| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Internal Medicine,
College of Pharmacy, and
Department of Pathology, University of Michigan, Ann Arbor, MI 48109;
Department of Internal Medicine, University of Pennsylvania, Philadelphia, PA 19104; and
¶ State University of New York Upstate Medical University at Syracuse and Cornell University College of Veterinary Medicine, Ithaca, NY 14850
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
) apoptosis occurs in patients with systemic lupus erythematosus (SLE) and is mediated, at least in part, by an autoreactive CD4+ T cell subset. Furthermore, autoreactive murine CD4+ T cells that kill syngeneic M in vitro induce a lupus-like disease in vivo. However, it is unclear whether increased M apoptosis in SLE per se is sufficient to accelerate/promote autoimmunity. We have investigated whether increased M apoptosis in vivo, induced by the administration of clodronate liposomes, can exacerbate the autoimmune phenotype in NZB x SWR (SNF1) lupus-prone mice, and induce autoantibody production in haplotype-matched BALB/c x DBA1 (DBF1) non-lupus-prone mice. Lupus-prone mice SNF1 mice that were treated with clodronate liposomes, but not mice treated with vehicle, developed significant increases in autoantibodies to dsDNA, nucleosomes, and the idiotypically related family of nephritic Abs IdLNF1, when compared with untreated SNF1 mice. Furthermore, clodronate treatment hastened the onset of proteinuria and worsened SNF1 lupus nephritis. When compared with vehicle-treated controls, clodronate-treated non-lupus-prone DBF1 mice developed significantly higher levels of anti-nucleosome and IdLNF1 Abs but did not develop lupus nephritis. We propose that M apoptosis contributes to the pathogenesis of autoantibody formation and organ damage through both an increase in the apoptotic load and impairment in the clearance of apoptotic material. This study suggests that mechanisms that induce scavenger cell apoptosis, such as death induced by autoreactive cytotoxic T cells observed in SLE, could play a pathogenic role and contribute to the severity of the disease. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
) (3, 4), both in vitro and in vivo (3, 5). This phenomenon correlates with disease activity (3). One of the mechanisms responsible for accelerating apoptosis of monocytes/M in lupus patients is an autoreactive CD4+ T cell subset that kills autologous monocytes in the absence of added Ag (3, 5). In addition, lupus sera can induce monocyte/M death by yet-uncharacterized mechanisms (6). Among scavenger-type cells in the body, M have been implicated in the key functions of uptake and clearance of apoptotic and necrotic cells, with phagocytosis mediated by a variety of cell surface molecules (7).
We propose that accelerated monocyte/M apoptosis in SLE increases the apoptotic load by reducing the clearance of apoptotic cells in addition to contributing to additional apoptotic material (1, 3). Previous studies suggest that high levels of apoptotic and necrotic cells, such as those seen in SLE, can also induce secondary in vivo apoptosis, with the M the most sensitive to die because of their scavenger activity (8). Cell death can therefore induce additional M apoptosis, which in turn leads to the release of nuclear contents, both from the engulfed cell and its own, in a cleaved form into the blood (8). However, whether the enhanced monocyte/M apoptosis observed in SLE plays a pathogenic role or represents an epiphenomenon has not been clarified.
The function of M in lymphoid tissues has been studied in vivo by using the liposome-mediated M depletion model (9, 10, 11). Injected liposomes containing the biphosphonate clodronate (dichloromethylene bisphosphonate) kill phagocytic cells by apoptosis, by the release of this compound following uptake into phagosomes. Liposomes are readily engulfed by M and, once delivered into phagocytic cells, lysosomal phospholipases disrupt the phospholipid bilayers of the liposomes resulting in the intracellular release of clodronate. The phosphonate accumulates intracellularly and, after exceeding a threshold concentration, the cell is irreversibly damaged and dies by apoptosis (12). In vivo, clodronate liposomes are taken selectively by M, predominantly from liver and spleen (10, 13, 14). In the mouse spleen, M close to the bloodstream can be depleted completely with little harm to other cell populations (9, 14).
Using this system, we have examined whether accelerated M apoptosis per se is sufficient to trigger autoimmunity in non-lupus-prone mice and augment autoimmunity and organ damage in lupus-prone mice. We now report that accelerated M apoptosis aggravates lupus nephritis and increases autoantibody formation in lupus-prone mice, and is associated with the development of autoantibodies in non-lupus-prone mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Six- to 8-wk-old female NZB, SWR, DBA1, and BALB/c mice were obtained from The Jackson Laboratory. The F1 progeny of NZB x SWR (SNF1) and BALB/c x DBA1 (DBF1) mice were bred and aged at the State University of New York Upstate Medical University at Syracuse, Department of Laboratory Animal Resources, and shipped to the Unit of Laboratory Animal Medicine, University of Michigan, where they were housed in a specific pathogen-free facility. SNF1 mice develop a lupus like-disease with autoantibody formation and severe nephritis (15). DBF1 were used as non-autoimmune haplotype-matched controls, which do not develop lupus-like features (15). Protocol was reviewed and approved by the University of Michigan’s Committee on Use and Care of Animals.
