T-cell-dependent antibody responses in the rat: forms and sources of keyhole limpet hemocyanin matter.

The T-cell-dependent antibody response (TDAR) is a functional assay used in immunopharmacology and immunotoxicology to assess ability to mount an antibody response to immunization. Keyhole limpet hemocyanin (KLH) is extensively used as the immunogen of choice in non-clinical and clinical settings. Native KLH is comprised of high molecular weight (HMW; 4-8 MDa) assemblies of KLH subunit dimers (> 600-800 kDa). It is not known how the different forms (HMW vs subunit) and manufacturing processes (commercial sources) may impact the nature of anti-KLH immune responses (e.g. magnitude and inter-animal variability). Anti-KLH IgM and IgG responses were studied in Sprague-Dawley rats immunized with different forms and commercial sources of KLH: 100 µg of HMW KLH from two different sources or subunit KLH from three different sources. Biophysical and biochemical analyses were conducted to characterize the KLH formulations. Anti-KLH IgM and IgG responses were measured using a proprietary indirect quantitative electrochemiluminescence immunoassay. The HMW KLH preparations showed a greater number of sub-visible particles (2-150 µm size range) than the subunit KLH preparations. All HMW KLH and all subunit KLH were equivalent on SEC (hydrodynamic volume), PAGE (size and charge), and SDS-PAGE (molecular radius). Robust primary and secondary anti-KLH responses were detected for both sources of HMW KLH. The subunit KLH immunizations resulted in lower IgG and IgM responses compared to the HMW KLH, with the exception of Stellar Biotechnologies subunit KLH that produced both robust primary and secondary responses, which approached the HMW KLH responses. Inter-animal variability for IgM and IgG responses was lower with HMW KLH than with subunit KLH. In conclusion, different forms and commercial sources of KLH were associated with different magnitudes and inter-animal variability in IgM and IgG responses, a critical finding to take into consideration when designing TDAR studies for robust immunotoxicology or immunopharmacology testing.

The effect of platelet releasate on wound healing in animal models.

The alpha granules of platelets contain growth factors that are important in wound healing. We found that a major effect of thrombin-induced human platelet releasates in animal models of wound healing is to enhance the development of granulation tissue and new connective tissue matrix. These studies provide further evidence that platelet-derived protein factors may be useful in treating full-thickness dermal wounds by increasing the rate of granulation tissue formation.

Reaction to injectable collagen: results in animal models and clinical use.

Since its commercial release, Zyderm collagen implant has been used to treat more than 200,000 subjects in the United States for soft-tissue contour defects and more than 250,000 patients internationally (including the United States). Approximately 3 percent of subjects' skin tested with Zyderm collagen experience localized hypersensitivity reactions to collagen, whereas approximately 1 percent of treated patients demonstrate symptoms of hypersensitivity at treatment sites. Of the latter treatment responses reported since the conclusion of clinical trials with Zyderm, 56 percent occurred following the first treatment, 28 percent following the second, 10 percent following the third, and 6 percent following subsequent exposures. The data indicate that most patients receive a median of three treatments (mean = 4.4) with Zyderm collagen, but most patients who are likely to develop sensitivity to Zyderm collagen appear to respond immunologically to the test implant or first treatment exposure. Examining these treatment responses, 45 percent of the patients reported an onset of symptoms within 10 days and 22 percent at more than 30 days following the last treatment with Zyderm collagen. Erythema was the sole symptom in 24 percent of cases, whereas erythema plus induration comprised an additional 42 percent. Antibodies against Zyderm collagen were detected in the sera of 88 percent of these subjects using an ELISA, but no reactivity was observed against human collagen. Sera from patients reporting only systemic symptoms were not found to have anticollagen antibodies. These data suggest that the relative risk of a hypersensitivity reaction to Zyderm collagen does not increase with multiple exposures, since patients who are going to develop an immune response to bovine collagen react with greatest frequency to initial injections of collagen. In animal models, Zyderm collagen was shown to be less immunogenic than other medical devices which are composed of bovine collagen. Specifically, comparative studies were conducted in which Zyderm collagen and hemostatic agents were implanted in the guinea pig subcutaneum: sera from animals treated with collagen-derived hemostatic devices possessed significant levels of anti-implant antibodies (titers greater than 640), whereas animals treated with Zyderm collagen mounted minimal responses (titers less than 40). Additional studies were conducted in which implant materials were compared in a guinea pig parietal (bony defect) model and in a rabbit hemostasis model: in both, Zyderm collagen demonstrated lower immunogenicity than commercial bovine collagen hemostatic agents. Histologic results from these studies showed Zyderm

Leukocyte complement: interleukin-like properties of factor Bb.

