Department of Immunology, Duke University Medical Center, Durham, North Carolina
Correspondence: Thomas Tedder, Box 3010, Department of Immunology, Room 353 Jones Building, Research Drive, Duke University Medical Center, Durham, NC 27710. E-mail thomas.tedder{at}duke.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: inflammation • vascular biology • leukocyte adhesion • cell trafficking • adhesion molecules
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
L-Selectin ligands on HEVs characteristically bear a sulfate-dependent carbohydrate epitope defined by the MECA-79 mAb [18 19 20 21 22 23 ]. Others have proposed that sialyl 6-sLex and 6'-sulfo-sLex, sulfated derivatives of the sLex tetrasaccharide, are structural components for L-selectin ligand recognition [23 , 24 ]. However, vascular endothelial cells do not normally express the MECA-79 antigen [14 ]. Nonetheless, sulfation of L-selectin ligands on vascular endothelial cells is required because chlorate treatment of cytokine-activated HUVEC and FucT-VII cDNA-transfected EA.hy926 cells (926-FtVII cells) eliminates L-selectin-mediated leukocyte attachment [14 ]. Membrane-bound sulfotransferases potentially involved in the generation of selectin ligands have been identified recently [25 26 27 28 29 30 31 32 33 ]. This new family of membrane-bound enzymes currently includes CHST1 (chondroitin 6-sulfotransferase, keratan sulfate galactose-6-sulfotransferase, KSST), CHST2 (6-sul-T and N-acetylglucosamine-6-O-sulfotransferase), CHST3 (chrondroitin 6-sulfotransferase), and CHST4 (HEC-GlcNAc6ST). Among this family of enzymes, CHST4 is reported to be specifically expressed by HEVs, while CHST2 is highly expressed by mouse HEVs [27 , 29 , 33 ]. CHST1, CHST2, and CHST3 are expressed by a broad range of other tissues [25 , 26 , 28 , 30 31 32 , 34 , 35 ]. However, the expression of the CHSTs by vascular endothelial cells and their involvement in the biosynthesis of vascular L-selectin ligands remain unclear. Therefore, we assessed the role of CHSTs in generating vascular L-selectin ligands.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(Genzyme,
Cambridge, MA) for 6 h.
Antibodies and Flow Cytometry Analysis
The anti-L-selectin (LAM1-3) mAb was produced and
purified as described elsewhere [36
, 38
].
The HECA-452 (anti-CLA), CSLEX-1 (anti-sLex), and MECA-79
mAb-producing hybridomas were obtained from the ATCC. Reactivity of the
MECA-79 mAb was verified by staining HEVs of mouse lymph node tissue
sections using standard immunochemical techniques (unpublished
results). The anti-PSGL-1 (PL-1; K. Moore, University of Oklahoma
Health Science Center, Oklahoma City, OK), anti-CD57 (HNK-1; M. Cooper,
University of Alabama in Birmingham, Birmingham, AL), anti-P-selectin
(G1; B. Furie, Beth Israel Medical Center, Boston, MA), and anti-VAP1
(TK8-14; S. Jalkanen, University of Turku, Turku, Finland) mAbs were
generous gifts. The mouse mAbs to human E-selectin (HAE-1f), vascular
cell adhesion molecule 1 (VCAM-1 [HAE-2d]), and intercellular
adhesion molecule 1 (ICAM-1 [HAE-4b]) were produced as previously
described [3
]. The anti-CD34 mAb (My10; Becton
Dickinson, San Jose, CA), anti-mouse CD25 mAb (rat immunoglobulin M
[IgM]), fluorescein isothiocyanate-conjugated goat anti-rat IgM
(anti-rat Ig [H + L]), and anti-mouse IgM antibodies were
purchased from Southern Biotechnology Associates (Birmingham, AL). All
mAbs were used at predetermined optimal concentrations for indirect
immunofluorescence staining. Staining and immunofluorescence analysis
were performed as previously described with a FACScan flow cytometer
(Becton Dickinson) [14
, 16
].
