KRAS is one of the most frequently mutated genes in cancer, but targeting KRAS with potent small molecules has proved to be difficult. Moreover, although inhibitors of BRAF and MEK, which are downstream targets of KRAS, have been developed, they have transient benefits only in patients with melanoma who have mutated BRAF. Therefore, an effective therapy for KRAS-driven tumours remains a pressing unmet medical need.
Experiments in mice indicate that simultaneous deletion of Mek1 and Mek2 or Erk1 and Erk2 leads to organ failure. By contrast, simultaneous deletion of Braf and Craf is tolerated in adult animals (Blasco, 2011). In addition, data from mouse models show that CRAF is required for the initiation of lung cancer by oncogenic KRAS (Blasco, 2011 & Karreth, 2011). Thus, optimised CRAF inhibitors might provide a well-tolerated therapy for lung cancer as long as established lung cancers remain dependent on CRAF for viability. To address this, we carried out a large shRNA validation study in which multiple shRNAs were used to assess the dependence of a panel of human non-small cell lung cancer (NSCLC) cells on KRAS and CRAF.
Detecting and Quantifying Synergy
13 KRAS mutant and 7 KRAS wild-type NSCLC cell lines were infected with lentiviruses containing doxycycline-inducible shRNAs targeting either KRAS or CRAF and their proliferation was assessed (Figure 1). A subset of the KRAS mutant NSCLC cell lines were sensitive to depletion of KRAS. Furthermore, a significant correlation was observed between cell lines that exhibited sensitivity to KRAS depletion and those that showed sensitivity to CRAF depletion.
Figure 1. Effects of shRNA mediated depletion of KRAS and CRAF in a panel of NSCLC cell lines. (A) A panel of 20 lung cancer cells lines that are either KRAS mutant or wild type were infected with multiple doxycycline (dox)-inducible shRNAs targeting either KRAS or CRAF. The anti-proliferative effects of KRAS or CRAF depletion were determined over a 168-hour period following dox induction of shRNA expression. A minimum shRNA-induced survival percentage was calculated for each line based on the maximum anti-proliferative effect observed over the course of the experiment compared with the un-induced control, which is expressed as the mean of three sequence independent shRNAs. (B) A significant (p=0.045) increase in the anti-proliferative effects of KRAS depletion was observed in KRAS mutant cell lines. However, no effect was observed following CRAF depletion. (C) A significant correlation was evident between the cell lines sensitive to KRAS depletion and those sensitive to CRAF depletion (r=05925, p=00059).
Validating CRAF as a potential drug target using shRNA
In order to further validate CRAF as a potential target, we more closely examined the effects of shRNAs targeting CRAF in specific NSCLC cell lines.
We initially used a CRAF cDNA add back approach in cells expressing CRAF shRNAs. However, a cDNA of CRAF containing silent mutations only partially rescued the anti-proliferative effects of two shRNAs targeting endogenous CRAF (Figure 2).
Figure 2. Partial cDNA rescue of the effects of shRNA mediated depletion of CRAF in Calu1 cells. (A) Calu1 cells containing the dox-inducible shRNAs targeting CRAF were stably transduced with lentiviruses expressing CRAF cDNA that was either refractory or non-refractory to shRNA mediated depletion. Cell proliferation assays over a 168-hr period show a partial rescue of the dox-induced phenotype in the cells expressing the refractory cDNA. (B) Immunofluorescent staining of CRAF (green) and the nuclei (red) of the Calu1 cells reveals a high degree of heterogeneity in the expression levels of the CRAF cDNA before and after dox treatment.
Western blot data (not shown) and immunofluorescence data showed that CRAF protein expression levels from the CRAF cDNA lentivirus were heterogeneous in the Calu1 cell population. Moreover, additional ambiguity arose from the fact that one of the five shRNAs targeting CRAF suppressed CRAF expression without inhibiting proliferation in the KRAS dependent cell lines (Figure 3).
