Abstract
Abstract 874
Fibrosis is a common complication of leukemia with a megakaryoblastic blast phenotype. Children with Down syndrome (DS) have an increased risk for two hematological disorders with a predominantly megakaryoblastic phenotype. First, approximately 10% of newborns with DS develop Transient Leukemia (TL; also termed Transient Myeloproliferative disorder, TMD), which is self-resolving in the majority but associated with life-threatening complications in approximately 15% of cases. Liver fibrosis and failure are predominant causes of fatal outcomes in these infants. Second, children with DS develop acute myeloid leukemia, predominantly with a megakaryoblastic phenotype, approximately 150-times more often than the general pediatric population during the first 4 years of life. Somatic mutations of GATA1 are specific for TL/TMD and myeloid leukemia of DS. We found significant liver fibrosis in NOD/SCID recipients of human DS myeloid cells. We are using this model to define the mechanisms that control the development of this potentially life-threatening complication.
Recipients were engrafted with primary human DS myeloid leukemia cells and cells of the human DS myeloid leukemia cell line GRW (established at our institution). U937 human monocytic leukemia cells, which lack trisomy 21, were used in control experiments. Cells were injected into the right femur of 8-week-old irradiated NOD/SCID mice, which had also been injected with anti-NK (anti-CD122) antibody. Phenotypic analysis at 4 to 6 weeks after transplantation used standard cytological, histological and flowcytometric methods (CD45, CD61 and CD34 antibodies were obtained from B&D). Affymetrix human Gene ST1.0 expression arrays were used to compare GRW and U937 cells. Specific gene expression was quantified by TaqMan real time RT-PCR relative to b-actin (Applied Biosystems).
Organ fibrosis in recipients of DS myeloid leukemia cells. Primary cells from patients with DS myeloid leukemia engrafted murine recipients and caused moderate bone marrow fibrosis. Transplantation of GRW human DS myeloid cells resulted in significant infiltration of recipient liver and bone marrow (33±10% of mononuclear cells, n=6). The frequency of the leukemia-initiating cells was 1/87,448 by limiting dilution. Marked fibrosis of the liver developed within 6 weeks of transplantation. Fibrosis-associated gene expression in DS myeloid leukemia cells. DS myeloid leukemia cells (GRW) expressed significantly more PDGFD and FGF17 transcripts than non-DS myeloid leukemia control cells (U937). PDGFD transcripts were 4.3-fold more abundant in GRW cells than U937 cells (p=2.67 × 10-7; n=4). FGF17 expression levels were moderately increased (1.24, p=0.07). Quantitative real time RT-PCR analysis confirmed expression of PDGFD in DS myeloid leukemia cells (GRW) (5.55 × 10-4PDFGD/b-actin transcripts; n=4) and did not detect measurable expression in control (U937) cells. FGF17 transcripts were approximately 100-fold more abundant in DS myeloid leukemia cells (GRW) cells compared to U937 controls (4.85 × 10-4 and 4.91 × 10-6 for FGF17/b-actin transcripts, n=4). Analysis of fibrosis-associated gene expression in recipient (murine) liver cells in response to transplanted DS myeloid leukemia cells (GRW) and functional analysis of the role of PDGFD and FGF17 expression in DS myeloid leukemia cells is underway.
Organ fibrosis is a significant cause of mortality of children with DS and megakaryoblastic disorders, particularly during the newborn period. Chemotherapy frequently is unable to achieve a timely reversal of this process. Better understanding of the mechanisms controlling the development of organ fibrosis is required to develop better anti-fibrotic intervention. The observed phenotype of marked liver fibrosis in murine recipients of the human DS myeloid leukemia cell line GRW provides an experimental tool to determine the components of the fibrogenic process both in infiltrating leukemia cells and recipient liver tissue. PDGFD and FGF17 expression by DS myeloid leukemia cells has been identified as possible inducer of organ fibrosis. The model can now be used to define tissue responses to infiltrating human DS myeloid leukemia cells in liver tissue of recipients and to determine the functional role of candidate genes in DS myeloid leukemia cells by altering expression levels prior to transplantation.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.