rournal of Autoimmunity (1989) 2,269-282 Investigations of Intrinsic Abnormalities in DNA-specific B Lymphocytes from Autoimmune Mice Marlene Aldo-Benson Department of Medicine, Division of Rheumatology, Indiana University School of Medicine, Indianapolis, IN 46223, USA In murine models of systemic lupus erythematosus and in many humans with SLE, antibodies against native DNA (dsDNA) are a major contributor to the pathogenesis of the disease. Loss of self-tolerance to the DNA antigen may be associated with B-cell defects or regulatory cell dysfunction. We have developed B-cell lines with specificity for the antigen DNA, from both the autoimmune BWF, mouse strain and from the non-autoimmune BALB/c strain, to use in the investigation of inherent B-cell defects in autoimmunity. Six BWF, cell lines and five BALB/c cell lines which are free of Thyl.2+ cells and esterase positive cells, and have between 35 and 89% rosetting with dsDNA-SRBC targets, have been propagated in vitro for 24-36 months. The cells are non-malignant, growth-factor dependent and have no antigen or mitogen in the growth medium. Lyt-1 positive cells are found in the cell lines, but Lyt-1 negative cells are also present. They respond to the antigen DNA-HRBC when EL-4 supernatant is present in culture, and the peak of the plaque-forming cell (PFC) response is the same for both strains. When cells from both strains are cultured with varying amounts of T-cell factors, there is no difference in spontaneous antibody-forming cell (AFC) forma- tion or in response to anti-mu stimulation between BWF, and BALB/c strains. BALB/c spleen cells do not respond to DNA-HRBC in this culture system, but BWF, spleen cells, as well as cell line cells from both strains, respond to this antigen. T cells from non-responding BALB/c spleen and responding BWF, spleen are able to suppress the immune response to DNA-HRBC of cell line B cells from both strains. Propagating B-cell lines in the presence of DNA for 2 weeks stimulates BWF, cell line cells, but suppresses the response of BALBfc cell lines to antigen. Introduction Anti-DNA antibodies are felt to be important in the pathogenesis of both human SLE and diseases resembling SLE in autoimmune strains of mice. Although B 269 0896-8411/89/030269+ 14 $03.00/O 0 1989 Academic Press Limited
Transcript
rournal of Autoimmunity (1989) 2,269-282
Investigations of Intrinsic Abnormalities in DNA-specific B Lymphocytes from Autoimmune
Mice
Marlene Aldo-Benson
Department of Medicine, Division of Rheumatology, Indiana University School of Medicine, Indianapolis, IN 46223, USA
In murine models of systemic lupus erythematosus and in many humans with SLE, antibodies against native DNA (dsDNA) are a major contributor to the pathogenesis of the disease. Loss of self-tolerance to the DNA antigen may be associated with B-cell defects or regulatory cell dysfunction. We have developed B-cell lines with specificity for the antigen DNA, from both the autoimmune BWF, mouse strain and from the non-autoimmune BALB/c strain, to use in the investigation of inherent B-cell defects in autoimmunity.
Six BWF, cell lines and five BALB/c cell lines which are free of Thyl.2+ cells and esterase positive cells, and have between 35 and 89% rosetting with dsDNA-SRBC targets, have been propagated in vitro for 24-36 months. The cells are non-malignant, growth-factor dependent and have no antigen or mitogen in the growth medium. Lyt-1 positive cells are found in the cell lines, but Lyt-1 negative cells are also present. They respond to the antigen DNA-HRBC when EL-4 supernatant is present in culture, and the peak of the plaque-forming cell (PFC) response is the same for both strains. When cells from both strains are cultured with varying amounts of T-cell factors, there is no difference in spontaneous antibody-forming cell (AFC) forma- tion or in response to anti-mu stimulation between BWF, and BALB/c strains.
BALB/c spleen cells do not respond to DNA-HRBC in this culture system, but BWF, spleen cells, as well as cell line cells from both strains, respond to this antigen. T cells from non-responding BALB/c spleen and responding BWF, spleen are able to suppress the immune response to DNA-HRBC of cell line B cells from both strains. Propagating B-cell lines in the presence of DNA for 2 weeks stimulates BWF, cell line cells, but suppresses the response of BALBfc cell lines to antigen.
