RAS Pathway Inhibitors Combined with Targeted Agents Are Active in Patient-Derived Spheroids with Oncogenic KRAS Variants from Multiple Cancer Types

Abstract

The Kirsten rat sarcoma viral oncogene homolog (KRAS) gene is among the most frequently altered genes in cancer, and the KRAS protein was long deemed undruggable. Recent strategies to target oncogenic KRAS have included both direct inhibition of the KRAS protein and indirect inhibition of its activity by targeting upstream and downstream signaling pathway mediators. A high-throughput screen of multicell-type tumor spheroids was designed to identify active combinations of targeted small molecules and KRAS pathway inhibitors. Inhibitors of the nonreceptor protein tyrosine phosphatase Src homology 2 domain-containing protein tyrosine phosphatase (SHP2) and the guanine nucleotide exchange factor Son of Sevenless homolog 1 (SOS1) were tested to evaluate indirect upstream pathway inhibition, whereas sotorasib directly inhibited the KRAS G12C variant. As single agents, sotorasib and the SHP2 inhibitor batoprotafib (TNO155) exhibited selectivity toward spheroids with KRAS G12C, whereas the SOS1 inhibitor BI-3406 showed varying activity across KRAS variants. Vertical inhibition of the rat sarcoma virus (RAS)/MEK/ERK pathway by targeting SHP2 or SOS1 and the downstream kinases MEK (trametinib) or ERK (temuterkib) was highly effective. Inhibition of upstream tyrosine receptor kinases with nintedanib in combination with batoprotafib or BI-3406 was also effective and, in combination with sotorasib, demonstrated synergy in spheroids harboring KRAS G12C. Dual inhibition of the RAS/MEK/ERK and PI3K/Ak strain transforming (AKT)/mTOR pathways by batoprotafib or sotorasib with either the mTORC1/2 inhibitor sapanisertib or the AKT inhibitor ipatasertib demonstrated combination activity, primarily in spheroids harboring KRAS G12C. The BCL-2 inhibitor venetoclax, in combination with sotorasib, batoprotafib, or BI-3406, resulted in additive and synergistic cytotoxicity. Lastly, concurrent inhibition of the KRAS pathway with sotorasib and batoprotafib demonstrated combination activity in spheroids containing KRAS G12C.

Significance: KRAS variants are oncogenic drivers for a range of human cancers. Multiple combinations of small-molecule agents that target RAS signaling were screened and reduced the viability of multicell-type tumor spheroids from a variety of human solid tumors. Combinations warranting further testing were identified.

Figure 1.

Single-agent activity of sotorasib, batoprotafib, and BI-3406 in multicell-type tumor spheroid models. A, Concentration–response graphs from sotorasib (left), batoprotafib (middle), and BI-3406 (right) as single agents in all 19 multicell-type tumor spheroid models. B, Heatmap of normalized AUC shown in A, with brown indicating low AUC values and blue indicating high AUC values. Annotation colors denote the KRAS genotype as indicated.

Figure 2.

Vertical inhibition of the RAS pathway by trametinib, temuterkib, or nintedanib in combination with sotorasib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of (A) trametinib, (B) temuterkib, or (C) nintedanib with sotorasib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 3.

PI3K/AKT/mTOR or apoptosis pathway targeted agents in combination with sotorasib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of (A) sapanisertib, (B) ipatasertib, or (C) venetoclax with sotorasib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 4.

Vertical inhibition of the RAS pathway with batoprotafib in combination with trametinib or temuterkib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of batoprotafib with (A) trametinib or (B) temuterkib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 5.

Vertical inhibition of the RAS pathway with BI-3406 in combination with trametinib or temuterkib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of BI-3406 with (A) trametinib or (B) temuterkib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 6.

Receptor tyrosine kinase inhibition with nintedanib in combination with batoprotafib or BI-3406. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of (A) batoprotafib or (B) BI-3406 with nintedanib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 7.

Dual pathway inhibition with batoprotafib in combination with sapanisertib or ipatasertib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of batoprotafib with (A) sapanisertib or (B) ipatasertib are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.

Figure 8.

Combination of venetoclax with batoprotafib or BI-3406 and the combination of batoprotafib with sotorasib. Concentration–response graphs (top, mean ± SD, n = 4 technical replicates) from combinations of (A) batoprotafib or (B) BI-3406 with venetoclax, as well as the combination of (C) batoprotafib with sotorasib, are shown, with corresponding Bliss independence scores from each combination’s concentration matrix (bottom, mean of n = 4 technical replicates) displayed numerically and as a heatmap (blue indicates synergy, gray indicates additivity, and brown indicates antagonism). The malignant cell line name, tumor type, and KRAS status are indicated above each set of graphs. WT, wild-type.