Jul 12, 1982 - hybridization between concanavalin A-activated BALB/c spleen cells and the AKR thymoma BW 5147. Media conditioned by hy- bridoma cells ...
Proc. Natl Acad. Sci. USA Vol. 79, pp. 7455-7459, December 1982 Immunology
B-cell growth factor: Distinction from T-cell growth factor and B-cell maturation factor (B lymphocytes/T-cell hybridomas)
TOMAS LEANDERSON*, ERIK LUNDGRENt, ERIK RUUTH*, HAKAN BORGt, HAKAN PERSSONt, AND ANTONIO COUTINHOt tLaboratory for Cell and Tissue Culture Research, *Institute of Pathology, and +Department of Immunology, UmeA University, S-901 87 UmeA, Sweden
Communicated by Niels K. Jerne, July 12, 1982
ABSTRACT A T-cell hybridoma was derived by somatic cell hybridization between concanavalin A-activated BALB/c spleen cells and the AKR thymoma BW 5147. Media conditioned by hybridoma cells, even at high dilutions (1:1,000) support the growth of lipopolysaccharide-stimulated B-cell blasts but not that of T-cell growth factor (TCGF)-reactive T-cells. This activity, herein designated B-cell growth factor (BCGF), has a Mr of =20,000 and it can readily be separated from TCGF (Mr "30,000) by gel filtration. BCGF is constitutively produced by the hybridoma cells, it is removed from conditioned media by incubation with target cells at +4°C, and it is equally effective on B-cell blasts carrying different major histocompatibility complex and Ig haplotypes. BCGF shows no T-cell replacing factor (TRF) activity, and it is poor in supporting the development of Ig-secreting plaque-forming cells in B-cell blast cultures. Terminal maturation, however, can be induced in BCGF-dependent blasts by addition of conditioned media from normal helper T cell cultures, suggesting that two distinct factors are involved in the helper cell-dependent growth and maturation of B lymphocytes.
We have initiated these studies by systematically screening hybridoma activities in different assays that measure either growth or maturation of all B lymphocytes, regardless of clonal specificity. We chose to study activated rather than resting B cells because of the -overwhelming evidence that initiation of cooperative B-cell responses requires direct cellular interaction (13, 14). In this report we describe a hybridoma clone secreting a factor that supports growth but not maturation of lipopolysaccharide (LPS)-activated B-cell blasts and is devoid of T-cell growth factor (TCGF) and T-cell replacing factor (TRF) activities.
phocytes.
MATERIALS AND METHODS Mice. BALB/c, C57BL/6, C3H/Hej, and C3H/Tif mice of both sexes were bred in our colony and used at 8-12 weeks of age. Cell Culture. BW 5147 cells and hybridoma cultures were kept in RPMI-1640 medium supplemented with 10% fetal calf serum, antibiotics, 50 ,uM 2-mercaptoethanol, and 0.01 M Hepes; they were cultivated at 37°C in a humidified atmosphere. Cell Fusion and Hybridoma Selection. Cell fusion was carried out as described (15). Cells were cloned by seeding cells at 1 or 0.2 cell per culture inmicrotiter plates with a feeder layer of peritoneal cells (limiting dilution). Adsorptions. BCGF-containing supernatant was adsorbed either on LPS-derived or concanavalin A (Con A)-activated blasts. Shortly thereafter, purified blasts were suspended in medium containing 10% fetal calf serum at a concentration of 108/ml and incubated on ice for 1 hr. The blasts were then spun, resuspended at the same concentration in the appropriate dilutions of either supernatant or medium, and kept on ice another hour. All supernatants were then tested in the BCGF assay as described below, in parallel with unadsorbed supernatants, at appropriate dilutions, and in the convenient mixtures. Gel Filtration. Hybridoma supernatant or Con A-conditioned media from normal spleen cells were precipitated at 80% saturation with ammonium sulfate on ice for 2 hr. The precipitate was spun down (10,000 rpm; 15 min) and resuspended in NaCl/0.01 M Hepes, pH 7.3. Samples (1 ml) were applied to a ACA 54 column (LKB, Bromma, Sweden) equilibrated with phosphate-buffered 0.3 M NaCl. The column was eluted with the same buffer at 4°C; fractions (2 ml) were collected, sterilized by filtration through 0.45-,um Millipore filters, and assayed for BCGF and TCGF activity at 1:60 to 1:120 dilutions in different experiments. Assays. The assay for BCGF was carried out on B-cell blasts
The publication costs ofthis article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
Abbreviations: BCGF, B-cell growth factor; BCMF, B-cell maturation factor; Con A, concanavalin A; LPS, lipopolysaccharide; PFC, plaqueforming cells; TCGF, T-cell growth factor; TRF, T-cell replacing factor.
