- Email: [email protected]
LPS augments human B-cell differentiation by direct stimulation of PWM-responsive B cells
CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
44, 259-271 (1987)
LPS Augments Human B-Cell Differentiation by Direct Stimulation of PWM-Responsive B Cells1 STEPHEN J. ANDERSON~ANDALEXANDER R. LAWTON~ Departments
of Microbiology
and Pediatrics, Vanderbilt University Nashville, Tennessee 37232
School
of Medicine,
Bacterial lipopolysaccharide (LPS) augments production of IgM and IgG by two- to seven-fold in cultures of peripheral blood lymphocytes (PBL) stimulated by pokeweed mitogen (PWM), but only if monocytes are rigorously depleted. When PBL were separated into adherent cell (AC), B-cell-enriched, and T-cell-enriched fractions, pulsed with LPS, and recombined in culture with PWM, increased generation of plasma cells was seen only in cultures containing LPS-treated B cells. This effect of LPS appears to be independent of soluble factors. Supematants from LPS-stimulated B cells or AC did not consistently increase PWM responses when cultured with fresh B cells in the presence of polymyxin B. Furthermore, pulsing of B cells with purified interleukin 1 from two different commercial sources failed to augment PWM-induced differentiation. When B cells were depleted of surface IgD (sIgD)-bearing cells by panning, no effect on LPSmediated augmentation of PWM-driven differentiation was seen. B cells were also fractionated by rosetting with mouse erythrocytes. Treatment of BMR+ cells with LPS did not induce them to respond to PWM, while treatment of BMR- cells with LPS augmented generation of plasma cells. These results indicate that LPS acts directly to augment differentiation of PWM-responsive B cells, rather than recruiting sIgD+, BMR+ cells to become PWM responsive. 0 1987 Academic PRSS. IX.
INTRODUCTION
Polyclonal B-cell activators have been used for in vitro study of B-cell differentiation in normal individuals and patients with a variety of diseases for more than a decade. The great majority of studies have used pokeweed mitogen (PWM)4 as the stimulatory agent (1, 2). The induction of B-cell differentiation by PWM is regulated by interacting sets of helper and suppressor T cells, mimicking B-cell respnses to T-dependent antigens (2-6). Because the PWM response is not restricted by histocompatibility differences 1 Supported by NIH Grant AI-17996. 2 Present address: Department of Immunology and Medical Microbiology, University of Florida, Gainesville, FL. 3 To whom all correspondence should be addressed. 4 Abbreviations used: LPS, lipopolysaccharide; PWM, pokeweed mitogen; PBL, peripheral blood lymphocytes; MNC, mononuclear cells; NAC, nonadherent cells; AC, adherent cells; sIg and cIg, surface and cytoplasmic immunoglobulin; BMR+, B cells forming rosettes with mouse erythrocytes; BMR-, B cells not forming rosettes with mouse erythrocytes; DI, differentiation index; ANAE, (Ynaphthyl acetate esterase; IL-l, interleukin 1; PBS, phosphate-buffered saline; FCS; fetal calf serum; PHA, phytohemagglutinin; Con A, concanavalin A; PFC, plaque-forming cell; EBV, Epstein-Barr virus; HBSS, Hanks’ balanced salt solution; ELISA, enzyme-linked immunosorbent assay; PC, plasma cells; MRBC, mouse red blood cells; ISC, immunoglobulin-secreting cells. 259 0090-1229187 $1.50 Copyright Q 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.
260
ANDERSON
AND
LAWTON
between B and T cells, cocultures of patient and normal cells may be used to dissect the cellular basis of differentiation defects. In such assays it has been demonstrated that B cells from the majority of the patients with common varied hypogammaglobulinemia (7), B cells from blood of normal newborns (6, 8), or bone marrow of adults (9), and monoclonal B cells from patients with chronic lymphocytic leukemia (10) have a very limited differentiation capacity with respect to circulating B cells from normal adults. It is now recognized that only a small fraction of the recirculating B-cell pool is capable of differentiating in response to PWM (11). These cells lack receptors for mouse erythrocytes (12), have little or no surface IgD (sIgD) (13), and are relatively large with low buoyant density (13, 14). Viewed from this perspective the “intrinsic defects of B-cell differentiation” observed in hypogammaglobulinemic patients and in the other settings mentioned above may simply reflect the absence of a minor population of preactivated PWM-responsive B cells among circulating lymphocytes. To better define the nature of these B-cell defects a method of inducing polyclonal differentiation of the majority of B cells, particularly those which do not respond to PWM, is needed. One approach is to use combinations of polyclonal activators which may act on different subpopulations of B cells. Bacterial lipopolysaccharide (LPS) is capable of stimulating human-B-cell differentiation directly (14- 16) and of augmenting the T-cell-dependent response to PWM (17). Observations of Dagg and Levitt suggested that LPS and PWM act on different subpopulations of B cells, distinguishable by buoyant density (14). We have investigated the ability of LPS to activate B cells which are otherwise unresponsive to PWM. Our results indicate that LPS augments PWM-induced differentiation by direct interaction with B cells, but that it does not stimulate a significant number of cells which bear sIgD or receptors for mouse erythrocytes to respond to PWM in the presence of T cells. MATERIALS
AND METHODS
Cell separation. Peripheral blood mononuclear cells (MNC) were isolated from blood of healthy adult donors by centrifugation on Ficoll-Hypaque gradients. Nonadherent cells (NAC) and adherent cells (AC) were separated by consecutive incubations on glass and plastic petri dishes for 1 hr at 37°C. AC were recovered by scraping the glass plates with a rubber policeman. Monocytes were identified by staining for a-naphthyl acetate esterase (ANAE) activity (Sigma Histozyme Kit 90-Al). NAC were separated into B-enriched (B cells) and T-enriched (T cells) fractions by one round of rosetting with 2-aminoethylisothiuronium bromidetreated sheep erythrocytes and centrifugation on Ficoll-Hypaque gradients (18). B-enriched cells contained 9.8 ? 14% ANAE+ (n = 13) cells and 12.2 -+ 9% OKT3+ cells (n = 9), respectively, after a single round of rosetting. Adherent cells contained 67 2 16% ANAE+ cells (n = 8). Depletion of slgD-bearing ceils. The B-cell fraction was depleted of sIgDbearing cells by panning on plastic petri dishes (Fisher 8-757-12) precoated with anti-human IgD (6 chain-specific, Southern Biotechnology Associates, Bir-
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
261
mingham, AL) (19). The plates were washed three times with PBS before use. B cells in PBS + 5% FCS at ~7 x 10‘Yml were incubated for 70 min at 4°C in 3 ml cell suspension per 100 x 15mm plate. Nonadherent cells (sIgD-) were decanted and the plates were rinsed twice with cold PBS + 1% FCS. This process was repeated once to adequately deplete sIgD-bearing cells. Separation of BMR- and BMR+ cells. B cells were fractionated according to their ability to form rosettes with neuraminidase-treated mouse erythrocytes (MRBC) as previously described (20). B cells and MRBC were mixed, and rosetting (BMR+) and nonrosetting (BMR-) cells were separated by centrifugation on Ficoll-Hypaque gradients. Rosettes were disrupted by hypotonic lysis of MRBC. Cells were washed twice in HBSS + 2% FCS before culture. Culture conditions. Cells were cultured in RPM1 1640 medium supplemented with 2.0 mM L-glutamine, 5 x low5 M 2-mercaptoethanol, 50 kg/ml gentamicin, and 10% FCS (complete medium). Triplicate cultures were established in 96-well microtiter plates with flat-bottom wells (Costar 3596). For the first set of experiments each well contained lo5 MNC or lo5 NAC in 0.2 ml complete medium. AC (5 x 103) were added to NAC cultures to ensure the presence of enough monocytes to support a response to PWM (21). For all subsequent experiments each well contained 5 x lo4 B cells, 5 x lo4 T cells, and 5 x lo3 AC in 0.2 ml complete medium. PWM was used at a 1:200 dilution of stock (Gibco lot No. lON9424, 13Kl331). Cultures were incubated at 37°C for 7 days in a humidified atmosphere with 5% CO, before assay. Pulsing of cells with LPS. In order to define the role of B cells, T cells, and monocytes in the response to LPS, cell populations were separated and pretreated with LPS before recombination and culturing with PWM as described above. To accomplish this “pulsing,” cells were resuspended at 106/ml in complete medium this LPS (Escherichia coli Olll:B4, Sigma Chemicals) and incubated in 24-well plates at 37°C. Control cells were incubated as described in complete medium without LPS. Pulsed cells were harvested and washed once before further use. Production of interleukin 1 (IL-I)-containing supernatants. To stimulate production of IL-l, B cells and AC were cultured in 24-well plates at 106/ml in complete medium with 1.0 l.&rnl LPS (22). After 48 hr, culture supernatants were harvested, centrifuged at 1500g for 15 min, and filtered (Millex GV 0.22 pm syringe filters, Millipore Corp.). Supernatants were stored at - 20°C. Purified human IL-l preparations were obtained from Genzyme Corp. and Collaborative Research, Inc. Supernatants were assayed for IL-1 activity using a thymocyte costimulation assay (22). Thymocytes from C57B1/6 or C3H/HeJ mice were cultured in complete medium with 5% FCS at 1.5 x lo6 cells/O.2 ml/well in microtiter plates with flat-bottom wells (Costar 3596). Cells were stimulated with 1.0 kg/ml PHA (PHA-P, Difco) or 0.25 pg/ml Con A (Calbiochem) and serial dilutions of test supernatants for 72 hr. Cultures were pulsed with 0.5 &i/well [3H]thymidine (New England Nuclear, Boston, MA) for the last 16 hr, harvested onto glass fibers filters, and counted in a Beckman liquid scintillation counter. For pulsing with IL-l-containing supernatants, B cells were resuspended in undiluted super-
262
ANDERSON
AND
LAWTON
natant at 106/ml and cultured in 24-well plates as described above. In some experiments, polymyxin B (Aerosporin. Burroughs Wellcome Co., Research Triangle Park, NC) was added at a final concentration of 20 ug/ml. Assay for differentiation. Differentiation was determined by measurement of secreted IgM and IgG by ELISA as described (20) or by enumeration of cells containing cytoplasmic immunoglobulin (cIg+) by immunofluorescence (3). For the latter, slides were examined with a Leitz Orthoplan microscope with epi-illumination. At least 500 cells or 20 cIg+ cells were counted per slide. The total number of plasma cells (PC) per well was determined using the formula -=PC well
fr. cIg+ cells x No. cells recovered from pooled wells No. wells
used several donors whose response to PWM varied considerably in magnitude. In order to facilitate comparison of such experiments, the results are expressed as differentiation indices. The differentiation index (DI) is defined by Differentiation
DI = Statistics.
