Capturing cell‐fate decisions from the molecular signatures of a receptor‐dependent signaling response

Open Access
Publisher Website
Google Scholar
Abstract
We examined responses of the B‐cell antigen receptor‐dependent intracellular signaling network to targeted perturbations induced through siRNA‐mediated depletion of select signaling intermediates. The constituent nodes displayed graded sensitivities, which resulted from the differential effects of perturbations on the kinetic and quantitative aspects of phosphorylation at each node. By taking the rate of initial phosphorylation, rate of subsequent dephosphorylation, and the total intensity of phosphorylation at each node as separate signaling parameters, we generated data‐driven models that accurately predicted the cellular responses of apoptosis, proliferation, and cytokine secretion. Importantly, the effects of perturbation on the primary target alone did not yield successful models. Rather, it also required incorporation of secondary effects on many other nodes. A significant feature of these models was that the three signaling parameters derived from each node functioned largely as independent entities, making distinctive contributions to the cellular response. Thus, the kinetic and quantitative features of phosphorylation at a node appear to play discrete roles during signal processing. ### Synopsis Our present study was designed with a two‐fold objective in mind. The first of these was to determine how the signal transduction network adapted to perturbations. It was anticipated that such studies would yield information on the basis of the structural plasticity of the signaling network. In addition to this, by examining the consequent effects on the cellular response, we also hoped to gain insights into how signal processing is achieved. For these experiments, we employed a murine B lymphoma cell line that could be stimulated through the B‐cell antigen receptor (BCR) with the F(ab)2 fragment of anti‐ IgG. Signaling from the BCR was then monitored in terms of the time‐dependent phosphorylation of a panel of 21 signaling intermediates. To dissect factors contributing to plasticity, we also employed siRNA to individually deplete cells of each of these signaling intermediates. In these latter experiments, we examined how each of these depletions influenced BCR‐dependent phosphorylation at the remaining signaling intermediates. Thus, collectively, these experiments yielded data on the behavior of each of these signaling intermediates (or nodes) under 21 distinct conditions of perturbation. A preliminary analysis of these results revealed that individual nodes displayed widely divergent sensitivities to these perturbations. Thus, there were cases such as Akt that were relatively resistant to these perturbations, while molecules such as JNK and PLCã proved to be highly sensitive. To obtain a more detailed characterization, we described the phosphorylation profile at each of the nodes in terms of three signaling parameters that combined both its kinetic and quantitative features. The three parameters were the rate of initial phosphorylation (or activation rate), the total area under the phosphorylation curve, and the rate of subsequent dephosphorylation from the peak value (decay rate). Taken together, these three signaling parameters provided a unique description of the phosphorylation profile at each node under each perturbation condition. An examination of how these parameters were influenced by the perturbations performed provided us with some novel insights ([Figure 2][1]). At one level, these three parameters were found to differentially contribute to the sensitivity of the various nodes to the perturbations. Further, each of these individual signaling parameters was also found to exhibit differential sensitivity depending upon the perturbation performed. These two features together defined nodal sensitivity as a multivariate property where the response to the various perturbations was expressed through combinatorial variations in the pattern of signaling parameter values generated at each of the nodes. In other words, the emergent properties of the signaling network appear to derive from modulations in relative contributions from the node‐specific signaling parameters. The above studies were also accompanied by parallel experiments wherein we examined the effects of these perturbations on the resulting BCR‐dependent cellular responses. The responses examined were cell proliferation, protection against Fas‐mediated apoptosis, and secretion of the cytokine IL‐2. A cursory examination revealed that the perturbations employed had a significant impact on all the three cellular response modes. To further probe and, possibly, extract causal relationships between the perturbation‐induced alterations in the behavior of the signaling network, and the corresponding cellular response profile, we undertook an exercise in data‐driven modeling. For this exercise, we employed a partial least square (PLS) regression analysis wherein the signaling parameters were taken as the independent variables ( X ) and the cell response profiles as the dependent variable set ( Y ). Our choice of an approach based on PLS modeling was based on the fact that it captures covariance between the independent and dependent variable groups, thereby extracting relationships that facilitate generation of a hypothesis. Our efforts proved successful in that we were able to generate PLS models that accurately captured those defining features of the signaling network that characterized the cellular responses of proliferation, apoptosis, and cytokine secretion ([Figure 4][2]). Importantly, these models survived stringent validation by accurately predicting the dependent variables from untrained data sets that were derived either from the siRNA of additional molecules or from the use of pharmacological inhibitors as perturbing agents. Importantly, we were unable to obtain satisfactory models by incorporating only the specific...