Abstract

Alternative pre-messenger RNA splicing remodels the human transcriptome in a spatiotemporal manner during normal development and differentiation. Here we explored the landscape of transcript diversity in the erythroid lineage by RNA-seq analysis of five highly purified populations of morphologically distinct human erythroblasts, representing the last four cell divisions before enucleation. In this unique differentiation system, we found evidence of an extensive and dynamic alternative splicing program encompassing genes with many diverse functions. Alternative splicing was particularly enriched in genes controlling cell cycle, organelle organization, chromatin function and RNA processing. Many alternative exons exhibited differentiation-associated switches in splicing efficiency, mostly in late-stage polychromatophilic and orthochromatophilic erythroblasts, in concert with extensive cellular remodeling that precedes enucleation. A subset of alternative splicing switches introduces premature translation termination codons into selected transcripts in a differentiation stage-specific manner, supporting the hypothesis that alternative splicing-coupled nonsense-mediated decay contributes to regulation of erythroid-expressed genes as a novel part of the overall differentiation program. We conclude that a highly dynamic alternative splicing program in terminally differentiating erythroblasts plays a major role in regulating gene expression to ensure synthesis of appropriate proteome at each stage as the cells remodel in preparation for production of mature red cells.

INTRODUCTION

Alternative pre-messenger RNA (mRNA) splicing enables individual genes to generate multiple protein products that differ structurally and functionally by insertion or deletion of important functional domains encoded by alternative exons. During normal development and differentiation, dynamic changes in the expression or activity of the splicing regulatory machinery coordinately modulate networks of alternative splicing events in a spatiotemporal manner. Post-transcriptional RNA processing can thereby modulate essential protein functions according to the physiological requirements of the cell by regulating coherent biological pathways (1). Conversely, network perturbations caused by mutations that alter splicing factor expression or activity underlie an array of complex human diseases (2). The experimental work supporting these concepts has been performed primarily in non-hematologic cell types. However, recent studies have revealed that normal T cells execute a complex splicing program (3) and that splicing factor mutations are associated with hematological cancers, including myelodysplasia (4–6), demonstrating the importance of alternative splicing in hematology. Characterization of the alternative splicing program in human erythroblasts undergoing terminal erythroid differentiation will reveal novel insights into RNA regulatory pathways that drive cell differentiation and provide a basis for identifying splicing defect in human erythroid diseases.

Alternative isoforms of various erythroid transcripts have been reported, and a modest number of differentiation stage-specific switches in alternative splicing patterns are known (7). Best studied mechanistically is the upregulation of splicing efficiency for protein 4.1R alternative exon 16, which encodes part of the spectrin–actin binding domain required for optimal assembly of a mechanically competent membrane skeleton (8,9), an essential structure for mature erythrocyte function. This exon is predominantly skipped in early erythroblasts but included efficiently in late erythropoiesis (10,11), and several RNA binding proteins/splicing factors that influence exon 16 splicing have been identified (12–15). Underscoring the importance of this splicing switch, failure to include exon 16 causes mechanically unstable red cells and aberrant elliptocytic phenotype with anemia (16,17).

To comprehensively explore the alternative splicing landscape during terminal erythroid differentiation, we used an RNA-seq strategy to analyze and compare transcriptomes of highly purified human erythroblasts cultured in vitro from CD34+ cord blood progenitors (18). RNA-seq allows a robust high-resolution assessment of the transcriptome, but there remain computational challenges in data interpretation. Using extensions of current transcript abundance estimation tools combined with non-parametric statistical methods, we found an extensive alternative splicing program that is significantly enriched in genes controlling cell cycle, organelle organization, chromatin function and RNA processing. Importantly, hundreds of these alternative splicing events are regulated in a differentiation stage-specific manner, with most switches in exon inclusion/exclusion efficiency occurring in late-stage erythroblasts concurrent with extensive cellular remodeling as erythroblasts transition from a highly proliferative state to a terminally differentiated state. Finally, we discovered a subset of splicing switches that introduce premature translation termination codons (PTCs), thus decreasing the proportion of full-length coding mRNAs and downregulating gene expression via induction of nonsense-mediated decay (NMD). Post-transcriptional RNA processing pathways may be regulated by this mechanism in late erythroid differentiation, as numerous RNA binding proteins exhibit elevated expression of PTC exons in the most mature erythroblasts. These studies support the central hypothesis that human erythroblasts execute a surprisingly complex, differentiation stage-specific alternative splicing program that is essential for normal proliferation and differentiation.

MATERIALS AND METHODS

Human erythroblast populations

CD34+ cells were enriched from cord blood, cultured under conditions that promote erythroblast differentiation, and further enriched for stage-specific erythroblast populations by fluorescence-activated cell sorting (FACS) exactly as described (18). By gating narrow windows, high cell purity is attained for five discrete erythroblast populations (>90%), representing the last four cell divisions in terminal erythropoiesis. In some experiments, the cell populations enriched for proerythroblasts (proE) (culture day 9) and orthochromatic erythroblasts (orthoEs) (culture day 16) were treated with 100 μg/ml cycloheximide for 4 h to inhibit nonsense-mediated decay of RNA transcripts containing PTCs.

RNA-seq analysis

Total RNA was extracted using an RNeasy Plus Mini Kit (QIAGEN) from sorted human erythroblasts at distinct stages of erythroid maturation. The integrity of the RNA was evaluated on an Agilent Bioanalyzer, and samples with RNA integrity numbers (RIN) >8.0 were prepared for sequencing. Poly(A)-tailed RNA was subsequently prepared by the Epigenomics Core, Cornell Medical College, using the mRNA Seq Sample Prep Kit (Illumina Inc., San Diego, CA, USA) and used to create libraries for HiSeq2000 sequencing (Illumina).

Hundreds of millions of 50-nt RNA-seq reads were obtained from each of the 15 erythroblast samples, representing five distinct stages of differentiation from three independent sorting experiments. RNA-seq fragments were aligned to the Ensembl-annotated transcriptome version 67 (19) [hg19 human genome build GRCh37 (20)] using the Bowtie aligner (21). Read statistics are given in Supplementary Table S1. Transcript-level estimates were obtained using the transcript abundance estimation tool, eXpress (22), and were expressed as fragments per kilobase of exon per million fragments mapped (FPKM) as described previously (23,24). For each exon of every Ensembl gene, transcripts from the gene were divided into two sets: the set of isoforms including the exon (termed inclusion) and the set of isoforms with an intronic region spanning it (termed overlap). Expression of individual exons in ‘exon-inclusion’ isoforms relative to the total expression of all isoforms was represented as percent spliced in (PSI) and computed by calculating the ratio FPKM (inclusion)/FPKM (inclusion, overlap). Because only ratios of reads were compared, a library size normalization step between samples was not needed.

To classify exons as differentially expressed between the proE and orthoE samples, we used the estimated PSI values with the samr R package (25), assuming continuous data. This resulted in a test similar to a t-test, but where the null distribution was generated by a permutation test non-parametrically from the data. In the same manner, we estimated the PSI of each exon in the basophilic (basoE) and polychromatophilic (polyE) erythroblast populations. We then considered any exon for which the inclusion plus overlap FPKM was in the lower quartile to be insufficiently expressed and discarded it from the analysis. We also removed any exon for which a contributing isoform had a contradiction in coverage by having long spanning regions of 0 coverage (see Supplementary Methods). From this list filtered for low gene expression, we also filtered out exons with minimal alternative splicing (PSI <0.05 or >0.95) at any differentiation stage, because empirically we found that this PSI filter greatly reduced the incidence of false positives. Data for the PSI-filtered set of splicing events are shown in Supplementary Table S2, and the larger unfiltered PSI data set is provided in Supplementary Table S3. The filtered list of alternatively spliced exons was used as input into GOrilla (Gene Ontology enRIchment anaLysis and visuaLizAtion tool) for Gene Ontology (GO) analysis (26) (Table 1) and for analyzing stage-specificity of splicing switches (Figure 5).

