Intronic and flanking sequences are required to silence enhancement of an embryonic beta-type globin gene.

In the course of studying regulatory elements that affect avian embryonic rho-globin gene expression, the multipotential hematopoietic cell line K562 was transiently transfected with various rho-globin gene constructs containing or lacking an avian erythroid enhancer element. Enhanced levels of rho gene expression were seen from those constructs containing an enhancer element and minimal 5' or 3' flanking rho sequences but were not seen from enhancer-containing constructs that included extensive 5' and 3' flanking sequences. Deletion analysis localized 5' and 3' "enhancer-silencing elements" to -2140 to -2000 and +1865 to +2180 relative to the mRNA cap site. A third element required for enhancer silencing was identified within the second intron of the rho gene. The treatment of K562 cells with hemin, which induces erythroid differentiation, partially alleviated the enhancer-silencing effect. The silencer elements were able to block enhancement from a murine erythroid enhancer, but not from a nonerythroid enhancer. Electrophoretic mobility shift assays demonstrated that the transcription factor YY1 is able to bind both the 5' and 3' enhancer silencer elements; a point mutation of the single overlapping YY1/NF-Y binding site in the 3' element completely abolished the enhancer-silencing effect. These results demonstrate a complex enhancer silencer that requires 5' flanking, intronic, and 3' flanking sequences for a single regulatory effect on a eukaryotic gene.

5'-flanking sequences mediate butyrate stimulation of embryonic globin gene expression in adult erythroid cells.

A stable transfection assay was used to test the mechanism by which embryonic globin gene transcription is stimulated in adult erythroid cells exposed to butyric acid and its analogs. To test the appropriate expression and inducibility of chicken globin genes in murine erythroleukemia (MEL) cells, an adult chicken beta-globin gene construct was stably transfected. The chicken beta-globin gene was found to be coregulated with the endogenous adult mouse alpha-globin gene following induction of erythroid differentiation of the transfected MEL cells by incubation with either 2% dimethyl sulfoxide (DMSO) or 1 mM sodium butyrate (NaB). In contrast, a stably transfected embryonic chicken beta-type globin gene, rho, was downregulated during DMSO-induced MEL cell differentiation. However, incubation with NaB, which induces MEL cell differentiation, or alpha-amino butyrate, which does not induce differentiation of MEL cells, resulted in markedly increased levels of transcription from the stably transfected rho gene. Analysis of histone modification showed that induction of rho gene expression was not correlated with increased bulk histone acetylation. A region of 5'-flanking sequence extending from -569 to -725 bp upstream of the rho gene cap site was found to be required for both downregulation of rho gene expression during DMSO-induced differentiation and upregulation by treatment with NaB or alpha-amino butyrate. These data are support for a novel mechanism by which butyrate compounds can alter cellular gene expression through specific DNA sequences. The results reported here are also evidence that 5'-flanking sequences are involved in the suppression of embryonic globin gene expression in terminally differentiated adult erythroid cells.

Amelioration of severe hereditary spherocytosis in nonablated adult mice by marrow transplantation.

Human severe hereditary spherocytosis (sHS) is life threatening and transfusion dependent. sHS is lethal within 6 days of birth for 99% of jaundiced (ja/ja) mice, making these mice excellent models for early therapeutic interventions. Nonablated ja/ja neonates simultaneously transfused and given intravenous injections of normal marrow become chimeric for donor cells. Significant improvement of red blood cell parameters occurs but is temporary because the donor marrow-derived cells gradually disappear from the circulation. The average lifespan, however, is increased to 8.7 months. We postulate that donor cells are diluted by rapidly proliferating host cells during postnatal growth. Here, we test this hypothesis by determining whether treatment of adults improves long-term therapy. Nonablated ja/ja adults rescued by a single neonatal transfusion were injected intravenously with 1 x 10(10) normal, genetically marked donor marrow cells/kg body weight. Donor cell implantation and blood parameters were monitored periodically and tissue histopathology was determined at necropsy.sHS recipients with 100% donor erythroid cells have significantly improved red blood cell counts throughout life when compared with ja/ja controls transfused once at birth. Total serum iron and bilirubin levels are corrected in ja/ja marrow recipients. Donor-implanted HS mice necropsied at 16 to 21 months of age have normal mean cell hemoglobin concentration and dramatically decreased tissue iron deposits. Reticulocyte counts but not red cell counts normalize, suggesting the HS mice reset their response to hypoxia. Nonablative transplantation performed after cessation of host postnatal red blood cell amplification can be therapeutic long term for transfusion-dependent hemolytic anemias.

Murine recessive hereditary spherocytosis, sph/sph, is caused by a mutation in the erythroid alpha-spectrin gene.