Immune responses are the result of extended clonal growth and maturation ofspecific precursors preexisting in low frequencies. Much work has led to the isolation and characterization of growth factors with specificities for cytotoxic T lymphocytes (1, 2). Ever since 1970, various factors have been reported to interact directly with B lymphocytes (3, 4). Recently, evidence has been presented that activated B cells can grow under the influence of "nonspecific" molecules produced by other cells (5, 6). However, there is no general agreement on the characteristics or physiologic significance of a B-cell growth factor (BCGF). This is most likely due to the widespread belief that antigens are the BCGF (7). In addition, most assays of B-cell responses do not measure growth directly but measure terminal differentiation of putative clonal progenies to antibody production. Another variable of this situation is the source of conditioned media containing competence factors. These are currently produced by complex cell mixtures or, in the best approaches, by mixtures of cloned T cells with ill-defined populations of "antigen-presenting cells. " Furthermore, it is now well established that a single cell type can produce various mediators with a wide range of target cell specificities and biological effects (8). Successful fusions between tumor cells of T origin and activated lymphocytes have been described (9), and established hybridoma clones with T helper cell characteristics have been reported (10-12). In view of these observations, it would appear feasible to select for hybrid lines showing only one type of biological activity-namely, either growth or maturation of B lym-
7455
7456
Immunology: Leanderson et aLPProc. Natl. Acad. Sci. USA 79 (1982)
purified by density centrifugation from spleen cell cultures stimulated with LPS (50 Ag/ml; Escherichia coli 055; B5, Difco) for 20-24 hr as described (6). Blasts were extensively washed and subcultured in microtiter plates at 5 x 104/ml (104 per culture) in the same medium as described above. Hybridoma supernatants were titrated into the assay wells, with medium or BW 5147-conditioned media as negative controls; restimulation with LPS provided the positive control. At various periods of subculture, the cells were pulsed with [3H]dThd (1 tCi per culture; 1 Ci = 3.7 x 1010 becquerels) for 4 hr, harvested, and processed for assay of radioactivity. The assay for B-cell maturation factor (BCMF) was essentially the same but total IgM-secreting plaque-forming cells (PFC) were determined in the protein A plaque assay (16). TCGF assays were performed on T-cell blasts (105/ml), obtained from Con A-stimulated cultures, by titrating the hybridoma supernatants into the assay wells with a standard TCGF preparation as positive control (17). Interferon assays were performed by titrating the supernatants on a subclone of mouse L-929 cells with vesicular stomatitis virus as challenge; NIH G 002-902-026 reference preparation was used as a standard. Assay for TRF activity was carried out by adding the hybridoma supernatant to Mishell-Dutton type cultures modified to microculture conditions (18) and using anti-Thy-i and complement-treated spleen cells (106 per culture). RESULTS Cell Fusion and Screening for Factor-Producing Cells. From a fusion between BW 5147 and splenocytes activated for 18 hr with Con A (5 Ag/ml), 9 growing cultures of 48 seeded were retrieved after the selection procedure. Supernatants from these nine cultures were tested for different biological activities (Table 1). Low interferon activity was detected in two supernatants; threshold BCGF activity was detected in three other supernatants. In this particular fusion, no cultures were identified that showed either TCGF or TRF activity. One of the hybridoma cultures from this experiment (AC 6) was cloned, and the clones isolated were screened for TCGF and BCGF production. Of 34 clones, 3 showed BCGF activity and 1 showed low, but detectable, TCGF activity. One of the three BCGF-producing clones was recloned; of 14 subelones tested, all showed constitutive production of BCGF activity and none showed TCGF activity. One of these (TUH 3) showed high production and favorable growth properties, and it was selected for further studies. The results of cloning indicated that, early after fusion, hybridoma cells are very unstable in regard to the production of biologically active material. Thus, only 3 of 34 clones isolated from a primary culture that should not be expected to contain a high.;multiplicity of different hybrids (9 positive among 48 cultures) were found to continue producing growth promoting Table 1. Screening of hybridoma culture supernatants for various biological activities Culture Primary: AC5 AB3
IFN*
TCGF
TRF
-
-
-
-
-
-
+
AC1
-
-
-
-
AD6
-
-
-
-
AB2
-
-
-
-
+ -
-
-
-
+
+
-
-
-
+
AC4 BB5 AA4 AC6 *
IFN, interferon.