The monocytic/M cells lines RAW and P388D1 were grown in DMEM with 4 mM L-glutamine adjusted to contain 1.5 g/L sodium bicarbonate and 4.5 g/L glucose and supplemented with 10% FBS.
Preparation of liposomes
Egg phosphatidylcholine and distearoyl (polyethylene glycol 2000) phosphatidylethanolamine (DS (PEG2000) PE) were purchased from Avanti Polar Lipids. Cholesterol was obtained from Sigma-Aldrich. Clodronate was a gift from Roche Diagnostics. All of the liposomes were composed of egg phosphatidylcholine, DS (PEG2000) PE, and cholesterol, at a molar ratio of 1.9:0.1:1 and prepared using a previously established method, with some modifications (16, 17). Briefly, a chloroform solution of the above lipids, containing 10 µmol of total phospholipids, was dried in a round-bottom tube under vacuum using a rotary evaporator for 2 h, to form a thin lipid film. The lipid film was hydrated with 1 ml of isotonic HEPES-buffered saline (HBS, 10 mM HEPES, 140 mM NaCl (pH 7.4)) and vortexed thoroughly to form the control liposome suspension (HBS liposomes, called vehicle liposomes). A 100 mM isotonic clodronate solution in HBS was used instead as the hydrating solution, to form the clodronate liposome suspension (clodronate liposomes). The resulting liposome suspensions were freeze-thawed four times and extruded four times, respectively, through 0.4- and 0.2-µm pore size Poretics polycarbonate filters (Osmonics Lab Products). Any un-encapsulated clodronate was removed from the liposome suspension by size exclusion chromatography using a 1 x 20-cm Sepharose CL-4B column (Pharmacia) containing isotonic HBS as the elution buffer. Liposomes were stored under argon at 4°C and used in experiments within 3 days after preparation. The liposomal phospholipid concentration was measured by a standard phosphate assay (18), and the lipid concentration of control (HBS) liposomes was matched to that of clodronate liposomes for each experiment. The mean diameter of each liposome formulation was assessed by quasi-elastic light scattering method using a Nicomp 380 ZLS Particle Sizer. The amount of clodronate in each liposome formulation was routinely measured from its absorbance peak at 226 nm, using a DU650 spectrophotometer (Beckman Instruments). The measured absorbance value was corrected for the absorbance due to the lipid component, and then converted into the amount of clodronate by normalizing with a clodronate calibration curve generated using at least three different clodronate concentrations.
In vitro characterization of M apoptosis
To verify that the clodronate liposome preparation used for the in vivo studies induced M apoptosis, we initially tested its potency on M cell lines in vitro. RAW and P388D1 cells were treated with varying dilutions (1/20 to 1/500) of clodronate liposomes or HBS liposomes for 24–48 h, or left untreated. Cells were harvested, resuspended in Annexin V binding buffer, and stained with 5 µl of Annexin V-FITC, and 10 µl of propidium iodide (PI) (BD Pharmingen), as described previously (19). The percentage of cells that underwent apoptosis or necrosis was determined using a FACScan flow cytometer (BD Biosciences). Cells were considered apoptotic when they stained positive for annexin V and negative for PI, and late apoptotic or early necrotic when they stained positive for both annexin V and PI.
Injection of liposomes and characterization of M apoptosis in vivo
To determine the kinetics of clodronate liposome-mediated M apoptosis, 8-wk-old lupus-prone and non-lupus-prone mice were given clodronate or HBS liposome injections and then euthanized at different time points. Splenic apoptotic M were identified in frozen sections by staining with rat anti-mouse CD11b (BD Pharmingen) and by TUNEL (Roche Molecular Biochemicals) following published methodology (20). The prevalence of apoptotic cells was assessed as the mean number of apoptotic cells/high-power field, and five to six high-power fields were analyzed in each mouse.
Once the kinetics of M apoptosis were established, 8-wk-old SNF1 mice and 14-wk-old DBF1 mice received bimonthly tail vein injections of 0.2 ml of either clodronate liposomes or HBS (vehicle) liposomes for a total of 10–15 injections/mouse. Sera and urine samples were collected periodically. SNF1 mice were euthanized between 30 and 32 wk of age, and DBF1 mice were euthanized at 40 wk of age. At euthanasia, organs were weighed and processed for histological examination. One kidney from each mouse was fixed in 4% paraformaldehyde and processed for conventional histology, while the other kidney and the spleen were embedded in OCT and frozen by immersion in liquid nitrogen.
