Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Blood First Edition Paper, prepublished online May 16, 2008; DOI 10.1182/blood-2008-03-145672 Mutations in EKLF/KLF1 form the molecular basis of the rare blood group In(Lu) phenotype. Belinda K. Singleton1, Nicholas M. Burton2, Carole Green1, R. Leo Brady2, David J. Anstee1. 1 Bristol Institute for Transfusion Sciences (BITS), National Blood Service, Bristol, United Kingdom; and 2Department of Biochemistry, University of Bristol, Bristol, United Kingdom Corresponding author: Belinda K. Singleton, Bristol Institute for Transfusion Sciences, Southmead Road, Bristol, BS10 5ND, United Kingdom. E-mail: [email protected] Telephone: +44 117 9912126 Fax: +44 117 9591660 Keywords: Transcription factor, erythropoiesis, Lutheran Blood Group, transcriptomics, EKLF Copyright © 2008 American Society of Hematology From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Abstract Comparison of normal erythroblasts and erythroblasts from individuals with the rare In(Lu) type of Lu(a-b-) blood group phenotype revealed increased transcription levels for 314 genes and reduced levels for 354 genes in In(Lu) cells. Many erythroid-specific genes (including ALAS2, SLC4A1) had reduced transcript levels suggesting the phenotype resulted from a transcription factor abnormality. A search for mutations in erythroid transcription factors revealed mutations in the promoter or coding sequence of EKLF in 21 of 24 individuals with the In(Lu) phenotype. In all cases the mutant EKLF allele occurred in the presence of a normal EKLF allele. Nine different loss of function mutations were identified. One mutation abolished a GATA1 binding site in the EKLF promoter (124T>C). Two mutations (Leu127X; Lys292X) resulted in premature termination codons, two (Pro190LeufsX47; Arg319GlufsX34) in frameshifts and four in amino acid substitution of conserved residues in zinc finger domain 1 (His299Tyr) or domain 2 (Arg328Leu; Arg328His; Arg331Gly). Individuals with the In(Lu) phenotype have no reported pathology indicating that one functional EKLF allele is sufficient to sustain human erythropoiesis. These data provide the first description of inactivating mutations in human EKLF and the first demonstration of a blood group phenotype resulting from mutations in a transcription factor. Introduction Erythroid differentiation is a highly regulated process involving numerous transcription factors including GATA1 and EKLF.1-3 GATA1 binds to EKLF to form a functional unit.4,5 Absence of GATA1 in mice causes embryonic death from severe anemia due to maturation arrest and apoptosis of erythroid progenitor cells.6 In mice, EKLF is not required for the proliferation of erythroid progenitors but is essential for terminal differentiation of erythroid cells because it activates essential erythroid genes.7,8 Mice lacking EKLF die in utero around embryonic day 14 from severe anemia associated with beta globin deficiency whereas heterozygous EKLF +/- mice develop normally.9 These data suggest homozygous inheritance of inactivating mutations affecting EKLF expression in man would be incompatible with post-embryonic life although heterozygous inheritance of inactivating EKLF mutations would be viable. Nevertheless, inactivating mutations in human EKLF have not been previously reported. The blood group phenotype Lu(a-b-) is known to arise from three distinct genetic backgrounds: homozygosity for inactivating mutations in the Lutheran gene,10,11 inheritance of an X-linked suppressor gene (XS2)12 and inheritance of an autosomal suppressor gene InLu.13 Inheritance of XS2 has been reported in only one family.12 In contrast, inheritance of InLu is relatively common, occurring with a frequency of about 1 in 3,000 blood donors in South-East England14 and 1 in 5,000 of South Wales blood donors.15 In order to elucidate the molecular mechanism giving rise to the In(Lu) type of Lu(a-b-) we compared the transcriptome of normal and In(Lu) erythroblasts at different stages of maturation and hypothesised that the differences observed could result from a mutated erythroid transcription factor. Sequencing of several erythroid transcription factor genes revealed heterozygous inheritance of loss of function mutations in EKLF in 21 of 24 In(Lu) donors. This report identifies the first mutations in human EKLF, explores their likely molecular consequences and demonstrates the viability of the EKLF+/- genotype in man. Materials and Methods Cell culture and flow cytometry Waste buffy coat material from anonymous blood donors was made available from the National Blood Service (Tooting, United Kingdom) and the Welsh Blood Service (Cardiff, United From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Kingdom). This provision complies with the Nuffield Council on Bioethics Guidance on Human Tissue Ethical and Legal Issues 1954, the Medical Research Council's Operational and Ethical Guidance on Human Tissue and Biological Samples for use in Research 2005, and the Royal College of Pathologists Transitional Guidelines to facilitate changes in procedures for handling "surplus" and archival material from human biological samples. Informed consent was provided according to the Declaration of Helsinki. Waste buffy coat material was provided with informed consent from four blood donors with the In(Lu) phenotype (InLu1-4) and four controls (Con1-4) matched for age, sex and ethnicity by the Welsh Blood Service, Cardiff and the National Blood Service, Tooting, and anonymised for the purposes of this study. CD34+ cells were isolated by positive selection with the MiniMACS magnetic bead system according to the manufacturer’s instructions (Miltenyi Biotec, Bisley, UK). Erythroid progenitor cells were obtained by ex vivo culture of CD34+ cells, as described in Fricke et al.16 In(Lu) and matched control cells were grown in parallel. Cells were harvested during culture for analysis by flow cytometry and for freezing and storage at -70oC in RNAlater (Ambion, Huntingdon, UK). Cells from an additional In(Lu) sample (InLu5) and an unmatched control (Con5) had previously been cultured and stored in RNAlater. Cultured erythroid progenitor cells were analysed by flow cytometry after labelling with mouse monoclonal antibodies: anti-RhAG (LA1818), anti-GPA (BRIC256), anti-Band 3 (BRIC6) and a blend of anti-Lub (BRIC108) and anti-Lu (BRIC221), detecting epitopes on domain 1 and domain 4 of the Lutheran glycoprotein respectively. All BRIC antibodies were provided by the International Blood Group Reference Laboratory (IBGRL, Bristol, UK). Isotype controls (Dako, Ely, UK) were tested in parallel. Bound antibody was detected by