Targeted drug delivery to AML by OFA/iLRP aptamer guided DNA nanostructure

Abstract

Acute myeloid leukemia (AML) has relatively poor clinical outcome. The primary treatment of AML is chemotherapy, which is often associated with severe adverse effects. Targeted drug delivery is a promising strategy to reduce the adverse effects of chemotherapy. Immature laminin receptor protein (OFA/iLRP) is a potential target for AML treatment. In this study, an OFA/iLRP aptamer (AB3) was optimized via sequence truncation to generate a shortened functional aptamer (AB3-2). Two AB3-2 aptamers were coupled together via sticky ends to form a DNA nanostructure (Apt-Couple) for targeted delivery of doxorubicin (Dox) to AML cells. Apt-Couple had an average size of 11.70 nm, and could selectively bind with OFA/iLRP-positive AML cells (HL-60). By intercalating Dox into the DNA structure of Apt-Couple, an Apt-Couple-Dox complex was formed and could carry approximately 18 Dox molecules. Moreover, Apt-Couple-Dox efficaciously destroyed OFA/iLRP-positive AML cells, but notably reduced the damage to control cells in vitro. Furthermore, in HL-60 bearing mice, Apt-Couple-Dox significantly improved the anti-malignancy efficacy compared with free Dox, and prolonged the survival of mice without raising systemic toxicity. These results indicate that the OFA/iLRP aptamer AB3-2 may serve as an AML-homing ligand, and that Apt-Couple-Dox has application potential in targeted therapy against AML.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-34487-1.

Introduction

AML is a severe threat to human health, accounting for approximately 80% of all adult acute leukemias¹. It is characterized by differentiation arrest and aberrant proliferation of myeloid blasts or progranulocytes². It is estimated that there were 158.4 thousand incident cases of AML worldwide in 2018³. Outcomes for AML patients are often described as “dire” or “poor”, despite advances in diagnosis and therapeutics. The five-year survival rate for patients with AML in the US is only 28.3%⁴. The standard treatment for most subtypes of AML at present is still chemotherapy. Standard intensive chemotherapy that combines cytarabine with anthracycline such as doxorubicin (Dox) has been in use for more than 40 years⁴. However, in addition to destroying AML cells, chemotherapeutics are also highly toxic to normal cells, leading to severe adverse effects, including nausea, vomiting, neutropenia, typhlitis, and potentially fatal cardiotoxicity^5,6. Moreover, more than half of AML patients are over 65 years at diagnosis⁴, and these older patients often have poor tolerance to cytotoxic drugs with increased adverse effects. Therefore, development of efficacious anti-AML chemotherapeutics with reduced toxicity is an unmet medical need.

A promising strategy to mitigate adverse effects is targeted therapy, which is often implemented with targeted drug delivery systems (TDDS). A typical TDDS usually consists of two components: tumor-homing ligands and anticancer drugs. When tumor-homing ligands bind to the targets, the drugs are endocytosed by the tumor cells. By selectively delivering cytotoxic drugs to tumor tissue, the adverse effects are reduced and the therapeutic efficacy is improved. Sacituzumab govitecan, an antibody-drug conjugate (ADC) targeting human trophoblast cell-surface antigen 2 (Trop-2), has been shown to significantly prolong progression-free and overall survival compared with single-agent chemotherapy among patients with triple-negative breast cancer⁷. Recently, several TDDSs in the form of ADCs have been approved by the FDA for clinical applications⁸. In addition, currently there are over 80 ADCs in clinical development⁹. These studies indicate that targeted delivery of chemotherapeutics to cancer cells is a promising strategy to improve clinical outcomes.

AML-targeted therapy requires molecular targets that are highly expressed on AML cells but under-expressed in normal myeloid cells. Oncofetal antigen/immature laminin receptor protein (OFA/iLRP) is a potentially important molecular target for AML treatment. The 37-kD OFA/iLRP is a species-conserved membrane protein¹⁰. It has been detected on the cell surface of various malignant tumors, including AML, colon cancer, fibrosarcoma, ovarian cancer, lung cancer, and cervical cancer^11–16. Siegel et al. found that all samples of leukemic cells obtained from AML patients exhibited significant OFA/iLRP expression¹¹. In contrast, OFA/iLRP is not detectable on differentiated adult cells¹⁷. Physiologically, OFA/iLRP is only expressed in embryonic and early fetal period, and disappears in later stages of development¹⁸. Several studies have demonstrated that OFA/iLRP may be a potential target for tumor immunotherapy and gene therapy^18,19. OFA/iLRP-specific cytotoxic T lymphocytes (CTLs) were capable of killing OFA/iLRP-expressing hematologic cells¹¹. Taken together, OFA/iLRP may potentially serve as an attractive molecular target for AML treatment.