Immunofluorescence microscopy of tissue sections
Frozen sections of spleen (7 µm) or kidney (4 µm) were cut using a cryostat, dried to a microscope slide, and stored at –20°C until stained. For analysis of M apoptosis in the spleen and kidney, sections were dried at 37°C for 1 h, fixed in 4% paraformaldehyde in PBS, and rinsed in 50 mM Tris (pH 8.0). Splenic and kidney M were identified based upon expression of the marker CD68, and cell death was evaluated by TUNEL assay. Briefly, the fixed sections were permeabilized in 0.1% Triton X-100 in 50 mM Tris for 5 min at room temperature and washed in distilled water, and apoptotic cells were labeled with tetramethyrhodamine-labeled dUTP and TdT (TMR Tunel Assay kit; Roche Biotech). Sections were rinsed once in 50 mM Tris, blocked for nonspecific binding of biotin and avidin with blocking kit from Zymed, and incubated with 0.01 mg/ml biotinylated-anti-CD68 (Serotec) at 4°C for 1 h. After washing twice with 50 mM Tris, the sections were incubated with 0.02 mg/ml fluorescein-conjugated streptavidin (Southern Biotechnology Associates) at 4°C for 1 h, washed twice with 50 mM Tris, and then incubated for 10 min at room temperature in a solution of 5 mM CuSO4 and 5 mM MnCl2 in 50 mM ammonium acetate to quench autofluorescence, rinsed briefly in distilled water, and coverslipped in aqueous-based mounting medium containing an antifade reagent (Prolong; Molecular Probes).
Ig deposition was evaluated in unfixed frozen kidney sections stained with 5 µg/ml FITC-conjugated goat anti-mouse IgM F(ab')2 (Southern Biotechnology Associates) and 10 µg/ml Cy3-conjugated sheep anti-mouse IgG F(ab')2 (Sigma-Aldrich). Sections were washed twice with 50 mM Tris, nuclei were counterstained with 10 µg/ml Hoechst 33346 (Molecular Probes) for 10 min, autofluorescence was quenched, and specimens were rinsed in distilled water and mounted in medium. C3 deposition was examined similarly by staining a kidney section with a 1/200 dilution of anti-C3-FITC (Cappel) for 1 h at 4°C. Glomerular staining was graded according to intensity on a 0 to 4+ scale (0, no staining; 4+, maximum intensity staining), and the pattern of staining (predominantly mesangial vs capillary) was recorded. Background was defined as the strongest level of staining observed in DBF1 mice, and only staining above background was considered positive. All images were acquired using a fluorescence microscope equipped with fluorescein, rhodamine, and DAPI filters (Olympus) and a digital camera. Image acquisition software settings were identical for all fluorescein or rhodamine images and processing was done using Adobe Photoshop, version 7.0.
Assessment of lupus nephritis
Proteinuria was measured at different time points during the liposome injections using Uristix 6 (Boehringer Mannheim). Proteinuria level was scored as follows: 0.5+, 15–30 mg/dl; 1+, 30 mg/dl; 2+, 100 mg/dl; 3+, 300 mg/dl; 4+, >2000 mg/dl. Kidney damage was evaluated in 4-µm paraffin-embedded sections using H&E and Masson trichrome staining. All samples were evaluated by a renal pathologist (P. D. Killen) in a blinded manner. Mice were scored as having glomerulonephritis only when >50% of their glomeruli showed a qualitative (segmental and/or global mesangial, or proliferative) lesion on multiple sections.
Assessment of organ damage
Lungs, heart, and brain were fixed in 4% paraformaldehyde, and 7-µm paraffin-embedded sections were stained with H&E. For spleen, frozen sections were used for H&E staining.
ELISA determination of total IgG, Abs to dsDNA, to ssDNA, to nucleosomes, and IdLNF1 Abs
Only autoantibodies with titers significantly elevated on more than one occasion throughout the study were considered for analysis (21). Total serum IgG concentrations were measured using commercially available kits from Alpha Diagnostic International following the manufacturer’s instructions. In brief, samples were diluted to 1/20,000. Twenty microliters of samples or standards (ranging from 0 to 1000 ng/ml) were added to precoated wells in combination with 80 µl of sample buffer. Plates were incubated for 1 h and washed five times, 100 µl of HRP-labeled anti-mouse IgG conjugate was added, and plates were incubated for 30 min. Plates were washed five times, and TMB substrate solution was added to each well. Plates were further incubated for 15 min, and 100 µl of stop solution was added into each well and mixed. Absorbance was measured at 450 nm using a microplate reader (Biotek). All samples were within the range of standards and in the linear portion of the curve.
Anti-ssDNA and anti-dsDNA Ab titers were determined by coating Immulon 4 plates with purified calf thymus ssDNA (Sigma-Aldrich) or dsDNA (purified pcdna3 vector plasmid; Promega) as described (21). HRP-conjugated goat anti-mouse polyvalent (IgG, IgM, IgA) Ab (Sigma-Aldrich) was used as the secondary Ab, and plates were developed with Sigma Fast tablets. Controls included identical determinations performed in the presence of 2.5 µg/ml purified dsDNA or ssDNA as competitive inhibitor. Results are presented as the mean of triplicate determinations on sera from individual mice. Positive control included pooled ascites from 
6-mo-old female MRL/lpr mice, which was a gift from Dr. G. Glick (University of Michigan, Ann Arbor, MI).
For detection of serum IdLNF1 Ig, purified IgG eluted from the kidneys of 6- to 8-mo-old female SNF1 mice was emulsified in CFA and used to immunize rabbits as described previously (22). Anti-allotypic and anti-rheumatoid factor reactivities were removed from the rabbit sera by adsorption with myeloma Ig of all isotypic classes, and IgG was then purified by protein A affinity chromatography. The specificity of the rabbit anti-IdLNF1 reagent was confirmed by ELISA using plates coated with either SNF1 kidney eluate Ig or control myeloma Ig, as described (22). Nonspecific binding was not detectable in the purified rabbit anti-IdLNF1 reagent, indicating that the adsorptions with myeloma Ig were sufficient to render the anti-IdLNF1 reagent Id specific.
