CDH17 nanobodies facilitate rapid imaging of gastric cancer and efficient delivery of immunotoxin

Abstract

Background: It is highly desirable to develop new therapeutic strategies for gastric cancer given the low survival rate despite improvement in the past decades. Cadherin 17 (CDH17) is a membrane protein highly expressed in cancers of digestive system. Nanobody represents a novel antibody format for cancer targeted imaging and drug delivery. Nanobody targeting CHD17 as an imaging probe and a delivery vehicle of toxin remains to be explored for its theragnostic potential in gastric cancer.

Methods: Naïve nanobody phage library was screened against CDH17 Domain 1-3 and identified nanobodies were extensively characterized with various assays. Nanobodies labeled with imaging probe were tested in vitro and in vivo for gastric cancer detection. A CDH17 Nanobody fused with toxin PE38 was evaluated for gastric cancer inhibition in vitro and in vivo.

Results: Two nanobodies (A1 and E8) against human CDH17 with high affinity and high specificity were successfully obtained. These nanobodies could specifically bind to CDH17 protein and CDH17-positive gastric cancer cells. E8 nanobody as a lead was extensively determined for tumor imaging and drug delivery. It could efficiently co-localize with CDH17-positive gastric cancer cells in zebrafish embryos and rapidly visualize the tumor mass in mice within 3 h when conjugated with imaging dyes. E8 nanobody fused with toxin PE38 showed excellent anti-tumor effect and remarkably improved the mice survival in cell-derived (CDX) and patient-derived xenograft (PDX) models. The immunotoxin also enhanced the anti-tumor effect of clinical drug 5-Fluorouracil.

Conclusions: The study presents a novel imaging and drug delivery strategy by targeting CDH17. CDH17 nanobody-based immunotoxin is potentially a promising therapeutic modality for clinical translation against gastric cancer.

Keywords: Cadherin-17; Gastric cancer; Immunotoxin; Jingbo Ma, Xiaolong Xu and Chunjin Fu are contribute equally to this work and share the first-authorship.; Nanobody; Targeted therapy.

Fig. 1

Recapitulation of CDH17 expression in gastric cancer samples and isolation of CDH17 nanobodies. a CDH17 RNA expression (TPM, RNAseq) in gastric cancers and normal stomach controls. n = 408 (tumors) and 211(controls). b CDH17 protein expression assessed by IHC in gastric TMA samples. n = 79. Scale bars, 100 μm. c Score percentage (left) and positivity rate (right) of CDH17 expression analyzed from b. d CDH17 protein expression in cell lines determined with western blot. e CDH17 immunostaining in cell membrane of IM95 and MKN45 cell lines. f SDS-PAGE gel analysis of recombinant CDH17 domain 1-3. g Nanobody screening against CDH17 domain 1-3 with phage display technology. h Sequences alignment of isolated A1 and E8 nanobodies with highlighted CDR regions

Fig. 2

Characterization of A1 and E8 nanobodies against CDH17. a SAD-PAGE analysis of purified nanobodies (left) and nanobody confirmation with HA antibody (middle) and His antibody (right). The molecular weight for three nanobodies ranged from 15 kDa to 18 kDa. b ELISA analysis of binding ability of A1 and E8 nanobodies to CDH17 domain 1-3 (n = 2). Data are representative of two independent experiments. c Determination of binding affinity to CDH17 protein by SPR analysis. The equilibrium dissociation constant K_D was 377 nM (A1) and 70.3 nM (E8) respectively. d Binding activity of A1 and E8 nanobodies in CDH17-positve cells (MKN45, IM95, TMK1 and AGS) assessed by fluorescent cell ELISA (n = 3). Both of nanobodies could recognize CDH17 protein expressed in cell membrane, while E8 nanobody shows a better performance. Data are expressed as mean ± SEM. e Validation of knockdown CDH17 with shRNA#3 in IM95 and MKN45 cells determined with western blot. f Binding specificity of E8 nanobody to CDH17 in CDH17-overexpressing and -knockdown cell lines respectively. E8 nanobody cannot obviously stain the cells with knockdown CDH17, indicating the great specificity of E8 nanobody to CDH17. g Internalization analysis of E8 nanobody with one-hour and three-hour incubations followed with HA-tag staining. An irrelevant nanobody as a control was used for all the assays above

