International Journal of Molecular Sciences

Article RAP-011 Rescues the Disease Phenotype in a Cellular Model of Congenital Dyserythropoietic Anemia Type II by Inhibiting the SMAD2-3 Pathway

Gianluca De Rosa 1,2 , Immacolata Andolfo 1,2,*, Roberta Marra 1,2, Francesco Manna 2, Barbara Eleni Rosato 1,2, Achille Iolascon 1,2 and Roberta Russo 1,2,* 1 Department of Molecular Medicine and Medical Biotechnologies, University of Naples Federico II, 80131 Naples, Italy; [email protected] (G.D.R.); [email protected] (R.M.); [email protected] (B.E.R.); [email protected] (A.I.) 2 Ceinge Biotecnologie Avanzate, 80145 Naples, Italy; [email protected] * Correspondence: [email protected] (I.A.); [email protected] (R.R.); Tel.: +39-081-3737736 (I.A.); +39-081-3737736 (R.R.)  Received: 9 July 2020; Accepted: 30 July 2020; Published: 4 August 2020 

Abstract: Congenital dyserythropoietic anemia type II (CDA II) is a hypo-productive anemia defined by ineffective erythropoiesis through maturation arrest of erythroid precursors. CDA II is an autosomal recessive disorder due to loss-of-function mutations in SEC23B. Currently, management of patients with CDA II is based on transfusions, splenectomy, or hematopoietic stem-cell transplantation. Several studies have highlighted benefits of ACE-011 (sotatercept) treatment of ineffective erythropoiesis, which acts as a ligand trap against growth differentiation factor (GDF)11. Herein, we show that GDF11 levels are increased in CDA II, which suggests sotatercept as a targeted therapy for treatment of these patients. Treatment of stable clones of SEC23B-silenced erythroleukemia K562 cells with the iron-containing porphyrin hemin plus GDF11 increased expression of pSMAD2 and reduced nuclear localization of the transcription factor GATA1, with subsequent reduced expression of erythroid differentiation markers. We demonstrate that treatment of these SEC23B-silenced K562 cells with RAP-011, a “murinized” ortholog of sotatercept, rescues the disease phenotype by restoring gene expression of erythroid markers through inhibition of the phosphorylated SMAD2 pathway. Our data also demonstrate the effect of RAP-011 treatment in reducing the expression of erythroferrone in vitro, thus suggesting a possible beneficial role of the use of sotatercept in the management of iron overload in patients with CDA II.

Keywords: congenital dyserythropoietic anemia type II; II ligand trap; in vitro drug treatment

1. Introduction Congenital dyserythropoietic anemias (CDAs) are a group of rare hereditary disorders that are defined by maturation arrest of the erythroid lineage, the main consequence of which is ineffective erythropoiesis, with the consequent reduced production of erythrocytes [1]. CDA type II (CDA II) is the most common form of CDAs, and it is caused by biallelic loss-of-function mutations in the SEC23B gene. SEC23B encodes a of the COPII complex that is involved in intracellular vesicle trafficking from the endoplasmic reticulum to the Golgi compartment [2,3]. CDA II is characterized by normocytic anemia of variable degree and relative reticulocytopenia, which are often accompanied by jaundice and splenomegaly, due to the hemolytic component. Bone-marrow examination of patients with CDA II has highlighted erythroid hyperplasia and the presence of bi-nucleated and multi-nucleated erythroblasts, due to maturation arrest [1,4]. The most harmful complication for patients with CDA II

Int. J. Mol. Sci. 2020, 21, 5577; doi:10.3390/ijms21155577 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2020, 21, 5577 2 of 11

De Rosa G et al. 2 of 11 is iron overload [5]. Indeed, ineffective erythropoiesis results in strong downregulation of the hepatic ironhormone absorption hepcidin, and and systemic this condition iron canoverload lead to, increasedwhich is iron mediated absorption byand the systemic erythroid iron hormone overload, erythroferronewhich is mediated (ERFE) by the[6]. erythroid hormone erythroferrone (ERFE) [6]. InIn the the last last years, years, members members of of the the tr transformingansforming growth growth factor-beta factor-beta (TGF- (TGF-ββ)) superfamily, superfamily, which which includesincludes activins activins (A (A and and B), B), grow growthth differentiation differentiation factors factors (GDFs) (GDFs) and and bone bone morphogenetic morphogenetic proteins (BMPs),(BMPs), have have been been studied studied as as potential potential regulators regulators of of erythropoiesis. erythropoiesis. In In particular, particular, GDF11 GDF11 (also (also known known asas BMP11) BMP11) has been investigatedinvestigated asas aa possiblepossible negative negative regulator regulator of of erythropoiesis. erythropoiesis. Indeed, Indeed, through through its itsbinding binding to activinto activin receptors receptors (ActR) (ActR) IIA IIA and and IIB, IIB, GDF11 GDF11 can can inhibit inhibit terminal terminal erythroid erythroid maturation maturation [7]. [7].Nevertheless, Nevertheless, arecent a recent study study excluded excluded that thatGDF11 GDF11 isis the only only effector effector of of TGF- TGF- inhibitioninhibition of lat latee erythropoiesiserythropoiesis in in mice, mice, but but it it may may contribute contribute toge togetherther with with other other related related cytokines to to ineffective ineffective erythropoiesiserythropoiesis [8]. [8]. ToTo date, date, treatments treatments for for patients patients with with CDA CDA II IIha haveve only only included included supportive supportive therapies, therapies, such such as transfusion,as transfusion, splenectomy splenectomy and and hematopoietic hematopoietic stem stem-cell-cell transplantation transplantation in in transfusion-dependent patientspatients [1,4,9,10]. [1,4,9,10]. Of Of note, note, two two inhibitors inhibitors of TGF- of TGF- pathway, pathway, ACE-011ACE-011 (sotatercept)(sotatercept) and ACE-536 ACE-536 (luspatercept),(luspatercept), areare ligand ligand traps traps for for ActRIIA ActRIIA and ActRIIB,and ActRIIB, respectively, respectively, and their and use their has beenuse associatedhas been associatedwith improved with improved hematologic hematologic parameters parameters in healthy in healthy post-menopausal post-menopausal women women [11]. Studies[11]. Studies with withthe “murinized” the “murinized” orthologs orthologs of sotatercept of sotatercept and luspatercept, and luspatercept, known as known RAP-011 as and RAP-011 RAP-536, and respectively, RAP-536, respectively,in a murine modelin a murine of β-thalassemia model of β resulted-thalassemia in increased resulted di infferentiation increased ofdifferentiation erythroid cells, of erythroid improved cells,anemia improved and reduced anemia iron and overload reduced in the iron treated overload mice [in11 ,the12]. treated Sotatercept mice and [11,12]. its murinized Sotatercept counterpart and its murinizedRAP-011 are counterpart chimeric proteins RAP-011 where are chimeric the extracellular proteins where domain the of extracellular the ActRIIA receptordomain of has the been ActRIIA fused receptorto the Fc has portion been fused of the to human the Fc (orportion mouse, of the respectively) human (or IgG1mouse, antibody respective [13ly)]. IgG1 RAP-011 antibody treatment [13]. RAP-011has already treatment been evaluated has already in abeenβ-thalassemia evaluated murinein a β-thalassemia model, as wellmurine as inmodel, a zebrafish as well model as in ofa zebrafishDiamond–Blackfan model of Diamond anemia [–7Blackfan,13]. anemia [7,13]. Here,Here, we we investigated investigated the the effects effects of ofRAP-011 RAP-011 treatment treatment in an in in an vitroin vitro CDACDA II model II model of the of erythroleukemiathe erythroleukemia K562 K562 cell line cell linestably stably silenced silenced for SEC23B for SEC23B expression.expression.

