Light and electron microscopy of the leucocytes of Crassostrea virginica (Mollusca: Pelecypoda)

Summary The fine structure of oyster leucocytes resembles to a great extent, that of typical eucaryotic cells. Organelles which have been described for the first time in this report are light granules, dense granules, protocentriole and X structure. Light microscopy reveals two morphological types of oyster leucocytes: agranular and granular. Based upon nuclear morphology and cytoplasmic compositions revealed in electron microscopy, at least three types of agranular and one type of granular cells are recognized.In the Giemsa-stained preparations, granular leucocytes exhibit three distinct types of cytoplasmic granules: refractile, dark blue, and pink, which presumably correspond to light granules Type A, B, and C seen in the electron micrographs. A granular leucocyte may contain one or more types of granules. Cytochemical investigations show that oyster leucocytes contain at least three hydrolytic enzymes: non-specific esterases, acid, and alkaline phosphatase. The latter two enzymes constitute 63% of the enzyme activity detected. These intracellular enzymes may be associated with the light granules and/or lysosome-like bodies.It is also demonstrated that the granular leucocyte population is significantly higher (P<0.001) in the oysters experimentally infected with Bacillus mycoides (72.19±4.71%) as contrasted with that of the controls (37.18±4.48%).Leucocytes in progressive stages of degeneration are also described.Contribution No. 71 from Marine Research Laboratory, University of Connecticut.The initial phase of this investigation was carried out at the Department of Zoology, Rutgers, The State University, New Brunswick, New Jersey, and supported by Public Health Service Research Grant AI-00781 from the National Institute of Allergy and Infectious Diseases of the National Institute of Health, awarded to Dr. L. A. Stauber. Supported by a grant from the University of Connecticut Research Foundation and Faculty Summer Fellowship to S. Y. Feng.     

Biochemical and functional characterisation of macrophage stimulating factors secreted by mitogen-induced goldfish kidney leucocytes

Mitogen-stimulated goldfish kidney leucocytes secrete a number of different macrophage activation factors (MAF) that induce profound physiological changes in macrophages. MAF produced by goldfish kidney leucocytes was characterised using fast performance liquid chromatography (FPLC) and bioassays that measured MAF-induced respiratory burst (RB) and nitric oxide (NO) responses of activated macrophages. Mitogen-induced fish kidney leucocyte supernatants were fractionated using gel permeation FPLC (GP-FPLC) and the ability of different fractions to induce NO or RB measured. A MAF of M(r) 50 kD, that induced a potent nitric oxide response in both a long-term goldfish macrophage cell line (GMCL) and in in vitro-derived fish kidney macrophages (IVDKM) was identified. The GP-FPLC partially purified 50 kD MAF activity occasionally induced significantly higher nitric oxide production than that of the crude MAF preparations. This increase in the NO-inducing activity was due to segregation of the 50 kD MAF from a novel macrophage deactivating molecule of M(r) 10-12 kD present in crude MAF preparations. This 10-12 kD molecule was shown to inhibit nitric oxide production in cytokine-activated goldfish macrophages. Mitogen-induced fish kidney leucocyte supernatants contained two distinct MAFs that induced the respiratory burst in GMCL and IVDKM: the 50 kD and 30 kD proteins. The partially purified 30 kD MAF primed goldfish macrophage for increased RB activity after only 6 h of treatment, and continued to augment the RB activity after 24 h of stimulation. In contrast, the GP-FPLC partially purified 50 kD molecule also primed the RB after only 6 h of stimulation, but subsequently deprimed the RB after 24 h of stimulation, an effect similar to that observed for crude MAF preparations. The 50 kD MAF activity was further purified using chromatofocusing FPLC (C-FPLC) using basic pH gradients and was shown to consist of two distinct NO-inducing molecules (> pI 9.3). Mitogen-stimulated fish kidney leucocytes secrete several factors that profoundly affect the anti-microbial responses of teleost macrophages and which undoubtedly are responsible for regulating teleost macrophage function in vivo.     

