Rapid cytologic diagnosis of choroidal malignant melanoma by vitreous smear.

The eye is an uncommon subject of cytopathological examination. However, cytopathologic examination may be required for definitive diagnosis in some cases, as malignant tumors of the eye may sometimes be difficult to distinguish clinically from benign disorders. We report a case of malignant melanoma (MM) of the choroid, in which vitrectomy was performed for the initial clinical diagnosis of vitreous hemorrhage. As the dense vitreous hemorrhage was gradually cleared during the vitrectomy, a choroidal mass was discovered and the vitreous fluid was procured for rapid cytologic diagnosis. We used a modified Shorr's stain that can be completed within several minutes. With this method, highly atypical, pleomorphic cancer cells, occasionally associated with melanin pigment granules, were demonstrated. These cytologic findings indicated a diagnosis of MM arising from the choroid. Histologic examination of the enucleated eye confirmed MM of epithelioid type. The advantage and indication of the rapid cytologic diagnosis is discussed.

The calcium-independent protein kinase C participates in an early process of CD3/CD28-mediated induction of thymocyte apoptosis.

Thymocyte negative selection eliminates self-reactive clones and involves both a T-cell receptor (TCR)/CD3-mediated signal and a costimulatory signal, which can be delivered via CD28. Anti-CD3/anti-CD28-triggered apoptosis in isolated CD4+CD8+ thymocytes in vitro provides a basic model for negative selection. Effects of isoform-selective and non-isoform-selective inhibitors of protein kinase C (PKC) on this apoptotic process suggest that activation of Ca2+-independent PKC isoforms during the first 2-3 hr of culture is essential for inducing apoptosis, and that Ca2+-dependent PKC isoforms may be influential, but not essential, for apoptosis. To assess the CD3/CD28-mediated activation of PKC in the apoptotic process, we prepared CD4+CD8+ thymocytes (without contamination with cells that had received negative or positive selection signals in vivo) by establishing TCR-transgenic mice with RAG-2-deficient and non-selecting major histocompatibility complex (MHC) backgrounds, in addition to a CD4+CD8+ thymocyte-enriched population from normal mice. Translocation of Ca2+-independent PKC from the cytosolic fraction to the particulate fraction of CD4+CD8+ thymocytes was induced by CD3/CD28-mediated stimulation, but not by CD3- or CD28-mediated stimulation alone, and peaked 2 hr after the start of culture. The kinase activity of the translocated Ca2+-independent PKC was dependent on cofactors in vitro, indicating that novel (n)PKC, but not atypical (a)PKC or a proteolytic PKC fragment, was responsible for the activity. Immunoblotting analysis indicated that the nPKC-theta isoform was the major contributor among nPKC isoforms, and that the classical (c)PKC-alpha isoform was the major contributor among cPKC isoforms. These results suggest that activation of nPKC (especially the theta isoform) in CD4+CD8+ thymocytes is involved in a pathway for negative selection.

Calcineurin activation protects T cells from glucocorticoid-induced apoptosis.

In T cell hybridomas, TCR/CD3 complex-mediated stimulation induces apoptosis but inhibits that induced by glucocorticoids. A combination of ionomycin (IM), a calcium ionophore, and PMA, a protein kinase C activator, mimics the effects of the TCR/CD3-mediated stimulation. Glucocorticoid-induced apoptosis is, however, markedly inhibited by IM alone, and less markedly by PMA alone. The immunosuppressant FK506 canceled the inhibition by IM but not that by PMA. As calcineurin (CN) is one of the target molecules of FK506, we examined whether CN activation might have an anti-apoptotic effect. BOG8, a T cell hybridoma, was stably transfected with a mutant CN catalytic subunit with Ca2+/calmodulin-independent, constitutive but FK506-sensitive phosphatase activity. The transfectant clones were fairly resistant to glucocorticoid-induced death. Their resistance, however, was hardly affected by FK506 when added simultaneously with glucocorticoid, but was lost after a prolonged preincubation with FK506. In the parent BOG8 cells, FK506 failed to cancel the inhibitory effect of IM on glucocorticoid-induced death when the addition of FK506 was delayed for 1 h or more. These results suggest that CN activation is required for the resistance only as an early event. The transfectant clones produced IL-2 but failed to undergo apoptosis upon stimulation with PMA alone, whereas apoptosis was induced by a combination of IM and PMA. These results suggest that activation-induced cell death may require a higher level of CN activity than IL-2 production or may require another Ca(2+)-dependent pathway.

Involvement of protein kinase C-epsilon in glucocorticoid-induced apoptosis in thymocytes.

