Introduction

The tumor microenvironment (TME) plays a pivotal role in the behavior and development of solid tumors.1 Notably, tumor-associated macrophages (TAMs) suppress the recruitment and activity of tumor-killing cells such as CD8+ T and natural killer (NK) cells in the TME.2–4 Consequently, reprogramming of the TME by targeting immunosuppressive cells has emerged as an attractive research paradigm and has garnered significant clinical interest.5

TAMs are major tumor-infiltrating immune cells that constitute up to 50% of the tumor mass.6 These highly plastic cells primarily exist as M1-TAMs and M2-TAMs. While M1-TAMs kill tumor cells or secrete immunostimulatory cytokines to enhance the cytotoxic function of other leucocytes, M2-TAMs interact with tumor cells to enhance the immunosuppressive environment within the TME and contribute to all aspects of tumor progression. A high density of M2-TAMs within the TME is associated with poor prognosis in solid tumors.7 Therefore, targeting M2-TAMs may be an attractive anticancer strategy. However, the lack of specific markers for M2-TAMs has hampered the development of selective drugs that target M2-TAMs.

CD18 (ITGB2) is a complex leucocyte-specific adhesion molecule that plays a crucial role in cell adhesion and migration. CD18 combines with α subunits (CD11a, CD11b, CD11c, and CD11d) and forms heterodimers such as lymphocyte factor 1 (LFA-1, CD11a/CD18), macrophage-1 antigen (Mac-1, CD11b/CD18), complement receptor 4 (CR4, CD11c/CD18), and CD11d/CD18.8 CD18 undergoes a conformational change from a bent closed conformation to an extended open conformation on activation, which is critical for leucocyte adhesion. Circulating leucocytes in the blood predominantly express CD18 in an inactive state. However, during the inflammatory response and in the presence of chemokines, CD18 is activated, leading to leucocyte adhesion.9 10 The TME contains abundant chemokines and extracellular matrix proteins that induce Mac-1 activation.11 12 High expression of CD18 has been shown to correlate negatively with patient survival and is associated with poor prognosis in various solid tumors.13–15 Therefore, CD18, with its open conformation, can be an ideal drug target for M2-TAMs in the TME and a promising strategy for treating solid tumors.

Peptide-drug conjugate (PDC) is a promising treatment method for targeted delivery of drugs to specific cells or tissues by chemically linking a peptide that selectively binds to target cells and a drug that induces cell death.16 17 In the context of cancer immunotherapy, PDCs are developing as powerful tools to selectively modulate the TME.18 Recently, PDCs are being developed that can specifically target M2-like TAMs, which play a key role in tumor progression and immune evasion. However, despite these promising advances, the field of PDC is still in its infancy, and there are many challenges to overcome.19 20 One such challenge is efficient transport of PDCs across biological barriers, which is critical for therapeutic efficacy.21 This requires further research into peptide carriers that could potentially be used to enhance PDC delivery.22 In previous studies, we developed a peptide (TAMpep) that preferentially binds to M2 macrophages.23 However, their binding mechanism was unclear.

This study aimed to identify the target and precise mechanism of action of PDC (TB511), comprising a TAM-binding peptide (TAMpep) and a pro-apoptotic peptide (dKLA), in solid tumor models including lung and colon cancer. The findings of this study could provide valuable insights for future clinical therapies. Specifically, comprehending the conformational changes of CD18 in M2 TAMs within the TME is crucial for elucidating its correlation with tumor progression and recognizing its potential as a significant target protein for cancer therapy; the findings of this study could provide valuable insights for future clinical therapies.

Materials and methodsPeptide synthesis

Peptides used in this study were synthesized by GenScript (Piscataway, New Jersey, USA) and are detailed in online supplemental table S1. All peptides were purified to greater than 95% purity. Moreover, the structure of TB511 was analyzed using Fourier-transform infrared spectroscopy and Circular Dichroism (CD) spectroscopy.

Immunofluorescence staining

The tumor tissues were fixed overnight in 10% neutral-buffered formalin. After embedding in paraffin, tumor tissues were cut into 5 µm of regular thickness. The slides were deparaffinized in xylene and dehydrated in ethanol and deionized water. For antigen retrieval, slides were incubated in sodium citrate buffer (pH 6.0) for 15 min in a microwave. The slides were then incubated with a peroxidase-blocking solution (Dako, Glostrup, Denmark) for 15 min and blocked with 5% bovine serum albumin. Slides were incubated with primary antibodies overnight, washed with Tris-buffered saline (TBS) containing 0.1% Tween20. Slides were then incubated with primary antibodies overnight at 4℃. Slides were then stained with FITC-labeled TAMpep and APC-conjugated anti-rabbit IgG secondary antibody (Invitrogen) at 37℃ for 1 hour. To visualize the nuclei, the cover glasses were mounted with Vectashield mounting medium (Vector Laboratories) with 4′,6-diamidino-2-phenylindole. Images were captured using the LSM 800 confocal laser-scanning microscope (Carl Zeiss, Oberkochen, Germany). The antibodies used are listed in online supplemental table S2.

