In support of this hypothesis, we found that CB naive CD4 T cells had reduced activation and impaired early Th1 differentiation compared with adult peripheral blood naive CD4 T cells aft
Trang 1Impaired Allogeneic Activation and T-helper 1
Differentiation of Human Cord Blood Naive
CD4 T Cells
Li Chen, Aileen C Cohen, David B Lewis
Department of Pediatrics and the Immunology Program, Stanford University School of Medicine,
Stanford, California
Correspondence and reprint requests: D B Lewis, MD, Division of Immunology and Transplantation Biology, Department of Pediatrics, Stanford University School of Medicine, CCSR Bldg, Room 2115b, 269 Campus Dr, Stanford, CA 94305 (e-mail: dblewis@stanford.edu).
Received June 25, 2005; accepted October 22, 2005
ABSTRACT
CD4 T cells, particularly those of the T-helper 1 (Th1) subset, are important effectors in alloimmune diseases, such as graft-versus-host disease, and in controlling infections with intracellular pathogens Thus, it is plausible that impaired neonatal CD4 T-cell immunity might contribute to the low incidence of acute graft-versus-host disease after allogeneic transplantation of hematopoietic stem cells using cord blood (CB) compared with adult sources of hematopoietic stem cells In support of this hypothesis, we found that CB naive CD4 T cells had reduced activation and impaired early Th1 differentiation compared with adult peripheral blood naive CD4 T cells after stimulation by allogeneic dendritic cells derived from adult monocytes Early Th1 polarization was dependent on interleukin-12 and CD154, and CB CD4 T cell/dendritic cell co-cultures had impaired expression of both proteins CB naive CD4 T cells had low basal levels of signal transduction and activation of transcription 4 messenger RNA and protein, and, after alloantigen stimulation, reduced inter-leukin-12-induced signal transduction and activation of transcription 4 tyrosine phosphorylation, compared with adult peripheral blood naive T cells Lastly, FoxP3 protein expression, a marker for regulatory CD25 high
CD4 T cells, was lower for naive CD4 T cells of CB compared with those of adult peripheral blood, which argued against increased T-regulatory activity as a mechanism for the decreased Th1 differentiation of CB CD4 T cells Together, these intrinsic limitations in T-cell activation and Th1 differentiation may impair the ability of T cells in CB and the neonate to respond to allogeneic or infectious challenges.
© 2006 American Society for Blood and Marrow Transplantation
KEY WORDS
Cord blood ● Naive CD4 T cell ● Dendritic cell ● Allogeneic activation ● T-helper 1 cells
INTRODUCTION
Acute graft-versus-host disease (GVHD), in which
immunocompetent transplanted donor T cells
recog-nize alloantigens presented by host cells[1,2], remains
a major barrier to successful transplantation of
alloge-neic hematopoietic stem cells After their infusion,
allogeneic donor T cells predominantly migrate to
lymphoid tissues where they become activated by
in-teracting with host or donor antigen-presenting cells
[3], particularly the dendritic cell (DC) [4] Recent
studies suggest that CD4 antigenically naive rather
than memory T cells are mainly activated in the re-cipient and are critical in initiating acute GVHD in vivo[5-7] Activated CD4 T cells then travel to non-lymphoid tissues, such as the skin, gastrointestinal tract, and liver, followed by CD8 T cells, where they cause tissue damage[7] as effector CD4 and CD8 T cells producing proinflammatory cytokines, such as interferon (IFN)-␥ and tumor necrosis factor-␣, and cytotoxins [8] The importance of effector CD4 T cells of the T-helper 1 (Th1) type, which produce IFN-␥ but not T-helper 2 cytokines, in the immuno-pathogenesis of acute GVHD, especially for intestinal pathology [9], is supported by a number of animal studies[1]
Li Chen and Aileen Cleary Cohen contributed equally to this report.
