Researchers at Stanford University and the University of California at San Francisco have developed a phosphospecific flow cytometry technique to profile signaling at the single-cell level, including molecules downstream of the granulocyte-macrophage colony-stimulating factor receptor and molecules closely associated with Ras signaling.
The cell populations will be evaluated for the presence of primary juvenile myelomonocytic leukemia, or JMML, cells with altered signaling behavior correlated with disease physiology.
The researchers hope that the technique can be refined and used to screen potential JMML therapies. Unlike other leukemias, the only treatment for JMML is a bone marrow transplant.
The technique, which was featured as the cover story in the Oct. 7 issue of Cancer Cell, involves permeabilizing the cell to allow antibodies to enter it and bind to signaling molecules within the cell, in this case, STAT5. JMML cells tend to proliferate in the presence of low concentrations of GM-CSF, whereas normal cells only respond to higher concentrations. GM-CSF activates the JAK-STAT pathway.
Although the JAK-STAT pathway had not previously been implicated in JMML, the researchers used an antibody that binds only to activated STAT5 to determine whether the cells of 12 patients with JMML displayed abnormally high levels of STAT5 in response to low doses of GM-CSF. Samples from 11 of the 12 patients did so, pointing toward the involvement of the STAT pathway in JMML.
In contrast to the JMML patient samples, seven out of eight normal bone marrow samples, as well as all eight bone marrow samples from patients with other, similar myeloproliferative disorders, maintained normal levels of activated STAT5 after receiving low doses of GM-CSF.
Mignon Loh, an associate professor of clinical pediatrics and hematology/oncology at UCSF and a corresponding author on the paper, spoke with CBA News this week about the work and its potential for use as a drug discovery tool.
Please give me a little background on this work.
Our collaborator, Garry Nolan [at Stanford], has developed a very nice addition to traditional flow cytometry. Flow cytometry is mostly relegated to looking at surface markers on cells. But they developed a methodology where you could poke holes in the cell membrane using a variety of different agents. We happened to use methanol for our purposes, but you could also use saponin or ethanol.
You can add phosphoepitopes that can seep through the holes and actually get a nice readout of certain molecules that may be activated by their phosphorylation status. You can still retain the surface markers, as well as get an idea of how the subsets of cells are signaling by looking at phosphoepitopes inside the cells, and assess what specific subsets of cells are doing.
How exactly does the assay work?
We take a patient sample and we isolate the mononuclear cells. We then expose those cells to stimuli, such as cytokines, which we did for this paper because JMML cells are hypersensitive to the cytokine GM-CSF. We then stopped the reaction at 15 minutes.
People can do time courses and can also expose cells to inhibitors, and look at the downstream effects of those targeted agents. For this report, the majority of the work that we did involved exposing the mononuclear cells to GM-CSF at increasing concentrations, and we had appropriate positive and negative controls for each of our runs.
We then stopped the reaction, fixed the cells with paraformaldehyde, and permeabilized the cells with methanol. We then stained the cells with surface antibodies and phosphoepitopes, after we rehydrated the cells.
We did find that one important step was that we had to rehydrate the cells in order to get the cell surface staining and the epitopes to work very well. We then added a cocktail of antibodies that look at cell surface markers and intracellular molecules, did the appropriate washes, prepared the cells, and then ran them on an LSR II flow cytometer equipped with 433 nm and 633 nm lasers. The LSR II is capable of looking at as many as 14 different colors. I think we did five or six colors at a time.
Then you just get a readout, and look at what the cells’ signals are as they pass through the lasers, to see what fluorochrome-conjugated antibodies they are picking up. You can get a nice readout of what kinds of cells are hyperphosphorylating STAT5. In our particular report, it was CD 14 positive, CD 33 positive, CD 38 dim, and CD 34 negative that were phosphorylating STAT5 in response to low doses of GM-CSF.
How does this pertain to compound-library screening?
In the paper, there is a figure that describes how we demonstrated the specificity of the JAK-STAT pathway by first exposing the cells to a chemical JAK2 inhibitor that you can buy off the shelf from Calbiochem, and we were also able to obtain a targeted inhibitor of the JAK-STAT pathway, XL019, from Exelixis, that is in clinical trials right now for adults with polycythemia vera. Patients with this disorder have been found to have a high incidence of JAK2-activating mutations, so they are being treated with a JAK2 inhibitor.
When you stick the JAK2 inhibitor on our primary JMML cells, you can actually abrogate the p-STAT5 response. So one way that you could look at the specificity of targeted agents is if you have a readout [that] works well on this particular assay, you can potentially look in patient cells before they get the drugs, to see if they will respond. So that is one potential application.
What about preclinically screening candidate compounds?
It depends on the system that you have. You could probably do that in cell lines very easily. We also have a lot of experience in Kevin Shannon’s lab [at UCSF], who I work very closely with, using accurate murine models of these diseases. So we either give the drugs to the mouse, or we give them ex vivo to the mouse bone marrow that is affected, and see if there is signaling.
But I cannot imagine why you could not do a larger drug screening assay with cell lines. Cell lines are a little difficult to work with, because they show great inhibition or great response, or they are this or that, because they are a little bit “tricked up.” They are maintained outside of the human body and they mutate a little bit, and they have adapted to their ex vivo state. They can, however, still be pretty powerful.
Is this something you plan to license or commercialize?
Yes, we have some documents drawn up between Stanford and UCSF, and we do have something registered with the Stanford and UCSF office of technology.
What do you see as the next step in this work?
For these patients, I am hoping that we can also demonstrate that the signaling phenotype is abrogated by the inhibition of the JAK2 inhibitor in stem or progenitor cells. Then I think we have some good evidence that we should try this particular compound in patients with JMML. Whereas before, I am not sure that we would have selected a JAK2 inhibitor to use in these patients. That is one application.
The other application is that we are really hoping to bring this test into a CLIA-approved environment, so that we can help use it as a diagnostic test for patients with this disorder.