High-affinity peptide ligand LXY30 for targeting α3β1 integrin in non-small cell lung cancer

Peptide-biotin and peptide-fluorescein isothiocyanate (FITC) were designed by attaching biotin or FITC to the side chain of Lys and two hydrophilic linkers between peptide and Lys (biotin) and Lys (FITC) on Rink amide MBHA resin as previously described [20]. Standard solid-phase peptide synthesis approach using Fmoc/tBu chemistry and 6-chloro-N-hydroxybenzotriazole (6-Cl HOBt)/1,3-diisopropylcarbodiimide (DIC) coupling was employed to synthesize the linear peptides and peptide conjugates on Rink amide resin (loading 0.52 mmol/g) (GL Biochem, Shanghai, China) as described [17, 19, 21]. The disulfide formation was achieved using CLEAR-OX^TM resin (Peptide International Inc, Louisville, KY) in 50% of 0.1 M ammonium acetate buffer in acetonitrile (ACN). The collected liquid was lyophilized to yield crude products that were purified by reversed-phase HPLC with a purity of > 95%. The identities of compounds were confirmed with matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS). Analytical HPLC was performed on a Waters 2996 HPLC system equipped with a 4.6 × 150 mm Waters Xterra MS C18 5.0 μm column and employed a 20-min gradient from 100% aqueous H₂O (0.1% trifluoroacetic acid, TFA) to 100% ACN (0.1% TFA) at a flow rate of 1.0 mL/min. Preparative HPLC was performed on a System Gold 126NMP solvent module (Beckman) with a C18 column (Vydac, 10 μm, 2.2 cm i.d. × 25 cm). A gradient elution of 0–60% B over 45 min, then 60–100% B over 5 min, and followed by 100% B for 5 min was used at a flow rate of 5 mL/min (solvent A, H₂O/0.1% TFA; B, acetonitrile/0.1% TFA).

Human NSCLC cell lines H1975 and A549 were obtained from American Type Culture Collection (Manassas, VA). H3255 was a gift from Dr. Pasi A. Janne (Dana-Farber Cancer Institute, Boston, MA). H3255 R#2 is the best-characterized resistant clone mimicking the acquired resistance model of EGFR-mutant NSCLC to first-generation EGFR TKI erlotinib [22]. Patient biospecimens were collected under an institutional review board (IRB)-approved protocol (Protocol No. 226210) at the University of California, Davis.

Our group recently isolated and comprehensively characterized LXY30-enriched, nanosized extracellular vesicles (EVs) from ovarian tumor cells with a composition reflecting the cells’ biological state [23]. The collected EVs were confirmed as exosomes according to the International Society for Extracellular Vesicles (ISEV) suggested standards [24, 25]. Standard differential centrifugation protocols were used to isolate exosomes from both cancer cell culture supernatant and pleural effusion [26, 27]. Tumor cells were cultured in exosome-free medium in T75-cm² flasks for 3 days until they reached a confluency of 70–80%. The media or supernatant from patient’s pleural effusion specimens was collected and centrifuged at 2,000g for 20 min followed by 10,000g for 30 min to remove the cellular debris. The resulting media or supernatant samples were filtered through a 0.22-µM filter (Millipore, Boston, MA), followed by being ultrafiltered through Amicon® Ultra 15 mL Centrifugal Filters (Millipore, Boston, MA) to enrich the exosomes. For the purification of circulating EVs from patients, we used a commercial exosome isolation kit, and exosome-enriched media were combined with 1/2 volume of Total Exosome Isolation Reagent (Thermo Fisher Scientific, Waltham, MA) and mixed well by vortexing or pipetting up and down until a homogenous solution was formed. The resulting solution was incubated at 4 °C overnight and centrifuged at 4 °C at 12,000×g for 1 h. The supernatant was discarded, and the purified EVs were resuspended in about 500 μL 1X PBS buffer and stored at − 80 °C until further analysis. These EVs were confirmed to be enriched in “exosome” type via flow cytometry, transmission electron microscopy (TEM) or nanoparticle tracking analysis (NTA), dynamic light scattering (DLS), and Western blots.