CHST and FucT-VII cDNA expression vectors
A CHST1 cDNA clone [28
] was amplified by
polymerase chain reaction (PCR) using the 5'-GCT CTA GAC ATG CAA TGT
TCC TGG AAG-3' and 5'-CCC TCG AGT CAC GAG AAG GGG CGG AAG-3' primer
set. A CHST2 cDNA clone [28
] was PCR amplified using the
5'-GCT CTA GAA TGA GCC GCA GCC CGC AGC GA-3' and 5'-GCG TCG AGA ACC CCT
TTT AGA GAC GGG G-3' primer set. The PCR products were cloned into the
BlueScript vector (Stratagene, La Jolla, CA) and sequenced. The CHST1
and CHST2 coding regions were excised using the XbaI and
XhoI restriction enzymes and cloned into the pcDNA3.1(-)
vector (Invitrogen, Carlsbad, CA) at the appropriate sites. An internal
ribosome entry site (IRES)-ß-gal fusion gene (from Y. Zhuang, Duke
University Medical Center, Durham, NC) was then inserted into the
XhoI site downstream of the CHST1 or CHST2 coding region,
resulting in new vector pcDNAIßST1 or pcDNAIßST2. A FucT-VII cDNA
(from B. Weston, University of North Carolina at Chapel Hill, NC) was
modified and used as described elsewhere [14
].
Generation of cDNA-transfected cell lines
EA.hy926 cells transfected with FucT-VII cDNA (926-FtVII) were
generated previously [14
]. In this study, EA.hy926 and
ECV-304 cells were transfected with either a FucT-VII cDNA expression
vector or a pcDNAIßST vector or with both vectors together using the
calcium phosphate method [3
, 14
].
Transfected cells were selected in Dulbecco’s modified Eagle’s medium
containing 10% FCS and G418 (0.5 mg/mL; Sigma). Surviving cells
expressing the HECA-452 antigen or ß-gal were isolated by
fluorescence-based cell sorting using a FACStar flow cytometer (Becton
Dickinson). Cells transfected with both FucT-VII and CHST cDNAs were
first sorted based on high levels of HECA-452 antigen expression.
Subsequently, cells were sorted based on ß-gal expression using the
fluorescein di-ß-D-galactopyranoside (FDG; Molecular
Probes, Eugene, OR) ß-gal substrate as previously described
[39
]. Briefly, the cells (1 x 106)
were resuspended in 50 µL of phosphate-buffered saline (PBS) at
37°C. FDG solution (50 µL; 2 mmol in 1% (v/v) dimethyl sulfoxide)
at 37°C was added to the cell suspension and mixed quickly. After
incubation at 37°C for 2.5 min, 10 v of ice-cold PBS containing
5% FCS were added to stop FDG loading. The cells were maintained on
ice for 30–60 min until flow cytometry analysis or cell sorting. In
additional experiments, cells transfected with both FucT-VII and CHST
cDNAs were isolated first based on increased ß-gal activity, with
subsequent isolation based on HECA-452 antigen expression. These cells
were phenotypically and functionally identical to the comparable cells
isolated as first described above (unpublished data).
Physiologic shear flow assays
In vitro flow chamber assays were performed as previously
described [16
]. Monolayers of EA.hy926 cells, ECV-304
cells, and their cDNA-transfected counterparts were grown to confluence
on 25-mm circular glass coverslips and mounted in a parallel-plate flow
chamber. 300-LAM cells (106 cells/mL) in flow medium (PBS
containing Ca2+-Mg2+
and 0.5% bovine serum albumin) were drawn through the chamber with a
syringe pump (Harvard Apparatus, Natick, MA) at defined flow rates.
Wall shear stress was varied at 1-min intervals by changing the flow
rate through the flow chamber. For antibody-blocking experiments,
300-LAM cells (in flow medium) were incubated with the LAM1-3 mAb (10
µg/mL) at room temperature for 15 min before a fivefold dilution with
flow medium. Rolling cells were observed using an inverted
phase-contrast microscope (Olympus Corporation, Lake Success, NY) and
videotaped using a charge-coupled device video camera (Hitachi Denshi,
Ltd., Tokyo, Japan), a Sony SuperVHS video recorder (model SVO-9500MD;
Sony Corp. of America, New York, NY), and an attached time-date
generator (Microimage Video Systems Co., Bechtelsville, PA).