2.2. Results Results

2.1.2.1. Ex Ex Vivo Vivo and InIn VitroVitro Quantitative Quantitative Evaluation Evaluation of GDF11 of GDF11 OnOn the the basis basis of of data data that that corre correlatelate increased increased GDF11 GDF11 levels levels to β to-thalassemia,β-thalassemia, we investigated we investigated the expressionthe expression of this of thiscytokine in healthy in healthy controls controls and and patients patients with with CDA CDA II. II. In In total, total, 15 15 healthy healthy controls controls andand 12 12 patients patients with with SEC23BSEC23B-related-related CDA CDA II II provided provided peripheral peripheral blood blood for for evaluation evaluation of of protein protein expressionexpression of GDF11. The ex vivo analysisanalysis atat thethe proteinprotein level level showed showed 2.2-fold 2.2-fold increased increased expression expression of ofsecreted secreted GDF11 GDF11 in in plasma plasma samples samples from from the the patients patients with with CDACDA IIII comparedcompared to the healthy controls controls (Figure(Figure 11a,b).a,b).

Figure 1. Ex vivo analysis of GDF11 expression. (a) Densitometry quantification of GDF11 expression Figurein plasma 1. Ex samples vivo analysis from of healthy GDF11 controls expression. (HCs; (a)n Densitometry= 15) and patients quantification with CDA of GDF11 II (n = expression12) on red ponceau-stained membrane. O.D., optical density. Data are means standard error. p-value by in plasma samples from healthy controls (HCs; n = 15) and patients ±with CDA II (n = 12) on red ponceau-stainedMann–Whitney tests.membrane. (b) Representative O.D., optical Westerndensity. Data blot forare GDF11means ± expression standard error. in plasma p-value samples by Mann from– Whitney3 healthy tests. controls (b) andRepresentative 3 patients with Western CDA II.blot for GDF11 expression in plasma samples from 3 healthy controls and 3 patients with CDA II.

Int. J. Mol. Sci. 2020, 21, 5577 3 of 11 De Rosa G et al. 3 of 11

K562 sh- SEC23B-74-74 andand K562 sh-SEC23B-70-70 cellscells thatthat were both stably silenced for SEC23B were induced to erythroid differentiation differentiation by hemin treatmenttreatment (Figure S1a,b). Five days days after after this this treatment, treatment, there waswas nono significant significant increase increase of of GDF11 GDF11 levels levels in thein the K562 K562 sh- SEC23Bsh-SEC23B-74 cells-74 cells compared compared to the to K562 the sh-CTRK562 sh-CTR cells, andcells, there and there was a was significant a significant marked marked increase increase in GDF11 in GDF11 in the in K562 the K562 sh-SEC23B sh-SEC23B-70 cells-70 (Figurecells (Figure S1c,d). S1c,d). Of note, Of thenote, two the silenced two silenced clones showed clones dishowedfferent behaviorsdifferent behaviors in terms of in expression terms of ofexpressionSEC23A ,of the SEC23A paralog, oftheSEC23B paralog. Indeed, of SEC23B at steady-state,. Indeed, at there steady-state, was significant there downregulationwas significant ofdownregulationSEC23A in the of K562 SEC23A sh- SEC23Bin the K562-70 cells, sh-SEC23B which-70 suggested cells, which a less-specific suggested a e ffless-specificect of the silencing effect of inthe the silencing K562 sh- in SEC23Bthe K562-70 sh- cellsSEC23B (Figure-70 cells S2). (Figure For this S2). reason, For this we reason, chose K562 we chose sh-SEC23B K562 -74sh-SEC23B cells for- the74 cells subsequent for the subsequent analyses. analyses.

2.2. SMAD2 Protein PhosphorylationPhosphorylation is Inhibited is Inhibited by RAP-011 by RAP-011 To reproducereproduce thethe microenvironmentmicroenvironment of the CDA II marrow, humanhuman recombinantrecombinant GDF11 was added to to the the cell cell medium medium at at different different times times (0.5, (0.5, 1, and 1, and 2 h) 2 after h) after two twodays daysof differentiation of differentiation with withhemin. hemin. There Therewas an was increase an increase in the in phosphorylation the phosphorylation of SMAD2 of SMAD2 in the in K562 the K562GDF11-treated GDF11-treated cells cellscompared compared to the non-treated to the non-treated cells (Figure cells 2a). (Figure Comp2a).arison Comparison of the GDF11-treated of the GDF11-treated cells and the GDF11 cells and+ RAP-011-treated the GDF11 + RAP-011-treated cells showed that cells the showed RAP-011 that trea thetment RAP-011 led treatmentto significant led toreduction significant in the reduction levels inof the levelsphosphorylated of the phosphorylated SMAD2 induced SMAD2 by GDF11 induced (Figure by GDF11 2b). (Figure2b).

