Low or absent SPARC expression in acute myeloid leukemia with MLL rearrangements is associated with sensitivity to growth inhibition by exogenous SPARC protein.

Secreted protein, acidic and rich in cysteine (SPARC), is a matricellular glycoprotein with growth-inhibitory and antiangiogenic functions. Although SPARC has been implicated as a tumor suppressor in humans, its function in normal or malignant hematopoiesis has not previously been studied. We found that the leukemic cells of AML patients with MLL gene rearrangements express low to undetectable amounts of SPARC whereas normal hematopoietic progenitors and most AML patients express this gene. SPARC RNA and protein levels were also low or undetectable in AML cell lines with MLL translocations. Consistent with its tumor suppressive effects in various solid tumor models, exogenous SPARC protein selectively reduced the growth of cell lines with MLL rearrangements by inhibiting cell cycle progression from G1 to S phase. The lack of SPARC expression in MLL-rearranged cell lines was associated with dense promoter methylation. However, we found no evidence of methylation-based silencing of SPARC in primary patient samples. Our results suggest that low or absent SPARC expression is a consistent feature of AML cells with MLL rearrangements and that SPARC may function as a tumor suppressor in this subset of patients. A potential role of exogenous SPARC in the therapy of MLL-rearranged AML warrants further investigation.

Genetic complexity of regulatory mutants defective for HLA class II gene expression.

MHC (called HLA in man) class II genes play an essential role in cell-mediated immunity. Absence of HLA class II Ag on B lymphocytes is the basis of some congenital immunodeficiencies (CID). We have studied CID by generating transient heterokaryons from cell lines of such patients, and we report that the mutations fall into four complementation groups. In addition, fusions with the HLA class II deletion mutant 721.180 indicate that the genetic defects for each group in HLA class II expression map outside the HLA class II region. A small HLA-DRA promoter fragment is sufficient to drive expression of a reporter gene in normal B cell lines, but expression from the same construct is clearly reduced in mutant cell lines representative of all four complementation groups. This confirms earlier results that indicate defective transcription of HLA class II genes in the class II- CID mutant cell lines. Analysis of proteins that bind to the DRA promoter in nuclear extracts of the mutants suggests that complexes recognizing distinct elements of the DRA promoter may be quantitatively decreased in different mutants. In addition, we show that nuclear extracts from two groups fail to transcribe a DRA promoter construct in vitro accurately reflecting their DRA- phenotypes. In contrast, nuclear extracts from another mutant, RJ2.2.5, transcribe the DRA construct, albeit at a reduced level. Finally, though cell lines from different groups complement each other in vivo, no complementation was observed by mixing extracts for transcription in vitro.

Regulation of alternative splicing in the generation of isoforms of the mouse Ly-5 (CD45) glycoprotein.

Alternative splicing generates various Ly-5 glycoprotein isoforms of the cell surface that typify different cell lineages and stages of hematopoietic differentiation in the mouse; exons 4-6 are incorporated to generate a B-cell isoform (B220) and excluded from a T-cell isoform (T200), the other coding exons (3 and 7-33) being shared. As a first step to understanding the mechanisms regulating Ly-5 alternative splicing, and thus determining Ly-5 isoforms, a minigene representing exons 3-7 was transfected into Ly-5-expressor T cells and B cells and into nonexpressor L cells for comparison of splicing patterns. We conclude that all the information required for faithful splice-site selection according to cell type is contained within the resulting pre-mRNA. The splicing pattern manifested by nonexpressor L cells may represent a default and nonregulated type. We postulate trans-acting factor(s) to account for the selection of appropriate exons, and we provide support for this interpretation from analysis of fused hybrid T-B cells, which exhibited B-cell specific Ly-5 transcripts. Splicing patterns were well conserved despite substantial disruption of constructs. However, extensive deletion analyses suggested that cis sequences flanking and within exon 6 affect the exclusion of that exon in T cells.