Search

It is widely accepted that the video frame size process is long range dependent (LRD), which has been indicated by a number of statistical tests. This approach suggests the use of LRD traffic models for traffic engineering, which in many cases leads to subexponential queue tails in the network buffers. Basically, there are two concepts that are considered as the origin of LRD in video modeling: self-similarity and heavy tailed level durations. In this paper we argue for a new approach, where the frame size sequence is described by three processes representing three explicit time scales, all of them having a physical meaning. We demonstrate that this concept serves as an explanation for the outcome of the LRD tests. The properties that are attributed to long range dependence in fact seem to be the result of non-stationarity of the large time scale process. We also show that the high frequency component is short range dependent, thus questioning the presence of long range dependence in video traffic.

We studied inactivation of HIV-1 by fresh sera of animals. We found that while fresh sera of humans and chimpanzees (among others) did not have antiviral activity, fresh sera of several other mammals, especially those of rodents and felines, showed a dose-dependent viral-inactivating property against cell-free HIV-1; these sera were also capable of inactivating virus preadsorbed to cells, similar to neutralizing antibody. The activity was destroyed by heating to 56 degrees C, required Ca2+ but no antiviral antibody, and therefore apparently involves the classical complement pathway. The activity could not be ascribed to any single fraction of sera separated on a size exclusion HPLC column. Mouse serum (the only one tested) also inactivated HIV-2. The data are consistent with classical C-mediated pathway inactivation of HIV-1 and HIV-2 by animal sera and is the first report of any "complement-like" antilentiviral serum factor. Elucidation of this mechanism may aid in understanding the lack of activity of human serum against HIV-1 and may prove useful in combined interventive strategies against HIV-1.

Methods for preparing chicken erythrocytes (CE) with C7 bound to their surface were devised by using either the classical pathway (CEA1-7) or an activated 56hu reagent (CE567) derived from inulin-treated human serum. These intermediates were used to study the lysis of CE by functionally purified C8 and C9 isolated from guinea pig and human serum. The results indicated that GPC9 was less efficient in lysing CEA1-8 or CE5678 than HuC9. This finding was observed irrespective of the species of C8 used. Experiments designed to analyze this difference indicated that there were two functionally distinct forms of C5b-8 that were randomly distributed among the cells but differed in their ability to generate a C lesion depending on the species of C9 used to complete the reaction. The implications of these results on the mechanism of generation of C lesions are briefly discussed.

EA3H-C4hu treated with 2-mercaptoethanol (MSH) or dithiothreitol (DTT) were analyzed for residual chains and hemolytic activity. MSH treatment resulted in no loss of the chains of C4b and no loss of reactivity with a polyclonal anti-human C4 rabbit antibody, and the cells did not lose their ability to generate SAC4b2a. DTT-treated cells lost about 70% of the gamma-chain and over 95% of C4b activity; there was no loss of alpha' and beta chains or the ability to react with anti-C4 antibody. Kinetic studies indicated that the remaining 30% of the gamma-chains could not be removed by prolonged incubation with DTT, implying heterogeneity of cell bound C4b. The data also imply that the gamma-chain is important in the generation of SAC4b2a.

After treatment with aminoethylisothiouronium bromide hydrobromide (AET), the ascites forms of the diethylnitrosamine-induced guinea pig hepatomas, line-1 and line-10, were susceptible or more susceptible to killing in vitro by certain combinations of tumor-specific or IgM anti-Forssman antibody and either human or guinea pig complement. Since AET could be toxic to either cell line, conditions of pH, concentration of AET, and duration of exposure of the cells to the reagent were determined that resulted in enhanced susceptibility without significantly affecting cell viability. Chicken erythrocytes (CE) also were tested and AET-treated cells found to be more susceptible to lysis by IgM anti-Forssman antibody and guinea pig complement. The enhancement apparently was not due to increased ability of AET-treated CE to fix antibody. In contrast to CE, the lytic susceptibility of sheep cells was not affected by AET treatment. In addition to AET, several drugs and enzymes that also can affect the susceptibility of the tumor cells to antibody and complement were tested and found to be ineffective against CE. Since AET reputedly acts directly on the cell surface, it seems reasonable to assume that the increased susceptibility of the tumor cells and CE to antibody and complement may result from a modification of the cell surface.