Tumor cells from human NSCLC cell lines, patient’s malignant pleural effusion, or PBMCs from patients with advanced NSCLC were collected, spun down, and resuspended in 10 mL of culture medium in a 10-cm Petri dish. For the whole-cell binding assay, 5 μL of beads coated with a known peptide sequence was washed sequentially with ethanol, water, and PBS. The beads were then incubated with suspended cells in the dish, and the entire dish was swirled at a speed of 40 rpm in an incubator at 37 °C and 5% CO₂. The plate was then examined under an inverted microscope every 15 min to check the cell binding. To determine the binding sensitivity of LXY30, A549 cells or malignant pleural effusion (PE) was subjected to a serial dilution (1:10⁵ or 1:10³, respectively) using 1 mL of supernatant of malignant pleural effusion from NSCLC patients, followed by incubation with ~ 250 TentaGel (90 μm, 0.26 mmol/g) (Rapp Polymere GmbH, Tϋbingen, Germany) beads coated with LXY30 or scrambled-LXY30 (S-LXY30) for 2 h before examination under microscope.

For the exosome-bead binding assay, 1.5 μg/μL A549, H1975, or patient tumor-derived exosomes in 200 μL were added into 1.5 mL tube followed by 100 TentaGel beads coated with LXY30 or S-LXY30 at 37 °C for 60 min, respectively. The exosome-beads were then washed three times in PBS. After the wash, Alexa Fluor® 647 mouse anti-human CD63 antibody (Biolegend, San Diego, CA) was added into the tube, incubating for 1 h and then washed three times in PBS. Next, A549 exosome-bead and H1975 exosome-bead binding were visualized under a LSM710 confocal fluorescence microscope (Zeiss, Germany).

Confluent (70–80%) human NSCLC cell lines and tumor cells isolated from patient pleural effusion were dissociated with 0.05% trypsin-EDTA and neutralized with culture medium. PBMCs were directly collected from the blood via Ficoll-Paque density gradient centrifugation. Each sample contained 3 × 10⁵ cells and was incubated with biotinylated peptides in 50 μL of PBS containing 10% FBS and 1 mM MnCl₂ for 30 min on ice. Each sample was washed three times with 1 mL of 1X PBS containing 1% FBS and incubated with a 1:500 dilution of streptavidin-PE (1 mg/mL) for 30 min on ice followed by a single wash with 1 mL of PBS containing 1% FBS. The final samples were analyzed by flow cytometry (Becton Dickinson Fortessa Flow Cytometer, San Jose, CA). Histogram analysis with mean fluorescence intensity (MFI) was analyzed using FlowJo 7.6.1 program (Ashland, OR).

Six-well plates were coated with 1500 μL of 20 μg/mL avidin (Thermo Fisher Scientific) and incubated for 1 h at 37 °C. Avidin-coated wells were washed three times with PBS and incubated with 1500 μL molar equivalents (2 μM) of D biotin (Thermo Fisher Scientific), LXY30-biotin, LXW64-biotin, or LXY30- and LXW64-biotin combo for 1 h. The wells were washed three times with PBS and blocked with 1% BSA (Thermo Fisher Scientific) for 1 h. After the wells were washed three times with PBS, 1 × 10⁵ H1975 cells were suspended in the maintenance medium, added to the wells, and incubated for 72 h at 37 °C, 5% CO₂. On the indicated time point (8 h, 24 h, 48 h, 72 h), trypsinized cells were counted by using hemocytometer for cell number analysis. After incubation for 72 h, one million cells were lysed in the Radio-Immunoprecipitation Assay buffer (Thermo Fisher, Waltham, MA). Thirty micrograms of lysates or exosome samples was separated by electrophoresis on 10% SDS-PAGE gels, transferred to nitrocellulose membranes, and probed with the following primary antibodies at 1:400 dilution: pEGFR Y1068, EGFR, pAKT S473, AKT (40D4), pMEK1/2 S217/221, MEK1/2 47E6, pSTAT3 Y705, and STAT3 124H6 (all from Cell Signaling Technology, Danvers, MA); CD63 (Santa sc-365604), integrin α3 (sc-374242), integrin β1 (sc-59829), integrin αV (sc-9969), and β-actin (sc-47778) (Santa Cruz Biotech). The secondary antibody was anti-mouse IgG, HRP-linked antibody (cell signaling, 1:500; #7076) or anti-rabbit IgG, HRP-linked antibody (cell signaling, 1:500; #7074). Densitometry was performed with Gel Doc^TM software (XR+ Imager, Bio-Rad, USA). The expression of each protein was normalized to β-actin in each sample.

