Alpha-1-Antitrypsin Enhances Primary Human Macrophage Immunity Against Non-tuberculous Mycobacteria

Rationale: The association between non-tuberculous mycobacterial lung disease and alpha-1-antitrypsin (AAT) deficiency is likely due, in part, to underlying emphysema or bronchiectasis. But there is increasing evidence that AAT itself enhances host immunity against microbial pathogens and thus deficiency could compromise host protection.

Objectives: The goal of this project is to determine if AAT could augment macrophage activity against non-tuberculous mycobacteria.

Methods: We compared the ability of monocyte-derived macrophages cultured in autologous plasma that were obtained immediately before and soon after AAT infusion—given to individuals with AAT deficiency—to control an ex vivo Mycobacterium intracellulare infection.

Measurements and Main Results: We found that compared to pre-AAT infused monocyte-derived macrophages plus plasma, macrophages, and contemporaneous plasma obtained after a session of AAT infusion were significantly better able to control M. intracellulare infection; the reduced bacterial burden was linked with greater phagosome-lysosome fusion and increased autophagosome formation/maturation, the latter due to AAT inhibition of both M. intracellulare–induced nuclear factor-kappa B activation and A20 expression. While there was a modest increase in apoptosis in the M. intracellulare-infected post-AAT infused macrophages and plasma, inhibiting caspase-3 in THP-1 cells, monocyte-derived macrophages, and alveolar macrophages unexpectedly reduced the M. intracellulare burden, indicating that apoptosis impairs macrophage control of M. intracellulare and that the host protective effects of AAT occurred despite inducing apoptosis.

Conclusion: AAT augments macrophage control of M. intracellulare infection through enhancing phagosome-lysosome fusion and autophagy.

Novel method for quantitative ANA measurement using near-infrared imaging

Antinuclear antibodies (ANA) have been detected in patients with systemic rheumatic diseases and are used in the screening and/or diagnosis of autoimmunity in patients as well as mouse models of systemic autoimmunity. Indirect immunofluorescence (IIF) on HEp-2 cells is the gold standard for ANA screening. However, its usefulness is limited in diagnosis, prognosis and monitoring of disease activity due to the lack of standardization in performing the technique, subjectivity in interpreting the results and the fact that it is only semi-quantitative. Various immunological techniques have been developed in an attempt to improve upon the method to quantify ANA, including enzyme-linked immunosorbent assays (ELISAs), line immunoassays (LIAs), multiplexed bead immunoassays and IIF on substrates other than HEp-2 cells. Yet IIF on HEp-2 cells remains the most common screening method for ANA. In this study, we describe a simple quantitative method to detect ANA which combines IIF on HEp-2 coated slides with analysis using a near-infrared imaging (NII) system. Using NII to determine ANA titer, 86.5% (32 of 37) of the titers for human patient samples were within 2 dilutions of those determined by IIF, which is the acceptable range for proficiency testing. Combining an initial screening for nuclear staining using microscopy with titration by NII resulted in 97.3% (36 of 37) of the titers detected to be within two dilutions of those determined by IIF. The NII method for quantitative ANA measurements using serum from both patients and mice with autoimmunity provides a fast, relatively simple, objective, sensitive and reproducible assay, which could easily be standardized for comparison between laboratories.

Immunologic response to administration of standardized dog allergen extract at differing doses

BACKGROUND

The immunologic response to immunotherapy with dog extract is not well characterized.

OBJECTIVE

The purpose of this study was to examine the immunologic response to 3 doses of dog extract expressed as their Can f 1 content.

METHODS

Cluster immunotherapy was administered to 28 patients with dog allergy who were randomly assigned to 1 of 4 treatment arms: placebo or acetone-precipitated extract containing 0.6 mug, 3.0 mug, or 15.0 mug Can f 1 per 0.5 mL maintenance dose. Studies included titrated skin prick tests, the late cutaneous response, titrated nasal challenge with dog extract, and serum allergen-specific IgE and IgG(4). Dog allergen-stimulated lymphocyte proliferation was performed with measurement of secreted cytokines by ELISA and of intracellular cytokines by flow cytometry.

