Activation of Grm1 expression by mutated BRaf (V600E) in vitro and in vivo

Abstract

Our laboratory previously showed that ectopic expression of Grm1 is sufficient to induce spontaneous melanoma formation with 100% penetrance in transgenic mouse model, TG-3, which harbors wild-type BRaf. Studies identified Grm1 expression in human melanoma cell lines and primary to secondary metastatic melanoma biopsies having wild-type or mutated BRaf, but not in normal melanocytes or benign nevi. Grm1 expression was detected in tissues from mice genetically engineered with inducible melanocyte-specific BRaf^V600E. Additionally, stable clones derived from introduction of exogenous BRaf^V600E in mouse melanocytes also showed Grm1 expression, which was not detected in the parental or empty vector-derived cells, suggesting that expression of BRaf^V600E could activate Grm1 expression. Despite aberrant Grm1 expression in the inducible, melanocyte-specific BRaf^V600E mice, no tumors formed. However, in older mice, the melanocytes underwent senescence, as demonstrated previously by others. It was proposed that upregulated p15 and TGFβ contributed to the senescence phenotype. In contrast, in older TG-3 mice the levels of p15 and TGFβ remained the same or lower. Taken together, these results suggest the temporal regulation on the expression of "oncogenes" such as Grm1 or BRaf^V600E is critical in the future fate of the cells. If BRaf^V600E is turned on first, Grm1 expression can be induced, but this is not sufficient to result in development of melanoma; the cells undergo senescence. In contrast, if ectopic expression of Grm1 is turned on first, then regardless of wild-type or mutated BRaf in the melanocytes melanoma development is the consequence.

Keywords: Grm1; melanoma; mutated BRaf (V600E); p15; senescence.

Figure 1

(A) Immunohistochemistry (IHC) with Grm1 antibody. (a) Ear sample was from BRaf^CA mouse (melanocytes with wild-type Braf before Cre-mediated recombination [29]). (b) BRaf^V600E PTEN null mouse ear (conditional melanocyte-specific tyrosinase regulated Cre recombinase and Braf^CA, in PTEN null background induced by the inducer, 4-hydroxytamoxifen (TAM) and led to mutated BRaf^V600E expression only in melanocytes [27], and (c) ear from melanoma prone TG-3, with aberrant Grm1 expression with wild-type BRaf [15, 16]. The mice were 4–5 months old and the IHC were performed with 8 mice from each genotype. The percentage of positive Grm1 staining in each genotype is expressed as the average of all 8 mice ± S.D. Unbiased quantitative assessment of IHC staining is completed using a digital Aperio ScanScope GL system and Aperio ImageScope software (v 10.1.3.2028) (Aperio Technologies Inc., Vista, CA) according to the manufacturer’s protocol. (B) Hyperpigmentation in skin around the ears and tails only in 4 months old BJB mice induced with 4-hydroxytamoxifen (TAM). BJB mice are derived from crossing a BRaf^CA mouse line [loxP-BRaf(V600E)-loxP] [27] with a B6CST (Cre^ERT2) mouse line, which harbors conditionally active Cre recombinase only in melanocytes [27]. (C) (a) An ear of BJB mouse that was not induced with TAM is Grm1-negative with very low Grm1 staining likely from other cell types not melanocytes. (b) Sample from an ear of BJB mouse induced with TAM show Grm1 staining, and (c) Sample from an ear of LLA (albino version of TG-3) mouse was used as positive control for Grm1 staining. IHC staining were performed with ears from five mice in each group and the percent of Grm1 positive is the average of all five mice ± S.D. Unbiased quantitative assessment of IHC staining is completed using a digital Aperio ScanScope GL system and Aperio ImageScope software (v 10.1.3.2028) (Aperio Technologies Inc., Vista, CA) according to the manufacturer’s protocol.

Figure 2

(A) BRaf^V600E expression in the ears of BJB mice was confirmed by Western immunoblot. BRaf^V600E/PTEN null transgenic mouse used as positive control (+), BJB 91 was not treated with TAM and used as negative control. The same membrane was used to probe for p16/INK4a, and α-tubulin used as loading control. (B) Same set of protein lysates from (A) was used in Western immunoblots for Grm1 expression in the ears of BJB mice. LLA transgenic mouse was used as positive control, BJB 91 was not treated with TAM and used as negative control, α-tubulin used as loading control. (C) PTEN expression was assessed by Western immunoblot, showing preserved expression in both TAM treated and non-treated samples. α-tubulin used as loading control. (D) Phosphorylated PDK1 [pPDK1] expression was assessed by Western immunoblots, and demonstrated similar expression in TAM treated and untreated samples. Total PDK1 [tPDK1] was used as a loading control. (E) β-galactosidase staining was used to assess cell senescence using frozen sections of 14-month-old, TAM-treated BJB mouse ears. BJB TAM induced samples (top panels) show positive staining, similar to etoposide-treated MCF7 positive control (bottom right panel), while 14-month-old BJB mouse ear not induced with TAM did not show positive β-galactosidase staining (bottom left panel).

Figure 3. Melan-a-muBRaf clones induce Grm1 expression and downstream signaling pathways

(A) Expression of BRaf^V600E in several stable melan-a-muBRaf clones was confirmed by Western immunoblot. PCIneo was used as vector control. (B) Several stable melan-a-muBRaf clones also demonstrated Grm1 expression by Western immunoblot. melan-a parent was used as negative control and α-tubulin as loading control. (C) Elevated basal levels of pAKT (T308) and (D) pERK1/2 were detected in two randomly selected melan-a-muBRaf clones 11 and 25 compared to the parental melan-a or MASS20 (melan-a with exogenous Grm1). Total AKT [tAKT] and total ERK1/2 [tERK] were used as loading controls. Blots were quantitated and shown as fold difference over parent control normalized to loading control.