adding RPE-conjugated F(ab′)2 anti-mouse globulin (Dako) and the geometric mean (FL2) was recorded for each test. Flow cytometry was performed on a FC500 flow cytometer and analysed with CXP software (Beckman Coulter, High Wycombe, UK). RNA isolationf A minimum of 1.5 x106 erythroid cells were harvested every few days from day 4 to day 14 of culture and frozen in RNALater (Ambion). Total RNA was isolated using the RNeasy Mini Kit (Qiagen, Crawley, UK), including an on-column DNaseI digestion, according to the manufacturer’s instructions. RNA yield was determined using the RiboGreen RNA Quantitation Kit (Molecular Probes, Invitrogen Ltd, Paisley, UK). Transcriptomics Two In(Lu) samples (InLu3 and InLu4) and their matched controls (Con3 and Con4) were analysed by microarrays using cells from day 6 and day 11 of culture. Cells from day 4 from InLu3 and Con3 were also analysed. Total RNA (8 - 10µg) was supplied to the University of Bristol Transcriptomics Facility. Quality of the total RNA samples and the resulting cRNA was checked using a 2100 Bioanalyser (Agilent Technologies, Stockport, UK). Fragmented biotinylated cRNA was prepared and 15µg hybridised to each of two HG-U133A GeneChips (i.e., two technical replicates of each) according to the manufacturer’s recommended protocols (Affymetrix, Santa Clara, CA, USA). Single Array Expression Analysis was performed using the Affymetrix GeneChip Operating Software (GCOS). A global scaling strategy was used to give a target intensity of 500 for each array. Arrays were only included for further analysis if they complied with the quality control metrics defined by Affymetrix.17 Data from all 20 arrays were filtered in Microsoft Excel (Microsoft, Seattle, WA, USA) to exclude all probe sets called either Absent or Marginal in all arrays. Control probe sets with the prefix AFFX were also removed prior to subsequent data analysis. Filtered data were imported into From www.bloodjournal.org by guest on June 15, 2017. For personal use only. GeneSifter software (VizX Labs, Seattle, WA, USA), transformed to a log2 scale and analysed to determine differentially expressed genes. Day 6 and day 11 data were analysed by balanced 2-way ANOVA. Day 4 data were analysed by unpaired t-test. A 1.5-fold change threshold and test statistic of P<.05 were used for both data sets. Gene Ontology terms showing enrichment were identified with the Z-score Report function of the software. Quantitative Real-Time PCR (Q-PCR) Total RNA (1ug) from each harvested erythroid cell sample was converted to cDNA using SuperScript II reverse transcriptase according to the manufacturer’s instructions (Invitrogen, Paisley, UK). Differentially expressed genes were analysed by quantitative real-time PCR (Q-PCR). Amplification of LU, CD44H, AHSP and SLC4A1 used primers designed with Primer Express software v2.0 (Applied Biosystems, Warrington, UK). Sequences were as follows: LU 5′CCCCCAACAAAGGGACACT-3′ and 5′-TTGCGATACCACGTGATCTTG-3′; CD44H 5′CACAGACAGAATCCCTGCTACCA-3′ and 5′-TCCCATGTGAGTCCATCTGAT-3′; AHSP 5′GTTAGACCTGAAGGCAGATGGC-3′ and 5′-GAGGATCATTGAAGACCTGCTGA-3′; SLC4A1 5′-GGATTTCTTCATTCAGGATACCTACAC-3′ and 5′-CCCCGGGCTGAGGAGTT-3′, and probe 5′FAM-AAACTCTCGGTGCCTGATGGCTTCAAG-3′TAMRA. Amplification of SLC4A1 required TaqMan Universal PCR Master Mix whereas the remaining amplifications used SYBR Green PCR Master Mix (Applied Biosystems). Primers for AHSP and SLC4A1 were kindly provided by Dr. K. Finning (IBGRL). ALAS2, HBB and PABPC1 gene amplifications were performed using TaqMan Universal PCR Master Mix and inventoried TaqMan Gene Expression Assays (Applied Biosystems). Q-PCR was performed on the ABI 7000 Sequence Detection System (Applied Biosystems) with the following conditions: 50oC for 2 minutes and 95oC for 10 minutes, followed by 45 cycles of 95oC for 15 seconds and 60oC for 60 seconds. All reactions were performed in triplicate. Target gene expression was normalised to PABPC1 expression and compared to a calibrator sample (UR/FL; cDNA prepared from equal amounts of human Universal Reference RNA and fetal liver RNA (Clontech, Mountain View, CA, USA)) using the the Pfaffl method of relative quantification as described in the Bio-Rad Real-Time PCR Applications Guide.18 PABPC1 was used as a control for gene amplification as it has been shown by microarrays to have consistent expression in our erythroid cultures (B.K.S., unpublished data, July 1, 2005). Genomic DNA Genomic DNA (gDNA) was extracted from 1 x 106 cells from each erythroid culture using the QIAamp DNA Blood Mini Kit (Qiagen, Crawley, UK). Additional samples from individuals with the In(Lu) phenotype (InLu6-24) were obtained from archived gDNA or prepared from archived cryopreserved red blood cells (available in-house). Control gDNA from 32 random blood donors (Con6-37) was kindly provided by Dr. V. Karamatic Crew (BITS). Polymerase Chain Reaction (PCR) PCR was used to amplify human EKLF cDNA (Entrez Accession number NM_006563.2) and exons from gDNA (Entrez Accession number NT_086897.1: complement of region 4090501 – 4093981). PCR primers were designed with Primer Express software v2.0 (Applied Biosystems). EKLF cDNA was amplified using 2 overlapping pairs of primers: 5′AGTTCACGAGGCAGCCGAG-3′ and 5′-CCGGGTCCCAAACAACTCA-3′; 5′GCACTTCCAGCTCTTCCGC-3′ and 5′-CTTGTCCCATCCCCAGTCACT-3′. Two additional primers were also used for sequencing: 5′-CCGAGACTCTGGGCGCATA-3′ and 5′ACCCAAAAGCCCAGCCAC-3′. PCR reactions contained 1U Expand DNA polymerase with 1.5 From www.bloodjournal.org by guest on June 15, 2017. For personal use only. mM MgCl2 and 1 x buffer (Expand High Fidelity PCR system, Roche, Mannheim, Germany), 200 mM dNTPs (GE Healthcare, Amersham, UK), 500 mM each primer (Sigma-Aldrich, Haverhill, UK) and 1 x PCRx Enhancer solution (Invitrogen). Amplification was performed using the GeneAmp PCR system 9700 (Applied Biosystems) with a Touchdown PCR profile as follows: 95oC for 5 minutes then 9 cycles of 95oC for 30 seconds, 70oC for 30 seconds, 72oC for1 minute, with a drop in annealing temperature of 2oC per cycle (to 54oC). This was followed by 35 cycles of 95oC for 30 seconds, 52oC for 30 seconds, 72oC for1 minute with a final extension at 72oC for 10 minutes. EKLF exon 1 and 286 bp of promoter sequence was amplified from gDNA using primers 5′TTACCCAGCACCTGGACCCT-3′ and 5′-GAACCTCAAACCCCTAGACCACC-3′ with the same conditions as for cDNA plus the addition of 2 x PCRx Enhancer solution (Invitrogen). EKLF exon 2 was amplified with 2 overlapping pairs of primers: 5′-GTGTCCAGCCCGCGATGT-3′ and 5′-CCGGGTCCCAAACAACTCA-3′; 5′-CCGAGACTCTGGGCGCATA-3′ and 5′GCCCTCTGCAACCCTTCTTC-3′ with the same conditions as