The BD211 autologous CD34⁺ hematopoietic stem cell injection, non-transduced stem cell, and its cryopreservation solution were sourced from BDGENE Therapeutics Co. Ltd. Human CD34⁺ hematopoietic stem cells were isolated from the peripheral blood of healthy donors following Filgrastim (Recombinant Human Granulocyte Colony-stimulating Factor) mobilization. On the day of use, the BD211 cells and non-transduced stem cells were thawed and maintained on ice until transplantation, with the entire transplantation procedure completed within 4 h after cell thawing. Viability was assessed using the trypan blue exclusion assay, with 78%–90% of the cells demonstrating viability. The hematopoietic differentiation potential and target gene expression of BD211 were confirmed in vitro through flow cytometry and PCR analysis (Figure 2).

Specific pathogen-free (SPF) grade NCG-X mice, evenly categorized between males and females and aged 7–9 weeks, were procured from Jiangsu Jicui Yaokang Biotechnology Co., Ltd. (GemPharmatech Co., Ltd., Nanjing, China). The mice were housed in individually ventilated cages, with the animal room temperature maintained at 22.0°C–26.0°C and relative humidity controlled at 40%–70%. The central air conditioning system provided ventilation at a rate of ≥15 times per hour, and the lighting was set to a 12-h light/12-h dark cycle. Cage ventilation was set at 30–80 times per hour. The facilities used in this study have been accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. All experimental animals were bred and used for scientific research purposes. All procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Guoke Saifu Hebei Pharmaceutical Technology Co., Ltd. (approval numbers: IACUC-2023-149 and IACUC-2023-220).

This study is part of the preclinical safety evaluation of the BD211 autologous CD34⁺ hematopoietic stem cell injection for treating TDT. The study design adhered to ICH guidelines M3 (R2) and relevant technical guidelines (M3R2 I, 2009; National Medical Products Administration, 2017; National Medical Products Administration, 2021).

Overall, 222 SPF-grade NCG-X mice, aged 7–9 weeks, were randomly assigned to four groups, each with an equal number of males and females. The groups included a vehicle control group (n = 40) that received 0.4 mL of Biolife Solutions CS10 cell cryopreservation solution, a mock control group (n = 52) that received 0.4 mL of non-transduced CD34⁺ hematopoietic stem cells (3 × 10⁶ cells/mL), and BD211 low-dose (n = 52) and high-dose groups (n = 78) that received 0.13 mL and 0.4 mL of BD211 (3 × 10⁶ cells/mL), respectively. The administered doses were 0, 1.2 × 10⁶, 4.0 × 10⁵, and 1.2 × 10⁶ cells per mouse, equivalent to 0, 6.0 × 10⁷, 2.0 × 10⁷, and 6.0 × 10⁷ cells/kg, respectively. Based on the results of previous studies, the low dose was considered the effective dose, while the high dose was anticipated to induce adverse reactions. Mice were administered a single intravenous injection via the tail vein. The first day of administration was designated as Day 1 (D1), and the mice were observed continuously for 13 weeks (D92).

Evaluations were conducted at multiple time points to assess various parameters. Throughout the experimental period, daily observations were made of the general condition of the mice. Body weight and food intake were measured weekly, and ophthalmological examinations were conducted on D0 and D90. On D2 (24 h post-administration), D57, and D92, assessments focused on in vivo erythroid differentiation of stem cells, including the differentiation efficiency of hCD235a⁺ and hCD71⁺ cells, along with the expression levels of the β^A−T87Q-globin target gene. Additionally, a more comprehensive series of analyses were performed on D2, D29, D57, and D92, encompassing stem cell stemness through lymphocyte and lymphocyte subset profiling, measurement of reconstitution efficiency in mouse, and cellular differentiation ratios in peripheral blood, spleen, and bone marrow. The analyses also included cytokine level analysis (IFN-γ, TNF-α, IL-2, IL-4, IL-6, and IL-10), tissue distribution (BD211 high-dose group only), vector copy number (VCN) quantification, and toxicokinetic parameters, including BD211 levels based on Woodchuck Hepatitis Virus Post-transcriptional Regulatory Element (WPRE) gene concentration and expression analysis of both β-globin and β^A−T87Q-globin genes. The solvent and mock control groups were tested only on D2 and D29. On D92, hematological, blood biochemical, gross dissection, organ weighing, bone marrow smear examination, histopathological examination, and oncogenicity assessment were performed (Figure 1).