Annexin V assay

THP-1 cells or Jurkat were incubated with 1 mM Mn2+ for 1 hour at 37℃ for CD18 activation. The activated cells were incubated with 1 µM TB511 for 1 hour at 37℃ and then cultured in serum-free medium for 24 hours. Cells were harvested and stained with APC-conjugated Annexin V (Invitrogen) for 15 min at RT. After centrifugation at 300×g for 5 min at 4℃, cells were treated with 7-AAD Viability Staining Solution (Invitrogen) containing BD Pharmingen Stain Buffer (FBS) (BD Bioscience). All data were acquired using FACSLyric (BD Bioscience) and analyzed using FACSuite software (BD Bioscience).

Live cell imaging fluorescence microscopy

To measure apoptosis induced by TB511, THP-1 cells were seeded at 2 × 104 cells/well in 96-well plates and differentiated into M0, M1, M2, or TAMs. Differentiated cells were treated with MitoTracker RedROX (Invitrogen) and a caspase 3/7 reagent for 30 min. Next, the cells were treated with TB511 (1 µM) and then immediately evaluated using restoration fluorescence microscopy (DeltaVison, New Jersey, USA).

Biotin pull-down assay

Lysates of M2-differentiated THP-1 cells were obtained using a Membrane Protein Extraction Kit (Thermo Fisher Scientific). Cell lysates were incubated with scrambled and TAMpep peptides (0, 0.001, 0.01, 0.1, 1, and 10 µg) and then with biotinylated TAMpep for 1 hour at 4℃ with shaking. After incubation, the peptide-bound membrane proteins were eluted using streptavidin Dynabeads (Thermo Fisher Scientific). Eluted proteins were loaded onto a 10% sodium dodecyl sulfate-polyacrylamide gel and transferred onto polyvinylidene difluoride membranes. The membrane was blocked and incubated with anti-CD18 antibody (LSbio) overnight at 4℃. After washing with TBS containing 25 mM Tris-Cl, 150 mM NaCl, and 0.05% Tween-20, the membranes were incubated with anti-rabbit horseradish peroxidase-conjugated secondary antibodies. Protein bands were visualized using a Western Blotting Detection Reagent Kit (Thermo Fisher Scientific) and analyzed using ImageJ software (NIH, Bethesda, Maryland, USA).

Surface plasmon resonance

The binding affinity of total CD18 (Prospec), the I-domain (Cloud-Clone), the cys-rich domain (Cloud-Clone), and recombinant Myelin Basic Protein (MBP; Cloud-Clone, #RPA539Mu01) to TAMpep was measured using surface plasmon resonance (SPR) on a Biacore T200 system (GE Healthcare, Freiburg, Germany). Before the experiment, the surface of the series S sensor chip SA was cleaned with injections of 1 M NaCl and 50 mM NaOH solutions, three times for 1 min each. TAMpep (10 nM) was immobilized on the SA chip surface at a flow rate of 10 µL/min. The surface was subsequently washed with 50% isopropanol, 1 M NaCl, and 50 mM NaOH to remove non-specific binding. For the analyte binding studies, total CD18, I-domain, cys-rich domain, and MBP protein were prepared at concentrations achieving saturated responses below 150 response units (RU). Each analyte was then serially diluted (twofold) four to five times. The multicycle analysis method was used to inject different concentrations of analytes at a flow rate of 30 µL/min to measure binding kinetics. Kinetic data were analyzed using BIA evaluation software (V.3.2.1, GE Healthcare), and the equilibrium dissociation constant (KD) was calculated as KD=kd/ka, where ka is the association rate constant and kd is the dissociation rate constant. Confidence in the results was validated by U-values (<15) for all analyses, except for total CD18.

Immunohistochemistry

For histological analysis, the brain, breast, colon, kidney, liver, lung, skin, and tumor tissues were fixed overnight in 10% neutral-buffered formalin. After embedding in paraffin, tumor tissues were cut into 4 µm of regular thickness. The slides were deparaffinized in xylene and dehydrated in ethanol and deionized water. For antigen retrieval, slides were incubated in sodium citrate buffer (pH 6.0) for 15 min in a microwave. The slides were then incubated with a peroxidase-blocking solution (Dako, Glostrup, Denmark, #S202386) for 15 min and blocked with 5% bovine serum albumin. Slides were incubated with primary antibodies overnight, washed with TBS containing 0.1% Tween20, incubated with an avidin-biotin complex kit (Vector Laboratories, Burlingame, CA, USA; #PK-6101), and visualized by incubation with diaminobenzidine-HCL (Vector laboratories, #SK4100). Nuclei were counterstained with hematoxylin. Slides were observed under a bright-field microscope (Nikon, Tokyo, Japan) and analyzed using ImageJ software (NIH). Antibodies used for immunohistochemistry (IHC) staining are listed in the online supplemental table S2.

