1. Art and Diseño

From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
Early Human Thymocyte Proliferation Is Regulated by an Externally
Controlled Autocrine Transforming Growth Factor-p1 Mechanism
By M. Djavad Mossalayi, Frank Mentz, Fateh Ouaaz, Ali H. Dalloul, Catherine Blanc, Patrice Debre,
and Francis W. Ruscetti
Early thymocytes undergo extensive proliferation after their
entry into the
thymus, but cellular interactions and cytokines
regulating this intrathymic step remain t o be determined.
We analyzed the effects of various T-cell growth factors and
cellular interactions on in vitro proliferation of early
CD2'CD3/TCR-CD4-CD8- (triple negative [TN]) human thymocytes. Freshly isolated TN cells were then assayed for
their growth capacity after incubation with CD21,111-monoclonal antibody (MoAb), recombinant human interleukin-2
(IL-21. IL-7, and/or IL-4. These cells displayed significant proliferative responses with IL-4, IL-7, or CD2-MoAb+lL-2. The
addition of recombinanttransforming
growth factor p
(TGFPI or autologous irradiated CD3'CD8+CD4- cells t o TN
cell cultures dramatically decreased their growthresponses
t o IL-2 and IL-7, whereas IL-4-induced proliferation was less
sensitive to growth inhibition. We thus asked whether the
CD8+ cell-derived inhibitory effect was due t o TGFp. The
addition of neutralizing anti-TGFP MoAb completely abolcell growth. Analysis
ished CD8' cell-derived inhibition of TN
of CD8+ cell-derived supernatants indicated that these cells
TGFPl production capacity, whereas TN cells sehad IOW
crete significantly high levels of TGFP1. Cell fixation studies
showed that TN cells were the source of the TGFP. TGFpl
released from TNcells was in the latent form thatbecame
the active inhibitory form through interaction of TN cells
with CD8' cells. Together, these data suggest a role for
TGFPl as an externally controlled, autocrine inhibaory factor for human early thymocytes, with a regulatory role in
thymic T-cell output.
0 1995 by The American Societyof Hematology.
RANSFORMING GROWTH factors P (TGFP) are
multifunctional homodimeric proteins that belong to
the structurally related polypeptides termed the TGF superfamily.',2 TGFPl is secreted by a variety of cell types, including B cells, T cells, macrophages, and platelets, predominantly or exclusively in a latent form. The in vivo
mechanism that controls activation of this latent form has
not yet been d e f i ~ ~ e dThe
. ~ . pleiotropic
~
activities of TGFPl
include regulation of cell proliferation and differentiation,
control of tissue repair, promotion of angiogenesis, and modulation of inflammatory responses.'~2,sTGFPl has also been
shown as a potent inhibitor of T-cell proliferation, both in
interleukin-2 (IL-2)- and IL-4-derived response^.^" On
some mature human T-cell subsets, TGFPl was also shown
to enhance the proliferative responses of immobilized CD3
monoclonal antibody (MoAb)-stimulated T cells."
Recently, TGFP from thymic epithelial cells was shown
to play a role in murine T-cell maturation.'' However, the
role of TGFPl on early human T-cell proliferation was not
clearly determined. T-cell differentiation occurs in the thymus where bone marrow-derived prothymocytes develop
into cells expressing TCWCD3, a prerequisite event for their
subsequent selection and function.14"*The first human intrathymic step is characterized by an extensive growth of early
prec~rsors,'"'~
whereas the role of soluble growth factors
and cell-cell interactions between various thymic cell subsets
in this phenomenon remain to be determined. In vitro, normal
unfractionated thymocytes respond poorly to mitogens or
antigens." Extensive programmed cell
abnormal
signal transduction capacity of double-positive cells
(CD4.+CD8+)after ligation of CD3 or CD2 antigen^,'^,'^ and
a role for suppressor cells2"have been suggested to explain
this phenomenon. By contrast to CD3+CD4+CD8+thymocytes, human triple-negative (TN) cells (CD3-CD4-CD8-)
possess comparatively higher proliferative capacity in
vitro4.s.x.12and have the potential to generate CD3/TCR+ cells
after appropriate in vitro c ~ n d i t i o n i n g . ' ~ " *In~vitro
~ ~ ~prolif~*
eration of these precursors can be observed following their
treatment with IL-7, IL-2, or IL-4. These cytokines are well
known as the most common physiologic growth factors for
T-cell lineage, including leukemic precursor ce11s.18.19~24~29
This report shows an important inhibitory effect of TGFPl
and autologous CD8+CD4- cells on early TN cell proliferation. TGFPl is also shown to be secreted by early thymocytes and accounts for the inhibitory effect of CD8+ cells.
T
Fromthe Molecular Immuno-Hematology Group, Pitie-SalpCtrikre Hospital, Paris, France; andthe Laboratory of Leukocyte
Biology, National Cancer Institute, Frederick, MD.
Submitted August 25, 1994; accepted February 6, 1995.
Supported by grants from the 'Centre Nationale de la Recherche
Scient$que, " and "Association pour la Recherche sur le Cancer.
Address reprint requests to M. Djavad Mossalayi, PhD, Groupe
d'lmmuno-Hkmatologie Molkculaire, CNRS URA625, H6pital de La
Pitie"Salp2triPre. 91, Bd de I'Hapital, 75013 Paris, France.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
0006-4971/95/8512-0030$3.00/0
"
3594
MATERIALS AND METHODS
The following fluorescein isothiocyanate (FITC)conjugated MoAbs were used: OKT6 (CDl), OKTl l (CD2), OKT3
(CD3), OKT4 (CD4), and OKT8 (CD8) (Ortho, Raritan, NJ); IOT14
(CD25), IOB4 (CD19), and IOTlO (CD57) (Immunotech, Marseille
Lumigny, France); Leu11 (CD16) and MY4 (CD14) (Coulter Clone,
Hialeah, E);and D44 MoAb (kind gift ofDr A. Bernard, Nice
Medical School, Nice, France)." CDS+D44+and CD8'DM- cells
were reported to comprise most functional cytotoxic and suppressor
cells, respectively." For double marker analysis CD4-RDUCD8FITCand CD3-RDlKDZ-FITC (Coulter Clone) were also used.