Manufacturer details for reagents and instruments are listed in Supplementary Tables S5, 6.

During the trial period, daily observations were conducted to record the appearance, dosing site, behavior, skin condition, respiration, mouth and nose, glandular secretion, fecal and urinary characteristics, and any instances of death. Body weight and food intake were measured weekly. Ophthalmic examinations were performed the day before administration and on D90, with a necropsy conducted on D92. Hematological analyses (including coagulation), biochemical analyses (including electrolytes), and histopathological examinations were performed. Gross dissections were carried out, and major organs were weighed.

Genomic DNA for VCN analysis was extracted from BD211-transduced CD34⁺ cells cultured for 14 days post-transduction, and whole blood and bone marrow samples were collected from NCG-X mice on D2, D29, D57, and D92 post-administration. Dual fluorescent quantitative PCR was performed using a real-time PCR instrument (Thermo Fisher Scientific, Quant Studio 5) to amplify the viral element WPRE and the human cellular reference gene RNase P using the primers and probe sequences provided in Supplementary Table S4. Thermocycling conditions were as follows: 37°C for 2 min, 95°C for 5 min, followed by 45 cycles of 95°C for 15 s and 60°C for 40 s. Fluorescence signals were monitored during the 60°C phase through the FAM (WPRE) and VIC (RNAseP) channels. WPRE-containing vector RNA undergoes reverse transcription during transduction and becomes integrated into the host genome. It thus serves as a molecular marker for lentiviral VCN quantification via duplex qPCR. Plasmid standards (Lenti-genome-Copies) containing WPRE (vector-specific) and RNAseP (endogenous control) sequences were serially diluted in Yeast tRNA solution with NCG mouse matrix blanks to generate standard curves (R² > 0.99). Experimental DNA samples were processed identically, with VCN determined by interpolating Ct values into the standard curve, enabling the normalization of vector copies to host genome equivalents. VCN was calculated using the formula: WPRE copy number×2/RNase P copy number.

The bone marrow cavities were subsequently flushed with DPBS utilizing a 1 mL syringe until the cavities appeared visibly devoid of residual marrow contents. The resultant cell suspension underwent filtration through a 70-µm cell strainer. A centrifugation process at 330 × g for 5 min at 4°C ensued to pellet the cells. The pelleted cells were subjected to treatment with red cell lysis buffer to eliminate red blood cells, followed by a washing step with DPBS. Subsequent experiments utilized the acquired bone marrow mononuclear cells.

Splenic mononuclear cell isolation involved the mechanical dissociation of spleens on a 70-µm strainer. The cell suspension was centrifuged at 330 g for 5 min at 4°C, after which red cell lysis and DPBS washing were performed to eradicate erythrocytes and debris.

Peripheral blood mononuclear cell isolation commenced with blood collection via retro-orbital puncture. Dual erythrocyte lysis cycles were implemented: blood samples were centrifuged at 400 g for 5 min at 4°C. The pellet was resuspended in red cell lysis buffer. This lysis procedure was repeated to ensure thorough removal of red blood cells. Ultimately, cells were washed with DPBS, and the resulting peripheral blood mononuclear cells were collected.

Single-cell suspensions of peripheral blood, spleen, and bone marrow were prepared as described in method 2.6. Cells were adjusted to 1 × 10⁶ cells/mL in DPBS. For staining, 100 µL aliquots were incubated with 100 µL Zombie Aqua Fixable Viability Kit for 15 min at room temperature, protected from light. After DPBS washing (300 × g, 5 min), cells were resuspended in 100 µL DPBS and stained with pre-titrated antibody cocktails (Supplementary Table S5) for 30 min. Unbound antibodies were removed via two sequential DPBS washes (300 × g, 5 min). Final pellets were resuspended in 300 µL DPBS and stored at 4°C protected from light until acquisition. Representative gating schematics are provided in Supplementary Figures S1–4.