doi:10.1016/j.bbmt.2005.10.027
Trang 2Allogeneic hematopoietic stem cell
transplanta-tion using umbilical vein cord blood (CB) is associated
with a decreased incidence and severity of acute
GVHD compared with transplantation using adult
peripheral blood (APB) or bone marrow cells[10,11]
Because CB contains a substantially higher
concentra-tion of naive T cells than either APB [12] or bone
marrow[13], these findings suggest that transplanted
CB naive T cells have a reduced capacity to mediate
alloimmune responses in vivo However, the
mecha-nisms that are responsible for decreased GVHD after
allogeneic CB transplantation remain unclear
Th1 effector cells are important for the control of
intracellular microbes, particularly bacteria[14]and
con-tributing to the immunopathology of acute GVHD[1]
The activation and differentiation of naive CD4 T cells
into Th1 effectors is initiated by engagement of
␣-T-cell receptor (TCR) with antigenic peptide/major
histocompatibility complex (MHC) class II complexes
[15] DCs express high levels of MHC class II and
costimulatory ligands, such as CD80 and CD86, and
are essential for activating naive CD4 T cells in
re-sponse to foreign antigens[16]and MHC alloantigens
in vivo [4] DCs promote Th1 differentiation by
se-creting cytokines, such as interleukin (IL)-12,
espe-cially after engagement of their CD40 molecule by
CD154 (CD40-ligand) on activated naive CD4 T cells
[17] Activated naive CD4 T cells also secrete IL-2
and acquire high-affinity receptors for 2 and
IL-12, which results in IL-2- and IL-12-dependent T-cell
proliferation [15,17] and facilitation of IFN-␥ gene
transcription[18]
The biologic effects of IL-12 are mainly
depen-dent on tyrosine phosphorylation of signal transducer
and activator of transcription 4 (STAT4) [19] that
occurs after IL-12 receptor engagement by its ligand
Key STAT4-dependent events include increased CD4
T-cell expression of IFN-␥ [20], which, in turn, may
enhance CD4 T-cell expression of T-bet, a transcription
factor that strongly favors Th1 and inhibits T-helper 2
development [21] IFN-␥ produced early after naive
CD4 T-cell activation may also favor Th1
differenti-ation by helping to maintain functional IL-12
recep-tors on the T-cell surface[20]
Regulatory CD4 T cells (Tregs) comprise a
well-characterized population of cells that express high
levels of CD25 and play an important role in the
negative control of inflammation and in the
mainte-nance of tolerance to self [22,23] Both human and
murine Tregs express high levels of the forkhead
fam-ily of transcription factors member FoxP3, whose
ex-pression is essential for programming Treg cell
develop-ment and function[24,25] CB from term pregnancies
has been shown to have equivalent numbers of CD4 T
cells that are CD25high as APB[26] These CD25high
CB T cells are able to suppress antigen-specific T-cell
responses in vitro[27,28], although to a lesser extent than CD25high CD4 T cells of APB[29]
The possibility that decreased acute GVHD seen
in CB transplantations might be caused by impaired Th1 differentiation of CB CD4 T cells is currently not resolved There is evidence that CD154 induction may be reduced on CB-derived T cells versus those derived from adult blood sources [30-33] However, others have demonstrated equivalent expression of CD154 after stimulation [34-36] Although some studies show equivalent levels of IFN-␥ by CB and APB T cells after stimulation [34,35], others have shown impaired IFN-␥ production[37,38], depending
on the conditions used for stimulation [39] Impor-tantly, none of the above studies analyze the early events of Th1 differentiation using purified CD4 naive
T cells and allogeneic stimulation with mature DCs, the interaction thought to initiate acute GVHD[1]
We show here that naive CD4 T cells of CB compared with those of APB have impaired early activation, IFN-␥ expression, and IL-12 responsive-ness after allogeneic stimulation with DCs, and an impaired ability to induce IL-12 production by DCs Thus, multiple intrinsic limitations in CB CD4 T-cell activation, which also limit DC function, impair the differentiation of this CD4 T-cell population into effector Th1 cells, and these limitations account, in part, for the decreased risk of acute GVHD after CB transplantation We further found that FoxP3 expres-sion, which is strongly associated with Treg cell func-tion, is decreased in CB naive CD4 T cells that are CD25high, suggesting that these limitations in Th1 differentiation are an intrinsic property of naive CB CD4 T cells rather than a result of an increase in the proportion of naive cells that are regulatory T cells
MATERIALS AND METHODS
CD4 T-Cell Purification
Naive (CD45RAhighCD45R0low) CD4 T cells were purified by negative selection from APB or CB mono-nuclear cells of full-term neonates prepared by Ficoll-Hypaque density gradient centrifugation of anticoagu-lated blood The mononuclear cells were incubated with RosetteSep Cocktail for CD4 T cells (Stemcell Technologies, Vancouver, British Columbia, Canada), followed by incubation with paramagnetic beads coated with monoclonal antibodies (mAbs) against non-T-cell markers, CD8-␣, and CD45R0 (Miltenyi, Auburn, Calif), and application to a magnetic-acti-vated cell sorter Final purity as assessed by flow cy-tometry was routinely more than 95%
Monocyte-Derived DCs
Monocytes were isolated from APB mononuclear cells using CD14 mAb-coated paramagnetic beads
Trang 3(Miltenyi) and positive selection using the
magnetic-activated cell sorter Monocytes were cultured (5.0⫻
105/mL) for 7 days in complete RPMI-1640 medium
