Advancing clinical response against glioblastoma: Evaluating SHP1705 CRY2 activator efficacy in preclinical models and safety in phase I trials

Abstract

Background

It has been reported that circadian clock components, Brain and Muscle ARNT-Like 1 (BMAL1) and Circadian Locomotor Output Cycles Kaput (CLOCK), are essential for glioblastoma (GBM) stem cell (GSC) biology and survival. Consequently, we developed a novel Cryptochrome (CRY) activator SHP1705, which inhibits BMAL1-CLOCK transcriptional activity.

Methods

We utilized GlioVis to determine which circadian genes are differentially expressed in non-tumor versus GBM tissues. We employed in vitro and in vivo methods to test the efficacy of SHP1705 against patient-derived GSCs and xenografts in comparison to earlier CRY activator scaffolds. We applied a novel REV-ERB agonist SR29065, which inhibits BMAL1 transcription, to determine whether targeting both negative limbs of the circadian transcription-translation feedback loop (TTFL) would yield synergistic effects against various GBM cells.

Results

SHP1705 is the first circadian clock-modulating compound to be found safe and well-tolerated in Phase I clinical trials. SHP1705 has increased selectivity for the CRY2 isoform and potency against GSC viability compared to previously published CRY activators, making it promising for applications in GBM where CRY2 levels are found to be low. SHP1705 prolonged survival in mice bearing GBM tumors established with GSCs. When combined with novel REV-ERB agonist SR29065, SHP1705 displayed synergy against multiple GSC lines and differentiated GSCs (DGCs).

Conclusions

We demonstrate the efficacy of SHP1705 against GSCs, which pose as a major source of chemoradiation resistance leading to poor GBM patient prognosis. Novel circadian clock compounds have high potential for targeting GBM as single agents or in combination with each other or current standard-of-care.

Key Points.

• SHP1705, a novel CRY2 activator, is the first clock compound that is well-tolerated in Phase 1 safety trials.

• It has significantly improved potency against GSCs and GBM tumors.

• REV-ERB agonist SR29065 and SHP1705 display synergistic effects against GSCs.

Importance of the Study.

SHP1705, a CRY2 activator has demonstrated to be safe and well-tolerated in Phase 1 trials with healthy individuals, has significantly improved preclinical efficacy against GSCs. Given that CRY2 is decreased in GBM tissues compared to CRY1, promoting CRY2 activity over CRY1 may be more efficacious in targeting GBM. In support of this, SHP1705 is significantly more potent than other previously published CRY activators including KL001, which has affinity for both CRYs. SHP1705 poses as an ideal candidate for applications in combination studies with standard-of-care as a novel GBM treatment paradigm.

Glioblastoma (GBM) is the most common primary brain tumor type with a poor prognosis and limited effective treatment options especially after recurrence.¹ GBM has a median survival time of 15 months with the current standard of care or the Stupp protocol.² This consists of an aggressive regimen of maximal surgical resection followed by concurrent temozolomide (TMZ) chemotherapy and radiation than adjuvant TMZ.³ One of the major challenges in treating GBM is the presence of GBM stem cells (GSCs), which have tumor-initiating properties, secrete angiogenic factors, induce an immune suppressive environment, are highly invasive, and are resistant to both chemotherapy and radiation. GSCs can remain in and well beyond the margins of the surgical cavity following tumor resection and drive tumor recurrence.^4–6

Core circadian clock components BMAL1 and CLOCK are essential to GSC biology. BMAL1 and CLOCK chromatin binding sites were shown to be increased compared to neural stem cells (NSCs), allowing for BMAL1-CLOCK-driven expression of stemness and metabolic genes in GSCs.⁷ BMAL1 and CLOCK drive expression of the chemokine Olfactomedin-Like 3 (OLFML3) in GSCs to recruit immune suppressive microglia and drive angiogenesis.^8,9 Perturbation of either BMAL1 or CLOCK genetically or pharmacologically resulted in specific cell death in GSCs that was not seen in differentiated GSCs (DGCs) or noncancerous cells; down regulated stemness markers, the citric acid (TCA) cycle genes, and OLFML3; induced apoptosis as indicated by cleavage of poly-ADP ribose polymerase (PARP); and decreased tumor volume and increased survival in GBM patient-derived xenograft (PDX) mouse models established with GSCs.^7–9

Circadian rhythms in mammals are generated by cycling in gene expression that is driven by BMAL1 and CLOCK transcriptional output in individual cells under tight control of a TTFL. This results in roughly a 24-h period in behavioral and physiological output. BMAL1 and CLOCK form a heterodimer (BMAL1-CLOCK) and converge upon the E-box motifs of clock-controlled gene (CCG) promoters to drive cyclical gene expression. BMAL1-CLOCK also control the expression of their positive and negative regulators. One TTFL loop of the clock consists of Cryptochrome 1/2 (CRY1/2) and Period 1/2 (PER1/2), which form a heterodimer that inhibits the BMAL1-CLOCK transcriptional loop. In another TTFL loop, REV-ERBα/β inhibits while retinoic acid-related orphan receptor α/β/γ (RORα/β/γ) promotes the transcription of BMAL1 by acting upon the ROR-element (RORE) in its promotor (Figure 1A). This TTFL network and their posttranslational modifications generate rhythmicity in key bodily functions.^10,11

Figure 1.

