K-Ras4A splice variant is widely expressed in cancer and uses a hybrid membrane-targeting motif

Abstract

The two products of the KRAS locus, K-Ras4A and K-Ras4B, are encoded by alternative fourth exons and therefore, possess distinct membrane-targeting sequences. The common activating mutations occur in exons 1 or 2 and therefore, render both splice variants oncogenic. K-Ras4A has been understudied, because it has been considered a minor splice variant. By priming off of the splice junction, we developed a quantitative RT-PCR assay for K-Ras4A and K-Ras4B message capable of measuring absolute amounts of the two transcripts. We found that K-Ras4A was widely expressed in 30 of 30 human cancer cell lines and amounts equal to K-Ras4B in 17 human colorectal tumors. Using splice variant-specific antibodies, we detected K-Ras4A protein in several tumor cell lines at a level equal to or greater than that of K-Ras4B. In addition to the CAAX motif, the C terminus of K-Ras4A contains a site of palmitoylation as well as a bipartite polybasic region. Although both were required for maximal efficiency, each of these could independently deliver K-Ras4A to the plasma membrane. Thus, among four Ras proteins, K-Ras4A is unique in possessing a dual membrane-targeting motif. We also found that, unlike K-Ras4B, K-Ras4A does not bind to the cytosolic chaperone δ-subunit of cGMP phosphodiesterase type 6 (PDE6δ). We conclude that efforts to develop anti-K-Ras drugs that interfere with membrane trafficking will have to take into account the distinct modes of targeting of the two K-Ras splice variants.

Keywords: K-Ras; Ras; alternate splicing; oncogene; palmitoylation.

Fig. 1.

K-Ras4A is expressed in human cancer. (A, Upper) Schematic of K-Ras4A and K-Ras4B mRNA sequences with locations of primer pairs used for RT-PCR indicated. Reverse primers were designed to prime in exon 4A or the exon 3/4B splice. (A, Lower) Representative agarose gel showing RT-PCR products using 4A- or 4B-specific primer pairs and the indicated cDNA template. (B and C) Quantification of K-Ras isoform transcripts. cDNA from total RNA extracts was generated from (B) indicated cancer cell lines or (C) fresh frozen colorectal tumors, and qPCR was performed using the primer pairs indicated in A. Results are graphed as percentage of K-Ras isoform transcript relative to total K-Ras mRNA. In B, asterisks indicate cell lines with mutated KRAS. In B and C, the data shown are the means of two to three independent determinations (error bars representing SD are shown when n = 3). (D, Upper) Validation of the specificity of an anti–K-Ras4A antibody. (D, Lower) Lysates from indicated cancer cell lines (HCT116, HT29, and CaCo-2 from colorectal carcinoma and T24, J82, and 5637 cells from urothelial carcinoma) were immunoprecipitated for Ras and immunoblotted using the K-Ras4A–specific or antitotal K-Ras antibodies. (E) Lysates of 10⁷ HT29 cells were immunoprecipitated for total Ras, and the indicated amounts of eluate (total volume = 133 µL) were immunoblotted for the indicated Ras isoforms alongside a standard curve generated with bacterially expressed 6xHis-tagged Ras proteins.

Fig. 2.

K-Ras4A is targeted to the PM, even in the absence of palmitoylation. (A) Sequences of the C-terminal HVRs of four Ras proteins. The farnesylated cysteine of the CAAX motif is indicated in yellow, whereas the cysteines serving as sites of palmitoylation are in green. For K-Ras4B, the polybasic domain is in red, with blue indicating the site of phosphorylation. (B) HEK293 cells were transfected with the indicated constructs and imaged live by confocal microscopy. (C) HEK293 cells were transfected with GFP-K-Ras4A or GFP-N-Ras and treated with 2-bromopalmitate (2-BP; 50 µM) or vehicle (DMSO) for 3 h before imaging live by confocal microscopy. All images are representative of the indicated percentage of >50 live fluorescent cells examined. (Scale bar: 10 µm.)

Fig. 3.

Polybasic domains contribute to K-Ras4A PM localization. (A) Sequences of the K-Ras4B and K-Ras4A C-terminal HVRs with PBRs (PB1 and PB2) underlined. Features are indicated as described in Fig. 2A. (B and C) HEK293 cells were transfected with the indicated constructs and imaged live by confocal microscopy. ∆PB1 and ∆PB2 mutations indicate 167–170 QLQQ and 182–185 QIQQ substitutions, respectively. ∆PB1+2 indicates combined substitutions at both PBRs. All images are representative of the indicated percentage of >50 live fluorescent cells examined. (Scale bar: 10 µm.)

Fig. 4.

PDE6δ fails to extract K-Ras4A from membranes. HEK293 cells were transfected with the indicated GFP-tagged Ras proteins (Right) alone or (Left and Center) along with mCherry-PDE6δ and imaged live 24 h later. Asterisks indicate cells in which PDE6δ extracted Ras from membranes. (Scale bar: 10 µm.)

Fig. 5.

Dual membrane-targeting signals support K-Ras4A signaling. (A) HeLa cells expressing the indicated GFP-K-Ras4A constructs were serum-starved overnight and then stimulated with EGF (10 ng/mL) for 3 min. GTP-bound Ras was affinity-purified from cell lysates with GST–Ras binding domain (RBD). Ras was detected by immunoblot with a pan-Ras antibody from both GST-RBD pull downs and lysates. Lower shows quantification of GTP loading, with results presented as fold change in GTP-bound K-Ras4A normalized to expression. Values are mean ± SEM (n = 5). (B) Lysates of HeLa cells expressing the indicated constitutively active GFP-K-Ras4A12V constructs were immunoblotted for phospho-ERK, total ERK, and Ras. Lower shows quantification of phospho-ERK presented as a ratio to total ERK normalized to GFP-Kras4A expression. Values are mean ± SEM (n = 4). (C) Anchorage-independent growth of NIH 3T3 fibroblasts stably expressing the indicated constructs. Results are shown as the percentages of cells that form colonies and plotted as mean ± SEM (n = 4). (D) PC12 cells were transfected with the indicated GFP-Ras constructs, with + palm indicating the palmitoylated form and − palm indicating palmitoylation-deficient mutants 180S, 181S, or 181,4SS for K-Ras4A, N-Ras, or H-Ras, respectively. After 3 d, GFP-positive cells were scored for neurites with lengths ≥1.5 times the diameter of the cell soma. Data plotted are percentages of cells with neurites (mean ± SEM; n = 3). (E) PC12 cells were transfected with the indicated GFP-K-Ras4A constructs and scored for neurite processes as in C. *P = 0.01; **P = 0.001 (Student’s t test).