Nanobody-targeted photodynamic therapy for the treatment of feline oral carcinoma: a step towards translation to the veterinary clinic

2.1 Cell lines and culture

The human cervical adenocarcinoma cell line HeLa and the human mammary adenocarcinoma cell line MCF7 were purchased from ATCC (ATCC CCL-2 and HTB-22). The feline oral squamous cell carcinoma cell lines SCCF1, SCCF2, and SCCF3 were kindly provided by Dr. Rosol (Ohio University). SCCF1 derives from a laryngeal SCC, SCCF2 from a bone-invasive gingival SCC, and SCCF3 from a lingual SCC [29, 30]. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose and ultraglutamine 1 (Lonza, Basel, Switzerland) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, Zwijndrecht, the Netherlands), 100 U/mL penicillin and 100 μg/mL streptomycin (Sigma-Aldrich). Cells were cultured at 37 °C and 5% CO₂.

2.2 Nanobody and conjugation to fluorophore/photosensitizer

NB_A was produced and purified from the periplasmic fraction of E. coli as previously described [31].

For fluorescence detection, NB_A was conjugated to the fluorophore Alexa Fluor 647 NHS ester (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s protocol, yielding a degree of conjugation (DOC), i.e. PS to NB ratio, of 1. Briefly, NB_A in PBS was incubated with the fluorophore for 2 h at room temperature, in a molar ratio of 1–4. Thereafter, free fluorophore was removed by size exclusion chromatography using three consecutive Zeba Spin Desalting Columns (Thermo Fisher Scientific, Perbio Science Nederland, Etten–Leur, the Netherlands).

For NB-PDT assays, NB_A was conjugated to the PS IRDye700DX (LI-COR, Biosciences, Lincoln, NE, USA). For this, the provider’s protocol was followed, controlling the conjugation conditions (i.e. time and temperature) to achieve NB-PS conjugates with DOC 0.5, 1, and 2.5, the latter referred to as NiBh. Free PS was removed using four consecutive Zeba Spin Desalting Columns. Further characterization of the NB-PS conjugates was performed as described in the Supplementary Materials.

2.3 Immunohistochemistry

Paraffin-embedded tissue blocks of feline OSCC were obtained from the archive of the Pathology division of the Veterinary Faculty at Utrecht University. Ten cases which contained OSCC as well as normal oral epithelium within the same tissue block were selected. For immunohistochemistry, 4 µm-thick tissue sections were mounted onto slides, deparaffinized and rehydrated. Antigen retrieval took place for 4 min at 37 °C with proteinase K (DAKO, Amstelveen, the Netherlands, cat no. S3020). Endogenous peroxidase was blocked with dual endogenous enzyme block (DAKO, S2003) for 10 min at room temperature and tissue was then blocked with PBS + 0.1% Tween (PBST) + 10% BSA for 1 h at room temperature. Tissue was incubated overnight at 4 °C with mouse anti-EGFR antibody (Thermo Fisher Scientific, MA5-13269) diluted 1:100 in PBST + 1% BSA. This primary antibody is a species cross-reactive antibody known to bind EGFR in dog, human, mouse, sheep, nonhuman primate and cat. Its detection was performed with the kit Envision + System-HRP anti-mouse (DAKO, K4401) for 30 min at room temperature. Staining was visualized by applying DAB chromogen (DAKO, K3468) and reaction stopped after 4 min. Slides were counterstained with hematoxylin. Canine skin was used as positive control tissue. As negative control, normal mouse IgG (Santa Cruz Biotechnology, Heidelberg, Germany, cat no. sc-2025) was used instead of the primary antibody. Immunoreactivity was evaluated as the product of percentage of positive tumor cells (1 ≤ 10%, 2 = 10–30%, 3 = 31–60%, or 4 ≥ 60%) and staining intensity (1 = low, 2 = moderate, and 3 = high), as in previous publications [22, 32]. Scores ≥2 were considered positive. To quantify staining intensity, ImageJ was used for deconvolution of the DAB color spectra and the optical density recorded in the regions of interest (tumor nests or basal layer of normal epithelium). Outcome was verified by an experienced veterinary pathologist (Guillaume C.M. Grinwis).

