Introduction
In β-thalassemias, mutations of the β-globin gene or its regulatory regions cause absence (β
0 phenotype) or reduced synthesis (β
+ phenotype) of β-globin chains [
1‐
4], and this impairment leads to an excess of the complementary α-globin chains [
1]. Ultimately, the precipitation of α-globin chains in excess promotes apoptosis of erythroid precursors in the bone marrow and at extra-medullary sites and shortens survival of red blood cells (BRCs) in the peripheral blood [
5‐
10]. The disease is associated with morbidity and mortality due to severe chronic anemia or treatment-related complications [
1].
Gene therapy is one of the possible approaches for the cure of β-thalassemia, following β-globin gene transfer in autologous hematopoietic stem cells (HSCs) [
11‐
13]. Retroviral- or lentiviral-mediated insertion of single and multiple copies of the β-globin gene in ErPCs has been reported in many studies to demonstrate the feasibility of the gene therapy approach for the cure of β-thalassemia [
11‐
17]. This approach, while straightforward in its principle, exhibits several critical issues, the major being the control of transgene expression, which needs to be: (a) erythroid-specific, (b) differentiation- and stage-restricted, (c) elevated, (d) position-independent, (f) sustained over time and (g) independent from the patient's genotype [
18‐
21]. Moreover, the gene transfer, in order to be effective, needs to target the majority of HSCs, limiting the vector copy number to minimize the possibility of insertional mutagenesis. The study of Hargrove et al. [
22] showed how lentiviral vector insertion can perturb the expression of endogenous genes in β-thalassemic hematopoietic cells [
22]. Despite this evidence, clinical trials based on gene therapy on β-thalassemic patients have been initiated, although this therapeutic intervention was used on a restricted number of patients.
In the most recent study, a 19-year-old patient who received gene therapy treatment no longer needs palliative transfusions, and 2 years later, seems healthy. Still, the unexpected observation of the expansion of a group of clonal cells with the same gene insertion (the
HMGA2 gene) suggests that a lentiviral vector can promote growth advantage of selected cells [
19]. The pattern is reminiscent of the gene therapy trial for the X-linked combined immunodeficiency disease (SCID) in which a retroviral vector triggered leukemia [
23]. Therefore, the potential of gene therapy is limited by the number of viral particles that can be safely incorporated into the genome. In fact, the integration of fewer molecules is necessary to avoid genotoxicity, and at the same time, a minimum number of integration is required to sustain elevated transgene expression [
15].
Our laboratory has been investigating the effects of several compounds on inducing the expression of γ-globin genes and the increase of fetal hemoglobin (HbF) synthesis [
24‐
28]. It is generally believed that even small increases of HbF synthesis could be beneficial to β-thalassemia patients [
29‐
31]. Although ErPCs from different patients might respond to a different extent to the same HbF inducer, our experience indicates that this effect is reproducible. The limitation of such approach is to reach levels of HbF clinically relevant. However, clinical trials with HbF inducers have been extensively investigated, using hydroxyurea (HU), thalidomide, and butyrates [
32‐
36]. Following this research field, several studies focusing on the mechanisms regulating reactivation of HbF production in humans have been reported [
27,
35,
37‐
39].
As for possible co-expression of γ-globin and β-globin genes, it should be considered that during ontogeny, two switches occur in β-like globin genes expression that reflect the changing oxygen requirements of the fetus, the second of which, from γ- to adult δ- and β-globin, occurs shortly after birth. Throughout the locus,
cis-acting elements are dynamically bound by
trans-acting proteins, including transcription factors, co-activators, repressors, and chromatin modifiers [
40‐
42]. Despite the complex hierarchy of events regulating gene expression during development, from extracellular signaling to transcriptional activation or repression, the expression of β-globin and γ-globin genes appears to be inversely regulated (i.e., high expression of γ-globin genes versus low expression of β-globin genes and vice versa) [
40,
41].
