Macrophage Colony-Stimulating Factor (CSF-1) Gene Expression in Human T-Lymphocyte Clones

By Marie-Martine Hallet, Vincent Praloran, Henri Vie, Marie-Alix Peyrat, Gordon Wong, JoAnn Witek-Giannotti, Jean-Paul Soulillou, and Jean-FranGois Moreau

Macrophage colony stimulating factor (CSF-1) is one of several cytokines that control the differentiation, survival, and proliferation of monocytes and macrophages. A set of 11 human T-cell clones, chosen for their phenotypic diversity, were tested for their ability to express CSF-1 mRNA. After 5 hours of stimulation with phorbol myristate acetate (PMA) + calcium ionophore (Cal), , all T-cell clones expressed a major 4-kb transcript, a less abundant 2-kb transcript, and several other minor species. This pattern of expression is typical for CSF-1 mRNAs. Furthermore, of the two alloreactive T-cell

HE MACROPHAGE COLONY stimulating factor T (CSF-1) belongs to a family of glycoprotein hematopoi- etic growth factors known as the colony stimulating factors (CSFS).’.~ In 1985, Kawasaki et al cloned a cDNA encoding the CSF-1 protein from the pancreatic tumor cell line, MIA-Paca-2. RNA blot analysis of this cell line has shown the existence of extensive heterogeneity in the size of CSF-1 transcripts.’ These different RNA species encode distinct but related forms of CSF-1 proteins, which appear to be transcribed from a single copy gene and arise from alter- nated splicing pathway^.^,'

The CSF family has proliferative and differentiating actions on granulocyte-macrophage (GM) progenitors and it is known that T lymphocytes, in addition to regulating the immune system, are also able to regulate the growth and differentiation of the hematopoietic cells through the re- lease of such soluble Nevertheless, production of CSF-1 by normal T cells has never been conclusively shown. It has been suggested by the presence of CSF-1 mRNAs in lectin-stimulated human peripheral blood lym- phocytes (PBLs) but no effort was made in these studies to eliminate the possibility that contaminating monocytes might be responsible for the CSF-1 production,” the pres- ence of a cDNA encoding this class of molecules from human T-cell lymphotropic virus (HTLV)-positive T-cell line after lectin stimulation (C10-MJ2)6 and the fact that CSF-1 is constitutively produced by the human lymphoblas- toid cell line (CEM-ON).’”

To investigate the possibility that CSF-1 might be ex- pressed by normal T cells, we have tested several different human T-lymphocyte clones from different origins for their

From the INSERM U.211, Facult.? de Mkdecine, Laboratoire d’Hkmatologie, Centre Hospitalier REgional, Nantes, France; and the Genetics Institute Inc, Cambridge, MA.

Submitted December 4, 1989; accepted October IO, 1990. Address reprint requests to M.M. Hallet, PhD, INSERM U.211,

Facult.? de Mbdecine, I rue Gaston Veil, 4403.5 Nantes Ceder, France. The publication costs of this article were defrayed in pari by page

charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section I734 solely to indicate this fact.

0 1991 by The American Society of Hematology. 0006-4971 I91 17704-0023$3.00/0

clones analyzed, only one showed a definitive message for CSF-1 on specific antigenic stimulation, but with delayed kinetics and less efficiency. Both conditions of stimulation induced the release of CSF-1 protein by T cells in the culture medium. Together, these findings demonstrate for the first time that normal T cells are able to produce CSF-1, previous reports being limited to two cases of tumoral cells of the T-cell lineage. 0 1991 by The American Society of Hematology.

ability to express CSF-1 mRNA and secrete CSF-1 protein after stimulation with several different agents.

MATERIAL AND METHODS

Cells. A panel of 11 human T-lymphocyte clones from different origins were used for this study. Five of them were derived from T cells infiltrating skin during acute graft-versus-host disease (GVHD) (Donor A), two of them were obtained from the PBLs of a patient with large granular lymphocyte disorder (LGLD) (Donor B)” and four alloreactive T-lymphocytes clones (ATLC) from a donor C were derived from mononucleated cells invading an irreversibly rejected human kidney allograft.’* The monoclonality of all these cultures, cloned by the limiting dilution technique,’? has been previously confirmed by analyses of T-cell receptor (TCR) genes rearrangements (data not s h o ~ n ) . ” . ’ ~

The phenotypic and functional characteristics of these clones are given in Table 1. They exhibit a wide range of surface marker phenotype (CD4-CD8-, CD4-CD8’, CD4’CD8-, CD4TD8’; cup or yS TCR; cytotoxic or not), and are considered to be representa- tive of the different T-lymphocyte subpopulations.

