BASE EDITING APPROACHES FOR THE TREATMENT OF BETAHEMOGLOBINOPATHIES

Abstract

The clinical history of β-hemoglobinopathies shows that the severity is mitigated by the synthesis of the fetal γ-globin in adulthood, typically associated with genetic variants the HBB cluster known as hereditary persistence of fetal hemoglobin (HPFH) mutations. The inventors identified that most of the known HPFH mutations in the γ-globin promoters (C>T, G>A, T>C or A>G) can be recapitulated using CBE- and ABE-mediatedbase-editing approaches. In particular, the inventors designed gRNAs that, when combined with CBEs or ABEs, generate HPFH mutations, and either disrupt binding sites for transcriptional repressors (-200 and -115 sites) or generate de novo DNA motifs recognized by transcriptional activators (e.g., -198 T>C, the -175 T>C and -113 A>G). It is noteworthy that a subset of the gRNAs targeting the -200 and the 115 regions are predicted to generate simultaneously HPFH mutations and also to make base changes other than HPFH mutations in or around the LRF and BCL11A binding sites, which might further reduce LRF and BCL11A occupancy. Accordingly, the present invention relates to base editing approaches for the treatment of β-hemoglobinopathies.

Claims

1. A method for increasing fetal hemoglobin content in a eukaryotic cell comprising contacting the eukaryotic cell with a gene editing platform comprising (a) at least one base-editing enzyme and (b) at least one guide RNA molecule for guiding the base-editing enzyme to at least one target sequence in the an HBG1 or HBG2 promoter, thereby editing said promoter and subsequently increasing the expression of gamma globin in said eukaryotic cell.

2. The method of claim 1 wherein the gene editing platform introduces the a -198T>C mutation in the HBG1 or HBG2 promoter so that the KFL1 activator binds to the HBG1 or HBG2 promoter.

3. The method of claim 1 wherein the gene editing platform introduces the a -175T>C mutation in the HBG1 or HBG2 promoter so that the TAL1 activator binds to the HBG1 or HBG2 promoter.

4. The method of claim 1 wherein the gene editing platform introduces a -113A>G mutation in the HBG1 or HBG2 promoter thereby permitting binding of a GATA1 activator to the HBG1 or HBG2 promoter.

5. The method of claim 1 wherein the gene editing platform edits a -200 region in the HBG1 or HBG2 promoter thereby disrupting a binding site for the LRF repressor.

6. The method of claim 5 wherein the gene editing platform introduces at least one mutation selected from the group consisting of -201C>T, -200C>T, -197C>T, -196C>T, -195C>T and -194C>T in the HBG1 or HBG2 promoter thereby disrupting a binding site for the LRF repressor.

7. The method of claim 1 wherein the gene editing edits a -115 region in the HBG1 or HBG2 promoter thereby disrupting a binding site for the BCL11A repressor.

8. The method of claim 7 wherein the gene editing platform introduces at least one mutation selected from the group consisting of -114C>T, -113C>T, -115C>T and -116C>T in the HBG1 or HBG2 promoter thereby disrupting a binding site for the BCL11A repressor.

9. The method of claim 1 wherein the eukaryotic cell is selected from the group consisting of hematopoietic progenitor cells, hematopoietic stem cells (HSCs), pluripotent cells and induced pluripotent stem cells (iPS)).

10. The method of claim 1 wherein the at least one base-editing enzyme comprises a nickase.

11. The method of claim 10 wherein the nickase comprises the amino acid sequence as set forth in SEQ ID NO: 3 or SEQ ID NO:33.

12. The method of claim 1 wherein the at least one base-editing enzyme is a cytidine deaminase or an adenosine deaminase.

13. The method of claim 12 wherein the cytidine deaminase or the adenosine deaminase comprises a variant of the amino acid sequence as set forth in SEQ ID NO:4-14.

14. The method of claim 1 wherein the at least one base-editing enzyme is ABEmax, AncBE4max, or evoCDA1-BE4max-NG.

15. The method of claim 1 wherein the at least one base-editing enzyme and the at least one guide RNA molecule is chosen according to Table B.

16. The method of claim 1 wherein a plurality of guide RNA molecules are designed for targeting a plurality of sequences in the HBG1 or HBG2 promoter.

17. The method of claim 1 wherein a plurality of base-editing enzymes and a plurality of guide RNA molecules are designed for targeting a plurality of sequences in the HBG1 or HBG2 promoter.

18. A method for increasing fetal hemoglobin levels in a subject in need thereof, comprising transplanting into the subject a therapeutically effective amount of a population of eukaryotic cells obtained by the method of claim 1.

19. The method of claim 18 wherein the subject has been diagnosed with a hemoglobinopathy.

20. The method of claim 9, wherein the pluripotent cells are embryonic stem (ES) cells.

21. The method of claim 10, wherein the nickase is a Cas9 nickase.

22. The method of claim 19 wherein the hemoglobinopathy is sickle cell disease or β-thalassemia.

Description

Figures

[0134] FIG. 1: HPFH mutations in the HBG½ promoters, in the β-globin locus. Schematic representation of the β-globin locus on chromosome 11, depicting the HBG2 and HBG1 genes and their promoters. The sequence of the HBG2 and HBG1 identical promoters (from -214 to -98 nucleotides upstream of the HBG TSS) is shown below. Black arrows indicate HPFH mutations described at HBG1 and/or HBG2 promoters. HPFH mutations that lead to the de novo creation of transcriptional activator (TAL1, KLF1 and GATA1) binding sites, are above the DNA sequence, while HPFH mutations that lead to the disruption of transcriptional repressor (LRF and BCL11A) binding sites are below the DNA sequence. HPFH mutations that can be generated by base editing are highlighted with rectangles. Ovals indicate transcriptional activators and repressors. Target sequences of the gRNAs designed to be used with base editing enzymes, are reported in the bottom part of the figure (highlighted with dark arrows) and aligned with the DNA sequence that they bind to. The target bases are highlighted in white bold and the rest of the protospacer sequence is highlighted in grey. The PAM (protospacer adjacent motif) for each gRNA is reported in black at the end of the arrows.

[0135] FIG. 2: Efficient base editing of the HBG½ promoters in K562 cells. (A-M) A-T to G-C (A-C, N) or C-G to T-A (D-M) base editing efficiency, calculated by the EditR software in samples subjected to Sanger sequencing. The base editing efficiency percentage was measured by subtracting the percentage of the base conversion in the control that was considered as background noise. On the top of each graph the target and the enzyme used are indicated. Data are expressed as mean±SD (n=2-3 biologically independent experiments). (O) The frequency of insertions and deletions (InDel), measured by TIDE analysis, is reported for the control and the base edited samples. (P) The frequency of the 4.9-kb deletion, measured by ddPCR, is reported for control and base edited samples, and for a positive control (DNA extracted from K562 cells edited with the canonical Cas9 nuclease). (Q) Binding site conversion after base editing as analyzed by DiffLogo.sup.(26). *Two different gRNAs (BCL11A_bs_1 and BCL11A_bs_2) were used in combination with the enzymes evoCDA1-BE4max-NG and evoFERNY-BE4max-NG (graphs E and F). Similarly, two different gRNAs (LRF_bs_1 and LRF_bs_2) were used in combination with CBE-SpRY enzyme (graphs L and M).

[0136] FIG. 3: Efficient base editing of the HBG½ promoters in HUDEP-2 cells. A-T to G-C (A-C) or C-G to T-A (D-H) base editing efficiency, calculated by the EditR software in samples subjected to Sanger sequencing. The base editing efficiency was calculated as described in FIG. 2 legend. On the top of each graph, the target and the enzyme used are indicated. (I) The frequency of insertions and deletions (InDel), measured by TIDE analysis, is reported for the control and the base edited samples. (J) Binding site conversion after base editing as analyzed by DiffLogo.sup.(26). *Two different gRNAs (BCL11A_bs_1 and BCL11A_bs_2) were used in combination with the enzyme evoCDA1-BE4max-NG and evoFERNY-BE4max-NG (graphs E and F).