index. Our experiments
response of cultures with LPS-pulsed cells to PWM response of cultures with no LPS-pulsed cells to PWM’
Data were analyzed using the Student t test. RESULTS
Initial experiments were performed to determine the effect of LPS on PWM-induced Ig production by human peripheral blood mononuclear cells. In our hands, LPS alone, in concentrations up to 100 ug/ml, did not stimulate significant levels of Ig production. In agreement with others, we found that coculture with LPS and PWM had an inconsistent effect on the responses of unfractionated MNC. This may be due to stimulation by LPS of large numbers of monocytes present in populations of human peripheral blood MNC (15, 16, 23). Effect of LPS on monocyte-depleted peripheral blood lymphocytes. Data in Table 1 are from experiments in which monocytes were rigorously depleted from human MNC by adherence to glass or plastic petri dishes. Unfractionated MNC contained 25-35% monocytes by a-naphthyl acetate esterase staining. After depletion, the nonadherent cells were ~5% ANAE+ . MNC or NAC were incubated with or without LPS, washed, and cultured with PWM. Secretion of both IgM and IgG were consistently enhanced in experiments where monocyte-depleted cells (NAC) were pulsed with LPS. Differentiation indices ranged from 1.8 to 6.1 for IgM production and 1.1 to 15.2 for IgG production. While MNC showed enhanced responses to PWM in some experiments when pulsed with LPS, the effect of LPS was inconsistent. Effect of LPS on separated B cells, T cells, and adherent cells. To determine with which cell population LPS interacts to stimulate B-cell differentiation, MNC were separated into NAC and AC, and NAC were further separated into B-cellenriched (B cells) and T-cell-enriched (T cells) fractions. Results of eight experiments in which B cells, T cells, and AC were separately incubated overnight with
LPS AUGMENTS
HUMAN
B-CELL
263
DIFFERENTIATION
TABLE 1 EFFECT OF LPS PULSE ON PWM-INDUCED IMMUNOGLOBUL~N PRODUCTION Immunoglobulin MNC Donor I 2 3 4 5 6
kM 298 314 370 980 580 11.5
production (t&ml)
MNC=Ps” I&
IN
516
270 448 325 340 786 2964
360 570 592 330 536
W 165 940 150 65 464 3972
NACLPS”
NAC I&f
W
298 ndb 660 400 582 1664
728 1364 374 140 210 4428
IgM
I&s
531 970 1670 1,612 1,970 10,316
1996 1464 4064 2130 886 4804
Note. MNC or NAC (103 were cultured with PWM in flat-bottom microtiter wells for 8 days. AC (5 x 10)) were added to cultures of NAC. Supernatants from triplicate cultures were pooled and assayed for secreted Ig by ELISA. 0 Cells were pulsed with 10 pg/ml LPS for 12 hr before culturing. b nd, none detectable.
or without LPS, washed, and recombined in microculture with PWM are summarized in Fig. 1. In all experiments, using several different donors, enhancement of PWM-stimulated plasma cell generation was seen only when the B-cell fractions were pulsed with LPS. LPS treatment of B cells increased the number of IgM plasma cells by an average of 4.7-fold (range, 2-9) and IgG plasma cells by an average of 3.3-fold (range, 2-6) compared to control cultures. There was no significant increase in plasma cells of either isotype in cultures reconstituted with LPS-treated T cells or AC. Mean DI values for LPS-treated B cells differed from those of LPS-treated T cells or AC at the 95% confidence level. Data from representative experiments presented in Table 2 show that increases in absolute numbers of plasma cells reflect increases in the percentage of cIg+ cells per culture and not merely recovery of greater numbers of cells. This observation holds true for all subsequent experiments as well. Optimal conditions for LPS pulsing. Optimal conditions for exposure of B cells to LPS were determined. In three of six experiments, pulsing of B cells for as brief a period as 1 hr caused enhancement of PWM-induced plasma cell generation of 15fold or greater. However, in all experiments, pulses of 12 hr or more gave the greatest degree of stimulation (data not shown). Experiments using LPS over a five-log concentration range revealed a threshold rather than a linear relationship between stimulation and concentration (Table 3). Stimulation was seen when B cells were pulsed with as little as 0.1 &ml LPS for 12-16 hr. and 10 kg/ml gave a consistently high response. Effect of LPS on small resting B cells. We wished to determine if the enhancement of PWM-induced plasma cell generation by LPS was due to recruitment of otherwise unresponsive B cells which bear receptors for mouse erythrocytes and surface IgD. In these experiments, B-cell subpopulations were isolated, separately pulsed with LPS, washed, and reconstituted with T cells and AC in microculture with PWM.
264
ANDERSON
AND LAWTON
Depletion of slgD-bearing cells. Table 4 summarizes data from experiments in which sIgD- B cells were prepared by two rounds of panning on anti-s-coated petri dishes. Unfractionated B cells contained approx 27% sIgD+ cells by direct immunofluorescence. The sIgD-depleted populations contained ~0.4% sIgD+ cells. In each of these experiments, removal of sIgD-bearing cells did not diminish the effect of pulsing with LPS. We were not able to recover sufficient numbers of sIgD+ cells from the panning plates to carry out the reciprocal experiments. Response of BMR- and BMR+ cells. Results of a representative experiment in which B cells were separated according to their ability to form rosettes with MRBC are shown in Fig. 2. The unfractionated B-cell populations contained approx 20% BMR+ cells, similar to the number of sIgD+ cells. After separation, the yield of BMR- cells and BMR+ cells was approximately 10: 1. In agreement with previous studies, BMR- cells gave a significant plasma cell response to PWM, while BMR+ cells failed to respond. Exposure of BMR- cells to an overnight pulse with LPS markedly enhanced generation of plasma cells of both IgM and IgG classes. Treatment of BMR+ cells with LPS caused a slight increase in plasma cell generation, but the level of their response, in absolute number of plasma cells, remained l-2 orders of magnitude below that of BMR- cells. Dif-
IgM Plasma
Cells
2
T CELLS
B CELLS LPS
PULSED
AC CELLS
FIG 1. Augmentation of B-cell differentiation by LPS is mediated by interaction with B cells. Bcell, T-cell, and adherent cell populations were cultured for 12-16 hr in complete medium with or without 10 &ml LPS. The cells were then washed and recombined in culture such that only one or none of the three populations had been exposed to LPS and cultured an additional 8 days with PWM. Each microwell contained 5 x 104 B, 5 x 104 T, and 5 x lo3 adherent cells. Cells from triplicate cultures were pooled and assayed for expression of cytoplasmic immunoglobulin. The results shown are the means + SD of differentiation indices, calculated as described under Materials and Methods, for eight separate experiments. Significant augmentation of the IgM and IgG response occurred only in culture containing LPS-pulsed B cells (P < 0.05).
