It was demonstrated previously that the cytolysis of murine viral lymphoma cells by anti-murine sarcoma virus (MSV) syngeneic T-killer lymphocytes was restricted by some products of the H-2 complex. The respective role of the products of different regions of the H-2 complex were studied with six H-2(b) and three H-2(d) lymphomas induced by five different type C viruses. They were tested in a classical chromium release test against anti-MSV T-killer cells obtained from different inbred strains of mice, including several H-2 recombinants. Tumors o£ the H-2(b) haplotype were lysed only when effectors and target cells have in common the D(b) region. On the contrary an identity limited to the K end of the H-2 complex is necessary and sufficient in the H-2(d) haplotype. An in vitro restimulation of the spleen cells with concanavalin A strongly increased the activity of in vivo-primed T lymphocytes but did not provide any response for in vivo-primed but nonresponder cells. Preincubation of the tumor cells with anti-H-2 sera abolished the lysis by syngeneic anti-MSV effector lymphocytes. The same results were obtained by preincubating the H-2(b) targets with anti-H-2D(b), or the H-2(d) target with anti-H-2K(d). Preincubation with anti-H-2K(b) or anti- H-2D(d) were ineffective. These results show that the T-killer/target cells interaction in the MSV system involved some products of the H-2 complex which might be different with the various H-2 haplotypes and could possibly vary according to the antigenic specificity. A specific association of a viral product with a normal cellular structure, directed by the H-2 region during the viral budding could explain the observed results.

Gross leukemia virus-induced cell surface antigen; /r, immune response; MHC, major histocompatibility complex; MLTC, syngenic mixed leukocyte-tumor cell culture; SV40, Simian virus-40.J. Exp. MEt).

Cytolytic T lymphocytes (CTL) specific for the virus-induced and leukemia-associated Friend, Moloney, Rauscher (FMR) antigen are easily detected in the spleens of primary and secondary stimulated H-2b or H-2d mice. They react, respectively, with H-2Db + FMR and H-2Kd + FMR; Dd and Kb never being involved. On the other hand, recombinant (KbDd) mice are relatively low responders that produce CTL only after secondary stimulation. Competition and blocking experiments with monospecific anti-H-2 antibodies have demonstrated that on the same H-2b tumor cells, C57BL/6 (H-2b) lymphocytes recognize Db + FMR, whereas B10.A(5R) lymphocytes recognize Kb + FMR, the restriction cannot, therefore be explained by a specific association of viral molecules with certain H-2 products. The CTL response of (B10 X 5R)F1 hybrids is (a) easily detected in primary reaction, the high responder anti-FMR phenotype being dominant and (b) directed against Db + FMR, F1 mice being low responder against Kb + FMR like the B10 parent. These results suggest that a D region-associated immune response gene controls the cell-mediated anti-FMR reaction, the best available H-2 + FMR antigenic association being chosen by CTL precursors.

SYNOPSIS Among the acute leukaemias of the granulocytic group, acute premyelocytic leukaemia is distinguished by the severity of its haemorrhages, the frequency of hypofibrinaemia, a rapidly fatal course, and an unusual cellular hyperplasia. Myelograms show an increased proportion (average 80%) of characteristic cells of large diameter, with numerous azurophil granules. The infiltration of other organs is inconstant. MATERIAL AND METHODS Acute premyelocytic leukaemia represents about 6-8 % of all cases of acute leukaemia seen at our Institute (Bernard, Boiron, Weil, Levy, Seligmann, and Najean 1962). Between 1956 and August 1962, 34 cases of acute premyelocytic leukaemia from a total of 497 cases of acute leukaemia were seen. We have chosen for this study 25 cases followed up until death with, in 12 cases, a full necropsy. The diagnostic criterion used was the presence of a very large proportion of cells of the premyelocyte type (over 50% and often 80 to 100%) in the bone marrow, associated with the clinical and haematological features of acute leukaemia. Cytological studies were made on bone marrow and blood smears stained by the May-Grunwald-Giemsa method. The histological features of the bone marrow and various viscera obtained at necropsy were examined in tissues fixed in Bouin's fluid and stained with haematooxylin-phloxin-saffron. HAEMATOLOGICAL STUDY The clinical characteristics of acute premyelocytic leukaemia were first described in a previous communication (Bernard, Mathe, Boulay, Ceoara, and Chome, 1959). Table I presents the principal haematological data on the present series of cases. Anaemia was generally intense with a red cell count at first examination of under 21 million per c.mm. in 18 cases. The leucocyte count was variable, but often diminished (less than 5,000 c.mm. in 17 cases). Premyelocytes were found in the blood in 21 of the cases, but their number showed considerable individual variations. The myelogram as a rule

A one-way mixed lymphocyte-tumor-cell reaction ( M L T R ) was applied to the study of anti-tumor cell-mediated immunity in M S V-infected mice. No stimulation of normal, non-immune lymphocytes by tumor cells was observed. Lymphoid cells which can be stimulated in vitro by various tumor cells of the same antigenic specificity as MSV-induced tumor cells were detected in the spleen and lymph nodes of MSV-infectedmice. These '' responder cells" were only found before the appearance of the tumor, and at a time ( 3 to 7 days after M S V inoculation) when no cytotoxic cells can be detected in the chromium-51 release test ( C R T ) . The reactivity of responder cells declines rapidly and remains undetectable thereafter, both when palpable tumors and cytotoxic cells are present and after tumor rejection, when none or very few CRT-killer cells are active. An inverse relationship between responder cells in M L T R and killer cells in C R T was also detected after allogeneic immunization; its significance is briefly discussed.