Modeling of protein-peptide interactions

Based on the sequence of the cysteine-rich region of CD18 (Uniport _p05107, cysteine-rich tandem repeat region: 449-617), a tertiary structure was constructed using the trRosetta algorithm (https://yanglab.nankai.edu.cn/trRosetta/). Docking simulations of TB511 and CD18 were performed using the CABS-dock server (http://biocomp.chem.uw.edu.pl/CABSdock/) with the CD18 PDB file generated using trRosetta and the amino acid sequence of TB511. Among the 10 models created from the docking simulation, a representative model was selected based on the results of alanine substitutions in the TB511 sequence.

Single particle tracking imaging

TB511-AuNRs were synthesized with AuNRs and indicated in online supplemental methods. THP-1 cells were seeded on 22×22 mm coverslips (No. 1, Deckglaser, Freiburg, Germany) and housed in Petri dishes. After differentiating into M2 macrophages, the concentration of TB511-AuNRs was adjusted to ~1.5×109 particles/mL. The cells were incubated with 20 µL of TB511-AuNRs for 1 hour, followed by incubation with MitoTracker Green (Invitrogen) and Hoechst 33 342 (Invitrogen) for 0.5 hour. An 18×18 mm coverslip was attached to the cell coverslip using double-sided tape. The space was filled with a cell culture medium to sustain the cells. Then, single particle tracking images were acquired with a 10 ms interval by integrated multidimensional light-sheet microscopy (iMLSM). An Olympus BX53 upright microscope was used to construct the iMLSM.24 ImageJ (NIH) and MATLAB (Mathworks, Torrance, California, USA) were used to analyze the acquired images. The 3D spatial trajectories (x, y, z) of individual peptide-AuNRs were super-localized with the astigmatism method.25 26 The rotations and orientations of individual peptide-AuNRs were calculated by fitting cos2φ for transverse SPR and sin2θ for longitudinal SPR.27

Animal study

To create a subcutaneous lung cancer mouse model, tumor cells were resuspended in a serum-free medium and mixed with 1:1 Matrigel (Corning, New York, USA). LLC cells were subcutaneously inoculated into the right flank of 6–8 weeks old C57BL/6 (Taconic Biosciences, New York, USA). When the tumor volume reached 50–100 mm3, the tumor-bearing mice were divided into two groups. TB511 has shown plasma stability for approximately 60 hours; therefore, it was administered every 3 days in the tumor mouse model. TB511 was injected subcutaneously at a dose of 200 nmol/kg every 3 days. Tumor diameters were measured using a digital caliper, and tumor volume was calculated according to the following formula: (length×width2)/2. All animal studies were terminated when the tumor volume reached <2000 mm3. For the orthotopic mouse model, the fur in the abdominal area of the mice was shaved with clippers, and the mice were anesthetized with isoflurane. The abdomen of the mice was excised using a sterile surgical scalpel, and CT-26 Luc cells were injected into the subserosal plane of the cecum. The surgical incision site was closed using sutures. 7 days after tumor inoculation, the mice were randomly divided into two groups. TB511 was injected subcutaneously every 3 days. D-luciferin (BioVision, Milpitas, California, USA) was injected intraperitoneally at a dose of 3 mg/mouse to track tumor growth. Mice were evaluated using the NightOWL LB 983 in vivo imaging system (Berthold Technologies, Bad Wildbad, Germany) once a week. All mice were sacrificed, and tumor tissues were isolated.

To assess whether CD18 is a target protein of TB511, BALB/c wild-type (WT) mice and CD18 knockout mice (C.129S7(B6)-Itgb2tm2Bay/AbhmJ, 6–8 weeks old) were purchased from the Jackson Laboratory. For the subcutaneous mouse model of lung cancer, LLC cells were inoculated with 1×106 cells per mouse into the right flank. After 7 days, TB511 was subcutaneously administered twice per week. The tumor volume was measured twice weekly using a digital caliper. The mice were euthanized after four injections. In an orthotopic mouse model of colorectal cancer (CRC), CT26-luc cells were injected into the subserosal plane of the cecum. After 7 days, TB511 was subcutaneously injected every 3 days and the tumor volume was measured weekly using an in vivo imaging system. The mice were sacrificed after 4 weeks, and the tumor tissues were extracted.

