integrated protein quality-control pathways regulate free -globin in

RED CELLS, IRON, AND ERYTHROPOIESIS

Integrated protein quality-control pathways regulate free �-globin inmurine �-thalassemiaEugene Khandros,1 Christopher S. Thom,1 Janine D’Souza,2 and Mitchell J. Weiss2

1Cell and Molecular Biology Graduate Group, University of Pennsylvania School of Medicine, Philadelphia, PA; and 2Division of Hematology, Children’s Hospitalof Philadelphia, Philadelphia, PA

Cells remove unstable polypeptidesthrough protein quality-control (PQC)pathways such as ubiquitin-mediated pro-teolysis and autophagy. In the presentstudy, we investigated how these path-ways are used in �-thalassemia, a com-mon hemoglobinopathy in which �-globin gene mutations cause the accumu-lation and precipitation of cytotoxic �-globin subunits. In �-thalassemic erythro-cyte precursors, free �-globin waspolyubiquitinated and degraded by theproteasome. These cells exhibited en-

hanced proteasome activity, and tran-scriptional profiling revealed coordinatedinduction of most proteasome subunitsthat was mediated by the stress-responsetranscription factor Nrf1. In isolated thalas-semic cells, short-term proteasome inhibi-tion blocked the degradation of free�-globin. In contrast, prolonged in vivotreatment of �-thalassemic mice with theproteasome inhibitor bortezomib did notenhance the accumulation of free �-globin. Rather, systemic proteasome inhi-bition activated compensatory proteo-

toxic stress-response mechanisms, in-cluding autophagy, which cooperated withubiquitin-mediated proteolysis to degradefree �-globin in erythroid cells. Our find-ings show that multiple interregulatedPQC responses degrade excess �-globin.Therefore, �-thalassemia fits into thebroader framework of protein-aggregationdisorders that use PQC pathways as cell-protective mechanisms. (Blood. 2012;119(22):5265-5275)

Introduction

The production of functional hemoglobin A (HbA) tetramers (�2�2)requires the coordinated synthesis and assembly of �- and �-globinprotein chains and iron-containing heme groups. Individually, allHbA components are toxic to RBCs and their precursors, asillustrated by �-thalassemias, a common hemoglobinopathy inwhich �-globin gene (HBB) mutations cause the buildup of free�-globin.1 These unpaired � chains initiate an oxidative damagecascade and form damaging precipitates that contribute largely tothe clinical problems associated with �-thalassemia.

The pathophysiology of �-thalassemia bears similarities to adiverse group of protein-aggregation diseases affecting multipleorgans (for review, see Khandros and Weiss2). These disorders,which include Parkinson disease, Alzheimer disease, Huntingtondisease, amyotrophic lateral sclerosis, and �1-antitrypsin defi-ciency, are caused by the accumulation of unstable, relativelyinsoluble proteins. It is believed that the affected cells can detoxifyand remove these damaging proteins via multiple interactingbiochemical pathways called protein quality-control (PQC) path-ways, but that disease ensues when such compensatory mecha-nisms are overwhelmed (for review, see Ciechanover and Brundin,3

Ding and Yin,4 and Jaeger and Wyss-Coray5). Cellular PQCsystems include molecular chaperones, ubiquitin-mediated proteol-ysis, and autophagy. Several lines of evidence suggest that�-thalassemic erythroid cells use PQC pathways to detoxify free�-globin (for review, see Khandros and Weiss2): (1) the clinicalseverity of �-thalassemia is proportional to the degree of �-globinexcess; (2) there is a threshold below which excess �-globin is lessharmful, as illustrated by subjects with the �-thalassemia trait, who

experience 50% reduced �-globin synthesis with minimal clinicalmanifestations or accumulation of �-globin precipitates; and(3) there is direct biochemical evidence that �-globin interacts withcellular PQC components.

Numerous studies have shown that normal and �-thalassemicerythroid precursors can balance globin ratios through selective�-chain proteolysis. Pulse-chase experiments using intact human�-thalassemic erythroid cells6-13 and cell lysates11,13,14 showed thatexcessive � chains are actively degraded and accumulate mainly inthe late stages of erythroid maturation, presumably as the proteo-lytic capacity becomes exceeded. The ubiquitin proteasome system(UPS; for review, see Ciechanover and Brundin3), originallycharacterized in reticulocyte lysates using denatured Hb as asubstrate,15 is responsible for physiologic degradation of nativeproteins and for removing misfolded proteins as part of the PQCpathway in all cells. Studies by Shaeffer et al showed that normaland �-thalassemic hemolysates can ubiquitinate and degradeexogenous �-globin,12,14 although the associated pathways remainlargely uncharacterized. RBC precursors also use autophagy, agroup of related processes in which targeted proteins or organellesare fused to lysosomes and degraded.16 For example, autophagy-related genes are up-regulated by the master erythroid transcriptionfactor GATA-1 during terminal erythropoiesis.17 During reticulo-cyte maturation, mitochondria are eliminated by “macroau-tophagy” (for review, see Ding and Yin4 and Mizushima et al16), aprocess in which cells form double-membrane vesicles (autophago-somes) around cytoplasmic contents for delivery to lysosomes.18-20

Interestingly, electron micrographs of �-thalassemic erythroblasts

Submitted December 8, 2011; accepted March 7, 2012. Prepublished online asBlood First Edition paper, March 16, 2012; DOI 10.1182/blood-2011-12-397729.

There is an Inside Blood commentary on this article in this issue.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page chargepayment. Therefore, and solely to indicate this fact, this article is herebymarked ‘‘advertisement’’ in accordance with 18 USC section 1734.

© 2012 by The American Society of Hematology

5265BLOOD, 31 MAY 2012 � VOLUME 119, NUMBER 22

For personal use only.on April 1, 2018. by guest www.bloodjournal.orgFrom

http://www.bloodjournal.org/

http://www.bloodjournal.org/site/subscriptions/ToS.xhtml

identify a subset of �-globin precipitates within lysosomes.2,21,22

More recent work indicates that autophagic processes are increasedin HbE/�-thalassemia.23

Most studies linking PQC to detoxification of free �-globin in�-thalassemia were performed before the development of newergenetic and pharmacologic approaches to interrogating the relevantmechanisms. Moreover, recent proof-of-principle studies haveshown that the induction of PQC can improve phenotypes ofvarious aggregation disorders in murine models. For example,pharmacologic activation of autophagy can attenuate liver damagein �1-antitrypsin deficiency,24 and genetic up-regulation of protea-some activity can alleviate proteotoxic heart disease.25 However,before such principles can be applied to �-thalassemia, therelevance and contributions of PQC pathways involved in neutral-izing free �-globin must be better defined. In the present study, weinvestigated this problem in a mouse model of �-thalassemia and inhuman patient cells, and found that the processes of UPS,autophagy, and heat shock all likely participate in the detoxificationof free �-globin and are coordinately regulated.

Methods

Mice

�-thalassemic (�-globinTh3/�) mice were kindly provided by OliverSmithies (University of North Carolina, Chapel Hill, NC),26 and werebackcrossed onto a C57BL/6J background for 9-10 generations. All animalexperiments were done in accordance with protocols approved by theInstitutional Animal Care and Use Committee of The Children’s Hospital ofPhiladelphia.

Human samples

Blood samples were collected from �-thalassemia patients with pretransfu-sion reticulocyte counts of 3%-10%) and healthy controls at the Children’sHospital of Philadelphia per research protocols approved by the localinstitutional review board. Written informed consent was obtained from allparticipants.

