| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
CUTTING EDGE |
*
Molecular Cardiobiology Program, Division of Cardiovascular Medicine, and the Raymond and Beverly Sackler Foundation Laboratory, Boyer Center for Molecular Medicine, Yale University School of Medicine, New Haven, CT 06536; and
Unit of Clinical Immunology, Department of Biological and Technical Research, San Raffaele Scientific Institute, Milan, Italy
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
-independent manner. Here, we demonstrate that NK cells induce
HLA-DR on the surface of a mutant cell line that is defective in
IFN--induced MHC class II expression. RNA analysis in these cells
and in a cell line that is defective in class II transactivator (CIITA)
demonstrates that NK cell-induced HLA-DR
mRNA expression is also
CIITA-independent. The Janus kinase-1-deficient cell line U4A expresses
HLA-DR mRNA in response to NK cell activation, and HLA-DR
promoter constructs transfected into these cells are induced by NK
cells but not IFN-. These data indicate that the IFN--independent
component of the target cell HLA-DR expression induced by lymphocyte
adhesion uses a signaling pathway that is distinct from the
IFN--dependent mechanism and also suggest that CIITA is not
required. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
and
lymphocyte adhesion (2, 3, 4). IFN- induction of MHCII requires new
protein synthesis, specifically of the class II transactivator (CIITA).
Mutations of CIITA have been identified in all MHCII-deficient cell
lines from bare lymphocyte syndrome complementation group A. CIITA
expression itself is induced by IFN- (5), apparently via the Janus
kinase (JAK)-1/STAT-1 pathway (6, 7, 8). Rather than binding to
specific DNA sequences of the MHCII promoters, CIITA appears to
interact directly with other transactivating factors (9, 10, 11).
NK cells are the most efficient cellular inducers of EC class II
molecules (3). Although NK cells are competent producers of
IFN-, NK-mediated HLA-DR induction at both the
transcriptional and membrane level can be achieved in an
IFN--independent manner (12). Direct NK cell to EC contact is
required, and the ß2 integrin/ICAM-1 adhesion pathway is
critical. Additionally, NK cells induce HLA-DR mRNA more rapidly
than does IFN- (12), indicating that new protein synthesis may not
be required.
Here, we provide evidence that the HLA-DR expression induced by NK
cells can be achieved not only in an IFN--independent manner,
but also in the absence of CIITA expression. HLA-DR mRNA and protein
were induced by NK cells in mutant cell lines that were defective in
MHCII and CIITA expression. Additionally, HLA-DR promoter constructs
and mRNA were induced in a cell line that was deficient in JAK-1, which
is a regulatory kinase that is essential for CIITA induction by
IFN-. Taken together, these results indicate a pathway of HLA-DR
induction by NK cells that is not only independent of the IFN-R, but
also of CIITA, which was previously believed to be essential for the
induction of MHCII molecules.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Anti-CD56 mAb was obtained from BioSource International
(Camarillo, CA), rIFN- was supplied by Collaborative Biomedical
Products (Bedford, MA), and neutralizing anti-IFN- was purchased
from Genzyme (Cambridge, MA). Oligonucleotide primers were obtained
from the W. M. Keck Foundation Biotechnology Resource Laboratory
at Yale University (New Haven, CT). All other Abs and
reagents used were as described previously (12).
Cell isolation and culture
NK cells were isolated from leukocyte-enriched products from single healthy adult donors; the products were obtained by leukopheresis that was performed at the Yale Pheresis Unit, as described previously (12). Immunodepletion (negative panning) led to NK cells that were >96% CD56+, CD16+, CD3-. Single-donor HUVECs were isolated and cultured as described previously (12) and used at less than passage 6. The HT-1080 cell line was obtained from American Type Culture Collection (Manassas, VA) (CCL-121) and maintained in Eagle’s MEM with nonessential amino acids and Earle’s balanced salt solution (Life Technologies, Gaithersburg, MD) supplemented with 10% FBS. The 2C4, G2A, G3A, and U4A cell lines were obtained from George R. Stark (Cleveland Clinic Foundation, Cleveland, OH) and maintained in high glucose DMEM (Life Technologies) supplemented with 10% FBS.
Membrane HLA-DR induction assays and flow cytometry
Flow cytometric analysis was performed as described previously (12, 13).
