Experimental elimination of tumor necrosis factor in low-dose endotoxin models has variable effects on survival.

Tumor necrosis factor alpha (TNF) is thought to play a major role in the pathogenesis of septic shock. Anti-TNF antibody was preadministered in low-dose endotoxin lethality models in which BALB/c mice were challenged with small amounts of lipopolysaccharide following their sensitization with either carrageenan (CAR) or D-galactosamine (D-GalN). Although the antibody virtually eliminated circulating TNF in both the CAR and the D-GalN models, only the D-GalN model mice were afforded survival, adding to a growing body of evidence that substances other than TNF play a key role in endotoxin-induced lethality. Further examination of sera from these mice showed a much greater elevation of interleukin-6 levels in the CAR-sensitized group than in the D-GalN-sensitized group.

Incorporation of bacterial lipopolysaccharide by human Leu-11a+ natural killer cells. Ultrastructural and functional correlations.

Lipopolysaccharides (LPS) of gram-negative bacteria are known to augment the ability of macrophages and natural killer (NK) cells to lyse susceptible target cells. In the present studies, we sought correlations between the ultrastructural changes and function of Leu-11a+ cells from the peripheral blood mononuclear cells which occurred as a result of their incorporation of LPS. We also studied the effect of LPS on the NK activity of purified Leu-11a+ cells. LPS samples from E. coli and Pseudomonas aeruginosa were utilized. A double-immunolabeling technique employing immunogold and immunoperoxidase markers was used to identify the Leu-11a+ cells with incorporated LPS. Pinocytosis, phagocytosis, and direct insertion into membrane bilayers appeared to be the routes for incorporation of LPS by Leu-11a+ cells. Subsequent ultrastructural effects included dilation of intracellular membrane compartments, formation of tubuloreticular inclusions and increased acid phosphatase activity. Opsonized Staphylococcus aureus were ingested both by control and LPS-treated Leu-11a+ cells; however, the number of Leu-11a+ cells with bacteria increased significantly as a result of LPS treatment. Autoradiography combined with immunolabeling showed no [3H] thymidine incorporation by either control or LPS-stimulated Leu-11a+ cells. NK cell-mediated cytotoxicity was significantly increased as compared with the control (p less than or equal to 0.02) when isolated Leu-11a+ cells were treated with LPS for 24 hours. This result suggests that LPS may have direct effect on the NK cell activity. Tests of peripheral blood mononuclear cell samples after LPS treatment showed up to a 2-fold increase in NK cytotoxicity and a dose-related increase of in vitro interferon production. This latter finding together with the ultrastructural observations of tubuloreticular inclusions in Leu-11a+ cells suggests that, in addition to any direct effect of LPS, cytotoxic activity could be indirectly augmented as a result of autocrine or paracrine interferon, or lymphokine production.

Immunoultrastructural studies of human NK cells: I. Ultracytochemistry and comparison with T cell subsets.

Human NK activity is known to be associated with a population of large granular lymphocytes (LGL) exhibiting several immunophenotypic surface markers including Leu-11a (NKP-15), Leu-7 (HNK-1), Leu-3a (T4), and Leu-2a (T8). Based upon correlation with cytolytic activity, Leu-11a is now considered the most specific antigenic marker for human NK cells. Present investigation compared the ultrastructure of cells expressing Leu-11a, Leu-7, Leu-3a, and Leu-2a, both in human peripheral blood lymphocytes (PBL) and the purified LGL fraction. Subcellular cytochemical reactions were investigated in Leu-7+ or Leu-11a+ PBL or LGL and in cells conjugated with K562 targets (indicating NK cytolytic potential). The surface markers, localized with monoclonal antibodies, were detected by immunoelectron microscopy by using direct or indirect avidin-biotin-peroxidase (ABC) or colloidal gold methods. A peroxidase-colloidal gold double-labeling system was used to identify subsets of Leu-7+ or Leu-11a+ cells. Previously described ultrastructural features of LGL including a villous surface, reniform nuclei, low nuclear/cytoplasm ratios, and abundant cytoplasm with vesicles, vacuoles, electron-dense granules, parallel tubular arrays (PTA), or paracrystalline inclusions were associated with Leu-7+, Leu-11a+, Leu-7+/Leu-11a+, Leu-7+/Leu-11a-, and Leu-7-/Leu-11a+ PBL or LGL. Results showed that the Leu-7+/Leu-11a+ cells were the most abundant NK cells in PBL. Lymphocyte subsets with Leu-3a or Leu-2a surface marker showed some ultrastructural features including PTA similar to Leu-7+ cells and Leu-11a+ cells, and their subsets. These T-cells appeared ultrastructurally more similar to the Leu-7+/Leu-11a- subset. Cytochemical studies showed that electron-dense cytoplasmic granules and PTA typical of the Leu-11a+ cells and Leu-7+ cells contained glycoprotein, acid phosphatase, and arylsulfatase. Large cytoplasmic vacuoles were heterogeneous and typically contained electron-dense material with DAB reactivity, membranous material, PTA, and/or paracrystalline inclusions. Glycoprotein, acid phosphatase, and arylsulfatase, and peroxidase reactive material were also found in these vacuoles. These features suggested that the vacuoles could be secondary lysosomes. The coexistence of intact PTA or degenerating PTA in the same vacuoles with paracrystalline inclusions suggested that the latter are possibly derived from PTA.

