1. Introduction
In previous studies, it was reported that approximately 1%-5% of acute lymphoblastic leukemia (ALL) patients who were not undergoing central nervous system (CNS)-directed prophylaxis would relapse with CNS-ALL [
1], [
2]. Currently, the main clinical treatment strategies for CNS leukemia are ① oral administration of glucocorticoids and dexamethasone, ② intravenous injection of high-dose methotrexate (MTX) combined with intrathecal injection, and ③ moderate radiotherapy [
3], [
4], [
5]. Although these approaches achieved a clinically significant short-term effect, the neurotoxicity has a long-term effect on patients [
6], [
7].
Antibody-based immunotherapy is a new method to improve CNS-ALL treatment. It has been reported that the use of chimeric antibodies against cluster of differentiation (CD) 19 and CD3, such as blinatumomab, effectively improves the clinical outcome in patients positive for minimal residual disease (MRD). However, blinatumomab causes significant toxicity, resulting in CNS events and cytokine release syndrome [
8], [
9]. Other antibodies, such as the CD19 monoclonal antibody CD19-DE [
10], CD38 antibody daratumumab [
11] and interleukin 7 receptor (IL7R) monoclonal antibody [
12], have also achieved good therapeutic effects in preclinical animal models. The limitation of these antibody therapies, however, is their inefficient delivery to the CNS. Another novel strategy is chimeric antigen receptor T (CAR-T) cell immunotherapy. A clinical trial of CD19-CAR-T cell therapy showed that this regimen achieved good outcomes in chemotherapy-resistant childhood and adult ALL, with a reduced rate of cytokine release syndrome [
12]. In another clinical trial on drug-resistant acute B-cell lymphoblastic leukemia (B-ALL), ALL cells were undetectable in the cerebrospinal fluid (CSF) of two CNS-ALL patients after CD19-CAR-T cell treatment [
13]. In an international study of CAR-T cell therapy for relapsed B-ALL with CNS involvement, 51 of 54 evaluable patients (94%) achieved complete response, but relapse occurred in 22 of 54 patients (41%) [
14]. Data from these clinical trials suggest that CAR-T therapy may be an effective treatment for CNS-ALL. However, existing clinical trials have used the different structure of CD19-CAR-T cell, resulting in inconsistent in efficacy, and the use of CAR-T cells is often accompanied by side effects such as cytokine release syndrome [
15]. Therefore, safer and more effective CNS-targeted therapies are urgently needed.
Integrin α6 is a transmembrane protein encoded by the human integrin α6 (
ITGA6) gene [
16]. Naturally, integrin α6 interacts with β1 or β4 to form heterodimers, namely, integrins α6β1 and α6β4, which are abnormally overexpressed in a variety of tumors [
17], [
18]. Integrins α6β1 and α6β4 are involved in the processes of tumor progression and immune escape, as well as the induction of tolerance to tumor chemotherapy and radiotherapy [
17], [
19], [
20], [
21], [
22], [
23], [
24]. Analysis of the Gene Expression Profiling Interactive Analysis (GEPIA) cancer database
† showed that the
ITGA6 gene was overexpressed in thirteen kinds of tumors, including acute myeloid leukemia (AML), which showed the highest tumor-to-normal tissue ratio of 360.75 among them (Table S1 in Appendix A). In a clinical study of relapsed AML, the expression of integrin α6 was significantly upregulated, suggesting that integrin α6 may act as an important therapeutic target for refractory AML [
21]. Furthermore, integrin α6 has been reported to be overexpressed in B-ALL, and the deletion of integrin α6 induced apoptosis in B-ALL cells [
25]. Moreover, integrin α6 has been demonstrated to interact with laminin to induce the migration of ALL cells into CSF, suggesting that integrin α6 expression directly promotes CNS-ALL disease progression [
26]. Another study confirmed that phosphoinositide 3-kinase (PI3K) inhibitors can reduce the expression of integrin α6, thus decreasing CNS involvement and prolonging the survival of B-ALL mice [
26]. Additionally, integrin α6 has been demonstrated to mediate B-ALL drug resistance [
21], [
25]. In summary, integrin α6 plays an important role in the migration of ALL cells toward the CNS, as well as in B-ALL drug resistance.
Previously, we identified an integrin α6-targeted peptide, CRWYDENAC (abbreviated RWY), by using phage display and next-generation sequencing techniques [
27]. Similar to the integrin αvβ3-targeting peptide arginine-glycine-asparticacid (RGD), the identified peptide RWY shows high-affinity to tumors through binding to integrin α6. The RWY peptide has been employed for molecular imaging and targeted therapy in several kinds of tumors, including nasopharyngeal carcinoma [
27], hepatocellular carcinoma (HCC) [
28], and intestinal carcinoma [
29] in mouse models and breast cancer in patients [
30]. Recently, we performed alanine scanning of the integrin α6-targeted peptide RWY and identified three key amino acids, namely, arginine (R), tryptophan (W), and tyrosine (Y), which mediated the peptide’s tumor-targeting ability [
31]. Moreover, we constructed an optimized integrin α6-targeted peptide, CRWYDANAC (abbreviated S5), which showed a twofold improvement in binding affinity to integrin α6 compared with the RWY peptide [
31]. Based on peptide S5, three molecular imaging probes, cyanine 5 (Cy5)-S5 for near-infrared fluorescence (NIRF), gadolinium (Gd)-S5 for magnetic resonance (MR), and
18F-S5 for positron emission tomography (PET) imaging, were developed and employed for CNS-ALL imaging [
32]. These previous studies demonstrated that integrin α6-targeted peptides are ideal vehicles to deliver imaging or treatment agents for CNS-ALL diagnosis and therapy.