[40]with 10% heat-inactivated human AB serum plus
recombinant human granulocyte
macrophage-colony-stimulating factor (25 ng/mL) (Immunex Corp, Seattle,
Wash) and recombinant IL-4 (20 ng/mL) (Peprotech,
Rocky Hill, NJ) Recombinant human tumor necrosis
factor-␣ (10 ng/mL) (Biosource, Camarillo, Calif) was
added after day 7, and mature monocyte-derived DCs
were harvested on day 9 and cryopreserved in 7% (vol/
vol) dimethylsulfoxide (Sigma Chemical, St Louis,
Mo) in heat-inactivated human AB serum Unless
oth-erwise indicated, equal numbers of DCs from 3
unre-lated donors were pooled together for allogeneic
stim-ulation
Allogeneic Stimulation of Naive CD4 T Cells
Equal numbers (2⫻ 105-1⫻ 106at 1.5⫻ 106/mL)
of adult or CB-derived naive CD4 T cells were
stim-ulated with allogeneic DCs (1.5⫻ 105/mL) in 200L
of complete RPMI-1640 medium in 96-well
flat-bot-tom plates for 24 to 72 hours For intracellular
cyto-kine staining, brefeldin A (10g/mL) (Sigma
Chem-ical) was added for the last 4 hours of a 48-hour
stimulation
Primary Antibodies
Mouse-antihuman biotinylated CD154 (clone
24-31) (Ancell, Bayport, Minn); phycoerythrin
(PE)-Cy5-CD4 (clone S3.5) (Caltag, Burlingame, Calif);
PE-CD69 (clone CH4, Caltag); fluorescein isothiocyanate
(FITC)-CD45RA (clone L48) (BD Biosciences,
Moun-tain View, Calif); allophycocyanin-CD45RO (clone
UCHL1) (BD Biosciences); PE-CD25 (clone
CD25-3G10) (Caltag); and FITC-IFN-␥ (clone 25723.11)
(BD Biosciences) mAbs were used for flow cytometric
analysis and cell isolation Appropriate biotinylated or
fluorochrome-conjugated mouse isotype control
mAbs (Caltag) were used as negative controls
Uncon-jugated mouse antihuman CD154 (clone, 5C8, IgG2a)
and IL-12 (clone 24910.1, IgG1) (R and D Systems,
Minneapolis, Minn) mAbs or equivalent
concentra-tions of sodium azide-free murine IgG1 and IgG2a
isotype controls (Sigma Chemical) were used for
neu-tralization experiments Rabbit antihuman STAT4
and phosphotyrosine (pY)-STAT4 antisera, and
pre-immune sera were purchased from Zymed, South San
Francisco, Calif FoxP3 antibody and isotype control
were purchased from eBioscience, San Diego, Calif
Cytokine Assays
Cell culture supernatants were analyzed for
cyto-kine content using enzyme-linked immunoassay kits
for IFN-␥, IL-2, IL-4, IL-10 (OptEIA, BD
Bio-sciences), and IL-12 (Quantikine HS immunoassay
kit, R and D Systems) The lower limit of sensitivity for each of the cytokine enzyme-linked immunosor-bent assays were as follows: IFN-␥, 1.0 pg/mL; IL-2, 1.0 pg/mL; IL-4, 2.0 pg/mL; IL-10, 2.0 pg/mL; and IL-12, 0.5 pg/mL
Cell Staining and Flow Cytometric Analysis
Staining for cell surface proteins and intracellular cytokine staining, poststaining fixation, and flow cyto-metric analysis were performed as previously described [40,41] The number of cells analyzed per condition by flow cytometry was no less than 10,000 naive CD4 T cells, and was often between 20,000 to 50,000 events, depending on the experiment FoxP3 staining was accomplished using anti-FoxP3 antibodies directly conjugated to FITC versus isotype controls (both from eBiosciences) These were used in the intracel-lular staining assay following a protocol provided by the manufacturer
Real-Time Reverse-Transcriptase Polymerase Chain Reaction Analysis
Total RNA was isolated, converted to comple-mentary DNA, and polymerase chain reaction ampli-fied in triplicate using an ABI Prism 5700 instrument (Applied Biosystems, Foster City, CA) and SYBR green dye (Applied Biosystems, Foster City, CA) as previously described [40] Oligonucleotide primer pairs used for amplification were human STAT4 [nucleotide (nt)1387F 5=-TGCCTCTATGGCCTGACCAT-3= and nt1438R 5=-TCACCACAGGCAATGAGCTG-3=] and previ-ously reported[40]18S ribosomal RNA primers The cycle threshold (Ct) value, the calculated polymerase chain reaction cycle at which products first became detectable, was determined using software (ABI) Ct values for 18S ribosomal RNA, which assessed the effectiveness of reverse transcription and polymerase chain reaction amplification, were used for normaliza-tion of the STAT4 transcript Ct values Levels of STAT4 messenger RNA (mRNA) relative to those found in freshly isolated naive CB CD4 T cells were calculated from Ct values using the formula 2⫺⌬Ct Total RNA from CD69hi CD4 T cells was isolated after 24 hours of stimulation with allogeneic DCs by incubating cultured cells with PE-conjugated CD69 mAb, and positively selecting CD69hi cells using PE-specific paramagnetic microbeads (Miltenyi) and a magnetic-activated cell sorter The final CD69hi cell fraction contained less than 0.1% DCs based on light microscopic cellular morphology
STAT4 Flow Cytometric Assays
Intracellular total STAT4 staining was performed
as previously described[42] For pY-STAT4 staining, naive CD4 T cells were allogeneically stimulated for
48 hours and then incubated with 20 ng/mL of
Trang 4re-combinant IL-12 (Genetics Institute, Boston, Mass)
for 20 minutes at 37°C Cells were treated with
fixa-tion reagent A (Fix and Perm, Caltag) for 2 to 3
minutes at room temperature, followed by the
addi-tion of an equal volume of ice-cold methanol for 10
minutes Cells were washed, permeabilized with
re-agent B (Caltag), stained at room temperature with
primary STAT4 antibodies or preimmune rabbit sera
for 30 minutes, followed by incubation with
FITC-conjugated goat antirabbit IgG (Caltag) for 20 minutes,
and then immediately analyzed by flow cytometry
Statistics
Data are presented as mean⫾ SEM P values for
differences between means were calculated using the
2-tailed, unpaired Student t test.