It has been previously shown that the activated form of Factor B (Factor Bb) of the alternative pathway of complement activation stimulates monocyte spreading and killing of xenogenic erythrocytes and staphylococci. Factor Bb also stimulates lymphocyte blastogenesis in vitro, and native (uncleaved) Factor B is a major constitutive product of murine macrophages. To evaluate the possible "monokine" or "lymphokine"-like properties of Factor Bb, a radioimmunoassay was developed to measure the quantities of Factor B in phytohemagglutinin (PHA)-mitogen-stimulated cultures of human peripheral blood mononuclear cells. Nonstimulated mononuclear cell cultures from human peripheral blood (containing 10-14% monocytes and greater than 85% lymphocytes) at a density of 3 X 10(6) cells/ml (in serum-free medium) released less than 7 X 10(-10) M/liter (60 ng/ml) of Factor B antigen in 24 hr at 37 degrees C, and when mononuclear cells were stimulated with PHA mitogen in serum-free medium, the levels of Factor B antigen in media at 24 hr were significantly higher 1-3 X 10(-8) M/liter (0.9-2.8 micrograms/ml). The molecular size of Factor B in these media was 50-65 kDa by immunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electrophoresis, a size appropriate for Factor Bb (60 kDa). Since pathological effects of macrophages in autoimmune disease may result from the release of lysosomal hydrolases, the effects of purified Factor Bb on mononuclear phagocytes were investigated in an in vitro system of murine peritoneal exudate macrophages. Factor Bb induced secretion of marker lysosomal hydrolases N-acetyl-beta-D-glucosaminidase (hexosaminidase) and beta-glucuronidase from thioglycollate-elicited murine peritoneal exudate macrophages in a dose-response and kinetic manner. Hydrolase release was induced in serum-free medium without a known particulate activator at a concentration of 80-200 nM (5-13 micrograms/ml) Factor Bb. Maximal release occurred in 3-5 hr at 37 degrees C and extracellular enzyme activity of hexosaminidase and glucuronidase increased as intracellular enzyme levels decreased, suggesting that Factor Bb triggers release of these enzymes from intracellular lysosomal pools. These results provide an example of a complement protein which is synthesized, released, and activated during mononuclear cell culture and which induces release of lysosomal enzymes from macrophages. In conventional terminology, Factor B or Factor Bb might be termed a "lymphokine," "monokine," or "interleukin".

Complement activation in type 1 human diabetes.

Complement activation was quantitated in serum and plasma of diabetic and normal subjects by sensitive competitive equilibrium radioimmunoassays (RIA) for C3a, C4a, C5a, Factor B, and a newly described C5 neoantigen (termed C5 activation antigen, and abbreviated C5-AA) in a stable 54-kDa fragment of C5. Plasma C3a levels were significantly elevated in 8 of 16 patients with newly diagnosed Type 1 diabetes (P less than 0.0005) with the mean C3a concentration for these patients being more than 10-times greater than the mean value of normal controls. C4a levels were also elevated in 2 of these patients (P less than 0.02), but C5a levels, although higher than normal, were not significantly increased. In contrast, the levels of C5-AA in the serum of all patients (11/11) with chronic Type 1 diabetes were significantly higher than in control Type 2 patients (noninsulin-dependent diabetes) (P less than 0.0005) and 4 of 7 patients with new onset insulin-dependent diabetes mellitus also had significantly higher levels of C5-AA than the Type 2 patients (P less than 0.01). The levels of Factor B in the serum of 5 of 9 patients with new onset diabetes were significantly higher than normal (P less than 0.0025). Five recent onset Type 1 diabetes patients were evaluated longitudinally for C3a, C4a, and C5a: in 3 the levels of C3a were elevated during new onset disease decreasing into the normal range during remission; in 2 of these patients C4a was also significantly elevated and the levels decreased during remission; and in 3 patients the levels of C5a were not significantly elevated but they decreased during remission. Purified human complement proteins and complement hemolytic assays were used to measure complement activation in serum during incubation with rat pancreatic islet cells. With diluted normal human serum, less than 20% of C3 or Factor B were consumed during 30 min at 37 degrees C, while with new onset Type 1 diabetic patient sera up to 90% of C3 and Factor B were consumed in 5/6 sera and 4/6 sera, respectively. These findings suggest (a) that complement activation fragments C3a, C4a, and C5a are generated in vivo in new onset Type 1 diabetes; (b) that both the classical and the alternative complement pathways may be activated; and (c) that this may result in a measurable activation of C5 generating biologically and immunologically active C5a and other C5 activation fragments.(ABSTRACT TRUNCATED AT 400 WORDS)