Table 1.

GO terms enriched in alternatively spliced erythroblast genes

Rank GO term Description P-value FDR q-value 
GO:0044260 Cellular macromolecule metabolic process 1.74E-11 1.57E-07 
GO:0006996 Organelle organization 1.73E-10 7.84E-07 
GO:0022402 Cell cycle process 4.01E-09 1.21E-05 
GO:0016043 Cellular component organization 5.48E-08 8.27E-05 
GO:0016568 Chromatin modification 1.41E-07 1.60E-04 
10 GO:0006325 Chromatin organization 1.61E-07 1.46E-04 
12 GO:0007049 Cell cycle 5.63E-07 4.25E-04 
13 GO:0080090 Regulation of primary metabolic process 7.68E-07 5.35E-04 
14 GO:0010468 Regulation of gene expression 8.07E-07 5.22E-04 
15 GO:0044265 Cellular macromolecule catabolic process 9.00E-07 5.43E-04 
18 GO:0051276 Chromosome organization 1.37E-06 6.90E-04 
19 GO:0000278 Mitotic cell cycle 1.72E-06 8.21E-04 
20 GO:0016569 Covalent chromatin modification 2.59E-06 1.17E-03 
21 GO:0016570 Histone modification 2.75E-06 1.19E-03 
22 GO:0090304 Nucleic acid metabolic process 2.97E-06 1.22E-03 
23 GO:0048285 Organelle fission 3.84E-06 1.51E-03 
30 GO:0000279 M phase 7.27E-06 2.19E-03 
33 GO:0010467 Gene expression 1.07E-05 2.94E-03 
34 GO:0000086 G2/M transition of mitotic cell cycle 1.14E-05 3.03E-03 
35 GO:0000226 Microtubule cytoskeleton organization 1.48E-05 3.81E-03 
39 GO:0006511 Ubiquitin-dependent protein catabolic process 1.86E-05 4.32E-03 
45 GO:0016070 RNA metabolic process 2.48E-05 4.98E-03 
50 GO:0016573 Histone acetylation 2.86E-05 5.18E-03 
56 GO:0000280 Nuclear division 5.01E-05 8.09E-03 
57 GO:0007067 Mitosis 5.01E-05 7.95E-03 
58 GO:0044772 Mitotic cell cycle phase transition 5.27E-05 8.22E-03 
60 GO:0000090 Mitotic anaphase 5.67E-05 8.55E-03 
62 GO:0051301 Cell division 8.95E-05 1.31E-02 
67 GO:0070925 Organelle assembly 1.25E-04 1.69E-02 
76 GO:0043484 Regulation of RNA splicing 1.90E-04 2.27E-02 
Rank GO term Description P-value FDR q-value 
GO:0044260 Cellular macromolecule metabolic process 1.74E-11 1.57E-07 
GO:0006996 Organelle organization 1.73E-10 7.84E-07 
GO:0022402 Cell cycle process 4.01E-09 1.21E-05 
GO:0016043 Cellular component organization 5.48E-08 8.27E-05 
GO:0016568 Chromatin modification 1.41E-07 1.60E-04 
10 GO:0006325 Chromatin organization 1.61E-07 1.46E-04 
12 GO:0007049 Cell cycle 5.63E-07 4.25E-04 
13 GO:0080090 Regulation of primary metabolic process 7.68E-07 5.35E-04 
14 GO:0010468 Regulation of gene expression 8.07E-07 5.22E-04 
15 GO:0044265 Cellular macromolecule catabolic process 9.00E-07 5.43E-04 
18 GO:0051276 Chromosome organization 1.37E-06 6.90E-04 
19 GO:0000278 Mitotic cell cycle 1.72E-06 8.21E-04 
20 GO:0016569 Covalent chromatin modification 2.59E-06 1.17E-03 
21 GO:0016570 Histone modification 2.75E-06 1.19E-03 
22 GO:0090304 Nucleic acid metabolic process 2.97E-06 1.22E-03 
23 GO:0048285 Organelle fission 3.84E-06 1.51E-03 
30 GO:0000279 M phase 7.27E-06 2.19E-03 
33 GO:0010467 Gene expression 1.07E-05 2.94E-03 
34 GO:0000086 G2/M transition of mitotic cell cycle 1.14E-05 3.03E-03 
35 GO:0000226 Microtubule cytoskeleton organization 1.48E-05 3.81E-03 
39 GO:0006511 Ubiquitin-dependent protein catabolic process 1.86E-05 4.32E-03 
45 GO:0016070 RNA metabolic process 2.48E-05 4.98E-03 
50 GO:0016573 Histone acetylation 2.86E-05 5.18E-03 
56 GO:0000280 Nuclear division 5.01E-05 8.09E-03 
57 GO:0007067 Mitosis 5.01E-05 7.95E-03 
58 GO:0044772 Mitotic cell cycle phase transition 5.27E-05 8.22E-03 
60 GO:0000090 Mitotic anaphase 5.67E-05 8.55E-03 
62 GO:0051301 Cell division 8.95E-05 1.31E-02 
67 GO:0070925 Organelle assembly 1.25E-04 1.69E-02 
76 GO:0043484 Regulation of RNA splicing 1.90E-04 2.27E-02 

Besides the computationally defined alternative splicing switches, a few additional examples were discovered in genes of interest by manual inspection of RNA-seq reads mapped to the human genome. Exons that did not computationally meet a ΔPSI threshold of 10% but showed strong evidence for stage-specific changes in the ratio of exon junction reads supporting inclusion versus exon exclusion reads were tested by reverse-transcriptase polymerase chain reaction (RT-PCR); examples of splicing switches detected in this manner are MEF2D, PICALM and MTMR3.

Validation of computationally predicted alternative exons

Many alternative splicing events were verified by manual inspection of mapped reads on the human genome browser, to filter out exons not supported by multiple exon-inclusion and exon-skipping reads. In selected cases, RT-PCR analysis was performed to visualize the amplified products corresponding to inclusion and skipping events. Total RNA was reverse-transcribed into cDNA using random primers and the Superscript III First Strand Synthesis System (Invitrogen), and PCR analysis was performed using AccuTaq polymerase (Sigma) using primers located in constitutive exons located upstream and downstream of each candidate alternative exon. Primer sequences are given in Supplementary Table S4. For initial RT-PCR validations, RNA from unsorted day 9 (proE-enriched) and day 16 (orthoE-enriched) cells was used; these results were subsequently confirmed with RNA purified from highly enriched (FACS-sorted) proEs and orthochromatophilic erythroblasts.

RESULTS

Differentiating erythroid cells execute a robust alternative splicing program

Highly purified human erythroblasts were obtained from in vitro cultures by FACS using differentiation stage-restricted cell surface markers as previously described (18). Stage-specific RNA-seq libraries were prepared from five morphologically distinct populations of successively more differentiated erythroblasts, each separated from the next by one cell division: proEs, early basophilic erythroblasts (e-basoE), late basophilic erythroblasts (l-basoE), polychromatophilic erythroblasts (polyE) and finally orthochromatophilic erythroblasts (orthoE), the last stage before enucleation (Figure 1). Three independent preparations of sorted cells were analyzed to provide biological replicates.

Figure 1.

Morphology of differentiation stage-specific erythroblast populations purified by FACS as described previously (18).