On D2, D57, and D92, femurs, tibias, and sternums were dissected to obtain bone marrow cells. Flow cytometry was used to assess the efficiency of erythroid maturation and differentiation by detecting hCD235a and hCD71 with the following gating strategy: mCD45⁻hCD45⁻hCD235a⁺CD71^- (mature erythrocytes), mCD45⁻hCD45⁻hCD71⁺CD235a⁻ (proerythroblasts), and mCD45⁻hCD45⁻hCD235a⁺CD71⁺ (intermediate erythroblasts).

On D2, D29, D57, and D92, peripheral blood, spleens, and bone marrow from femurs, tibias, and sternums were dissected. The analysis of myeloid and lymphoid cell differentiation cell populations was conducted as follows: T cells (hCD45⁺hCD3⁺), CD8⁺ T cells (hCD45⁺hCD3⁺hCD8⁺), CD4⁺ T cells (hCD45⁺hCD3⁺hCD4⁺), monocytes (hCD45⁺hCD13⁺), B cells (hCD45⁺hCD19⁺), and NK cells (hCD45⁺hCD56⁺). This analysis was conducted to evaluate myeloid and lymphoid cell differentiation. Additionally, hCD45 was measured in peripheral blood to calculate the peripheral blood chimerism rate.

All assays were performed using a DxFLEX cytometer, and data analysis was conducted using CytExpert for DxFLEX.

Frozen cells were rapidly thawed at 37°C and diluted into 5 mL IMDM medium, followed by centrifugation at 500 g for 5 min. Viable cells were resuspended in IMDM, adjusted to 1.1 × 10⁴ cells/mL using trypan blue exclusion counting, and serially diluted in pre-warmed MethoCult H4435 Enriched Medium at a 1:10 ratio to achieve a final density of 1 × 10³ cells/mL. After vigorous vortex mixing (2–5 min) and 10-min deaeration at room temperature, 1 mL aliquots were plated in triplicate onto 6-well plates. Plates were gently tilted to ensure even distribution and incubated for 14 days at 37°C/5% CO₂ under humidified conditions. Colony enumeration was performed manually using inverted microscopy, with morphological criteria distinguishing colony types: burst-forming unit erythroid (BFU-E), colony-forming unit granulocyte (CFU-G), macrophage (CFU-M), granulocyte-macrophage (CFU-GM), and multipotential granulocyte-erythrocyte-macrophage-megakaryocyte (CFU-GEMM) as defined by the manufacturer.

Serum cytokines were quantified using the V-PLEX^® Proinflammatory Panel 1 Kit on a QuickPlex SQ 120 platform. Samples (2-fold diluted in Diluent 41) underwent 2 h incubation at 25°C with orbital shaking (600 rpm). Post-washing (PBS-T, 3×), SULFO-TAG-conjugated detection antibodies were added (1 h incubation), followed by Read Buffer T application and electrochemiluminescence signal measurement. Analytes included IFN-γ (0.37–1,500 pg/mL), IL-2 (0.37–1,530 pg/mL), IL-4 (0.06–258 pg/mL), IL-6 (0.18–720 pg/mL), IL-10 (0.09–384 pg/mL), and TNF-α (0.09–362 pg/mL). Measurements below detection thresholds were excluded per manufacturer specifications.

Tissue sections (4 µm) were deparaffinized in xylene, rehydrated through graded ethanol, and stained with Mayer’s hematoxylin (10 min). After differentiation (1% acid ethanol, 1 s) and bluing (tap water, 7 min), cytoplasmic counterstaining was performed with eosin Y (5 min). Sections were dehydrated, cleared in xylene, and mounted with neutral resin for brightfield microscopy.