A direct binding ELISA was used to detect IdLNF1 Abs as previously described (22). Briefly, Immulon I plates (Dynatech) were coated overnight at 4°C with 1 µg/ml rabbit anti-IdLNF1 IgG in 0.05 M borate buffer (pH 9.5), washed twice with PBS-1% Tween (PBS-T) and twice with PBS, and blocked for 1 h at room temperature with PBS-1% BSA. The plates were again washed twice with PBS-T, twice with PBS, and 50 µl of diluted serum was added per well in duplicate. A standard curve was constructed using pooled aged (>24-wk-old) SNF1 female serum with 2-fold dilutions ranging from 1/50 to 1/6400 in PBS-T. The plates were incubated overnight, washed twice with PBS-T and twice with PBS, and incubated with 50 µl of goat anti-mouse IgG conjugated to alkaline phosphatase (Boehringer Mannheim) for 2 h at room temperature. After two washes with PBS-T and two washes with PBS, 50 µl of p-nitrophenyl phosphate solution (Sigma-Aldrich) was added for 15 min, and substrate development corresponding to bound anti-mouse-IgG was determined by measuring absorbance at 405 nm using an ELISA microplate reader (Bio-Rad). A standard curve for IdLNF1 was constructed using the absorbance values from the diluted SNF1 sera, with the 1/50 dilution of SNF1 corresponding to 50 U of IdLNF1 Ig. The IdLNF1 concentration in the test sera was determined from this standard curve, and the data were expressed as IdLNF1 units.
To verify that the reactivity was due to idiotypic recognition of IdLNF1 and not rheumatoid factor activity present in the sera of mice prone to autoimmune disorders, ELISA plates coated with 1 µg/ml normal rabbit IgG instead of anti- IdLNF1 IgG, were incubated with pooled sera from 8-wk-old BALB/c mice, SNF1 or DBF1 test sera. SNF1 and DBF1 sera yielded OD405 of <0.2, values that were significantly lower than those attained using anti-IdLNF1-coated plates (range, 0.38–2.0), and similar to those of BALB/c sera. In contrast, MRL/lpr mouse sera generated OD values >1.2, consistent with the high titers of rheumatoid factors found in this strain of autoimmune-prone mice. These results excluded the possibility that the activity of the IdLNF1 assay could be secondary to rheumatoid factor activity in SNF1 or DBF1.
Detection of anti-nucleosome Abs was also performed by ELISA as described previously (23, 24). In brief, mononucleosomes were obtained by micrococcal nuclease digestion of isolated mouse liver nuclei and subsequent purification with a 5–20% sucrose gradient as described (25). The sucrose fraction containing only mononucleosomes was confirmed by SDS-gel electrophoresis; the integrity of mononucleosomes was evaluated by the content of core histones via sulfuric acid extraction. Poly-L-lysine-treated plates were then coated with 1 µg/ml mononucleosomes. Mice sera were tested at a 1/1000 dilution and adsorbed Abs were detected using biotinylated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories) and alkaline phosphatase-conjugated streptavidin. Serum from a diseased MRL/lpr mouse with a high titer of autoantibodies was also tested at 10 different dilutions from 1/250 to 1/128,000, as a positive control.
Cytokine determination
Serum TGF-
1 was measured by ELISA using a Quantikine mouse/rat/porcine TGF-1 immunoassay (R&D Systems) following the manufacturer’s instructions. TNF-
was determined using an ELISA kit from PeproTech. In brief, ELISA plates were coated overnight at room temperature with 1 µg/ml affinity-purified goat anti-mouse TNF- in PBS. The wells were washed, blocked with 1% BSA in PBS, and incubated for 2 h with samples and murine TNF- standards. Plates were washed and incubated for 2 h with 0.25 µg of biotinylated goat Ag-affinity-purified anti-murine TNF-. Samples were washed, incubated with avidin peroxidase (1:2000; Sigma-Aldrich) for 30 min, washed, and developed with ABTS substrate solution (Sigma-Aldrich) at room temperature. The plate was read at 405 nm with wavelength correction set at 650 nm.
Statistical analysis
Statistical significance was assessed using Student’s t test or Mann-Whitney U test unless otherwise noted, and data are given as mean ± SEM unless otherwise noted. Values of p were considered statistically significant at <0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
apoptosis in vitro and in vivo
Previous studies have demonstrated that clodronate-induced apoptosis is specific for M (9) and is not detected in other cells, including T and B lymphocytes and neutrophils (16, 17, 26, 27). To verify that clodronate preparations induce significant M apoptosis, M cells lines P388D1 and RAW were incubated with different concentrations of vehicle or clodronate liposomes for 24–48 h. Clodronate liposomes, but not vehicle liposomes, caused significant P388D1 apoptosis, as measured by annexin V/PI staining (Fig. 1A). Similar results were obtained using RAW cells (vehicle liposomes, 15 ± 3.5%; clodronate liposomes, 76 ± 5.6%; untreated cells, 11 ± 4%; results representing percentage of apoptotic cells ± SEM of three independent experiments; p < 0.05 when comparing clodronate-treated cells with vehicle-treated or untreated cells; p = NS between vehicle-treated and untreated cells).