Generation of humanized tumor mice model

SID mice (NOD-Prkdcem1BeakIL2rgem1Break, 4-week-old) were purchased from the GemBioscience (Chungbuk, Korea).28 Mice were irradiated with 1.5 Gy of γ-rays and were intravenously injected with human CD34+ hematopoietic stem cells (2×105 cells/mouse, Lonza, Basel, Switzerland). After 11 weeks, mouse peripheral blood was obtained, and the engraftment of mature human white blood cells (human CD45+ cells) was confirmed using flow cytometry. Humanized mice were considered to have more than 25% human CD45 cells in their peripheral blood; the hCD45+ cell population accounted for 37.32%±13.44% in 16 male mice and 42.63%±10.41% in 15 female mice 11 weeks after engraftment. A549, a human lung cancer cell line, and PANC1, a human pancreatic cancer cell line, were subcutaneously injected into the right flank of humanized mice. When the tumor volume reached 50–100 mm3, PBS or TB511 was subcutaneously injected, and the tumor volume was measured every 3 days. At the end of the animal study, the mice were sacrificed, and tumor tissues were extracted to analyze the changes in immune cells in the TME.

Flow cytometry analysis

Tumor tissues were minced into thin pieces using a MACS dissociator (Milteny Biotec, Auburn, California, USA) and incubated with 1 U/mL DNase I (Roche, Indianapolis, Indiana, USA) and 1 mg/mL collagenase D (Roche) in RPMI 1640 serum-free media for 30 min at 37°C. The tissue was filtered through a 100 µm nylon mesh strainer. The RBCs in the filtered single cells were lysed with BD Pharm lyse buffer (BD Bioscience, California, USA). The single cells were then stained with BD Pharmingen Stain Buffer (BD Bioscience) with antibodies, which are listed in online supplemental table S3. The cells were detected using a FACS Lyric System (BD Bioscience) and analyzed using FlowJo software (BD Bioscience).

To evaluate CD18-activated macrophages in tissues, the brain, breast, colon, kidney, liver, lung, skin, and tumor tissues were made into single cells. The single cells were then stained with antibodies, which are listed in online supplemental table S3. The cells were detected using a FACS Lyric System (BD Bioscience) and analyzed using FlowJo software (BD Bioscience).

For assessing CD18 and KIM127 target levels, flow cytometric analysis was performed using a BD FACSLyric flow cytometer (BD Bioscience; equipped with six lasers—510, 525, 575, 774, 603, and 421 nm) with FlowJo software (BD Bioscience). To estimate the absolute number of membrane-bound CD18 and KIM127 molecules, we used Quantibrite beads (information provided with Quantibrite PE beads). The analysis of Quantibrite PE calibration beads is Log10 values for PE fluorescence geometric means and for values of PE molecules per were calculated. Using the values obtained, we plotted the log10 values for the number of PE molecules per bead against the log10 values for fluorescence intensity. Given that the PE/antibody ratio was 1:1, PE/antibody values could be converted to antibody/cell values, which represented the number of receptors per cell. We gated the populations to be analyzed on the basis of indices of forward and side scattering (FSC-A/SSC-A dot plot) that were located within the leucocytic regions. Subsequently, we gated subpopulations of CD4+T lymphocytes or CD8+ T lymphocytes and CD11b+macrophage (dot plots APC-A/FITC-A and APC-A/PE-Cy7-A) on the basis of markers of the different subpopulations. A minimum of 10 000 events were gated.

Spatial transcriptomics

Paraffin-embedded blocks were carefully sectioned at a thickness of 5 µm and mounted onto a Xenium slide. The sectioned slide was incubated for 3 hours at 42°C on a thermal cycler (C1000 96-well, Bio-Rad) for drying. Deparaffinization and de-crosslinking were performed. After prime hybridization and RNase treatment, the sectioned slide was hybridized with the Xenium human 5 k pan-tissue probes (10 x Genomics) at 50°C overnight (~24 hours). The sectioned slide was then washed, and ligation was performed at 42°C for 30 min. Amplification enhancement and amplification were immediately processed on the thermal cycler (C1000 96-well, Bio-Rad). The sectioned slide was then washed, and cell segmentation staining probes were applied to the slide at 4°C overnight (~24 hours). Following stain enhancement and washing, autofluorescence quenching and nuclei staining were performed on the sectioned slide. Subsequently, the sectioned slide was loaded onto the Xenium analyzer (XOA V.3.0.2, 10x Genomics) to begin fluorescence signal detection. After the Xenium run, raw data was preprocessed with xeniumranger V.3.0.0.