Cell analysis and sorting were performed using FACstar (Becton
Dickinson, Mountain View, CA).To eliminate their fluorescent
background for the reanalysis of their surface markers, sorted cells
were incubated for 24 hours before staining. For cell activation,
purified CD3-MoAb (OKT3; Ortho) and comitogenic CD21+lrlMoAb (X1 1 and D66 clones, gifts of Dr Alain Bernard)" were used.
Cell preparations. Thymus fragments were obtained from children (t3 years old) undergoing corrective cardiac surgery. Various
thymic cell subsets were obtained as described in detail elseBriefly, major thymic cell subsets were isolated from
Antibodies.
Blood, Vol 85, No 12 (June 15). 1995: pp 3594-3601
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
TGFO AND THYMOCYTE GROWTH
3595
Table 1. Surface Marker Analysis of Various Thymocyte Subsets After Their Purification
Thymic Subsets
Total thymocytes
CD3-CD4-CD8ITN)
CD3+CD4'CD8'
CD3+CD4+CDE-(CD4+)
CD3TD4-CD87CD8')
CD7
CD2
CD3
CD4
CD8
CD4,8*
Percentage of Thymocyte
95
96
96
94
97
95
95
93
93
94
78
2
82
85
84
4
97
96
2
85
1
96
2
98
77
2
97
2
2
>95
4
83
6
5
82
CD2' thymoctes and their various subsets were analyzed by direct immunofluorescence and flow cytometry. Values are the percentage
positive cells (mean from 3 experiments).
* Double positive cells.
of
and unfractionated thymocytes displayed significantly lower
growth ability than other thymic cell subsets ( P < .001;
Table 2). TN thymocytes, as predicted from their surface
phenotype (Table l), did not respond to CD3 ligation. Despite their h e t e r ~ g e n e i t y , ' ~TN
"~.~
~ had significant and
cells
constant proliferative responses in the presence of IL-4, IL7, or IL-2. Maximal proliferation of TN and CD4+CD8cells required CD2-MoAb and IL-2, whereas CD8TD4cell growth was higher with IL-4 (compared with CD8' cell
response to CD2-MoAb + IL-2). Finally, optimal proliferation of TN and CD4+CD8- cells required IL-2 and CD2MoAb. In contrast, whereas CD2-MoAb increased the proliferation of CD8+CD4- cells, optimal proliferation of these
cells was seen with IL-4 stimulation, which was not affected
by CD2-MoAb.
Recombinant TGFPl inhibits the proliferation of TN thymocytes. TGFPis well known as a potent regulator of
hematopoietic cell
Recombinant TGFPl was therefore added to early thymocytes cultured in the presence of
CD2-MoAb+IL-2, IL-7, or IL-4. A dose-dependent inhibition of the proliferative responses was observed (Fig l),
which indicates that TGFPl displays a high inhibitory effect
on E-2- or IL-7-dependent thymocytes growth ( P < .001),
whereas IL-4-dependent TN cell proliferation was less sensitive to TGFPl addition (P < .05). This finding indicates
that these cytokines induced distinct proliferation pathways
in human early thymocytes. TGFP effect can be reversed
through the addition of neutralizing anti-TGFP MoAb to the
cultures but not through the addition of an isotype-matched
control (CD19-MoAb; Fig 1).
Mature CD8+ thymocytes downregulate the proliferation
of TN thymocytes. Cellular interactions regulating early
thymocyte growth remain ill defined. However, we have
previously shown that agar colony formation by prothymocytes was inhibited by CD8+4- thymic subset." We assayed
here the effect of CD8+ as well as other major thymic cell
subsets on the growth of TN cells stimulated with CD2RESULTS
MoAb+IL-2. Data in Fig 2 show the ability of CD8' cells
.007) in
to inhibit the proliferation of TN thymocytes ( P
Definition of growth requirement of TN early thymocytes.
a dose-dependent manner, whereas TN, CD4+, and
TN thymocytes were assayed for their proliferative response
after their incubation with comitogenic CD21+III-M~Abr CD4+CD8' cell subsets displayed no such inhibitory effect.
CD3-MoAb, IL-2, IL-7, and/or IL-4 and compared with
Higher amounts of exogenous IL-2 (500 to 1,OOO U / d ) or
other major thymocyte subsets. Optimal concentrations of
CD2-MoAb (up to 100 pg/mL) did not significantly overthese reagents were previously
Results in
come this inhibitory effect (not shown). This finding indiTable 2 indicate that thymocytes varied in their response to
cated that the CD8' cell-derived effect was not caused by
the above physiologic signals. Double-positive (CD4+CD8+)
IL-2 or CD2-MoAb consumption.
CD2' (AET-sheep erythrocyte') cells after their labeling with CD4PE and CD8-FITC MoAb and subsequent cell sorting. Using this
procedure, CDCCD8-, CD4+CD8+,
CD4+CD8(CD4+),
and
CD4-CD8' (CD8+)subsets were obtained. CD2+ thymocytes were
also treated with CD4, CD8, and CD3 MoAb and CD3-CD4-CD8TN, fortriple-negativefraction).Each
cellsweresorted(termed
thymiccellsubsetwasfirstanalyzedforsurfacemarkers.Results
in Table 1 indicate that the purity of these fractions was greaterthan
90%. By contrast to other thymic cells,a significant percentage(3%
to 6%) of TN cells expressed IL-2 receptor,CD25 Labeling of TN
cells with both CD7 and CD2 showed greaterthan 98% cell reactivity. No cell reactivity was observed with MoAbs for CD14, CD19,
CD16, or CD57 antigens.