Introduction
Anti-DNA antibodies are felt to be important in the pathogenesis of both human SLE and diseases resembling SLE in autoimmune strains of mice. Although B
lymphocytes specific for the antigen DNA are present in normal humans, anti-DNA antibody is at very low levels; however, patients with active SLE have an increased number of B cells specific for the antigen DNA, and may have circulating anti-DNA antibodies [l-3]. Similarly, in a murine model of SLE, most of the (NZB x NZW) Fl hybrid (BWF,) mice develop anti-double-stranded DNA (dsDNA) antibodies and also develop nephritis [4]. Native DNA-anti-DNA im- mune complexes have been implicated in glomerulonephritis of both BWF, mice and humans [5,6]. It is still unclear whether the production of this autoantibody is due to abnormality in the B or regulatory cells. Thus, the study of the B cells pro- ducing anti-DNA antibodies in BWF, mice may enhance our understanding of the mechanisms for loss of tolerance to dsDNA in this type of autoimmune disease.
We are using continuous B-cell lines specific for the antigen DNA from both the autoimmune BWF, mouse strain and the non-autoimmune BALB/c mouse strain to study B-cell abnormalities in autoimmunity. These cells are non-malignant and do not require antigen or mitogen in the culture medium for growth. However, they are dependent upon the growth factors contained in supernatant from EL-4 lymphoma or recombinant IL-4 for continued growth. Cells from both BWF, and BALB/c cell lines can be stimulated by the antigen DNA bound to horse red blood cells (DNA- HRBC) but the antigenic response requires the presence of T-cell factors from EL-4 supernatants, and is accentuated by the presence of irradiated thymus filler cells. The anti-DNA response is DNA specific. These B-cell lines from both autoimmune and non-autoimmune strains were used to study the loss of self tolerance and auto- immunity. BWF, B-cell lines were not more sensitive to T-cell factors than BALB/c cells and both B cells could be suppressed by T cells from both BWF, and BALB/c spleens. Propagating these cell lines in the presence of DNA suppressed the BALB/c response to DNA antigen but stimulated the BWF, B cells.
Materials and methods
Animals
Spleens from 18 to 20 week-old BWF, mice obtained from Jackson Laboratories, Bar Harbour, Maine and 12-16 week-old BALB/c mice obtained from Harlan Industries, Indianapolis, Indiana were used to prepare DNA-specific cell lines and as a source of T lymphocytes. Four to five week-old Harlan Sprague Dawley rats, used as a source of thymus filler cells, were also obtained from Harlan Industries.
Preparation of DNA-erythrocytes
Either sheep or horse erythrocytes were coated with DNA (single or double stranded) [7]. Briefly, 200 ug/ml of calf thymus dsDNA (Sigma Chemical, St. Louis, MO) is mixed with 0.8 ml of 1.2% saline and 0.1 ml of CrC1,.6H,O at 20 mg/ml. To this mixture 0.2 ml of a 50% red blood cell (RBC) solution was added and mixed at room temperature for 5 min. The cells were then washed and used in either a rosette assay or a Cunningham PFC slide assay, or as antigen.
Enhancement of DNA binding cells
Spleen cells from both BWF, and BALB/c mice were enriched for numbers of B cells specific for the antigen DNA by panning on DNA-coated petri dishes. These dishes
DNA-specific I3 lymphocytes in autoimmunity 271
were prepared by the method of Sasaki et al. [8]. Briefly petri dishes were precoated with 4 mg protamine sulfate, and 40 mg calf thymus dsDNA (Sigma Chem. Co., St. Louis, MO) was added to each dish. After 2 h incubation at room temperature dishes were washed with phosphate buffered saline (PBS) and spleen cell suspensions were added. After 1 h incubation at 37°C plates were washed three times with warm RPM1 1640 to remove unbound cells. Bound cells were eluted with cold PBS con- taining 50); FCS. The percentage of DNA-specific cells in the enriched population was determined by rosetting with DNA-coated sheep erythrocytes.
Propagation of cell lines
The enriched cell population was dispersed and grown in soft agar using the method of Pillai and Scott [9] which has been modified as previously described [lo]. After the colonies had grown to macroscopic size they were picked from the soft agar with the aid of a dissecting microscope, and each colony was placed in one well of a 96-well microtiter plate in Dulbecco’s Modified Eagles Medium (DME) containing 20°, supernatant from EL-4 thymoma. When cells in each well had become confluent, they were harvested and placed in a single well of a 24-well tissue culture plate. When cells had again grown to confluency they were tested for the presence of T cells by cytotoxicity with anti-thy 1.2, (Ortho Diagnostics, Rariton, NJ) and for macro- phages using an esterase stain as previously described [ 111. They were tested for relative DNA specificity using a standard rosetting assay with DNA-SRBC as tar- gets. Cell lines with no T cells or macrophages which were highly DNA-specific were propagated in the 24-well tissue culture plate, and then transferred to 25 cm2 tissue culture flasks. Fresh DME with 20% EL-4 supernatant was supplied every 3 d, and when cells from a single flask became confluent they were transferred into two tissue culture flasks.
