Search

The data presented in this paper support the hypothesis that unresponsiveness to autologous thyroglobulin (Tg) exists in the T cells and responsiveness exists in the B cells. Such a conclusion is based on the results of antigen-binding studies where few, if any, thymocytes recognized syngeneic Tg. Comparable numbers of antigen-binding lymphocytes for syngeneic Tg were found in the spleens of normal intact mice and of nude mice. The latter fact suggested that B cells exist which recognize self-constituents. From antigen-suicide experiments, a clearer picture of the susceptibility of B cells to iodinated self-antigen and of the obligatory role of antibody in the induction of lesions was developed. Only bone marrow cells (B cells) were affected by [(125)I]syngeneic Tg, in which case the incidence of lesions was diminished. From adoptive transfer experiments, the results demonstrate that unresponsiveness may be terminated by immunization with a mixture of heterologous (cross-reacting) Tg's. In this situation T cells are required since a B-cell reconstituted host failed to make antibody (plaque-forming cells) and to develop lesions. T cells in this form of an unresponsive state may recognize determinants on the heterologous Tg unrelated to autologous Tg and as such stimulate the normal complement of B cells to produce antibody that both reacts with autologous and heterologous Tg.

Although nonresponder, H-2(s) and H-2(q), mice fail to develop GAT-specific PFC responses to GAT, they do develop GAT-specific PFC responses when stimulated by GAT complexed to an immunogenic carrier such as methylated bovine serum albumin. The studies described in this paper show that injection of nonresponder mice with GAT specifically decreases their ability to develop anti-GAT PFC responses to a subsequent challenge with GAT-MBSA. Addition of GAT to cultures of spleen cells from nonresponder mice also prevents development of the GAT-specific PFC responses stimulated by GAT-MBSA. Thus, interaction of nonresponder spleen cells with GAT leads to the induction of unresponsiveness in vivo and in vitro. Various parameters of the tolerance induction have been investigated and described. A comparison of the effects of GAT on B cells indicates that nonresponder B cells are more readily rendered unresponsive by soluble GAT than are responder B cells. The significance of these data for our understanding of Ir gene regulation of the immune response is discussed.

Two new methods are described for making chromosomal spreads of single antibody-forming cells. The first depends on the controlled rupture of cells in small microdroplets through the use of a mild detergent and application of a mechanical stress on the cell. The second is a microadaptation of the conventional Ford technique. Both methods have a success rate of over 50%, though the quality of chromosomal spreads obtained is generally not as good as with conventional methods. These techniques have been applied to an analysis of cell to cell interaction in adoptive immune responses, using the full syngeneic transfer system provided by the use of CBA and CBA/T6T6 donor-recipient combinations. When neonatally thymectomized mice were restored to adequate immune responsiveness to sheep erythrocytes by injections of either thymus cells or thoracic duct lymphocytes, it was shown that all the actual dividing antibody-forming cells were not of donor but of host origin. When lethally irradiated mice were injected with chromosomally marked but syngeneic mixtures of thymus and bone marrow cells, a rather feeble adoptive immune response ensued; all the antibody-forming cells identified were of bone marrow origin. When mixtures of bone marrow cells and thoracic duct lymphocytes were used, immune restoration was much more effective, and over three-quarters of the antibody-forming mitotic figures carried the bone marrow donor chromosomal marker. The results were deemed to be consistent with the conclusions derived in the previous paper of this series, namely that thymus contains some, but a small number only of antigen-reactive cells (ARC), bone marrow contains antibody-forming cell precursors (AFCP) but no ARC, and thoracic duct lymph contains both ARC and AFCP with a probable predominance of the former. A vigorous immune response to sheep erythrocytes probably requires a collaboration between the two cell lineages, involving proliferation first of the ARC and then of the AFCP. The results stressed that the use of large numbers of pure thoracic duct lymphocytes in adoptive transfer work could lead to good adoptive immune responses, but that such results should not be construed as evidence against cell collaboration hypotheses. Some possible further uses of single cell chromosome techniques were briefly discussed.