LPS AUGMENTS
HUMAN
B-CELL
265
DIFFERENTIATION
TABLE 2 LPS AUGMENTATION OF PWM-INDUCED PLASMA CELL GENERATION % cIg+ cells LPS pulse
I@
PC/well x lO-3 W
I&f
I&
Experiment
1
None B cells T cells AC
1.4 5.4 1.2 1.0
1.3 4.3 0.4 0.4
1.3 23.4 5.8 3.5
6.8 18.7 1.9 1.4
Experiment
2
None B cells T cells AC
4.0 22.0 4.0 3.0
4.0 11.0 3.0 2.0
6.2 34.5 6.7 4.1
6.2 17.3 5.0 2.7
Experiment
3
None B cells T cells AC
9.0 25.0 9.0 12.0
8.0 17.0 7.0 12.0
10.3 36.3 12.0 14.7
9.1 24.1 9.3 14.7
Note. Culture conditions are as described in the legend to Fig. 1.
ferentiation indices for IgM and IgG plasma cells generated in cultures of BMRcells were as high or higher than those of unfractionated B cells or sIgD- B cells. Together these results indicate that LPS augments differentiation of B cells in the subset already responsive to PWM, and not small resting B cells. Pulsing with supernatants from LPS-stimulated B cells or monocytes. We wished to determine if the effect of LPS in these experiments was mediated by a soluble factor produced by monocytes contaminating the B-cell preparations or by the B cells themselves. For these experiments, we isolated B cells and pulsed them with supernatants from LPS-stimulated B cells or AC. Polymyxin B was added during the pulse to block the effects of residual LPS in the supernatants. Results of these experiments, presented in Table 5, show that some supernatants TABLE 3 AUGMENTATIONOFPWM-INDUCEDPLASMACELLGENERATIONBYVARYINGCONCENTRATIONSOF LPSDURINGTHEPULSE Differentiation
index
Expt. 1
LPS pulse (i&ml)
I@
0.01 0.1 1.0 10.0 100.0
3.3 4.0 3.0 8.6 nd
Expt. 2 W 2.9 4.4 3.7 18.5 nd
nd” 10.7 nd 9.3 11.7
nd 7.5 nd 9.3 7.0
Note. B cells were pulsed 12-16 hr with varying concentrations of LPS, washed, and resuspended in fresh medium. Culture conditions were as described in the legend to Fig. 1. a nd, not determined.
266
ANDERSON
AND TABLE
B
LAWTON 4
CELLS LACKING sIgD HAVE AN AUGMENTED RESPONSE TO PWM AFTER PULSING WITH LPS Differentiation Expt. 1 2 3
LPS-treated cells B-enriched sIgD-depleted B-enriched sIgD-depleted B-enriched sIgD-depleted
index
IN
I&
1.2 2.6 5.6 3.0 9.1 43.5
1.3 2.7 3.2 4.9 6.2 3.1
Note. B-enriched cells, or the same population after two cycles of panning on anti-IgD-coated plates, were treated with 10 &ml LPS for 12-16 hr, washed, and cultured as described in the legend to Fig. 1.
contain an activity which enhances B-cell differentiation to the same degree as LPS. This activity is different from LPS, since it is not inhibited by polymyxin B. Role of IL-l in LPS-induced enhancement of B-cell differentiation. IL-1 is known to augment B-cell differentiation and is produced by both B cells and monocytes as a result of LPS stimulation (22, 24-26). When the supernatants used in the previous experiments were assayed for IL-l activity by the thymocyte costimulation assay, all were found to contain IL-1 activity in the range of 3-10 U/ml as compared to IL-1 from Collaborative Research, Inc. (data not shown).
IaM IgM
IgG IaG BMR-BMRB CELL
IgM NM
IgG I@ BMR; BMR+
FRACTION
FIG. 2. Response of BMR- and BMR+ cells to LPS stimulation. B cells fractionated by rosetting with MRBCs were treated as described in the legend to Fig. 1. Results are expressed as number of plasma cells/well. Two other experiments of this type gave similar results.
LPS AUGMENTS
HUMAN
B-CELL
TABLE
267
DIFFERENTIATION
5
SUPERNATANTSFROMLPS-PULSEDB CELLSVARIABLYAFFECTPWM-INDUCEDDIFFERENTIATION Differentiation
index
Pulse
Polymyxin B”
I&f
W
1
LPS LPS B cell SNb
+ +
3.5 1.6 0.2
3.2 1.0 0.1
2
LPS LPS AC SNc
+ +
4.6 1.4 5.0
4.4 1.3 4.7
3
LPS LPS B cell SNb AC SN’
+ + +
2.1 1.0 1.1 0.8
4.8 1.0 1.7 0.8
4
LPS B cell SNd AC SN’
+ +
4.0 5.1 1.5
3.6 2.2 0.1
Expt.
Nore. B cells were pulsed 12-16 hr with undiluted supematant (SNl or 10 &ml LPS in the presence or absence of Polymyxin B, washed, and resuspended in fresh medium. Culture conditions were as described in the legend to Fig. 1. a 20 pglml. b B cells cultured with 10 &ml LPS, 16 hr. c AC cultured with 1 &ml LPS, 60 hr. d B cells cultured with 1 pg/ml LPS, 48 hr. e AC cultured with 1 p,g/ml LPS, 48 hr.
Therefore, we wished to determine if purified IL-l could substitute for LPS augmentation of plasma cell generation. Table 6 shows the results of experiments in which BMR- and BMR+ cells were pulsed with purified human IL-l preparations at concentrations of 3-10 U/ml. These concentrations increased the proliferative response of mouse thymocytes to Con A by five- to seven-fold (data not shown). In no case did purified IL-l cause significant augmentation of plasma cell generation similar to LPS. DISCUSSION
LPS is a potent polyclonal activator of mouse splenic B cells, but has inconsistent effects in cultures of human PBL (15-17). We found that LPS consistently enhanced IgM and IgG production by human PBL stimulated with PWM. In agreement with observations of others (E-17), the enhancement was observed only after rigorous depletion of monocytes. Zabala and Lipsky examined the effect of LPS on the response of human PBL to PWM (17). In an extensive 3-year study involving 59 different donors, LPS augmented PWM-induced PFC responses in the majority, suppressed responses in some, and had no effect in others. In their experiments, PBL were cocultured
268
ANDERSON
AND LAWTON
TABLE PURIFIED
IL-l
FAILS TO AUGMENT
6 THE B-CELL
RESPONSE
Plasma cells/well
TO PWM
X lo-’
BMRExpt.
BMR+
Pulse
IN
Ii&
0 LPS IL- 1”
28.9 131.1 28.7
21.6 77.4 22.2
1.4 0.5 0.6
0.4 0.5 0.6
0 LPS IL-16
132.0 209.8 105.8
81.5 104.9 64.3
1.1 1.1 Cl.2
Cl.1 2.1 <1.2
0 LPS IL-l’
Cl.2 1.3 co.9
1.2 10.5 1.7
1.1 Cl.7
<0.5 <1.7 <2.0
I@
<2.0
I&
Note. BMR- or BMR+ cells were pulsed 12-16 hr, washed, and resuspended in fresh medium. Culture conditions are as described in the legend to Figure 1. DGenzyme Corp. human IL-l, 1% (v/v). b Collaborative Research Inc. human IL-l, 3 U/ml. c Collaborative Research Inc. human IL-l, 4.5 U/ml.