described for cDNA amplification. Exon 3 was amplified using primers 5′-TGCGGCAAGAGCTACACCA-3′ and 5′CTTGTCCCATCCCCAGTCACT-3′ and required the addition of 2 x PCRx Enhancer solution (Invitrogen). Primers described for the amplification of EKLF cDNA were also used as gDNA sequencing primers, as appropriate. gDNA derived from random blood donors was amplified as described except for the following modifications: exon 1 amplification used 0.625 U GoTaq Flexi DNA polymerase with its own buffer (Promega, Southampton, UK) and without PCRx Enhancer solution, and exons 2 and 3 used 1 U AccuPrime GC-Rich DNA polymerase with Buffer A (Invitrogen), also without PCRx Enhancer solution. PCR products were purified using either the Qiaex II Gel Extraction Kit or QiaQuick PCR Purification Kit (Qiagen) and sequenced on both strands using the ABI-PRISM 3100 Applied Biosystem automated DNA sequencer (Applied Biosystems). Details of primer sequences and PCR conditions used for the amplification and sequencing of other erythroid transcription factor genes (GATA1, FOG1, GFI1B, TAL1, SPI1, MYB, MZF1) are available on request. Homology modeling An homology model was constructed of the EKLF zinc finger domains in complex with a duplex oligodeoxynucleotide (5'-TTCCACACCCT-3'), based on the crystal structure of a Zif268-DNA complex (PDB ID 1AAY).19 The protein sequences are 44% amino acid identical with one insertion of two residues in EKLF. The initial protein model was constructed using Modeller20 to mutate the Zif268 sequence to EKLF; all other molecular modeling operations were performed with SYBYL7.3 (Tripos, St Louis, MO). The initial DNA model was constructed by manual mutation of DNA from the Zif268 structure. The two initial models were merged and steric clashes between protein and DNA were relieved by manual alteration of amino acid sidechain conformations. Three layers of water were then added to the complex and subsequently iteratively minimized to convergence using the Amber99 force-field21 while constraining the co-ordinates of the macromolecular atoms. During subsequent rounds of energy minimization the positional constraints on the macromolecular components were incrementally removed from the protein sidechains, then the protein backbone, and finally the DNA. Molecular geometry was assessed with MolProbity22 and NUCheck.23 Areas with poor geometry were manually adjusted and the solvation and energy minimization procedures repeated. Models of three EKLF variants (Arg328Leu, Arg328His and Arg331Gly) were constructed by mutation of the native EKLF model and repetition of the minimization process. All final models did not contain any Ramachandran outliers and were ranked by MolProbity within the 85th percentile of crystallographic structures determined at approximately 2 Å resolution. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Results Transcription is altered in erythroblasts from cultures of In(Lu) cells Transcriptional differences between control and In(Lu) erythroid cells at days 6 (pronormoblasts) and 11 (normoblasts) of ex vivo culture were determined by microarray analysis using RNA obtained from cultured erythroblasts. A balanced 2-way ANOVA was used to look for genes differentially expressed between the two cell types and between the two days tested. The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE10584. Analysis of control and In(Lu) transcript levels showed 373 genes with increased levels between day 6 and day 11 and 389 with decreased levels. The same genes were affected in control and In(Lu) erythroblasts. Selected results are displayed in Supplemental Table S1. As expected, many genes involved in hemoglobin synthesis or encoding blood group antigens were found to increase their expression. Genes involved in antigen processing and presentation showed decreased expression, in keeping with the flow cytometric findings of Southcott et al.24 Further inspection of the data revealed significant differences between control and In(Lu) erythroblasts with respect to the magnitude of transcript levels. In In(Lu) samples, 314 genes had higher expression than controls and 354 had lower expression at day 6 and/or at day 11. Gene expression between days 6 and 11 of cultured In(Lu) cells showed the same qualitative pattern seen with controls but marked differences regarding the levels of gene expression (Table 1). Genes with decreased expression during erythropoiesis (e.g., the major histocompatibility complex (HLA) genes) showed higher expression in the In(Lu) cells while erythroid-specific genes (e.g. delta-aminolevulinate synthase 2 (ALAS2), alpha hemoglobin stablising protein (AHSP), and glycophorin A (GPA)) showed lower expression, suggesting that the pattern of gene expression in In(Lu) cells is less capable of promoting erythroid maturation. Flow cytometric analysis of In(Lu) cells (donor InLu2) from day 8 of culture demonstrated reduced surface expression of GPA, Band 3 and Lutheran (mean fluorescence intensity 106.4, 1.9 and 0.6 respectively compared with control (Con2) values of 172.5, 5.6 and 11.6 respectively) suggesting reduced transcript levels are reflected in the cell surface phenotype. The differences between In(Lu) and control cells were most apparent on day 6 where almost all erythroid-specific genes show a difference between the In(Lu) and the control cells. The only exceptions are the genes for the basal cell adhesion molecule (BCAM) which carries the Lutheran blood group antigens and CD44 which carries the Indian blood group antigens. These genes show a greater difference change at day 11. This is of particular interest since the expression of these blood group antigens on the surface of red blood cells is known to be suppressed in individuals with the In(Lu) phenotype.13,25,26 We reasoned that if gene expression profiles at day 6 revealed a relative reduction in transcription of erythroid genes in In(Lu) cells then this effect might be even more apparent earlier in the cultures. cRNA derived from day 4 cells from one pair of control and In(Lu) cultures were hybridised to arrays (2 technical replicates of each). Microarray data were analysed by unpaired ttest and the results revealed 378 genes showing at least a 1.5-fold difference in expression between the two cell types (P<.05). Selected results are shown in Supplemental Table S2. The 159 genes upregulated in In(Lu) cells at day 4 again involved the defense response and antigen processing and presentation. Two hundred and nineteen genes were down-regulated and, as at day 6, these contained genes for hemoglobin synthesis and erythroid-specific factors. The transcript for ALAS2 showed a