In this series of papers it has been shown that the immune response of mice to sheep erythrocytes requires the participation of two classes of lymphoid cells. Thymus-derived cells initially react with antigen and then interact with another class of cells, the antibody-forming cell precursors, to cause their differentiation to antibody-forming cells. Antilymphocyte globulin depressed the ability of mice to respond to sheep erythrocytes. This effect was more marked when the antigen was injected intraperitoneally than intravenously, and occurred only when the antilymphocyte globulin was given before or simultaneously with antigen. Injection of thymus cells restored to near normal the ability to respond to an intravenous injection of sheep erythrocytes. Spleen cells from antilymphocyte globulin-treated mice gave a weak adoptive immune response in irradiated recipients. The addition of thymus cells however enabled a response similar to that given by normal spleen cells. When thymectomized irradiated recipients were used, normal spleen cells continued to give a higher response to a challenge of sheep erythrocytes at 2 and 4 wk postirradiation than did spleen cells from ALG-treated donors. This result is more consistent with the notion that thymus-derived target cells are eliminated, rather than temporarily inactivated, by antilymphocyte globulin. These findings suggest that, in vivo, antilymphocyte globulin acts selectively on the thymus-derived antigen-reactive cells.

A combined morphological and metabolic study has been made of the lymphoid cells in the blood during the immune response in man. Similar changes were observed in both primary and secondary responses to a number of different microbial antigens. The cellular response involved an increase in numbers of three types of cell; hyperbasophilic medium lymphocytes, plasma cells, and large lymphoid cells. The large lymphoid cells were about 20 micro in diameter with large nuclei, prominent nucleoli, and an intensely basophilic cytoplasm with numerous polyribosomes. About 30% of these cells were in the DNA synthetic phase of cell growth. Electron microscopy has shown that many of the basophilic medium-sized cells have sufficient well-organized endoplasmic reticulum to be included in the plasma cell series. The hyperbasophilic cells labeled more heavily with tritiated uridine and tritiated leucine than the normal small and medium lymphocytes from the peripheral blood of patients not under antigenic stimulation. The evidence in this paper supports the argument that the atypical mononuclear cells first described by Türk and others in the blood of patients with infections are immunoblasts, plasma cells, and other reactive lymphoid cells representing a circulating population of lymphoid cells derived from lymphoid tissue responding to antigenic stimulation. The presence of such cells may be a valuable indication that an immunological reaction is in progress when direct proof is lacking.

The effects of a single injection of a bacterial endotoxin on the cellular changes of a primary immune response to a standard dose of sheep red blood cells were studied in the spleens and mesenteric lymph nodes of mice. Daily histological comparisons of these organs in mice, injected with endotoxin, or with antigen, or both, showed that endotoxin given simultaneously with sheep red blood cells, as antigen, significantly enhanced all of the cellular changes that appear in the mesenteric lymph nodes and spleens of mice that form antibody when that antigen is given alone. First, in the white pulp of the spleens and cortical regions of the nodes, there appeared an early and excessive proliferation of the large pyroninophilic cells which seems to be responsible for the earliest formation of antibody, as judged by this work and that of others cited in the body of the paper. Polymorphonuclear cells invaded the spleens of these animals early after simultaneous challenge with antigen and endotoxin, and in far greater numbers than have ever been seen in mice given the same antigen without endotoxin. "Activated" germinal centers formed in the lymphoid tissue either 1 day before the appearance of antibody in the blood stream or on the same day, and they became larger than in the mice given antigen only. On the other hand, these specific and characteristic cellular changes failed to appear in mice prevented from forming any antibody at all by injections of endotoxin given 2 days before the antigenic challenge. These findings are discussed in the light provided by data from recent reports of others as well as in the light of the accompanying paper (1) which demonstrated not only the enhancement of antibody formation following simultaneous injections of antigen and endotoxin, as already known, but a totally unexpected, complete suppression of its formation when endotoxin was given 2 days before antigen.

High and low zone antibody tolerance to bacterial flagellin can be induced in adult strain W Wistar rats by multiple injections of a cyanogen bromide (CNBr) digest of flagellin at two widely spaced dose levels. Intermediate doses of the CNBr digest produce enhanced antibody titers to flagellin rather than antibody tolerance. Studies reported in this paper revealed that both high and low zone antibody tolerance to flagellin were accompanied by heightened levels of delayed-type hypersensitivity. Conversely, when enhancement of the antibody response occurred, suppression of delayed hypersensitivity was observed. This inverse relationship between humoral and cell-mediated immunity was very striking in strain W Wistar rats but was not quite so clear-cut in another strain of Wistar rats (strain J). Strain J rats were resistant to the induction of antibody tolerance and gave higher immunological responses to flagellin than strain W animals. In addition, it was observed that, in contrast to adult tolerance, administration of the CNBr digest to neonatal rats induced complete tolerance at the level of both humoral and cell-mediated immunity. These findings were discussed in the light of earlier studies with flagellin and provide further evidence for a previously described hypothesis.