with PWM and a high dose (50 I&ml) of LPS for the entire culture period. Under these conditions, depletion of monocytes or T cells abrogated LPS-stimulated augmentation of PWM responses. The authors attributed the effect of monocytes to the production of IL-l by several circumstantial arguments, the chief one being that IL-I produced by LPS-stimulated monocytes is known to augment PFC responses. However, no experiments assessing the role of IL-1 were done. In contrast, our results suggest that LPS acts directly on B cells. The effects of T cell and monocyte depletion observed by Zabala and Lipsky (17) are not surprising since no response to PWM is seen in the absence of T cells and monocytes. By pulsing each cell population separately with LPS, we were able to more specifically address the role of each cell type in LPS-mediated stimulation. We found that augmentation occurred only when B cells, and not T cells or monocytes, were pulsed with LPS. Furthermore, pulsing B cells for 12 hr with a low dose (10 pg/ml) of LPS gave consistent augmentation of PWM-induced B-cell differentiation in all donors tested, despite varying degrees of T cell and monocyte contamination. While it is well documented that IL-1 enhances B-cell differentiation (17, 22, 26), our results indicate that LPS acts directly on B cells, independent of IL-l production. Supernatants from LPS-stimulated B cells or AC, treated with polymyxin B to remove LPS activity, augmented plasma cell generation in some cases but not others. These supernatants contained IL-l activity as demonstrated by the thymocyte costimulation assay. However, B cells pulsed with purified human IL-1 did not show enhanced differentiation. We conclude that LPS interacts di-
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
269
rectly with B cells to enhance differentiation. While LPS may also induce the production of a soluble factor(s) with the capacity to augment plasma cell generation, this factor is apparently not IL-l. Dagg and Levitt reported that B cells differentiating in response to LPS could be partially separated from those responding to PWM on density gradients (14). PWM-responsive precursors were of lower buoyant density than cells responding to LPS. Other studies indicate that PWM responders lack receptors for mouse erythrocytes (12) and lack surface IgD (13). These observations suggested that LPS might act on small resting B cells, rendering them responsive to PWM and T-cell-derived differentiation signals. Our results do not support this hypothesis. When B cells were depleted of surface IgD-bearing cells no effect n LPS-stimulated augmentation was observed. B cells rosetting with mouse erythrocytes responded minimally to PWM, confirming earlier results. BMR+ cells pulsed with LPS still did not respond to PWM, while differentiation of BMR- cells was greatly augmented by LPS treatment. Therefore it appears that those B cells whose differentiation is enhanced by LPS belong to the population which contains PWM-responsive precursors. The failure of LPS to render PWM-unresponsive B cells receptive to differentiation signals provided by PWM and T cells is reminiscent of results obtained using Cowan I Staphylococcus aureus (SAC) as a costimulator in the response to PWM (20). SAC stimulates T-cell-independent proliferation and T-cell-dependent differentiation of B cells. BMR+ cells proliferate in response to SAC but do not differentiate to Ig secretion in the presence of PWM and T cells. BMR- cells proliferate little, but contain the precursors for plasma cells. These observations highlight some problems with current models for B-cell activation and differentiation. It is assumed that small resting B cells, activated by appropriate signals, enlarge, become responsive to growth factors, enter the cell cycle, and eventually become receptive to T-cell-derived differentiation factors. While small resting B cells may be activated by anti-p or SAC such that they proliferate when exposed to B-cell growth factor(s) (27), and preactivated large B cells differentiate in response to B-cell growth and differentiation factors (28), there is little or no direct evidence that small resting B-cell populations from human peripheral blood (or tonsil) can undergo this entire sequence in vitro. Falkoff et al. reported that small resting B cells could differentiate when treated with SAC and exposed to T-cell-derived differentiation factors (29). Likewise, Jelinek et al. observed that sIgD+ B cells produced IgM when stimulated with SAC and supernatant from activated T cells (30). However, in both these cases, the magnitude of the response was negligible. In the former study, the frequency of ISC generated in cultures of small B cells was approx l/100; in the latter, the response of either sIgD+ or sIgD- cells to SAC was lo-fold less than the response of sIgD- cells to PWM + T cells. Surface IgD+ cells did not respond to PWM even in the presence of T cells. In a report by Layton et al. small resting B cells from mouse spleen could be induced to differentiate by LPS, but no combination of anti-Ig and soluble T-cell factors could induce differentiation in this subpopulation (31).
270
ANDERSON
AND LAWTON
The second problem highlighted by these results is the inadequacy of current in vitro B-cell differentiation assays in humans. Assays using PWM do not address the function of a major population of sIgM+ sIgD+ B lymphocytes. Other polyclonal activators such as LPS and SAC appear likewise unable to stimulate these cells. Epstein-Barr virus (EBV) is also a polyclonal inducer of Ig secretion by human B cells and has been thought to infect primarily high-density (small) B cells (32). However, in a recent study utilizing a limiting dilution technique with high cloning efficiency, EBV-responsive precursors were found to heterogeneous (33). IgM precursors were large sIgM+ sIgD+ cells, while IgG and IgA precursors were small, sIgD- , and expressed sIgG or sIgA, respectively. The technique used in this study has considerable potential for the study of B cells from immunodeficient patients, as the cloning efficiency was 3-30%. Nevertheless, the need for an assay which can address the majority population of recirculating B cells is clear. ACKNOWLEDGMENT The authors thank Ms. Melanie Rardin for her expert assistance in preparing this manuscript.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
Wu, L. Y. F., Lawton, A. R., and Cooper. M. D., J. C/in. Invest. 52, 3180, 1973. Janossy, G., and Greaves, M., Transplant. Rev. 24, 177, 1975. Keightley, R. G., Cooper, M. D., and Lawton. A. R., J. Zmmunol. 117, 1538, 1976. Fauci, A. S., Pratt, K. R., and Whelan, G., J. Immunol. 117, 2100, 1976. Siegal, F. I?, and Siegal, M., J. Immunol. 118, 642, 1977. Hayward, A. R., and Lawton, A. R., J. Immunol. 119, 1213, 1977. Broom, B. C., DeLaConcha, E. G.. Webster, A. D. B.. Janossy, G. J., and Asherson, G. L., Clin. Exp. Immunol. 23, 73, 1976. Moller, G. (Ed.), Zmmunol. Rev. 57, 1981. Pereira, R. S., Wilkins, S. R., Webster, A. D., Asherson, G.L., and Platts-Mills, T. A. E., Birth Defects, Orig. Artic. Ser. 19, 165, 1983. Yoshizaki, K.. Nakagawa. T.. Kaieda, T., Muraguchi. A., Yamamura, Y., and Kishimoto, T., J. Immunol. 128, 1296, 1982. Stevens, R. H., Macy, E., and Thiele, C. J., Stand. J. Immunol. 14, 449, 1981. Lucivero, G., Lawton, A. R., and Cooper, M. D., Clin. Exp. Immunol. 45, 18.5, 1981. Kuritani, T., and Cooper, M. D., J. Exp. Med. 155, 1561, 1982. Dagg, M. K., and Levitt, D., Clin. Zmmunol. Immunopathol. 21, 39, 1981. Levitt, D., Duber-Stull, D., and Lawton, A. R., Clin. Zmmunol. Zmmunopathol. 16, 309, 1981. Sager, D. S., and Jasin, H. E., Clin. Exp. Zmmunol. 47, 645, 1981. Zabala, C., and Lipsky, P. E., J. Zmmunol. 129, 2496, 1982. Saxon, A., Feldhaus, J., and Robbins, R. A., J. Immunol. Methods 12, 285, 1976. Kuritani, T., and Cooper, M. D., J. Exp. Med. 155, 389, 1982. Ito, S., and Lawton, A. R., J. Zmmunol. 133, 1891, 1984. Rosenberg, S. A., and Lipsky, P E., J. Zmmunol. 122, 926, 1979. Matsushima, K., Procopio, A., Abe, H., Scala, G., Ortaldo, J. R., and Oppenheim, J. J., J. fmmunol. 135, 1132, 1985. Elmer, J. J., and Spagnuolo, P. J., J. Immunol. 123, 2689, 1979. Wood, D. D.. J. Immunol. 123, 2400, 1979. Wood, D. D., and Cameron, P. M., J. Zmmunol. 121, 53. 1978. Rosenberg, S. A., and Lipsky, P. E.. J. Zmmunol. 126, 1341, 1981. Kehrl, J. H., Muraguchi, A., Butler, J. L., Falkoff. R. J. M., and Fauci. A. S., Immunol. Rev. 78, 75, 1984.
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
271
28. Kishimoto, T., Yoshizaki, K., Kimoto, M., Okada, M., Kuritani, T., Kikutani, H., Shimizu, K., Nakagawa, T., Nakagawa, N., M&i, Y., Kishi, H., Fukunaga, K., Yoshikubo, T., and Taga, T., Immunol. Rev. 18, 97, 1984. 29. Falkoff, R. J. M., Butler, J. L., Dinarello, C. A., and Fauci, A.. S., J. Immunol. 133, 692, 1984. 30. Jelinek, D. F., Splawski, J. B., and Lipsky, P. E., J. Immunol. 136, 83, 1986. 31. Layton, J. E., Krammer, P. H., Hamaoka, T., Uhr, J. W., and Vitetta, E. S., J. MO/. Cell. Immunol. 2, 155, 1985. 32. Aman, I?. Ehlin-Henriksson, B., and Klein, G., J. Exp. Med. 159, 208, 1984. 33. Chan, M. A., Stein, L. D., and Dosch, H. M., J. Immunol. 136, 106, 1986. Received September 5, 1986; accepted with revision April 13, 1987.