Figure 2. Effects of RAP-011 on SMAD2 pathway. (a) Representative Western blots and statistical Figure 2. Effects of RAP-011 on SMAD2 pathway. (a) Representative Western blots and statistical analysis based on densitometric quantification of three replicates of phosphorylated SMAD2 (pSMAD2) analysis based on densitometric quantification of three replicates of phosphorylated SMAD2 for K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 for 0 (NT), 0.5, 1 and 2 h (pSMAD2) for K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 for 0 (NT), 0.5, (normalized to total SMAD2/3). (b) Representative Western blots and statistical analysis based on 1 and 2 h (normalized to total SMAD2/3). (b) Representative Western blots and statistical analysis densitometric quantification of three replicates of phosphorylated SMAD2 (pSMAD2) for K562 sh-CTR based on densitometric quantification of three replicates of phosphorylated SMAD2 (pSMAD2) for and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011 (fold-change of K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011 (fold- pSMAD2 for GDF11 + RAP-011 compared to GDF11). Data are means standard deviations. *, p < 0.05; change of pSMAD2 for GDF11 + RAP-011 compared to GDF11). Data are± means ± standard deviations. **, p < 0.01 (Student t-tests). *, p < 0.05; **, p < 0.01 (Student t-tests). 2.3. RAP-011 Treatment Induces Nuclear Translocation of the Transcription Factor GATA1 2.3. RAP-011 Treatment Induces Nuclear Translocation of the Transcription Factor GATA1 To investigate the mechanism through which RAP-011 improved this erythroid differentiation, subcellularTo investigate fractionation the mechanism of the K562 through sh-CTR which and RAP-011 sh-SEC23B improved-74 cells this was erythroid performed. differentiation, There was increasedsubcellular expression fractionation of GATA1 of the inK562 the sh-CTR nuclear and compartment sh-SEC23B after-74 thecells GDF11 was performed.+ RAP-011 There treatment was forincreased both the expression K562 sh-CTR of GATA1 and K562in the sh- nuclearSEC23B co-74mpartment cells, but after in particularthe GDF11 for + RAP-011 this SEC23B treatment-silenced for K562both the cell K562 clone, sh-CTR which showedand K562 a significantsh-SEC23B 3.4-fold-74 cells, increase. but in particular Similarly, for there this was SEC23B increased-silenced expression K562 ofcell the clone, molecular which showed chaperone a significant HSP70 in 3.4-fold the nuclear increase. fractions Similarly, of the there GDF11 was+ increasedRAP-011-treated expression cells. of Conversely,the molecular immunoblotting chaperone HSP70 highlighted in the nuclear significantly fractions decreased of the expression GDF11 + ofRAP-011-treated SMAD4, the nuclear cells. Conversely, immunoblotting highlighted significantly decreased expression of SMAD4, the nuclear mediator of phospho-SMAD2 signaling, in both the K562 sh-CTR and K562 sh-SEC23B-74 cells treated with GDF11 + RAP-011, compared to those treated with GDF11 alone (Figure 3a,b).

Int. J. Mol. Sci. 2020, 21, 5577 4 of 11

De Rosa G et al. 4 of 11 mediator of phospho-SMAD2 signaling, in both the K562 sh-CTR and K562 sh-SEC23B-74 cells treated Dewith Rosa GDF11 G et al. + RAP-011, compared to those treated with GDF11 alone (Figure3a,b). 4 of 11

Figure 3. GATA1 nuclear localization after RAP-011 treatment. (a) Representative Western blot of Figure 3. GATA1 nuclear localization after RAP-011 treatment. (a) Representative Western blot of three Figurethree replicates3. GATA1 of nuclear GATA1, localization HSP70 (HSP-73, after RAP-011 constitutive treatment. isoform; (a) HSP-72,Representative inducible Western isoform) blot and of replicates of GATA1, HSP70 (HSP-73, constitutive isoform; HSP-72, inducible isoform) and SMAD4 threeSMAD4 replicates localization of GATA1, in the nuclear HSP70 comp (HSP-73,artment consti in K562tutive sh-CTR isoform; and HSP-72, K562 sh- inducibleSEC23B-74 isoform) (sh-74) andcells localization in the nuclear compartment in K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated SMAD4treated withlocalization GDF11 in and the GDF11 nuclear + comp RAP-011artment (normalized in K562 sh-CTR to TBP). and (b )K562 Densitometry sh-SEC23B quantification-74 (sh-74) cells of with GDF11 and GDF11 + RAP-011 (normalized to TBP). (b) Densitometry quantification of Western treatedWestern with blots GDF11 of GATA1, and GDF11 HSP70, + RAP-011and SMAD4 (normalized localization to TBP). in the ( bnuclear) Densitometry compartment quantification in K562 sh-of blots of GATA1, HSP70, and SMAD4 localization in the nuclear compartment in K562 sh-CTR and K562 WesternCTR and blots K562 of sh- GATA1,SEC23B HSP70,-74 (sh-74) and cells SMAD4 treated loca withlization GDF11 in theand nuclear GDF11 compartment+ RAP-011 (normalized in K562 sh- to sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011 (normalized to TBP). Data are CTRTBP). and Data K562 are sh- meansSEC23B ± standard-74 (sh-74) deviations. cells treated •, pwith ≤ 0.05 GDF11 (sh-CTR and +GDF11 GDF11 + vsRAP-011 sh-74 +(normalized GDF11); °, pto ≤ means standard deviations. , p 0.05 (sh-CTR + GDF11 vs. sh-74 + GDF11); ◦, p 0.05, ◦◦, p < 0.01 TBP).0.05, °°,Data± p < are 0.01 means (sh-74 ± + standard GDF11• vs deviations.≤ sh-74 + GDF11/RAP-011) •, p ≤ 0.05 (sh-CTR (Student + GDF11 t-tests). vs sh-74≤ + GDF11); °, p ≤ (sh-74 + GDF11 vs. sh-74 + GDF11/RAP-011) (Student t-tests). 0.05, °°, p < 0.01 (sh-74 + GDF11 vs sh-74 + GDF11/RAP-011) (Student t-tests). 2.4. GATA1 Nuclear Translocation Promoted by RAP-011 Restores Gene Expression of Erythroid 2.4. GATA1 Nuclear Translocation Promoted by RAP-011 Restores Gene Expression of Erythroid Markers 2.4.Markers GATA1 Nuclear Translocation Promoted by RAP-011 Restores Gene Expression of Erythroid The responses to the GDF11 and RAP-011 treatments were then analyzed in terms of the expression MarkersThe responses to the GDF11 and RAP-011 treatments were then analyzed in terms of the profiles of the different involved in erythroid differentiation, apoptosis and the GDF11-response expression profiles of the different genes involved in erythroid differentiation, apoptosis and the pathways.The responses The erythroid to the diGDF11fferentiation and RAP-011 markers treatmKLF1,entsABCB6 were, ALAS2, then analyzedand HBG inwere terms markedly of the GDF11-response pathways. The erythroid differentiation markers KLF1, ABCB6, ALAS2, and HBG expressionincreased inprofiles the K562 of sh-theSEC23B different-74 genes cells treated involved with in GDF11erythroid+ RAP-011, differenti comparedation, apoptosis to those and treated the were markedly increased in the K562 sh-SEC23B-74 cells treated with GDF11 + RAP-011, compared GDF11-responsewith GDF11 alone. pathways. Moreover, The there erythroid was rescued differentiation expression markers of BCL2 KLF1in the, ABCB6 K562, sh-ALAS2,SEC23B and-74 HBG cells to those treated with GDF11 alone. Moreover, there was rescued expression of BCL2 in the K562 sh- weretreated markedly with GDF11 increased+ RAP-011, in the K562 compared sh-SEC23B to those-74 treatedcells treated with GDF11with GDF11 alone. + Conversely,RAP-011, compared the K562 SEC23B-74 cells treated with GDF11 + RAP-011, compared to those treated with GDF11 alone. tosh- thoseSEC23B treated-74 cells with showed GDF11 alone. downregulation Moreover, ofthereBAX wasand rescuedBAD expressionafter GDF11 of +BCL2RAP-011 in the treatment. K562 sh- Conversely, the K562 sh-SEC23B-74 cells showed downregulation of BAX and BAD after GDF11 + SEC23BFinally,-74 expression cells treated of the with activin GDF11 receptor + RAP-011, genes was comp alsoared downregulated to those treated in the with K562 GDF11 sh-SEC23B alone.-74 RAP-011 treatment. Finally, expression of the activin receptor genes was also downregulated in the Conversely,cells treated the with K562 GDF11 sh-SEC23B+ RAP-011-74 (Figurecells showed4a,b). downregulation of BAX and BAD after GDF11 + RAP-011K562 sh- SEC23Btreatment.-74 Finally,cells treated expression with GDF11 of the +acti RAP-011vin receptor (Figure genes 4a,b). was also downregulated in the K562 sh-SEC23B-74 cells treated with GDF11 + RAP-011 (Figure 4a,b).

