Introduction: The Vel– phenotype is a rare blood group, and it is challenging for identifying this phenotype due to limited available reagents. Moreover, there are relatively few studies on genomic editing of erythroid antigens and generation of knockout (KO) cell lines at present. Methods: To identify the high-efficiency small-guiding RNA (sgRNA) sequence, candidate sgRNAs were transfected into HEK 293T cells and analyzed using Sanger sequencing. Following this, the high-efficiency sgRNA was transfected into K562 cells using lentivirus transduction to generate KO Vel blood group gene cells. The expression of the Vel protein was detected using Western blot on single-cell clones. Additionally, flow cytometry was used to detect the erythroid markers CD235a and CD71. Hemoglobin quantification and Giemsa staining were also performed to evaluate the erythroid differentiation of KO clones induced by hemin. Results: The high-efficiency sgRNA was successfully obtained and used for CRISPR-Cas9 editing in K562 cells. After limiting dilution and screening, two KO clones had either deleted 2 or 4 bases and showed no expression of the Vel protein. In the hemin-induced KO clone, there was a significant difference in erythroid marker and hemoglobin quantification compared to untreated cells. The morphological changes were also observed for the hemin-induced KO clone. Conclusion: In this study, a highly efficient sgRNA was screened out and used to generate Vel erythroid antigen KO single-cell clones in K562 cells. The edited cells could then be induced to undergo erythroid differentiation with the use of hemin.
The Vel blood group antigen is a single-pass membrane protein with 78 amino acids encoded by gene small integral membrane protein 1 (SMIM1). The SMIM1 gene consists of four exons with exons 3 and 4 coding functional protein. Evidence presently shows that regulating the expression of the gene depends on genetic signatures, such as the homozygous or heterozygous 17-base pair (bp) deletion and other single nucleotide polymorphisms. A homozygous 17-bp deletion in exon 3 is responsible for a truncation during mRNA translation resulting in a Vel negative (Vel−) phenotype [1‒3].
Some Vel− individuals produce alloantibody through transfusion, which could lyse Vel+ erythrocytes causing hemolytic transfusion reactions [4, 5]. The frequency of Vel− individuals with the 17-bp deletion has been shown in Caucasians (0.025%) and in Brazilians (0.021%) [6, 7]. In several studies in China, researchers identified 2 individuals with Vel− phenotype out of 6,153 donors in Shanghai . One individual out of 9,122 donors in Jiangsu and 14 individuals out of 3,328 donors in Xinjiang were identified with a heterozygous c.64_80del allele [9, 10]. Although the presence of the Vel antigen has important clinical implications, the availability of anti-Vel antibody reagents and red blood cells (RBCs) from Vel− individuals is currently limited.
For enhancement transfusion compatibility, CRISPR/Cas9-mediated (clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) genomic editing has been reported to modify the blood group genes [11, 12]. According to previous studies, we could generate an SMIM1 gene knockout (KO) cell line, resulting in a Vel− phenotype due to the characteristic of the Vel protein with a single-pass membrane and the molecular mechanism of Vel− phenotype for allele deletion.
In this study, we present an alternative approach for generating erythroid antigen KO clones through CRISPR/Cas9 editing of the K562 cell line. Given the low transfection efficiency of K562 cells, we propose a method that involves using HEK 293T cells to screen for highly efficient editing small-guiding RNA (sgRNA) sequences, which can then be transfected into K562 cells to generate KO clones (shown in Fig. 1). Here, we initially demonstrate the feasibility of generating cell lines completely deficient in the Vel antigen in K562 cells using CRISPR/Cas9-mediated genome editing. The cells that lack of the Vel blood group gene possess erythroid differentiation abilities.
Materials and Methods
Cells and Erythroid Differentiation
K562 cells were cultured in the RPMI 1640 medium with 10% fetal bovine serum (FBS) at 37°C and 5% CO2. HEK 293T cells were cultured in the DMEM medium with 10% FBS at the same cultural condition. Hemin (Sangon Biotech) was prepared with dimethyl sulfoxide at 10 mm concentration and diluted in culture medium to the required test concentration. K562 cells were incubated with hemin in gradient concentration (0 µm, 30 µm, 40 µm, 50 µm, and 60 µm) for 48, 72, and 96 h. In each separate assay, after being treated for 48 h, the medium was replaced with fresh medium containing hemin.