Mice studies were performed according to an Institutional Animal Care and Use Committee (IACUC)-approved protocol (Protocol No. 20080) at the University of California, Davis. Female athymic nude mice (nu/nu), obtained from Harlan (Indianapolis, IN) at 5–6 weeks of age, were injected with 5 × 10⁶ of H3255, H1975, or A549 cells subcutaneously in the right flank. Patient-derived xenograft (PDX) model was generated from a patient with metastatic squamous cell lung cancer [28] implanted into the flank of NSG mice at age of 5–6 weeks old. For the intracranial implantation, 2.5 × 10⁵ cells in 5 μL PBS were injected into the right striatum area of the mouse with the aid of a mouse stereotactic instrument (Stoelting Co, Wood Dale, IL, USA). When the subcutaneous tumors reached 0.5–1.0 cm in diameter or 21–28 days after implantation, the mean size of intracranial xenograft tumors was 0.2–0.5 cm in diameter. Mice bearing NSCLC tumors were subjected to in vivo and ex vivo imaging studies.

Biotinylated peptide-SA-Cy5.5 (1.8 nmol), prepared by mixing 7.2 nmol of biotinylated peptide with 1.8 nmol of streptavidin-Cy5.5 in PBS overnight at 4 °C, was injected via the tail vein in an anesthetized mouse before imaging. Animals were placed in the supine, prone, or lateral position. Images were acquired by a Kodak IS2000MM image station (Rochester, NY) with a 625/20 band-pass excitation filter, 700WA/35 band-pass emission filter, and 150 W quartz. Halogen lamp light source was set to maximum. Images were captured at different time points with a CCD camera set at F stop = 0, FOV = 150, and FP = 0. Mean fluorescence intensity (MFI) was calculated by drawing the region of interest (ROI) of the mouse tumor using the Kodak ID 3.6 software. For ex vivo imaging, the mice were euthanized, and their organs were excised for imaging.

Cryosections of tumor xenografts in 10 μm thickness were fixed with 4% paraformadehyde for 20 min. After rinsing with deionized water, the slides were stained with hematoxylin for 5 s, rinsed in tap water, dipped in eosin for 30 s, and then dehydrated for 5 s with 95% ethanol, 5 s with 100% ethanol, and 15 s with xylene. The slides were covered with Permount solution and coverslips and examined with fluorescence microscope (IX81; Olympus) (image software: Metaphore).

For assessing the expression of targeted integrin in the xenograft tumors, 10-μm-thick slides of xenograft tumors were stained with mouse anti-α3 integrin antibody at 1:200 for 2 h. After washing with PBS, the slides were incubated with chicken anti-mouse Alexa 594 (Thermo Fisher Scientific, Waltham, MA) at 1:1000 for 30 min and washed with PBS. The slides were then covered by mounting solution with DAPI and evaluated under a fluorescence microscope (IX81; Olympus) (image software: Metaphore). For the cell line uptake experiment, A549 cells adhering on the bottom of chamber slides were incubated with 1 μM biotinylated LXY30 streptavidin-Alexa 594 conjugations for 2 h then observed under confocal fluorescence microscope (LSM710; Zeiss). For the H3255 intracranial xenograft microscopy, 10 μM cryosections of intracranial H3255 tumor after injection with LXY30-biotin/streptavidin-Cy5.5 complex were fixed in acetone at − 20 °C for 20 min. After washing with PBS, the sections were mounted and observed under the fluorescence microscope.

Descriptive statistics for continuous and categorical variables were stratified by binding to each integrin subtype or marker. All data are shown as mean ± standard deviation (SD) with at least 3 independent measurements. The two-sample t test was used for continuous variables. All analyses were conducted using SAS, university edition 2.5 9.4 M4 (SAS Institute, Cary, NC), and figures were made using GraphPrism software (Version 7.03). All statistical tests were two sided, and a p value less than 0.05 was considered statistically significant.

The datasets generated and/or analyzed during this study, as well as the computer code used to perform statistical analysis, are available from the corresponding authors on reasonable request. Supplementary information about IHC procedures, exome sequencing, and generation of exome-derived variables are available in supplementary methods.