Figure 1.

Morphology of differentiation stage-specific erythroblast populations purified by FACS as described previously (18).

Alternative splicing events in differentiating erythroblasts were identified following the strategy shown in Figure 2. A total of 1.5 billion reads representing the transcriptomes of all 15 erythroblast samples were mapped to the Ensembl annotated transcriptome and analyzed using numerous tools to derive exon-level alternative splicing information as described in ‘Methods’ section. The initial set of 243 464 candidate alternative exons was filtered to remove low-expression genes, and exons with minimal alternative splicing, to derive a data set of 3784 alternative exons in 2250 genes (Supplementary Table S2). Validation of computationally predicted alternative splicing events was performed by manual inspection of RNA-seq reads mapped to the human genome and by RT-PCR analysis of the relevant transcript regions. Figure 3 illustrates representative alternative splicing events in genes involved in transcriptional regulation (NFE2L1, MEF2D, SOX6, ATRX, FOXM1, MAX), DNA and histone modifications (DNMT1, MBD1, CARM1, KDM4B) and signal transduction pathways (MAP2K7, MAP3K3, MAP3K7). The majority of these alternative exons did not substantially alter their splicing efficiency between proEs and orthoEs; i.e. similar results were observed by computational and experimental analyses independent of differentiation stage. However, alternative exons in NFE2L1, MEF2D and CARM1 were among those observed to splice differently as the cells matured (discussed below in Figure 4). Of interest, SOX6 and ATRX are important regulators of globin gene expression (27–29). These results demonstrate that differentiating erythroblasts execute an extensive alternative splicing program that can regulate protein structure and function via control of pre-mRNA splicing; moreover, when the affected proteins are transcription factors, important secondary effects on transcript levels of other genes might amplify downstream cellular consequences.

Figure 2.

Flow chart showing the process by which major alternative splicing events were detected in discrete populations of erythroblasts at specific stages of their terminal differentiation.

Figure 2.

Flow chart showing the process by which major alternative splicing events were detected in discrete populations of erythroblasts at specific stages of their terminal differentiation.

Figure 3.

RT-PCR validation of alternative splicing events in human erythroblasts. Shown are amplification products containing (filled circles) or lacking (empty circles) alternative exons amplified from erythroblast transcripts. Gene names are indicated above each lane.

Figure 3.

RT-PCR validation of alternative splicing events in human erythroblasts. Shown are amplification products containing (filled circles) or lacking (empty circles) alternative exons amplified from erythroblast transcripts. Gene names are indicated above each lane.

Figure 4.

Differentiation stage-specific alternative splicing switches in human erythroblasts. (A) Alternative exons that exhibit stage-specific upregulation in splicing efficiency. Gels show splicing changes assessed by RT-PCR using primers located in flanking constitutive exons. Amplification products from proE RNA and orthoE RNA are shown in the left and right lanes, respectively. Gene name is indicated above each gel, while the calculated PSI is shown below each lane. Arrowheads indicate exon inclusion products. Asterisk indicates genes for which PCR results indicate larger splicing changes than were predicted bioinformatically. (B) Alternative splicing event that exhibits a differentiation-associated decrease in PSI as cell matures. (C) A differentiation-associated switch involving mutually exclusive exons of 100 and 91 nt, respectively. The size difference allows electrophoretic separation of the alternative products.

Figure 4.

Differentiation stage-specific alternative splicing switches in human erythroblasts. (A) Alternative exons that exhibit stage-specific upregulation in splicing efficiency. Gels show splicing changes assessed by RT-PCR using primers located in flanking constitutive exons. Amplification products from proE RNA and orthoE RNA are shown in the left and right lanes, respectively. Gene name is indicated above each gel, while the calculated PSI is shown below each lane. Arrowheads indicate exon inclusion products. Asterisk indicates genes for which PCR results indicate larger splicing changes than were predicted bioinformatically. (B) Alternative splicing event that exhibits a differentiation-associated decrease in PSI as cell matures. (C) A differentiation-associated switch involving mutually exclusive exons of 100 and 91 nt, respectively. The size difference allows electrophoretic separation of the alternative products.

To investigate more broadly which biological processes are impacted by the erythroid splicing program, we performed GO analysis. GO terms significantly enriched among genes expressing major alternative splicing events, relative to all genes expressed (above the lowest quartile) in erythroblasts, are shown in Table 1. The top terms included numerous descriptors associated with chromatin structure and function, cell cycle regulation, organelle organization and regulation of RNA splicing. Among the read-validated alternative splicing events, we found at least 24 differentially expressed exons in 22 genes encoding enzymes that participate in histone modifications (30) (Supplementary Table S5), and at least 65 high-confidence alternative splicing events were detected in 48 genes encoding RNA binding proteins (Supplementary Table S6). As indicated in the tables, many of these alternative splicing events are not annotated among Refseq transcripts, and ∼40% are predicted to introduce PTCs that induce NMD. All of these splicing events passed low-expression and low-alternative splicing filters, and all were validated by RNA-seq read mapping. These observations suggest a high level of complexity in regulation of erythroid gene expression, and a strong potential that these splicing events have downstream effects on the abundance and structure of transcripts of numerous other genes. Thus, important biological pathways express a rich transcript diversity during late erythropoiesis.

Stage-specific switches in alternative splicing remodel the transcriptome during late erythroid differentiation

Alternative splicing switches can play key roles during differentiation and development by selective up- or downregulation of individual exons that alter function of the encoded protein(s). To determine the extent of splicing transitions during erythroid differentiation, we first compared PSI values for alternative exons in early (proE) and late (orthoE) erythroblast populations. In all, 1166 exons for which the difference in PSI values (ΔPSI) exceeded 10% were identified by this approach; 266 of these exhibited ΔPSI values exceeding 25%. A subset of splicing switches that alter protein coding domains was validated by RT-PCR using primers in the flanking constitutive exons (Figure 4). The previously described switch in protein 4.1R (EPB41) pre-mRNA splicing served as a positive control; as expected, splicing of exon 16 exhibited a substantial increase in PSI in orthoE cells compared with a much lower splicing efficiency in proE (Figure 4A). Similar to 4.1R, many of the newly discovered switches involve upregulation of splicing in orthoE (Figure 4A), but examples of downregulation (Figure 4B) and switching between mutually exclusive exons (Figure 4C) were also observed. For some of these, e.g. NDEL1, USO1, MEF2D and MTMR3, the ‘new’ isoforms with potentially new functions can be generated in orthoE that were absent or poorly expressed in proE. Except for USO1, PICALM1 and LGALS9, most of the genes with upregulated coding exons were also upregulated at the gene expression level, relative to global gene expression levels, further supporting the idea that the new isoforms might play important functions in late erythroblasts.

Consistent with a role for alternative splicing in important erythroid pathways, several validated switches in alternative splicing map to genes that function in cytoskeletal organization, cell division and chromatin function (Table 2). One of the biggest switches occurred in the NDEL1 (nuclear distribution factor E-homolog-like1) gene that is important for nuclear migration and nucleokinesis (31), while a less dramatic switch was predicted in survivin (BIRC5), a cytokinesis factor with a novel role in erythroblast enucleation (32,33). In the H2AFY gene, relative expression of mutually exclusive 100- and 91-nt exons reversed almost completely between proE and orthoE (Figure 4A). This switch alters the structure and function of the modified histone, macroH2A1, which can replace the canonical H2A to generate functional differences in chromatin activity (34–36). Consistent with the H2AFY splicing switch in terminal erythroid differentiation, differences in the relative abundance of the two isoforms have been reported in other tissues rich in proliferating versus differentiating cells (37). Other switches of likely functional importance occurred in kinases having roles in chromosome segregation and cytokinesis [NEK1, (38–40)] and microtubule- and actin-based cytoskeletal networks [MARK2, (41)] in a clathrin assembly gene (PICALM) associated with anemia and iron metabolism abnormalities in mouse mutants (42,43); and the mitotic assembly protein NUMA1.