The first component of human complement was separated from C1-INH by sucrose linear gradient ultracentrifugation. Activation of C1 was studied in the absence and presence of immune complexes; activation was monitored by SDS-PAGE and Western blot. When the partially purified native C1 preparation was incubated at 37 degrees C without immune complexes, activated C1s appeared after 30 min in the case of eightfold dilution with respect to the original serum, and after 45 min with 32-fold dilution. Kinetics of appearance of activated C1r was the same as that of activated C1s. From the following results, we concluded that spontaneous activation may be partially due to proteolytic enzymes contaminating the preparation: 1) a nonspecific protease inhibitor, PMSF, completely inhibited spontaneous activation but did not inhibit the activation of C1 by immune complexes; 2) alpha 2-macroglobulin partially inhibited spontaneous activation, and 3) although spontaneous activation in the absence of PMSF was relatively slow, activated C1 accelerated spontaneous activation that was completely blocked by C1-INH. In contrast to spontaneous activation, the partially purified native C1 was rapidly activated by immune complexes: within 5 min almost all C1 was activated by rabbit IgG anti-human IgM-human IgM complexes. These results support conclusions derived from activation studies when using native C1 and hemolytic assays, and do not support those derived from the activation studies with reconstituted C1 and SDS-PAGE analysis. We suggest that the contradictions can be resolved if one assumes that C1 activation can be both an intra- and intermolecular process; which process dominates is determined by the state of C1 and by experimental conditions.

We investigated the dissociation by fluid phase hapten of IgG antibodies bound to cell surface hapten in the presence and absence of anti-IgG antibodies. Dissociation was quantitated with fluid phase hapten, preventing reassociation of the anti-hapten antibodies. More than 90% of the anti-hapten IgG alone was prevented from reassociation by low concentrations of fluid phase hapten (nanogram to microgram range). In contrast, no dissociation of some IgG-anti-IgG complexes could be measured even at 24 hr incubation in the presence of very large excess of fluid phase hapten (100 mg/ml). We excluded aggregate formation between anti-hapten antibodies due to cross-linking by anti-antibodies as a cause for decreased dissociability by 1) performing the experiments in large excess of anti-antibody, 2) showing that the phenomenon was independent of anti-hapten antibody density, 3) showing that decreased dissociation also occurred at 4 degrees C, and 4) showing that aggregation by protein A did induce decreased dissociability, albeit three orders of magnitude lower than the anti-antibody. It was concluded that anti-antibody directly affected the "avidity" of cell hapten bound anti-hapten IgG in an unknown manner.

It has been known for many years that IgG-C4 complexes are generated during the activation of the classical complement pathway. Information regarding binding of C4 by IgM is not available. We studied the binding of human C4 by natural human and rabbit anti-hapten IgM under conditions in which the IgM was bound to hapten coupled to sheep red cells. We eluted the IgM by excess fluid phase hapten, reattached the eluted IgM to red cells coupled with hapten, and transferred the IgM from red cell-IgM-C4 complexes to red cells coupled with hapten. Analysis of the cells by radioimmunoassay and/or by hemolytic assays showed that no C4 was eluted from the cells with the IgM, that no C4 transferred from cell to cell with the IgM, and that the C4 remaining on the cells was hemolytically active. We concluded that in contrast to IgG, natural human and rabbit IgM at a cell surface failed to bind C4.

Rabbit anti-methotrexate (MTX) IgG antibody was used to study the effect of MTX density on sheep red cells on the lysis of cells by complement (C). Under conditions in which the density of IgG did not vary, the number of lytic IgG complexes remained approximately the same when the average distance between MTX molecules was between 5 and 7 nm, but decreased rapidly with an average distance less than 10 nm. The results were interpreted to mean that in addition to aggregation for IgG to acquire maximal C in fixing and activating properties, the angle of the Fab arms of the molecule must not be larger than about 50 to 60 degrees. The data also suggested IgG molecules with Fab angles of less than about 30 degrees might also not be able to induce C-mediated cell lysis.