Immunoultrastructural studies of human NK cells: II. Effector-target cell binding and phagocytosis.

The binding of NK cells to a target cell appears to be a necessary step for NK cell-mediated cytolysis. In this report, we demonstrated effector-target binding by immunoelectron microscopy by using monoclonal antibodies against NK cells (Leu-7, Leu-11a) and T-cell subsets (Leu-2a/T8, Leu-3a/T4). The surfaces of NK and K562 cells were characterized by antitransferrin receptor antibody and various lectins. In addition, the controversial phagocytic activity of NK cells was studied by incubation of peripheral blood mononuclear cells with opsonized Staphylococcus aureus and labeling with anti-Leu-7 or anti-Leu-11a antibody. Results showed that only Leu-11a+ cells displayed a broad cell-to-cell contact with the target by a shallow intercellular interdigitation of cytoplasmic projections, while Leu-7+, Leu-2a+, or Leu3a+ cells showed only a partial contact with target without interdigitation. The Leu-11a+ cells were frequently observed in small clusters and in close association with monocytes. Cluster formation and association with monocytes were not observed in other NK and T-cell immunophenotypes. In Leu-11a+ cells conjugated with target cells, membrane-bound granules, small vesicles, parallel tubular arrays, Golgi apparatus, endoplasmic reticulum, and small vacuoles were evident and concentrated toward the target. The surface of NK cells was intensely stained for glycoprotein by chromic acid-phosphotungstic acid, whereas target cells were not stained. Transferrin receptors were stained only on the surface of target cells. Only the lectins RCA and UEA labeled the surfaces of both NK and target cells. Phagocytic vacuoles containing cell debris or fragments and ingested bacteria were found in the cytoplasm of Leu-11a+ cells but not in Leu-7+ cells. NK cells were also found within the cytoplasm of K562 target cells. All these findings suggest that Leu-11a+ cells are the true functional NK cells involved in NK cell-mediated cytolysis, phagocytosis, and emperipolesis. Therefore, the NK cell is probably "a phagocyte in lymphocyte's clothing." The presence of peroxidase in the small vesicles of NK cells and endocytotic vesicles of target cells at the effector-target contact area indicates that cytolytic enzymes or factors derived from NK cells may be transported into the target by endocytosis.

Evidence that Lyb-5 is a differentiation antigen in normal and xid mice.

These studies addressed the question of whether the Lyb-5 determinant on mouse B cells might represent a marker for a unique lineage of B cells as has been thought, or rather, a differentiation antigen. Previous studies have addressed this question by treating splenic B cells with anti-Lyb-5 + complement and then culturing them with various antigens. In this study, we first immunized in vivo or in vitro and then determined susceptibility to anti-Lyb-5. These studies demonstrate that a substantial proportion of antibody-forming cells produced in response to immunization with sheep red blood cells, trinitrophenyl (TNP)-keyhole limpet hemocyanin, TNP-Brucella abortus, TNP-lipopolysaccharide, as well as TNP-Ficoll, are eliminated by anti-Lyb-5 + complement treatment at the end of culture. Some generation of Lyb-5+ antibody-forming cells occurred in the last 24 hr of culture. These studies suggest that the Lyb-5 marker represents a differentiation antigen on relatively mature B cells.