greater reduction at day 4 than at the later days (12.5-fold at day 4 compared to 2-fold at day 6 and no change at day 11). ALAS2 is an enzyme that catalyses the first step of heme synthesis. This first step is rate-limiting and so the reduction in expression may be expected to have severe From www.bloodjournal.org by guest on June 15, 2017. For personal use only. consequences for heme and hence hemoglobin production. At day 4 reduced expression of alpha and beta globin transcripts is also apparent. This was not seen at the later timepoints. Furthermore, expression of RNA encoding the erythrocyte anion exchanger (SLC4A1, also known as Band 3) also shows a much greater reduction at day 4 compared with later days (i.e. 18.8-fold down at day 4, 3.1-fold at day 6 and only 1.1-fold at day 11). The greater differences in expression of these erythroid-specific genes at the earlier timepoint are consistent with our initial observation that erythroid cells from individuals with the In(Lu) phenotype are less able to promote erythroid maturation at the transcriptional level. Validation of microarray data by quantitative real-time PCR (Q-PCR) In order to validate our microarray findings and to obtain more comprehensive data regarding temporal changes in expression of selected genes, we performed Q-PCR on cDNA from four pairs of control and In(Lu) cultures, and included a greater number of timepoints (Figure 1 and Supplemental Figure S1). In control cells, LU RNA is expressed at very low levels during early days of culture, rising by day 8 and peaking at day 11. The Lutheran antigens can be detected on the surface of cultured erythroblasts at around day 7 to 8.24 In In(Lu) cells, LU appears to be transcriptionally repressed (a maximum of 40-fold at day 8; Figure 1 and Supplemental Figure S1), although some expression is detected at all days with a slight increase at day 11 suggesting that the gene is not completely silenced. Although red blood cells from individuals with the In(Lu) phenotype type serologically as Lu(a-b-), in many cases adsorption/elution techniques have shown the presence of very weak Lutheran antigen expression.27 The hemopoietic form of CD4426,28 shows consistently higher expression in the control cells than in the In(Lu) cells but the difference is much smaller than for LU (a maximum of 3-fold at day 14). Four additional erythroid genes tested by Q-PCR (AHSP, SLC4A1, ALAS2 and beta globin (HBB)) share similar patterns of expression. In control cells there is an increase in expression from day 4 to day 8 or 11 and then a reduction at day 14. The In(Lu) cells show a similar pattern but expression is lower at all but the last timepoint. For HBB, expression is always lower in the In(Lu) cells. For all 4 genes, the difference in expression between control and In(Lu) cells is greatest at day 4: 3.2-fold for AHSP, 17.1-fold for SLC4A1, 4.4-fold for ALAS2 and 5.2-fold for HBB (Figure 1 and Supplemental Figure S1). These results are consistent with the microarray findings and our hypothesis that In(Lu) erythroblasts have a general abnormality of transcription with respect to erythroid genes. In(Lu) erythroblasts have mutations in erythroid transcription factor EKLF. Since abnormal transcript levels were observed for multiple erythroid genes in In(Lu) type erythroblasts we looked for evidence of altered erythroid transcription factors. We amplified and sequenced exons from two control and two In(Lu) samples (gDNA or cDNA) for genes encoding several transcription factors functioning in erythropoiesis (GATA1, FOG1 (90% covered), EKLF, GFI1B, MYB, MZF1 (80% covered), TAL1, SPI1 (PU.1)). Only EKLF showed the presence of mutations unique to In(Lu) samples. We sequenced the three EKLF exons from 3 more cDNA samples (derived from In(Lu) erythroid cultures),18 more gDNA samples typed as Lu(a-b-) In(Lu) type and one unusual Lu(a-b-) (InLu24), derived from archived gDNA or archived blood samples (available in house). Mutations were detected in all but three of the samples (Table 2). In all cases the mutated allele was expressed with a normal EKLF allele. Nine different mutations were detected. Eight UK individuals had a single base insertion in exon 3 (Arg319GlufsX34). Six individuals from two Spanish families had a single base deletion in exon 2 (Pro190LeufsX47). Six of the nine mutations occur in that part of the EKLF gene encoding the zinc finger domains (Figure 2). One (Arg319GlufsX34) is predicted to cause a frameshift and another (Lys292X ) encodes a premature termination codon; all others result in single amino acid From www.bloodjournal.org by guest on June 15, 2017. For personal use only. substitutions (His299Tyr, Arg328Leu, Arg328His and Arg331Gly). The three mutations outside of the zinc finger domains result in a premature termination codon (Leu127X), a frameshift (Pro190LeufsX47) and an altered GATA1 binding site in the EKLF promoter (-124T>C). Two known single nucleotide polymorphisms (SNPs; dbSNP rs2072597 and dbSNP rs2072596) were also detected (Table 2). Interestingly, a third reported SNP (dbSNP rs35891202) has a single base deletion at the beginning of exon 3 which would be predicted to result in a frameshift affecting the entire zinc finger region. This SNP may therefore represent an EKLF mutation in another individual with the In(Lu) phenotype. We also sequenced all EKLF exons and the promoter region in 5 cDNA samples from control erythroid cultures and 32 control gDNAs from random blood donors. None of the mutations found in the In(Lu) samples were seen in the controls. EKLF mutations in InLu create a non-functional allele Mutation 1 disrupts a GATA1 binding site in the EKLF promoter (Table 2). The other mutations identified are located in the EKLF coding region. Several mutations must render the affected EKLF allele non-functional. Mutations 2 and 4 convert Leu127 and Lys292 respectively to termination codons resulting in translated proteins lacking all zinc finger (ZF) domains. Mutation 3 is a base deletion in the codon for Pro190 while mutation 6 is a base insertion in the codon for Arg319. Both mutations would alter the subsequent reading frame with consequent loss of all ZF domains (mutation 3) or loss of ZF domains 2 and 3 (mutation 6). Mutations 5, 7, 8 and 9 cause amino acid changes in or near the consensus sequence of a ZF domain. Mutation 5, which changes His299 (ZF1) to Tyr, is expected to diminish the binding of zinc which is integral to the overall protein conformation. Mutations 7 and 8 lead to substitution of Arg328 (ZF2) by His and Leu respectively. These changes are expected to disrupt the normal mode of DNA binding in which Arg328 