|
death was detected in the spleen within 24 h postinjection (Fig. 1B). Significant apoptosis was restricted to the red pulp and marginal zone areas (Fig. 1, D and F). In addition, clodronate-treated mice displayed decreased numbers of CD11b+ M in the red pulp and marginal zone areas (Fig. 1H), again consistent with previous studies demonstrating that red pulp and marginal zone M decrease with a single treatment with liposome-encapsulated clodronate within 24 h postadministration (29). No differences in degree of M apoptosis induced by clodronate treatment were detected between lupus-prone and non-lupus-prone mice (data not shown). Similarly, we saw no differences in the degree of M apoptosis between mice treated with HBS liposomes and untreated mice. Autoimmune and non-autoimmune mice develop higher levels of autoantibodies after clodronate treatment
Autoimmune-prone mice.
SNF1 mice that received clodronate liposomes developed levels of anti-dsDNA Abs that were significantly higher than untreated (not shown) and vehicle-treated SNF1 mice (Fig. 2A). The increases were first observed at 8 wk postinitial clodronate injection when mice were 
16 wk old and were sustained throughout the study, even 4 wk after the last injection and just before euthanasia. Titers of anti-nucleosome Abs were also higher in the clodronate-treated lupus-prone mice starting at 12 wk after the first injection when mice were 20 wk old and were also sustained throughout the study, (Fig. 2B). The pathogenic Abs deposited in the renal lesions of SNF1 mice bear idiotypic determinants collectively referred to as IdLNF1 (30). These Abs are primarily IgG and cationic in charge, characteristic of nephritogenic Abs. Increased levels of IdLNF1 Abs were transiently observed in clodronate-treated SNF1, appeared rapidly following the initiation of clodronate injections (5 wk after first injection) when mice were 13 wk old, and then decreased to levels below those of untreated or vehicle-treated mice by 13 wk after the first injection when mice were 21 wk old (Fig. 2C). This pattern of increased IdLNF1 Ab expression is consistent with previous reports in which the ratio of CD4+ IdLNF1-reactive to CD8+ IdLNF1-reactive T cell increases allowing for enhanced IdLNF1 IgG production before deposition in the kidney and onset of nephritis. As deposition occurs, the serum levels of IdLNF1 decrease (31). No significant elevations were seen in total IgG or ssDNA Abs when clodronate-treated SNF1 mice were compared with HBS-treated or untreated SNF1 mice (data not shown).
|
24 wk old and elevations were sustained until euthanasia. Indeed, higher levels were detected even 4 wk after the final clodronate injection. Significant increases in IdLNF1 Ab levels were first detected at 10 wk postinitial injection when mice were 25 wk old and lasted for 4 additional weeks. Afterward, levels decreased to similar or lower levels to the ones seen in the vehicle-treated mice. There were no differences in levels of total IgG, anti-ssDNA and anti-dsDNA Abs between the HBS-treated and clodronate-treated DBF1 mice (data not shown). Thus, administration of clodronate liposomes to the non-lupus-prone DBF1 is sufficient to induce the generation of specific autoantibodies. As expected, the levels of all autoantibodies were significantly higher in the group of SNF1 mice relative to DBF1 mice even after clodronate treatment (p < 0.05; Figs. 2 and 3 and data not shown). Total IgG levels ranged from 6 to 9 mg/ml in the DBF1 mice and from 9.5 to 12 mg/ml in the SNF1 mice, and no significant differences were observed between treated and untreated mice.
|
apoptosis induces increased severity of lupus nephritis
SNF1 mice develop a lethal accelerated glomerulonephritis that peaks at 24–30 wk of age, which is due to the deposition of immune complexes, including anti-dsDNA Abs and IdLNF1 Abs (15, 32). Administration of clodronate liposomes to SNF1 mice significantly hastened the onset of nephrotic range proteinuria by several weeks relative to untreated or vehicle-treated mice (Fig. 4). This acceleration in severe proteinuria development was accompanied by a readily apparent increase in the severity of lupus nephritis by histology (Fig. 5). Indeed, when compared with untreated or vehicle-treated mice, clodronate-treated mice displayed marked increase in mesangial cellularity, mesangial matrix, cellular crescents, and capillary loop deposits, as well as an increase in focal perivascular interstitial infiltrates (Fig. 5). There was a marked increase in thickness of the capillary loops with subendothelial fuchsinophilic deposits in addition to the expansion of the mesangium. There were no significant differences in the severity of nephritis between the vehicle-treated and untreated SNF1 mice. All SNF1 mice displayed significant and equivalent deposition of immune complexes, irrespective of treatment (Fig. 6, C and D). Likewise, clodronate-treated, vehicle-treated, or untreated SNF1 mice had similar levels of CD68+ M infiltrates within their glomeruli (Fig. 6, G and H). M apoptosis was not observed in the glomeruli of SNF1 mice, indicated by the absence of TUNEL-positive cells, suggesting that the increase in organ damage was not due to increased cell death within the glomeruli. There were no differences in the histology of spleen, heart, brain, and lungs between untreated, vehicle-treated, and clodronate-treated mice (data not shown).
|
|
|
infiltrates regardless of treatment (Fig. 6, A and B, and E and F, respectively). Similarly, no histological changes were seen in the spleen, lungs, heart, or brain of DBF1 mice.