Table 2.

Alternative splicing switches that alter coding domains

gene Exon size ΔPSI (proE to orthoE) Gene function (biological process) 
EPB41 63 +50 Cortical actin cytoskeleton 
EPB49 66 +34 Actin cytoskeleton organization 
NDEL1 35 +46 Centrosome localization 
NEK1 132 +42 Cell division; mitosis 
NUMA1 42 −18 Establishment of mitotic spindle orientation 
BIRC5 69 +11 Cell division; cytokinesis 
MARK2 162 +34 Regulation of cytoskeleton organization; protein serine/threonine kinase activity 
MTMR3 27 −2a Peptidyl-tyrosine dephosphorylation; phosphatidylinositol dephosphorylation 
CARM1 69 +22 Methyltransferase activity 
NFE2L1 90 +39 Sequence-specific DNA binding transcription factor activity 
H2AFY 100b −41 Chromatin modification 
H2AFY 91b +35 Chromatin modification 
MEF2D 24 +7a Regulation of transcription from RNA polymerase II promoter 
PICALM 24 +9a Clathrin coat assembly; hemopoiesis 
BNIP2 36 +27 Apoptotic process 
USO1 21 +42 Intracellular protein transport 
HPS1 99 +16 Lysosome organization 
LGALS9 96 +47 Carbohydrate binding 
gene Exon size ΔPSI (proE to orthoE) Gene function (biological process) 
EPB41 63 +50 Cortical actin cytoskeleton 
EPB49 66 +34 Actin cytoskeleton organization 
NDEL1 35 +46 Centrosome localization 
NEK1 132 +42 Cell division; mitosis 
NUMA1 42 −18 Establishment of mitotic spindle orientation 
BIRC5 69 +11 Cell division; cytokinesis 
MARK2 162 +34 Regulation of cytoskeleton organization; protein serine/threonine kinase activity 
MTMR3 27 −2a Peptidyl-tyrosine dephosphorylation; phosphatidylinositol dephosphorylation 
CARM1 69 +22 Methyltransferase activity 
NFE2L1 90 +39 Sequence-specific DNA binding transcription factor activity 
H2AFY 100b −41 Chromatin modification 
H2AFY 91b +35 Chromatin modification 
MEF2D 24 +7a Regulation of transcription from RNA polymerase II promoter 
PICALM 24 +9a Clathrin coat assembly; hemopoiesis 
BNIP2 36 +27 Apoptotic process 
USO1 21 +42 Intracellular protein transport 
HPS1 99 +16 Lysosome organization 
LGALS9 96 +47 Carbohydrate binding 

aRead map data and RT-PCR indicate stronger switch than predicted by computational analysis.

bMutually exclusive exons.

To explore the hypothesis that alternative splicing networks can impact erythroblast remodeling, we examined the differentiation stage-specificity of splicing transitions. PSI as a function of erythroblast stage is shown in Figure 5A for several PCR-validated alternative exons that exhibited substantially increased inclusion between proE and orthoE (ΔPSI values ≥34). For these genes, most of the change in alternative splicing occurred in the last two (most mature) erythroblast populations, although lesser transitions in splicing efficiency occurred at earlier stages. Extending the analysis to the broader set of alternative exons, we next determined ΔPSI values across each of the four-stage transitions: proE to early basoE, early to late basoE, late basoE to polyE and polyE to orthoE. Figure 5B shows that ∼155 stage-specific splicing switches with ΔPSI > 25 were identified in these cultures, and the great majority of these occurred during the last two cell divisions. Of note, this is a conservative estimate that counts only exons having PSI values between 5 and 95 in all 15 samples assayed. These results demonstrate that most splicing transitions occur in late erythroblasts, temporally correlated with extensive cellular remodeling as the cells prepare to enucleate, and supporting the hypothesis that splicing changes play a determinative role in promoting late erythroid differentiation.

Figure 5.

Stage-specificity of splicing switches during erythroblast differentiation. (A) Splicing efficiency of selected splicing switches across erythroblast populations. (B) Broader analysis of alternative exons exhibiting a splicing switch during erythroblast maturation. Number of exons with |ΔPSI| > 25 is represented by bar height at the indicated stage transitions.

Figure 5.

Stage-specificity of splicing switches during erythroblast differentiation. (A) Splicing efficiency of selected splicing switches across erythroblast populations. (B) Broader analysis of alternative exons exhibiting a splicing switch during erythroblast maturation. Number of exons with |ΔPSI| > 25 is represented by bar height at the indicated stage transitions.

Alternative splicing coupled to NMD regulates expression of selected RNA processing factors

Alternative splicing coupled to NMD (AS-NMD) can regulate gene expression by splicing-mediated introduction of PTCs that induce nonsense-mediated degradation of the affected transcripts. PTC-containing transcripts can be generated by inclusion of exons that encode stop codons or alter translational reading frame (PTC-upon-exon inclusion), or by exon skipping events that alter reading frame (PTC-upon-exon skipping). Having noted that a number of RNA binding proteins express PTC transcripts in erythroblasts (Supplementary Table S4), we wondered whether AS-NMD was functioning to maintain homeostatic control of expression as described in previous studies (44–46). Instead, PSI and gene expression data indicated a different situation characterized by substantial stage-specific changes in relative expression of PTC transcripts (Figure 6). Increased proportions of PTC-containing transcript isoforms were expressed in late erythroblasts by either increases in exon inclusion or by increases in exon skipping (Figure 6A). Many of the predicted increases are substantial; e.g. the SNRNP70 gene switches from PSI = 13 to PSI = 43 for the smaller PTC-upon-inclusion exon. Similarly, well-known AS-NMD targets in the SR protein family including SRSF2, SRSF3, SRSF6 and SRSF7 also showed large ΔPSI switches from proE to orthoE (ΔPSI values of +12 to +47). Other classes of AS-NMD candidates include hnRNP proteins (D, L, LL and M), small nuclear ribonucleic particle proteins (SNRPA1, SNRNP70 and U2AF1), alternative splicing regulators [TRA2A (45), TRA2B] and a 3' processing factor (CPSF4). In a few cases, an increased proportion of PTC transcripts in late erythroblasts can occur in the context of PTC-upon-exon skipping events (Figure 6A, lower panel). SNRPA1, for example, expressed mainly full-length transcripts in proE as indicated by PSI = 94, but the decrease to PSI = 53 in orthoE predicts substantial expression of PTC isoforms in orthoE. However, it is important to note that some known [SRSF11, SNRPB, (45)] or candidate (TIAL1 and HNRNPD) PTC splicing events showed no evidence of stage-specific regulation (Figure 6B, upper panel), and a few (RHD, BCL2L12, CLK1 and CLK4) were regulated so as to decrease the proportion of PTC isoforms in orthoE (Figure 6B, lower panel). RT-PCR validation of splicing switches involving PTC exons enriched in orthoE or proE is provided in Figure 7. We propose that splicing-mediated generation of non-productive (PTC-containing) transcripts is an important post-transcriptional regulatory mechanism for controlling a selected subset of RNA processing factors in late-stage erythroblasts.

Figure 6.