Cell cultures.Totalthymocytesortheirvarioussubsetswere
incubated (1 to 5 X 104 cells/100 pUwell) in 96-well microtiter in
human ABRPM1medium(Tebu,Paris,France)containing20%
sera, and various combinations ofthe following reagents were added
tothecultures:comitogenicCD2,+,,,-MoAb(U400ascitesfrom
each), CD3-MoAb (OKT3 clone, 20
pg/mL), recombinant IL-2(100
U/mL; BoehringerMannheim,Meylan,France),IL-7
(IO ng/mL;
Calbiochem,SanDiego,CA),IL-4
(10 ng/mL; a giftfromDr J.
Banchereau, Schering Plough, Dardilly, France), TGFPl (rTGFP1;
Bristol-Myers-Squibb, Seattle,WA), neutralizing anti-TGFP MoAb
(clone1-Dl1016,neutralizing both TGFPlandTGFP2 forms):'
IOB4 MoAb as isotype-matched control (CD19, clone BC3; Immunotech), and polyclonal anti-TNFa antibody (Tebu).All human sera
used were pretested to ensure that there were no inhibitory effects
on thymocyte proliferation. Tritiated thymidine (1 pCi/well; CEA,
Les Ulis, France) was incorporated at day4 and radioactivity uptake
was measured 18 hours later. Maximal growth responses were observed after thisincubationperiod,assuggested
by our previous
~tudy.'~.'~
TGFPI was quantified using a specific radioimmunoassay (NENDupont, Paris, France) as recommended by the manufacturer. This
assay detects more than 30 pg TGFPl in cell supernatants and measures active as well as latent TGFP. For some experiments the cells
were irradiated with 3,000 rad or fixed
with paraformaldehyde before
cultures.
Statistics. Various resultswereanalyzedandcomparedusing
the Student's t-test for paired data.
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
3596
MOSSALAYI
ET AL
Table 2. Prolieration Capacity of Various ThymicCell Subsets
~
~~~
~
~
Thymic Cells
None
IL-2
IL-2 + CD2 MoAb
IL-2 + CD3 MoAb
TN(CD2')
CD2*CD4+CD8'
CD2'CD4'
CD2+CD8'
E'(CD2')
860
1,214
2,072
1,200
1,196
5,327
2,001
3,022
1,845
2,134
27.149
6,800
30,324
8,865
8,453
5,535
6,549
25,068
8,657
7,932
IL-4
IL-4
15,059
2,820
7,995
17,962
5,232
+ CD2 MoAb
16,245
1,998
8,658
16,258
6,525
IL-7
IL-7 + CD2 MoAb
6,632
1,352
3,932
1,502
2,397
7,212
NT
NT
NT
3,022
Thymocyte subsets were incubated (lO'/lOO pUwell) with rlL-2 (100 U/mL), CD3 MoAb (20 pg/mL), comitogenic CD~MII-MOA~
(V400 ascites
from each), rlL-7 (10 ng/mL), and rlL-4 (10 ng/mL). Tritiated thymidine was added at day 4 and radioactivity uptake was measured 18 hours
later. Results represent the mean values from nine distinct thymocyte preparations, each performed in triplicate (SE <11%).
Abbreviation: NT, not tested.
Phenohpical characterization of downregulating CD8'
cells. The data above led us to investigate the surface phenotype of the suppressor CD8' cells. CD2TD8' thymocytes are a heterogeneous cell population and contain the
precursors of functionally distinct peripheral T lymphocyte~.'~"'Using negative selection procedure (by positive
cell elimination), we attempted to further define the cell
subset responsible for the inhibition of TN cell proliferation.
CD8' cells were therefore treated with CD3-, CD4-,
CD57-, or D44-MoAb; negative cells were then sorted and
assayed for their suppressive activity. After treatment with
above MoAb, CD8' cells contained less than 2% cell positivity with the MoAb used for cell elimination. D44 MoAb
was used because it was previously described to recognize
functionally distinct peripheral bloodCD8'
cell subsets:
CD8'D44' cells are mostly cytotoxic, whereas CD8'DUlymphocytes comprise cells that suppress Ig secretion by B
lymphocytes.3' Although a minor subset of CD8' cells (2%
to 4%), CD44' cells were also sorted and assayed in this
study. Data in Fig 3A show the effect of various cell depletions on the ability of CD8' cells to suppress the proliferation
of TN cells. CD4'.CD57'.
or D44'cell depletion from
CD8' cells didnot affect their inhibitory effect, whereas
CD3' cell elimination completely abolished CD8' cell-derived effect. These findings indicate that the suppressor cells
are likely to be CD3'CD8+CD4-CD57-DU-. Consequently, we have positively sorted these cells (2% to 4%
CD2-mAb
+
11-2
total thymocytes) and assayed them for their ability to suppress TN cell growth. Data in Fig 3B confirm the highinhibitory effect of both irradiated and nonirradiated CD8' cells
on both IL-2- and IL-7-mediated proliferation of TN cells
and indicate that suppression was not caused by CD8' cell
irradiation. Together, these data clearly show
that
CD3TD8' thymocytes comprise cells that downregulate
the proliferation of TN thymocytes.
Anti-TGFpl MoAh reverses CD8' cell-derived inhibiton
effects. To clarify the mechanism ofCD8'
cell-derived
suppression, we first asked whether an inhibitory cytokine
was involved in this phenomenon. In regard to the similarities between CD8' cell-derived and TGFPI-derived effects,
we have analyzed the role of TGFPl in CD8' cell-derived
inhibition of TN cell proliferation. TN thymocytes were then
cultured in the presence of CD2-MoAb+IL-2. irradiated
CD8' cells, neutralizing anti-TGFPI MoAb, anti-TNFa Ab,
andor an isotype-matched control (IOB4). Interestingly,
only anti-TGFPI addition to these cocultures significantly
reversed the inhibitory effect of CD8' cells ( P < .001; Fig
4A). This effect was dose-dependent and was not observed
with control isotype-matched MoAb (Fig 4B).
