K27-linked ubiquitination of BRAF by ITCH engages cytokine response to maintain MEK-ERK signaling

Abstract

BRAF plays an indispensable role in activating the MEK/ERK pathway to drive tumorigenesis. Receptor tyrosine kinase and RAS-mediated BRAF activation have been extensively characterized, however, it remains undefined how BRAF function is fine-tuned by stimuli other than growth factors. Here, we report that in response to proinflammatory cytokines, BRAF is subjected to lysine 27-linked poly-ubiquitination in melanoma cells by the ITCH ubiquitin E3 ligase. Lysine 27-linked ubiquitination of BRAF recruits PP2A to antagonize the S365 phosphorylation and disrupts the inhibitory interaction with 14-3-3, leading to sustained BRAF activation and subsequent elevation of the MEK/ERK signaling. Physiologically, proinflammatory cytokines activate ITCH to maintain BRAF activity and to promote proliferation and invasion of melanoma cells, whereas the ubiquitination-deficient BRAF mutant displays compromised kinase activity and reduced tumorigenicity. Collectively, our study reveals a pivotal role for ITCH-mediated BRAF ubiquitination in coordinating the signals between cytokines and the MAPK pathway activation in melanoma cells.

Fig. 1

ITCH promotes K27-linked polyubiquitination of BRAF. a BRAF was ubiquitinated via the K27 and K29 linkages. Immunoblot (IB) analysis of whole-cell lysates (WCL) and immunoprecipitates (IP) derived from 293T cells transfected with Flag-BRAF and the indicated HA-ubiquitin (UB) constructs. b ITCH promoted BRAF ubiquitination in cells. IB analysis of WCL and Ni-NTA (Ni-nitrilotriacetic acid) affinity precipitates derived from 293T cells transfected with HA-BRAF, His-UB, and the indicated Flag-tagged NEDD4 family constructs. c ITCH promoted BRAF ubiquitination in vitro. Flag-BRAF protein immunopurified from 293T cells was incubated with GST-ITCH proteins, as well as E1, E2, and ubiquitin proteins at 30 °C for the indicated time period before SDS-PAGE and IB analyses. d WT- but not C832S(CS)-ITCH promoted BRAF ubiquitination in vitro. e BRAF was ubiquitinated via the K27- and K29 linkages. IB analysis of WCL and Ni-NTA affinity precipitates derived from 293T cells transfected with Flag-BRAF and the indicated Myc-ITCH, WT, or K-only His-UB constructs. f K27R-ubiquitin failed to facilitate BRAF ubiquitination. IB analysis of WCL and Ni-NTA affinity precipitates derived from 293T cells transfected with Flag-BRAF and the indicated Myc-ITCH, WT, or KR His-UB constructs. g BRAF was polyubiquitinated via the K27-linkage. IB analysis of WCL and Ni-NTA affinity precipitates derived from 293T cells transfected with Flag-BRAF and the indicated Myc-ITCH, His-WT-UB, or His-K27-UB constructs. h ITCH promoted BRAF K27-linked polyubiquitination in vitro. i K27R-UB failed to facilitate ITCH-mediated BRAF polyubiquitination in vitro. j A schematic illustration of the lysine residues identified as the BRAF ubiquitinated sites. k BRAF-KR mutants displayed compromised ubiquitination in cells. IB analysis of WCL and Ni-NTA affinity precipitates derived from 293T cells transfected with the indicated Flag-BRAF, Myc-ITCH, and His-UB constructs. l BRAF-KR mutants displayed compromised ubiquitination in vitro. Various Flag-BRAF WT and KR mutants immunopurified from 293T cells were incubated with bacterially purified GST-ITCH and the E1, E2, and ubiquitin proteins as indicated. The reaction was performed at 30 °C for 60 min and followed by SDS-PAGE and IB analyses