For immunohistochemistry with fluorescence detection, slides were subjected to the same steps as explained above. However, after incubation with the primary antibody, goat antimouse Alexa 555 (Invitrogen, A21424) was used as the secondary antibody (1:200 in PBST + 1% BSA) for 1 h at room temperature. Alternatively, to detect EGFR with NB_A instead, 10 nM of directly labeled NB_A-Alexa 647 were incubated overnight at 4 °C. Slides were counterstained with DAPI (Roche, Basel, Switzerland). Images were taken with a confocal laser scanning microscope (Carl Zeiss Microscopy GmbH, Germany, LSM700) using a plan-apochromat 63x/1.40 Oil DIC and 20x/0.8 M27 objectives.

2.4 Binding assay

To assess the apparent binding affinity (K_D) of NBs to the target in its natural conformation, EGFR-expressing human and feline cells were seeded in 96-well plates (10,000 cells per well) and incubated at 37 °C. The next day, cells were incubated for 2 h at 4 °C with a concentration range of NB (0.2–100 nM) in binding medium (DMEM without phenol red, supplemented with 25 mM HEPES and 1% BSA, pH 7.2). Unbound NB was washed off and cells fixed with 4% PFA (Merck, Haarlem, the Netherlands) for 10 min at room temperature. Bound NB was detected by incubating with rabbit anti VHH antibody (QVQ, Utrecht, the Netherlands, cat no. QE19) for 1 h at room temperature (1:1000 in PBS + 1% BSA) followed by goat anti rabbit IRDye800CW (LI-COR) for 1 h at room temperature (1:2000 in PBS + 1% BSA). Fluorescence at 800 nm was detected with an Odyssey infrared scanner (LI-COR). To assess the apparent binding affinity of the conjugates NB_A-PS and NiBh, the same procedure was followed, but plates were directly scanned at 700 nm after washing off the unbound conjugate. Data was analyzed with GraphPad Prism software (GraphPad, La Jolla, CA, USA) and K_D values determined using a nonlinear fit with one-site specific binding. The K_D, i.e. apparent dissociation constant, corresponds to the concentration of NB at which half of the total EGFR molecules are associated with NB.

The binding of NiBh was also assessed in the presence of EGF. For this, 10 nM of NiBh was incubated with the cells in the absence or presence of an equimolar concentration of EGF or a 10× molar excess of EGF (R&D Systems, Minneapolis, MN, USA, cat no. 236-EG).

2.5 Flow cytometry

Cells were added in a U-bottom 96 well-plate (10⁵ cells per well), washed once with PBS + 1% BSA and incubated for 45 min at 4 °C with 20 nM of mouse anti EGFR antibody. After washing, secondary antibody goat antimouse Alexa 488 diluted 1:200 was added and incubated with the cells for 30 min at 4 °C. Alternatively, to detect EGFR with NB_A instead, 40 nM of labeled NB_A-Alexa 647 was incubated with the cells for 30 min at 4 °C. Unstained controls and samples stained with secondary antibody only were taken along for each cell line. Measurements were performed with a FACS Canto II (BD Biosciences, San Jose, CA, USA) and further analyzed with FlowLogic software (Inivai Technologies). The fluorescence intensity of both fluorophores was normalized to compare both in a single graph.

Resection material from two cats with OSCC was taken with informed owner consent at the University Clinic for Companion Animal Health (Utrecht University). To perform flow cytometry with primary feline OSCC tissue, the tissue was first cut into small fragments (∼5 mm³) and incubated for 30 min at 37 °C with 0.125% trypsin in Advanced DMEM/F12 (Life Technologies, Carlsbad, CA, USA) supplemented with 1× GlutaMAX, 100 U/mL penicillin–streptomycin and 10 mM HEPES (Life Technologies). The resulting cell suspension was filtered through a 100 μm cell strainer, followed by a 70 μm cell strainer. Red blood cells were lysed by incubating with a homemade RBC lysis buffer for 5 min on ice. Cell suspension was washed and subsequently stained for flow cytometry as described above for the cell lines.