In this study, we identified an optimized integrin α6-targeted peptide with nanomolar affinity to integrins α6β1 and α6β4 using peptide scanning techniques. Based on this peptide, we developed an integrin α6-targeted self-assembling nanopeptide
D(RWYD)-
D(KLAKLAK)
2-G
D(FFY) (abbreviated RD-KLA-Gffy), which consists of the newly identified integrin α6-targeted peptide
D(RWYD) (abbreviated RD), the proapoptotic peptide
D(KLAKLAK)
2 (abbreviated KLA) [
33] and the self-assembling peptide G
D(FFY) (abbreviated Gffy) [
34]. RD-KLA-Gffy induced apoptosis of leukemia cells
in vitro and
in vivo, and thus showed an antileukemia effect when used as a single agent or combined with MTX, a traditional chemotherapeutic drug for ALL therapy.
2. Materials and methods
2.1. Cell culture
HCC cell lines, including HCC-LM3, Huh7, and Hep3B; leukemia cell lines, including Nalm6, Jurkat, and K562; and the ovarian cancer cell line HeLa, were purchased from the American Type Culture Collection (ATCC; USA). HCC-LM3, Huh7, Hep3B, and HeLa cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM); and Nalm6, Jurkat, and K562 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal bovine serum and 0.5% penicillin-streptomycin. To further verify the binding affinities of the peptides to cells, HCC-LM3 and HeLa cells were stably transfected with ITGA6 knockout (KO) lentivirus using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) genome editing system (OBiO Technology (Shanghai) Co., Ltd., China). To facilitate tumor monitoring, Nalm6 cells were stably transfected with luciferase (OBiO Technology). All cells were regularly authenticated and confirmed to be mycoplasma-negative.
2.2. Synthesis of peptide
All peptides involved in this study were synthesized by the Chinese Peptide Company (China), which provided complete quality control reports. The amino acid sequences of all peptides involved in this study are as follows: cyclic peptides CRWYDENAC (abbreviated RWY), CAWYDENAC, CRAYDENAC, CRWADENAC, CRWYAENAC (abbreviated M4), CRWYDANAC (abbreviated S5), CRWYDEAAC (abbreviated M5), RWYDENA, RWYDANA, RWYDEN, RWYDAN, RWYDE, and RWYDA; linear peptides RWYD, RWY, and D(RWYD) (abbreviated RD); fluorescent peptides CRWYDENAC-PEG4-K-fluorescein isothiocyanate (FITC) (abbreviated FITC-RWY) and RWYD-PEG4-K-FITC (abbreviated FITC-RWYD); nanopeptides and control peptides D(RWYD)-D(KLAKLAK)2-GD(FFY) (abbreviated RD-KLA-Gffy), D(CRWYAENAC)-D(KLAKLAK)2-GD(FFY) (abbreviated M4-KLA-Gffy), D(CRWYDEAAC)-D(KLAKLAK)2-GD(FFY) (abbreviated M5-KLA-Gffy), D(CRWYDENAC)-D(KLAKLAK)2-GD(FFY) (abbreviated RWY-KLA-Gffy), D(KLAKLAK)2-GD(FFY) (abbreviated KLA-Gffy), D(RWYD)-D(KLAKLAK)2 (abbreviated RD-KLA), D(KLAKLAK)2 (abbreviated KLA), and D(RWYD)-GD(FFY) (abbreviated RD-Gffy).
2.3. Microscale thermophoresis (MST) assay
In the MST assay, fluorescence labeling of integrin α6β4 or α6β1 recombinant protein (#CT069-H2508H or #CT013-H2508H; Sino Biological, China) was performed with the reactive dye RED-tris-NTA (#MO-L018; NanoTemper Technologies, Germany), which reacts with the primary amines of a protein to form a stable dye-protein conjugate. Analytes with different concentrations of peptides (Chinese Peptide Company) were incubated with labeled protein for 5 min at 25 °C, and then the mixtures in the tubes were absorbed into capillaries (#MO-K022; NanoTemper Technologies) for sample detection by Monolith NT.115 (NanoTemper Technologies) and final analysis.
2.4. Molecular docking
The crystal structure of integrin α6 was obtained by SWISS-MODEL homology modelling [
35], and optimized by adding exclusive hydrogens and missing atoms. The three-dimensional structure of the peptide was created by the PEPstrMOD server [
36], [
37], and unbiased docking to the optimal structure of integrin α6 was performed by using the ZDOCK server [
38]. Molecular operating environment (MOE) software was used to analyze the docking mode between the protein and peptide, and the best docking mode was selected by combining the indices of binding force and interaction site [
39].
2.5. Flow cytometry
To detect the expression of integrin α6 in cells, cells were seeded at a density of 1 × 106 cells per milliliter in 6-well plates and then stained with 10 µL integrin α6 antibody per 1 × 106 cells (vFAB13501P; R&D Systems, USA) and immunoglobulin G2A (IgG2A) antibody (#IC006P; R&D Systems) at 4 °C. After incubation for 30 min, the cells were washed three times with phosphate buffered saline (PBS) and analyzed on a CytoFLEX S cytometer (Beckman Coulter, USA). To investigate the binding of peptide RD to cells, cells were seeded at a density of 1 × 106 cells per milliliter in a 6-well plate and then stained with 1 μmol∙L−1 FITC-peptide (Chinese Peptide Company) at 37 °C. After incubation for 60 min, the cells were washed three times with PBS and analyzed on a CytoFLEX S cytometer. To verify the principle by which the nanopeptide kills leukemia cells, cells were seeded at a density of 1 × 106 cells per milliliter in 6-well plates and then incubated with 5 μmol·L−1 peptides at 37 °C. After incubation for 60 min, the cells were washed three times with PBS and stained with Annexin V/PI (#BMS500FI; Invitrogen, USA) at 4 °C. After incubation for 30 min, the cells were washed three times with PBS and analyzed on a CytoFLEX S cytometer.