High-Efficient sgRNA Screening
The coding sequence of the SMIM1 gene was obtained from the NCBI (National Center for Biotechnology Information) database. Subsequently, five sgRNA sequences were designed targeting exon 3 of the SMIM1 gene using an online sgRNA website (http://crispr.dfci.harvard.edu) (shown in online suppl. Fig. 1; Table 1; for all online suppl. material, see https://doi.org/10.1159/000534012;). The sgRNA sequences were synthesized by Azenta Life Sciences and subsequently cloned into the PX458 vector using T4 DNA ligase (Beyotime, China) at the BbsI restriction enzyme site, targeting the SMIM1 gene and expressing green fluorescent protein (GFP). The resulting ligation products were transformed into DHα competent Escherichia coli cells. After transformation and propagation, the plasmid DNA was extracted using the EndoFree Maxi Plasmid Kit V2 (TIANGEN, China) and subsequently sequenced by BGI Genomics company.
Five sgRNA screening plasmids were then transfected into HEK 293T cells in equal amounts separately. The transfection was performed in a 24-well plate using Xfect transfection reagent (TaKaRa, Japan) when the cell density reached 70–80%. For the transfection, 1 μg of plasmid was mixed with 25 μL of Xfect reaction buffer and 0.3 μL of Xfect polymer. The mixture was incubated for 10 min at room temperature and then added to the cells. After 4 h, the cells were replaced with fresh medium. The same method of transfection was utilized to compare the efficiency of plasmid PX458 transfection in HEK 293T and K562 cells.
After 48 h of transfection, genomic DNA was extracted from the cells using the TGuide cells/tissue genomic DNA kit (TIANGEN). DNA fragments including exon 3 and exon 4 as coding sequences were amplified using the PCR primers previously published . The final reaction volume was 50 μL containing 100 ng of genomic DNA, 0.8 μmol primers, 2.5 mmol of each dNTP, 2.5 U of DNA polymerase (Hot Start Version, TaKaRa LA Taq) with the buffer supplied, and sterile water. The cycling conditions for amplification were as follows: initial denaturation at 94°C for 5 min, 30 cycles at 94°C for 40 s, 60°C for 50 s, 72°C for 30 s, and a final extension step for 7 min at 72°C. Sanger sequencing was detected by BGI Genomics Company. Sequencing results were aligned using SnapGene software and the insertion and deletion (indel) were analyzed using the Synthego website (https://ice.synthego.com).
The sgRNA sequence with high editing efficiency was cloned into the vector lentiCRISPRv2 expressing Cas9 protein obtaining from Addgene (USA). Plasmid construction was performed using the methods mentioned above. A second-generation packaging system with plasmid psPAX2 encoding the structural protein and plasmid pMD2.G encoding the protein coat of the lentivirus was utilized to package the lentivirus. HEK 293T cells were cultured in 10 cm dishes the day prior to experimentation. When the cell density reached 70–80%, the Xfect transfection reagent was used to co-transfect 12 µg of lentiCRISPRv2 with sgRNA, 6 µg of psPAX2, and 3 µg of pMD2.G. After 72 h of transfection, the virus supernatant was collected and filtered through a 0.45 μm filter (Millex-HV). Subsequently, the virus supernatant was centrifuged at 25,000 rpm for 2 h at 4°C and then resuspended in RPMI 1640 medium with the appropriate volume. The lentiviral titer test card (Biodragon, China) was used to estimate the virus titer.
K562 cells were transduced with lentivirus at a multiplicity of infection value of 20 using polybrene (final concentration 8 μg/mL) from Yeasen Biotech, China. The cells were then cultured in a 12-well plate and centrifuged at room temperature at 900 × g for 30 min to enhance transduction efficiency. After 48 h, cells were cultured with puromycin (concentration 1 μg/mL) (Yeasen Biotech, China) for selection.
Single-Cell Clone Screening
Single-cell cloning and sorting were performed by SONY SH800 Cell Sorter in 96-well plate. Clones were detectable by microscopy after 2 to 3 days to check and mark each well that contains just a single colony. These colonies were transferred into 24-well plates for expansion.