Tumor-targeting ligands with high affinity and specificity are crucial to constructing an effective TDDS. Besides antibodies, aptamers may also be employed as tumor-targeting ligands. Aptamers are selected single-stranded oligonucleotides with low immunogenicity and target-binding affinity comparable to antibodies²⁰. Aptamers may have better tumor penetration due to small molecular size (5 nm or smaller)²¹. Moreover, aptamers can be chemically synthesized and modified at low cost²². Indeed, an aptamer-based drug (Macugen^®) has been approved by the FDA for age-related macular degeneration treatment²³. Consequently, aptamers are recognized as promising agents for developing novel tumor-targeting therapies.

In our previous work, we selected an aptamer (AB3) against OFA/iLRP²⁴. The aptamer could preferentially bind to OFA/iLRP-positive AML cells such as HL-60. Moreover, AB3-Dox complex could achieve targeted killing of OFA/iLRP-positive AML cells in vitro, while reducing the damage to control cells. However, it remained unknown whether this OFA/iLRP aptamer could be utilized to construct an AML-targeted drug delivery system for in vivo application. In this study, we established a murine xenograft model using human AML HL-60 cells. The AB3 aptamer was optimized into a shorter form (AB3-2), and a DNA nanostructure (Apt-Couple) was built with two AB3-2 aptamers assembled together via sticky ends. By intercalating Dox into the DNA structure of Apt-Couple, an Apt-Couple-Dox complex was formed. We now report that, compared with free doxorubicin, Apt-Couple-Dox significantly improved antitumor efficacy in vivo without increasing systemic toxicity.

Results

Sequence optimization of aptamer AB3

To minimize the aptamer AB3 to its core functional sequence, the aptamer was optimized through sequence truncation. The full-length AB3 sequence was analyzed using RNAStructure5.0 software. Using this computer-guided approach, nucleotides were systematically removed from the 5’ and 3’ ends so that the stem-loop structure was preserved as much as possible. Three candidates (AB3-1, AB3-2, and AB3-3) were generated, with lengths of 44nt, 33nt, and 16nt, respectively (Fig. 1a). To evaluate whether these truncated aptamer candidates retain the ability to bind to OFA/iLRP, beads coated with OFA/iLRP structure were incubated with FAM-labeled AB3, AB3-1, AB3-2, or AB3-3, washed, and analyzed by flow cytometry. As presented in Fig. 1b, AB3-1 and AB3-2 showed strong binding to the OFA/iLRP-coated beads, whereas AB3-3 had a weak binding to the target. The results indicated that AB3-1 and AB3-2 retained the capability to bind with OFA/iLRP structure, whereas AB3-3 did not, presumably because the sequence of AB3-3 was too short to form a functional 3D structure. Given that AB3-1 and AB3-2 had similar target-binding properties, the shorter AB3-2 was chosen for further study.

Fig. 1.

Truncation and binding analysis of aptamer candidates. (a) Schematic of the aptamer truncation process. Secondary structures of AB3 and the truncated aptamers were predicted using RNAstructure5.0 software. Red crosses indicated the truncation sites. (b) Binding of truncated aptamer candidates to OFA/iLRP structure. FAM-labeled AB3, AB3-1, AB3-2, or AB3-3 were incubated with OFA/iLRP-epitope coated beads and analyzed by flow cytometry. Red curves represented the fluorescent signals of the aptamer candidates; black curves represented the control fluorescence signals generated by FAM-labeled polyT sequence.

Although the truncated aptamer AB3-2 retained binding capability to OFA/iLRP structure, it was still unknown whether AB3-2 could also bind to OFA/iLRP-positive cancer cells. To address this issue, FAM-labeled AB3-2 was incubated with OFA/iLRP-positive cells (HL-60, Jurkat, and Ramos) or OFA/iLRP-negative control cells (PBMC). Prior studies have established that OFA/iLRP is over-expressed on the surface of AML cell line (HL-60), T lymphocytic leukemia cell line (Jurkat), and B-cell lymphoma cell line (Ramos), but not expressed on the surface of PBMC¹¹. As presented in Fig. 2, in OFA/iLRP-positive cells, AB3-2 generated strong fluorescent signals (Fig. 2a-c), while in OFA/iLRP-negative control cells, AB3-2 generated weak signals (Fig. 2d). The results indicated that aptamer AB3-2 could still selectively bind with OFA/iLRP-positive cancer cells.

Fig. 2.

Binding of aptamer AB3-2 to OFA/iLRP-positive and -negative cells. FAM-labeled AB3-2 was incubated with OFA/iLRP-positive cells HL-60 (a), Jurkat (b), or Ramos (c), and OFA/iLRP-negative cells PBMC (d). The cells were subsequently analyzed by flow cytometry. The red curves represented the fluorescent signals of the aptamer and the black curves represented the control fluorescence signals generated by FAM-labeled polyT sequence.