2.6. Physical characterization of the nanopeptides
The nanopeptides were synthesized and analyzed by high-performance liquid chromatography (HPLC) and mass spectrometry (MS) by the Chinese Peptide Company. The critical aggregation concentration (CAC) values were determined by dynamic light scattering (DLS) (Malvern Paralytical, UK). Solutions containing different concentrations of compounds were tested, and the light scattering intensities were recorded for final analysis. The zeta potentials and sizes of the nanopeptides were determined by DLS. Transmission electron microscopy (TEM) images of the nanopeptides were acquired with a TEM microscope (FEI Tecnai Spirit TEM T12; Thermo Fisher Scientific, USA) at 100 000-fold magnification.
2.7. Cytotoxicity of the nanopeptides
To determine the cytotoxicity of the nanopeptides, cells were seeded at a density of 1 × 106 cells per milliliter in a 96-well plate and then incubated with different concentrations of peptides (Chinese Peptide Company) at 37 °C. After incubation for 3 h, the cells were stained with cell counting kit-8 (CCK-8) (#C0039; Beyotime, China). Three hours later, the cells were analyzed by ultraviolet (UV) spectroscopy (Synergy H1; BioTek, USA). To monitor the cytotoxicity of the peptides to cells in real time, cells were seeded at a density of 5 × 104 cells per milliliter in a 96-well plate. Then, the cells were incubated with 5 μmol∙L−1 peptides accompanied by 2.5 μmol∙L−1 SYTOX Green nucleic acid stain (#S7020; Invitrogen) and photographed every 10 min using a cell real-time monitoring system (IncuCyte S3; Essen Bioscience, USA).
2.8. Protein extraction and Western blotting
After incubation with 5 μmol∙L−1 peptides for 3 h, the cells were washed with PBS. Proteins were extracted and dissolved in radio immunoprecipitation assay (RIPA) lysis buffer (#P0013B; Beyotime). Protein concentrations were determined by the Bradford assay (#23225; Thermo Fisher Scientific). The proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes, which were then incubated with antibodies overnight at 4 °C, including anti-caspase-3 (1: 2000; #ab32351; Abcam, UK), anti-cleaved caspase-3 (1:500; #ab32042; Abcam), anti-caspase-8 (1: 2000; #ab227430; Abcam), anti-poly(ADP-ribose) polymerase-1 (PARP1) (1: 1000; #ab191217; Abcam), anti-cleaved PARP1 (1: 1000; #ab32064; Abcam), anti-B cell lymphoma-2 (BCL-2) (1: 1000; #15071; Cell Signaling Technology, USA), and anti-β-tubulin (1: 1000; #2146; Cell Signaling Technology). Finally, goat anti-rabbit horseradish peroxidase-labeled immunoglobulin G (IgG-HRP) antibody (1: 10 000; #G-21234; Thermo Fisher Scientific) and goat anti-mouse IgG-HRP antibody (1: 10 000; #31430; Thermo Fisher Scientific) were used as secondary antibodies.
2.9. Mouse engraftment
The mouse experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Sun Yat-sen University Cancer Center (IACUC approval number L025501202204008). To generate xenograft mice, 6-week-old female BALB/c-nude mice (Shanghai Model Organisms Center, China) were inoculated subcutaneously with 1 × 10
6 Nalm6-luciferase-enhanced green fluorescent protein (luc-EGFP) cells. Similarly, to generate CNS-ALL mice [
26], 6-week-old female nonobese diabetic (NOD)/severe combined immunodeficient (SCID)/interleukin 2 receptor gamma chain (IL-2Rγ) (null) (NSG) mice (Shanghai Model Organisms Center, China) were inoculated intravenously with 1 × 10
6 Nalm6-luc-EGFP cells. When Nalm6-luc-EGFP cells could be observed in the head (meninge) by luminescence imaging approximately eight days later, the CNS-ALL mouse model was considered to have been successfully constructed.
2.10. NIRF imaging
To verify the tumor accumulation of the nanopeptide, RD-KLA-Gffy was labeled to display red fluorescence with an EZLabel protein Cy5 labeling kit (#K839-5; BioVision, USA), and high-purity Cy5-labeled RD-KLA-Gffy (abbreviated Cy5-RD-KLA-Gffy) and Cy5-labeled KLA-Gffy (abbreviated Cy5-KLA-Gffy) were obtained by molecular sieving (General Electric, USA). NIRF imaging was performed using an in vivo imaging system (IVIS) spectrum (PerkinElmer, USA) at 1, 12, 24, and 48 h after 0.5 mg·kg−1−1 Cy5-RD-KLA-Gffy was injected intravenously into subcutaneous xenograft mice or CNS-ALL mice. Luminescence imaging was performed starting 5 min after intraperitoneal injection of 150 mg⋅kg−1 luciferin (#P1041; Promega Corporation, USA), and NIRF images were acquired with excitation at 640 nm and emission at 680 nm. After NIRF imaging, the mice were euthanized by cervical dislocation and immediately dissected. Luminescence and NIRF signals from the tumor and vital organs were recorded.
2.11. Immunofluorescence, hematoxylin and eosin (HE) staining, and immunohistochemistry (IHC)
The brains of the CNS-ALL mice were cut into frozen sections, stained with 1 μg∙mL
−1 4′,6-diamidino-2-phenylindole (DAPI) (#28718-90-3; Sigma-Aldrich, USA) and mounted on slides with ProLong Gold Antifade (P26930; Invitrogen). Fluorescence images were captured under a confocal microscopy confocal laser scanning system (Pannoramic MIDI; 3DHISTECH, Hungary). In the fluorescence images, blue fluorescence represents the nucleus, green fluorescence represents Nalm6-luc-EGFP cells, and red fluorescence represents the nanopeptide Cy5-RD-KLA-Gffy. The frozen sections were also stained with HE according to routine histological procedures. Furthermore, frozen sections were used for IHC with an anti-integrin α6 antibody (1: 150; #ab181551; Abcam). IHC was performed following the conventional procedure we reported previously [
28], and images were observed under a microscope (CSU-W1; Nikon, Japan).