Summary Sheep erythrocytes were coupled with trinitophenyl sulphonate, sensitized with anti-TNP (or-DNP) IgM monoclonal antibodies, and exposed to components of the classical pathway of complement activation. When human fibronectin (FN) was added after C1q, but before addition of C1r and C1s (subunits of the first complement component), inhibition of haemolytic activity was observed which was strictly dependent upon the dose of FN. When FN was added after addition of C1 (reconstituted from C1q, C1r and C1s), the haemolytic activity of complement was not affected by the presence of FN. These data suggest that FN binds on C1q by interfering with C1r and C1s fixation. In addition, FN was unable to displace the activated subcomponents (C1r and C1s) from their binding site on C1q. When using other systems (sheep erythrocytes sensitized with anti-Forssman IgM monoclonal antibodies), the quantity of FN required to inhibit complement haemolytic activity was greater than in the TNP system. In normal plasma, there is a 50-fold excess of FN compared to free C1q.

Spontaneous activation of C1r in the presence of EDTA was examined by a Western blot. Partially purified native C1r was prepared by ultracentrifugation of fresh serum in 10 to 30% sucrose gradient; final concentration of C1r was one-sixth of the original serum. C1(-)-INH was not detectable by a single radial immunodiffusion (less than 0.5% of serum). The results demonstrated that 1) the rate of spontaneous activation of C1r was slow (less than 10% in 30 min); 2) it was concentration-dependent; 3) it was enhanced by activated C1r; and 4) it was almost completely suppressed by serine protease inhibitors up to 1 h. These results were inconsistent with an intramolecular autoactivation model of C1r in the fluid phase and suggested intermolecular activation by contaminating protease or activated C1r.

The intermediate product EAC1-7 released hemoglobin when incubated with high concentrations of GPC8 in the absence of C9. The reaction failed to reach an end point within 8 hr at 37 degrees C, and analysis of the kinetics indicated that it did not conform to the one-hit theory of immune hemolysis, and was not, therefore, the result of C9 contamination of the C8 preparation. The release of 86Rb from labeled EAC1-7 incubated at 30 degrees C with limiting C8 and excess C9 was paralleled, within 5 to 10 min, by release of hemoglobin. In the absence of C9 and with higher concentrations of C8, 86Rb was released rapidly, but hemoglobin release was delayed for several hours. Addition of excess C9 to concentrations of C8, which did not alone cause 86Rb release, resulted in substantial release of the isotope. These observations indicate that C9 acts by producing a distinct lesion in the cell membrane rather than by accelerating the release of hemoglobin from the C8-initiated 86Rb-releasing lesions. It is concluded that 86Rb release cannot be used as a reliable indicator of cell lysis and that C8- and C8/C9-mediated hemolysis are the result of mechanistically different processes.

The lack of direct proportionality between the number of SAC14 formed and the concentration of serum C1 was ascribed to the dissociation of C1 molecules into their subcomponents. We have obtained evidence, however, that this non-linear relationship is due to the dissociation from the C1 molecule of a hitherto unrecognized factor necessary for the activation of bound C1. Determination of the hemolytic activity of EAC14 before and after treatment with trypsin demonstrated that EAC14 contained two populations of SAC14, one in which C1 was activable by itself and by trypsin and another in which C1 was activable by trypsin but not by itself. A factor necessary for activation of C1, missing in non-self-activable C1, was designated “X” and the C1 molecules lacking this factor were designated (C1 - X). The finding that the apparent C1 content of whole serum was dependent on EAC4 concentration indicated that the activating factor “X” was bound independently from C1 and (C1 - X).

We present evidence for an activation step in the C1 molecule that occurs after the binding of C1 and before the appearance of C1̄ activity. Evidence for this internal activation step was obtained by the use of two different inhibitors of C1̄: C1̄-inactivator (I1) and a pentosan-poly-sulfo-ester (Sp54). Sp54 and I1 do not inhibit the uptake of C1. Sp54 inhibits a population of bound C1 not inhibitable by I1. C1 taken up in the presence of Sp54 are not inhibitable either by I1 or by Sp54. C1 taken up in the presence of I1 are inhibitable by Sp54 but not by I1. Sp54 is capable of inhibiting the hemolytic action of EAC1q4 while I1 is not. The data suggests that the internal activation step involves the subcomponent C1q.