Preparation and characterization of Apt-Couple

Since the optimized aptamer AB3-2 could bind to OFA/iLRP-positive AML cells, we constructed an AB3-2 functionalized DNA nanostructure for AML-targeted delivery of doxorubicin, which is one of the most extensively used drugs in AML chemotherapy. DNA nanostructure was employed here, because doxorubicin could intercalate into GC- or CG-rich double-stranded DNA^25,26. Moreover, DNA nanostructure had good biocompatibility and biodegradability, and could be easily fabricated in a self-assembled manner. Here in this study, the DNA structure consisted of two 98-nt single-strand DNA (ssDNA) chains. Each chain contained an AB3-2 aptamer at the 3’ end, connected via a 5-nt poly-C linker to a 60-nt segment with GC- or CG-rich sticky ends that were complementary to those on the other chain, enabling self-assembly into the DNA nanostructure. The predicted secondary structure of the complex (Apt-Couple) was shown in Fig. 3a. To evaluate whether the two ssDNA chains were assembled into the complex, electrophoresis was performed. As shown in Fig. 3b, the results confirmed the successful formation of the Apt-Couple complex.

Fig. 3.

Preparation and characterization of the DNA nanostructure Apt-Couple. (a) Predicted secondary structure of Apt-Couple generated by the RNAStructure5.0 software. (b) Cropped electrophoretic gel for verifying the formation of Apt-Couple. Lanes 1–3 represented the single strand DNA1, single strand DNA2, and the resultant Apt-Couple complex, respectively. The original gel is presented in Supplementary Fig. 1. (c) Size distribution of Apt-Couple per DLS. (d) TEM image of Apt-Couple. Arrows point to individual nanostructures.

The size of a TDDS is critical for its therapeutic efficacy, because particles smaller than 10 nm are susceptible to renal clearance²⁷. To characterize the nanostructure, dynamic light scattering (DLS) and transmission electron microscopy (TEM) were employed. As shown in Fig. 3c, DLS analysis revealed that the Apt-Couple had an average size of 11.70 ± 1.68 nm. This result was supported by TEM imaging (Fig. 3d), which revealed particles with corresponding dimensions. Taken together, these results indicated that the size of Apt-Couple was above the renal clearance threshold (approximately 10 nm)²⁷.

Serum stability of Apt-Couple

A primary obstacle to the in vivo application of DNA nanostructures as drug carriers is their degradation by nucleases, which are abundant in serum. Several chemical methods have been developed to enhance the nuclease-resistance of DNA strands²⁸. Among them, phosphorothioate modification of the DNA backbone is known to confer nuclease resistance²⁹. In this study, the first nucleotide at the 5’ end of each Apt-Couple ssDNA was modified with phosphorothioate. To evaluate the serum stability, phosphorothioate-modified and non-modified Apt-Couples were incubated in 50% serum for various durations, and analyzed by electrophoresis. As presented in Fig. 4a, in the presence of serum, phosphorothioate-modified Apt-Couple persisted longer than its unmodified counterpart. The results indicated that phosphorothioate modification could enhance the nuclease resistance and serum stability of Apt-Couple, making the modified form a superior drug carrier for in vivo applications. Therefore, phosphorothioate-modified Apt-Couple was utilized in subsequent experiments.

Fig. 4.

The phosphorothioate-modified Apt-Couple exhibited enhanced serum stability and specific binding to target cells. (a) Cropped electrophoretic gel of Apt-Couple that had been incubated in 50% FBS for various time periods (N: non-modified Apt-Couple; P: phosphorothioate-modified Apt-Couple). The original gel is presented in Supplementary Fig. 2. (b) Binding of Apt-Couple and AB3-2 to OFA/iLRP-positive and -negative cells. Each Apt-Couple and each AB3-2 were labeled with a single FAM molecule. OFA/iLRP-positive cells HL-60 (left panel) and OFA/iLRP-negative cells PBMC (right panel) were incubated with FAM-labeled polyT sequence (blank curves), AB3-2 (blue curves), or Apt-Couple (red curves), washed, and analyzed by flow cytometry.

Binding of Apt-Couple to target AML cells

To evaluate whether the aptamers in Apt-Couple could still bind with the target cells, flow cytometry experiments were conducted. To ensure a fair comparison, each Apt-Couple or free AB3-2 aptamer was labeled with a single FAM molecule. OFA/iLRP-positive HL-60 cells or OFA/iLRP-negative PBMC were incubated with FAM-labeled Apt-Couple, AB3-2, or polyT that served as a negative control, and analyzed by flow cytometry. As shown in Fig. 4b, in HL-60 cells, Apt-Couple generated a stronger fluorescent signal than AB3-2, while in PBMC, both Apt-Couple and AB3-2 generated very weak signals. The results indicated that, compared to the monovalent AB3-2 aptamer, the bivalent Apt-Couple had a higher affinity for OFA/iLRP-positive HL-60 cells.

Drug-loading capacity of Apt-Couple

Doxorubicin (Dox) is one of the most extensively used drugs in AML chemotherapy⁴. In this study, Dox was loaded into Apt-Couple, based on the mechanism that Dox tends to intercalate into double-stranded DNA²⁵. To evaluate the drug-loading capacity of Apt-Couple, a fixed concentration of doxorubicin was mixed with increasing concentrations of Apt-Couple. The fluorescence spectrum of the mixture was analyzed, based on the principle that the red fluorescence of doxorubicin is quenched when the drug intercalates into dsDNA³⁰. As illustrated in Fig. 5a, with increasing Apt-Couple/doxorubicin ratio, the fluorescence intensity of the mixture gradually decreased. When the molar ratio of Apt-Couple to Dox increased to 1:18, the fluorescence intensity plateaued at its minimum, indicating that most Dox molecules had intercalated into Apt-Couple under this condition. The results demonstrated that one Apt-Couple could load approximately 18 Dox molecules, forming the Apt-Couple-Dox complex.

