The researchers, from both the US and Europe, say they managed to genetically modify the Lactococcus lactis bacteria so that it produces cyanovirin - a drug that has blocked HIV infection during tests on monkeys and is set for human trials next year.
L. lactis produces lactic acid and so is used widely to make cheese and yoghurt. Some probiotic dairy products, mainly yoghurts, which claim to benefit gut health also contain different strains of the bacteria.
Bharat Ramratnam, one of the researchers and an HIV expert at Brown Medical School, told sister site, DairyReporter.com, that the team had definitively shown that L. lactis, "can be used as a vehicle for drug delivery."
This, he said, meant that strains of so-called 'friendly bacteria' could be used as transport in the fight against certain other infectious diseases. "There are a lot of molecules with anti-viral properties, but there is currently no way of delivering them to the human body."
However, the team only focused on how to block the sexual transmission of HIV, which means that the live bacteria would have to be capable of moving from the gut down to the vagina.
L. lactis is already found naturally in both the human gut and vaginal wall, suggesting that any strain introduced would have a good chance of survival.
And, Ramratnam said there was also evidence the bacteria could move from one to the other, but: "there are hundreds of strains and not much is known about any of these. All of this depends on choosing the right strain."
The researchers emphasised that, even if HIV-blocking foods were a little way off, L. lactis containing cyanovirin may still be manufactured as a drug.
Either method could be a much cheaper way of combating the fast-spreading HIV virus. The World Health Organisation estimated that 40m adults and children were living with HIV in 2005, with 4.9m new infections during the year.
Ramratnam said the team first needed to test and perfect their use of L. lactis as a delivery vehicle for cyanovirin.
He said the researchers needed to look at whether the L. lactis would still work in the same way after being genetically modified. Secondly, if it does, they must assess whether the bacteria will send out enough of the cyanovirin compound to halt HIV infection.
Lastly, but not least, Ramratnam said the team needed to be sure that cyanovirin was safe. "No one has ever applied this to animals for months at a time," he said, adding the testing may not be complete for two or three years.
Initial tests have shown that cyanovirin blocks a receptor the HIV uses to infect cells by binding itself to sugar molecules attached to the HIV virus.