2.12. Survival experiment
As described for the mouse engraftment experiment, when the CNS-ALL mouse model was successfully constructed, the mice were divided into groups with 10 mice per group. The mice were intravenously treated with drugs every two days. The concentration of RD-KLA-Gffy and control peptides was 5 mg∙kg−1. The concentration of MTX was 2 mg∙kg−1. The body weights and survival curves of the CNS-ALL mice were monitored. Nalm6-luc-EGFP cell metastasis was detected by luminescence imaging.
2.13. Statistical analysis
All statistical analyses were performed using GraphPad Prism software (Version 8.3.0.538). Statistical evaluations were performed using either one-way analysis of variance (ANOVA) or two-way ANOVA, and the results are shown as the mean ± standard deviation (SD). Statistical analyses of the survival experiments were performed using log-rank tests. P values are denoted as follows: no significance (ns), P > 0.05; *P < 0.05; **P < 0.01; and ***P < 0.001.
3. Results
3.1. Identification of the high-affinity integrin α6-targeted peptide
Although the integrin α6-targeted peptides RWY and S5 can both be highly enriched in tumor tissues [
27], [
28], [
31], their binding affinities to integrins α6β1 and α6β4 are still relatively low (micromolar range). To obtain optimized integrin α6-targeted peptides with improved binding affinity to integrins α6β1 and α6β4, three peptide scanning techniques, namely, alanine scanning, truncation, and D-substitution, were performed (
Fig. 1(a)). Beginning with the previously identified integrin α6-targeted peptides RWY and S5, we designed and synthesized a series of mutant peptides by using alanine scanning and truncation techniques. Then, we determined the binding affinity of the candidate peptides to integrin α6β4 by using MST analysis, and the dissociation constants (
Kd values) of RWY and S5 were (6.97 ± 1.44) μmol∙L
−1 and (4.34 ± 3.42) μmol∙L
−1, respectively (
Table 1). Among these mutant peptides, the linear mutant peptide RWYD demonstrated the highest binding affinity to integrin α6β4 (
Kd = (21.82 ± 3.86) nmol∙L
−1), which was approximately 319-fold more potent than that of the peptide RWY (
Fig. 1(b)). Moreover, two cyclic mutant peptides, M4 and M5, also showed significantly improved affinities with enhancements of 17- and 11-fold, respectively (
Table 1; Fig. S1 in Appendix A), compared to RWY. Additionally, the binding affinity of the linear peptide RWYD to integrin α6β1 was (23.95 ± 6.10) nmol∙L
−1, which was approximately 235-fold more potent than that of the peptide RWY (
Kd = (5.63 ± 1.01) μmol∙L
−1) (
Fig. 1(b)). In summary, these results demonstrate that the linear peptide RWYD is an ideal integrin α6-targeted peptide with nanomolar binding affinities to integrin α6β4 (
Kd = (21.82 ± 3.86) nmol·L
−1) and integrin α6β1 (
Kd = (23.95 ± 6.10) nmol·L
−1).
Furthermore, to determine the key amino acids that mediate the binding of the peptide RWYD to integrin α6, molecular docking analysis was performed. The interaction sites of the RWYD and RWY peptides on integrin α6 are highlighted (
Fig. 1(c)). RWYD formed three hydrogen bonds (R1-S331, Y3-Y450, and D4-H110) and one aromatic hydrophobic interaction (W2-Y273). In comparison, the previously identified peptide RWY formed two hydrogen bonds (R2-S331 and Y4-R111) and one aromatic hydrophobic interaction (W3-F23). Compared to peptide RWY, RWYD showed one more binding interaction (D4-H110), which may explain why the binding affinity of peptide RWYD to integrins α6β4 and α6β1 is stronger than that of peptide RWY.
3.2. Binding specificity of the optimized integrin α6-targeted peptide
To determine whether the integrin α6 expression level plays an important role in affecting the affinity of the optimized integrin α6-targeted peptide RWYD correlates, two FITC-labeled peptides, FITC-RWY and FITC-RWYD, were synthesized, and their affinity to three HCC cell lines with different expression levels of integrin α6 was examined. The expression levels of integrin α6 were first determined using an α6 antibody, and the data showed that the expression of integrin α6 was high in HCC-LM3 cells, moderate in Huh7 cells, and low in Hep3B cells (
Fig. 2(a)). In line with the different expression levels of integrin α6, the cell binding ability of the peptide FITC-RWYD was high in HCC-LM3 cells, moderate in Huh7 cells, and low in Hep3B cells (
Figs. 2(b)-(d)). Comparatively, in cell lines with high and moderate expression of integrin α6 (HCC-LM3 and Huh7), the cell binding ability of FITC-RWYD was significantly higher than that of peptide FITC-RWY (
Fig. 2(e)). However, in Hep3B cells with low integrin α6 expression, there was no significant difference between the cell binding abilities of peptides FITC-RWYD and FITC-RWY (
Fig. 2(e)). Additionally, we explored the cell binding abilities of FITC-RWYD to three leukemia cell lines, including Nalm6, Jurkat, and K562. Similar to HCC cells, FITC-RWYD bound more strongly than FITC-RWY to Nalm6 and Jurkat cells, which exhibit high expression of integrin α6. Moreover, there was no significant difference between the peptides FITC-RWYD and FITC-RWY in cell affinity to K562 cells, which have low integrin α6 expression (Fig. S2 in Appendix A). These results demonstrated that the cell binding ability of the optimized integrin α6-targeted peptide RWYD is correlated with the expression level of integrin α6.
To further determine the binding specificity of the peptide RWYD, the CRISPR/Cas9 genome editing system was used to knock out the
ITGA6 gene in the HCC cell line HCC-LM3 and the ovarian cancer cell line HeLa, both of which exhibit high expression of integrin α6 (
Fig. 2(f)). As expected, in
ITGA6 knockout HCC-LM3-KOα6 and HeLa-KOα6 cells, the cell binding ability of the peptide FITC-RWYD was completely abolished (
Figs. 2(g)-(i)). These results clarified that the cell binding ability of the optimized integrin α6-targeted peptide RWYD mainly depended on the expression of integrin α6.