Two different RBC samples, carrying the VEL*01/VEL*01 and VEL*01/VEL*01N.01 genotypes, were screened out from blood donors using an in-house method and used as controls in WB experiments. The genotypes of the two samples were confirmed by Sanger sequencing. RBC ghosts were prepared as previously published . 1 × 106 of KO clones were washed 2 times with phosphate-buffered solution (PBS). 100 µL of radioimmunoprecipitation assay (RIPA) buffer mixed with 1 mm of phenylmethanesulfonyl fluoride (Beyotime, China) were added into KO clones incubating for 30 min on ice to lysis cells. Protein concentration was measured by Enhanced BCA protein assay kit (Beyotime, China). In this study, 5 µg of denatured protein from RBCs and 15 µg of denatured protein from KO clones were separated using a 15% sodium dodecyl sulfate-polyacrylamide gel. Following transfer to a PVDF membrane (Immobilon-E, Merck Millipore Ltd), the membrane was blocked and shaken in a 3% bovine serum albumin solution for 1 h. The membrane was incubated with the rabbit antibody against SMIM1 (ProteinTech, Cat.: 27849-1-AP) at 4°C overnight. Then it was washed in PBS with Tween 20 (PBST) and incubated with horseradish peroxidase (HRP) conjugated to anti-rabbit immunoglobulin G (ProteinTech, Cat.: SA00001-2) for 1 h. After washing with PBST, bands were visualized by Super ECL detection reagent (Yeasen, China). Subsequently, images were obtained using the ChemiDoc MP Imaging System (Bio-Rad). Finally, the bands of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were incubated with primary antibody (HUABIO, Cat.: M1310-2) and secondary antibody labeled HRP (Beyotime, Cat.: A0216). Bands of β-actin were incubated with primary antibody (Beyotime, Cat.: AF5003) and secondary antibody labeled HRP (Proteintech, Cat.: SA00001-2).
1 × 106 cells were washed with PBS and 2% of FBS and were incubated with antibodies for 30 minutes on ice. After washing twice, samples were detected on CytoFLEX LX (Beckman Coulter) and analyzed with CytExpert software. The following antibodies were used: PE anti-human CD235a (Glycophorin A) (Santa Cruz) and APC anti-human CD71 (transferrin receptor) (BD Biosciences).
For hemoglobin quantification, 8 × 105 of K562 wild-type cells and KO clone 1 were harvested and washed twice with PBS. Cell lysates were dissolved and centrifuged in 160 µL of deionized distilled water. The concentration of hemoglobin was determined and calculated using a QuantiChrom hemoglobin assay kit (BioAssay Systems) according to the manufacturer’s instructions. Hemoglobin quantification was obtained at 400 nm OD on a microplate reader (Spark™ 10M, TECAN, Switzerland).
K562 cells were treated with 50 µm hemin for 72 h. A 96-well plate was treated with Retronectin (TaKaRa) for 30 min at room temperature. Then, 1 × 105 cells were added into the 96-well plate and centrifuged at 2,500 rpm for 5 min. The cells adhered to the bottom of the plate were fixed with methanol for 10 min and stained with Giemsa (Sigma-Aldrich) for 5 min. The solution was removed and washed 5 times with distilled deionized water. The stained cells were examined and captured using BioTek Cytation7.
Experimental data are presented in mean ± SD as described in the figure legends. Data were analyzed by multiple t tests using GraphPad software. p < 0.05 was considered statistically significant. Asterisks are used to indicate significance: ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. NS, not significant.
Selection of sgRNA with High Editing Efficiency
To investigate the transfection efficiency, plasmids expressing GFP with sgRNA sequences were transfected into K562 and HEK 293T cells. Flow cytometry analysis revealed a significantly higher positive rate of GFP in HEK 293T cells compared to K562 cells (shown in Fig. 2a). Therefore, HEK 293T cells were chosen for screening sgRNA sequences with high editing efficiency. The sgRNA sequences were then cloned into vector PX458 and transfected into HEK 293T cells. The flow cytometry analysis (shown in online suppl. Fig. 2) detected the expression of GFP, while the results of indel allele are analyzed in Table 1. The findings revealed that sgRNA sequences exhibited varying editing efficiencies despite having similar transfection efficiency. Thus, HEK 293T cells could serve as a useful screening tool for selecting efficient sgRNA sequences, thereby reducing both time and costs.
GFP%: the transfection efficiency was determined by the GFP level.
Indel%: the editing efficiency was evaluated using indel analysis.
In this study, sgRNA4 (5′-CGTCCTCCCACCTACTATAG-3′) was chosen for further experiments based on its editing efficiency. Efficient targeting of SMIM1 gene in K562 cells was achieved through transduction with lentivirus containing Cas9 protein and sgRNA4. After selecting with puromycin, genomic DNA was extracted from the CRISPR/Cas9-mediated K562 cells. The fragment of exon 3 in SMIM1 gene was sequenced, resulting in a multiple KO at the sgRNA cutting site (shown in Fig. 2b). Simultaneously, exon 4 sequences did not exhibit mutation (shown in online suppl. Fig. 3). The finding suggested that CRISPR/Cas9-mediated genome editing successfully modified the SMIM1 gene encoding Vel blood group and did not affect other coding sequences.