To our knowledge, no attempt has been made to verify whether induction of HbF and HbA might be obtained using cells isolated from homozygous patients unable to produce β-globin mRNA. This combined approach might have a synergistic effect, by adding HbA synthesis to high levels of HbF, which ameliorate the clinical parameters of β-thalassemia patients [
28‐
31]. On the other hand, at least in theory, β-globin mRNA production following gene therapy approaches might interfere with γ-globin gene expression. The aim of this paper was to verify whether the HbF inducer mithramycin (MTH) stimulates the production of HbF in erythroid cells treated with a lentiviral vector carrying a therapeutic β-globin gene. The gene transfer vector utilized in this study (T9W) [
17], is a third-generation lentivirus that has been obtained by modifying TNS9, a lentiviral vector of second generation. TNS9 was successfully utilized in mice to cure and rescue thalassemic mice affected by thalassemia intermedia and major, respectively [
11,
14].
We analyzed the effect of MTH on K562 cell clones carrying the enhanced green fluorescent protein (eGFP) and the red fluorescent protein (RFP) driven by the γ-globin and β-globin promoter, respectively [
43,
44]. We analyzed the effect of MTH on several transgenic K562 cell lines harboring different copies of the human β-globin gene to verify the possible co-expression of β-globin and γ-globin mRNAs, and the possible production of HbA and HbF under stimulation with MTH. K562 cells appear to be particularly useful in this context, since wild-type K562 cells do not express the endogenous β-globin gene, being exclusively committed to the expression of embryo-fetal globin genes [
47,
48]. We then treated erythroid precursor cells (ErPCs) [
49,
50] from β-thalassemic individuals with both MTH and the therapeutic T9W vector and analyzed globin mRNA expression by RT-PCR and production of HbF and HbA by HPLC.
Materials and methods
Lentiviral vector and chemical inducers
The T9W vector was generated by modifying TNS9, with the aim of increasing its safety and efficiency. For this purpose, the 3′ long terminal repeat (3′ LTR) was disarmed by deleting the U3 region (self inactivating-LTR or SIN-LTR) [
17]. The deletion, which includes the TATA box and all the major determinants responsible for regulating the HIV-1 promoter, abolished the LTR promoter activity, but did not affect vector titers or transgene expression in vitro (data not shown). The
cis-acting woodchuck post-regulatory element (WPRE) was also introduced [
17]. Mithramycin (MTH) was purchased from Sigma (St.Louis, MO, USA) [
23].
Vector production and titration
Viral stocks were generated by co-transfection of the gene transfer plasmid T9W together with the envelope plasmid (VSV-G), the packaging plasmid (pMDLgpRRE), and the pRSV-REV plasmid into 293 T cells. An aliquot (5 x 106) of 293 T cells was seeded into cell culture dishes (10 cm) 24 h prior to transfection in Iscove's modification of Eagle's medium (I-MEM, CAMBREX-Biowhittaker, Europe) with 10% fetal bovine serum (FBS, Biowest, Nuaillé, France) 100 U/ml penicillin, and 100 mg/ml streptomycin (Pen-Strep, CAMBREX-Biowhittaker, Europe), at 37°C under 5% CO2. The culture medium was changed 2 h prior to transfection. The precipitate was formed by adding the plasmids to a volume of 450 μl of 0.1× TE (0.1× TE is 10 mM Tris plus 1 mM EDTA) and 50 μl of 2 M CaCl2, then adding 500 μl of 2× HEPES-buffered saline (281 mM NaCl, 100 mM HEPES, 1.5 mM Na2HPO4) drop-wise, and then the precipitate was mixed and immediately added to the cultures. The medium (10 ml) was replaced after 16 h. Viral supernatant was collected at 24 and 48 h, cleared by low speed centrifugation, and filtered through cellulose acetate filters (0.2 μm). Following concentration by ultracentrifugation, serial dilution of concentrated virus (5; 0.5; and 0.05 μl, respectively) were used to infect 1 × 105 NIH 3T3 cells in 1 ml of transfection buffer complemented with polybrene (Chemicon International, Millipore, Billerica, MA, USA) at a final concentration of 8 mg/ml. Genomic DNA was extracted after 3 days (Quiagen kit, Hilden, Germany). The multiplicity of infection (MOI) was calculated using the following formula: number of cells (1 × 105) × dilution factor × VCN, measured via real-time PCR, using oligos for WPRE and ID genes (see below).
