Lectin‐based phototherapy targeting cell surface glycans for pancreatic cancer

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is resistant to current treatments but lectin‐based therapy targeting cell surface glycans could be a promising new horizon. Here, we report a novel lectin‐based phototherapy (Lec‐PT) that combines the PDAC targeting ability of rBC2LCN lectin to a photoabsorber, IRDye700DX (rBC2‐IR700), resulting in a novel and highly specific near‐infrared, light‐activated, anti‐PDAC therapy. Lec‐PT cytotoxicity was first verified in vitro with a human PDAC cell line, Capan‐1, indicating that rBC2‐IR700 is only cytotoxic upon cellular binding and exposure to near‐infrared light. The therapeutic efficacy of Lec‐PT was subsequently verified in vivo using cell lines and patient‐derived, subcutaneous xenografting into nude mice. Significant accumulation of rBC2‐IR700 occurs as early as 2 hours postintravenous administration while cytotoxicity is only achieved upon exposure to near‐infrared light. Repeated treatments further slowed tumor growth. Lec‐PT was also assessed for off‐target toxicity in the orthotopic xenograft model. Shielding of intraperitoneal organs from near‐infrared light minimized off‐target toxicity. Using readily available components, Lec‐PT specifically targeted pancreatic cancer with high reproducibility and on‐target, inducible toxicity. Rapid clinical development of this method is promising as a new modality for treatment of pancreatic cancer.

What's new?

Antibody‐conjugated therapies have shown insufficient efficacy against pancreatic cancer. In our study, the authors engineered and tested a lectin‐photoabsorber conjugate that targeted cell‐line and patient‐derived xenograft‐based pancreatic cancer. The conjugate reduced tumor size after near‐infrared light treatment and exhibited minor off‐target effects in a mouse model. The study offers foundational in vitro and in vivo evidence for a novel phototherapy that targets cell surface glycans using a lectin‐based carrier system. With all components readily available, a therapeutic protocol that could greatly enhance treatment of refractory pancreatic cancer could be rapidly developed in the clinic.

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

CRE

creatinine

EGFR

epidermal growth factor receptor

GFP

green fluorescent protein

HNSCC

head and neck squamous cell cancer

Lec‐PT

lectin‐based phototherapy

NIR

near‐infrared

PDAC

pancreatic ductal adenocarcinoma

PDX

patient‐derived xenograft

PI

propidium iodide

PIT

photoimmunotherapy

ROI

region of interest

1. INTRODUCTION

Pancreatic cancers, especially pancreatic ductal carcinoma (PDAC), carry a dismal survival rate (5‐year survival rate around 10%). A key reason is the locally advanced or nonresectable stages at which most patients are diagnosed, ¹ even an apparent R0 resection following radical treatment often retains residual disease that eventually leads to relapse. With the rising trend in PDAC‐related mortality ¹ and the lackluster breakthrough in early detection and therapies for PDAC, ² , ³ a radically different approach is required.

A promising and underexplored approach is to target the glycan‐enriched outer layer of cancer cells known as the glycocalyx. ⁴ These carbohydrate‐based surface molecules undergo diverse aberrant glycosylation processes in cancer cells, including sialylation, fucosylation, truncation of O‐glycans, and branching of N‐ and O‐linked glycans. ⁵ , ⁶ , ⁷ Thus, the outer surface could be attractive for therapeutic targeting.

We have previously demonstrated that fucosylated glycans (H type‐3 motifs) are expressed in pancreatic cancers ⁸ and a recombinant lectin (AiLecS1), called rBC2LCN, ⁹ binds to these glycans specifically. ¹⁰ Furthermore, we also found no evidence of hemagglutination which results from introduction of exogenous lectins to the blood. ¹¹

Exploiting these unique properties, we previously employed rBC2LCN as a PDAC‐targeting agent to specifically deliver the catalytic part of Pseudomonas exotoxin A (containing domains Ib, II, and III [PE38]), demonstrating that this rBC2LCN‐PE38 fusion protein reduced tumor volumes in mouse models of pancreatic cancer. ¹⁰ Furthermore, in spite of lower binding affinities of glycan‐lectin interactions, often in the range of 1‐10⁻⁶ M vs 0.1‐10⁻⁹ M for antibody‐antigen interactions, ¹² the presence of multiple surface glycans (membrane proteins and glycolipids), plus glycan branching, offers a comparatively target‐rich binding environment for “swarm” attacks as opposed to single‐epitope antibody‐antigen binding. However, as with therapeutic antibodies targeting tumor‐associated antigens, minimization of off‐target toxicity is paramount, since we have reported cross reactivity of rBC2LCN to noncancerous cells, including renal tubular epithelial cells. ¹⁰

Assuming a two‐step process for the lectin‐payload conjugate, the use of lectins as a carrier of cytocidal drugs to cancer cells is more realistic. Once bound to the tumor surface, activation of the payload is required in a manner that does not harm healthy tissue. For this purpose, modification of near‐infrared photoimmunotherapy (NIR‐PIT), a newly developed, molecularly targeted cancer phototherapy based on conjugating photoabsorber dyes to antibodies, could offer such capabilities. ¹³ IRdye700DX (IR700) is a nontoxic and nonionizing water‐soluble phthalocyanine dye which exerts strong cytotoxicity upon activation with NIR (Excitation peak at 689 nm) light only when associated with cellular membranes. ¹⁴ Previously reported antibody‐IR700 conjugates, injected intravenously, bind to targets expressed on the surface of cancer cells and, upon exposure to NIR light, induces physical stress within the cellular membrane, leading to necrotic cell death. ¹⁵ , ¹⁶ , ¹⁷ We therefore hypothesized that IR700 would be useful as a payload for rBC2LCN lectin‐based carrier system, termed lectin‐based phototherapy (Lec‐PT). Here, we validate both the cellular and systemic on‐ and off‐target effects of our novel Lec‐PT approach in a PDAC xenograft mouse model with the expectation of evolution toward clinical use for currently refractory pancreatic cancer.