Interacting cells (tethering and rolling) were quantified at each shear
rate by analysis of videotapes in which 300-LAM cells interacting with
the adherent cell monolayer were counted in 16 randomly chosen
0.16-mm2 fields near the center line of the flow chamber.
For calculating velocities, the distance that individual rolling cells
traveled between two time points was measured, converted into actual
distance, and divided by the elapsed time.
RNA isolation and Northern blot analysis
Cytoplasmic RNA free of DNA contamination was isolated from each
cell type using an RNeasy Mini Kit (Qiagen, Inc., Chatsworth, CA)
according to the manufacturer’s instructions. RNA concentrations were
determined by ultraviolet absorbance. RNA (30 µg/lane) was
electrophoresed through 1% agarose-formaldehyde gels and transferred
to nitrocellulose filters. CHST1, CHST2, and ß-actin probes were
generated by labeling PCR-generated DNA fragments with
[-32P]dATP using the Prime-a-Gene® labeling system
with random hexadeoxynucleotide primers (Promega, Madison, WI). After
overnight hybridization at 65°C, the filters were washed once at room
temperature for 10 min in 1x saline sodium citrate containing 0.1%
sodium dodecyl sulfate and twice at 60°C for 20 min in 0.1x saline
sodium citrate containing 0.1% sodium dodecyl sulfate. Hybridization
was visualized using a Storm 680 Phosphorimager (Molecular Dynamics,
Sunnyvale, CA).
PCR analysis of sulfotransferase expression
Semiquantitative reverse transcription (RT)-PCR analysis was
performed using equal amounts of RNA (
2 µg) from each cell type
for RT reactions as described elsewhere [35
,
40
]. Briefly, cDNA over a range of dilutions (neat to
1:100) was PCR amplified using predetermined optimal conditions. PCR
amplification of CHST1 cDNA used sense 5'-CAC GCG CAG CGG CTC CTC CTT
CGT-3' and antisense 5'-GCC AGG TCC TCG TAG CGC ACC G-3' primers to
generate a 713-bp fragment. CHST2 amplification used sense 5'-GGG CGC
AAC CTC ACC ACG-3' and antisense 5'-CCA CGA AAG GCT TGG AGG
AGG-3'—primers to generate a 690-bp fragment. CHST3 amplification used
sense 5'-CAA CCA GCA GGG CAA CAT CT-3' and antisense 5'-CCC TAC GTG ACC
CAG AAG G-3' primers to generate a 980-bp fragment. CHST4 amplification
used sense 5'-GTG GTG GAG AAG GCC TGC CG-3' and antisense 5'-ACC CTC
TTA GTG GAT TTG CT-3' primers to generate a 680-bp fragment. Primers
for ß-actin were sense 5'-ATG TTT GAG ACC TTC AAC AC-3' and antisense
5'-CAC GTC ACA CTT CAT GAT GG-3', which generate a 495-bp fragment.
Primers for hypoxanthine phosphoribosyltransferase (HPRT) were sense
5'-CCT GCT GGA TTA CAT CAA AGC AC-3' and antisense 5'-TCC AAC ACT TCG
TGG GGT CCT-3', which generate a 300-bp fragment. As controls, RNA from
each cell type equal to the amount of RNA used for cDNA generation was
used in PCR reactions without RT. PCR signals were not detected in any
of the control reactions, ruling out the possibility that the PCR
products were generated from contaminating genomic DNA.
PCR conditions for the CHSTs were 94°C for 2 min, then 30 cycles of 94°C for 1 min, 58°C for 1 min, and 72°C for 40 s. PCR conditions for ß-actin were 94°C for 2 min, then 25 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 30 s. PCR conditions for HPRT were 94°C for 2 min, then 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 20 s. PCR products were electrophoresed on agarose gels and visualized by ethidium bromide staining. Relative PCR product concentrations were assessed by scanning densitometry and analyzed using NIH Image software (version 1.60). The relative intensities of the HPRT bands generated from each cDNA dilution were used to generate dose-response curves for each cDNA aliquot. In each case, the intensity of PCR products generated at a 1:10 or 1:20 dilution of cDNA was in the midpoint of the linear range where band intensity was proportional to the amount of input cDNA. RNA isolated from HL-60 and COS cells served as positive controls for CHST1, CHST2, CHST3, and CHST4 amplification by PCR as previously described [35 ].