Growth factors
Supernatant was prepared from the EL-4 lymphoma as previously described [lo]. Briefly, EL-4 cells were harvested from the peritoneum of mineral oil-primed mice and placed in tissue culture in RPM1 1640 without fetal calf serum. The cells at a density of 2 x 106/ml were stimulated with 2 pg/ml PHA (Sigma Chemical Co., St. Louis, MO.). After 24 h in culture the cells were washed, the supernatant dis- carded and the cells placed back in culture for the remaining 48 h. At the end of this 48 h, the supernatant was harvested and stored in aliquots at -70°C. Supernatant prepared in this way has no IL-2 activity as determined by CTLL assay, but has IL-4 activity. Recombinant IL-4 was obtained from Genzyme (Boston, MA).
Antigen or anti-mu stimulation of lymphocytes
Cell line cells ( 105) or splenic lymphocytes were placed in each well of a 96-well tissue culture plate along with 50 ~1 of 0.01 y. DNA-HRBC in DME with 10% fetal calf serum. Either 10% EL-4 supematant or 20% EL-4 supematant and 1 x lo* irra- diated rat thymus cells (2,000 rads) were added to each well and the cells allowed to incubate for 4 d. In some cases irradiated syngeneic spleen cells were used as filler
272 M. Aldo-Benson
cells, and gave similar immune responses, but for economy the rat thymus cells were used in most cases. On day 5 the cells were harvested and a Cunningham slide plaque-forming cell-assay was done using DNA-SRBC as target cells. [ 111. Initially, background PFC of uncoated SRBC was also determined simultaneously.
For stimulation of cell lines with anti-mu antibody, only large sIg+ cells were used, so that the relative number of antibody-forming cells could be compared for BALB/c and BWF, cell lines. These cell lines were placed on a Percoll gradient and cells from the 20-50% fractions were harvested and washed. We have previously shown that 70% of the cells from this fraction are in Go. sIg+ cells ( 104) were placed in each well of the microculture plate with varying amounts of EL-4 supernatant and usually 25 pg/ml goat anti-mouse IgM F(ab),. No filler cells were used. After 4 d in culture the antibody-forming cells were enumerated using a reverse PFC assay with Staph protein-A coated sheep red blood cells (SRBC) as targets. Anti-mouse IgG, IgA and IgM (GAM), which recognizes all murine isotypes, was used as developing antisera.
Preparation of T lymphocytes
Spleen cells were obtained from 20-week-old BALB/c or BWF, mice and panned on petri dishes coated with rabbit anti-mouse Ig. The eluted cells were passed over a G- 10 column to remove macrophages and an aliquot of the remaining cells was washed and stained with fluorescein anti-Thy 1.2 and esterase to determine the purity of the population. T cells were always 95% Thy 1.2+ and less than 2% esterase positive.
Assay for surface Lyt-I
sIgM+ cells from each cell line were separated on a Percoll gradient and washed. Cells were incubated with fluorescein labeled anti-Lyt-1 (Becton Dickinson, Mountain View, CA), washed and analyzed for percent positive cells on Coulter Epics IV fluorescence activated cell sorter, using 525 wavelength with narrow pass band filter.
Statistics
Mean PFC for cell lines from each strain was determined from the value of 4-6 separate determinations. Comparisons of the response of either cell line to anti-mu were done using a repeated-measures analysis of variance and the Tokey multiple comparison procedure to avoid Type I error. All experiments were done with multiple cell lines but in some cases representative experiments are presented.
Results
Spleen cells from BWF, and BALB/c mice showed up to 30% DNA-specific rosette- forming cells (RFC) after panning on DNA coated plates. After soft agar propagation colonies were picked and placed in a single well of a 96-well microtiter plate, the cell lines which grew to confluency were examined for the presence of T cells and esterase-positive cells. Cell lines which were found to have less than 1 o/0 T cells or
DNA-spkd& B fj&frocytes in autoimmunity 273
Table 1. DNA-speci$c cell lines
Cell lines % T cells’ Esterase + cells DNA rosettes’ Lyt- 1
‘Determined by cytotoxicity with anti-Thy 1.2. 2Expressed as Ou of total cells.
esterase-positive cells were then examined using the rosetting assay with DNA- SRBC as targets to determine which cell lines had a high specificity for the DNA antigen. This procedure was done on three separate occasions for each mouse strain. Six cell lines showed a high specificity for DNA (Table 1). Cell lines from BALB/c which had fewer than 1% T cells or macrophages were also obtained by similar procedures. Five of these showed a high specificity for DNA (Table 1). The percent RFC in each cell line varies in a cyclic fashion similar to that described for DNP specific cell lines, but maximum RFC are shown here [12, 131. Those BALB/c and BWF, cell lines which are listed in Table 1 were used in the remaining studies.