We previously reported that LXY30 binds to the α3 integrin on the surface of a panel of live NSCLC cell lines with variable affinities and entered the cells via endocytosis [20]. We further characterized the binding specificity of LXY30 to a representative α3 integrin-expressing A549 cells. As illustrated in Fig. 1, A549 cells are specifically bound to LXY30-FITC (Fig. 1a) but not to S-LXY30-FITC (Fig. 1b) as assessed by flow cytometry (Fig. 1c) and fluorescence microscopy (Fig. 1d). LXY30-FITC conjugate detected α3 integrin expression in 7 out of 8 (87.5%) NSCLC tumor cells isolated from patient’s biofluids (malignant pleural or pericardial effusion or plasma) (Table 1). The sensitivity was tested by serial dilutions of A549 cells, and LXY30 could detect individual A549 cells in 10⁵ dilutions (data not shown). We further showed that LXY30 but not S-LXY30 is specifically bound to α3 integrin-expressing exosomes from the supernatant of NSCLC cell lines or patient’s biofluids (malignant pleural or pericardial effusion or plasma) in 9 out of 10 (90%) patients (Table 1). Representative fluorescence images from the supernatant of 2 NSCLC cell lines (A549 and H1975) and a patient’s malignant pleural effusion are shown in Fig. 1e. The expression of α3 and β1 integrin subunits was found in exosomes isolated from the supernatant of 4 lung adenocarcinoma cell lines A549, H1975, H3255, and H1650 and a patient pericardial effusion by Western blots using exosomal (CD63) and cellular-specific (β-actin) markers (Fig. 1f). The size and morphology of exosomes (EXO) as nanosize EVs were confirmed by dynamic light scattering (DLS) and nanoparticle tracking analysis (NTA) (Fig. 2a) and transmission electron microscopy (TEM) (Fig. 2b), respectively. We analyzed the yield of DNA isolated from the EV and EXO (Fig. 2c). While there was no significant difference between the yield of DNA isolated from S-LXY30-enriched and LXY30-enriched EVs (4.0 ± 0.7 vs 3.0 ± 1.2 ng/μL, p = 0.074), the yield of DNA was 3.4-fold higher in LXY30-enriched exosomes than in S-LXY30-enriched exosomes (3.4 ± 0.7 vs 1.0 ± 0.2 ng/μL, p = 0.014). The same driven mutations (KRAS G12S in A549, EGFR L858R and T790 M in H1975) were detected in the DNA isolated from LXY30-enriched exosomes and tumor cells by polymerase chain reaction (PCR) and Sanger sequencing (Fig. 2d). Using 10 patient samples with known plasma EGFR mutations by a clinical next-generation sequencing assay, we could detect the same EGFR mutations in LXY30-enriched exosomes using Sanger sequencing when the mutation allelic frequency (MAF) was high (> 50%) but not low (< 10%) although the threshold was not determined.

We further detected circulating tumor cells using LXY30 in 7 out of 8 (87.5%) patient’s pleural effusion. Figure 3 illustrates that representative tumor cells are isolated from malignant pleural effusion from a NSCLC patient using TentaGel beads coated with LXY30 but not S-LXY30 after 2- and 5-h incubation (Fig. 3a, left panel). LXY30 did not bind the inflammatory cells in the pleural effusion, such as macrophage and monocytes. Also, peripheral mononuclear blood cells (PBMCs) isolated from the same patients (N = 8) did not bind to TentaGel beads coated with either LXY30 or S-LXY30 (Fig. 3a, right panel). The specificity of binding was further confirmed by flow cytometry, in which FITC-labeled LXY30 specifically bound to tumor cells present in malignant pleural effusion but not the PBMCs in the blood of the same patient (Fig. 3b). We further showed that LXY30 sensitively detected individual, live pleural tumor cells in 1:1000 dilution using the supernatant from a patient’s malignant pleural effusion (Fig. 3c). On average, LXY30 could enriched malignant tumor cells by twofolds in these 8 patients tested (Fig. 3d). In a patient whose tumor cells were present in ~ 5% of cells in cell block, LXY30-bound beads were able to enrich the cellularity of tumor cells to > 20% in cell block (Fig. 3e), which enabled the successful detection of genomic alterations by next-generation sequencing assay.