activity for B lymphocytes. The results from the subeloning, on the other hand, suggest that, after 6-8 weeks, producing hybrids are relatively stable. As shown below, the hybrid described here has maintained biological activity for >6 months, both after freezing and thawing and upon continuous culture. The BCGF-producing hybridoma (TUH 3) used in these experiments expressed both alleles of Thy-i, demonstrating that it stems from fusion of BW 5147 with a BALB/c T cell. These results also demonstrate that the productions of growth-supporting activities for activated T and B lymphocytes segregate in different clones and therefore appear to be distinct molecular entities. BCGF: Growth-Promoting Activity for Activated B Cells. As shown previously (19), activated B cell blasts require the continuous presence of a mitotic stimulus for maintenance of growth; this stimulus can be provided by nonspecific factors generated by unrelated helper cell activity (5, 6). We have characterized the ability of media conditioned by hybridoma cells to maintain activated B cell blasts (LPS-derived) in exponential growth. LPS-activated blasts did not maintain growth when re.cultured in normal medium, but they did so if reexposed to LPS or to conditioned medium from TUH 3 hybridoma (Fig. 1). Blast cell growth often was comparable to that induced by LPS, but it also could be considerably lower. This variation most likely is due to heterogeneity in the LPS blast preparation, and it could indicate that not all cycling B cells are reactive to the hybridoma-derived activity. Alternatively, the lower response observed with hybridoma supernatants might be due to depletion ofnutrients in the spent medium. We conclude from these results that TUH 3 cells constitutively produce factors that maintain growth in activated B cells, a biological activity we call BCGF. Fig. 2 shows a number ofexperiments in which various TUH 3 supernatants were titrated in parallel with media conditioned by BW 5147 cells or by a similar hybridoma devoid of BCGF activity. High concentrations of TUH 3 supernatants often were somewhat inhibitory (see below) but between 12% and 1% (vol/ Exp. 1
Exp. 2
A
0
Exp. 3 ,
A~~~~~
104-4
$. 0.
0 AL
tH
A
Al 0
c0
o
0
A
\ 0
BCGF 103 , 0
1
3 1 3 1 Days after subculture
3
FIG. 1. B-cell blasts (104 per culture) purified from 24-hr LPSstimulated cultures were reactivated with medium (o), LPS (o), or 5% TUH 3 supernatant (A).
Immunology: Leanderson et al. A
Proc. Natl. Acad. Sci. USA 79 (1982) C
B
7457
D S 0
a)
s.
104
0.
0
0
104 0 0
*
3
0
103
256 1.5 50 12.5 3
6 1.5 03 256 1.5 12.5 3 0.75 50 12.5 3 Supernatant in medium, % (vol/vol)
vol), a plateau of activity was observed. The activity decreased higher dilutions to reach background levels at 0.02-0.01% (not shown). BCGF activity was characteristic of the supernatants from the TUH 3 cells because neither BW 5147 nor an irrelevant hybridoma produced growth-supporting activity. TUH 3 cells kept in continuous culture produced BCGF activity similar to that obtained from cells frozen and thawed at different
at
times.