Isolation and analysis of insoluble globin fractions

Analysis of globin precipitates in circulating erythrocyte membrane skel-etons was performed as described previously.27 For fetal liver cultures, cellswere lysed in RIPA buffer containing 1mM DTT, 1mM PMSF, and proteaseinhibitor cocktail (Sigma-Aldrich). Insoluble fractions were collected bycentrifugation at 16 000g, washed, and solubilized by boiling in 2�Laemmli buffer (Sigma-Aldrich). See supplemental Methods (available onthe Blood Web site; see the Supplemental Materials link at the top of theonline article) for detailed denaturing immunoprecipitation and Westernblotting protocols.

Reticulocyte pulse-chase analysis

Freshly collected mouse erythrocytes and reticulocytes were used forpulse-chase experiments with 35S-labeled methionine and cysteine (Perkin-Elmer),27 as described in supplemental Methods. Where indicated, chasemedium contained 100�M chloroquine (Sigma-Aldrich), 0.5�M epoxomi-cin, 10�M MG132 (Enzo Life Sciences), or 0.1% DMSO as a control.

Fetal liver cultures

Fetal livers were collected from embryonic day 14.5 embryos from crossesof wild-type or Th3/� � Th3/� mice. Embryos were genotyped anderythroid precursors were isolated from individual embryos using theEasySep hematopoietic progenitor enrichment kit (StemCell Technologies)

Figure 1. Insoluble �-globin accumulates after protea-some inhibition in �-thalassemic erythroid cells.(A) Coomassie-stained triton acetic acid urea gel show-ing �-globin precipitates in Th3/� erythrocyte mem-branes. The marker lane M shows purified � and �-globins. (B) Th3/� reticulocytes were pulse labeled with35S-methionine and 35S-cysteine and chased with unla-beled amino acids for the indicated periods of time with orwithout proteasome inhibitors. Radiolabeled soluble andinsoluble globins were isolated from equal numbers ofcells, fractionated by triton acetic acid urea gel electropho-resis, and visualized by autoradiography. (C) Quantifica-tion of autoradiographs from panel B; n � 3 mice/group.*P � .05 versus DMSO. (D) �-globin Western blots ofsoluble and insoluble fractions from mouse fetal livererythroid cultures (48 hours differentiation) of wild-type,Th3/�, and Th3/Th3 erythroblasts. (E) �-globin Westernblots of soluble and insoluble fractions from mouse fetalliver erythroid cultures (72 hours of differentiation) treatedwith vehicle or MG132 (1 or 10�M) for 12 hours.

5266 KHANDROS et al BLOOD, 31 MAY 2012 � VOLUME 119, NUMBER 22




supplemented with biotin-conjugated CD71 Ab (BioLegend). PurifiedCD71�Lin� cells were cultured in differentiation medium consisting ofIMDM with 10% FCS, 10% PDS, 2mM L-glutamine, 10�M 1-thioglycerol,1% penicillin/streptomycin, 5% PFHM-II medium, and 5 U/mL of erythro-poietin (Amgen). For retroviral infections, cells were cultured in expansionmedium consisting of StemPro34 medium (Invitrogen) supplemented with2mM L-glutamine, 1% penicillin/streptomycin, 10�M 1-thioglycerol, 1�Mdexamethasone, 0.5 U/mL of erythropoietin, and 1% murine SCF-conditioned medium.28 After 72 hours of expansion, cells were washed andresuspended in differentiation medium. For Nrf1 and Nrf2 inductionexperiments, cells in differentiation medium were treated for 24 hourswith 0.1�M MG132 or 1�M R-sulforaphane (LKT laboratories). Forproteasome-inhibition studies, cells were incubated for 12 hours with 1 or10�M MG132.

Retroviral shRNA delivery

shRNA constructs were purchased from OpenBiosystems (see supplemen-tal Methods) and selected hairpins cloned into the MSCV-PIG (puromycin-IRES-GFP) vector. Cells (5 � 104) were spinfected at 1300g with 50 �L ofretroviral supernatant and 8 �g/mL of polybrene for 90 minutes at 30°C.

Flow cytometry

Flow cytometry staining is described in supplemental Methods. Forproteasome activity quantification, cells were first stained with 1�MMV151 (Chemical Proteomics Reagents, Leiden Institute for Chemistry,Leiden, The Netherlands) for 4 hours in culture medium.29 Human patient

erythrocytes were similarly treated, but were stained with Hoescht33342and thiazole orange (ReticCOUNT; BD Biosciences) instead of specificAbs. Cells were analyzed on an LSRII or an LSRFortessa instrument (BDBiosciences) maintained by the Flow Cytometry Core Laboratory at TheChildren’s Hospital of Philadelphia Research Institute.

Microarray analysis

CD71�Ter119�FSChigh cells were sorted from E14.5 fetal livers of Th3/� � Th3/� mouse crosses (3 embryos per genotype) using a FACSAriaIIcell sorter (BD Biosciences). Cells were sorted directly into TRIzol LSreagent (Invitrogen) and RNA was prepared using the RNeasy kit (QIA-GEN). Samples were processed for microarray analysis using the MouseGene 1.0ST Array (Affymetrix) by the microarray core facility at theUniversity of Pennsylvania. Microarray data reported herein were submit-ted to the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/44)as accession number GSE34125. Further details are provided in supplemen-tal Methods.

Bortezomib treatment and hematologic analysis

Mice were treated by IP injection of 0.25, 0.5, or 1.0 mg/kg of bortezomib(Velcade; Millennium Pharmaceuticals) in normal saline or of saline controlevery 3 days. Blood was collected by submandibular bleeding, anticoagu-lated with EDTA, and analyzed on a Hemavet HV950FS analyzer (DrewScientific). Reticulocyte counts were done using Retic-COUNT reagent(BD Biosciences). BM erythroid precursors were purified from mice using

Figure 2. Free �-globin is ubiquitinated in �-thalassemicerythroid cells. Immunoprecipitation-Western blot anal-ysis of solubilized �-globin aggregates from Th3/� eryth-rocytes. Membrane-associated �-globin was isolatedfrom mouse Th3/� (A,C) or human �-thalassemia major(B,D) erythrocytes and solubilized in SDS. Samples wereimmunoprecipitated with anti–�-globin (A-B) or anti-ubiquitin (C-D) Abs and analyzed by Western blottingusing the indicated Abs. Input fractions and preimmuneserum (IgG) immunoprecipitation are included as con-trols. Immunoprecipitation reactions containing purified�-globin or �-globin competitors demonstrateAb specificity.

DEGRADATION OF FREE �-GLOBIN IN �-THALASSEMIA 5267BLOOD, 31 MAY 2012 � VOLUME 119, NUMBER 22




Figure 3. Proteasome components are up-regulated in �-thalassemic erythroblasts. (A) Transcriptome analysis was performed on FACS-purified, developmentalstage–matched fetal liver erythroblasts from wild-type, heterozygous (Th3/�), and homozygous (Th3/Th3) thalassemic mouse embryos (supplemental Figure 1). Gene SetEnrichment Analysis (GSEA) reveals the induction of proteasome subunit mRNAs in Th3/� (left) and Th3/Th3 (right) erythroblasts compared with wild-type controls.(B) Relative expression levels of proteasome subunit mRNAs in �-thalassemic erythroblasts normalized to wild-type controls. (C) Proteasome activities in fetal liver erythroidcultures using the fluorescent proteasome activity indicator MV151. Samples were costained for expression of the erythroid-specific antigen Ter119 and for DNA usingHoescht33342 to distinguish nucleated and enucleated erythroid cells. (D) MV151 mean fluorescence intensity (MFI) for nucleated (top) and enucleated (bottom) erythroidcells (Ter119�) in 48-hour erythroid cultures from wild-type and Th3/� fetal livers; n � 4 embryos/genotype. (E) MV151 MFI for circulating human reticulocytes(Hoescht33342�, thiazole orange�) from control or �-thalassemia major patients. Values are normalized to controls.





the EasySep PE Selection Kit (StemCell Technologies) with Ter119-PE Ab(BioLegend).