INTRODUCTION: Spectrin, a heterodimer of alpha- and beta-subunits, is the major protein component of the red blood cell membrane skeleton. The mouse mutation, sph, causes an alpha-spectrin-deficient hereditary spherocytosis with the severe phenotype typical of recessive hereditary spherocytosis in humans. The sph mutation maps to the erythroid alpha-spectrin locus, Spna1, on Chromosome 1. MATERIALS AND METHODS: Scanning electron microscopy, osmotic gradient ektacytometry, cDNA cloning, RT-PCR, nucleic acid sequencing, and Northern blot analyses were used to characterize the wild type and sph alleles of the Spna1 locus. RESULTS: Our results confirm the spherocytic nature of sph/sph red blood cells and document a mild spherocytic transition in the +/sph heterozygotes. Sequencing of the full length coding region of the Spna1 wild type allele from the C57BL/6J strain of mice reveals a 2414 residue deduced amino acid sequence that shows the typical 106-amino-acid repeat structure previously described for other members of the spectrin protein family. Sequence analysis of RT-PCR clones from sph/sph alpha-spectrin mRNA identified a single base deletion in repeat 5 that would cause a frame shift and premature termination of the protein. This deletion was confirmed in sph/sph genomic DNA. Northern blot analyses of the distribution of Spna1 mRNA in non-erythroid tissues detects the expression of 8, 2.5 and 2.0 kb transcripts in adult heart. CONCLUSION: These results predict the heart as an additional site where alpha-spectrin mutations may produce a phenotype and raise the possibility that a novel functional class of small alpha-spectrin isoforms may exist.

Thrombosis in heritable hemolytic disorders.

Thromboses are a serious complication in patients with sickle cell disease, paroxysmal nocturnal hemoglobinuria, beta-thalassemia major, or thalassemia intermedia. Despite prophylaxis, thrombotic events can continue and can result in severe physical or mental debilitation or death of the patient. The fact that thrombosis does not occur in all patients with hemolytic anemias suggests that multiple factors interact to cause the coagulation crisis. Genetic modifiers, associated diseases, nutritional status, infections, environment, and treatment modalities are variables implicated in thrombophilia. The complexity confounds attempts to identify single causative agents in humans with hemolytic anemias. In the past year, mutations in putative genetic modifiers of the coagulation response have been examined as risk factors in patients with a history of thromboses; red cell binding sites on endothelial cells have been identified; and mouse models of thrombogenesis that permit experimental manipulation of single factors on a defined genetic background have been described.

Erythroid phosphatidyl serine exposure is not predictive of thrombotic risk in mice with hemolytic anemia.

Thrombosis is a major complication of human hemolytic anemias such as sickle cell disease, thalassemia, and severe hereditary spherocytosis (HS). Mice with severe HS and severe hereditary elliptocytosis (HE) also suffer from thrombosis, with incidences ranging from 15 and 22% in beta-spectrin- and ankyrin-deficient mice, respectively, to 85 to 100% in alpha-spectrin-deficient and band 3 knockout mice. A contributing factor to thrombosis could be loss of phospholipid asymmetry of the mutant red blood cells (RBCs), with concomitant exposure of the aminophospholipid phosphatidylserine (PS). Increased PS exposure occurs in RBCs from sickle cell and thalassemia patients and in RBCs from band 3-deficient mice. To determine if increased PS exposure correlates with thrombotic risk in HS and HE mice with ankyrin, beta-spectrin, and alpha-spectrin deficiencies, measurements of FITC-labeled annexin V binding to externalized PS on RBCs were performed. PS exposure is elevated in all mice with HS and HE, but the percentage of RBCs with exposed PS does not correlate with thrombotic risk in these mice.

Reduced incidence of thrombosis in mice with hereditary spherocytosis following neonatal treatment with normal hematopoietic cells.

Thrombosis is a life-threatening complication of hemolytic anemia in humans. Cardiac thrombi are present in all adult alpha-spectrin-deficient (sph/sph) mice with severe hereditary spherocytosis, providing a model for events preceding thrombosis. The current study evaluated (1) the timing of thrombosis initiation and (2) the effect of postnatal transplantation of normal cells on life span and thrombotic incidence in adult mice. Thrombi are detected histologically following necropsy in untreated sph/sph mice of various ages and are not observed until 6 weeks of age. Thrombotic incidence increases from 50% at 6 to 7 weeks of age to 100% at 9 weeks of age. As a potential therapy, nonablated sph/sph neonates were transfused with either genetically marked normal peripheral blood (PB), bone marrow (BM), or both and assessed for donor cells and thrombosis. A single transfusion of PB, with or without BM, significantly increases the percentage of sph/sph mice that survive to weaning (4 weeks of age). Replacement in all sph/sph recipients is limited to red blood cells (RBCs). RBCs derived from donor PB are lost within 5 weeks. PB plus BM prolongs high-level donor PB cell production better than BM alone. Thrombotic incidence is significantly reduced in all sph/sph mice treated with PB, BM, or both. Hence, the presence of normal blood cells in the peripheral circulation of neonatal and adult sph/sph mice rescues the former and abrogates the development of thrombosis in the latter. (Blood. 2001;97:3972-3975)

Thrombosis and secondary hemochromatosis play major roles in the pathogenesis of jaundiced and spherocytic mice, murine models for hereditary spherocytosis.

