LFA-1-Dependent HuR Nuclear Export and Cytokine mRNA
Stabilization in T Cell Activation1
Jin Gene Wang,* Mark Collinge,* Vinod Ramgolam,* Oran Ayalon,2* Xinhao Cynthia Fan,3†
Ruggero Pardi,‡and Jeffrey R. Bender4*
Lymphokine gene expression is a precisely regulated process in T cell-mediated immune responses. In this study we demonstrate
that engagement of the ?2integrin LFA-1 in human peripheral T cells markedly extends the half-life of TNF-?, GM-CSF, and IL-3
mRNA, as well as a chimeric ?-globin mRNA reporter construct containing a strongly destabilizing class II AU-rich element from
the GM-CSF mRNA 3?-untranslated region. This integrin-enhanced mRNA stability leads to augmented protein production, as
determined by TNF-? ELISPOT assays. Furthermore, T cell stimulation by LFA-1 promotes rapid nuclear-to-cytoplasmic trans-
location of the mRNA-stabilizing protein HuR, which in turn is capable of binding an AU-rich element sequence in vitro. Abro-
gation of HuR function by use of inhibitory peptides, or marked reduction of HuR levels by RNA interference, prevents LFA-1
engagement-mediated stabilization of T cell TNF-? or IFN-? transcripts, respectively. Thus, HuR-mediated mRNA stabilization,
stimulated by integrin engagement and controlled at the level of HuR nuclear export, is critically involved in T cell
activation. The Journal of Immunology, 2006, 176: 2105–2113.
lowing TCR engagement with antigenic peptides presented in the
context of APC MHC molecules. Additional receptors provide
TCR complementary signals essential for effective T cell activa-
tion, such that the full repertoire of T cell-mediated events can
occur. CD28, the most extensively characterized T cell costimu-
lation receptor, activates independent signaling pathways and con-
sequently synergizes with TCR signaling to enhance immune re-
sponses (1, 2). Other accessory T cell membrane molecules
include ?2integrin adhesion receptors, most notably LFA-1. By
engagement with ICAMs, LFA-1 provides a strong adhesive force
to promote T cell-APC conjugate formation and greatly stabilize
this interaction. In addition, LFA-1 has the ability to transduce a
variety of transmembrane signals, including calcium mobilization
(3), phospholipase C-?1 up-regulation (4), protein kinase C acti-
vation (5), and cytoskeletal rearrangement (6), all of which may
directly affect T cell activation. Recently, LFA-1 engagement has
been shown to impart potent “coactivation” (in cooperation with
lymphocytes are the central regulatory cells of the im-
mune response and require distinct signals for activation.
An Ag-specific signal is delivered through the TCR, fol-
CD3-TCR engagement) signals, with both common and distinct
properties as those achieved by CD28 (7–10).
Many critical T cell functions are mediated by cytokine production,
T cell cytokine transcripts are intrinsically unstable and, hence, pro-
duction of the soluble protein products is limited by rapid turnover of
their mRNA (13, 14). Several key cytokine mRNAs contain AU-rich
elements (ARE)5in their 3?-untranslated region (UTR). Conse-
quently, they are rapidly degraded after transcription. Induced resis-
tance to degradation, that is, mRNA stabilization, thus becomes a
crucial regulatory step in the control of cytokine production (15, 16),
the molecular basis for which is largely unknown. The costimulatory
signals provided by CD28 not only enhance induced cytokine gene
activation, but also have the ability to stabilize key cytokine tran-
scripts (17). However, the effect of adhesion receptor engagement on
T cell cytokine mRNA half-life has not been characterized. In our
recent T cell activation studies, we found that, along with TCR-CD3
engagement, the adhesion receptor LFA-1 provides coactivation sig-
nals resulting in the surface expression of the activation Ag urokinase
plasminogen activator receptor (uPAR) (18). uPAR mRNA is short-
lived and contains a class II AREs in its 3?-UTR, similar to cytokine
AREs LFA-1 engagement leads to uPAR mRNA half-life elongation,
as well as stabilization of a chimeric mRNA bearing the uPAR 3?-
UTR (19). These results led us to investigate whether LFA-1 coacti-
vation could stabilize T cell cytokine transcripts, which contain sim-
stabilization. In this study we show that LFA-1 engagement by mAbs
results in prolonged half-life of cytokine transcripts bearing typical
class II AREs (20). HuR, a constitutive nuclear protein highly ex-
pressed in the resting T cell nucleus, has specific ARE binding affinity
and has been reported to stabilize cytoplasmic mRNA (21, 22).
LFA-1, as well as CD28, stimulation resulted in a rapid nuclear-to-
cytoplasmic translocation of HuR, which is then capable of binding to
AREs. We discuss this stimulated nuclear export as a central feature
of T cell activation.
*Sections of Cardiovascular Medicine and Immunobiology, Vascular Biology and
Transplant Program, Boyer Center for Molecular Medicine, Raymond and Beverly
Sackler Foundation Cardiovascular Laboratory and†Department of Biophysics and
Biochemistry, Howard Hughes Medical Institute, Yale University School of Medi-
cine, New Haven, CT 06536; and‡Department of Molecular Pathology, Universita `
Vita-Salute School of Medicine, San Raffaele Scientific Institute, Milan, Italy
Received for publication July 6, 2005. Accepted for publication November 22, 2005.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported by National Institutes of Health Grant HL43331 and a
Raymond and Beverly Sackler Foundation Award (to J.R.B.) and by grants from
Associazione Italiana per la Ricerca sul Cancro and Telethon (to R.P.).
2Current address: OraDel Medical, Katzrin 12900, Israel.
3Current address: Department of Radiology, University of Michigan Hospital, Ann
Arbor, MI 48105.
4Address correspondence and reprint requests to Dr. Jeffrey R. Bender, Sections of
Cardiovascular Medicine and Immunobiology, Yale University School of Medicine,
The Anlyan Center S469, 333 Cedar Street, New Haven, CT 06520. E-mail address:
5Abbreviations used in this paper: ARE, AU-rich element; UTR, untranslated region;
uPAR, urokinase plasminogen activator receptor; siRNA, small interference RNA;
DAPI, 4?,6?-diamidino-2-phenylindole; AEC, 3-amino-9-ethyl carbazole.
The Journal of Immunology
Copyright © 2006 by The American Association of Immunologists, Inc.0022-1767/06/$02.00
Materials and Methods
After leukopheresis of healthy blood donors, human peripheral lympho-
cytes were freshly isolated, and purified T cells (?97% CD3?) were iso-
lated by negative immunoselection as has been described (18). The human
Jurkat T cell leukemia line was obtained from the American Type Culture
Collection. Purified murine anti-human CD18 mAb (clone TS1/18) and
murine anti-human CD3 mAb (clone UCHT-1) were purchased from En-
dogen and Immunotech, respectively. Murine anti-human CD11a mAb
(clone TS1/22) purified from ascites was generated in our laboratory (Yale
University, New Haven, CT). Purified murine anti-human CD28 mAb
(clone 9.3) was a gift from P. Linsley (Bristol-Myers Squibb, Seattle, WA).
Murine anti-human LFA-3 mAb (clone TS2/9) purified from ascites was
also generated in our laboratory. Rabbit anti-human HuR polyclonal Ab
was provided by J. Steitz (Yale University, New Haven, CT). Rabbit anti-
human AUF-1 was a gift of G. Brewer (Wake Forest University, Winston-
Salem, NC). Human TNF-? capture Ab, biotinylated anti-human TNF-?
detection Ab, and streptavidin-HRP were obtained from BD Pharmingen.