Differential expression of full-length and PTC transcript isoforms in differentiating erythroblasts. (A) RNA binding protein genes that increase expression of PTC isoforms in orthoE by exon inclusion (above) or by exon skipping (below). Exons in red contain stop codons (indicated by asterisks). PSI values for the PTC exons in proE and orthoE are shown at the right. (B) Genes in which PTC isoforms are more abundant in proE (above) or are not regulated in late erythroblasts (below).

Figure 6.

Differential expression of full-length and PTC transcript isoforms in differentiating erythroblasts. (A) RNA binding protein genes that increase expression of PTC isoforms in orthoE by exon inclusion (above) or by exon skipping (below). Exons in red contain stop codons (indicated by asterisks). PSI values for the PTC exons in proE and orthoE are shown at the right. (B) Genes in which PTC isoforms are more abundant in proE (above) or are not regulated in late erythroblasts (below).

Figure 7.

Validation of stage-specific PTC splicing events in human erythroblasts. (A) Transcripts in which PTC isoforms are more abundant in late erythroblasts due to upregulation of PTC-exons; (B) transcripts in which PTC transcripts are relatively more abundant in early erythroblasts due to exon skipping events that alter translational reading frame. Gene names and differentiation stage are indicated above each lane, while calculated PSI values are shown below each lane. Arrowheads indicate PCR bands representing PTC isoforms. * indicates genes for which PCR results indicate larger splicing changes than were predicted bioinformatically.

Figure 7.

Validation of stage-specific PTC splicing events in human erythroblasts. (A) Transcripts in which PTC isoforms are more abundant in late erythroblasts due to upregulation of PTC-exons; (B) transcripts in which PTC transcripts are relatively more abundant in early erythroblasts due to exon skipping events that alter translational reading frame. Gene names and differentiation stage are indicated above each lane, while calculated PSI values are shown below each lane. Arrowheads indicate PCR bands representing PTC isoforms. * indicates genes for which PCR results indicate larger splicing changes than were predicted bioinformatically.

In principle, elevated expression of PTC transcripts could be due to increased production (splicing regulation) or to decreased degradation (NMD regulation). We reasoned that simple loss of NMD activity could not explain maturation-associated increases in PTC exons, as only selected PTC exons were upregulated in orthoE and a few were actually downregulated in these cells. This hypothesis was investigated experimentally by assaying whether inhibition of NMD could affect steady-state levels of PTC exons. If NMD is active in normal orthoE, then steady-state PSI values for PTC exons should increase due to selective stabilization of these transcripts in cycloheximide-treated erythroblasts inhibited for NMD. Figure 8 shows that PTC-upon-inclusion exons in SNRNP70 and HNRPLL had substantially elevated PSI values in NMD-inhibited cells, with PSI of a 72-nt exon in SNRNP70 increasing moderately in proE-enriched day 9 erythroblasts and dramatically in orthoE-enriched day 16 erythroblasts (upper panel). NMD inhibition also increased the proportion of PTC transcripts from the BCL1L12 and CLK4 genes that were generated by exon skipping and were more enriched in proEs (Figure 8, lower panel). We conclude that the NMD machinery is still active in late erythroblasts and therefore that the accumulation of PTC isoforms in these normal cells is not primarily due to loss of NMD, but instead must be due, at least in part, to regulation at the level of alternative splicing. Moreover, as the ‘missing’ quantity of PTC isoforms in normal cells must have been degraded, we propose that alternative splicing coupled to NMD (AS-NMD) is a mechanism for downregulating gene expression in late erythroblasts, analogous to the intron retention mechanism used by late-stage granulocytes to downregulate selected genes (47). Consistent with this, 9 of 11 RBP genes with upregulated PTC events (82%) were downregulated at the gene expression level relative to total mRNAs in late erythroblasts. In contrast, among genes shown in Figure 4 with validated upregulation of non-PTC exons in orthoE, 12 of 15 genes (80%) exhibited increases in gene expression level.

Figure 8.

Effects of NMD inhibition on expression of PTC splicing events. (A) PTC transcripts produced by exon inclusion events were assayed in proE-enriched day 9 and polyE/orthoE-enriched day 16 erythroblast cultures without (−) or with (+) cycloheximide treatment. Cycloheximide dramatically increased the expression of PTC isoforms, especially at day 16, as expected if their abundance in untreated cells was limited by active NMD machinery. For SNRNP70, PCR detects the smaller of the two PTC exons depicted in Figure 7A, B. Cycloheximide effects on PTC-upon-exon skipping events. As in (A), cycloheximide increased expression of PTC isoforms, as expected, if their abundance in untreated cells was limited by NMD. The higher proportion of full-length transcripts predicted in late erythroblasts was maintained ±NMD inhibition. Gene names and culture stage are indicated above each lane, while PSI values shown below each lane were measured by densitometric analysis of stained gels. Arrowheads indicate PCR bands representing PTC isoforms.

Figure 8.

Effects of NMD inhibition on expression of PTC splicing events. (A) PTC transcripts produced by exon inclusion events were assayed in proE-enriched day 9 and polyE/orthoE-enriched day 16 erythroblast cultures without (−) or with (+) cycloheximide treatment. Cycloheximide dramatically increased the expression of PTC isoforms, especially at day 16, as expected if their abundance in untreated cells was limited by active NMD machinery. For SNRNP70, PCR detects the smaller of the two PTC exons depicted in Figure 7A, B. Cycloheximide effects on PTC-upon-exon skipping events. As in (A), cycloheximide increased expression of PTC isoforms, as expected, if their abundance in untreated cells was limited by NMD. The higher proportion of full-length transcripts predicted in late erythroblasts was maintained ±NMD inhibition. Gene names and culture stage are indicated above each lane, while PSI values shown below each lane were measured by densitometric analysis of stained gels. Arrowheads indicate PCR bands representing PTC isoforms.

DISCUSSION

The erythroblast system provides a unique opportunity to study the role of post-transcriptional RNA processing in a differentiation stage-specific manner during the extensive cellular remodeling that characterizes terminal differentiation. We found a robust alternative splicing program enriched in genes that function in cell cycle regulation, organelle organization, chromatin structure and function and RNA processing, all of which undergo tremendous changes before enucleation. Most importantly, dynamic switches in alternative exon usage remodel the late erythroid transcriptome so as to generate new mRNA isoforms or to alter proportions of existing isoforms. These data support the hypothesis that a highly structured alternative splicing program is executed during terminal erythroid differentiation to regulate key biological processes as the cells prepare to enucleate and form mature red cells.

Previously, proper alternative splicing of protein 4.1 (EPB41) pre-mRNA was shown to be essential for generating a red cell membrane with the specialized mechanical properties required for survival in the circulation (48). The finding of abundant alternative splicing events in erythroblasts suggests many additional contributions to erythroid-specific functions. For example, regulators of alpha- and beta-globin gene expression such as ATRX and its interacting partners MECP2 and macroH2A1, as well as BCL11A and SOX6 all express multiple transcript isoforms whose importance has not been sufficiently evaluated. The GO categories enriched in alternative splicing events suggest additional roles in regulation of chromatin condensation, autophagy and enucleation in late erythroblasts, all key and unique events during late stages of terminal erythroid differentiation. We anticipate that some of these alternative splicing switches might be shared by other cell types undergoing the transition from proliferation to terminal differentiation; e.g. differences in relative abundance of H2AFY isoforms in tissues rich in proliferating versus differentiating cells (37) mirror the changes observed in erythroblasts. The erythroid differentiation system may be ideal for studying such splicing events.