2.6 Immunofluorescence on cells

Ten thousand cells per well were seeded in 16 wells Lab-Tek Chamber Slides (Thermo Fisher Scientific, 178599) and incubated at 37 °C. The next day, cells were washed with binding medium and costained with 20 nM mouse antiEGFR antibody and 40 nM NB_A-Alexa 647 in binding medium. Incubation took place for 1.5 h at 4 °C and, subsequently, the cells were fixed with 4% PFA for 10 min at room temperature. The secondary antibody goat antimouse Alexa 488 (Invitrogen, A11029) was added for 1 h at room temperature (1:200 in PBS + 1% BSA). Cells were stained with DAPI and imaged with a confocal microscope.

2.7 Nanobody-targeted photodynamic therapy

NB-PDT was performed as previously described [1]. Briefly, cells were seeded in 96-well plates (10,000 cells/well) one day before the assay and incubated at 37 °C. The next day, cells were incubated with a concentration range of NB-PS conjugate (0.78–100 nM) for 30 min at 37 °C in PDT medium, i.e. DMEM without phenol red and L-glutamine (Lonza) supplemented with 10% FBS and antibiotics. Unbound conjugate was washed off and the plate was scanned with the Odyssey scanner at 700 nm to detect association of the conjugate with cells. Thereafter, cells were illuminated with 7 mW/cm² for 59 min (25 J/cm²) using a 690 nm laser (Modulight ML7700, Tampere, Finland). Fluence rate was monitored with an Orion/PD optometer (Ophir Optronics, Jerusalem, Israel). After illumination, the plates were placed back at 37 °C.

NB-PDT was also performed with an excess of NB_A, i.e. under competing conditions, with some minor adjustments of the above-mentioned protocol. Briefly, before adding NiBh (50 nM), cells were preincubated for 10 min at 37 °C with 10× or 50× molar excess of NB_A. The unconjugated NB was present as well during the incubation with NiBh. In a different experiment, NB-PDT was performed in the presence of 25 mM or 50 mM sodium azide, as a quencher of singlet oxygen [33]. In this case, sodium azide was added to the cells only during the illumination time.

2.8 Cell viability and cell death after NB-PDT

One day after NB-PDT, cells were incubated with Alamar Blue reagent (Bio-Rad, Hercules, CA, USA) to assess viability, according to the protocol of the manufacturer. Fluorescence was measured with a FLUOstar Optima microplate reader (BMG Labtech, Ortenberg, Germany) and results expressed as cell viability in percentage relative to untreated cells. The median lethal dose (LD50), i.e. concentration of conjugate to achieve 50% of cell death, was determined using GraphPad Prism software with a log (inhibitor) vs normalized response fit.

Alternatively, 2 and 24 h after NB-PDT, live and dead cells were distinguished by staining with calcein AM (invitrogen) and propidium iodide (PI) (invitrogen) at a final dilution of 1:2000 and 1:1000, respectively, for 10 min at 37 °C. Cells were imaged with an EVOS microscope (Thermo Fisher Scientific) using transmitted light, a GFP cube for calcein AM, and an RFP cube for PI.

2.9 Co-cultures and specificity assays

SCCF2 and MCF7 cells were brought in suspension and labeled with the cell tracking dyes ViaFluor 405 and ViaFluor 488 (Biotium, Amsterdam, the Netherlands, cat no. 30068 and 30086), respectively, according to the provider’s protocol. Both cell lines were mixed in a ratio 1:1, seeded in 16 wells Lab-Tek Chamber Slides (Thermo Fisher Scientific) at a density of 10,000 cells per well and cultured at 37 °C. Cells were used the next day for further assays.

The cocultures were incubated with 50 nM of NB_A-Alexa 647 in culture medium for 30 min at 37 °C. Alternatively, NB_A-Alexa 647 was incubated for 1 h at 4 °C in binding medium. Thereafter, cells were fixed, stained with DAPI and imaged with a confocal microscope.

In a different assay, cocultures were treated with NB-PDT using 50 nM of NiBh. Cells were placed back in the incubator and, 2 h after illumination, cells were stained with PI (1:1000) and directly imaged with a confocal microscope.