Recombinant human ICAM-1/Fc chimera and recombinant MCP-1 were
obtained from R&D Systems. Goat anti-human IgG, Fc?-specific was pur-
chased from Jackson ImmunoResearch Laboratories. Anti-human TATA
box-binding protein Ab (clone 17) was purchased from Transduction Lab-
oratories. DRB (5,6-dichloro-1-?-D-ribobenzimidazole) and cycloheximide
were purchased from Sigma-Aldrich. Plasmid pBBB was provided by J.
Belasco (Harvard Medical School, Boston, MA), into which was cloned the
Small interference RNA (siRNA) duplexes were synthesized by Qiagen.
The duplex specifically targeting HuR was 5?-aaGAGUGAAGGAGUU
GAAACU-3? and corresponds to nt 1135–1153 of the human HuR cDNA
sequence, within the 3?-UTR. The scrambled HuR control siRNA duplex
was 5?-aaGCCAAUUCAUCAGCAAUGG-3?, as described (23). Oligonu-
cleotide primers used for quantitative real-time PCR and peptides used for
blocking HuR translocation were synthesized by the W.M. Keck Biotech-
nology Resource Laboratory (Yale University). The primers used were as
follows: IFN-? sense 5?-GTCGCCAGCAGCTAAAACAGG-3? and anti-
sense 5?-TGCAGGCAGGACAACCATTACT-3?; TNF-? sense 5?-GTCA
CAA-3?; and GAPDH sense 5?-ACCAGCCCCAGCAAGAGCACAAG-3?
and antisense 5?-TTCAAGGGGTCTACATGGCAACTG-3?. The antenna-
pedia peptides AP-NES (nuclear export sequence) and AP-HNS (HuR nu-
cleocytoplasmic shuttling sequence), and scrambled control peptides used
in HuR translocation blocking experiments were as described (24).
T cell activation and transfection
For activations using Abs, cells were incubated with either single or com-
binations of the listed mAbs on ice for 20 min: anti-CD3 (0.1 ?g/107cells),
anti-CD11a (1.5 ?g/107cells), anti-CD18 (1.5 ?g/107cells), and anti-
CD28 (1.5 ?g/107cells). Excess Abs were washed out, and cells were
plated on goat anti-mouse Ab-coated plates at 4°C for 60 min. The bound
cells were cultured in medium at 37°C for the indicated time periods. For
mRNA degradation assays, DRB (0.2 mM) was added to the medium 3 h
after Ab cross-linking. For activation of cells using recombinant ICAM-
1-Fc, petri dishes were coated for 1 h with 10 ?g/ml goat anti-human
IgG-Fc in 50 mM Tris (pH 9.5), followed by blocking for 1 h with calcium/
magnesium-free PBS containing 2% dialyzed FBS. Dishes were then in-
cubated overnight at 4°C with the calcium/magnesium-free PBS containing
2% dialyzed FBS, which contained 100 ng/ml recombinant human
ICAM-1. Cells were resuspended at 4 ? 106cells/ml in LFA-1 activation
buffer (100 mM Tris-HCl (pH 7.5), 0.9% NaCl, 2 mM MnCl2, 2 mM
MgCl2, 5 mM D-glucose, 1.5% BSA) before adding to ICAM-1-coated
dishes. Cells were incubated for 45 min at 37°C/5% CO2before replacing
the LFA-1 activation buffer with warm RPMI medium/10% FBS with or
without PMA. For plasmid transfections, 108Jurkat T cells were incubated
at 37°C with 20 ?g of plasmid DNA in TS buffer (25 mM Tris-HCl (pH
7.4), 137 mM NaCl, 5 mM KCl, 0.7 mM CaCl2, 0.5 mM MgCl2) in the
presence of 200 ?g/ml DEAE-dextran for 20 min. Cells were washed, kept
in complete medium for 24 h, and serum-starved for another 24 h.
Electroporation of Jurkat cells with siRNA duplexes was performed
using 107cells/500 ?l in serum-free Opti-MEM I medium (Invitrogen Life
Technologies) containing 400 nM siRNA duplex. Cells were electropo-
rated at 500 ?F, 0.4 kV using a Bio-Rad Gene Pulser and immediately
transferred to 2.5 ml of RPMI 1640 medium (Invitrogen Life Technolo-
gies) containing 10% FBS. After 24 h, cells were added to 7 ml of RPMI
medium containing 10% FBS and allowed to incubate for an additional
24 h, at which time viable cells were recovered by centrifugation over a
cushion of Histopaque 1077 (Sigma-Aldrich) and subjected to a second
round of transfection. Experiments were performed 48 h after the second
round of transfection.
Northern blot analysis and RNase protection assay
A total of 15 ?g of total RNA was subjected to Northern blot analysis as
described (19). Normalization was performed by densitometric analysis of
the same filters hybridized with a probe for GAPDH. For RNase protection
assays, rabbit ?-globin, and chloramphenicol acetyltransferase probes were
synthesized by using a MAXI script in vitro Transcription kit (Ambion).
Full-length probes were gel-purified and hybridized with 15 ?g of total
RNA, using an RNase protection assay II kit (Ambion). Protected frag-
ments were separated on 5% acrylamide/8 M urea gels and quantitated by
densitometric analysis of ?-globin signal, normalized for chloramphenicol
Quantitative real-time PCR analysis of gene expression
Total RNA was isolated from cells using the RNeasy Mini kit (Qiagen) and
1 ?g reverse transcribed using an iScript cDNA synthesis kit (Bio-Rad),
according to the manufacturer’s protocols. The resulting cDNA template
was subjected to real-time PCR analysis by a Quantitect SYBR Green PCR
kit (Qiagen) using an Opticon DNA Engine 2 (MJ Research) and the fol-
lowing cycling parameters: 95°C/4 min; then 50 cycles of 95°C/30 s,
56°C/1 min, and 72°C/1 min. IFN-? or TNF-? mRNA levels were nor-
malized to GAPDH levels for each sample run in duplicate.
Cell fractions were obtained by resuspending harvested cells in buffer A
(10 mM Tris (pH 7.4), 5 mM MgCl2, 1.5 mM KOAc, and 2 mM DTT),
rocking at 4°C for 20 min followed by centrifugation at 14,000 rpm. Cy-
toplasmic extracts were collected from the supernatant. The nuclear pellets
were washed twice and resuspended in buffer B (20 mM HEPES (pH 7.9),
0.42 M KCl, 0.5 mM DTT, 0.2 mM EDTA, and 25% glycerol). Nuclear
lysates were centrifuged at 14,000 rpm, and nuclear proteins were collected
from the supernatant, followed by dialysis against buffer C (20 mM HEPES
(pH 7.9), 0.1 M KCl, 0.5 mM DTT, 0.2 mM EDTA, and 20% glycerol).
Total amount of protein was determined by the Bio-Rad Bradford method
and an equal amount of protein was subjected to SDS-PAGE. Membranes
were stained with anti-HuR Ab (dilution 1/3000) and anti-TATA box-
binding protein Ab (dilution 1/500), and signals were generated by the ECL
detection method (New England Biolabs).
Binding reaction and RNA mobility shift assay
RNA oligonucleotides (AUUU)5A, corresponding to a region of the GM-
CSF 3?-UTR, and control (AGGU)5A were synthesized (New England
Biolabs), and (AUUU)5A was labeled with [32P]ATP by T4 polynucleotide
kinase (New England Biolabs). Protein extracts were incubated with probes
at room temperature for 20 min in 20 ?l of buffer containing 5 ?g of yeast
RNA, 40 mM KCl, 10 mM HEPES (pH 7.9), 3 mM MgCl2, 1 mM DTT,
and 5% glycerol. The reaction mixtures were then separated by electro-
phoresis on nondenaturing 5% polyacrylamide gels containing 5% glycerol
in 0.25? TBE buffer (Tris-borate-EDTA) at 4°C.
Glass coverslips were coated with goat anti-murine Ig Ab overnight.