Generation of Single KO (SMIM1 KO) Cell Clones
To obtain a single-cell clone with KO SMIM1 gene, the CRISPR/Cas9-mediated K562 cells were sorted. Thirty-three of these clones were detectable by microscopy and were subsequently expanded. DNA sequencing was performed to confirm the presence of indels. After analysis, some clones did not exhibit any insertions or deletions of bases, while others showed deletions that were multiples of three or were non-single-cell clones. Interestingly, two clones displayed a single peak in chromatograph analysis and demonstrated the deletion of either two or four bases, ultimately leading to homozygous frameshift deletions (shown in Fig. 3a). The alignment of deduced amino acid sequences from the two clones with wild-type Vel revealed a substitution of amino acid 10 with a stop codon, resulting in a truncated protein sequence (shown in Fig. 3b). Based on this finding, we hypothesized that KO clone 1 and KO clone 2 may cause inactivation of the Vel protein function.
Expression of Vel Proteins in KO SMIM1 Cell Lines
After editing the SMIM1 gene in a molecular level, the effect on the protein KO was evaluated by WB. To evaluate the specificity of the antibody and the consistency of Vel antigen expression in K562 cells with erythrocytes, we conducted pretests on Vel+ phenotype samples using homozygous VEL*01/VEL*01 blood samples and blood samples carrying heterozygous c.64_80del alleles (VEL*01/VEL*01N.01), which were then subjected to WB analysis. The results indicated that the antibody was able to specifically detect the erythrocyte antigen Vel, and the protein expression level of the antigen in heterozygous samples was found to be weaker (shown in Fig. 4a). A significant difference was found between K562 WT and two single KO SMIM1 clones. In brief, a normal Vel protein expression level was observed for K562 WT. However, the results of two KO clones edited by CRISPR/Cas9 system showed no Vel protein expression (shown in Fig. 4b). This result illustrated the feasibility to KO the SMIM1 gene encoding Vel protein in the K562 cell line by CRISPR/Cas9-mediated genome editing.
Evaluation of the Erythroid Differentiation Ability
To evaluate the potential of K562 KO cell lines to differentiate into erythroid, we treated KO-1 with hemin and measured the expression of CD235a and CD71, hemoglobin content, as well as morphological changes. After exposing to varying concentrations of hemin for a period of 48–96 h, the expression of CD235a was observed to increase while the expression of CD71 decreased (shown in Fig. 5a, b). KO-1 treated with 50 μm hemin exhibited significant changes in hemoglobin levels as compared to untreated KO-1 clones. Meanwhile, KO-1 clones which knocked out the SMIM1 gene showed similar hemoglobin concentrations to that of unedited K562 parallel control cells (shown in Fig. 5c). We examined the morphology of unedited K562 cells and KO-1 clones treated with 0 µm or 50 µm hemin for 72 h. Our results showed that KO-1 clones treated with 50 μm hemin exhibited morphological changes and nuclear features consistent with erythroid differentiation. These changes were similar to those observed in unedited K562 cells, as confirmed by Giemsa staining (shown in Fig. 6). These results suggest that K562 cells that have undergone CRISPR-Cas9 genomic editing can induce erythroid differentiation in response to chemical agents like hemin.
The concept of modifying the expression of blood group antigens to improve the transfusion compatibility of RBCs is not new. Converted ABO blood groups have been reported by using glycosidases to mask antigens . In addition, reduction of the expression of blood group antigen in adult hematopoietic stem cells by lentiviral vector-mediated shRNA is possible. However, the knockdown is incomplete and repeated shRNA transduction is required to obtain a number of modified cells . Genomic editing technology has led to the use of TALENs (transcription activator-like effector nucleases) to KO the RHD gene and convert RhD-positive cells into RhD-negative cells . However, the application of TALENs technology is dependent on the recognition between protein and DNA sequences. Rebuilding protein modules to target different DNA sequences is a complex process . Recently, researchers at the University of Bristol have used the CRISPR/Cas9 system to KO the common blood group antigens: ABO, Rh, Kell, Duffy, and GPB, which lead to transfusion incompatibility in erythroid cells to reconstitute the phenotype of blood group antigens .
