Checkpoint kinase 1 inhibitor + low‐dose hydroxyurea efficiently kills BRAF inhibitor‐ and immune checkpoint inhibitor‐resistant melanomas

Treatment of melanomas with targeted and immunotherapies has proven effective, but resistance to both treatments is a common outcome leaving a high proportion of patients without effective alternative treatment options. Replication stress is a common feature of melanomas, and this is effectively targeted using a combination of checkpoint kinase 1 (CHK1) inhibitor and low‐dose hydroxyurea (LDHU). This combination also promotes inflammatory and anti‐tumour immune responses in vivo. Melanoma cell lines resistant to BRAF inhibitor (BRAFi) or immune checkpoint inhibitors (ICI) retain their sensitivity to CHK1i + LDHU, with sensitivity similar to that of parental tumours. In vivo, BRAFi‐resistant and BRAFi‐sensitive parental tumours produce an identical immune response with treatment.


ZENG et al.
We show that resistance to either targeted therapy or immunotherapy had a little effect on sensitivity to our CHK1 inhibitor + low-dose HU treatment (CHK1i + LDHU) in vitro and in vivo and had no negative effect on the immune response promoted by this treatment in vivo.
To generate BRAFi-resistant cells, YUMMUV1.7 and A2058 cells were grown continuously in increasing concentrations of dabrafenib up to 4 μM.Once cells were proliferating normally in 4 μM dabrafenib, they were grown on drug for another 2 weeks, and then a dose response of dabrafenib was performed to confirm resistance status.

| Dose response of sensitive/resistant lines
All cell lines were seeded in 96-well plates and incubated in a 37°C incubator overnight, then treated with increasing concentrations of the dabrafenib or CHK1i (SRA737, Sierra Oncology) in combination with 0.2 mM HU (Sigma).Cells were assessed for viability after treatment for 3 days using resazurin (Sigma).Assays were performed in quadruplicate.

Significance
The lack of effective alternative treatments for patients who develop resistance to targeted and/or immunotherapies is a significant problem for a large proportion of melanoma patients.CHK1i + LDHU is an effective treatment in a high proportion of melanomas and retains this efficacy in current treatment-resistant tumours.This suggests that CHK1i + LDHU could be an effective alternative therapy in otherwise treatment-resistant melanoma patients.

TA B L E 1
Characteristic of the melanoma cell lines used.

| Mouse tumour assays
Experiments were performed with approval from The University of Queensland Animal Ethics Committee (2017/AE000211).Syngeneic mouse YUMMUV1.7 tumours were established in C57BL/6J mice and treated with CHK1i (SRA737) + LDHU as described previously (Proctor et al., 2021).Tumours were excised on Day 10 after treatment and processed for staining with Live/Dead Aqua (Thermo Human tumours were xenografted onto nude mice, and tumour growth was followed as described previously (Oo et al., 2019).Probes) to mark dead cells was also added, and cells were imaged every 4 h for up to 5 days using an IncuCyte S3 live cell imaging system.Analysis was performed using IncuCyte software to identify total number of cells and percentage of cells positive for Sytox Green staining.

| Statistical analysis
All statistical analyses were done using GraphPad Prism 9. Bar graphs display mean values and standard deviation.For pairs of sensitive/resistant lines, unpaired Student's t-test were performed with p value < .05.