Monoclonal Ab-treated cells were loaded on the coverslips and incubated
at 4°C for 1 h. Adherent cells were cultured for indicated periods, fixed
with 3% paraformaldehyde, and permeabilized with 0.1% Triton X-100.
Cells were blocked with 10% normal goat serum, followed by staining with
purified anti-HuR (dilution 1/100) and cyanine 3-conjugated goat anti-
rabbit Ab (dilution 1/100; Jackson ImmunoResearch Laboratories) sequen-
tially. Cells were also costained with 0.0005% 4?,6?-diamidino-2-phenylin-
dole (DAPI; Sigma-Aldrich). Stained samples were visualized and
photographed on a Microphot immunofluorescence microscope (Nikon).
MultiScreen Immobilon-P 96-well plates (Millipore) were coated with hu-
man TNF-? capture Ab alone (5 ?g/ml), TNF-? capture Ab plus recom-
binant ICAM-1 (5 and 2 ?g/ml, respectively), or TNF-? capture Ab plus
recombinant MCP-1 (5 and 2 ?g/ml, respectively) by diluting in PBS,
adding 100 ?l/well, and incubating overnight at 4°C. Plates were washed
twice with assay diluent (PBS containing 10% FBS) and blocked for 2 h at
room temperature with assay diluent. Peripheral T cells were coated with
or without 0.002 ?g of anti-CD3 per 1 ? 106cells, as previously described,
and 1 ? 104cells in 100 ?l of complete culture medium added to each
well. Following incubation for 16 h at 37°C, cells were aspirated, the wells
2106 LFA-1 EFFECTS ON RNA-BINDING PROTEINS IN T CELLS
washed twice with distilled water then three times with buffer I (PBS con-
taining 0.05% Tween 20). One hundred microliters of biotinylated anti-
human TNF-? detection Ab (2 ?g/ml in assay diluent) was added to each
well, and incubated for 2 h. Wells were washed three times in buffer I, and
then 100 ?l of streptavidin-HRP (1/100 in assay diluent) added for 1 h.
After washing four times with buffer I followed by three times with PBS,
plates were developed with AEC substrate reagent (3-amino-9-ethyl car-
bazole; BD Pharmingen). Plates were read on a CTL automatic ELISPOT
reader and analyzed using Immunospot 3.1 software (CTL). All samples
were run in triplicate within each experiment.
LFA-1 engagement enhances cytokine mRNA induction
The short half-life of cytokine mRNAs are largely attributed to the
AU-rich sequences in their 3?-UTR. The AU-rich sequences in the
3?-UTR of TNF-?, GM-CSF, and IL-3 are shown in Fig. 1A. The
AREs in these transcripts contain multiple copies of a core
nonameric sequence UUAUUUA(U/A(U/A)) within a highly U-
rich region (13, 25, 26), which has been characterized as a class II
degradation sequence. To address the effect of LFA-1-mediated
coactivation on these cytokine transcripts, freshly isolated human
T cells were activated with an established immunoadherence pro-
tocol (19), using mAbs against the TCR-CD3 complex and LFA-1.
RNA harvested from resting, CD3-activated, or CD3-activated
plus LFA-1-activated cells was subjected to Northern blot analysis,
to assess the effect of these various conditions on steady-state cy-
tokine transcript levels (Fig. 1B). There was no detectable TNF-?,
GM-CSF, or IL-3 mRNA in quiescent T cells. In cells treated with
anti-CD3 mAb, there was a minor induction of all three cytokine
mRNAs at 2 h after treatment, and minimal detection at 11 h. In
contrast, cells coactivated with anti-LFA-1, in addition to anti-
CD3, accumulated higher levels of cytokine mRNA, which per-
sisted throughout the 11-h assay period. This suggests that LFA-1
engagement either increases the transcription rate or delays the
rapid decay of these cytokine mRNAs. In cells treated with anti-
LFA-1 alone, no detectable cytokine mRNA induction was ob-
served over this same time period (data not shown).
LFA-1 engagement stabilizes short-lived cytokine mRNAs with
consequent-enhanced protein production
Cytokine mRNA stability was measured using DRB, a specific
RNA polymerase II inhibitor, as a agent to arrest transcription
(27). In the singly CD3-activated samples, TNF-? mRNA levels
fell to background within 45 min of DRB addition (Fig. 2A). GM-
CSF and IL-3 mRNA levels fell to 24 and 33%, respectively (Fig.
2, B and C). In contrast, transcripts of all three cytokines were
remarkably stable in LFA-1-coactivated T cells, with no change in
mRNA levels during the 45 min post-DRB period. Similar results
were obtained using another transcription inhibitor, actinomycin D
(data not shown). Although an additional regulatory effect on tran-
scription was not evaluated and cannot be excluded, these findings
demonstrate that LFA-1 engagement stabilizes TNF-?, GM-CSF,
and IL-3 mRNA. This result was not generalized to all ARE-con-
taining transcripts, as LFA-1-mediated coactivation had little effect
on c-myc mRNA degradation (data not shown).
Although it would be most common for the noted enhanced
transcript stability and higher mRNA levels to correlate with
greater protein production, ELISPOT assays were performed to
document that correlation. Freshly isolated, human peripheral
blood T cells were activated with anti-CD3, following which they
were added to ELISPOT wells coated with anti-TNF-? capture Ab
alone, or with a combination of anti-TNF-? and recombinant
ICAM-1. Fig. 3 demonstrates that CD3 engagement alone results
in a mild increase in TNF-? production. However, with the addi-
tion of LFA-1 engagement (adhesion to recombinant ICAM-1),
there is an 8-fold increase in TNF-? production, compared with
CD3 stimulation alone. Recombinant MCP-1, which was added to
the wells in a similar fashion to ICAM-1 and used as a control
recombinant protein, failed to augment the small CD3-induced in-
crease in TNF-? production. The ELISPOT-determined protein
levels correlated with TNF-? mRNA stability, as in Fig. 2A. We
have previously demonstrated similar effects of LFA-1-mediated
coactivation on T cell IL-2 mRNA stability and consequent en-
hanced protein levels (7, 28). These data suggest that coactivation-
mediated stabilization of ARE-bearing cytokine transcripts via
LFA-1 does, in fact, result in increased production of functional
gene products, some of which are major mediators of adaptive
LFA-1 engagement stabilizes a chimeric class II-bearing
The typical class II ARE in cytokine mRNAs are important mRNA
destabilization cis-acting elements in 3?-UTRs. Insertion of the
GM-CSF ARE into the 3?-UTR of the rabbit ?-globin reporter
gene causes the otherwise stable mRNA to become highly unstable
in vivo (13). To assess the effect of LFA-1 engagement on a highly
unstable reporter transcript, a region of the GM-CSF 3?-UTR, se-
quence (AUUU)4A, corresponding to the second bolded GM-CSF
ARE highlighted in Fig. 1A, was introduced into the 3?-UTR of the
rabbit ?-globin gene with a c-fos serum-inducible promoter (29).
A, Human mRNA sequences in the 3?-UTR of TNF-?, GM-CSF, and IL-3.
The AU-rich sequences are presented with the nonameric sequences in
bold. B, Effect of LFA-1 engagement on cytokine mRNA induction. Hu-
man peripheral T cells were immunostimulated and bound via anti-CD3 or
anti-CD3 plus anti-LFA-1 mAbs for the indicated times. Total RNA was
harvested for Northern blot analysis, using probes for TNF-?, GM-CSF,
IL-3, and GAPDH. These results are representative of those obtained in
five separate experiments.
Effect of LFA-1 engagement on cytokine mRNA induction.
