino acids downstream of the tyrosine (e.g., in SLAM or
IL6-R -chain). Thus, Sinclair (15) proposed to redefine
the consensus sequence of ITIM by V/I/L/SxYxxL/V/I/S. This
consideration has been supported recently by the finding that
tyrosine-phosphorylated peptides of the putative MAFA-ITIM sequence
SIYSTL bound the phosphatases SHP-1, SHP-2, and SHIP
(16). This suggests that SIYSTL in MAFAfunctions as
classical ITIM motif, which might be involved in the capacity of MAFA
to inhibit the IgE-mediated mast cell degranulation
(17).
Therefore, we speculated that the amino acid stretch SIYEMV in CDw92
could link CDw92 to ITIM-mediated signal transduction cascades. To test
this, we ectopically expressed CDw92 in the murine mast cell CPII that
was stably transfected with a TNF- reporter gene (18),
and we analyzed whether incubation by mAbs VIM15 or VIM15b inhibits
IgE-mediated activation of this gene in these cells. Application of the
mAbs neither cross-linked (using rabbit anti mouse Ig
F(ab')2) nor in monomeric form led to a
significant inhibition of the Fc
RI-dependent transcriptional
induction of TNF- compared with an isotype control mAb. In addition,
no differences were observed between the CDw92-transfected CPII mast
cells vs the parental cell line. To exclude a lack of an ITIM-based
negative regulating signaling cascade in this particular cell line, the
effect of cross-linking of the FcRI to the endogenous
ITIM-containing FcRII was studied (19). In this
experimental setting, TNF- transcription was reduced to 50% (data
not shown). Interestingly, it was found that CDw92 coprecipitated with
protein kinases (20), which are required for
phosphorylation of ITIM motifs (21). Thus, it is at the
moment not clear whether this ITIM-related sequence plays a role in the
function of CDw92.
Comparison of the CDw92 protein sequence
Comparison of the CDw92 protein sequence using the basic local
alignment search tool program at the National Center for Biotechnology
Information (Bethesda, MD) revealed 99.1% identity to the human
choline transporter-like protein 1 (hCTL1). The differences to the
protein sequence of hCTL1 are in Cys69 vs Arg,
Val79 vs Ile, Ala644 vs
Ser, and Ile651 vs Leu, and there are two
additional amino acids (Lys653 and
Arg654) in the C terminus of hCTL1 (Fig. 7).
|
), a KR difference of only 16. In any case,
each prediction favors a 10-transmembrane segment arrangement for
hCTL1/CDw92. Ten-transmembrane helical models are discussed for several
transporters such as the bicarbonate transporter (PFAM database,
http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?PF00955), which is the
principal regulator of pH in animal cells; the sodium:solute symporters
(PFAM database
http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?PF00474), which mediate
the intake of a wide variety of molecules with the concomitant uptake
of sodium ions; or the glucose 6-phosphate transporter
(23). Thus, the structural feature of hCTL1/CDw92 together
with the finding that expression of Torpedo marmorata CTL1
increased the high-affinity uptake of choline in a yeast mutant
suggests a transporter function of hCTL1/CDw92, in particular for
choline, in human cells. To analyze a potential function of CDw92 as choline transporter, we assessed in a first experiment uptake of 14C-choline chloride by CDw92 transductants in comparison with the untransduced parental cells. We found a reproducible 10–15% increased incorporation of 14C-choline by the CDw92-transduced mast cell line CPII, but not by CDw92 BW5147 mouse thymoma transductants. The enhanced 14C-choline uptake was only seen 60 min after onset of the culture and not at other time points. Therefore, this experiment does give a further hint but does not clearly attest to CDw92 as a choline transporter. Inasmuch as several choline transporters do exist (22) that might be able to substitute each other, the revelation of CDw92 as choline transporter requires a separate set of carefully designed experiments.
Comparison of the CDw92 nucleotide sequence
A comparison of the nucleotide sequences of CDw92 and hCTL1
revealed that the discrepancies in the amino acid sequences resulted
from the following base changes: codon TGT
(Cys69) vs CGT (Arg69),
codon GTA (Val79) vs ATA
(Ile79), and codon GCC
(Ala644) vs TCC (Ser644).
