Site-specific conversion of a pro-fusion protein toxin

Comment on Abi-Habib et al, page 2143

Substituting the native furin endoprotease activation site with a urokinase plasminogen activator recognition sequence, Abi-Habib and colleagues have constructed a diphtheria toxin–related granulocyte-macrophage colony-stimulating factor fusion protein toxin with increased utility and specificity toward acute myelogeneous leukemic cells.

Protein engineering of bacterial protein toxins has been focused largely on substitution of the native receptor-binding domain of the toxin with surrogate ligands in order to selectively redirect toxicity toward those cells that display the appropriate surface target receptor. While there has been extensive work in this field, to date the only fusion protein toxin that has been approved for human clinical use is diptheria fusion protein targeting interleukin-2 receptor (DAB₃₈₉IL-2 [ONTAK]).^1-3  Abi-Habib and colleagues, in their current article, have taken a significant step forward in the field of therapeutic fusion toxins. In this instance, replacement of the native diphtheria toxin receptor-binding domain with granulocyte-macrophage colony-stimulating factor (GM-CSF) targets the fusion protein toxin toward the GM-CSF receptor on acute myeloid leukemia (AML) blasts. While this genetic construct is highly toxic toward AML cells in vitro, the normal distribution of the GM-CSF receptor in vivo is sufficiently broad that a second layer of specificity was engineered into the fusion protein toxin.

The diphtheria toxin–related fusion protein toxins follow a highly defined route of entry into the cell.⁴  The intoxication process begins with receptor binding; the fusion protein toxin is then internalized into the cell by receptor-mediated endocytosis. In the lumen of the early endosome, the endoproteinase furin introduces a “nick” in the α-carbon backbone in a protease-sensitive loop that connects the catalytic (C) and transmembrane (T) domains. This nicking event is essential for the subsequent membrane translocation and release of the C-domain into the cytosol of target cells. By replacement of the furin recognition site (RVRRSV) with the urokinase plasminogen activator (uPA) cleavage site (GSGRSA), Abi-Habib and colleagues have constructed a “pro-fusion protein toxin.” Their study clearly shows that the targeted action of the modified construct, DTU2GMCSF, strongly correlates with the combined presence of the uPA receptor, uPA, and the GM-CSF receptor.

The development of new fusion protein toxins as potential therapeutic agents is still in its infancy. The challenge remains to construct biologics that retain a high level of specificity toward cell surface determinants that are largely restricted to malignant cells in order to minimize damage to normal cells. Since uPA is often up-regulated in particular malignancies, the introduction of the uPA cleavage site into DTU2GMCSF confers a second level of specificity and shifts the balance of action toward the malignant cell population. We await with interest the initiation of and findings from clinical studies with this pro-fusion protein toxin as we anticipate that the coupling of cell-specific drug delivery with site-specific conversion from protoxin to toxin will greatly expand the utility and therapeutic margin of these agents. The DTU2GMCSF described by these authors offers a promising, novel, therapeutic approach for patients with AML resistant to standard treatment. ▪