[0137] FIG. 4: HbF de-repression after generation of the TAL1 and KLF1 binding sites in HUDEP-2 cells. (A) Frequency of HbF-expressing cells (as determined by flow cytometry), in Glycophorin A (GPA).sup.high populations at day 0 (D0) and day 9 (D9) of erythroid differentiation. The base editing efficiency is indicated on the top of each black bar (D0). (B) RT-qPCR analysis of β- and γ-globin mRNA levels at D0, 6 and 9 of erythroid differentiation. β- and γ-globin mRNA expression was normalized to α-globin mRNA and expressed as percentage of the total β-like globins. (C). Expression of γ- (.sup.Gγ- + .sup.Aγ-) and β-globin chains measured by RP-HPLC. β-like globin chain expression was normalized to α-globin. (D) Analysis of HbF and HbA by cation-exchange HPLC. We calculated the percentage of each Hb type over the total Hb tetramers. (E-G) Flow-cytometry analysis of the late erythroid marker GYPA (E) and of the early erythroid markers CD36 (F) and CD71 (G) at D0 and D9 of HUDEP-2 differentiation.

[0138] FIG. 5: HbF reactivation upon disruption of the BCL11A binding site in HUDEP-2 cells. (A-C) Flow-cytometry analysis of the late erythroid marker GYPA (A) and of the early erythroid markers CD36 (B) and CD71 (C) at D0 and D9 of HUDEP-2 differentiation. (D) Frequency of HbF-expressing cells (as determined by flow cytometry), in GPA.sup.high populations at day 0 (D0) and day 9 (D9) of erythroid differentiation. The base editing efficiency is indicated on the top of each black bar (D0). (E) RT-qPCR analysis of β- and γ-globin mRNA levels at D0, 6 and 9 of erythroid differentiation. β- and γ-globin mRNA expression was normalized to α-globin mRNA and expressed as percentage of the total β-like globins. (F) Expression of γ-(.sup.Gγ- + .sup.Aγ-) and β-globin chains measured by RP-HPLC. β-like globin expression was normalized to α-globin. (G) Analysis of HbF and HbA by cation-exchange HPLC. We calculated the percentage of each Hb type over the total Hb tetramers.

[0139] FIG. 6: HbF increase upon disruption of the LRF binding site in HUDEP-2 and HSPCs. (A) Frequency of HbF-expressing cells (as determined by flow cytometry) in GPA.sup.high populations in undifferentiated HUDEP-2 cells. (B) To edit the LRF binding site, HSPCs were transfected with two plasmids expressing the AncBE4max_NAA and the LRF_bs_1 gRNA, respectively. Base editing efficiency in the LRF binding site, as calculated by the EditR software in HSPC samples subjected to Sanger sequencing. The base editing efficiency percentage was calculated as described in FIG. 2 legend. The frequency of insertions and deletions (InDel), measured by TIDE analysis, is reported for the control and the base edited samples. (C) Binding site conversion after base editing in HSPCs as analyzed by DiffLogo.sup.(26). (D) Quantification of HbF and HbA by cation-exchange HPLC in a bulk population of control and HBG-edited BFU-E colonies. We plotted the percentage of each Hb type over the total Hb tetramers. Control cells include samples transfected either with TE buffer or with the AncBE4max_NAA plasmid only, or with the AncBE4max_NAA plasmid and a gRNA targeting an unrelated to the β-globin locus site (AAVS1 site; Weber et al., Sc. Advances, 2020).

[0140] FIG. 7: Base editing efficiency by enzymes with low RNA off-target activity, in the HBG½ promoters, in K562 cells. A-T to G-C (A-C) or C-G to T-A (D) base editing efficiency, calculated by the EditR software in samples subjected to Sanger sequencing. The base editing efficiency percentage was calculated as described in FIG. 2 legend. On the top of each graph, the target and the enzyme used are indicated. For each target, we compared the classical base editing enzyme and the base editing enzyme with lower RNA off-target activity.

[0141] FIG. 8: Testing CBEs to disrupt the LRF binding site in SCD HSPCs. C-G to T-A (A and B) base editing efficiency, calculated by the EditR software in samples subjected to Sanger sequencing. (C) The frequency of insertions and deletions (InDel), measured by TIDE analysis, is reported for base edited samples. (D) RT-qPCR analysis of β.sup.S- and γ-globin mRNA levels at Day 13 of erythroid differentiation. β.sup.S- and γ-globin mRNA expression was normalized to α-globin mRNA and expressed as percentage of the total β-like globins. (E) Expression of γ- (.sup.Gγ-+ .sup.Aγ-) and β.sup.S-globin chains measured by RP-HPLC. β-like globin chain expression was normalized to α-globin. (F) Analysis of HbF and HbS by cation-exchange HPLC. We calculated the percentage of each Hb type over the total Hb tetramers. (G) Frequency of HbF-expressing cells (as determined by flow cytometry), in Glycophorin A (GPA).sup.high populations at Day 19 of erythroid differentiation. (H) Frequency of non-sickle cells upon O.sub.2 deprivation in mock-transfected control and base edited samples. (D-F and H) Below each graph, the base editing efficiency (BE%), the percentage of HbF over the total Hb tetramers (HbF%), as measured by cation-exchange HPLC and the frequency of HbF-expressing cells, as determined by flow cytometry (F-cells%) are indicated for each sample. Donor number n=1.

[0142] FIG. 9: Testing ABEs to disrupt the LRF binding site or create the KLF1 binding site in SCD HSPCs. A-T to G-C (A and D) base editing efficiency, calculated by the EditR software in samples subjected to Sanger sequencing. (E) The frequency of insertions and deletions (InDel), measured by TIDE analysis, is reported for base edited samples. (F) RT-qPCR analysis of β.sup.s- and γ-globin mRNA levels at Day 13 of erythroid differentiation. β.sup.S- and γ-globin mRNA expression was normalized to α-globin mRNA and expressed as percentage of the total β-like globins. (G) Expression of γ- (.sup.Gγ- + .sup.Aγ-) and β.sup.S-globin chains measured by RP-HPLC. β-like globin chain expression was normalized to α-globin. (H) Analysis of HbF and HbS by cation-exchange HPLC. We calculated the percentage of each Hb type over the total Hb tetramers. (I) Frequency of non-sickle cells upon O.sub.2 deprivation in control (mock-transfected and edited in an unrelated AAVSI1 locus) and base edited samples. (J) Frequency of HbF-expressing cells (as determined by flow cytometry), in Glycophorin A (GPA).sup.high populations at Day 19 of erythroid differentiation. (F-I) Below each graph, the base editing efficiency (BE%), the percentage of HbF over the total Hb tetramers (HbF%), as measured by cation-exchange HPLC and the frequency of HbF-expressing cells, as determined by flow cytometry (F-cells%) are indicated for each sample. Donor number n=1.

[0143] FIG. 10: RNA-mediated base editing at the HBG½ promoters in K562 cells and SCD HSPCs. (A-H) A-T to G-C or C-G to T-A base editing efficiency in K562 and HSPCs. The base editing efficiency was calculated by the EditR software in K562 (A-E) and SCD HSPC (F-H) samples subjected to Sanger sequencing. On the top of each graph the enzyme and the gRNA used are indicated.

EXAMPLE

Methods

Cell Line Culture

[0144] K562 were maintained in RPMI 1640 (Lonza) containing glutamine and supplemented with 10% fetal bovine serum (Lonza), 2 mM Hepes (Life Technologies), 100 nM sodium pyruvate (Life Technologies), and penicillin and streptomycin (Life Technologies). HUDEP-2 cells were cultured in StemSpan SFEM (Stem Cell Technologies), supplemented with 1 .Math.g/mL doxycycline (Sigma), 10.sup.-6 M dexamethasone (Sigma), 100 ng/mL human stem cell factor (SCF) (Peprotech), 3IU/mL erythropoietin (Necker Hospital Pharmacy), L-glutamine (Life Technologies) and penicillin/streptomycin. HUDEP-2 cells were differentiated in Iscove’s Modified Dulbecco’s Medium (IMDM) (Life Technologies) supplemented with 330 .Math.g/mL holo-transferrin (Sigma), 10 .Math.g/mL recombinant human insulin (Sigma), 2IU/mL heparin (Sigma), 5% human AB serum, 3IU/mL erythropoietin, 100 ng/mL human SCF, 1 .Math.g/mL doxycycline, 1% L-glutamine, and 1% penicillin/streptomycin for 9 days. Flow cytometry analysis of CD36, CD71 and GYPA surface markers and a standard May-Grumwald Giemsa staining were performed to monitor erythroid differentiation.