3.1 EGFR expression in feline oral squamous cell carcinoma and surrounding normal oral epithelium

As EGFR has been described to be overexpressed in feline OSCC [22, 23], immunohistochemistry was performed on 10 cases of feline OSCC to investigate the expression of this protein in neoplastic cells and to elucidate its presence in cells of the surrounding normal oral mucosa (epithelium and stroma). EGFR presence, with predominant cell membrane localization, was detected in neoplastic cells forming nests and trabecula, and not found to be expressed in the surrounding stroma cells (Figure 1(a), right). On the other hand, membranous EGFR expression was also detected in the normal adjacent oral epithelium, more prominently in the basal cell layer and decreasing towards the outer, more differentiated epithelial layers (Figure 1(a), left). Interestingly, a similar pattern of differentiation and loss of EGFR expression was observed towards the core of large neoplastic nests which typically contain more differentiated neoplastic cells that can even show keratinization. For all 10 studied cases, EGFR positivity in the basal epithelial layer of preexisting normal epithelium was in the same range as in the neoplastic nests (Figure 1(b)). Accordingly, nine of the ten investigated cases expressed intermediate to high EGFR levels in the neoplastic nests, but EGFR was not overexpressed compared to the expression observed in the basal layer of the epithelium (Figure 1(c)). Altogether, these observations support that EGFR is a relevant target due to its expression in tumor cells and its absence in stroma.

Figure 1:

EGFR expression in feline OSCC and surrounding tissue. Immunohistochemistry was performed in 10 cases of OSCC with normal adjacent oral epithelium.

(a) Representative images are shown of neoplastic nests compared to normal neighboring epithelium (left) and neoplastic nests surrounded by stroma (right), for two cases with strong (top), and intermediate (bottom) EGFR expression. Scale bar, 100 µm (left images) and 50 µm (right images). (b) Quantification of the EGFR signal in neoplastic nests (tumor), mucosal stroma (stroma), and basal layer of the adjacent oral epithelium (epithelium) per case. (c) Immunoreactivity score and classification of each case based on EGFR intensity calculated in B and % of positive tumor cells.

3.2 Species cross-reactivity of NB_A to EGFR in human and feline cells

Having identified EGFR as a promising molecular target in feline oral carcinoma, we first confirmed the presence of membrane EGFR on three feline OSCC cell lines (SCCF1, SCCF2, and SCCF3), in comparison to two well-characterized human cancer cell lines (HeLa and MCF7) (Figure S1a). This was performed with flow cytometry using a commercial EGFR-targeting, species cross-reactive antibody. Thereafter, the binding affinity of a panel of NBs originally selected against human EGFR (with binding affinities, K _D, <10 nM) was assessed on SCCF1 cells. From this NB panel, NB_A was identified as the most promising candidate due to its high binding affinity (K _D ∼ 0.44 nM) for feline cells (Figure 2(a)). To further investigate the species cross-reactivity of NB_A and its EGFR specificity, its binding to the three feline OSCC cell lines and the two human cancer cell lines was evaluated using flow cytometry and immunofluorescence. As a reference, the commercial EGFR-targeting antibody was used. NB_A was first conjugated to the fluorophore Alexa 647 (with binding affinity comparable to NB_A, data not shown) to enable direct detection in the assays. The three feline OSCC cell lines had EGFR levels in the range of HeLa cells, while the fluorescence signal was minimal on the low EGFR-expressing MCF7 cells (Figure 2(b) and (c)). The same EGFR expression trend and pattern on the different cell lines was detected by the reference antibody, indicating the capability of NB_A to bind both human and feline EGFR (Figure 2(b) and (c)). Further supporting this, EGFR was also detected by NB_A (and the commercial antibody) on feline OSCC tissue (Figure 2(d)), predominantly at the basal epithelium and neoplastic nests. Interestingly, EGFR levels on primary feline OSCC tissue were in the same range as HeLa cells (Figure 2(e)) and, therefore, comparable to the feline OSCC cell lines used in this study. Quantification of the number of membrane EGFR molecules on the used human and feline cell lines indicated around 175,000 EGFR molecules per cell, while MCF7 cells (10,000 receptors per cell [34]) were below the detectable limit (Figure S1b).

Figure 2:

EGFR expression on human and feline cells and tissue detected by NB_A.