Figure 4.4. Gene expression profilingprofiling inin RAP-011-treatedRAP-011-treated cells.cells. ( a) Heat map forfor expressionexpression profilingprofiling for each of the cell clones (as indicated) treated with GDF11 or GDF11 + RAP-011. Fold-changes of Figurefor each 4. ofGene the cellexpression clones (as profiling indicated) in RAP-011-treatedtreated with GDF11 cells. or (GDF11a) Heat + map RAP-011. for expression Fold-changes profiling of sh- sh-CTR and sh-SEC23B-74 cells treated with GDF11 were calculated on sh-CTR and sh-SEC23B-74 forCTR each and of sh- theSEC23B cell clones-74 cells(as indicated) treated with treated GDF11 with were GDF11 calculated or GDF11 on +sh-CTR RAP-011. and Fold-changes sh-SEC23B-74 of cells sh- cells treated with vehicle. Fold-changes of sh-SEC23B-74 cells treated with GDF11 + RAP-011 were CTRtreated and withsh-SEC23B vehicle.-74 Fold-changes cells treated with of sh- GDF11SEC23B were-74 calculated cells treated on sh-CTRwith GDF11 and sh- +SEC23B RAP-011-74 cellswere calculated on sh-SEC23B-74 cells treated with GDF11. Gene expression: green, low; grey, medium; red, treatedcalculated with on vehicle. sh-SEC23B Fold-changes-74 cells treated of sh- withSEC23B GDF11.-74 cellsGene treated expression: with green,GDF11 low; + RAP-011grey, medium; were high. A gray-scale version of the heat map is shown in Figure S3. (b) Relative expression of KLF1, calculatedred, high. Aon gray-scale sh-SEC23B version-74 cells of thtreatede heat withmap GDF11.is shown Gene in Figure expression: S3. (b) Relativegreen, low; expression grey, medium; of KLF1 , red,ABCB6 high., HBG A gray-scale, BAD, BCL2 version and of AVR2A the heat genes map is in shown K562 sh-CTRin Figure and S3. K562(b) Relative sh-SEC23B expression-74 (sh-74) of KLF1 cells, ABCB6treated, withHBG GDF11, BAD, andBCL2 GDF11 and AVR2A + RAP-011 genes (normalized in K562 sh-CTR to GAPDH and gene).K562 sh- DataSEC23B are means-74 (sh-74) ± standard cells treateddeviations with of GDF11 three andexperiments. GDF11 + RAP-011p-value by(normalized ANOVA totest, GAPDH internal gene). post-hoc Data correctionare means by± standard Tukey’s deviations of three experiments. p-value by ANOVA test, internal post-hoc correction by Tukey’s

Int. J. Mol. Sci. 2020, 21, 5577 5 of 11

ABCB6, HBG, BAD, BCL2 and AVR2A genes in K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011 (normalized to GAPDH gene). Data are means standard ± De Rosadeviations G et al. of three experiments. p-value by ANOVA test, internal post-hoc correction by Tukey’s5 of 11

multiple comparison tests. , p < 0.05; , p < 0.01 (vehicle vs. GDF11 treated cells); ◦, p < 0.05; multiple comparison tests.•, •p <0.05; ••, ••p <0.01 (vehicle vs GDF11 treated cells); °, p < 0.05; °°, p < ◦◦, p < 0.01 (GDF11 treated cells vs. GDF11 + RAP-011 treated cells). 0.01 (GDF11 treated cells vs GDF11 + RAP-011 treated cells).

2.5. RAP-011 Treatment Impairs Erythroferrone Expression 2.5. RAP-011 Treatment Impairs Erythroferrone Expression ToTo investigate the effects effects of RA RAP-011P-011 on the expression of ERFE, it was analyzed at at both both gene gene ERFE SEC23B andand protein protein levels. levels. There There wa wass strong strong upregulation upregulation of of ERFE expressionexpression in in the the K562 K562 sh- sh-SEC23B-74-74 cells cells treatedtreated with with GDF11. GDF11. Conversely, Conversely, ad additiondition of of RAP-011 RAP-011 also also resulted resulted in in significantly significantly marked marked reduction reduction ERFE SEC23B ofof ERFEexpression expression in thesein these K562 sh-K562 sh--74SEC23B cells (Figure-74 cells5a). (Figure Accordingly, 5a). markedAccordingly, downregulation marked downregulationof the ERFE protein of the levels ERFE was protein seen in levels these was K562 seen sh- SEC23Bin these-74 K562 cells sh- followingSEC23B-74 the cells GDF11 following+ RAP-011 the GDF11treatment + RAP-011 (Figure5 treatmentb). (Figure 5b).