Our studies suggest that a second major function for the erythroid alternative splicing program is post-transcriptional downregulation of gene expression in erythroblasts. We hypothesize that NMD-associated splicing switches, similar to switches in coding exons, are driven by changes in expression of splicing regulatory proteins during terminal erythroid differentiation that can up- or downregulate expression of PTC exons, according to physiological requirements for the encoded proteins. The marked changes in steady-state levels of PTC-containing transcripts during differentiation indicate that the mechanism regulating their production must be distinct from homeostatic (often autoregulatory) control mechanisms previously ascribed to AS-NMD [(49) and references therein]. By maintaining active machinery for both alternative splicing and NMD, erythroblasts could coordinate these processes to limit expression in proliferating proE of differentiation functions required for remodeling in orthoE, while also downregulating in orthoE proliferative functions not needed in late erythroblasts. Differentiation-associated switches in PTC exons have already been proposed to operate in differentiating neurons, as, for example, changes in expression of splicing factor PTB can regulate switches in both PTC and non-PTC splicing events (50,51). The full impact of such pathways on erythroid biology may extend well beyond RNA binding proteins, as a substantial fraction of alternative splicing events affecting histone modifying enzymes (∼42%, Supplementary Table S1) and many DNA binding proteins (results not shown) were predicted to induce NMD. Moreover, the current study likely underestimated or entirely missed PTC splicing switches that generate rapidly degraded transcripts, or splicing switches that resulted in retained introns; the latter occurs during late stages of granulocyte differentiation where they can suppress expression of selected genes (47). In fact, inspection of RNA reads mapping to the SF3B1 gene revealed increasing abundance of an intron retention event in late-stage erythroblasts (results not shown) that predicts reduced expression of this critical splicing factor implicated in myelodysplasia (4,52). Future studies will explore these potential expansions of AS-NMD by RNA-seq analysis of NMD-inhibited erythroblasts.

In conclusion, coordinated temporal control of alternative splicing plays a critical role in regulating structure and function of the erythroid transcriptome, likely having a major impact on orderly differentiation of erythroblasts. Analogous cell type-specific or differentiation-associated alternative splicing programs that are critical for cell identity have been identified in non-erythroid cell types including neurons, muscle cells, epithelial cells and T lymphocytes (3,53–55). Conversely, abnormal splicing factor expression as a consequence of human genetic disease or as a result of genetic knockdown in model organisms can profoundly alter splicing networks and have adverse phenotypic effects. Disturbance of programmed splicing networks underlies many complex human genetic diseases including myotonic dystrophy, the autoimmune neurologic disease paraneoplastic opsoclonus-myoclonus ataxia, facioscapulohumeral muscular dystrophy and amyotrophic lateral sclerosis (53,56–59). Moreover, a number of hematological cancers including myelodysplasia can be caused by mutations in splicing factor genes (4,52,60,61), and it has been shown that erythroblasts from myelodysplasia patients exhibit disordered erythropoiesis (62). Understanding the erythroblast splicing program will provide new insights into normal differentiation mechanisms and also into disordered differentiation in disease states.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online.

FUNDING

The National Institutes of Health (NIH) [DK094699 and DK032094]. Director, Office of Science, and Office of Biological & Environmental Research of the US Department of Energy under Contract No. DE-AC02-05CH1123. Funding for open access charge: NIH [DK094699].

Conflict of interest statement. None declared.

ACKNOWLEDGEMENTS

J.G.C., L.P., N.M. and X.A. designed the research; H.P., M.P., S.L.G., D.G. and J.L. performed research and analyzed data; and J.G.C., N.M., X.A., L.P. and H.P. wrote the article.

REFERENCES

1
Licatalosi
DD
Darnell
RB
RNA processing and its regulation: global insights into biological networks
Nat. Rev. Genet.
 , 
2010
, vol. 
11
 (pg. 
75
-
87
)
2
Singh
RK
Cooper
TA
Pre-mRNA splicing in disease and therapeutics
Trends Mol. Med.
 , 
2012
, vol. 
18
 (pg. 
472
-
482
)
3
Martinez
NM
Pan
Q
Cole
BS
Yarosh
CA
Babcock
GA
Heyd
F
Zhu
W
Ajith
S
Blencowe
BJ
Lynch
KW
Alternative splicing networks regulated by signaling in human T cells
RNA
 , 
2012
, vol. 
18
 (pg. 
1029
-
1040
)
4
Yoshida
K
Sanada
M
Shiraishi
Y
Nowak
D
Nagata
Y
Yamamoto
R
Sato
Y
Sato-Otsubo
A
Kon
A
Nagasaki
M
, et al.  . 
Frequent pathway mutations of splicing machinery in myelodysplasia
Nature
 , 
2011
, vol. 
478
 (pg. 
64
-
69
)
5
Cazzola
M
Rossi
M
Malcovati
L
Biologic and clinical significance of somatic mutations of SF3B1 in myeloid and lymphoid neoplasms
Blood
 , 
2013
, vol. 
121
 (pg. 
260
-
269
)
6
Wan
Y
Wu
CJ
SF3B1 mutations in chronic lymphocytic leukemia
Blood
 , 
2013
, vol. 
121
 (pg. 
4627
-
4634
)
7
Yamamoto
ML
Clark
TA
Gee
SL
Kang
JA
Schweitzer
AC
Wickrema
A
Conboy
JG
Alternative pre-mRNA splicing switches modulate gene expression in late erythropoiesis
Blood
 , 
2009
, vol. 
113
 (pg. 
3363
-
3370
)
8
Discher
D
Parra
M
Conboy
JG
Mohandas
N
Mechanochemistry of the alternatively spliced spectrin-actin binding domain in membrane skeletal protein 4.1
J. Biol. Chem.
 , 
1993
, vol. 
268
 (pg. 
7186
-
7195
)
9
Horne
WC
Huang
SC
Becker
PS
Tang
TK
Benz
EJ
Jr
Tissue-specific alternative splicing of protein 4.1 inserts an exon necessary for formation of the ternary complex with erythrocyte spectrin and F-actin
Blood
 , 
1993
, vol. 
82
 (pg. 
2558
-
2563
)
10
Chasis
JA
Coulombel
L
Conboy
J
McGee
S
Andrews
K
Kan
YW
Mohandas
N
Differentiation-associated switches in protein 4.1 expression. Synthesis of multiple structural isoforms during normal human erythropoiesis
J. Clin. Invest.
 , 
1993
, vol. 
91
 (pg. 
329
-
338
)
11
Tang
TK
Leto
TL
Correas
I
Alonso
MA
Marchesi
VT
Benz
EJ
Jr
Selective expression of an erythroid-specific isoform of protein 4.1
Proc. Natl Acad. Sci. USA
 , 
1988
, vol. 
85
 (pg. 
3713
-
3717
)
12
Hou
VC
Lersch
R
Gee
SL
Ponthier
JL
Lo
AJ
Wu
M
Turck
CW
Koury
M
Krainer
AR
Mayeda
A
, et al.  . 
Decrease in hnRNP A/B expression during erythropoiesis mediates a pre-mRNA splicing switch
EMBO J.
 , 
2002
, vol. 
21
 (pg. 
6195
-
6204
)
13
Ponthier
JL
Schluepen
C
Chen
W
Lersch
RA
Gee
SL
Hou
VC
Lo
AJ
Short
SA
Chasis
JA
Winkelmann
JC
, et al.  . 
Fox-2 splicing factor binds to a conserved intron motif to promote inclusion of protein 4.1R alternative exon 16
J. Biol. Chem.
 , 
2006
, vol. 
281
 (pg. 
12468
-
12474
)
14
Yang
G
Huang
SC
Wu
JY
Benz
EJ
Jr
An erythroid differentiation-specific splicing switch in protein 4.1R mediated by the interaction of SF2/ASF with an exonic splicing enhancer
Blood
 , 
2005
, vol. 
105
 (pg. 
2146
-
2153
)
15
Yang
G
Huang
SC
Wu
JY
Benz
EJ
Jr
Regulated Fox-2 isoform expression mediates protein 4.1R splicing during erythroid differentiation
Blood
 , 
2008
, vol. 
111
 (pg. 
392
-
401
)
16
Conboy
JG
Shitamoto
R
Parra
M
Winardi
R
Kabra
A
Smith
J
Mohandas
N
Hereditary elliptocytosis due to both qualitative and quantitative defects in membrane skeletal protein 4.1
Blood
 , 
1991
, vol. 
78
 (pg. 
2438
-
2443
)
17
Marchesi
SL
Conboy
J
Agre
P
Letsinger
JT
Marchesi
VT
Speicher
DW
Mohandas
N
Molecular analysis of insertion/deletion mutations in protein 4.1 in elliptocytosis. I. Biochemical identification of rearrangements in the spectrin/actin binding domain and functional characterizations
J. Clin. Invest.
 , 
1990
, vol. 
86
 (pg. 
516
-
523
)
18
Hu
J
Liu
J
Xue
F
Halverson
G
Reid
M
Guo
A
Chen
L
Raza
A
Galili
N
Jaffray
J
, et al.  . 
Isolation and functional characterization of human erythroblasts at distinct stages: implications for understanding of normal and disordered erythropoiesis in vivo
Blood
 , 
2013
, vol. 
121
 (pg. 
3246
-
3253
)
19
Flicek
P
Ahmed
I
Amode
MR
Barrell
D
Beal
K
Brent
S
Carvalho-Silva
D
Clapham
P
Coates
G
Fairley
S
, et al.  . 
Ensembl 2013
Nucleic Acids Res.
 , 
2013
, vol. 
41
 (pg. 
D48
-
D55
)
20
Church
DM
Schneider
VA
Graves
T
Auger
K
Cunningham
F
Bouk
N
Chen
HC
Agarwala
R
McLaren
WM
Ritchie
GR
, et al.  . 
Modernizing reference genome assemblies
PLoS Biol.
 , 
2011
, vol. 
9
 pg. 
e1001091
 