Because the antigens recognized by these cells are not known, the GVHD and LGLD T-lymphocyte cultures were maintained in the presence of recom- binant interleukin-2 (rIL-2) under polyclonal activation: lo3 to 5 X

lo? cellslwell were plated in microtiter plates (Falcon, Grenoble, France) in the presence of 5 x lo4 irradiated (30 Gy) alloreactive PBLs and 5 x 10’ cells from an irradiated (30 Gy) B-lymphoblas- toid cell line (BLCL), in RPMI 1640 supplemented with 10% human serum, pure rIL-2 (1 nmol/L) (Biogen, Geneva, Switzer- land), indomethacin (1 pLgimL) (Sigma, St Louis, MO) and leucoagglutinin A (1 pgimL) (Pharmacia, Upsala, Sweden).

ATLC cultures were maintained by weekly addition of irradiated (30 Gy) kidney-donor-derived BLCL (D-BLCL)’* at a 1:l ratio in RPMI 1640 supplemented with 10% human serum and pure rIL-2 (1 nmolk).

Responding cells (5 x 10‘imL) in complete medium were stimu- lated either with 20 ngimL Phorbol ester (PMA; Sigma, ref. 81-39) plus 2 nmol/L Calcium Ionophore, A23187 (CaI; Sigma, ref C75-22) or, for ATLC cultures, with irradiated D-BLCL (30 Gy) at a ratio of 1:1, after 6 to 8 days of culture, at which time irradiated allogeneic stimulator cells were absent from cultures. Cells and their media were then harvested at different points of the time course for RNA blot and CSF-1 protein release analyses.

For CSF-1 production per day, cells from T-lymphocyte clone 2F7 were washed twice 7 days after antigenic stimulation and seeded at 0.5 x lo6 cellslml with irradiated (10,000 rads) D-BLCL (2.5 x 10‘ImL) in RPMI medium supplemented with recombinant human IL-2 (rhu IL-2) (1 nmol/L) and 10% human serum (identified as culture medium). At day 4, cultures were diluted 1 to

Culture conditions and cell stimulations.

780 Blood, Vol77, No 4 (February 15), 1991: pp 780-786

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

CSF-1 GENE EXPRESSION IN T LYMPHOCYTES

Table 1. Phenotypic and Functional Characterization of T-cell Clones From GVHD (Donor A), From PBLs From LGLD (Donor E), and From ATLC (Donor C)

Cytotoxic Donor Clone CD2 CD3 CD4 CD8 WT3 1 TCRG1 STCSl TiyA Potential

A A A A A B B C C C C

TM2 TM4 TM6 TM8 TM15 Y 8 Y l O 2B11 4AS 2F7 4 w

+ + + + + + + + + + +

+ + + + + + + + + + +

+ + + + + - - + + + +

4 with fresh culture medium to prevent cell death because of medium exhaustion. Each day the cultures were washed twice, reseeded in fresh culture medium at 0.5 X lo6 cells/mL, and further incubated for 24 hours. The spun supernatant was stored at -20°C for further CSF-1 production per day analysis.

Total cellular RNA was isolated from 30 x lo6 cells by using the guanidium isothiocyanate lysis p roced~re ’~ followed by cesium chloride density gradient centrifugation. Equal amounts of total RNA (0.015 mg as determined at 260-nm wavelength) were denatured for 15 minutes at 60°C in electrophore- sis buffer (20 mmol/L morphomolinopropane sulfonic acid, 5 mmol/L sodium acetate, 1 mmol/L EDTA) containing 6% formal- dehyde (vol/vol) and 50% formamide (vohol), quick chilled on ice and then size-fractionated by electrophoresis through 1% agarose gel containing 6% formaldehyde. Blotting was performed on Gene screen plus filter (NEN, Boston, MA) according to the supplier’s instructions. After baking at 80°C for 2 hours and prehybridization, filters were hybridized for 20 hours at 42°C with ”P-labeled probe in a buffer containing 50% formamide (vohol); 1% sodium dodecyl sulfate (SDS) (wthol), 900 mmol/L NaCl, 50 mmol/L phosphate buffer pH 6.5,2 mmol/L EDTA, 0.08% (wthol) each of bovine serum albumin (BSA), ficoll and polyvinylpyrolidone, 4% dextran sulfate and 0.2 mg/mL herringdenatured sperm DNA. Filters were washed twice for 10 minutes at room temperature and twice for 10 minutes at 65°C in the presence of 0.2 XSSC, 0.1% SDS, and then exposed for 7 hours to 2 days at -80°C on Kodak (Rochester, NY) X - 0 mat X-ray filter using intensifying screens (Cronex, Dupont, USA). The CSF-1 coding region from plasmid 3ACSFRLMCSF was used as a hybridization probe. An actin genomic subclone pHA 4-1 derived from A HA-4 clone“ was used as a house keeping gene hybridization probe to normalize differ- ences in loading amounts of RNA and to verify uniform transfer efficiency. Inserts were multiprimed using alpha [”PI dCTP (ref PB 10205, specific activities: 3,000 Ci/nmoVL; Amersham, Les Ulis, France). Their specific activities were 1 to 3 X lo9 cpm/mg x For rehybridization, probes were washed out with a boiling solution of 0.01% SDS in 0.01 XSSC according to the supplier’s instructions.