3.3. Characteristics of integrin α6-targeted self-assembling nanopeptide RD-KLA-Gffy
Considering that right-handed amino acids are usually more stable than left-handed amino acids
in vivo, the D-substitution technique was performed. The MST results showed that the binding affinities of the right-handed peptide RD to integrin α6β4 (
Kd = (30.29 ± 2.83) nmol·L
−1) and α6β1 (
Kd = (19.23 ± 3.04) nmol·L
−1) were similar to those of the left-handed peptide RWYD to integrins α6β4 (
Kd = (21.82 ± 3.86) nmol·L
−1 and α6β1 (
Kd = (23.95 ± 6.10) nmol L
−1) (
Table 2; Fig. S3 in Appendix A). Therefore, the right-handed integrin α6-targeted peptide RD was used in the following study.
Previously, based on integrin α6-targeted peptide S5, three integrin α6-targeted peptide-based molecular probes were developed and employed for the accurate detection of CNS-ALL in a mouse model using different imaging strategies, such as Cy5-S5 for NIRF, Gd-S5 for MR imaging, and
18F-S5 for PET imaging [
32]. These integrin α6-targeted probes were highly enriched in CNS-ALL lesions in the mouse models, suggesting that the use of integrin α6-targeted peptides is valuable for CNS-ALL targeting. Therefore, an integrin α6-targeted self-assembling nanopeptide RD-KLA-Gffy consisting of an integrin α6-targeted peptide RD, a proapoptotic peptide KLA [
33], and a self-assembling tetrapeptide Gffy [
34] was designed (
Fig. 3(a)). Among the different elements of RD-KLA-Gffy, integrin α6 and the tumor-targeting peptide RD mediated tumor enrichment, the proapoptotic peptide KLA induced tumor cell apoptosis, and the backbone of the nanopeptide Gffy facilitated self-assembly into RD-KLA-Gffy nanoparticles. The synthesis process of the nanopeptide RD-KLA-Gffy is shown in Fig. S4 in Appendix A. The nanopeptide was purified and identified using HPLC chromatogram and MS spectrum (Fig. S5 in Appendix A). The basic nanocharacteristics of RD-KLA-Gffy were examined by DLS. The CAC value of RD-KLA-Gffy was 2.52 μmol·L
−1 (
Fig. 3(b)), and its zeta potential was 7.07 mV (
Fig. 3(c)), which indicates a high tendency to assemble. Its hydrodynamic diameter was 21.8 nm (
Fig. 3(d)), as measured by DLS, similar to the data detected by TEM images (
Fig. 3(e)). Moreover, the affinities of RD-KLA-Gffy for integrins α6β4 and α6β1 were (70.6 ± 3.4) nmol·L
−1 and (25.7 ± 4.3) nmol·L
−1, respectively, which were slightly lower than those of the single peptide RD (
Kd = (30.29 ± 2.83) nmol·L
−1 for integrin α6β4 and
Kd = (19.23 ± 3.04) nmol·L
−1 for integrin α6β1) (
Fig. 3(f)).
3.4. Antitumor effect of the integrin α6-targeted self-assembling nanopeptide RD-KLA-Gffy
In addition to RD-KLA-Gffy, we synthesized other nanopeptides: M4-KLA-Gffy, M5-KLA-Gffy, and RWY-KLA-Gffy, as controls (Fig. S6(a) in Appendix A). To determine their antitumor effects, cell viability assays were performed using CCK-8. The data showed that all nanopeptides had dose-dependent cytotoxicity, among which RD-KLA-Gffy was the most toxic to Nalm6 cells, with a half maximal inhibitory concentration (IC50) value of 3.15 μmol·L−1 (Fig. S6(b) in Appendix A). Then, we incubated Nalm6 cells with RD-KLA-Gffy and the peptide controls at a concentration of 5 μmol·L−1, along with SYTOX green nucleic acid staining, and photographed the cells every ten minutes. The results showed that RD-KLA-Gffy had the greatest toxicity, with a significant difference from the control, over the first forty minutes (Figs. S6(c) and (d) in Appendix A).
We also detected the toxicity of these nanopeptides in other tumor cells that highly express integrin α6, such as leukemia cell Jurkat, nasopharyngeal carcinoma cells HNE1 and S18, esophageal carcinoma cells KYSE30 and KYSE450, colorectal carcinoma cell SW620, and cervical carcinoma cell HeLa (Fig. S7(a) in Appendix A). The results showed that the IC50 values of RD-KLA-Gffy in these tumor cells were less than 10 μmol·L−1, while many IC50 values of M4-KLA-Gffy, M5-KLA-Gffy, and RWY-KLA-Gffy were higher than 10 μmol·L−1, indicating that the cytotoxicity of these nanopeptides to tumor cells was positively correlated with the binding ability of the targeting peptides (Fig. S7(b) and Table S2 in Appendix A).
Then, we constructed several peptide controls, such as KLA-Gffy, RD-KLA, KLA, and RD-Gffy (
Fig. 4(a)), and determined their toxicity in Nalm6 cells. The data showed that the IC
50 of RD-KLA-Gffy was the lowest (5.12 μmol·L
−1), which was approximately one-third that of KLA-Gffy (14.17 μmol·L
−1) and one-quarter that of RD-KLA (22.09 μmol·L
−1) (
Fig. 4(b)), indicating that RD-KLA-Gffy possessed the strongest cytotoxicity. Moreover, after treatment for 150 min, the percentage of SYTOX-positive Nalm6 cells was 2.39% in the PBS control group, 9.09% in the KLA control group, 2.54% in the RD-Gffy control group, 25.50% in the RD-KLA control group, and 54.61% in the KLA-Gffy control group, while in the RD-KLA-Gffy group, approximately 91.78% of cells were positive (
Figs. 4(c) and
(d)). In Jurkat cells, RD-KLA-Gffy also showed fast and potent cytotoxicity (Figs. S8(a) and (b) in Appendix A).
