A dual reporter mouse model of the human β-globin locus: applications and limitations

Abstract

The human β-globin locus contains the β-like globin genes (i.e. fetal γ-globin and adult β-globin), which heterotetramerize with α-globin subunits to form fetal or adult hemoglobin. Thalassemia is one of the commonest inherited disorders in the world, which results in quantitative defects of the globins, based on a number of genome variations found in the globin gene clusters. Hereditary persistence of fetal hemoglobin (HPFH) also caused by similar types of genomic alterations can compensate for the loss of adult hemoglobin. Understanding the regulation of the human γ-globin gene expression is a challenge for the treatment of thalassemia. A mouse model that facilitates high-throughput assays would simplify such studies. We have generated a transgenic dual reporter mouse model by tagging the γ- and β-globin genes with GFP and DsRed fluorescent proteins respectively in the endogenous human β-globin locus. Erythroid cell lines derived from this mouse model were tested for their capacity to reactivate the γ-globin gene. Here, we discuss the applications and limitations of this fluorescent reporter model to study the genetic basis of red blood cell disorders and the potential use of such model systems in high-throughput screens for hemoglobinopathies therapeutics.

Figure 1. Modification of the human β-globin locus and generation of transgenic dual reporter mouse lines.

(A) The human β-globin locus (SceI flanked) PAC used for the modifications made in the γ- and β-globin genes. GFP and DsRed were introduced in the ATG (+1) position of the transcripts followed by a stop codon (*). (B) Schematic representation of the GPA-GFP construct. The grey numbered stretches of the cartoon (1–114, 834–1234 bp) represent the glycophorin-A cDNA and the green stretch represents the GFP cDNA (114–834). The bilayer represents the transmembrane part of the protein, thus the GFP is expressed in the extracellular part of the fusion GPA-GFP protein. (C) Representative picture of K562 cells transfected with the γGFP/βDsRed modified human β-globin locus to check expression of γ-globin and flow cytometry analysis of GFP expression in 12.5dpc embryonic blood of γGFP/βDsRed transgenic embryos (left). Representative picture of fetal liver cells transduced with the γGPA-GFP construct to check expression of GFP protein in the plasma membrane and flow cytometry analysis of GFP expression in 12.5dpc embryonic blood of γGPA-GFP/βDsRed transgenic embryos (right). Mean fluorescence intensity (MFI) ratio is indicated in both graphs. (D) Southern blot of both mouse transgenic lines (γGFP/βDsRed and γGPA-GFP/βDsRed). Tail genomic DNA was digested with SacI restriction enzyme and hybridized with cosLCRε (left) and cosγγδβ (right), as previously described . Lane 1: γGPA-GFP/βDsRed tail DNA, Lanes 2, 3: mouse line PAC8 carrying the human β-globin locus and Lane 4: γGFP/βDsRed tail DNA. Symbol ▹ indicates end fragments, ▸ HGG1 3.6 Kb SacI fragment, ▸▸ HGG1-GFP 4.3 Kb SacI fragment, ▸▸▸ HGG1-GPA-GFP 4.9 Kb SacI fragment, ⧫ β-DsRed modification (16.4 to 17 Kb fragment). (E) S1 nuclease protection analysis of mouse globin expression of WT and transgenic mice at different developmental stages as indicated.

Figure 2. Analysis of GFP (γ-globin) expression during development.

(A) Histogram overlay of embryonic blood from transgenic and WT embryos in the GFP axis. The percentages of positive GFP cells and primitive cells for each developmental stage are included in the table underneath (SSC: side scatter). Differences in the MFI of WT cells amongst the three histograms shown are due to the developmental stage and the gated cells plotted (i.e. SSC^high in the third histogram). (B) Dot plot depicting embryonic blood of representative γGPA-GFP/βDsRed transgenic mice at 14.5dpc in which the GFP⁺ population is depicted as green dots against the Forward (FSC) and Side Scatter (SSC). GFP⁺ cells (8.62±2.68%) are SSC^high, i.e. primitive cells. (C) qPCR analysis of GFP expression in WT and transgenic fetal liver cells at 11.5dpc and 14.5dpc. RFE is relative fold enrichment. Average and standard deviation derived from three mice per group is depicted. T-test was performed to calculate the p values.

Figure 3. Analysis of DsRed (β-globin) expression during development.

(A) Flow cytometry analysis of fetal liver of γGPA-GFP/βDsRed transgenic mice during development. Arrowhead at 11.5dpc and 12.5dpc indicates the DsRed positive population. Representative data are depicted. (B) qPCR analysis of DsRed expression in WT and transgenic fetal liver cells at 11.5dpc and 14.5dpc. RFE is relative fold enrichment. Average and standard deviation derived from three mice per group is depicted. T-test was performed to calculate the p values. (C) Histogram overlays of DsRed and GFP expression in 14.5dpc fetal liver cells cultured for 2 days in hanging drops. DsRed expression is detected in transgenic cells differentiated in vitro when compared to WT while GFP is not. Representative data are depicted.

Figure 4. In vivo treatment of transgenic mice with 5-Azacytidine.

(A) Graph representing absolute numbers of GFP⁺ cells/10⁶ events measured gated in blood of γGPA-GFP/βDsRed mice treated (PHZ+AZA, red dots) or mock treated (PBS, white dots). (B) Graph representing absolute numbers of GFP⁺ cells/10⁵ events measured after 2 days of bone marrow hanging drop culture (BM HD D2) derived from γGPA-GFP/βDsRed mice treated (PHZ+AZA, orange dots) or mock treated (PBS, white dots). A logarithmic scale is used to better visualize the distribution of values found (each dot represents a mouse). The average per group is depicted as a black line. P values were calculated from the Log transformed data with a T-test.

Figure 5. Characterization of the transgenic dual reporter fetal liver cell lines.

(A) Flow cytometry analysis of transgenic fetal liver cell lines before and after differentiation. Histograms against forward scatter and erythroid surface markers CD117 (cKit) and CD71 (transferrin receptor) are depicted. (B) Flow cytometry analysis of transgenic fetal liver cell lines before and after differentiation. Histograms against DsRed and GFP are depicted. (C) Representative pictures taken during erythroid differentiation of transgenic fetal liver cell lines. Arrows indicate spontaneously differentiating cells expressing DsRed protein (left) and differentiated cells with much smaller size that are not as bright as the bigger ones, as a consequence of the continuous production of endogenous hemoglobin and subsequent quenching of cytoplasmic DsRed fluorescent signal (right).

Figure 6. Knockdown assays in transgenic dual reporter fetal liver cell lines.

(A) Flow cytometry analysis of the knockdown of cMyb, Bcl11a, Hdac3 and Fop in the γGFP/βDsRed cell line. The same vector with a non-specific shRNA sequence was used as a control. Percentages of cells positive for GFP (upper panel) and DsRed (lower panel) are shown. Contour plots show gated live cells. (B) Western blots of the knockdown experiments in protein extracts of transduced cells. Equal numbers of cells are loaded on each lane. φ, empty vector control extracts; KD, knockdown extracts.