Modeling the dynamics of hypoxia inducible factor-1{alpha} (HIF-1{alpha}) within single cells and 3D cell culture systems

HIF (Hypoxia Inducible Factor) is an oxygen-regulated transcription factor that mediates the intracellular response to hypoxia in human cells. There is increasing evidence that cell signaling pathways encode temporal information, and thus cell fate may be determined by the dynamics of protein levels. We have developed a mathematical model to describe the transient dynamics of the HIF-1{\alpha} protein measured in single cells subjected to hypoxic shock. The essential characteristics of these data are modeled with a system of differential equations describing the feedback inhibition between HIF-1{\alpha} and Prolyl Hydroxylases (PHD) oxygen sensors. Heterogeneity in the single-cell data is accounted for through parameter variation in the model. We previously identified the PHD2 isoform as the main PHD responsible for controlling the HIF-1{\alpha} transient response, and make here testable predictions regarding HIF-1{\alpha} dynamics subject to repetitive hypoxic pulses. The model is further developed to describe the dynamics of HIF-1{\alpha} in cells cultured as 3D spheroids, with oxygen dynamics parameterized using experimental measurements of oxygen within spheroids. We show that the dynamics of HIF-1{\alpha} and transcriptional targets of HIF-1{\alpha} display a non-monotone response to the oxygen dynamics. Specifically we demonstrate that the dynamic transient behavior of HIF-1{\alpha} results in differential dynamics in transcriptional targets.

💡 Research Summary

The authors present a comprehensive mathematical framework that captures the transient dynamics of hypoxia‑inducible factor‑1α (HIF‑1α) in both single cells and three‑dimensional (3D) spheroid cultures. The core of the model is a negative‑feedback loop between HIF‑1α and prolyl hydroxylases (PHDs), which act as oxygen sensors. HIF‑1α synthesis, PHD‑mediated hydroxylation, and proteasomal degradation are each described by distinct kinetic constants (k_syn, k_oh, k_deg). The hydroxylation rate is assumed to increase with both intracellular oxygen concentration and HIF‑1α level, thereby creating a rapid rise of HIF‑1α after a hypoxic shock followed by a swift decline once PHD activity catches up. This simple set of ordinary differential equations reproduces the characteristic “peak‑decay” profile observed experimentally.

To account for cell‑to‑cell variability, the authors fit the model to time‑lapse fluorescence data from hundreds of individual cells using Bayesian inference, extracting a distribution of parameter values. The analysis reveals that variability in the PHD2 isoform’s expression and catalytic efficiency is the dominant source of heterogeneity: cells with higher PHD2 activity display lower HIF‑1α peaks and shorter durations. This finding confirms earlier experimental work that identified PHD2 as the principal regulator of the HIF‑1α transient response.

The model is then extended to 3D spheroids. Oxygen diffusion and consumption are modeled by solving the radial diffusion‑consumption equation, yielding spatially resolved oxygen profiles that match measurements obtained with oxygen‑sensitive probes. These profiles serve as time‑varying inputs to the HIF‑1α‑PHD system for each radial position. Simulations show that HIF‑1α dynamics in the spheroid are non‑monotonic with respect to the external oxygen swing: cells near the periphery experience brief HIF‑1α spikes, whereas cells in the core sustain elevated HIF‑1α levels for longer periods. Importantly, the model incorporates a transcriptional delay (τ) between HIF‑1α accumulation and the expression of downstream targets such as VEGF, GLUT1, and EPO. The resulting target gene dynamics differ markedly from the HIF‑1α profile, especially under repetitive hypoxic pulses. The authors predict that increasing the frequency or duration of hypoxic pulses amplifies target gene expression more than it does the HIF‑1α peak, suggesting that temporal coding of hypoxic signals can selectively bias downstream transcriptional programs.

Experimental validation is proposed through (i) repeated hypoxia‑reoxygenation cycles (e.g., 5 min hypoxia followed by 10 min normoxia) with simultaneous live‑cell imaging of HIF‑1α and PHD2, (ii) pharmacological inhibition of PHD2 using dimethyloxalylglycine (DMOG) to test the predicted enlargement of HIF‑1α peaks, and (iii) measurement of target gene mRNA/protein levels after defined pulse regimes. Sensitivity analysis further identifies k_oh (oxygen‑dependent hydroxylation) and k_deg (degradation) as the parameters most influencing peak height and duration, highlighting them as potential therapeutic leverage points.

Overall, the study delivers a unified, experimentally anchored model that links oxygen dynamics to HIF‑1α signaling and downstream transcription in both monolayer and 3D contexts. By demonstrating how the temporal pattern of hypoxia, rather than just its magnitude, shapes cellular responses, the work provides a valuable quantitative tool for investigating hypoxia‑related pathologies such as cancer, ischemic disease, and for designing hypoxia‑targeted therapies.