2. MATERIALS AND METHODS

2.1. Reagents

Water soluble, silica‐phthalocyanine derivative, IRDye 700DX NHS ester (IR700) was purchased from LI‐COR Biosciences (Cat# 929‐70 011) while rBC2LCN lectin was purchased from FUJIFILM Wako Pure Chemical Corporation (Cat# 025‐18 063). All other reagents were of reagent grade.

2.2. Synthesis of IR700‐conjugated rBC2LCN

The rBC2LCN lectin (160 μg, 10 nmol) was incubated for 2 hours at room temperature with IR700 (19.5 μg, 10 nmol) in 0.1 mol/L Na₂HPO₄ (pH 8.6) (Figure 1A). Sephadex G‐25 Fine (Cytiva, Cat# 17003202) was used to purify the mixture. Protein concentration and conjugation ratio were calculated from absorbance at 280 and 689 nm with UV‐Vis spectroscopy (Nanodrop 1000; Thermo Fisher Scientific), according to the following formula described in the manufacturer's instructions detailed below:

$$\text{Concentration of rBC2 protein}\left(\mathrm{M}\right)=\left(A_{280}-\left[0.095\times A_{689}\right]\right)/\left({Ɛ}_{\text{protein}}\times 0.1\right).$$

$$\text{Concentration of}\mathrm{IR}700\left(\mathrm{M}\right)=A_{689}/\left({Ɛ}_{\mathrm{dye}}\times 0.1\right).$$

$${Ɛ}_{\text{protein}}=13856/\mathrm{M}\mathrm{cm},{Ɛ}_{\mathrm{dye}}=165000/\mathrm{M}\mathrm{cm}.$$

The correction factor at A ₂₈₀, 0.095, was provided by the supplier (LI‐COR Biosciences).

FIGURE 1.

Validation and specificity of rBC2‐IR700. (A) The rBC2LCN lectin was conjugated with IR700. (B) Representative absorption spectra of rBC2‐IR700. The concentrations of rBC2LCN lectin and IR700 were calculated from absorbance at 280 and 689 nm, respectively. (C) SDS‐PAGE validation of rBC2‐IR700 labeling with rBC2LCN lectin as a control. (D) In vitro rBC2‐IR700 binding specificity was evaluated by flow cytometry. (E) Transmitted and fluorescence microscopy images of Capan‐1 and SUIT‐2 cells after incubation with rBC2‐IR700 for 3 hours. High fluorescence intensities are visible only in Capan‐1 cells (H type‐3 positive). Scale bars = 20 μm. (F) Representative images of histochemical lectin staining. The staining patterns of three clinical PDAC samples featuring well‐ (well), moderately‐ (mod) and poorly‐ (por) differentiated specimens are shown. Hematoxylin and eosin (H&E) staining was used for confirmation of PDAC sample classification. Positive staining with rBC2‐IR700 was of a similar pattern to rBC2LCN‐HRP. Scale bars = 100 μm

IR700 conjugated to rBC2LCN lectin is hereafter described as rBC2‐IR700. As a quality control for the conjugate, SDS‐PAGE was performed with a 15% to 25% gradient polyacrylamide gel (DRC, Cat# NTH‐7G6HP). After separating the proteins, the gel was imaged with a FUSION FX7.EDGE (Vilber) using a 680 nm laser for excitation and 740 nm filter for emission. Nonlabeled rBC2LCN lectin was used as a control. The gel was then stained with Instant Blue Coomassie Protein Stain (Abcam, Cat# ab119211) for equal loading of rBC2LCN and rBC2‐IR700 and digitally scanned.

2.3. Cell lines and culture conditions

Capan‐1, a H type‐3‐positive human pancreatic cancer cell line, ⁸ was obtained from American Type Culture Collection (ATCC; Cat# HTB‐79, CVCL_0237). SUIT‐2, a H type‐3‐negative human pancreatic cancer cell line, was obtained from Japanese Collection of Research Bioresources (JCRB) Cell Bank (Cat# JCRB1094, CVCL_3172). All human cell lines have been authenticated using STR profiling within the last 3 years. All experiments were performed with mycoplasma‐free cells. Capan‐1 cells were cultured in Iscove's modified Dulbecco's medium (FUJIFILM Wako Pure Chemical Corporation, Cat# 098‐06465) supplemented with 20% fetal bovine serum (FBS; Thermo Fisher Scientific, Cat# 10270‐106) and containing 1% Penicillin‐Streptomycin (FUJIFILM Wako Pure Chemical Corporation, Cat# 168‐23191). SUIT‐2 cells were cultured in Dulbecco's modified Eagle's medium (FUJIFILM Wako Pure Chemical Corporation, Cat# 044‐29765) supplemented with 10% FBS and containing 1% penicillin‐streptomycin. All cells were maintained at 37°C in a humidified incubator in 5% CO₂.

2.3.1. Capan‐1‐GFP‐luc cells

Capan‐1 cell line ectopically expressing GFP (green fluorescent protein)‐luciferase fusion (Capan‐1‐GFP‐luc) was generated through transduction of Capan‐1 with GFP‐luciferase‐encoding lentivirus (pGF1‐CMV). Capan‐1 cells were infected with the virus (a gift from Yasuyuki Kida) in the presence of polybrene (10 μg/mL) for 24 hours. At 24 hours post infection, cells were observed using a fluorescent microscope to visualize the GFP expressed by pGF1‐CMV. At 72 hours after infection, cells were harvested and FACS sorted for GFP‐positive cells. To verify the sorted cells are expressing the GFP‐luc, fluorescent images were acquired using a 450 to 490 nm excitation filter and a 500 to 550 nm band‐pass emission filter corresponding to GFP. Transmitted light images were also acquired (Figure S1).

2.4. Flow cytometry

To confirm retention of rBC2‐IR700 binding specificity, the IR 700 fluorescence signal in H type‐3‐positive Capan‐1 and H‐type‐3‐negative SUIT‐2 cells were measured on a flow cytometer after incubation with rBC2‐IR700 (SH800S; Sony Corporation). Capan‐1 cells and SUIT‐2 cells (2 × 10⁵) were seeded into 35 mm dishes and incubated for 24 hours. Medium was replaced with fresh culture medium containing 1 μg/mL of rBC2‐IR700 and incubated for 3 hours at 37°C.