Immunoelectron microscopic identification of human NK cells by FITC-conjugated anti-Leu-11a and biotinylated anti-Leu-7 antibodies.

Human natural killer (NK) cells have been reported to express various surface antigens. The majority and the most functionally potent NK cells are Leu-11a (NKP-15) positive cells. Only a small number of functional NK cells express Leu-7 (HNK-1) antigen. In the present study, we have established techniques for immunoelectron microscopic identification of NK cells by mouse monoclonal FITC-conjugated anti-Leu-11a and biotinylated anti-Leu-7 antibodies. Ficoll-Hypaque-isolated peripheral blood lymphocytes (PBL) were reacted with the specific antibodies before or after fixation in a 1% glutaraldehyde/1% paraformaldehyde fixative. Prefixation labeling of viable cells with the antibodies was carried out at 4 degrees C or 37 degrees C. Cells prelabeled with anti-Leu-11a antibody were reacted with secondary antibodies either before or after fixation. Anti-Leu-7 antibody was stained directly via an avidin-biotin-peroxidase (ABC) system, anti-Leu-11a antibody was stained indirectly by the ABC immunoperoxidase procedure via a biotinylated anti-mouse IgG secondary antibody or by a 10 nm or 40 nm colloidal gold-labeled anti-mouse IgG antibody. Results indicate that Leu-7 antigen could be localized by incubation with the specific antibody either before or after 20 min fixation; however, Leu-11a antigen was totally abrogated following the same fixation procedure. The Leu-11a antigen was well stained by the methods of prefixation labeling of cells with anti-Leu-11a antibody and incubation with a biotinylated secondary antibody and the ABC system after fixation. With respect to colloidal gold labeling, better results were obtained when cells were reacted with the gold-labeled antibodies immediately after incubation with anti-Leu-11a antibody but before fixation. Ultrastructurally both Leu-7 positive (+) and Leu-11a positive (+) cells shared common ultrastructural features associated with large granular lymphocytes. Using the above described techniques, we found approximately 2-5% Leu-7+ and 9-15% Leu-11a+ cells in the PBL of healthy donors. The overall results suggest that Leu-11a antigen is more sensitive to glutaraldehyde/paraformaldehyde fixation than Leu-7, since it can be localized only by prefixation labeling procedures; the ABC immunoperoxidase procedure is an ideal technique for labeling NK cells for light and electron microscopic enumeration; the immunogold method provides an adequate technique for labeling NK cells which are designated for ultracytochemical studies.

Light and electron microscopic alterations in spinal cord myelin sheaths after decompression sickness.

Pathological examination of spinal cords from animals subjected to experimental decompression sickness (DCS) was undertaken in an attempt to explain the disparate response to treatment observed. Eight experimental animals, four undived control animals, and two dived but untreated animals were perfusion fixed, and the spinal cords were removed. Light microscopy of toluidine blue stained, ultrathin sections from dived animals demonstrated a distinctive widened myelin sheath showing a banded pattern of myelin disruption. This pattern was confirmed by electron microscopy and showed the separation to be between abutting double layers of myelin. Artifactual changes were also present in dived and undived animals. These previously unreported changes may be caused by DCS. They are compatible with the major mechanisms proposed in the pathophysiology of spinal cord DCS and may also account for the response to treatment seen in our experimental animals. It is suggested that these findings may also explain the response to treatment seen in patients, together with the formation of late lesions described in the spinal cords of long-term survivors of DCS.

The role of complement receptor positive and complement receptor negative B cells in the primary and secondary immune response to thymus independent type 2 and thymus dependent antigens.