HSPC Purification and Culture

[0145] We obtained human cord blood (CB) CD34.sup.+ HSPCs from healthy donors and human non-mobilized peripheral blood CD34.sup.+ HSPCs from SCD patients. CB samples eligible for research purposes were obtained because of a convention with the CB bank of Saint Louis Hospital (Paris, France). SCD samples eligible for research purposes were obtained because of a convention with “Hôpital Necker-Enfants malades” Hospital (Paris, France). Written informed consent was obtained from all adult subjects. All experiments were performed in accordance with the Declaration of Helsinki. The study was approved by the regional investigational review board (reference: DC 2014-2272, CPP Ile-de-France II “Hôpital Necker-Enfants malades”). HSPCs were purified by immunomagnetic selection with AutoMACS (Miltenyi Biotec) after immunostaining with the CD34 MicroBead Kit (Miltenyi Biotec). Forty-eight hours before transfection, CD34.sup.+ cells (10.sup.6cells/ml) were thawed and cultured in the “HSPC medium” containing StemSpan (STEMCELL Technologies) supplemented with penicillin/streptomycin (Gibco), 250 nM StemRegenin1 (STEMCELL Technologies), and the following recombinant human cytokines (PeproTech): human stem cell factor (SCF) (300 ng/ml), Flt-3L (300 ng/ml), thrombopoietin (TPO) (100 ng/ml), and interleukin-3 (IL-3) (60 ng/ml).

Plasmids

[0146] Plasmids used in this study include pCMV_ABEmax_P2A_GFP (Addgene #112101), pCMV_AncBE4max_P2A_GFP (Addgene #112100), pBT374 (Addgene #125615), pBT372 (Addgene #125613), pCMV-ABEmaxAW (Addgene #125647), pMJ920 (Addgene #42234), ABE8e (Addgene #138489), pCMV-BE4max-NRCH (Addgene #136920), pCAG-CBE4max-SpG-P2A-EGFP (RTW4552) (Addgene #139998) and pCAG-CBE4max-SpRY-P2A-EGFP (RTW5133) (Addgene #139999). The AncBE4max_NAA plasmid was created by replacing the PAM interaction domain of the SpCas9n with the one of the SmaCas9, while the plasmid AncBE4max_R33A/K34A was created by inserting the mutations R33A and K34A in the APOBEC1 domain of the AncBE4max plasmid (Addgene #112094). A DNA fragment (3′UTR+poly-A) containing two copies of the 3′ untranslated region (UTR) of the HBB gene and a poly-A sequence of 96 adenines was purchased by Genscript (gene synthesis). Similarly, a DNA fragment containing the uridine depleted coding sequence of pCAG-CBE4max-SpRY-P2A-EGFP was generated (CBE-SpRY_U-delp). The CBE-SpRY-OPT plasmid was generated by inserting the 3′UTR+poly-A fragment in pCAG-CBE4max-SpRY-P2A-EGFP and by replacing CBE4max-SpRY with the CBE-SpRY_U-delp fragment. Furthermore, we replaced the CAG synthetic promoter of CBE-SpRY plasmid with a T7 promoter. The ABE-SpRY-OPT plasmid was created by inserting the 3′UTR+poly-A fragment in pCMV-T7-ABEmax(7.10)- SpRY-P2A-EGFP (RTW5025) (Addgene #140003).

gRNA Design

[0147] For the gRNA expression plasmid construction, oligonucleotides were annealed to create the gRNA protospacer and the duplexes were ligated into Bbs I-digested MA128 plasmid (provided by M. Amendola, Genethon, France).

TABLE-US-00018 gRNA target sequences. gRNA Target sequence (5′ to 3′) Position (hg19) Strand TAL1_bs_1 ATATTTGCATTGAGATAGTGTGG (SEQ ID NO : 15) chr11: 5271260-5271279 (HBG1) + chr11: 5276184-5276203 (HBG2) KLF1_bs_1 GTGGGGAAGGGGCCCCCAAGAGG (SEQ ID NO : 16) chr11: 5271279-5271298 (HBG1) + chr11: 5276203-5276222 (HBG2) BCL11A_bs_1 CTTGACCAATAGCCTTGACAAGG (SEQ ID NO : 17) chr11: 5271188-5271207 (HBG1) - chr11: 5276112-5276131 (HBG2) BCL11_A­bs­_2 TTGACCAATAGCCTTGACAAGG (SEQ ID NO 18) chr11: 5271187-5271206 (HBG1) - chr11: 5276111-5276130 (HBG2) LRF_bs_1 CCTTCCCCACACTATCTCAATG (SEQ ID NO 19) chr11: 5271269-5271288 (HBG1) - chr11: 5276193-5276212 (HBG2) LRF_bs_2 GCCCCTTCCCCACACTATCTCAA (SEQ ID NO :20) chr11: 5271272-5271291 (HBG1) - chr11: 5276196-5276215 (HBG2) PAM motif is highlighted in bold.

mRNA in Vitro Transcription

[0148] 10 .Math.g of CBE-SpRY-OPT, ABE-SpRY-OPT or ABE8e expressing plasmids were digested overnight with 20 Units of a restriction enzyme that cuts once right after the poly-A tail (AflII for CBE-SpRY and ABE-SpRY and SapI for ABE8e). The linearized plasmids were purified using a PCR purification kit (QIAGEN #28106) and were eluted in 30 .Math.l of DNase/RNase-free water. 1 .Math.g of linearized plasmid was used as template for the in vitro transcription reaction (MEGAscript, Ambion #AM1334). The in vitro transcription protocol was modified as follows.

[0149] The GTP nucleotide solution was used at a final concentration of 3.0 mM instead of 7.5 mM and the anti-reverse cap analog N7-Methyl-3′-O-Methyl-Guanosine-5′-Triphosphate-5′-Guanosine (ARCA, Trilink #N-7003) was used at a final concentration of 12.0 mM resulting in a final ratio of Cap:GTP of 4:1 that allows efficient capping of the mRNA. The incubation time for the in vitro reaction was reduced to 30 minutes. After DNaseI treatment (MEGAscript, Ambion #AM1334), the ABE8e mRNA was poly-A tailed according to the manufacturer’s protocol (Poly-A tailing kit, Ambion #AM1350). mRNA was precipitated using lithium chloride and resuspended in TE buffer in a final volume that allowed to achieve a concentration of ≥ 1 .Math.g/.Math.l. The mRNA quality was checked using Bioanalyzer (Agilent).

Plasmid Transfection

[0150] K562 and HUDEP-2 cells (10.sup.6 cells/condition) were transfected with 3.6 .Math.g of a base editing enzyme expressing plasmid and 1.2 .Math.g of gRNA-containing plasmid. For base editing enzyme plasmids that do not expressing GFP, we co-transfected 250 ng of a GFPmax expressing plasmid (Lonza). We used AMAXA Cell Line Nucleofector Kit V (VCA-1003) and U-16 and L-29 programs (Nucleofector II) for K562 and HUDEP-2, respectively. For K562 cells, transfection efficiency was evaluated by flow cytometry, using the Fortessa X20 (BD Biosciences) flow cytometer 18 h after transfection and cells were maintained in culture for at least 3 days and at day 3 genomic DNA extraction was performed. GFP.sup.+ HUDEP-2 cells were sorted 18h after transfection using SH800 Cell Sorter (Sony Biotechnology) and sorted cells were expanded in culture. 3 days after transfection, genomic DNA extraction was performed. CD34.sup.+ HSPCs (10.sup.6 cells/condition) were transfected with 3.6 .Math.g of the enzyme-expressing plasmid and 2.4 .Math.g of the gRNA-containing plasmid. For base editing enzyme plasmids that do not express GFP, we co-transfected 250 ng of a GFPmax expressing plasmid (Lonza). We used AMAXA Human CD34 Cell Nucleofector Kit (VPA-1003) and U-08 program (Nucleofector II). 18 h after transfection, GFP.sup.+ CD34.sup.+ HSPCs were sorted using SH800 Cell Sorter (Sony Biotechnology) and either plated at a concentration of 500,000/mL in cytokine-enriched HSPC medium (described above) for at least 6 days and then genomic DNA extraction was performed, or were differentiated in mature RBCs using a 3-phase erythroid differentiation protocol (Weber, Frati Science Advances 2020) of up to 20 days. Genomic DNA extraction was performed at Day 6, total RNA extraction was performed at Day 13 and functional analyses to evaluate the HbF expression (flow cytometry, RP-HPLC, CE-HPLC and a sickling assay) were conducted at day 19.