Mice were euthanized on D2, D29, D57, and D92 following the injection. Blood (with EDTA anticoagulant) and tissues, including bone marrow, spinal cord, brain, lungs, heart, liver, spleen, kidneys, skeletal muscle, joints, stomach, duodenum, jejunum, colon, epididymis, ovaries, or whole blood, were collected from each mouse. Samples were temporarily stored on ice and quickly transferred to conditions below −70°C for tissue distribution analysis. Genomic DNA was extracted using the MagaBio Plus Universal Genomic DNA Purification Kit II. Quantitative PCR was employed to analyze the WPRE gene DNA in mouse tissues to determine the in vivo distribution of BD211. The PCR parameters were set to 37°C for 2 min, 95°C for 5 min, followed by 45 cycles of 95°C for 15 s and 60°C for 40 s. A standard curve was generated by linear regression of the WPRE plasmid Ct and concentration log values. The actual sample Ct values were substituted into the standard curve to calculate the WPRE DNA content in whole blood and tissue samples. The primer sequence information is detailed in Supplementary Table S4.

Total RNA was extracted from tissues using the MagaBio Plus Total RNA Purification Kit II (BSC69L1E/BSC69M1E, BioFlux) and from whole blood and bone marrow using the Whole Blood Total RNA Kit (5201050, Hangzhou Xingjing Biotech Co., Ltd.). Quantitative PCR was used to analyze the human β^A-globin, wild-type β^A-globin, and GAPDH mRNA in mouse tissues. The parameters were set to 95°C 15 min; 95°C 10 s, 60°C 20 s, 72°C 20 s, and 40 cycles. The Ct values for the samples were obtained using the instrument software. The relative expression levels were calculated as follows:

WT/GAPDH (wild-type β^A-globin/GAPDH) = Ct_WT/Ct _GAPDH×100%;

T87Q/GAPDH (β^A−T87Q-globin/GAPDH) = Ct_T87Q/Ct _GAPDH×100%

All preserved tissues and organs from the vehicle control, mock control, and high-dose BD211 groups underwent microscopic examination for proliferative or neoplastic changes using H&E staining. Should tumors be detected in high-dose animals, corresponding organs from low-dose groups would have been analyzed.

Statistical analyses were performed using SPSS Statistics for Windows (v.23.0, IBM Corp., Armonk, N.Y., United States). Measurement data, including body weight, food intake, hematological indicators, serum biochemical indicators, and organ weight and coefficient, were presented as mean ± standard deviation (x ± SD). One-way analyses of variance (ANOVAs) were used to assess significance. If a statistically significant difference was found among the groups, further pairwise comparisons were conducted. Levene’s test was used to determine the homogeneity of variance: if groups were homoscedastic, the least significant difference method was applied, and if heteroscedastic, the Games-Howell test was used for pairwise comparisons. Count data, such as behavioral and pathological incidences, were described as fractions, and non-parametric tests (Kruskal–Wallis test) were employed for significance analysis. When a statistically significant difference was observed among the overall groups, differences between pairwise groups were compared. The P-value of <0.05 was considered statistically significant.

The colony-forming cell assay is an in vitro method used to assess the proliferation and differentiation capabilities of hematopoietic stem cells (Kronstein-Wiedemann and Tonn, 2019). Both the mock control and test groups were cultured using semi-solid culture media for 14 days and successfully differentiated into 5 colony types: erythrocytic burst-forming unit, granulocyte colony-forming unit, macrophage colony-forming unit, granulocyte-macrophage colony-forming units, and granulocyte, erythrocyte, macrophage, and megakaryocyte colony-forming units (Figures 2A,B).

During the directed differentiation of CD34⁺ hematopoietic stem cells into erythrocytes, the expression of CD71 gradually decreased, while that of CD235a increased. Early erythroid precursors (proerythroblasts) primarily exhibited a CD71⁺CD235a⁻, intermediate erythroid precursors (basophilic, polychromatophilic, and orthochromatic erythroblasts) exhibited a CD71⁺CD235a⁺ and mature erythrocytes (reticulocytes and mature red blood cells) predominantly displayed CD71^-CD235a⁺ markers. Both mock control cells and four batches of test cells successfully differentiated into erythroid cells in vitro (Figure 2C).

