Defining an EPOR- regulated transcriptome for primary progenitors, including Tnfr-sf13c as a novel mediator of EPO- dependent erythroblast formation

Abstract

Certain concepts concerning EPO/EPOR action modes have been challenged by in vivo studies: Bcl-x levels are elevated in maturing erythroblasts, but not in their progenitors; truncated EPOR alleles that lack a major p85/PI3K recruitment site nonetheless promote polycythemia; and Erk1 disruption unexpectedly bolsters erythropoiesis. To discover novel EPO/EPOR action routes, global transcriptome analyses presently are applied to interrogate EPO/EPOR effects on primary bone marrow-derived CFUe-like progenitors. Overall, 160 EPO/EPOR target transcripts were significantly modulated 2-to 21.8-fold. A unique set of EPO-regulated survival factors included Lyl1, Gas5, Pim3, Pim1, Bim, Trib3 and Serpina 3g. EPO/EPOR-modulated cell cycle mediators included Cdc25a, Btg3, Cyclin-d2, p27-kip1, Cyclin-g2 and CyclinB1-IP-1. EPO regulation of signal transduction factors was also interestingly complex. For example, not only Socs3 plus Socs2 but also Spred2, Spred1 and Eaf1 were EPO-induced as negative-feedback components. Socs2, plus five additional targets, further proved to comprise new EPOR/Jak2/Stat5 response genes (which are important for erythropoiesis during anemia). Among receptors, an atypical TNF-receptor Tnfr-sf13c was up-modulated >5-fold by EPO. Functionally, Tnfr-sf13c ligation proved to both promote proerythroblast survival, and substantially enhance erythroblast formation. The EPOR therefore engages a sophisticated set of transcriptome response circuits, with Tnfr-sf13c deployed as one novel positive regulator of proerythroblast formation.

Figure 1. Defining the EPO/EPOR- regulated transcriptome in primary bone marrow CFUe-like erythroid progenitor cells.

A] Short term expansion of primary bone marrow erythroid progenitors, and isolation of a CFUe- like cohort – Approaches employed for the expansion and purification of Kit^posCD71^highTer119^neg stage E1 progenitor cells are outlined. B, C] For isolated stage E1 cells, flow cytometric features are illustrated, together with representative May-Grumswald cytospin preparations. D] For CFUe-like progenitors (stage E1) purity also was assessed based on initial overall transcriptome profiling (n = 3 per stage), with relative mean expression intensities illustrated for select erythroid (Tfrc, Klf1), lymphoid (CD19, B220, IL2r-a, Tcr-a) and myeloid (Mac1, Csfr1) marker transcripts. E] Distributions and variance of overall hybridization signal intensities among control and EPO- challenged treatment groups. F] Heat-map signatures of expressed transcripts (for all independent quadruplicate samples) for [−] EPO vs [+] EPO exposed CFUe-like stage E1 progenitors. Here, a fold cut-off of >1 was used, and signal strengths for 365 transcripts (probe sets) are illustrated. G] In this volcano plot, the dividing line (and vertical color-code) denote significant p-values (0.05 cut-off) for EPO- modulated transcripts (−log₁₀ ordinate scale). The x-axis displays transcripts based on their fold-change due to EPO (log₂ scale). H] Overall scatter plot of significantly modulated EPO/EPOR- response genes.

Figure 2. K-Means clustering of EPO/EPOR- modulated gene sets, and identification of novel EPOR/JAK2/STAT5 targets.

A] The EPOR is diagrammed, together with subdomains that engage canonical Ras/Raf/Mek/Erk, Stat5, and PI3K/Akt response pathways. B] K-means clustering of four predominantly patterned EPO/EPOR– response gene subsets C] Proposed nature of “cluster-4” EPO- response genes as EPOR-PY343/Stat5 targets. D] Quantitative RT-PCR analyses of novel “cluster-4” EPO/EPOR response genes for isolated CFUe- like progenitors expressing wild-type, PY site-deficient, or PY343 Stat5- coupled EPOR alleles (wt, EPOR-HM, EPOR-H). In RT-PCR, normalization was to beta-actin. Values are means of duplicate assays.