NK cell panning
After recovering the NK/target cell coculture from plates with trypsinization, the panning of NK cells away from target cells was performed by immunodepletion with anti-CD16, -CD45, and -CD56 mAbs as described previously (12). RNA was purified from target cells that had been separated from NK cells and was subjected to RT-PCR (see below) with primers specific for CD45 to ensure complete NK cell depletion.
RNA analysis
Total cellular RNA was isolated using TRIzol reagent (Life
Technologies) and the phase lock gel system (5 Prime–3 Prime, Boulder,
CO). RT-PCR was performed using 1 µg of total RNA, Superscript II
RNaseH- reverse transcriptase (Life Technologies), and
oligo(dT)15 primer (Promega, Madison, WI) according
to the manufacturers’ instructions. PCR was performed as described
previously (12) using the following primer pairs: HLA-DR as
described previously (12); human CIITA, nucleotides (nt) 2881–2901 and
3331–3351; human CD45, nt 224–244 and 1059–1079; and human
glyceraldehyde-3-phosphate dehydrogenase (GAPDH), nt 12–37 and
971–994. Ribonuclease protection assay (RPA) probes were generated
using a MAXIscript kit (Ambion, Austin, TX). Both CIITA (nt 3137–3496)
and HLA-DR (nt 119–592) fragments in pBluescript (Stratagene, La
Jolla, CA) were transcribed using T7 polymerase in the presence of
[-32P]uridine triphosphate (800 Ci/mmol) (DuPont New
England Nuclear, Wilmington, DE). RPA was performed using total
cellular RNA and a RPAII kit (Ambion).
Plasmid constructs
The HLA-DR promoter human growth hormone (HGH)
constructs that were kindly provided by Richard Flavell (Yale
University) have been used by us previously (12). The luciferase
reporter pGL2 control vector (Promega) was cotransfected as a
normalization control.
Transient transfections
Cells that had been plated to 50% confluence were transfected
for 8 h with 10 µg of each plasmid using a calcium phosphate
transfection protocol (14). After glycerol shock, cells were fed fresh
medium and left overnight to recover. Fresh medium, medium containing
100 U/ml of IFN-, or allogeneic NK cells were added and incubated
for 48 h. The HGH secreted into the medium was determined using a
125I-HGH radioimmunoassay kit (Nichols Institute, San Juan
Capistrano, CA), and the HGH that normalized to luciferase activity in
cell extracts was determined using a luciferase assay system (Promega)
and a Lumat LB 9501 (Berthold, Wildbad, Germany)
luminometer.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
-independent, adhesion-dependent
manner (12). To investigate the involved signaling pathways, we took
advantage of a series of cell lines that had been rendered
MHCII-deficient by chemical mutagenesis (15). The cell line 2C4, which
was derived from the human fibrosarcoma cell line HT-1080, was
mutagenized with ICR-191; clones were selected that do not
express surface MHCII molecules either at rest or in response to
IFN-. One such cell line, G2A, shows partial HLA-DR promoter
occupancy, and MHCII expression can be complemented by an
overexpression of CIITA (K. Wright and J. Ting, personal
communication). We treated 2C4 and G2A with either 250 U/ml of
IFN- or allogeneic NK cells (20:1 NK:responder ratio) for 72 h,
followed by membrane HLA-DR analysis by flow cytometry. Figure 1
shows that both IFN- and NK cells
induce the surface expression of HLA-DR in the parental cell line 2C4.
As described earlier, IFN- failed to induce HLA-DR in the G2A cell
line, but G2A retained NK cell responsiveness with a significant
induction of membrane HLA-DR when cocultured with NK cells. This
finding is consistent with previous observations of
IFN--independent, endothelial HLA-DR induction by NK cells and also
provided a system whereby this phenomenon could be analyzed at the
level of transcription.
