By David N. Leff

Like a heavily-armed squadron of paratroopers that drops down to fight off an invading enemy, in advance of defensive ground forces, the immune system deploys interferon, well ahead of defensive antibodies, to confront and destroy viruses before they can infect a human cell or body.

As research virologist Peter Palese, at New York¿s Mt. Sinai School of Medicine, puts this rule of engagement: ¿The normal idea is that all through evolution, a virus infects a cell, or host organism. That angers the host¿s immune system, so it makes interferon [IFN] as an antiviral response. Then, if the virus has no way of overcoming that interferon, it will lose the engagement, it will be inhibited.¿

From this basic encounter Palese and his co-senior author, Adolfo Garcia-Sastre, have evolved an antiviral vaccine concept reported in the current Proceedings of the National Academy of Sciences (PNAS; Early Edition, Vol. 1, No. 9), released Mar. 21, 2000. Its title: ¿Influenza A and B viruses expressing altered NS1 proteins: A vaccine approach.¿

Their paper makes the point that ¿influenza virus is a formidable pathogen, killing thousands of people a year in the U.S. alone.¿ But it adds that their strategy ¿may be applicable to the rational generation of vaccines for other viruses with defined IFN antagonists.¿

¿Now many viruses ¿ and that¿s sort of the new concept here ¿ particularly smaller RNA viruses such as the influenzas,¿ Palese told BioWorld Today, ¿have not been known to have an anti-interferon antagonist activity. It was not known for the influenza virus. And our finding that it does is new. We have identified this viral NS1 protein, which has interferon antagonist activity.

¿We have previously shown,¿ he went on, ¿that a virus lacking NS1 ¿ that anti-interferon protein ¿ can not grow in normal mice or normal cells. But if the interferon response is knocked out in a transgenic mouse, that virus is perfectly happy replicating, and killing its mammalian host. That NS1 protein is not needed in an interferon-minus mouse, or an INF-minus cell.

¿So that¿s one extreme,¿ Palese observed. ¿An influenza virus that does not have an interferon antagonist will be completely attenuated, weak, unable to replicate in a normal host that makes interferon, but can grow in an interferon-minus host. In a normal influenza infection, the virus with a functional NS1 protein in its genome will overcome the host¿s interferon response ¿ either completely or partially.¿

Engineering A Happy Medium

Palese and his co-authors have modified the NS1 protein by genetic engineering techniques, and have found, he believes, ¿a happy intermediate, where the virus can overcome a little bit of the interferon response, but can¿t kill the host. That virus is not very virulent, not very pathogenic. On the other hand it can replicate to titers high enough to induce an immune response that will be protective against the infection.¿

To generate their middle-way flu virus, the team cut away just over half of the NS1 protein¿s 230-amino-acid chain. The first 99 residues comprised a deletion mutant containing the active site for binding the virus to its host¿s double-stranded RNA. Serendipity dictated this happy discovery: ¿We did not want that virus,¿ Palese recalled. ¿It just happened that we got it, and we were pretty unhappy until we realized it was an interesting species. We showed that this mutant virus is highly attenuated ¿ at least 1,000 - to 10,000-fold ¿ but it induced a very good protective immune response.¿

To demonstrate this effect, he recounted, ¿We put the virus up the noses of mice. The animals got infected, but because the virus is attenuated, weakened, not very pathogenic, not very virulent, they didn¿t get sick. And four weeks later, we challenged them with a virulent virus. Those that had never seen the immunizing pathogen died, whereas immunized mice were protected.

¿Unlike many flu vaccines, Palese noted, ¿this one generated both humoral and cellular immune responses. But what is also important is that it induced a mucosal immunity in the respiratory tract, site of viral activity. This would be an added benefit to current flu vaccine administration.¿

He emphasized, ¿This is a laboratory finding, using mice. Clearly, additional experiments would be needed to make it advisable to put it into people.¿

Ferreting Out New Flu Virus Models

Looking toward further preclinical trials, higher up the mammalian ladder, Palese observed ¿A primate model is not a very good one for influenza. In general, with HIV vaccines, obviously one has to go into primates, particularly because there is no other good animal model. With influenza there are ¿ for instance, ferrets.

¿In preliminary experiments here, we have indications that influenza B viruses work the same way as influenza A. Namely, B also has an NS1 protein. Even though there is no sequence identity between them, B has similar anti-IFN functions. So that has to be explored, because any influenza vaccine for humans would clearly require that both A and B viruses be in that formulation.¿

(Influenza A is somewhat more virulent than B. It is more apt to change its antigenic spots by mutating, and gives rise to more lethal flu outbreaks.)

¿The university has filed patent applications,¿ Palese observed, ¿claiming the making of vaccines ¿ in the case of influenza ¿ modulating NS1, or more generally, the anti-interferon viral response, not only for flu but for other viruses as well. Clearly, we are interested in identifying viruses that have unsuspected anti-interferon activity.

¿And you could make the other argument that if one has a super anti-interferon protein, which cuts down the host interferon response completely, then you could see that that virus would win ¿ very rapidly.¿

Meanwhile, he concluded: ¿We would welcome expressions of interest in further developing this approach for vaccinating humans.¿