HumanTN thymocytes produce high TGFPl levels. The
effects of recombinant TGFPl on TN cell growth and the
ability of anti-TGFP1 MoAb to reverse the inhibitory effect
of CD8' cells led us to quantify the TGFPI levels produced
by these cells as well as other thymic subsets. Unexpectedly,
IL-7
IL-4
T
1
1
I
20000
T
10000
n
"
~~
10 5
2.5
1 0.5 0.1
0
10 10
~
0
~~
5
10
rTGFp (ng/ml)
0
5
10
Fig 1. Dose-dependent inhibitory effect of TGFpl
on early thymocyte proliferationin response t o CD2MoAb+lL-2, IL-7, or IL4 (the same concentrations
used in Table 21. This effect could bereversed by the
addition of anti-TGFp1 MoAb (20 pglmL) and not
by isotype-matched CD19 MoAb (1084. 20 pglmL).
Shown are thymidine uptake at day 5 per 10' TN
thymocytes seeded. Values are the mean + SD from
four experiments.
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
3597
TGFB AND THYMOCYTE GROWTH
completely abolished the release of the suppressor factor in
these cocultures. CD8' cell addition to TN cell supernatants
also converted from inactive to active TGFPI, which then
inhibited TN cell growth. Incubation of TN-SN with CD8'
cells did not significantly increase TGFPl levels, as measured by radioimmunoassay (data not shown). Finally. the
suppressor effect of CD8' + TN cell-derived supernatants
was reversedthroughtheaddition
of anti-TGFPI MoAb
(Table 4). Together, these data indicate that TN cells are
indeed the source of TGFPl that is converted to an active
form after their interaction with CD8' cells.
1 00'
I
75
50
25
DISCUSSION
0
I
CD8+
TN
CD4+
CD4+8+
The presentstudy attempted to dissect thegrowthand
the activation capacities of human early thymocytes, their
interactions with other thymic cells, and the effect of TGFPl
on their in vitro growth response. Severalnovelfindings
IRRADIATED THYMOCYTES ADDED
TO THE CULTURES ( ~ 1 0 3 )
Fig 2. Inhibitory effect of CD8' thymocyte addition on the proliferation of TN cellsubset. Purified CD4+(CD8-),CD8+(CD4-),CD4'CD8',
and TN thymocytes were irradiated
at 3,000 rad andadded t o TN cells
cultured (lO"/mL) in the presence of CD2,+,,,-MoAb and IL-2. Tritiated
thymidine (1pCilwell) was added on day 4 and radioactivity uptake
by thecells was measured 18 hours later. Results show thepercentage of growth inhibition as compared with thymocytes cultured in
the absence of irradiated cells (Table 2). Values are the mean + SD
from three distinct thymocytepreparations.
CD8' cells were poor producers of TGFPl (<0.8 ng/mL),
even after their in vitro activation (Table 3). By contrast,
TN thymocytes produced significantly high TGFPl levels,
as did activated CD4' cells. However, this TGFPl was not
active because these cells proliferated in vitro in the presence
of the same growth factors (Table 2). Finally, we have obtained similar data from three different TN cell preparations
containing, respectively, 95%, 97%, and 99% CD2' cells,
which suggest that contaminating non-T cells maynot account for TGFPl production. Figure 5 shows that TGFPl
production is dependent on TN cell numbers added to the
cultures. These data indicate that TN thymocytes differ from
other thymic subsets in their high TGFPl releasing capacity
in the absence of apparent in vitro activation.
Meckanism of CD8' cell-derived inhibition of early tkymocyte prol(feration: role for TGFPI. We here analyzed
the mechanism by which CD8' cells mediate the inhibitory
effect on TN cells. In contrast to the cells themselves, the
addition of CD8' cell supernatants had no effect on IL-2or IL-7-dependent early thymocyte growth (Table 4). This
finding indicates that these cells did not produce biologically
active TGFPI. Because supernatants from cocultures of
CD8' and TN cells displayed important inhibitory effect
(Table 4), we thus analyzed the role of each cell subset in
inhibitory factor production. CD8' or TN thymocytes were
thenfixed by paraformaldehyde and cocultured withthe
other subset. As shown in Table 4, CD8' cell fixation did
not affect TN growth inhibition, whereas TN cell fixation
A
x 108 CPMllO4 TN CELLS
2
10
CD@
None
CDWCDI'
CDWCD67'
B
Cdb Added to
the cultures:
30-
20
l
J-
I
x 10s CPMllOJ TN CELLS
20
10
30
None
I-
+
lo*
2.10s
=
6x 10s
104
+
11-2
2.104
Fig 3. Phenotypical characterization of suppressive CDB'CD4thymocytes. (A) Various CD8' cell subsets were sorted, irradiated
(3,000rads), and added (10'/well) t o TN thymocyte cultures, together
with CD2-MoAb+lL-2. Of the total CD8* cells: CD4- cells = 96% to
99%;CD57= 96Yo t o 99%;
CD"
= 95% t o 98%;CD3= 196
t o 3%; and CD44' = 1% t o 4%. Values are the mean 2 SD from
three
distinct
thymuses. (B) Suppressor thymic
cell
subset
(CD8+CD4-CD57-D44XD3+)was directly sorted from thymocytes
and assayed for its inhibitory activities on the growth of cells
TN in
the presence of IL-7 or CD2-MoAb+lL-2 (as in Table 2). CD8' cells
were added after irradiation atincreasing numbers or without
irradiation at 10' cells/well. Values are the mean SD from four different
cell preparations.
+
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
3598
MOSSALAYI ETAL
-E
10000
n
-
1500
\
1
f I
W
P
I
I
sow
c3,
T
CD1 9-MoAb
10
0
20
30
CELL PROLIFERATION(cPm103)
0
x103 TN Thymocytes I ML
Fig 5. Dose-dependent production of TGFpl from TN cells. Cells
were incubated (lo' t o 1O6/mL) in culture medium containing CD2MoAb+lL-Z. TGFpl levels were quantified in cell supernatants after
48 hours of incubation. In this case, TN cells had 97% CD2* cells.