On the cell intermediate EAC142, made with limited C1, excess C4 and excess C2, more than one (up to six) hemolytically effective complexes may be present at a single SAC1, i.e., complexes cluster at a singl hemolytice site. These cells lyse more rapidly, utilize C-EDTA more efficiently and have more C3 convertase activity per hemolytic site than EAC142 made with excess C1 and C4 and limited C2, although both cell types lyse to the same extent with excess C-EDTA. Multiple complexes per hemolytic site were generated with guinea pig and with human C1, C4 and C2.

The hemolytic efficiency of rabbit-anti-Forssman IgM antibody depends on the source of complement components. With whole guinea pig complement as few as one in 10 IgM molecules was hemolytic; it was found that with purified guinea pig C1, C4, and C2 components and with rat late-acting components, all IgM molecules were hemolytic. Evidence indicates that the block occurred as a result of generating at least two different cell-bound C4 molecules, one of which was not competent with homologous late-acting components.

The ability of IgG antibodies with different hapten specificities to fix C1 and activate C as a function of hapten density on a red cell surface was investigated. Rabbit anti-methotrexate and anti-folinic acid IgG antibodies in a mixture were highly efficient in fixing C1 and activating C when cells carried simultaneously high levels of both haptens. We wished to find out whether in a C-activating IgG complex both IgG molecules had to be in a form that could activate C1. By reducing hapten density of one of the haptens on a double labeled cell, complexes were generated where only one in a pair of IgG molecules was in the activating form; such a pair had the same activating efficiency as a pair in which both IgG molecules were in the activating form. It was concluded that cooperative activation of C in C1-binding IgG complexes required only one IgG in the complex to be in the activating form.

DELAGE et al.1 recently reported that L-asparaginase (L-A) in vitro inhibits human and guinea-pig whole serum complement and that L-A is a potent activator of C1. They therefore proposed that their results may “give a clue to the severe anaphylactic reactions noted in some patients on administration of the drug”. Because guinea-pig serum contains L-A in high quantities and Delage et al. used reagents and methods that make their data difficult to interpret on a molecular basis, we have reinvestigated the effect of L-A preparations on human and guinea-pig complement.

The mechanism of the terminal steps in the lysis of antibody-sensitized tumor cells by complement (TAC) was studied. It was shown that once complement has reacted, lysis proceeded even in the absence of fluid-phase complement. Transformation of TAC to dead cells was found to be at least a two-step process: one of the steps was temperature dependent whereas the other was reversibly inhibited by EDTA. In analogy to the hemolytic system, TAC has been designated T.

Lysis of sheep erythrocytes (E) sensitized with anti-Forssman antiserum (EA) is inhibited by the action of concanavalin A (Con A) on whole guinea pig complement (GPC). The degree of inhibition observed for a given quantity of GPC was dependent on the Con A concentration. Specifically, Con A inhibits the activity of the early acting complement components C1 and C2 in the fluid phase, but has no significant effect on lysis once these components are bound to EA. Results of tmax experiments performed in the presence or absence of Con A showed that inhibition of C2 activity results from a direct interaction between Con A and C2 and not from a decreased number of effective EAC14 sites. Furthermore, since Con A pretreated or untreated EAC14 cells had the same tmax value, Con A and C2 apparently do not compete for the same binding site on the indicator cells. The lectin has no observable effect on either fluid phase or cell-bound C4 activity. Under similar conditions, wheat germ or soy bean agglutinin, leucoagglutinin or pokeweed mitogen did not inhibit hemolysis.

A comparison of the events leading from T to dead cell and E to E ghost and hemoglobin was made. Transformation of T to dead cells was inhibitable by EDTA, NaF, histamine, DMSO, cytochalasin B and 3'5' cAMP. Transformation of E to E ghost and hemoglobin was only inhibited by EDTA.

