Neurons developing out ot Mesoderm?!

Snuck into the very end of this, otherwise very interesting article on neurobiology of cephalopods and moths, is this little passage:

As for flies, Tublitz outlined a tantalizing question, as yet unanswered, that has continued to take flight out of his lab for the last decade. Scientists for years, he said, have held "one hard rule" about what constitutes a neuron -- that a neuron cell always arises from the ectoderm of a developing embryo. However, a discovery in Drosophila -- fruit flies -- has softened that assumption.

Cells arising from the mesoderm rest in a layer on top of the fruit fly's nervous system, Tublitz explained. "These cells have all of the properties of nerve cells." A slide shown during his talk displayed a long list of characteristics most often applied, with only few exceptions, to neurons. "Are these mesodermal cells nerve cells? I can't answer that question conclusively, but we have generated data that suggest the answer may be 'yes'."

More like this

My (very disjointed) knowledge of embryology and ESCs tells me that the neural lineage is pretty much the default program of differentiation. So unless you go out of your way to make something different, neurons is all you are going to get from early embryonic cells in vitro. This is mammals of course. But unless I am totally confused here, it shouldn't be too surprising that under the normal circumstances invertebrates make neurones out of non-ectodermal cells.