NEW YORK -- Scientists have produced rudimentary mouse sperm from stem cells in the laboratory, a step that may lead to a treatment for infertile men.

If the technique pans out in people, doctors might someday be able to turn skin cells from a man into sperm that can pass along his DNA to his offspring.

But experts warned that the technology would have to overcome some hurdles first.

The mouse work is reported by Chinese scientists in an article released Thursday by the journal Cell Stem Cell. The technique is now being tested in monkeys, a senior author said.

Qi Zhou of the Chinese Academy of Sciences in Beijing added in an email that the research has "a long way to go" before it could be used for infertile men.

The scientists began with mouse embryonic stem cells, which are found in embryos and can develop into any kind of cell in the body. In the lab these cells were exposed to chemicals to nudge them toward becoming sperm. While previous research has also generated sperm precursors in this way, these precursors then had to be transplanted into the testicles of mice to develop further.

The new technique is an advance, experts said, because it eliminates the need for transplants. Instead, the researchers put the sperm precursors in a lab dish containing testicle cells.

Although the precursors never became fully mature sperm, they developed far enough to fertilize eggs. They were injected into 379 eggs; nine baby mice resulted.

The results are convincing, said John Schimenti of Cornell University in Ithaca, New York, who didn't participate in the work. He said he thinks the major payoff will come in basic research into sperm development.

But he and another expert, Renee Reijo Pera of Montana State University, said they thought the technique might someday be adapted to more directly help some infertile men.

In that case, the process would begin with skin cells or other cells from the patient, which could be converted into stem cells bearing the patient's DNA. These stem cells would then be put through the process.

But there would be challenges in applying the mouse technique to people, such as finding an alternative to using testicle cells from newborns, Reijo Pera said.

She also said the mouse results would have to be confirmed by other labs first, a standard caveat for research. And given the practical limits on availability of eggs from women, the technique would have to be made more efficient in producing births, she said.