Growth factors for activated B cells have been described as major histocompatibility complex-unrestricted and nonspecific (5, 6). In our experiments measuring the activity of a clonally produced factor on polyclonally activated B cells, it was evident that BCGF is not antigen specific and it acted equally well in syngeneic conditions as on targets differing both in major his-
tocompatibility complex and Ig haplotypes (Table 2). BCGF and TCGF Activities Are Carried by Distinct Molecular Entities. TCGF is the product of activated T-helper cells that appears to be a selective growth factor for cytotoxic T lymphocytes (20, 21). Because the TUH 3 hybridoma is derived from populations containing TCGF-producing cells that have previously been used to derive TCGF-producing hybrids (11), it was important to analyze the TUH 3-conditioned media for TCGF activity. The same supernatant preparation that showed high mitotic activity for B-cell blasts failed to maintain proliferation of TCGF-reactive, activated T cells (Table 3). Since mixing experiments excluded mutually inhibitory activities in Table 2. Lack of genetic restriction in the activity of BCGF for various types of B-cell blasts Donor blasts Haplotypes Response, cpm/culture Strain MHC Ig-1 Medium 5% TUH 3 sup. LPS BALB/c d a 2,142 10,514 11,503 1 k C3H/Tif 1,575 10,527 8,030 b b C57BL/6 1,650 14,365 15,279 Blasts purified after 24 hr of LPS stimulation of spleen cells of the indicated strains were recultured as indicated (104 cells per culture). All cultures were pulsed with 1 puCi of [3H]dThd per culture for 4 hr on day 2 of subculture. The results, shown as cpm, represent the mean responses upon restimulation of triplicate cultures. MHC, major histocompatibility complex; Ig-1, Ig-1 heavy chain.
12.5
6 1.5 0.3 3 0.75
FIG. 2. Proliferation of LPS-activated B-cell.blasts recultured (104 cells perculture) in the presence of medium (o), LPS (e), or the indicated mixture of conditioned media prepared from TUH 3 hybridoma cells ( 0, inA, C, and D), an irrelevant hybridoma isolated from the same fusion protocol (A, in B), or BW 5147 thymoma cells (a, in C andD). Supernatants were prepared from TUH 3 cells continuously maintained in culture (A) early after fusion (A) or 6 months later (C) or frozen and thawed ( 0 ) at two different time points (D).
the two supernatants, these results demonstrate that BCGF andTCGF are distinct molecular entities as previously suggested by some reports but denied by others (10, 11). Preliminary experiments aiming at the biochemical characterization of BCGF had revealed that all biological activity could be recovered from sizing columns in a single peak of Mr =20,000. Because murine TCGF has been previously assigned a Mr of 30,000 (20), we sought to separate these two biological activities biochemically. We mixed equal volumes of BCGFcontaining hybridoma supernatant and TCGF-rich Con A-conditioned medium. This mixture was then chromatographed, and the profile of activities recovered from the eluent is shown in Fig. 3. BCGF and TCGF activities eluted at two distinct peaks corresponding to Mrs -20,000 and =30,000, respectively. These observations establish that the B-lymphocyte and cytotoxic T-lymphocyte specific growth factors are two distinct molecular entities. Table 3. Target cell specificity of BCGF and TCGF Growth, cpm/culture T-cell blasts B-cell blasts Restimulation Exp. 2 Exp. 1 Exp. 2 Exp. 1 LPS (50 ,ug/ml) 26,560 ND ND 29,758 TUH 3 supernatant (5%)* 23,523 1,050 434 27,047 TCGF CM (35%)t 4,564 42,233 66,155 2,024 Medium 3,350 1,106 628 4,422 TUH 3 supernatant (5%) + TCGF CM (25%) ND 19,547 ND 52,500 T-cell blasts (2 x 104 per culture) were activated with Con A for 48 hr; B-cell blasts (104 per culture) were stimulated with LPS for 24 hr. The cells were restimulated as shown and pulsed for 4 hr with 1 tkCi of [3H]dThd on day 2 of culture. Results are means of triplicate cultures. ND, not done; CM, conditioned medium. * Conditioned medium from TUH 3 hybridoma cells (105/ml) in exponential growth for 24 hr. t Standard TCGF-containing conditioned medium (supernatant from 24-hr cultures of 5 x 106 rat spleen cells per ml with 5 pug of Con A per ml) supplemented with 10 mg/ml of a-methyl D-mannoside (10 mg/ml) for inactivation of the lectin.