Certain metabolic inhibitors or chemotherapeutic agents that increase the susceptibility of line-1 or line-10 guinea pig hepatoma cells to humoral immune attack were studied for their effects on the ability of the cells to synthesize lipids. A direct correlation was found between the drug-induced increase in sensitivity to antibody-C mediated killing and the inhibition of the ability of the cells to incorporate acetate, glycerol, and fatty acids into complex cellular lipids. Drug-treated cells recultured in drug-free medium regained their resistance to antibody-C mediated killing; these cells recovered their ability for complex lipid synthesis at this time. Thin layer chromatography of CHCl3:CH3OH lipid extracts from these cells indicated that the drug-induced increase in susceptibility to humoral immune attack correlated with the inhibition of acetate, glycerol, and fatty acid incorporation into cardiolipin and triglyceride in line-10 cells and the inhibition of incorporation of these compounds into cardiolipin alone in line-1 cells. No direct correlation was found between the sensitivity of the cells to humoral immune attack and the ability of the cells to incorporate precursors of lipid synthesis into other lipid moieties (sphyngomyelin, phosphatidyl serine, phosphatidyl choline, phosphatidyl glycerol, or cholesterol esters). The synthesis of cardiolipin and triglycerides, therefore, appears to be associated with the mechanism whereby these tumor cells resist antibody-C mediated killing.

Agents that increase (certain metabolic inhibitors, chemotherapeutic agents, and x-irradiation), decrease (hormones), or have no effect (hyperthermia) on the susceptibility of line-1 and line-10 guinea pig hepatoma cells to humoral immune attack were studied for their effects on the ability of these tumor cells to synthesize macromolecules. A correlation was found between the drug-induced increase in sensitivity of these cells to antibody-C mediated killing and the loss of their ability to incorporate fatty acids into complex cellular lipids. Similarly, the hormone-induced increase in resistance of the cells to killing was accompanied by an enhancement in complex lipid synthesis by these cells was also observed after the cells were exposed to physical means of insult (x-irradiation or hyperthermia). No correlation was found between the sensitivity of the cells to antibody-C mediated killing and their ability to synthesize DNA, RNA, protein, or complex carbohydrate, or their capacity for de novo lipid synthesis as measured by incorporation of acetate and glycerol into cellular macromolecules. The assembly of free fatty acids into complex lipid moieties is therefore proposed to be of fundamental importance for the ability of the tumor cells to resist humoral immune killing.

Sheep erythrocytes pretreated with supernatants of mouse spleen or thymus cells become resistant to lysis by guinea pig complement. The inhibitory activity (IA) reduces the utilization of C2 by EAC14. Because IA binds to the surface of sheep erythrocytes and does not inhibit C1 irreversibly, it is probably a hitherto undescribed inhibitor of complement.

The inhibitory effects of 0.1 M EDTA on the lysis of E prepared by incubating EA with whole GPC was studied. At high end point lysis (greater than 70%) 0.1 M EDTA failed to prevent hemoglobin release whereas at lower end point (less than 60%) 0.1 M EDTA was effective. In all cases hemoglobin release was inhibited by 25% BSA. When E were prepared by incubating EAC1-8 with C9, similar results were obtained. In this system the difference in the ability of 0.1 M EDTA to inhibit hemoglobin release at high or low end point lysis could not be correlated with the low end point lysis could not be correlated with the number of lesions/cell but appeared to be related to the C9 to SAC1-8 ratio. With limiting SAC1-8 and excess C9, E were produced from which hemoglobin release could not be prevented by 0.1 M EDTA whereas at lower C9 to SAC1-8 ratios hemoglobin release was prevented by 0.1 M EDTA. These differences most probably reflect functionally different sized transmembrane channels that were produced at different C9 to SAC1-8 ratios.