Experiments in rabbits were designed to test the two unproven assumptions of the hypothesis proposed in the companion paper (1): that Original Antigenic Sin is fundamentally a restricted anamnestic response, and that there exists a trapping mechanism capable of deflecting antigen from one kind of cell and guiding it to another. It is shown that whole-body X-irradiation, sufficient to abolish primary but not secondary production of antibodies, leaves all manifestations of the Original Antigenic Sin untouched. This proves the first assumption. Primary immune animals challenged with very large doses of a related antigen produce an immediate response of cross-reactive antibodies, followed by a standard primary response to the challenging antigen. When boosted with an appropriate mixture of both antigens, the response is of standard secondary type to the homologous antigen, followed by a standard primary response to the crossreacting antigen. When animals are primarily vaccinated with a mixture of two related antigens, small booster doses of either will stimulate a standard secondary response only. When such animals are given very large booster doses of either antigen, the response is a compound of a homologous secondary and of an Original Antigenic Sin-type against the related antigen. Each of these findings demonstrates a corollary of the second assumption. The results are discussed in terms of the limitations they impose on theories concerned with the production of antibodies.

The DNA-synthesizing capacity of single antibody-forming cells was tested by a combination of micromanipulatory and autoradiographic techniques. Rats were immunized with S. adelaide flagellin, a protein antigen known to contain significant contamination with somatic (O) antigen. Single cells from secondarily immunized rats were tested for production of anti-H and anti-O antibodies by previously described and newer techniques. Positive antibody producers were transferred onto clean dry slides by micromanipulation, and autoradiographs were performed. When rats had received tritiated thymidine 1 hour before killing, labeling of antibody-forming cells was taken to imply that the cell was preparing for further mitotic division. It was found that on the 2nd and 3rd day of a secondary response, many of the antibody-producing cells in the nodes (chiefly plasmablasts) were incorporating tritiated thymidine. At the height of the cellular response, however, at 4 and 5 days, the majority of active antibody producers (chiefly mature plasma cells) were incapable of DNA synthesis. There appeared to be an inverse relationship between the antibody-forming and DNA-synthesizing capacities of the cell population under study; as more of the cells studied formed detectable antibody, fewer of them incorporated the DNA precursor. The age of plasma cells was also studied. Animals were killed at the height of the cellular immune response, having previously received an injection of tritiated thymidine 1 to 48 hours before killing; i.e., at 63 to 110 hours after their secondary stimulus. As the interval between isotope injection and killing increased, the proportion of antibody-forming cells showing labeling increased. With an interval of 30 hours, about half the antibody-forming cells were labeled and of 48 hours, over 95 per cent were labeled. This was taken as evidence that, few, if any, antibody-forming cells found at the height of a secondary response were more than 48 hours old. On the basis of these experiments and those reported in the accompanying paper, a simplified scheme showing the development of an antibody-forming clone in the secondary response was proposed.

The adoptive transfer of 2,4-dinitrophenyl(DNP)-keyhole limpet hemocyanin(KLH)-primed lymphocytes into a heavily irradiated allogeneic recipient permits the development of a secondary anti-DNP antibody response to DNP-bovine gamma globulin(BGG) whether or not the irradiated allogeneic host possesses BGG-specific helper T cells. This "allogeneic effect" has been demonstrated to result from the capacity of residual, apparently radioresistant, T cells in the irradiated host to exert an active effect on the transferred histoincompatible B lymphocytes. This conclusion derives from two corroborative experiments. In the first, an allogeneic effect was shown to occur on DNP-primed F(1) spleen cells that had been adoptively transferred to irradiated parental recipients; the second experiment demonstrated the development of an allogeneic effect on anti-theta-treated, DNP-specific donor cells transferred to irradiated allogeneic hosts. These results emphasize the extreme caution required in designing and interpreting experiments that may involve adoptive cell transfers into histoincompatible hosts, and illustrate why such models are unsuitable for investigation of the question of physiologic cooperative interactions between T and B lymphocytes. Suitable approaches are described in the accompanying paper.