Fig. 2

BRAF interacts with ITCH through its kinase domain. a ITCH bound to RAF isoforms. Immunoblot (IB) analysis of whole-cell lysates (WCL) and immunoprecipitates (IP) derived from 293T cells transfected with HA-ITCH together with the indicated Flag-RAF constructs. b ITCH promoted ubiquitination of RAF proteins in cells. IB analysis of WCL and Ni-NTA (Ni-nitrilotriacetic acid) affinity precipitates derived from 293T cells transfected with the indicated Myc-ITCH, His-UB, and Flag-RAF constructs. c Endogenous BRAF bound to ITCH. IB analysis of WCL and anti-BRAF IP derived from WM1346 cells. d In vitro transcribed and translated HA-BRAF (IVT-HA-BRAF) bound to purified recombinant GST-ITCH. e ITCH bound to the BRAF kinase domain. IB analysis of WCL and IP derived from 293T cells transfected with HA-ITCH together with the indicated Flag-BRAF constructs. f BRAF bound to the WW domain of ITCH in vitro. In vitro transcribed and translated HA-BRAF was incubated with the indicated recombinant GST proteins before washing and being resolved by SDS-PAGE. g A schematic illustration of the proline-rich motifs in the BRAF kinase domain that might mediate ITCH–BRAF interaction. h A structural illustration of the putative spatial location of the candidate ITCH-binding motifs in a reported BRAF crystal structure (PDB ID: 1UWH). i Mutation of ITCH-binding motifs abrogated ITCH–BRAF interaction in cells. IB analysis of WCL and IP derived from 293T cells transfected with HA-ITCH together with the indicated Flag-BRAF constructs. j Mutation of ITCH-binding motifs abolished ITCH-mediated BRAF ubiquitination in cells. IB analysis of WCL and Ni-NTA affinity precipitates derived from 293T cells transfected with the indicated Myc-ITCH, His-UB, and Flag-BRAF constructs

Fig. 3

Depletion of ITCH attenuates BRAF activity and suppresses melanoma cell growth. a Immunoblot (IB) analysis of whole-cell lysates (WCL) derived from WT and Itch^-/- MEFs. b IB analysis of WCL derived from WT and Itch^−/− MEFs treated with 50 ng^∙ml⁻¹ TNFα for the indicated time period. c, d IB analysis of WCL derived from WM3918 (c) and WM1346 (d) cells infected with the indicated shITCH lentiviral constructs; a scrambled shRNA construct (shScr) was used as the negative control. e IB analysis of WCL derived from WM1346 cells stably expressing EZ-Tet-pLKO-shITCH which allows doxycycline-induced depletion of ITCH. Cells were treated with 1 μg^∙ml⁻¹ doxycycline (Dox), as indicated for 48 h before harvest. f Depletion of ITCH abolished endogenous BRAF ubiquitination. IB analysis of WCL and anti-BRAF immunoprecipitates (IP) derived from WM1346 cells infected with shScr or shITCH lentiviral construct. g, h WT- but not enzymatic dead C832S (CS)-ITCH rescued decreased p-MEK and p-ERK upon ITCH depletion. IB analysis of WCL derived from MEFs (g) and WM1346 (h) cells infected with the indicated viral constructs. Murine ITCH cDNA was used in the reconstitution experiments. i–k WM1346 cells generated in (d, e, h) were subjected to clonogenic survival assays in RPMI-1640 media supplemented with 10% FBS for 14 days. Crystal violet was used to stain the formed colonies (i–k), and representative pictures were shown from three independent experiments. l, m Tumor growth curves (l) and pictures at the end point (m) for the xenograft experiments with the WM1346 cells generated in (e) were inoculated subcutaneously. In each flank of six nude mice, 1 × 10⁶ cells were injected. The visible tumors were measured at the indicated days. The arrow indicates the day doxycycline administration was started. Error bars represent ± SEM (n = 6). IB analysis of WCL derived from the tumor samples generated in (l, m)

Fig. 4

Proinflammatory cytokines engage ITCH to promote BRAF ubiquitination and subsequent activation. a, b Immunoblot (IB) analysis of whole-cell lysates (WCL) derived from WM3918 (a) and WM1346 (b) cells treated with 50 ng^∙ml⁻¹ TNFα for the indicated time period. c Endogenous BRAF ubiquitination was increased after TNFα treatment. IB analysis of WCL and anti-BRAF immunoprecipitates (IP) derived from WM3918 cells treated with TNFα for 20 min. d IB analysis of WCL derived from shScr- and shITCH-WM3918 cells. e Endogenous BRAF ubiquitination was elevated in WT, but not Itch^-/-. f IB analysis of WCL derived from WT and Jnk1^-/-;Jnk2^-/- MEFs. g IB analysis of WCL derived from WM3918 cells co-infected with shJNK1 and shJNK2 lentiviral constructs. h WM3918 cells were subjected to cell proliferation assays for 12 days. Cell viability was determined at the indicated time points. The viability was calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test. i Illustration of the coculture experiment of M2-differentiated THP1 cells and melanoma cells. j, k Coculture with M2- differentiated THP1 cell stimulated WM3918 cells growth. The colony numbers (j) were calculated as mean ± SD (n = 3), *P < 0.05; Student’s t test (k). l Coculture with M2-differentiated THP1 cell activated the MEK/ERK signaling in WM3918 cells. IB analysis of WCL derived from WM3918 cells of the coculture experiment as described in (j, k). m IB analysis of WCL derived from melan-a cells stably expressing GFP, WT-ITCH of the constitutively active 3D-ITCH. n, o Melan-a cells generated in (m) were subjected to clonogenic survival assays without TPA for 14 days. The colony numbers were calculated as mean ± SD (n = 3), *P < 0.05; Student’s t test (o). p, q Melan-a cells generated in (m) were transduced with shScr or shPten lentiviral constructs followed by soft agar colony-formation assays without TPA for 21 days (p). The colony numbers were calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test (q)