As noted above, clodronate induced splenic M apoptosis and decreased the number of tissue-resident M within 24 h both in lupus-prone and non-lupus-prone mice (Fig. 1). As previously described in the SNF1 model, the spleens from these mice showed significant disruption of the typical H&E staining pattern distinguishing red and white pulp (data not shown) (32). M in the DBF1 and SNF1 spleens were localized predominantly within the red pulp and marginal zones as determined by staining with anti-CD68 Ab (Fig. 7), and the pattern of disruption of the splenic architecture in SNF1 mice was also appreciated using this method. Previous studies have shown that, after an initial significant splenic M depletion at 24–48 h following clodronate administration, levels of these cells gradually rise and might actually exceed initial levels within 7 days postclodronate administration (29, 33, 34). Because at least 2 wk elapsed from the final administration of clodronate to euthanasia, confirming previous reports (33, 34), there was no decrease of splenic M or increase in M apoptosis when spleens were analyzed at that specific time point (Fig. 7 and data not shown). This finding suggests that, while the increase in the level of autoantibodies was maintained in the clodronate-treated mice, repopulation of the splenic M population occurred once clodronate injections were discontinued before euthanasia.
|
cytokines that may be relevant to kidney damage in SLE, TGF- and TNF-. There were no significant differences in levels of TGF- between clodronate-treated and vehicle-treated DBF1 and SNF1 mice. TNF- levels were decreased in both SNF1 and DBF1 mice treated with clodronate relative to vehicle-treated mice (mean ± SEM, 405 ± 42 vs 490 ± 55 pg/ml for SNF1 mice and 208 ± 32 vs 360 ± 28 pg/ml for DBF1 mice, respectively), but statistical significance was attained only in the DBF1 mice (p = 0.04). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
is defective in SLE (24, 41, 44). Although a current model proposes that cells that die as part of the normal cell turnover are rapidly phagocytosed by scavenger M without inflammation or immune activation (7, 45, 46), recent data suggest that, under certain circumstances, apoptotic cells may also promote proinflammatory immune responses by eliciting the release of chemokines and proapoptotic molecules from M (47), or caspase-1-mediated release of inflammatory cytokines from dying cells (48, 49, 50). Therefore, apoptotic cells may be immunogenic in vivo, and this immunogenicity appears regulated by different factors, including the apoptotic load, the APCs involved in engulfment, processing and presentation of Ags contained in the apoptotic cells, and the cytokines released in the microenvironment (50). Apoptotic cell death also generates oxidatively modified moieties, which can induce autoimmune and local inflammatory responses by recruiting monocytes via monocyte-endothelial cell interactions (51). Apoptotic material represents a source of autoantigens, because intracellular and nuclear Ags are modified during apoptosis and translocated into the cell membrane (39). Indeed, nucleosomes are generated primarily by apoptosis and are considered one of the main autoantigens to which Abs are produced in human and murine lupus (2, 52, 53). Nucleosomes are also present in kidney eluates of lupus-prone mice with protenuria (2). Recognition of nucleosomes, with Ag spreading to other chromatin components, may be the initiating event leading to the development of other autoantibodies. Although nucleosomes appear to be ignored by immunocompetent cells under normal physiological conditions, the development of CD4+ T cells primed to nucleosomal Ags in lupus-prone mice precedes other abnormalities that develop (23, 54), and apoptotic material presented to autoreactive lymphocytes by APCs can promote autoimmunity. This may be especially relevant in settings of excessive cellular apoptosis whereby the scavenger system becomes overwhelmed.