CSF-1 release was quantified accord- ing to a previously described procedure” with slight modifica- tion~.’~.’’ Iodinated human rCSF-1 (generous gift of Dr D. Kreum- wieh, Behringwerke Marburg, Germany) was used instead of CSF-1 purified from urine. Briefly, standard CSF-1 preparation, samples to be tested, rabbit anti-CSF-1 serum18 and iodinated CSF-1 were all diluted in normal rabbit serum (NRS) (1:lO in PBS). Two hundred microliters of diluted NRS and the samples to be tested were mixed in 250 p,L of anti-CSF-1 antiserum (1:500) and incubated at 20°C in a fully humidified atmosphere for 24

RNA blot anafysis.

CSF-1 radioimmunoassay.

hours. Fifty microliters of lZsI rhuCSF-1 (10s cpm/tube) were then added to all tubes for 24 hours. After this time, antigen-antibody complexes were precipitated by the addition of 4 mL of PEG (18%) and then centrifugated for 30 minutes at 3,000g. Supernatants were discarded and the tubes were drained off. The pellets were counted in a gamma counter. In this assay, one unit of CSF-1 is equivalent to 0,44 fmol.

Murine bone marrow assays were performed as previously described,2a using bone marrow from femurs of 8- to 10-week-old Balb/c mice. Cells 2 x lo4 were plated in 0.1-mL cultures in microtiter wells in Iscoves medium containing 1.2% methylcellulose (Dow Chemical Company, Woburn, MA), 30% heat-inactivated fetal calf serum (Hazelton, Lanexa, KS), 1% BSA (GIBCO, Grand Island, NY), and either 10% or 2% condi- tioned medium. Each point was assayed in triplicate. Colonies were counted through an inverted microscope on day 7, and were considered to be macrophage colonies on the basis of their morphology.

Alternatively, colony assays were performed in collagen gels as described.” After 7 days of culture, the gels were dried, fixed, and stained with May-Griinwald-Giemsa or butyrate esterases for identification of monocytic colonies.

Murine bone marrow colony assays.

RESULTS

Expression of CSF-1 transcripts in human T cells exposed to phorbol ester + calcium ionophore or to antigenic stimulation. Total RNA extracted from CD4-CD8- y6 TCR cell clones from LGLD, from CD4TD8- and CD4- CD8’ aP TCR cell clones from GVHD (Fig 1) and from CD4’CD8-, CD4-CD8+, and CD4TD8’ ap TCR ATLCs (Figs 2 and 3) were analyzed by RNA blots. The RNA blot analysis showed significant amounts of M-CSF messages in total cellular RNA isolated from the T cell clones 5 or 6 hours after stimulation with PMA + CaI. All cell clones expressed (Fig 2) a major transcript around 4 kb in size, a less abundant message of small size around 2 kb, and several other minor mRNA species.

From Fig 3, it can be seen that hybridization of mRNA from specific-antigen-stimulated 4W ATLC, with the M-CSF-specific cDNA probe, failed to show any significant induction of the expression of M-CSF gene. This result is in contrast to that obtained with the 2F7 ATLC, where the induction of multiple M-CSF transcripts were evident 11 hours after antigenic challenge. In this last case, despite the signal being smaller compared with the previous situation

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

782 HALLET ET AL

TM4 TM6 TM8 TM15 nnnn o t B o ( o o ( o o ( o

7 6 710 TM2 nnn 0 t B o t B o m CSF-1

28S-

23s-

18s-

16s -

ACTIN - where PMA + CaI were used to activate T cells, both the 4-kb and the 2-kb mRNAs were clearly induced by such antigenic stimulation. Similarly, RNA obtained from the pancreatic carcinoma cells MIA-Paca-2 exhibited multiple transcripts after PMA stimulation because of alternative splicing of exon six with 3' untranslated sequences.'