Factor B, the complement alternative pathway serine proteinase, is a major constitutive protein synthesized and secreted by resident and elicited mouse macrophages.

Factor B, the complement alternative pathway serine proteinase, a class III gene product of the major histocompatibility complex, is a major constitutive secretion product of mouse mononuclear phagocytes. This glycoprotein was synthesized and secreted by macrophages as a doublet of Mr 90,000 and 93,000 polypeptides that were immunoprecipitable with antibodies raised to human serum factor B, and that were indistinguishable from plasma factor B by immunoreactivity, peptide mapping, and molecular weight. Macrophage factor B was cleaved and activated to factor Bb- and Ba-like fragments by factor D and cobra venom factor. Some conversion of macrophage factor B to Bb-sized fragments occurred spontaneously in the conditioned culture medium after several hours. Factor B represented approximately 0.5% of newly synthesized protein and 4-6% of the secreted protein of resident peritoneal macrophages and macrophages elicited with thioglycollate broth, pyran copolymer, NaIO4, bacillus Calmette-Guerin, or Corynebacterium parvum. We detected synthesis of factor B immediately upon explanting these macrophages in culture; synthesis continued for several days in culture. The rate of secretion of factor B, as a proportion of total protein secretion in culture, remained constant with time. By radioimmunoassay, factor B antigens accumulated in the 24-h macrophage-conditioned culture medium at 2-10 nM, and was present in cell lysates at 4-8 nmol per 10(6) cells. We detected synthesis of factor B in bone marrow-derived macrophages as early as 5 d of culture. The P388D1 macrophage line synthesized factor B, but mouse L cells did not. In contrast, apolipoprotein E, another secreted protein of macrophages, was secreted by resident and thioglycollate-elicited macrophages but not by freshly harvested pyran copolymer-activated macrophages. Its synthesis was initiated at day 9 in culture of bone marrow-derived macrophages. These data support the classification of factor B as a constitutive biosynthetic and secreted protein of immature and mature macrophages in various states of activation. Production of factor B was modulated by treatment of macrophages in vivo or in culture with bacterial lipopolysaccharide endotoxin, which increased the synthesis, secretion, and accumulation of factor B up to 11-fold.

Immunological events in new onset diabetes.

Data from our laboratory are reviewed showing that both cell-mediated cytotoxicity, complement-dependent antibody-mediated cytotoxicity, antibody-dependent cellular cytotoxicity, and complement-augmented antibody-dependent cellular cytotoxicity may provide mechanisms which kill pancreatic islets during pathogenesis of insulin-dependent (type I) diabetes. Xenogenic test systems employing dispersed rat islet target cells were employed. Most of the findings were reversible during clinical remission of diabetes.

Leukocyte complement: a possible role for C5 in lymphocyte stimulation.

The results presented here show that Fab' antibody fragments directed to complement proteins C5, C6, and C7 inhibit lymphocyte stimulation in mixed lymphocyte culture (MLC) by up to 65%, as determined by decreased incorporation of 3H-thymidine. Lymphocyte stimulation induced by PHA-mitogen was also inhibited up to 100% by anti-C5 Fab'. Specificity of these reactions was established by the findings that goat anti-C5 or murine hybridoma anti-C5 both inhibited MLC; the inhibitory activity of anti-C5 Fab' was absorbed with highly purified C5 (but not with C3), and antibody directed to C3 did not inhibit lymphocyte stimulation by MLC or PHA. The effects of anti-C5 were exerted in a nontoxic manner. Cleavage of lymphocyte associated C5 with factor B (Bb) or with trypsin resulted in stimulation of lymphocyte thymidine incorporation. Purified C5a was found to induce lymphocyte stimulation in serum-free medium in pulse-chase types of experiments. Anti-C6 and C7 Fab' also inhibited lymphocyte stimulation induced in one-way MLC. These results suggest that C5, C5a, and/or C6 and C7 may play a role in triggering of lymphocyte blastogenesis.