(a) A binding assay was performed with representative NBs, previously selected against human EGFR, using feline SCCF1 cells. The graph displays the normalized binding curves of several NBs to SCCF1 cells. The apparent binding affinities (K _D) of the different NBs are shown. (b) and (c) Membrane EGFR was detected on three feline OSCC cell lines (SCCF1, SCCF2 and SCCF3) and two human cancer cell lines (HeLa and MCF7). (b) Membrane EGFR levels measured by flow cytometry using a commercial antiEGFR antibody or NB_A-Alexa 647, expressed as median fluorescence intensity and normalized based on HeLa cells. (c) Confocal microscope images of each cell line co-stained for membrane EGFR using NB_A-Alexa 647 (red, middle panel) and a commercial species cross-reactive antibody (green, bottom panel). Nuclei were stained with DAPI (blue). Merged images are shown on the top panels. Scale bar, 15 µm. (d) Feline OSCC tissue sections stained with NB_A-Alexa 647 (left) or a commercial antiEGFR antibody (right). Images show an overview of the OSCC tissue including adjacent normal oral epithelium, and a close-up of the tumor cells. Scale bar, 100 µm (left images) and 15 µm (right images). (e) Membrane EGFR was detected on HeLa cells and primary tissue samples of two feline OSCC cases (T1 and T2). EGFR levels were measured by flow cytometry using a commercial antiEGFR antibody, expressed as median fluorescence intensity (MFI).

3.3 Characterization of NiBh as an agent for EGFR-targeted PDT

NB_A was conjugated to the PS aiming to obtain conjugates with a different DOC, i.e. PS to NB ratios, in order to increase the PS density on tumor cells and, thus, cytotoxicity upon illumination. Accordingly, three conjugates were synthesized with 0.5, 1, or 2.5 PS molecules per NB molecule, named NB_A-PS(0.5), NB_A-PS(1), and NiBh, respectively. The conjugates were characterized in terms of purity by SDS-PAGE and the PS absorbance spectra was acquired, showing that the absorbance properties of the PS were not affected when conjugated to the NB, regardless of the DOC (Figure S2). Although the apparent binding affinities of the conjugates to both human and feline cells were slightly affected with an increasing degree of PS modification, the affinities still remained in the low nanomolar range (Figure 3(a)). Furthermore, the specificity of the conjugates for (feline) EGFR was verified by using EGFR knockdown SCCF1 cells, which resulted in a considerably reduced binding of the conjugate (Figure S3). When employing these conjugates for NB-PDT, it is evident that the use of NiBh resulted in the highest fluorescence (or density) of PS associated with the cells (Figure 3(b)) and, after illumination, NiBh was the only conjugate that induced significant cytotoxicity to the feline cells (Figure 3(c)).

Figure 3:

Binding affinity and potency of different NB_A-PS conjugates.

(a) A binding assay was performed with NB_A, NB_A-PS(0.5), NB_A-PS(1), and NiBh using HeLa (top) and SCCF1 cells (bottom). The graphs display the saturation binding curves of each targeting molecule per cell line. Apparent binding affinity (K _D) values of each NB and conjugate are shown. (b) To perform NB-PDT, a concentration range of NB_A-PS(0.5), NB_A-PS(1) or NiBh was incubated with SCCF2 cells for 30 min at 37 °C. The graph shows the PS signal on cells detected after washing off unbound conjugate, right before illumination. (c) The viability of the cells was assessed 24 h after illumination and expressed in percentage relative to nontreated cells.