Real-time RT-PCR

Quantitative real-time RT-PCR used the standard curve method with SYBRGreen dye on an ABI 7900 real-time machine (PE Applied Biosystems).Target gene expression was normalized to the average of Actb and Hprtvalues. For proteasome subunit PCR, multiple randomly selected subunitswere examined, and the average fold change is presented. Primer sequencesare described in supplemental Methods.

Statistical methods

Statistical analysis was performed using GraphPad Prism Version 4.0 soft-ware. All multigroup comparisons were done using 1-way ANOVA.Comparisons between 2 groups were done using the Student t test.

Results

Insoluble �-globin accumulates in proteasome-inhibited�-thalassemic reticulocytes.

We investigated whether pharmacologic inhibition of the UPSimpairs �-globin turnover and accumulation in �-thalassemicerythroid cells. We used Th3 mutant mice in which both the �1(Hbb1) and �2 (Hbb2) adult globin genes are deleted.26 Homozy-gous mutants (Th3/Th3) die in utero or perinatally of severeanemia. Heterozygous animals (Th3/�) are viable and exhibitmicrocytic hypochromic anemia with accumulation of insoluble�-globin chains in erythroid precursors and in the membranes ofcirculating erythrocytes (Figure 1A). In general, the extent of�-globin precipitation within the latter reflects the severity of�-thalassemia and can be used as a surrogate marker to gaugedisease modifiers.27,30,31

Pulse radiolabeling, followed by chase with unlabeled aminoacids, demonstrated that insoluble �-globin was degraded with ahalf-life of approximately 2 hours in murine Th3/� reticulocytes(Figure 1B-C). Similar degradation of membrane-associated in-soluble �-globin has been observed in human samples.12,13 Treat-ment with 2 different proteasome inhibitors, MG132 and epoxomi-cin, during the chase period decreased the turnover of �-globinaggregates, indicating that �-globin is degraded at least in part byubiquitin-mediated proteolysis (Figure 1B-C).

To study �-globin degradation further, we used a mouse fetalliver erythroid culture system in which purified erythroid progeni-tors mature synchronously over 72 hours.32 In this system, culturedmurine Th3/Th3 cells accumulate insoluble �-globin and begin todie at approximately 36-48 hours of differentiation, whereas theTh3/� cells mature at a slightly delayed rate relative to wild-typecells. Remarkably, after 48 hours of culture, Th3/� erythroblastscontained relatively little insoluble �-globin compared with Th3/Th3 cells (Figure 1D), which is consistent with the presence ofcompensatory proteolytic mechanisms as observed in humanthalassemic erythroblasts.7,9,12 Treatment with MG132 caused adose-dependent accumulation of insoluble �-globin chains inTh3/� cultured erythroblasts and to a lesser extent in control cells(Figure 1E top 2 panels). The latter finding suggests that the UPSmay degrade the free � chains that are synthesized at slight excessrelative to � chains during normal erythropoiesis.33 SQSTM1/p62,which serves as a marker of insoluble protein aggregates instructures called aggresomes34 (see “Discussion”), also copurifiedwith this fraction after proteasome inhibition in wild-type andTh3/� cells.

Free �-globin is polyubiquitinated in �-thalassemic cells

Previous studies of globin chain removal by the UPS have focusedprimarily on degradation of destabilized globins in vivo orartificially denatured Hbs in reticulocyte lysates. Shaeffer et alshowed that exogenous purified �-globin chains are ubiquitinatedand degraded by the proteasome in �-thalassemic patient hemoly-sates,14 but ubiquitination of endogenous �-globin has not beendemonstrated previously. We investigated whether �-globin isactually ubiquitinated in vivo in thalassemic cells. We useddenaturing immunoprecipitation from samples boiled in 1% SDS topurify �-globin chains from native circulating �-thalassemicerythroid cells under conditions that prevent noncovalent protein-protein associations during immunoprecipitation. Because mostsoluble �-globin in Th3/� RBCs exists in stable complexes with�-globin as HbA, we performed the immunoprecipitation experi-ments using material from �-globin–enriched insoluble fractions.Immunoprecipitation by �-globin–specific Ab, followed by West-ern blot analysis using an anti-ubiquitin Ab, revealed high-molecular-weight polyubiquitinated �-globin in RBCs from Th3/�mice (Figure 2A) and a �-thalassemia major human patient (Figure2B). Inclusion of competitor �-globin, but not �-globin, during theimmunoprecipitation blocked the recovery of polyubiquitinatedproteins, demonstrating Ab specificity. Denaturing immunoprecipi-tation using an Ab that recognizes mono- and polyubiquitinatedproteins, but not free ubiquitin, followed by �-globin Westernblotting, also revealed ubiquitinated �-globin in mouse (Figure 2C)and human (Figure 2D) thalassemic RBCs. Ubiquitinated �-globinwas not detected in wild-type RBCs (not shown). Therefore,�-globin is polyubiquitinated in mouse and human �-thalassemicRBCs in vivo.

Proteasome subunits are up-regulated in �-thalassemicerythroblasts

Total proteolytic activity and �-globin–specific proteolysis areincreased in �-thalassemic erythroblasts relative to normal patientsamples,10,11 but the identity of the proteases and the associatedmechanisms have not been determined. We compared gene-expression profiles of flow cytometry–purified erythroblasts fromembryonic day 14.5 fetal livers of wild-type, Th3/�, and Th3/Th3mice. We examined Ter119�CD71�FSChigh cells to compare devel-opmental stage-matched samples at a point when Hb synthesisbecomes active and when Th3/Th3 homozygous erythroblastsremain viable (supplemental Figure 1A). Hierarchical clustering ofsignificantly up-regulated or down-regulated (P � .01 by ANOVA)genes grouped mice on the basis of genotype and showed thatwild-type and Th3/� cells are more similar to each other than toTh3/Th3 cells (supplemental Figure 1B). There were relatively fewchanges in Th3/� cells compared with wild-type samples, with anR2 value of 0.9925 for �/� versus Th3/�, compared with 0.8930for �/� versus Th3/Th3, suggesting that heterozygous mutanterythroblasts compensate for �-globin synthetic excess, therebyminimizing its impact on gene expression.

We performed Gene Set Enrichment Analysis (GSEA) to assesswhether transcripts that are differentially expressed in �-thalassemiaare enriched for specific biologic properties or functions.35 Themost highly enriched gene set in both Th3/� and Th3/Th3erythroblasts was the proteasome pathway (Figure 3A), which wasoverrepresented in �-thalassemia at a high statistical significance(normalized enrichment scores of 2.44 for Th3/� and 2.11 forTh3/Th3, with false discovery rate q-values less than 0.0001).Messenger RNAs encoding all subunits of the catalytically active





20S proteasome complex, and almost all subunits of the regulatory19S complex, were up-regulated in an �-globin–dosage dependentmanner (Figure 3B). The � subunits specific for the immunoprotea-some that is involved in the generation of peptides for MHCdisplay were not altered. We used MV151, a fluorescent protea-some inhibitor that binds specifically to active proteasome sub-units,29 to determine whether proteasome activity is enhanced in�-thalassemic cells (Figure 3C). Costaining of cultured fetal livererythroid cells with MV151 and stage-specific markers revealedelevated proteasome activities in Th3/� cells, in both immature

nucleated erythroblasts (Ter119�Hoechst33342�) and enucleatedreticulocytes (Ter119�Hoechst33342�). The magnitude of theincrease was similar to the transcript changes shown in Figure 3B.We were not able to examine Th3/Th3 cultures because of hightoxicity of MV151 in these cells. We also used MV151 to analyzeproteasome activity in circulating reticulocytes from human �-thalassemia major patients, whose reticulocytes also had increasedproteasome activity relative to reticulocytes from healthy controls(Figure 3E). These data indicate that up-regulation of proteasomesubunit transcripts in �-thalassemia leads to the production of newfunctional proteasomes.