Human monocyte spreading induced by activated factor B of the complement alternative pathway: differential effects of Fab' and F(ab')2 antibody fragments directed to C5, C6, and C7.

Human peripheral blood mononuclear phagocytes are induced by activated Factor B (Bb) of the complement alternative pathway to undergo morphological shape changes in vitro which have been described as "spreading." The spreading reaction induced by Bb has previously been shown to depend upon the enzymatic activity of Bb and to be inhibited by Fab' antibody fragments directed to C5 (but not anti-C3 Fab'). The possibility that Bb may exert its effect on monocytes by initiating assembly of terminal complement complexes comprised of C5b, 6, 7, C5b-8, or C5b-9 was addressed in the present study. The effects were tested of Fab' and F(ab')2 antibody fragments directed to C5, C6, C7, and C8 and to neoantigens expressed in the assembling terminal complement complexes on the monocyte spreading reaction induced by Bb. Differential effects of monovalent Fab' and divalent F(ab')2 antibody fragments were observed. Anti-C5, C6, and C7 Fab' were found to inhibit the spreading reaction induced by Bb in an immunologically specific manner. Divalent F(ab')2 fragments directed to these same proteins (but not to C3, C4, C8, or C9) induced monocyte spreading in the complete absence of Bb or other recognized inducing agents. Monocyte spreading induced by hybridoma immunoglobulin (Ig) directed to C5 and C7 was found to be correlated with the binding of 10(6) molecules Ig per cell. These findings support the notion that C5, C6, and C7 (or an analogous system of cellular proteins) are associated with the surface of human peripheral blood monocytes and that these proteins may play a role in certain reactions by which mononuclear phagocytes are induced to altered states of cellular physiology.

The leukocyte complement system.

The existence of a leukocyte complement system is suggested by the findings that certain lymphocytes and mononuclear phagocytes activate complement, express membrane receptors specific for complement activation fragments, and synthesize complement proteins. In this review, attention is directed toward studies indicating that complement proteins may be expressed on the surface of lymphocytes and monocytes, the mechanisms by which complement is activated by leukocytes, and the possible role(s) of this cellular complement system in processes by which lymphocytes are stimulated and monocytes are activated to increased cellular activity.

Human monocyte spreading induced by factor Bb of the alternative pathway of complement activation. A possible role for C5 in monocyte spreading.

The central serine esterase of the alternative pathway of complement (APC) activation, activated factor B (Bb), has been shown recently to induce murine macrophages and human monocytes to become spread on a glass substrata. It has also been established that to induce the spreading reaction, the catalytic site of the Bb enzyme must be structurally intact since treatment of Bb with heat (56 degrees C for 30 min) or diisopropylfluorophosphate (10(-3) M) destroyed both enzymatic and spreading activities. In the C3b,Bb complex, Bb exhibits restricted substrate specificity for C3 and C5. With this in mind, the role of C3 and C5 in the monocyte spreading reaction was explored in the present study. Expression of C3 and C5 on the surface of human peripheral blood monocytes was investigated by the direct fluorescent antibody technique employing fluorescein isothiocyanate-conjugated anti-C3 or C5 F(ab')2 antibody fragments. It was found that C3 and C5 were present on 6 +/- 7% of freshly prepared monocytes and that expression of C5, but not C3, increased to 70 +/- 6% when monocytes were incubated for 3 d in serum-free medium. Biosynthesis of C5 was indicated when it was found that under serum-free conditions, monocytes incorporated [3H]leucine into immunoprecipitable C5 with an apparent mol wt of 180,000 on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The role of C3 and C5 in the monocyte spreading reaction induced by factor Bb was explored by testing for the ability of anti-C3 and anti-C5 Fab' antibody fragments to block monocyte spreading. It was found that anti-C5 Fab' inhibited by up to 100% the 3-h human monocyte spreading reaction induced by Bb; in contrast, anti-C3 Fab' or anti-C4 Fab' inhibited by less than 10%. That the inhibitory effect of anti-C5 Fab' was exerted directly on the monocyte was established when it was found that the 3-h monocyte spreading reaction was significantly inhibited by pretreating monocytes with anti-C5 Fab' for 20 min and then washing before the addition of Bb. The specificity of the inhibitory effect of anti-C5 Fab' was established by quantitatively absorbing the antibody fragments with polyacrylamide gel-purified C5 antigen: greater than 4 microgram of C5 absorbed by 100% the inhibitory activity of 10-20 microgram of anti-C5 Fab'. That factor Bb exerted its effect on monocytes by interacting directly with cell surface C5 was indicated when it was found that purified C5 inhibited the monocyte spreading reaction induced by Bb; greater than 25 microgram of C5 inhibited by 100% the spreading reaction induced by 3 microgram factor Bb.