SHP1705 displays significantly improved anti-GSC effects compared to earlier CRY activator scaffolds. (A) Mechanism of the mammalian clock and CRY activators. BMAL1 and CLOCK form a heterodimer and converge upon the E-box motifs of clock-controlled gene promoters to drive their transcription. BMAL1 and CLOCK also drive the expression of their positive and negative regulators. CRYs and PERs form a heterodimer and inhibit the transcriptional activity of BMAL1-CLOCK. RORα/β/γ promote while REV-ERBα/β repress the expression of BMAL1 by acting upon the RORE of its promoter. CRY activators prevent FBXL3-mediated ubiquitination of CRY1/2 proteins and their downstream degradation. Consequently, CRY1/2 can continue to form a heterodimer with PER1/2 and repress BMAL1-CLOCK transcriptional activity. (B)  CRY2 mRNA expression of non-tumor and GBM tissues from adult TCGA_GBM HG-U133A data plotted and analyzed with GlioVis. Cell viability analysis of NM cells (solid blue lines), GSCs (solid lines), or DGCs (dotted lines of matched GSC solid lines) following (C) SHP1705, (D) KL001, (E) SHP656, or (F) SHP1703 treatment for 3 days (n = 3-4 biologically independent samples). (G) Heatmap of summarized IC₅₀ values following CRY activator treatment for 3 days in control NM cells, GSCs, and DGCs.

BMAL1 and CLOCK are transcription factors that, to date, have been difficult to directly target pharmacologically. However, small molecules have been developed that indirectly modulate their function or transcription by directly engaging with their negative regulators CRY and REV-ERB.^11–13 One such type of circadian clock compounds that have been developed by our group are the CRY activators. These were identified through a series of high throughput luciferase-based reporter assays that monitored changes in Bmal1 (Bmal1-dLuc) and Per2 (Per2-dLuc) gene expression cycling. CRY activators compete with F-Box and leucine-rich repeat protein (FBXL3) in the flavin adenine dinucleotide (FAD) binding pockets of CRY1/2, therefore preventing ubiquitination and downstream proteasome degradation (Figure 1A). KL001 was the first reported CRY activator and, since then, bioavailable CRY isoform-selective scaffolds, such as SHP656 and SHP1703, have since been developed. We have previously shown that these compounds have anti-GSC effects in both in vitro and in vivo assays, and SHP656 and SHP1703, in particular, are selective for CRY2.^7,14–17

CRY2 was found to be dysregulated in a glioma rat model and sensitize cells to apoptosis following radiation.¹⁸ These studies suggest that pharmacologically perpetuating CRY2 function can have favorable anti-cancer effects against GSCs and GBM tumors. Consequently, in this study, we focused specifically on investigating the efficacy of a novel CRY2 activator scaffold, SHP1705.

Materials and Methods

Cell Culture

Derivation of GSC 387, GSC 3565, MGG 4, MGG 8, MGG 31, and hGBM18 FMC MP1 are detailed in the Supplementary Material¹⁹. They were cultured in Neurobasal-A medium without phenol (Gibco, Cat. #12349-015), B27 supplement without Vitamin-A (Life Technologies, Cat. #12587-010), 1% penicillin/streptomycin (10,000 U/mL, Gibco, Cat. #15140-122), 1% GlutaMAX^TM (100X, Gibco, Cat. #35050-61), 1% Sodium Pyruvate (100 mM, Gibco, Cat. #11360-070), 0.02 µg/mL EGF (0.5 mg/mL dissolved in PBS, R&D, Cat. #236-EG-01M), and 0.02 µg/mL FGF (0.5 mg/mL dissolved in PBS, Cat. #4114-TC-01M) (complete Neurobasal-A media). NM cells were cultured in 1:1 ratio of Dubecco’s Modified Eagle Medium (DMEM) (Gibco, Cat. #11995-065), 10% fetal bovine serum (FBS, R&D Systems, Cat. #S11150), and 1% penicillin/streptomycin (10,000 U/mL, Gibco, Cat. #15140-122) and complete Neurobasal-A media. DGCs were differentiated from GSCs and maintained by adding 10% FBS (R&D Systems, Cat. #S11150) to complete Neurobasal-A media for at least 48 h. Cells were all incubated at 37°C and 5% CO₂. Cells were kept at less than 10 passages, and cells that were used for in vivo studies were kept below 5 passages.

Cell Viability Assays

Cells were plated in 96-well black, clear, flat bottom plates (Falcon, Cat. #353219) at 1,000 cells/100 μL/well. SHP1705 or SR29065 was dissolved in DMSO (Sigma-Aldrich, Cat. #D2650-100ML) to make a stock concentration of 10 mM. SHP1705 and/or SR29065 was added at the desired concentrations accordingly to the plate (final ≥ 1% DMSO). Cells were incubated at 37°C and 5% CO₂ for 3–4 days with clock compound. 50 μL of Cell-Titer Glo (Promega, Cat. # G7573) was added to each well at the time of reading, and the plate was incubated at room temperature and protected from light for 15 min. Luminescence was read as a function of ATP levels by the Tecan Infinite Pro M200 plate reader. Values were plotted in GraphPad Prism to generate cell viability curves and determine IC₅₀ values.

qPCR

Cells were plated in a 6-well plate at 1 × 10⁶ cells/well with the appropriate cell medium and the indicated concentrations of SHP1705 for 24 h. The cell lysates were then collected and snap frozen. RNA was extracted with the Qiagen RNA extraction kit as instructed with the manufacturer's guidelines. RNA was reversed transcribed to cDNA with the iScript^TM cDNA Synthesis Kit (Bio-Rad, Cat. #170-8891). RT-qPCR was performed with the Bio-Rad CFX Opus 384 Real-Time PCR System using the SsoAdvanced Universal SYBR^® Green Supermix (Bio-Rad, Cat. #1725274) Cycle conditions as follows: 95°C (30 s), 95°C (10 s), 60°C (60 s), repeat Steps 2-3 39X, 65°C (5 s), and increase to 95°C (0.5°C/cycle). Primers were purchased from Integrated DNA Technologies and sequences (5' → 3') are listed below.