2107The Journal of Immunology
This GM-CSF ARE region contains two complete, overlapping
nonameric sequences typical of class II AREs. Human Jurkat T
cells, used in our previous study to demonstrate LFA-1-mediated
stabilization of endogenous uPAR mRNA (19), were transfected
with the chimeric pBBB-GM-CSF ARE reporter construct, serum-
starved for 24 h, then serum-repleted to transiently induce tran-
scription of the chimeric mRNA. The rapidly serum-induced
?-globin mRNA, normally stable for over 24 h, decayed to basal
levels within 6 h, as a consequence of the inserted GM-CSF ARE
(Fig. 4). However, LFA-1 stimulation markedly stabilized the un-
stable ?-globin chimeric mRNA, with minimal change in steady-
state levels at 6 h. Because a nuclear run-on experiment previously
excluded the possibility of LFA-1-mediated c-fos promoter trans-
activation (19), the sustained reporter mRNA level must be a result
of LFA-1 engagement imparting stabilization signals to overlap-
ping nonameric AREs.
LFA-1 and CD28 engagement promote HuR nuclear-to-
The stability of many mRNAs appears to be regulated by ARE
binding of proteins that either promote or inhibit degradation. This
idea raises the possibility that LFA-1 engagement generates sig-
nals that modulate ARE-binding protein interactions and, conse-
quently, control mRNA decay. The ELAV-like Hu proteins are
RNA-binding proteins that play a major role in the development of
the mammalian nervous system. One such family member, HuR, is
abundantly and constitutively expressed in the T cell nucleus, and
binds avidly to the (AUUU)4A sequence (30) as well as AREs in
c-fos and IL-3 (31).
HuR was recently defined as a nuclear-to-cytoplasmic shuttling
protein, with cytoplasmic HuR levels paralleling its stabilization
effect on a reporter construct. To evaluate whether LFA-1 signal-
ing provides a stimulus to HuR translocation, indirect immunoflu-
orescence microscopy was performed in LFA-1-, CD28-, and
LFA-3-engaged peripheral T cells. DAPI staining was performed
for nuclear definition. In unstimulated (data not shown) and LFA-
3-stimulated (Fig. 5A) cells, HuR remained exclusively nuclear.
F-actin staining confirmed that sufficient cell spreading had oc-
curred in the LFA-3-engaged cells (data not shown), such that
cytoplasmic HuR would have been easily detectable if present. In
LFA-1-activated cells, perinuclear HuR started to appear within 15
min of stimulation. At 45 min, HuR was diffusely distributed in a
punctuate staining pattern throughout the cytoplasm. LFA-1-stim-
ulated HuR cytoplasmic localization was not inhibited by cyclo-
heximide (data not shown), indicating that this induced compart-
mentalization is not a consequence of HuR neosynthesis, but rather
rapid transport from the nucleus. To address whether this feature of
T cell activation is specific to LFA-1, monoclonal anti-CD28 was
used in identical fashion. Fig. 5A demonstrates the same pattern of
HuR redistribution in CD28-stimulated cells. CD3 activation had
no effect on HuR distribution, nor did it modulate LFA-1-induced
translocation (data not shown).
To biochemically confirm this stimulated HuR redistribution,
nuclear and cytoplasmic cell fractions were carefully separated and
cytokine mRNA degradation. A–C, Human pe-
ripheral T lymphocytes were immunostimulated
and bound via either anti-CD3 or anti-CD3 plus
anti-LFA-1 mAbs for 3 h, after which the tran-
scriptional inhibitor DRB was added (t ? 0) and
RNA harvested at the indicated times for North-
ern analysis using the noted probes. The accom-
panying degradation curves represent measured
RNA densitometric units as a percentage of time
(at t ? 0) counts, normalized to GAPDH RNA
signals. These results are representative of those
obtained in three separate experiments.
Effect of LFA-1 engagement on
2108 LFA-1 EFFECTS ON RNA-BINDING PROTEINS IN T CELLS
HuR Western blot analysis performed. Exclusion of nuclear ma-
terial was determined by the absence of the nuclear Ag TATA
box-binding protein in cytoplasmic extracts (32) (Fig. 5B). As ex-
pected, nuclear HuR was present in untreated, LFA-3-, LFA-1-,
and CD28-activated cells. However, cytoplasmic HuR was easily
detectable only in extracts obtained from LFA-1- and CD28-stim-
ulated, but not LFA-3-engaged cells (Fig. 5B). These findings are
consistent with the immunofluorescent microscopic analysis and
confirm that activation through either LFA-1 or CD28 has the abil-
ity to promote nuclear export of HuR into the T cell cytoplasm.
Redistributed cytoplasmic HuR binds to class II AREs in vitro
To determine whether translocated, cytoplasmic HuR in LFA-1-
activated (or CD28-activated) T cells is functional to bind AREs,
purified cytoplasmic extracts were incubated with a32P-labeled
(AUUU)5A sequence (which contains three overlapping class II
nonamers), and run on a native acrylamide gel. This sequence is
identical with a portion of the GM-CSF 3? ARE shown in Fig. 1A,
and an identical GM-CSF-derived probe has been demonstrated to
bind HuR in T cell cytoplasmic extracts (33). Fig. 6A displays the
formation of an RNA protein complex using cytoplasmic extracts
obtained from LFA-1-, CD28-, but not LFA-3-activated cells. Pre-
incubation of the extract with anti-HuR Ab (Fig. 6, A and B), but
not anti-AUF-1 (another ARE-binding protein) (34) (Fig. 6B),
blocked the complex formation, demonstrating that HuR is the pre-
dominant induced protein contained within this complex. Excess
wild-type ([AUUU]5) unlabeled probe competitively inhibited bind-
ing of HuR to the labeled probe, whereas excess mutated ([AGGU]5)
probe did not block complex formation. These findings demonstrate
that the translocated HuR, mobilized by LFA-1 (or CD28) engage-
ment, has sequence-specific binding activity to AREs.
Blocking HuR nucleocytoplasmic shuttling inhibits LFA-1-
mediated cytokine mRNA stabilization
Given that LFA-1 engagement triggers HuR nuclear export and
stabilization of cytokine transcripts bearing potentially HuR-bind-
ing sequences, we hypothesized that blocking HuR translocation
would diminish the LFA-1 effects on mRNA stability. To this end,
two peptides known to inhibit HuR shuttling (24) were used. AP-
HNS represents a region of the shuttling domain of HuR recog-
nized by transportin 2, whereas AP-NES corresponds to a nuclear
Human peripheral T lymphocytes were treated with or without anti-CD3,
and 1 ? 104cells added per well of a 96-well plate, which had been coated
with anti-human TNF-? capture Ab, in the presence or absence of recom-
binant ICAM-1 or recombinant MCP-1 as a control. After 16 h, plates were
processed for detection of TNF-? protein secretion by immune ELISPOT.
The histograms represent data from two experiments, each run in triplicate,
and representative well images are shown at bottom.
Effect of LFA-1 engagement on TNF-? protein production.
bearing chimeric RNA. Jurkat T cells were cotransfected with the chimeric
rabbit ?-globin/GM-CSF 3?-ARE, and normalization chloramphenicol
acetyltransferase (CAT) constructs followed by serum induction in the ab-
sence (untreated) or presence (?LFA-1-treated) of anti-LFA-1 mAbs, after
which RNA was harvested for RNase protection assay. ?-Globin RNase
protection assay signals were densitometrically analyzed and represented
as the percentage of counts at 2-h time point, normalized to chloramphen-
icol acetyltransferase RNA signals.
Effect of LFA-1 engagement on an unstable, class II ARE-
HuR localization. A, Human peripheral T lymphocytes were immunostimu-
lated and adhered to glass coverslips via the indicated mAbs, and incubated
in medium at 37°C for 45 min, after which they were fixed and perme-
abilized. Immunofluorescent costaining was performed with rabbit anti-
HuR/goat anti-rabbit Ig-cyanine 3 and DAPI for nuclear definition. Indi-
vidual images were overlaid and merges displayed as noted. Magnification,
?1200. B, HuR and TATA box-binding protein (TBP) immunoblots were
performed on cytoplasmic and nuclear extracts obtained from LFA-3-,
LFA-1-, or CD28-stimulated (at 45 min) peripheral T cells.