RESULTS

There was a significant dose-dependent response in suppression of titrated skin prick tests and suppression of the late cutaneous response. There was a significant increase from baseline in dog-specific IgG(4) in both the high-dose and low-dose groups and a dose-dependent suppression of secreted TNF-alpha and increase in secreted TGF-beta. There was a dose-dependent trend in suppression of secreted IL-4 with a significant decrease from baseline in the high-dose group. There were no significant changes in symptom scores; lymphocyte proliferation; secreted IFN-gamma, IL-10, or IL-5; or intracellular cytokine production.

CONCLUSION

The dose-response in immunologic parameters after immunotherapy with dog extract is similar to that previously demonstrated with cat extract.

CLINICAL IMPLICATIONS

The greatest and most consistent response is seen with a dose containing 15 mug Can f 1.

Binding of rat and human surfactant proteins A and D to Aspergillus fumigatus conidia

Surfactant proteins A (SP-A) and D (SP-D) are thought to play important roles in pulmonary host defense. We investigated the interactions of rat and human SP-A and SP-D with Aspergillus fumigatus conidia. Rat SP-D but not rat SP-A bound the conidia, and the binding was inhibited by EDTA, mannose, glucose, maltose, and inositol. Binding studies using a mutant recombinant rat SP-D with altered carbohydrate recognition but normal structural organization clearly established a role for the carbohydrate recognition domain in binding to conidia. However, neither rat SP-A nor SP-D increased the association of fluorescein isothiocyanate-labeled conidia with rat alveolar macrophages as determined by flow cytometry. Both human SP-A (isolated from normal and alveolar proteinosis lungs) and SP-D (recombinant protein and protein isolated from alveolar proteinosis lungs) bound the conidia. These data indicate that important differences exist between rat and human SP-A in binding to certain fungi. Human SP-A and SP-D binding to conidia was also examined in the presence of hydrophobic surfactant components (HSC), containing both the phospholipid and hydrophobic proteins of surfactant. We found that HSC inhibited but did not eliminate human SP-A binding to Aspergillus conidia. In contrast, the SP-D binding to conidia was unaffected by HSC. These findings indicate that SP-D plays a major role in the recognition of Aspergillus conidia in alveolar fluid.

Staphylococcus aureus isolated from Kawasaki disease patients hyper-releases extracellular protein A

S. aureus isolates from patients with Kawasaki disease (KD) release high levels of extracellular protein A (SpA), as compared to S. aureus in other diseases. The molecular weight of this released protein A is about 70 kDa. Extracellular KD SpA purified by affinity chromatography possessed the same amino acid sequence at the NH2-terminal IgG binding region and the same antigenic specificity as recombinant and cell-wall-bound SpA preparations. The size of DNA fragments containing the spa gene from S. aureus KD strains was 160-165 kb. All of these DNA fragments contained the igb portion encoding the IgG-binding region of KD SpA. Significantly higher molecular size of the SpA molecules hyper-released in the stationary-phase culture and the lack of production of other exo-proteins allow us to speculate that S. aureus isolated from patients with KD have mutations occurring in the agr locus.

Prediction of early-onset asthma in genetically at-risk children

The W.T. Grant Foundation Asthma Risk Study was designed to prospectively examine children who were considered at a genetically increased risk for the development of asthma. The respective contributions of 11 potential risk factors, both environmental and biological, were assessed in order to determine their relative roles in affecting the early onset of asthma. This is a report of an inception cohort of children born to asthmatic mothers and followed for a 3-year period. All 150 families were recruited from the general community and living within 2 h of the National Jewish Center for Immunology and Respiratory Medicine (Denver, CO). Mothers in the index risk sample had been previously diagnosed with asthma and were recruited during their pregnancy through physician referrals and media solicitation. The index sample of 150 families was 92% Caucasian and predominantly middle class. The mean age of mothers was 29.3 years, and of fathers, 31.1 years. The main outcome was the determination of the early onset of asthma and its association with quantified risk factors. By age 3 years, 14 of the 150 children had developed asthma. Frequent illness, IgE levels at age 6 months, parenting difficulties, and early eczema were significantly associated with the onset of asthma (P = 0.003, P = 0.006, P = 0.01, and P = 0.03, respectively). Only frequent illness, elevated serum IgE levels, and parenting difficulties entered a predictive model where they were independently related to the development of asthma.