Transcription factor Nrf1 induces proteasome subunit mRNAsin normal and �-thalassemic erythroblasts

Proteasome subunit genes contain antioxidant response elements(AREs) that are bound by members of the NFE2-like proteinfamily, Nrf1 and Nrf2, which accumulate in cells during specificstresses.36-38 Typically, oxidative stress stabilizes Nrf2 and proteo-toxic stress/proteasome inhibition stabilizes Nrf1.37,38 We exam-ined how these transcription factors regulate erythroid proteasomesubunit gene expression. First, we investigated whether activationof Nrf1 or Nrf2 up-regulates proteasome subunits in wild-type fetalliver erythroblasts. Treatment with the Nrf2 activator sul-foraphane39 induced the Nrf2-specific target gene Nqo1 (Figure4A), but had no effect on proteasome subunit mRNAs (Figure 4B).In contrast, treatment with a low dose (0.1�M) of the proteasomeinhibitor MG132 induced proteasome subunit mRNAs (Figure 4B)in a manner similar to what occurs in nonerythroid cells throughNrf1-mediated effects.37,38 MG132 did not up-regulate Nqo1 mRNA(Figure 4A), indicating that Nrf2 was not activated. Proteasomesubunit mRNA induction by MG132 was blocked entirely byshRNA-mediated knock-down of Nrf1 (Figure 4C-D). These dataindicate that activation of Nrf1 by MG132 induces proteasomesubunit gene expression, whereas activation of Nrf2 by sul-foraphane does not. Nrf2 shRNA partially blocked proteasomesubunit mRNA induction by MG132 (Figure 4C,E-F), but thiseffect may be indirect because Nrf1 expression was also inhibited(Figure 4D). Proteasome inhibition by MG132 also activatesheat-shock responses with up-regulation of molecular chaperones(Figure 4G).40 Incubation at 42°C for 1 hour to induce heat shockincreased Hsp105 transcripts (Figure 4G), but failed to up-regulateproteasome subunits (Figure 4C). Therefore, the induction ofproteasome subunits by proteasome inhibition is independent ofheat-shock responses.

We also investigated whether proteasome subunit mRNAup-regulation in �-thalassemic erythroblasts was attenuated byshRNAs against Nrf1 or Nrf2. Only shRNAs against Nrf1 blockedthe up-regulation of proteasome subunits in Th3/Th3 cells (Figure4H and supplemental Figure 2). Somewhat paradoxically, Nrf2knock-down increased proteasome subunit gene expression specifi-cally in Th3/Th3 cells (Figure 4H and supplemental Figure 2). Thismay reflect competition between Nrf1 and Nrf2 for binding toAREs within proteasome subunit genes and/or for obligate het-erodimerization partner Maf proteins.36 In this case, suppression ofNrf2 would favor binding of the more potent activator Nrf1 toAREs within proteasome subunit genes to further induce theirexpression. Overall, our findings indicate that Nrf1 largely medi-ates the induction of proteasomal subunit mRNAs in normalerythroblasts exposed to proteotoxic stress by MG132 and in�-thalassemic erythroblasts, which experience proteotoxic stressthrough the accumulation of free �-globin.

Figure 4. Proteasome subunit up-regulation in �-thalassemia is Nrf1 depen-dent. (A-B) Wild-type murine fetal liver erythroid cultures were treated for 24 hourswith 1�M sulforaphane or 0.1�M MG132 to activate Nrf2 or Nrf1, respectively.(A) Expression of Nqo1, an Nrf2 target gene; n � 4 embryos/group. ***P � .001versus control (CTRL). (B) Proteasome subunit mRNA expression (average foldchange of 3 randomly selected subunits) normalized to control. (C-G) Wild-type fetalliver erythroid precursors were infected with retroviruses encoding control (anti-luciferase, Luc), Nrf1, or Nrf2 targeted shRNAs and differentiated for 24 hours with orwithout 0.1�M MG132. As an additional control, Luc shRNA–infected cells were alsoheat-shocked at 42°C for 1 hour and allowed to recover for 1 hour at 37°C (HS).Transcript expression was normalized to �-actin and Hprt mRNAs and comparedamong experimental groups with control-treated Luc shRNA samples assigned anarbitrary value of 1.0. Proteasome subunit expression is shown as the average foldchange of an expanded panel of 9 proteasome subunit mRNAs (C), Nrf1 (D), Nrf2(E), Nqo1 (F), and Hsp105 (G); n � 3 embryos per group. **P � .01 versus Luc Ctrl;***P � .001 versus Luc Ctrl; ##P � .01 versus Luc MG132; ###P � .001 versus �/�Luc. (H) Proteasome subunit mRNA expression in wild-type or Th3/Th3 fetal livererythroblasts infected with retroviruses expressing shRNAs targeting Nrf1, Nrf2, orluciferase (Luc). Analysis was performed after 48 hours of expansion and 44 hours ofdifferentiation. Data shown are the average fold change for 9 subunit transcripts;n � 3 embryos/group. *P � .05 versus �/� Luc; **P � .01 versus �/� Luc,#P � .05 versus Th3/Th3 Luc; ## P � .001 versus �/� Luc.





In vivo proteasome inhibition impairs erythropoiesis inwild-type and �-thalassemic mice

To determine the in vivo importance of the UPS in modulating�-globin toxicity, we treated �-thalassemic mice with the protea-some inhibitor bortezomib.41 Proteasome inhibition is potentiallytoxic to all cells and has been shown to impair normal erythroiddifferentiation in vitro.42 However, we reasoned that if �-thalassemicerythroblasts are dependent on the UPS for degradation of excess�-globin, then affected mice may exhibit increased bortezomib-induced toxicities compared with wild-type animals. We firstconfirmed that bortezomib can impair �-globin turnover ex vivo inTh3/� mouse reticulocyte pulse-chase experiments (supplementalFigure 3). We then conducted a dose-escalation trial in whichwild-type or Th3/� mice were treated with bortezomib at dosessimilar to those used in human patients (Figure 5A).43 The timingof treatment at each dose was 2-3 weeks, which representsapproximately one half-life for wild-type murine RBCs. Examina-tion of circulating RBCs 24 hours after systemic treatment withbortezomib revealed an approximately 25% inhibition of protea-some activity at 0.25 mg/kg and approximately 30%-40% inhibi-tion at 1.0 mg/kg (Figure 5B), which is consistent with previousfindings.44 Lower doses of drug (0.25 and 0.5 mg/kg) producedminimal effects on RBC parameters (supplemental Table 1). At1.0 mg/kg of bortezomib, RBC counts and Hb decreased by similarextents in both wild-type and Th3/� mice (Figure 5C-D andsupplemental Table 2). Similar results were obtained with a 3-weektreatment at 1.0 mg/kg without dose escalation (data not shown).Bortezomib treatment caused an increase in spleen size (Figure 5E)and expansion of early erythroid precursors in the BM and spleens

of both wild-type and Th3/� mice, although the effects were moreprominent in wild-type animals (supplemental Figure 4).31,32

Bortezomib did not alter the turnover of insoluble �-globin in(Th3/�) reticulocytes or increase the levels of insoluble globins inwild-type or Th3/� RBCs (Figure 5F-G). Overall, UPS inhibitionby bortezomib generally impaired erythropoiesis with subtle differ-ences in drug effects on wild-type versus Th3/� mice. However,bortezomib did not cause the accumulation of �-globin precipitatesor enhanced toxicity in the �-thalassemic animals.