Fig. 5.

Evaluation of the drug-loading capacity of Apt-Couple and the targeted drug delivery ability of Apt-Couple-Dox. (a) Fluorescence spectra of Dox solution mixed with increasing molar ratios of Apt-Couple (Apt-Couple/Dox ratios from top to bottom: 0, 1:1000, 1:100, 1:50, 1:30, 1:20, 1:18, and 1:1). The fluorescence generated by doxorubicin was substantially quenched at the Apt-Couple/Dox ratio of 1:18 (designated with a thick red line). (b) Cellular uptake of doxorubicin by HL-60 AML cells and PBMC control cells. Cells were treated with free Dox (upper panel) or Apt-Couple-Dox (lower panel) for 2 h, and evaluated by confocal microscopy. The red fluorescence was generated by doxorubicin. The nuclei were stained blue with DAPI.

Apt-Couple selectively delivered doxorubicin to OFA//iLRP-positive AML cells

To investigate whether the Apt-Couple could selectively deliver Dox to OFA/iLRP-positive AML cells, Apt-Couple-Dox or free Dox was incubated separately with OFA/iLRP-positive AML cells (HL-60) or OFA/iLRP-negative control cells (PBMC), and analyzed by confocal microscopy. As shown in Fig. 5b, after treatment with free Dox, both HL-60 and PBMC exhibited strong red fluorescence, indicating that free Dox non-selectively entered both types of cells. However, after treatment with Apt-Couple-Dox, HL-60 exhibited much stronger fluorescence than PBMC, indicating that Dox in Apt-Couple-Dox preferentially entered HL-60 but not the control cells. The results demonstrated that Apt-Couple-Dox selectively delivered Dox into OFA/iLRP-positive AML cells.

Mechanism of cellular uptake for Apt-Couple

The internalization pathways of aptamers are generally considered to involve endocytosis into lysosomes and macropinocytosis into cytoplasm³¹. To investigate the potential internalization mechanisms of Apt-Couple-Dox, FAM-labeled Apt-Couple was incubated with HL-60 cells, in which lysosomes had been pre-stained with LysoTracker Red. Confocal microscopy studies revealed that Apt-Couple predominantly colocalized with lysosomes, as shown by the merged yellow signal (from red and green) in Fig. 6 (upper panel). Moreover, treatment with Dynasore (an endocytosis inhibitor) significantly inhibited Apt-Couple internalization (middle panel). In contrast, treatment with Cytochalasin D (an actin polymerization inhibitor that disrupts macropinocytosis) had no obvious effect on Apt-Couple internalization (lower panel). The results demonstrated that Apt-Couple entered cells mainly via a dynamin-dependent but actin-independent endocytic pathyway.

Fig. 6.

Determination of the Apt-Couple internalization pathway by confocal fluorescence microscopy. HL-60 cells were treated with FAM-labeled Apt-Couple (100 nM) alone (Control) or in combination with Dynasore (100 µM) or Cytochalasin D (2 µM) for 2 h at 37 °C. Cellular localization was visualized by confocal microscopy. Red fluorescence was generated by LysoTracker Red. The nuclei were stained blue with DAPI.

Apt-Couple-Dox generated a targeted cytotoxicity against OFA/iLRP-positive AML cells in vitro

The above data showed that Apt-Couple-Dox could selectively deliver Dox into OFA/iLRP-positive AML cells. However, it was still unknown whether Apt-Couple-Dox could selectively destroy these AML cells. To address this issue, OFA/iLRP-positive HL-60 cells and OFA/iLRP-negative PBMC were incubated with Apt-Couple, free Dox, or Apt-Couple-Dox, and MTS assay was used to evaluate the cell viability. As shown in Fig. 7a,b, free Dox efficiently killed both HL-60 cells and PBMC. In contrast, Apt-Couple-Dox efficiently killed HL-60 cells but reduced the damage to PBMC. Furthermore, Apt-Couple exhibited no significant cytotoxicity to either HL-60 cells or PBMC, indicating that the empty vector Apt-Couple was relatively nontoxic to both cell types. Taken together, these results demonstrated that Apt-Couple-Dox could generate a targeted cytotoxicity against the OFA/iLRP-positive AML cells in vitro.

Fig. 7.