Both complement receptor positive (CR+) and complement receptor negative (CR-) B cells have been shown to be involved in the primary immune response to PC-Hy (phosphocholine conjugated hemocyanin), a thymus dependent (TD) antigen which preferentially induces antibody secretion in Lyb-5+ B cells during a primary adoptive transfer assay. CR+ and CR- B cells also responded in a primary adoptive transfer assay to TNP-Ficoll, a thymus independent type 2 (TI-2) antigen which activates only Lyb-5+ B cells. When the secondary immune response to PC-Hy and TNP-Ficoll were analyzed, it was found that most of the immune memory to both antigens was present in the CR- B cell subset. The CR- B cell subset also dominated the secondary immune response to PC-Hy in immune defective (CBA/N X DBA/2N)F1 male mice. These data indicate that CR- B cells dominate the memory response in both the Lyb-5+ and Lyb-5- B cell subsets of normal and xid immune defective mice and suggest that Lyb-5+ and Lyb-5- B cells can be subdivided into CR+ and CR- subsets.

Analysis of B-cell subpopulations. I. Relationships among splenic xid, Ly 1+, and Lyb 5+ B cells.

The phenotypic and functional relationships among the various B-cell subsets is of importance for better understanding studies of the immune system. In this report, the realm of B cells encompassed by Lyb 5, Ly 1, and xid has been examined through the use of the alloantiserum anti-Lyb 5, the unique functional properties of Ly 1+ B cells, and the spontaneous autoantibody producing congenic xid mice, NZB.xid. By functional and phenotypic analysis, we have shown that (i) Lyb 5- B cells and B cells of xid mice are largely, but not completely, overlapping; (ii) xid spleen cells contain a population which is Lyb 5+; (iii) Ly 1+ B cells fall largely in the Lyb 5+ compartment; and (iv) the autoantibody-producing Ly 1+ B cells are predominantly Lyb 5+.

Expression of Lyb-5 and Mls determinants by Peyer's patch B lymphocytes of X-linked immunodeficient mice.

Treatment of the Peyer's patch (PP) B cells from X-linked immunodeficient (xid) (CBA/N X DBA/2)F1 male mice with anti-Lyb-5 plus complement kills approximately 75% of these cells, although xid spleen B cells are unaffected. Cytotoxic depletion of Lyb-5+ cells renders the xid PP B cell population unable to generate an in vitro direct plaque-forming cell response to the TI-2 antigen TNP-Ficoll. In addition, the B cell-enriched population from the PP of xid (CBA/N X CBA/J)F1 male mice were strong stimulators of proliferation in an M1s-defined MLR, thereby demonstrating that they also express the M1s antigen(s). Two cell surface antigens associated with mature B cells are therefore expressed by xid PP B cells, suggesting that the TNP-Ficoll responsive cells in this population are mature B cell.

Morphological and physiological responses of the lungs of dogs to acute decompression.

The lung's response to decompression was studied in dogs anesthetized with pentobarbital sodium. Arterial pressure, hematocrit, right ventricular pressure, left ventricular end-diastolic pressure (LVEDP), dynamic compliance (CL), pulmonary resistance (RL), and arterial PO2, PCO2, and pH were measured prior to and for 3 h after a simulated air dive to 300 feet of seawater. Bronchoscopy was performed predive and at 3 h postdive. At 3 h animals were killed, and sections of lung were excised for histological examination. The decompression profile used regularly produced pulmonary hypertension, systemic hypotension, hemoconcentration, and arterial hypoxemia. CL fell in all but one dived animal. RL was more variable but remained unchanged postdive in most animals. The decompression stress did not alter the bronchoscopic and histological appearance of the airway mucosa. Pulmonary edema was regularly observed in histological sections and occurred without elevations of LVEDP. We concluded that noncardiac pulmonary edema is the principal response of the lung to decompression stress.

Effect of intravenous lidocaine on experimental spinal cord injury.

A series of experiments was conducted to study the effect of systemic intravenous administration of lidocaine on neurological recovery after acute experimental spinal cord injury in cats. The spinal cord was injured by the rapid inflation of an epidural balloon at T-6. The physiological integrity of the spinal cord ceased within 2 seconds in all animals, as demonstrated by acute disappearance of the somatosensory evoked response (SER). There was essentially no return of the SER in the five untreated animals when monitored for 4 hours post-injury. All of the pathological specimens from these animals revealed severe central cord hemorrhage. Intravenous lidocaine was begun 15 minutes after the injury in five animals. Three of these animals had significant return of the SER. The pathological specimens from the lidocaine-treated animals revealed either mild or moderate central cord hemorrhage. The results of this experiment suggest that systemic lidocaine administration has a significant beneficial effect in the treatment of acute spinal cord injury.