Statistics

All data are presented as the mean and SEMs. Statistical significance was analyzed using unpaired t-test and one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test, and two-way ANOVA followed by Bonferroni post hoc comparisons using Prism software (GraphPad Software, California, USA). A p<0.05 was considered statistically significant.

ResultsExpression of activated CD18 in M2 TAMs

Numerous studies have highlighted a significant association between CD18 and the progression of solid tumors, particularly in relation to the infiltration of M2 macrophages into tumor tissues.13–15 We investigated the expression levels of CD18 in macrophage subtypes. Both gene and protein expression of CD18 were significantly higher in M2 macrophages and TAMs compared with M0 and M1 macrophages (figure 1A–C). To determine the correlation between CD18 expression and tumor progression, we conducted an analysis using the Tumor IMmune Estimation Resource analysis and a tumor tissue array. In a human tissue array, the level of CD18 was significantly higher in tumor tissues than in normal tissues, including the breast, colon, lung, liver, kidney, and skin (figure 1D). Analysis of gene expression patterns in various solid tumor tissues revealed that the expression of ITGB2/CD18 genes and marker genes of M2 macrophages (CD163, AGR and TGFb) were upregulated, and there was a significant correlation between the expression of ITGB2/CD18 and MRC1, one of the markers of M2 macrophages (online supplemental figure 1A,B). CD18, known to effectively mediate leucocyte trafficking and immune response, transitions from an inactive to an active form through a conformational change, thereby exhibiting a high affinity for ligands29 (figure 1E). To determine the difference in the number of activated CD18 molecules per cell in human immune cells, CD4, CD8 T cells, neutrophils, and macrophages derived from human PBMCs were quantified by conjugating PE to the KIM127 antibody, which is known to specifically bind to the activated CD18. As a result, TAMs and M2 macrophages showed higher CD18 expression compared with other immune cells (figure 1F). Activated CD18 showed the highest level in TAMs compared with other cells as well as M2 macrophages (figure 1G). Additionally, the THP-1 derived TAMs exhibited an increased activation level of CD18 among the macrophage lineage cells (online supplemental figure 1C,D). Furthermore, the activation level of CD18 (pan CD18 and KIM127-double positive) in human tissues was significantly elevated in tumor tissues compared with normal tissues in the breast, colon, lung, and liver (figure 1H). Therefore, these results suggest that activated CD18 is highly expressed in M2-TAMs derived from the TME.

Figure 1

Association of activated CD18 and M2 TAMs in progressed human tumor tissues. (A) The relative mRNA expression levels of CD18 in M0, M1, M2, and M2-TAM macrophages were measured using qRT-PCR. (B) The protein expression levels of CD18 in M0, M1, M2, and M2-TAM were determined using western blotting. (C) Immunostaining images of CD18+ cells in M0, M1, M2, and M2-TAM macrophages were marked with CD18 (red) and nuclei (DAPI, blue). (D) Expression of CD18 in various normal and cancerous tissues, including the breast, lung, liver, kidney, and colon was measured using tumor tissue microarrays. CD18 expression was quantified per area (μm2); scale bar, 500 µm. (E) Schematic diagram illustrating the inactive and active forms of the CD18 protein. (F, G) The degree of expression or activated CD18 on CD4, CD8 T cells, neutrophils, M1, and M2 macrophages derived human PBMCs was measured by staining with anti-CD18 antibody and KIM127 and using flow cytometry. (H) Double-positive staining with pan-CD18 (anti-CD18 antibody-PE) and KIM127-FITC indicates activation of CD18 in normal and tumor tissues, including the breast, colon, lung, and liver. Nuclei (DAPI) were counterstained. Arrows indicate double-positive cells. All data are presented as the mean±SEM. n=3–5, *p<0.05, **p<0.01, ***p<0.001. TAMs, tumor-associated macrophages.