Figure 5. ERFE gene and protein expression in RAP-011-treated cells. (a) Quantification of ERFE in K562 sh-CTR and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011. Data are Figure 5. ERFE gene and protein expression in RAP-011-treated cells. (a) Quantification of ERFE in means standard deviations of three experiments. p-value by ANOVA test, internal post-hoc correction K562 sh-CTR± and K562 sh-SEC23B-74 (sh-74) cells treated with GDF11 and GDF11 + RAP-011. Data by Tukey’s multiple comparison tests. , p < 0.05; , p < 0.01 (vehicle vs. GDF11 treated cells); are means ± standard deviations of three• experiments.•• p-value by ANOVA test, internal post-hoc ◦◦, p < 0.01 (GDF11 treated cells vs. GDF11 + RAP-011 treated cells). (b) Representative Western blot of correction by Tukey’s multiple comparison tests. •, p < 0.05; ••, p < 0.01 (vehicle vs GDF11 treated three replicates and densitometry quantification of ERFE expression in K562 sh-CTR and sh-SEC23B-74 cells); °°, p < 0 .01 (GDF11 treated cells vs GDF11 + RAP-011 treated cells). (b) Representative Western (sh-74) cells treated with GDF11 and GDF11 + RAP-011 (normalized to GAPDH). Data are means blot of three replicates and densitometry quantification of ERFE expression in K562 sh-CTR and sh- standard deviations. p value by Student t-test; , p < 0.05 (sh-CTR + GDF11 vs. sh-74 + GDF11); SEC23B± -74 (sh-74) cells treated with GDF11 and GDF• 11 + RAP-011 (normalized to GAPDH). Data are ◦◦, p < 0.01 (sh-74 + GDF11 vs. sh-74 + GDF11 + RAP-011). means ± standard deviations. p value by Student t-test; •, p < 0.05 (sh-CTR + GDF11 vs sh-74 + GDF11); 3. Discussion°°, p < 0.01 (sh-74 + GDF11 vs sh-74 + GDF11 + RAP-011). Blood transfusion therapy or treatments with erythropoiesis-stimulating agents such as 3. Discussion recombinant erythropoietin are the present front-line therapies for anemia associated with ineffective erythropoiesis.Blood transfusion However, therapy both treatments or treatments have side with effects, erythropoiesis-stimulati and they are not alwaysng eff ective.agents Therefore,such as recombinantthere is the clinical erythropoietin need for are novel the compoundspresent front-lin withe ditherapiesfferent mechanisms for anemia associated of action towith those ineffective that are erythropoiesis.already available. However, both treatments have side effects, and they are not always effective. Therefore,Recently, there two is the inhibitors clinical ofneed TGF- for pathway,novel compounds sotatercept with and different luspatercept, mechanisms have been of action evaluated to those for thatthe treatmentare already of available. hereditary anemias, such as β-thalassemia and Diamond–Blackfan anemia [7,13]. SotaterceptRecently, antagonizes two inhibitors GDF11, of TGF- which is a path negativeway, sotatercept regulator of and erythropoiesis, luspatercept, as have well been as several evaluated other formembers the treatment of the TGF- of herediβ superfamilytary anemias, that such signal as through β-thalassemia ActRIIA and [14 Diamond–16]. Of note,–Blackfan aberrant anemia expression [7,13]. Sotaterceptof GDF11 was antagonizes demonstrated GDF11, in awhichβ-thalassemia is a negative murine regulator model of [7 ].erythropoiesis, Although the earlyas well reports as several have otheridentified members GDF11 of as the the TGF- primaryβ superfamily target of ligand that trapssignal [7 ,through12], a recent ActRIIA study [14 demonstrated–16]. Of note, that aberrant Hbbth3/+ expressionand Hbb+/+ ofmice GDF11 deleted was for demonstratedGdf11 are still in able a β to-thalassemia respond to themurine ligand model trap RAP-536, [7]. Although thus suggesting the early reportsthat Gdf11 have is notidentified the only GDF11 effector as of TGF-the primary inhibition ta ofrget late of erythropoiesis ligand traps in [7,12], mice [8a]. recent Nevertheless, study demonstrated that Hbbth3/+ and Hbb+/+ mice deleted for Gdf11 are still able to respond to the ligand trap RAP-536, thus suggesting that Gdf11 is not the only effector of TGF- inhibition of late erythropoiesis in mice [8]. Nevertheless, this is in agreement with the observation that the ligand traps are able to bind also other members of the TGF- family, including GDF8 and activin B [11]. Additionally, we cannot exclude a different genetic or epigenetic regulation in human and mouse

Int. J. Mol. Sci. 2020, 21, 5577 6 of 11 this is in agreement with the observation that the ligand traps are able to bind also other members of the TGF- family, including GDF8 and activin B [11]. Additionally, we cannot exclude a different genetic or epigenetic regulation in human and mouse genes, as described for similar erythroid regulators as GDF15 [17]. It is conceivable to hypothesize that GDF11 is one of the players involved in the regulation of terminal erythroid differentiation. Recently, a phase II clinical trial with sotatercept was carried out with patients with β-thalassemia, which produced encouraging results [18]. As β-thalassemia and CDA II show similar pathophysiologies, we first investigated the GDF11 levels in patients with CDA II. This analysis highlighted overexpression of GDF11 in the patients with CDA II, compared to healthy controls, which identified this biomarker as a possible therapeutic target for CDA II. Given the lack of a reliable mouse models for CDA II [19], we used the previously developed in vitro model of K562 cells stably silenced for SEC23B [5] to investigate the efficacy of RAP-011 treatment. Unlike ex vivo analysis, we did not observe any marked increase in GDF11 levels in these SEC23B-silenced K562 cells when they were induced to erythroid differentiation by hemin, compared to the control K562 cells. This difference might be due to the absence of systemic production of GDF11 in the K562 cell line. Given the increased expression of GDF11 in ineffective erythropoiesis, we treated these K562 cells with this cytokine to simulate the pathological context and analyze the effects of the RAP-011 treatment in our cellular model. It has already been demonstrated that GDF11 acts through its binding to ActRIIA or ActRIIB, with the consequent phosphorylation of the intracellular mediator SMAD2 and activation of the SMAD2/ SMAD3/ SMAD4 complex. In our in vitro cell system, GDF11 treatment of both control K562 cells and SEC23B-silenced K562 cells resulted in increased phosphorylation of SMAD2 at different times, as expected, whereas the combined treatments with GDF11 and RAP-011 produced inhibition of the GDF11-ActR binding that led to reduced phosphorylation of SMAD2. To understand how inhibition of the SMAD2 pathway improves erythroid survival, we investigated the role of the transcription factor GATA1 during this RAP-011 treatment. Arlet and colleagues demonstrated that, in β-thalassemia, the cytosol to nucleus translocation of GATA1 mediated by HSP70 is counteracted by the increased need of HSP70 for folding of denatured proteins. This process induces caspase-3-mediated cleavage of GATA1, and the consequent impairment of erythroid gene expression, end-stage maturation arrest and apoptosis [20]. Accordingly, we observed decreased expression of GATA1 and HSP70 in the nuclear compartment of cells treated with GDF11 alone versus with GDF11 plus RAP-011. Moreover, we observed decreased SMAD4 in the nuclear compartment of these RAP-011-treated cells, which confirmed that the RAP-011 treatment inhibited the GDF11-activated pathway [21]. We observed the downregulation of various erythroid markers in SEC23B-silenced K562 cells treated with GDF11. This was in agreement with Dussiot and colleagues, who demonstrated that treatment with recombinant GDF11 of bone marrow- and spleen-derived erythroblasts blocked terminal erythroblast maturation in thalassemic cells [7]. Thus, to evaluate the effects of RAP-011 on the transcription factor activity of GATA1, we analyzed gene expression of the several GATA1-induced erythroid markers: HBG, KLF1, ALAS2 and ABCB6. In agreement with the enhanced nuclear translocation of GATA1, we observed increased expression of these erythroid marker genes in the SEC23B-silenced K562 cells treated with RAP-011. Furthermore, to determine whether the addition of RAP-011 treatment restored caspase-3-mediated cleavage of GATA1, we also evaluated gene expression of proapoptotic and antiapoptotic genes, namely BAX, BAD and BCL2. Accordingly, after the RAP-011 treatment, we observed reduced expression of both BAX and BAD and overexpression of BCL2. These effects were directly correlated to the RAP-011 treatment. Indeed, the expression levels of the ActRs (types I, IB, IIA, and IIB) were reduced in the RAP-011-treated cells. Iron overload due to reduced expression of hepatic hormone hepcidin is one of the main hallmarks of CDA II. As key erythroid regulator of pathological suppression of hepcidin expression, ERFE is overexpressed in CDA II patients and plays an important role in abnormal erythropoiesis [5]. Indeed, Int. J. Mol. Sci. 2020, 21, 5577 7 of 11 we recently described a low-frequency ERFE variant that is recurrent in patients with CDA II with a severe phenotype and that is associated with increased ERFE expression [22]. Thus, we investigated the effect of RAP-011 on ERFE expression as a biomarker of ineffective erythropoiesis. We observed increased ERFE expression in the SEC23B-silenced K562 cells after the GDF11 treatment. In the cells treated with GDF11 plus RAP-011, this GDF11-induced increased ERFE expression was blocked by addition of RAP-011 treatment and further reduced. This suggests that treatment with RAP-011 might have effects on the iron overload by reducing the levels of ERFE. These results are also in agreement with the phase II clinical trial based on sotatercept for patients with β-thalassemia [18]. Indeed, both transfusion-dependent and non-transfusion-dependent patients showed good responses in terms of increased hemoglobin levels, with reduced red blood cell transfusions for the transfusion-dependent patients, and finally also reduced systemic iron overload [18]. Here, we demonstrated that treatment with RAP-011, a murinized analog of sotatercept, can rescue the disease phenotype in GDF11-treated SEC23B-silenced K562 cells by restoring the expression of erythroid marker genes through inhibition of the phosphorylated SMAD2 pathway. Indeed, inhibition of GDF11-signaling pathway appears to translate into more intense transcriptional activity of these erythroid markers, which might allow undifferentiated erythroblasts to overcome the maturation arrest. These data also demonstrate the beneficial role of RAP-011 treatment for reduction of expression of ERFE, which again supports the use of sotatercept in the management of iron overload for patients with CDA II.