After verifying the cytotoxicity of RD-KLA-Gffy in ALL cells, we explored the mechanism underlying the killing process. Considering that the nanopeptide contains the proapoptotic peptide KLA, we measured apoptotic markers using annexin V/propidium iodide (annexin V/PI) staining and flow cytometry analysis, as well as Western blotting assays. Flow cytometry analysis showed that 98.88% of the Nalm6 cells exhibited spreading characteristic of late-stage apoptosis after 3 h of RD-KLA-Gffy treatment. In addition, 89.23% of KLA-Gffy-treated cells and 61.58% of RD-KLA-treated cells underwent apoptosis, but the cell status in the RD-Gffy, KLA, and PBS control groups remained nearly unchanged (
Fig. 4(e)). In addition, RD-KLA-Gffy induced apoptosis in 99.98% of Jurkat cells (Fig. S8(c) in Appendix A). Subsequently, we examined the changes in the expression of several main apoptotic proteins by Western blotting and found that the expression of caspase-3, caspase-8, PARP, and BCL-2 was downregulated, while the expression of cleaved caspase-3 and cleaved PARP was upregulated (
Fig. 4(f)). Taken together, these results suggest that the leukemia cytotoxicity of the nanopeptide RD-KLA-Gffy was mainly induced by a proapoptotic effect.
3.5. The tumor-targeting ability of the integrin α6-targeted self-assembling nanopeptide RD-KLA-Gffy
To verify the antileukemic effect of the RD-KLA-Gffy nanopeptide
in vivo, we labeled the nanopeptide with Cy5 fluorescence using a Cy5-labeling kit and purified it using molecular sieving (Fig. S9 in Appendix A). Then, Cy5-RD-KLA-Gffy was injected into Nalm6-luc-EGFP tumor-bearing mice through the tail vein. Forty-eight hours later, Cy5-RD-KLA-Gffy was significantly enriched in the tumor, while the uptake in the skin or metabolic organs, such as the intestine, kidney, and bladder, was low (Fig. S10(a) in Appendix A). Moreover, the red fluorescence of Cy5-RD-KLA-Gffy and luminescence of Nalm6-luc-EGFP cells showed strong colocalization in the subcutaneous tumors. Subsequently, the mice were euthanized, and the main organs and tumors were collected and analyzed. The data also showed that Cy5-RD-KLA-Gffy was mainly distributed in tumor tissues and colocalized with luminescent Nalm6-luc-EGFP cells (
Fig. 5(a); Fig. S10(b) in Appendix A). In addition, the distribution of Nalm6-luc-EGFP cells in the head and spine of CNS-ALL mice bearing Nalm6-luc-EGFP (Fig. S11 in Appendix A) was observed by luminescence, and these cells colocalized with Cy5-RD-KLA-Gffy, indicating that Cy5-RD-KLA-Gffy could specifically target CNS-ALL (
Fig. 5(b)). In addition, Cy5-KLA-Gffy showed slight colocalization with Nalm6-luc-EGFP cells (Fig. S12 in Appendix A), but this colocalization efficiency was much weaker than that of Cy5-RD-KLA-Gffy. Fluorescence microscopy scanning of the brains of CNS-ALL mice showed that the red fluorescence of Cy5-RD-KLA-Gffy and the green fluorescence of Nalm6-luc-EGFP were highly colocalized in the meninges (
Fig. 5(c)). IHC confirmed the overexpression of integrin α6 within the CNS-ALL lesions (
Fig. 5(d)). Taken together, the results in the Nalm6-luc-EGFP subcutaneous tumor mouse model and the meningeal metastatic tumor mouse model further verified the specific targeting ability of the self-assembling nanopeptide RD-KLA-Gffy to leukemia cells, indicating ideal applicability to CNS-ALL targeted therapy.
3.6. Therapeutic effect of the nanopeptide RD-KLA-Gffy in CNS-ALL mice
Generally, peptide drugs show good biocompatibility, few adverse reactions, and low biological toxicity [
40], [
41]. Toxicity analysis
in vivo was performed during RD-KLA-Gffy treatment. The data showed that RD-KLA-Gffy treatment did not cause significant changes in body weight (Fig. S13(a) in Appendix A), and the blood biochemical indices also remained within a normal range (Table S3 in Appendix A). In addition, there were no obvious pathological lesions in the main organs, indicating that systemic side effects during RD-KLA-Gffy treatment were rare (Fig. S13(b) in Appendix A).
To explore the biological and therapeutic significance of RD-KLA-Gffy
in vivo, we examined the survival of Nalm6-luc-EGFP tumor-burdened mice that received RD-KLA-Gffy and peptide control treatment (
Fig. 6(a)). After twenty days of administration, the body weights of mice in the saline and KLA-Gffy groups decreased because of disease progression. Conversely, the body weight of mice in the RD-KLA-Gffy treatment group remained normal, suggesting that RD-KLA-Gffy suppressed disease progression (
Fig. 6(b)). IVIS imaging also showed that the luminescence signal of Nalm6-luc-EGFP in the RD-KLA-Gffy group was significantly lower than that in the control groups on the twentieth day of administration (
Figs. 6(c) and
(d); Fig. S14 in Appendix A). In addition, the median survival of the mice in the saline, RD-KLA, KLA-Gffy, and RD-KLA-Gffy groups was 14, 17, 19, and 22 d, respectively, indicating that the nanopeptide RD-KLA-Gffy effectively prolonged the survival of CNS-ALL mice (
Fig. 6(e)). To further verify the therapeutic use of RD-KLA-Gffy, we treated mice with a combination of RD-KLA-Gffy and MTX. NIRF imaging showed that Nalm6-luc-EGFP metastasis was significantly suppressed in the combination group compared with the RD-KLA-Gffy and MTX groups (
Fig. 6(f)). In addition, the median survival of the mice in the saline, RD-KLA-Gffy, MTX, and RD-KLA-Gffy + MTX groups was 17, 22, 21, and 30 d, respectively (
Fig. 6(g)). Taken together, these data obtained in CNS-ALL mice demonstrated that RD-KLA-Gffy had promise for therapeutic application as a single agent or in combination with MTX.