18S rRNA Forward: GCTTAAATTTGACTCAACACGGGA

18S rRNA Reverse: AGCTATCAATCTGTCAATCCTGTC

PER2 Forward: TACGCTGGCCACCTTGAAGTA

PER2 Reverse: CACATCGTGAGGCGCCAGGA

U2OS Circadian Assays

U2OS Bmal1-dLuc and Per2-dLuc cells were generated as previously published.^20,21 Cells were maintained in DMEM (Gibco, Cat. #11995-065), 10% FBS (R&D Systems, Cat. #S11150), and 1% penicillin/streptomycin (10,000 U/mL, Gibco, Cat. #15140-122) at 37°C and 5% CO₂. Cells were plated at 10,000 cells/100 μL/well in a 96-well white, flat bottom plate (Corning, Cat. #3917) and incubated for 24 h. Media was changed to an explant media (DMEM (2X, Gibco, Cat. #12800-017), 1% penicillin/streptomycin (10,000 U/mL, Gibco, Cat. #15140-122), 1% GlutaMAX^TM (100X, Gibco, Cat. #35050-61), 1 mM D-Luciferin, potassium salt (PerkinElmer, Cat. #122799), 400 μM NaOH (100 mM, Fluka Cat. #72079-100ML), 10% FBS (R&D Systems, Cat. #S11150), and Cell Culture Grade Water (Corning, Cat. #25-055-CM). 2X DMEM was made mixing DMEM powder (Gibco, Cat. #12800-017), 10 mL HEPES (1 M, Gibco, Cat. #15630-080), 5 mL sodium bicarbonate (7.5%, Gibco, Cat. #25080-094), and 485 mL Cell Culture Grade Water (Corning, Cat. #25-055-CM) and sterilizing it with a vacuum filtration cup. SHP1705 or SR29065 was dissolved in DMSO (Sigma-Aldrich, Cat. #D2650-100ML) to make a stock concentration of 10 mM. SHP1705 and/or SR29065 was added at the desired concentrations to the explant media accordingly to the plate (final ≥ 0.2% DMSO). Plates were covered with a clear plastic film and maintained at 37^°C in the Tecan Infinite Pro M200 plate reader, where luminescence was read every 2 h for 120 h. Values were plotted in Excel to generate graphs. Amplitude, period, and phase were calculated with MultiCycle (Actimetrics) and 2-way ANOVA was performed for statistical analysis.

Per2::Luc Repression Assay

Wild type, Cry1/Cry2 double knockout, Cry1 knockout, and Cry2 knockout fibroblasts harboring a Per2::Luc knock-in reporter²² were plated on a white, solid-bottom 384-well plate and cultured for 2 days to reach confluency, followed by the application of 500 nL of compounds (final 0.7% DMSO). After 2 days, the medium was replaced with BrightGlo (Promega, Cat. #E2620), and luminescence was recorded in Cytation3 plate reader (BioTek).

Per2::Luc Cell Circadian Assay

5,000 wild type, Cry1/Cry2 double knockout, Cry1 knockout, and Cry2 knockout fibroblasts harboring a Per2::Luc knock-in reporter were plated on a white, solid-bottom 384-well plate and cultured for 2 days to reach confluency. The cells were treated with forskolin (final 10 µM) for 2 h. Then, the medium was replaced with explant medium [DMEM (Gibco, Cat. #12800-017) supplemented with 2% B27 (Gibco), 10 mM HEPES, 0.38 mg/mL sodium bicarbonate, 0.29 mg/mL L-glutamine, 100 units/mL penicillin, 100 µg/mL streptomycin, and 0.2 mM luciferin; pH 7.2] containing various concentrations of compounds (final 0.2% DMSO), and luminescence was recorded every 30 min for 5 days in a luminometer LumiCEC (Churitsu).

Protein Thermal Shift Assay

CRY1(PHR) and CRY2(PHR) recombinant proteins²³ were diluted to 2 µM with DSF buffer (20 mM HEPES-NaOH, 150 mM NaCl, 2 mM DTT; pH 7.5) and dispensed into a 384-well white PCR plate (Bio-Rad, Cat. #MSP3852) at 17 µL per well, followed by the application of 1 µL of compounds (final 5% DMSO). The mixtures were incubated at room temperature with gentle shaking for 60 min. 2 µL of 50X SYPRO Orange in DSF buffer (final 5X SYPRO Orange) was added, and thermal denaturation was performed using a real-time PCR detection system, CFX384 Touch (Bio-Rad).

Quantification

For reporter activity repression, the half maximal inhibitory concentrations (IC₅₀ or log[IC₅₀]) were obtained by sigmoidal dose–response fitting of dilution series data with Prism software (GraphPad Software). Circadian period was determined from luminescence rhythms by a curve fitting program MultiCycle (Actimetrics). Data from the first day were excluded from analysis because of transient changes in luminescence upon medium exchange. Concentrations for 4 hrs period-lengthening (EC_4h) were obtained by exponential growth fitting of dilution series data with Prism software. In thermal shift assays, the first derivative of the fluorescence intensity was plotted as a function of temperature (dF/dT), and the highest peak of the curve was defined as the melting temperature. The half maximal effective concentrations (EC₅₀ or log[EC₅₀]) were obtained by sigmoidal dose–response fitting of dilution series data with Prism software.

Intracranial Tumor Formation

All mouse procedures were performed under animal protocols that were approved by the Institutional Animal Care and Use Committees at University of Southern California, Children’s Hospital Los Angeles, and the University of Pittsburgh. Intracranial implantation of GSCs into 8-week-old female NOD.Cg-Prkdc^scid ll2rg^tm1Wjl/SzJ (NSG, The Jackson Laboratory) mice were performed via stereotactic method and 20,000 cells were injected 2 mm posterior to bregma, 1 mm laterally, and 2 mm deep.