CD18 as a target protein of TB511 in M2-TAMs

We developed a PDC (TB511) comprizing TAMpep, a peptide with preferential binding affinity for M2 macrophages, and dKLA, a pro-apoptotic peptide, linked via a GGGGS linker (online supplemental figure 2).23 However, the precise mechanism by which TB511 selectively targets M2 macrophages remains undisclosed. To elucidate the mechanism of action of TB511, we verified the target proteins of TB511 using immunoprecipitation and a proteomics-based analysis. As a result, 21 cell membrane proteins were identified in M0, M1, and M2 macrophages using LC-MS/MS. Among these proteins, the ITGB2/CD18 protein exhibited preferential binding to biotin-labeled TAMpep (carrier peptide of TB511) in M2 macrophages compared with M0 and M1 macrophages (online supplemental figure 3). The binding of the CD18 protein with TAMpep in cell membrane proteins of M2 macrophages was blocked by preincubation with non-biotinylated TAMpep in a concentration-dependent manner, whereas preincubation with scrambled peptides was not affecting immunoprecipitation (figure 2A). Colocalization of TAMpep and CD18 on M2 macrophages was confirmed by confocal microscopy (figure 2B). To identify the binding domain of CD18, we measured the affinity of TAMpep with the extracellular domains of CD18: total CD18, I-domain, and cysteine (cys)-rich domain (figure 2C). As a result, TAMpep exhibited a high binding affinity to the cys-rich domain protein compared with I-domain and total CD18 proteins (figure 2D). In the SPR analysis, TAMpep showed higher affinity in the cys-rich domain (KD:0.185 nM) compared with total CD18 (KD:194.8 nM), I-domain (KD:4.701 nM), and I-domain taq protein, MBP protein (KD:3.144 nM). Notably, other proteins except cysteine-rich protein exhibited low RU values (<70) (figure 2E). We also predicted the binding sites of TAMpep to CD18 using peptide-protein modeling (CABS-dock server). While the binding of TAMpep to the total CD18 sequence of the bent (closed) form was not predicted, the leucine6 residue of TAMPep binds to leucine (528), threonine (538), leucine (556), cysteine (557) and phenylalanine (558) of the cysteine-rich domain of the open form, and the tryptophan12 residue predicted to bind to glutamic acid (466), isoleucine (469), cystein (470) and arginine (471) (figure 2F). To determine the binding site of TAMpep for CD18, we confirmed the binding affinity of the TAMpep, in which amino acids of L6 and W12 were substituted with alanine, on M2 macrophages and cys-rich domain protein. Peptides substituted with alanine showed a markedly reduced affinity to M2 macrophages and cys-rich domain protein (figure 2G,H). Therefore, our results suggest that TAMpep of TB511 binds to the cysteine-rich domain of activated CD18, and in particular, Lys6 and Trp12 of TAMpep are the key binding residues.

Figure 2

CD18 as a target protein of TB511 in M2-TAMs. (A) M2 macrophage lysates treated with scramble or TAMpep peptides (0–10 ug) were incubated with biotinylated TAMpep for 1 hour at RT. Biotinylated TAMpep-bound proteins were eluted by a pull-down assay, followed by measuring CD18 protein levels via western blotting. (B) Immunofluorescence staining showed the colocalization of TAMpep (green) and CD18 (red). Nuclei are counterstained with DAPI (blue); Scale bar, 20 µm. (C) Schematic diagram illustrating the inactive and active forms of the CD18 protein. (D) The affinity of total CD18, I-domain, and Cys-rich domain with TAMpep was evaluated by ELISA. (E) The binding kinetics of total CD18, I-domain, and Cys-rich domain with TAMpep were measured using a Biacore T-200. The equilibrium dissociation constant (K/D) was calculated. (F) Docking simulation of TB511 and CD18 (cysteine-rich region) by the CABS-dock server. (G) Assessment of FITC-TAMpep and alanine-substituted peptides binding to M2 macrophages or (H) to CD18 protein (Cys-rich domain) via flow cytometry or ELISA, respectively. The absorbance of each sample in the ELISA was determined using a fluorescence microplate reader and normalized to its intrinsic fluorescence value. All data are presented as the mean±SEM. *p<0.05, **p<0.01, ***p<0.001. AMs, tumor-associated macrophages.

TB511 targeting activated CD18 of M2 macrophages

To confirm whether TAMpep, the peptide carrier of TB511, preferentially attaches to M2 macrophages, TAMpep was synthesized by conjugation with FITC. TAMpep showed higher binding ability in M2 macrophages and TAMs compared with M0 and M1 type macrophages (figure 3A). TB511 comprised TAMpep and dKLA and showed cytotoxicity (IC50: 4.925 µM) against M2 macrophages, but TAMpep did not show cytotoxicity (figure 3B). To assess the impact of TB511 on apoptosis, we measured the activation of caspase 3 in each macrophage subtype using live-cell imaging. Activation of caspase 3/7 by TB511 was detected within 5 min and peaked at 15 min after incubation with TB511 in M2-TAMs. Although TB511 affected M2 macrophages, its effect was less pronounced. Notably, caspase 3/7 activation was not observed in M0 or M1 macrophages and was only mildly affected in M2 macrophages following TB511 treatment (figure 3C,D; online supplemental movie S1). Furthermore, the double positive expression of CD11b and TB511 in human tissues was significantly elevated in tumor tissues compared with normal tissues in the breast, lung, liver, kidney, and colon (figure 3E). To investigate whether TB511 selectively targets macrophages expressing activated CD18, we induced the activation of CD18 in T cells (Jurkat) and macrophages (THP-1) using Mn2+ and subsequently measured the cytotoxicity of TB511. The number of positive cells for the activated CD18 was higher in macrophages than in T cells when treated with Mn2+ (figure 3F). On treatment with TB511, the number of apoptotic cells significantly increased in activated macrophages, whereas the number of apoptotic cells in activated T cells remained unchanged compared with inactivated cells (figure 3G). Moreover, TB511 induced more apoptosis in macrophages than in other immune cells such as T cells, NK cells, dendritic cells, and neutrophils obtained from the spleen of mouse (figure 3H). Therefore, these results suggest that TB511 induces apoptosis by primarily targeting M2-TAMs, which exhibit high expression levels of activated CD18.