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
44, 259-271 (1987)
LPS Augments Human B-Cell Differentiation by Direct Stimulation of PWM-Responsive B Cells1 STEPHEN J. ANDERSON~ANDALEXANDER R. LAWTON~ Departments
of Microbiology
and Pediatrics, Vanderbilt University Nashville, Tennessee 37232
School
of Medicine,
Bacterial lipopolysaccharide (LPS) augments production of IgM and IgG by two- to seven-fold in cultures of peripheral blood lymphocytes (PBL) stimulated by pokeweed mitogen (PWM), but only if monocytes are rigorously depleted. When PBL were separated into adherent cell (AC), B-cell-enriched, and T-cell-enriched fractions, pulsed with LPS, and recombined in culture with PWM, increased generation of plasma cells was seen only in cultures containing LPS-treated B cells. This effect of LPS appears to be independent of soluble factors. Supematants from LPS-stimulated B cells or AC did not consistently increase PWM responses when cultured with fresh B cells in the presence of polymyxin B. Furthermore, pulsing of B cells with purified interleukin 1 from two different commercial sources failed to augment PWM-induced differentiation. When B cells were depleted of surface IgD (sIgD)-bearing cells by panning, no effect on LPSmediated augmentation of PWM-driven differentiation was seen. B cells were also fractionated by rosetting with mouse erythrocytes. Treatment of BMR+ cells with LPS did not induce them to respond to PWM, while treatment of BMR- cells with LPS augmented generation of plasma cells. These results indicate that LPS acts directly to augment differentiation of PWM-responsive B cells, rather than recruiting sIgD+, BMR+ cells to become PWM responsive. 0 1987 Academic PRSS. IX.
INTRODUCTION
Polyclonal B-cell activators have been used for in vitro study of B-cell differentiation in normal individuals and patients with a variety of diseases for more than a decade. The great majority of studies have used pokeweed mitogen (PWM)4 as the stimulatory agent (1, 2). The induction of B-cell differentiation by PWM is regulated by interacting sets of helper and suppressor T cells, mimicking B-cell respnses to T-dependent antigens (2-6). Because the PWM response is not restricted by histocompatibility differences 1 Supported by NIH Grant AI-17996. 2 Present address: Department of Immunology and Medical Microbiology, University of Florida, Gainesville, FL. 3 To whom all correspondence should be addressed. 4 Abbreviations used: LPS, lipopolysaccharide; PWM, pokeweed mitogen; PBL, peripheral blood lymphocytes; MNC, mononuclear cells; NAC, nonadherent cells; AC, adherent cells; sIg and cIg, surface and cytoplasmic immunoglobulin; BMR+, B cells forming rosettes with mouse erythrocytes; BMR-, B cells not forming rosettes with mouse erythrocytes; DI, differentiation index; ANAE, (Ynaphthyl acetate esterase; IL-l, interleukin 1; PBS, phosphate-buffered saline; FCS; fetal calf serum; PHA, phytohemagglutinin; Con A, concanavalin A; PFC, plaque-forming cell; EBV, Epstein-Barr virus; HBSS, Hanks’ balanced salt solution; ELISA, enzyme-linked immunosorbent assay; PC, plasma cells; MRBC, mouse red blood cells; ISC, immunoglobulin-secreting cells. 259 0090-1229187 $1.50 Copyright Q 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.
260
ANDERSON
AND
LAWTON
between B and T cells, cocultures of patient and normal cells may be used to dissect the cellular basis of differentiation defects. In such assays it has been demonstrated that B cells from the majority of the patients with common varied hypogammaglobulinemia (7), B cells from blood of normal newborns (6, 8), or bone marrow of adults (9), and monoclonal B cells from patients with chronic lymphocytic leukemia (10) have a very limited differentiation capacity with respect to circulating B cells from normal adults. It is now recognized that only a small fraction of the recirculating B-cell pool is capable of differentiating in response to PWM (11). These cells lack receptors for mouse erythrocytes (12), have little or no surface IgD (sIgD) (13), and are relatively large with low buoyant density (13, 14). Viewed from this perspective the “intrinsic defects of B-cell differentiation” observed in hypogammaglobulinemic patients and in the other settings mentioned above may simply reflect the absence of a minor population of preactivated PWM-responsive B cells among circulating lymphocytes. To better define the nature of these B-cell defects a method of inducing polyclonal differentiation of the majority of B cells, particularly those which do not respond to PWM, is needed. One approach is to use combinations of polyclonal activators which may act on different subpopulations of B cells. Bacterial lipopolysaccharide (LPS) is capable of stimulating human-B-cell differentiation directly (14- 16) and of augmenting the T-cell-dependent response to PWM (17). Observations of Dagg and Levitt suggested that LPS and PWM act on different subpopulations of B cells, distinguishable by buoyant density (14). We have investigated the ability of LPS to activate B cells which are otherwise unresponsive to PWM. Our results indicate that LPS augments PWM-induced differentiation by direct interaction with B cells, but that it does not stimulate a significant number of cells which bear sIgD or receptors for mouse erythrocytes to respond to PWM in the presence of T cells. MATERIALS
AND METHODS
Cell separation. Peripheral blood mononuclear cells (MNC) were isolated from blood of healthy adult donors by centrifugation on Ficoll-Hypaque gradients. Nonadherent cells (NAC) and adherent cells (AC) were separated by consecutive incubations on glass and plastic petri dishes for 1 hr at 37°C. AC were recovered by scraping the glass plates with a rubber policeman. Monocytes were identified by staining for a-naphthyl acetate esterase (ANAE) activity (Sigma Histozyme Kit 90-Al). NAC were separated into B-enriched (B cells) and T-enriched (T cells) fractions by one round of rosetting with 2-aminoethylisothiuronium bromidetreated sheep erythrocytes and centrifugation on Ficoll-Hypaque gradients (18). B-enriched cells contained 9.8 ? 14% ANAE+ (n = 13) cells and 12.2 -+ 9% OKT3+ cells (n = 9), respectively, after a single round of rosetting. Adherent cells contained 67 2 16% ANAE+ cells (n = 8). Depletion of slgD-bearing ceils. The B-cell fraction was depleted of sIgDbearing cells by panning on plastic petri dishes (Fisher 8-757-12) precoated with anti-human IgD (6 chain-specific, Southern Biotechnology Associates, Bir-
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
261
mingham, AL) (19). The plates were washed three times with PBS before use. B cells in PBS + 5% FCS at ~7 x 10‘Yml were incubated for 70 min at 4°C in 3 ml cell suspension per 100 x 15mm plate. Nonadherent cells (sIgD-) were decanted and the plates were rinsed twice with cold PBS + 1% FCS. This process was repeated once to adequately deplete sIgD-bearing cells. Separation of BMR- and BMR+ cells. B cells were fractionated according to their ability to form rosettes with neuraminidase-treated mouse erythrocytes (MRBC) as previously described (20). B cells and MRBC were mixed, and rosetting (BMR+) and nonrosetting (BMR-) cells were separated by centrifugation on Ficoll-Hypaque gradients. Rosettes were disrupted by hypotonic lysis of MRBC. Cells were washed twice in HBSS + 2% FCS before culture. Culture conditions. Cells were cultured in RPM1 1640 medium supplemented with 2.0 mM L-glutamine, 5 x low5 M 2-mercaptoethanol, 50 kg/ml gentamicin, and 10% FCS (complete medium). Triplicate cultures were established in 96-well microtiter plates with flat-bottom wells (Costar 3596). For the first set of experiments each well contained lo5 MNC or lo5 NAC in 0.2 ml complete medium. AC (5 x 103) were added to NAC cultures to ensure the presence of enough monocytes to support a response to PWM (21). For all subsequent experiments each well contained 5 x lo4 B cells, 5 x lo4 T cells, and 5 x lo3 AC in 0.2 ml complete medium. PWM was used at a 1:200 dilution of stock (Gibco lot No. lON9424, 13Kl331). Cultures were incubated at 37°C for 7 days in a humidified atmosphere with 5% CO, before assay. Pulsing of cells with LPS. In order to define the role of B cells, T cells, and monocytes in the response to LPS, cell populations were separated and pretreated with LPS before recombination and culturing with PWM as described above. To accomplish this “pulsing,” cells were resuspended at 106/ml in complete medium this LPS (Escherichia coli Olll:B4, Sigma Chemicals) and incubated in 24-well plates at 37°C. Control cells were incubated as described in complete medium without LPS. Pulsed cells were harvested and washed once before further use. Production of interleukin 1 (IL-I)-containing supernatants. To stimulate production of IL-l, B cells and AC were cultured in 24-well plates at 106/ml in complete medium with 1.0 l.&rnl LPS (22). After 48 hr, culture supernatants were harvested, centrifuged at 1500g for 15 min, and filtered (Millex GV 0.22 pm syringe filters, Millipore Corp.). Supernatants were stored at - 20°C. Purified human IL-l preparations were obtained from Genzyme Corp. and Collaborative Research, Inc. Supernatants were assayed for IL-1 activity using a thymocyte costimulation assay (22). Thymocytes from C57B1/6 or C3H/HeJ mice were cultured in complete medium with 5% FCS at 1.5 x lo6 cells/O.2 ml/well in microtiter plates with flat-bottom wells (Costar 3596). Cells were stimulated with 1.0 kg/ml PHA (PHA-P, Difco) or 0.25 pg/ml Con A (Calbiochem) and serial dilutions of test supernatants for 72 hr. Cultures were pulsed with 0.5 &i/well [3H]thymidine (New England Nuclear, Boston, MA) for the last 16 hr, harvested onto glass fibers filters, and counted in a Beckman liquid scintillation counter. For pulsing with IL-l-containing supernatants, B cells were resuspended in undiluted super-
262
ANDERSON
AND
LAWTON
natant at 106/ml and cultured in 24-well plates as described above. In some experiments, polymyxin B (Aerosporin. Burroughs Wellcome Co., Research Triangle Park, NC) was added at a final concentration of 20 ug/ml. Assay for differentiation. Differentiation was determined by measurement of secreted IgM and IgG by ELISA as described (20) or by enumeration of cells containing cytoplasmic immunoglobulin (cIg+) by immunofluorescence (3). For the latter, slides were examined with a Leitz Orthoplan microscope with epi-illumination. At least 500 cells or 20 cIg+ cells were counted per slide. The total number of plasma cells (PC) per well was determined using the formula -=PC well
fr. cIg+ cells x No. cells recovered from pooled wells No. wells
used several donors whose response to PWM varied considerably in magnitude. In order to facilitate comparison of such experiments, the results are expressed as differentiation indices. The differentiation index (DI) is defined by Differentiation
DI = Statistics.