|
gene
transcription is not directly and immediately induced by IFN-, but
does require new protein synthesis, specifically of CIITA (reviewed in
1 . The kinetics of EC HLA-DR mRNA induction by allogeneic NK
cells are faster than those by IFN- (12), which led us to
investigate whether newly synthesized CIITA is required for HLA-DR
induction by NK cells. We again took advantage of the class II-negative
G2A cell line and of an additional cell line, G3A, that had been
generated by the same means; G3A is CIITA and class II negative and is
complemented to wild-type IFN--mediated MHCII inducibility by
expression of rCIITA (16). HUVECs and 2C4, G2A, and G3A cells
were incubated for 8 h in the presence of either 100 U/ml of
IFN- or NK cells (12:1 NK:responder ratio). For the NK cell
treatments, the responder cells were pretreated with anti-IFN-R
Ab, and anti-IFN--neutralizing Abs were maintained in the
coculture. Following the treatments, NK cells were removed through Ab
panning; RNA was isolated from the target cells and analyzed for CIITA
and HLA-DR mRNA expression by RPA (Fig. 2A). CIITA and HLA-DR mRNA
was absent in control resting cells, except for a trace amount of
HLA-DR in the cell line 2C4. As expected, IFN- significantly
up-regulated HLA-DR and CIITA mRNA in HUVECs and 2C4. In the G2A and
G3A cell lines, there was a small but detectable level of HLA-DR
mRNA induction by IFN-; however, this level approximated the level
of basal expression in the 2C4 cell line. A low level of CIITA mRNA was
also detected in the cells that were treated with IFN-. All cell
lines that were treated with NK cells showed a dramatic induction of
HLA-DR mRNA over basal levels. In the 2C4, G2A, and G3A cell lines,
this increased expression of HLA-DR was not accompanied by a
parallel increase in CIITA mRNA; no CIITA signal was detectable, even
with prolonged exposure. There is a barely detectable CIITA mRNA band
in NK-treated HUVECs that is consistent with previous observations that
NK cells have an IFN--dependent component of MHCII induction (12).
However, the induction of HLA-DR in this sample exceeds that seen
for the IFN--treated sample, in which there is substantial induction
of CIITA mRNA. This induction is not due to rapid CIITA mRNA
turnover or to unusual kinetics of induction, as no significant
CIITA transcript levels were observed over a wide time range
(1–48 h) (data not shown). These results indicate that HLA-DR mRNA
can be induced in target cells by NK cells in a manner that is
independent of CIITA. To fully demonstrate the absence of
CIITA message in NK cell-induced target cells, the
more sensitive but less quantitative RT-PCR method was used. The
results are shown in Figure 2B and confirm the pattern of
CIITA and HLA-DR expression that was seen by RPA. No CIITA mRNA was
detected by PCR in 2C4, G2A, or G3A that had been treated with NK
cells. PCR using primers that were specific for CD45 was negative,
confirming that NK mRNA was absent from the samples. The
IFN--independent component of the NK-mediated endothelial class II
mRNA induction is variable, as described previously (12). Although
HLA-DR mRNA is induced by NK cells in the absence of CIITA in most
experiments, CIITA mRNA is sometimes noted depending upon the NK cell
and EC donors, since these experiments are conducted with single-donor
cells for both allogeneic NK cells and ECs. Figure 2C
demonstrates that the IFN--neutralizing Abs used in the NK cell
activation experiments completely inhibit EC CIITA and HLA-DR mRNA
induction by IFN- as determined by PCR. Adding dextran sulfate to
ECs suppresses the induction of CIITA and MHCII gene expression by
IFN- (17). We are able to reproduce these results and show that NK
cells bind to vascular ECs in the presence of dextran sulfate; in
addition, HLA-DR mRNA is induced (data not shown).
|
Since it appears that the IFN--independent induction of HLA-DR by
NK cells is also independent of CIITA, we investigated other molecules
involved in IFN- signaling. JAK-1 and JAK-2 associate directly with
the IFN-R; upon phosphorylation, these kinases are
able to phosphorylate and activate STAT-1 (reviewed in
21 . STAT-1 then translocates to the nucleus and binds a
-activation sequence (GAS) or GAS-like elements in the promoters of
IFN--activated primary response genes. At least one of the promoters
for CIITA contains a GAS-like element (19, 20). Suppression of
STAT-1 expression inhibits IFN--stimulated CIITA mRNA expression
(8). Since our coculture experiments used anti-IFN- and
anti-IFN-R Abs and since CIITA was not induced in target cells
that were treated with NK cells, we investigated whether JAK-1 was
required. The U4A cell line (22) is defective in IFN- signaling and
expresses truncated JAK-1 mRNA and protein (23). We used U4A in our
coculture system to determine whether JAK-1 was required to obtain
HLA-DR induction by NK cells. IFN- (100 U/ml) or NK cells (5:1
NK:responder ratio) were added to confluent cultures of the U4A cell
line or the parental cell line HT-1080 and incubated for 16 h.
