By David N. Leff

Bone marrow (osso buco) and calf's brains (cervelles) are gourmet delicacies of Italian and French cuisine, respectively. Americans may fancy them as acquired tastes.

The biological nexus between bone and brain seems distant and unnatural, but two research papers in last week's issue of Science, dated Dec. 1, 2000, span that apparent gap. Their titles tell why and how:

"From marrow to brain: Expression of neuronal phenotypes in adult mice," report scientists at Stanford University. And researchers at NINDS, the National Institute of Neurological Disorders and Stroke, weigh in with: "Turning blood into brain: Cells bearing neuronal antigens generated in vivo from bone marrow."

Long bones, which serve as scaffolding for the mammalian body, are semi-hollow shafts, stuffed with constantly regenerating blood cells - the process of hematopoiesis. Topping the list of these bone-created cells are erythrocytes, the red blood cells that freight oxygen and nutrients all over the body. Then there is the numerous family of immune-system cells - the T lymphocytes and B cells that produce antibodies and killer cells - along with macrophages, mast cells, granulocytes, megakaryocytes and others.

This multiservice army of blood cells is commanded, indeed recruited, by a single general - the bone marrow stem cell, whose commission was traditionally limited to hematopoiesis. "Until a few years ago," stated a related Science commentary, "scientists did not think mammals produced any new neurons at all after childhood - much less that foreign bone marrow cells could be coaxed into such a feat." Previous research had succeeded, under certain culture conditions, in morphing bone marrow stem cell progeny into neuron-resembling cells, but only in culture. Now the two back-to-back Science papers are the first to show that this process can be made to happen in living animals.

Can Bone Marrow Stem Cells Really Go Neuronal?

Senior author of the Stanford paper is gene therapist Helen Blau, who chairs the university's Department of Molecular Pharmacology. Predoctoral cell biologist Timothy Brazelton is its lead author. "Our biggest finding," he told BioWorld Today, "was simply that cells derived from outside the central nervous system of brain and spinal cord could be transformed into cells of a neuronal fate. Both our paper and the one from NINDS reported that this could occur in vitro. Both showed that bone marrow cells could be induced to express neuron-specific genes.

"For us," Brazelton added, "the most unexpected finding was simply that this plasticity existed. Our other big surprise was that these cells from the bone marrow could differentiate into a neuronal phenotype. Originally we were looking for the participation of bone marrow-derived astrocytes - scarring of the brain after injury. And actually we're still working on that. But in the meantime we noticed that some cells, with their green-glowing bone-marrow markers, looked surprisingly like neurons, and that's what we reported for the Science paper."

He and his co-authors enlisted transgenic mice that expressed the green fluorescence protein (GFP), a standard cell marker. The team then lethally irradiated these animals, to destroy most of their own bone marrow, then injected transplanted unfractionated murine marrow via their tail veins.

"Four weeks after transplantation," Brazelton recounted, "we ended up with mice that had only the hematopoietic compartment labeled with fluorescent protein. Then we evaluated their brains for the appearance of neurons expressing GFP. We used confocal microscopy, a method that allows us to take very thin optical sections of tissues to absolutely confirm that our brain markers as well as neuron-specific proteins are glowing in one and the same cell."

What they saw in those brain sections were hundreds of marrow-derived cells that expressed two types of gene products typical of neurons - NeuN, 200-kilodalton neurofilament, and class III b-tubulin. "NeuN," Brazelton explained, "is a nuclear protein found in many neurons and it has cytoplasmic distribution in a subset of neurons. The tubulins are expressed in the neuronal cytoskeleton. The primary result," he summed up, "was simply that there were very distinct cells with neuron-like morphology, expressing neuron-specific proteins and the GFP gene, indicating that they had been derived from the bone marrow compartment. We used antibodies to identify our cytoskeletal proteins, and these antibodies essentially lit up all the neurons in our tissue sections.

"The fact that this biological process exists," Brazelton suggested, "means that it may be tweakable. That is, we may be able to modify it for therapeutic purposes. It could have a huge impact potentially on the treatment of all kinds of brain injuries - trauma, stroke - as well as genetic disorders such as Huntington's disease or Parkinson's disease [PD].

"Alzheimer's is a little more complicated," he went on, "because we don't really understand its etiology. In contrast, for example, in PD we know there's a distinct loss of a very specific neuronal phenotype - dopaminergic neurons. Fetal transplants can replace that lost cell with another cell that has a similar function. And it has actually alleviated some PD symptoms of the disease in many patients. We hope that our cells would have the ability to do something similar - with fewer drawbacks."

Stanford, NINDS Research Dovetails

Brazelton compared the companion NINDS paper with Stanford's. "There are two important differences," he commented, "but I think theirs complements ours. They actually had a very similar approach in general, in the sense that they did a bone arrow transplant, then identified bone marrow-derived cells expressing neuronal phenotypes.

"The difference between our two papers," he continued, "is that they used an adult donor and neonatal recipient. We used adult donors and adult recipients. This is important because it suggests the process could occur throughout adult life.

"The thrust of our research," Brazelton pointed out, "is that we've identified a new game - a new process that occurs - but we don't understand the rules of that game, nor any of the intermediate steps in this process. And what we really want to deal with is being able to understand all the single rules that these cells are following to make this cellular phase transformation.

"At some point," he concluded, "we'd love to embark on efforts to produce a therapeutic product. But that's really several years away."