ETO and AML1 phosphoproteins are expressed in CD34+ hematopoietic progenitors: implications for t(8;21) leukemogenesis and monitoring residual disease

To study acute myelogenous leukemia 1 (AML1) transcription factor, ETO protein, and t(8;21) AML chimeric AML1/ ETO protein in normal hematopoiesis and in leukemia, we raised rabbit antisera to a bacterially expressed polypeptide containing amino acid residues 1 to 220 of ETO and to synthetic peptides extending from residues 528 to 548 of ETO and 32 to 50 of AML1. The latter was selected to have little chance of cross-reactivity with other members of the PEBP2 alpha family. With affinity-purified reagents, we observed immunofluorescent staining for both AML1 and ETO in the nucleus of HEL, K562, and Kasumi-1 leukemic cell lines, the last from a t(8;21) AML. Biochemical analysis confirmed specificity of the antibodies and the nuclear localization of the antigens, the latter being exclusive for AML1 and primary for ETO. Immunoprecipitations of metabolically labeled 32P-proteins from Kasumi-1 cells show that AML1 and ETO are phosphorylated on serine and threonine. Investigations with normal bone marrow reveal AML1 and ETO are coexpressed in megakaryocytes and that each is expressed in a portion of the approximately 10-microns-diameter cells residing there. Using a CD34+ enriched population mobilized to peripheral blood, we found AML1 and, unexpectedly, ETO present in these cells. Because of this, we conclude that the expression of ETO in hematopoietic cells is not by itself leukemogenic. Also, because ETO would not be exclusively expressed as part of chimeric AML1/ETO in leukemic patients, its presence cannot be used to monitor t(8;21) AML residual disease.

The ETO portion of acute myeloid leukemia t(8;21) fusion transcript encodes a highly evolutionarily conserved, putative transcription factor

The 8;21 translocation, t(8;21)(q22;q22.3), is seen only in acute myelogenous leukemia and is characteristically associated with the M2 subtype. Subsequent to our identification of the t(8;21) breakpoint region on chromosome 21, we reported that the translocation results in the fusion of the AML1 gene on chromosome 21 with a novel gene on chromosome 8 which we called ETO (for eight twenty-one). Recently, the AML1 portion of the fusion protein has been shown to correspond to the DNA-binding and dimerization domains of the mouse gene, polyoma enhancer binding protein 2 alpha B (pebp 2 alpha B). We report here the complete sequence of the ETO portion of the fusion transcript as compiled from complementary DNAs from a t(8;21) AML patient and compare this with the ETO sequence from a mouse brain transcript. The deduced amino acid sequences are 99% identical. ETO has several features consistent with it being a transcription factor. The ETO sequence is different from the portion of PEBP 2 alpha B it replaces in the AML1/ETO fusion protein, except for their common high content of proline, serine, and threonine residues. Because neither the putative zinc fingers nor the TAF110 homology domain of ETO is present in PEBP2 alpha B, one might expect functional differences in the ability of AML1/ETO protein to affect the levels of transcription of genes normally regulated to some degree by AML1 (PEBP2 alpha B) during myeloid differentiation. The relatively high levels of ETO in developing brain suggest that it could be involved in the regulation of some aspect of neural proliferation or differentiation.

Sequence of cDNA for rat cystathionine gamma-lyase and comparison of deduced amino acid sequence with related Escherichia coli enzymes

A cDNA clone for cystathionine gamma-lyase was isolated from a rat cDNA library in lambda gt11 by screening with a monospecific antiserum. The identity of this clone, containing 600 bp proximal to the 3'-end of the gene, was confirmed by positive hybridization selection. Northern-blot hybridization showed the expected higher abundance of the corresponding mRNA in liver than in brain. Two further cDNA clones from a plasmid pcD library were isolated by colony hybridization with the first clone and were found to contain inserts of 1600 and 1850 bp. One of these was confirmed as encoding cystathionine gamma-lyase by hybridization with two independent pools of oligodeoxynucleotides corresponding to partial amino acid sequence information for cystathionine gamma-lyase. The other clone (estimated to represent all but 8% of the 5'-end of the mRNA) was sequenced and its deduced amino acid sequence showed similarity to those of the Escherichia coli enzymes cystathionine beta-lyase and cystathionine gamma-synthase throughout its length, especially to that of the latter.