Cytotoxicity generated by Apt-Couple-Dox or free Dox in vitro. HL-60 AML cells (a) and PBMC control cells (b) were incubated with Apt-Couple, free Dox, or Apt-Couple-Dox for 3 h, washed, and cultured for another 48 h. Cell viability was evaluated by MTS assay (mean ± SD, n = 6). The star indicates a statistically significant difference (p < 0.05). (c) Estimation of IC₅₀ values for HL-60 cells by free Dox or Apt-Couple-Dox. HL-60 cells were incubated with free Dox or Apt-Couple-Dox for 3 h, at the equivalent Dox concentration of 0, 0.005, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, or 2 µM, respectively. MTS assay was performed to evaluate the average cell viability (n = 6). (d) Estimation of IC₅₀ values for PBMC by free Dox or Apt-Couple-Dox. PBMC control cells were incubated with free Dox or Apt-Couple-Dox for 3 h, at the equivalent Dox concentration of 0, 0.01, 0.05, 0.1, 0.5, 1, 2, 3, 5, 10, 20, or 40 µM, and assayed for average cell viability (n = 6).

To further characterize the pharmacological property of Apt-Couple-Dox, the IC₅₀ was evaluated. HL-60 cells and PBMC were treated with free Dox or Apt-Couple-Dox at increasing concentrations, and the cell viability was measured by MTS assay. As shown in Fig. 7c, free Dox and Apt-Couple-Dox exhibited comparable IC₅₀ values in target HL-60 cells (0.45 µM vs. 0.52 µM). In contrast, Apt-Couple-Dox demonstrated a significantly higher IC₅₀ in control PBMC than free Dox (21.63 µM vs. 2.07 µM) (Fig. 7d). Collectively, these findings demonstrated that Apt-Couple-Dox generated preferential cytotoxicity against the target AML cells, and mitigated the damage to the control cells.

Apt-Couple-Dox generated superior antitumor efficacy in vivo

To evaluate the in vivo antimalignancy efficacy of Apt-Couple-Dox, a xenograft model was established by inoculating human AML cells (HL-60) into NOD/SCID mice. Mice bearing HL-60 tumors were treated intravenously with PBS, free Dox (1.5 mg/kg), Apt-Couple, or Apt-Couple-Dox (1.5 mg/kg Dox equivalent) for a total of three doses. As shown in Fig. 8a, Apt-Couple treatment showed no obvious antitumor effects, with tumor growth comparable to the PBS group. However, free Dox inhibited tumor growth to a certain degree. Importantly, Apt-Couple-Dox generated the most potent antitumor efficacy, which was statistically superior to that of free Dox. Moreover, body weight measurements revealed no significant differences between the Apt-Couple-Dox group and free Dox group (Fig. 8b), indicating that Apt-Couple-Dox did not generate extra systemic toxicity. The influences of free Dox vs. Apt-Couple-Dox on animal survival were also compared (Fig. 8c). Both treatments prolonged the survival; however, Apt-Couple-Dox was more efficacious compared with free Dox. The results suggested that aptamer AB3-2 might serve as a tumor-homing ligand for targeted drug delivery to AML cells, and that Apt-Couple-Dox might have application potential in development of novel chemotherapeutics against AML.

Fig. 8.

Tumor inhibition by Dox, Apt-Couple, or Apt-Couple-Dox. Tumor-bearing NOD/SCID mice were treated intravenously with PBS (control), free Dox (1.5 mg/kg), Apt-Couple (143.63 nmol/kg), or Apt-Couple-Dox (with Apt-Couple of 143.63 nmol/kg and Dox of 1.5 mg/kg) on days 10, 13, and 20 post-inoculation. Tumor growth curves (a), body weight (b), and survival rates (c) of HL60-bearing NOD/SCID mice were recorded and shown (n = 8, mean ± SEM). Single star indicates statistically significant differences (p < 0.05) between the PBS and the free Dox groups, and double stars indicate statistically significant differences (p < 0.05) between the free Dox and the Apt-Couple-Dox groups.

Discussion

OFA/iLRP is a potential therapeutic target because it is highly expressed on the surface of AML cells vs. normal myeloid cells. In this study, we optimized a previously developed OFA/iLRP aptamer through sequence truncation, and obtained a shortened aptamer AB3-2. This aptamer retained the capability to recognize and bind with OFA/iLRP-positive AML cells (Figs. 1 and 2). To develop a TDDS for AML, a DNA nanostructure (Apt-Couple) was constructed with two AB3-2 aptamers coupled together via sticky ends. Apt-Couple had an average diameter of 11.70 nm (Fig. 3). Phosphorothioate modification of Apt-Couple improved its nuclease resistance and serum stability (Fig. 4a). Similar to AB3-2 aptamer, Apt-Couple could recognize and bind with OFA/iLRP-positive AML cells (Fig. 4b). By intercalating doxorubicin into the DNA structure of Apt-Couple, the TDDS (Apt-Couple-Dox) was formed (Fig. 5a). Compared with free Dox, Apt-Couple-Dox could selectively deliver doxorubicin into OFA/iLRP-positive AML cells (HL-60) (Fig. 5b), and generated targeted cytotoxicity in vitro (Fig. 7). Moreover, in animal studies, Apt-Couple-Dox markedly enhanced HL-60 inhibition and prolonged overall survival vs. free Dox, without raising the systemic toxicity (Fig. 8). These data indicated that aptamer AB3-2-based TDDS might have application potential in chemotherapy against AML.