2.5. Live‐cell imaging

To detect the glycan‐specific localization of rBC2‐IR700, live‐cell imaging was performed via confocal laser microscopy (TCS SP8; Leica Microsystems). Capan‐1 cells and SUIT‐2 cells (1 × 10⁵) were seeded on cover‐glass‐bottomed dishes and incubated for 24 hours. After that, rBC2‐IR700 was added to the culture medium at 1 μg/mL and incubated at 37°C. The cells were washed with PBS after incubation. To visualize the cellular localization of rBC2‐IR700, cell membranes were imaged with CellMask Green Plasma Membrane Stain (Thermo Fisher Scientific, Cat# C37608) which was added to the culture medium and incubated for 10 minutes at 37°C. The cells were then washed with PBS and fluorescent images were acquired using a 638 nm excitation laser and a detector set at 676 to 724 nm to detect IR700. Transmitted light images were also acquired.

2.6. Lectin histochemistry

To verify the binding of rBC2‐IR700 to human clinical pancreatic cancer tissues, fluorescence histochemistry was performed. Fresh, frozen tissues were sliced into 3 μm serial sections. Following blocking with normal serum, the tissues were incubated with 1 μg/mL of rBC2‐IR700 for 1 hour at room temperature. Fluorescent images were obtained using a BZX710 microscope (Keyence Corporation). Horseradish peroxidase‐labeled rBC2LCN was also applied to the serial sections and visualized by applying the chromogen diaminobenzidine. Endogenous peroxidase activity was blocked with 3% H₂O₂ and methanol. Adjacent sections were used for hematoxylin and eosin (H&E) staining.

2.7. In vitro Lec‐PT

2.7.1. LIVE/DEAD staining

To assess the kinetics of the cytotoxic effect of Lec‐PIT following NIR irradiation, a Live/Dead Cell Staining Kit II (PromoCell GmbH, Cat# PK‐CA707‐30002) was used. Capan‐1 cells (1 × 10⁵) were seeded on cover‐glass‐bottomed dishes and incubated for 24 hours. rBC2‐IR700 was then added to the culture medium at 1 μg/mL and incubated for 3 hours at 37°C The cells were then washed with PBS and LIVE/DEAD reagents were added to the dish. Serial microscopic images were obtained by confocal microscopy (TCS SP8; Leica Microsystems) with a 638 nm excitation laser.

2.7.2. Cytotoxicity assay

The cytotoxic effects of Lec‐PT with rBC2‐IR700 were determined by propidium iodide (PI; Dojindo Laboratories, Cat# 343‐07461) incorporation, which happens when plasma membrane loses its integrity, and analyzed using flow cytometry. Capan‐1 cells and SUIT‐2 cells (2 × 10⁵) were seeded into 35 mm dishes and incubated for 24 hours. Medium was replaced with fresh culture medium containing 1 μg/mL of rBC2‐IR700 and incubated at 37°C. After washing with PBS, the cells were exposed to NIR light at a power density of 50 mW/cm². At 1 hour after NIR irradiation, the cells were collected and resuspended. PI was then added in the cell suspension (final concentration 2 μg/mL) and incubated at room temperature for 30 minutes before flow cytometry.

Luciferase activity was used to quantitatively measure the in vitro cytotoxic effects of Lec‐PT with Capan‐1‐GFP‐luc cells. A total of 1 × 10⁴ cells per well were seeded into 96‐well plates and incubated for 24 hours. rBC2‐IR700 was then added to the culture medium at 1 μg/mL and incubated for 3 hours at 37°C. After washing with PBS, the cells were exposed to NIR light at a power density of 20 mW/cm². Then, 150 μg/mL of D‐luciferin‐containing media (IVISbrite D‐Luciferin, RadiJect; PerkinElmer Inc., Cat# 770504) was added to cells 1 hour after Lec‐PT. After that, the cells were examined using a bioluminescence plate reader (Varioskan LUX; Thermo Fisher Scientific).

The cells were irradiated with an NIR light‐emitting diode which emits light at a wavelength of 670 to 710 nm (L690‐66‐60; Ushio‐Epitex). Power density was measured with an optical power meter (PM100; Thorlabs).

2.8. Experimental animals

Female BALB/c nude mice and CB17/Icr‐scid/SCID mice (aged 6 weeks, weighting approximately 20 g) were purchased from CLEA Japan (Tokyo, Japan). Mice were housed and maintained in specific‐pathogen‐free conditions according to animal use guidelines of the University of Tsukuba (Ibaraki, Japan). During the procedure, mice were anesthetized with inhaled isoflurane. Tumor volumes were determined using the following formula: (width)² × (length)/2.

2.8.1. Subcutaneous cell xenograft model

Capan‐1 cells (2 × 10⁶) and SUIT‐2 cells (2 × 10⁶) were injected subcutaneously in the right dorsum of nude mice. Tumors were studied after they reached volumes of approximately 50 mm³.

2.8.2. Subcutaneous patient‐derived xenograft model

Pancreatic cancer patient tumor tissues were obtained at surgery, cut into 2 mm cubic fragments, and transplanted subcutaneously into SCID mice. These subcutaneous tumors were then passaged subcutaneously into the right dorsum of nude mice. Tumors were studied after they reached volumes of approximately 80 mm³.

