The Ink4a/Arf locus operates as a regulator of the circadian clock modulating RAS activity

The mammalian circadian clock and the cell cycle are two major biological oscillators whose coupling influences cell fate decisions. In the present study, we use a model-driven experimental approach to investigate the interplay between clock and cell cycle components and the dysregulatory effects of RAS on this coupled system. In particular, we focus on the Ink4a/Arf locus as one of the bridging clock-cell cycle elements. Upon perturbations by the rat sarcoma viral oncogene (RAS), differential effects on the circadian phenotype were observed in wild-type and Ink4a/Arf knock-out mouse embryonic fibroblasts (MEFs), which could be reproduced by our modelling simulations and correlated with opposing cell cycle fate decisions. Interestingly, the observed changes can be attributed to in silico phase shifts in the expression of core-clock elements. A genome-wide analysis revealed a set of differentially expressed genes that form an intricate network with the circadian system with enriched pathways involved in opposing cell cycle phenotypes. In addition, a machine learning approach complemented by cell cycle analysis classified the observed cell cycle fate decisions as dependent on Ink4a/Arf and the oncogene RAS and highlighted a putative fine-tuning role of Bmal1 as an elicitor of such processes, ultimately resulting in increased cell proliferation in the Ink4a/Arf knock-out scenario. This indicates that the dysregulation of the core-clock might work as an enhancer of RAS-mediated regulation of the cell cycle. Our combined in silico and in vitro approach highlights the important role of the circadian clock as an Ink4a/Arf-dependent modulator of oncogene-induced cell fate decisions, reinforcing its function as a tumour-suppressor and the close interplay between the clock and the cell cycle network.

inhibits the subsequent phosphorylation of RB1 (Figure 2A) [5]. In this model, we use CDK to represent all CDKs inhibited by INK4a, namely CDK4 and CDK6.
It has been shown that the expression of ARF, another protein encoded by the CDKN2A locus, can be activated by MYC (Figure 2B) [6]. Even though it is not clear if this activation is achieved through a direct binding to the promoter of the Arf gene, it is common to model the interaction using Hill-type kinetics [7]. Accumulated ARF stabilizes p53 by binding to MDM2, a E3 ubiquitin ligase targeting p53 in the nucleus ( Figure 2B) [8].

The INK4a/RB/E2F pathway and its regulation of Bmal
In order to interpret the circadian phenotype of INK4a/ARF-knockout MEFs, it is necessary to extend the model with a feedback from INK4a and ARF to the core circadian clock. For this, we used the INK4a-CDK/CycD-Rb-E2F pathway (Figure 3). The transcription factor MYC directly induces the synthesis of Cdk4 [9]. CDK4 and another cyclin D-dependent kinase, CDK6, form an active complex with CycD and play an important role in the phosphorylation of RB1, the key regulator of the E2F family of transcription factors. Once RB1 is phosphorylated, active

El-Athman et al.
Supporting Information S1 Text -Model description, design and analysis 5 E2F will be released from the RB1/E2F complex [10][11][12]. MotifMap, a database of candidate regulatory motif sites in humans, reports that several E2F activators such as E2F1, E2F2, and E2F3a can potentially bind to the promoter of Bmal1 to activate its transcription [13]. On the other hand, the formation of the CDKs/CycD complex is inhibited by INK4a, which has a negative effect on RB1 phosphorylation and reinforces the inhibition of E2F [5]. MYC also promotes the transcription of the three E2Fs [14,15]. In this model, we used E2F to represent the three activators belonging to E2F family, i.e. E2F1, E2F2, and E2F3a. The heterodimer MYC:MAX has also been reported to bind to E-boxes and thereby to influence the circadian clock either by inducing REV-ERBα to dampen the expression and oscillation of BMAL1 [16] or by direct repression of BMAL1 and CLOCK via MIZ1 [17]. Moreover, MYC has been reported to repress Per1 transcriptional activation by CLOCK/BMAL1 via competitive targeting of E-box sequences of the Per1 promoter [18]. In the model, this connection is included implicitly via the Bmal inhibition rate.
In addition, the tumor suppressor protein p53 inhibits the phosphorylation of RB1 via the p21/p27-CDK/CycE-RB1 pathway. Both p21 and p27 are inhibitors of the cyclin E-dependent kinase CDK2, which regulates RB1 phosphorylation and E2F activity synergistically with CDK4/CycD and CDK6/CycD, thus influencing Bmal transcription [19,20]. The transcription of p21 is induced by p53 [21]. To reduce the complexity, the effect of the p53-p21/p27-CDK/CycE arm was modelled as a negative correlation between p53 and the enzymatic activity of CDK/CycE (Figure 3).