In view of the risk of introduction of yellow fever into South-East Asia, comparative studies have been made of yellow fever vaccination in Malayan volunteers with a high prevalence of antibody to related viruses and in volunteers without related antibody. In a previous paper the neutralizing antibody responses of these volunteers were reported. The present paper describes the haemagglutinin-inhibiting (HI) antibody responses of the same groups of volunteers and discusses the relationship of these responses to the neutralizing antibody responses.The HI responses to yellow fever following vaccination closely paralleled the neutralizing antibody responses whether vaccination was subcutaneous or by multiple puncture. Volunteers with a high level of YF HI antibody due to infection with other group B viruses were found to be less likely to show a significant YF HI response than those without antibody. 90% of HI responses could be detected by the 21st day after vaccination.As with neutralizing antibody responses, volunteers given vaccine doses of 50-500 mouse intracerebral LD(50) subcutaneously gave greater responses than those given higher doses.

Streptococcal pyrogenic exotoxin (SPE) was shown previously to be a potent immunosuppressant. This paper presents data documenting that immunosuppression frequently was followed by elevated antibody and plaque-forming cell levels as noted in our earlier report. The data are interpreted as a differential effect of SPE upon suppressor T cells and secreting B cells.

Previous studies suggested that an active immune response was partially responsible for maintaining immunological unresponsiveness to sheep erythrocytes. Measurement of the plaque-forming (antibody-releasing) cell response proved to be a sensitive indicator of an immune response to sheep erythrocytes in the absence of detectable circulating antibody to the antigen. The present studies were undertaken to determine whether an active immune process, measured by the plaque-forming cell response, was partially responsible for induction and maintenance of tolerance. Rats injected intraperitoneally with large doses of sheep erythrocytes beginning at the day of birth develop tolerance to the antigen. In this paper, the plaque-forming cell and antibody response to sheep erythrocytes was characterized for rats receiving a single antigen injection at various ages and for rats which received repeated antigen injections as adults. The dose of antigen was the same as that used to produce tolerance; the injection schedule for repeated immunizations was also the same as that used to produce tolerance. Rats receiving a single antigen injection on the day of birth or at age 7 days had no measurable response to the antigen. Rats receiving a single antigen injection at age 17 days and sacrificed 4 days later had an unequivocal response to the antigen. The spleens had about one-tenth as many plaque-forming cells as spleens of adult animals immunized similarly, but the antibody titers were as high as titers for adult animals. Presumably the high titers of these young animals resulted from the high ratio of plaque-forming cells to body weight and blood volume. Adult animals receiving a single antigen injection had a peak or near peak plaque-forming cell response 4 days after immunization; at this time, sera contained high titers of 19S antibody and the numbers of plaque-forming cells in spleens correlated reasonably well with circulating antibody titers. 7S antibody appeared in serum 5 or 6 days after immunization. The numbers of plaque-forming cells declined progressively 2 and 3 weeks after immunization. Repeated twice weekly, injections of the antigen in adult rats produced a marked decline and then stabilization of numbers of plaque-forming cells in spleens. Although the numbers of plaque-forming cells were fewer, titers of 19S and 7S antibody stabilized at high levels. A progressive recovery of the plaque-forming cell response and a rise in antibody titer occurred when the interval between the last 2 injections was increased from 3 to 10, 17, or 32 days. These findings suggested that repeated closely spaced antigen injections interfered with either cell division or maturation of antibody-forming cells. As the interval between injections was increased, additional antibody-forming cells matured or were formed through cell division. Thus, relatively constant antigenic stimulation provided a mechanism for controlling or limiting the response of antibody-forming cells. In the following paper, it will be shown that this mechanism operating together with a homeostatic mechanism which prevents induction of new antibody-forming cells serves to regulate the immune response to repeated injections of antigen.

Rats thymectomized at birth gained weight and otherwise developed normally, but were found to be very susceptible to intercurrent infections. Both Arthus reactivity and delayed hypersensitivity to BSA were markedly impaired in rats thymectomized during the first week of life and significantly impaired in rats thymectomized as late as 3 weeks after birth. The inhibition of Arthus reactivity in thymectomized rats was well correlated with their failure to develop significant titers of precipitating or hemagglutinating antibody. However, natural heteroagglutinin titers were not altered in these animals, and no abnormality of serum proteins, including gamma-globulin could be detected by paper electrophoresis. The loss of immunologic activity could not be corrected by injecting homogenates of spleen or thymus before and during the sensitization period. Splenectomy at birth did not influence Arthus or delayed reactivity.