The cell lines described in Table 1 have similar characteristics to those described for dinitrophenyl (DNP)-specific continuous B-cell lines [ 121. They are composed of 10-20 micron diameter surface immunoglobulin-bearing cells and small (5-10 pm) cells which are negative for surface immunoglobulin. The percent of sIgM+ cells varies in a cyclic fashion from as low as 20% to as high as 80% of total cells. They are non-malignant and will not grow as tumors in irradiated syngeneic recipients. They require EL-4 supernatant for continuous growth but will proliferate only with recombinant IL-4 and fetal calf serum in the medium. Antibody against IL-4 inhibits their growth. There is no antigen or mitogen present in the continuous growth media.
Several of the BWF, and the BALB/c cell lines have a significant percentage of cells which stain for Lyt-1 antigen. BWF, 13L and 4 had the highest percentages of Lyt-1 cells while BW 24 and 22 had lower percentage of positive cells. Of the BALB/c cell lines tested, 1D and 5B had the highest percentage of positive cells (Table 1). A representative fluorescence activated cell sorter (FACS) histogram of BALB/c 5B is presented in Figure 1. A DNP-specific cell line from BWF, strain did not have Lyt+ cells.
To determine the requirements for antigen-specific stimulation of these cell lines, several cell lines from either strain were cultured in the presence of DNA-HRBC both with and without EL-4 growth factor present. A representative experiment is shown in Table 2. Irradiated thymus filler cells were present in all cultures. Antigen
274 M. Mdo-Benson
0
Figure 1. FACS histogram of cell line 5B stained with Fl-anti-Lyt-1. There is some background autofluorescence of the cell line which is most prominent between channel 5 and 20. The integral for calculation of percent cells binding Lyt-1 is determined between channel 20 and 70 to minimize the contribution of background autofluorescence. 25.6% of cells stained with Fl-anti-Lyt-1. ( -) Lyt-1 fluorescence; (. . .) autofluorescence.
Table 2. Requirement for T-cell factors in immune response to DNA-HRBC
‘20% supematant from EL-4 thymoma. *DNA-SBBC used as target cells.
without EL-4 supernatant gives a background number of anti-DNA plaque-forming cells. However, addition of EL-4 supernatant significantly increases the PFC response. Filler cells alone do not give a PFC response (data not shown). The percent of sIg+ cells increases after antigenic stimulation.
To determine the specificity of the PFC response, cell-line cells were cultured with DNA-HRBC with EL-4 supernatant and filler cells. On day 5 a PFC assay was done using SRBC, DNA-coated SRBC, or dinitrophenyl (DNP)-coated SRBC as targets. The results of a representative experiment (Figure 2) demonstrate that a significant increase in plaque-forming cell-response above background occurs only with DNA- HRBC or DNA-SRBC as targets. There is no increase above background with HRBC or DNP-SRBC, nor is there any response above background to HRBC targets (data not shown). The SRBC background was subsequently performed with all PFC assays in all cell lines and was always negligible. The PFC response to DNA-HRBC could be inhibited by adding free DNA to the plaquing mixture
DNA-specifk B lymphocytes in autoimmunity 275
14
2
SRBC DNA-HRBC DNA-SRBC TNP-SRBC
Torget cet Is
Figure 2. Specificity of PFC response in cell-line lymphocytes. DNA-specific B-cell lines were cultured alone or with DNA antigen and the PFC response determined on day 5 of culture using different target cells. The PFC response in antigen-stimulated cells increased only when DNA targets were used. 0, unstimulated cells; q , DNA-HRBC, stimulated cells.
- t 26 ----c BALE/C
% -+t-- BWF,
”
3 4 5 6
,Doys after oddOlon of ontlgen
- ! Figure 3. Kinetics of direct PFC response. DNA-specific cell lines from either BWF, (--x--) or
BALB/c (-•-) mice were stimulated with DNA-HRBC antigen on day zero, and the PFC assay was done using DNA-SRBC on days 3 through 6. The peak response is the same for both BWF, and BALB/c cell lines.