In Vivo SHP1705 Administration

Seven days following cell implantation, mice were randomized and divided into treatment groups of 1% CMC vehicle control, 10 mg/kg SHP1705, or 30 mg/kg SHP1705. The reagents were administered via oral gavage once a day until the manifestation of neurological signs (ie, hunched back, severe weight loss, ataxia, and spinning). To maximize inhibition of BMAL1-CLOCK transcriptional activity, mice were treated in the morning as Bmal1 expression is at its peak in the morning in rodents.²⁴

BLI Imaging

Mice were injected intraperitoneally with 200 μL of Beetle Luciferin Potassium Salt (5 mg/mL dissolved in PBS, Promega, Cat. #E1605). Mice were anesthetized with isoflurane (5%, 100% oxygen). Fifteen minutes after injection, mice were placed in the IVIS Spectrum BLI system for imaging and anesthesia was maintained with isoflurane (2.5%, 100% oxygen). Luminescence was quantified with Living Image and data was plotted in Excel. Statistical analysis was performed with 2-way ANOVA.

Survival Analysis

Once mice began to experience neurological symptoms (ie, hunched back, significant weight loss, circling, etc.) Survival data were plotted, and statistical analysis was performed with log-rank (Mantel-Cox) test in GraphPad Prism.

SHP1705 and SR29065 Synergy Assays

Cell viability assays were conducted as mentioned above. Relative percentages of cell viability were plotted into Excel to generate a table of SHP1705 vs. SR29065 treatments. Excel sheets were uploaded onto the web application SynergyFinder3.0 (https://synergyfinder.fimm.fi/) to visualize dose–response data and calculate the synergy score.²⁵ Detection of outliers and LL4 curve fitting was performed by the program.

Results

CRY2 Is Elevated in GBM Tissues Compared to Non-cancerous Samples

We previously reported the anti-GSC effects of the CRY2-selective compound SHP1703, which showed improved efficacy compared to KL001, which does not display a CRY isoform selectivity.^14,15,17 To interrogate the molecular basis for this result, we compared clock gene levels in GBM tissue compared to non-tumor tissue in the Cancer Genome Atlas (TCGA) and CCGA datasets. We found that CRY2 mRNA levels are decreased in GBM tissue while CRY1 is increased (Figure 1B, Supplementary Figure 2A). The differences in CRY2 and CRY1 levels could explain the increased efficacy of SHP1703 compared to KL001 against GSCs and provide a rationale for CRY2-selective modulation in GBM. We hypothesized that stabilizing and elevating CRY2 protein levels could improve the ability of CRY2 to inhibit BMAL1-CLOCK transcriptional activity in GSCs and have a greater anti-GSC effect compared to targeting CRY1.

Phase 1 Study Indicates that SHP1705 Is Safe and Well-tolerated in Humans

To further improve upon the efficacy and bioavailability of existing CRY2-targeting compounds, we developed a novel scaffold, SHP1705 (Supplementary Figure 1A), based on SHP656 and SHP1703, for oral administration to patients.¹⁷

We conducted a Phase I randomized, double-blind, placebo-controlled study with healthy human volunteers. In a single ascending dose (SAD) phase of the trial, a total of 56 subjects were randomized and treated, with 42 receiving single oral doses of SHP1705 and 14 receiving the placebo. In the multiple ascending dose (MAD) phase, a total of 16 subjects were randomized and treated, with 12 receiving multiple oral doses of SHP1705 and 4 receiving placebos. Both single and multiple doses of SHP1705 were well tolerated in this study. In the SAD portion of the study, the most frequently reported treatment-emergent adverse events (TEAE) overall was headache, which occurred in 5 subjects who received SHP1705 and 2 subjects who received placebo. Diarrhea and nausea were reported by 3 subjects who received SHP1705. In the MAD portion of the study, 3 subjects who received SHP1705 had reported TEAEs and none of the TEAEs were reported by more than one subject each. Subject 2205 in the Treatment I group (25 mg SHP1705 tablets, fed) discontinued the study early because of a TEAE of ventricular extrasystoles, which was considered unrelated to the study drug. There were no apparent trends in the TEAE frequency or type by dose, dose frequency (single or multiple doses), formulation (tablet or capsule), fed or fasted state, or time of dose (morning or evening). Likewise, there were no treatment-related trends noted for changes in clinical laboratory parameters, vital signs, physical examinations, or electrocardiograms. The summary of single-dose SHP1705 pharmacokinetic (PK) parameters were measured for 500 mg oral treatment for the max concentration of SHP1705 (C_max), time of maximal SHP1705 (T_max), exposure to the SHP1705 (AUC_0-t), and half-life of SHP1705 (T_1/2) (Supplementary Figure 1B).

Analysis of biomarker data showed that expression of specific BMAL1-CLOCK target genes, PER1 and D site albumin promoter binding protein (DBP), were altered at the highest dosage of SHP1705 (500 mg) administration in the morning, supporting target engagement (Supplementary Figure 1C). There were no changes in other clock genes, such PER2 and REV-ERBα, likely due to the presence of an intact clock that takes in signals from external zeitgebers still entrains the master and peripheral clocks.¹¹ Expression levels of non-circadian, housekeeping genes were not altered (Supplementary Figure 1D) The blood transcriptome is partially under circadian control and can therefore provide information about circadian rhythmicity in the SCN and peripheral clocks.^26–29 Blood transcriptome-based biomarkers genes (Supplementary Figure 1E) were generally not significantly altered by SHP1705 dosing, although some trends may have been masked by high variability in the data.

