Selective targeted delivery of the TNF-alpha receptor p75 and uteroglobin to the vasculature of inflamed tissues: a preliminary report

Abstract

Background: Ligand-targeted approaches have proven successful in improving the therapeutic index of a number of drugs. We hypothesized that the specific targeting of TNF-alpha antagonists to inflamed tissues could increase drug efficacy and reduce side effects.

Results: Using uteroglobin (UG), a potent anti-inflammatory protein, as a scaffold, we prepared a bispecific tetravalent molecule consisting of the extracellular ligand-binding portion of the human TNF-alpha receptor P75 (TNFRII) and the scFv L19. L19 binds to the ED-B containing fibronectin isoform (B-FN), which is expressed only during angiogenesis processes and during tissue remodeling. B-FN has also been demonstrated in the pannus in rheumatoid arthritis. L19-UG-TNFRII is a stable, soluble homodimeric protein that maintains the activities of both moieties: the immuno-reactivity of L19 and the capability of TNFRII to inhibit TNF-alpha. In vivo bio-distribution studies demonstrated that the molecule selectively accumulated on B-FN containing tissues, showing a very fast clearance from the blood but a very long residence time on B-FN containing tissues. Despite the very fast clearance from the blood, this fusion protein was able to significantly improve the severe symptomatology of arthritis in collagen antibody-induced arthritis (CAIA) mouse model.

Conclusions: The recombinant protein described here, able to selectively deliver the TNF-alpha antagonist TNFRII to inflamed tissues, could yield important contributions for the therapy of degenerative inflammatory diseases.

Figure 1

L19-hUG-TNFRII characterization. A) Scheme of the cDNA of L19-UG-TNFRII including the leader peptide; CMV cytomegalovirus promoter. B) Schematic representation of homodimeric L19-UG-TNFRII. C) SDS-PAGE analysis of purified L19-UG-TNFRII under reducing (R) and non-reducing (NR) conditions; on the left the molecular masses of the standards in kDa. D) Size exclusion chromatography profile (Superdex 200) of purified L19-UG-TNFRII; mAU, milli-absorbance units.

Figure 2

In vitro L19-hUG-TNFRII biological activity. A) Inhibitory activity of TNF-alpha cytotoxicity by different concentrations of L19-UG-TNFRII using LM mouse fibroblasts treated with 2 pM TNF-alpha. B) L19-UG-TNFRII bound to ED-B is able to inhibit TNF-alpha cytotoxicity (in situ inhibition). The cytotoxic activity of 25 pM TNF-alpha was evaluated on LM cells using plates pre-coated with the recombinant FN fragment 7.ED-B.8.9 and pre-incubated with different concentrations of L19-UG-TNFRII (5-3125 pM). To assess the role of the L19 component the experiment was also carried out mixing the L19-UG-TNFRII with a large excess of ED-B to inhibit of the scFv L19 moiety. After washing out the unbound fusion protein, TNF-alpha was inhibited only by L19-UG-TNFRII bound to the recombinant FN fragment with which the plates were coated. The means ± S.D. are reported.

Figure 3

In vivo L19-hUG-TNFRII bio-distribution and therapy properties. A-B) Bio-distribution of the radio-iodinated L19-UG-TNFRII in F9 teratocarcinoma-bearing mice. A) The %ID/g in the tumour and in the blood at the indicated times after i.v. injection of the radio-iodinated L19-UG-TNFRII are shown (mean ± S.D.). B) The tumour to blood ratio of the %ID/g at different times after injection of the radio-iodinated protein are plotted. C) In vivo therapy experiments with L19-UG-TNFRII in CAIA animal models. The median arthritis scores determined in treated and control mice at day 3 to 10 from the injection of anti-collagen antibodies are indicated. The p values determined using the non parametric Mann-Whitney U test are indicated. Bars indicate the standard error of the mean.