A number of metabolic inhibitors and chemotherapeutic agents have been found to increase the sensitivity of a chemically induced guinea pig hepatoma (line 1) to killing by antibody and complement. We have investigated whether the mechanism whereby these drugs increase sensitivity to killing is attributable to their primary action of inhibiting DNA, RNA, or protein synthesis. Line 1 cells incubated for 1, 4, or 17 hr with actinomycin D (25 microng/ml), adriamycin (40 microng/ml), or puromycin (5 micron/ml) or with 5-fold lower concentrations of these drugs were maximally inhibited (greater than 90%) in their ability to synthesize DNA, RNA, and protein within 1 hr. However, only cells incubated for 17 hr with the high concentrations of drugs showed increased sensitivity to killing by antibody and complement. Line 1 cells incubated with high concentrations of these drugs of 17 hr, washed, and resuspended in drug-free medium recovered their resistance to killing by antibody and complement within 4 hr. These cells ever after culture for 24 hr in drug-free medium did not regain their ability to synthesize DNA, RNA, or protein. A similar lack of correlation between synthesis of these macromolecules and sensitivity to antibody-complement-mediated killing was observed after the cells were treated with physical agents that inhibit macromolecular synthesis. Both heat-treated and X-irradiated cells were inhibited in their ability to synthesize DNA, RNA, and protein immediately after treatment; however, only X-irradiated cells (6 and 16 hr postirradiation) were increased in their sensitivity to antibody-complement-mediated killing. Our data show that the ability of line 1 tumor cells to resist humoral immune attack does not depend solely on their ability to synthesize DNA, RNA, or protein.

Antibody-sensitized line-1 or line-10 tumor cells treated with GPC (TAC) incorporated fatty acids into complex cellular lipids and released increased amounts of fatty acids within 5 to 10 min after the addition of GPC as compared to control cells. This effect was dependent on the concentration of GPC used; however, under conditions where the cells were not killed, the enhanced synthesis and release of lipids were not dependent on the antibody concentration used to sensitize the cells. Treatment of the cells with antibody alone, GPC alone, or antibody plus heat-inactivated GPC did not result in enhanced synthesis or release of lipids. No enhancement in DNA, RNA, or protein synthesis in TAC was noted. Line-1 cells, which can be killed by GPC when sensitized with excess anti-Forssman IgM antibody, demonstrated enhanced lipid synthesis within 1 to 3 min after the addition of GPC to the antibody-sensitized cells, before measurable killing of the cells had occurred. This effect persisted in the surviving cells when tested 5 and 10 min after the formation of TAC. Addition of GPC deficient in C4 to antibody-sensitized cells did not result in enhanced lipid synthesis or release. These data suggest that the synthesis of macromolecules of which lipids are a major component is of central importance for the ability of the cells to resist antibody-GPC mediated attack.

The inhibitory effect of metal salts on the formation and transformation of E to ghosts was studied. Ferrous and ferric salts inhibited the binding of C9 to EAC1-8 cells but had no other inhibitory role in the reaction. Uranly, copper, and zinc salts inhibited the transformation of the EAC1-9inserted intermediate to E ghosts. This inhibition occurred at a stage prior to detectable damage as measured by 86Rb or hemoglobin release and at a step prior to that inhibitable by high molarity EDTA. Consequently a further step in the reaction sequence of E to ghosts was identified. The reaction sequence leading from EAC1-9 to ghosts can be summarized as follows: formula: (see text).

Transformation of T to dead cells was prevented by 3'5' cAMP. The effect of 3'5' cAMP was dose, time, and temperature dependent. T washed free of 3'5' cAMP after short-term incubation proceeded to die to the same extent as control cells. After 3 hr of incubation of T with 3'5' cAMP the level of killing was significantly reduced. The 3'5' cyclic nucleotides of uridine, guanine, cytosine, and thymidine and the 2'3' cyclic adenosine nucleotide were not effective. It was concluded that prolonged treatment of T with 3'5' cAMP either irreversibly blocked the damage-producing process or facilitated the reapir of damaged sites.

Antigenic similarity between cell-bound human and guinea pig C4 was detected with antisera specific for each of these molecules. No cross-reaction could be detected between C4 of normal guinea pig or human sera or after treatment of the sera with antigen-antibody C1 complexes. The cross-reactivity between cell-bound human and guinea pig C4 may reflect conformational changes in the C4 molecule as a result of binding of the cell surface.