RESULTS
Reduced Naive CB CD4 T-Cell Expression of CD69
and CD154 after Stimulation by Allogeneic DCs
To determine if CB and APB naive CD4 T cells
differed in their ability to respond to optimal antigenic
stimulation, these ells were co-cultured with mono-cyte-derived DCs from unrelated adult blood donors DCs were selected as allogeneic stimulators because of their physiologic relevance as antigen-presenting cells for naive CD4 T-cell activation in vivo[16], including
in the context of GVHD[4], and their potency as acti-vators, which enhanced our ability to analyze T-cell responses using flow cytometry Naive CD4 T cells co-cultured individually with 3 different allogeneic
DC lines were first analyzed for their expression of CD69 and CD154, which are induced early after ac-tivation [43,44] Each DC line induced a substantial proportion of APB naive CD4 T cells to express CD69 (Figure 1A) Pooling of the DC lines from 3 unrelated individuals consistently induced greater amounts of CD69 than any of the individual lines, consistent with the pool providing a greater number
of MHC class II alloantigenic determinants CB naive CD4 T cells under these conditions had a significantly lower percentage of CD69⫹cells than APB naive cells (CB and APB CD4 T-cell percentages of 6.60⫾ 1.35 and 13.08⫾ 2.68, respectively, n ⫽ 6, P ⫽ 005) The
amount of CD69 per cell was similar based on mean
Figure 1.Reduced activation of CB naive CD4 T cells by allogeneic DCs CD69 (A) and CD154 (B) expression on APB and CB naive CD4
T cells after incubation with or without allogeneic DCs for 24 hours Cells were analyzed by flow cytometric analysis for CD4, CD45RA, and CD154 or CD69 DC1, DC2, and DC3 represent DC lines derived from 3 unrelated healthy adult blood donors, respectively, and DC1 ⫹2⫹3 indicates pool of these 3 DC lines Events shown are gated for CD45RA⫹lymphocytes, with boxed numbers indicating percentage of positive cells Results shown are representative of at least 6 independent experiments.
Trang 5fluorescent index (MFI) measurements (CB and APB
CD4 T-cell MFI of 237⫾ 19 and 282 ⫾ 18,
respec-tively, P⫽ 12) In contrast, CB naive CD4 T cells had
a significant reduction in both the percentage of
CD154 positive cells (Figure 1B; CB and APB CD4
T-cell values of 2.9⫾ 0.4 and 5.9 ⫾ 0.7, respectively,
n⫽ 7, P ⫽ 001) and the level of CD154 per positive
cell (CB and APB CD4 T-cell MFI values of 134⫾ 18
and 325⫾ 52, P ⫽ 002) In subsequent experiments,
a pool of 3 DC lines was used to maximize the CD4
T-cell response to alloantigens, and to limit any
po-tential impact of chance matching of MHC class II
alleles between stimulator and responder cells
Reduced IFN- ␥, IL-2, and IL-12 Production by
Cultures of CB CD4 T Cells with Allogeneic DCs
The cytokine content of supernatants from co-cultures of CB or APB naive CD4 T cells with allo-geneic DCs was analyzed As expected, IFN-␥, IL-2, IL-4, IL-10, and IL-12 were not detected in superna-tants of naive CD4 T cells or DCs cultured alone (Figure 2A) Co-cultures of APB naive CD4 T cells with DCs produced substantially greater amounts of IFN-␥, IL-2, and IL-12 compared with those contain-ing CB CD4 T cells (Figure 2A) Peak levels of IL-2 and IFN-␥ were achieved at 48 and 72 hours of
incu-Figure 2.Reduced cytokine production by co-cultures of CB naive CD4 T cells with allogeneic DCs A, IFN-␥, IL-2, IL-4, IL-10, and IL-12 content in cell culture supernatants by enzyme-linked immunosorbent assay (ELISA) of naive CD4 T cells from APB (open bars) or CB (closed
bars) after 48 hours of co-culture with allogeneic DCs *P⬍ 05 compared with adult CD4 T-cell co-cultures B, Kinetics of IFN-␥ and IL-2 accumulation in supernatants of ABP (solid line) or CB (dashed line) naive CD4 T-cell co-cultures with allogeneic DCs by ELISA Results are representative of 3 independent experiments.