(DNA-HRBC stimulated cell line cells 25,362 &3,223 l?FC/106 with free DNA 4,859 + 240 PFC/106). Stimulation of DNA cell lines with an irrelevant antigen such as DNP-Ficoll did not increase the PFC response to DNA-SRBC targets (back- ground PFC 5,574 f 1,382; DNP-Ficoll stimulated cells 5,149&946 PFC/106 cells).
Kinetics of the immune response to DNA-antigen in both strains of mice were determined by culturing cells in the presence of DNA-HRBC, EL-4 supernatant and filler cells and assaying for anti-DNA-PFC on day 3 through 6 after initiation of the culture. The response peaks on day 5 in both instances (Figure 3). This is consistent for several cell lines from each strain. When indirect PFC assay was done using DNA-HRBC targets and either anti-IgG or anti-IgA as developing antiserum, we found that the peak IgG response to DNA antigen occurred on day 6. An IgA response was not induced.
Figure 4. Response of varying cell lines to dsDNA and ssDNA. A single-cell line was placed in culture with either native DNA-SRBC (dsDNA-SRBC) or denatured DNA-SRBC (ssDNA-SRBC) as antigen. Cells from each culture were assayed for PFC against both dsDNA and ssDNA coated targets on day 5. Stimulating by both antigens and assays for both targets for a single cell line were done in the same experiment, but each cell line was assayed on a separate day. m, ssDNA-SRBC target; @, dsDNA-SRBC target.
Cell lines from both strains were stimulated in vitro with single stranded DNA (ssDNA) bound to HRBC, and the PFC response determined on day 5 with both ssDNA-SRBC and dsDNA-SRBC as targets. As controls the same cell lines were also stimulated with dsDNA-HRBC and PFC assayed against both targets. Three of the cell lines responded only to dsDNA antigen and produced only antibody which bound to dsDNA targets (cell lines BW24, BW13 and BALB/c 4B) (Figure 4). One of the BALB/c cell lines responded to both antigens, and produced antibody reacting with both targets suggesting that the specificity of the response was to an epitope which is cross-reactive to both dsDNA and ssDNA.
TocomparetheimmuneresponseofcellsfromBWF, andBALB/ccelllinesweused only cells which were sIg+. Since they were from the same fraction of the Percoll gradient and we have previously shown that cells from this fraction are mainly in Go stage of the cell cycle, (manuscript submitted) it is likely that cells from both groups were at Go at the start of the cultures. Because the percentage of cells specific for DNA is not the same in each cell line, we stimulated these cells using25 ug/ml of goat anti-mu antibody, [F(ab)2] which was found to be optimal for antibody formation in BALB/c cell lines. There was no DNA antigen in these cultures and filler cells were not required for a response. We measured antibody-forming cells using a reverse PFC assay with anti-GAMantiserum and Staph A-coated SRBC as targets. Varying amounts of EL-4 supernatant were added to the above cultures and 5 cell lines from each strain were compared with respect to spontaneous AFC and response to anti mu. There was no difference in either spontaneous AFC or the response to anti-mu between either strain at any concentration of EL-4 supernatant (Table 3). No cell line within a strain was significantly different from the others in that strain (data not shown).
Spleen cells from BALB/c and BWF, mice were placed in culture with DNA- HRBC antigen and at the same time the anti-DNA response of cell-line cells was
Table 3. ESfect of T-cellfactors on slg’ B cells from BALB/c and BWF, mice
tested using the same culture system. BALB/c spleen cells did not produce an anti- DNA response (Table 4) while BALB/c cell line cells consistently responded to this antigen. The BWF, spleen cells responded well to antigen as did the BWF, cell line cells.
To test the possibility that the difference in response of splenic B cells is due to a difference in activity of suppressor T cells (Ts) in spleens of these two strains, studies were done in which T cells were added to cell line B cells at the initiation of the culture with antigen. T cells were obtained from BALB/c, BWF, or NZB spleens. An aliquot of the spleen cell suspension from which T cells were obtained was always cultured with antigen at the time of the culture with added T cells. The T cells were added to lo5 cell-line B cells in varying numbers at the start of culture with DNA- HFUX antigen, and the anti-DNA response was compared to cell-line cells cultured alone. Nine percent BALB/c T cells effectively suppressed anti-DNA PFC of both BALB/c and BWF, B cell lines (Table 5). Three percent T cells suppressed BALB/c cells in most cases and also suppressed 2 BWF, cell lines. Similarly only 3”/, T cells from BWF, mice were able effectively to suppress anti-DNA response of BWF, B-cell lines. Hz-compatible T cells from autoimmune NZB mice were also able to
278 M. Aldo-Benson
Table 5. BALBIc ceil-line response to DNA-HRBC with or without addition of T cells
IT cells separated from spleen 95“~ Thy 1.2; less than 2% esterase + ‘No rat thymus filler cells added.
suppress BALB/c B-cell lines. Irradiation of the T cells from all three strains reversed the suppression and 95% B cells from these spleens were not suppressive (data not shown).