The intermediate product EAC1-8 released cytoplasmic components as a result of at least two sequential reactions after its interaction with C9. Binding of C9 to EAC1-8 occurred in a few minutes even at 0 degrees C. Trypsinization of EAC1-9 prepared and held at low temperature resulted in nullification of the potential hemolysis of these cells. A brief incubation at 30 or 37 degrees resulted in the formation of an intermediate whose hemolytic potential could not be nullified by trypsin. The failure of trypsin to nullify hemolysis was attributed to the insertion of C9 into the cell membrane. Studies on the effec of EDTA or low temperature suggested that the reported temperature-dependent step in E* formation described by Frank et al. was the insertion of C9. The results of the studies with 86Rb-labeled EAC1-8 indicated that a transmembrane channel was not formed until after the C9 had been inserted and a further reaction or reactions had occurred.

We reported inhibition of growth of primary rat mammary carcinomas after infusions of tumor-bearer plasma absorbed with Protein A-Sepharose or inactivated CNBr Sepharose. Absorbed plasmas were depleted of the third component of complement (C3) (other complement components defined similarly) and C5 but not C1, C4, or C2. These results suggested that activation of the alternative pathway of complement might be involved in the observed antitumor effects. To test this concept sera were treated with ethylenedinitrilotetraacetic acid or [ethylenebis(oxyethylenenitrilo)]tetraacetic acid before absorption with Protein A-Sepharose. Ethylenedinitrilotetraacetic acid, by chelating calcium and magnesium, prevents activation of both the alternative and classical complement pathways. [Ethylenebis(oxyethylenenitrilo)]tetraacetic acid, by chelating calcium but not magnesium, permits activation of the alternative pathway but inhibits activation of the classical complement pathway. Sera in the presence or absence of chelating agent were absorbed with Protein A-Sepharose twice at room temperature. After absorption calcium was added to the sera. Rats were treated by i.v. injection of sera twice a week for 2 weeks. Measurements of tumor size were made weekly for 5-7 weeks and then tumor weight was determined. Groups were compared both for size of index and total tumors. The results can be summarized as follows: tumor-bearer sera before absorption did not inhibit the growth of rat primary mammary carcinomas; tumor-bearer sera after absorption with Protein A-Sepharose showed significant consumption of C3 and did inhibit tumor growth; tumor-bearer sera absorbed in the presence of ethylenedinitrilotetraacetic acid did not show a decrease in C3 functional activity and did not inhibit tumor growth; tumor-bearer sera absorbed in the presence of [ethylenebis(oxyethylenenitrilo)]tetraacetic acid did show a decrease in C3 functional activity and did inhibit tumor growth; sera from normal adult female rats after absorption with Protein A-Sepharose did inhibit tumor growth. The results are consistent with a role for the alternative pathway of complement in the inhibition of growth of rat primary mammary carcinomas observed after treatment with absorbed sera.

The activation of classical C pathway by immune complexes depends on the binding and activation of C1, the first component of C. The Ig in the complex must be of the right class and in the right configuration to accomplish the conversion of precursor (zymogen) C1s to C1s, the active enzyme, whose substrates are C4 and C2. The primary question discussed in this paper is evidence that indicates that epitope distribution and density is a major factor in controlling the configuration of antibodies which in turn controls the activation of bound C1. This evidence confirms earlier findings that binding of C1 is a necessary but not sufficient condition for activating C1.

Lysis of autologous sheep red cells (labeled with methotrexate as the target hapten) by neat autologous serum was studied. The results indicated that the efficiency of lysis under physiological concentrations of cells and serum was about the same as lysis of the same target cells sensitized with autologous antibody and guinea pig serum as the source of complement. The data also indicate that the one-hit mechanism of immune lysis operates under near physiological conditions.

Line 1, a chemically induced guinea pig hepatoma, is susceptible to killing by anti-Forssman immunoglobulin M antibody and guinea pig complement. When these tumor cells are pretreated with insulin, L-epinephrine, hydrocortisone, or prednisolone, the cells show a marked reduction in their susceptibility to antibody-complement-mediated killing within 15 to 60 min; this effect reverses within 4 hr in the continued presence of hormone. Maximal binding of the hormones to the line 1 cells was observed within 60 min. However, the hormones remained bound to the cells after 4 hr of incubation, suggesting that line 1 cells incubated in the continued presence of hormone revert to the susceptible state despite the persistence of cell-bound hormone. Hormone-treated tumor cells, washed free of hormone and reincubated in hormone-free medium, lost nearly all their bound hormone within 15 to 30 min of washing. These cells, however, remained resistant to antibody-complement-mediated killing for up to 2 hr after washing. Line 1 cells, reverted in the continued presence of hormone, remained susceptible to killing by antibody and guinea pig complement after reexposure to the same, but not to a different, hormone. Hormone-treated cells reverted after prolonged incubation in hormone-free media; however, they were rendered resistant to killing after reexposure to the same hormone. The temporary refractoriness of reverted cells to further hormone stimulation was not due to an inability of the cells to bind hormone.