Early preferential stimulation of gamma delta T cells by TNF-alpha

Although recent findings indicate that gamma delta T cells influence both early innate and Ag-specific adaptive host responses, it has remained unclear what triggers gamma delta T cell reactivity. Investigating very early T cell activation in mouse and human models of bacterial infection, we measured CD69 expression as an indicator of early cellular activation. Both murine alpha beta and gamma delta T cells responded polyclonally to systemic bacterial infections, and to LPS. However, gamma delta T cells responded more strongly to the bacteria and to LPS. In vitro LPS-stimulated human T cells showed a similar differential response pattern. We identified TNF-alpha as mediator of the early differential T cell activation, and of differential proliferative responses. The stronger response of gamma delta T cells to TNF-alpha was correlated with higher inducible expression levels of TNF-Rp75. Among unstimulated splenocytes, more gamma delta T cells than alpha beta T cells expressed CD44 at high levels. The data suggest that TNF-Rp75 determines the differential T cell reactivity, and that most gamma delta T cells in the normal spleen are present in a presensitized state. As TNF-alpha stimulates activated T cells, it may early preferentially connect gamma delta T cell functions with those of cells that produce this cytokine, including activated innate effector cells and Ag-stimulated T lymphocytes.

Early Preferential Stimulation of γδ T Cells by TNF-α

Although recent findings indicate that γδ T cells influence both early innate and Ag-specific adaptive host responses, it has remained unclear what triggers γδ T cell reactivity. Investigating very early T cell activation in mouse and human models of bacterial infection, we measured CD69 expression as an indicator of early cellular activation. Both murine αβ and γδ T cells responded polyclonally to systemic bacterial infections, and to LPS. However, γδ T cells responded more strongly to the bacteria and to LPS. In vitro LPS-stimulated human T cells showed a similar differential response pattern. We identified TNF-α as mediator of the early differential T cell activation, and of differential proliferative responses. The stronger response of γδ T cells to TNF-α was correlated with higher inducible expression levels of TNF-Rp75. Among unstimulated splenocytes, more γδ T cells than αβ T cells expressed CD44 at high levels. The data suggest that TNF-Rp75 determines the differential T cell reactivity, and that most γδ T cells in the normal spleen are present in a presensitized state. As TNF-α stimulates activated T cells, it may early preferentially connect γδ T cell functions with those of cells that produce this cytokine, including activated innate effector cells and Ag-stimulated T lymphocytes.

Comparison of exhaled nitric oxide, serum eosinophilic cationic protein, and soluble interleukin-2 receptor in exacerbations of pediatric asthma

The hypotheses tested in this study were that during acute asthma exacerbations (1) exhaled nitric oxide concentrations [eNO] are a more sensitive, noninvasive indicator of asthma disease activity than serum markers of inflammation such as eosinophil cationic protein (ECP) or soluble interleukin 2 receptor (sIL2R), and (2) elevated [eNO] are reduced after treatment with glucocorticoids (GC). Peak eNO levels were measured by chemiluminescence during slow expiration. Seven asthmatic subjects (mean age 11 yrs; mean morning FEV1 65% predicted) receiving inhaled GC, and with no radiographic evidence of acute sinusitis, were studied before and after a course of oral GC. Measurements of [eNO], ECP and sIL2R levels, and FEV1% were obtained before and after a course of GC. Six atopic nonasthmatic subjects (mean age 12 years; mean FEV1 94% predicted) and seven normal subjects (mean age 13 years; mean FEV1 100% predicted) were studied. The mean peak [eNO] level (parts per billion: ppb) for the asthma subjects before treatment (52 +/- 5 ppb SEM) was greater than the value for both nonasthmatic atopic and normal subjects (16 +/- 2 ppb and 14 +/- 2 ppb SEM, respectively; P < 0.0001). There was no significant difference in ECP or sIL2R values between asthmatic subjects and either atopic or normal subjects (P > 0.05). Baseline pre-GC treatment ECP levels in the asthmatic subjects were significantly higher (P < 0.002) than post-GC treatment values. The mean peak [eNO] level in the asthmatic subjects declined after oral GC treatment to 14 +/- 1 ppb (P < 0.0002) and was less than 2 ppb different from either control group (P > 0.75). We conclude that [eNO] is a more sensitive marker of asthma disease activity than ECP and sIL2R levels. In addition, [eNO] appears to be a more useful indicator of the beneficial response to GC therapy than these other measurements in pediatric asthma.