Although new treatment modalities have emerged in recent years, AML treatment still heavily depends on conventional chemotherapy. CAR-T for AML faces the challenge of lacking truly AML-specific molecular targets. Some of the candidates, such as CD33, CD123 and CLEC12A, are not restricted to malignant cells, leading to potential “on-target, off-tumor” toxicity³². While combining chemotherapy with immune checkpoint blockade (ICB) by PD-1/PD-L1 antibodies can lower the probability of disease relapse, this strategy still needs chemotherapy in the first place³³. Moreover, FLT-3 or Bcl-2 inhibitors usually need to be combined with chemotherapy to improve clinical outcome³⁴. These developments indicate that cytotoxic chemotherapy maintains a pivotal role in AML treatment, despite the emergence of modern AML therapeutics.

One major obstacle to long-term chemotherapy is the severe adverse effects associated with cytotoxic agents. The general chemotherapeutic strategy for AML has remained largely unchanged for over 3 decades. Specifically, the first phase is induction chemotherapy (continuous-infusion of cytarabine with anthracycline), which is a short and intensive course, typically lasting for about a week. If complete remission is achieved, consolidation chemotherapy is administered in the second phase⁴. The most significant drawback of chemotherapy, however, is the severe toxicity associated with cytotoxic drugs. This is particularly a problem for patients over 65, who account for more than half of the adult AML population. Currently, reducing the toxicity of AML chemotherapy remains an unmet medical need.

Targeted tumor therapy in theory can reduce chemotherapy-related side effects by selectively delivering cytotoxic agents to malignant cells. For targeted AML treatment, gemtuzumab ozogamicin, an ADC comprising an anti-CD33 antibody and calicheamicin, has been approved by the FDA for treating CD33-positive AML³⁵. However, since CD33 is also expressed in normal myeloid cells, gemtuzumab ozogamicin may generate severe adverse effects at times, limiting its application³⁶. Therefore, it is still necessary to explore new therapeutic targets for AML. OFA/iLRP appears a promising molecular target for AML treatment, because it is over-expressed in most AML cells but under-expressed in normal myeloid cells. This discrepancy in expression makes it possible to deliver more cytotoxic drugs to AML cells by targeting OFA/iLRP. Prior to this study, however, the idea had not been tested in animal models. Here we found that OFA/iLRP aptamer enabled selective delivery of Dox to AML cells. Specifically, compared with free Dox, Apt-Couple-Dox generated a targeted killing of HL-60 cells in vitro (Fig. 7), and exhibited stronger therapeutic efficacy in vivo (Fig. 8). The results indicate that OFA/iLRP may potentially serve as a viable molecular target for delivery of cytotoxic agents to AML cells.

Most TDDS approved by the FDA are ADCs, which consist of tumor-targeting antibodies conjugated to cytotoxic drugs. ADCs have been approved by FDA for treatment of a variety of malignancies, including acute lymphoblastic leukemia (ALL), HER2-positive breast cancer, Hodgkin’s lymphoma, systemic anaplastic large cell lym-phoma, and triple negative breast cancer⁹. However, since each antibody conjugates only a few drug molecules, ADCs have a relatively low drug-loading capacity, and thus need to carry extremely potent cytotoxic agents to achieve proper efficacy³⁷. Moreover, ADC production requires special chemical linkers and is quite complicated³⁸. In this study, an aptamer-guided DNA nanostructure was employed to build a TDDS. Compared with ADCs, DNA nanostructure has certain advantages, including high drug-loading capacity, good biocompatibility, and low production cost. In particular, Apt-Couple-Dox can be formed in a self-assembled manner, which is relatively simple and potentially scalable for mass production. Previous studies have reported that DNA nanostructures elicited a modest pro-inflammatory immune response that diminished over time in vivo³⁹. Nevertheless, Apt-Couple-Dox showed significantly stronger antitumor efficacy than free doxorubicin in vivo (Fig. 8), indicating that the potential immunogenicity of Apt-Couple did not compromise the efficacy significantly. Furthermore, given that doxorubicin is widely utilized to treat leukemia, the Apt-Couple-Dox is quite appropriate as a targeted therapy for AML.

The superior anti-AML efficacy of Apt-Couple-Dox may be attributed to several mechanisms. First, Apt-Couple-Dox could not enter normal cells, because cell membrane and DNA were both negatively charged and tend to repel each other. However, Apt-Couple-Dox could selectively bind to OFA/iLRP-expressing AML cells (Fig. 4b), overcoming the repelling force and increasing the drug delivery to these cells (Fig. 5b). Second, phosphorothioate modification of Apt-Couple improved the nuclease resistance and serum stability of the DNA complex (Fig. 4a), thus prolonging its circulating time. Third, the doxorubicin in Apt-Couple-Dox probably had a sustained release into circulation, increasing the drug’s lifetime vs. free doxorubicin. Fourth, the Apt-Couple had a size (11.7 nm) that was appropriate for in vivo application, since nanostructures smaller than 10 nm would be rapidly cleared by the kidneys, and those above 200 nm would be captured by the reticuloendothelial system (RES)⁴⁰. All these factors might have contributed to the enhanced antitumor efficacy of Apt-Couple-Dox vs. free Dox.

