![Figure 1. Gfi1aa and wild-type gfi1b transcript levels are severely reduced in primitive erythroblasts of qmc551 homozygous and gfi1b MO-injected zebrafish embryos, respectively. (A) Diagram summarizing the experimental strategy of the transcriptome analysis. (B) Image of a 20-hpf gfi1aaqmc551het zebrafish embryo taken on a fluorescent dissecting microscope with a fluorescein isothiocyanate filter set. Brackets highlight GFP+ cells in the intermediate cell mass (ICM) and in the posterior blood island (PBI). (C) Quality checks performed on the sequencing reads using the SeqMonk DNA Sequence Analyzer. The analysis gives the percentage of reads that map to annotated genes and the percentage of the latter that are located in exons, as well as the percentage of reads that match ribosomal RNA (rRNA) or mitochondrial DNA (MT) sequences. As expected for paired-end reads, 50% of the reads represent the sense strand. (D) Tracks showing the number of raw base read counts over the nucleotide sequences of exons 2 to 7 of gfi1aa in the representative samples Het1 and Hom3. These 2 samples had similar overall numbers of sequence reads (63.1 and 64.7 million reads, respectively) and can, therefore, be compared without additional normalization. (E) Tracks showing the raw base counts over the sequence of gfi1b in representative samples Het1, HetMO3, Hom3, and HomMO1. For the morphant samples HetMO1-3 and HomMO1-3, raw base counts over 10 nucleotides at the end of exon 3 (EE3), the beginning of exon 4 (BE4), the end of exon 4 (EE4), and the start of exon 5 (BE5) were averaged. The ratios of BE4/EE3 and EE4/BE5 were calculated and averaged. Similarly, the ratios of BE5/EE3 were determined and averaged. The average ratios are shown above the arrows in the figure. They give a good estimate of the ratio of gfi1b transcripts with and without exon 4 in the 2 morphant samples. To generate the images shown in panels D-E, the BAM files were imported into the Integrated Genome Viewer (IGV_2.3.72). (F-H) Bar charts comparing the average normalized transcript levels (read per kilobase of transcript per 1 million mapped reads [RPKM values]) for gfi1b (F), gfi1aa (exons 2-7, G), and qmc551:gfp (gfp sequence only, H) in 551het, 551hetMO, 551hom, and 551homMO samples. Sequence reads over exon 1 of gfi1aa were excluded as they cannot be allocated unambiguously to the gfi1aa or the qmc551:gfp transcript. On GraphPad Prism, Welch’s t test was used to examine the statistical significance of the differences between the mean expression values (*P<.05; **P<.01; ***P<.001; ****P<.0001). FAC, fluorescence-activated cell.](https://ash.silverchair-cdn.com/ash/content_public/journal/bloodadvances/2/20/10.1182_bloodadvances.2018020156/4/advances020156f1.png?Expires=1724921639&Signature=ikGRpRqUDXjFjnEt4NREP6lRBttc1LhA~M59rhMdb2HuacjK0OOZYZDIRnhj~QR~NTTbvAArCvPPttIaq1cWzpZXPvogXVHc450K5nCinKrYDxEQQOSH~Rl8DhJ-d0Q4fet6zlzM9Km5eC7OEM~d0m0QDIoTxpEC7Ow2WWSzcZSnDs~bcsp1Fj~JxkjBZxwyGW6~WEofFmtASZr0Mua~lwBJURGGe3iYEyKwAaVdM0VCDwo0lshMAebcoQLyxJQ-6chT4S5iLjdK~Bz4jCpMH4DReZ4mA5IMgwlds~Agi4y1tOMt90ZgHlAgzRgMkL-gBcCuYtERmqeAnz4U7tNcng__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Gfi1aa and wild-type gfi1b transcript levels are severely reduced in primitive erythroblasts of qmc551 homozygous and gfi1b MO-injected zebrafish embryos, respectively. (A) Diagram summarizing the experimental strategy of the transcriptome analysis. (B) Image of a 20-hpf gfi1aaqmc551het zebrafish embryo taken on a fluorescent dissecting microscope with a fluorescein isothiocyanate filter set. Brackets highlight GFP+ cells in the intermediate cell mass (ICM) and in the posterior blood island (PBI). (C) Quality checks performed on the sequencing reads using the SeqMonk DNA Sequence Analyzer. The analysis gives the percentage of reads that map to annotated genes and the percentage of the latter that are located in exons, as well as the percentage of reads that match ribosomal RNA (rRNA) or mitochondrial DNA (MT) sequences. As expected for paired-end reads, 50% of the reads represent the sense strand. (D) Tracks showing the number of raw base read counts over the nucleotide sequences of exons 2 to 7 of gfi1aa in the representative samples Het1 and Hom3. These 2 samples had similar overall numbers of sequence reads (63.1 and 64.7 million reads, respectively) and can, therefore, be compared without additional normalization. (E) Tracks showing the raw base counts over the sequence of gfi1b in representative samples Het1, HetMO3, Hom3, and HomMO1. For the morphant samples HetMO1-3 and HomMO1-3, raw base counts over 10 nucleotides at the end of exon 3 (EE3), the beginning of exon 4 (BE4), the end of exon 4 (EE4), and the start of exon 5 (BE5) were averaged. The ratios of BE4/EE3 and EE4/BE5 were calculated and averaged. Similarly, the ratios of BE5/EE3 were determined and averaged. The average ratios are shown above the arrows in the figure. They give a good estimate of the ratio of gfi1b transcripts with and without exon 4 in the 2 morphant samples. To generate the images shown in panels D-E, the BAM files were imported into the Integrated Genome Viewer (IGV_2.3.72). (F-H) Bar charts comparing the average normalized transcript levels (read per kilobase of transcript per 1 million mapped reads [RPKM values]) for gfi1b (F), gfi1aa (exons 2-7, G), and qmc551:gfp (gfp sequence only, H) in 551het, 551hetMO, 551hom, and 551homMO samples. Sequence reads over exon 1 of gfi1aa were excluded as they cannot be allocated unambiguously to the gfi1aa or the qmc551:gfp transcript. On GraphPad Prism, Welch’s t test was used to examine the statistical significance of the differences between the mean expression values (*P<.05; **P<.01; ***P<.001; ****P<.0001). FAC, fluorescence-activated cell.