Trang 6bation, respectively, in APB or CB CD4 T-cell
co-cultures with DCs (Figure 2B) Substantially lower
levels of IFN-␥ were also observed in CB CD4 T
cell/DC cultures compared with cultures of ABP naive
CD4 T cell/DC cultures for at least 120 hours of
incubation (Figure 2B) This suggests that the events
of T-cell activation required for de novo IL-2 and
IFN-␥ mRNA expression by CD4 T cells, e.g.,
acti-vation of nuclear factor of activated T-cell proteins
[45], occurred in CB CD4 T cells without a major lag The delayed appearance of IFN-␥ relative to IL-2 that was observed was also consistent with most IFN-␥ expression, but not IL-2 expression, requiring chro-matin remodeling linked with cell division[18] Reduced intracellular accumulation of IFN-␥ by
CB naive CD4 T cells after co-culture with allogeneic DCs was also evident (Figure 3A), with both the fre-quency of IFN-␥⫹ cells and the amount of IFN-␥
Figure 3.IFN- ␥ production by naive CD4 T cells after allogeneic DC stimulation is dependent on IL-12 and CD154 A, APB or CB naive CD4 T cells were incubated with allogeneic DCs for 48 hours in presence or absence of 10 g/mL of neutralizing CD154 or IL-12 p70 mAb
or isotype-matched control mAbs Intracellular IFN- ␥ accumulation in APB (top) or CB (bottom) naive CD4 T cells is shown Events shown are CD45RA high lymphocytes, with boxed numbers indicating percentage and MFI of IFN- ␥ ⫹ cells B, IFN- ␥ or IL-12 levels in supernatants
of co-cultures of CD4 T cells and DCs AB or CB naive CD4⫹T cells (3 ⫻ 10 5 ) were incubated with allogeneic DCs (3 ⫻ 10 4 ) in presence
or absence of anti-CD154 antibody (5c8), or anti-IL-12 or their isotype control antibodies IgG2a, or IgG1, at 10 g/mL After 48 hours of culture, supernatants were harvested and measured for IFN-␥ or IL-12 content by enzyme-linked immunosorbent assay *P ⬍ 05 compared
with isotype control mAb incubation Results shown are representative of 3 independent experiments.
Trang 7produced per positive cell, based on MFI
measure-ments, being substantially lower compared with APB
naive CD4 T cells No IFN-␥ accumulation by DCs
could be detected in these co-cultures (data not
shown), although DCs are a potential source of this
cytokine [46] We were also unable to detect either
IL-2 or IL-12 expression by co-cultured cells using intracellular cytokine staining, most likely because of the relatively low levels of these proteins per cell However, CD4 T cells were likely the major source of IL-2 and DCs the major source of IL-12, based on previous studies by others[15,17]
Cultures of CB CD4 T Cells and Allogeneic DCs Produce IL-10 but Not IL-4
The production of IL-4 by either APB or CB CD4 T-cell co-cultures with DCs was low to undetectable (Figure 2A) Thus, there was no evidence for CB CD4
T cells being prone to T-helper 2 polarization after optimal stimulation with mature DCs, although this might still apply to less optimal activation conditions [47] APB and CB CD4 T-cell co-cultures with DCs both produced readily detectable and similar amounts of IL-10 (Figure 2A), which has multiple antagonistic ef-fects on the generation of Th1 immunity[48] As was the case for IL-2 and IL-12, intracellular staining for IL-10 was not detectable so that the relative contri-bution of DCs and CD4 T cells to IL-10 production could not be determined
Role of Decreased CD154 and IL-12 in Limiting
CB CD4 Th1 Polarization
Previous studies have shown that CD154 expres-sion and IL-12 are critical for the in vivo accumulation
of human Th1 memory cells [40,49] In co-cultures
of APB naive CD4 T cells with allogeneic DCs, we found that specific neutralization of either CD154 or IL-12 reduced the frequency of CD4 T cells that expressed IFN-␥ intracellularly (Figure 3A) and the level of IFN-␥ protein in cell culture supernatants (Figure 3B) and this occurred to a similar extent with either APB or CB-derived naive CD4 T cells IL-12 levels were also specifically and markedly decreased in cultures containing neutralizing CD154 mAb (Figure
Figure 4.Reduced STAT4 expression and IL-12-induced tyrosine phosphorylation of STAT4 by CB naive CD4 T cells A, STAT4 mRNA levels in CB and APB naive CD4 T cells immediately after isolation or in naive CD4 T cells expressing CD69 after 24 hours of alloantigenic stimulation Transcript level for freshly isolated CB naive CD4 T cells was arbitrarily assigned value of 1.0 for compar-ison Ct values were normalized using 18S ribosomal RNA as
internal control *P⬍ 05 compared with corresponding CB CD4 T-cell subset B, Intracellular STAT4 protein in APB (top) or CB naive (bottom) CD4 T cells, with number in left upper corner indicating percentage of cells staining positive by flow cytometry C, Flow cytometric analysis of intracellular pY-STAT4 expression by allogeneically primed CB and APB naive CD4 T cells after brief treatment with IL-12 Events shown are for CD45RA⫹ CD4 T cells, with numbers indicating percentage of positive cells For all panels, one of 3 independent experiments with similar results is shown.