RNA Transfection

[0151] K562 cells (2×10.sup.5 cells/condition) were transfected with 2.0 .Math.g of a base editor-expressing mRNA and a synthetic gRNA containing chemical modifications (2′-O-Methyl at 3 first and last bases, 3′ phosphorothioate bonds between first 3 and last 2 bases) purchased from Synthego, at a final concentration of 1.5 .Math.M. We used the P3 Primary Cell 4D- Nucleofector X Kit S (Lonza) and the CA137 program (Nucleofector 4D). Cells transfected with TE buffer served as negative control. RNA-transfected K562 cells were maintained in culture for at least 3 days prior to genomic DNA extraction and base editing analysis.

[0152] CD34.sup.+ HSPCs (2×10.sup.5 cells/condition) were transfected with 3.0 .Math.g of a base editor-expressing mRNA and a synthetic gRNA containing chemical modifications (2′-O-Methyl at 3 first and last bases, 3′ phosphorothioate bonds between first 3 and last 2 bases) purchased from Synthego, at a final concentration of 4.6 .Math.M. We used the P3 Primary Cell 4D- Nucleofector X Kit S (Lonza) and the CA137 program (Nucleofector 4D). Cells transfected with TE buffer or with the base editor mRNA only, or with the base editor mRNA and a gRNA targeting the AAVS1 locus, served as negative controls. RNA-transfected HSPCs were plated at a concentration of 500,000/mL in the HSPC medium (described above) and cultured for at least 6 days prior to genomic DNA extraction and base editing analysis.

Evaluation of Editing Efficiency

[0153] Genomic DNA was extracted from control and edited cells using PURE LINK Genomic DNA Mini kit (LifeTechnologies) following manufacturer’s instructions, 3 days post-transfection for K562 and HUDEP2 cells and 6 days post-trasnfection for CD34.sup.+ HSPCs. To evaluate base editing efficiency at gRNA target sites, we performed PCR followed by Sanger sequencing and EditR analysis (EditR: A Method to Quantify Base Editing from Sanger Sequencing).sup.(27). TIDE analysis (Tracking of InDels by Decomposition) was also performed in order to evaluate the percentage of insertion and deletion (InDels) in base edited samples.sup.(28).

TABLE-US-00019 Primers used to detect base editing and InDels events. Amplified region F/R Sequence (5′ to 3′) HBG1 + HBG2 promoters F AAAAACGGCTGACAAAAGAAGTCCTGGT AT (SEQ ID NO :21) R ATAACCTCAGACGTTCCAGAAGCGAGTGT G (SEQ ID NO :22)

[0154] Digital Droplet PCR (ddPCR) was performed using EvaGreen mix or primer/probe mix (Biorad) to quantify the frequency of the 4.9-kb deletion by amplifying the HBG1-HBG2 intervening region I or II respectively. Short (~1 min) elongation time allowed the PCR amplification of the genomic region harboring the deletion. Control primers annealing to a genomic region on the same chromosome (chr 11) or to hALB (chr 4) were used as DNA loading control respectively.

TABLE-US-00020 Primers used for ddPCR. Amplified region F/R Sequence (5′ to 3′) HBG1-HBG2 intervening region I F GTTTTAAAACAACAAAAATGAGGGAAA GA (SEQ ID NO :23) R GTTGCTTTATAGGATTTTTCACTACAC (SEQ ID NO :24) Chr11 control region F CCCTTCCGAGAGGATTTAGG (SEQ ID NO :25) R AGTCGGGATCTGAACAATGG (SEQ ID NO :26) HBG1-HBG2 intervening region II F ACGGATAAGTAGATATTGAGGTAAGC (SEQ ID NO :54) R GTCTCTTTCAGTTAGCAGTGG (SEQ ID NO :55) hALB F ACTCATGGGAGCTGCTGGTT (SEQ ID NO :56) R GCTGTCATCTCTTGTGGGCTG SEQ ID NO : 57) F, forward primer; R, reverse primer.

Flow Cytometry Analysis

[0155] Differentiated HUDEP-2 were fixed and permeabilized using BD Cytofix/Cytoperm solution (BD Pharmingen) and stained with an antibody recognizing HbF (APC-conjugated anti HbF antibody, MHF05, Life Technologies). Flow cytometry analysis of CD36, CD71 and GYPA erythroid surface markers was performed using a V450-conjugated anti-CD36 antibody (561535, BD Horizon), a FITC-conjugated anti-CD71 antibody (555536, BD Pharmingen) and a PE-Cy7-conjugated anti- GYPA antibody (563666, BD Pharmingen). SCD RBCs differentiated from control and edited HSPCs were fixed with 0.05% glutaraldehyde. permeabilized with 0.1% TRITON X-100 and stained with an antibody recognizing HbF (FITC-conjugated anti HbF antibody, clone 2D12 552829 BD). Flow cytometry analysis of CD36, CD71, GYPA, BAND3 and α4-Integrin erythroid surface markers was performed using a V450-conjugated anti-CD36 antibody (561535, BD Horizon), a FITC-conjugated anti-CD71 antibody (555536, BD Pharmingen), a PE-Cy7-conjugated anti-GYPA antibody (563666, BD Pharmingen), a PE-cpnjugated anti-BAND3 antibody (9439, IBGRL) and a APC-conjugated anti-CD49d antibody (559881, BD). Flow cytometry analysis of DRAQ5 (enucleation) and 7AAD (viability) was performed using anti-double stranded DNA dyes (65-0880-96, Invitrogen and 559925, BD, respectively). Flow cytometry analyses were performed using Fortessa X20 (BD Biosciences) or Gallios flow cytometers. Data were analyzed using the FlowJo (BD Biosciences) or KALUZA software.

Colony-Forming Cell (CFC) Assay

[0156] HSPCs were plated at a concentration of 1×10.sup.3 cells/mL in methylcellulose-containing medium (GFH4435, Stem Cell Technologies) under conditions supporting erythroid and granulomonocytic differentiation. BFU-E and CFU-GM colonies were scored after 14 days. BFU-Es and CFU-GMs were randomly picked and collected as bulk populations (containing at least 30 colonies) to evaluate the hemoglobin expression by CE-HPLC.

RT-qPCR Analysis of Globin Transcripts

[0157] Total RNA was extracted from HUDEP-2 (at day 0, 6 and 9 of differentiation) and erythroid cells differentiated from SCD HSPCs (at day 13) using RNeasy micro kit (QIAGEN), following manufacturer’s instructions. Mature transcripts were reverse-transcribed using SuperScript First-Strand Synthesis System for RT-qPCR (Invitrogen) with oligo (dT) primers. RT-qPCR was performed using iTaq universal SYBR Green master mix (Biorad) and a Viia7 Real-Time PCR system (ThermoFisher Scientific).

TABLE-US-00021 Primers used for RT-qPCR. Amplified region F/R Sequence (5′ to 3′) HBA F CGGTCAACTTCAAGCTCCTAA (SEQ ID NO :27) R ACAGAAGCCAGGAACTTGTC (SEQ ID NO :28) HBB F GCAAGGTGAACGTGGATGAAGT (SEQ ID NO :29) R TAACAGCATCAGGAGTGGACAGA (SEQ ID NO :30) HBG1+HBG2 F CCTGTCCTCTGCCTCTGCC (SEQ ID NO :31) R GGATTGCCAAAACGGTCAC (SEQ ID NO :32) F, forward primer; R, reverse primer.