index. Our experiments
response of cultures with LPS-pulsed cells to PWM response of cultures with no LPS-pulsed cells to PWM’
Data were analyzed using the Student t test. RESULTS
Initial experiments were performed to determine the effect of LPS on PWM-induced Ig production by human peripheral blood mononuclear cells. In our hands, LPS alone, in concentrations up to 100 ug/ml, did not stimulate significant levels of Ig production. In agreement with others, we found that coculture with LPS and PWM had an inconsistent effect on the responses of unfractionated MNC. This may be due to stimulation by LPS of large numbers of monocytes present in populations of human peripheral blood MNC (15, 16, 23). Effect of LPS on monocyte-depleted peripheral blood lymphocytes. Data in Table 1 are from experiments in which monocytes were rigorously depleted from human MNC by adherence to glass or plastic petri dishes. Unfractionated MNC contained 25-35% monocytes by a-naphthyl acetate esterase staining. After depletion, the nonadherent cells were ~5% ANAE+ . MNC or NAC were incubated with or without LPS, washed, and cultured with PWM. Secretion of both IgM and IgG were consistently enhanced in experiments where monocyte-depleted cells (NAC) were pulsed with LPS. Differentiation indices ranged from 1.8 to 6.1 for IgM production and 1.1 to 15.2 for IgG production. While MNC showed enhanced responses to PWM in some experiments when pulsed with LPS, the effect of LPS was inconsistent. Effect of LPS on separated B cells, T cells, and adherent cells. To determine with which cell population LPS interacts to stimulate B-cell differentiation, MNC were separated into NAC and AC, and NAC were further separated into B-cellenriched (B cells) and T-cell-enriched (T cells) fractions. Results of eight experiments in which B cells, T cells, and AC were separately incubated overnight with
LPS AUGMENTS
HUMAN
B-CELL
263
DIFFERENTIATION
TABLE 1 EFFECT OF LPS PULSE ON PWM-INDUCED IMMUNOGLOBUL~N PRODUCTION Immunoglobulin MNC Donor I 2 3 4 5 6
kM 298 314 370 980 580 11.5
production (t&ml)
MNC=Ps” I&
IN
516
270 448 325 340 786 2964
360 570 592 330 536
W 165 940 150 65 464 3972
NACLPS”
NAC I&f
W
298 ndb 660 400 582 1664
728 1364 374 140 210 4428
IgM
I&s
531 970 1670 1,612 1,970 10,316
1996 1464 4064 2130 886 4804
Note. MNC or NAC (103 were cultured with PWM in flat-bottom microtiter wells for 8 days. AC (5 x 10)) were added to cultures of NAC. Supernatants from triplicate cultures were pooled and assayed for secreted Ig by ELISA. 0 Cells were pulsed with 10 pg/ml LPS for 12 hr before culturing. b nd, none detectable.
or without LPS, washed, and recombined in microculture with PWM are summarized in Fig. 1. In all experiments, using several different donors, enhancement of PWM-stimulated plasma cell generation was seen only when the B-cell fractions were pulsed with LPS. LPS treatment of B cells increased the number of IgM plasma cells by an average of 4.7-fold (range, 2-9) and IgG plasma cells by an average of 3.3-fold (range, 2-6) compared to control cultures. There was no significant increase in plasma cells of either isotype in cultures reconstituted with LPS-treated T cells or AC. Mean DI values for LPS-treated B cells differed from those of LPS-treated T cells or AC at the 95% confidence level. Data from representative experiments presented in Table 2 show that increases in absolute numbers of plasma cells reflect increases in the percentage of cIg+ cells per culture and not merely recovery of greater numbers of cells. This observation holds true for all subsequent experiments as well. Optimal conditions for LPS pulsing. Optimal conditions for exposure of B cells to LPS were determined. In three of six experiments, pulsing of B cells for as brief a period as 1 hr caused enhancement of PWM-induced plasma cell generation of 15fold or greater. However, in all experiments, pulses of 12 hr or more gave the greatest degree of stimulation (data not shown). Experiments using LPS over a five-log concentration range revealed a threshold rather than a linear relationship between stimulation and concentration (Table 3). Stimulation was seen when B cells were pulsed with as little as 0.1 &ml LPS for 12-16 hr. and 10 kg/ml gave a consistently high response. Effect of LPS on small resting B cells. We wished to determine if the enhancement of PWM-induced plasma cell generation by LPS was due to recruitment of otherwise unresponsive B cells which bear receptors for mouse erythrocytes and surface IgD. In these experiments, B-cell subpopulations were isolated, separately pulsed with LPS, washed, and reconstituted with T cells and AC in microculture with PWM.
264
ANDERSON
AND LAWTON
Depletion of slgD-bearing cells. Table 4 summarizes data from experiments in which sIgD- B cells were prepared by two rounds of panning on anti-s-coated petri dishes. Unfractionated B cells contained approx 27% sIgD+ cells by direct immunofluorescence. The sIgD-depleted populations contained ~0.4% sIgD+ cells. In each of these experiments, removal of sIgD-bearing cells did not diminish the effect of pulsing with LPS. We were not able to recover sufficient numbers of sIgD+ cells from the panning plates to carry out the reciprocal experiments. Response of BMR- and BMR+ cells. Results of a representative experiment in which B cells were separated according to their ability to form rosettes with MRBC are shown in Fig. 2. The unfractionated B-cell populations contained approx 20% BMR+ cells, similar to the number of sIgD+ cells. After separation, the yield of BMR- cells and BMR+ cells was approximately 10: 1. In agreement with previous studies, BMR- cells gave a significant plasma cell response to PWM, while BMR+ cells failed to respond. Exposure of BMR- cells to an overnight pulse with LPS markedly enhanced generation of plasma cells of both IgM and IgG classes. Treatment of BMR+ cells with LPS caused a slight increase in plasma cell generation, but the level of their response, in absolute number of plasma cells, remained l-2 orders of magnitude below that of BMR- cells. Dif-
IgM Plasma
Cells
2
T CELLS
B CELLS LPS
PULSED
AC CELLS
FIG 1. Augmentation of B-cell differentiation by LPS is mediated by interaction with B cells. Bcell, T-cell, and adherent cell populations were cultured for 12-16 hr in complete medium with or without 10 &ml LPS. The cells were then washed and recombined in culture such that only one or none of the three populations had been exposed to LPS and cultured an additional 8 days with PWM. Each microwell contained 5 x 104 B, 5 x 104 T, and 5 x lo3 adherent cells. Cells from triplicate cultures were pooled and assayed for expression of cytoplasmic immunoglobulin. The results shown are the means + SD of differentiation indices, calculated as described under Materials and Methods, for eight separate experiments. Significant augmentation of the IgM and IgG response occurred only in culture containing LPS-pulsed B cells (P < 0.05).
