The Human Myeloperoxidase Promoter Directs Hematopoietic Cell-Specific Transcription in Transgenic Mice.

The myeloperoxidase (MPO) gene is expressed specifically in immature myeloid cells and not to a detectable extent in other cell types or in more mature myeloid cells. Furthermore, the MPO gene is actively transcribed both in normal myeloblasts (MPO protein constitutes about 5% of total protein in granulocytes) and in the myeloblasts of the majority of acute myeloid leukemias, but not in lymphoid leukemias or non-myeloid tumors. This has led to use of MPO activity for many years as the defining cytochemical marker for myeloid differentiation in anatomic pathology. We were the first to purify the human MPO promoter and to dissect many of its regulatory elements. Later, we showed the existence of three distinct human MPO promoters. More recently, we demonstrated that only one of these, which we term “P1”, is responsible for most of the physiologic MPO transcription of the human MPO gene. While until recently, purified human (or murine) MPO promoter constructs have exhibited only partial tissue specificity, we recently obtained a human MPO promoter construct whose specificity in cell culture mimics that of MPO gene expression seen in vivo. Comparison of many different promoter constructs reveals optimal, myeloid-specific activity for a construct extending from bp −4193 in the 5′-flanking region of the MPO gene to bp +15, just downstream from the transcription start site. Inclusion of additional downstream sequences did not increase promoter activity or specificity. We have now developed transgenic murine lines in which 4.2 kb of human MPO proximal 5′ flanking region DNA was linked to an adjacent downstream Renilla luciferase reporter (MPORLUC). A series of founder mouse lines were screened by PCR assays of DNA extracted from the tail vein for the presence of the MPO promoter-Renilla luciferase construct. Five founder mouse lines were positive for the transgene by the PCR assay. Transgenic and control mice (founders or offspring) were sacrificed and all possible tissues were examined for Renilla luciferase activity by an enzymatic assay (Promega). High-level Renilla luciferase activity was observed in two founder lines but only one of these founders produced offspring. In each of multiple different mice belonging to the founder line which was successfully mated, high-level Renilla luciferase activity was evident in bone marrow of femur and to a variable degree in vertebral bone marrow. In mice younger than four months old, a modest-level of luciferase activity was also observed in spleen. Luciferase activity was at or near background levels in thymus, heart, lung, liver, kidney, stomach, colon, bladder, brain, skeletal muscle, skin and small intestine in all of the MPORLUC transgenic mice. All tissues in wild type mice lacked measurable Renilla luciferase activity. Interestingly, there has been a slightly higher level of deaths in the luciferase-positive transgenic mice than in controls, suggesting an as yet uncharacterized toxic effect of the transgene. These results indicate that the human MPO promoter construct used in these studies exhibits tissue specificity of activity. Such a promoter might be useful for future gene targeting studies or for gene therapy.