Immunoelectron microscopic application of monoclonal antibodies for identification of lymphocyte subsets bearing tubuloreticular inclusions or parallel tubular arrays.

Reactions of mouse monoclonal antibodies with T-cell, B-cell, or NK cell antigens were localized for electron microscopy by the formation of avidin-biotin-peroxidase complexes using direct or indirect labeling techniques. The techniques proved advantageous for identification of lymphocyte subsets bearing two different types of cytoplasmic structures: tubuloreticular inclusions (TRI) which have been related to interferon treatment or to diseases involving systemic immune dysfunctions, and parallel tubular arrays (PTA) which may be normal structures. Mononuclear cell samples were isolated from the peripheral blood of patients with systematic lupus erythematosus (SLE), acquired immunodeficiency syndrome (AIDS), or chronic hepatitis B (CHB). Results depended upon the fixation, the type of specific antibody, and the concentrations employed. Brief fixation in 1% glutaraldehyde/1% paraformaldehyde proved useful for stable preservation of both the inclusion fine structure and surface antigen activity, even after prolonged storage in buffer. The T-cell subset antigens (Leu-2a and Leu-3a) were more labile than the pan-T-cell antigen (Leu-1), the B-cell surface antigen (Leu-10), or the NK cell antigen (Leu-7). Localization of the former was improved by a two-step labeling procedure with primary and secondary antibodies. Both TRI and PTA were identified in T cells. Either type of inclusion could be found in the suppressor/cytotoxic or helper/inducer T-cell subsets. TRI also were found in a few anti-Leu-10 reactive cells. A crystalline form of PTA was identified in anti-Leu-1 and anti-Leu-7 reactive cells.

Contribution of Lyb 5+ and Lyb 5- B cells to the primary and secondary phosphocholine-specific antibody response.

Due to a mutation on their X-chromosome, CBA/N mice lack the Lyb-5+ subset of B cells. The loss of this B cell subset results in a profound alteration in the immune response of these mice to the hapten phosphocholine (PC). Thus, when these mice are immunized with high doses of PC-KLH (200 micrograms) in CFA, they: 1) fail to produce IgM anti-PC antibodies; 2) produce little or no anti-PC antibody bearing the normally predominant T15-idiotype; and 3) produce IgG anti-PC antibodies only late in the primary response. In order to more fully delineate this defect in responsiveness to PC, the splenic focus assay was used to analyze Lyb-5- B cell precursors from both normal and immune defective mice. Lyb-5- cells were obtained from normal (CBA/N x DBA/2)F1 (CD) female spleens by treatment with anti-Lyb-5 serum and complement. These normal Lyb-5- cells and Lyb-5- cells from immune defective CD male mice were stimulated in vitro with either PC-Hy or TNP-Hy in the presence of Hy-primed T helper cells. The results demonstrate that primary Lyb-5- PC-specific B cells fail to respond in the splenic focus assay, while secondary Lyb-5- PC-specific precursors respond normally, and that both primary and secondary Lyb-5- TNP-specific precursors respond in the splenic focus assay. These data suggest that Lyb-5- PC-specific precursors must differentiate into memory cells before they can be activated to secrete antibody, and they also indicate that the Lyb-5- B cell subset may be composed of two subsets with different activation requirements.

Functional heterogeneity of the Lyb-5- B cell subpopulation: mutant xid B cells and normal Lyb-5- B cells differ in their responsiveness to phenol-extracted lipopolysaccharide.