Apoptosis is induced in immature thymocytes by physiological peak levels of glucocorticoid hormones, especially in murine and rat cells. Endogenous glucocorticoids may have some role in thymic selection. Glucocorticoid-induced thymocyte apoptosis appears to be dependent on protein kinase C (PKC), since it is inhibited by PKC inhibitors. PKC is a family of closely related enzymes, consisting of Ca(2+)-dependent (PKC-alpha, -beta I, -beta II, and -gamma) and Ca(2+)-independent (PKC-delta, -epsilon, -eta (L), -theta, -zeta, and -lambda) isozymes. In the present study, we analyzed the role of PKC in glucocorticoid-induced apoptosis in murine thymocytes and found that glucocorticoid selectively induces an increase in Ca(2+)-independent PKC activity in the particulate fraction of immature thymocytes but not in that of mature T cells. The increase as well as the apoptosis was inhibited by actinomycin D, cycloheximide, or the glucocorticoid receptor antagonist, RU 38486. Immunoblotting studies revealed the selective translocation of PKC-epsilon from the cytosolic fraction to the particulate fraction upon glucocorticoid treatment. These results suggest that glucocorticoid-induced apoptosis in immature thymocytes involves glucocorticoid receptor-mediated and selective activation of PKC-epsilon through de novo synthesis of macromolecules.

Early mobilization of Ca2+ is not required for glucocorticoid-induced apoptosis in thymocytes.

Contradictory results have been reported on the question of the role of Ca2+ in glucocorticoid-induced apoptosis in thymocytes. To resolve this problem, we investigated the effect of dexamethasone, a synthetic glucocorticoid, on intracellular Ca2+ concentration ([Ca2+]i), by microscopic fluorometry that enables us to monitor real-time [Ca2+]i of cells loaded with fura-2, a fluorescent Ca2+ indicator, on a single cell basis. The results indicated that dexamethasone does not induce an increase in [Ca2+]i above control level both in murine and rat thymocytes at least for 1 h after the start of the culture. We also investigated whether the depletion of extracellular Ca2+ with EGTA or buffering intracellular Ca2+ with quin-2/AM inhibited glucocorticoid-induced apoptosis as reported on rat thymocytes. Dexamethasone-induced apoptosis in both murine and rat thymocytes, however, was not inhibited by EGTA. High concentrations (25 microM and over) of quin-2/AM inhibited DNA fragmentation, but failed to inhibit cytolysis. Calmodulin inhibitors, trifluoperazine and calmidazolium, also inhibited DNA fragmentation as reported, although they markedly enhanced cytolysis. Therefore, glucocorticoid-induced death is not inhibited by quin-2/AM or calmodulin inhibitors. Furthermore, we have previously found that a proper combination of the calcium ionophore, ionomycin, and the protein kinase activator, PMA, inhibits corticosterone-induced apoptosis. These results suggest that an early increase in [Ca2+]i is neither induced by glucocorticoids nor responsible for glucocorticoid-induced apoptosis in thymocytes.

Retinoic acids inhibit activation-induced apoptosis in T cell hybridomas and thymocytes.

Apoptosis is induced in immature thymocytes and T cell hybridomas upon stimulation via the TCR/CD3 complex. This phenomenon appears to be related to negative selection of T cell clones in the thymus. In T cell hybridomas, it has been shown that glucocorticoids inhibit TCR/CD3-mediated apoptosis, whereas glucocorticoids alone induce apoptosis. All-trans-retinoic acid (RA) at 0.1 to 10 microM also inhibited TCR/CD3-mediated apoptosis assessed by DNA fragmentation and cytolysis, but RA alone hardly induced apoptosis. RA enhanced the effects of glucocorticoids to induce apoptosis and to inhibit TCR/CD3-mediated apoptosis. TCR/CD3-mediated stimulation can be mimicked by the combination of ionomycin, a calcium ionophore, and PMA, an activator of protein kinase C, and the combination-induced DNA fragmentation was also inhibited by RA. RA, however, failed to inhibit the combination-induced increase in intracellular Ca2+ concentration or the combination-induced translocation of protein kinase C from the cytosolic fraction to the particulate fraction. Time course studies of RA addition into the culture indicated that a 3- to 6-h delay in the addition of RA did not reduce its inhibitory effect on anti-CD3-induced DNA fragmentation. These results suggest that RA interferes with the apoptotic process at some point after its initiation stage. It has been suggested that negative selection involves not only TCR/CD3-mediated signals but also LFA-1-mediated signals. RA at 0.01 to 1 microM significantly inhibited the induction of thymocyte apoptosis by co-immobilized mAb to CD3 and LFA-1 molecules. RA by itself hardly induced apoptosis, but enhanced glucocorticoid-induced apoptosis. The results suggest that thymic selection might be influenced by RA at near-physiologic concentrations. The receptors of glucocorticoids and RA belong to the erbA oncogene-related steroid hormone receptor superfamily. Thyroid hormones and 1 alpha,25-dihydroxy vitamin D3, whose receptors also belong to the superfamily, failed to modulate apoptosis in both T cell hybridomas and thymocytes.

