The circadian clock, a fundamental biological regulator, governs crucial cellular processes. Circadian-based therapeutic strategies are gaining recognition for enhancing treatment efficacy and minimizing side effects by aligning drug administration with the circadian rhythm. However, identifying optimal treatment timings remains challenging. This study introduces a high-throughput approach combining live-cell imaging and advanced data analysis to deeply phenotype cancer cell models. The approach evaluates circadian rhythms, growth dynamics, and drug responses across different times of day. This allows for the identification of optimal treatment windows, responsive cell types, and effective drug combinations. Furthermore, it utilizes multiple computational tools to pinpoint cellular and genetic factors that influence time-of-day drug sensitivity. The method's adaptability to diverse biological models broadens its applicability. The ultimate goal is to optimize anti-cancer drug treatments by leveraging circadian rhythms, promising improved outcomes and transformative treatment strategies. The circadian clock's disruption is linked to multiple cancer subtypes, affecting processes like cell proliferation, immune response, DNA repair, and apoptosis. Existing studies demonstrate that administering chemotherapeutic agents according to circadian rhythms alters their efficacy, but a standardized method for optimal timing is lacking. This research aims to fill this gap by providing a comprehensive methodology to determine the optimal time-of-day for drug administration, identify the most responsive cell subtypes, and uncover the cellular mechanisms governing time-of-day drug sensitivity.

The literature extensively documents the circadian clock's influence on various physiological and behavioral processes across different organisms. In mammals, a hierarchical organization ensures coordinated rhythms at cellular and organismal levels. Studies in primates and mice reveal rhythmic tissue-specific expression of protein-coding genes. These clock-controlled genes regulate key processes like metabolism, cell proliferation, immune response, DNA repair, and apoptosis. Disruption of the circadian system is implicated in cancer development and progression, linked to hallmarks such as sustained proliferation and metastasis. Patients with mutations in circadian clock genes often exhibit lower survival rates. Furthermore, the circadian clock interacts directly with therapeutic targets, influencing drug responses. Existing research demonstrates the impact of circadian rhythm-aligned drug administration on chemotherapeutic efficacy, but a standardized method for identifying optimal treatment times remains elusive. This study builds upon this existing knowledge to develop a comprehensive and standardized approach.

The study employs a multi-faceted approach involving experimental and computational techniques. First, a deep circadian phenotyping is performed on a panel of cancer and healthy tissue cell models. This involves integrating recordings of circadian clock activity with comprehensive time-series analysis techniques (autocorrelation, continuous wavelet transform, and multiresolution analysis) to quantify circadian clock strength. Three complementary techniques are used to capture different aspects of circadian clock dynamics: autocorrelation for stable temporal features; continuous wavelet transform for time-dependent amplitude and period changes; and multiresolution analysis for multi-scale features, ensuring comprehensive understanding of signal dynamics. Next, growth characteristics and drug sensitivities are evaluated across a spectrum of drugs and cell lines using time-resolved live-cell imaging, tracking cell nuclei and confluency. This dynamic approach captures temporally evolving effects of drugs, avoiding the limitations of single time-point measurements. A normalized growth rate inhibition (GR) approach is used to obtain robust drug sensitivity metrics, including GEC₅₀, GR₅₀, GRinf, Hill coefficient, and GRAOC. To screen for time-of-day drug sensitivities, a streamlined experimental strategy minimizes workload while increasing throughput, reproducibility, and accuracy. A three-step circadian clock resetting protocol using dexamethasone pulses is implemented. Drugs are administered at their estimated half-effective doses at various times relative to the circadian cycle. Live-cell imaging tracks drug responses over several days. Drug responses are determined by comparing cell counts at treatment and evaluation times to account for different initial cell densities. The maximum variability in relative responses is quantified as the ToD Maximum Range (ToDMR). Computational analysis includes linear regression, dominance analysis, and determinant ranking to investigate how circadian clock strength, growth dynamics, and drug sensitivity parameters influence ToDMR values. Linear correlation analysis, linear discriminant analysis (LDA), and principal component analysis (PCA) are used to explore the relationship between the expression patterns of core circadian clock genes and ToDMR values. Finally, a chronotherapeutic index is defined to rank cellular models and drug agents based on their potential benefits from circadian-aligned treatments.

The study reveals heterogeneous circadian clock phenotypes across different cancer and healthy cell models, with some cancer cells exhibiting robust circadian rhythms. Analysis of growth dynamics in triple-negative breast cancer (TNBC) cell lines shows substantial variability in growth rates and doubling times. Drug sensitivity analysis using a normalized growth rate inhibition (GR) approach reveals drug-dependent and metric-specific sensitivity. Time-of-day (ToD) drug sensitivity profiles vary significantly across drugs and cell lines, with some exhibiting up to 30% response differences throughout the day. The ToD Maximum Range (ToDMR) is identified as a key parameter for time-of-day sensitivity. Analysis shows that circadian clock metrics, particularly Bmall amplitude, are strongly correlated with ToDMR values, while growth and drug sensitivity metrics show weaker correlations. Dominance analysis reveals that Bmall amplitude is the most significant factor influencing ToDMR for some drugs, while other drugs show a more homogenous distribution of contributions. Analysis of core circadian clock gene expression shows limited associations between individual gene expression levels and ToDMR values, but LDA and PCA reveal that specific gene combinations can contribute to discriminating between high and low ToD-dependent sensitivity, varying by drug. The study introduces a chronotherapeutic index that ranks cancer cell models and drugs based on their differential time-of-day sensitivity compared to non-malignant cells. The index highlights treatment times that maximize cancer toxicity while minimizing harm to healthy tissues.

The findings challenge the common assumption that most cancer cells have weak circadian clocks, highlighting the importance of model-specific characterization of circadian strength. The diverse drug sensitivity metrics and the lack of strong correlations among them challenge the traditional binary classification of sensitive versus resistant models. The substantial variability in time-of-day sensitivity underscores the potential benefits of chronotherapy, and the chronotherapeutic index offers a valuable tool for identifying optimal treatment schedules by considering the differential sensitivity between cancer and healthy cells. The study reveals complex interactions between circadian clock dynamics, growth patterns, and drug sensitivities in shaping time-of-day drug responses. The varying contributions of these factors depending on the specific drug highlights the need for a multifaceted approach. While the direct impact of individual core clock genes on ToD sensitivity is limited, the collective influence of these genes is significant, hinting at a complex interplay within the circadian clock network. Future research could focus on the role of the cell cycle and drug target expression levels in shaping ToD sensitivity profiles.

This study presents a comprehensive high-throughput method for identifying optimal drug treatment times in cancer, considering both circadian rhythms and differential sensitivity between cancer and healthy tissues. The findings underscore the potential for chronotherapy to improve cancer treatment outcomes. Future research could extend this approach to more complex models like 3D organoids and animal models, paving the way for personalized chronotherapeutic strategies.

The study primarily utilizes in vitro models, which may not fully reflect the complexities of in vivo settings. While robust methods were employed, the potential impact of random reporter insertions on circadian clock estimations should be considered. Although the study uses population-based recordings of non-clonal cell lines, it is also important to acknowledge the limitations in generalizability to individual patient heterogeneity.