Furthermore, a silent point mutation in codon GCC vs GCT
(Ala7) was found. Whether these differences
reflect a polymorphism or are the result of a cloning artifact is not
clear. However, replacement of the C-terminal codons ATT
(Ile651) and AAG (Lys652)
of CDw92 vs CTG (Leu651), AAG
(Lys652), AAA (Lys653), and
AGG (Arg654) of hCTL1, and the completely
unrelated 3'-untranslated regions show that despite the strong
homology, CDw92 and hCTL1 are not identical (Fig. 8). Nonetheless, both sequence-stretches
downstream of Met650 match 100% to segments
contained in the genomic clone RP11-287A8 (accession no. AL161627,
version GI:10443402). This clone was constructed from chromosome 9 and
was sequenced by the Sanger Center Chromosome 9 Mapping Group
(Hinxton Hall, UK; http://www.sanger.ac.uk/HGP/Chr9). Although the
CDw92 3' segment downstream of Met650 is
located between bp 8010 and 8528 of clone RP11-287A8, the hCTL1 3'
segment is separated by 48,000 bp and is located between bp 56,387
and 56,610 (Fig. 9). Both 3' segments are
flanked by splice junction consensus sequences. This shows that the
extreme C termini and the 3'-untranslated regions are determined by
individual exons and that the CDw92 and the hCTL1 clone are splice
variants. To check whether other 3' exons do exist, we compared the
sequence section between the last common exon of hCTL1/CDw92 and the 3'
exon of CDw92 (bp 3609–8009 of genomic clone RP11-287A8; Fig. 9) as well as the 48,000 bp spacing the 3' exons of CDw92
and hCTL1 (bp 8,529–56,386 of genomic clone RP11-287A8; Fig. 9) with
expressed sequence tag (EST) entries contained in the sequence
databases. We found a number of different ESTs matching this part of
the sequence. Basically, one can watch this cluster of ESTs
(http://www.ensembl.org/Homo_sapiens/contigview?chr=9&vc_start=
107335209&vc_end= 107435209&x=46&y=7) at the Human Genome Server
Ensembl. This finding suggests that further splice variants with
different C termini and 3'-untranslated regions do exist.
Indeed, four different but overlapping EST clones (accession nos.
AV71362, AU133344, BE787530, and AW408761) clearly define a third
variant with the C terminus Ala-Ser-Gly-Ala-Ser-Ser-Ala-COOH.
Interestingly, serines located two residues from the C terminus (-2
position) as found in the latter variant were shown to be critical in
C-terminal peptide motifs recognized by PDZ domains (24).
Two C-terminal variants of the sodium bicarbonate cotransporter NBC4
that is also a multimembrane protein have been identified recently
(25). One of these variants also contains in the extreme C
terminus a consensus PDZ-interaction domain, a motif that appears to be
crucial for linking proteins within signaling cascades
(26). The 3'-untranslated regions of the messages of
various proteins were shown to be implicated in the stability of mRNA
and the regulation of translation (27, 28, 29, 30, 31, 32). The fact
that hCTL1/CDw92 employs at least three individual exons for
transcription of the C terminus and the 3'-untranslated region suggests
that alternate usage of these sequences is a critical level to control
hCTL1/CDw92 protein expression and function.
|
|
, in this part, the
sequences of the hCTL1 and CDw92 clones differ considerably. One reason
for this might be that 175 of the 209 residues forming the
5'-untranslated end of the CDw92 clone presented in this study are
either G or C. This high G and C content (83%) caused considerable
difficulty in sequencing this region, probably because of stable
secondary structures known to be formed by GC-rich sequence stretches
(33). Such stable secondary structures result in stalling
of DNA polymerase and reverse transcriptase. However, our
5'-untranslated sequence is 100% identical with the base pair segment
133,821–134,030 of genomic clone RP1-5K7 (accession no. AC016666,
version AC016666.2 GI: 7230031). Thus, the differences in the 5' end of
hCTL1 compared with CDw92 are cloning artifacts. In the genomic clone
RP11-5K7, the 5'-untranslated sequence is followed by the first 36
residues of the open reading frame of hCTL1/CDw92. Then, this sequence
stretch (245 bp) stops at a splice junction consensus sequence, which
shows that this sequence belongs to exon 1 of hCTL1/CDw92. We were not
able to clarify yet whether this is the complete exon or only part of
exon 1. Because of the problematic 5' sequence, despite the length of
209 bp, it is not clear whether the CDw92 clone shown in this study
contains the entire 5'-untranslated end of the hCTL1/CDw92 transcript.