2.9. In vivo and ex vivo fluorescence imaging

Subcutaneous cell xenograft mice and patient‐derived xenograft (PDX) mice were used. After intravenous injection of 20 μg of rBC2‐IR700 via the tail vein, serial ventral and dorsal fluorescent images of IR700 were obtained with an IVIS Spectrum (PerkinElmer Inc.) using a 675 nm excitation filter and a 720 nm emission filter at 0, 1, 2, 4, 6, 12, 24, 48, 72, 96, 120, 144, and 168‐hour timepoints (n = 5). To determine the optimal dosage of rBC2‐IR700 that gave optimal discrimination between the intended tumor region against background, 10 μg, 20 μg, and 40 μg of rBC2‐IR700 were administered and fluorescent images of IR700 were obtained (n = 3 each). Region(s) of Interest (ROI) were placed on the tumor, kidneys and liver. ROIs were also placed in the adjacent nontumor region as background (left dorsum). The average fluorescence intensity of each ROI was measured. Fluorescence intensities of tumor‐to‐background ratios were also calculated. For ex vivo imaging, each organ (heart, lung, liver, spleen, kidneys, pancreas, bladder, peritoneum, and organs of gastrointestinal tract) and tumor were resected from the mouse at 6 hours after administration of rBC2‐IR700 (20 μg) and imaged using the IVIS Spectrum. To evaluate autofluorescence of each organ, ex vivo images were also acquired at 6 hours after administration of 20 μg of rBC2LCN lectin.

2.10. In vivo Lec‐PT

2.10.1. Single treatment for subcutaneous model

Cell xenograft mice were randomized into four groups (n = 6 each): (i) no treatment (control); (ii) 20 μg of rBC2‐IR700 intravenous (i.v.), no NIR light exposure (rBC2‐IR700); (iii) NIR light exposure only, NIR light exposure administered at 100 J/cm² (NIR light); and (iv) 20 μg of rBC2‐IR700 i.v., NIR light exposure administered at 100 J/ cm² after injection (Lec‐PT). PDX mice were divided into 2 groups (n = 6 each): (i) control and (ii) Lec‐PT, as described above. The tumor volume was measured three times a week and the subcutaneous tumors were excised for weight measurements on day 15.

2.10.2. Repeated treatment for subcutaneous model

Cell xenograft mice were randomized into four groups (n = 6 each): (i) control; (ii) rBC2‐IR700; (iii) NIR light; and (iv) Lec‐PT, as described above. These therapies were performed every week for 3 weeks. The tumor volume was measured three times a week, and the subcutaneous tumors were excised for weight measurements on day 28.

2.10.3. Assessment for off‐target injuries from Lec‐PT

Nude mice were divided into four groups (n = 4 each): (i) no treatment (control); (ii) NIR light irradiated with no shield (no shield); (iii) NIR light irradiated with both kidneys shielded (kidney shield); and (iv) NIR light irradiated with liver shielded (liver shield). NIR light (100 J/cm²) was used to irradiate the entire abdomen under laparotomy at 6 hours following rBC2‐IR700 (20 μg) i.v. injection. Aluminum foil was used for shielding each organ. Whole blood was collected 24 hours after single Lec‐PT and hematological parameters were measured. The collected organs were histologically evaluated using 3‐μm slide sections of formalin‐fixed, paraffin‐embedded tissues.

2.10.4. Orthotopic model of pancreatic cancer

Nude mice were anesthetized with isoflurane, a laparotomy was created using a left lateral abdominal incision, and the pancreas was exteriorized. Two tumor pieces (approximately 2 mm in diameter) obtained from subcutaneous Capan‐1‐GFP‐luc tumors were transplanted into the body of the pancreas and secured with surgical sutures.

2.10.5. Treatment for the orthotopic model of pancreatic cancer

Pancreatic orthotopic mice were divided into two groups (n = 7 each): (i) control and (ii) Lec‐PT, as described above. NIR light was directed at the tumor intraperitoneally under laparotomy with the rest of the organs shielded by aluminum foil. Therapeutic effects were evaluated by quantifying the luciferase activity of orthotopic tumors three times a week. D‐luciferin (30 mg/mL, 100 μL) was injected intraperitoneally and the mice were analyzed via bioluminescence imaging (IVIS spectrum; PerkinElmer Inc.) for luciferase activity (photons/sec/cm²/sr). ROIs were set on the entire tumors and average luciferase activity of each ROI was measured. The activity on the day of Lec‐PT was set to 100%. To evaluate the side effects of Lec‐PT, body weight changes for all mice were measured during the experimental period. In addition, whole blood and major organs (heart, lung, liver, spleen, pancreas, kidney, stomach, duodenum, small intestine, and colon) were collected on day 15. Hematological parameters were measured, including complete blood count (white blood cells [WBC], red blood cells [RBC], hemoglobin [HGB], hematocrit [HCT], and platelets [PLT]), creatinine (CRE), aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin (BIL), blood urea nitrogen (BUN) and amylase (AMY). The organs collected were histologically evaluated using 3‐μm slide sections of formalin‐fixed, paraffin‐embedded tissues.

2.11. Statistical analysis

Data are expressed as means ± SEM. One‐way ANOVA, followed by Tukey post hoc comparison for multiple‐group comparisons, and Student's T‐test for two‐group comparisons were performed to determine the statistical significance of the differences by using GraphPad Prism 9. The threshold for statistical significance was set at P < .05.

3. RESULTS

3.1. Conjugation of rBC2LCN with IR700 and affinity of rBC2‐IR700 to target glycans

The IR700 to rBC2LCN conjugation ratio was determined from absorbance at 280 and 689 nm and approximated to be 1:0.5 (rBC2:IR700). Representative data are shown in Figure 1B (Supplementary Methods). The rBC2‐IR700 product was confirmed by SDS‐PAGE, with a sharp band around 16 kDa exhibiting clear fluorescence intensity (Figure 1C). After a 3‐hour incubation, rBC2‐IR700 bound to H type‐3‐positive Capan‐1 cells was visible via flow cytometry (Figure 1D) and fluorescence microscopy (Figure 1E). On the other hand, rBC2‐IR700 showed no binding to H type‐3‐negative SUIT‐2 cells. The staining patterns of three clinical PDAC samples featuring well‐ (well), moderately (mod) and poorly (por) differentiated specimens are shown in Figure 1F. Positive staining with rBC2‐IR700 was of a similar pattern to rBC2LCN‐HRP. These results were similar to our previous report that tumor tissue regions, but not surrounding stromal components, stained strongly with rBC2LCN after analysis of 69 human clinical PDAC specimens, which included all differentiation statuses. ¹⁰ These results indicate that rBC2‐IR700 retains reactivity to PDAC cells.