Generation of K562 cell clones transduced with the pCCL.Promβ.HcRed1.Promγ.EGFP lentiviral vector
For determining the activity of the γ-globin and β-globin promoters under different treatment conditions, we modified the pCCL.PGK.GFP.WPRE construct, in which a constitutive expression of the eGFP gene is driven by the human phosphoglycerokinase gene promoter [
43]. The PGK sequence was replaced by the γ-globin promoter to drive the expression of the eGFP gene. Additionally, we cloned into this construct the red fluorescent protein (RFP) gene together with the regulatory LCR and β-globin promoter elements. Human K562 cells [
47] were used to obtain stable transfectants. In this system, increase of green eGFP signal will be consistent with a γ-globin gene promoter driven activity; on the contrary, increase of the far-red FP signal will be associated with β-globin promoter activity [
43,
44]. For determining the promoter activity, cells were seeded at 12500 cells/ml and treated with MTH. After 5 days of culture, cells were assayed for fluorescent protein expression. For the determination of fluorescence intensity using the FACScan™ Flow Cytometer (Becton Dickinson, Franklin Lakes, NJ, USA), cells were harvested and washed; then, 10,000 cells were analyzed using the fl1 and the fl3 channels to detect green and red fluorescence, respectively, and analyses were carried out by using the Cell Quest (Becton Dickinson) software.
Human K562 cell clones carrying the human β-globin gene
For the generation of stable K562 clones integrating human β-globin gene, the pCCL.β-globin.PGW vector was used [
43,
44]. Transduction was carried out by plating 10
6 K562 cells in 9.5-cm
2 dishes with 45% RPMI and 45% I-MDM (Iscove's Modified Dulbecco's Medium, CAMBREX—Biowhittaker Europe), 10% FBS, 2 mM
l-glutamine (CAMBREX—Biowhittaker Europe, Milan, Italy), 100 U/ml penicillin, and 100 mg/ml streptomycin in humified atmosphere of 5% CO
2/air and adding the decided volume of the viral supernatant. In order to facilitate cell infection, 10 μl of the 800 μg/μl transduction agent polybrene (Chemicon International, Millipore, Billerica, MA, USA) was added to the K562 cells plated, which were subsequently cultured in a 5% CO
2 incubator. After 7 days, cells were cloned by limiting dilutions and GFP-producing clones identified under a fluorescence microscope and further characterized. Cell cultures were maintained in humified atmosphere of 5% CO
2/air in RPMI 1640 medium (SIGMA, St Louis, MO, USA) supplemented with 10% fetal bovine serum, 50 U/ml penicillin, and 50 mg/ml streptomycin. Cell growth was studied by with a ZF Coulter Counter (Coulter Electronics, Hialeah, FL, USA).
In vitro culture of erythroid progenitors from β-thalassemia patients
Blood samples of patients were collected following receiving informed consent. The two-phase liquid culture procedure was employed as previously described [
49,
50]. Mononuclear cells were isolated from peripheral blood samples by Ficoll–Hypaque density gradient centrifugation and seeded in α-minimal essential medium (α-MEM, SIGMA) supplemented with 10% FBS (Celbio, Milano, Italy), 1 μg/ml cyclosporine A (Sandoz, Basel, Switzerland), and 10% conditioned medium from the 5637 bladder carcinoma cell line. The cultures were incubated at 37°C, under an atmosphere of 5% CO
2 in air, with extra humidity. After 7 days incubation in this phase I culture, the non-adherent cells were harvested, washed, and then cultured in fresh medium composed of α-MEM (SIGMA), 30% FBS (Celbio), 1% deionized bovine serum albumin (BSA, SIGMA), 10
-5 M β-mercaptoethanol (SIGMA), 2 mM
l-glutamine (SIGMA), 10
−6 M dexamethasone (SIGMA), and 1 U/ml human recombinant erythropoietin (EPO) (Tebu-bio, Magenta, Milano, Italy) and stem cell factor (SCF, BioSource International, Camarillo, CA, USA). This part of the culture is referred to as phase II [
47]. Erythroid differentiation was determined by counting benzidine positive cells after suspending the cells in a solution containing 0.2% benzidine in 0.5 M glacial acetic acid, 10% H
2O
2 [
49]. Treatment with MTH was carried out by adding the appropriate drug concentrations at the beginning of the experiment (cells were usually seeded at 10
6 cells/ml). For analysis of hemoglobins, cells were harvested, washed once with phosphate-buffered saline (PBS), and the pellets were processed in lysis buffer (0.01% sodium dodecyl sulphate). After spinning for 1 min in a microcentrifuge, the supernatant was collected and stored at 4°C.