Systemic proteasome inhibition activates alternate PQCpathways

The failure of bortezomib to enhance �-globin accumulation in�-thalassemic mice contrasts with our findings that proteasomeinhibition blocks degradation of free �-globin in erythroid culturesand in reticulocytes. In vivo, early-stage erythroid progenitors mayactivate alternate PQC pathways in response to systemic protea-some inhibition. To investigate this, we analyzed purified Ter119�

erythroid precursors from the BM of bortezomib-treated wild-typeand Th3/� mice. Proteasome inhibition activates heat-shock factortranscription factors that induce the expression of heat-shockproteins (HSPs), molecular chaperones that bind misfolded pro-teins to stabilize their structures and prevent precipitation.40 Inwild-type fetal liver erythroblasts, proteasome inhibition inducedmRNAs encoding Hsp105 (Figure 4G) and Hsp90aa1 (not shown).Without bortezomib treatment, numerous HSP mRNAs trendedtoward up-regulation in Th3/� erythroblasts, likely reflectingproteotoxic stress and activation of the heat-shock factor pathway(Figure 6A). However, a significantly greater HSP response

Figure 5. Systemic proteasome inhibition by bortezomib impairsboth thalassemic and normal erythropoiesis in vivo. (A) Scheme forbortezomib dosing in wild-type and Th3/� mice. (B) Proteasome activity inerythrocytes from mice 24 hours after treatment with 1.0 mg/kg ofbortezomib. Activity, normalized to total protein, was measured by fluores-cence release from Suc-LLVY-AMC proteasome substrate and normalizedto control treated mice; n � 3 mice/group. **P � .01 versus control(CTRL). (C) RBC counts of �/� or Th3/� mice treated with vehicle or1.0 mg/kg of bortezomib for 2 weeks; n � 10 mice/group. (D) RBC countsof �/� or Th3/� mice treated with vehicle or 1.0 mg/kg of bortezomib for5 weeks; n � 6 mice/group. *P � .05; **P � .01; ***P � .001. (E) Spleenweight normalized to total body weight in bortezomib-treated and controlmice n � 6 mice/group. *P � .05; ***P � .001. (F) Pulse-chase analysis ofinsoluble �-globin in reticulocytes from control or bortezomib-treatedTh3/� mice. (G) Coomassie-stained insoluble globin aggregates (top)from equal numbers of circulating erythrocytes from wild-type or Th3/�mice treated with vehicle or bortezomib at 1.0 mg/kg. Soluble fractions areincluded as loading controls (bottom).





resulted from systemic treatment with bortezomib in both wild-type and Th3/� erythroid precursors. Therefore, one compensatoryresponse to proteasome inhibition is the induction of HSPs, whichmay alleviate some of the toxic effects of free �-globin in�-thalassemia.

Autophagy is active during normal erythropoiesis17 and hasbeen shown to degrade unstable proteins in numerous diseases.5 Inthe present study, we investigated whether autophagy compensatesfor proteasome inhibition during normal and thalassemic erythro-poiesis. We used Western blotting to examine the autophagosomemarker LC3b in purified Ter119� BM cells (Figure 6B). LC3blevels, including the phosphatidylethanolamine-conjugated (typeII) form indicative of active autophagy, were increased in Th3/�erythroblasts compared with wild-type, similar to what wasreported for human thalassemic cells.23 Systemic bortezomibtreatment increased LC3b in both genotypes. Therefore, the highestlevels of autophagy were observed in Th3/� mice treated with theproteasome inhibitor. We also investigated whether autophagyparticipates in the turnover of �-globin aggregates in �-thalassemia.In pulse-chase studies of Th3/� mouse reticulocytes, ATP deple-tion (which inhibits both the UPS and autophagy) reduced thedegradation of �-globin aggregates to a greater extent thanproteasome inhibition alone (supplemental Figure 5). Inhibition oflysosomal acidification with chloroquine inhibited �-globin degra-dation to a similar extent as proteasome inhibition, whereas bothinhibitors together produced additive effects (Figure 6C-D). Theseresults indicate that in �-thalassemia, excess �-globin is degradedby autophagy, and that this process is enhanced by inhibition of theUPS. Therefore, both HSP and autophagy pathways are induced bybortezomib in wild-type and Th3/� mice. The induction of thesePQC components likely explains why systemic proteasome inhibi-

tion does not enhance the accumulation of precipitated �-globin inTh3/� mice (Figure 7).

Discussion

More than 40 years ago, Fessas, Bank, Nathan, Weatherall, andothers demonstrated that �-thalassemia is caused by globin chainimbalance and that the accumulation of cytotoxic free �-globin is aprimary determinant of disease pathophysiology.45-48 Additionalearly studies, interpreted in light of more recent work, indicatestriking similarities between �-thalassemia and a class of diseasescalled “protein-aggregation disorders” (for review, see Khandrosand Weiss2). Common features include the accumulation of un-stable, misfolded proteins that can be detoxified to some extent bycellular PQC systems, with disease ensuing when protectivemechanisms are overwhelmed. It is likely that lessons learned fromrecent studies of protein-aggregation disorders, including �1 anti-trypsin deficiency,24 cardiomyopathy, myocardial ischemia,25 andsome neurodegenerative diseases,5 can be exploited to betterunderstand and treat �-thalassemia. Conversely, defining how�-globin is detoxified in �-thalassemia may further elucidate PQCmechanisms in nonerythroid protein-aggregation disorders. In thepresent study, we investigated further how �-thalassemia fits intoemerging paradigms of protein-aggregation disorders. We showthat numerous interregulated PQC pathways, including the UPS,autophagy, and HSP responses, are used to detoxify and removefree �-globin in �-thalassemic erythroid cells. Furthermore, wedemonstrate for the first time that the UPS is regulated dynamicallyat the transcriptional level in �-thalassemic erythroblasts through a

Figure 6. Systemic proteasome inhibition activates alternate �-glo-bin detoxification pathways. (A) Real-time RT-PCR quantification ofHSP mRNA expression in Ter119� erythroblasts from wild-type or Th3/�mice treated with vehicle (CTRL) or bortezomib (BOR). Expression isnormalized to �-actin and Hprt mRNA levels. Relative expression betweendifferent experimental groups is shown with vehicle-treated wild-type miceassigned an arbitrary value of 1.0. Data are shown for Hspa1a (i), Hspa1b(ii), Hsp90aa1 (iii), and Hsp105 (iv); n � 4 mice/group. *P � .05; *P � .01;***P � .001. (B) Purified BM Ter119� cells from mice treated with bort-ezomib (BOR, 1.0 mg/kg for 14 days) or vehicle (CTRL) were analyzed byWestern blotting for the autophagosome marker LC3b. Two forms of LC3bare indicated: the unmodified form (I) and the phosphatidylethanolamine-conjugated form (II), which indicates active autophagosomes. �-actinexpression was examined as a loading control. Mean SEM LC3b-IIsignal normalized to �-actin is shown for 4 mice for each group, withwild-type control treated mice set at 1. (C) Reticulocytes from �-thalassemic(Th3/�) mice were labeled with 35S-cysteine and 35S-methionine and chasedwith unlabeled amino acids in the presence or absence of proteasome(MG132, 10�M) or lysosome (CQ, chloroquine, 100�M) inhibitors. Soluble andinsoluble fractions were purified and analyzed for labeled �-globin by tritonacetic acid urea gel electrophoresis, followed by autoradiography. (D) Quantifi-cation of autoradiographs from panel C; n � 3 mice. *P � .001 versus control;#P � .001 versus MG or CQ. Note that the drugs inhibit the loss of insoluble�-globin additively.