Using CSF-I probe with high specific radioactivity (3 x lo9 cpm/kg) and after prolonged exposure, we could observe only the 4-kb message in D-BLCL after irradiation. How- ever, in the same gel and autoradiograph, the comparison with a T-lymphocyte clone (y6, taken 6 hours after stimula- tion with PMA + CaI) clearly showed a stronger 4-kb signal together with additional bands and a major one around 2-kb for the cell clone. Finally, a weak signal is present in stimulated clones at day 6 to 7 following irradiated D-BLCL addition to the culture. This signal could be observed in the 11 clones tested after a prolonged autoradiography expo- sure and is likely to be caused by the persistence of CSF-1 transcripts for several days after activation.

- 28s

- 23s Fig 1. Northern-blot analysis of 0.015 mg of total RNA ob- - 18s tained from LGLD T-lymphocyte

-16s clones (y6 and y10) and from GVHD T-lymphocyte clones (TM2, TM4, TM6, TM8, and TMl5) unstimulated (0 ) or 6 hours (6) PMA + Cat stimulated. Hybridiza- tion with BACSFRI-MCSF probe (2 days' exposure) and with actin (24 hours' exposure).

Because CSF-1 activity has been already detected in concentrated supernatants of EBV-B cell lines:' we further attempted to evaluate the amount of CSF-1 protein effec- tively present in the culture medium of D-BLCL-stimu- lated T-cell clones.

In some initial experiments, the secretion of CSF-1 by the human ATLCs, 2F7 and 4W, was evaluated by collecting culture supema- tants at 5, 11, 24, 48, 72, and 96 hours of culture after specific-antigenic stimulation with D-BLCL and assaying for CSF activity in a day-7 murine methylcellulose colony assay. In this murine bone marrow assay human GM-CSF and IL-3 have no effect. In addition, RNA blot analysis of the same responder cells with G-CSF-specific probe did not show any signal either after PMA + CaI or after antigenic stimulation (data not shown), therefore precluding any involvement of the G-CSF protein in the growth of murine bone marrow colonies, which were exclusively found to be of the macrophage phenotype. Figure 4 shows that CSF

CSF-I protein production by T-cell clones.

2B11 4AS 2F7

28s-

23S-

18s-

16s-

Fig 2. Northern-blot analysis of 0.015 mg of total RNA obtained from 281 1,4AS, and 2F7 ATLC unstim- ulated (0). 5 hours (5), and 24 hours (24) PMA + Cal stimulated. Hybridization with BACSFR1-MCSF probe (7 hours' exposure) and with actin probe (24 hours' exposure).

ACTIN

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

CSF-1 GENE EXPRESSION IN T LYMPHOCWES 783

4w 2F7 I I

CSF-1 28s

232

181 16$

Fig 3. Northern-blot analysis of 0.007 mg of total RNA ob- tained from 4W and 2F7 ATLC unstimulated (0). 5 hours (5-A), and 24 hours (24-A) stimulated with PMA + Cal and 5 hours (5-8). 11 hours (11-B), and 24 hours (24-8) stimulated with ina- diated donor BLCL. Hybridira- tion with 3ACSFRl-MCSF probe (overnight exposure) and with actin probe (24 hours’ exposure).

ACTIN-

6 I

Y)

activity began to accumulate in the 2F7 culture supernatant by 24 hours after addition of stimulators and the level continued to increase after 96 hours. CSF activity 4W culture medium became significantly detectable 48 hours poststimulation, and further increased. These results fit well with the situation observed at the RNA level for the 4W clone, which did not clearly exhibit a CSF-1 transcript 24 hours after antigenic stimulation, suggesting delayed kinetics of activation. CSF-1 activity was readily detectable at 5 hours after PMA + CaI stimulation of both ATLCs, with no further increase until 24 hours.