B
.
h
0
04
0
. 10. - 2 0. -3 0 . -4 0 . SO
- . >
pglml MoAb added
Fig 4. Dose-dependent reversion of CD8' cell-derived inhibition
by anti-TGFp MoAb. (A) TN thymocytes were cultured in the presence of CDZ-MoAb+lL-2, with or without CD8' thymocytes (5,0001
well), anti-TGFp-MoAb, anti-TNFu Ab, or an isotype-matched
control,
the 019-MoAb.Values are the mean tritiated thymidine uptake from
five distinct experiments, each performed in triplicate (SE < 139b).
(B) TN cells were culturedin the presence of CD8' cells and increasing
concentrations of anti-TGFp MoAb or isotype-matched106-4 MoAb.
Values are the mean from three distinct thvmuses(SD < 13%).
emerged from these data. ( I ) Human thymocytes differ in
their proliferation potential and sensitivity to IL-2, IL-7, and
IL-4, which maybe related to their developmental stages.
(2) The growth of human TN thymocytes is downregulated
Table 3. TGFp Production by Various Thymocyte Subsets
TGFL?l (nglmU10' cells)
Supernatants FromCells Incubated With
Thymocytes
None
Total
TN
CD4'
CD8'
0.6
1.4
3.2
4.1 10.1
0.6
0.5
0.6
CD2 MoAb
CD3 MoAb
0.9
1.6
5.4
0.6
0.6
5.0
0.5
CD2 MoAb + IL-2
2.1
Values are TGFBl levels quantified by radioimmunoassay in 48hour supernatants from various thymocyte subsets. Values reflect
representative data from one experiment (SD <15%) of three.
by a subset of autologous CD8'cells.an
effect requiring
cell-cell interactions and reversed by MoAb to TGFPI. (3)
Recombinant TGFPl dramatically inhibited the proliferation
of human TN thymocytes. (4) TN cells produce high TGFD 1
levels that, after contact withCD8'cells.convertedinto
biologically active form.
TGFD is a member of a highly conserved family of polypeptide factors regulating cell growth and differentiation. It
is secreted by most human cells and selectively inhibits CSFinduced growth of both murine and human immature hematopoietic cells.' TGFDl also acts as an important immunomodulatory protein5 for cells of the immune system as it
regulates the proliferation of ~'.".''' and B lympho~ytes.'~.~~
inhibits Ig release by B cell^,^^.^" and modulates cytokine
production by T lymphocyte^.^'.^' Some of these effects
seems to be mediatedthrough counteracting IL-l funct i o n ~by
~ .inhibiting
~~
IL- I R expression."
The role of TGFP on human early T-cell development
remains to be defined. The intrathymic sojourn of T lymphocytes constitutes an essential step in the generation of immunocompetent cells. Bonemarrowprothymocytesmigrate
through the thymus, where theyundergo extensive expansion
together with the acquirement of various surface antigens
including CD3/TCR.'"" These events allow subsequent positive and negative selection of appropriate clones. Present
work shows that, although heterogeneou~,'~"~
early
CD3-CD4-CD8- (TN) thymocytes display a significantly
higher proliferation potential than most thymic cells (Table
2). The absence of their response through CD3 cross-linking
further confirmed the CD3- phenotype of these cells (Tables
1 and 2). High growth potential of TN thymocytes correlates
with the fact that they belong to the proliferative compartment of thymic outer cortex.I5In vivo, early thymocyte activation maybe initiatedthrough CD2-triggering by LFA3 antigen on epithelial cell^.'"^^.^^ Early thymocytes also
proliferated in response to IL-4 and IL-7, corroborating with
previous reports on the role of these cytokines in early Tcell d e v e l ~ p m e n t . ' ~ . ~ ~ " ~
Our data showthe presence of a limitedpopulation of
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
TGFB
3599
Table 4. Cell-Cell Interaction Requirement for the Generation of Inhibitory Factor From TN Cells
TN Cells Cultured With*
(CD8'
CD2 MoAb
1. None
2.CD8' cells (104/well)
TN)-SN
3.
4. (CD8'
TN)-SN + anti-TGFP
5. CD8+-SN
6. TN-SN
7. (CDl+-fixed TN)-SN
8.(CD8' + TN-fixed"
9. (TN-SN +13,429
CDB')-SN
10. (TN-SN + CD8')-SN
anti-TGFP
22,652
5,253
10,267
20,839
21,142
21,425
12,124
22,119
+ IL-2
5,241t
t 2.147
5 1,502
-C 2,125
t 1,247
5 2,555
5 1,855
2 1,186
IT 2,452
20,259 5 3,214
+
+
+
+
Inhibition
IL-7
-
8,965 2 1,242
-
Yes (77%)$
Yes (55%)
No (8%)
No (7%)
NO (5%)
Yes (46%)
N o (2%)
Yes (41%)
NO (11%)
1,284 t 254
Yes (86%)
4,239
8,856
9,623
8,858
4,969
7,967
t 2,027
t 862
2 1,005
-t 212
Yes (53%)
No (1%)
No (0%)
No (1%)
Yes (45%)
No (11%)
5,436 5 259
8,002 t 2,351
Yes (39%)
No (11%)
5
2 323
2 1,245
Inhibition
+
* TN cells were cultured (104/100pL) in the presence of CD2 MoAb IL-2 and (1) none; (2) CD8+ cells; (3) 24-hour supernatants (SN) from
CD8' + TN cells; (4) CD8'
TN cell SN and anti-TGFP1 MoAb; (5) CD8' cell-derived SN; (6) TN cell-derived SN; (7) SN from CD8' cells fixed
with paraformaldehyde + TN cell cocultures; (8) SN from CD8' + fixed TN cell cocultures; (9) 24-hour SN from TN cells incubated with CD8'
cells for an additional 24 hours; (10) same as 9 anti-TGFP MoAb. SN were added at 20% final dilution.
t Thymidine uptake per lo* TN thymocytes cultured as in Table 2.