Effect of LFA-1 and CD28 engagement on intracellular
2109 The Journal of Immunology
export signal recognized by the export protein CRM1. Both pep-
tides are made as fusion partners with the homeodomain of Dro-
sophila antennapedia protein to render them cell permeable. Si-
multaneous exposure of cells to both peptides has been
demonstrated to block HuR shuttling (24). The human T cell line
HSB-2 was used for these studies rather than human T cells or
Jurkat cells because we found it to be more resistant to the mild
documented cytotoxic effects of these peptides (24). HSB-2 cells
were pretreated with 1 ?M of each inhibitory peptide, or scram-
bled control peptides, before PMA activation for 3 h in the pres-
ence or absence of recombinant human ICAM-1. PMA was used,
rather than anti-CD3, in these experiments because HSB-2 mem-
brane CD3 levels are low, and thus PMA is a more efficient TNF-?
gene transcriptional activator. Following DRB-mediated transcrip-
tional arrest, TNF-? mRNA levels were determined by quantita-
tive real-time PCR. As shown in Fig. 7, ICAM-1 (i.e., LFA-1
engagement) induced stabilization of TNF-? mRNA in the pres-
ence of control peptides, relative to cells plated on poly-L-lysine.
However, this stabilization effect was greatly reduced in cells pre-
treated with HuR inhibitory peptides, clearly demonstrating a role
for functional HuR in LFA-1-mediated enhanced cytokine expres-
sion. Transcript levels at time t ? 0 were unaffected by the pair of
inhibitory peptides, demonstrating the lack of toxic effect. The
HuR inhibitory peptides, but not the control peptides, partially in-
hibited LFA-1-stimulated HuR nuclear-to-cytoplasmic transloca-
tion (data not shown). Although HuR nuclear-to-cytoplasmic shut-
tlingin response totranscriptional
actinomycin D) has been described in 3T3 cells (35), we have not
observed this phenomenon in T cells.
Knockdown of HuR expression inhibits LFA-1-mediated cytokine
To unequivocally determine whether HuR is required in LFA-1-
induced T cell cytokine mRNA stabilization we attempted specific
deletion of HuR in vitro using siRNA methodology. A specific
siRNA duplex was designed which is located within the 3?-UTR of
HuR mRNA, and showed no homology to other members of the
ELAV family of molecules. Using two rounds of transfection by
electroporation, depletion of ?90% of HuR protein was achieved
in Jurkat T cells, as determined by immunoblotting (Fig. 8A). A
control, scrambled siRNA duplex showed no reduction of HuR
protein compared with mock-transfected Jurkat cells.
class II ARE-binding activity in vitro. A, After 45 min stimulation with the
indicated mAbs, peripheral T cell cytoplasmic extracts were recovered and
incubated with the32P-labeled oligonucleotide (AUUU)5A, corresponding
to the GM-CSF 3?-ARE, in the absence (?) or presence of 250-fold excess
of unlabeled (AUUU)5A (WT) or of mutated (AGGU)5A (MU) oligonu-
cleotide. HuR denotes Ab blocking assay in which cytoplasmic extracts
were anti-HuR-treated before incubation with labeled probe. Reaction mix-
tures were separated on a nondenaturing polyacrylamide gel. B, Cytoplas-
mic extracts from LFA-1-activated T cells were pretreated with either no
Ab (?) or anti-HuR (?HuR) or anti-AUF-1 (?AUF1) Abs, then incubated
with32P-labeled oligonucleotide (AUUU)5A. Reaction mixtures were sep-
arated on a nondenaturing polyacrylamide gel.
Effect of LFA-1 and CD28 engagement on cytoplasmic
mediated TNF-? mRNA stabilization. HSB-2 cells were pretreated with 1
?M each of antennapedia peptides (AP-HNS and AP-NES) (Pep), 1 ?M
each control scrambled peptide (cPep), or with no peptide for 1 h at 37°C.
Cells were then plated onto dishes coated with either recombinant human
ICAM-1 (sICAM-1) or poly-L-lysine, and activated for 3 h with 320 nM
PMA, after which the transcriptional inhibitor DRB was added (t ? 0).
RNA was harvested from cells at the indicated times, and TNF-? mRNA
levels determined by quantitative real-time PCR, using GAPDH mRNA
levels as an internal control.
Effect of HuR inhibitory peptides on LFA-1 engagement-
2110LFA-1 EFFECTS ON RNA-BINDING PROTEINS IN T CELLS
Cell extracts were blotted in parallel with Abs to AUF-1, an-
other mRNA binding protein, and as shown in Fig. 8A, expression
of none of the four isoforms of AUF-1 was affected by siRNA
treatment, verifying the specificity of the duplex selected. In ad-
dition, surface expression of LFA-1, determined by flow cytom-
etry, was unaffected by siRNA treatment (data not shown). Jurkat
cells were used in these experiments because the efficiency of HuR
knockdown was much higher than that observed with either freshly
isolated, or HSB-2, T cells. The Jurkat line used expresses IFN-?
in response to anti-CD3 treatment, but is a poor producer of
TNF-?. We thus investigated IFN-?, another ARE-containing cy-
tokine transcript because we have previously described the LFA-
1-mediated mRNA stabilization for this transcript (19). Control
siRNA- and HuR siRNA-transfected cells were subjected to Ab-
mediated activation using either anti-CD3 alone or anti-CD3 plus
anti-LFA-1 mAbs, and IFN-? mRNA decay analyzed by quanti-
tative real-time PCR, following addition of DRB. As demonstrated
in Fig. 8B, depletion of cellular HuR abrogated the LFA-1-medi-
ated stabilization of IFN-? mRNA. These findings demonstrate
that HuR is a required component of LFA-1-mediated RNA
mRNA stabilization induced by LFA-1 engagement appears to be-
long to a general mechanism whereby accessory, coactivator, or
costimulator transmembrane signaling enhances T cell activation
responses. Of note is that the induced stabilization of these cyto-
kine mRNAs appears more dramatic than that observed for uPAR
(19). uPAR mRNA is intrinsically less unstable than the noted
cytokine transcripts, likely due to the greater number of class II
degradation motifs expressed in the cytokine 3?-UTRs. Thus, the
stabilization imparted through these motifs is quantitatively greater
for the less stable transcripts. LFA-1 coactivatory engagement
does not stabilize all labile activation transcripts, as this effect had
little impact on the half-life of induced c-myc mRNA. In contrast
with the vigorous class II degradation sequences within all those
cytokine transcripts studies, c-myc has scattered, pentameric se-
quences with typical class I ARE features. It is possible that LFA-
1-mediated signals, notably resulting in HuR translocation and
consequential RNA stabilization, are more effectively targeted to
class II rather than class I sequences. This idea seems unlikely, as
HuR has been shown, in fact, to bind the 3?-UTR of c-myc RNA
(36). Rather, there are multiple cis-elements within the c-myc tran-
script that are degradation targets and control its half-life, includ-
ing sequences within the 5?-UTR, as well as within exons 2 and 3
(37). Thus, the catabolic regulation of any given RNA species can
be complex. The regulatory mechanisms induced by coactivatory
or costimulatory engagement may affect only one component of
this control, as is likely the case for c-myc.
The possibility exists that both transcriptional and posttranscrip-
tional regulation are involved in coactivation-induced gene expres-
sion. A direct effect on transcription initiated at the IL-2 enhancer
was reported to be responsible for CD28 costimulation-enhanced
production of IL-2 mRNA (38). The LFA-1 or CD28 effect on
transcription may be dominant during the early phase of cytokine
mRNA accumulation, whereas the stabilization becomes more im-
portant when the rapid decay follows (16). LFA-1 engagement
may well influence enhanced gene expression at both levels.