The potential role of gastroesophageal reflux in the pathogenesis of food-induced wheezing

Severe reactive airways disease (RAD) in children is frequently associated with gastroesophageal reflux or food allergy. However a relationship between these two confounding factors has yet to be investigated. We postulate that, in certain patients with micro-aspiration of gastric contents into the airways, food allergens sensitize T cells in the peribronchial lymphoid tissue and induce the production of food-specific IgE antibodies that sensitize airway cells. Subsequent exposure to these food allergens might then induce IgE dependent mediator release from mast cells as well as T cell and eosinophil activation, thus contributing to airway inflammation and RAD. In the current report, we describe the case of a patient with severe asthma who had food allergy and gastroesophageal reflux whose clinical findings support this hypothesis. We also provide additional evidence for a high rate of food sensitization in patients with bronchopulmonary dysplasia (BPD), RAD and GER. We conclude that additional studies are warranted to examine the possibility that patients who have RAD and GER require an evaluation for food allergy.

The production of interferon-gamma in response to a major peanut allergy, Ara h II correlates with serum levels of IgE anti-Ara h II

The current study was undertaken to examine the potential role of T cells in the pathogenesis of peanut allergy. Peripheral blood mononuclear cells (PBMCs) from patients with peanut allergy, patients with asthma, and nonatopic normal control subjects were assessed for proliferation after stimulation with a 17 kd major peanut allergen (Ara h II), ovalbumin, casein, soy, and Candida albicans. We found that Ara h II and C. albicans induced significantly higher levels of proliferation than ovalbumin, casein, and soy. Because interferon-gamma (IFN-gamma) and interleukin-4 (IL-4) play critical roles in IgE regulation, we assessed the production of these cytokines after stimulation with C. albicans and Ara h II. C. albicans stimulated similar levels of IFN-gamma in all three study groups. In contrast, after stimulation with Ara h II, culture supernatants from PBMCs of subjects with peanut allergy contained significantly lower levels of IFN-gamma than did the PBMCs of the two control groups (p = 0.02). More important, there was a significant (p = 0.05) inverse correlation between the serum IgE anti-Ara h II levels and IFN-gamma production by PBMCs from the respective peanut-allergic patients. IL-4 protein was not detected in culture supernatants of PBMCs stimulated with Ara h II. However, amplification of cytokine gene transcripts by polymerase chain reaction did demonstrate IL-4 expression in Ara h II-stimulated PBMCs from both patients with peanut allergy and control subjects. These data suggest that the level of IFN-gamma production in response to Ara h II may be an important factor in determining the development of peanut-specific IgE responses.

Presence of IgE antibodies to staphylococcal exotoxins on the skin of patients with atopic dermatitis. Evidence for a new group of allergens

In the current study, we investigated whether Staphylococcus aureus grown from affected skin of atopic dermatitis (AD) patients secreted identifiable toxins that could act as allergens to induce IgE-mediated basophil histamine release. The secreted toxins of S. aureus grown from AD patients were identified by ELISA using antibodies specific for staphylococcal enterotoxin (SE) exfoliative toxin (ET), or toxic shock syndrome toxin (TSST-1). S. aureus isolates from 24 of 42 AD patients secreted identifiable toxins with SEA, SEB, and TSST accounting for 92% of the isolates. 32 of 56 AD sera (57%) tested contained significant levels of IgE primarily to SEA, SEB, and/or TSST. In contrast, although SEA, SEB, or TSST secreting S. aureus could be recovered from the skin of psoriasis patients, their sera did not contain IgE antitoxins. Freshly isolated basophils from 10 AD patients released 5-59% of total histamine in response to SEA, SEB, or TSST-1 but only with toxins to which patients had specific IgE. Basophils from eight other AD patients and six normal controls who had no IgE antitoxin failed to demonstrate toxin-induced basophil histamine release. Stripped basophils sensitized with three AD sera containing IgE to toxin released 15-41% of total basophil histamine only when exposed to the relevant toxin, but not to other toxins. Sensitization of basophils with AD sera lacking IgE antitoxin did not result in release of histamine to any of the toxins tested. These data indicate that a subset of patients with AD mount an IgE response to SEs that can be grown from their skin. These toxins may exacerbate AD by activating mast cells, basophils, and/or other Fc epsilon-receptor bearing cells armed with the relevant IgE antitoxin.

