Los Angeles ? Malaria kills nearly 1 million people a year, but it has a weakness — to infect humans, it needs mosquitoes. In a potential step toward eradicating the disease, researchers reported this week that they have developed a genetically engineered breed of mosquito that cannot be infected by the malaria-causing parasite.

Genetically modified mosquitoes are far from ready for use in the field, but the researchers achieved an unprecedented 100 percent blockage of the Plasmodium parasite, highlighting the promise of this approach.

The team, led by entomologist Michael Riehle at the University of Arizona, created the mosquitoes by changing a single gene, one involved in the production of insulin. To test the effect of that change, researchers injected 90 of the mosquitoes with the malaria parasite. Ten days later, at a point when normal mosquitoes would have bellies full of parasites, they didn’t find a single one.

This is the first instance of a genetic modification that completely blocked development of a malaria parasite that can infect humans. The research was reported online Thursday in the journal PLoS Pathogens.

Total blockage is not only impressive — it’s biologically crucial. Other groups have achieved 90 percent to 95 percent blockage by combining multiple genetic alterations, but such less-than-perfect protection could allow the parasite to evolve around the mosquito’s blocking mechanism, in much the same way that drug-resistant bacteria can arise if antibiotic treatments aren’t completed.

The researchers used a slightly modified version of a gene that influences the mosquito’s life span, as well as its immune response. By targeting this gene, they were aiming to make insects that died young so that the malaria parasite didn’t have the 16 days in the mosquito gut it needs to mature.

It worked — the mosquitoes on average lived about 20 percent shorter lives than normal mosquitoes.