Role of cell-to-cell variability in activating a positive feedback antiviral response in human dendritic cells

Abstract

In the first few hours following Newcastle disease viral infection of human monocyte-derived dendritic cells, the induction of IFNB1 is extremely low and the secreted type I interferon response is below the limits of ELISA assay. However, many interferon-induced genes are activated at this time, for example DDX58 (RIGI), which in response to viral RNA induces IFNB1. We investigated whether the early induction of IFNBI in only a small percentage of infected cells leads to low level IFN secretion that then induces IFN-responsive genes in all cells. We developed an agent-based mathematical model to explore the IFNBI and DDX58 temporal dynamics. Simulations showed that a small number of early responder cells provide a mechanism for efficient and controlled activation of the DDX58-IFNBI positive feedback loop. The model predicted distributions of single cell responses that were confirmed by single cell mRNA measurements. The results suggest that large cell-to-cell variation plays an important role in the early innate immune response, and that the variability is essential for the efficient activation of the IFNB1 based feedback loop.

Figure 1. Time course of IFNB1 and DDX58 induction.

Measurement of IFNB1 (solid line) and DDX58 (dashed line) expression in human DCs at 1, 2, 4, 6, 8, 10, 12, 14, 16, and 18 hours following NDV infection. Percent of maximal induction was measured by microarray, compared to non-infected control, and shows that half-maximal induction of DDX58 occurs hours prior to IFNB1 half-maximal induction.

Figure 2. DDX58 and IFNB1 mRNA expression level.

2A. Total mRNA copy number. DDX58 and IFNB1 mRNA were measured by quantitative real-time PCR and the relative IFNB1 and DDX58 expression levels were normalized to ACTB. The columns show the gene expression level 7 hrs post treatment [NDV only (NDV), NDV plus blocking antibodies (NDV+Abs), not infected (ni), not infected with blocking antibodies (ni+Abs)]. Left panel: IFNB1. Right panel: DDX58. Error bars represent measurement error. 2B. Single cell mRNA expression. DDX58 and IFNB1 mRNA in individual DCs were measured 7 hrs post treatment by the hemi-nested PCR protocol illustrated in Supplementary Figure S4. The copy numbers of both DDX58 and IFNB1 in single DCs (normalized to ACTB) were determined for all cells with detectable expression of the mRNAs. Each symbol shows the gene expression in a single cell. The treatments were same as those in Figure 2A , but with uninfected DCs labeled as control (Ctr). Left panel: IFNB1. Right panel: DDX58.

Figure 3. Single DC simulations with and without IFN-blocking antibodies.

3A. Time course of the average copy number of IFNB1 (solid line) and of DDX58 (dashed line) obtained from the simulation with the ABM model. 3B. Scatter plots display IFNB1 (X axis) and DDX58 (Y axis) mRNAs copy numbers in each single DC as simulated with the ABM model. Simulations without blocking antibodies (upper panel) or with blocking antibodies (lower panels) are shown at 6 hrs (left) and 11 hrs (right) post infection.

Figure 4. Experimental results in individual DCs with and without blocking antibodies.

The scatter plots display IFNB1 (X axis) and DDX58 (Y axis) copy numbers in individual cells, as determined by the hemi-nested single cell qRT-PCR, and normalized to ACTB. 4A. IFNB1 and DDX58 expression in single DCs with no NDV infection and no antibody blockage. 4B. IFNB1 and DDX58 expression in single DCs following NDV infection. In the same order as in Figure 3B, experiments without blocking antibodies (upper panels) or with blocking antibodies (lower panels) are shown at 6 hrs (left) and 11 hrs (right) post infection. Note that the dataset for 6 hours was obtained from donor 1, while the datasets for 11 hours and uninfected control were obtained from donor 2.

Figure 5. Single cell simulations of early responder DCs.

Simulation scatter plots of IFNB1 (X axis) and of DDX58 (Y axis) copy numbers in individual DCs at 1, 2, 3 and 4 hours post infection. The simulations were performed with 1000 cells in order to show a sufficient number of cells responding to infection at early time points.

Figure 6. Phase space trajectories of individual cells in simulations.

Each plot follows a single cell at 10 minute intervals, and plots the number of bound receptors vs. the number of IFNB1 messages in that cell at that time point. A. An uninfected cell cannot produce IFNB1 message, and thus exhibits only an increase in the number of bound receptors. B. A late responder cell is characterized by receptors binding before the cell produces IFNB1 messages, and is thus activated only through paracrine signaling. C. An early responder cell produces considerable amounts of IFNB1 message prior to significant receptor binding, suggesting an autocrine activation. The ratio between early and late responders in the simulation is 7∶124. D. A cell that shows a late responder trajectory in the simulation without antibodies (solid line), but changes to a trajectory suggesting autocrine activation when the simulation includes antibodies (dashed line).

Figure 7. Single DC simulation with high and low cell-to-cell variation.

Time course of the average copy number of IFNB1 per infected cell (solid line) and of DDX58 (dashed line) obtained from the simulation with the ABM model. The lines marked by “Using Fitted Variance” are drawn from a simulation with parameters fitted to the experimental data, and are identical to the ones shown in Fig. 3A. The lines marked by “Using Reduced Variance” result from a simulation in which the variance of the initial DDX58 concentration was reduced 10-fold. In order to ensure a similar number of early responder cells, the sensitivity of IFNB1 to RIG-I concentration was increased by more than 30-fold. The reduction in variability leads to results that do not account for the observed delay in IFN induction relative to DDX58 induction. Furthermore the highly variable system generates low levels of early interferon signaling that can initiate antiviral responses without being prone to later high and potentially toxic levels of interferon secretion.

Figure 8. Processes, Descriptions and Rates.

The processes, descriptions and rates for each of the six reactions in the stochastic simulations.