Alpha beta T-lymphocyte depleted mice, a model for gamma delta T-lymphocyte functional studies

Adult mice can be depleted of essentially all mature alpha beta T lymphocytes by chronic treatment with the framework-recognizing, pan-specific anti-TCR alpha beta mAb, H57-597. Similar findings have been reported in rats, gamma delta cell populations remain essentially unaltered in size and reactivity. Suppression of alpha beta T-cell development results in the loss of alloantigen reactivity and of B-cell help, suggesting that gamma delta and alpha beta populations differ in their functional capabilities. Indirect effects of the antibody treatment include quantitative changes in splenic B cells, as well as reduced sizes and weights of experimental animals. alpha beta-suppressed mice and rats may provide model systems for studies on gamma delta cell function in vivo.

Possible links between immune system and stress response: the role of gamma delta T lymphocytes

Heterologous heat shock proteins (hsps) are antigens in many infectious diseases involving bacteria, parasites and perhaps even fungi. T and B lymphocytes recognize multiple epitopes on these proteins. Recently, lymphocytes expressing gamma delta T cell receptors (gamma delta cells) were also found to react with hsps that are members of the 60 kiloDalton (kDa) family. The response of gamma delta cells to hsp 60 differs from that of alpha beta T cells and B cells in several ways: the frequency of reactive gamma delta clones is high. Many clones recognize the same portion of this protein instead of scattered antigenic epitopes; and most gamma delta cells that react with the mycobacterial hsp-60 homolog are also stimulated by the autologous homolog. Perhaps, such (self) hsp-reactive gamma delta populations function by distinguishing stressed from not stressed states in autologous cells and tissues, rather than by discriminating 'self' and 'non-self'.

Elastic properties of the excised lungs of NZB/W mice and their correlation with histopathologic changes

Lung volume and the pressure-volume (PV) relationships of the lung were determined in excised lungs of an animal model of human systemic lupus erythematosus, which develops in NZB/W mice, and correlated with histopathologic changes. In young, 4-month-old NZB/W mice, and in nonautoimmune BALB/c mice at any age, there were minimal histologic changes or alterations in lung function. The NZB/W mice developed histologic changes with aging. Perivascular, peribronchial, and interstitial changes developed and were associated with a significant reduction in total lung capacity as well as changes in the pressure-volume characteristics of the lung, which are compatible with a restrictive process.

Specific Removal of Bovine Serum Albumin (BSA) Antibodies in Vivo by Extracorporeal Circulation over BSA Immobilized on Nylon Microcapsules

Studies were undertaken to determine whether BSA, once immobilized on activated nylon microspheres, would be capable of specifically removing circulating BSA antibody in vitro and in vivo in an extracorporeal circulation system in dogs. Nylon microspheres were prepared and, after gentle hydrolysis and glutaraldehyde treatment, demonstrated a retention of up to 34.5 mg of BSA. In vitro studies showed that BSA immobilized on microspheres removed a significant percentage of BSA-binding activity. For in vivo studies, an extracorporeal circulation system was established and mongrel dogs were then injected with anti-BSA and anti-HSA antibodies. After an equilibration period, BSA microspheres were introduced into the extracorporeal circulation system. After the insertion of BSA microcapsules, serum exhibited a sharp reduction in BSA binding over the next 15 min, with a more gradual diminution over the ensuing 60 to 90 min. There was no significant reduction in anti-HSA binding over the same time frame. This effect could not be attributed to release of BSA from the microspheres since no 125I-BSA was detected in the serum or organs of the dogs at the conclusion of the experiments. After extracorporeal circulation over nylon microspheres, there were only minimal changes in the host's hematocrit or leukocyte counts and no significant thrombotic material or cellular debris was recoverable on the capsules. These data suggest that antigen immobilized on nylon microspheres may specifically withdraw circulating antibodies in vivo with minimal release of its antigenic material and little alteration in the host's hematologic status.