Large granular lymphocytes provide an accessory function in the in vitro development of influenza A virus-specific cytotoxic T cells

We report that large granular lymphocytes (LGL) have an accessory function in the development of cytotoxic T cells (Tc) through the production of soluble factor(s). LGL and T cells were separated on Percoll gradients and the ability of the separated and of the recombined LGL and T cells to generate influenza A virus-specific Tc activity was measured. When stimulated by virus-infected, irradiated, adherent cells, neither LGL nor T cells cultured separately produced Tc activity. When they were co-cultured, however, even if separated by a 0.22-micron pore size membrane, Tc responses were readily generated from the small T cell precursors and natural killer activity was maintained in the LGL. Thus, LGL were required as accessory cells for Tc responses to occur and the effect was mediated by a soluble factor(s). alpha-Interferon (IFN) was produced in cultures containing LGL and/or stimulating adherent cells, whereas gamma-IFN was only produced in cultures containing both LGL and T cells. Therefore, neither alpha- nor gamma-IFN appeared to be the LGL produced soluble factor that mediated the accessory effect of LGL on Tc responses.     

Chemotactic effect of human recombinant interleukin 2 on mouse activated large granular lymphocytes

Human recombinant interleukin 2 (hrIL-2) was demonstrated in vitro to be chemotactic for mouse large granular lymphocytes (LGL) activated in vivo by virus infection. Peritoneal exudate cells harvested from virus-infected mice were used as a source of LGL. LGL collected from mouse hepatitis virus-infected mice at 3 days postinfection were a source for NK 1.1 positive natural killer (NK)/LGL. LGL collected from mice treated with antiserum to gangliotetraosylceramide and infected with lymphocytic choriomeningitis virus for 7 days were used as a source for Lyt-2 positive cytotoxic T lymphocytes (CTL)/LGL. Both NK/LGL and CTL/LGL responded chemotactically to hrIL-2, purified IFN-beta, and to crude cell-free washout fluids collected from the peritoneal cavity of virus-infected mice. hrIL-2 had chemotactic activity for virus-elicited granular and agranular lymphocytes but did not attract the contaminating macrophages, in contrast to IFN-beta, which displayed chemotactic activity for virus-elicited granular and agranular lymphocytes as well as macrophages. The migration to hrIL-2 was inhibited by a monoclonal antibody (7D4) to the IL-2 receptor, but treatment with 7D4 did not affect migration in response to IFN-beta. Microscopic examination of Wright's-Giemsa-stained migrated NK/LGL and CTL/LGL revealed that the majority of migrated LGL in either LGL population had a blast cell morphology (enlarged cells with rich basophilic cytoplasm). The frequency of cells bearing the LGL morphology within the virus-elicited nonadherent peritoneal exudate cell population was on incubation in vitro, stabilized by either hrIL-2 or IFN-beta. These data suggest that another important immunomodulating function of IL-2 may be to attract activated NK/LGL and CTL/LGL to sites of inflammation.     

Large granular lymphocytes are central cells in the interleukin-2-dependent differentiation pathway of natural killer cells

Peripheral blood low-density cells were sorted, with respect to their ability to accumulate the lysosomotropic agent mepacrine (Mep), into lysosome-rich (Mep+) and lysosome-poor (Mep-) cell populations. Cells of large granular lymphocyte (LGL) morphology and phenotype were found in the Mep+ but not in the Mep- cell population. The latter cells lacked any natural killer (NK) activity. Cultures of the Mep- cells resulted in the appearance of cells showing K-562 lytic activity, LGL morphology and CD16 and/or Leu-7 positivity. This process was facilitated by the supplementation of the culture with recombinant human interleukin-2 (rIL-2). Mep+ cells retested after 7 days of culture showed a decline in the fraction of granular (LGL and Mep+) cells. This decrease was less pronounced but also seen in rIL-2-supplemented cultures. In spite of the lower number of typical LGL, Mep+ cells cultured with rIL-2 were mostly large but scarcely granular; rIL-2-activated K-562 killing (rIL-2 AK) of originally Mep+ cells was much higher than K-562 lytic activity of these cells at the beginning of the culture, and as compared to rIL-2 AK of Mep- cells. From this finding it is apparent that the most active rIL-2 AK cells originate from low-density granular (Mep+) cells (LGL) and, therefore, we propose to call them 'giant' NK cells. Furthermore, in the presence of rIL-2, LGL differentiate from agranular (Mep-) low-density cells. In view of these data, LGL appear to be resting cells on the differentiation pathway of NK cells.     