The guinea pig hepatoma (line-1) treated with anti-Forssman antibody (TA) and GPC sequentially released 86Rb, 14C from 14C aminoidobutyric acid and failed to exclude trypan blue. Incubation of TA with fluid phase GPC for 1 min caused maximal 86Rb release; however, if the GPC was removed at this time, the cells were not subsequently killed. Using a number of naturally occurring human sera deficient in a complement component we have shown 86Rb release requires the binding of the complement components 1 through 8, but there was no absolute requirement for C9. Irreversible damage to the cell as measured by 14C AIB release or uptake of trypan blue required the complete sequence of complete sequence of complement components. These observations indicate that 86Rb release is not a relible indicator cytotoxicity.

Line-1 diethylnitrosamine-induced guinea pig tumor cells can be rendered sensitive to killing by rabbit anti-Forssman IgM antibody plus guinea pig complement (GPC) or antitumor antibody plus GPC following prolonged incubation (17 hr ) of the cells with one of several metabolic inhibitors. Compared to control cells, these cells have been shown to be inhibited in their ability to incorporate fatty acids into complex cellular lipids, which suggested that lipid synthesis is of fundamental importance for the ability of the tumor cells to resist humoral immune killing. In this study, drug-treated cells that were rendered sensitive to killing by anti-Forssman antibody plus GPC, but not antitumor antibody plus GPC, were inhibited in their incorporation of saturated (palmitic or stearic acid), but not an unsaturated, fatty acid (linoleic acid). These data suggested that the fatty acid composition of specific lipids may also be important for the resistance of these tumor cells to killing by antibody and complement.

Concanavalin A (Con A) was found to inhibit the killing of antibody-sensitized line-1 tumor cells (TA) by guinea pig complement (GPC) but not by human complement (HuC). Other plant lectins (wheat germ, leucoagglutinin, and pokeweed mitogen) were also tested but Con A was the only lectin found to inhibit antibody-GPC-mediated killing. The inhibitory effect of Con A was observed when the GPC was mixed with Con A or when the antibody-sensitized cells were pretreated with Con A (TA-Con A) before the addition of GPC. The effect could be reversed by treatment of such cells with alpha-D-methylglucopyranoside or by incubation at 37 degrees C for approximately 2 hr. Con A appeared to act by preventing the binding of the first component of GPC (GPC1) to antibody-sensitized tumor cells. Differences in the binding of the first component of HuC (HuC1) and GPC1 to TA-Con A suggested that a difference in the binding site for HuC1 and GPC1 might exist. There was no difference in the number of GPC1 molecules fixed to antibody-sensitized sheep erythrocytes (EA) or EA treated with Con A in experiments using the same antibody as used with the tumor cells and the same Con A preparation. It would consequently appear that the inhibitory effect of Con A on the binding of GPC1 to TA is not due solely to an interaction of Con A with the antibody.

Line-1 guinea pig hepatoma cells are susceptible to killing by anti-Forssman IgM antibody plus guinea pig complement (GPC). When these tumor cells are incubated with insulin, epinephrine, hydrocortisone, or prednisolone, the cells show a marked reduction in their susceptibility to antibody-C-killing. If the ability of the cells to synthesize DNA, RNA, and protein is impaired by pretreatment with metabolic inhibitors, x-irradiation, or culture in nutrient-deficient media, the hormones are no longer effective in rendering the cells resistant to killing. If only DNA synthesis is impaired, but not RNA and protein synthesis, the hormones are effective. The inability of cells inhibited in their macromolecular synthesis to be rendered resistant to killing after hormone treatment is not due to an inability of the cells to bind hormone.