Cell-line B-cells responsive to dinitrophenyl (DNP) can be inhibited from a re- sponse to DNP antigen by propagation for 14 d with small amounts of antigen [13]. We next tested whether progagation of these DNA-specific B cells in the presence of DNA could have an effect on the response of these B cells to antigen. Cell lines from each strain were grown in Dulbecco’s Modified Eagle’s Medium (DME) with FCS but varying amounts of DNA were added to the medium for 14 d. Cells were then washed extensively to remove DNA and then placed in culture with filler cells with or without antigen. After 5 d the PFC response was assayed. BWF, B-cell lines propa- gated in the presence of 5 pg/ml DNA showed a marked spontaneous increase in PFC which was not appreciably augmented by the presence of antigen (Figure 5). BALB/c B-cell lines did not show an increase in spontaneous PFC when propagated in 5 pg/ ml DNA, and stimulation with antigen did not significantly increase the PFC above background in this group.
Discussion
In this report we describe BALB/c and BWF, cell lines specific for DNA. This may be a useful model for studying B-cell characteristics which might account for the loss of tolerance to DNA in the autoimmune strains. Since their method of isolation and propagation is the same as that used for DNP-specific cells, it is not surprising that the growth characteristics of these cell lines are the same as those of previously reported cell lines [ 121.
The cell lines can be propagated using recombinant interleukin-4 and do not require other lymphokines once they have been established. However, we cannot be sure that the other lymphokines are not necessary for their initial growth. Since our
DNA-specific B lymphocytes in autoimmunity 279
75
$ 0 50 ” m = 25 8
w V
,‘j 0 w/ml 2 W/ml 5pg/ml
25
Opg/ml 2 pg/m I 5 pg/ml
dsDNA m culture
Figure 5. Effect of DNA in culture medium on responsiveness of cell-line lymphocytes. Cell-line lymphocytes were propagated with the DME and growth factors but native DNA was added to aliquots of each cell line at the start of culture and maintained in culture for 2 weeks. At the end of 2 weeks all cells were washed extensively and cells from each group placed in culture with filler cells either with or without antigen. In BWF, 13L propagation with both 2 ug/ml and 5 pg/ml DNA stimulated backgroundPFC above control (0 ug/ml). Further addition of antigen (DNA-SRBC) did not increase the response. BALB/c 7D was stimulated by propagation with 2 ug/ml of DNA but addition of antigen suppressed these cells. Presence of 5 ug/ml DNA for 2 weeks inhibited a response above background controls. 0, background; H ,
DNA-SRBC.
EL-4 supernatant has no IL-2 activity, it is certain that IL-2 is not required for initial growth of these cells. The specific interleukins required for antibody secretion after stimulation by antigen or anti-mu are presently under investigation. However, we have established that filler cells are only important in antigen-induced immune response and are not required for anti-mu induced response. This suggests that filler cells function as antigen-presenting cells. Syngeneic spleen cells function as effi- ciently as rat thymus filler cells, indicating that xenogeneic stimulation does not play a part in induction of antibody response.
The Lyt-1 antigen is present on at least some of the cells from all lines tested. Since there are no cells bearing Thy 1.2 in these cell lines the Lyt-1 + cells are most likely in the B-cell compartment. Thus, the characteristics of these cells are at least partially representative of the Lyt- 1 -positive B cells, which are said to be the major secretors of IgM autoantibodies in murine models of lupus [ 141. It is possible that only the Lyt-negative cells are responding to antigen. However, the cell lines with a larger proportion of Lyt-1 -bearing cells have the same responses as those with fewer Lyt- l- positive cells, which argues against this possibility. There are several possible explanations for cell lines having both Lyt+ and Lyt- cells. Each line may be a mixture of several clones which are heterogeneous with respect to Lyt-1 status, or some of the cells may have low density of Lyt-1, which cannot be detected. There is no evidence that the cell lines are clones.