Leucocyte myosin and its location in the cell.

The intracellular location of the binding site of antibody against purified myosin prepared from equine leucocytes was investigated in neutrophils and lymphocytes by electron microscopy using peroxidase-labelled antibody method. The myosin extracted from equine leucocytes could bind skeletal muscle F-actin and the formed complex showed the biophysical and biochemical properties and electron microscopic appearance of actomyosin. On immunodiffusion, the leucocyte myosin formed a single precipitin line with its antibody prepared in rabbits. The antibody also formed single precipitin lines with myosins from lymphocytes and thrombocytes, fusing with each other. The antibody against the leucocyte myosin did not react with myosins from skeletal or arterial smooth muscle. The specificity of the antibody was further established by determination of K+-EDTA-activated ATPase activity remained in the supernate of antigen-antibody mixture. Under electron microscope, the intracellular immunoreactive products of peroxidase labelled antibody were found in cytoplasm of neutrophils and lymphocytes incubated with antibody against leucocyte myosin, but not in neutrophils or lymphocytes treated with IgG from normal rabbits.     

Isolation of leucocytes from the anterior kidney and spleen of rainbow trout in a self-generating density gradient

Separation of leucocytes from tissues and enrichment of specific types of leucocytes are essential first steps in studies of leucocyte function. We describe a simple and rapid method for separating and enriching leucocytes from the anterior kidney and spleen of rainbow trout. Leucocytes were separated on self-generating density gradients of Sepracell-MN, a colloidal silica-based medium. Recovery of leucocytes from the gradients was near 100% and was not affected by procedural variables such as cell suspension: Sepracell-MN ratio (separation ratio), initial temperature, centrifugation time, or number of cells per gradient. Two bands of leucocytes formed at separation ratios of 1 : 1, 1:1.5, 1:2 and 1 : 3. Recovery of selected leucocyte types could be maximized, and contamination by other cell types reduced, by selection of the appropriate separation ratio and cell band. Recovered leucocytes were responsive in assays of functional capability (adherence, phagocytosis, superoxide anion production). Leucocyte populations that adhered to glass were enriched for macrophages, but some neutrophils and lymphocytes (paricularly spleen lymphocytes) were also adherent. The most abundant leucocytes in the anterior kidney and spleen of the experimental fish were lymphocytes (respectively, 47 and 88%, including thrombocytes), neutrophils (35 and 7%), and macrophages (11 and 3%).     

Fractionation, morphological and functional characterization of effector cells responsible for human natural killer activity against cell-line targets

Human natural killer cells cytotoxic against cell-line target cells (NK-CLT) were isolated and characterized by utilizing adsorption-elution of the effector cells from the K-562 target cells. The cell associated with the cytotoxicity was a large lymphocyte with pale and characteristically granular cytoplasm. Thus, its morphology was identical with that of the large granular lymphocyte (LGL) previously shown to be the principal cytotoxic NK cell against fetal fibroblasts (NK-FF). The association of LGL with natural killer activity was verified with contact analysis from mixtures of unfractionated effector cells and target cells, which revealed that the number of contact of LGL with K-562 was correlated to the level of the individually expressed intensity of natural cytotoxicity. The ANAE-staining distribution of LGL was intensively positive with granular or diffuse staining pattern. In direct surface marker analysis LGL were E-rosette forming but, in contrast to NK-FF, heterogenous in regard to the Fc receptors. During in vitro incubation after elution from the target cells, the cytotoxic activity of LGL increased several fold. Also, the presence of K-562 among unfractionated effector cells caused an augmentation of cytotoxicity. This phenomenon was not observed as a result of effector cell-fetal fibroblast coculturing. Evidence from fetal fibroblast adsorption-elution and aggregated IgG blocking experiments suggested that the LGL with strong expression of Fc receptors were initially cytotoxic “mature” NK-cells, whereas the LGL with a weak expression of Fc receptors were initially noncytotoxic, but contact with K-562 “augmented” or “recruited” them to nonselective cytotoxicity.     