Nrf1 stress-response pathway, suggesting a new potential therapeu-tic target for �-thalassemia.

Previous studies have shown that free � chains can be ubiquiti-nated in cell-free systems or in reticulocyte lysates in vitro14 andthat ATP-dependent degradation of � chains occurs in �-thalasse-mic patient reticulocytes.12 In the present study, we used a mousemodel of �-thalassemia to establish a simple system for followingthe fate of excess �-globin in live cells and show that newlysynthesized excess �-globin chains rapidly associate with aninsoluble stromal fraction and are degraded, mirroring earlierfindings in human cells.12 Both proteasomal and lysosomal inhibi-tors alter turnover of insoluble � chains, implicating the UPS andautophagy in this process. Most likely, autophagy mediates directturnover of insoluble �-globin, whereas the UPS degrades soluble�-globin, which accumulates and shifts to the insoluble fractionafter proteasome inhibition in isolated reticulocytes. This interpre-tation of the current data is consistent with electron microscopyanalysis of human �-thalassemic erythroblasts, in which putativeelectron-dense �-globin inclusions were observed being engulfedby and within lysosomes.21,22 In addition, our denaturing immuno-precipitation experiments detected insoluble polyubiquitinated�-globin chains in �-thalassemic erythrocytes, even without protea-some inhibitor treatment. These findings show for the first time thatfree �-globin is polyubiquitinated in vivo and suggest that the formof �-globin recognized by ubiquitin ligases is misfolded and/orunstable. Most likely, a proportion of polyubiquitinated �-globinprecipitates in �-thalassemic erythroblasts because proteasomaldegradation systems are saturated.

Our results indicate that in �-thalassemia, unstable �-globindoes not form static precipitates, but rather exists in dynamicsubcellular fractions that interact with both UPS and autophagypathways. These features raise the possibility that free �-globininteracts physically with the aggresome, a recently describedintracellular structure in which unstable proteins colocalize withPQC machinery, specialized adapter proteins, and the cytoskel-eton.49 Further, electron micrographs of �-thalassemic erythro-blasts demonstrate that some precipitated �-globin is perinuclear,ubiquitin-associated, and pericentriolar, all features of ag-gresomes.21,22,50 It is believed that aggresomes sequester abnormalproteins to minimize cellular damage and provide a staging area forchaperone-mediated refolding, proteasome degradation, or bulkautophagy (macroautophagy) through scaffolding proteins such asSQSTM1/p62 that interact with the autophagosome protein LC3.34,49

Therefore, we speculate that thalassemic erythroblasts degradeprecipitated free �-globin via aggresome-mediated macroautophagy.

Previous work suggests that �-thalassemic erythroblasts exhibitincreased capacity for protein degradation,10,11 but the identity ofthe proteases and the mechanisms of their increased activity are notknown. We have demonstrated herein that within �-thalassemiccells, most proteasome subunits are coordinately up-regulatedaccording to the dosage of free �-globin. Whereas it has beenshown previously that proteasome subunit gene transcriptionincreases in response to pharmacologic proteasome inhibition,37,38

similar effects have not been demonstrated in any disease process,but are predicted to occur in nonerythroid protein-aggregationdisorders according to the present findings. Proteasome subunitgenes contain AREs that are recognized by a family of cap-n-collarbasic leucine zipper (CNC-bZip) transcription factors, includingNrf1 (TCF11) and Nrf2, which activate distinct and overlappingsets of stress-response genes.36-38 The TCF11 isoform of Nrf1up-regulates proteasome subunit genes in cell lines treated withlow doses of proteasome inhibitor. A role for Nrf2 in proteasome

gene regulation has been suggested, but is less well defined.51,52

Mice lacking Nrf1 in neurons exhibit reduced brain proteasomeactivity and neurodegeneration.53 Global deletion of Nrf1 leads toembryonic lethality from anemia, but this defect is not cellautonomous, so the role of Nrf1 in erythroid development, and inproteasome function therein, remains unknown.54 We used drugsand shRNAs to show that Nrf1 mediates erythroid proteasomeup-regulation in �-thalassemia. TCF11, the human isoform of Nrf1known to activate proteasome subunit gene transcription, isnormally constitutively degraded via the proteasome.38 Therefore,proteasome inhibition stabilizes TCF11, which translocates to thenucleus and activates transcription via AREs. Mechanistically, thislikely occurs through proteotoxic stress and proteasome “clogging”by aggregated � subunits in �-thalassemia. Pharmacologic up-regulation of Nrf1 activity may be a potential route to furtherincrease proteasome activity in thalassemic mice and patients infuture studies.

Systemic bortezomib impaired erythropoiesis in a dose-dependent fashion, which is consistent with in vivo studies ofproteasome inhibition42 and reports of anemia as a drug toxicity inpatients.55 Contrary to our initial expectations, bortezomib did notproduce a disproportionately adverse effect or enhance �-globinaccumulation in �-thalassemic mice. This may be explained by ourobservations that multiple erythroid PQC pathways are activatedadditively by �-thalassemia and systemic protease inhibition.These induced pathways include an HSP response, which producesmolecular chaperone proteins that bind denatured “client” proteinsto either promote their refolding or target them for degradation (forreview, see Weiss and Dos Santos56). In addition, thalassemia andbortezomib induced the autophagosome marker LC3b additively,suggesting increased autophagic flux. Moreover, free �-globin isdegraded by autophagy in thalassemic reticulocytes (Figure 6). Ourfindings are consistent with the concept that PQC pathways areinterrelated and interactive. Proteasome inhibition induces bothHSPs40 and macroautophagy57,58 by causing the accumulation of

Figure 7. Model of �-globin detoxification pathways in �-thalassemia. Excessfree �-globin is unstable and cytotoxic to RBC precursors and mature RBCs.Unstable �-globin is polyubiquitinated and degraded via the proteasome. If ubiquitin-proteasome activity is insufficient, �-globin forms relatively insoluble aggregates thatserve as a substrate for macroautophagy. Chaperones may be involved in refolding�-globin or in targeting excess �-globin for degradation. Molecular cross-talk existsbetween these pathways; for example, inhibition of proteasome activity results in theaccumulation of unfolded proteins, activation of stress pathways, and consequentinduction of autophagy and heat-shock/molecular chaperone responses.





denatured proteins, which activate various cellular unfolded pro-tein responses. These compensatory pathways are more likely to beinduced in vivo in early-stage erythroid precursors during therelatively long time course of systemic bortezomib administrationcompared with brief ex vivo protease inhibitor treatment of isolatedlate-stage fetal liver erythroblasts or transcriptionally inert reticulo-cytes. In the future, it will be of interest to investigate how in vivosuppression of autophagy affects �-thalassemia. The use of lyso-somal inhibitors such as chloroquine is complicated by thepotential for autophagy-independent effects, such as interferencewith iron metabolism, which would impair erythropoiesis. Mostlikely, erythroid-specific suppression of autophagy pathways viagenetic manipulation in mice will be the best initial approach toinvestigate whether �-thalassemia is sensitive to the loss of thispathway with or without proteasome inhibition.