| RE SULTS AND D ISCUSS I ON
We assessed the efficacy of CHK1i + LDHU treatment in a panel of melanoma cell lines with resistance to BRAF inhibitor (BRAFi) acquired either in vitro or from patients who have developed resistance to indicated treatment (Table 1).These represent different mechanisms of acquisition and resistance.This assessment included four sets of cell lines with pre-treatment and resistant pairs (or multiple resistant clones for SKMEL28).The parental cell lines were shown to be sensitive to BRAFi (dabrafenib) with IC 50 < 0.1 μM, whereas the resistant lines were >10 fold less sensitive (Figure 1).
When sensitivity to CHK1i (SRA737) + LDHU was assessed in these BRAFi-sensitive and BRAFi-resistant pairs, both lines were inhibited similarly (Figures 1, 2a).To further define whether tumour cells from patients that had acquired resistance to BRAFi or ICIs are sensitive to CHK1i + LDHU, six more cell lines were assessed, including two melanoma cell lines derived from patients with acquired BRAFi resistance and four cell lines from patients who had acquired resistance to different ICIs (Table 1).All of these cell lines displayed similar sensitivity to CHK1i + LDHU (Figure 2b,c), and five out of six displayed IC 50 values around 2 μM for CHK1i (Figure 2d-f; Data S1) which is below the 6 μM Cmax found in patients (Plummer et al., 2019).One cell line, WMD-009, did not achieve 50% growth inhibition (Figure 2b).
To assess whether patient-derived BRAFi-sensitive and BRAFi- When the R2 BRAFi-resistant tumours were treated with the normal regimen of CHK1i + LDHU, it was found to have sensitivity to the treatment (Figure 3).
treatment in both BRAFi-sensitive and BRAFi-resistant models (Fig- ure 3c).The NK cell population (CD45 ) also increased to a similar degree in both models, whereas there was no effect on the recruitment of CD4 + or CD8 + T cell populations to the tumour sites, as reported previously (Proctor et al., 2021).The BRAFi-resistant YUMMUV1.7 line retained its resistance to BRAFi during CHK1i + LDHU treatment, as a line grown from the remaining tumour material after treatment showed reduced sensitivity to killing by BRAFi (dabrafenib), similar to BRAFiresistant A2058 line (Data S1; Figure 4).
Together, these data indicated that the majority of melanoma cell lines tested are sensitive to the CHK1i + LDHU combination at concentrations of CHK1i that are readily achievable in patients.
Sensitivity to the combination is not significantly affected by acquired resistance to either BRAFi, BRAFi + MEKi combinations, or resistance to immunotherapy.Resistance to BRAFi has been acquired through a variety of mechanisms that result in reactivation of the MAPK signalling pathway (Czarnecka et al., 2020) and increased replication stress (Yang et al., 2021).This suggests that BRAFi resistance may increase sensitivity to treatments targeting the replication stress such as the CHK1i + LDHU combination, although that was not observed here.Several mechanisms of resistance to immunotherapy have been identified including reduced MHC I expression (Lee et al., 2020).The cell lines from ICI-resistant patients used here have reduced MHC I expression and were associated with increased AXL expression and dedifferentiation.However, inhibition of AXL increased DNA damage and replication stress (Ramkumar et al., 2021).Thus, the mechanism by which MHC I expression is reduced is more likely to be involved in tolerance of replication stress which could reduce sensitivity to treatments targeting replication such as WEE1, ATR and CHK1 inhibitors, although it did not reduce sensitivity to CHK1i + LDHU.The majority of BRAFi and ICI treatment-resistant tumours demonstrated to be sensitive to the CHK1i + LDHU combination in this study mirrors our previous studies that reported >70% of melanomas were sensitive to this combination (Oo et al., 2019).This high proportion of sensitive melanomas to CHK1i + LDHU and their remained sensitivity after acquiring resistance to current clinical treatments suggests that CHK1i + LDHU combination could be a useful treatment in patients that have exhausted current therapy options.
Time-lapse cell viability killing assay YUMMUV1.7 (Parent and DabR Tx, a BRAFi-resistant line grown from the remaining tumour material after CHK1i + LDHU treatment) and A2058 (Parent and DabR, a BRAFi-resistant line generated inhouse) cells were seeded into 96-well plates and treated with or without 1 μM BRAFi (dabrafenib).Sytox Green (250 nM; Molecular resistant melanomas were sensitive to CHK1i + LDHU treatment in vivo, SUM-030 parental and R2 resistant lines were grown as F I G U R E 1 Dose response of the indicated melanoma cell linesdabrafenib-sensitive parental and acquired resistance pairs-to dabrafenib or CHK1i (SRA737) with a constant 0.2 mM HU: (a) YUMMUV1.7 (mouse), (b) A2058 (human), (c) WM164 (human).Cell viability was measured by resazurin assay, and the data are the mean and SD of triplicate determinations.These are representative of duplicate experiments.line grew as a tumour, and the parental line only produced a barely palpable tumour by 60 days after implanting.

Finally
, we have previously reported that CHK1i + LDHU triggers an immune response in vivo characterised by increased natural killer (NK) cells and NK-T cell (NKT) infiltration into syngeneic mouse melanomas with treatment (Proctor et al., 2021).Using the BRAFi-sensitive and BRAFi-resistant YUMMUV1.7 pairs developed in-house, CHK1i + LDHU treatment inhibited tumour growth in an identical manner regardless of their sensitivity to BRAFi (Figure 3a,b) and promoted a very similar pattern of immune cell infiltration (Figure 3c).As observed previously, the major immune response to CHK1i + LDHU was the high level of NKT cells infiltrating into tumours, identified as CD45 + CD3 + TCRβ + NK1.1 + CD4 − CD8 − NKp46 − .This population increased significantly after CHK1i + LDHU F I G U R E 2 Dose response of the indicated cell lines-(a) SKMEL28 human melanoma cell lines, either parental or three acquired BRAFi-resistant clones, SMU-030 patient-derived cell lines pre-treatment (Pre) and two clones from progression on combination BRAFi + MEKi (Table 1), (b) patient-derived BRAFiresistant melanoma cell lines, (c) patientderived ICI-resistant lines-to CHK1i (SRA737) + 0.2 mM HU. Cell viability was measured by resazurin assay, and the data are the mean and SD of triplicate determinations.These are representative of at least three separate experiments.(d-f): IC50 values from the experiments shown in Figures 1, 2a-c.

F
I G U R E 3 SMU-030 R2 human BRAFi-resistant melanoma was established in immune compromised mice.When a tumour had reached >100 mm 3 , treatment as in A was commenced (Day 24).Tumour growth is presented as fold increase over Day 24.The * indicated p value < .05 at individual time points.a): Syngeneic mouse melanoma YUMMUV1.7 was established in immune competent mice, either BRAFi-sensitive parental (Par) or resistant (Res).Once a tumour had reached >100 mm 3 , mice were treated with vehicle control or CHK1i (SRA737) + LDHU, as previously described (Proctor et al., 2021).Tumour growth was measured and reported, and the fold increases over Day 8 (first day of treatment).The * indicated p value < .05 at individual time points.(b): The tumour volume at the end of treatment.(c): The immune cell profile in the tumour microenvironment was assessed 10 days after drug treatment started.The abundance of the indicated major immune cell types assessed, CD4 + and CD8 + T cells, NK and NKT cells, was shown as number of immune cells/10 6 tumour cells for each mouse.