Novel antitenascin antibody with increased tumour localisation for Pretargeted Antibody-Guided RadioImmunoTherapy (PAGRIT)

Abstract

The Pretargeted Antibody-Guided RadioImmunoTherapy (PAGRIT) method is based on intravenous, sequential administration of a biotinylated antibody, avidin/streptavidin and (90)Y-labelled biotin. The hybridoma clone producing the monoclonal antitenascin antibody BC4, previously used for clinical applications, was found not suitable for further development because of the production of an additional, nonfunctional light chain. In order to solve this problem, the new cST2146 hybridoma clone was generated. The monoclonal antibody ST2146, produced by this hybridoma, having the same specificity as BC4 but lacking the nonfunctional light chain, was characterised. ST2146 was found able to bind human tenascin at an epitope strictly related, if not identical, to the antigenic epitope of BC4. It showed, compared to BC4, higher affinity and immunoreactivity and similar selectivity by immunohistochemistry. Biodistribution studies of biotinylated ST2146 and three other monoclonal antitenascin antibodies showed for ST2146 the highest and more specific tumour localisation in HT29-grafted nude mice. On the overall, ST2146 appears to be a good alternative to BC4 for further clinical development of PAGRIT.

Figure 1

Schematic representation of human tenascin-C, pTn28 and A–D recombinant fragments and strategy to generate a BC4-like antibody.

Figure 2

Hydroxylapatite chromatography of ST2146 (A) and BC4 (B). SDS–PAGE analysis under reducing conditions of ST2146 not digested (C, lane 1) or digested with sialidase (C, lane 2) and BC4 peaks 1, 2 or 3 from hydroxylapatite chromatography (D, P1, P2 and P3, respectively). Note: The dotted chain was found to be not functional. Therefore, divalent BC4 corresponds only to peak 3.

Figure 3

(A) Western blot analysis of BC2 (i), BC4 (ii) and ST2146 (iii) antibodies. A: pTn28; B: A–D fragment; C: tenascin. (B) Competitive ELISA. Biotinylated ST2146 was mixed with increasing concentrations of ST2146, BC4 or a not related IgG1 as competitors and plated on tenascin-coated plates. Binding was measured after addition of horseradish peroxidase–streptavidin and related chromogenic substrate.

Figure 4

ELISA plates coated with tenascin at 5 μg ml⁻¹ (A) or 0.5 μg ml⁻¹ (B). BC4 peaks 1–3 were obtained by subjecting the protein A eluted BC4 to hydroxylapatite chromatography as shown in Figure 2B.

Figure 5

BiaCore analysis of ST2146, BC4, BC2 and ST1897 on tenascin-coated chip. (A) Sensograms of indicated Mabs. All antibodies were injected at concentrations of 500, 250, 62.5, 15.6 and 3.9 nM for 60 s. The dissociation rates of the Mabs were determined over a time of 120 s. (B) Comparison of the binding and dissociation of the indicated Mabs at an injected concentration of 500 nM. (C) k_a1, k_d1, K_D1 and χ² of the fitted curves.

Figure 6

Immunohistochemistry on serial sections of breast, liver, lung and rectum carcinomas with BC4 and ST2146 antibodies. × 20 magnification.

Figure 7

Biodistribution of ¹²⁵I-labelled ST2146, BC4, BC2, ST1897 and normal mouse IgG. Data represent the mean values, ±standard error, of the % of injected dose per gram of tissue (%i.d. g⁻¹) from five animals at 48 or 72 h. P-values were obtained vs normal mouse IgG by Student's t-test.

Figure 8

Tumour/nontumour ratios of ¹²⁵I-labelled ST2146, BC4, BC2, ST1897 and normal mouse IgG at 48 and 72 h after administration.