The results of the present study indicate that �-thalassemicerythroid cells compensate for �-globin excess through multiple,functionally interconnected PQC pathways, and that decreased fluxthrough a single pathway can be compensated for by enhancedactivity of the others. Up to a certain threshold level, combinedPQC pathways can detoxify most excess �-globin to minimizeclinical phenotypes. Therefore, �-thalassemia, and potentiallyother hemoglobinopathies, can be studied using the same tools andmodels currently being applied to a broad range of protein-aggregation disorders. Given the high degree of clinical variabilityin �-thalassemic patients, some of which cannot be explained by�-globin gene mutations or compensatory -globin expression, itmay be informative to examine genetic variation of PQC pathwaysin these patients. Moreover, improved understanding of howerythroblasts modulate PQC components in response to the accumu-lation of free �-globin may illustrate new pharmacologic andgenetic means with which to up-regulate the activity of these

pathways for treating �-thalassemia and other protein-aggregationdisorders affecting nonerythroid tissues.

Acknowledgments

The authors thank Don Baldwin at the University of Pennsylvaniamicroarray core facility and Zhe Zhang at the Children’s Hospitalof Philadelphia Bioinformatics Core facility for assistance withdesign and analysis of the microarray experiments, and GerdBlobel, David Nathan, and Vijay Sankaran for thoughtful com-ments on the manuscript.

This work was supported by the National Institutes of Health(grants DK061692, HL087427, and P30DK090969 to M.J.W.).E.K. was a trainee under a National Heart, Lung, and BloodInstitute Medical Scientist Training Program grant(3T32GM007170-35S1). The DiGaetano family also providedgenerous support.

Authorship

Contribution: E.K. designed and conducted the experiments,analyzed the data, and wrote the manuscript; C.S.T. designedand conducted the experiments and analyzed the data; J.D.conducted the experiments and analyzed the data; and M.J.W.designed the experiments, analyzed the data, and wrote themanuscript.

Conflict-of-interest disclosure: The authors declare no com-peting financial interests.

Correspondence: Mitchell J. Weiss, Division of Hematology,Children’s Hospital of Philadelphia, 316B Abramson ResearchCenter, Philadelphia, PA 19104; e-mail: [email protected].

References

1. Steinberg MH. Disorders of Hemoglobin: Genet-ics, Pathophysiology, and Clinical Management.2nd Ed. Cambridge, United Kingdom: CambridgeUniversity Press; 2009.

2. Khandros E, Weiss MJ. Protein quality controlduring erythropoiesis and hemoglobin synthesis.Hematol Oncol Clin North Am. 2010;24(6):1071-1088.

3. Ciechanover A, Brundin P. The ubiquitin protea-some system in neurodegenerative diseases:sometimes the chicken, sometimes the egg. Neu-ron. 2003;40(2):427-446.

4. Ding W-X, Yin X-M. Sorting, recognition and acti-vation of the misfolded protein degradation path-ways through macroautophagy and the protea-some. Autophagy. 2008;4(2):141-150.

5. Jaeger PA, Wyss-Coray T. All-you-can-eat: au-tophagy in neurodegeneration and neuroprotec-tion. Mol Neurodegener. 2009;4:16.

6. Wood WG, Stamatoyannopoulos G. Globin syn-thesis in fractionated normoblasts of beta-thalas-semia heterozygotes. J Clin Invest. 1975;55(3):567-578.

7. Bank A, O’Donnell JV. Intracellular loss of freealpha chains in beta thalassemia. Nature. 1969;222(5190):295-296.

8. Clegg JB, Weatherall DJ. Haemoglobin synthesisduring erythroid maturation in beta-thalassaemia.Nat New Biol. 1972;240(101):190-192.

9. Kan YW, Nathan DG, Lodish HF. Equal synthesisof alpha- and beta-globin chains in erythroid pre-cursors in heterozygous beta-thalassemia. J ClinInvest. 1972;51(7):1906-1909.

10. Braverman AS, Lester D. Evidence for increased

proteolysis in intact beta thalassemia erythroidcells. Hemoglobin. 1981;5(6):549-564.

11. Loukopoulos D, Karoulias A, Fessas P. Proteoly-sis in thalassemia: studies with protease inhibi-tors. Ann N Y Acad Sci. 1980;344:323-335.

12. Shaeffer JR. Turnover of excess hemoglobin al-pha chains in beta-thalassemic cells is ATP-de-pendent. J Biol Chem. 1983;258(21):13172-13177.

13. Vettore L, De Matteis MC, Di Iorio EE,Winterhalter KH. Erythrocytic proteases: prefer-ential degradation of alpha hemoglobin chains.Acta Haematol. 1983;70(1):35-42.

14. Shaeffer JR. ATP-dependent proteolysis of hemo-globin alpha chains in beta-thalassemic hemoly-sates is ubiquitin-dependent. J Biol Chem. 1988;263(27):13663-13669.

15. Ciehanover A, Hod Y, Hershko A. A heat-stablepolypeptide component of an ATP-dependentproteolytic system from reticulocytes. BiochemBiophys Res Commun. 1978;81(4):1100-1105.

16. Mizushima N, Yoshimori T, Levine B. Methods inmammalian autophagy research. Cell. 2010;140(3):313-326.

17. Kang YA, Sanalkumar R, O’Geen H, et al. Au-tophagy driven by a master regulator of hemato-poiesis. Mol Cell Biol. 2012;32(1):226-239.

18. Schweers RL, Zhang J, Randall MS, et al. NIX isrequired for programmed mitochondrial clearanceduring reticulocyte maturation. Proc Natl Acad SciU S A. 2007;104(49):19500-19505.

19. Zhang J, Ney PA. Autophagy-dependent and -in-dependent mechanisms of mitochondrial clear-ance during reticulocyte maturation. Autophagy.2009;5(7):1064-1065.

20. Sandoval H, Thiagarajan P, Dasgupta SK, et al.Essential role for Nix in autophagic maturation oferythroid cells. Nature. 2008;454(7201):232-235.

21. Wickramasinghe SN, Bush V. Observations onthe ultrastructure of erythropoietic cells and re-ticulum cells in the bone marrow of patients withhomozygous beta-thalassaemia. Br J Haematol.1975;30(4):395-399.

22. Wickramasinghe SN, Hughes M. Ultrastructuralstudies of erythropoiesis in beta-thalassaemiatrait. Br J Haematol. 1980;46(3):401-407.

23. Lithanatudom P, Wannatung T, Leecharoenkiat A,Svasti S, Fucharoen S, Smith DR. Enhanced acti-vation of autophagy in beta-thalassemia/Hb Eerythroblasts during erythropoiesis. Ann Hematol.2011;90(7):747-758.

24. Hidvegi T, Ewing M, Hale P, et al. An autophagy-enhancing drug promotes degradation of mutantalpha1-antitrypsin Z and reduces hepatic fibrosis.Science. 2010;329(5988):229-232.

25. Li J, Horak KM, Su H, Sanbe A, Robbins J,Wang X. Enhancement of proteasomal functionprotects against cardiac proteinopathy and isch-emia/reperfusion injury in mice. J Clin Invest.2011;121(9):3689-3700.

26. Yang B, Kirby S, Lewis J, Detloff PJ, Maeda N,Smithies O. A mouse model for beta 0-thalasse-mia. Proc Natl Acad Sci U S A. 1995;92(25):11608-11612.