Statistical analysis
Data were expressed as means ± SE. The
Student’s t-test was used to determine the significance of
differences in sample population means.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (46K): [in a new window] |
Figure 1. Expression of CHST transcripts by vascular and nonvascular cells. For
Northern blot analysis of CHST1 and CHST2 expression by cell lines,
filters bearing 30 µg of RNA from each cell type were hybridized
with radiolabeled CHST1, CHST2, or ß-actin cDNA probes (A).
Hybridization was visualized using a phosphorimager. Ribosomal RNA
migration is shown as an indicator of size. To analyze CHST expression
by resting and activated HUVECs, we performed semiquantitative RT-PCR
amplification of total RNA prepared from equivalent numbers of HUVEC
cells cultured in medium alone or containing TNF- (100 U/ml) for
6 h (B). CHST expression by EA.hy926 and ECV-304 cells was
analyzed as follows (C): for controls, HPRT (not shown) or ß-actin
primers were used to amplify both mRNA (ß + RNA negative
control) and its cDNA product (ß + cDNA positive control).
Semiquantitative RT-PCR products were visualized by electrophoresis and
ethidium bromide staining and represent results from four separate
experiments. The migration of standard size markers is shown.
|
treatment enhanced CHST2
mRNA expression by 2.8-fold (n = 2) and induced
expression of CHST4 mRNA, whereas CHST1 and CHST3 mRNA levels did not
change significantly. In addition, EA.hy926 cells consistently
expressed three- to fivefold-higher levels of CHST1 and CHST2
transcripts than did ECV-304 cells, whereas CHST3 mRNA expression was
equivalent in both cell lines (n = 2; Fig. 1C
). CHST4
transcripts were not detected in either cell line. Thus, vascular
endothelial cells may express multiple CHSTs, and inflammatory
mediators may influence their transcription.
EA.hy926 and ECV-304 cells express functional
L-selectin ligands
Functional L-selectin ligand expression by EA.hy926
and ECV-304 cells was quantified by assessing their ability to support
rolling of 300.19 cells stably transfected with L-selectin
cDNA (300-LAM) during in vitro flow chamber assays [16
].
300-LAM cells were used for these studies, because native 300.19 cells
do not attach to or roll on EA.hy926 and ECV-304 cells or
cytokine-activated endothelial cell monolayers [14
,
16
, 41
]. Moreover, the use of 300.19 cells
allows examination of L-selectin function in isolation
since 300.19 cells do not express most of the known vascular adhesion
molecules [42
, 43
].
Neither EA.hy926 nor ECV-304 cells supported 300-LAM cell rolling (Fig. 2B ), consistent with their lack of HECA-452 antigen expression (Fig. 2A , Table 1 ). However, transfection of EA.hy926 and ECV-304 cells with FucT-VII cDNA (resulting in 926-FtVII and 304-FtVII cells, respectively) induced HECA-452 antigen expression (Fig. 2A , Table 1 ) and 300-LAM cell rolling (Fig. 2B) . The majority of 300-LAM cells in contact with 926-FtVII or 304-FtVII monolayers remained in contact for their entire transit across the field of view, and there was no stationary adhesion of 300-LAM cells to the cell monolayers. 300-LAM cells rolled at velocities significantly lower than the theoretical velocity of a lymphocyte not interacting with the endothelial cell surface (526 µm/s at 1.85 dynes/cm2) [44 ]. 926-FtVII cells supported leukocyte tethering and rolling over a broad range of shear stress levels (0.25–2.5 dynes/cm2), while 304-FtVII cells supported tethering and rolling only within a narrow range of shear stress levels (0.25–1.25 dynes/cm2). Similarly, 304-FtVII cells supported the rolling of significantly fewer 300-LAM cells when compared with 926-FtVII cells at the same shear stresses. 300-LAM cell rolling on 926-FtVII and 304-FtVII cells was completely abolished by blocking L-selectin function with the LAM1-3 mAb (Fig. 2B) . Therefore, 926-FtVII cells appeared more efficient than 304-FtVII cells in supporting transient interactions (tethering) between L-selectin and its ligands and in mediating sustained rolling. It is unlikely that this results from FucT-VII expression differences by 926-FtVII and 304-FtVII cells because both cell lines displayed similar HECA-452 antigen levels (Fig. 2A , Table 1 ). In addition, 926-FtVII and 304-FtVII cells were similar in their lack of expression of other reported ligands for L-selectin and other adhesion receptors (Table 1) . Therefore, the differential abilities of 926-FtVII and 304-FtVII cells to support L-selectin-mediated leukocyte rolling suggested a correlation between L-selectin ligand activity and endogenous cellular expression of CHST1 and CHST2 transcripts.