4. Materials and Methods

4.1. Patients Twelve patients with CDA II and 15 age- and gender-matched healthy controls were enrolled in the study. CDA II diagnosis was based on clinical findings and biochemical and molecular analyses (Table S1), as previously reported [4,5,22,23]. The Naples University Ethical Committee (protocol number: 252/18, September 2018) approved the collection of the patient data from the Medical Genetics Ambulatory in Naples (DAIMedLab, “Federico II” University, Naples, Italy). Samples from the patients were obtained after signed informed consent, and according to the Declaration of Helsinki.

4.2. Production of Lentiviral Particles and Infection of the K562 Cell Line Knock-down of SEC23B expression was obtained through infection of lentiviral particles that targeted human SEC23B, in human myeloid leukemia K562 cells, as previously described [5]. Briefly, we used pGIPZ Lentiviral shRNAmir targeting human SEC23B from Open Biosystems (Horizon Discovery, Waterbeach, UK). We used two different sh-RNAs for SEC23B (V3LHS_357970, sh-70; V3LHS_357974, sh-74) (Table S2). A non-silencing pGIPZ Lentiviral shRNAmir was used as control (RHS4346, sh-CTR). HEK-293T were transfected by 10 µg of sh-RNA plasmid DNA, 30 µL of Trans-Lentiviral Packaging Mix (Open Biosystems Horizon Discovery, Waterbeach, UK) and 25 µL of TransFectin (BioRad, Milan, Italy) in 10-mm plate. The supernatants (10 mL for points) were harvested after 24 h, centrifuged at low speed to remove cell debris and filtered through a 0.45-µm filter. After 48 h of incubation, the transduced cells were examined microscopically for the presence of TurboGFP expression (90–95%). The lentiviral particles of sh-CTR, sh-70 and sh-74 (50 MOI) were used to infect K562 cell line. After 48 h of infection, the cells were maintained in puromycin (0.5 mg/mL) for 2 weeks and then analyzed for GFP+ expression and sorted by cell sorting flow cytometry assay. The sorted GFP+ cells were then assayed for SEC23B expression to verify the stability of the produced clones.

4.3. Cell Culture and RAP-011 Treatment The wild-type (control) K562 cells and SEC23B-silenced K562 stable clones were maintained in RPMI medium (Sigma Aldrich, Milan, Italy) supplemented with 10% fetal bovine serum (Sigma Aldrich, Milan, Italy) and grown in a humidified 5% CO2 incubator at 37 ◦C. Erythroid differentiation of the K562 Int. J. Mol. Sci. 2020, 21, 5577 8 of 11 cells was performed as previously described [24]. Briefly, 50 µM of the iron-containing porphyrin hemin (Sigma Aldrich, Milan, Italy) was added to the culture medium containing the wild-type (control) K562 cells (Sh-CTR) and the SEC23B-silenced K562 cell clones (Sh-SEC23B-70 and Sh-SEC23B-74 cells) (at 4 105 cells/mL). Cell samples were collected at specific times: before hemin addition, on day × 0 and on days 2 and 5 after hemin addition. Differentiation was assessed by FACS detection of the transferrin receptor 1 (CD71) [24]. Recombinant human GDF11 protein (1958-GD; R&D Systems, Minneapolis, MN, USA) was used at 50 ng/mL, with RAP-011 at 0.05 g/L (Celgene Corporation, Summit, NJ, USA).

4.4. Gene Expression Analysis

4.4.1. RNA Isolation and Reverse Transcription. Total RNA was extracted from peripheral blood cells and K562 cells using Trizol reagent (Life Technologies, Waltham, MA, USA). Synthesis of cDNA from total RNA (1 µg) was performed using cDNA synthesis kits (Life Technologies, Waltham, MA, USA Roche).

4.4.2. Quantitative Real-Time PCR Analysis. Quantitative real-time PCR (qRT-PCR) analysis was carried out using Power SYBR Green PCR Master Mix (Life Technologies, Waltham, MA, USA) to evaluate gene expression of the GDF11, SEC23B, GATA1, KLF1, ABCB6, ALAS2, HBG, BCL-2, BAX, BAD, ACVR1, ACVR1B, ACVR2A, ACVR2B and ERFE genes (Table S2). The samples were amplified (7900HT Sequence Detection System; Applied Biosystems, Foster City, CA, USA) using standard cycling conditions. The primers were designed with the Primer Express 2.1 software (Applied Biosystems, Foster City, CA, USA). β-Actin and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were used as internal controls. Relative ∆Ct gene expression was calculated using the 2− method [25], with the fold change determined using the ratio between the gene expression and the internal control.