The binding of a monoclonal antibody reactive with pp60v-src to the rat CNS both in vitro and in vivo: evidence that the epitope is present intracellularly as well as being associated with a number of antigenically related polypeptides located externally in the plasma membrane only in the synaptic region

A monoclonal antibody, F4, has been produced which reacts with an epitope possessing an unusual subcellular distribution. It binds to the external surface of the neuronal plasma membrane only in the region of the synapse. This is evidenced by binding of F4 to presynaptic terminals in unfixed cultures of rat cerebellum and to preparations of unfixed synaptosomes. In addition, much larger amounts of the epitope are present intracellularly. In fixed nervous tissue, the epitope is found in many neurons, and is associated mainly with presynaptic plasma membranes, synaptic vesicles, postsynaptic densities (cerebral cortex and hippocampus, but not cerebellum), rough endoplasmic reticulum, and the Golgi apparatus. The epitope is especially abundant in large neurons (e.g. pyramidal cells). Similar amounts of epitope are present in the chromaffin cells of the adrenal medulla. It is also expressed in ependymal cells in the brain, and in epithelial cells present in ducts of the medulla, but not cortex, of the kidney. However, the epitope is not found in glial cells in the brain, or in either liver, spleen, skeletal muscle, or testes. F4 is not species specific, as it binds to postmortem adult human cerebral cortex and neonatal cerebellum in a manner as described for the rat. It also binds to homogenates of brains of fish, chicken and mouse. The appearance of the epitope during development of the cerebellum in vivo and in vitro occurs in parallel with the differentiation of neurons and formation of synapses, though small amounts are also present in neuronal precursor cells. The F4 antibody can detect nanogram amounts of pp60v-src on immunodots. The strength of this reaction is high enough that F4 can be used to demonstrate pp60v-src-like immunoreactivity in Rous Sarcoma virus-transformed chick embryo fibroblasts. However, present evidence suggests that it may be premature to assign the immunocytochemical reactivity of F4 in the brain exclusively to pp60c-src. This conclusion is based on the fact that F4 reacts with several polypeptides from synaptic plasma membranes on Western blots of renaturing, two-dimensional gels that are dissimilar in size to pp60c-src, and from the fact that it can cross-react, albeit weakly, with several other serine protein kinases in an immunodot assay. Appreciation of this cross-reactivity, and of the evolutionary conservation of the epitope, as well as its sensitivity to denaturation, has led to our working hypothesis that F4 binds to a conformational epitope present on several polypeptides that may be most perfectly represented by some aspect of the catalytic domain of tyrosine protein kinases.

Quantitative electrophoretic transfer of polypeptides from SDS polyacrylamide gels to nitrocellulose sheets: a method for their re-use in immunoautoradiographic detection of antigens

Polypeptides (233 micrograms) from brain synaptosomes resolved in SDS 7.5%--15% polyacrylamide gradient gels were electrophoretically transferred quantitatively to nitrocellulose sheets at 3 V/cm for 21 h in the following buffer: 25 mM Tris-192 mM glycine, pH 8.3/20% methanol/0.1% SDS. After immunoautoradiographic detection of antigens on this nitrocellulose replica with either rabbit anti-rat cerebrum immunoglobulins or a mouse monoclonal antibody to a rat synaptosomal protein, the antibody and [125I]protein A were removed from the replica by treatment with 8 M urea, 0.1 M 2-mercaptoethanol, and 5 mg/ml BSA at 60 degrees C for 1 h. The replicas were successfully re-used by re-exposing them to the antibody and [125I]protein A.

Distribution of fucosyl-conjugates in rat cerebellar cells in vitro: binding patterns of Lotus tetragonolobus and Ulex europeus lectins

Light microscope radioautographic procedures were employed to investigate the binding patterns of lectins derived from Lotus tetragonolobus (Lotus A) and Ulex europeus (UEA I) to the surfaces of rat cerebellar cells maintained in dispersed cell culture. Lotus A bound extensively to the surfaces of neuronal cell somata and neuronal processes at all time points investigated; the lectin also bound to nonneuronal cell surfaces, but to an extent much less than that for neurons. Lectin binding was completely inhibited by the presence of 0.1 M alpha-L-fucose in the reaction medium. Conversely, UEA I did not appear to bind to any significant extent to the surfaces of cerebellar cells in vitro at any stage of maintenance. The incorporation of [3H]fucose into fucosyl-conjugates by cerebellar neurons in culture was investigated using light microscope radioautography and polyacrylamide gel electrophoretic methods. Radioautographs indicated that the fucose came to be distributed throughout neuronal cell somata and processes and that, with increasing time, the grain density appeared to be highest over areas of neuropil and fascicles of cellular processes. The electrophoresis data demonstrated that, among others, the label was incorporated into prominent classes of polypeptides having nominal molecular weights of 400,000, 125,000, 91,000 and 46,000 daltons.