Specific removal of bovine serum albumin (BSA) antibodies in vivo by extracorporeal circulation over BSA immobilized on nylon microcapsules

Studies were undertaken to determine whether BSA, once immobilized on activated nylon microspheres, would be capable of specifically removing circulating BSA antibody in vitro and in vivo in an extracorporeal circulation system in dogs. Nylon microspheres were prepared and, after gentle hydrolysis and glutaraldehyde treatment, demonstrated a retention of up to 34.5 mg of BSA. In vitro studies showed that BSA immobilized on microspheres removed a significant percentage of BSA-binding activity. For in vivo studies, an extracorporeal circulation system was established and mongrel dogs were then injected with anti-BSA and anti-HSA antibodies. After an equilibration period, BSA microspheres were introduced into the extracorporeal circulation system. After the insertion of BSA microcapsules, serum exhibited a sharp reduction in BSA binding over the next 15 min, with a more gradual diminution over the ensuing 60 to 90 min. There was no significant reduction in anti-HSA binding over the same time frame. This effect could not be attributed to release of BSA from the microspheres since no 125I-BSA was detected in the serum or organs of the dogs at the conclusion of the experiments. After extracorporeal circulation over nylon microspheres, there were only minimal changes in the host's hematocrit or leukocyte counts and no significant thrombotic material or cellular debris was recoverable on the capsules. These data suggest that antigen immobilized on nylon microspheres may specifically withdraw circulating antibodies in vivo with minimal release of its antigenic material and little alteration in the host's hematologic status.

Degradation of circulating DNA by extracorporeal circulation over nuclease immobilized on nylon microcapsules

Studies were undertaken to determine whether deoxyribonuclease I, (DNase I) once immobilized on activated nylon microspheres, would be capable of degrading circulating DNA in vitro and in vivo in an extracorporeal circulation system in dogs. Nylon microspheres were prepared and after gentle hydrolysis and glutaraldehyde treatment, demonstrated a retention of up to 4.73 mg of Dnase I. In vitro studies showed that DNase I immobilized on microspheres degreded a significant percentage of 125I-native DNA (nDNA) within 15 min. Mongrel dogs were injected with 125I-nDNA and a variation in initial t 1/2 in individual animals was observed. Therefore, for experimental studies, 125I-nDNA was injected and decay was recorded during a control period in which untreated microcapsules were utilized in the extracorporeal system. DNase I microspheres were then introduced into the extracorporeal circuit which resulted in an acceleration of degradation of acid precipitable 125I-nDNA. When 200 mug of unlabeled DNA with 125I-nDNA was injected, a similar augmentation of DNA degradation was noted after extracorporeal circulation over DNase I microcapsules. This effect could not be attributed to release of DNase I from the microspheres since no 131I-DNase was detected in the serum or organs of the dogs at the conclusion of the experiments. 125I-nDNA:anti-DNA complexes were passively injected into dogs and after a similar control period of circulation over untreated microcapsules. DNase I microspheres were introduced. Results showed a rapid acceleration in the degradation rate of 125I-nDNA:anti-DNA complexes precipitable with (NH4)2SO4. Extracorporeal circulation over nylon microspheres resulted in no significant alteration of the host's hematocrit or platelet count, and little residual cellular debris on the microcapsules. These data suggest that DNAase immobilized on nylon microspheres may have a potential role in the specific therapy of systemic lupus erythematosus, when it is desirable to hydrolyze DNA circulating free or in combination with antibody.

Specific removal of DNA antibodies in vivo with an extracorporeal immuno-adsorbent

The selective removal of circulating antibody specific for DNA was affected with an immuno-absorbent consisting of DNA-cellulose incorporated into agar gel. Antisera to DNA obtained from patient with systemic lupus erythematosus was circulated in vitro through experimental and control columns by a dual channel haemodialysis pump and serial aliquots were withdrawn and assayed for antibody. A 65% reduction in DNA binding of serum was achieved at a flow rate of 210 ml/min over a 4-hr period with no release of 125I-labelled DNA from the column into the serum. For in vivo studies, 2-6-3-8 kg rabbits were actively immunized with methylated bovine serum albumin conjugated to single-stranded DNA (MBSA-ssDNA). Whole blood was pumped from the femoral artery through an immuno-adsorbent composed of ssDNA-cellulose in an agar matrix. Results showed significant reductions in ssDNA binding activity over various periods after connection of the rabbit's circulation to the immuno-adsorbent with only minimal changes in BSA binding during the same period. Little release of incorporated 125I-labelled DNA from the column as assayed in the blood and tissues of the experimental animals occurred during the procedure. The immuno-adsorbent columns showed no residual cellular debris or thrombotic material. These results suggest that this immuno-adsorbent system may be used to specifically withdraw circulating DNA antibodies in vivo. Such a system may have clinical potential for specific therapy of systemic lupus erythematosus.