In addition to exon 1 and the three exons identified by genomic clone
RP11-287A8 (Fig. 9), we found 12 additional hCTL1/CDw92 exons when
browsing the human genome sequence using the Map Viewer via the home
page of the National Center for Biotechnology Information. Thus, both
cDNAs of hCTL1 and CDw92 consist of 16 exons. Not only do complex and extremely stable secondary structures within 5'-untranslated regions stall enzymes used for cloning, but so does the 43S ribosome subunit when scanning the transcript for the AUG start codon in vivo. Thus, proteins whose transcripts form secondary structures within the 5' end must initiate translation through mechanisms different to the ribosome scanning or cap-dependent mechanism, which is initiated by the eIF-4F protein complex that facilitates binding of the 43S to mRNA (32). One of these alternative mechanisms capable of bypassing the impediment of RNA secondary structures in the 5'-untranslated region is internal ribosome entry. It appeared recently that this interplay between RNA secondary structures and internal ribosome entry guarantees translation of key cellular proteins even under conditions in which overall cellular protein synthesis, including formation of the eIF-4F complex, is compromised. Growth arrest, heat shock, stress, apoptosis, hypoxia, and virus infection are examples of adverse conditions diminishing eIF-4F complex activity. By internal ribosome entry, c-myc was efficiently translated under apoptosis (34) and the angiogenic growth factor VEGF under hypoxic conditions (35). Inasmuch as internal ribosome entry was shown to be responsible for translation of c-jun that has a similar GC-rich sequence as hCTL1/CDw92 (33), it is tempting to speculate that this mechanism also contributes to the regulation of expression of hCTL1/CDw92. In respect to a putative role of hCTL1/CDw92 in transport of choline for phospholipid and/or acetylcholine synthesis, one can envisage that efficient translation of hCTL1/CDw92 must also be secured under unfavorable cellular conditions.
Analysis of CDw92 expression upon cell activation/differentiation
Because both the 5' end and the 3' end of the hCTL1/CDw92 message
suggested a complexity of levels for controlling hCTL1/CDw92
expression, we were keen on studying regulation of CDw92 upon cell
activation/differentiation. First, we analyzed the binding of the CDw92
mAbs to differentially activated PBLs in comparison with nontreated
cells. No evidence of altered CDw92 expression was detected upon
stimulation with PHA or ionomycin. Then, we analyzed HUVECs and dermal
human microvascular endothelial cells when treated with TNF-; these
cells also did not change their CDw92 staining pattern upon stimulation
(Fig. 10), which is in agreement with a
previous report (36). Granulocytes also did not change
CDw92 surface expression by cultivation with PMA (37) and
monocytes upon differentiation to Mo-DCs using GM-CSF plus IL-4
(38). Finally, we examined CDw92 expression on
differentially activated peripheral blood monocytes and Mo-DCs. We
cultured both types of these cells for 1 day in the presence of LPS,
LPS/TNF-, IFN-, IFN-, calcium ionophore, or ionomycin. On
monocytes, CDw92 expression was not modified by these stimuli (data not
shown). Also, Mo-DCs did not respond in terms of CDw92 expression to
LPS, LPS/TNF-, IL-10, IFN-, or IFN- (Fig. 11 and data not shown). However, when
we treated Mo-DCs for 1 day with ionomycin or calcium ionophore, CDw92
expression was down-regulated. Interestingly, however, CDw92 was
reinduced when ionomycin- or calcium ionophore-treated Mo-DCs were
cultured in the presence of IL-10 (Fig. 11).
|
|
). However, only the Ca2+-releasing agent
caused down-regulation of CDw92 demonstrating that different maturation
states are induced by ionomycin and LPS in DCs. Modulation of DC function by CDw92 mAbs
Based on this selective regulation of CDw92 on Mo-DCs, we
assumed a particular function of CDw92 on these cells. To study this,
we used our CDw92 mAbs as surrogate ligands and we analyzed their
influence on the function of Mo-DCs. The most interesting finding
obtained by these experiments was the 2.8-fold augmentation of
LPS-induced production of IL-10 (1167 ± 388 vs 411 ± 41
pg/ml; n = 3) by one of our CDw92 mAbs, VIM15b, with
Mo-DCs. VIM15, the second CDw92 mAb, had no influence. The effect of
mAb VIM15b appeared to be specific for IL-10 and LPS-pretreated Mo-DCs,
as it did not change production of TNF-, IL-1, and IL-12-p40 in
these cells or untreated Mo-DCs. Furthermore, both CDw92 mAbs had no
significant effect on production of these cytokines on monocytes
pretreated with or without LPS, IFN-, or LPS plus IFN- (Fig. 12 and data not shown).
