NEW YORK (GenomeWeb) – Researchers from the Moffitt Cancer Center and the University of Colorado have developed a map of anaplastic lymphoma kinase signaling in lung cancer and used it to identify proteins that could further sensitize tumor cells to ALK inhibition.
Detailed in a paper published this week in Science Signaling, the study identified 464 proteins linked to ALK signaling and 64 and 9 proteins that when knocked down by small hairpin RNA sensitized tumor cells to the ALK inhibitors crizotinib and alectinib, respectively.
More generally, the study provides the field with a large amount of proteomic data on protein networks involved in the ALK rearrangement in lung cancer and a number of potential targets for follow up, Eric Haura, a Moffitt researcher and senior author on the paper, told GenomeWeb.
Around 5 percent of lung cancer patients have the EML4-ALK rearrangement, in which the ALK gene is fused to the gene for echinoderm microtubule-associated protein-like 4. A proportion of these patients benefit from treatment with ALK inhibitors, but essentially all go on to develop resistance to these drugs.
A better understanding of ALK signaling could help researchers identify ways to combat this resistance, Haura noted. By targeting other proteins involved in ALK function, researchers might further sensitize cells to treatment with ALK inhibitors, improving their performance in patients who have responded to these drugs and perhaps extending the benefit to patients with the EML4-ALK fusion who have not seen significant benefit from such treatment.
Studying the EML4-ALK rearrangement in the non-small cell lung cancer H3122 cell line, the researchers generated a map of ALK signaling using a combination of mass spec techniques. First, they tagged the EML4-ALK protein and then used affinity purification mass spectrometry to pull down this protein and identify its direct interactors. They also used three known interactors of EML4-ALK — SHC1, GRB2, and PIK3R2 — as additional bait proteins, probing H3122 cells with these molecules and then identifying their interactors via mass spec. The EML4-ALK, SHC1, GRB2, and PIK3R2 pulldowns identified, respectively, 84, 96, 64, and 62 unique proteins from the H3122 cells.
Haura and his colleagues then followed up these affinity purification mass spec experiments with a broader tyrosine phosphoproteomic analysis looking at H3122 cells with and without exposure to crizotinib to identify proteins involved in ALK signaling but with less direct interaction with that system's core nodes.
Combining this data with existing literature on interactions between the identified proteins, the researchers generated a network consisting of 464 proteins and 4,443 interactions. They then used shRNA to knock down 407 proteins in this network, one at a time, investigating which proteins, when knocked down, sensitized H3122 cells to ALK inhibitor treatment.
Such an approach was necessary given the complexity of protein signaling, Haura said, noting that "there is a lot of systems biology and network biology in action that you may not be able to predict."
"Maybe if you understood the intricate mathematics of each protein in the network and its binding characteristics and the phosphorylation reactions you could create a mathematical model of this, but I just don't think the data is there to do that," he said. Better then, "to just go ahead and knock out each individual protein one-by-one and ask who ends up being a sensitizer."
The data generated in the study could help future researchers better model the system, Haura said. "I think somebody could now say, 'OK, if I know the landscape of the signaling network surrounding the ALK rearrangement, and I know which of these nodes is important in the signal transduction or survival, I might now be able to create a simplified model and produce some insights into why those nodes are important and what that might mean for creating new sensitizers.'"
One limitation of the study, Haura noted, is that the researchers looked at the H3122 cells at one point in inhibitor treatment, meaning they would not necessarily observe signaling changes and resistance mechanisms that could emerge over the course of longer-term treatment.
They did analyze ALK rearranged cells that had developed resistance and found that the proteins that sensitized the non-resistant cells to inhibitor treatment also sensitized the resistant cells, suggesting that these sensitizing targets might remain effective across cells with various forms of ALK inhibitor resistance.
"It may be that the cells retain the essential network that they need to transduce their signals, and [the sensitizing proteins] may be something that is reproducible despite different forms of resistance," Haura said, though he added that he and his colleagues had not yet established that this is, in fact, the case.
Looking ahead, Haura said that he and his colleagues are now working with outside collaborators to look more in depth at the mechanisms of resistance to ALK inhibition.
He said they also plan to design proximity ligation assays in order to visualize and quantify protein complexes involved in ALK signaling in disease tissues, with the aim of potentially developing diagnostics.
"We are very interested in trying to use what we have learned through this mapping exercise to create novel diagnostics [looking at] ALK and other receptor tyrosine kinases," he said.
Another aim they hope to pursue based on the Science Signaling work is the development of drugs that interfere with protein-protein interactions and disrupt protein complexes, Haura said, noting that he and his colleagues hope to begin working with chemists on the design of such agents.