SHP1705 Displays Significantly Improved Anti-GSC Effects Compared to Earlier CRY Activator Scaffolds

We treated numerous GSCs with differences in TMZ sensitivity (Supplementary Table 1) and found that SHP1705 is more efficacious against GSCs compared to non-malignant (NM) neuronal cultures, indicating potential for a therapeutic window. When the GSCs were differentiated (DGCs), they lost their sensitivity to SHP1705 treatment which further supports previous reports that it is truly the cancer stem cell population that is susceptible to clock perturbation (Figure 1C). Loss of stemness was indicated by lack of neurosphere formation and plate-adherence (Supplementary Figure 2B). We treated multiple GSCs, DGCs, and NM cells for 2 different days before accessing cell viability given that cell lines either respond to circadian clock compound treatment at 3 and 4 days. SHP1705 generally displayed increased efficacy against GSCs compared to earlier scaffolds KL001, SHP656, and SHP1703, with the TMZ-sensitive line MGG 31 showing the highest increased sensitivity to SHP1705 out of all the GSCs tested (Figure 1C–G, Supplementary Figure 2C–G, Supplementary Figure 3).

SHP1705 Modulates Circadian Gene Expression and Output

To confirm increased repression of BMAL1-CLOCK transcriptional activity, we examined expression of CCGs following SHP1705 treatment. SHP1705 significantly decreased expression of PER2, which is under direct control of BMAL1-CLOCK (Figure 2A) in GSCs compared to non-cancerous NM cells following 24 h of treatment. Treatment of SHP1705 decreased cell counts and neurosphere size and count in GSCs compared to NM290 (Supplementary Figure 4). We assessed circadian gene cycling in osteosarcoma U2OS luciferase reporter cells. Dosages were chosen starting at the concentration that reduced GSC viability to 50% (IC₅₀) (Figure 1C, 1G). We found that SHP1705 displayed period lengthening and arrhythmicity in higher concentrations in the U2OS Bmal1-dLuc cells (Figure 2B). SHP1705 reduced amplitude of the luciferase reporter at even the lowest doses of treatment and caused arrhythmicity in the remaining dosages in Per2-dLuc reporter cells (Figure 2C). These results indicate that SHP1705 represses GSC growth and BMAL1-CLOCK activity and, consequently, CCG output.

Figure 2.

SHP1705 directly modulates circadian gene expression. (A)  PER2 mRNA expression in NM 290, GSC 3565, and MGG 31 was measured by qPCR following 24 hrs of DMSO control or SHP1705 treatment at 1.25, 2.5, or 5 μM (n = 3 biologically independent samples). Fold change was normalized to the 18S rRNA house-keeping gene and then DMSO control. Statistical analysis was performed with 2-way ANOVA. *P < .05, **P < .01, ***P < .001 (B) U2OS Bmal1-dLuc or (C)  Per2-dLuc reporter cells were treated with increasing doses of SHP1705 (n = 3–4 biologically independent samples). Luminescence was read every 2 h for 120 h. Luciferase intensity (amplitude), period, and phase was measured and plotted below Statistical analysis was performed with 2-way ANOVA *P < .05, **P < .01, ***P < .001

SHP1705 Is Selective for the CRY2 Isoform

We next aimed to confirm SHP1705’s selectivity for CRY2, like its predecessors SHP656 and SHP1703.¹⁵ We analyzed the effects of KL001, SHP656, SHP1703, and SHP1705 on endogenous CRY1 and CRY2 activity to repress Per2 expression in fibroblasts from Per2::Luc knock-in mice. SHP1705 repressed Per2::Luc reporter activity in wild type cells containing both Cry1/2 to a greater extent than either KL001, SHP656, or SHP1703 (Figure 3A  Top Left, 3B). There was no repression of reporter activity in Cry1/2 double knockout cells suggesting that the effects of all the tested CRY activators are indeed CRY-dependent (Figure 3A  Top Right, 3B). In Cry1 knockout cells, while KL001 was less effective than intact cells, SHP1705 was still able to repress Per2 reporter activity and to a greater extent than SHP656 or SHP1703 (Figure 3A  Bottom Left, 3B). In Cry2 knockout cells, SHP1705, SHP656, and SHP1703 showed a reduced overall repression of Per2::Luc reporter activity compared to wild type calls, with no differences in activity between these compounds (Figure 3A, Bottom Right, 3B). In Cry1/2 wild type cells, SHP1705 showed a greater period lengthening effect compared to either SHP1703 or KL001 (Figure 3C, Top Left, 3D). SHP1705 was able to lengthen the period in Cry1 knockout and Cry2 knockout cells, but this was to a greater or lesser extent, respectively, compared to Cry1/2 wild type cells (Figure 3C, Bottom, 3D). The effects observed in Cry1 knockout cells could be contributed to Cry2 compensation and upregulation. We conducted a thermal shift assay to evaluate direct interaction with the photolyase homology region (PHR) of CRY1 and CRY2. We found that SHP1705 exhibited a preference for CRY2 compared to KL001 and, even, SHP1703 (Figure 3E–F). These results confirm SHP1705 activity is mediated by selectively interacting with CRY2 and highlights its increased affinity for CRY2 compared to earlier scaffolds.

Figure 3.