A previous report suggests increased expression of multiple integrin subtypes in metastatic tumor cells compared to those of normal cells [29]. We further determined the integrin expression using the three most potent integrin ligands available (Additional file 1: Figure S1) on the surface of the EGFR-mutant NSCLC cells. Figure 4a illustrates that EGFR-mutant H3255 cells are bound to integrin ligand LXY30 (for α3β1 integrin) and LXW64 (for αvβ3 integrin) with high affinity but not to LLP2A (for α4β1 integrin) using the whole-cell binding assay. Consistent with previous reports [30, 31], we observed that erlotinib-resistant H3255 R#2 cells have increased the expression of all these three integrins compared to their parental erlotinib-sensitive, EGFR-mutant H3255 cells. The binding specificity of these cells to different peptide ligands was further confirmed by flow cytometry (Fig. 4b) and semi-quantified (Fig. 4c). As LLP2A could also bind to activated lymphocytes in PBMCs from cancer patients in clinical remission (data not shown), we further characterized the function of LXY30 and LXW64 in NSCLC in this study.

Integrins and EGFR share common downstream signaling pathways mediating the lung cancer initiation, growth, metastasis, and survival [32, 33]. LXY30 and LXW64 bind to the α3β1 and αvβ3 integrins, respectively, and activate EGFR and downstream signaling molecules (Fig. 5a). Compared to the untreated cells, LXY30 and LXY64 bind to erlotinib-resistant H1975 cells and activate the expression of α3, β1, and αv integrins; EGFR; and downstream signaling molecules either independently or in concert (Fig. 5b, c). The effect of LXY30 and/or LXW64 on the growth and proliferation of H1975 cells is shown by morphology (Additional file 2: Figure S2A) and cell counts (Additional file 2: Figure S2B) at 8 h, 48 h, and 72 h. Studies on other NSCLC cell lines revealed that the effect of LXY30 and LXW64 might be cell line dependent (Table 1), although the underlying mechanisms remain to be defined.

The in vivo tumor-targeting effect of LXY30 was evaluated in a subcutaneous and intracranial mouse models generated from EGFR-mutant lung adenocarcinoma H3255 cells by optical imaging (Fig. 6). We found that LXY30-biotin/streptavidin-Cy5.5 dye complex was accumulated in the subcutaneous and intracranial xenografts of H3255 cells (Fig 6a, b, respectively). In both subcutaneous and intracranial xenografts, there was about twofold higher uptakes of SA-Cy5.5 dye with LXY30 complexed compared to those without LXY30 in the imaging complex (Fig. 6c, d, respectively). The expression of α3 integrin was detected by fluorescence stain in tumor cells but not surrounding normal tissues in subcutaneous and intracranial xenografts (Fig. 6e, left panel). The histopathology of subcutaneous and intracranial xenografts was confirmed by H&E stain (Fig. 6e, right panel). The in vivo targeting effect of LXY30 was also shown in subcutaneous xenograft models of H1975 (EGFR L858R/T790 M mutations) and A549 (KRAS G12S mutation) (Fig. 7).

The above studies were performed in lung adenocarcinoma models. There are very few lung squamous cell models for research use. We analyzed the RNA levels of α3 integrin (INTA3) gene in 517 lung adenocarcinoma (LUAD) and 502 lung squamous cell carcinoma (LUSC) tumor samples using the transcriptome expression data from lung cancer patients in The Cancer Genome Atlas (TCGA) database (http://​www.​cancergenome.​nih.​gov). We found that the expression of α3 integrin levels was significantly higher in LUAD than in LUSC (134.73 vs 54.22, p < 0.001) (Fig. 8a). High ITGA3 expression was associated with poorer prognosis in LUSC (Fig. 8c) but not in LUAD (Fig. 8b).

We thus tested the in vivo targeting effect of LXY30 on a PDX model that was generated from a patient with metastatic lung squamous cell cancer [28]. Histopathological assessment showed that the tumors were freshly formed without any treatment effect [28]. The LXY30-biotin-streptavidin (SA)-Cy5.5 dye complex had significantly increased uptake compared to SA-Cy5.5 dye complex in tumor xenografts as shown in in vivo (Fig. 9a) and ex vivo imaging (Fig. 9b, c). The expression of α3 integrin was confirmed by fluorescence stain (Fig 9d), similar to these xenografts derived from human lung adenocarcinoma cell lines (Fig. 6 and Fig. 7).