CSF-1 radioimmunoassay (RIA) on 2F7 culture superna- tants confirmed that the macrophage-stimulating activity detected in colony assays was indeed caused by immunolog- ically reactive CSF-1 protein. The per day production of CSF-1 protein by 2F7 cells following stimulation with irradiated D-BLCL were serially measured by a this RIA. Results of one experiment are given in Table 2. Culture medium exhibited a backgroung of 120 U/mL, which has been substrated from data. CSF-1 released by 2F7 cells on D-BLCL challenge reached a maximum at day 2 and thereafter decreased to a plateau level until day 7, underlin- ing the close relationship between gene transcription and protein secretion. As a control, we assessed CSF-1 produc- tion under the same conditions by irradiated D-BLCL. Although irradiated D-BLCL produced the protein, it cannot alone account for the total release of CSF-1 in these cultures. To verify that the CSF-1 detected by radioimmu- noassay was biologically active, 2F7-conditioned media were tested in a collagen murine colony assay. Concen- trated supernatants of 2F7 cultures activated with irradi- ated D-BLCL stimulated murine progenitor cells to grow specific anti-rhu CSF-1 rabbit antiserum, which by itself

had no effect on colony growth, significantly reduced the macrophage colony-forming ability of these preparations (Table 3). This observation demonstrates a major involve- ment of CSF-1 derived mainly from T cells in these experiments. In this set of experiments we have checked the monocytic nature of the colonies by staining for butyrate esterase.

DISCUSSION

In the present paper we report on the CSF-1 gene expression and the corresponding protein production by 11 T-lymphocyte clones from different origins exhibiting a large phenotypic diversity.

The biologic effects of CSF-1 are mainly restricted to the mononuclear phagocytic lineage and the specific cell recep- tor that mediates this effect, the protooncogene product c-fms, occur almost exclusively on cells of the same lineage. In humans, the biologically active molecule is found in urine and normal sera. It is produced by non-T-cell tumor cell lines such as MIA-Paca-2 (pancreatic carcinoma),23 GCT-C (metastatic fribrous histiocytoma),” and TPA30.1 (SV40-transformed trophoblast cells)> In addition, it has been recently reported that treatment of mouse bone marrow cultures with culture supernatants from ovarian cancer cells stimulated the production of macrophages colonies and that this effect was because of CSF-1 produc- tion by tumor cells. Conversely, the presence of phagocytic and adherent cells increased the cloning efficiencies of ovarian tumor cells in semisolid medium, whereas their depletion lead to a decreased cloning efficiency demonstrat- ing complex interactions between ovarian tumor cells and normal, phagocytic, tumor-infiltrating ~elIs.2~ In normal

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

784 HALLET ET AL

2 I 3 LL 0 .I-

O

0, L

n E 3 C

L

0 5 11 24 48 72 96

Time (hours)

.I-

0 L P) n E

0 5 11 24 48 72 96

Time (hours)

Fig 4. Time-course CSF-1 accumulation in the supernatant of 2F7 and 4W ATLCs after stimulation by PMA + Cal(0) and with irradiated D-BLCL (A). The CSF-1 level is measured by the number of murine bone marrow colonies (CFU-M) per 2 x 10' plated, which proliferate in response to 1:lO dilutions of the supernatants. All the growing colonies were macrophages. Lymphocyte supernatants stimulated by PMA + Cal were dialysed before assay. Data points represent means of triplicate determinations and standard error of mean (SEM) is given. The negative control (at time 0) is D-BLCL cell culture superna- tant.

physiologic circumstances CSF-1 has been implicated in placental developmentz5 and several normal cell types, including monocytes/macrophages, fibroblasts, stromal cells, and endothelial cells produce CSF-1 either constitutively or after tim mu la ti on.^"^^ Until recently, the T-cell lineage was thought to be incapable of CSF-1 production.

The concept that T cells do not produce CSF-1 has recently been questioned. Recent reports have shown that PMA-stimulated PBLs either produced the protein"." or expressed the mRNA for this factor.6 However, in these studies, no details were presented on the likely contamina- tion of the T-cell population by monocytes and, therefore, no definitive conclusion could be drawn concerning the ability of T lymphocytes to produce CSF-1. Nonetheless, it is clear that some T-cell lines of leukemic origin either transcribed this gene or secrete the protein in the culture supernatant. This result has been reported for C10-MJ2

Table 2. CSF-1 Production Per Day by 2F7 ATLC Stimulated With Irradiated D-BLCL and by Irradiated D-BLCL Alone

Productions Per Day of CSF-1 IUImLI

Day After 2F7 Antigenic + Irradiated

Stimulation D-BLCL Irradiated

D-BLCL

495 600 250 280 340 240 210

40 110

0 30 40 60 74

CSF-1 protein was quantified by a radioimmunoassay as described in Materials and Methods and expressed in units per milliliter (1U = 0.4 fmol). The culture medium exhibited a background of 120 U/mL, which has been subtracted from data.