In the present study, responses stimulated by phenol-extracted lipopolysaccharide (LPS(phenol)) and butanol-extracted LPS (LPS(butanol)) were used to assess the possibility that xid B cells might not be identical to the Lyb-5- B cells present in normal mice. It was found that xid B cells responded well only to LPS(butanol) whereas normal B cells responded well to both LPS(butanol) and LPS(phenol). Thus, LPS(butanol) appeared to be a TI-1 antigen and LPS(phenol) appeared to be a TI-2 antigen. In contrast to classical TI-2 responses, however, responses stimulated by LPS(phenol) did not exhibit a stringent requirement for accessory cells. Furthermore, if LPS(phenol) were a classical TI-2 antigen, it should only activate Lyb-5+ B cells. To determine if the responsiveness of normal B cells to LPS(phenol) were due, at least in part, to the stimulation of normal Lyb-5- B cells, the responsiveness of normal neonatal B cells and normal adult B cells that had been pretreated with anti-Lyb-5.1 + C was assessed. It was found that both normal neonatal B cells and normal adult Lyb-5- B cells did respond well to LPS(phenol). Thus, even though LPS(phenol) does not stimulate xid B cells, these data demonstrate that LPS(phenol) is different from other TI-2 antigens. More importantly, these data also demonstrate that xid B cells and normal Lyb-5- B cells are not identical. It is hypothesized that the normal Lyb-5- B cell subpopulation is heterogeneous, consisting of an Lyb-5(1)- and an Lyb-5(2)-B cell subset with the xid mutation blocking the differentiation of Lyb-5(1)-B cells into Lyb-5(2)-B cells.

Analysis of the B cell subpopulations influenced by allogeneic effect factor. I. MHC restricted enhancement of b cell responses to thymic-independent antigens, types 1 and 2, in normal and CBA/N mice.

The ability of allogeneic effect factor (AEF) to enhance the in vitro antibody responses of murine T cell-depleted B cell populations to thymic-dependent (TD) antigens also occurs with thymic-independent (TI) antigens. AEF significantly enhanced the in vitro antibody responses of normal T cell-depleted spleen cell populations to both 2,4,6-trinitophenylated Brucella abortus (TNP-BA), a TI type 1 antigen, and 2,4,6-trinitrophenylated aminoethylcarbamylmethyl-Ficoll (TNP-Ficoll), a TI type 2 antigen. The enhancing activity of AEF on TI responses was analyzed further in immune-defective CBA/N mice that possess only Lyb-5- B cells, and thus fail to respond to TNP-Ficoll, but do respond to TNP-BA and sheep red blood cells (SRBC). AEF was able to reconstitute the in vitro antibody responses of CBA/NT cell-depleted B cell populations to SRBC and to augment the T cell-depleted response to TNP-BA. These data suggest that AEF can enhance the activity of both Lyb-5+ as well as Lyb-5- B cells in their responses to both type 1 and type 2 TI antigens, and to the TD antigen SRBC. This enhancing effect of AEF on the TD and TI responses was MHC restricted for both Lyb-5+ as well as Lyb-5- spleen cells.

Induction of a B-lymphocyte receptor for a T cell-replacing factor by the crosslinking of surface IgD.

The observation that anti-immunoglobulin antibodies and T cell-replacing factor (TRF) have a synergistic effect on the stimulation of B lymphocytes to differentiate into antibody-secreting cells suggested to us the possibility that the crosslinking of B-cell surface immunoglobulin by antigen or anti-immunoglobulin antibody might induce the expression of a B-cell receptor for TRF. In order to test this possibility we studied whether spleen cells from mice injected with 400-800 micrograms of anti-IgD antibody 1-3 days before sacrifice had an enhanced capacity to adsorb TRF activity from a partially purified culture supernatant of concanavalin A-stimulated mouse spleen cells. We found that spleen cells from euthymic or congenitally athymic mice injected 24-30 hr prior to sacrifice with either affinity-purified goat anti-mouse IgD antibody or a monoclonal allospecific anti-IgD antibody had greater than 100 times the TRF adsorptive capacity of spleen cells from control mice. In contrast, spleen cells from anti-IgD treated DBA/2Ha mice, which have been shown to have an immune defect associated with the lack of a TRF receptor, were unable to adsorb TRF activity from concanavalin A-stimulated helper supernatants. This suggests that the crosslinking of B-cell surface immunoglobulin may induce B lymphocytes to express a receptor or receptors for TRF and thus enhance B-cell responsiveness to this helper factor.