Regulation of T lymphocyte apoptosis. Signals for the antagonism between activation- and glucocorticoid-induced death.

Apoptotic death can be induced in T cell hybridomas by glucocorticoids or the stimulation via the TCR/CD3 complex. The two apoptotic processes are mutually antagonistic. We have previously proposed that positive selection of thymocytes for the formation of the T cell repertoire might be based on a similar mechanism. We analyzed the TCR/CD3-mediated signals essential for the regulation of apoptosis in T cell hybridomas. We suggest that both an increase in the intracellular Ca2+ level and an activation of protein kinase C are essential for the TCR/CD3-mediated apoptosis, because we obtained the following results: 1) either reduction of extracellular Ca2+ concentration or addition of a protein kinase inhibitor, 1-(5-isoquinolkinelsulfonyl)-2-methylpiperazine or N-(2-(methylamino)ethyl)-5-isoquinolinesulfonamide, inhibited anti-CD3-induced but not dexamethasone-induced DNA fragmentation. 2) The combination of ionomycin and PMA, but neither one alone nor the combination of ionomycin and cyclic nucleotide analogs, induced DNA fragmentation. On the contrary, we suggest that only an increase in the intracellular Ca2+ level is essential for the inhibition of glucocorticoid-induced apoptosis, because ionomycin alone as well as the combination of ionomycin and PMA inhibited dexamethasone- but not anti-CD3-induced DNA fragmentation.

Lymphokine-activated killer (LAK) activity against autologous malignant tumors of the skin.

When peripheral blood mononuclear cells (PBMC) cultured with interleukin 2 (IL-2) develop the ability to lyse fresh tumor cells, such activity is referred to as lymphokine-activated killer (LAK) activity. In this paper, we examined LAK activity against the autologous skin tumors, malignant melanoma (MM), squamous cell carcinoma (SCC), and basal cell epithelioma (BCE), which have distinct clinical characteristics. Similar levels of LAK activity against Daudi 1A4, an NK resistant cell line were significantly obtained from all cancer patients. However, different levels of LAK activity against autologous tumor cells were found from three kinds of cancer patients using mixtures of autologous tumors and LAK. The levels of cytotoxicity were 29.8 +/- 7.0% in five MM, 15.9 +/- 4.9% in seven SCC, and 4.0 +/- 2.3% in five BCE patients at an effector/target ratio of 50/1. Allogeneic MM targets were lysed by LAK from all three types of cancer patients at similar levels, implying that LAK is not restricted to major histocompatibility complex (MHC) antigens. These results indicate that the levels of autologous LAK activity were significantly associated with the magnitude of clinical malignancy of the tumor targets, and suggested the selective lysis of tumor targets by LAK. NK activity of cancer patients bearing tumors was also examined prior to therapy. The levels of NK activity observed in MM patients were considerably lower than those in two other cancer patients.

Generation and characterization of lymphokine-activated killer cells against fresh human leukemia cells.

Lymphokine-activated killer (LAK) cells generated from 15 acute leukemia patients in remission showed significant levels of cytotoxicity against Daudi 1A4, a natural killer-resistant cell line. This indicates that lymphocytes of leukemia patients in remission could respond to interleukin-2 to generate conventional LAK cells. However, LAK cells caused lysis of autologous leukemia cells at considerably lower levels in seven out of the 15 patients, with the exception of one case (48.6% cytolysis). None of the remaining eight patients exhibited LAK activity against autologous leukemia cells. On the other hand, patients' LAK could lyse allogeneic leukemia cells including those resistant to autologous LAK. Thus, patients' LAK seem not to be defective in lysis of leukemia cells. In the cold target competition analysis, the binding of patients' LAK to leukemia cells could be inhibited by autologous and allogeneic leukemia cell competitors, implying that almost all leukemia cells could be recognized by patients' LAK. Most LAK cells from normal donors showed significant lysis of allogeneic leukemia cells, but some leukemia cells were found to be resistant to lysis. LAK cells against both leukemia cells and Daudi 1A4 were phenotypically heterogenous, and were predominantly observed in the T3- fraction in the precursor phase. In the effector phase, whereas LAK activity against leukemia cells was also predominantly shown in the T cell-depleted fraction, similar levels of LAK activity against Daudi 1A4 were found in both the T cell-depleted and -enriched fractions.

The interaction of recombinant IL-2 with human resting lymphocytes: blocking effects of monoclonal antibodies to IL-2 receptors.