3.7. Sketch map
This study identified the high-affinity integrin α6-targeting peptides RWYD and RD through alanine scanning, truncation, molecular docking, and D-substitution. Based on the peptide, we designed an antileukemia self-assembled nanopeptide RD-KLA-Gffy and verified the specificity of this nanopeptide for CNS-ALL lesions. Our data demonstrated that RD-KLA-Gffy has a potent antileukemia effect through inducing cancer cell apoptosis, and the survival of CNS-ALL mice could be significantly prolonged by treatment with RD-KLA-Gffy (
Fig. 7).
4. Discussion
Integrin α6 plays an important role in leukemia. As determined from gene expression profiling studies of children with B-ALL, integrin α6 is commonly overexpressed in B-ALL and represents a potentially useful marker for the immunophenotypic detection of MRD [
42]. A comprehensive clinical study of integrins in ALL indicated that integrin α6, but not integrin α4, is associated with persistent MRD [
43]. Leukemia cells mainly exist in the meninges that cover the brain and spinal cord, and invade the brain parenchyma only in advanced stages. Integrin α6-laminin interactions mediate the migration of ALL cells toward the CSF [
26]. Treatment with the PI3K inhibitor GS-649443 or specific α6 integrin-neutralizing antibodies reduced integrin α6 expression in cultured B-ALL cells, reduced CNS involvement and prolonged the survival of B-ALL xenografted animals [
26]. The US Food and Drug Administration (FDA)-approved pan-PI3K inhibitor copanlisib downregulated B-ALL α6 integrin expression, decreased CNS invasion, inhibited bone marrow (BM) disease progression, and improved the response to chemotherapy in leukemic mice [
44]. A phase I clinical trial of copanlisib is currently underway to investigate the effects of integrin α6 expression and lymphocyte proliferation in adult patients with refractory or relapsed B-ALL (NCT04803123). Various other integrins have been linked to CNS-ALL, but no drugs targeting integrins or their downstream signaling pathways have progressed to clinical testing.
In a previous study, we reported the integrin α6-targeted peptides RWY and S5, which were employed for molecular imaging in CNS-ALL mice [
32]. Although RWY and S5 already have considerable tumor-targeting effects, both still have relatively low affinities for integrin α6 (micromolar range). To find optimized integrin α6-targeted peptides with enhanced binding affinity for integrin α6, three peptide scanning techniques, alanine scanning, truncation, and D-substitution, were used. Alanine scanning systemically replaces each peptide residue with alanine. Alanine scanning can determine permissive positions within the sequence for rational optimization because replacement with alanine is expected to have little impact on structure, and to be neutral in structure-activity relationship studies. Alanine scanning has been used extensively, particularly with the cell-permeable peptide TAT-RasGAP317-326 [
45], a vasoactive intestinal peptide [
46], and sunflower trypsin inhibitor-1 [
47]. Furthermore, truncation can identify biologically active peptides with reduced size that can be synthesized at lower cost. Truncation was applied to the antimicrobial peptide anoplin [
48], neurokinin-1 receptor tachykinin peptide substance P [
49], and secretin receptor analogous peptide [
50]. Additionally, D-substitution replaces amino acids with their enantiomers (an
L-amino acid is replaced with its corresponding
D-amino acid). D-substitution can improve the biostability and efficiency of peptide drugs. D-substitution was applied with T-cell receptor contact residues [
51], the antimicrobial peptide D enantiomer of LL-37 (D-LL-37) [
52], and peptide inhibitors of p53-murine double minute 2 and X (p53-MDM2/MDMX) interactions [
53]. Therefore, using alanine scanning and truncation techniques, we systemically designed a series of mutant peptides based on RWY and S5, and evaluated their affinities to integrin α6β4 by MST assay (
Table 1). Surprisingly, the linear peptide RWYD had nanomolar affinity for integrin α6, an increase of approximately 319-fold compared to that of the peptide RWY ((21.82 ± 3.86) nmol·L
−1 vs (6.97 ± 1.44) μmol·L
−1). Moreover, we examined the affinities of the peptides for integrin α6 by molecular docking (
Fig. 1(c)) and flow cytometry (
Fig. 2). Additionally,
D-amino acids are usually more stable than
L-amino acids; thus, the
D-amino acid sequence RD was used for the development of the self-assembling nanopeptide RD-KLA-Gffy.
Anticancer bioactive peptides (ACPs) have potential applications in cancer therapeutics [
54], [
55]. To increase their specificity and reduce their toxicity, ACPs are usually conjugated with tumor-targeting peptides. The proapoptotic peptide KLA is a well-known ACP that was first reported by Ellerby et al. [
56]. KLA is nontoxic to normal cells; however, once it is introduced into cancer cells by cell-penetrating peptides (CPPs), it can destroy the mitochondrial membrane, leading to the release of cytochrome C, promoting the cleavage and activation of caspase-3, and ultimately promoting cellular apoptosis. KLA has also been tested for toxicity in mice and appeared nontoxic over a 3-month period [
56]. In addition, KLA was conjugated with the well-known integrin αvβ3-targeting peptide RGD to form a tumor-targeting ACP with antitumor effects and low systemic toxicity [
56], [
57], [
58]. Conversely, high-dose MTX can cause significant toxicity, especially acute kidney injury (AKI), in 2%-12% of patients [
59].