 View larger version (37K): [in a new window] |
Figure 2. EA.hy926 and ECV-304 cells expressed functional L-selectin
ligands after transfection with FucT-VII cDNA. HECA-452 antigen
expression by 926, 304, 926-FtVII, and 304-FtVII cells was assessed by
indirect immunofluorescence staining with flow cytometry analysis (A).
Solid lines indicate HECA-452 mAb staining, while isotype-matched
control mAb staining is indicated by dashed lines. Flow cytometry
histograms are shown on a four-decade log scale. To examine the
tethering and rolling of 300-LAM cells on monolayers of EA.hy926,
ECV-304, 926-FtVII and 304-FtVII cells, flow was initiated in these in
vitro parallel-plate flow chamber assays at 3.0 dynes/cm2,
and shear stress was reduced as indicated at 1-min intervals by
changing the flow rate through the flow chamber (B). 300-LAM cells
preincubated with either medium or the LAM1-3 mAb (10 µg/mL) were
diluted to 106 cells/mL and perfused through the flow
chamber. Values represent means ± SE of results from
three experiments. In each experiment, the mean number of 300-LAM cells
interacting with the adherent cell monolayer was counted in 16 randomly
chosen fields at each shear stress.
|
|
View this table: [in a new window] |
Table 1. Adhesion Molecule Expression by EA.hy926 and ECV-304 Cell Lines
|
Effects of CHST2 overexpression by 926-FtVII cells
Because CHST2 expression was found to be up-regulated in HUVECs
after TNF- treatment, the relationship between sulfotransferase
expression levels and L-selectin ligand activity was
assessed by increasing CHST2 expression in EA.hy926 cells. To monitor
sulfotransferase transcript levels in transfected cells, bicistronic
CHST2 expression constructs were generated by placing an IRES element
and a ß-gal gene downstream of CHST2 cDNA in the pcDNA3.1(-) vector
(Fig. 3A
). Thereby, ß-gal levels correlated linearly with CHST
transcription levels because ß-gal translation was initiated via an
IRES site within CHST transcripts. EA.hy926 cells were cotransfected
with a CHST2/ß-gal-encoding vector and a conventional FucT-VII cDNA
expression vector. Repeated sorting of cDNA-transfected EA.hy926 cells
resulted in the isolation of EA.hy926 and 926-FtVII cells with elevated
ß-gal activity that were termed 926-CHST2 and 926-FtVII-CHST2 cells,
respectively (Fig. 3B)
. On average, ß-gal activity in 926-CHST2 cells
was 20-fold higher (n = 2) than in EA.hy926 cells,
while 926-FtVII-CHST2 cells expressed fourfold higher ß-gal activity
(n = 3). By contrast, repeated sorting of mock
transfected EA.hy926 cells for high ß-gal activity did not select for
EA.hy926 cells with ß-gal activity higher than the levels endogenous
to parental EA.hy926 cells (data not shown). Enhanced CHST2 expression
by EA.hy926 cells was further verified by semiquantitative RT-PCR
amplification of cDNA from each transfected cell line. 926-CHST2 and
926-FtVII-CHST2 cells expressed four- to fivefold higher levels of
CHST2 transcripts than parental EA.hy926 cells (n = 4;
Fig. 3C
). HECA-452 or CSLEX1 antigen expression levels were not
influenced by enhanced CHST2 expression or FDG-based cell sorting (Fig. 3
; data not shown).