Several lines of evidence suggest that subsets of resting lymphocytes naturally express interleukin-2 receptors (IL-2.R). Recombinant IL-2 (rIL-2) induced the enhancement of natural killer (NK) activity, the generation of activated killer (AK) cells, the proliferation of resting lymphocytes, and the production of interferon-gamma (IFN-gamma) in lymphocyte cultures. The subsets of lymphocytes mediating these responses appeared to be heterogeneous, but reside predominantly in nylon wool-passed non-T, non-B cells ("null cells" or T3- cells); in response to rIL-2, only Leu 11+T3- cells showed enhanced NK activity, and both Leu 11+T3- and Leu11-T3- cells showed predominantly AK activity, proliferation and production of IFN-gamma. These findings suggest that the T3- fraction (null cell fraction) contains predominantly cells expressing IL-2.R at the resting state. Unlike the case with activated T cells, however, none of these responses was blocked by any of three monoclonal antibodies to IL-2.R, including anti-Tac antibody at any dilution. These results indicate that IL-2.R on the resting T3- cells possess unique biological features compared to those on activated T or B cells. A most likely explanation is that T3- cells possess higher affinity IL-2.R than activated T or B cells. Other possibilities are also discussed.

The capacity of antigen-presenting cells is fully preserved in childhood cancer patients.

T cells from 19 out of 25 childhood cancer patients showed impaired proliferative responses to purified protein derivatives (PPD)-pulsed antigen-presenting cells (APC) although all of the patients had been immunized with BCG. To test whether such low responsiveness of T cells results from the dysfunction of T cells or from that of APC, the experiment was designed to assess the proliferative response of T cells from patients or their parents to PPD-pulsed APC from patients or parents. These combinations seem to be suitable to assess the activity of T cells or APC since at least partial identity of HLA-D/DR antigens is required for T cell-APC interactions. Although T cells from patients who showed low responsiveness to PPD failed to respond even to PPD-pulsed APC from parents, T cells from parents were able to respond to PPD-pulsed APC from patients as well as to autologous APC. These observations strongly suggest that the low responsiveness to PPD in childhood cancer patients results from the dysfunction of T cells, and the capacity of APC is fully preserved. In other words, it appears that the capacity of APC is not impaired by chemotherapy, neoplastic cells, or other factors. Suppressor T cells appeared not to be involved in such dysfunction of T cells.

Cream, a new coat-color mutant in the musk shrew.

A coat-color mutant was found in the wild musk shrew (Suncus murinus, Insectivora). Five musk shrews with gray pelage, the common coat color of this species, were captured in the village of Tambum near Jakarta, Indonesia. Two males and two females were transported to Japan and mated. Matings between one male and two females segregated several cream-colored offspring, a color that had never been seen before in this species. From the pedigree record and data on mating experiments, it was confirmed that this mutant coat color was expressed in the homozygote by an autosomal recessive gene designated cr, and at least three of the four wild shrews examined were carriers of this gene. The cr gene was associated with failure of normal pigmentation in the pelage and skin. The mutant shrews also showed some behavioral abnormalities.

Genetic variants found in plasma alpha-amylase and erythrocyte carbonic anhydrase in the musk shrew (Suncus murinus).

Genetic variants of plasma alpha-amylase and erythrocyte carbonic anhydrase in the musk shrew (Suncus murinus) were found by electrophoreses using cellulose acetate plates. It was demonstrated that phenotypic differences of alpha-amylase are controlled by two codominant alleles (Amy-1a and Amy-1b) at a single autosomal locus (Amy-1). The segregation data of the carbonic anhydrase phenotypes in the progeny supported the genetic theory of two codominant alleles (Car-1a and Car-1b) at a single autosomal locus (Car-1). The data suggested that there was no close linkage between the two loci, Amy-1 and Car-1. The Car-1 locus was fixed with one of the two alleles in each of the four lines, i.e. Nag, Oki , Tar and Jak originating from wild animals captured in Nagasaki and Naha and Tarama Island, Okinawa, Japan, and in Jakarta, Indonesia, respectively. Oki and Tar lines still showed segregation of the two alleles at the Amy-1 locus.

Genetic polymorphism of the vitamin D binding protein (Gc) in the musk shrew (Suncus murinus).

In the musk shrew (Suncus murinus), the electrophoretic bands in the post-albumin region were identified as vitamin D binding protein (Gc) by the [3H] vitamin D3 binding method. Three Gc phenotypes were distinguished from each other: a single faster band (Gc-A), a single slower band (Gc-B) and the double bands (Gc-AB). Results of mating experiments indicated that the Gc-A and Gc-B are controlled by two codominant alleles, Gca and Gcb at an autosomal locus (Gc), respectively. It was noticed that, in the Gc-AB phenotypes, the Gc-B band was constantly more intense than the Gc-A band in the protein staining. The same tendency was also observed between the homozygous Gc-A and Gc-B bands, and further, radioactivity of the Gc-B bound with [3H] vitamin D3 was about twofold higher than that of the Gc-A. These results suggest that the Gcb yields its protein product twofold more than the Gca. No cross-reaction between the shrew proteins and a rabbit anti-human Gc protein was observed.