The CRISPR/Cas9 system has been widely used in genome editing of plants and animals due to its simple operation and high editing efficiency. Only 20 bp of the sgRNA sequence target different gene sites, depending on the base recognition. Once Cas9 cleaves double-stranded DNA, and the cell activates the DNA repair mechanism of nonhomologous end joining, which inserts or deletes allele fragments to change the reading frame leading to gene KO . Although the CRISPR/Cas9 system results in gene KO and site-specific modification, the efficiency based on homologous recombination mechanism is low. Furthermore, single nucleotide mutations are common in rare blood groups. Therefore, we can also modify the blood group antigen phenotypes using base editing, which can convert bps independent of DNA double-strand breakage [19, 20]. It is expected to establish cell models of rare blood group antigens caused by single nucleotide polymorphisms with greatly efficiency of precise genomic editing. However, there have been limited studies on genome editing of the erythroid antigen gene. In the field of hematology, editing technologies have been primarily utilized in the treatment of diseases, such as sickle cell disease and β-thalassemia [21, 22].
In this study, we obtain two single-cell KO clones of SMIM1 gene with 2-bp or 4-bp deletion leading to a truncation during mRNA translation. Fortunately, the Vel protein expression determined by Western blot is consistent with the molecular characteristic. Therefore, single-cell clones deficient in Vel blood group antigen was established through editing the SMIMI1 gene using K562 cells. In our study, we were able to show that the SMIM1 KO cell line expressed the erythroid marker and synthesized hemoglobin in response to hemin treatment. Furthermore, we observed morphological changes. Recently, it has been reported that knocking down the SMIM1 gene in zebrafish leads to a slight decrease in the average hemoglobin content of RBCs . However, our study aimed to verify whether the KO cells could undergo erythroid differentiation by measuring the relative change of hemoglobin concentration. Although our results did not show a mild reduction compared to previous studies , this discrepancy may be attributed to variations in species and cell lines. And recently, we did not find a correlation between hemoglobin reduction and the function of the SMIM1 gene in humans. In addition, SMIM1 is a potential diagnostic marker for intervertebral disc degeneration through bioinformatics and machine learning methods . The cell model with Vel− phenotype could be used in biochemical functions, physiological implications, hematological studies, and applications due to the rarity of Vel− individuals.
In this study, a method for editing RBC antigens on K562 cells was established. The resulting Vel KO clones can serve as negative control cells for identifying high-frequency antibodies or for further in vitro functional studies. Previous reports of erythroid genomic editing have primarily been performed using immortalized adult erythroid line (Bristol Erythroid Line Adult; BEL-A) and human umbilical cord blood-derived erythroid progenitor cell lines, which are more complex and expensive [24, 25]. K562 cells, on the other hand, are widely used in erythropoiesis research, can be induced to differentiate into erythroid cells by inducers, and are easy to obtain and culture [26, 27]. Therefore, our proposed method for screening high-efficiency sgRNAs and performing erythroid genomic editing in K562 cells offers a cheaper and more convenient alternative.
While our study provides valuable insights, there are some limitations that must be acknowledged. First, detection of Vel antigen on the surface of live KO clones did not performed due to lack of serological or flow cytometry antibodies. Additionally, there were no clinical samples available to verify its application. Second, the erythroid-induced differentiation of the KO clones in this experiment was not sufficient. Further optimization may be necessary, either by optimizing experimental conditions or by using other induction reagents such as cabozantinib . Lastly, it is important to note that the K562 cell line, being an erythroleukemia cell line, has a slightly different natural expression of blood group antigens compared to erythrocytes. As a result, its application for erythroid editing may be limited.
In conclusion, we have demonstrated the feasibility of creating a K562 cell line that lacks the Vel antigen through CRISPR/Cas9-mediated blood group genomic editing. Our results provide an alternative choice for the development of diagnostic reagents with complex matching requirements and for the generation of customizable RBCs for individuals who lack the Vel antigen in the future.
Statement of Ethics
The materials and cell lines used in this study were obtained from the supplier. Ethical approval for the use of this is not required in accordance with local guidelines.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
This study was financially supported by the National Natural Science Foundation of China (grant no. 81970168).
Jiaxuan Yang designed the study, performed the assays, analyzed the data, and wrote the paper. Aijing Li contributed to Sanger sequencing. Minghao Li contributed to Western blot. Shulin Ruan contributed to genomic DNA extraction. Luyi Ye designed the study and revised the manuscript. All authors reviewed the data, provided comments, and approved the final manuscript.
Data Availability Statement
All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.