Following the removal of NK cells by Ab-mediated panning, target cell
RNA was isolated and subjected to RT-PCR (Fig. 3A). No basal HLA-DR
message was detected in either cell line, but IFN- was able to
induce both CIITA and HLA-DR message in the parental cell line
HT-1080; neither message was detected in the JAK-1-deficient U4A.
However, NK cells induced HLA-DR mRNA in both the HT-1080 and U4A
cell lines; CIITA message was not detected in either case. It appears,
therefore, that HLA-DR mRNA can be induced by NK cells in the
absence of both CIITA and JAK-1, which indicates a completely divergent
pathway of IFN--independent induction of HLA-DR by NK cells. To
confirm this observation, a HGH construct containing a portion of the
HLA-DR promoter (-269 HLA-DR) was transfected into these
cell lines along with a luciferase normalization plasmid; cells were
then activated with either NK cells or IFN-. The results are
shown in Figure 3B. In both the HT-1080 and U4A cell lines,
there is always substantial basal HGH expression from this promoter
construct. In the HT-1080 cell line, IFN- and NK cells (5:1
NK:responder ratio) induced the promoter (6.2- and 3.6-fold,
respectively). However, in the U4A cell line, IFN- was unable to
induce the expression of this promoter construct; there was a 2.5-fold
induction with NK cells. These results are in agreement with the RT-PCR
data shown in Figure 3A.
|
in
target cells independently of CIITA using a pathway that is completely
divergent from IFN- induction, since Abs that neutralize IFN- and
also block IFN-R have no effect, and JAK-1 kinase, which is
essential for the IFN- response, can be circumvented. Whether
another trans-activating factor is induced by NK cell
activation and can act independently of CIITA or whether a more direct
signaling event is involved is under investigation. HLA-DR mRNA
induction is more rapid with NK cell stimulation compared with the
induction observed with IFN- stimulation (12). Perhaps a more direct
signaling event is involved, with posttranslational modification of a
preexisting but as yet unknown factor. Direct contact through
ß2 integrin/ICAM-1 adhesion is required to achieve
IFN--independent induction (12). ICAM-1 cross-linking can induce
tyrosine phosphorylation events (24, 25) and induce
MHCII membrane expression on A20 cells (26). NK cells may trigger
ICAM-1-mediated signal transduction via adhesion, resulting in
HLA-DR gene activation.
According to DNA footprinting analysis, the HLA-DR promoter in the
G3A cell line (CIITA-defective) is completely bare; however, this
promoter is partially occupied in the G2A cell line (K. Wright and J.
Ting, personal communication). Efforts are underway to compare
IFN-- and NK cell-induced footprints in these mutant cell lines as
well as in ECs in an effort to identify distinct regions of the
HLA-DR promoter that is involved in gene activation by NK cells.
CIITA knockout mice express MHCII molecules in the interdigitating
reticular cells of the thymus (27). An isotype-specific activator of
HLA-DQ that can act independently of CIITA has been proposed (28, 29).
There is precedent, therefore, for the endogenous expression
of MHCII molecules in the absence of CIITA. Although we
describe an IFN--independent pathway, accessory pathways of MHCII
induction by IFN- have been described previously (30) that involve
non-Janus protein tyrosine kinases. Given the easily
detectable HLA-DR gene activation in the U4A cell line, it is
possible that NK cell adhesion results in the activation of a similar,
non-JAK-mediated kinase pathway. It is also worth noting that CIITA has
a nt-binding motif that appears to be critical for its function
(31). Other GTP-binding proteins may be induced in ECs by
lymphocyte adhesion, by replacing CIITA in its protein-protein
interaction role, or through other common roles of GTP-binding
proteins, including signal transduction and protein transport.