Peritoneal cells (PC) from normal, unimmunized mice were placed in ultra-thin monolayer cultures containing carboxymethylcellulose (CMC), sheep red blood cells (SRBC), and complement, and tested for the appearance of plaques of lysis. The behavior of PC from young male mice and from female mice that had given birth to several litters (retired breeder mice) was studied. It was found that cells from spleen, mesenteric lymph node, thymus, bone marrow, thoracic duct lymph, or Peyer's patches could not form plaques in the CMC microcultures. Also, various combinations of these cells did not lead to plaque formation. When cells from any of these sources were mixed with PC, there was either no effect or an actual inhibition of plaque formation, the plaque counts being lower than would have been expected from the number of PC present in the mixture. Optimal plaque formation by peritoneal cells was found to be dependent on an optimal cell concentration, this optimum being around 5 x 10(6)/ml for young male mice and 0.5 x 10(6)/ml for retired breeders. Inhibition of plaque formation was found with either supra- or suboptimal cell concentrations. The inhibition by excess cell concentration may have been a simple nutritional or nonspecific overcrowding effect, as it could also be induced by an addition of an excess of spleen or lymph node cells. The failure of more dilute PC preparations to give adequate numbers of plaques appeared to be more specific, as plaque numbers could not be restored to normal by addition of spleen cells. The suggestion was that some cell to cell interaction between PC was involved. This dependence on cell concentration was not seen with immunized spleen PFC. Plaque appearance could be specifically and reversibly suppressed by placing PC in a medium containing rabbit anti-mouse IgM serum. Anti-IgG serum had no such effect. These experiments strengthened our view, expressed in the accompanying paper, that plaque formation was due to the formation of IgM, hemolytic antibody to SRBC by the PC. Metabolic inhibitors were incorporated into monolayer cultures and had different effects with the different types of PFC used. In the case of spleen cells from mice actively immunized against SRBC 4 days before killing, actinomycin D had no effect on plaque counts and puromycin reduced plaque numbers by a factor of 2. In the case of PC from young male mice, actinomycin D in concentrations above 0.01 microg/ml caused reductions down to < 2% of control values in plaque counts, and puromycin (10 microg/ml) had a similar effect. The PC from retired breeder mice occupied an intermediate position between the two cases just discussed. A compartment of cells, equal to about one-fifth of the total normal PFC compartment, was identified as resistant to high concentrations of either actinomycin D or puromycin, being similar in these respects to PFC from spleens of intentionally preimmunized mice. The mitotic poison, Colcemid, did not affect plaque counts in any situation tested. The theoretical implications of these results are briefly discussed.

Transfer of live lymphoid cells from BGG-immunized strain 2 guinea pigs into syngeneic animals primed with DNP-OVA prepares the recipients for a markedly enhanced secondary anti-DNP antibody response upon challenge with DNP-BGG. This phenomenon has been demonstrated when the recipients were challenged 1 day after cell transfer, but it was considerably more striking when an interval of 6 days was allowed between transfer of cells and challenge with antigen. As demonstrated in the preceding paper (6), BGG preimmunization enhanced anti-DNP antibody responses to challenge with DNP-BGG. An analysis of the characteristics of substances to which preimmunization was effective in enhancing subsequent anti-hapten responses was made. It was shown that preimmunization of strain 13 guinea pigs with a copolymer of glutamic acid and lysine (GL), to which these animals are genetically unable to accord an immune response, failed to prepare them for an enhanced anti-DNP response to DNP-GL. Tolerance to BGG specifically abrogated the capacity of subsequent BGG immunization to prepare DNP-OVA primed rabbits for an enhanced anti-DNP response to DNP-BGG. Sensitization of animals to haptens by preimmunization with hapten-protein conjugates failed to prepare them for enhanced primary or secondary antibody responses to other determinants associated with that hapten on a different carrier. These studies indicate that the enhancing effect of carrier preimmunization reflects a cooperative interaction between lymphoid cells specific for carrier and cells specialized for haptenic determinants. Furthermore, the capacity of a substance to behave as a carrier requires more than its possession of a variety of antigenic determinants.