|
In conclusion, these results indicate a constitutive and relatively
stable expression of hCTL1/CDw92 on various cells of the hemopoietic
system, but a subtle regulation by immunomodulating substances,
ionomycin, and IL-10 on DCs. Ionomycin is known as strong activator of
DC immune responses, up-regulating Ag-presenting and costimulatory
molecules (39 and Fig. 11), whereas IL-10 negatively
controls DC responses (40, 41, 42). The specific regulation by
these agents, together with the structural information including the
potential ITIM motif and the observed involvement in IL-10 expression
by DCs, suggests that hCTL1/CDw92 might not only act as a mere
transporter of choline for membrane phospholipid synthesis of immune
cells, but may also be implicated in specific regulation of immune
functions, in particular negative signaling pathways.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 S.W. and A.S. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Hannes Stockinger, Institute of Immunology, Vienna International Research Cooperation Center at Novartis Forschungsinstitut, Brunner Strasse 59, A-1235 Vienna, Austria. E-mail address: Hannes.Stockinger{at}univie.ac.at
4 Abbreviations used in this paper: DC, dendritic cell; Mo-DC, monocyte-derived DC; TMD, transmembrane domain; ITIM, immunoreceptor tyrosine-based inhibition motif; MAFA, mast-cell function-associated Ag; hCTL1, human choline transporter-like protein 1; EST, expressed sequence tag.
Received for publication December 20, 2000. Accepted for publication September 13, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
RI-mediated activation of mast cells by 2,3,4- trihydropyrimidino[2,1-a]isoquinolines. J. Med. Chem. 41:1050.[Medline]
RIIB, regulates negatively BCR-, TCR-, and FcR- dependent cell activation. Immunity 3:635.[Medline]
receptor I-associated lyn-dependent phosphorylation of Fc receptor IIB during negative regulation of mast cell activation. J. Immunol. 160:1647.This article has been cited by other articles:
|
|
 |
|
  V. Michel and M. Bakovic The solute carrier 44A1 is a mitochondrial protein and mediates choline transport FASEB J, August 1, 2009; 23(8): 2749 - 2758. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Bles, M. Horckmans, A. Lefort, F. Libert, P. Macours, H. El Housni, F. Marteau, J.-M. Boeynaems, and D. Communi Gene Expression Profiling Defines ATP as a Key Regulator of Human Dendritic Cell Functions J. Immunol., September 15, 2007; 179(6): 3550 - 3558. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Yuan, A. Tie, M. Tarnopolsky, and M. Bakovic Genomic organization, promoter activity, and expression of the human choline transporter-like protein 1 Physiol Genomics, September 14, 2006; 26(1): 76 - 90. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Michel, Z. Yuan, S. Ramsubir, and M. Bakovic Choline Transport for Phospholipid Synthesis. Experimental Biology and Medicine, May 1, 2006; 231(5): 490 - 504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Fullerton, L. Wagner, Z. Yuan, and M. Bakovic Impaired trafficking of choline transporter-like protein-1 at plasma membrane and inhibition of choline transport in THP-1 monocyte-derived macrophages Am J Physiol Cell Physiol, April 1, 2006; 290(4): C1230 - C1238. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kirchberger, O. Majdic, P. Steinberger, S. Bluml, K. Pfistershammer, G. Zlabinger, L. Deszcz, E. Kuechler, W. Knapp, and J. Stockl Human Rhinoviruses Inhibit the Accessory Function of Dendritic Cells by Inducing Sialoadhesin and B7-H1 Expression J. Immunol., July 15, 2005; 175(2): 1145 - 1152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Guillemin and B. J. Brew Microglia, macrophages, perivascular macrophages, and pericytes: a review of function and identification J. Leukoc. Biol., March 1, 2004; 75(3): 388 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Selenko-Gebauer, O. Majdic, A. Szekeres, G. Hofler, E. Guthann, U. Korthauer, G. Zlabinger, P. Steinberger, W. F. Pickl, H. Stockinger, et al. B7-H1 (Programmed Death-1 Ligand) on Dendritic Cells Is Involved in the Induction and Maintenance of T Cell Anergy J. Immunol., April 1, 2003; 170(7): 3637 - 3644. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
IL-10). CDw92 expression was assessed by
indirect immunofluorescence and flow cytometry using mAb VIM15 (filled
areas). mAb VIAP was included as control (open areas). Data are
representative of four separate experiments.