Control of mRNA degradation is a complex process that can be
regulated, in part, by extracellular stimuli (39). The half-life of a
particular transcript can be a major determinant of the cellular
activation phenotype. The majority of existing studies on mecha-
nisms of RNA stability have been performed either in transfection/
overexpression or cell-free systems. Our T cell activation experi-
mental system provides a viable model with which to study
regulation of mRNA decay in a physiologic setting. In our previ-
ous work, we provided direct evidence that transmembrane sig-
naling via ?2integrin can alter T cell-activated gene expression
levels by attenuating the turnover of uPAR and IFN-? transcripts
(19). This finding is consistent with the emerging concept that a
major role of T cell coactivation and costimulation is to stabilize
effector cytokine mRNAs. CD28 engagement has been shown to
stabilize IL-2 (7), IFN-?, TNF-?, and GM-CSF RNAs (17), but
not other T cell activation transcripts without typical class II deg-
radation domains, such as c-myc and the IL-2R (40).
As noted, regulation of the stability of ARE-containing mRNAs
is complex and much remains to be elucidated. However, it is clear
that control of mRNA decay involves the interaction of mRNA
molecules with a number of regulatory proteins, including both
stabilizing factors, for example HuR, and destabilizing factors, for
example tristetraprolin and AUF-1. Although there are likely many
regulatory proteins, the roles of which remain to be determined,
HuR is emerging as a key ARE-binding, mRNA stabilizing pro-
tein. In contrast, tristetraprolin has been demonstrated to bind and
display destabilizing activity on transcripts for IL-3 (41) and
TNF-? and GM-CSF (42). In addition, other proteins are known to
bind the 3?-UTR of selective cytokine transcripts, consequently
regulating gene expression. For example, TIA-1 binds the 3?-UTR
ARE of TNF-? (43) and acts as a translational silencer, restricting
protein production from existing transcripts. The TNF-? 3?-UTR
also contains other non-ARE regulatory elements (44, 45), includ-
ing a constitutive decay element located 80 nt downstream of the
ARE and targets the mRNA transcript for rapid decay.
IFN-? mRNA stabilization. A, Jurkat cells were transfected with an siRNA
duplex specifically targeting HuR, a control scrambled duplex, or with no
siRNA, and 5 ?g of cell protein extract subjected to immunoblotting with
either anti-HuR or anti-AUF-1 Abs. B, Cells transfected as in A were im-
munostimulated and bound via either anti-CD3 (?CD3) or anti-CD3 plus anti-
LFA-1 mAbs. Following activation, DRB was added (t ? 0), RNA harvested
at the indicated times, and IFN-? mRNA levels determined by quantitative
real-time PCR, using GAPDH mRNA levels as an internal control.
Effect of HuR depletion on LFA-1 engagement-mediated
2111The Journal of Immunology
Regulation of these complex interactions is likely to be tran-
script-, stimulus-, and cell-specific. It is possible that signaling
mediated by LFA-1 regulates stabilization of ARE-containing cy-
tokine transcripts not only through HuR, but also through modu-
lation of destabilizing factor activity. Although the signaling
mechanisms mediated by LFA-1, and the subsequent cytokine
mRNA-protein complex formation, are actively being investi-
gated, LFA-1-mediated HuR nuclear export is clearly a crucial step
in the stabilization process. It is noteworthy that LFA-1 engage-
ment also induces HuR translocation and TNF-? transcript stabi-
lization in cells of the monocyte/macrophage lineage (D. Smith, G.
Gao, and J. R. Bender, unpublished observations). Although the
transcriptional activation stimulus would be different, it is likely
that cell adhesion mediated through leukocyte integrins represents
a common mechanism of enhanced cytokine gene expression.
It has been suggested that the contributory roles of LFA-1 and
CD28 to T cell activation overlap but are qualitatively and quan-
titatively distinct. LFA-1 appears to facilitate T cell activation by
lowering the amounts of Ag necessary for activation, whereas
CD28 reduces the required number of triggered TCRs and allows
T cell activation by low affinity ligands (46). Our observations
suggest that these two T cell membrane molecules can serve sim-
ilar activating functions by promoting transport of RNA-binding
proteins and RNA stabilization. This redundancy may be critical in
settings in which Ag presentation is performed by semiprofes-
sional APCs, such as human endothelial cells, which have ample
levels of the ?2integrin ligand ICAM-1 (and ICAM-2), but are
B7-negative. Furthermore, although the final effector pathway de-
scribed in this work is common to CD28 and LFA-1 engagement,
this event may be achieved through qualitatively distinct signaling
pathways. This conclusion is suggested by our previous work dem-
onstrating the requirement for an intact actin-based cytoskeleton to
achieve LFA-1-, but not CD28-induced, stabilization of IL-2
mRNA in a cyclosporine-resistant manner (7). Finally, although
induced HuR translocation and cytokine mRNA stabilization is a
common feature of activation between LFA-1 and CD28, these
membrane receptors certainly can achieve distinct effects.
Several other components of heterogeneous nuclear ribonucle-
oprotein particles, such as A1 (47) and AU-A (48), have been
found to shuttle between the nucleus and cytoplasm. The potential
importance of coactivator-driven HuR transport is underscored by
the fact that this class of ELAV proteins does bind specifically to
ARE and can alter the fate of bound mRNAs (49–51). In the
pathologic setting of hypoxia, in which the production of vascular
endothelial growth factor is a major stimulus to angiogenesis, vas-
cular endothelial growth factor mRNA half-life is prolonged. HuR
can bind to an AU-rich sequence of vascular endothelial growth
factor mRNA, perhaps as a complex with poly(A)-binding protein-
interacting protein 2 (52) and when overexpressed, stabilizes vas-
cular endothelial growth factor mRNA in hypoxic conditions (51).
Our data are not only consistent with these previous descriptions of
HuR function, but also provide a potential mechanistic link between
cell activation pathways. That is, our results suggest that LFA-1-
driven (and CD28-driven) HuR translocation is a key element of the
underlying activation mRNA stabilization mechanism.
Posttranscriptional control of activation of gene expression re-
quires the dynamic regulation of RNA export and degradation.
Mammalian cells are thought to contain a very limited set of RNA
endonucleolytic enzymes (mRNases), with much less specific tar-
get sequence restriction than endo-DNases. Therefore, although
some mRNAs, especially those with AREs, may be inherently
more susceptible to RNase-mediated cleavage than others, the sta-
bility of many RNAs is likely determined by binding of proteins
that either positively or negatively modify the accessibility of the
target recognition site (53). Furthermore, although incompletely
understood, ribonucleoproteins of complex composition, rather
than naked RNAs, are nuclear export substrates and hence, this
result is another level of posttranscriptional gene regulation in
which RNA protein interactions must occur. The levels at which
HuR control the fate of activation transcripts are an area of active
investigation. We are presently studying the LFA-1-generated sig-
nals that promote HuR translocation, their biochemical and struc-
tural effects on HuR, and whether these alterations in HuR struc-
ture and function direct RNA export as well as stabilization.
The use of mitogenic levels of anti-CD3 in the activation of
mouse splenocytes has been demonstrated to significantly increase
HuR protein expression (35), whereas submitogenic levels do not,
unless a costimulus (CD28) is provided. These effects on HuR
protein level were observed over a period of 1–2 days, and likely
represent a change in the proliferation status of the cells. Indeed,
HuR translocates from the nucleus to the cytoplasm during the G1
phase of the cell cycle (35). Seko et al. (54) have also demon-
strated that TCR signaling can lead to HuR translocation to the
cytoplasm in mouse T cell lines, but these studies used quantities
of anti-TCR Abs consistent with the mitogenic doses used in other
studies (35, 55). Similarly, Raghavan et al. (33) proposed that HuR
is localized within the cytoplasm of unstimulated human T cells.