Patients with SLE have evidence of accelerated monocyte/M apoptosis, a phenomenon that correlates with disease activity (3). Increased monocyte/M apoptosis has been considered secondary to a number of different factors. Indeed, SLE serum can induce M apoptosis and impair M clearance by uncharacterized mechanisms (4). Lupus patients also have an autoreactive CD4+ T cell subset that kills autologous monocytes in the absence of added Ag (3, 5). In this case, autoreactivity is defined by the capacity to kill autologous APCs in the absence of Ag (5). This type of autoreactivity has been described in the DNA hypomethylation model (55) and in the model of LFA-1-overexpressing lupus T cells (5). In addition, T cells observed in chronic graft-vs-host disease also cause a lupus-like disease by responding to host MHC molecules in vivo (56). Indeed, this phenomenon is MHC restricted, and it is not found in healthy controls or using allogeneic lupus monocytes (3, 5). We have reported that these autoreactive lupus T cells have increased expression of the apoptotic ligands TRAIL, TWEAK, and Fas ligand and that blockade of these apoptotic pathways significantly decreases monocyte cytotoxicity, with the residual cytotoxicity attributable to the perforin pathway (3, 57). The increased expression of these proapoptotic molecules appears to play a significant role in the induction of APC cytotoxicity in SLE, and this phenomenon might be important in inducing autoimmunity (3, 57). Indeed, it appears that cells killed by cytotoxic T cell-assisted apoptosis are preferential sources of autoantigens (58) and that autoreactive T cells play an important role in inducing this increase in the apoptotic load (3). When Th2 and Th1 cell lines are induced to overexpress the -integrin LFA-1, they become autoreactive and are cytolytic to M without Ag (59). These cells also induce anti-DNA Abs in vivo, suggesting that this killing of M could contribute to the autoantibody response by providing a source of autoantigen (21, 55, 60, 61). Furthermore, when autoreactive T cell clones are transferred to syngeneic female recipients, these cells (but not non-autoreactive T cells) preferentially localize to the red pulp M in the spleen (B. C. Richardson, unpublished observations), suggesting that autoreactive T cells interact with splenic M in vivo, inducing apoptosis. These findings are consistent with increased M apoptosis contributing to the development of a lupus-like disease in mice. Additional evidence indicates that accelerated apoptosis of both freshly isolated monocytes and monocytes maturing to M is associated with in vitro impairment in the interaction between apoptotic cells and M (42). As such, this accelerated apoptosis dramatically influences the efficiency of clearance of the remaining M in vitro. Upon restoration of M numbers, SLE M can interact with apoptotic cells similarly as cells from normal controls (42). These results also suggest that impaired apoptotic clearance in SLE may be secondary, at least in part, to an accelerated apoptosis of M that under normal circumstances would engulf apoptotic material and prevent propagation of autoimmune responses. Supporting these findings, anti-TNF therapy for patients with rheumatoid arthritis induces selective M apoptosis (62). This therapy is known to induce autoantibodies and in some cases overt SLE. Although it has been proposed that the main mechanism for this complication is promotion of Th2 responses when TNF- is blocked, an additional possibility is that increased M apoptosis contributes to autoantibody formation (62). Supporting this hypothesis, accumulation of plasma nucleosomes upon treatment with anti-TNF- Abs has been reported (63).
Despite all these studies, it has been unclear whether accelerated monocyte/M apoptosis is important in the pathogenesis of lupus or represents an epiphenomenon. We now report that accelerated M apoptosis in lupus-prone mice results in increased and sustained elevations of anti-nucleosome and anti-dsDNA Abs, hastens proteinuria onset, and increases severity of lupus nephritis. In addition, we report elevation of the autoantibody IdLNF1. This is relevant, because all of these autoantibodies have been proposed to be pathogenic in kidney damage in murine SLE (15, 23, 30, 64, 65, 66, 67, 68, 69, 70), and their significant increases in the clodronate-treated mice further supports the notion that they were involved in worsening renal disease. The mechanisms by which accelerated M apoptosis induces worsening nephritis in lupus-prone mice are not entirely clear. The most likely explanation is that increased levels of pathogenic Abs leads to increased deposition of immune complexes in the kidney. This is substantiated by the fact that nephritogenic IdLNF1 Abs were initially increased in the clodronate-treated group and then dropped below the levels seen in the nontreated mice at around the time when clinical nephritis developed, suggesting deposition in the kidneys, as described in previous work (31). It is likely that we did not detect differences in the amount of renal immune complexes because these organs were examined late in the course of the disease, when end-stage had occurred in all the lupus-prone mice and maximum immune complex deposition had been achieved. It is also possible that local effects mediated by cytokines could be involved in the increased severity of the disease and future studies will address this possibility. However, no significant changes in the serum levels of specific cytokines involved in lupus nephritis were detected in the SNF1 mice.
In non-lupus-prone mice, accelerated M apoptosis led to elevations of a variety of autoantibodies, but the kidneys of these mice displayed no evidence of autoimmunity, suggesting that the genetic background and other immunological factors might be crucial for the development of autoimmune manifestations, subsequent to accelerated M death. In addition, the levels of autoantibodies seen in the non-lupus-prone mice, albeit increased in the clodronate-treated group, might not be sufficient to promote organ damage mediated by immune complexes. Our data on non-autoimmune prone mice are consistent with what other groups have reported, where exposure to increased apoptotic cells is not a sufficient explanation for induction of high-titer, high-affinity, pathogenic autoantibodies without the right genetic background. Indeed, given our results, it is likely that additional immunoregulatory defects would be required for the full induction of systemic autoimmune disorders (35).