27. Yu X, Kong Y, Dore LC, et al. An erythroid chaper-one that facilitates folding of alpha-globin sub-units for hemoglobin synthesis. J Clin Invest.2007;117(7):1856-1865.





28. England SJ, McGrath KE, Frame JM, Palis J. Im-mature erythroblasts with extensive ex vivo self-renewal capacity emerge from the early mamma-lian fetus. Blood. 2011;117(9):2708-2717.

29. Verdoes M, Florea BI, Menendez-Benito V, et al.A fluorescent broad-spectrum proteasome inhibi-tor for labeling proteasomes in vitro and in vivo.Chem Biol. 2006;13(11):1217-1226.

30. Kong Y, Zhou S, Kihm AJ, et al. Loss of alpha-hemoglobin-stabilizing protein impairs erythropoi-esis and exacerbates beta-thalassemia. J ClinInvest. 2004;114(10):1457-1466.

31. Gardenghi S, Ramos P, Marongiu MF, et al. Hep-cidin as a therapeutic tool to limit iron overloadand improve anemia in beta-thalassemic mice.J Clin Invest. 2010;120(12):4466-4477.

32. Liu Y, Pop R, Sadegh C, Brugnara C, Haase VH,Socolovsky M. Suppression of Fas-FasL coex-pression by erythropoietin mediates erythroblastexpansion during the erythropoietic stress re-sponse in vivo. Blood. 2006;108(1):123-133.

33. Gill FM, Schwartz E. Free alpha-globin pool inhuman bone marrow. J Clin Invest. 1973;52(12):3057-3063.

34. Bjørkøy G, Lamark T, Brech A, et al. p62/SQSTM1 forms protein aggregates degraded byautophagy and has a protective effect on hunting-tin-induced cell death. J Cell Biol. 2005;171(4):603-614.

35. Subramanian A, Tamayo P, Mootha VK, et al.Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wideexpression profiles. Proc Natl Acad Sci U S A.2005;102(43):15545-15550.

36. Ohtsuji M, Katsuoka F, Kobayashi A, Aburatani H,Hayes JD, Yamamoto M. Nrf1 and Nrf2 play dis-tinct roles in activation of antioxidant responseelement-dependent genes. J Biol Chem. 2008;283(48):33554-33562.

37. Radhakrishnan SK, Lee CS, Young P, Beskow A,Chan JY, Deshaies RJ. Transcription factor Nrf1mediates the proteasome recovery pathway after

proteasome inhibition in mammalian cells. MolCell. 2010;38(1):17-28.

38. Steffen J, Seeger M, Koch A, Kruger E. Protea-somal degradation is transcriptionally controlledby TCF11 via an ERAD-dependent feedbackloop. Mol Cell. 2010;40(1):147-158.

39. Zhang Y, Talalay P, Cho CG, Posner GH. A majorinducer of anticarcinogenic protective enzymesfrom broccoli: isolation and elucidation of struc-ture. Proc Natl Acad Sci U S A. 1992;89(6):2399-2403.

40. Kim HJ, Joo HJ, Kim YH, et al. Systemic analysisof heat shock response induced by heat shockand a proteasome inhibitor MG132. PLoS One.2011;6(6):e20252.

41. Testa U. Proteasome inhibitors in cancer therapy.Curr Drug Targets. 2009;10(10):968-981.

42. Chen CY, Pajak L, Tamburlin J, Bofinger D,Koury ST. The effect of proteasome inhibitors onmammalian erythroid terminal differentiation. ExpHematol. 2002;30(7):634-639.

43. Kane RC, Bross PF, Farrell AT, Pazdur R. Vel-cade: U.S. FDA approval for the treatment of mul-tiple myeloma progressing on prior therapy. On-cologist. 2003;8(6):508-513.

44. Luker GD, Pica CM, Song J, Luker KE,Piwnica-Worms D. Imaging 26S proteasome ac-tivity and inhibition in living mice. Nat Med. 2003;9(7):969-973.

45. Fessas P. Inclusions of hemoglobin erythroblastsand erythrocytes of thalassemia. Blood. 1963;21:21-32.

46. Weatherall DJ, Clegg JB, Naughton MA. Globinsynthesis in thalassaemia: an in vitro study. Na-ture. 1965;208(5015):1061-1065.

47. Nathan DG, Gunn RB. Thalassemia: the conse-quences of unbalanced hemoglobin synthesis.Am J Med. 1966;41(5):815-830.

48. Bank A. Hemoglobin synthesis in beta-thalasse-mia: the properties of the free alpha-chains. J ClinInvest. 1968;47(4):860-866.

49. Kopito RR. Aggresomes, inclusion bodies andprotein aggregation. Trends Cell Biol. 2000;10(12):524-530.

50. Wickramasinghe SN, Lee MJ. Evidence that theubiquitin proteolytic pathway is involved in thedegradation of precipitated globin chains in thala-ssaemia. Br J Haematol. 1998;101(2):245-250.

51. Kapeta S, Chondrogianni N, Gonos ES. Nuclearerythroid factor 2-mediated proteasome activa-tion delays senescence in human fibroblasts.J Biol Chem. 2010;285(11):8171-8184.

52. Kwak M-K, Wakabayashi N, Greenlaw JL,Yamamoto M, Kensler TW. Antioxidants enhancemammalian proteasome expression through theKeap1-Nrf2 signaling pathway. Mol Cell Biol.2003;23(23):8786-8794.

53. Lee CS, Lee C, Hu T, et al. Loss of nuclear factorE2-related factor 1 in the brain leads to dysregu-lation of proteasome gene expression and neuro-degeneration. Proc Natl Acad Sci U S A. 2011;108(20):8408-8413.

54. Chan JY, Kwong M, Lu R, et al. Targeted dis-ruption of the ubiquitous CNC-bZIP transcrip-tion factor, Nrf-1, results in anemia and embry-onic lethality in mice. EMBO J. 1998;17(6):1779-1787.

55. Dy GK, Thomas JP, Wilding G, et al. A phase Iand pharmacologic trial of two schedules of theproteasome inhibitor, PS-341 (bortezomib, vel-cade), in patients with advanced cancer. ClinCancer Res. 2005;11(9):3410-3416.

56. Weiss MJ, Dos Santos CO. Chaperoning erythro-poiesis. Blood. 2009;113(10):2136-2144.

57. Wu WK, Sakamoto KM, Milani M, et al. Macroau-tophagy modulates cellular response to protea-some inhibitors in cancer therapy. Drug ResistUpdat. 2010;13(3):87-92.

58. Zhu K, Dunner K Jr, McConkey DJ. Proteasomeinhibitors activate autophagy as a cytoprotectiveresponse in human prostate cancer cells. Onco-gene. 2010;29(3):451-462.





online March 16, 2012 originally publisheddoi:10.1182/blood-2011-12-397729

2012 119: 5265-5275

Eugene Khandros, Christopher S. Thom, Janine D'Souza and Mitchell J. Weiss

-thalassemiaβmurine -globin inαIntegrated protein quality-control pathways regulate free

http://www.bloodjournal.org/content/119/22/5265.full.htmlUpdated information and services can be found at:

(865 articles)Red Cells, Iron, and Erythropoiesis Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

Copyright 2011 by The American Society of Hematology; all rights reserved.of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society


http://www.bloodjournal.org/content/119/22/5265.full.html

http://www.bloodjournal.org/cgi/collection/iron_red_cells_erythropoiesis

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests

http://www.bloodjournal.org/site/misc/rights.xhtml#reprints

http://www.bloodjournal.org/site/subscriptions/index.xhtml





integrated protein quality-control pathways regulate free -globin in

Documents