Lymphokine-activated killer cells in rats. IV. Developmental relationships among large agranular lymphocytes, large granular lymphocytes, and lymphokine-activated killer cells

The developmental relationships among large agranular lymphocytes (LAL) large granular lymphocytes (LGL) and the activation of these cells into lymphokine-activated killer (LAK) cells by rIL-2 was investigated. Highly enriched populations of LAL were isolated from Fischer 344 spleen cells by a combination of nylon-wool filtration (to remove B cells and macrophages), treatment with a pan T cell antibody plus complement (to remove T cells) and incubation in L-leucine methyl ester (to remove LGL). The resultant cells were highly enriched in morphologically identifiable LAL which expressed asialo GM1 and partially expressed the OX8 surface marker. The enriched LAL did not contain detectable NK cytotoxic activity, did not express pan T cell (OX19), Ia, Ig, or laminin surface markers and contained less than 0.2% LGL. Incubation of LAL in a low dose of rIL-2 (100 U/ml) induced the generation of LGL having NK activity within 24 h of culture. Longer culture periods (48 h) resulted in a continued increase in the percentage of LGL and higher levels of NK activity. However, with this low dose of rIL-2, little or no LAK activity (i.e., reactivity against NK-resistant target cells) was generated. With a high dose of rIL-2 (500 U/ml), LAL responded by first generating LGL with NK activity (within 24 h), with subsequent generation of LAK activity by 48 h. Evidence that the development of granular lymphocytes from LAL was responsible first for NK activity and then LAK activity was demonstrated by depletion of the generated granular NK or LAK effector cells by second treatments with L-leucine methyl ester. Concomitant with the induction of LGL with NK or LAK activity, rIL-2 also caused LGL to proliferate and expand four- to five-fold in 48 h. This occurred in the presence of high or low dose rIL-2. These results indicate that LAL are the precursors of LGL/NK cells, that LAL, LGL/NK cells and LAK cells appear to represent sequential developmental or activation stages and that LAL may comprise major source of LAK progenitors in lymphoid populations having few LGL or mature active NK cells.     

Production of hemopoietic growth factors and gamma-interferon by large granular lymphocytes from patients with T gamma lymphocytosis

Two patients with T gamma lymphocytosis in whom expanded large granular lymphocyte (LGL) populations were detected in peripheral blood and bone marrow are described. The surface antigen phenotypes of the LGL from these patients were similar with a major portion of cells carrying T3, T8, T11 and Leu7 markers. However, whereas fresh LGL from both patients demonstrated antibody-dependent cell cytotoxicity (ADCC), natural killer (NK) cell function was present in one case but absent in the other. Supernatants from enriched suspensions of the LGL unstimulated by exogenous antigen or mitogens were shown to contain significant amounts of hemopoietic growth factors colony-stimulating activity (CSA) and burst-promoting activity (BPA). In one case gamma-interferon was also detected. This study contributes to the accumulating evidence that LGL are able to generate factors which have the capacity to influence the proliferation of hemopoietic progenitor cells in vitro.     

An electron microscope study of the circulating leucocytes of the marine mussel,Mytilus edulis

Circulating leucocytes of the mussel, Mytilus edulis, were studied by electron microscopy. Based on morphological criteria, the leucocytes were classified as agranular and granular leucocytes, dependent upon the presence or absence of specific granules in their cytoplasm. Furthermore, the existing literature is being critically revised, and it is suggested that agranular and granular leucocytes might belong to the same cell line.     