3.2. In vitro cytotoxic effect of Lec‐PT

First, we analyzed the effect of rBC2‐IR700 combined with NIR light exposure, termed Lec‐PT, on Capan‐1 cells in vitro. After Lec‐PT (following a 3‐hour incubation with rBC2‐IR700), cellular swelling and bleb formation were observed in Capan‐1 cells. Furthermore, green fluorescence (Calcein‐AM), representative of live cells, gradually disappeared and, at the same time, red fluorescence (ethidium homodimer), indicative of dead cells, was observed to internalize into the cells. These morphologic changes were observed within 30 minutes (Figure 2A, Supplementary Video), indicating rapid induction of cell death, following cell membrane damage by Lec‐PT.

FIGURE 2.

Evaluation of Lec‐PT in vitro cytotoxic effects. (A) Microscopic images (transmitted and fluorescence) of LIVE/DEAD staining in Capan‐1 cells after 3 hours of rBC2‐IR700 incubation. Morphologic changes, as well as green (Calcein‐AM) and red (ethidium homodimer) fluorescence, were evaluated upon excitation with NIR light. Scale bars = 20 μm. (B) Cytotoxicity in Capan‐1 cells (H type‐3 positive) induced by Lec‐PT was measured by PI staining and flow cytometry (n = 3; **P < .01 vs untreated control; ***P < .001 vs untreated control, by Student's T‐test). (C) Cytotoxicity in SUIT‐2 cells (H type‐3 negative) induced by NIR light exposure was measured by PI staining and flow cytometry (n = 3). (D) Cytotoxicity of IR700 dye alone and rBC2LCN lectin alone in Capan‐1 cells (n = 3). (E) Time course fluorescence microscopic images of Capan‐1 cells after incubation with rBC2‐IR700. Scale bars = 5 μm. (F) Cytotoxicity in Capan‐1 cells induced by Lec‐PT compared to incubation time (n = 3)

The cytotoxic activity of Lec‐PT on Capan‐1 cells was then analyzed by PI staining and flow cytometry. The percentage of cell death increased in a light‐dose dependent manner (Figure 2B) while no effect was observed in H type‐3‐negative SUIT‐2 cells (Figure 2C). NIR irradiation caused no significant cytotoxicity in the absence of rBC2‐IR700 while rBC2‐IR700 alone without NIR light also was not cytotoxic. The IR700 dye alone (62.5 nM) and rBC2LCN alone (62.5 nM), both of which were equivalent in concentration to that used for the rBC2‐IR700 conjugates described above, showed no cytotoxicity with Capan‐1 cells (Figure 2D). We also acquired time course microscopic images to visualize the cellular localization of the rBC2‐IR700. IR700 fluorescence was detected primarily on cell surfaces after 1 hour of incubation and intracellularly at 3 and 6 hours (Figures 2E and S2), indicating a gradual internalization of rBC2‐IR700. There were no significant differences in NIR‐induced cytotoxicity over the 1‐to‐6‐hour incubation period with rBC2‐IR700 (Figure 2F), indicating that Lec‐PT effectiveness is based on rBC2‐IR700 cellular binding.

3.3. Target tumor accumulation and biodistribution of rBC2‐IR700 in vivo

We then analyzed target tumor accumulation of rBC2‐IR700 on a Capan‐1 cell xenograft mouse model in vivo. High fluorescence intensities were detected in Capan‐1 tumor cells within 2 hours after rBC2‐IR700 administration and remained high for almost an entire day (Figure 3A,B) while no obvious accumulation of rBC2‐IR700 was observed in H type‐3‐negative SUIT‐2 tumor cells of similar volume (Figure S3A). This confirmed specific delivery of rBC2‐IR700 to Capan‐1 tumors even in vivo. Similarly, high fluorescence intensity was also detected in kidneys and liver immediately after administration of rBC2‐IR700 and decreased rapidly within a few hours (Figure 3A,B). Fluorescence intensity in Capan‐1 tumors increased in a rBC2‐IR700 dose‐dependent manner while tumor‐to‐background ratios plateaued at 20 μg of rBC2‐IR700 (Figure S3B,C). Since the best possible therapeutic index relies on high rBC2‐IR700 in tumors but as low as possible fluorescence in normal tissues, we thus decided to irradiate with NIR light at 6 hours after rBC2‐IR700 administration (Lec‐PT).

FIGURE 3.

In vivo and ex vivo fluorescence imaging in a Capan‐1 cell xenograft mouse model. (A) In vivo rBC2‐IR700 fluorescence serial imaging of Capan‐1 cell xenograft mice. The white arrows indicate high fluorescence intensities in Capan‐1 tumors from the right dorsum. Yellow arrows indicate kidneys. ROIs were placed on the tumor (right flank; black circle), background (left flank; blue circle), kidneys (both left and right: yellow circle). and liver (upper abdomen; red circle). (B) Fluorescence time course in tumors, backgrounds, livers, and kidneys after rBC2‐IR700 administration (n = 5). An arrow indicates the 6‐hour time point. (C) Ex vivo fluorescence imaging of Capan‐1 cell xenograft mouse at 6 hours after rBC2‐IR700 administration. Fluorescence intensities in the heart, lungs, liver, spleen, kidneys, pancreas, bladder, peritoneum, gastrointestinal tract and tumor are represented as a heatmap scale

Ex vivo examinations were also performed to assess the biodistribution of rBC2‐IR700 at 6 hours after administration. A high fluorescence intensity was observed in tumors, kidneys and livers, consistent with the result of our in vivo imaging study (Figures 3C and S3D).