Activated factor B (Bb) of the alternative pathway of complement activation cleaves and activates plasminogen.

Activated Factor B (Bb), the central serine esterase of the alternative pathway of complement activation, exhibits restricted substrate specificity in the complement system for C3 and C5. The results presented here indicate that Bb can cleave and activate plasminogen in an experimental system containing purified plasminogen and Bb; complement cellular intermediate bearing the Bb-enzyme; or cobra venom factor-stabilized Bb-enzyme (CVF,Bb). Cleavage of plasminogen by Factor Bb generated 2 disulfide-linked polypeptides with apparent m.w. of 64,000 and 25,000 to 32,000 (SDS-PAGE). Complement cellular intermediates containing the C3b, Bb-enzyme cleave 40 to 80% of 4.5 micrograms of 125I-labeled plasminogen during 30 min of incubation at 37 degrees C; native Factor B was inactive; and anti-Factor Blg inhibited by 100% the plasminogen cleavage mediated by complement cellular intermediates bearing the Bb-enzyme. Fibrinolytic activity was detected in plasminogen activator (PA) assays when purified plasminogen and 125I-labeled fibrin tubes were incubated with Bb, CVF, Bb, or complement cellular intermediates bearing the C3b,Bb-enzyme: 10 micrograms Bb released 40 to 65% of the 125I-fibrin released by 5 micrograms urokinase in 4 hr at 37 degrees C. Plasminogen activator activity of the C3b,Bb-enzyme was found to be regulated in serum. At dilutions of NHS 1:50, the PA-activity of 1.6 micrograms Bb was 100% inhibited, and at a 1:250 dilution, 50% inhibition was observed. This report describes a novel activity for the Bb-enzyme, which constitutes the C3/C5-convertase of the alternative pathway of complement activation.

The effects of mild reduction on the structure and function of C3.

Reduction and alkylation of C3 in nondenaturing buffers by 2 mM DTT (1 hr, 23 degrees C) (C3 R/A) results in a > 99% loss of C3 hemolytic activity. In contrast, sham alkylated protein, or protein allowed to reoxidize from the mildly reduced state, showed full activity. The loss in lytic activity correlated with the alkylation of approximately 6 liberated sulfhydryls, all of which arose from intrachain disulfides in the alpha polypeptide chain of C3. Data from CD experiments indicate that these bonds are essential for maintaining the native conformation of the protein. The loss of hemolytic activity of C3 R/A was neither due to its inability to be cleaved by C3 convertases nor to a dysfunction of the cleavage-induced labile membrane binding site as cells bearing C3 convertases firmly bound C3 R/A to an extent 80% that of native C3. The hemolytic defect in C3 R/A was found to result from its inability, once deposited on C42-bearing cells, to bind C5 and thereby participate in the formation of a C5 convertase. Unlike the effect on hemolytic activity, C3 R/A showed only a partial loss (3-fold) in its ability to bind to C3 receptors on monocytes. Taken together, these studies differentiate in molecular terms several of the functional sites in C3 and stress the significance of intra-alpha-chain disulfides in maintaining the conformation in activated C3 required for binding of C5.

Neoantigen of the complement membrane attack complex of cytotoxic human peripheral blood lymphocytes.