21
Langmead
B
Trapnell
C
Pop
M
Salzberg
SL
Ultrafast and memory–efficient alignment of short DNA sequences to the human genome
Genome Biol.
 , 
2009
, vol. 
10
 pg. 
R25
 
22
Roberts
A
Pachter
L
Streaming fragment assignment for real-time analysis of sequencing experiments
Nat. Methods
 , 
2013
, vol. 
10
 (pg. 
71
-
73
)
23
Mortazavi
A
Williams
BA
McCue
K
Schaeffer
L
Wold
B
Mapping and quantifying mammalian transcriptomes by RNA-Seq
Nat. Methods
 , 
2008
, vol. 
5
 (pg. 
621
-
628
)
24
Trapnell
C
Williams
BA
Pertea
G
Mortazavi
A
Kwan
G
van Baren
MJ
Salzberg
SL
Wold
BJ
Pachter
L
Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation
Nat. Biotechnol.
 , 
2010
, vol. 
28
 (pg. 
511
-
515
)
25
Tusher
VG
Tibshirani
R
Chu
G
Significance analysis of microarrays applied to the ionizing radiation response
Proc. Natl Acad. Sci. USA
 , 
2001
, vol. 
98
 (pg. 
5116
-
5121
)
26
Eden
E
Navon
R
Steinfeld
I
Lipson
D
Yakhini
Z
GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists
BMC Bioinformatics
 , 
2009
, vol. 
10
 pg. 
48
 
27
Xu
J
Sankaran
VG
Ni
M
Menne
TF
Puram
RV
Kim
W
Orkin
SH
Transcriptional silencing of {gamma}-globin by BCL11A involves long-range interactions and cooperation with SOX6
Genes Dev.
 , 
2010
, vol. 
24
 (pg. 
783
-
798
)
28
Yi
Z
Cohen-Barak
O
Hagiwara
N
Kingsley
PD
Fuchs
DA
Erickson
DT
Epner
EM
Palis
J
Brilliant
MH
Sox6 directly silences epsilon globin expression in definitive erythropoiesis
PLoS Genet.
 , 
2006
, vol. 
2
 pg. 
e14
 
29
Clynes
D
Higgs
DR
Gibbons
RJ
The chromatin remodeller ATRX: a repeat offender in human disease
Trends Biochem. Sci.
 , 
2013
, vol. 
38
 (pg. 
461
-
466
)
30
Allis
CD
Berger
SL
Cote
J
Dent
S
Jenuwien
T
Kouzarides
T
Pillus
L
Reinberg
D
Shi
Y
Shiekhattar
R
, et al.  . 
New nomenclature for chromatin-modifying enzymes
Cell
 , 
2007
, vol. 
131
 (pg. 
633
-
636
)
31
McKenney
RJ
Weil
SJ
Scherer
J
Vallee
RB
Mutually exclusive cytoplasmic dynein regulation by NudE-Lis1 and dynactin
J. Biol. Chem.
 , 
2011
, vol. 
286
 (pg. 
39615
-
39622
)
32
Keerthivasan
G
Liu
H
Gump
JM
Dowdy
SF
Wickrema
A
Crispino
JD
A novel role for survivin in erythroblast enucleation
Haematologica
 , 
2012
, vol. 
97
 (pg. 
1471
-
1479
)
33
Miura
K
Fujibuchi
W
Unno
M
Splice variants in apoptotic pathway
Exp. Oncol.
 , 
2012
, vol. 
34
 (pg. 
212
-
217
)
34
Talbert
PB
Henikoff
S
Histone variants—ancient wrap artists of the epigenome
Nat. Rev. Mol. Cell Biol.
 , 
2010
, vol. 
11
 (pg. 
264
-
275
)
35
Gamble
MJ
Frizzell
KM
Yang
C
Krishnakumar
R
Kraus
WL
The histone variant macroH2A1 marks repressed autosomal chromatin, but protects a subset of its target genes from silencing
Genes Dev.
 , 
2010
, vol. 
24
 (pg. 
21
-
32
)
36
Buschbeck
M
Uribesalgo
I
Wibowo
I
Rue
P
Martin
D
Gutierrez
A
Morey
L
Guigo
R
Lopez-Schier
H
Di Croce
L
The histone variant macroH2A is an epigenetic regulator of key developmental genes
Nat. Struct. Mol. Biol.
 , 
2009
, vol. 
16
 (pg. 
1074
-
1079
)
37
Pehrson
JR
Costanzi
C
Dharia
C
Developmental and tissue expression patterns of histone macroH2A1 subtypes
J. Cell Biochem.
 , 
1997
, vol. 
65
 (pg. 
107
-
113
)
38
Chen
Y
Chen
CF
Chiang
HC
Pena
M
Polci
R
Wei
RL
Edwards
RA
Hansel
DE
Chen
PL
Riley
DJ
Mutation of NIMA-related kinase 1 (NEK1) leads to chromosome instability
Mol Cancer
 , 
2011
, vol. 
10
 pg. 
5
 
39
White
MC
Quarmby
LM
The NIMA-family kinase, Nek1 affects the stability of centrosomes and ciliogenesis
BMC Cell Biol.
 , 
2008
, vol. 
9
 pg. 
29
 