A basal membrane-like structure surrounding tumour nodules may prevent intraepithelial leucocyte infiltration in colorectal cancer

Epithelial tumours consist of an epithelial compartment and a stromal compartment, which are sometimes separated by a basal membrane-like structure. We sought to determine whether these factors have prognostic value in 84 curatively resected stage II and III colorectal cancer by immunohistochemically staining tumours for leucocytes (CD45) and extracellular matrix, and to assess the presence of a basal membrane-like structure. Leucocyte infiltration was also assessed in hematoxylin-eosin (HE) stained sections. Most leucocytes were located in the tumour stroma. A relatively high intraepithelial leucocyte infiltration was significantly correlated with a lower level of tumour recurrence (P=0.03) and a longer disease-free survival (P=0.05), whereas leucocytes located in the tumour stroma (P=0.92) or at the advancing margin (p=0.06) were not. Intraepithelial leucocyte infiltration was also significantly correlated with leucocyte infiltration in the tumour stroma (P=0.02) and at the advancing tumour margin (P=0.005), and as assessed in HE-stained tumour sections (P=0.05), but each of these parameters on its own did not have a prognostic value in predicting disease-free survival. Moreover, the presence of a basal membrane-like structure surrounding the tumour epithelium was inversely correlated with the number of intraepithelial leucocytes (P=0.05), suggesting that this membrane-like structure functions as a barrier to intraepithelial leucocyte infiltration. We conclude that leucocytes must be in the direct vicinity of tumour cells to affect tumour growth. The presence of an extracellular matrix barrier seems to prevent this interaction.     

Modulation of adhesion molecules on human large granular lymphocytes by interleukin-2 in vivo and in vitro.

The modulation of adhesion molecules on human large granular lymphocytes (LGL) by interleukin (IL)-2 was investigated both in vivo and in vitro. Intercellular adhesion molecule-1 (ICAM-1; CD54) expression increased on LGL of cancer patients receiving IL-2 adoptive immunotherapy. ICAM-1 expression on LGL isolated by Percoll gradient centrifugation, LGL purified, and expanded by adherence to plastic surfaces and LGL identified by Leu 19 (CD56) monoclonal antibody were increased significantly in response to IL-2 in vitro. Exposure of LGL to IL-1, interferon (IFN)-gamma, and tumor necrosis factor (TNF) in vitro did not induce ICAM-1. The expression of LFA-1 (CD11a/CD18), a receptor for ICAM-1, and other leukocyte adhesion molecules, including Mac-1 (CD11b/CD18) and p150,95 (CD11c/CD18), was only maintained by IL-2. IL-2 induction of ICAM-1 and the maintenance of CD18 complex expression on small lymphocytes separated by Percoll gradients were similar to that on LGL. We conclude that IL-2 enhances the expression of ICAM-1 on multiple human lymphocyte populations including LGL effectors. Expression of the CD18 complex on LGL does not appear to be highly regulated by IL-2. These findings may have implications relevant to the role of these adhesion molecules in the activities of LGL modulated by IL-2.     

One-signal requirement for interferon-gamma production by human large granular lymphocytes

This laboratory has been investigating IFN-gamma gene expression by highly purified human large granular lymphocytes (LGL) and T cells. We report here that within 1 hr after interleukin 2 (IL 2) treatment of freshly isolated human LGL, IFN-gamma mRNA can be detected, with IFN-gamma protein in the culture medium within 4 to 6 hr of treatment. CD3- Leu-11+ LGL require only a single signal for IFN-gamma production because phytohemagglutinin (PHA), phorbol myristate acetate (PMA), IL 2, or ionomycin can each independently induce IFN-gamma production. In addition, PHA and ionomycin (but not IL 2) show significant synergy with PMA as a stimulus to LGL. In contrast, CD3+ T cells require two stimuli for high levels of IFN-gamma production, and not only are PMA plus ionomycin or PHA synergistic, but in addition, IL 2 and PHA demonstrate some synergy. Furthermore, we have found by fractionation of peripheral blood lymphocytes that IL 2-induced IFN-gamma production is associated with the LGL population and not T cells. These results indicate that with certain stimuli, LGL may be the predominant source of IFN-gamma from peripheral blood lymphocytes.     