RP-HPLC Analysis of Globin Chains

[0158] RP-HPLC analysis was performed using a NexeraX2 SIL-30AC chromatograph and the LC Solution software (Shimadzu). Globin chains were separated by HPLC using a 250×4.6 mm, 3.6 .Math.m Aeris Widepore column (Phenomenex). Samples were eluted with a gradient mixture of solution A (water/acetonitrile/trifluoroacetic acid, 95:5:0.1) and solution B (water/acetonitrile/trifluoroacetic acid, 5:95:0.1). The absorbance was measured at 220 nm.

CE-HPLC Analysis of Hemoglobin Tetramers

[0159] Cation-exchange HPLC analysis was performed using a NexeraX2 SIL-30AC chromatograph and the LC Solution software (Shimadzu). Hemoglobin tetramers were separated by HPLC using a 2 cation-exchange column (PolyCAT A, PolyLC, Columbia, MD). Samples were eluted with a gradient mixture of solution A (20 mM bis Tris, 2 mM KCN, pH=6.5) and solution B (20 mM bis Tris, 2 mM KCN, 250 mM NaCl, pH=6.8). The absorbance was measured at 415 nm.

Sickling Assay

[0160] At the end of the erythroid differentiation, mature RBCs derived from SCD HSPCs were incubated under hypoxia conditions (0% O.sub.2) and the time course of sickling was monitored in real time by video microscopy, capturing images every 20 min for at least 60 min using an AxioObserver Z1 microscope (Zeiss) and a 40× objective. Images of the same fields were taken throughout all stages and processed with ImageJ to determine the percentage of non-sickle RBCs per field of acquisition in the total RBC population. More than 400 cells were counted per condition.

Results

Efficient Base Editing in the HBG Promoters Leads to the De Novo Generation of Binding Sites for TAL1 and KLF1 Transcriptional Activators

[0161] The -175 T>C HPFH mutation has been shown to recruit TAL1 transcription activator to the HBG promoters.sup.(6), while the -198 T>C HPFH mutation recruits the KLF1 transcriptional activator.sup.(7). We used gRNAs (TAL1_bs_1 and KLF1_bs_1) (FIG. 1) that can target bases in positions -175 and -198 in the HBG promoters. For these gRNAs, the target bases are in position 3 and 7, respectively. Transfection of the erythroleukemia cell line K562 with the ABEmax_GFP plasmid and the TAL1_bs_1 or KLF1_bs_1 gRNA plasmid, led to an A>G conversion (T>C in the opposite strand) with efficiencies of 61.0% and 30.3% respectively, in the bulk cell populations (FIGS. 2A, 2B and 2Q).

Generation of TAL1 and KLF1 Binding Sites Leads to Γ-Globin De-Repression

[0162] K562 cells express mainly γ-globin and for this reason they cannot be used as a model to measure γ-globin de-repression. Hence, we employed the HUDEP-2 adult erythroid progenitor cell line to evaluate γ-globin de-repression following the creation of the TAL1 and KLF1 activator binding sites. After plasmid transfection, with the ABEmax_GFP plasmid and either the TAL1_bs_1, or the KLF1_bs_1 gRNA plasmid, GFP.sup.+ HUDEP2 cells were sorted and expanded. The base editing efficiency was 65% and 45% at position -175 and -198, respectively (FIGS. 3A, 3B and 3J). Control and edited cells were differentiated in mature erythrocytes. Flow cytometry analysis of cells edited at -175 and -198 positions, revealed a high frequency of HbF-expressing cells (76.0% and 80.3% at day 0, and 85.0% and 88.1% at day 9 of differentiation), while in control populations transfected only with the ABEmax_GFP plasmid, the HbF-expressing cells were around 3.0% (FIG. 4A). Accordingly, we observed an increased production of γ-globin transcripts and a parallel decrease of the adult β-globin mRNAs (FIG. 4B). Reverse phase high-performance liquid chromatography (RP-HPLC) confirmed the significant increase in γ-globin with concomitant decrease of β-globin production (FIG. 4C). Generating the binding sites of TAL1 and KLF1 resulted in high HbF levels accounting for up to 53.5% and 30.0% respectively of the total Hb as determined by cation-exchange HPLC (CE-HPLC) (FIG. 4D). The base editing of the HBG½ promoters did not alter erythroid cell differentiation, as assessed by flow cytometry analysis of the erythroid markers GYPA, CD36 and CD71 (FIGS. 4E-4G).

Base Editing in the -115 Region of the HBG Promoters Leads to the Disruption of the BCL11A Transcriptional Repressor Binding Site and the Simultaneous Generation of a Binding Site for GATA1 Transcriptional Activator

[0163] HPFH mutations on the -115 region of the HBG promoters cause elevated HbF expression by disrupting the binding site of the BCL11A transcriptional repressor (-114 C>T) or by creating a binding site for GATA1 transcriptional activator (-113 A>G). We designed gRNAs (BCL11A_bs_1 and BCL11A_bs_2) that can be used, either by ABE or CBE, in order to create these HPFH mutations (-114 C>T and -113 A>G) and additional HPFH-like mutations (-116 A>G and -115 C>T) (FIG. 1). The target bases fall into the canonical base editing window. In particular, the -116, -115, -114 and -113 bases are in positions 5-8 and 4-7 for the gRNAs BCL11A_bs_1 and BCL11A_bs_2 respectively. After transfection of the K562 cell line with the ABEmax_GFP and gRNA BCL11A_bs_1 plasmid, we succeeded to obtain the A>G conversions in position -116 and -113 (40.7% and 34.3% respectively) (FIGS. 2C and 2Q). Transfection of the same gRNA BCL11A_bs_1 plasmid with the AncBE4max_GFP plasmid, led to the C>T conversion in positions -115 and -114 (34.8% and 28.5% respectively) (FIGS. 2D and 2Q). In an effort to expand the enzyme options for disrupting the BCL11A transcriptional repressor binding site, we transfected the K562 cell line with the evoCDA1-BE4max-NG or the evoFERNY-BE4max-NG plasmid, two enzymes that recognize NG PAM, in combination with the two gRNAs targeting the BCL11A binding site (BCL11A_bs_1 and BCL11A_bs_2). All of the 4 different combinations led to the same C>T conversions (-115 C>T and -114 C>T) with efficiencies ranging from 18 to 40.5% (evoCDA1-BE4max-NG -BCL11A_bs_1; 40.5% -115 C>T and 31.5% -114 C>T, evoCDA1-BE4max-NG -BCL11A_bs_2; 30.0% -115 C>T and 22.0% -114 C>T, evoFERNY-BE4max-NG -BCL11A_bs_1; 22.0% -115 C>T and 21.5% -114 C>T and evoFERNY-BE4max-NG -BCL11A_bs_2; 18.0% -115 C>T and 19.0% -114 C>T) (FIGS. 2E, 2F and 2Q).

