A new approach to antibody-based therapies might one day allow clinicians to target all strains of Ebola with a broad-spectrum drug.

In early research, the "Trojan horse" antibodies were effective in preventing all five varieties of Ebola from infecting human cells, according to Kartik Chandran, PhD, of Albert Einstein College of Medicine in New York City, and colleagues.

And the approach protected mice from a lethal dose of the Zaire Ebola virus -- the variety responsible for the recent African outbreak, Chandran and colleagues reported in Science.

Much work remains to establish the clinical utility of the Trojan horse antibodies in humans, Chandran told MedPage Today, including further work in rodent models and non-human primates before human testing.

But at least in principle the work overcomes an important barrier -- the need to develop a drug that targets each strain separately. The most promising therapy currently is the three-antibody cocktail dubbed ZMapp, which has been shown to reverse the course of Ebola infection in macaques and is under study in humans.

But ZMapp is specific for the Zaire strain of Ebola and a broad-spectrum drug would be an important step forward, Chandran said.

He added that the Trojan horse approach might also be used against other filoviruses, such as Marburg and Lassa fever, as well as possibly some of the flaviviruses, such as Zika and chikungunya.

The research "looks really interesting," commented Thomas Geisbert, PhD, of the University of Texas Medical Branch at Galveston.

"But there have been so many things that have protected mice against Ebola that don't work in monkeys," he told MedPage Today by email. "So unless there is any non-human primate data, it is hard to really say much at this point other than it is an interesting approach."

Geisbert and colleagues are working on treatments, vaccines, and animal models for hemorrhagic fevers, such as Ebola and Marburg.

The key cellular step in infection with Ebola and other filoviruses, Chandran and colleagues noted, is a viral interaction with a receptor deep inside the cell.

That interaction -- between the viral glycoprotein and the Neimann-Pick C1 receptor (NPC1) -- is "shielded" from most antibodies because it takes place in late endosomes, membrane-bound vesicles in the cellular cytoplasm, the researchers noted.

To overcome that shielding, Chandran's group conceived of a bi-specific antibody approach -- a delivery antibody that binds to a region of the viral glycoprotein that is common to all five Ebola strains and an associated antibody that blocks the important interaction between the virus and NPC1.

An unusual feature of infection by all filoviruses, they noted, is that the glycoprotein undergoes cleavage within the endosome to reveal "cryptic epitopes" that bind to NPC1, offering two choices to prevent the interaction: block either NPC1 or the newly revealed viral proteins.

Importantly, neither option is available until the virus is in the cell and has undergone the cleavage. But the researchers had in hand a human antibody, MR72, that binds to the cleaved glycoprotein's NPC1 receptor binding site, while another, mAb-548, blocks human NPC1 itself.

To deliver the antibodies, the researchers used an antibody dubbed FVM09, which binds to a "cap" on the glycoproteins of all the Ebola varieties -- a cap that is later stripped off in the endosomes.

In essence, he said, the FVM089 antibody binds to the virus and tricks it into carrying the seeds of its own destruction into the cell -- just as the Trojans carried the mythical wooden horse stuffed with Greek soldiers into their city and behind their defenses.

In test-tube experiments with human cells, both complexes -- FVM09-MR72 and FVM09-mAb-548 -- prevented infection by all Ebola strains, the researchers reported.

In murine cells, FVM09-MR72 was also highly effective, while FVM09- mAb-548 was less so, largely because it would not bind properly to the mouse version of NPC1, they noted.

Paralleling that finding, they found that in mice given a lethal dose of Zaire Ebola, FVM09-MR72 prevented the death of 70% of the animals, with no significant benefit for FVM09-mAb-548.

Nevertheless, Chandran said he expects both will work well in non-human primates, whose NPC1 is identical with the human version, and eventually in humans.

But he added wryly, the experiment still has to be done: "My expectations and a subway token will get you a ride on the subway."

The next step, he said, is to make enough of the antibody complexes to move forward into more expensive experiments in non-human primates.