Trang 83B), indicating that IL-12 production was dependent,
in part, on CD154 expression
Decreased STAT4 Expression and Activation
in Naive CB CD4 T Cells
We next determined if decreased expression of
STAT4, which plays a critical role in Th1
differenti-ation [20,21], might contribute to reduced CB CD4
T-cell IFN-␥ production after allogeneic stimulation
Freshly purified CB naive CD4 T cells had STAT4
transcript levels that were approximately 20-fold
lower than those of APB naive CD4 T cells (Figure
4A) These marked differences in STAT4 mRNA
ex-pression were also observed at the protein level (
Fig-ure 4B) To determine if STAT4 expression was
in-creased by allogeneic priming, total RNA from CB or
APB naive CD4 T cells that were CD69⫹ after
co-culture with allogeneic DCs was analyzed Although
priming consistently increased STAT4 mRNA
ex-pression for both CD4 T-cell types relative to their
basal levels, primed CB CD4 T cells still had reduced
levels of STAT4 transcripts (Figure 4A) and protein
(data not shown) compared with APB cells These
differences in STAT4 expression were of functional
sig-nificance, in that allogeneically primed CB naive CD4 T
cells had a substantially decreased expression of
pY-STAT4 after IL-12 treatment compared with similarly
cultured and IL-12-treated APB naive CD4 T cells (
Fig-ure 4C) No pY-STAT4 was detectable in cultured CD4
T cells in the absence of IL-12 treatment
Reduced FoxP3 Expression in CD25 high CB Naive
CD4 T Cells
CD25high CD4 T cells include a population of
Tregs that are known to be important in preventing
autoimmune disease, in mediating peripheral trans-plantation tolerance, and in dampening the immune response to endogenous bacterial flora and pathogens [22,23] Thus, it is plausible that these cells may also limit the activation of and cytokine production by naive CD4 T cells in acute GVHD To determine whether an increased proportion of Tregs among na-ive CD4 T cells in CB might limit their ability to differentiate into Th1 cells, we first examined the percentage of naive CD4 T cells that expressed CD25
in CB versus APB A slightly lower percentage of CB naive CD4 T cells expressed high levels of CD25 versus APB naive CD4 T cells (Figure 5A) Because FoxP3 expression has been strongly correlated to Treg function[24,25], we next examined FoxP3 protein level
in APB versus CB naive CD4 T cells Although a sub-stantial fraction of CD25highnaive CD4 T cells in both
CB and APB expressed FoxP3 protein, this percentage for APB CD25highCD4 naive T cells was almost twice that of the analogous CD25high CD4 naive T-cell population in CB (Figure 5B) Neither APB nor CB-derived CD4 T cells that were CD25⫺/lowexpressed FoxP3 protein (Figure 5C) Because FoxP3 expression
is thought to tightly correlate with Treg development and function, these data suggest that increased Treg activity within the CB naive CD4 T-cell population does not play a role in limiting activation and differ-entiation into Th1 cells compared with APB naive CD4 T cells
DISCUSSION
The transplantation of allogeneic CB is associated with a lower incidence and severity of acute GVHD compared with transplantation using allogeneic
pe-Figure 5.Decreased FoxP3 expression by naive CD4 T cells of the CD25 high subset from CB compared with APB Flow cytometric analysis
of naive (CD45RA high ) CD4 T cells from APB (top) and CB (bottom) are shown A, CD4 versus CD25 surface expression, with percentage CD25⫺/lowcells (low right quadrant) and percentage of CD25 high cells and MFI for CD25 of CD25 high population (upper right quadrant) indicated B, Intracellular FoxP3 expression in freshly isolated CD25 high naive CD4 T cells, with percentage of positive cells indicated Filled histogram (isotype control) is compared with solid line (FoxP3) C, Intracellular FoxP3 expression in freshly isolated CD25⫺/lownaive CD4
T cells Filled histogram (isotype control) is compared with solid line (FoxP3), and percentage of positive cells is indicated.