Fig. 3

Gastric cancer imaging ex vivo and in vivo by CDH17 nanobody E8. a E8 nanobody (green) co-localization with CDH17-positve MKN45 cells (red) in zebrafish embryos. The appearance of zebrafish embryos (left) and co-localization of nanobody with cells (right). Dashed circles indicated the areas for quantification. b Quantification of co-localization (yellow) ratio to total red cells (n = 10, ****p < 0.0001, two-tailed student’s t test). Data are present as mean ± SEM. c and e Imaging of in vivo tumor-bearing mice with IR-800-labelled nanobodies in different time points (c, n = 3). Quantification analysis indicated that E8 nanobody in tumors produced significantly stronger signals as compared with control nanobody at each time point (e, n = 3, *p < 0.05, **p < 0.01, two-tailed student’s t-test). d and f Ex vivo imaging of major organs dissected from in vivo imaged mice in c (d, n = 3). Imaging quantification disclosed the strongest signals in E8-treated tumor tissues than all the control organs from both groups (f, n = 3, **p < 0.01, two-tailed student’s t-test). g Nanobody tissue distribution 12 hours after intravenous administration. Scale bars:50 μm. E8 nanobody could specifically accumulate into CDH17-positive tumor mass. Liver tissues showed some weak staining due to unspecific phagocytosis

Fig. 4

Activity evaluation of E8-PE38 immunotoxin in vitro and in vivo. a SDS-PAGE analysis of purified E8 nanobody (16 kDa), E8-PE38 (60 kDa) and Con-PE38 (60 kDa). b SPR analysis of E8-PE38 binding to CDH17. The K_D was 86.87 nM. c Cell viability detection after treatment with E8 nanobody alone, E8-PE38 and Con-PE38 in MKN45, TMK1, AGS and IM95 cells (n = 3, ****p < 0.0001, two-way ANOVA). Dashed line indicated the IC50 for E8-PE38 immunotoxin. d Schema of animal treatment schedule. e MKN45 tumor growth curves with the treatment of PBS, 0.4 or 0.6 mg/kg E8-PE38 (n = 4-5 per group, **P < 0.01, ***P < 0.001, two-way ANOVA). Mice were euthanized when tumor size reached 2000 mm³. f Individual tumor growth curves for three groups in e. g Body weight during the treatment from three groups in e. h Survival curves for treated mice in e (**p < 0.01, Log-rank (Mantel-Cox) test)

Fig. 5

Toxin is the determinant causing tumor inhibition by E8-PE38. a SDS-PAGE analysis of purified E8 (16 kDa), E8-PE38 (60 kDa) and E8-PE38 mut (60 kDa). b Cell viability assay for MKN45 and TMK1 cells treated with E8, E8-PE38 and E8-PE38 mut (n = 3, ****p < 0.0001 compared with E8 or E8-PE38 mut, two-way ANNOVA). E8 nanobody alone and E8-PE38 mut did not show cytotoxic effect on cell proliferation. c MKN45 tumor growth curves from mice treated with PBS, E8-PE38 (0.6 mg/kg), E8-PE38 mut (0.6 mg/kg) or equal molar E8 nanobody (n = 5-6, ****p < 0.001 as compared with all three control groups, two-way ANOVA). E8 alone or E8-PE38 mut did not have any tumor inhibitory effect. E8-PE38 significantly suppressed tumor growth. d Tumor weight from treated mice at the end of treatment in b (n = 5-6, ***p < 0.001, one-way ANOVA). e Body weight during the treatment from four groups in c. f and g Ki67 immunostaining in tumor tissues collected from c. Tumor tissues treated with E8-PE38 markedly inhibited the Ki67 expression and no obvious change was detection in three control groups (n = 5-6, *p < 0.05, ***p < 0.001, one-way ANOVA). Scale bars: 200 μm. h Tumor growth curves from TMK1 tumors treated with PBS, E8-PE38 (0.4 mg/kg), E8-PE38 mut (0.4 mg/kg) or equal molar E8 nanobody (n = 5, ****p < 0.0001 as compared with all three control groups, two-way ANOVA). i Survival curves of treated mice from h (n = 5, **p < 0.01, Log-rank (Mantel-Cox) test). j Tumor growth curves from MNK45 tumor bearing mice received the treatment with PBS, 5-FU (25 mg/kg), E8-PE38 (0.4 mg/kg), and combination therapy (5-FU + E8-PE38) (n = 6, ***p < 0.001, ****p < 0.0001, two-way ANOVA). k Survival curves of tumor-bearing mice treated in j (n=6, **p<0.01, ***p<0.001, Log-rank (Mantel-Cox) test)

Fig. 6

Anti-tumor effect of E8-PE38 in a PDX model. a CDH17 expression in a gastric PDX tumor. Scale bars: 40 μm. b Schema of treatment schedule in the PDX model. c Tumor growth curves from the treated PDX model with PBS, 0.4 or 0.6 mg/kg E8-PE38 immunotoxin (n = 6-7, ****p < 0.0001, two-way ANOVA). Both of dosages significantly inhibited the tumor growth. Higher dose of immunotoxin showed better anti-tumor efficacy. d Individual tumor growth curves for all three groups in c. e Body weight of the treated mice in c. f Survival curves of PDX mice treated in c (n = 6-7, *p < 0.05, **p < 0.01, ***p < 0.001, Log-rank (Mantel-Cox) test)