Accelerated M apoptosis might be particularly important in autoimmunity for a number of reasons. First, an increase in apoptotic load may overwhelm the normal mechanisms of apoptotic cell clearance. Second, accelerated M death also impairs clearance of apoptotic material. These two factors may act in concert to greatly enhance autoimmunity in SLE. For example, when levels of apoptosis exceed the clearance capacity, as is the case in SLE, the apoptotic cells are allowed to become necrotic and release intracellular components to the circulation. Previous studies have suggested that high levels of apoptotic and necrotic cells can further promote a secondary round of apoptosis, particularly in the M due to its scavenger activity, which results in the release of nuclear contents both from the engulfed cell and the affected M (8). Therefore, M apoptosis is crucial in the generation of circulating nucleosomes. One potential consequence of enhanced M apoptosis is that the immunostimulatory dendritic cell may then be permitted to take up and present autoantigenic material and initiate proinflammatory pathways. Favoring this hypothesis, histological analysis of lymph nodes in a subgroup of SLE patients showed that apoptotic cells are not properly cleared by tingible body M of the germinal centers. Consequently, nuclear autoantigens bind to follicular dendritic cells and may thus provide survival signals for autoreactive B cells (71). In addition to capturing and presenting exogenous Ags, immature as well as mature dendritic cells are capable of phagocytosing endogenous apoptotic cells and presenting self-Ags to T cells (72, 73, 74, 75), particularly when dendritic cells are exposed to high doses of apoptotic cells (76). This phenomenon could be important because dendritic cells, but not M, efficiently cross-present Ag derived from apoptotic cells to autologous CTLs and activate autoreactive T cells in the lymph node. Furthermore, an increase in proinflammatory cytokines during secondary necrosis may favor dendritic cell maturation and migration to secondary lymphatic tissues, where activation of a specific immune response would take place. Importantly, a number of previous studies have shown that clodronate liposome administration does not affect dendritic cell function or in vivo viability (13, 14, 33). Because mature dendritic cells that have migrated to the spleen are significantly less efficient at uptaking Ag, it is expected that they will be much less susceptible to damage by the clodronate. Upon Ag engulfment, M normally provide suppressor cytokine signals to nearby cells, such as TGF- and IL-10. Under specific circumstances, such as a cytokine-driven proinflammatory environment or hypothetically when M are undergoing accelerated cell death, mature dendritic cells could then gain the necessary tools to drive an autoimmune response. Furthermore, apoptotic monocytic cell lines in the presence of autoantibodies can cause production of a clearly immunostimulatory cytokine, IFN-, which among other functions can direct monocyte differentiation into dendritic cells rather than M (77, 78).
Liposomal clodronate appears to have a very selective effect because neutrophils and lymphocytes have not been found to be directly affected by the drug (79). In our study, we did not find significant changes of pro- and anti-inflammatory cytokines in blood in the SNF1 mice. This might be explained by the fact that, unlike other methods that induce M death, clodronate liposomes do not lead to the secretion of proinflammatory cytokines by the dying M (80). This might explain the lack of significant increases, and even a slight decrease, in TNF- seen in the mice that we studied. TGF- is a cytokine secreted by M that are engulfing apoptotic material (81), is up-regulated in lupus nephritis, and plays a role in end-stage renal disease in SLE (82). We did not see an increase in the levels of this cytokine in serum, probably reflecting a decrease in the numbers of M that were capable of engulfing apoptotic material.
Additional studies have shown that clodronate administration causes significant decreases of marginal zone M as well as red pulp M within 24 h, and thereafter, these areas gradually repopulate to even higher levels than before the injection (28, 34, 83, 84, 85, 86). Furthermore, tissue M in the lungs and kidneys are not depleted (33). Similarly, in our study, the absolute number of M was not reduced when mice were examined >2 wk after the last clodronate administration, but significant M apoptosis was detected when spleens were harvested 24 h after clodronate administration. Therefore, an insult that contributes to prolonged M apoptosis (such as autoreactive T cells or a cytokine) would be necessary to induce a chronic increase in apoptotic load and a decrease in clearance mechanisms. Because the bone marrow produces >10 billion monocytes/day (87, 88), it is very likely that the total M number in vivo does not diminish significantly but that the increased apoptotic load impairs the ability of new M to process apoptotic material in an efficient manner. Interestingly, previous work has also shown that M incubated with clodronate liposomes release DNA into the medium after undergoing apoptosis, increasing the apoptotic load (8).
We propose that M apoptosis contributes to the pathogenesis of autoantibody formation and organ damage. We hypothesize that this is secondary to both an increase in the apoptotic load and impairment in the clearance of apoptotic material. This study suggests that autoreactive cytotoxic T cells and other mechanisms that contribute to scavenger cell apoptosis in SLE, could play a pathogenic role and contribute to the severity of the disease.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by U.S. Public Health Service Grants AR048235 and AR48310, Anthony S. Gramer Fund in Inflammation Research, and Lupus Research Institute (all to M.J.K.). R.C. was supported by the Arthritis Foundation. E.E.L. was supported by the Lupus Foundation of America. This research is also supported (in part) by the National Institutes of Health through the University of Michigan’s Cancer Center Support Grant (P30 CA46592) and the George M. O’Brien Kidney Research Center (P50 DK039255). 
2 Current address: Office of Generic Drugs, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Rockville, MD 20857.
3 Address correspondence and reprint requests to Dr. Mariana J. Kaplan, Division of Rheumatology, University of Michigan, 5520 Medical Sciences Research Building 1, 1150 West Medical Center Drive, Ann Arbor, MI 48109-0680. E-mail address: makaplan{at}umich.edu
4 Abbreviations used in this paper: SLE, systemic lupus erythematosus; M, macrophage; PI, propidium iodide.
Received for publication May 26, 2005. Accepted for publication November 4, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
c double-knockout mice after treatment with CL2MDP liposomes. Exp. Hematol. 32: 1118-1125. [Medline]
