Laboratory tests found that the virus targeted key cells involved in brain development and then destroyed or disabled them, they said.
The findings are the first concrete evidence of a link between the mosquito-borne virus and microcephaly, which until now had been circumstantial, said Guo-li Ming, a professor of neurology at Johns Hopkins' Institute for Cell Engineering, and a co-leader of the research.
"Studies of foetuses and babies with the telltale small brains and heads of microcephaly in Zika-affected areas have found abnormalities in the cortex, and Zika virus has been found in the foetal tissue," he said in a statement.
The first -- known as human neural progenitor cells (hNPCs) -- is crucial for the development of the cortex, or outer layer, of foetal brains.
Also Read
Damage to these cells, which eventually differentiate into mature neurons, would be consistent with the brain defects caused by microcephaly.
The other two types of cells were stem cells and neurons.
As predicted, Zika virus attacked the human neural progenitor cells. Within three days of exposure, 90 percent were infected, and nearly a third had died.
Furthermore, the genes needed to fight viruses failed to activate, which was a highly unusual outcome.
By comparison, the other two types of human cells were relatively unharmed.
"Our results clearly demonstrate that Zika can directly infect hNPCs in vitro with high efficiency," the study concluded.
"It is very telling that the cells that form the cortex are potentially susceptible to the virus," Ming added.
The findings, published in the journal Cell Stem Cell, may help to identify drugs that protect these vulnerable cells or reduce infections after they occur.
By itself, Zika is typically no more threatening than a bad cold or a mild case of the flu. Sometimes there are no symptoms at all.