Rate of inhibition compared with that of TN cells cultured with medium alone (none). Yes indicates significant inhibition with P < .01; no
indicates that no significant inhibition was observed.
+
+
*
thymic CD8+ cells that inhibited the proliferation of early
TN precursors. Phenotypical characterizationof these suppressor cells indicated that theyare CD3+CM-CD8+CD57-D44.
Such a phenotype was previously reported as a characteristic
of PBL-derived T lymphocytessuppressing Igproduction
by B lymphocytes
and
to
be distinct
from
NK cells
(CD3-CD57+CD16+)and cytolytlc (CD8+D44+)cells?' Interestingly, anti-TGFPl MoAb completely reversed the inhibitory
effect of CD8+ cells. The failure of supernatants from CD8+
cellstosuppressthymocyteproliferationsuggestedthatthe
interaction between these thymocytesand TN cells is required
for the activationof TGFP1. Further analysis showed the presence ofhigh TGFP1 levels in supernatantsfrom TN cells,
whereasCD8'
cellsproducedcomparativelylow
TGFPl
amounts (Table 3). CD8+ cells induced the conversion of TN
cell-derived TGFPl from latent to active form, leading us to
postulate that the mature cell pool in the thymus may have a
feedback controlon the proliferation of early precursor (Fig5).
Even though we feel most of the TGFP is made by CD2+ TN
cells, we cannot rule out
a contribution by other cell types such
as thymic epithelial cells.
Recently, Takahama et a l l 3 reported the role of TGFP
in murine CD4-CD8I0 cell differentiation into CD4+CD8+
thymocytes through a paracrine mechanism. This study did
not address the effect of TGFP on human TN cell differentiation; however, it differs from the work of Takahama et all3
in that it shows an autocrine source for TGFP. Our data are
supported by previous reports that point to the autocrine
inhibitory effect of TGFP on
and murine4' hematopoietic stem cells. In addition, by contrast to Takahama et
we have reversed TN inhibition by antibody to TGFP,
further supporting the direct involvement of this cytokine.
However, our data do not exclude a role for paracrine TGFP
on human TN cell proliferation.
TGFP is generally secreted in a latent formb composed of
a homodimer of 105 kD of which the c-terminal 112 amino
acids of each chain form the mature active 25-kD cytokines.
In vitro, the release of active TGFP from latent complex can
be facilitated by enzymatic or physicochemical treat-
ment.3.4.49
As in this study, it was previously reported that
coculture of two cell types, pericytes and endothelial cells,
produced active TGFP, whereas culture of either one alone
would produce only the latent form?' Meanwhile, molecular
mechanism(s) for TGFPl activation and the TGFPI-derived
suppression of TN cell growth remain to be established. In
various human cell lineages, TGFP is shown to inhibit the
proliferation by delaying or arresting progression through
the late portion of G1Recent
studies onthe intracellular
target for TGFP have demonstrated that it prevents the assembly of cyclin E-Cdk2 and subsequent accumulation of
cyclin-E-associated kinase activity.53TGFP may cause a
general inhibition of cyclin-CDK interactions during G1
leading to the inhibition of retinoblastoma protein phosphorylati~n.".'~In recent years, various reports pointto the
involvement of repressor genes in the regulation of hematopoietic cell de~eloprnent.~'
Relatively high TGFPl levels
produced by TN cells suggest a role for these genes in regulating early T-cell development. TGFPl may thus repress
early T-cell proliferation until their entry into the thymus,
where an appropriate microenvironment inactivates TGFP 1
and allows their proliferation. Data showed an enhancement
of the in vitro proliferation of early hematopoietic precursor
cells after the addition of antisense to TGFP during their
cult~res.''~.~~
Accordingly, disregulation of continuous repression by TGFPl or overcoming its effects may thus lead
to leukemic transformation of T-cell precursors. Such TGFPrelated function was recently suggested for the development
of a case of human T-cell lymphomas6and some nonhematopoietic t ~ m o r s . "Experiments
~ ~ ~ ~ ~ ~ are nowin progress to
assay this hypothesis in acute T-cell leukemia patients.
ACKNOWLEDGMENT
We thank Drs S. Saeland, A. Bernard, P.C.L. Bevereley, J. Banchereau, and J.C. Lecron for their giftsof cytokines and MoAb, and
M. Benhamou for critical reading of the manuscript.
REFERENCES
1. Keller JR. Ruscetti F W : Transforming growth factor P (TGFP)
and its role in hematopoiesis. Int J Cell Cloning 10:2, 1992
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
3600
2. Sporn MB, Roberts AB: Transforming growth factor-@:Recent
progress and new chalenges. J Cell Biol 119:1017, 1992
3. Lyons RM, Gentry LE, F'urchino AF, Moses HL: Mechanism
of activation of latent recombinant transforming growth factor h l
by plasmin. J Cell Biol I10:1361, 1990
4. Massagut J: The transforming growth factor-P family. Annu
Rev Cell Biol 6597, 1990
5. Ruscetti W ,Varesio L, Ochoa A, Ortaldo J: Pleiotropic effects
of transforming growth factor-@on cells of the immune system. Ann
NY Acad Sci 685:488, 1993
6. Kehrle JH, Roberts AB, Wakefield LW, Jakowlew S, AlvarezMon M, Rerynck R, Sporn MB, Fauci AS: Production of transforming growth factor-@by human T lymphocytes and its potential
role in the regulation of T cell growth. J Exp Med 163:1037, 1986
7. Wahl SM, Hunt DA, Wong HL, Dougherty S, McCartneyFrancis N, Wahl LW, Ellingsworth L, Schmidt JA, Hall J, Roberts
AB, Sporn MB: Transforming growth factor p is a potent immunosuppressive agent that inhibits IL-I -dependent lymphocyte proliferation. J Immunol 140:3026, 1988
8. Ruegemer JJ, Ho SN, Augustine JA, Schlager JW, Bell MP,
McKean DJ, Abraham RT: Regulatory effects of transforming
growth factor-@on IL-2- and IL-4-dependent T cell cycle progression. J Immunol 144:1797, 1990
9. Ahuja SS, Paliogianni F, Yarnada H, Balow JE, Boumpas DT:
Effect of transforming growth factor-p on early and late activation
events in human T cells. J Immunol 150:3109, 1993
10. Stoeck M, Ruegg C, Miescher S, Carrel S, CoxD,von
Fliedner V, Alkan S: Comparison of the immunosuppressive properties of milk growth factor and transforming growth factor-p1 and
02. J Immunol 43:3258, 1989
11. Fox FE, Ford HC, Douglas R, Cherian S, Nowell PC: Evidence that TGF-P can inhibit T-lymphocyte proliferation through
paracrine and autocrine mechanisms. Cell lmmunol 150:45, 1993
12. de Jong R, van Lier RAW, Ruscetti F W , Schmitt C, Debr6
P, Mossalayi MD: Differential effect of transforming growth factorp1 on the activation of human naive and memory CD4' T lyrnphocytes. lnt Immunol 6:631, 1994
13. Takahama Y, Letterio JJ, Suzuki H, Farr AJ, Singer A: Early
progression of thymocytes along the CD4/CD8 developmental pathway is regulated by a subset of thymic cells expressing transforming
growth factor p, J Exp Med 179:1495, 1994
14. Reinherz EL, Kung PC, Goldstein G, Levy RH, Schlossman
SF: Discrete stages of human intrathymic differentiation. Analysis
of normal thymocytes and leukemic lymphoblasts of T-cell lineage.