Trang 9ripheral blood or bone marrow from adults as sources
of hematopoietic stem cells [10,11] Recent studies
suggest that the naive rather than memory/effector
CD4 T cells are an important determinant of acute
GVHD [5-7] and, therefore, that naive T cells are
important precursors of effector cells producing Th1
cytokines and cytotoxins that are implicated in this
disease [1,2] Because CB contains a substantially
higher concentration of naive CD4 and CD8 T cells
expressing a diverse surface␣-TCR repertoire [50]
than either ABP [12] or bone marrow [13], these
findings suggest that the functional capacity of naive T
cells in CB to mediate GVHD after transplantation
may be limited compared with naive T cells from
adult sources In this study, we used allogeneic DCs, a
cell type that is highly effective in naive T-cell
acti-vation[16]and that is likely to play a central role in
acute GVHD [4], to activate highly purified naive
CD4 T cells from CB or APB in vitro Our results
reveal multiple T-cell intrinsic mechanisms that are
likely to limit the acquisition of Th1 effector function
in vivo by naive CD4 T cells in CB after their
trans-plantation into an allogeneic recipient
The major difference between our study and
pre-vious observations that compare CD154 or cytokine
expression using APB or CB T cells[34,35,37,38] is
the demonstration of impairment in early
allostimu-lation events using mature DC popuallostimu-lations that are
co-cultured with purified naive CD4 T cells rather
than a mixture of naive and memory cells A
compar-ison of purified naive CD4 T cells from APB versus
CB is important because memory T cells have a
sub-stantially lower activation threshold and mediate rapid
effector responses compared with naive T cells [51]
We also focused on naive CD4 T cells, because this
subset appears to be critical in the initiation of acute
GVHD immunopathogenesis [5-7] Another
differ-ence from earlier studies using allogeneic stimulation
with DCs [34,35] is that we analyzed our cells and
culture supernatants for activation-dependent
pro-teins and responsiveness to IL-12 at a shorter interval
after stimulation (i.e., 24-120 hours) rather than
longer periods In contrast to other studies[34,35], we
also did not perform restimulation of alloantigen
primed CD4 T cells with pharmacologic stimuli, such
as calcium ionophore and phorbol ester Our focus
was to determine whether there were differences in
allogeneic activation and early Th1 differentiation
be-tween naive CD4 T cells from CB versus APB that
might have clinical relevance to the early events in the
initiation of acute GVHD
Despite their diverse␣-TCR repertoire, CB
na-ive CD4 T cells had a significantly lower percentage
of cells expressing CD69 than APB naive CD4 T cells
after allogeneic DC stimulation The level of CD69
per cell was similar for both naive CD4 T-cell types
In contrast, both the frequency and the amount of
CD154 per cell were significantly reduced on CB naive CD4 T cells compared with APB cells These results obtained using a physiologically relevant anti-gen-presenting cell population for␣-T-cell stimula-tion support previous observastimula-tions in which engage-ment of the␣-TCR/CD3 complex using mAbs and bacterial superantigens was used, and which also showed that CB naive CD4 T cells had substantially greater limitations in CD154 than in CD69 expression compared with APB CD4 T cells[33] These differ-ences in the capacity of CB CD4 T cells for the expression of CD69 versus CD154 are consistent with
an early bifurcation in the signal transduction path-ways that lead to the induction of de novo transcrip-tion of the CD69 and CD154 genes after T-cell acti-vation: CD69 gene expression is mainly dependent on phorbol ester-inducible events[52], such as activation of ras and protein kinase C[53], while CD154 gene tran-scription requires a combination of both calcium-depen-dent and ras/protein kinase C-mediated signals[45] Although the role of CD69 in T-cell activation or effector function remains poorly understood, CD154 expression has previously been shown to play multiple roles in naive CD4 T-cell activation that are mediated,
in large part, by engagement of CD40 on DCs[44] These include increasing DC expression of costimu-latory molecules and the production of IL-12 p70 [54] A decrease in these CD154-dependent mechanisms
is also likely to account for the absence of antigen-specific memory CD4 T cells secreting IFN-␥ in human beings with genetic deficiency of CD154[49] Using a neutralizing antibody to CD154, we confirmed that the DC-derived IL-12 p70 production and early IFN-␥ expression by APB naive CD4 T cells was CD154-dependent after allogeneic stimulation with DCs CD154 neutralization also reduced the relatively low levels of production of IL-12 by DCs and IFN-␥
by naive CD4 T cells in CB CD4 T-cell/DC co-cultures, indicating a CD154-dependent pathway for Th1 differentiation also applied to CB CD4 T cells in
an allogeneic context Studies by others have also demonstrated the importance of CD154[55]and IL-12/STAT4[9] in the pathogenesis of acute GVHD Together, these results indicate that reduced CD154 expression by CB naive CD4 T cells after allostimu-lation and its impact on recipient DC cytokine pro-duction plays an important role in limiting early Th1 cell differentiation of cells involved in acute GVHD
We found that allogeneic priming of CB or APB naive CD4 T cells resulted in increased STAT4 mRNA and protein expression To our knowledge, this striking up-regulation of STAT4 protein expres-sion by T-cell activation has not been described pre-viously, and determining the signal transduction and molecular mechanisms by which this is achieved will