NiBh, the NB_A-PS conjugate with highest payload, was selected as the best agent for EGFR-targeting NB-PDT on feline OSCC cells. The special feature of this conjugate is its high binding affinity to both human and feline cells (K_D ∼ 7–10 nM) while having DOC 2.5 (Figure 4(a)). In addition, the specific binding of NiBh to feline EGFR was maintained, as evidenced by a reduced NiBh binding in the presence of EGF in a concentration dependent manner (Figure 4(b)). The potency of NiBh as an agent for NB-PDT was evaluated using the panel of human and feline cell lines (Figure 4(c)). HeLa, SCCF2, and SCCF3 cells were effectively killed after treatment with comparable low nanomolar LD50 values (HeLa, 15.9 ± 3.7 nM; SCCF2, 17.8 ± 2.2 nM; SCCF3, 26 ± 3.5 nM). In line with the very low EGFR expression on MCF7 cells, these cells were only slightly affected at the highest concentrations of the conjugate. On the other hand, the use of NiBh for NB-PDT on SCCF1 cells resulted in only slight cytotoxicity, contrary to the other cell lines with comparable EGFR expression. To further investigate whether the induced cytotoxicity relies on the specific binding of NiBh to EGFR, NB-PDT was performed in the presence of an excess of unconjugated NB_A. Accordingly, the cytotoxicity decreased under these competing conditions (Figure 4(d)). Moreover, the NB-PDT effect could be inhibited in the presence of sodium azide (NaN₃), a singlet oxygen quencher [33] (Figure 4(d)). To visualize the distinct NB-PDT effect on moderate and low EGFR-expressing cells, SCCF2, and MCF7 cells were stained with fluorescent dyes denoting live/dead cells after NB-PDT. Already early after NB-PDT (2 h), dead SCCF2 cells were clearly distinguishable and their number increased over time, while only a very small number of dead MCF7 cells could be detected one day after treatment (Figure 4(e)).

Figure 4:

Specific binding and potency of NiBh.

(a) A binding assay was performed with NiBh using the panel of human and feline cell lines (HeLa, SCCF1, SCCF2, SCCF3, and MCF7 cells). The graph displays the saturation binding curves of NiBh to each cell line from which K_D values were calculated. (b) Binding of NiBh (10 nM) to SCCF2 cells in the absence or presence of equimolar concentration and 10× molar excess of EGF. The fluorescence corresponding to bound NiBh under each condition is displayed in the graph. (c) NB-PDT using NiBh was performed on the panel of cell lines and, one day later, viability was assessed. The graph shows viability curves for each cell line, in percentage relative to untreated cells. (d) NB-PDT (50 nM NiBh) was performed on SCCF2 cells in the absence or presence of a molar excess of unconjugated NB_A. NB-PDT was also performed in the presence of sodium azide (NaN₃). A control consisting of cells exposed to light, but no conjugate, was included. Cell viability was assessed the next day and represented as percentage relative to untreated cells. (e) NB-PDT (50 nM NiBh) was performed on SCCF2 or MCF7 cells and cells stained with calcein and propidium iodide to visualize live (green) and dead (red) cells, respectively. A control in which light was applied, but no conjugate, was taken along. Images were taken 2 and 24 h after NB-PDT. Scale bar, 20 µm.

3.4 Specific tumor cell killing mediated by NB-PDT using NiBh

To investigate the specificity of NiBh for NB-PDT in a representative and biologically relevant setting, co-cultures were set up consisting of feline OSCC cells (SCCF2), as neoplastic nests, and low EGFR-expressing cells (MCF7) mimicking the surrounding stroma of the oral mucosa. First, NB_A-Alexa 647 was employed to visualize its differential binding (4 °C) to both cell lines. As anticipated, NB_A was found on the membrane of the feline carcinoma cells, while the detected signal was minimal on the surrounding low EGFR-expressing cells (Figure 5(a), top). The next step was to address the accumulation of NB_A-Alexa 647 in cells after an incubation time mimicking the incubation with NB-PS conjugate used for the NB-PDT studies (30 min at 37 °C). This revealed a predominant accumulation of the NB in the feline tumor cells both membrane-bound and internalized (Figure 5(a), bottom). Lastly, NB-PDT using NiBh was performed on the cocultures and dead cells were visualized via PI staining. Most feline neoplastic cells were killed shortly after treatment (2 h), whereas neighboring low EGFR-expressing cells were left unaffected (Figure 5(b)).

Figure 5:

Specificity of NB_A and NiBh in cocultures. SCCF2 cells were labeled with ViaFluor 405 (blue), MCF7 cells with ViaFluor 488 (green), and co-cultured in a 1:1 ratio.

(a) Confocal images of cells incubated with 50 nM of NB_A-Alexa 647 (red) for 1 h at 4 °C (top) or 30 min at 37 °C (bottom). Scale bar, 15 µm. (b) Confocal images of cells 2 h after NB-PDT with 50 nM of NiBh. Cells were stained with propidium iodide to distinguish dead cells (red). Scale bar, 15 µm.