 View larger version (27K): [in a new window] |
Figure 3. Enhanced CHST2 expression by transfected EA.hy926 cells. The structure
of the bicistronic CHST expression vector that encodes
ß-galactosidase (ß-gal) via an IRES element was determined (A).
Abbreviations: IRES, internal ribosome entry site; ß-gal, bacterial
ß-galactosidase gene; BGA pA, bovine growth hormone gene
polyadenylation signal sequence; SV40 Ori, simian virus 40 origin of
replication; Neor, neomycin-resistant gene; SV40 pA, simian
virus 40 polyadenylation signal sequence; Ori, ColE1 origin of
replication; Ampr, ampicillin-resistant gene; Pcmv,
cytomegalovirus promoter sequence. HECA-452 antigen and ß-gal
expression by cDNA-transfected cells was also determined (B). The
indicated cell lines (above columns) were assessed for HECA-452 antigen
expression (solid lines) by indirect immunofluorescence staining or
were loaded with FDG (solid lines) to assess ß-gal expression levels,
with subsequent flow cytometry analysis. Dashed histograms represent
cell staining with an isotype-matched control mAb or the fluorescence
intensity of cells not loaded with FDG. Results represent those
obtained in three separate experiments and are shown on a four-decade
log scale. To determine CHST2 mRNA expression, semiquantitative RT-PCR
analysis was carried out with serial dilutions (as indicated) of cDNA
prepared from RNA of the indicated cell lines (C). PCR products were
visualized by electrophoresis with ethidium bromide staining. Results
represent those from at least two separate experiments.
|
 View larger version (29K): [in a new window] |
Figure 4. Effect of enhanced CHST2 expression by EA.hy926 cells on leukocyte
rolling. Rolling of 300-LAM cells on monolayers of cDNA-transfected
EA.hy926 cells during in vitro flow chamber assays was assessed as in
Fig. 2B
(A). Values represent means ± SE of results
from three experiments. Velocities of rolling leukocytes were
determined (B). Velocity measurements are from one of three experiments
with similar results. Each symbol represents the velocity of an
individual 300-LAM cell plotted in rank order at 1 dyne/cm2
of shear stress.
|
 View larger version (41K): [in a new window] |
Figure 5. Enhanced CHST1 and CHST2 expression by ECV-304 cells. HECA-452, CSLEX1,
and MECA-79 antigen and ß-gal expression by cDNA-transfected ECV-304
cells are shown (A). Each cell line was assessed for cell surface
antigen expression (solid lines) by indirect immunofluorescence
staining or was loaded with FDG to assess ß-gal expression levels,
with subsequent flow cytometry analysis. Dashed histograms represent
cell staining with an isotype-matched control mAb or the fluorescence
intensity of cells not loaded with FDG. Results represent three
separate experiments and are shown on a four-decade log scale. CHST1
and CHST2 mRNA expression by ECV-304 transfectants was determined by
semiquantitative RT-PCR analysis performed as described for Fig. 3C
(B). PCR amplification of mRNA samples is shown as a negative control.
Results represent two separate experiments.
|
 View larger version (31K): [in a new window] |
Figure 6. Effect of enhanced CHST1 and CHST2 expression by ECV-304 cells on
leukocyte rolling. Rolling of 300-LAM cells on ECV-304 cell monolayers
during in vitro flow chamber assays was assessed as in Fig. 2B
(A, B).