LPS AUGMENTS
HUMAN
B-CELL
265
DIFFERENTIATION
TABLE 2 LPS AUGMENTATION OF PWM-INDUCED PLASMA CELL GENERATION % cIg+ cells LPS pulse
I@
PC/well x lO-3 W
I&f
I&
Experiment
1
None B cells T cells AC
1.4 5.4 1.2 1.0
1.3 4.3 0.4 0.4
1.3 23.4 5.8 3.5
6.8 18.7 1.9 1.4
Experiment
2
None B cells T cells AC
4.0 22.0 4.0 3.0
4.0 11.0 3.0 2.0
6.2 34.5 6.7 4.1
6.2 17.3 5.0 2.7
Experiment
3
None B cells T cells AC
9.0 25.0 9.0 12.0
8.0 17.0 7.0 12.0
10.3 36.3 12.0 14.7
9.1 24.1 9.3 14.7
Note. Culture conditions are as described in the legend to Fig. 1.
ferentiation indices for IgM and IgG plasma cells generated in cultures of BMRcells were as high or higher than those of unfractionated B cells or sIgD- B cells. Together these results indicate that LPS augments differentiation of B cells in the subset already responsive to PWM, and not small resting B cells. Pulsing with supernatants from LPS-stimulated B cells or monocytes. We wished to determine if the effect of LPS in these experiments was mediated by a soluble factor produced by monocytes contaminating the B-cell preparations or by the B cells themselves. For these experiments, we isolated B cells and pulsed them with supernatants from LPS-stimulated B cells or AC. Polymyxin B was added during the pulse to block the effects of residual LPS in the supernatants. Results of these experiments, presented in Table 5, show that some supernatants TABLE 3 AUGMENTATIONOFPWM-INDUCEDPLASMACELLGENERATIONBYVARYINGCONCENTRATIONSOF LPSDURINGTHEPULSE Differentiation
index
Expt. 1
LPS pulse (i&ml)
I@
0.01 0.1 1.0 10.0 100.0
3.3 4.0 3.0 8.6 nd
Expt. 2 W 2.9 4.4 3.7 18.5 nd
nd” 10.7 nd 9.3 11.7
nd 7.5 nd 9.3 7.0
Note. B cells were pulsed 12-16 hr with varying concentrations of LPS, washed, and resuspended in fresh medium. Culture conditions were as described in the legend to Fig. 1. a nd, not determined.
266
ANDERSON
AND TABLE
B
LAWTON 4
CELLS LACKING sIgD HAVE AN AUGMENTED RESPONSE TO PWM AFTER PULSING WITH LPS Differentiation Expt. 1 2 3
LPS-treated cells B-enriched sIgD-depleted B-enriched sIgD-depleted B-enriched sIgD-depleted
index
IN
I&
1.2 2.6 5.6 3.0 9.1 43.5
1.3 2.7 3.2 4.9 6.2 3.1
Note. B-enriched cells, or the same population after two cycles of panning on anti-IgD-coated plates, were treated with 10 &ml LPS for 12-16 hr, washed, and cultured as described in the legend to Fig. 1.
contain an activity which enhances B-cell differentiation to the same degree as LPS. This activity is different from LPS, since it is not inhibited by polymyxin B. Role of IL-l in LPS-induced enhancement of B-cell differentiation. IL-1 is known to augment B-cell differentiation and is produced by both B cells and monocytes as a result of LPS stimulation (22, 24-26). When the supernatants used in the previous experiments were assayed for IL-l activity by the thymocyte costimulation assay, all were found to contain IL-1 activity in the range of 3-10 U/ml as compared to IL-1 from Collaborative Research, Inc. (data not shown).
IaM IgM
IgG IaG BMR-BMRB CELL
IgM NM
IgG I@ BMR; BMR+
FRACTION
FIG. 2. Response of BMR- and BMR+ cells to LPS stimulation. B cells fractionated by rosetting with MRBCs were treated as described in the legend to Fig. 1. Results are expressed as number of plasma cells/well. Two other experiments of this type gave similar results.
LPS AUGMENTS
HUMAN
B-CELL
TABLE
267
DIFFERENTIATION
5
SUPERNATANTSFROMLPS-PULSEDB CELLSVARIABLYAFFECTPWM-INDUCEDDIFFERENTIATION Differentiation
index
Pulse
Polymyxin B”
I&f
W
1
LPS LPS B cell SNb
+ +
3.5 1.6 0.2
3.2 1.0 0.1
2
LPS LPS AC SNc
+ +
4.6 1.4 5.0
4.4 1.3 4.7
3
LPS LPS B cell SNb AC SN’
+ + +
2.1 1.0 1.1 0.8
4.8 1.0 1.7 0.8
4
LPS B cell SNd AC SN’
+ +
4.0 5.1 1.5
3.6 2.2 0.1
Expt.
Nore. B cells were pulsed 12-16 hr with undiluted supematant (SNl or 10 &ml LPS in the presence or absence of Polymyxin B, washed, and resuspended in fresh medium. Culture conditions were as described in the legend to Fig. 1. a 20 pglml. b B cells cultured with 10 &ml LPS, 16 hr. c AC cultured with 1 &ml LPS, 60 hr. d B cells cultured with 1 pg/ml LPS, 48 hr. e AC cultured with 1 p,g/ml LPS, 48 hr.
Therefore, we wished to determine if purified IL-l could substitute for LPS augmentation of plasma cell generation. Table 6 shows the results of experiments in which BMR- and BMR+ cells were pulsed with purified human IL-l preparations at concentrations of 3-10 U/ml. These concentrations increased the proliferative response of mouse thymocytes to Con A by five- to seven-fold (data not shown). In no case did purified IL-l cause significant augmentation of plasma cell generation similar to LPS. DISCUSSION
LPS is a potent polyclonal activator of mouse splenic B cells, but has inconsistent effects in cultures of human PBL (15-17). We found that LPS consistently enhanced IgM and IgG production by human PBL stimulated with PWM. In agreement with observations of others (E-17), the enhancement was observed only after rigorous depletion of monocytes. Zabala and Lipsky examined the effect of LPS on the response of human PBL to PWM (17). In an extensive 3-year study involving 59 different donors, LPS augmented PWM-induced PFC responses in the majority, suppressed responses in some, and had no effect in others. In their experiments, PBL were cocultured
268
ANDERSON
AND LAWTON
TABLE PURIFIED
IL-l
FAILS TO AUGMENT
6 THE B-CELL
RESPONSE
Plasma cells/well
TO PWM
X lo-’
BMRExpt.
BMR+
Pulse
IN
Ii&
0 LPS IL- 1”
28.9 131.1 28.7
21.6 77.4 22.2
1.4 0.5 0.6
0.4 0.5 0.6
0 LPS IL-16
132.0 209.8 105.8
81.5 104.9 64.3
1.1 1.1 Cl.2
Cl.1 2.1 <1.2
0 LPS IL-l’
Cl.2 1.3 co.9
1.2 10.5 1.7
1.1 Cl.7
<0.5 <1.7 <2.0
I@
<2.0
I&
Note. BMR- or BMR+ cells were pulsed 12-16 hr, washed, and resuspended in fresh medium. Culture conditions are as described in the legend to Figure 1. DGenzyme Corp. human IL-l, 1% (v/v). b Collaborative Research Inc. human IL-l, 3 U/ml. c Collaborative Research Inc. human IL-l, 4.5 U/ml.