Irradiated D-BLCL (10,000 rads) at 2.5 x 10' cells/mL were incubated in 5 mL culture medium in plastic culture flasks. Once every 24 hours the incubation medium was removed and stored frozen (-20°C) before RIA. The cells were washed twice (PBS) and resuspended in fresh culture medium and incubation continued for a further 24 hours. This procedure was repeated over a period of 7 days.

cells, an HTLV-1-positive T-cell line, following lectin stimulation6 and also for a clone of the lymphoblastoid T-cell line CEM-ON, which constitutively secreted CSF-1 in sufficient quantity to allow its identification through amino-terminal microsequencing." However, in this latter case it is not known whether the CSF-1 production by CEM-ON truely reflects a normal behaviour of T cells during the immune response because the factor production was not enhanced after lectin stimulation. These two pieces of data suggested that T cells produce CSF-1 that would, in addition to regulating the immune and/or inflammatory systems, regulate growth and differentiation of hematopoie- tic cells.

CSF-1 expression by malignant T-cell lines prompted us to test normal T-cell clones for the ability to produce this cytokine. For this purpose, we exploited alloreactive T-cell clones that we had previously generated from kidney allograft infiltrating cells (ATLCs)'* from GVHD skin patients or LGLD" to examine the potential CSF-1 gene transcription from these T-cell clones representing a variety

Table 3. Monocytic Colony Formation Induced by 2F7-Conditioned Medium

Conditions Normal Rabbit Rabbit of Stimulation Serum 1:100* Anti-CSF-l 1:100*

Irradiated B cells: 24 h 33 lr 3 2 2 1 Irradiated B cells: 72 h 27 ? 2 0

rhuCSF-1 Behring (500 U/mL) 32 ? 5 3 2 1

"Murine bone marrow cells, 1.2 x lo4, were plated in 0.3 mL collagen culture in 24 wells containing 10% of 2F7-conditioned medium, previ- ously concentrated 5 to 10 times on Minicon macrosolute concentrator (Amicon, Danvers, MA). Cultures were incubated with either NRS or anti-CSF-1 antiserum at a 1 :IO0 final dilution as describedtg for 7 days at 37°C with 5% CO, in air.

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

CSF-1 GENE EXPRESSION IN T LYMPHOCYTES 785

of distinct phenotypes. Of the 11 clones tested all contained significant amounts of CSF-1 mRNA when stimulated by PMA + CaI. The peak of CSF-1 mRNA appeared at about 5 hours after stimulation and rapidly decreased thereafter. Although this type of stimulation is not “physiological” in nature, it does, however, demonstrate that whatever the T-cell surface phenotype, there is no block of the CSF-1 gene transcription. This result is not the case for G-CSF, for example, with the same panel of T-cell clones under identical culture conditions (data not shown). For ATLCs where the molecular targets are known,3o the specific antigenic challenge also led to mRNA induction but with slower kinetics and the peak expression was found around 11 hours postaddition of the irradiated stimulators. Eight days after the last stimulation with lectin or antigen, the cultures expressed a low level of IL-2 receptors at their surface and ceased pr~liferating.~’ In this “resting” state, the 4-kb transcript was still detectable, suggesting either a long half-life for these messenger RNAs or a constitutive transcription of the CSF-1 gene at low rate, a situation similar to that mentioned earlier for monocytes.29

After both specific antigenic stimulation and PMA + CaI exposure, the transcript accumulation preceded the release of a M-CSF activity by ATLCs, as tested in the murine standard bone marrow assay. However, for one ATLC (4W) the Northern blot was not conclusive for the presence of CSF-1 transcripts 24 hours after specific allogenic stimu- lation. Nevertheless, CSF activity was detectable in culture supernatant at 48, 72, and 96 hours, suggesting, therefore, delayed transcription of the CSF-1 message in these cells. This colony stimulating activity was identified as immunolog- ically and biologically active CSF-1 by a specific RIA, by the exclusive monocytic colony growth, and by its abrogation by a specific anti-CSF-1 rabbit antiserum. Recently, spontane- ously outgrown Epstein Barr Virus (EBV)-cell lines and normal activated B lymphocytes have been shown to pro-