Figure 3. Subsets of functional targets within the EPO/EPOR response transcriptome, including survival and cell cycle factors.

A] Frequencies of EPO/EPOR- modulated targets are defined within eleven functional sub-categories. B, D] Among survival factors, EPOR ligation significantly modulated seven prime factors (6 induced, plus Bim repressed). C, E] Cell cycle factors as modulated by EPO/EPOR ligation included three involved in phase- G1 progression; and three that regulate phase G2.

Figure 4. EPO/EPOR – modulation of cytokines plus receptors, and negative feedback factors.

A] Cytokines/ receptors modulated due to EPOR ligation included Tnfr-sf13c (BAFF receptor), Cmtm6 (transmembrane chemokine receptor like superfamily-6), Lrp8 (cholesterol co-receptor), Gdf3 (a BMP antagonist), and Oncostatin-M (a new indirect mediator of iron transport). B] Among negative-feedback factors, EPOR ligation up modulated five prime factors as Spred2, Spred1, Edf1, Socs2, and Socs3. C] For Tnfr-sf13c, associated factors and pathways are outlined (based on Ingenuity algorithms). D] For Spred-1 and -2, associated factors and pathways are outlined.

Figure 5. Tnfr-sf13c expression during EPO-dependent (pro)erythroblast development, and erythropoietic effects of BAFF.

A] EPO efficiently induces Tnfr-sf13c expression in CFUe-like progenitors, proerythroblasts, and Ter119^pos erythroblasts (stages E-1, -2, and -3, respectively)- Stage E1, E2, and E3 EPC's were isolated; cultured for 5.5 hours in the absence of EPO; and then EPO-challenged frequencies of Ter119^pos stage E3 erythroblasts were determined (by flow cytometry) (means +/− SE, n = 4). for 90 minutes (4U/mL). RNA was isolated, reverse-transcribed and used in quantitative RT-PCR analyses. B] Tnfr-sf13c expression peaks within stage-E2 erythroblasts. Tnfr-sf13c levels also were determined within stage E1, E2, and E3 EPC's directly upon isolation. C] EPO/EPOR- induction of Tnfr-sf13c depends, in part, upon Stat5 engagement. Stage E1 bone marrow EPC's were expanded and isolated from wild-type mice, and mice harboring EPOR-HM or EPOR-H alleles. For each, EPO induction of Tnfr-sf13c expression was then assayed (by RT-PCR) (mean +/− SE, n = 3 per group). D] BAFF- dependent inhibition of apoptosis among primary bone marrow proerythroblasts. EPCs were expanded from wild-type bone marrow. Stage E1 plus E2 cells were then isolated, cultured for 5.5 hours without EPO, exposed to EPO for 2 hours, and then washed thrice. Cells then were cultured for 15 hours in the presence (or absence) of BAFF ligand (1.2 ug/mL) or EPO (0.06 U/mL). Levels of annexin v- positive cells then were determined (via flow cytometry) (means +/− SE, n = 4). E] BAFF ligation of Tnfr-sf13c substantially promotes the formation of maturing Ter119^pos erythroblasts. Ter119^neg EPC's were expanded and isolated. Cells were then cultured in the presence, or absence of BAFF ligand (1.2 ug/mL) or EPO (0.06 U/mL) and at 15 hours, frequencies of Ter119^pos erythroblasts were determined (by flow cytometry). Graphed values are means +/− SE.

Figure 6. Model for EPO\EPOR deployment of Tnfr-sf13c as an agent for proerythroblast survival, and erythroblast formation.

In stage E1 EPC's and stage E2 pro-erythroblasts, elevated EPO levels heighten Tnfr-sf13c expression. Tnfr-sf13c ligation via stromal cell Baff then results in heightened survival of pro-erythroblasts, and increased formation of Ter119^pos stage E3 erythroblasts.