7458
Immunology: Leanderson et al.
20 -
67 43
Proc. Natl. Acad. Sci. USA 79 (1982) 25
13.7
co X 10-
A
x
.
A
.
AAA1
.....A.... .........
pF-1 ce --. Q
10
1
20
30 Fraction
40
50
c).
'FIG. 3. Gel filtration on ACA 54 of mixture of equal volumes of conditioned media from TUH 3 hybridoma and Con A-stimulated normal mouse spleen cells. Fractions were assayed in parallel on B-cell blasts (A) and T-cell blasts (A) with medium (0, 0); TCGF-containing medium (.), and LPS (mnas controls. Mr standards (shown x 10-3) were bovine serum albumin (67,000), ovalbumin (43,000), chymotrypsinogen A (25,000), and RNase A (13,700) all obtained from Pharmacia (Uppsala, Sweden).
BCGF is Specifically Adsorbed onto Activated B Lymphocytes. The B-cell specificity of the TUH 3 activity was further probed by specific binding of the growth-promoting activity to target cells. As shown previously, TCGF can be adsorbed readily on target cytotoxic T-cell blasts (22) which express 1,000-fold more specific receptors for growth factors than do irrelevant cells (23). BCGF-containing supernatants were incubated in the cold with the same number ofeither LPS-activated B-cell blasts or Con A-stimulated T cells. There was a specific loss of growthpromoting activity upon incubation with activated B cells but notwith activated T cells (Fig. 4). This was not due to production of inhibitory substances by target B cells because mixtures of unadsorbed BCGF-containing supernatants and medium incubated with activated B cells under identical conditions retained full activity. On the other hand, removal of BCGF from TUH 3 supernatants revealed a marked inhibitory effect that also must be present in crude supernatants but is masked by 0
W 4-
0
1
Is4 cO
0
103 O 0.25 0.075 0.01 0.125 0.03
0.5
FIG. 4. Appropriate dilutions of TUH 3.supernatant were incubated in the cold for 1 hr with either B or T blasts and then titrated in the BCGF assay. The data show the proliferation of 104 LPS-derived blasts after 48 hr of reculture with medium (o), LPS (e), unadsorbed TUH 3 supernatant (A),' TUH 3 supernatant adsorbed on B-cell blasts (0), TUH 3 supernatant adsorbed on T-cell blasts (*), and unadsorbed TUH 3 supernatant mixed with medium adsorbed on B-cell blasts
Supernatant, % (vol/vol) (0).
1 3 1 3 Days after subculture
FIG. 5. Ig-secreting PFC upon restimulation of LPS-activated B-cell blasts with LPS (e), 10% TUH 3 supernatant (A), or medium (o). (Left) At 104 cells per culture. (Right) At 2 x 104 cells per culture. Parallel proliferative assays on day 2 of restimulation yielded (cpm per culture): for 104 cells per culture, medium, 1,497; LPS, 5 413; and BCGF, 4,178; for 2 x 0wcells perculture, medium, 1,150; LPS, 17,742; and BCGF, 7,293.
the stimulatory effects of BCGF. The nature of this inhibitory activity in crude supernatants is unknown, but its presence might explain why optimal dilutions of TUH 3 supernatant are always