3.4. Therapeutic effect of Lec‐PT in PDAC mouse models

3.4.1. Lec‐PT for subcutaneous model

We examined the effect of Lec‐PT by single treatment in a subcutaneous Capan‐1 xenograft mice model. Based on the result of in vivo fluorescence imaging, the mice of the Lec‐PT group were exposed to NIR light (100 J/cm²) at 6 hours following rBC2‐IR700 (20 μg) injection. In these Capan‐1 cell xenograft mice, tumor growth was significantly inhibited in the Lec‐PT group (P < .05 vs other control groups; Figure 4A,B). The effect for PDX mice was then also evaluated. We first confirmed the accumulation of rBC2‐IR700 in PDX tumors similar to Capan‐1 tumors (Figure S4) and found that tumor growth was significantly inhibited in the Lec‐PT treatment group (P < .001 vs control group; Figure 4C,D).

FIGURE 4.

In vivo therapeutic effects of Lec‐PT in a subcutaneous mouse model. Arrows in (A, C and E) indicate one treatment session. (A and B) Changes in Capan‐1 tumor volume during the single treatment experiment period and tumor weights excised from cell xenograft mice after a single treatment (n = 6; *P < .05 vs other control groups, by Tukey test with ANOVA). (C and D) Changes in PDX tumor volume during a single Lec‐PT experiment period and tumor weights excised from PDX mice after single Lec‐PT (n = 6; **P < .01 vs control group; ***P < .001 vs control group, by Student's T‐test). (E and F) Changes in tumor volumes during the repeated treatment experiment period and tumor weights excised from cell xenograft mice after repeated treatment (n = 6; ***P < .001 vs other control groups, by Tukey test with ANOVA). (G) Resected tumors from cell xenograft mice at day 28 after repeated treatment

Next, we examined repeated treatment effects in the subcutaneous Capan‐1 xenograft mice model. Even though a single dose of Lec‐PT treatment significantly inhibited tumor growth, regrowth was eventually seen around 7 days after Lec‐PT. Therefore, we checked repeated Lec‐PT for tumor reduction persistence with our Capan‐1 xenograft mice model. In this experimental regimen, the Lec‐PT group mice were treated as described above and repeated weekly up to 3 weeks total. Tumor growth was significantly inhibited in the Lec‐PT group and this effect was maintained over an extended period (P < .001 vs other control groups; Figure 4E‐G).

3.4.2. Assessment for off‐target injury by Lec‐PT

The biodistribution studies revealed significant accumulation of rBC2‐IR700 in the liver and the kidneys, risking unintended damage to these organs if exposed to NIR light. Conversely, no off‐target injury is expected if these organs are not exposed to NIR irradiation. As a proof of concept, we assessed whether simply shielding these organs from whole‐abdomen NIR irradiation during laparotomy could mitigate unintended damage. Blood CRE and AST/ ALT levels were measured as readouts of kidney and liver injury. Although CRE values showed significant increases in the no shield group compared to controls (P < .05; 0.93 mg/dL [±0.26 mg/dL] vs 0.063 mg/dL [±0.013 mg/dL]), differences were not significant between the kidney shield (0.23 mg/dL [±0.073 mg/dL]) and control groups (Figure 5A). AST and ALT values were significantly higher in the no shield group vs control group (P < .01; 17 000 U/L [±3000 U/L] vs 48 U/L [±1.3 U/L], P < .01; 22 000 U/L [±5400 U/L] vs 23 U/L [±2.1 U/L], respectively) while the differences in AST and ALT values were not significant between the liver shield (610 U/L [±230 U/L], 610 U/L [±250 U/L], respectively) and control groups (Figure 5B). In the no shield group, acute renal tubular necrosis was observed in the kidney while acute necrosis could be seen in the irradiated area of the liver based on histological analyses (Figure S5). Although a low fluorescence signal (Figure 3C) was observed in ex vivo images of normal pancreas, there was no observable damage based on histological analyses (Figure S6), likely reflective of very low background accumulation in a normal pancreas (Figure S3D). These results support the concept of the enhanced specificity of rBC2‐IR700 and demonstrated that, irrespective of the degree of accumulation at unintended sites, cytotoxicity is only achieved in areas directly exposed to NIR irradiation. Therefore, we performed subsequent orthotopic pancreatic tumor experiments with off‐target, intraperitoneal organs shielded in lieu of a precision laser system.

FIGURE 5.

Assessment for off‐target injury and in vivo therapeutic effects of Lec‐PT for pancreatic orthotopic Capan‐1‐GFP‐luc tumors. Arrows in (D and F) indicate one treatment session. (A) Hematological assessment of creatinine (CRE) (Shield n = 4, Control n = 4; *P < .05 vs control group, by Student's T‐test). (B) Hematological assessment of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) (Shield n = 4, Control n = 4; **P < .01 vs control group, by Student's T‐test). (C) In vivo bioluminescence images (BLI) of orthotopic pancreatic tumors during the experiment period. (D) Change of BLI activity of orthotopic pancreatic tumors over the experimental period (n = 7; **P < .01 vs control group, by Student's T‐test). (E) Hematological CRE, AST and ALT findings after Lec‐PT for the orthotopic model of pancreatic cancer (n = 7). (F) Body weight changes in mice during the experimental period. (n = 7; ***P < .001 vs control group, by Student's T‐test)

3.4.3. Lec‐PT in an orthotopic model of pancreatic cancer with shielding of off‐target organs

To measure in vivo tumor volume over time, we performed orthotopic engraftment of Capan‐1 transduced with a GFP‐T2A‐luciferase reporter that we confirmed to retain 100% reactivity toward rBC2‐IR700 (Figure S6A). Correlation between Capan‐1‐GFP‐luc cell number and bioluminescent intensity was also confirmed (Figure S6B). Furthermore, responsiveness to Lec‐PT (using luciferase activity of remaining viable cells as a readout) in a NIR light dose‐dependent manner similar to parental Capan‐1 (Figure 2B) was verified (Figure S6C). Based on these validation experiments, we established that Capan‐1‐GFP‐luc bioluminescence is a suitable proxy for tumor volume.