Figure 3. shCRAF#784 is effective at depleting CRAF but does not result in a sustained anti-proliferative phenotype. (A) Five dox-inducible shRNAs targeting CRAF were transduced into A549 cells. Four out of five of the shRNAs showed similar anti-proliferative effects. However, shCRAF#784 did not affect proliferation. (B) Western blot data indicate that shCRAF#784 was equally efficient at depleting CRAF protein compared with the other shRNAs. (C) shCRAF#784 also showed an atypical phenotype in a number of lines in the NSCLC panel irrespective of the level of depletion achieved.
An engineered A549 cell line that has reduced basal levels of CRAF expression has an increased sensitivity to shRNA-mediated CRAF depletion. This suggests that a greater reduction of CRAF than can be achieved with shRNA alone might indeed lead to substantial antiproliferation effects (Figure 4).
Figure 4. Increased sensitivity to shCRAF mediated depletion in an A549 cell line clone expressing reduced levels of CRAF protein. (A) In A549 cells, the CRAF locus was targeted with a minigene that produces a cDNA designed to be refractory to the shCRAF#911. An A549(CRAF+/+/sh911R) clone (KD) was identified that expressed the minigene at the mRNA level and expressed reduced levels of CRAF protein and was non-refractory to the shCRAF#911. Expression of the shRNAs targeting CRAF in this CRAF+/+/sh911R clone resulted in greater levels of depletion as compared to the parental line. (B) The A549 CRAF+/+/sh911R clone showed increased sensitivity to the anti-proliferative effects of the CRAF shRNAs as compared to the parental line.
Validating CRAF as a potential drug target using CRISPR-Cas9
As the shRNA data were unable to fully validate CRAF as a target in NSCLC cell lines, but the data from the A549 KD cell line suggested a promising effect, we chose to use CRISPR-Cas9 mediated gene knockout to further investigate KRAS and CRAF.
As greater sensitivity to increased CRAF depletion was observed in the A549 engineered line, it follows that complete loss of CRAF expression may induce cell death in KRAS mutant NSCLC cell lines. These data support the hypothesis that targeting CRAF in KRAS mutant cell lines could have a therapeutic effect.
Figure 5. Disruption of CRAF by sgRNA results in an anti-proliferative phenotype in A549 cells. Five guide RNA sequences that resided upstream of the kinase domain were designed using our in-house Guidebook algorithm. Each guide RNA sequence was cloned separately into an ‘all-in-one’ lentiviral vector (pLentiCRISPR (v2); Sanjana et al, 2014) which contains the individual sgRNA, Cas-9 enzyme and puromycin resistance gene. Separate viruses were produced for each individual sgRNA and titre determined using the methodology described previously (Sanjana et al, 2014). Each individual virus was spinfected separately into A549 cells, prior to selection for 7 days in puromycin. Cells were plated at 50 cells per well in a 384-well plate and observed over a 220-hour period using an IncucyteZOOMTM. A contrast based confluence algorithm (processing definition) was established for each cell line and applied to estimate cell confluence in each of the collected images. Data was exported into GraphPad Prism Software and plots of monolayer confluence vs time generated.
- These data support the hypothesis that CRAF is required for the maintenance of KRAS-driven lung cancers.
- Target validation of the shRNAs results has been confounded by the cDNA rescue approaches and one shRNA that did not produce an anti-proliferation effect despite depleting CRAF expression. Both of these results reflect established technical difficulties in using cDNA and shRNA approaches (Escheverri, 2006).
- An A549 clone with reduced levels of CRAF expression allowed the observation that greater CRAF protein depletion lead to greater anti-proliferative effects.
- CRISPR-Cas9 approaches may provide validation tools for targets, such as kinases, where RNA interference technologies provide ambiguous results.
- Blasco, R.B. et al. C-Raf, but not B-Raf, is essential for development of K-Ras oncogene-driven non-small cell lung carcinoma. Cancer Cell 19, 652-63 (2011)
- Karreth F.A. et al. C-Raf is required for the initiation of lung cancer by K-RasG12D. Cancer Discov. 1, 128–136 (2011)
- Sanjana N.E et al. Improved vectors and genome-wide libraries for CRISPR screening. Nat. Methods 11, 783-4 (2014)
- Echeverri, C.J. et al. Minimizing the risk of reporting false positives in large-scale RNAi screens. Nat. Methods 3, 777-9 (2006)