El-Athman et al.
Supporting Information S1 Text -Model description, design and analysis

The ARF/MDM2/p53 pathway and its regulation of Per
The ARF/MDM2/p53/Per pathway is a feedback from ARF to the core circadian clock ( Figure   4). The expression of ARF can be activated by MYC [6]. Accumulated ARF associates with MDM2 and leads to rapid degradation of MDM2, thereby inhibiting the MDM2-mediated degradation of p53 and promoting p53 stabilisation and accumulation [22]. Recent data showed that there is a p53 response element located in the promoter region of the Per2 gene which overlaps with E-box cis-elements crucial for CLOCK/BMAL-mediated Per2 transcription [23]. The binding of p53 strongly represses the transcription of Per2 by competing with CLOCK/BMAL for binding to the Per2 promoter [23], as a result p53 and Per2 are out-of-phase

Additional model analysis
To further explore the effect of RAS on the circadian clock in silico, we compared the Bmal phenotypes and the corresponding changes in period length after the perturbation by different levels of RAS overexpression represented by the parameter ktt<1 (Figure 6). When measuring the period for the first six peaks (five periods) after introducing the perturbation of RAS (represented by ktt<1), the same trend could be observed as for measuring the first three periods (Figure 7). Furthermore, we simulated the Bmal phenotype of the Ink4a/Arf -/system following an inhibition of RAS (represented by ktt=1.2) which resulted in a longer period (Figure 8) as was also observed in our experimental data (Figure S1C,E).
We additionally investigated the importance of the INK4a/RB1/E2F1 pathway (module 1) and the ARF/MDM2/p53 pathway (module 2) in reproducing the effect of RAS overexpression on the Bmal period by either uncoupling them from the core-clock system or by setting their expression to their constitutive average value (Figure 9).
In the model, we measured the period in the transient region of the simulations. This is in agreement with our RT-qPCR data in IMR-90 cells on day 5 and 11 after overexpression of RAS.
The data show that despite the assumed stability of retrovirus-mediated Hras overexpression, the expression level of Hras display some biological noise: it first strongly increases (day 5) and then decreases again (Figure 10). In silico expression data show that upon simulation of RAS overexpression, the Ink4a/Arf +/+ system acquires a longer and Ink4a/Arf -/system a shorter period compared to the corresponding simulated wild type system. The period was measured for a transient region, defined as the mean of the time between the first six peaks (five periods) after introducing the perturbation of RAS (represented by ktt<1).

El-Athman et al.
Supporting Information S1 Text -Model description, design and analysis 11

Figure 9: Modular analysis of Bmal expression level after perturbation by different levels of RAS.
The importance of the INK4a/RB1/E2F1 pathway (module 1) and the ARF/MDM2/p53 pathway (module 2) in influencing the circadian period is analysed by simulating different scenarios in silico.
The simulated Bmal expression profiles show phase-shifted oscillations that cause differing effects following the perturbation by RAS (represented by ktt<1). A) Module 1 is decoupled from the coreclock or B) the oscillatory expression of its connective component E2FN is clamped to its constitutive average value. C) Module 2 is decoupled from the core-clock or D) the oscillatory expression of its connective component p53N is clamped to its constitutive average value.