However, the purity of their cytoplasmic fractions was not rigor-
ously assessed. In contrast, in pure cytoplasmic fractions free of
nuclear material, from unstimulated T cells, we do not detect HuR
in the cytoplasm by immunoblotting or in intact cells by immu-
nofluorescence. In addition, CD3 activation alone does not lead to
the translocation of HuR from the nucleus to the cytoplasm under
our experimental conditions (data not shown), again in contrast to
the observations of Raghavan et al. (33). This difference between
our experiments and those earlier described is likely due to our
much lower, submitogenic concentrations of anti-CD3 used and to
the significantly shorter activation periods. The LFA-1-mediated
HuR translocation described in this study clearly requires no pri-
mary CD3 stimulus.
Relevant to our experimental system, signaling through TCR-
CD3 determines the primary specificity of an immune reaction,
and initiates transcription of specific T cell activation genes. In the
absence of coactivation, the induced mRNAs are rapidly degraded,
before reaching a functionally significant expression level. With
the conjugation of cooperative, accessory receptor ligand pairs, the
exact nature of which depends upon the interacting cell types,
functionally important mRNAs with different classes of ARE deg-
radation sequences are selectively stabilized, thereby bringing a
second level of specificity to T cell activation. Our data support a
model for LFA-1-regulated (and CD28-regulated) HuR-mediated cy-
tokine mRNA stabilization. Dissecting the roles that adhesion recep-
tors and classical costimulator molecules play in facilitating the as-
sembly of RNA protein complexes and transport, thereby promoting
stabilization of key T cell activation transcripts, will provide impor-
tant clues to the molecular basis of T cell effector responses.
We thank Joan Steitz and Imed Gallouzi for numerous helpful discussions
and valuable reagents. We express gratitude to Lynn O’Donnell for assis-
tance with cell culture, Rita Girdzis for performing leukopheresis, and
Dana Brenckle for manuscript assistance. We thank Wendy Walker and
Daniel Goldstein for assistance with ELISPOT experiments. We thank all
those who provided generous gifts of valuable reagents.
The authors have no financial conflict of interest.
2112 LFA-1 EFFECTS ON RNA-BINDING PROTEINS IN T CELLS
References Download full-text
1. Ledbetter, J. A., C. H. June, L. S. Grosmaire, and P. S. Rabinovitch. 1987.
Crosslinking of surface antigens causes mobilization of intracellular ionized cal-
cium in T lymphocytes. Proc. Natl. Acad. Sci. USA 84: 1384–1388.
2. Truitt, K. E., G. B. Mills, C. W. Turck, and J. B. Imboden. 1994. SH2-dependent
association of phosphatidylinositol 3?-kinase 85-kDa regulatory subunit with the
interleukin-2 receptor ? chain. J. Biol Chem. 269: 5937–5943.
3. Pardi, R., J. R. Bender, C. Dettori, E. Giannazza, and E. G. Engleman. 1989.
Heterogeneous distribution and transmembrane signaling properties of lympho-
cyte function-associated antigen (LFA-1) in human lymphocyte subsets. J. Im-
munol. 143: 3157–3166.
4. Kanner, S. B., L. S. Grosmaire, J. A. Ledbetter, and N. K. Damle. 1993. ?2-
integrin LFA-1 signaling through phospholipase C-?1 activation. Proc. Natl.
Acad. Sci. USA 90: 7099–7103.
5. Nakamura, S., and Y. Nishizuka. 1994. Lipid mediators and protein kinase C
activation for the intracellular signaling network. J. Biochem. 115: 1029–1034.
6. Kupfer, A., and S. J. Singer. 1989. The specific interaction of helper T cells and
antigen-presenting B cells. IV. Membrane and cytoskeletal reorganizations in the
bound T cell as a function of antigen dose. J. Exp. Med. 170: 1697–1713.
7. Geginat, J., G. Bossi, J. R. Bender, and R. Pardi. 1999. Anchorage dependence of
mitogen-induced G1to S transition in primary T lymphocytes. J. Immunol. 162:
8. Ni, H. T., M. J. Deeths, W. Li, D. L. Mueller, and M. F. Mescher. 1999. Signaling
pathways activated by leukocyte function-associated Ag-1-dependent costimula-
tion. J. Immunol. 162: 5183–5189.
9. Jenks, S. A., and J. Miller. 2000. Inhibition of IL-4 responses after T cell priming
in the context of LFA-1 costimulation is not reversed by restimulation in the
presence of CD28 costimulation. J. Immunol. 164: 72–78.
10. Salomon, B., and J. A. Bluestone. 1998. LFA-1 interaction with ICAM-1 and
ICAM-2 regulates Th2 cytokine production. J. Immunol. 161: 5138–5142.
11. Crabtree, G. R. 1989. Contingent genetic regulatory events in T lymphocyte
activation. Science 243: 355–361.
12. Janeway, C. A., Jr., and K. Bottomly. 1994. Signals and signs for lymphocyte
responses. Cell 76: 275–285.
13. Shaw, G., and R. Kamen. 1986. A conserved AU sequence from the 3? untrans-
lated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46:
14. Stoecklin, G., S. Hahn, and C. Moroni. 1994. Functional hierarchy of AUUUA
motifs in mediating rapid interleukin-3 mRNA decay. J. Biol. Chem. 269:
15. Wodnar-Filipowicz, A., and C. Moroni. 1990. Regulation of interleukin 3 mRNA
expression in mast cells occurs at the posttranscriptional level and is mediated by
calcium ions. Proc. Natl. Acad. Sci. USA 87: 777–781.
16. Umlauf, S. W., B. Beverly, O. Lantz, and R. H. Schwartz. 1995. Regulation of
interleukin 2 gene expression by CD28 costimulation in mouse T-cell clones:
both nuclear and cytoplasmic RNAs are regulated with complex kinetics. Mol.
Cell. Biol. 15: 3197–3205.
17. Raghavan, A., R. L. Ogilvie, C. Reilly, M. L. Abelson, S. Raghavan,
J. Vasdewani, M. Krathwohl, and P. R. Bohjanen. 2002. Genome-wide analysis
of mRNA decay in resting and activated primary human T lymphocytes. Nucleic
Acids Res. 30: 5529–5538.
18. Bianchi, E., E. Ferrero, F. Fazioli, F. Mangili, J. Wang, J. R. Bender, F. Blasi, and
R. Pardi. 1996. Integrin-dependent induction of functional urokinase receptors in
primary T lymphocytes. J. Clin. Invest. 98: 1133–1141.
19. Wang, G. J., M. Collinge, F. Blasi, R. Pardi, and J. R. Bender. 1998. Posttran-
scriptional regulation of urokinase plasminogen activator receptor messenger
RNA levels by leukocyte integrin engagement. Proc. Natl. Acad. Sci. USA 95:
20. Chen, C. Y., and A. B. Shyu. 1995. AU-rich elements: characterization and im-
portance in mRNA degradation. Trends Biochem. Sci. 20: 465–470.
21. Fan, X. C., and J. A. Steitz. 1998. Overexpression of HuR, a nuclear-cytoplasmic
shuttling protein, increases the in vivo stability of ARE-containing mRNAs.
EMBO J. 17: 3448–3460.
22. Peng, S. S., C. Y. Chen, N. Xu, and A. B. Shyu. 1998. RNA stabilization by the
AU-rich element binding protein, HuR, an ELAV protein. EMBO J. 17:
23. van der Giessen, K., S. Di-Marco, E. Clair, and I. E. Gallouzi. 2003. RNAi-
mediated HuR depletion leads to the inhibition of muscle cell differentiation.
J. Biol. Chem. 278: 47119–47128.
24. Gallouzi, I. E., and J. A. Steitz. 2001. Delineation of mRNA export pathways by
the use of cell-permeable peptides. [Published erratum appears in 2002 Science
296: 47.] Science 294: 1895–1901.