be of interest Our data confirm a previous report of decreased STAT4 transcript levels in CB T cells by
Trang 10microarray analysis[56] We furthered these findings
by showing that CB naive CD4 T cells had
signifi-cantly lower basal levels of STAT4 protein than APB
naive CD4 T cells, and a relatively reduced level of
STAT4 protein by CB CD4 T cells persisted after
allogeneic priming This reduced expression of
STAT4 may have contributed to the decreased
fre-quency of IL-12-induced pY-STAT4 in naive CB
CD4 T cells compared with APB cells after allogeneic
priming These results do not exclude the possibility
that allogeneic priming was less effective in CB naive
CD4 T cells in up-regulating IL-12 receptor surface
expression, although our inability to flow
cytometri-cally detect the two components of the IL-12
recep-tor, IL-12R1 or IL-122, on allogeneically primed
CD4 T cells precluded testing this possibility
Regardless of the relative contribution of receptor
versus postreceptor limitations in IL-12 signaling, our
results suggest that not only decreased IL-12
produc-tion but also reduced STAT4-mediated events limit
the Th1 differentiation of naive CB CD4 T cells by
alloantigen Because STAT4 appears to play an
im-portant role in acute GVHD[9], both reduced IL-12
production and a diminished capacity to mediate
STAT4-dependent events may contribute to the
re-duced risk of acute GVHD after transplantation with
allogeneic CB
IFN-␥ appears to up-regulate expression of T-bet, a
master transcription factor that promotes Th1
differen-tiation, although the precise mechanism by which T-bet
acts remains controversial IFN-␥ may also act directly
on activated naive CD4 T cells to maintain IL-12R2
expression, and this effect is dependent on STAT1
[20] Therefore, it is plausible that reduced early
IFN-␥ secretion by naive CB CD4 T cells after
allo-geneic stimulation might act in an autocrine or
para-crine manner to reduce Th1 differentiation compared
with ABP CD4 T cells
IL-10 is an anti-inflammatory cytokine known to
decrease tumor necrosis factor-␣ and IFN-␥
produc-tion and aid in graft tolerance, possibly through the
production of regulatory T-cell populations [57] In
addition, IL-10 production by donor T cells has been
shown to reduce the severity of acute GVHD in
ani-mal models[58] Moreover, others have demonstrated
increased levels of IL-10 from CB naive CD4 T cells
that have been stimulated with anti-CD3 mAb with or
without anti-CD28 mAb and IL-2 for 6 days[59] or
after phorbol myristate acetate and ionomycin
stimu-lation after a 7-day co-culture with mature
monocyte-derived DCs[35] These observations raised the
pos-sibility that regulatory T cells may be increased in the
CB population based on this cytokine profile
How-ever, we found that co-cultures of DCs with naive
CD4 T cells derived from either CB or APB produced
similar low but detectable amounts of IL-10 after 48
hours of co-culture It is possible that IL-10
produc-tion differs based on the mode of stimulaproduc-tion (DCs versus anti-CD3 mAb versus phorbol ester/ionomy-cin), or that increased incubation time is necessary for optimum IL-10 expression, and that IL-10 may play a role in inhibiting later events in acute GVHD Regardless of the cellular source, the higher IL-10:IFN-␥ ratio for CB compared with the APB naive CD4 T-cell co-cultures with DCs, which is a result of the reduced secretion of IFN-␥ in the CB co-cultures, might limit IFN-␥-dependent functions For example, IL-10 may reduce IFN-␥-mediated gene expression
by suppressing STAT1 tyrosine phosphorylation[60],
it is plausible that this high ratio might limit IFN- ␥-dependent enhancement of Th1 differentiation, such
as augmented IL-12 receptor expression, which also requires STAT1[20]
We considered the possibility that decreased priming of CB naive CD4 T cells toward Th1 re-sponses might be a result of increased proportion of regulatory T cells in this naive cell population com-pared with that of ABP Although our data support prior observations that CB and APB have similar numbers of naive CD4 T cells that are CD25high[26], the observation that a smaller fraction of these express FoxP3 protein suggests that the naive CD4 T-cell compartment of CB has a lower concentration of bona fide Tregs than naive CD4 T cells of APB Together, these results suggest that the decreased Th1 differen-tiation observed after allogeneic stimulation of naive CD4 T cells from CB is unlikely to be a result of increased Treg activity in this cell population, but rather defects that are intrinsic to naive CD4 T cells that are CD25⫺/low
The results of these studies are also of interest in defining limitations in CB CD4 T-cell immunity, par-ticularly by Th1 cells, in response to infection or vaccination For example, the human neonate and infant are highly vulnerable to severe or persistent infection with herpesviruses, such as herpes simplex virus and cytomegalovirus, and these susceptibilities are associated with decreased and delayed acquisition
of antigen-specific CD4 T-cell immunity, including Th1 function, compared with adults[41,61] Because recent studies support the idea that␣-TCR engage-ment by complexes of either peptide/allogeneic MHC
or foreign peptide/self-MHC are structurally similar [62], it is likely that these limitations of CD4 T-cell activation by allogeneic stimulation also apply to ac-tivation by foreign peptides bound to self-MHC mol-ecules, and contribute to the reduced CD4 T-cell responses of the neonate to intracellular infections, such as with herpesviruses[39]
In summary, our results reveal multiple mecha-nisms–including reduced expression of IFN-␥, IL-2, IL-12, CD154, and a reduced capacity for IL-12-dependent STAT4 tyrosine phosphorylation–that limit the activation of CB naive CD4 T cells and their