Non-NGG Base Editors Can Disrupt the LRF Transcriptional Repressor Binding Site By Editing Multiple Bases and Creating HPFH and HPFH-Like Mutations

[0164] The -200 region of the HBG promoters contains different HPFH mutations associated with high expression of γ-globin in adult life. The majority of these mutations de-repress the HBG genes by reducing the binding capacity of the LRF transcriptional repressor via the disruption of its binding site. In the LRF binding site, there are 8 cytosines and theoretically all of them can be targeted by base editing in order to be converted in thymines. Consequently, it is possible to create multiple HPFH mutations and additional mutations that could induce an HPFH-like phenotype by impairing the LRF binding capacity (FIG. 1). The absence of the canonical SpyCas9 NGG PAM close to the LRF binding site prompted us to generate base editing enzymes containing non-NGG Cas9 variants that allowed the editing of this site. This variant of Cas9 recognizes an NAA PAM, which is ideal for targeting the LRF binding site, as it allows the designing of a gRNA (LRF_bs_2) that places the target bases - 8 cytosines - in position 2-11. After the PAM-interacting domain exchange, the resulting enzyme (called AncBE4max_NAA) was transfected as plasmid in the K562 cell line in combination with the LRF_bs_2 gRNA plasmid. We were able to modify 7 out of the 8 cytosines of the LRF binding site with efficiencies of up to 37.0% (8.7% -202 C>T; 20.3% -201 C>T; 37.0% -200 C>T; 30.7% -197 C>T; 27.7% -196 C>T; 16.3% -195 C>T; 5.7% -194 C>T) (FIGS. 2G and 2Q). One more gRNA (LRF _bs_1) was designed so as to target the LRF binding site using the evoCDA1-BE4max-NG or the evoFERNY-BE4max-NG enzyme (FIG. 1). With these combinations we can target 6 out of 8 cytosines of the motif and these 6 cytosines are in position 1-8. Transfection of the K562 cell line with the LRF_bs_1 gRNA plasmid and either the evoCDA1-BE4max-NG or the evoFERNY-BE4max-NG enzyme plasmid revealed efficiencies that ranged from 8.0 to 28.5% for the evoCDA1-BE4max-NG enzyme (8.0% -201 C>T; 22.5% -200 C>T; 28.0% -197 C>T; 27.5% -196 C>T; 28.5% -195 C>T; 21.0% -194 C>T) and from 15.0 to 32.5% for the evoFERNY-BE4max-NG enzyme (28.0% -197 C>T; 32.5% -196 C>T; 25.5% -195 C>T; 15.0% -194 C>T) (FIGS. 2H, 2I and 2Q). The same gRNAs were tested with more efficient non-NGG base editors, such as CBE-NRCH, CBE-SpG and CBE-SpRY. When combined with LRF_bs_1 gRNA, CBE-NRCH, CBE-SpG and CBE-SpRY led to C>T efficiencies up to 50.3%, 43.7% and 46.3% respectively, upon plasmid transfection in K562 cells (FIGS. 2J-L and 2Q). Given the PAMless nature of CBE-SpRY, this base editor was combined with LRF_bs_2 gRNA and outperformed all the above-mentioned enzymes hitting efficiencies of 58.0%. Importantly thanks to its wide editing window, CBE-SpRY converted all the cytosines of the -200 region (FIGS. 2M and 2Q). Finally, an ABE8e enzyme plasmid was co-transfected in K562 cells with the KLF1_bs_1 gRNA and led to A>G modifications of both the -198 A:T bp and the -199 A:T bp with efficiencies of up to 72.7%, resulting in LRF binding site disruption (FIGS. 2N and 2Q).

Disruption of the BCL11A and LRF Transcriptional Repressor Binding Sites by Base Editing Leads to HbF Reactivation

[0165] The previously reported base editing approaches for targeting the -115 and -200 regions of the HBG promoters, by creating HPFH and HPFH-like mutations, were tested in the HUDEP-2 cell line in order to evaluate the γ-globin de-repression, after disrupting the BCL11A and LRF transcriptional repressor binding sites, and/or after creating a binding site for GATA1 transcriptional activator. For both regions, plasmids expressing 4 different enzymes (ABEmax_GFP, AncBE4max_GFP, evoCDA1-BE4max-NG and evoFERNY-BE4max-NG) were individually transfected in combination with single gRNA-expressing plasmids (BCL11A_bs_1, BCL11A_bs2, LRF_bs_1 and LRF_bs_2) in HUDEP-2 cells. For plasmids that do not express GFP (evoCDA1-BE4max-NG and evoFERNY-BE4max-NG), a small amount of a GFPmax-expressing plasmid was co-transfected. After transfection, GFP.sup.+ cells were FACS-sorted, expanded in culture and differentiated in erythrocytes.

[0166] Editing the BCL11A binding site with the ABEmax_GFP enzyme and the BCL11A_bs_1 gRNA, led to -116 A>G and -113 A>G conversions, with a frequency of 56.0% and 57.0%, respectively, disrupting the BCL11A binding site and simultaneously creating a GATA1 binding site (FIGS. 3C and 3J). Using the AncBE4max_GFP enzyme combined with the BCL11A_bs_1 gRNA, we succeeded in generating HPFH (38.0% -114 C>T) and HPFH-like (47.0% -115 C>T) mutations (FIGS. 3D and 3J). Base editing by the non-NGG PAM enzymes, (evoCDA1-BE4max-NG and evoFERNY-BE4max-NG) in combination with either the BCL11A_bs_1 or the BCL11A_bs_2 gRNA, was effective only using evoCDA1-BE4max-NG that led to -115 C>T and -114 C>T conversions (40.0% and 27.0%, respectively) with the BCL11A_bs_1 gRNA and -115 C>T and -114 C>T conversions (24.0% and 15.0%, respectively) with the BCL11A_bs_2 gRNA (FIGS. 3E, 3F and 3J). Base editing of the BCL11A binding site with the above-mentioned enzymes did not affect the erythroid differentiation, as assessed by flow cytometry analysis of the erythroid markers GYPA, CD36 and CD71 (FIGS. 5A-5C). Flow cytometry analysis showed an increased frequency of HbF-expressing cells (up to 87.8%) (FIG. 5D). RTqPCR analysis, in accordance with the flow cytometry data, revealed elevated production of γ-globin transcripts with a concomitant decrease of β-globin transcripts production (FIG. 5E). RP-HPLC and CE-HPLC analysis confirmed these data, with HbF representing up to 31.8% of the total Hb (FIGS. 5F-5G).

[0167] The use of evoCDA1-BE4max-NG and evoFERNY-BE4max-NG enzymes, combined with the LRF_bs_1 gRNA, to target the LRF repressor binding site, led to base editing efficiencies of up to 24.0% for evoCDA1-BE4max-NG enzyme (4.0% -201 C>T; 24.0% -197 C>T; 23.0% -196 C>T; 23.0% -195 C>T; 17.0% -194 C>T) and up to 20.0% for evoFERNY-BE4max-NG enzyme (20.0% -197 C>T; 20.0% -196 C>T; 5.0% -195 C>T; 1.0% -194 C>T) (FIGS. 3G, 3H and 3J), with a concomitant increase of the HbF-expressing cells to 12.9% (3.5% in the non-edited control cells) and 21.4% (6.7% in the non-edited control cells) respectively (FIG. 6A). We then used primary cord blood-derived CD34.sup.+ hematopoietic stem/progenitor cells (HSPCs). CD34+ cells were transfected with the AncBE4max_NAA, the LRF_bs_1 gRNA and the GFPmax plasmids and sorted for GFP expression. Cells were either maintained in liquid culture or plated in a semi-solid medium allowing the growth and differentiation of erythroid and granulocyte-monocyte progenitors (colony forming unit assay). 6 days post-transfection, we obtained an 19.0% efficiency of -200 C>T conversion in the liquid culture (FIGS. 6B-6C). 14 days post-transfection cation-exchange HPLC analysis was performed to detect hemoglobin tetramers in the burst-forming unit-erythroid (BFU-E) colonies derived from erythroid progenitors (FIG. 6D). This analysis revealed an increase in HbF level of 11.02% (68.22% in the control samples and 79.24% in the edited sample) (FIG. 6D). Altogether these data show that the novel AncBE4max_NAA base editing enzyme can target the LRF transcriptional repressor binding site in primary HSPCs and increase the HbF expression in their erythroid progeny.

InDels and Large Deletions Were Barely Detectable in Base-Edited Cells

[0168] One of the safety issues that emerges with the usage of CRISPR/Cas9 nuclease is the creation of Insertions and Deletions (InDels) in the genome after the generation of double strand breaks. Though, with the base editing system we can overcome this issue, as base editors contain an inactivated Cas9 nuclease. We wanted to verify that we do not create double strand breaks in the genome with the base editing enzymes that we employed. For this reason, we amplified the target regions by PCR, and performed Sanger sequencing followed by TIDE analysis.sup.(28) for base-edited and control K562 samples (FIGS. 2O, 3I and FIG. 6B). For almost all of the samples, we detected no InDels, except for cells transfected with ABE8e and KLF1_bs_1 gRNA plasmids, and evoCDA1-BE4max-NG and LRF_bs_2 gRNA plasmid, showing an average of 18.4% and 22.0% of InDels respectively (FIG. 2O).