with PWM and a high dose (50 I&ml) of LPS for the entire culture period. Under these conditions, depletion of monocytes or T cells abrogated LPS-stimulated augmentation of PWM responses. The authors attributed the effect of monocytes to the production of IL-l by several circumstantial arguments, the chief one being that IL-I produced by LPS-stimulated monocytes is known to augment PFC responses. However, no experiments assessing the role of IL-1 were done. In contrast, our results suggest that LPS acts directly on B cells. The effects of T cell and monocyte depletion observed by Zabala and Lipsky (17) are not surprising since no response to PWM is seen in the absence of T cells and monocytes. By pulsing each cell population separately with LPS, we were able to more specifically address the role of each cell type in LPS-mediated stimulation. We found that augmentation occurred only when B cells, and not T cells or monocytes, were pulsed with LPS. Furthermore, pulsing B cells for 12 hr with a low dose (10 pg/ml) of LPS gave consistent augmentation of PWM-induced B-cell differentiation in all donors tested, despite varying degrees of T cell and monocyte contamination. While it is well documented that IL-1 enhances B-cell differentiation (17, 22, 26), our results indicate that LPS acts directly on B cells, independent of IL-l production. Supernatants from LPS-stimulated B cells or AC, treated with polymyxin B to remove LPS activity, augmented plasma cell generation in some cases but not others. These supernatants contained IL-l activity as demonstrated by the thymocyte costimulation assay. However, B cells pulsed with purified human IL-1 did not show enhanced differentiation. We conclude that LPS interacts di-
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
269
rectly with B cells to enhance differentiation. While LPS may also induce the production of a soluble factor(s) with the capacity to augment plasma cell generation, this factor is apparently not IL-l. Dagg and Levitt reported that B cells differentiating in response to LPS could be partially separated from those responding to PWM on density gradients (14). PWM-responsive precursors were of lower buoyant density than cells responding to LPS. Other studies indicate that PWM responders lack receptors for mouse erythrocytes (12) and lack surface IgD (13). These observations suggested that LPS might act on small resting B cells, rendering them responsive to PWM and T-cell-derived differentiation signals. Our results do not support this hypothesis. When B cells were depleted of surface IgD-bearing cells no effect n LPS-stimulated augmentation was observed. B cells rosetting with mouse erythrocytes responded minimally to PWM, confirming earlier results. BMR+ cells pulsed with LPS still did not respond to PWM, while differentiation of BMR- cells was greatly augmented by LPS treatment. Therefore it appears that those B cells whose differentiation is enhanced by LPS belong to the population which contains PWM-responsive precursors. The failure of LPS to render PWM-unresponsive B cells receptive to differentiation signals provided by PWM and T cells is reminiscent of results obtained using Cowan I Staphylococcus aureus (SAC) as a costimulator in the response to PWM (20). SAC stimulates T-cell-independent proliferation and T-cell-dependent differentiation of B cells. BMR+ cells proliferate in response to SAC but do not differentiate to Ig secretion in the presence of PWM and T cells. BMR- cells proliferate little, but contain the precursors for plasma cells. These observations highlight some problems with current models for B-cell activation and differentiation. It is assumed that small resting B cells, activated by appropriate signals, enlarge, become responsive to growth factors, enter the cell cycle, and eventually become receptive to T-cell-derived differentiation factors. While small resting B cells may be activated by anti-p or SAC such that they proliferate when exposed to B-cell growth factor(s) (27), and preactivated large B cells differentiate in response to B-cell growth and differentiation factors (28), there is little or no direct evidence that small resting B-cell populations from human peripheral blood (or tonsil) can undergo this entire sequence in vitro. Falkoff et al. reported that small resting B cells could differentiate when treated with SAC and exposed to T-cell-derived differentiation factors (29). Likewise, Jelinek et al. observed that sIgD+ B cells produced IgM when stimulated with SAC and supernatant from activated T cells (30). However, in both these cases, the magnitude of the response was negligible. In the former study, the frequency of ISC generated in cultures of small B cells was approx l/100; in the latter, the response of either sIgD+ or sIgD- cells to SAC was lo-fold less than the response of sIgD- cells to PWM + T cells. Surface IgD+ cells did not respond to PWM even in the presence of T cells. In a report by Layton et al. small resting B cells from mouse spleen could be induced to differentiate by LPS, but no combination of anti-Ig and soluble T-cell factors could induce differentiation in this subpopulation (31).
270
ANDERSON
AND LAWTON
The second problem highlighted by these results is the inadequacy of current in vitro B-cell differentiation assays in humans. Assays using PWM do not address the function of a major population of sIgM+ sIgD+ B lymphocytes. Other polyclonal activators such as LPS and SAC appear likewise unable to stimulate these cells. Epstein-Barr virus (EBV) is also a polyclonal inducer of Ig secretion by human B cells and has been thought to infect primarily high-density (small) B cells (32). However, in a recent study utilizing a limiting dilution technique with high cloning efficiency, EBV-responsive precursors were found to heterogeneous (33). IgM precursors were large sIgM+ sIgD+ cells, while IgG and IgA precursors were small, sIgD- , and expressed sIgG or sIgA, respectively. The technique used in this study has considerable potential for the study of B cells from immunodeficient patients, as the cloning efficiency was 3-30%. Nevertheless, the need for an assay which can address the majority population of recirculating B cells is clear. ACKNOWLEDGMENT The authors thank Ms. Melanie Rardin for her expert assistance in preparing this manuscript.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
Wu, L. Y. F., Lawton, A. R., and Cooper. M. D., J. C/in. Invest. 52, 3180, 1973. Janossy, G., and Greaves, M., Transplant. Rev. 24, 177, 1975. Keightley, R. G., Cooper, M. D., and Lawton. A. R., J. Zmmunol. 117, 1538, 1976. Fauci, A. S., Pratt, K. R., and Whelan, G., J. Immunol. 117, 2100, 1976. Siegal, F. I?, and Siegal, M., J. Immunol. 118, 642, 1977. Hayward, A. R., and Lawton, A. R., J. Immunol. 119, 1213, 1977. Broom, B. C., DeLaConcha, E. G.. Webster, A. D. B.. Janossy, G. J., and Asherson, G. L., Clin. Exp. Immunol. 23, 73, 1976. Moller, G. (Ed.), Zmmunol. Rev. 57, 1981. Pereira, R. S., Wilkins, S. R., Webster, A. D., Asherson, G.L., and Platts-Mills, T. A. E., Birth Defects, Orig. Artic. Ser. 19, 165, 1983. Yoshizaki, K.. Nakagawa. T.. Kaieda, T., Muraguchi. A., Yamamura, Y., and Kishimoto, T., J. Immunol. 128, 1296, 1982. Stevens, R. H., Macy, E., and Thiele, C. J., Stand. J. Immunol. 14, 449, 1981. Lucivero, G., Lawton, A. R., and Cooper, M. D., Clin. Exp. Immunol. 45, 18.5, 1981. Kuritani, T., and Cooper, M. D., J. Exp. Med. 155, 1561, 1982. Dagg, M. K., and Levitt, D., Clin. Zmmunol. Immunopathol. 21, 39, 1981. Levitt, D., Duber-Stull, D., and Lawton, A. R., Clin. Zmmunol. Zmmunopathol. 16, 309, 1981. Sager, D. S., and Jasin, H. E., Clin. Exp. Zmmunol. 47, 645, 1981. Zabala, C., and Lipsky, P. E., J. Zmmunol. 129, 2496, 1982. Saxon, A., Feldhaus, J., and Robbins, R. A., J. Immunol. Methods 12, 285, 1976. Kuritani, T., and Cooper, M. D., J. Exp. Med. 155, 389, 1982. Ito, S., and Lawton, A. R., J. Zmmunol. 133, 1891, 1984. Rosenberg, S. A., and Lipsky, P E., J. Zmmunol. 122, 926, 1979. Matsushima, K., Procopio, A., Abe, H., Scala, G., Ortaldo, J. R., and Oppenheim, J. J., J. fmmunol. 135, 1132, 1985. Elmer, J. J., and Spagnuolo, P. J., J. Immunol. 123, 2689, 1979. Wood, D. D.. J. Immunol. 123, 2400, 1979. Wood, D. D., and Cameron, P. M., J. Zmmunol. 121, 53. 1978. Rosenberg, S. A., and Lipsky, P. E.. J. Zmmunol. 126, 1341, 1981. Kehrl, J. H., Muraguchi, A., Butler, J. L., Falkoff. R. J. M., and Fauci. A. S., Immunol. Rev. 78, 75, 1984.
LPS AUGMENTS
HUMAN
B-CELL
DIFFERENTIATION
271
28. Kishimoto, T., Yoshizaki, K., Kimoto, M., Okada, M., Kuritani, T., Kikutani, H., Shimizu, K., Nakagawa, T., Nakagawa, N., M&i, Y., Kishi, H., Fukunaga, K., Yoshikubo, T., and Taga, T., Immunol. Rev. 18, 97, 1984. 29. Falkoff, R. J. M., Butler, J. L., Dinarello, C. A., and Fauci, A.. S., J. Immunol. 133, 692, 1984. 30. Jelinek, D. F., Splawski, J. B., and Lipsky, P. E., J. Immunol. 136, 83, 1986. 31. Layton, J. E., Krammer, P. H., Hamaoka, T., Uhr, J. W., and Vitetta, E. S., J. MO/. Cell. Immunol. 2, 155, 1985. 32. Aman, I?. Ehlin-Henriksson, B., and Klein, G., J. Exp. Med. 159, 208, 1984. 33. Chan, M. A., Stein, L. D., and Dosch, H. M., J. Immunol. 136, 106, 1986. Received September 5, 1986; accepted with revision April 13, 1987.