Proc Natl Acad Sci USA 77:1588, 1980
15. Janossy G , Bofill M, Trejdosiewicz LK, Willox HN, Chilosi
M: Cellular differentiation of lymphoid subpopulations and their
microenvironments in the human thymus. Curr Top Pathol 75:89,
1987
16. Haynes BF, Denning SM, Singer KH, Kurtzherg J: Ontogeny
of T-cell precursors: A model for the initial stage of human T-cell
development. Immunol Today 1037, 1989
17. Mossalayi MD, Dehrt P: Human early T cell differentiation.
Res Immunol 145:119, 1994
18. Spits H:Early stages in human and mouse T-cell development. Cum Opin Biol 6:212, 1994
19. Campana D, Janossy G: Proliferation of normal and malignant
human immature lymphoid cells. Blood 71:1201, 1988
20. Dalloul AH, Mossalayi MD, Bertho JM, Courboin G, Debrt
P: Characterization of colony-forming cells in the human thymus:
Evidence for a suppressor mechanism. Exp Hematol 17:774, 1989
2 1. Jenkinson E, Kingston R, Smith CA, Williams GT, Owen JJ:
Antigen induced apoptosis in developing T cells: A mechanism for
negative selection of the T cell receptor repertoire. Nature 19:2175,
1989
MOSSALAYI ET AL
22. Fowlkes BJ, Pardoll DM: Molecular and cellular events of T
cell development. Adv Immunol 44:207, 1989
23. Mossalayi MD,DalloulAH, Bertho JM, Bismuth G, Blanc
C, DehreP: Stage specific phosphoinositides turnover capacity of
human intrathymic T cells following CD2-triggering. Biochem Biophys Res Commun 168:665, 1990
24. Dalloul AH, Mossalayi MD, Dellagi K, Bertho JM, Dehrt P:
Factor requirements for activation and proliferation steps of human
CD2+CD3-CD4-CD8early thymocytes. Eur J lmmunol
19: 1985, 1989
25. Torihio ML, de La Hera A, Borst J, Marcos M, Marquez C,
Alonso JM, Barckna A, Martinez-A C: Involvement of the IL2 pathway in the rearrangement and expression of both alp and y/6 T cell
receptor genes in human T cell precursors. J Exp Med 168:2231,
1988
26. Dalloul AH, Fourcade C, Debrt P, Mossalayi MD: The induction of prothymocytes differentiation by supernatants from thymic
epithelial cells. Eur J Immunol 21:2633, 1991
27. Barckna A, Torihio ML, Pezzi L, Martinez-A C: A role for
interleukin-4 in the differentiation of mature T cell receptor rb/a@
cells from human intrathymic T cell precursors. J Exp Med 172:439,
1990
28. Kraft DL, Weissman IL, Waller EK: Differentiation of
CD3-4-8- human fetal thymocytes in vivo: Characterization of a
CD3-4'8- intermediate. J Exp Med 178:265, 1993
29. Welch PA, Namen AE, Goodwin RC, Armitage R, Cooper
MD: Human IL-7: A novel T cell growth factor. J Immunol
143:3562, 1989
30. Huet S, Boumsell L, Raynal B, Degos L, Dausset J, Bernard
A: Role of T cell activation for HLA class I molecules from accessory cells: Further distinction between activation signals delivered
to T-cell via CD2 and CD3 molecules. Proc NatlAcad Sci USA
84:7222, 1987
3 1. Calvo CF, Bernard A, Huet S, Leroy E, Boumsell L, Senik
A: Regulation of immunoglobulin synthesis by human T cell subsets
as defined by anti-D44 monoclonal antibody within the CD4' and
CD8' subpopulations. J Immunol 136:1144, 1986
32. Dasch JR, Pace DR, Waegell W, Inenaga D, Ellingsworth L:
Monoclonal antibodies recognizing transforming growth factor-@:
Bioactivity neutralization and transforming growth factor p2 affinity
purification. J Immunol 142:1536, 1989
33. Mentz F, Ouaaz F, Michel A, Blanc C, Herv6 P, Bismuth G,
Dehr6 P, Merle-B6ral H, Mossalayi MD: Maturation of acute T
lymphoblastic leukemia cells following CD2-ligation and subsequent
treatment with interleukin-2. Blood 84: 1182, 1994
34. Terstappen LW, Huang S, Picker LJ: Flow cytometric assessment of human T-cell differentiation in thymus and bone marrow.