is predicted to form two central hydrogen bonds with a guanine base (Figure 3). Mutation 9 substitutes Arg331 with Gly. In the model, Arg331makes a water-mediated interaction with a phosphate of the DNA backbone; this interaction is not possible when Gly is incorporated at this position (Figure 3). Discussion The molecular basis of the In(Lu) type of Lu(a-b-) blood group phenotype has not previously been established. Many families exhibiting this phenotype have been studied using serological methods and apparent dominant inheritance observed - hence InLu (inhibitor of Lutheran) to describe the gene responsible.14,15,33,34 The phenotype is of interest because its characteristic reduced red cell surface expression of antigens from several genetically independent blood group systems (i, P1 and Inb) in addition to Lutheran system antigens suggests abnormality in a pathway common to the biosynthesis of these blood group-active molecules (reviewed in Daniels35). The Inb antigen is carried on CD44 and defined by its amino acid sequence.25,26 In contrast the i and P1 antigens are carbohydrate antigens.36,37 In order to explore the molecular basis of this phenotype we compared gene expression profiles of normal and In(Lu) cultured erythroblasts. The results showed a clear difference between the two cell types with respect to transcript levels of LU and CD44 (Table 1, Figure 1), consistent with the serological profile of In(Lu) red cells. Lutheran expression occurs late during ex vivo erythropoiesis24 and the markedly reduced levels of LU transcript appear indicative of incomplete erythroid maturation. Further inspection of the gene expression data revealed a reduction in transcript levels of erythroid genes in general. This was particularly noticeable when erythroblasts from day 4 of culture were examined where marked reductions in transcript levels for ALAS2, globin genes and SLC4A1 (Band 3) were apparent (Supplemental Table S2, Figure 1 and Supplemental Figure S1). These data prompted us to search for mutations in erythroid transcription From www.bloodjournal.org by guest on June 15, 2017. For personal use only. factor genes in DNA from blood donors with the In(Lu) phenotype. Initial analysis of DNA from two individuals of the In(Lu) type revealed mutations in EKLF. Subsequently, we were able to obtain In(Lu) DNA from 24 individuals. Our data showing the occurrence of mutations in the promoter or coding region of EKLF in all but 3 of these individuals and the absence of any of these mutations in individuals of normal Lutheran phenotype provide persuasive evidence that the EKLF mutations are responsible for the In(Lu) phenotype in these individuals. All 21 In(Lu) samples in which we found EKLF mutations had the mutations on one allele whereas the other EKLF allele had a normal sequence (Table 2). EKLF comprises a proline-rich N-terminal region (amino acids 1-275) containing a transactivation domain and a C-terminal region (amino acids 276-358) containing three zinc finger domains essential for binding to the DNA consensus sequence 5'-CCACACCCT-3' (Figure 2).29,38,39 In mice, EKLF is expressed at high levels only in erythroid cells and abundant evidence supports the view that it is a key regulator of erythroid-specific gene expression (reviewed in Bieker40). EKLF knockout mice die at around embryonic day 14 due to severe anemia resulting from the failure to switch from embryonic to adult beta-globin expression.2,3 Murine EKLF is also required for the transcription of heme biosynthesis genes such as ALAS2 and AHSP, and for genes encoding erythroid membrane proteins including EPB49, DARC and ERMAP.7,8 Comparison of the gene expression profiles observed for In(Lu) with those reported for mice deficient in EKLF7,8 reveals numerous similarities with respect to reduced expression of EKLF-dependent erythroid genes (Table 1, Figure 1). Less information is available regarding transcript levels in EKLF +/- mice but beta globin and TER119 (murine glycophorin A) are expressed at intermediate levels and in this respect are similar to human In(Lu) cells (Figure 1).8,9 The 21 In(Lu) individuals studied exhibited 9 different EKLF mutations. Mutation 1 (-124T>C; Table 2) causes the loss of one of three possible GATA1 binding sites in the human EKLF promoter. Analysis of mouse EKLF promoter activity by transient transfection in MEL cells of promoter constructs in which one or other of three GATA1 binding sites were disrupted showed that one site was critical for EKLF expression while inactivation of the other sites had a relatively minor effect.41 In this context it is interesting to note that red cells from the individual with the promoter mutation (InLu24) had Lutheran antigens and Inb antigen expression weaker than normal red cells but noticeably stronger than most In(Lu) samples (J. Poole, IBGRL, e-mail, February 4, 2008) suggesting that some expression of EKLF occurs from the mutant allele in the absence of this GATA1 site. Eight different mutations were found in the EKLF coding sequence of individuals with the In(Lu) phenotype. Eight individuals had the Arg319GlufsX34 mutation and 6 had the Pro190LeufsX47 mutation; it seems likely these groups share common ancestors in the UK and Spain respectively. All other mutations were single examples. The In(Lu) phenotype appears to be benign with no associated pathology and little evidence of erythroid abnormality reported. The red cells of two In(Lu) siblings with normal hematological indices had a normal half life although they showed increased hemolysis on storage at 4oC.42 In one family the In(Lu) phenotype was associated with acanthocytosis.43 It would be of interest to examine EKLF in the latter family to see if the reported acanthocytosis is linked to a unique mutation in EKLF. By analogy with EKLF “knock-out” mice it would be predicted that inheritance of two InLu alleles would result in embryonic death and in this context it is interesting to note that none of the many reported families with In(Lu) type red cells have progeny of proven homozygosity with respect to InLu. Of particular interest is the NEA family described by Gibson34 in which a first cousin marriage produced 5 children all of whom had the In(Lu) phenotype. Four of these children had children of their own (who were not of the In(Lu) phenotype), proving that they were heterozygous. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. The presence of a mutated EKLF allele is easily detected by observing the almost complete absence of Lutheran blood group antigens on the red cells of affected individuals since most individuals with the Lu(a-b-) phenotype will be of the In(Lu) type. The dramatic reduction in Lutheran expression observed in In(Lu) cells may be related to its transcriptional activation being at a later stage in erythropoiesis than other erythroid-specific genes. This suggests that EKLF availability in In(Lu) becomes limiting at later stages of erythroid maturation (Figure 1, Supplemental Figure S1). In this context it is interesting to note that EKLF protein content declines as J2E cells mature in response to erythropoietin under conditions where globin synthesis is increasing rapidly.44 Spadaccini et al44 suggest lower levels of EKLF are needed to maintain transcription once initial activation has occurred. If this observation also applies to human erythropoeisis it may be that levels of EKLF in In(Lu) are too low for transcriptional activation of LU at day 8 of culture. Several potential EKLF binding sites have been identified in the LU promoter.45 We conclude that the In(Lu) phenotype is most likely caused by inheritance of a loss of function mutation on one allele of EKLF. Transcription of EKLF from the single normal allele results in reduced levels of EKLF and as a consequence, a general reduction in transcription of erythroid genes ensues. The absence of mutation in the EKLF coding sequence in 3 In(Lu) samples (InLu8, 9 and 23) suggests either the presence of a mutation in part of the gene not sequenced (enhancer or distal promoter regions), or the possibility of other causes of the In(Lu) phenotype. In several In(Lu) individuals the mutated copy of EKLF can only produce non-functional protein and so they are representative of the EKLF +/- phenotype. It can be inferred that this phenotype is physiologically benign in man as it is in mice. In other cases mutations change single amino acids in the region of zinc finger domains 1 and 2 and so, by inference, alter residues critical for normal EKLF function. The diversity of mutations within a small number of In(Lu) individuals is consistent with sporadic mutation. The In(Lu) phenotype therefore results from reduced transcription (as previously postulated42) rather than the action of a dominant inhibitor gene. These data provide the first evidence of naturally occuring mutations affecting human EKLF and the first description of transcription factor mutations directly responsible for a specific blood group phenotype. Acknowledgements This work was supported by the Department of Health (England). We thank Heather Davies and Alan Gray for the Lu(a-b-) samples and their controls, Peter Martin and Joanna Summers-Yeo for DNA sequencing, Jane Coghill for the microarray hybridisation service, and Vanja Karamatic Crew and Kirstin Finning for providing useful reagents. We are grateful to Geoff Daniels, Frances Spring and Anna Ribera for providing archived samples. We also thank Joyce Poole and Nicole Warke for serology data and Kevin Gaston and Lucio Luzzatto for helpful advice and discussions. Contribution: B.K.S. designed and conducted experiments and wrote the paper; N.M.B. and C.G. conducted experiments; R.L.B. designed experiments; D.J.A. designed experiments and wrote the paper. Conflict of interest disclosure: The authors declare no competing financial interests. Correspondence: Belinda K. Singleton, Bristol Institute for Transfusion Sciences, Southmead Road, Bristol, BS10 5ND, United Kingdom; e-mail: [email protected] From www.bloodjournal.org by guest on June 15, 2017. For personal use only. References 1. Welch JJ, Watts JA, Vakoc CR, et al. Global regulation of erythroid gene expression by transcription factor GATA-1. Blood. 2004;104:3136-3147. 2. Nuez B, Michalovich D, Bygrave A, Ploemacher R, Grosveld F. Defective haematopoiesis in fetal liver resulting from inactivation of the EKLF gene. Nature. 1995;375:316-318. 3. Perkins AC, Sharpe AH, Orkin SH. Lethal β-thalassaemia in mice lacking the erythroid CACCC-transcription factor EKLF. Nature. 1995;375:318-322. 4. Merika M, Orkin SH. Functional synergy and physical interactions of the erythroid transcription factor GATA-1 with the Krüppel family proteins SP1 and EKLF. Mol Cell Biol. 1995;15:2437-2447. 5. Gregory RC, Taxman DJ, Seshasayee D, Kensinger MH, Bieker JJ, Wojchowski DM. Functional interaction of GATA1 with erythroid Krüppel-like factor and Sp1 at defined erythroid promoters. Blood. 1996;87:1793-1801. 6. Weiss MJ, Orkin SH. Transcription factor GATA-1 permits survival and maturation of erythroid precursors by preventing apoptosis. Proc Natl Acad Sci U S A. 1995;92:96239627. 7. Drissen R, von Lindern M, Kolbus A, et al. The erythroid phenotype of EKLF-null mice: defects in hemoglobin metabolism and membrane stability. Mol Cell Biol. 2005;25:52055214. 8. Hodge D, Coghill E, Keys J, et al. A global role for EKLF in definitive and primitive erythropoiesis. Blood. 2006;107:3359-3370. 9. Wijgerde M, Gribnau J, Trimborn T, et al. The role of EKLF in human beta-globin gene competition. Genes Dev. 1996;10:2894-2902. 10. Brown F, Simpson S, Cornwall S, Moore BP, Oyen R, Marsh WL. The recessive Lu(a-b-) phenotype: a family study. Vox Sang. 1974;26:259-264. 11. Karamatic Crew V, Mallinson G, Green C, et al. Different inactivating mutations in the LU genes of three individuals with the Lutheran-null phenotype. Transfusion. 2007;47:492-498. 12. Norman PC, Tippett P, Beal RW. An Lu(a-b-) phenotype caused by an X-linked recessive gene. Vox Sang. 1986;51:49-52. 13. Taliano V, Guévin R-M, Tippett P. The genetics of a dominant inhibitor of the Lutheran antigens. Vox Sang. 1973;24:42-47. 14. Shaw MA, Leak MR, Daniels GL, Tippett P. The rare Lutheran blood group phenotype Lu(a-b-): a genetic study. Ann Hum Genet. 1984;48:229-237. 15. Rowe GP, Gale SA, Daniels GL, Green CA, Tippett P. A study on Lu-null families in South Wales. Ann Hum Genet. 1992;56:267-272. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. 16. Fricke B, Parsons SF, Knöpfle G, von Düring M, Stewart GW. Stomatin is mis-trafficked in the erythrocytes of overhydrated hereditary stomatocytosis, and is absent from normal primitive yolk sac-derived erythrocytes. Br J Hematol. 2005;131:265-277. 17. Affymetrix. GeneChip expression analysis: data analysis fundamentals. http://www.affymetrix.com/support/downloads/manuals/data_analysis_fundamentals_manu al.pdf. Accessed January 31, 2008. 18. Real-Time PCR Applications Guide. Bio-Rad Laboratories, Inc.; 2005. 19. Elrod-Erickson M, Rould MA, Nekludova L, Pabo CO. Zif268 protein-DNA complex refined at 1.6A: a model system for understanding zinc finger-DNA interactions. Structure. 1996;4:1171-1180. 20. Sali A, Blundell TL. Comparative protein modelling by satisfaction of spatial restraints. J Mol Biol. 1993;234:779-815. 