Lec‐PT‐treated mice with orthotopic pancreatic tumors demonstrated significantly lower bioluminescence (P < .01 vs control group; Figure 5C,D). With intraperitoneal organ shielding implemented, no hematological parameters associated with liver and kidney damage (CRE, AST, and ALT) differed significantly between the control and Lec‐PT groups (0.14 mg/dL [±0.016 mg/dL] vs 0.11 mg/dL [±0.0091 mg/dL], 56 U/L [±2.7 U/L] vs 52 U/L [±2.1 U/L], 25 U/L [±1.8 U/L] vs 23 U/L [±3.7 U/L], respectively) (Figure 5E). Similarly, histological examination of intraperitoneal organs and other hematological parameters did not reveal any changes indicative of therapy‐induced damage or dysfunction (Figure S7A,B). Although body weights significantly decreased temporarily after Lec‐PT (P < .001 vs control group), likely a consequence of the open laparotomy procedure, the mice recovered weight over the experimental period (Figure 5F).

4. DISCUSSION

In our study, we engineered and tested a lectin‐photoabsorber conjugate that targeted cell‐line and PDX‐based pancreatic cancer, reducing tumor size after NIR light treatment and exhibiting only minor off‐target effects in a mouse model.

The concept of lectin as a cancer targeting agent was developed from necessity as antibody‐based systems have not shown definitive efficacy against pancreatic cancer. ² , ³ For antibody‐antigen interactions, reports of antibody‐based PIT for pancreatic cancer have already been reported, but without sufficient success in pancreatic malignancies to spur definitive clinical development. ¹⁸ , ¹⁹ NIR‐PIT has since been developed to target epidermal growth factor receptor (EGFR) and human epidermal growth factor receptor‐2 (HER2) ¹³ with additional target development for a variety of antigen‐specific transmembrane proteins. ²⁰ , ²¹ , ²² , ²³ Cetuximab‐conjugated IR700 against EGFR is being tested in an ongoing Phase III clinical trial in recalcitrant/inoperable head and neck squamous cell cancer (HNSCC) while NIR‐PIT received approval for clinical application by Japanese medical authorities for recurrent HNSCC. EGFR is an attractive and frequently exploited target because of its important role in human cancer. ²⁴ Unfortunately, the positivity proportion of EGFR in pancreatic cancer is relatively limited (62%‐69%). ²⁵ , ²⁶ , ²⁷ Therefore, to circumvent restrictions from the relatively limited EGFR‐positive nature of PDAC, we previously reported rBC2LCN as a reactive lectin, ¹⁰ indicating that a lectin‐binding, fucosylated glycan epitope could be an attractive therapeutic target for PDAC. ⁸ , ²⁸ , ²⁹ Promisingly, rBC2LCN lectin has been recently shown to have positive reactivity to diverse cancers of differing tissue origins, ³⁰ , ³¹ , ³² , ³³ suggesting a potential, generalized application of rBC2‐IR700 when no cancer‐specific cell surface receptors or clinically approved antibodies are available.

Our in vitro work, in which the Lec‐PT system successfully bound to Capan‐1 cells but did not induce death until NIR exposure, showed the potential of naturally derived lectins in targeting specific surface tumor cell glycans, similarly to reports of glycoprotein binding to angiogenesis‐ and extracellular matrix‐specific factors that mediate cell‐cell adhesion. ³⁴ However, the primary issue with any therapy (antibody, lectin, etc) is not only target specificity but achieving a sufficient binding threshold and duration to facilitate cargo delivery and therapeutic effect. In these cases, the lower binding affinity (single epitope interaction) would be compensated for by a target‐rich environment as multivalent binding across high‐density cell surface glycans increases the avidity (total binding potential across all interactive surfaces). ³⁵ We showed this in our in vivo work, where such effective multivalent binding (evidenced by tumoricidal activity upon NIR similar to reported antibody‐photoabsorber conjugates targeting EGFR) implies that the specificity of the lectin‐glycan interaction imparts both sufficient binding and duration to create a phototherapeutic effect. It is noteworthy that rBC2‐IR700 shows stronger accumulation in PDX tumors than Capan‐1 xenograft tumors, indicating that relative rBC2 reactivity is nonuniform across samples. Indeed, we have previously reported that rBC2LCN has strong reactivity toward the majority of human PDAC, albeit with differing degrees of avidity. ¹⁰ Given that most human PDAC cell lines are rBC2LCN negative despite the fact that most clinical PDAC are rBC2LCN positive, it is likely that cell lines maintained in vitro for extended periods of time have diverged from normal glycan expression patterns due to the absence of microenvironmental cues. Nevertheless, in our study, we showed that rBC2‐IR700 Lec‐PT is effective against both Capan‐1 and PDX tumors. Therefore, it is highly plausible that Lec‐PT will show even greater therapeutic efficacy toward clinical PDAC. Furthermore, lectins may possibly penetrate even into tumor parenchyma due to their lower binding affinity. It has previously been reported that, as the affinity of antibodies for their target antigen increases, their distribution becomes more perivascular in nature. ³⁶ Thus, lectins could improve intratumoral distribution of the conjugate and lead to more effective treatment, ³⁷ especially for PDAC characterized by dense, fibrotic tissue.

In practice, Lec‐PT should provide clinical benefit for patients with locally advanced or borderline resectable pancreatic cancers which are high risk for margin‐positive resection. For these patients, neoadjuvant treatment, including chemotherapy or chemoradiation, is applied to reduce the tumor load and increase the likelihood of R0 resections that translate directly into increased survival rates. ³⁸ However, due to systemic toxicity, neoadjuvant therapy is usually limited to patients with a good performance status. We previously proposed the use of rBC2‐PE38 lectin‐exotoxin conjugate to target PDAC with implication for its use as a neoadjuvant therapy. However, due to the nonspecific nature of that exotoxin, the 100% lethal dose (LD100) in the equivalent animal study was determined to be 10 μg rBC2. ¹⁰ In our study, we sought to improve specificity and reduce unintended side‐effects to widen human applications of this rBC2‐based neoadjuvant therapy. For direct comparison, we adopted 20 μg rBC2, since the approximate 50% labeling efficiency of rBC2 to IR700DX equates to 10 μg of therapeutic rBC2 with close to 100% conjugation efficiency. This both renders the same effective dose as the previous study and establishes a proof of concept that Lec‐PT functions in a tumoricidal fashion without systemic toxicity. In fact, our results indicate that rBC2‐IR700 accumulate in the tumor in a dose dependent manner but not the systemic background. Although there is some dose‐dependent accumulation of rBC2‐IR700 in the liver and kidneys (organs associated with drug clearance), we demonstrated no observable unintended damage provided that these organs are not irradiated by NIR light. Thus, Lec‐PT could be a new clinical modality for preoperative treatment of pancreatic cancer with no noticeable systemic toxicity, leading to a better general condition for the operation. While our preclinical animal experiments utilized an open laparotomy approach for NIR light irradiation using an LED light source, a laparoscopic approach using an NIR laser system (BioBlade; Rakuten Medical Inc., San Diego, California, clinically approved by the Japanese Pharmaceuticals and Medical Devices Agency) should be adopted for minimally invasive clinical use.

