Molecular dissection of effector mechanisms of RAS-mediated resistance to anti-EGFR antibody therapy.

Monoclonal antibodies targeting the epidermal growth factor receptor (EGFR), cetuximab and panitumumab, are a mainstay of metastatic colorectal cancer (mCRC) treatment. However, a significant number of patients suffer from primary or acquired resistance. RAS mutations are negative predictors of clinical efficacy of anti-EGFR antibodies in patients with mCRC. Oncogenic RAS activates the MAPK and PI3K/AKT pathways, which are considered the main effectors of resistance. However, the relative impact of these pathways in RAS-mutant CRC is less defined. A better mechanistic understanding of RAS-mediated resistance may guide development of rational intervention strategies. To this end we developed cancer models for functional dissection of resistance to anti-EGFR therapy in vitro and in vivo. To selectively activate MAPK- or AKT-signaling we expressed conditionally activatable RAF-1 and AKT in cancer cells. We found that either pathway independently protected sensitive cancer models against anti-EGFR antibody treatment in vitro and in vivo. RAF-1- and AKT-mediated resistance was associated with increased expression of anti-apoptotic BCL-2 proteins. Biomarkers of MAPK and PI3K/AKT pathway activation correlated with inferior outcome in a cohort of mCRC patients receiving cetuximab-based therapy. Dual pharmacologic inhibition of PI3K and MEK successfully sensitized primary resistant CRC models to anti-EGFR therapy. In conclusion, combined targeting of MAPK and PI3K/AKT signaling, but not single pathways, may be required to enhance the efficacy of anti-EGFR antibody therapy in patients with RAS-mutated CRC as well as in RAS wild type tumors with clinical resistance.

59Fe-distribution in conditional ferritin-H-deleted mice.

The objective was to explore how ferritin-H deletion influences (59)Fe-distribution and excretion-kinetics in mice. Kinetics of (59)Fe-release from organs, whole-body excretion, and distribution-kinetics of intravenously injected (59)Fe trace amounts were compared in iron-deficient and iron-replete mice with (Fth(Δ/Δ)) and without (Fth(lox/lox)) conditional Mx-Cre-induced ferritin-H deletion. (59)Fe was released from spleen and liver beginning on day 2 and day 5 after ferritin-H deletion, respectively, but was not excreted from the body. Plasma-(59)Fe was cleared significantly faster in iron-deficient Fth(Δ/Δ)-mice than in iron-adequate Fth(lox/lox)-controls. (59)Fe-distribution showed a transient peak (e.g., in heart, kidney, muscle) in Fth(lox/lox) control mice, but not in ferritin-H-deleted Fth(Δ/Δ) mice 24 hours after (59)Fe injection. (59)Fe uptake into the liver and spleen was significantly lower in iron-deficient Fth(Δ/Δ) than in Fth(lox/lox) mice 24 hours and 7 days after injection, respectively, and rapidly appeared in circulating erythrocytes instead. The rate of (59)Fe release after ferritin-H deletion supports earlier data on ferritin turnover in mammals; released (59)Fe is not excreted from the body. Instead, (59)Fe is channeled into erythropoiesis and circulating erythrocytes significantly more extensively and faster. Along with a lack of transient interim (59)Fe storage (e.g., in the heart and kidney), this finding is evidence for ferritin-related iron storage-capacity affecting rate and extent of iron utilization.