Tumor necrosis factor induction by Candida albicans from human natural killer cells and monocytes

Other investigators have previously reported that TNF has been induced from macrophages by bacteria and, more recently, from NK cells by certain tumor cells. Sendai virus has also been reported to induce TNF from macrophages. We report here that an opportunistic fungi, Candida albicans, can also induce TNF, not only from human monocytes, but also from Percoll-fractionated large granular lymphocytes (LGL) which mediate NK function. Incubation of monocytes of LGL with C. albicans for 8 h was sufficient for detection of TNF release and peak induction was observed at 24 h. Induction of TNF from LGL did not require the participation of monocytes or T cells because treatment of the LGL with CD14 or CD15 to eliminate contaminating monocytes and CD3, CD4, or CD8 to eliminate contaminating T cells did not decrease the level of TNF produced from the treated LGL. Small T cells recovered from the denser fractions of the Percoll gradient had no ability to produce TNF, even when 10% monocytes were added to the T cells to provide accessory function. The phenotype of the TNF-producing LGL was CD2+, CD11+, CD16+, NKH1+, LEU7-. The TNF produced by both monocytes and LGL was neutralized by specific monoclonal and polyclonal anti-TNF but not by monoclonal antilymphotoxin. These results indicate that TNF production is a normal response of monocytes and LGL to stimulation by fungi such as C. albicans and that the release of TNF may be related to its ability to activate effector function to control Candida growth, which we have shown earlier for neutrophils with TNF.     

Analysis of T cell receptors in highly purified rat and human large granular lymphocytes (LGL): lack of functional 1.3 kb beta-chain mRNA

Recent reports have suggested that the T cell receptor for antigen is somehow involved in the cytotoxicity of natural killer (NK) cells. However, we now report that highly purified, freshly isolated large granular lymphocytes (LGL) from both the human and rat, as well as LGL cultured in the presence of recombinant IL 2, express only the 1.0 kb beta-chain mRNA. The lack of a 1.3 kb mRNA, indicative of a functional beta-chain rearrangement, strongly suggests that a functional T cell receptor beta-chain is absent in freshly isolated LGL, thus making it extremely unlikely that this molecule is involved in target cell recognition by NK cells. These results also suggest that LGL are derived from a lineage distinct from T cells or develop before a functional rearrangement of the T cell receptor beta-chain.     

In vitro migration of rat large granular lymphocytes

Rat peripheral blood large granular lymphocytes (LGL) were isolated by fractionation on discontinuous Percoll gradients. LGL migration was studied using nitrocellulose filters. Rat LGLs migrated into nitrocellulose filters in response to N-formyl-methionyl-leucyl-phenylalanine (f-MLP), casein, and serum components. Percoll-enriched high-density lymphocytes had small, but significant, migratory capacity in response to stimuli under these conditions. Removal of OX-19+ contaminating cells by panning confirmed the migratory capability of rat LGL/NK cells under these conditions. Checkerboard analysis of the LGL response to chemoattractants revealed that induction of migration involved chemokinesis although a chemotactic component was also discernible. The prompt migration of rat LGL in response to different stimuli is consistent with the hypothesis that these cells may represent one of the first easily mobilizable lines of resistance against noxious agents. In the rat combined in vitro/in vivo studies may provide a better understanding of the regulation of LGL recruitment and extravasation.     