Blood 79:666, 1992
35. Sing GK, Keller JR, Ellingsworth LR, Ruscetti F W : Transforming growth factor @ selectively inhibits normal and leukemic
human bone marrow cell growth in vitro. Blood 72:1504, 1989
36. Ellingsworth LR, Nakayama D, Segarini P, Dasch J, Carrilo
P, Waegell W: Transforming growth factor @S are equipotent growth
inhibitors of interleukin-1 -induced thymocyte proliferation. Cell
Immunol 114:41, 1988
37. Kehrle JH, Roberts AB, Wakefield LM, Jakowlew S, Spom
MB, Fauci AS: Transforming growth factor p is an important immunomodulatory protein for human B lymphocytes. J Immunol
137:3855, 1986
38. Smeland EB, Blomhoff HK, Hoke H, Ruud E, Beiske K,
Funderud S, Godal T, Ohlsson R: Transforming growth factor p
(TGFp) inhibits G1 to S transition, hut not activation of human B
lymphocytes. Exp Cell Res 171:213, 1987
39. Islam KB, Nilsson L, Sideras P, Hammarstrom L, Smith CI:
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
TGFP AND THYMOCYTE GROWTH
TGF-p1 induces germ-line transcripts of bothIgAsubclassesin
human B lymphocytes. Int Immunol 3:1099, 1991
40. Wu CY, Brinkmann V, Cox D, Heusser C, Delespesse G:
Modulation of IgE synthesis by transforming growth factor 8. Clin
Immunol Immunopathol62:277, 1992
41. Swain S , Huson G, Tonkonogy S , Weinberg A: Transforming
growth factor-p and L4 cause helper T cell precursors to develop
into distinct effector helper cells that differ in lymphokine secretion
pattern and cell surface phenotype. J Immunol 147:2991, 1991
42. Fargeas C, Wu CY,Nakajima T, Cox D,Nutman T, Delespesse G: Differential effect of transforming growth factor (lon
the synthesis of Thl- and Th2-like lymphokines by human T lymphocytes. Eur J Immunol 22:2173, 1992
43. Dubois CM, Ruscetti I T , Palaszynski EW, Falk LA, Oppenheim JJ, Keller JR: Transforming growthfactor is a potent inhibitor
of interleukin 1 ( L - l ) receptor expression: Proposed mechanism of
inhibition of IL-1 action. J Exp Med 172:737, 1990
44. Denning SH, Kurtzberg J, Le FT, TuckDT, Singer KH,
Haynes BF: Human thymicepithelial cells directly induce activation
of autologousimmaturethymocytes.ProcNatlAcad
Sci USA
85:3125, 1988
45. Denning SM, Dustin ML, Springer TA, Singer KH, Haynes
BF: Purified lymphocyte function associated antigen-3 activates human thymocytes via the CD2 pathway. J Immunol 141:2980, 1988
46. Hatzfeld J, Li M,BrownEL,SookedeoH,Levelsque
JP,
O'Toole T, Gurney C,Clark SC, Hatzfeld A: Release of early hematopoietic progenitors from quiescence by anti-sense transforming
growth factor 8 1 or Rb oligonucleotides. J Exp Med 174:925, 1991
47. Li M, Cardoso A, SansilvestriP,Hatzfeld A, Brown EL,
Sookdeo H, Levesque JP, Clark SC, Hatzfeld J: Additive effects of
steel factor and anti-sense TGF-p1 oligodeoxynucleotides on CD34+
hematopoietic progenitor cells. Leukemia 8:441, 1994
3601
48. Ploemacher RE, van Soest PL, Boudewijn A: Autocrine transforming growth factor @lblocks colony formation and progenitor
cellgeneration by hematopoieticstemcellsstimulatedwithsteel
factor. Stem Cells 11:336, 1993
49. Huber D, Philippe J, Fontana A: Protease inhibitors interfere
with the transforming growthfactor-P-dependent but not the transforming growth factor-p-independent pathway of tumor cell-mediated immunosuppression. J Immunol 148:277, 1992
50. Antonelli-Orlidge A, Saunders KB, Smith SR, D'Amore PA:
An activated transforming growthfactor p is produced by cocultures
of endothelial cells and pericytes. Proc Natl Acad
Sci USA 86:4544,
1989
51. Laiho M, DeCaprio JA, Ludlow J W , Livingston DM, Massague J: Growth inhibitionby TGF-p linked to suppression of retinoblastoma protein phosphorylation. Cell 62: 175, 1990
52. Howe PH, Drraetta G, Leof EB: Transforming growth factor
p1inhibition of P34*" phosphorylation and histoneH1 kinase activity associated withGUS-phase growth arrest. Mol Cell Biol 11:1185,
1991
53. Koff A, Ohtsuki M, PolyakK, Robert JM,Massagu.5 J: Negative regulation of G1 in mammalian cells: Inhibition of cyclin Edependent kinase by TGF-p. Science 260536, 1993
54. Ewen ME, Sluss HK, Whitehouse LL, Livingston DM: TGFO
inhibition of Cdk4synthesis is linkedtocell
cycle arrest.Cell
74:1009, 1993
55. Haber DA, Bernstein A: Rate-limiting steps: The genetics of
pediatric cancers. Cell 64:5, 1991
56. Kadin ME, Cavaille-Col1
MW, Gem R, Massagu.5J, Cheifetz
S, George D: Loss of receptors for transforming growth factor b in
human T-cell malignancies. Proc Natl AcadSci USA 91:6002, 1994
57. Marx J: How cells cycletowardcancer.Science263:319,
1994
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
1995 85: 3594-3601
Early human thymocyte proliferation is regulated by an externally
controlled autocrine transforming growth factor-beta 1 mechanism
MD Mossalayi, F Mentz, F Ouaaz, AH Dalloul, C Blanc, P Debre and FW Ruscetti
Updated information and services can be found at:
http://www.bloodjournal.org/content/85/12/3594.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.