SHP1705 is selective for the CRY2. (A) Effects on Per2::Luc knock-in reporter activity in wild type, Cry1/Cry2 double knockout, Cry1 knockout, and Cry2 knockout MEFs. Changes in luminescence intensity compared to DMSO control are shown (n = 3–4 biologically independent samples). (B) Concentrations for 50% inhibition (plog[IC₅₀] that represents -log[IC₅₀]) are plotted. Statistical analysis was performed with 2-way ANOVA by Tukey’s multiple comparisons *P < .05, **P < .01, ***P < .001. (C) Effects on circadian period in Per2::Luc knock-in MEFs. Changes in period compared to DMSO control are shown (n = 4–12 biologically independent samples). (D) Concentrations of a 4-h period lengthening (plog[EC_4hr] that represents -log[EC_4hr]) are plotted. Statistical analysis was performed with 2-way ANOVA by Tukey’s multiple comparisons *P < .05, **P < .01, ***P < .001. (E) Interaction with CRY1(PHR) and CRY2(PHR) in vitro. Changes in denaturing temperatures of recombinant CRY(PHR) proteins in the presence of various concentrations of compounds compared to DMSO control are shown (n = 3 biologically independent samples). (F) Concentrations for 50% stabilization (plog[EC₅₀] that represents -log[EC₅₀]) are plotted. Statistical analysis was performed with 2-way ANOVA by Sidak’s multiple comparisons *P < .05, **P < .01, ***P < .001

SHP1705 Delays GBM Tumor Growth Rate and Increases Survival

To interrogate the effects of SHP1705 against GBM tumors, we established GBM patient-derived xenografts (PDX) initiated using different GSCs (Figure 4A). SHP1705 was well-tolerated and did not appear to disrupt overall circadian function as there was no evidence of significant weight gain or loss (Supplementary Figure 4). Use of hGBM8 Firefly luciferase mCherry (FMC) Mouse Passaged 1 (MP1) cells allowed for visual monitoring of tumor growth via bioluminescence imaging (BLI) and showed that, although not statistically significant, SHP1705 had a trend of decreasing the tumor growth rate in a dose-dependent manner compared to vehicle control (Figure 4B). In mice bearing tumors formed from either T387 GSC or T3565 GSC patient-derived lines (GSC 387 or 3565 transduced with a luciferase reporter), treatment with SHP1705 conferred longer survival times compared to mice given vehicle control (Figure 4C–D). Mice treated with SHP1705 survived longer than those treated with SHP656 (Figure 4D). These results provide further support for CRY2 activators as a novel paradigm for GBM treatment and SHP1705’s superior in vivo efficacy compared to earlier scaffolds.

Figure 4.

SHP1705 delays GBM tumor growth rate and increases survival. (A) in vivo pipeline for GBM PDX tumor establishment. 2 × 10⁴ GSCs containing a constitutive luciferase reporter are injected into 8-week-old NSG mice. Seven days after the surgery, mice are imaged with the Xenogen BLI system and are imaged thereafter once a week. Mice are randomized into treatment groups based on the 7-day-post-surgery luciferase signal and given vehicle control or SHP1705 once every morning, every day. Mice are euthanized once neurological symptoms appear. Image was created with BioRender. (B) Top: Bioluminescence images of representative mice injected with hGBM8 FMC MP1 and treated with for 1% CMC vehicle control, SHP1705 (10 mg/kg), and SHP1705 (30 mg/kg). Bottom: Quantification of radiance (p/sec/cm²/sr) (n = 10–11 mice/group) indicates size of tumor as a function of luciferase signal. (C) Kaplan–Meier survival curves of GBM PDX tumors established with T387 GSCs and treated with 1% CMC vehicle or 10 or 30 mg/kg of SHP1705. Statistical analysis was performed with log-rank (Mantel-Cox) test. (D) Kaplan–Meier survival curves of GBM PDX tumors established with T3565 GSCs and treated with 1% CMC vehicle, 10 mg/kg of SHP656, or 10 or 30 mg/kg of SHP1705 (n = 5 mice/group). Statistical analysis was performed with log-rank (Mantel-Cox) test.

SHP1705 Synergizes with Novel REV-ERB Agonist SR29065

Given that monotherapy treatment with SHP1705 or combination treatment using earlier REV-ERB agonists and CRY activators were effective against GSCs, we decided to interrogate the combination of SHP1705 with a novel REV-ERB agonist, SR29065.^7,30 REV-ERB agonists promote REV-ERBs’ suppression of BMAL1 transcription by increasing engagement of REV-ERBα/β with the Nuclear CoRepressor (NCoR) and Histone Deacetylase 3 (HDAC3) (Figure 5A).¹² SR29065 is a novel structure that is not related to earlier REV-ERB agonists SR9009 and SR9011, which have been shown to have REV-ERB independent effects.³¹ SR29065 is selective for REV-ERBα and has demonstrated to provide therapeutic benefit in a mouse model of autoimmune disorders.³⁰ We first tested SR29065 as a monotherapy against GSCs. Not only did SR29065 display specific effects against GSCs compared to NM cells and DGCs, like CRY activators, but it also had lower IC₅₀’s against GSCs compared to SR9009 and SR9011 (Figure 5B–E, Supplementary Figure 5A–D, Supplementary Figure 7). SR29065 altered the amplitude and resulted in arrhythmicity at higher doses in our luciferase reporter activity in U2OS Bmal1-dLuc cells, indicating it indeed affects CCG output (Supplementary Figure 8A). Although not statistically significant, SR29065 decreased the amplitude of luciferase intensity at higher dosages of treatment in the U2OS Per2-dLuc reporter cells (Supplementary Figure 8B). Overall, CRY activators were found to reduce Per2 (Figure 2C) intensity, while REV-ERB agonists reduces Bmal1 intensity of the luciferase reporters (Supplementary Figure 8B), as predicted from their mechanism of action.

Figure 5.