In conclusion, OFA/iLRP is highly expressed on most AML cells vs. normal myeloid cells, and may serve as a potential therapeutic target. In this study, we optimized an OFA/iLRP aptamer and constructed an aptamer-guided DNA nanostructure for targeted delivery of doxorubicin to AML cells. Compared with free doxorubicin, Apt-Couple-Dox significantly enhanced therapeutic efficacy in vivo, demonstrating its application potential in AML-targeted therapy.

Methods

Cell culture

HL-60 (human acute promyelocytic leukemia), Ramos (Burkitt’s Lymphoma), and Jurkat (human acute T lymphocyte leukemia) were obtained from the Cell Center of Chinese Academy of Medical Sciences (Beijing, China). Peripheral Blood Mononuclear Cells (PBMC) were isolated from healthy donors by using lymphocyte separation medium (TBD, Tianjin, China) following the manufacturer’s protocol. All donors were required to sign an informed consent form according to procedures approved by the Ethics Committee at Chinese Academy of Medical Sciences and Peking Union Medical College, and all methods were conducted in accordance with the Declaration of Helsinki. All cells were incubated in RPMI-1640 medium (Bioind, Israel) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 µg/mL streptomycin. Cells were grown in a humid atmosphere with 5% CO₂ at 37 °C.

Animals

The animal study and procedures were approved by the Ethics Committee of Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences, and all methods were performed in accordance with the relevant guidelines and regulations. Additionally, this study also adhered to the ARRIVE guidelines and the American Veterinary Medical Association (AVMA) Guidelines for Animal Euthanasia (2020). Endpoints for survival curves of mice bearing HL-60 were death or euthanization by carbon dioxide inhalation followed by cervical dislocation when they met euthanasia criteria, including large tumors, weight loss, and general ill health.

NOD/SCID female mice (6–8 weeks old, 21–23 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All mice were fed a standard diet and water.

Synthesis of DNAs

All DNAs were synthesized by Invitrogen (Shanghai, China). The sequence used in this study is shown as follows:

AB3: 5’-TGCGTGTGTAGTGTGTCTGTTGTTTGTATTGTTGTCTATCCTCTTAGGGATTTGGGCGG-3'.

AB3-1: 5’-TAGTGTGTCTGTTGTTTGTATTGTTGTCTATCCTCTTAGGGATT-3’.

AB3-2: 5’-TTGTTTGTATTGTTGTCTATCCTCTTAGGGATT-3’.

AB3-3: 5’-TATCCTCTTAGGGATT-3’.

Single strand DNA 1 (ssDNA 1):

5’-AGCTGCAGCAGCAGCTGCTGCAGCTGCA GCAGCTGCTGCAGCAGCAGCTGCAGCTGCT

GCCCCCCTTGTTTGTATTGTTGTCTATCCTCTTAGGGATT-3’.

Single strand DNA 2 (ssDNA 2):

5’-GCAGCAGCTGCAGCTGCTGCTGCAGCAG CTGCTGCAGCTGCAGCAGCTGCTGCT

GCAGCTCCCCCTTGTTTGTATTGTTGTCTATCCTCTTAGGGATT-3’.

Notes: The italicized parts indicated the sequence of AB3-2 aptamer.

Immobilization of targets on magnetic beads

The OFA/iLRP peptide (NQIQAAFREPR) of at least 95% purity (Bootech, Shanghai, China) was immobilized to epoxy magnetic beads (BaseLine ChromTech, Tianjin, China) by cross-linking of epoxy and amine groups. The epoxy magnetic beads (2 × 10⁵) were washed thrice with 1 mL of Phosphate Buffered Saline (PBS), resuspended in 400 µL of Carbonate Buffer solution (CBS, PH 10.7) containing 4 µg of OFA/iLRP peptide, and incubated at room temperature with gentle stirring for 12 h. After washing thrice with 500 µL of PBS, the beads were stored at 4℃ in PBS.

Preparation of Apt-Couple

Two single-strand DNAs were mixed in saline solution (0.9% NaCl, 2 mM MgCl₂) at an equal molar ratio. The mixture was incubated at 95℃ for 5 min, and then cooled slowly to room temperature at a rate of 0.1℃/s.

Characterization of Apt-Couple

Agarose gel electrophoresis was utilized to evaluate the formation of Apt-Couple. The gel was prepared in TBE buffer with 2.5% (w/v) agarose and GelStain (TransGen Biotech, Beijing, China). Electrophoresis was performed at 120 V for 30 min in TBE buffer, and the DNA bands were visualized under UV light. The average size and zeta-potential of Apt-Couple were evaluated by dynamic light scattering (Malvern Zetasizer Nano ZS, UK). The morphology of nanoparticles was observed with transmission electron microscope (TEM) (JEM-1400plus, JEOL, Tokyo, Japan) after the samples were deposited on a copper grid and stained with 1% uranyl acetate solution.