40
Patil
M
Pabla
N
Ding
HF
Dong
Z
Nek1 interacts with Ku80 to assist chromatin loading of replication factors and S-phase progression
Cell Cycle
 , 
2013
, vol. 
12
 (pg. 
2608
-
2616
)
41
Yamahashi
Y
Saito
Y
Murata-Kamiya
N
Hatakeyama
M
Polarity-regulating kinase partitioning-defective 1b (PAR1b) phosphorylates guanine nucleotide exchange factor H1 (GEF-H1) to regulate RhoA-dependent actin cytoskeletal reorganization
J. Biol. Chem.
 , 
2011
, vol. 
286
 (pg. 
44576
-
44584
)
42
Klebig
ML
Wall
MD
Potter
MD
Rowe
EL
Carpenter
DA
Rinchik
EM
Mutations in the clathrin-assembly gene Picalm are responsible for the hematopoietic and iron metabolism abnormalities in fit1 mice
Proc. Natl Acad. Sci. USA
 , 
2003
, vol. 
100
 (pg. 
8360
-
8365
)
43
Scotland
PB
Heath
JL
Conway
AE
Porter
NB
Armstrong
MB
Walker
JA
Klebig
ML
Lavau
CP
Wechsler
DS
The PICALM protein plays a key role in iron homeostasis and cell proliferation
PLoS One
 , 
2012
, vol. 
7
 pg. 
e44252
 
44
Lareau
LF
Inada
M
Green
RE
Wengrod
JC
Brenner
SE
Unproductive splicing of SR genes associated with highly conserved and ultraconserved DNA elements
Nature
 , 
2007
, vol. 
446
 (pg. 
926
-
929
)
45
Saltzman
AL
Kim
YK
Pan
Q
Fagnani
MM
Maquat
LE
Blencowe
BJ
Regulation of multiple core spliceosomal proteins by alternative splicing-coupled nonsense-mediated mRNA decay
Mol. Cell Biol.
 , 
2008
, vol. 
28
 (pg. 
4320
-
4330
)
46
Ni
JZ
Grate
L
Donohue
JP
Preston
C
Nobida
N
O'Brien
G
Shiue
L
Clark
TA
Blume
JE
Ares
M
Jr
Ultraconserved elements are associated with homeostatic control of splicing regulators by alternative splicing and nonsense-mediated decay
Genes Dev.
 , 
2007
, vol. 
21
 (pg. 
708
-
718
)
47
Wong
JJ
Ritchie
W
Ebner
OA
Selbach
M
Wong
JW
Huang
Y
Gao
D
Pinello
N
Gonzalez
M
Baidya
K
, et al.  . 
Orchestrated intron retention regulates normal granulocyte differentiation
Cell
 , 
2013
, vol. 
154
 (pg. 
583
-
595
)
48
Discher
DE
Winardi
R
Schischmanoff
PO
Parra
M
Conboy
JG
Mohandas
N
Mechanochemistry of protein 4.1's spectrin-actin-binding domain: ternary complex interactions, membrane binding, network integration, structural strengthening
J. Cell Biol.
 , 
1995
, vol. 
130
 (pg. 
897
-
907
)
49
Kalsotra
A
Cooper
TA
Functional consequences of developmentally regulated alternative splicing
Nat. Rev. Genet.
 , 
2011
, vol. 
12
 (pg. 
715
-
729
)
50
Licatalosi
DD
Yano
M
Fak
JJ
Mele
A
Grabinski
SE
Zhang
C
Darnell
RB
Ptbp2 represses adult-specific splicing to regulate the generation of neuronal precursors in the embryonic brain
Genes Dev.
 , 
2012
, vol. 
26
 (pg. 
1626
-
1642
)
51
Zheng
S
Gray
EE
Chawla
G
Porse
BT
O'Dell
TJ
Black
DL
PSD-95 is post-transcriptionally repressed during early neural development by PTBP1 and PTBP2
Nat. Neurosci.
 , 
2012
, vol. 
15
 (pg. 
381
-
388
)
52
Papaemmanuil
E
Cazzola
M
Boultwood
J
Malcovati
L
Vyas
P
Bowen
D
Pellagatti
A
Wainscoat
JS
Hellstrom-Lindberg
E
Gambacorti-Passerini
C
, et al.  . 
Somatic SF3B1 mutation in myelodysplasia with ring sideroblasts
N. Engl. J. Med.
 , 
2011
, vol. 
365
 (pg. 
1384
-
1395
)
53
Ule
J
Jensen
KB
Ruggiu
M
Mele
A
Ule
A
Darnell
RB
CLIP identifies Nova-regulated RNA networks in the brain
Science
 , 
2003
, vol. 
302
 (pg. 
1212
-
1215
)
54
Bland
CS
Wang
ET
Vu
A
David
MP
Castle
JC
Johnson
JM
Burge
CB
Cooper
TA
Global regulation of alternative splicing during myogenic differentiation
Nucleic Acids Res.
 , 
2010
, vol. 
38
 (pg. 
7651
-
7664
)
55
Warzecha
CC
Jiang
P
Amirikian
K
Dittmar
KA
Lu
H
Shen
S
Guo
W
Xing
Y
Carstens
RP
An ESRP-regulated splicing programme is abrogated during the epithelial-mesenchymal transition
EMBO J.
 , 
2010
, vol. 
29
 (pg. 
3286
-
3300
)
56
Kanadia
RN
Johnstone
KA
Mankodi
A
Lungu
C
Thornton
CA
Esson
D
Timmers
AM
Hauswirth
WW
Swanson
MS
A muscleblind knockout model for myotonic dystrophy
Science
 , 
2003
, vol. 
302
 (pg. 
1978
-
1980
)
57
Orengo
JP
Chambon
P
Metzger
D
Mosier
DR
Snipes
GJ
Cooper
TA
Expanded CTG repeats within the DMPK 3' UTR causes severe skeletal muscle wasting in an inducible mouse model for myotonic dystrophy
Proc. Natl Acad. Sci. USA
 , 
2008
, vol. 
105
 (pg. 
2646
-
2651
)
58
Pistoni
M
Shiue
L
Cline
MS
Bortolanza
S
Neguembor
MV
Xynos
A
Ares
M
Jr
Gabellini
D
Rbfox1 Downregulation and Altered Calpain 3 splicing by FRG1 in a mouse model of Facioscapulohumeral Muscular Dystrophy (FSHD)
PLoS Genet.
 , 
2013
, vol. 
9
 pg. 
e1003186
 
59
Polymenidou
M
Lagier-Tourenne
C
Hutt
KR
Bennett
CF
Cleveland
DW
Yeo
GW
Misregulated RNA processing in amyotrophic lateral sclerosis
Brain Res.
 , 
2012
, vol. 
1462
 (pg. 
3
-
15
)
60
Je
EM
Yoo
NJ
Kim
YJ
Kim
MS
Lee
SH
Mutational analysis of splicing machinery genes SF3B1, U2AF1 and SRSF2 in myelodysplasia and other common tumors
Int. J. Cancer
 , 
2013
, vol. 
133
 (pg. 
260
-
265
)
61
Zhang
SJ
Rampal
R
Manshouri
T
Patel
J
Mensah
N
Kayserian
A
Hricik
T
Heguy
A
Hedvat
C
Gonen
M
, et al.  . 
Genetic analysis of patients with leukemic transformation of myeloproliferative neoplasms shows recurrent SRSF2 mutations that are associated with adverse outcome
Blood
 , 
2012
, vol. 
119
 (pg. 
4480
-
4485
)
62
Liu
J
Zhang
J
Ginzburg
Y
Li
H
Xue
F
De Franceschi
L
Chasis
JA
Mohandas
N
An
X
Quantitative analysis of murine terminal erythroid differentiation in vivo: novel method to study normal and disordered erythropoiesis
Blood
 , 
2013
, vol. 
121
 (pg. 
e43
-
49
)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

Comments

0 Comments