4.5. Protein Expression Analysis

4.5.1. Cell Extracts Proteins were extracted from the cells using RIPA lysis buffer in the presence of a protease inhibitor cocktail (Roche, Rotkreuz, Switzerland). Total protein extracts were analyzed by SDS–PAGE, transferred to polyvinylidene difluoride membranes (BioRad, Milan, Italy), and then incubated with the required combinations of the following antibodies: rabbit anti-GDF11 (1:500; ab124721; Abcam, Cambridge, UK); rabbit anti-SEC23B (1:500; SAB2102104; Sigma Aldrich, Milan, Italy); mouse anti-GATA1 (1:500; H00002623-M06; Abnova, Taipei, Taiwan); rabbit anti-pSMAD2 (1:500; 43108; Technology, Danvers, MA, USA); rabbit anti-SMAD 2/3 (1:1000; 95678; Cell Signaling Technology); mouse anti-HSP70 (1:5000; SAB4200714; Sigma Aldrich, Milan, Italy); rabbit anti-SMAD4 (1:1000; ab215968; Abcam, Cambridge, UK); and rabbit anti-FAM132B (1:200; NBP2-57732; Novus Biologicals, Centennial, CO, USA). Rabbit anti-GAPDH (1:1000; 2118, Cell Signaling Technology, Danvers, MA, USA) and anti-TATA binding protein (TBP; 1:1000; ab51841; Abcam, Cambridge, UK) antibodies were used as the controls for equal loading of total and nuclear protein. Semi-quantitative analysis of protein expression was performed as previously described [26]. The bands were quantified using the Quantity One software (BioRad, Milan, Italy), to obtain integrated optical densities, which were then normalized to GAPDH or TBP.

4.5.2. Secreted Proteins Expression of secreted proteins was analyzed by loading plasma samples (from the healthy controls and patients) onto SDS–PAGE, followed by transfer to polyvinylidene difluoride membranes Int. J. Mol. Sci. 2020, 21, 5577 9 of 11 and incubation with the anti-GDF11 antibody (1:500; ab124721; Abcam, Cambridge, UK). The data were normalized through Ponceau red staining of the blots.

4.6. Subcellular Fractionation Subcellular fractionation of the nuclear and cytosolic proteins was performed according to the Schreiber method [27]. Briefly, the harvested cells were washed twice with ice-cold phosphate-buffered saline and homogenized in ice-cold buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 1 mM EDTA, 0.5 mM dithiothreitol, 10% [v/v] glycerol, 1 mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail [Roche, Rotkreuz, Switzerland]). The suspensions were then repeatedly passed through the needle (26 gauge) of a syringe and then centrifuged at 800 g for 5 min at 4 C × ◦ to separate the cytosol from the nuclear pellet. The nuclear pellet was resuspended in the same lysis buffer in the presence of 3 M KCl, with this nuclear extract stored in ice for 1 h and then centrifuged at 16,000 g for 30 min at 4 C. × ◦ 4.7. Statistical Analysis The statistical significances of the differences in protein and gene expression were assessed using Student’s t-tests or Mann–Whitney tests. Statistical significances of multiple comparisons were calculated using ANOVA, and post-hoc correction was performed using Tukey’s multiple comparison tests. A two-sided p 0.05 was considered statistically significant. ≤ Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/21/15/ 5577/s1. Author Contributions: Conceptualization, G.D.R., R.R. and I.A.; gene expression analysis, F.M.; cellular model, I.A.; protein expression analysis, B.E.R. and R.M.; statistical analysis, R.R.; writing—original draft preparation, G.D.R., R.R. and I.A.; writing—review and editing, R.M. and I.A.; supervision, A.I.; and funding acquisition, R.R. and A.I. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Spanish foundation “Ramón Areces” (CIVP18A1857-CoDysAn project to Achille Iolascon); by Programme STAR, financially supported by UniNA and Compagnia di San Paolo (to Roberta Russo); and by Junior Research Grant 2018, 3978026 of European Hematology Association (EHA) to Immacolata Andolfo. Acknowledgments: This study was supported by Celgene Corp., which provided supplies of the RAP-011 only. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations

CDA II Congenital Dyserythropoietic Anemia type II GDF11 Growth Differentiation Factor 11 CDAs Congenital dyserythropoietic anemias ERFE Erythroferrone TGF-β Transforming Growth Factor beta GDFs Growth Differentiation Factors BMPs Bone Morphogenetic Proteins ActR Activin Receptors HCs Healthy Controls CTR Control NT Non-treated Hbb Hemoglobin subunit beta

References

1. Iolascon, A.; Andolfo , I.; Russo, R. Congenital Dyserythropoietic Anemias. Blood 2020.[CrossRef] 2. Schwarz, K.; Iolascon, A.; Verissimo, F.; Trede, N.S.; Horsley, W.; Chen, W.; Paw, B.H.; Hopfner, K.-P.; Holzmann, K.; Russo, R.; et al. Mutations affecting the secretory COPII coat component SEC23B cause congenital dyserythropoietic anemia type II. Nat. Genet. 2009, 41, 936–940. [CrossRef][PubMed] Int. J. Mol. Sci. 2020, 21, 5577 10 of 11