Shepel, Michael (U.S. Army Biological Laboratories, Fort Detrick, Frederick, Md.) and Maxwell R. Klugerman. Effect of adjuvants on antibody response of rabbits inoculated with Venezuelan equine encephalomyelitis virus. J. Bacteriol. 85:1150-1155. 1963.-Hemagglutination-inhibition, neutralization, and complement-fixation tests were performed on sera of rabbits inoculated with Venezuelan equine encephalomyelitis (VEE) virus in combination with Freund's adjuvants and in Hank's salt solution. This study indicated that the complete adjuvants (i.e., with mycobacteria) considerably increased the antibody response to VEE virus. Mycobacterium butyricum (M. smegmatis) appeared to be more effective than M. tuberculosis H37Ra. In the absence of mycobacteria, the response was much less pronounced. Paper electrophoretic studies of the antisera demonstrated a marked increase in gamma-globulin production, an increase in the beta-globulin, and an increase in total protein as the result of adding VEE virus to the complete adjuvants. A decrease in the albumin fraction appeared to be caused by the complete adjuvants rather than by the VEE virus itself. The incomplete adjuvant (without mycobacteria) plus virus contributed little, if any, stimulation toward the production of gamma-globulin, nor did it appear to affect the serum-albumin levels.

Regulated intravenous doses of quinidine were given to patients with the antibody of quinidine purpura to produce controlled thrombocytopenia without clinical sequelae. The degree of thrombocytopenia and the rate at which it developed were dependent on the relative plasma concentration of quinidine and antibody. By relating in vivo changes in platelet levels to concurrent in vitro tests for antibody activity and to quantitative relationships between reactants determined in Papers I and III of this series, it was concluded that the amount of antibody which attaches to platelets when thrombocytopenia develops is insufficient to cause complement fixation or platelet agglutination. Platelets do not appear to be destroyed directly by reaction with antibody in vivo. The minimal amount of antibody which does attach to platelets in vivo appears to increase their susceptibility to the usual mechanisms of sequestration. Megakaryocytes and blood vessels do not appear to be affected directly by the antibody which causes quinidine purpura, and hemorrhagic manifestations of the disease appear to be consequent to changes in platelets alone. A safe method of performing in vivo tests for the presence of an antibody of drug purpura is described. The implications of the present work in idiopathic thrombocytopenic purpura are discussed.

Antirabies vaccines produced from infected brains of adult mammals have always had the potentiality of causing post-vaccinal paralysis or allergic encephalitis in man. Attempts in recent years either to remove the paralytic factor from brain-tissue vaccines or to use as the virus source infected tissue other than nervous tissue (e.g., chick embryos) have usually resulted in a substantial reduction of the specific antirabies potency.The authors' laboratory had previously developed a vaccine made from infected suckling-mouse brains in which the virus was inactivated by ultraviolet irradiation. This vaccine was found highly potent in animal tests and low in organ-specific antigens. Others have found the brains of newborn mammals to be free of the allergic encephalitic factor. The studies reported in this paper show that the antirabies antibody responses to a 14-dose course of this suckling-mouse-brain vaccine in children are at a high level even when the vaccine is used at a 1% tissue concentration. There was no evidence of deleterious reactions to this treatment in 31 children.It is concluded that these results justify a long-term trial of this vaccine for antirabies prophylaxis in man.

Infection of mammalian cells with herpes simplex virus (HSV) results in the production of a number of virus-induced soluble antigens. Immunodiffusion analyses of the soluble antigen mixture (SAM) obtained from HSV-infected KB or BHK cells revealed at least six well-defined immunoprecipitin bands. Calcium phosphate chromatography (Brushite) was employed to separate one immunoprecipitin (designated CP-1) from the remaining viral and host antigens. We conclude that CP-1 is a viral-specific antigen because (i) specific antiserum, which had been repeatedly absorbed with uninfected cell extracts or serum components, still retained the capacity to react in gel diffusion with CP-1 antigen; (ii) anti-CP-1 serum reacted in gel diffusion with SAM, yielding one precipitin band in identity with the band formed against human gamma globulin; (iii) the CP-1 fraction stimulated the production of HSV-neutralizing antibody of high capacity. The last observation suggests that fraction CP-1 contains a biologically active structural component of the virus which is associated with the envelope. The CP-1 immunoprecipitin was separated from SAM by an alternative method by using a cyanogen bromide-linked immunosorbent prepared from anti-CP-1 gamma globulin. The observation that the CP-1 antigen isolated from the immunosorbent effectively blocked serum-neutralizing activity provided further evidence that neutralizing antibody was directed against CP-1. Acrylamide gel electrophoresis and immunological experiments suggest that the CP-1 antigen is in part a glycoprotein. The finding that CP-1 contains only one antigenic component of the virus will permit future biological studies to be made with a monoprecipitin antiserum. In addition, the techniques described in this paper represent initial steps in the purification of HSV antigens.