Compared with linear peptides and cyclic peptides, self-assembled peptides have the advantages of good stability, targeted or controlled release, and immune enhancement [
60]. It is well-known that linear peptides are less stable than cyclic peptides, and cyclic peptides are more stable but more expensive to synthesize. The peptide self-assembly strategy can fix peptides into a specific shape in a particular environment, and increase the stability of peptides to some extent [
61]. It has also been reported that self-assembled peptides can act as adjuvants to enhance both cellular and humoral immunity [
62]. The tetrapeptide Gffy is a self-assembling peptide composed of four amino acids in the sequence
L-glycine,
D-phenylalanine,
D-phenylalanine, and
D-tyrosine. Gffy can be linked to ACPs to form water-soluble self-assembling nanoparticles, which can effectively enhance the stability of ACPs, reduce the plasma clearance rate, and promote the internalization of drugs into tumor cells [
34]. Gffy has been employed for the treatment of lung and breast cancer [
34], [
63]. Here, we recombined the proapoptotic peptide KLA, the self-assembling peptide Gffy, and the optimized integrin α6-targeted peptide RD to design the self-assembling nanopeptide RD-KLA-Gffy, which showed highly efficient antileukemia activity by inducing apoptosis (
Fig. 4). RD-KLA-Gffy was specifically enriched in leukemia lesions (
Fig. 5) and prolonged the survival of CNS-ALL mice. Importantly, the combined effect of RD-KLA-Gffy and MTX was superior to their individual therapeutic effects (
Fig. 6). Therefore, RD-KLA-Gffy can effectively improve the survival of CNS-ALL mice on the basis of the traditional clinical treatment method (MTX-based chemotherapy).
Based on the marked effect of the nanopeptide RD-KLA-Gffy combined with MTX in the treatment of CNS-ALL, we can directly conjugate RD-KLA-Gffy with MTX through a linker to form a peptide-drug conjugate (PDC), which may simplify the drug administration process and improve the therapeutic effect. Antibody-drug conjugates (ADCs) of MTX have been shown to be effective in the treatment of canine B-cell lymphoma [
64]. Compared with ADCs, PDCs have the advantages of smaller molecular weight, stronger tumor penetration, and lower immunogenicity [
65], while also crossing the blood-brain barrier into the CNS more easily [
66]. Furthermore, we can conjugate RD-KLA-Gffy or RD-Gffy with a 1,4,7,10-tetraazacyclododecane-N,N′,N″,N‴-tetraacetic acid (DOTA) chelator to form radionuclide-drug conjugates (RDCs). These RDCs can be labeled with gallium-68 for the noninvasive detection of CNS-ALL using PET imaging or labeled with lutetium-177 for the treatment of CNS-ALL. We already have considerable experience in the field of RDCs, such as the second integrin α6-targeted peptide S5, which was used for PET imaging in CNS-ALL mice [
32], and extra domain B containing fibronectin (EDB-FN) targeted probes for surgical navigation, PET imaging, and therapy in thyroid cancer [
67]. Therefore, the nanopeptides RD-KLA-Gffy or RD-Gffy have great application prospects as tools for the development of specific payload delivery systems, which might be applied for noninvasive imaging and targeted therapy with low toxicity.
This study has several limitations. First, the in vivo experiments were performed with only one cell line, Nalm6. Nalm6 was derived from a patient who suffered CNS disease relapse, and the CNS-ALL mouse model obtained by the introduction of Nalm6 is a reproducible and widely accepted model. Therefore, we used Nalm6 in the initial in vivo experiments to verify the therapeutic efficacy of the nanopeptide. However, other cell lines might be less sensitive to caspase-mediated apoptosis; therefore, we should perform mouse experiments with other ALL cell lines, such as RCH-ACV and LAX7R. Second, because lymphoblasts from patients with relapsed ALL can better reflect the clinical therapeutic effect, more experiments need to be performed before the first human trials. Third, our study describes the pathway of RD-KLA-Gffy into cells using integrin α6-mediated internalization, and the enhanced permeability and retention (EPR) effect of nanoparticles, but we have not elucidated how RD-KLA-Gffy is released and degraded. How does it escape from the endosome/lysosome? To address the above questions, we will systematically perform more experiments and report them in our next study.
5. Conclusions
In summary, we optimized a high-affinity integrin α6-targeted peptide RD and further designed an antileukemia self-assembling nanopeptide RD-KLA-Gffy. This nanopeptide induced apoptosis in leukemia cells, specifically enriched in CNS-ALL lesions, and prolonged the survival of CNS-ALL mice when used alone or combined with MTX. Thus, the integrin α6-targeted self-assembling nanopeptide RD-KLA-Gffy has promising applications in the treatment of CNS-ALL.
Acknowledgments
We thank Man-Zhi Li, Bo-Yu Yuan, Jing Wang, and Yong-Jian Peng for their excellent technical assistance. This work has been supported by grants from the National Natural Science Foundation of China (81972531, 82373175, 82102775, and 82002466), the Major Scientific and Technological Projects of Guangdong Province (2019B020202002), and the Young Talents Program of Sun Yat-sen University Cancer Center (YTP-SYSUCC-0067). The authenticity of this study has been validated by uploading the key raw data onto the Research Data Deposit (RDD) public platform†, with the approval RDD number of RDDB2023462053.
Authors’ contribution
Jia-Cong Ye, Wan-Qiong Li, Mei-Ling Chen, and Qian-Kun Shi collected, analyzed, and interpreted the data obtained from all experiments. Hua Wang, Xin-Ling Li, Ying-He Li, Jie Yang, Qiao-Li Wang, and Fang Hu provided resources and assistance. Jia-Cong Ye, Wan-Qiong Li, and Guo-Kai Feng wrote the manuscript. Yan-Feng Gao, Shu-Wen Liu, Mu-Sheng Zeng, and Guo-Kai Feng supervised the experiments and reviewed the manuscript. All the authors have read and approved the final manuscript.
Compliance with ethics guidelines
Jia-Cong Ye, Wan-Qiong Li, Mei-Ling Chen, Qian-Kun Shi, Hua Wang, Xin-Ling Li, Ying-He Li, Jie Yang, Qiao-Li Wang, Fang Hu, Yan-Feng Gao, Shu-Wen Liu, Mu-Sheng Zeng, and Guo-Kai Feng declare that they have no conflict of interest or financial conflicts to disclose.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2023.11.012.
PDF
(5893KB)