Here, we provide the first evidence that inducible MHCII (HLA-DR)
molecules can be expressed independently of CIITA. While it is likely
that CIITA is required for IFN- induction of MHCII molecules (5), it
can no longer be viewed as an essential regulator of MHCII in all
cellular contexts.
|
Acknowledgments |
|---|
-signaling mutant cell lines and Kenneth Wright and Jenny Ting
for helpful discussions. We also thank Lynn O’Donnell for expert
technical assistance and both Louise Benson and Gwen Davis for cell
culture support. We are grateful to the Milford General Hospital
Delivery Room for providing umbilical cords and to the Yale Pheresis
Unit and Rita Girdzis for assistance with leukopheresis. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Mark Collinge, Boyer Center for Molecular Medicine, Yale University School of Medicine, 295 Congress Avenue, New Haven, CT 06536-0812. E-mail address:
3 Abbreviations used in this paper: MHCII, MHC class II; EC, endothelial cell; CIITA, class II transactivator; JAK, Janus kinase; nt, nucleotide(s); GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RPA, ribonuclease protection assay; HGH, human growth hormone; GAS, -activation sequence.
Received for publication April 17, 1998. Accepted for publication June 19, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
mediated by the transactivator gene CIITA. Science 265:106. expression is involved in IFN- induction of the class II transactivator and class II MHC genes. J. Immunol. 157:1559.[Abstract]
-dependent and -independent mechanisms. J. Immunol. 154:3222.[Abstract]
. Proc. Natl. Acad. Sci. USA 90:2880. mutants reveals that defects in CIITA or RFX result in defective class II MHC and Ii gene induction. Immunity 1:687.[Medline]
-induced MHC class II gene expression is suppressed in endothelial cells by dextran sulfate. J. Immunol. 157:864.[Abstract]
receptor: a paradigm for cytokine receptor signaling. Annu. Rev. Immunol. 15:563.[Medline]
and interferons. Proc. Natl. Acad. Sci. USA 88:11455./ß and - signal transduction. Nature 366:129.[Medline]
action. J. Immunol. 158:1095.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  S.-D. Chou, A. N. H. Khan, W. J. Magner, and T. B. Tomasi Histone acetylation regulates the cell type specific CIITA promoters, MHC class II expression and antigen presentation in tumor cells Int. Immunol., November 1, 2005; 17(11): 1483 - 1494. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Muczynski, D. M. Ekle, D. M. Coder, and S. K. Anderson Normal Human Kidney HLA-DR-Expressing Renal Microvascular Endothelial Cells: Characterization, Isolation, and Regulation of MHC Class II Expression J. Am. Soc. Nephrol., May 1, 2003; 14(5): 1336 - 1348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Li, M. P. Haynes, and J. R. Bender Plasma membrane localization and function of the estrogen receptor alpha variant (ER46) in human endothelial cells PNAS, April 15, 2003; 100(8): 4807 - 4812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Johnson Locus-Specific Constitutive and Cytokine-Induced HLA Class I Gene Expression J. Immunol., February 15, 2003; 170(4): 1894 - 1902. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Jazaeri, C. J. Yee, C. Sotiriou, K. R. Brantley, J. Boyd, and E. T. Liu Gene Expression Profiles of BRCA1-Linked, BRCA2-Linked, and Sporadic Ovarian Cancers J Natl Cancer Inst, July 3, 2002; 94(13): 990 - 1000. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Magner, A. L. Kazim, C. Stewart, M. A. Romano, G. Catalano, C. Grande, N. Keiser, F. Santaniello, and T. B. Tomasi Activation of MHC Class I, II, and CD40 Gene Expression by Histone Deacetylase Inhibitors J. Immunol., December 15, 2000; 165(12): 7017 - 7024. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Harton and J. P.-Y. Ting Class II Transactivator: Mastering the Art of Major Histocompatibility Complex Expression Mol. Cell. Biol., September 1, 2000; 20(17): 6185 - 6194. [Full Text]
|
|
|
|
 |
|
 R. Chiari, G. Hames, V. Stroobant, C. Texier, B. Maillère, T. Boon, and P. G. Coulie Identification of a Tumor-specific Shared Antigen Derived From an Eph Receptor and Presented to CD4 T Cells on HLA Class II Molecules Cancer Res., September 1, 2000; 60(17): 4855 - 4863. [Abstract] [Full Text]
|
|
|
|
|
|
Y. Itoh-Lindstrom, J. F. Piskurich, N. J. Felix, Y. Wang, W. J. Brickey, J. L. Platt, B. H. Koller, and J. P.-Y. Ting Reduced IL-4-, Lipopolysaccharide-, and IFN-{gamma}-Induced MHC Class II Expression in Mice Lacking Class II Transactivator Due to Targeted Deletion of the GTP-Binding Domain J. Immunol., September 1, 1999; 163(5): 2425 - 2431. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