Novel REV-ERB agonist SR29065 has increased potency against GSCs. (A) Mechanism of REV-ERB agonists. REV-ERB agonists increase binding of REV-ERBα/β with their binding partners, NCoR and HDAC3, to increase repression of BMAL1 transcription. Cell viability following (B) SR29065, (C) SR9009, or (D) SR9011 treatment for 3 days. (E) Heatmap of summarized IC₅₀ values following CRY activator treatment in control NM cells, GSCs, and DGCs.

We then combined SHP1705 and SR29065 and treated several matched GSC and DGC lines with the combination to generate synergy plots and zero interaction potency (ZIP) synergy scores for each cell line. The ZIP scoring model is used to determine the change in potency of dose–response curves between single drugs and drug combinations. A synergy score that is less than −10 means that the 2 drugs are antagonistic. One that is from −10 to 10, the 2 drugs are likely to be additive but the closer the score is to 0, the less confidence there is with regards to synergy or antagonism. Lastly, a score that greater than 10 means that the 2 drugs are synergistic.^25,32 Most of the matched GSC and DGC cell lines showed synergy when the 2 compounds were combined (Figure 6A–E). However, the combined treatment did not display synergistic effects against DGC 4, or against either the TMZ resistant line MGG 31 or its differentiated counterpart DGC 31 (Figure 6F–H). This suggests that either monotherapy with either SHP1705 or SR29065 alone in these cells is sufficient to cause anti-GSC effects or the high sensitivity to SR29065 combined with SHP1705 yields a synergistic effect that is occurring earlier than 3 days and the remaining analyzed cells are resistant populations. Further experiments will have to be done across a time course to capture the full synergy panel across all cell lines. These results are encouraging in that they generate a hypothesis that SHP1705 could potentiate the standard of care (radiation plus TMZ) in TMZ-resistant cells/tumors too. Taken together, these results demonstrate that targeting the 2 negative limbs of the circadian clock can provide combination effects against GSCs and differentiated GBM cells.

Figure 6.

SHP1705 synergizes with novel REV-ERB agonist SR29065. (A) GSC 387, (B) DGC 387, (C) GSC 3565, (D) DGC 3565, (E) MGG 4, (F) DGC 4, (G) MGG 31, or (H) DGC 31 was treated with both SHP1705 and SR29065 for 4 days (n = 3–4 biologically independent samples). Cell viability results were plotted with SynergyFinder3.0.

Discussion

The current state of GBM treatment is discouraging, with treatments after tumor recurrence only offering mostly palliative care, as the current SOC of radiation with concomitant TMZ (chemoradiation) still only offers a median survival of 14–16 months.³³ GBM is highly infiltrative meaning surgeons must balance both maximal surgical resection without compromising neurological functions.³⁴ Additionally, the presence of GBM stem cells makes treatment even more challenging due to their existence well beyond tumor margins and resistance to chemoradiation amongst their ability to initiate new tumors, secrete angiogenic factors, and promote an immune-suppressive environment by increasing microglia infiltration.^5,11

The circadian clock has emerged in recent years as a novel therapeutic target in GBM that specifically addresses the survival of the GSC population as well targeting the tumor-supporting environment.^7–9,11,13 The core clock protein BMAL1 has an increased chromatin-binding pattern that overlaps with marks of active transcription in GSCs compared to noncancerous neural stem cells. This increased clock-mediated transcriptional control, which is not observed in other GBM or neural-related cells, may allow targeting for targeting of clock components.⁷

We probed this hypothesis by investigating the novel CRY activator SHP1705, which has been found to be safe and well-tolerated in healthy humans in Phase 1 trials (Supplement Figure 1B–D). The results from our in vitro cell viability assays suggest that, like genetic perturbation, pharmacological targeting of BMAL1-CLOCK transcriptional output selectivity attenuates GSC survival and is not observed in non-stem cell cancer cells nor non-cancerous cells (Figure 1C, Supplementary Figure 2C). This further supports our hypothesis and other reports that have demonstrated that GSCs are uniquely reliant on the clock-controlled transcription. Furthermore, this novel CRY activator scaffold has increased potency compared to earlier scaffolds, KL001, SHP656, and SHP1703 (Figure 1C–G, Supplementary Figure 2C–F) making it an ideal candidate to explore further in preclinical models in combination with standard of care (SOC) and in clinical trials with GBM patients.

To confirm the mechanism of action of SHP1705, we examined luciferase reporter activity following treatment of SHP1705. SHP1705 treatment resulted in increased repression of BMAL1-CLOCK transcriptional activity. Treatment of SHP1705 decreased the expression of BMAL1-CLOCK E-box target gene, PER2, in GSCs and had minimal effects on noncancerous NM cells (Figure 2A). Additionally, although SHP1705 had a strong effect on BMAL1-CLOCK transcriptional activity (Figure 2A), the cell viability effects of the compound occur much later (Figure 1C–G), suggesting a need for further investigation of the downstream transcriptional targets of BMAL1-CLOCK that affects GSC biology. This effect was observed directly through the dose-dependent decrease in intensity of the luciferase levels in the Per2-dLuc cells (Figure 2B). It was observed indirectly through the increase in the intensity of the luciferase reporter and elongation of periodicity of the luciferase reporter rhythms in the Bmal1-dLuc cells, which suggests that Nr1d1/2 (REV-ERBs) expression is reduced, resulting in decreased repression of Bmal1 transcription (Figure 2C).