Hemoglobin, the primary component of erythrocytes, is composed of three proteins, with hemoglobin A (HbA) being the most abundant, accounting for over 95% of total hemoglobin in the human body. Flow cytometry analysis revealed that both mock control cells and four batches of test cells successfully detected HbA-positive cells in vitro, with positivity exceeding 90% (Figure 2D).

The relative expression levels of WT/GAPDH in mock control cells and four batches of BD211 cells significantly increased with prolonged in vitro differentiation time. Similarly, the relative expression levels of T87Q/GAPDH in the four batches of BD211 cells also significantly increased. T87Q/GAPDH expression was not detected in mock control cells throughout the experiment (Figures 2E,F).

The mock control cells did not exhibit positivity for the BD211 vector, whereas all four batches of test groups showed BD211 vector positivity, with VCN ranging from 2.84 to 3.63 (Figure 2G). The VCN values for all four batches of test groups were within the safety range of 1–5 copies/cell in accordance with relevant guidelines (U. S. Food and Drug, 2024).

These results demonstrate that BD211 successfully differentiated into various stages of human erythroid cells in vitro and expressed β^A−T87Q-globin during the directed differentiation process.

During the trial period, mortality or near-death events occurred in all groups, with proportions of 17.50%, 19.23%, 11.54%, and 14.10%, respectively. No dose-dependent relationship was observed, and these events were not attributed to the test substance (Figure 3A).

Throughout the study, there were no significant differences in body weight gain between the BD211 low- and high-dose groups and the solvent control group for both male and female mice. Body weight increased over time in all groups (Figures 3B,C). Food consumption in BD211 groups fluctuated within normal ranges, with no significant changes noted. A transient increase in food consumption was observed in male mice of the mock control group, which was not considered toxicologically significant (Figures 3D,E).

At the end of the 13-week dosing period (D92), the hematological parameters of the mock control and BD211 groups, including red blood cell indices (RBC, HGB, HCT, MCHC, and RET), white blood cell indices (WBC, NEUT, LYMPH, MONO), and platelets (PLT), showed a decrease in absolute values compared to the solvent control group, exhibiting a dose-dependent relationship with the human CD34⁺ hematopoietic stem cell dose (Table 1). No biochemically significant changes were observed in the BD211 low- and high-dose groups (Supplementary Table S1).

From D29 onwards, the percentage of hCD3⁺, hCD13⁺, hCD19⁺, hCD34⁺, and hCD56⁺ in the bone marrow of both the mock control and BD211 groups increased in a dose-dependent manner, reaching the highest level by D92 (Table 2). Similarly, starting from D29, the percentage of hCD3⁺, hCD13⁺, and hCD19⁺ in the spleen of the mock control and BD211 groups increased in a dose-dependent manner, reaching the highest level by D92. The percentages of other indicators (hCD4⁺, hCD8⁺, hCD56⁺) did not show significant changes compared to their D2 levels (Table 3).

From D2 to D92, the levels of hCD3⁺, hCD4⁺, hCD8⁺, hCD4⁺/hCD8⁺, hCD56⁺, hCD13⁺, and hCD19⁺ in peripheral blood of the mock control and BD211 groups remained very low level, with no significant increase. This was likely due to the residual immune function of NCG-X mice, as suggested by background information from the animal supplier (unpublished), which indicated that monocytes or macrophages in the tissues could phagocytose and clear the xenogeneic human terminal blood cells released into the peripheral blood after the reconstruction and differentiation of human CD34⁺ hematopoietic stem cells (Supplementary Table S2).

Both the mock control and BD211 groups successfully differentiated into erythroid cells in the bone marrow, with a significant increase in the percentage of hCD71⁺hCD235a⁺ cells, indicating a high proportion of immature erythrocytes, consistent with in vitro differentiation results (Figures 4A–C).

Starting from D29, the percentage of hCD45⁺ cells and their proportion in peripheral blood increased in a dose-dependent manner for both the mock control and BD211 groups, reaching the highest level by D92. On D92, the hCD45⁺ chimerism rates for the mock, BD211 low-dose, and high-dose groups were 73.10%, 43.24%, and 58.85%, respectively (Figure 4D).