Jaundiced mice, ja/ja, suffer from a severe hemolytic anemia caused by a complete deficiency of erythroid beta-spectrin. We used these mice as a model to investigate the pathophysiological consequences of the deficiency, including the effects in the nonerythroid tissues where this protein is expressed. Because the ja/ja mice rarely survive beyond the fourth postnatal day, methods were assessed for extending lifespan into adulthood. Neonatal transfusion increased lifespan to a mean of 3.7 months, allowing a more complete characterization of the pathophysiology. Blood parameters and histopathology of the jaundiced mouse were compared with that from spherocytic mice, which have a hemolytic anemia caused by deficiency of erythroid alpha-spectrin, yet can survive the postnatal period transfusion free. The adult jaundiced and spherocytic mice present with greatly decreased hematocrit and red blood cell counts, reticulocytosis, and bilirubinemia, leading secondarily to hepatosplenomegaly and cardiomegaly. Jaundiced and spherocytic mice were analyzed histopathologically between 1.0 and 9.5 months of age. Interestingly, the complete absence of erythroid beta-spectrin in jaundiced mice leads to no detectable structural defects in brain, cardiac, or skeletal muscles. However, fibrotic lesions and lymphocytic infiltration were observed in cardiac tissue from 4 of 13 jaundiced mice and 15 of 15 spherocytic mice, and thrombi were detected at either the atrioventricular valves or within the atria of 2 of 13 jaundiced mice and 15 of 15 spherocytic mice. In addition, all affected mice had a progressive renal hemosiderosis concurrent with hydronephrosis and glomerulonephritis. The severity of the renal disease and its presence in all moribund mice suggests kidney failure rather than the fibrotic heart lesions as the major cause of death in these mice.

Hematopoietic cells from -spectrin-deficient mice are sufficient to induce thrombotic events in hematopoietically ablated recipients.

Thrombotic events are life-threatening complications of human hemolytic anemias such as paroxysmal nocturnal hemoglobinuria, sickle cell disease, and thalassemia. It is not clear whether these events are solely influenced by aberrant hematopoietic cells or also involve aberrant nonhematopoietic cells. Spherocytosis mutant (Spna1(sph)/Spna1(sph); for simplicity referred to as sph/sph) mice develop a severe hemolytic anemia postnatally due to deficiencies in -spectrin in erythroid and other as yet incompletely defined nonerythroid tissues. Thrombotic lesions occur in all adult sph/sph mice, thus providing a hematopoietically stressed model in which to assess putative causes of thrombus formation. To determine whether hematopoietic cells from sph/sph mice are sufficient to initiate thrombi, bone marrow from sph/sph or +/+ mice was transplanted into mice with no hemolytic anemia. One set of recipients was lethally irradiated; the other set was genetically stem cell deficient. All mice implanted with sph/sph marrow, but not +/+ marrow, developed severe anemia and histopathology typical of sph/sph mice. Histological analyses of marrow recipients showed that thrombi were present in the recipients of sph/sph marrow, but not +/+ marrow. The results indicate that the -spectrin-deficient hematopoietic cells of sph/sph mice are the primary causative agents of the thrombotic events.

Defective spectrin integrity and neonatal thrombosis in the first mouse model for severe hereditary elliptocytosis.

Mutations affecting the conversion of spectrin dimers to tetramers result in hereditary elliptocytosis (HE), whereas a deficiency of human erythroid alpha- or beta-spectrin results in hereditary spherocytosis (HS). All spontaneous mutant mice with cytoskeletal deficiencies of spectrin reported to date have HS. Here, the first spontaneous mouse mutant, sph(Dem)/ sph(Dem), with severe HE is described. The sph(Dem) mutation is the insertion of an intracisternal A particle element in intron 10 of the erythroid alpha-spectrin gene. This causes exon skipping, the in-frame deletion of 46 amino acids from repeat 5 of alpha-spectrin and alters spectrin dimer/tetramer stability and osmotic fragility. The disease is more severe in sph(Dem)/sph(Dem) neonates than in alpha-spectrin-deficient mice with HS. Thrombosis and infarction are not, as in the HS mice, limited to adults but occur soon after birth. Genetic background differences that exist between HE and HS mice are suspect, along with red blood cell morphology differences, as modifiers of thrombosis timing. sph(Dem)/sph(Dem) mice provide a unique model for analyzing spectrin dimer- to-tetramer conversion and identifying factors that influence thrombosis.