By Kate Kelland, Health and Science Correspondent

SEATTLE (Reuters) - He keeps them in warm, comfortable bug dorms, feeds them on meals of human blood with the occasional sugar water snack and lives in awe of their killing power.

Seattle-based research scientist Stefan Kappe says mosquitoes are the most dangerous animals in the world.

Which is probably why when his laboratory colleagues slice their heads off with miniscule needle-like scalpels and squeeze them with tweezers to extract early forms of the malaria parasite from their saliva glands, he feels no concern about cruelty to animals.

Kappe has spent his working life trying to figure out how this tiny malaria-carrying insect can inflict so much death and disease on humans, and what he and his team can do to stop it.

According to the World Health Organisation, malaria kills a child every 45 seconds in Africa and costs that continent's economy $12 billion a year.

FORMIDABLE PREDATORS

Kappe, molecular biologist and expert in parasitology who trained first in Germany then the United States, has no doubt the killer parasitic disease will one day be wiped out across the world, but acknowledges it's a tough fight.

"They are formidable little predators," he says as he looks through the mesh window on one of the mosquito bug dorms at his

Seattle BioMed laboratory and insectary. A handwritten sticker on this dorm says "fed."

"They are uniquely adapted to take blood meals, and unfortunately infectious diseases have taken a ride along with this ability of the mosquito to bite you and take your blood," Kappe says.

Until this week, some of the world's best scientific minds had failed to make an effective vaccine against malaria -- or any parasitic disease for that matter -- and Kappe says eradication can't be achieved without one.

RTS,S was recognized as the first effective malaria vaccine on Tuesday when scientists released data showing it halved the risk of children getting the disease in a large Africa trial.

"Right now malaria vaccine development stands at a very interesting point because we have a partially effective vaccine in RTS,S," Kappe told Reuters.

Experts stressed that RTS,S -- developed by the British drugmaker GlaxoSmithKline and the non-profit PATH Malaria Vaccine Initiative -- will be no quick fix.

At around a 50 percent protection rate, the new shot is less effective against malaria than other vaccines are against common infections like polio and measles.

"The RTS,S vaccine will always stand as the first really successful vaccine that can partially protect against malaria," Kappe said. "But to eradicate the disease -- and that is our goal -- you need a vaccine that protects 90 to 100 percent. So we have to build on RTS,S."

To do that Kappe's team is taking various routes -- most of which involve breeding large numbers of these dangerous animals in warm, soupy trays in what he calls the "swamp room."

After dissecting them, modifying them, breeding more generations and then allowing them to drink malaria-infected blood from a skin-like covered cup, he sets them on brave human trial volunteers who agree to be bitten in the name of science.

Seattle BioMed is a non-profit research institute that works on research to eliminate the world's most devastating infectious diseases, funded by the U.S. National Institutes of Health, the Bill and Melinda Gates Foundation and around 500 other donors.

One of the institute's approaches to creating a vaccine centers around immature forms of the malaria parasite called sporozoites, which are carried in the saliva glands of female malarial mosquitoes and transferred into humans when they bite.

The process of infection with malaria takes a complex path, starting in the human victim's blood and moving into the liver.

Inside the liver, the sporozoites change form and then grow and divide into thousands of merozoites. These in turn burst out from the liver cells and back into the blood.

Once back in the blood, the merozoites multiply in red blood cells, again burst out and produce more parasites, eventually damaging the brain and lungs, causing fever, chills, anemia and, in severe cases, death.

DELETING THE PARASITE'S GENES

Kappe's team is seeking to interrupt this process at a critical stage and has found a way of genetically modifying the sporozoites to delete key genes from their DNA, so that while they still make it into the liver where they trigger a strong immune response, they are also genetically programed to die off there.

"What we're interested in is preventing the liver-stage parasite from completing its development," explains Ashley Vaughan, a molecular geneticist working with Kappe.

"If you have enough sporozoites going to the liver and stopping there, they will alert your immune system that your liver is seeing a large amount of malaria, which would then generate a protective response.

"So if you then get bitten by a mosquito carrying natural malaria, the parasite would go to your liver, that same response would be triggered and your immune system would kill it.

"This would mean you'd never get a blood-stage infection, and never get sick."

In tests on mice, the so-called genetically attenuated whole parasite (GAP) experimental vaccine has proven 100 percent protective, 100 percent of the time, Kappe and Vaughan said.

And in the first early-stage human trials, where six volunteers agreed to be bitten first by a "vaccine mosquito" carrying genetically modified parasites and then by one with natural malaria, five out of six were protected.

Kappe is worried by the sixth case -- where the trial volunteer went on to develop malaria caused by the parasites in the vaccine failing to stop developing at the right stage.

In the trial, the volunteer was of course immediately treated and cured with anti-malarial drugs, but for the GAP experimental shot to be developed any further down the path to a potentially useful product, the team still has much work to do.

"What we have to do now is learn how to make it safer, and learn how we would be able to manufacture it on a larger scale," said Kappe.

For the moment the manufacturing process is very hands on, and a little gruesome.

Working with microscopes in a laboratory next to the "swamp room," scientists Heather Kain and Will Betz take each mosquito at a time, soak it in an ethanol solution, slice its head off, squeeze its thorax to get the saliva glands out, and then cut open each gland to harvest the sporozoites.

For every potential vaccine dose, the researchers need around 10,000 sporozoites, and all those and more can come from a single mosquito. As Kappe says, "it's hard to imagine making millions of doses" by hand.

"On a good day I can dissect around 200 mosquitoes an hour," says Betz. "But it takes a steady hand."