Figure 4. Concurrent alteration in BRaf function/expression and Grm1 expression

(A) A reduction in Grm1 expression was detected in representative melan-a-muBRaf clone 25 when treated with a small molecule inhibitor PLX4032 for 24 hrs at 0.1 μM or 0.3 μM, or for 48 hrs at 0.3 μM. MASS20 was used as positive control for Grm1 expression, melan-a parent as negative control, and α-tubulin as loading control. Blot was quantitated as shown as a fold difference over non-treated sample normalized to loading control. (B) Melan-a-muBRaf (V600E) cells were transfected with Tet^R and siBRAF^V600E-Tet^O plasmids to generate inducible melan-a siBRaf^V600E stable clones. Doxycycline [doxy] was applied for four days to induce siRNA expression. Western immunoblots confirmed suppression of BRAF^V600E induced by doxy. Control sample [ctl] not treated with doxy; α-tubulin was used as loading control. Blots were quantitated and shown as fold difference over non-treated sample normalized to loading control. (C) Two representative clones showed subsequent decrease in Grm1 expression with knockdown of mutated BRaf^V600E in Western immunoblot. Blots were quantitated and shown as fold differences over parent control normalized to α-tubulin as loading control. (D) Melan-a-muBRaf (V600E) clones were transfected with Tet^R and siWTBRaf-Tet^O plasmids to generate inducible melan-a siWTBRaf stable clones. Doxycycline [doxy] was applied for four days to induce siRNA activity. Western immunoblots confirmed suppression of wild-type BRaf [WTBRaf] induced by 5 μg/ml or 4 μg/ml doxy. Control sample [ctl] not treated with doxy, α-tubulin used as loading control. Blot was quantitated and shown as fold differences over parent control normalized to loading control. (E) Silencing of WTBRaf led to a similar degree of down-regulation of Grm1 expression compared to MASS20 positive control. α-tubulin used as loading control. Blots were quantitated and shown as fold differences over parent control normalized to loading control.

Figure 5. PTEN functionality may be altered in melan-a-muBRaf clones

(A) Western immunoblot showed PTEN expression in stable melan-a mutated BRaf^V600E clones that are Grm1-positive, melan-a parent was used as control and α-tubulin as loading control. Blots were quantitated and shown as fold difference over parent control normalized to loading control. (B) Western immunoblot shows expression of pPDK1 in melan-a mutated BRaf^V600E clones and very little in melan-a parent. Membrane was stripped and re-probed with total PDK1 as loading control. Blots were quantitated and shown as fold difference over parent control normalized to loading control. (C) Five different melan-a mutated BRaf^V600E clones treated with increasing concentrations of PTEN inhibitor bpV (phen) (1 μM, 2 μM and 5 μM) for 15 and 30 minutes showed pAKT levels increasing in a dose dependent manner. Membrane was stripped and re-probed with total AKT as loading control. An example of melan-a-muBRaf clone-22 is shown. Blots were quantitated and shown as fold difference over loading control.

Figure 6. Grm1 induced by mutated BRaf is functional

(A) L-quisqualate [Q] (10 μM) used to stimulate a representative melan-a-muBRaf clone led to increased levels of downstream pAKT and pERK1/2. Membrane was stripped and re-probed with total AKT or total ERK as loading controls. Blots were quantitated and shown as fold differences over loading control. (B) Western immunoblot showed expression of phosphorylated IGF-1R [pIGF-1R] is expressed in melan-a-muBRaf clones, but not in melan-a parent. Membrane was stripped and re-probed with total IGF-1R as loading control. Blots were quantitated and shown as fold difference over parent control normalized to loading control. (C) Melan-a-muBRaf was treated with Grm1 agonist, Q, for up to 30 min. pIGF-1R levels were modulated by Grm1 stimulator. The membranes were stripped and re-probed with total IGF-1R as loading control; blots were quantitated and shown as fold differences over treated sample at time 0 normalized to loading control.

Figure 7. Activation of PI3K/AKT pathway downstream of Grm1 induced by mutated BRAF^V600E is mediated through IGF-1R via Src

(A) Western immunoblot shows knockdown of pAKT expression in a representative melan-a-muBRaf clone 11 after treatment with IGF-1R inhibitor OSI-906 at increasing concentrations (10 μM or 2 5 μM) for 24 or 48 hours. Membrane was stripped and re-probed with total AKT as loading control; blots ere quantitated and shown as fold differences over non-treated control normalized to loading control. We performed this study with three different mutated BRaf clones (clone 11, 20 and 25) treated at three independent times. (B) Western immunoblot shows a representative melan-a-muBRAF clone 25 treated with an Src inhibitor, PP2 at 5 μM or 10 μM for 15 minutes, with subsequently lower expression of pAKT. The membrane was stripped and re-probed with total AKT as loading control; blots were quantitated and shown as fold differences over non-treated control normalized to loading control. Three independent melan-a-muBRaf clones were used (clones 11, 21 and 25) and the study was performed at three different times.

Figure 8

(A) TGFβ induced p15 in mutated BRaf but not wild-type BRaf mice. (A) Western immunoblots on ear protein lysates prepared from three-or ten-month old BJB mice, either not treated (NT) or treated with 4-hydroxytamoxifen (TAM, 15 mg/ml) and probed with antibodies to p15, α-tubulin, TGFβ and β-actin. (B) Western immunoblots on ear protein lysates prepared from three- or ten-month old TG-3 mice, and probed with antibodies to p15, α-tubulin, and TGFβ. All blots were quantitated and shown as fold differences normalized to loading control.