Figure 3

TB511 targeting activated CD18 of M2 macrophages. (A) THP-1 cells were differentiated into MO, M1, or M2 macrophages. Cells were treated with FITC-conjugated TAMpep. The affinity of TAMpep and cells was measured using flow cytometry. (B) The M2 macrophages were treated with TAMpep or TB511 by concentration from 0.01 uM to 20 uM. Cell viability was measured by CCK-8 assays. All data are presented as the mean±SEM. n=4 ***p<0.001. (C, D) Activation of caspase 3/7 (green) was observed using live-cell imaging fluorescence microscopy. The mitochondria were counterstained with MitoTracker (red). The total intensity of the area is graphed. (E) The colocalization of TB511 (yellow) with CD11b+cells (green) was assessed via confocal microscopy in both normal and tumor tissues, including those from breast, lung, liver, kidney, and prostate. Scale bar, 100 µm. (F) Jurkat or THP-1 cells were treated with 1 mM Mn2+ for 1 hour at 37°C. The degree of activated CD18 in cells was stained with mAb KIM127 and analyzed using flow cytometry. (G) Activated Jurkat or THP-1 cells were treated with TB511 (1 uM) for 1 hour. Cell viability was measured by flow cytometry using Annexin V (APC) and 7AAD staining. (H) Macrophages (F4/80+cells), T cells (CD3+cells), NK (NKp46+cells), dendritic cells (CD11c+cells), and neutrophils (Gr-1+cells) were treated with 1 µM TB511. The cell death was measured by Annexin V staining and was analyzed using flow cytometry. All data are presented as the mean±SEM. n=3–5, *p<0.05, **p<0.01, ***p<0.001.

Endocytosis and endosome escape of TB511 into M2 macrophages

Recently, the elucidation of mechanisms of action for the cytoplasmic delivery efficiency of drugs using cell-penetrating molecules and endosomal escape has emerged as a significant area of research.30 To monitor the dynamics of the cell membrane and intracellular transport of TB511 in M2 macrophages, we conjugated TB511 (TB511-AuNRs) with carboxylic acid-functionalized gold nanorods. These were then tracked using astigmatism imaging and polarized light-sheet illumination, allowing us to resolve the 3D spatial trajectory and rotational behavior of TB511-AuNRs in M2 macrophages. Differential interference contrast and super-resolution radial fluctuations images revealed the intracellular distribution of mitochondria (green) and nuclei (blue) in M2 macrophages (figure 4A). On contact with the cell membrane, the lateral movements and rotations of TB511-AuNRS were limited due to the viscosity of the cell membrane and the interaction between them.31 The gradual loss of displacements in x, y, and z directions, along with scattering intensity fluctuations, indicated the binding of TB511-AuNRs to the receptors on the cell membrane (figure 4B,C). At certain times, TB511-AuNRs exhibited reduced spatial movement and slow orientation changes, suggesting that TB511-AuNRs were enveloped by the cell membrane to assemble the endocytic machinery.32 Following these quiet periods, noticeable spatial movement in the x, y, and z directions was observed, indicating that the TB511-AuNRs were detached from the membrane and transported intracellularly. The representative rotational angles and corresponding scattering images of TB511-AuNRs in M2 macrophages provide insight into the rotational states of TB511-AuNRs during their internalization processes (figure 4D; online supplemental movie S2). To investigate the cellular uptake of TB511, M2 macrophages were treated with endocytosis inhibitors: chlorpromazine (CPZ) to inhibit clathrin-mediated endocytosis (CME), methyl-β-cyclodextrin (MβCD) to inhibit caveolae-mediated endocytosis, and cytochalasin D (CyD) to inhibit macropinocytosis (MP). Both CPZ and CyD, but not MβCD, inhibited TB511 uptake, thereby decreasing TB511-induced apoptosis (figure 4E,F). Additionally, M2 macrophages were treated with bafilomycin A1 (Baf-A1) and chloroquine (CQ) to inhibit the acidification of early endosomes and prevent the maturation and fusion of endosomes and lysosomes. Subsequently, we evaluated the colocalization of early endosome antigen 1 (EEA1) and TB511-Cy5 in endosomes. Treatment with Baf-A1 and CQ increased the colocalization of EEA1 and TB511-Cy5 in endosomes and significantly decreased TB511-induced apoptosis (figure 4G,H). Therefore, these results suggest that TB511 is internalized into cells via CME and micropinocytosis and subsequently exits from the endosomes into the cytoplasm.