Transduction of erythroid precursors (ErPC) from β°-thalassemia patients with a lentiviral vector carrying the human β-globin gene
Mock control cells were compared to samples treated with MTH or T9W, separately. Moreover, an aliquot of the cells transduced with T9W were also treated with MTH. ErPCs were infected with serial dilutions of the virus, starting from multiplicity of infection (MOI) equal to 0.3 and multiples of it. ErPCs were routinely infected at the beginning of phase 2, when erythropoietin was administered to the cells to promote their erythroid commitment.
RNA Isolation and RT-qPCR analysis
K562 clones and erythroid precursor cells were collected by centrifugation at 1,200 rpm for 5 min at 4°C, washed in PBS, lysed in 1 ml of TRIZOL® Reagent (GIBCO-Invitrogen-Life Technologies), according to the manufacturer's instructions. The isolated RNA was washed once with cold 75% ethanol, dried, and dissolved in diethylpyrocarbonate treated water before use. For gene expression analysis, 1 μg of total RNA was reverse transcribed by using random hexamers. Quantitative real-time PCR assay was carried out using gene-specific double fluorescently labeled probes in a 7700 Sequence Detection System version 1.7 (Applied Biosystems, Warrington Cheshire, UK) as described elsewhere [
23,
24,
46]. The nucleotide sequences used for real-time PCR analysis are reported in Table
1. For real-time PCR analysis, we used as reference gene the endogenous control human GAPDH kit (Applied Biosystems). The fluorescent reporter and the quencher of the GAPDH probe were VIC and 6-carboxy-
N,
N,
N′,
N′-tetramethylrhodamine (TAMRA), respectively.
Table 1
Primers and probes for RT-qPCR
α-Globin |
Forward primer | 5′-CACGCGCACAAGCTTCG-3′ |
Reverse primer | 5′-AGGGTCACCAGCAGGCAGT-3′ |
Probe | 5′-FAM-TGGACCCGGTCAACTTCAAGCTCCT-TAMRA-3′ |
β-Globin |
Forward primer | 5′-CAAGAAAGTGCTCGGTGCCT-3′ |
Reverse primer | 5′-GCAAAGGTGCCCTTGAGGT-3′ |
Probe | 5′-FAM-TAGTGATGGCCTGGCTCACCTGGA-TAMRA-3′ |
γ-Globin |
Forward primer | 5′-TGGCAAGAAGGTGCTGACTTC-3′ |
Reverse primer | 5′-TCACTCAGCTGGGCAAAGC-3′ |
Probe | 5′-FAM-TGGGAGATGCCATAAAGCACCTGC-TAMRA-3′ |
Human erythroid precursor cells were harvested, washed once with PBS, and the pellets were lysed in lysis buffer (0.01% sodium dodecyl sulphate). After incubation on ice for 15 min, and centrifugation for 5 min at 14,000 rpm in a microcentrifuge, the supernatant was separated from the membrane debris and injected. Hb proteins present in the lysates were separated by cation-exchange HPLC [
24], using a Beckman Coulter instrument System Gold 126 Solvent Module-166 Detector. Hemoglobins were separated using a Syncropak CCM 103/25 (250 × 4.6 mm) column, samples were eluted in a solvent gradient utilizing aqueous sodium acetate–BisTris–KCN buffers and detection was performed at 415 nm. The standard controls were the purified HbA (SIGMA, St Louis, MO, USA) and HbF (Alpha Wassermann, Milano, Italy) [
23].
Statistical analysis
The statistical significance of difference between treatments was analyzed, when appropriate, using one-way analysis of variance (ANOVA) and the Student–Newman Keul's test. p values lower than 0.01 were considered statistically significant.
Discussion
The aim of the present investigation was to verify whether a co-treatment of ErPCs from β-thalassemia patients with lentiviral-mediated gene transfer and with inducers of fetal hemoglobin could be effective to stimulate HbF and HbA production and abolish the excess of α-globin accumulation in erythroid cells. To address this question, we employed the T9W lentiviral vector for the infection and mithramycin as HbF inducer. We previously reported the efficacy of these tools for gene therapy purposes [
17] or HbF induction [
23]. In this study, all thalassemic specimens were obtained from homozygous β
039-thalassemia patients.