Values represent means ± SE of results from three
experiments. Asterisks indicate that 300-LAM cell rolling on
304-FtVII-CHST1 or 304-FtVII-CHST2 monolayers was significantly
different from rolling on 304-FtVII monolayers (P <
0.001). Velocities of leukocytes rolling on ECV-304 cell monolayers
were determined (C). Each symbol represents the velocity of an
individual 300-LAM cell plotted in rank order at 1 dyne/cm2
of shear stress. These results represent three separate experiments.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
926-FtVII cells constitutively expressed CHST1, CHST2, and CHST3 transcripts and supported leukocyte rolling via L-selectin (Fig. 1 and 2) . It is surprising that further increases in CHST2 expression by 926-FtVII cells had little effect on leukocyte rolling via L-selectin (Fig. 3 and 4) . This result may be explained by their relatively high expression of endogenous CHST2 as well as CHST1 mRNA (Fig. 1) , which may already reach appropriate levels for optimal L-selectin ligand synthesis (Fig. 4) . Consistent with this, EA.hy926 cells constitutively expressed CHST1 and CHST2 transcripts at levels significantly higher than those found in TNF-activated HUVECs. Moreover, 304-FtVII-CHST2 cells expressed CHST2 transcripts at levels threefold higher than those found in parental EA.hy926 cells and at levels equivalent to those expressed by 926-FtVII-CHST2 cells, which may explain why 300-LAM cell rolling on 304-FtVII-CHST2 cells was equivalent to 300-LAM cell rolling on 926-FtVII and 926-FtVII-CHST2 cells (Fig. 4 and 6B) . Because each CHST protein has enzymatic activity, small changes in CHST transcription levels may be functionally significant. Thereby, increased CHST2 expression by cytokine-activated HUVECs may be physiologically relevant (Fig. 1) . Cytokine-induced expression of CHST4 mRNA by HUVECs suggests that CHST4 may also contribute to L-selectin ligand generation (Fig. 1B) . However, the current study demonstrates that CHST4 expression was not required for the generation of functional L-selectin ligands in EA.hy926 or ECV-304 cells. Thus, functional redundancy or overlapping specificity of the CHSTs may promote optimal L-selectin ligand generation.
Expression of CHST1 and CHST2 by vascular endothelial cells is consistent with our current understanding of the biochemical nature of L-selectin ligands on HEV. The four known sulfotransferases expressed by vascular endothelial cells demonstrate overlapping abilities to catalyze sulfation of chondroitin, keratan sulfate, and sialyl lactosamine oligosaccharides [25 26 27 , 29 30 31 32 33 34 , 45 ]. CHST1 exhibits galactose-6-sulfotransferase activity, while CHST2 and CHST4 display GlcNAc-6-sulfotransferase activity [29 , 30 ]. These activities correlate with the two major sulfated carbohydrate groups found on putative L-selectin ligands: 6-sulfo sLex and 6'-sulfo sLex [22 , 23 ]. Since increased expression of either CHST1 or CHST2 by 304-FtVII cells enhanced L-selectin ligand activity, these results are consistent with both 6-sulfo sLex and 6'-sulfo sLex groups as functional components of vascular L-selectin-ligands. The biochemical specificity of vascular L-selectin ligand sulfation awaits further characterization and identification of the vascular L-selectin ligands. Identification of the specific proteins mediating L-selectin-dependent binding in HUVECs and endothelial cell lines is required because a large number of endothelial cell proteins are sulfated. This information will also allow a direct assessment of how the CHSTs regulate ligand interactions.
The current study extends a recent report showing that CHST2 and FucT-VII cotransfection of ECV-304 cells induces L-selectin ligand expression as detected in a nonstatic binding assay [17 ]. However, our study assessed the effect of either enhanced CHST1 or CHST2 expression on leukocyte rolling and attachment via L-selectin under physiological flow conditions. Nonetheless, significant differences in results were observed between the two studies. Differences in sensitivity of the assay systems may explain why 304-FtVII cells supported sustained leukocyte rolling in the current studies (Fig. 2B) , although L-selectin-mediated attachment to 304-FtVII cells was not detected in the nonstatic binding assays used in the earlier studies unless the cells were also transfected with CHST2 cDNA [17 ]. In addition, ECV-304 cells transfected with CHST2 and FucT-VII cDNAs were reported to express the MECA-79 antigen in the previous study [17 ]. However, none of the 304-CHST1, 304-CHST2, 304-FtVII-CHST1, or 304-FtVII-CHST2 cells expressed detectable MECA-79 antigen in the current studies (Fig. 5 , Table 1 ). Similarly, HUVECs, EA.hy926 cells, COS cells, neutrophils, KG-1a cells, and HL-60 cells express FucT-VII and endogenous CHSTs, yet do not display detectable MECA-79 antigen [14 , 35 ]. In addition, Chinese hamster ovary (CHO) cells transfected with CHST4 or mouse CHST2 cDNAs do not express the MECA-79 epitope [33 ]. It remains possible that MECA-79 antigen expression requires a high threshold of sulfation that was not achieved in our cDNA-transfected cells or those of others [14 , 33 , 35 ]. Nonetheless, numerous studies have shown that MECA-79 antigen expression is neither sufficient nor required for vascular L-selectin ligand activity [12 , 14 , 46 ].