Figure 4

Endocytosis and endosome escape of TB511 into M2 macrophages. (A) DIC and fluorescence-based SRRF images of M2-differentiated THP-1 macrophages stained with mitochondria (green) and nuclei (blue), with TB511-AuNRs endocytosis occurring in the white frame. Red and yellow dotted lines indicate the contours of the cell and nucleus, respectively. (B) 3D tracking trajectory of TB511-AuNRs in the same region (white frame in (A)) of live single M2 macrophages using iMLSM. The circular dot denotes the starting point, and the red triangle denotes the movement endpoint. (C) Corresponding displacements of TB511-AuNRs in the x, y, and z-directions. Description of different rotational states of single TB511-AuNRs on the upper surface of live single M2 macrophages using iMLSM. (D) Description of the rotational angle and corresponding scattering image of TB511-AuNRs in M2 macrophages. (E) M2 macrophages were pretreated with CPZ, CyD, and MβCD, followed by TB511-Cy5 (20 µM) treatment and anti-EEA 1 antibody staining, revealing TB511 (red) and endosomes (green) colocalization. (F) M2 macrophages treated with CPZ, CyD, and MβCD, followed by TB511 (1 µM), underwent CCK8 assays to assess cell viability. (G) M2 macrophages, pretreated with Baf-A1 or CQ, were treated with TB511-Cy5 and stained with anti-EEA 1 antibody. Colocalization of TB511-Cy5 (red) and endosomes (green) was observed using confocal microscopy. (H) M2 macrophages pretreated with Baf-A1 or CQ were treated with TB511 (1 µM) for 1 hour. Apoptosis of cells induced by TB511 was measured using CCK8 assays. All data are shown as the mean±SEM. n=3, **p<0.01, ***p<0.001. DIC, differential interference contrast; iMLSM, integrated multidimensional light-sheet microscopy; LSPR, longitudinal surface plasmon resonance; SRRF, super-resolution radial fluctuations; TSPR, transverse surface plasmon resonance.

Abrogation of the antitumoral effect of TB511 by CD18 deficiency

To determine whether TB511 induces apoptosis in M2 macrophages through CD18, we inhibited CD18 expression in M2 macrophages using CD18/sgRNA in cas9 stable THP-1 cells and anti-CD18 antibody. Consequently, the cytotoxicity of TB511 was significantly diminished in CD18/sgRNA-transfected cells compared with mock-transfected cells (online supplemental figure 4A,B). Furthermore, blocking of CD18 using antibodies attenuated TB511-induced apoptosis in M2 macrophages, accompanied by a downregulation in the gene expression of caspase 3, 8, and 9 (online supplemental figure 4C,D). We also found that the colocalization of TB511 with mitochondria, including the dKLA peptide that binds to mitochondrial membranes to induce cell death, was inhibited by anti-CD18 antibodies (online supplemental figure 4E). Moreover, antibody-mediated blockade of CD18 reduced TB511-induced expression of cytochrome c and caspase 3 in M2 macrophages (online supplemental figure 4F,G). To verify whether TB511 exerts its antitumor effect via CD18 of TAMs, we examined the efficacy of TB511 using CD18-deficient mice. As a result, the cytotoxicity of TB511 was diminished in bone marrow-derived macrophages obtained from CD18-KO mice, and the expression of cleaved-caspase 3 was also suppressed (figure 5A,B). In CRC model, TB511 significantly reduced the tumor size in WT mice, but these effects were not observed in CD18-KO mice (figure 5C). The number of M2-TAMs (F4/80+ CD206+/CD45+ cells) significantly decreased in WT tissues, whereas the populations of M2-TAMs (F4/80+ CD206+/CD45+ cells) did not change in response to TB511 in CD18-KO mice (figure 5D). In non-small cell lung cancer (NSCLC) model, the antitumoral effects of TB511 were only evident in the WT mice and not in CD18-KO mice (figure 5E). Like the results in CRC mouse model, M2-TAMs were reduced by TB5111 only in WT mice, with no change observed in CD18 KO mice (figure 5F). Based on these findings, we suggest that CD18 plays a crucial role in the action of TB511 for targeting M2 TAMs in TME of CRC and NSCLC and may be a potent target protein for cancer treatment.