[0169] Another issue that arises when editing the β-globin locus with the Cas9 nuclease, is the simultaneous cleavage of the identical HBG½ promoters resulting in the deletion of the intervening 4.9-kb genomic region and loss of the HBG2 gene. Therefore, we tested if the 4.9-kb deletion was present in base-edited K562 samples. In accordance with the InDel profile of base edited samples (FIG. 2O), we observed a low frequency of the 4.9-kb deletion only in ABEmax-, ABE8e- and evoFERNY-BE4max-NG-treated samples (4.9%, 4.9% and 3.2%, respectively; FIG. 2P). As a positive control for the 4.9-kb deletion, we used DNA extracted from K562 cells edited with the canonical Cas9 nuclease (FIG. 2P).

Enzymes With Low RNA Off-Target Activity Can Be Used to Target the HBG Repressors and Activators Binding Sites

[0170] Base editing enzymes can cause off-target editing of the cellular RNA that mostly is gRNA-independent. Mutations in the deaminase of the base editing enzyme can minimize RNA off-target editing. For the adenine base editors, these mutations are the E59A in the TadA domain and the V106W in the TadA* domain and the enzyme that carries these mutations is called ABEmaxAW.sup.(29). The mutations for the cytosine base editors are the R33A and the K34A in the APOBEC1 domain.sup.(30). By inserting these mutations in the AncBE4max enzyme, we created the AncBE4max_R33A/K34A base editing enzyme. Our purpose was to verify if these enzymes, with low RNA off-target editing, could be used to create the TAL1 and KLF1 activator binding sites or disrupt the BCL11A repressor binding site. Transfection of K562 cells with the ABEmaxAW plasmid and the TAL1_bs_1 or the KLF1_bs_1 gRNA plasmid led to -175 T>C (A>G to the opposite strand) and T>C (A>G to the opposite strand) conversion, with a frequency of 25.0% and 17.0%, respectively (FIGS. 7A and 7B). Targeting the BCL11A binding site with the ABEmaxAW in combination with the BCL11A_bs_1 gRNA plasmid, caused 36.0% -116 A>G and 33.0% -113 A>G conversions (FIG. 7C), while targeting this site with the same gRNA and the AncBE4max_R33A/K34A enzyme led to -115 C>T and -114 C>T conversions, with a frequency of 29.0% and 24.0%, respectively (FIG. 7D). Altogether, these data demonstrate that these safer versions of base editing enzymes can be used to efficiently target transcriptional activator or repressor binding sites in the HBG promoters.

Disruption of the LRF Repressor Binding Site in the HBG Promoters by CBEs in SCD HSPCs Leads to HbF Reactivation and Rescues the Sickling Phenotype

[0171] To prove the efficacy of our base editing approaches as therapeutic strategies for the treatment of SCD, we transfected base editor- and gRNA-expressing plasmids in primary human adult non-mobilized SCD HSPCs. In particular, plasmids expressing CBE enzymes (CBE-NRCH, CBE-SpG-GFP, CBE-SpRY-GFP) were individually transfected in combination with single gRNA-expressing plasmids (LRF_bs_1, LRF_bs_2). To enrich for edited cells, either we used plasmids that express base editor-GFP fusions (CBE-SpG-GFP, CBE-SpRY-GFP) or we co-transfected base editor- (CBE-NRCH) and GFPmax-expressing plasmids. GFP.sup.high cells were FACS-sorted differentiated toward the erythroid lineage using a 3-phase protocol.

[0172] Base editing efficiency was measured in erythroblasts at the end of the first phase of erythroid differentiation (Day 6). Samples treated with CBEs and LRF_bs_1 gRNA, converting 4C in the LRF binding site (LRF 4C), showed editing efficiencies of ~22.4% (26.8%, 23.8% and 16.5% with CBE-NRCH, CBE-SpG and CBE-SpRY respectively) (FIG. 8A). All the cytosines of the LRF binding site were converted into T in CBE-SpRY- and LRF_bs_2-treated samples (LRF 8C) with efficiencies of up to 25.5% (FIG. 8B). TIDE analysis in base-edited samples confirmed the absence of InDels (FIG. 8C).

[0173] We then differentiated bulk populations of edited erythroblasts into mature RBCs to evaluate HbF expression and the recovery of the sickling cell phenotype. The erythroid differentiation was similar between control and CBE-treated samples, as measured by flow cytometry analysis of late (CD36, CD71 and α4-Integrin) and early (GYPA and BAND3) erythroid markers along the differentiation (data not shown). The enucleation rate was similar between groups at different time points throughout the differentiation and at the end of the last phase reached more than 90% in all samples (data not shown). LRF 4C and LRF 8C CBE-treated samples showed fetal hemoglobin reactivation at both at mRNA and protein level, as measured by RT-qPCR, RP-HPLC (FIGS. 8D and 8E) and CE-HPLC (22.3% and 9.1% HbF in edited samples and control samples, respectively; FIG. 8F). Flow cytometry analysis revealed a high frequency of HbF-expressing RBCs (42.9%, 64.3% and 70.0% in control, LRF 4C and LRF 8C samples, respectively; FIG. 8G). To evaluate the effect of HbF reactivation on the sickling phenotype, we incubated mature RBCs under hypoxic conditions inducing HbS polymerization. Interestingly, editing of either 4 or 8 C of the LRF binding site ameliorated the sickling phenotype (23.2%, 52.5%- and 58.4% of non-sickle cells in control, LRF 4C and LRF 8C samples respectively) (FIG. 8H). Overall, these data demonstrate that base editing of the HBG½ promoters by CBEs can lead to HbF reactivation and rescue the sickling phenotype of RBCs differentiated from SCD patient HSPCs.

Disruption of the LRF Repressor Binding Site or Creation of the KLF1 Activator in the HBG Promoters by ABEs in SCD HSPCs Leads to HbF Reactivation and Rescues the Sickling Phenotype

[0174] ABEs can be also used in order to disrupt the LRF transcriptional repressor binding site or to create the KLF1 transcriptional activator binding site. We performed the same set of experiments described in the previous paragraph for CBEs in primary human adult non-mobilized SCD HSPCs to demonstrate ABEs′ therapeutical potential. More specifically, plasmids expressing ABEmax-GFP or ABE8e were individually transfected in combination with single gRNA-expressing plasmids (KLF1_bs_1 gRNA or AAVS1 gRNA targeting the unrelated AAVS1 locus; Weber et al., Sc. Advances, 2020). To enrich for edited cells, we used a plasmid expressing the ABEmax-GFP fusion protein or we co-transfected ABE8e- and GFPmax-expressing plasmids. After transfection, GFP.sup.medium and GFP.sup.high cells were FACS-sorted to obtain cell populations with a variety of editing efficiencies. Sorted cells were fully differentiated into mature RBCs using a 3-phase protocol.

[0175] Base editing efficiency was measured in erythroblasts at the end of the first phase of erythroid differentiation (Day 6). ABEmax (generating a KLF1 binding site, KLF1)- and ABE8e (converting the 2 T of the LRF binding site; LRF 2T) treated samples showed efficiencies that ranged from 41.0% to 52.3%, and 56.5% to 76.0% in the GFP.sup.medium and GFP.sup.high bulk populations, respectively (FIGS. 9A-9D). TIDE analysis in the base edited samples confirmed absence of InDels for ABEmax-treated cells (FIG. 9E), while a moderate InDel frequency of 7.9% and 14.8% was detected in GFP.sup.medium and GFP.sup.high ABE8e-treated samples, respectively (FIG. 9E).