The first set of results demonstrated that the T9W lentiviral vector was able to induce β-globin gene expression and protein synthesis in ErPCs from homozygous β
039-thalassemia patients, confirming observations previously reported [
17]. As depicted in Fig.
3, this result was consistently reproduced. However, after infection, all samples continue to synthesize some, although reduced, free α-globin chains, indicating that the excess of α-globin content was not completely abolished.
The major finding of this manuscript is the formal demonstration that forced expression of β-globin and γ-globin genes, respectively, by gene transfer and HbF induction, profoundly improves hemoglobin synthesis among β-thalassemic ErPC cells. This is achieved by reducing or eliminating free α-globin chain aggregates, indicating that this approach can correct clinically relevant parameters in treated cells. In this respect, the restoration of a balance between α-globin and β-like globin chains (here γ-globins and β-globins) is associated with clear amelioration of the phenotype of thalassemic cells [
52‐
54]. In this paper, we focused on the effects of co-treatment with MTH and the T9W lentiviral vector on cells derived from homozygous β
039-thalassemia patients, and similar findings were observed in preliminary experiments performed using ErPCs from two β
+39-thalassemia patients carrying a β
039/β
+IVSI-110 genotype (Zuccato et al., unpublished observations).
We believe that these data emphasize the clinical relevance of combining gene therapy with HbF induction for the cure of β-thalassemia. An elevated expression of fetal hemoglobin is beneficial to patients affected by thalassemia intermedia [
28‐
34]. Several experiments conducted in animal models [
55] as well as in patients treated with HbF inducers, support the use of HbF induction [
28‐
31,
33]. Recently, Ehsani et al. [
33] showed that a 6-month treatment of 16 transfusion-independent thalassemia intermedia patients with a 20 mg/kg/day dose of HU 4 days per week produced a significant increase of HbF resulting in the amelioration of hematological parameters. While a larger sample size study is needed to validate our data, these results appear to be well in agreement with independent reports showing a dramatic response of several β-thalassemia patients to HU-mediated induction of HbF [
28,
32‐
34]. Relevant to the issues covered in this paper, Musallam et al. [
56] analyzed the association between HbF levels and morbidity in β-thalassemia intermedia on a cohort of 63 untransfused patients who had also never received HbF induction therapy. There was a strong negative correlation between the HbF level and the total number of morbidities [
56].
Our results suggest that the combination of gene therapy with HbF induction (GT/HbF strategy) might be very useful to eliminate the excess of α-globins detected by the HPLC analyses of ErPCs from β-thalassemia patients. This appears to be a major goal in therapeutic intervention on β-thalassemic erythroid cells and, if reached, is expected to ameliorate the physiological parameter of treated cells.
In conclusion, our data suggest that the GT/HbF strategy, employing the co-treatment of target erythroid precursor cells with a lentiviral vector carrying a therapeutic β-globin gene and the HbF inducer mithramycin, leads to forced de novo accumulation of HbA and increased production of HbF, ultimately suppressing the excess of free α-globin chains. These results might be relevant for establishing a protocol maximizing the production of clinically therapeutic hemoglobins in thalassemic ErPCs.
In addition, our findings strongly support the need of further studies employing co-treatment with gene-therapy lentiviral vectors and other fetal hemoglobin inducers (including DNA-based HbF inducers). These studies are crucial since mithramycin is a chemotherapeutic agent which might cause toxicity if used in life-long treatments. However, less toxic mithramycin analogs have been recently described which exhibit better pharmacokinetics and tolerance [
57]. Finally, the results here presented support the use of vectors carrying the β-globin gene together with sequences enabling the production of HbF. Recently, Wilber et al. [
58] showed that a lentiviral vector encoding a short-hairpin RNA targeting the γ-globin gene repressor BCL11A was able to increase HbF levels from 33% to 45% in β-thalassemic erythroid cells, without compromising erythroid differentiation. On the basis of these findings and on the results here presented, novel vectors carrying, in addition to the therapeutic β-globin gene, sequences driving the production of shRNAs targeting mRNA encoding a repressor of human γ-globin gene transcription would be of interest, since they are expected to force β-globin gene transcription together with reactivation of γ-globin genes and HbF production.