25. Lagnado, C. A., C. Y. Brown, and G. J. Goodall. 1994. AUUUA is not sufficient
to promote poly(A) shortening and degradation of an mRNA: the functional
sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol. Cell.
Biol. 14: 7984–7995.
26. Zubiaga, A. M., J. G. Belasco, and M. E. Greenberg. 1995. The nonamer
UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degra-
dation. Mol. Cell. Biol. 15: 2219–2230.
27. Raju, U., C. Koumenis, M. Nunez-Regueiro, and A. Eskin. 1991. Alteration of
the phase and period of a circadian oscillator by a reversible transcription inhib-
itor. Science 253: 673–675.
28. Geginat, J., B. Clissi, M. Moro, P. Dellabona, J. R. Bender, and R. Pardi. 2000.
CD28 and LFA-1 contribute to cyclosporin A-resistant T cell growth by stabi-
lizing the IL-2 mRNA through distinct signaling pathways. Eur. J. Immunol. 30:
29. Shyu, A. B., M. E. Greenberg, and J. G. Belasco. 1989. The c-fos transcript is
targeted for rapid decay by two distinct mRNA degradation pathways. Genes
Dev. 3: 60–72.
30. Myer, V. E., X. C. Fan, and J. A. Steitz. 1997. Identification of HuR as a protein
implicated in AUUUA-mediated mRNA decay. EMBO J. 16: 2130–2139.
31. Ma, W. J., S. Cheng, C. Campbell, A. Wright, and H. Furneaux. 1996. Cloning
and characterization of HuR, a ubiquitously expressed Elav-like protein. J. Biol.
Chem. 271: 8144–8151.
32. Burley, S. K. 1996. The TATA box binding protein. Curr. Opin. Struct. Biol. 6:
33. Raghavan, A., R. L. Robison, J. McNabb, C. R. Miller, D. A. Williams, and
P. R. Bohjanen. 2001. HuA and tristetraprolin are induced following T cell ac-
tivation and display distinct but overlapping RNA binding specificities. J. Biol.
Chem. 276: 47958–47965.
34. DeMaria, C. T., and G. Brewer. 1996. AUF1 binding affinity to A?U-rich ele-
ments correlates with rapid mRNA degradation. J. Biol. Chem. 271:
35. Atasoy, U., J. Watson, D. Patel, and J. D. Keene. 1998. ELAV protein HuA
(HuR) can redistribute between nucleus and cytoplasm and is upregulated during
serum stimulation and T cell activation. J. Cell Sci. 111: 3145–3156.
36. Lafon, I., F. Carballe `s, G. Brewer, M. Poiret, and D. Morello. 1998. Develop-
mental expression of AUF1 and HuR, two c-myc mRNA binding proteins. On-
cogene 16: 3413–3421.
37. Yeilding, N. M., and W. M. Lee. 1997. Coding elements in exons 2 and 3 target
c-myc mRNA downregulation during myogenic differentiation. Mol. Cell. Biol.
38. Fraser, J. D., B. A. Irving, G. R. Crabtree, and A. Weiss. 1991. Regulation of
interleukin-2 gene enhancer activity by the T cell accessory molecule CD28.
Science 251: 313–316.
39. Jackson, R. J. 1993. Cytoplasmic regulation of mRNA function: the importance
of the 3? untranslated region. Cell 74: 9–14.
40. Lindstein, T., C. H. June, J. A. Ledbetter, G. Stella, and C. B. Thompson. 1989.
Regulation of lymphokine messenger RNA stability by a surface-mediated T cell
activation pathway. Science 244: 339–343.
41. Stoecklin, G., X. F. Ming, R. Looser, and C. Moroni. 2000. Somatic mRNA
turnover mutants implicate tristetraprolin in the interleukin-3 mRNA degradation
pathway. Mol. Cell. Biol. 20: 3753–3763.
42. Carballo, E., W. S. Lai, and P. J. Blackshear. 2000. Evidence that tristetraprolin
is a physiological regulator of granulocyte-macrophage colony-stimulating factor
messenger RNA deadenylation and stability. Blood 95: 1891–1899.
43. Piecyk, M., S. Wax, A. R. Beck, N. Kedersha, M. Gupta, B. Maritim, S. Chen,
C. Gueydan, V. Kruys, M. Streuli, and P. Anderson. 2000. TIA-1 is a transla-
tional silencer that selectively regulates the expression of TNF-?. EMBO J. 19:
44. Hel, Z., S. Di Marco, and D. Radzioch. 1998. Characterization of the RNA bind-
ing proteins forming complexes with a novel putative regulatory region in the
3?-UTR of TNF-? mRNA. Nucleic Acids Res. 26: 2803–2812.
45. Stoecklin, G., M. Lu, B. Rattenbacher, and C. Moroni. 2003. A constitutive decay
element promotes tumor necrosis factor ? mRNA degradation via an AU-rich
element-independent pathway. Mol. Cell. Biol. 23: 3506–3515.
46. Bachmann, M. F., K. McKall-Faienza, R. Schmits, D. Bouchard, J. Beach,
D. E. Speiser, T. W. Mak, and P. S. Ohashi. 1997. Distinct roles for LFA-1 and
CD28 during activation of naive T cells: adhesion versus costimulation. Immunity
47. Pin ˜ol-Roma, S., and G. Dreyfuss. 1992. Shuttling of pre-mRNA binding proteins
between nucleus and cytoplasm. Nature 355: 730–732.
48. Katz, D. A., N. G. Theodorakis, D. W. Cleveland, T. Lindsten, and
C. B. Thompson. 1994. AU-A, an RNA-binding activity distinct from hnRNP A1,
is selective for AUUUA repeats and shuttles between the nucleus and the cyto-
plasm. Nucleic Acids Res. 22: 238–246.
49. Antic, D., and J. D. Keene. 1998. Messenger ribonucleoprotein complexes con-
taining human ELAV proteins: interactions with cytoskeleton and translational
apparatus. J. Cell Sci. 111: 183–197.
50. Jain, R. G., L. G. Andrews, K. M. McGowan, P. H. Pekala, and J. D. Keene.
1997. Ectopic expression of Hel-N1, an RNA-binding protein, increases glucose
transporter (GLUT1) expression in 3T3–L1 adipocytes. Mol. Cell. Biol. 17:
51. Levy, N. S., S. Chung, H. Furneaux, and A. P. Levy. 1998. Hypoxic stabilization
of vascular endothelial growth factor mRNA by the RNA-binding protein HuR.
J. Biol. Chem. 273: 6417–6423.
52. Onesto, C., E. Berra, R. Grepin, and G. Pages. 2004. Poly(A)-binding protein-
interacting protein 2, a strong regulator of vascular endothelial growth factor
mRNA. J. Biol. Chem. 279: 34217–34226.
53. Rajagopalan, L. E., and J. S. Malter. 1994. Modulation of granulocyte-macroph-
age colony-stimulating factor mRNA stability in vitro by the adenosine-uridine
binding factor. J. Biol. Chem. 269: 23882–23888.
54. Seko, Y., H. Azmi, R. Fariss, and J. A. Ragheb. 2004. Selective cytoplasmic
translocation of HuR and site-specific binding to the interleukin-2 mRNA are not
sufficient for CD28-mediated stabilization of the mRNA. J. Biol. Chem. 279:
55. Thompson, C. B., T. Lindsten, J. A. Ledbetter, S. L. Kunkel, H. A. Young,
S. G. Emerson, J. M. Leiden, and C. H. June. 1989. CD28 activation pathway
regulates the production of multiple T-cell-derived lymphokines/cytokines. Proc.
Natl. Acad. Sci. USA 86: 1333–1337.
2113The Journal of Immunology