On D29, occasional expression of β-globin mRNA and β^A−T87Q-globin mRNA was observed in the bone marrow in the BD211 high-dose group (Figures 5E,F). By D57 and D92, widespread expression of β-globin mRNA was observed in the mock group, whereas β-globin mRNA and β^A−T87Q-globin mRNA in BD211 groups were detected in the bone marrow (Figures 4E,F).

In the high-dose group, WPRE was detected in the bone marrow of only a few mice (3/12) on D2, with a VCN value of 5.63. From D29 onwards, WPRE was consistently detected in the bone marrow and blood of all mice, with VCN values ranging from 2.07 to 2.84. No significant inter-sample or gender differences in VCN values were observed, and VCN values remained relatively stable throughout the study (Figure 4G).

During the study, human-specific cytokine levels of IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ were mostly below the limit of quantification for all groups at each testing point, failing to meet the requirements for statistical analysis (data not shown).

From D2 to D92, the detection rate and genomic concentration of BD211 (representing BD211 and its differentiated nucleated cells in toxicokinetic and tissue distribution studies, with the WPRE gene as the detection marker) increased with the extension of observation time. In the early stages of administration (D2 to D57), the concentration of BD211 in the blood of high-dose animals was significantly higher than that of the low-dose group. However, as time progressed, by D92, the BD211 concentrations in the blood of both groups tended to converge (Figure 5A). The expression trends of β^A−T87Q-globin and β-globin mRNA were essentially consistent with the distribution trend of BD211. Over time, the concentration and detection rate of BD211 in the blood and the detection rate and concentration levels of β^A−T87Q-globin and β-globin mRNA exhibited a clear increasing trend within a certain range (Figures 5B,C).

On D2, BD211 was predominantly detected in the bone marrow, lungs, liver, spleen, femur, and stomach, with concentrations ranging from 1.06 × 10² to 4.71 × 10³ copies/μg gDNA. Over time, the concentration and distribution range of BD211 gradually increased. By D29, it was distributed in multiple organs, with the highest concentration in the bone marrow, followed by the elbow joint, while concentrations in other tissues were relatively lower. The overall detection concentration range was 1.12 × 10² to 5.34 × 10⁵ copies/μg gDNA, with specific ranges of 2.07 × 10⁴ to 5.34 × 10⁵ copies/μg gDNA in the bone marrow and 3.54 × 10³ to 2.12 × 10⁵ copies/μg gDNA in the elbow joint. The distribution trend of BD211 in various tissues and organs at D57 and D92 was consistent with that of D29. The concentration growth trend from D57 to D92 was not significant, suggesting a possible steady state (Figure 5D).

The changes in β^A−T87Q-globin mRNA and β-globin mRNA were largely consistent. They were primarily detected in tissues, such as the lungs, elbow joint, bone marrow, and spinal cord. Over time, the range of detected tissues gradually increased until almost all tissues were involved (Figures 5E,F).

At D92, gross pathological examination of animals in the BD211 groups revealed no gross pathological changes related to the test substance nor any proliferative or tumorigenic changes associated with BD211.

The spleen weight, as well as the organ-to-body and organ-to-brain coefficients, were significantly elevated in male animals of the BD211 high-dose and mock control groups (Table 4). In both male and female mice of the BD211 high-dose and mock groups, an increase in the number of bone marrow (sternum) cells, thymic lymphocytes, and cells surrounding the splenic arteries was observed. Additionally, extramedullary hematopoiesis was observed in male animals. In the BD211 low-dose group, both male and female animals exhibited increased bone marrow (sternum) cells, and extramedullary hematopoiesis was noted in male animals (Figure 6). These histopathological changes were due to the engraftment, reconstitution, and differentiation of human CD34⁺ hematopoietic stem cells in NCG-X mice rather than specific changes induced by the test substance BD211.

Bone marrow smears at D92 showed active bone marrow proliferation in both the mock and BD211 groups, with a significant increase in the proportion of lymphocytes and a relative decrease in the number of granulocytes and erythrocytes, consistent with the trends observed in hematological changes (Supplementary Table S3).