The current study of vascular L-selectin ligand sulfation complements recent studies focused on HEV-associated L-selectin ligands. Transfection of CHST1 and CHST4 cDNAs into CHO cells expressing CD34 and FucT-VII cDNAs generates ligands bound by L-selectin–IgM fusion proteins [29 ]. In this system, CHST1 and CHST4 synergize to enhance binding of this chimeric fusion protein [29 ]. Similarly, recombinant GlyCAM-1–IgG chimeras produced in COS cells overexpressing CHST1, CHST2, or CHST4 support increased L-selectin–dependent rolling of lymphocytes and Jurkat T cells in parallel-plate flow chambers by three- to sixfold [47 ]. In this system, COS cell transfection with a CHST1 cDNA generated optimal L-selectin ligands, although additive effects were not observed when the COS cells were cotransfected with combinations of CHST cDNAs [47 ]. In a separate study, CHST4 expression caused CHO cells transfected with CD34 and FucT-VII cDNAs to roll at slower velocities in parallel-plate flow chambers coated with an L-selectin–IgG fusion protein [33 ]. It was surprising that CHST2 expression by CHO cells transfected with CD34 and FucT-VII cDNAs did not affect rolling velocities in this assay system. These and other data led the authors of that study to propose that CHST4 is the only sulfotransferase relative to CHST1, CHST2, and CHST3 that generates L-selectin ligands such as 6-sulfo sLex in core 2-branched oligosaccharides attached to CD34 [33 ]. However, the results described above and in the current study, in combination with the broad expression patterns of the CHSTs, suggest that CHST1, CHST2, and CHST3 may also contribute to the generation of physiologic L-selectin ligands.
In summary, these results suggest that endothelial cells can use CHST1 and CHST2 for L-selectin ligand synthesis. Multiple CHSTs may be capable of generating functional ligands and overlapping enzymatic activities may be necessary for optimal ligand generation. Appropriate sialylation and fucosylation are also critical for the generation of HECA-452 mAb-defined sLex determinants that are essential components of vascular and leukocyte L-selectin ligands [14 , 35 ]. Increased CHST expression may synergize with increased FucT-VII mRNA expression in activated vascular endothelial cells [14 , 17 ] to generate optimal L-selectin ligands at sites of inflammation. In addition, functional vascular L-selectin ligands may require special protein scaffolds [14 , 37 ]. Therefore, regulation of sulfation, fucosylation, sialylation, and transcription of appropriate acceptor molecules provides critical points for modifying L-selectin ligand function. Each of these enzymatic pathways represents a potentially important target for therapeutic intervention.
Received August 21, 2000; revised November 6, 2000; accepted November 25, 2000.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(1,3) fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L-, E-, and P-selectin ligand biosynthesis Cell 86,643-653[Medline](1,3)fucosyltransferase Fuc-TVII in lymphoid aggregate high endothelial venules correlates with expression of L-selectin ligands J. Biol. Chem. 271,8250-82591
3 fusosyltransferase VII and newly cloned GlcNAcß:6-sulfotransferase cDNA Proc. Natl. Acad. Sci. USA 96,4530-4535This article has been cited by other articles:
 |
|
 |
|
  D. D. Marathe, E. V. Chandrasekaran, J. T. Y. Lau, K. L. Matta, and S. Neelamegham Systems-level studies of glycosyltransferase gene expression and enzyme activity that are associated with the selectin binding function of human leukocytes FASEB J, December 1, 2008; 22(12): 4154 - 4167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. da Costa Martins, J.-J. Garcia-Vallejo, J. V. van Thienen, M. Fernandez-Borja, J. M. van Gils, C. Beckers, A. J. Horrevoets, P. L. Hordijk, and J.-J. Zwaginga P-Selectin Glycoprotein Ligand-1 Is Expressed on Endothelial Cells and Mediates Monocyte Adhesion to Activated Endothelium Arterioscler. Thromb. Vasc. Biol., May 1, 2007; 27(5): 1023 - 1029. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