The effect of dexamethasone on the occurrence and ultrastructure of rodlet cells in goldfish

In the bulbus arteriosus of goldfish Carassius auratus a subendothelial 'reticular space' was comprised of reticular-like cells, neutrophils, medium-sized granular leucocytes, phagocytes, endothelial cells and reticulo-endothelial cells; herein, rodlet cells (RCs) accumulated and probably matured. Four classes of RCs were identified: (1) Those occurring in the reticular space and contacting reticular-like cells; (2) those present in the reticular space contacting reticular-like and endothelial cells; here the RCs were discharging; (3) those occurring in the bulbar lumen anchored to lining endothelial cells or other RCs; (4) those free in the bulbar lumen. There was no significant difference in the number of RCs among control and saline- and dexamethasone-treated fish, but significant differences were found in RCs' class partition and in their discharge modality. In dexamethasone-treated goldfish class (3) RCs were almost absent (2% v. 10% of the other experimental groups), moreover 35% of the RCs developed a 'bleb' discharge modality rather than the most typical 'koronenartigen konfiguration' which developed in the vast majority (90%) of the other experimental groups.     

Modulation of the locomotory capacity of human large granular lymphocytes

The regulation of the migratory capacity of Percoll-purified large granular lymphocytes (LGL) into nitrocellulose filters was studied in a 2-hr assay with the use of modified Boyden chambers. Compounds that stimulate the natural killer cytotoxic function of LGL, such as interferons (natural beta, recombinant alpha A, recombinant hybrid alpha A/D, recombinant gamma), recombinant interleukin-2, and inactivated streptococci (OK 432), augmented the capacity of LGL to penetrate into filters spontaneously in the absence of chemoattractants in the lower compartment of the chamber. These compounds did not increase the LGL responsiveness to chemoattractants. Phorbol 12-myristate 13-acetate did not appreciably affect the locomotory capacity of LGL but augmented their cytotoxic activity. Thus the cytotoxic function and locomotion of LGL in response to biological response modifiers can be dissociated.     

Generation of large granular T lymphocytes in vivo during viral infection

Cytolytic lymphocytes were isolated from the spleens of lymphocytic choriomeningitis virus (LCMV)-infected mice and were characterized in regards to function, cell size, antigen phenotype, and cell morphology. Only 2% of the Lyt-2+ cells from uninfected mice were large granular lymphocytes (LGL), whereas 21% of the Lyt-2+ cells isolated 7 days postinfection were LGL. The day 7 Lyt-2+ populations contained all of the LCMV-specific, class I histocompatibility antigen-restricted cytotoxic T lymphocyte (CTL) activity, but no natural killer (NK) cell activity. The NK cell activity was consistently recovered in Lyt-2- populations isolated from both control mice and mice on day 7 postinfection. The LGL isolated on day 7 postinfection were concluded to be predominantly T cells and not NK cells because 1) the proportions of LGL in fractionated cell populations 7 days postinfection correlated with levels of CTL-mediated lysis but not NK cell-mediated lysis, 2) they were recovered in the Lyt-2+ population, and 3) antibody to asialo GM1, known to eliminate NK cell-mediated lysis but not T cell-mediated lysis, dramatically reduced NK cell LGL numbers in vivo on day 3 postinfection but only marginally affected LGL numbers on day 7. Virus-induced inflammation elicited a 50-fold increase in LGL numbers in the peritoneum on day 7 postinfection. The peritoneal exudate LGL were also associated with CTL activity and were resistant to treatment with antibody to asialo GM1. These results indicate that in vivo-generated CTL have the morphology of LGL and that the appearance of cytoplasmic granules correlates with the ability of cells to mediate lysis. To focus on cells being stimulated during infections, activated blast cells were separated from small resting cells by centrifugal elutriation. Coincidental with the peak in overall spleen leukocyte cytotoxic activity, the peaks of blast NK cells and CTL were at days 3 and 7 postinfection respectively. More than 50% of the blast lymphocytes isolated on either day 3 or day 7 postinfection were LGL. The CTL activity in the blast populations on day 7 postinfection was mediated by Lyt-2+ cells, and 37 to 64% of these Lyt-2+ blast cells were LGL. Cytolytic NK cell and CTL LGL could not be distinguished by morphology or by cell densities, because they overlapped in low density Percoll gradient fractions. Since this technique has been used to enrich for LGL, these data indicate that heterogeneity in LGL populations may result from the presence of both CTL and NK cell LGL.