Neoantigenic determinants (neoAg) specific for the assembling membrane attack complex (MAC) of complement were detected by immunofluorescence microscopy on the surface of cytotoxic lymphocytes during the antibody-dependent cellular cytotoxicity (ADCC) reaction. This study employed antibody-sensitized chicken erythrocytes as target cells, human peripheral blood lymphocytes as effector cells, and RITC-conjugated rabbit F(ab')2-anti-neoAg. NeoAg was present on 60% of ADCC plaque-forming lymphocytes (PFL). Eight out of 182 neoAg-positive PFL were observed in direct contact with their target cells. In these cases MAC-specific neoAg was visualized at the zone of contact between the cells. Anti-neoAg Ig was found to inhibit ADCC plaque assays up to 62%; and 51Cr-release assays up to 79%. Stimulation of lymphocytes by PHA or mixed lymphocyte culture increased the expression of neoAg. In the case of PHA, increased neoAg expression was correlated with an increased incorporation of 14C-leucine into C5, C6, C7, and C8 antigens, which was detected by immunodiffusion and autoradiography.

Leukocyte complement: neoantigens of the membrane attack complex on the surface of human leukocytes prepared from defibrinated blood.

The neoantigenic determinants (neoAg) which have been identified in the human C5b-9 membranolytic C complex were detected here by the direct fluorescent antibody technique on the surface of 27 +/- 11% of viable peripheral blood leukocytes (PBL). The cells were prepared from defibrinated blood by sedimentation on Ficoll-Hypaque. Specificity of the antisera was established by quantitative inhibition of the fluorescent staining reaction, and of agglutination of EAC1-7, by highly purified C5b-9 complex. No inhibition was observed with fresh normal human serum. The majority of the PBL with surface neoAg was found in the B lymphocyte subpopulation that failed to form rosettes with sheep erythrocytes. NeoAg on B lymphocytes was removed to differing degrees by trypsin, papain, or pepsin treatment, and by maintaining the cells at 4 degree C for 20 hr in serum-free medium. The individual components, C5, C6, C7, C8, and C9, were also detected on the surface of PBL. With differential fluorescent stains, C5 and neoAg as well as C8 and neoAg could be detected on the same cells. The results indicate that viable B lymphocytes prepared from defibrinated blood, have the components of the membrane attack complex of C on their surface. The concomitant occurrence of the neoAg indicates that these proteins are present at least in part in the form of the assembled terminal complex.

Neoantigen of the membrane attack complex of human complement: Occurrence on peripheral blood leukocytes from patients with systemic lupus erythematosus.

Since complement activation and hematological abnormalities occur in systemic lupus erythematosus (SLE), the present study is an investigation of whether the membrane attack complex of complement might be bound to peripheral blood leukocytes (PBL) in vivo. Assembly of the membrane attack complex results in the generation of neoantigen (neoAg) which is complex-specific and not expressed by any of the individual complement proteins. FITC antiserum specific to neoAg was employed to detect the membrane attack complex on PBL from 7 normal donors, 12 patients with SLE, and 2 patients with rheumatoid arthritis (RA): 3 +/- 1% of normal, 25 +/- 13% of SLE, and 23 +/- 11% of RA PBL were positive. The majority of the neoAg positive PBL in SLE were polymorphonuclear neutrophils (PMN) as shown by adherence to plastic, phagocytosis of carbonyl iron, and differential cell counts. The PBL were greater than 98% viable as indicated by the trypan blue exclusion technique. These observations strongly suggest that the membrane attack complex may be bound to viable PBL in patients with SLE and RA, and further raise the possibility that the membrane attack complex, may have a function other than lysis.

Leukocyte complement: assembly of the membrane attack complex of complement by human peripheral blood leukocytes in the presence and absence of serum.

The specific neoantigenic determinants (neoAg) that are indicative of the assembled C5b-9 C complex are generated on the surface of peripheral blood leukocytes (PBL) during collection and processing of blood. Formation of neoAg on PBL could be prevented by collecting blood directly into 20 mM EDTA and, could be induced in vitro by adding autologous serum to isolated PBL that lacked neoAg. When neoAg was induced by the addition of serum containing 125I-labeled C8, the C8 was incorporated into a 23S complex which could be eluted from PBL. A mechanism for neoAg formation on PBL independent of exogenous serum factors was detected when PBL were placed into culture in serum-free medium. Results with metabolic inhibitors and 14C-leucine suggest that PBL can synthesize C5 and assemble the C5b-9 complex. The possible relevance of these findings to the understanding of mechanisms of cell-mediated cytotoxicity is discussed.