21. Wang J, Cieplak P, Kollman PA. How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J Comput Chem. 2000; 21:1049-1074. 22. Lovell SC, Davis IW, Arendall WB 3rd et al. Structure validation by Calpha geometry: phi, psi and Cbeta deviation. Proteins. 2003;50:437-450. 23. Feng Z, Westbrook J, Berman HM. Report NDB-407: NUCheck. New Brunswick, NJ: Rutgers University;1998. 24. Southcott MJ, Tanner MJ, Anstee DJ. The expression of human blood group antigens during erythropoiesis in a cell culture system. Blood. 1999;93:4425-4435. 25. Telen MJ, Eisenbarth GS, Haynes BF. Human erythrocyte antigens: regulation of expression of a novel erythrocyte surface antigen by the inhibitor Lutheran In(Lu) gene. J Clin Invest. 1983;71:1878-1886. 26. Spring FA, Dalchau R, Daniels GL, et al. The Ina and Inb blood group antigens are located on a glycoprotein of 80 000 MW (the CDw44 glycoprotein) whose expression is influenced by the In(Lu) gene. Immunology. 1988;64:37-43. 27. Stanbury A, Francis B. The Lu(a-b-) phenotype: an additional example. Vox Sang. 1967;13:441-443. 28. Screaton GR, Bell MV, Jackson DG, Cornelis FB, Gerth U, Bell JI. Genomic structure of DNA encoding the lymphocyte homing receptor CD44 reveals at least 12 alternatively spliced exons. Proc Natl Acad Sci USA. 1992;89:12160-12164. 29. Feng WC, Southwood CM, Bieker JJ. Analyses of β-thalassemia mutant DNA interactions with erythroid Krüppel-like factor (EKLF), an erythroid cell-specific transcription factor. J Biol Chem. 1994;269:1493-1500. 30. National Center for Biotechnology Information. HomoloGene. http://www.ncbi.nlm.nih.gov/sites/homologene. Accessed December 18, 2007. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. 31. Klevit RE. Recognition of DNA by Cys2, His2 zinc fingers. Science. 1991;253:1367-1395. 32. Berg JM. Proposed structure for the zinc-binding domains from transcription factor IIIA and related proteins. Proc Natl Acad Sci USA. 1988;85:99-102. 33. Race RR, Sanger R. Blood Groups in Man, 6th edition, Oxford: Blackwell Scientific Publications; 1975:267-272. 34. Gibson T. Two kindred with the rare dominant inhibitor of the lutheran and p1 red cell antigens. Hum Hered. 1976;26:171-174. 35. Daniels G. Human Blood Groups. Oxford: Blackwell Science; 2002. 36. Feizi T. The blood group Ii system: a carbohydrate antigen system defined by naturally monoclonal or oligoclonal auto-antibodies of man. Immunol Commun. 1981;10:127-156. 37. Cory HT, Yates AD, Donald ASR, Watkins WM, Morgan WTJ. The nature of the human blood group P1 determinant. Biochem Biophys Res Comm. 1974;61:1289-1296. 38. Miller IJ, Bieker JJ. A novel, erythroid cell-specific murine transcription factor that binds to the CACCC element and is related to the Krüppel family of nuclear proteins. Mol Cell Biol. 1993;13:2776-2786. 39. Bieker JJ, Southwood CM. The erythroid Krüppel-like factor transactivation domain is a critical component for cell-specific inducibility of a β-globin promoter. Mol Cell Biol. 1995;15:852-860. 40. Bieker JJ. Probing the onset and regulation of erythroid cell-specific gene expression. Mt Sinai J Med. 2005;72:333-338. 41. Crossley M, Tsang AP, Bieker JJ, Orkin SH. Regulation of the erythroid Krüppel-like factor (EKLF) gene promoter by the erythroid transcription factor GATA-1. J Biol Chem. 1994;269:15440-15444. 42. Ballas SK, Marcolina MJ, Crawford MN. In vitro storage and in vivo survival studies of red cells from persons with the In(Lu) gene. Transfusion. 1992;32:607-611. 43. Udden MM, Umeda M, Hirano Y, Marcus DM. New abnormalities in the morphology, cell surface receptors, and electrolyte metabolism of In(Lu) erythrocytes. Blood. 1987;69:52-57. 44. Spadaccini A, Tilbrook PA, Sarna MK, Crossley M, Bieker JJ, Klinken SP. Transcription factor erythroid Krüppel-like factor (EKLF) is essential for the erythropoietin-induced hemoglobin production but not for proliferation, viability or morphological maturation. J Biol Chem. 1998;273:23793-23798. 45. El Nemer W, Rahuel C, Colin Y, Gane P, Cartron JP, Le Van Kim C. Organisation of the human LU gene and molecular basis of the Lua/Lub blood group polymorphism. Blood. 1997;89:4608-4616. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Table 1: Selected genes that are differentially expressed between In(Lu) and control cells Gene Symbol Gene Title Antigen processing and presentation ARTS-1 Type 1 tumor necrosis factor receptor shedding aminopeptidase regulator HLA-A Major histocompatibility complex, class I, A HLA-B Major histocompatibility complex, class I, B HLA-DPB1 Major histocompatibility complex, class II, DP beta 1 HLA-DRA Major histocompatibility complex, class II, DR alpha HLA-DRB1 Major histocompatibility complex, class II, DR beta 1 HLA-DRB4 Major histocompatibility complex, class II, DR beta 4 HLA-G HLA-G histocompatibility antigen, class I, G Defense response ECRP Eosinophil cationic related protein ELA2 Elastase 2, neutrophil FCER1A Fc fragment of IgE, high affinity I, receptor for; alpha polypeptide IL4R Interleukin 4 receptor MPO Myeloperoxidase NCF4 Neutrophil cytosolic factor 4, 40kDa PRG3 Proteoglycan 3 RNASE2 Ribonuclease, RNase A family, 2 (liver, eosinophil-derived neurotoxin) TPSAB1 Tryptase alpha/beta 1 TPSAB1 / Tryptase alpha/beta 1 / tryptase beta 2 TPSB2 TPSB2 Tryptase beta 2 Heme / globin synthesis AHSP Alpha hemoglobin stabilising protein ALAD Aminolevulinate, delta-, dehydratase ALAS2 Aminolevulinate, delta-, synthase 2 (sideroblastic/hypochromic anemia) HBE1 Hemoglobin, epsilon 1 UROD Uroporphyrinogen decarboxylase Blood group antigens / erythroid factors ANK1 Ankyrin 1, erythrocytic AQP1 Aquaporin 1 (Colton blood group) ARG1 Arginase, liver ART4 ADP-ribosyltransferase 4 (Dombrock blood group) BCAM Basal cell adhesion molecule (Lutheran blood group) BSG Basigin (Ok blood group) CD44 CD44 molecule (Indian blood group) DARC Duffy blood group, chemokine receptor EPB42 Erythrocyte membrane protein band 4.2 EPOR Erythropoietin receptor ERMAP Erythroblast membrane-associated protein (Scianna blood group) GYPA Glycophorin A (MNS blood group) NFE2 Nuclear factor (erythroid-derived 2), 45kDa PKLR Pyruvate kinase, liver and RBC SLC4A1 Solute carrier family 4, anion exchanger, member 1 (erythrocyte membrane protein band 3, Diego blood group) TMOD1 Tropomodulin 1 Day 6 fold change Day 11 fold change 4.6 1.5 1.1 1.8 2.0 1.9 1.7 1.2 2.4 1.6 2.1 1.6 1.8 1.3 1.0 1.8 2.5 3.4 1.8 2.2 3.0 2.1 1.5 1.7 4.6 4.5 3.0 2.8 1.2 2.2 1.6 1.5 1.6 1.8 3.9 4.0 5.0 3.5 -1.8 -2.1 -2.1 -1.8 -1.6 -1.1 -1.3 -1.0 -1.3 -1.2 -1.6 -3.0 -2.3 -2.6 -1.5 -1.7 -1.7 -1.7 -1.8 -1.9 -1.5 -1.5 -1.5 -2.6 -3.1 -1.2 -1.7 -1.6 -1.3 -3.5 -1.3 -1.9 -1.3 -1.1 -1.3 -1.4 -1.1 -1.2 -1.7 -1.1 -2.0 -1.3 Differentially expressed genes are grouped according to gene ontology terms showing enrichment. Separate values are shown for fold differences in expression (In(Lu) value divided by control value) in cells at day 6 and at day 11. Genes showing higher expression in In(Lu) cells than in controls have positive values; genes showing lower expression in In(Lu) cells than in controls have negative values. Average values have been taken for genes that are represented more than once on the arrays. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Table 2: Sequence variants detected in the EKLF gene in erythroid cells of the In(Lu) phenotype Variant Sample type name 1 InLu24 2 InLu10 3 InLu17 InLu18 InLu19 InLu20 InLu21 InLu22 4 InLu6 5 InLu16 6 InLu1 InLu3 InLu4 InLu5 InLu7 InLu13 InLu14 InLu15 7 InLu11 8 InLu12 9 InLu2 10 InLu8 InLu9 InLu23 □ DNA change□ Deduced protein change† [=]+[-124T>C] ‡ [=]+[380T>A] [=]+[569delC] [=]+[?] [=]+[Leu127X] [=]+[Pro190LeufsX47] [=]+[874A>T] [=]+[895C>T] [=]+[954dupG] [=]+[Lys292X] [=]+[His299Tyr] [=]+[Arg319GlufsX34] [=]+[983G>T] [=]+[983G>A] [=]+[991C>G] [=]+[?] [=]+[Arg328Leu] [=]+[Arg328His] [=]+[Arg331Gly] [=]+[?] dbSNP rs2072597 T/C T/C C/C C/C T/C T/C C/C T/C T/T T/C T/T T/T T/T T/T T/C T/T T/C T/C T/T C/C T/C T/C T/C T/T dbSNP rs2072596 T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/T T/C T/T T/C T/T T/T DNA changes are numbered relative to the translation initiation codon (where A is +1) in the reference cDNA sequence (Entrez Gene Accession number NM_006563.2). [=] indicates the presence of the wild-type allele. [?] indicates an unknown change. †Deduced protein changes are numbered relative to the translation initiator (where Met is +1) in the reference protein sequence (Entrez Protein Accession number NP_006554.1). [=] indicates presence of wild-type protein. [?] indicates unknown effect. ‡Promoter sequence can be obtained from the reference gDNA sequence (Entrez Gene Accession number NT_086897.1: complement of region 4090501 – 4093981). From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Figure legends Figure 1. Validation of selected microarray results by Q-PCR. Changes in target gene expression during ex vivo erythropoiesis are depicted. The vertical axis represents the target gene expression normalised to PABPC1 expression and relative to the UR/FL calibrator (as described in “Materials and Methods”). Black lines indicate average expression in control erythroblasts (n=4); gray lines indicate average expression in In(Lu) erythroblasts (n=4). Figure 2. Sequence variants located in the zinc finger domains of EKLF. Alignment of the zinc finger domains (ZF1 – ZF3) from EKLF is compared with ZIF268 (based on Feng et al29(Fig4)). Protein sequences of genes identified as putative homologs of EKLF by NCBI’s HomoloGene30 are from Homo sapiens (h), Pan troglodytes (p), Canis lupus familiaris (c), Mus musculus (m) and Rattus norvegicus (r). Boxed residues indicate XYZ amino acids required for DNA sequence recognition;31 bold residues show matches to the classical CC/HH motif;32 and circled numbers indicate positions of sequence variant types as described in Table 2. Figure 3. Homology model of the EKLF-DNA complex. (A) Overall structure of the complex. DNA is shown as an orange molecular surface; protein is shown as a gray Cα cartoon. The amino terminus (Ala278) is labelled N; the carboxy terminus (Leu362) is labelled C. The side chains of mutated residues are highlighted as sticks with the carbon atoms of Arg328 and Arg331 colored green and cyan, respectively. (B) Close-up of the ZF2 domain highlighting protein-DNA interactions involving mutated residues. DNA is shown as sticks colored by atom type. Protein is displayed as in (A). The bound water molecule is shown as a red sphere, and putative hydrogen bonds are shown as dashed yellow lines. From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Figure 1 CD44H 60 Expression relative to UR/FL (x100) Expression relative to UR/FL (x1000) LU 90 80 70 60 50 40 30 20 10 0 50 40 30 20 10 0 4 6 8 10 12 14 4 6 Days in culture AHSP 120 Expression relative to UR/FL (x100) Expression relative to UR/FL (x10) 14 SLC4A1 25 20 15 10 5 0 100 80 60 40 20 0 4 6 8 10 12 Days in culture 14 4 6 ALAS2 8 10 12 Days in culture 14 HBB 14 80 12 70 Expression relative to UR/FL (x10) Expression relative to UR/FL (x10) 8 10 12 Days in culture 10 8 6 4 2 0 60 50 40 30 20 10 0 4 6 8 10 Days in culture 12 14 4 6 8 10 Days in culture 12 14 From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Figure 2 4 5 278 ZF1 A A A A A P H H H H H Y T T T T T A C C C C C C A A T G G P H H H H H V P P P E E E G G G G G S C C C C C C G G G G G D K K K K K R S S S S S R Y Y Y Y Y F T T T S T S K K K K K R S S S S S S S S S S S D H H H H H E L L L L L L K K K K K T A A A A A R L L L L L I 8 7 6 H H H H H H R R R R R R T T T T T I H H H H H H T T T T T T G G G G G G E E E E E Q K K K K K K hEKLF pEKLF cEKLF mEKLF rEKLF ZIF268 K K K K K T H H H H H H T T T T T T G G G G G G Q Q Q H H E R R R R R K hEKLF pEKLF cEKLF mEKLF rEKLF ZIF268 9 308 ZF2 P P P P P P Y Y Y Y Y F A A A A A Q C C C C C C T T T S S W W W W W R E E D D D I G G G G G C C C C C C G G G D N M W W W W W R R R R R R N F F F F F F A A A A A S R R R R R R S S S S S S D D D D D F E E E E E H L L L L L L T T T T T T R R R R R T H H H H H H Y Y Y Y Y I R R R R R R F F F F F F R R R C C A C C C C C C Q Q Q G G D L L L L L I C C C C C C P P P P P G R R R R R R A A A A A K F F F F F F S S S S S A R R R R R R S S S S S S D D D D D D H H H H H E L L L L L R A A A A A K L L L L L R H H H H H H M M M M M T K K K K K K R R R R R I H H H H H H 338 ZF3 P P P P P P 362 L L L L L hEKLF pEKLF cEKLF mEKLF rEKLF ZIF268 From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Figure 3 From www.bloodjournal.org by guest on June 15, 2017. For personal use only. Prepublished online May 16, 2008; doi:10.1182/blood-2008-03-145672 Mutations in EKLF/KLF1 form the molecular basis of the rare blood group In(Lu) phenotype Belinda K. Singleton, Nicholas M. Burton, Carole Green, R. Leo Brady and David J. Anstee Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Advance online articles have been peer reviewed and accepted for publication but have not yet appeared in the paper journal (edited, typeset versions may be posted when available prior to final publication). Advance online articles are citable and establish publication priority; they are indexed by PubMed from initial publication. Citations to Advance online articles must include digital object identifier (DOIs) and date of initial publication. Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.