Serum stability of Apt-Couple

Phosphorothioate-modified and non-modified Apt-Couple (1 µM) were incubated in 50% FBS at 37℃ for various durations, followed by heating at 95℃ for 5 min. Then agarose gel electrophoresis was employed to evaluate the degradation of the DNA nanostructures.

Flow cytometric analysis

OFA/iLRP peptide-coated magnetic beads (2 × 10⁵) were incubated with 40 pmol FAM-labeled ssDNA in 200 µL PBS for 30 min at room temperature, washed twice with 200 µL PBS, resuspended in 200 µL PBS, and analyzed by flow cytometry (Accuri C6 Flow Cytometer, BD).

To evaluate the cell-binding capacity of Apt-Couple, cells (2 × 10⁵) were incubated with 80 pmol FAM-labeled poly-T, monovalent aptamer, or Apt-Couple in 300 µL PBS for 30 min at room temperature, then washed twice with 200 µL PBS, resuspended in 200 µL PBS, and analyzed by flow cytometry.

Drug-loading capacity of Apt-Couple

A fixed concentration of Dox (3 nM) was incubated with Apt-Couple at varying Dox/carrier molar ratios (0, 0.001, 0.01, 0.02, 0.033, 0.05, 0.056, and 1) for 1 h in a 96-well black plate. Synergy4 analyzer was used to read the plate for assessing the fluorescence spectrum (λEx = 488 nm, λEm = 520–700 nm).

Cellular drug uptake studies

HL-60 or PBMC were incubated with Dox or Apt-Couple-Dox at an equivalent dose of Dox (2 µM) for 2 h at 37℃, washed twice with PBS, and fixed with 4% formaldehyde for 10 min at 4℃. The cells were washed twice with PBS and stained with 10 µL of DAPI (1 µg/ml) for 5 min. The cells were mounted onto a glass slide, covered with a coverslip, and imaged using a confocal fluorescence scanning microscopy (Perkin Elmer Ultraview, Perkin, Waltham, MA, USA).

Internalization pathway of Apt-Couple

To examine the internalization pathway of Apt-Couple, HL-60 cells were incubated with FAM-labeled Apt-Couple (100 nM) along with Dynasore (100 µM) or Cytochalasin D (2 µM) for 2 h at 37℃. FAM-labeled Apt-Couple (100 nM) alone was used as control. After washing with PBS, the fixed cells were visualized by confocal fluorescence microscopy (Perkin Elmer Ultraview, Perkin, Waltham, MA, USA).

In vitro cytotoxicity studies

HL-60 or PBMC cells (5 × 10⁴ cells per well) were seeded in 96-well plates, treated with Dox, Apt-Couple, or Apt-Couple-Dox at an equivalent dose of Dox (HL-60: 0.5 µM; PBMC: 1.8 µM) at 37℃ for 3 h, washed thrice with PBS, and cultured for an additional 48 h. Cell viability was determined by MTS assay (Promega, US) according to the standard protocol suggested by the manufacturer.

Evaluation of IC₅₀

HL-60 or PBMC cells were cultured in 96-well plates as described before and incubated with free Dox or Apt-Couple-Dox at 37℃ for 3 h. HL-60 cells were treated with equivalent Dox concentrations of 0, 0.005, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, or 2 µM. PBMC cells were treated with equivalent Dox concentrations of 0, 0.01, 0.05, 0.1, 0.5, 1, 2, 3, 5, 10, 20, or 40 µM. The cells were washed thrice with PBS, and cultured for 48 h. Cell viability was determined by MTS assay.

In vivo tumor Inhibition study

For an animal AML model, 1 × 10⁷ HL-60 cells suspended in 100 µL PBS were injected subcutaneously on the right rear flank of NOD/SCID mice. When tumor diameter reached about 5 mm, mice were randomly divided into four groups. The mice were treated with PBS, free Dox, Apt-Couple, or Apt-Couple-Dox via intravenous injection on days 10, 13, and 20 post-inoculation. Tumor size and body weight were measured every 2 days during the experiment. Tumor volume was calculated using the formula: volume = (length × width²) / 2.

Statistical analysis

Statistical analyses were performed with GraphPad Prism 5 software (La Jolla, CA, USA). One-way ANOVA with Fisher’s least significant difference (LSD) comparison test was used for statistical calculation. p < 0.05 was regarded as statistically significant. All data are presented as mean value with standard deviation indicated (mean ± SD).

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

X-D.Y. designed and supervised the experiments; Y.A. performed the experiments; Y.A. analyzed the data; F.Y. , X.L. and Q.J. provided technical support; X-D.Y. and Y.A. wrote the manuscript; All authors read and approved the final manuscript.

Data availability

All data generated or analysed during this study are included in this published article and its Supplementary Information files.

Declarations

Competing interests

The authors declare no competing interests.