duce CSF-1.” We were thus concerned by the fact that the CSF-1 activity we detected in our cultures of pure T-cells stimulated with irradiated D-BLCL could have been due to release by B-cell line. Two lines of evidence argue against this hypothesis. First, the comparison between 24 hours serial analysis of the CSF-1 secretion of antigenically stimulated T-cell clones and irradiated D-BLCL CSF-1 secretion measured by radioimmunoassay shows that irradi- ated D-BLCL alone produced only a low level of this factor and that the presence of 2F7 lymphocytes is predominantly responsible for the CSF-1 release observed in the superna- tant of T cells after antigenic stimulation. Second, the 2F7 clone (CD8+) used in this set of experiments is highly cytotoxic against this BLCL line. It gives 30% to 40% of specific chromium release in a standard 4-hour assay at effectorharget ratio of l/1.l2 In this context, the above mentioned results showing some CSF-1 production by irradiated D-BLCL are clearly a maximal estimates because in these experiments the cytotoxic T cells were absent. However, the kinetics of the CSF-1 release with both stimulations is different, PMA + CaI inducing a rapid transcription of CSF-1 gene. The property of T cells to transcribe and produce CSF-1 after activation may be important in vivo as well. The ability of T cells to produce CSF-1 provides yet another mechanism for lymphocytes to regulate cells of the mononuclear/phagocytic series. In this respect, it is conceivable that CSF-1 production by the activated T cells present at the site of the rejected allograft might participate in the rejection process. However, this conclusion cannot be drawn from the present experiments in the case of GVHD and LGLD because we could not test these T-cell clones for CSF-1 production stimulated with their specific antigens because they are currently unknown and, second, the cells from which these clones were derived did not bear (skin biopsies) any activation marker (IL-2 receptor [IL-2-R], Class I1 antigens) in vivo.

REFERENCES

1. Metcalf D: The Hematopoietic Colony Stimulating Factors. New York, NY, Elsevier, 1984

2. Metcalf D: The granulocyte-macrophage colony stimulating factors. Science 229:16, 1985

3. Metcalf D: The molecular biology and functions of the granulocyte-macrophage colony stimulating factors. Blood 67:257, 1986

4. Ralph P, Warren MK, Ladner MB, Kawasaki ES, Boosman A, White TJ: Molecular and biological properties of human macrophage growth factor, CSF-1. Cold Spring Harb Symp Quant Biol51:679,1986

5. Kawasaki ES, Ladner MB, Wang AM, Van Arsdell J, Warren MK, Coyne MY, Schweickart VL, Lee MT, Wilson KJ, Boosman A, Stanley ER, David PR, Mark F: Molecular cloning of a complementary DNA encoding human macrophage-specific colony stimulating factor (CSF-1). Science 230:291,1985

6. Wong GG, Temple PA, Leary AC, Witek-Giannotti JS, Yang YC, Ciarletta AB, Chung M, Murtha P, Kriz R, Kaufman RJ, Ferenz CR, Sibley BS, Turner KJ, Hewick RM, Clark SC, Yanai N, Yokota H, Yamada M, Saito M, Motoyoshi K, Takaku F: Human CSF-1: Molecular cloning and expression of 4-Kb cDNA encoding the human urinary protein. Science 235:1504, 1987

7. Ladner MB, Martin GA, Noble JA, Nikoloff DM, Tal R, Kawasaki ES, White TJ: Human CSF-1: Gene structure and alternative splicing of mRNA precursors. EMBO J 6:2693,1987

8. Griffin JD, Meuer CS, Schlossman SF, Reinherz EL: T cell regulation of myelopoiesis: Analysis at a clonal level. J Immunol 133:1863,1984

9. Metcalf D: The molecular control of cell division, differentia- tion commitment and maturation in haemopoietic cells. Nature 339:27, 1989

10. Masayuki T, Hong YM, Yasuda S, Takano M, Kawai K, Nakai S, Hirai Y: Macrophage colony-stimulating factor is pro- duced by human T lymphoblastoid cell line, CEM-ON Identifica- tion by amino-terminal acid sequence analysis. Biochem Biophys Res Commun 152:1401,1988

11. Vi6 H, Chevalier S, Garand R, Moisan JP, Praloran V, Devilder MC, Moreau JF, Soulillou JP: Clonal expansion of lymphocytes bearing the second (gamma-delta) T cell receptor in a patient with large granular lymphocyte disorder. Blood 74:285, 1989

12. Moreau JF, Bonneville M, Peyrat MA, Godard A, Jacques Y, Desgranges C, Soulillou JP: T lymphocyte cloning from rejected

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

HALLET ET AL

human kidney allografts. Growth frequency and functional/ phenotypic analysis. J Clin Invest 78:874, 1986

13. Moreau JF, Miller RG: Growth at limiting dilution of human T cell colonies from T cell depleted peripheral blood leukocytes. J Immunol1301139,1983