280 M. Aldo-Benson
The cell lines used in these experiments have a high percentage of cells which rosette with dsDNA SRBC. These cell lines produce a small number of PFC against SRBC or HRBC themselves. This background is most likely artifactual, and not due to actual anti-SRBC or HRBC antibodies, since stimulation of cell lines by DNA- HRBC or DNA-SRBC did not raise the anti-SRBC or HRBC PFC response. Stimulation with DNA-erythrocytes increased only the PFC response against DNA- coated target cells, indicating that this is a specific response.
The rationale for using cell-line B lymphocytes is to eliminate as far as possible any host or regulatory effects from the B-cell response, so that differences intrinsic to the B cell can be examined. Additionally, manipulation of cells with varying regulatory factors can be more readily achieved with an easily available source of B cells which show similar characteristics from one experiment to another. One of the drawbacks of using cell lines for studies of B-cell function in general is that we cannot always be sure that the response of a single cell line will be representative of DNA-specific B cells in general. However, testing several different cell lines can help to minimize this problem. Another problem with using cell lines is that cells which have been in culture for long periods may be different from cells found in vivo. However, this may also be the advantage of this system since any differences found in the cells may be inherent to the cell itself.
The fact that B-cell lines from non-autoimmune BALB/c mice can be stimulated by DNA antigen in vitro while the spleen cells do not respond to antigen, suggests that culturing these cells for long periods abrogates their self tolerance. One explanation for BALB/c cell-line responsiveness may be the loss of Ts influence. Restoration of non-responsiveness by even small numbers of splenic T cells supports this idea. Another explanation is that in vitro propagation removes from the cells significant levels of their self-antigen DNA and thus the newly developing cells are not tolerant. Levels of DNA due to cell death in normal propagations of cell lines are very low. This explanation would be compatible with both Triplett’s experiments in self toler- ance and with the studies of Nossal et al. who have shown that cells which develop sIg receptors in the presence of anti-receptor antibody are unresponsive to stimulation [ 15,161. This explanation is supported by the experiments in which propagation of BALB/c cell-line B cells with 5 pg/ml DNA inhibits their ability to respond to DNA antigen. Normal humans do have free circulating DNA, but there is some contro- versy about the levels found in normals [ 171. However, patients with active SLE can have up to 4 l.tg/ml of free circulating DNA [ 181 Thus, the role of free DNA in the regulation of DNA response is still uncertain. Of course it is possible that a combination of these and other unidentified factors are responsible for this responsiveness in BALB/c cell-line B cells, in contrast to unresponsive BALB/c spleen cells.
The loss of normal antigenic non-specific suppressor activity from the spleens of older BWF, or NZB mice is a well recognized phenomenon [ 191. This loss of sup- pression is one of several explanations for autoantibody production in these mouse strains. An excess of T-cell help as well as functional abnormalities of T-cell derived interleukins have also been described as factors responsible for autoimmunity [ 19- 211. Thus, it is interesting to note that BWF, spleen cells from 20-week-old mice which respond well to DNA antigen in vitro do have T cells which are capable of suppressing the response of cell-line B cells. This suggests that BWF, spleen-cell
DNA-specific B lymphocytes in autoimmunity 281
autoimmunity is not due to an absence of Ts. Additionally, since BWF, B cells can be suppressed by Ts from both BALB/c and BWF, spleens, it seems likely that they are capable of receiving signals for suppression. It is not clear whether these T cells are antigen-specific or non-specific. This experimental system used relatively few T cells and also provided extraneous T-cell help which was necessary for a response. Thus, the effect of excess T-cell help or interleukins as factors in autoimmunity could not be evaluated in these experiments. Thus, the role of the T cell in autoimmunity is still unclear.
The response of BWF, DNA-specific B-cell lines to propagation for 2 weeks with DNA is in striking contrast to the effect of propagation with DNA on B cells from non-autoimmune mice. The BWF, cells show an increase in number of anti-DNA- producing cells both with and without added antigen, while non-autoimmune B-cell lines have a decreased antigenic response to DNA antigen. Previous studies in our laboratory have shown that cell lines specific for DNP hapten from both BWF, and BALB/c strains have a decreased anti-DNP response when propagated for several weeks with DNP-MGG [22]. The response of the autoimmune-derived B cells to free DNA suggests an intrinsic difference in these cells compared to BALB/c cells. This difference may be specific for this antigen. Of course this cannot be extrapolated to events which occur in viva in BWF, cells at this time, but further studies may help to determine whether this is one of the factors in autoimmunity of these mice.
Acknowledgements
The author thanks Mrs Linda Pratt for her excellent technical assistance, Mrs Donna Faulkner for typing the manuscript, and Dr Barry Katz for assistance with statistical analysis.