Given that highly perfused organs (such as the liver) absorb fluorescent light, the possibility of underestimating rBC2‐IR700 accumulation further highlights the benefits of using lasers for precise, targeted irradiation in any follow‐up studies and/or Phase 0 clinical trials. Moreover, a study that directly compared LEDs and lasers for in vivo NIR‐PIT applications determined that lasers outperform LEDs in every aspect, including greater selectivity and minimal cytotoxicity. ³⁹ Therefore, while we shielded organs in our in vivo experiments to demonstrate the feasibility of our rBC2‐IR700‐based, NIR light‐activated cytocide, shield‐free treatments using precision laser light systems will be the definitive approach for clinical use.

We also explored the utility of single vs repeated treatments using Lec‐PT and found that a regimen of repeated treatments (1) did not induce significant additional injuries (data not shown) and (2) maintained reductions in tumor volumes over the entire experimental period. However, a notable limitation of phototherapy for pancreatic cancer in a repeated usage case would be light delivery since laparoscopic NIR still requires general anesthesia. Less‐invasive developments in NIR delivery, such as a flexible catheter with light emitting diodes (LEDs) that can be inserted into the pancreatic duct under endoscopic retrograde cholangiopancreatography, would be preferable for repeat treatment cases. ⁴⁰ Another limitation of our study is potential immunogenicity in humans due to the xenobiotic (bacterial) nature of the rBC2LCN lectin. A promising way to avoid such immunogenicity would be to use endogenous lectins as payload carriers. Although 150 to 200 endogenous lectins are known to exist in the human organism, their in vivo dynamics, especially their tumor directivity, are not well understood. ⁴¹ Exploitation of such endogenous lectins with tumor‐reactive capacity will become a core foundation of Lec‐PT without immunogenicity. Alternatively, the rapidly developing field of lectin engineering may hopefully be able to produce nonimmunogenic and tumor‐directed artificial lectins. ⁴² , ⁴³

Individualized adjunct care, featuring lectins with low side effects and the possibility to safely repeat treatments, will certainly facilitate persistent reductions in tumor volume as well as enhance survival and quality of life for pancreatic cancer patients.

5. CONCLUSION

Here, we reported Lec‐PT a novel combination of PDAC‐targeting lectin rBC2LCN and NIR photoabsorber IR700DX that is effective and specific against PDAC xenografts in a mouse model. Our study serves as a proof of concept to demonstrate that lectin can serve as a cancer targeting molecule in place of antibodies used in NIR‐PIT. The frequently aberrant glycocalyx expression in cancer thus actualizes the utility of Lec‐PT, especially in cancer types with no approved cell surface antibodies. Our in vitro data also revealed comparable changes in cellular morphology to those reported for NIR‐PIT, suggesting that Lec‐PT may also elicit anticancer immune responses even in immune cold cancers like PDAC and warrant further examination.

AUTHOR CONTRIBUTIONS

Conceptualization: Yukihito Kuroda, Tatsuya Oda, Hiroaki Tateno. Methodology: Yukihito Kuroda, Tatsuya Oda, Osamu Shimomura, Pakavarin Louphrasitthiphol, Takahito Nakajima, Hiroaki Tateno. Validation: Yukihito Kuroda, Tatsuya Oda, Pakavarin Louphrasitthiphol. Formal analysis: Yukihito Kuroda. Investigation: Yukihito Kuroda, Kinji Furuya, Tomoaki Furuta, Hiromitsu Nakahashi. Resources: Tatsuya Oda, Yoshihiro Miyazaki, Takahito Nakajima, Hiroaki Tateno. Writing—original draft: Yukihito Kuroda, Tatsuya Oda, Pakavarin Louphrasitthiphol, Bryan J. Mathis. Writing—review & editing: Yukihito Kuroda, Tatsuya Oda, Pakavarin Louphrasitthiphol, Bryan J. Mathis. Visualization: Yukihito Kuroda, Tatsuya Oda, Shinji Hashimoto, Yoshimasa Akashi. Supervision: Tatsuya Oda, Pakavarin Louphrasitthiphol, Hiroaki Tateno. Project administration: Tatsuya Oda. Funding acquisition: Tatsuya Oda, Osamu Shimomura. The work reported in the paper has been performed by the authors, unless clearly specified in the text.

CONFLICT OF INTEREST

The authors declare no potential conflicts of interest.

ETHICS STATEMENT

Pancreatic cancer tissue samples were taken from patients who gave informed consent. All procedures were approved by the Institutional Review Board of the University of Tsukuba Hospital (IRV code: H28‐90). All animal experiments performed in the study were approved by the University of Tsukuba Animal Ethics Committee (Authorization Number: 21‐212).

Supporting information

Appendix S1. Supporting Information.

Video S1. Real‐time fluorescence microscope images of LIVE/DEAD staining in Capan‐1 cells.

Kuroda Y, Oda T, Shimomura O, et al. Lectin‐based phototherapy targeting cell surface glycans for pancreatic cancer. Int J Cancer. 2023;152(7):1425‐1437. doi: 10.1002/ijc.34362

DATA AVAILABILITY STATEMENT

The data that support the findings of our study are available from the corresponding author upon reasonable request.