14. Bonneville M, Moisan JP, Moreau JF, Bouyge I, Hallet MM, Lefranc MP, Soulillou JP: TRG alpha, beta and gamma gene rearrangements in human alloreactive T cell clones extracted from a rejected kidney. Transplant Proc 20:196,1988

15. Chirgwin JM, Przybyla AE, Mc Donald RJ, Rutter WJ: Isolation of biologically active ribonuclease acid from sources enriched in ribonuclease. Biochemistry 185249,1979

16. Khalili K, Weinmann R: Actin mRNAS in Hela cells. Stabilization after adenovirus infection. J Mol Biol 180:1007, 1984

17. Das SK, Stanley ER, Guilbert LJ, Forman LW: Human colony stimulating factor (CSF-1) radioimmunoassay: Resolution of three subclasses of human CSFs. Blood 58:630,1981

18. Gilbert HS, Praloran V, Stanley ER: Increased circulating CSF-1 (M-CSF) in myeloproliferative disease: Association with myeloid metaplasia and peripheral bone marrow extension. Blood 74:1231, 1989

19. Praloran V, Martin-Thouvenin V, Berrada L, Moreau A, Lanotte M: Intralymphnode injection of human monocyte macro- phage colony-stimulating factor (CSF-1) as a method of obtaining high titer anti-CSF-1 antibodies. Eur Cytokine Net (in press)

20. Wong GG, Witek-Giannotti J, Temple PA, Kriz R, Ferenz C, Hewick RM, Clark SC, Ikebuchi K, Ogawa M: Stimulation of murine hemopoietic colony formation by human IL-6. J Immunol 140:3040,1988

21. Lanotte M: Terminal differentiation of hematopoietic cell clones cultured in tridimensional collagen matrix: In situ cell morphology and enzyme histochemistry study. Biol Cell 50:107, 1984

22. Reisbach G, Sindermann J, Kremer JP, Hiiltner L, Wolf H,

Dormer P: Macrophage colony-stimulating factor (CSF-1) is ex- pressed by spontaneously outgrown EBV-B cell lines and activated normal B lymphocyte. Blood 74:959,1989

23. Wu MC, Cini JK, Yunis AA: Purification of a colony stimulating factor from cultured pancreatic carcinoma cells. J Biol Chem 254:226,1979

24. Ramakrishnan S, Xu FJ, Brandt SJ, Niedel JE, Bast RC, Brown F L Constitutive production of M-CSF by human ovarian and breast cancer cell lines. J Clin Invest 83:921, 1989

25. Pollard JW, Bartocci A, Arceci R, Orlofsky A, Ladner MB, Stanley ER: Apparent role of macrophage growth factor, CSF-1, in placental development. Nature 330484,1987

26. Vellenga E, Rambaldi A, Ernst TJ, Ostapovicz D, Griffin JD: Independent regulation of M-CSF and G-CSF gene expression in human monocytes. Blood 71:1529,1988

27. Fibbe WE, Van Damme J, Billiau A, Duinkerken N, Lwrvink E, Ralph P, Altrock BW, Kanshansky K, Willenze R, Falkenburg JHF: Human fibroblasts produce G-CSF, M-CSF and GM-CSF following stimulation by IL-1 and poly (r1)-poly (rc). Blood 72:860,1988

28. Fibbe WE, Van Damme J, Billiau A, Goselink HM, Voogt PJ, Van Eeden G, Ralph P, Altrock B, Falkenburg JHF: IL-1 induces human marrow stromal cells in long-term culture to produce G-CSF and M-CSF. Blood 271:430,1988

29. Sieff C A Hematopoietic growth factors. J Clin Invest 79:1549, 1987

30. Bonneville M, Moreau JF, Blockland E, Pool J, Moisan JP, Goulmy E, Soulillou JP: T lymphocyte cloning from rejected human kidney allograft: Recognition repertoire of alloreactive T cell clones. J Immunol141:4187, 1988

31. Jacques Y, Le Mauff B, Godard A, Olive D, Moreau JF, Soulillou JP: Regulation of interleukin 2 receptor expression on a human cytotoxic T lymphocyte clone: Synergism between alloanti- genic stimulation and interleukin 2. J Immunol 136:1693, 1986

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom

1991 77: 780-786  

MM Hallet, V Praloran, H Vie, MA Peyrat, G Wong, J Witek-Giannotti, JP Soulillou and JF Moreau human T- lymphocyte clonesMacrophage colony-stimulating factor (CSF-1) gene expression in 

http://www.bloodjournal.org/content/77/4/780.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American

For personal use only.on October 3, 2017. by guest www.bloodjournal.orgFrom