Distribution of the LETS protein (fibronectin) in rat cerebellum. An in vitro and in vivo developmental study

The distribution of the large, external, transformation-sensitive (LETS; fibronectin) protein was investigated in rat cerebellum, both in vitro and in vivo, by biochemical and immunocytochemical methods. Biochemical analyses indicated that LETS protein is not demonstrable on the surfaces of cerebellar neurons from postnatal rats maintained in cell culture for varying periods of time, but is present on the surfaces of at least some fraction of the total nonneuronal cell population in vitro. Indirect immunofluorescence microscopy with an anti-LETS antiserum substantiated these observations and further indicated that LETS-bearing cells of cerebellum maintained in vitro are probably of endothelial and fibroblastic origin. The LETS protein is arranged in a reticular network of filaments spanning the surfaces of the cells, and the filaments are often extensively interdigitated with each other. At all stages of development investigated (two days postnatal to adult) LETS antigen was observed in vivo to be primarily localized in the meninges covering the surface of the cerebellum and between folia, and in the walls of blood vessels within the tissue. Neuroblasts and neurons of the external and internal granule layers of the cerebellum, respectively, were negative for the presence of LETS antigen.

Cell surface proteins of differentiating rat cerebellar cells maintained in dispersed cell culture

The complement of cell surface proteins of differentiating rat cerebellar neurons maintained in dispersed cell cultures was evaluated by the combined methods of lactoperoxidase-catalyzed radioiodination and miniature sodium dodecyl sulfate polyacrylamide gel electrophoresis. Pure neuronal cell samples were obtained by direct dissection of such cells from the surfaces of the culture vessels. The electrophoretic profiles of cell surface proteins (CSP) from neurons maintained for 1, 7, 14, 21 and 28 days in vitro (DIV) were compared. It could be seen that while the pattern was relatively constant between 7 and 28 DIV, several differences existed between profiles from 1 and 7 DIV. The earliest stage was characterized by a prominent polypeptide triplet having molecular weight of approximately 48,000, 56,000 and 71,000 daltons. Later stages of development (7-28 DIV) were characterized by a greater number of high molecular weight surface polypeptides relative to the earliest stage. Comparison of electrophoretic profiles of CSPs of neurons and nonneuronal cells indicated that a number of polypeptides were not shared by these classes of cells.

Axonal transport of the Ca2+-dependent protein modulator of 3':5'-cyclic-AMP phosphodiesterase in the rabbit visual system

Water-soluble proteins were extracted from individual retinas, optic nerves, combined optic tracts and lateral geniculate bodies, and superior colliculi of rabbits at 1, 3, and 18 days after injection of [3H]leucine into the right eye. The Ca2+-dependent protein modulator of 3':5'-cyclic-AMP phosphodiesterase (calmodulin) was isolated from these samples by a two-step polyacrylamide gel electrophoresis procedure. An analysis of the radioactivity incorporated into the total soluble proteins and the calmodulin revealed that most of the calmodulin was axonally transported at a slow rate (2--4 mm/day) and represented about 0.45% of the total transported soluble protein.

Investigation of the axonal transport of three acidic, soluble proteins (14-3-2, 14-3-3, and S-100) in the rabbit visual system

The question of whether three acidic, water-soluble proteins (14-3-2, 14-3-3, and S-100, the first and last known to be brain-specific) are axonally transported was investigated in the rabbit visual system. The water-soluble proteins were obtained from individual optic nerves, combined optic tracts and lateral geniculate bodies, superior colliculi, and, in some instances, retinas at various times (1--56 days) after monocular injections of [3H]leucine. These proteins were separated by a two-step polyacrylamide gel electrophoresis procedure that isolated 14-3-2, 14-3-3, and S-100 almost uncontaminated by other radioactivity. The isolated 14-3-2 and S-100 were demonstrated to be approx. 90% pure by a new method based on retarding the migration of these proteins by immunoadsorption during the first step of electrophoresis. An analysis of the radioactive labeling of the total soluble proteins (TSP) and the isolated acidic proteins revealed that: (1) S-100 was not axonally transported; (2) both 14-3-2 and 14-3-3 were part of one of the slow components of axonal transport (2--4 mm/day); (3) the radioactivity of 14-3-2 and 14-3-3 represented about 2.7% and 3.2%, respectively, of the radioactivity incorporated into the axonally transported TSP; (4) the ultimate distributions of the radioactively labeled 14-3-2 and 14-3-3 were the same (about 70% of each destined for the superior colliculus) and differed from that of the TSP; and (5) the rates of catabolism of the axonally transported 14-3-2 and 14-3-3 were slightly greater than that of the TSP, with half-lives for 14-3-2 and 14-3-3 estimated to be 11 and 10 days, respectively.