[0176] Differentiation of bulk populations of edited erythroblasts into mature RBCs was performed to evaluate HbF expression and recovery of the sickling cell phenotype. The erythroid differentiation was similar between control and ABE-treated samples, as measured by flow cytometry analysis of late (CD36, CD71 and α4-Integrin) and early (GYPA and BAND3) erythroid markers along the differentiation (data not shown). The enucleation rate was similar between groups at different time points throughout the differentiation and at the end of the last phase reached more than 90% in all samples (data not shown). ABE-treated samples, bearing either the KLF1 binding site or the LRF 2T profile, expressed high HbF levels (66.3% and 62.6% respectively), as measured by CE-HPLC (FIG. 9H). These results were confirmed by RT-qPCR and RP-HPLC at mRNA and single globin chain levels (FIGS. 9F and 9G). Flow cytometry analysis revealed a high frequency of HbF-expressing RBCs (60.4%, ≥94.2% and ≥81.4% in control, ABEmax- and ABE8e-treated samples, respectively) (FIG. 9J). A sickling assay was performed in control and edited samples. High frequencies of corrected cells were observed for ABEmax- and ABE8e-treated samples (14.7%, 75.7% and 60.0% of non-sickle cells in control, ABEmax- and ABE8e-treated samples, respectively) (FIG. 9I). Overall, this study shows that either disrupting the LRF repressor binding site or creating the KLF1 activator binding site in the -200 region of the HBG½ promoters using ABEs leads to fetal hemoglobin reactivation and rescues the sickling phenotype in RBCs differentiated from SCD patient HSPCs.

Efficient RNA-Mediated Editing of the HBG½ Promoters in K562 Cells and SCD HSPCs

[0177] To establish a clinically relevant method to deliver the base editing system in primary HSPCs and achieve a high editing efficiency coupled with minimal toxicity, we optimized a protocol based on transfection of mRNA encoding base editors and synthetic modified gRNAs. First, we optimised the plasmids encoding CBE-SpRY and ABE-SpRY for in vitro transcription and mRNA production. In particular, we inserted two copies of the 3′ untranslated region (UTR) of the HBB gene (which has been shown to increase the half-life of mRNA and improve protein levels.sup.31-33) and a poly-A sequence after the 3′ UTR to further stabilize the mRNA.sup.34 in CBE-SpRY and ABE-SpRY constructs.

[0178] Next, we performed in vitro mRNA transcription using CBE-SpRY-OPT, ABE-SpRY-OPT and ABE8e plasmids. In K562 cells, transfection of CBE-SpRY, ABE-SpRY and ABE8e mRNAs together with LRF_bs_2, KLF1_bs_1 or BCL11A_bs_1 synthetic modified gRNAs led to high base editing efficiencies, demonstrating that we generated fully functional mRNAs. In particular, transfection of CBE-SpRY mRNA in combination with LRF_bs_2 or BCL11A_bs_1 gRNA resulted in 87.0% and 83.0% of C>T conversion, respectively (FIGS. 10A and 10B). Similarly, transfection of ABE-SpRY mRNA and KLF1_bs_ or BCL11A_bs_1 gRNA resulted in 55.0% and 39.0% of A>G conversion, respectively. Finally, co-transfection of ABE8e mRNA and BCL11A_bs_1 gRNA led to 88.0% of base editing efficiency (FIGS. 10C-10E).

[0179] CBE-SpRY and ABE8e mRNAs were transfected also in SCD HSPCs in combination with chemically modified single gRNAs. CBE-SpRY mRNA coupled with LRF_bs_1 or LRF_bs_2 gRNA led to 51.0% and 61.0% of C>T conversion respectively, while ABE8e mRNA coupled with KLF1_bs_1 gRNA led to 75.0% A>G conversion (FIGS. 10F-10H). These results demonstrate that RNA-mediated delivery of base editors allows efficient targeting of the HBG½ promoters in SCD HSPCs.

REFERENCES

[0180] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.

TABLE-US-00022 1 Taher, Weatherall, and Cappellini, ‘Thalassaemia’. 2 Kato et al., ‘Sickle Cell Disease’. 3 Chandrakasan and Malik, ‘Gene Therapy for Hemoglobinopathies’. 4 Cavazzana, Antoniani, and Miccio, ‘Gene Therapy for β-Hemoglobinopathies’. 5 Forget, ‘Molecular Basis of Hereditary Persistence of Fetal Hemoglobin’. 6 Wienert et al., ‘Editing the Genome to Introduce a Beneficial Naturally Occurring Mutation Associated with Increased Fetal Globin’. 7 Wienert et al., ‘KLF1 Drives the Expression of Fetal Hemoglobin in British HPFH’. 8 Martyn et al., ‘A Natural Regulatory Mutation in the Proximal Promoter Elevates Fetal Globin Expression by Creating a de Novo GATA1 Site’, 1. 9 Martyn, Quinlan, and Crossley, ‘The Regulation of Human Globin Promoters by CCAAT Box Elements and the Recruitment of NF-Y′. 10 Antoniani et al., ‘Induction of Fetal Hemoglobin Synthesis by CRISPR/Cas9-Mediated Editing of the Human β-Globin Locus’. 11 Weber, Frati et al., ‘Editing a γ-Globin Repressor Binding Site Restores Fetal Hemoglobin Synthesis and Corrects the Sickle Cell Disease Phenotype’. 12 Truong et al., ‘Microhomology-Mediated End Joining and Homologous Recombination Share the Initial End Resection Step to Repair DNA Double-Strand Breaks in Mammalian Cells’. 13 Milyavsky et al., ‘A Distinctive DNA Damage Response in Human Hematopoietic Stem Cells Reveals an Apoptosis-Independent Role for P53 in Self-Renewal’. 14 Cromer et al., ‘Global Transcriptional Response to CRISPR/Cas9-AAV6-Based Genome Editing in CD34+ Hematopoietic Stem and Progenitor Cells’. 15 Haapaniemi et al., ‘CRISPR-Cas9 Genome Editing Induces a P53-Mediated DNA Damage Response’. 16 Kosicki, Tomberg, and Bradley, ‘Repair of Double-Strand Breaks Induced by CRISPR-Cas9 Leads to Large Deletions and Complex Rearrangements’. 17 Gaudelli et al., ‘Programmable Base Editing of A.Math.T to G.Math.C in Genomic DNA without DNA Cleavage’. 18 Rees and Liu, ‘Base editing: precision chemistry on the genome and transcriptome of living cells’. 19 Yeh et al., ‘In Vivo Base Editing of Post-Mitotic Sensory Cells’. 20 Masuda et al., ‘Transcription Factors LRF and BCL11A Independently Repress Expression of Fetal Hemoglobin’. 21 Koblan et al., ‘Improving Cytidine and Adenine Base Editors by Expression Optimization and Ancestral Reconstruction’. 22 Li et al., ‘Reactivation of γ-Globin in Adult β-YAC Mice after Ex Vivo and in Vivo Hematopoietic Stem Cell Genome Editing’. 23 Behera et al., ‘Exploiting Genetic Variation to Uncover Rules of Transcription Factor Binding and Chromatin Accessibility’. 24 Brinkman et al., ‘Easy Quantitative Assessment of Genome Editing by Sequence Trace Decomposition’ . 25 Kurita et al., ‘Establishment of Immortalized Human Erythroid Progenitor Cell Lines Able to Produce Enucleated Red Blood Cells’. 26 Nettling et al., ‘DiffLogo: a comparative visualization of sequence motifs’. 27 Kluesner et al., ‘EditR: A Method to Quantify Base Editing from Sanger Sequencing’. 28 Brinkman et al., ‘Easy Quantitative Assessment of Genome Editing by Sequence Trace Decomposition’. 29 Rees et al., ‘Analysis and minimization of cellular RNA editing by DNA adenine base editors’. 30 Grünewald et al., ‘Transcriptome-wide off-target RNA editing induced by CRISPR-guided DNA base editors’ . 31 Karikó, K., Kuo, A. & Barnathan, E. Overexpression of urokinase receptor in mammalian cells following administration of the in vitro transcribed encoding mRNA. Gene Ther. 6, 1092-1100 (1999). 32 Ross, J. & Sullivan, T. D. Half-lives of beta and gamma globin messenger RNAs and of protein synthetic capacity in cultured human reticulocytes. Blood 66, 1149-1154 (1985). 33 Holtkamp, S. et al. Modification of antigen-encoding RNA increases stability, translational efficacy, and T-cell stimulatory capacity of dendritic cells. Blood 108, 4009-4017 (2006). 34 Gallie, D. R. The cap and poly(A) tail function synergistically to regulate mRNA translational efficiency. Genes Dev. 5, 2108-2116 (1991).