We characterized the preferential interaction of SHP1705 with CRY2. Through a series of cell-based experiments including Per2::Luc luciferase assays and protein thermal shift assays, we found that like its predecessors, SHP1705 has isoform selectivity for CRY2 over CRY1. Furthermore, SHP1705 displays even stronger preference for CRY2 than SHP656 and SHP1703 (Figure 3). Because CRY2 mRNA levels are decreased in GBM tumors compared to normal tissue, specifically stabilizing CRY2 protein levels over CRY1 may prove to be especially beneficial for GBM treatment (Figure 1B,  Supplementary Figure 1A). Future studies may further probe the difference between CRY2 overexpression versus prevention of protein turnover via SHP1705 treatment. We anticipate, however, overexpression of CRY2 will very similarly attenuate GSC viability given that we have previously reported that CRY1 overexpression decreased GSC proliferation.⁷ Additionally, future studies may explore potential differences in response to SHP1705 in GSCs with high or low levels of CRY1/2.

We show in this study that SHP1705 treatment in vivo reduces the tumor growth rate of GBM PDX tumors and increases survival time of our mouse model (Figure 4). The lack of significant weight gain or loss in low and higher doses of SHP1705 along with the Phase 1 studies further supports that SHP1705 does not result in cytotoxicity. SHP1705 did not adversely disrupt overall circadian rhythms under the light–dark cycle condition as indicated by the lack of significant weight gain (Supplementary Figures 1B–D  and 5).^35,36

REV-ERB agonists have previously been reported to show anti-GSC efficacy. SHP1705’s single agent efficacy predicts that combining CRY activators with a REV-ERB agonist might yield synergistic effects on the circadian clock TTFL by suppressing further BMAL1 transcription and, consequently, GSC viability. We probed the novel REV-ERB agonist SR29065 in combination with SHP1705 to determine whether targeting both feedback loops of the clock will result in synergistic effects against GSCs.⁷ Like SHP1705, SR29065 displays selectivity against GSCs over normal neuronal cells and DGCs (Figure 5B–D, Supplementary Figures 6–7). In luciferase reporter assays, SR29065 displayed its mechanism of action in increasing REV-ERB-mediated repression of Bmal1 transcription through the observed decrease in luciferase intensity in both Bmal1-dLuc and Per2-dLuc reporter cells and the periodicity elongation in Bmal1-dLuc cells (Supplement Figure 8). It was previously shown that SR29065 reduced REV-ERBα target genes Bmal1 and Clock in mice livers following treatment compared to vehicle.³⁰ We found that combinations of SHP1705 and SR29065 displayed synergistic effects in multiple GSC and DGC lines, supporting our hypothesis that simultaneously targeting both feedback loops of the clock potentiate clock-mediated anti-GBM effects (Figure 6A–E). However, some GSC and DGC cell lines did not show synergistic effects, most notably in a TMZ-resistant cell line (Figure 6F–H). Combined dosing may support the growth of clock compound treatment-resistant subpopulations. Combining either clock compound with TMZ and radiation may sensitize these cells to standard-of-care therapy. Since we see that differentiated GSCs lose their specific sensitivity to clock modulation, it will likely be important to combine any potential clock-targeting therapy with radiation and TMZ, which will select for the sensitive GSC population.

These results highlight the single agent efficacy of SHP1705 and that can be combined with other clock compounds, such as SR29065, or even chemoradiation as a novel GBM treatment paradigm that targets GSCs. As GSCs present a major obstacle in the treatment of GBM by promoting tumor recurrence, clock perturbation via the administration of orally bioavailable, well-tolerated compounds can provide an easily accessible yet effective treatment option for patients with minimal side effects. As the SHP1705 and SR29065 has been shown to be as effective as genetic perturbation of BMAL1 and CLOCK, future experiments will need to be conducted to validate the effects of these compounds on the TME and angiogenesis. With increased understanding of how response to chemoradiation changes according to time-of-day administration, we can take advantage of different modalities of chronomedicine to improve upon GBM patient outcomes.^13,37–40 We can do so by combining 2 methods: targeting circadian clock components directly and timing dosing of treatments to align with the maximal effects. BMAL1-CLOCK are required not only for cell autonomous survival and stemness of GSCs but also play an integral part in angiogenesis and immunosuppression in the GBM TME.^7–9 Therefore, targeting the clock has multiple beneficial effects in GBM treatment and this now needs to be extensively explored in combination treatment paradigms with current SOC. Beyond this, chronomedicine can be utilized to widen personalized and precision-based medicine that is currently heavily reliant on genetic testing, which is not yet widely available nor affordable for all patient populations.

Supplementary material

Supplementary material is available online at Neuro-Oncology (https://academic.oup.com/neuro-oncology).

Funding

This study was supported by the NINDS F31 NS120654 (P.C.); NCI R01 CA238662 (J.N.R. and S.A.K.); Charlie Teo Foundation (S.A.K.); Japan Society for the Promotion of Science 21H04766, 24H00554, and 24H02266 (T.H.); the Astellas Foundation for Research on Metabolic Disorders (T.H.); S.A.K and J.N.R. receive sponsored research support from Synchronicity Pharma.

Conflict of interest statement. Steve A. Kay sits on Synchronicity Pharma’s Scientific Advisory Board as the Founder and Director. Both Steve A. Kay and Jeremy N. Rich receive research support from Synchronicity Pharma. There are no other potential conflicts of interests for the other authors.

Authorship Statement

P.C., S.A.K., and J.N.R. conceived the project and main conceptual ideas. P.C., Y.N., and T.H. designed and carried out the in vitro experiments. P.C., Q.W., A.H. S.M., G.K., and R.A.M. designed and carried out the animal experiments. J.C. provided SHP1705 and supervised the Phase 1 study. J.W. generated the R script for the heat maps. T.K. and L.A.S. synthesized SR29065 and provided experimental methods for the compound. S.A.K. and J.N.R. supervised the findings of the work. P.C. and S.A.K. wrote the manuscript. All authors discussed the results and contributed to the final version of the manuscript.

Data Availability

All data will be made available upon request to the corresponding author.