Supported by USPHS Grant No. AA 07519, AI 21066, AM 7448, AM 20582, and Arthritis Foundation Center Grant and a Grant from the Grace M. Showalter Trust.
1.
2.
3.
4.
5.
6.
7.
8.
References
Bankhurst, A. D. and R. C. Williams. 1976. Cellular origin of autoimmunity-A perplex- ing question. Am. 3. Med. 61: 303-306 Weinstein, A., B. Bardwell, B. Stone, C. Tibbetts, and N. F. Rothfield. 1983. Antibodies to native-DNA and serum complement levels; Application to diagnosis and classification of systemic lupus erythematosus. Am. 3. Med. 74: 206-214 Bankhurst, A. D. and R. C. Williams. 1975. Identification of DNA binding lymphocytes in patients with systemic lupus erythematosus. 3. Clin. Invest. 56: 1378-1383 Lambert, P. H. and F. B. Dixon. 1968. Pathogenesis of the glomerulonephritis of the NZB/W mice. 3. Exp. Med. 127: 507-522 Dixon, F. B., M. B. A. Oldstone, and G. Tonietti. 1971. Pathogenesis of immune complex glomerulonephritis of New Zealand mice. 3. Exp. Med. 134: 66s-72s Koffler, D., P. H. Schur, and H. G. Kunkel. 1967. Immunologic studies concerning the nephritis of systemic lupus erythematosus. 3. Exp. Med. 26: 607-623 Sasaki, T. 1981. In Methods in Enzymology. J. J. Langone and H. VanVunakis, eds. Academic Press, NY. pp. 206 Sasaki, T., F. Endo, M. Mikami, Y. Sekiguchi, K. Tada, Y. Ono, N. Ishida, and K. Yoshinaga. 1984: Establishment of human monoclonal anti DNA antibody producing cell lines. 3. Immunol. Methods 72: 157-165
282 M. Aldo-Benson
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Pillai, P. S. and D. W. Scott. 1981. Hapten specific murine colony forming B cells: In Vitro response of colonies to fluoresceinated thymus independent antigens. J. Zmmunol. 126: 18821886 Aldo-Benson, M. A. and L. M. Scheiderer. 1983. Long-term growth of lines of murine dinitrophenyl specific B lymphocytes in vitro. 3. Exp. Med. 157: 342-347 Cunningham, A. J. and A. Szenberg, 1968. Further improvements in the plaque tech- nique for detecting single antibody forming cells. Immunology 14: 599-608 Aldo-Benson, M. A., L. M. Scheiderer, and F. E. Dwulet. 1986. DNP-specific continu- ous B-cell lines as a model system for studying B-cell activation and tolerance. Eur. 3. Zmmunol. 16: 69-74 Aldo-Benson, M. A. 1986. An in vitro model for clonal anergy in continuously growing antigen specific B-cell lines. Cell. Zmmunol. 101: 391-402 Hayakawa, K., R. Hardy, D. R. Parks, and L. A. Herzenberg. 1983. The Ly-1 B-cell population in normal, immunodefective, and autoimmune mice. 3. Exp. Med. 157: 202-2 18 Triplett, E. L. 1962. On the mechanism of immunologic self recognition.3. Zmmunol. 89: 505-510 Nossal, G. J. V. 1983. Cellular mechanism of immunologic tolerance. Ann. Rev. Zmmunol. 1: 33-62 Steinman, C. R. 1975. Free DNA in serum and plasma from normal adults.3. Clin. Invest. 56: 512-515 Raptis, L. and H. A. Menad. 1980. Quantitation and characterization of plasma DNA in normals and patients with SLE. 3. Clin. Invest. 66: 1391-1399 Steinberg, A. D., D. P. Huston, J. D. Taurog, J. S. Cowdery, and E. S. Ravache. 1981. The cellular and genetic basis of murine lupus. Zmmunol. Rev. 55: 121-154 Laskin, C. A., G. Haddad, and C. A. Soloninka. 1986. The regulatory role of NZB T lymphocytes in the production of anti DNA antibodies in vitro. 3. Zmmunol. 137: 1867-1873. T. J. Santoro, W. R. Benjamin, J. J. Oppenheim, and A. D. Steinberg. 1983. The cellular basis for immune interferon production in autoimmune MRL-lpr/Zpr mice. 3. Zmmunol. 131: 265-268 Aldo-Benson, M. A. and M. Brooks. 1986. Defects in antigen specific immune tolerance in continuous B-cell lines from autoimmune mice. 3. Clin. Invest. 78: 784-789