3. Russo, R.; Esposito, M.R.; Iolascon, A. Inherited hematological disorders due to defects in coat protein (COP)II complex. Am. J. Hematol. 2012, 88, 135–140. [CrossRef][PubMed] 4. Russo, R.; Gambale, A.; Langella, C.; Andolfo, I.; Unal, S.; Iolascon, A. Retrospective cohort study of 205 cases with congenital dyserythropoietic anemia type II: Definition of clinical and molecular spectrum and identification of new diagnostic scores. Am. J. Hematol. 2014, 89, E169–E175. [CrossRef][PubMed] 5. Russo, R.; Andolfo, I.; Manna, F.; De Rosa, G.; De Falco, L.; Gambale, A.; Bruno, M.; Mattè, A.; Ricchi, P.; Girelli, M.; et al. Increased levels of ERFE-encoding FAM132B in patients with congenital dyserythropoietic anemia type II. Blood 2016, 128, 1899–1902. [CrossRef] 6. Kautz, L.; Jung, G.; Nemeth, E.; Ganz, T. Erythroferrone contributes to recovery from anemia of inflammation. Blood 2014, 124, 2569–2574. [CrossRef] 7. Dussiot, M.; Maciel, T.; Fricot, A.; Chartier, C.; Negre, O.; Veiga, J.; Grapton, D.; Paubelle, E.; Payen, E.; Beuzard, Y.; et al. An activin receptor IIA ligand trap corrects ineffective erythropoiesis in beta-thalassemia. Nat. Med. 2014, 20, 398–407. [CrossRef] 8. Guerra, A.; Oikonomidou, P.R.; Sinha, S.; Zhang, J.; Presti, V.L.; Hamilton, C.R.; Breda, L.; Casu, C.; La, P.; Martins, A.C.; et al. Lack of Gdf11 does not improve anemia or prevent the activity of RAP-536 in a mouse model of beta-thalassemia. Blood 2019, 134, 568–572. [CrossRef] 9. Unal, S.; Russo, R.; Gümrük, F.; Ku¸skonmaz,B.; Çetin, M.; Sayli, T.; Tavil, B.; Langella, C.; Iolascon, A.; Uckan-Cetinkaya, D. Successful hematopoietic transplantation in a patient with congenital dyserythropoietic anemia type II. Pediatr. Transplant. 2014, 18, E130–E133. [CrossRef] 10. Uygun, V.; Russo, R.; Karasu, G.; Dalo˘glu,H.; Iolascon, A.; Ye¸silipek,A. Transplantation in Congenital Dyserythropetic Anemia Type II. J. Pediatr. Hematol. Oncol. 2020, 42, e507–e510. [CrossRef] 11. Ruckle, J.; Jacobs, M.; Kramer, W.; E Pearsall, A.; Kumar, R.; Underwood, K.W.; Seehra, J.; Yang, Y.; Condon, C.H.; Sherman, M.L. Single-Dose, Randomized, Double-Blind, Placebo-Controlled Study of ACE-011 (ActRIIA-IgG1) in Postmenopausal Women. J. Bone Miner. Res. 2009, 24, 744–752. [CrossRef] [PubMed] 12. Suragani, R.N.; Cadena, S.M.; Cawley, S.M.; Sako, D.; Mitchell, D.; Li, R.; Davies, M.V.; Alexander, M.J.; Devine, M.; Loveday, K.S.; et al. Transforming growth factor-beta superfamily ligand trap ACE-536 corrects anemia by promoting late-stage erythropoiesis. Nat. Med. 2014, 20, 408–414. [CrossRef][PubMed] 13. E Ear, J.; Huang, H.; Wilson, T.; Tehrani, Z.; Lindgren, A.; Sung, V.; Laadem, A.; Daniel, T.O.; Chopra, R.; Lin, S. RAP-011 improves erythropoiesis in zebrafish model of Diamond-Blackfan anemia through antagonizing lefty1. Blood 2015, 126, 880–890. [CrossRef][PubMed] 14. Iancu-Rubin, C.; Mosoyan, G.; Wang, J.; Kraus, T.; Sung, V.; Hoffman, R. Stromal cell-mediated inhibition of erythropoiesis can be attenuated by Sotatercept (ACE-011), an activin receptor type II ligand trap. Exp. Hematol. 2013, 41, 155–166.e17. [CrossRef][PubMed] 15. Langdon, J.M.; Barkataki, S.; Berger, A.E.; Cheadle, C.; Xue, Q.-L.; Sung, V.; Roy, C.N. RAP-011, an activin receptor ligand trap, increases hemoglobin concentration in hepcidin transgenic mice. Am. J. Hematol. 2014, 90, 8–14. [CrossRef] 16. Raje, N.; Vallet, S. Sotatercept, a soluble activin receptor type 2A IgG-Fc fusion protein for the treatment of anemia and bone loss. Curr. Opin. Mol. Ther. 2010, 12, 586–597. 17. Ulirsch, J.C.; Lacy, J.N.; An, X.; Mohandas, N.; Mikkelsen, T.S.; Sankaran, V.G. Altered Chromatin Occupancy of Master Regulators Underlies Evolutionary Divergence in the Transcriptional Landscape of Erythroid Differentiation. PLoS Genet. 2014, 10, e1004890. [CrossRef] 18. Cappellini, M.D.; Porter, J.; Origa, R.; Forni, G.L.; Voskaridou, E.; Galactéros, F.; Taher, A.T.; Arlet, J.B.; Ribeil, J.-A.; Garbowski, M.; et al. Sotatercept, a novel transforming growth factor beta ligand trap, improves anemia in beta-thalassemia: A phase II, open-label, dose-finding study. Haematologica 2019, 104, 477–484. [CrossRef] 19. Khoriaty, R.; Vasievich, M.P.; Jones, M.; Everett, L.; Chase, J.; Tao, J.; Siemieniak, D.; Zhang, B.; Maillard, I.; Ginsburg, D. Absence of a Red Blood Cell Phenotype in Mice with Hematopoietic Deficiency of SEC23B. Mol. Cell. Biol. 2014, 34, 3721–3734. [CrossRef] 20. Arlet, J.B.; Ribeil, J.-A.; Guillem, F.; Negre, O.; Hazoume, A.; Marcion, G.; Beuzard, Y.; Dussiot, M.; Moura, I.C.; Demarest, S.; et al. HSP70 sequestration by free alpha-globin promotes ineffective erythropoiesis in beta-thalassaemia. Nature 2014, 514, 242–246. [CrossRef] Int. J. Mol. Sci. 2020, 21, 5577 11 of 11

21. Lucarelli, P.; Schilling, M.; Kreutz, C.; Vlasov, A.; Boehm, M.E.; Iwamoto, N.; Steiert, B.; Lattermann, S.; Wäsch, M.; Stepath, M.; et al. Resolving the Combinatorial Complexity of Smad Protein Complex Formation and Its Link to Gene Expression. Cell Syst. 2018, 6, 75–89.e11. [CrossRef][PubMed] 22. Russo, R.; Esposito, M.R.; Asci, R.; Gambale, A.; Perrotta, S.; Ramenghi, U.; Forni, G.L.; Uygun, V.; Delaunay, J.; Iolascon, A. Mutational spectrum in congenital dyserythropoietic anemia type II: Identification of 19 novel variants in SEC23B gene. Am. J. Hematol. 2010, 85, 915–920. [CrossRef][PubMed] 23. Russo, R.; Gambale, A.; Esposito, M.R.; Serra, M.L.; Troiano, A.; De Maggio, I.; Capasso, M.; Luzzatto, L.; Delaunay, J.; Tamary, H.; et al. Two founder mutations in the SEC23B gene account for the relatively high frequency of CDA II in the Italian population. Am. J. Hematol. 2011, 86, 727–732. [CrossRef][PubMed] 24. Russo, R.; Marra, R.; Andolfo, I.; De Rosa, G.; Rosato, B.E.; Manna, F.; Gambale, A.; Raia, M.; Unal, S.; Barella, S.; et al. Characterization of Two Cases of Congenital Dyserythropoietic Anemia Type I Shed Light on the Uncharacterized C15orf41 Protein. Front. Physiol. 2019, 10, 621. [CrossRef][PubMed] 25. Russo, R.; Langella, C.; Esposito, M.R.; Gambale, A.; Vitiello, F.; Vallefuoco, F.; Ek, T.; Yang, E.; Iolascon, A. Hypomorphic mutations of SEC23B gene account for mild phenotypes of congenital dyserythropoietic anemia type II. Blood Cells Mol. Dis. 2013, 51, 17–21. [CrossRef] 26. Russo, R.; Andolfo, I.; Gambale, A.; De Rosa, G.; Manna, F.; Arillo, A.; Wandroo, F.; Bisconte, M.G.; Iolascon, A. GATA1 erythroid-specific regulation of SEC23B expression and its implication in the pathogenesis of congenital dyserythropoietic anemia type II. Haematologica 2017, 102, e371–e374. [CrossRef] 27. Schreiber, E.; Harshman, K.; Kemler, I.; Malipiero, U.; Schaffner, W.; Fontana, A. Astrocytes and glioblastoma cells express novel octamer-DNA binding proteins distinct from the ubiquitous Oct-1 and B cell type Oct-2 proteins. Nucleic Acids Res. 1990, 18, 5495–5503. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).