Fig. 5

Ubiquitination-deficient BRAF mutants exhibit compromised activity. a Immunoblot (IB) analysis of whole-cell lysates (WCL) derived from HEK293 cells transfected with the indicated Flag-tagged WT-BRAF or BRAF-KR mutants. b In vitro kinase assays showing that immunopurified Flag-WT-BRAF and BRAF-KR proteins displayed similar activity to promote the phosphorylation of GST-MEK1. BRAF/3KR: K698R + K699R + K700R; BRAF/5KR: K164R + K473R + K570R + K698R + K700R. c, d Kinase kinetics of WT-BRAF and 5KR-BRAF (c) as well as unmodified and ubiquitinated BRAF proteins (d). Initial rates were measured at various concentrations of ATP using the continuous assay. The rates were replotted against substrate concentration and fit to the Michaelis–Menten equation. e IB analysis of WCL derived from HEK293 cells transfected with the indicated Flag-tagged WT-BRAF or ITCH-binding-deficient BRAF mutants. f In vitro kinase assays showing that immunopurified Flag-WT-BRAF and ITCH-binding-deficient BRAF proteins displayed similar activity to promote the phosphorylation of GST-MEK1. g IB analysis of WCL derived from HEK293 cells stably expressing WT-BRAF or 5KR-BRAF treated with 50 ng^∙ml⁻¹ TNFα for the indicated time period

Fig. 6

Ubiquitination of BRAF disrupts the inhibitory interaction between BRAF and 14–3–3 proteins. a Ubiquitinated BRAF displayed reduced binding with 14–3–3 in vitro. In vitro ubiquitination assays using immunopurified Flag-BRAF proteins were performed followed by incubation of Flag-BRAF with 14–3–3. The anti-Flag immunoprecipitates (IP) were subjected to SDS-PAGE and immunoblot (IB) analysis. b 5KR-BRAF displayed stronger binding to 14–3–3 compared with WT-BRAF. c Binding between endogenous BRAF and 14–3–3 was reduced upon TNFα treatment. d S365A-BRAF and S729A-BRAF mutants exhibited a distinct affinity with 14–3–3 after ubiquitination. e Illustration of the proposed models for the disruption of BRAF-14–3–3 interaction by K27-linked ubiquitination. f PPP2CA and PPP2R2A subunits of the PP2A protein family specifically interacted with BRAF. g Ubiquitinated BRAF displayed increased binding with PPP2R2A in vitro. h In vitro phosphatase assay showing that compared with the catalytic PPP2CA subunit, the PP2A complex was more active in dephosphorylating p-S365-BRAF. Immunopurified Flag-BRAF proteins were incubated with the indicated PP2A phosphatases for the indicated time period before SDS-PAGE and IB analysis. i Phosphatase kinetics of PPP2CA and the PP2A complex in dephosphorylating BRAF. Initial rates were measured at various concentrations of Flag-BRAF protein using the continuous assay. The rates were replotted against substrate concentration and fit to the Michaelis–Menten equation. j In vitro phosphatase assay showing that ubiquitinated BRAF was a better substrate for the PP2A complex. Immunopurified Flag-BRAF were subjected to in vitro ubiquitination as described in Fig. 1c. Unmodified and ubiquitinated proteins were incubated with the indicated PP2A phosphatases for 60 min. The free phosphate was measured using the malachite green detection reagent, normalized, and calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test. k, l IB analysis of WCL derived from WM3918 cells infected with the indicated shPPP2CA (k) or shPPP2R2A (l) lentiviral constructs. m IB analysis of WCL derived from WM3918 cells infected with shScr or shPPP2R2A lentiviral constructs. n IB analysis of WCL derived from HEK293 cells transfected with the indicated Flag-tagged WT-BRAF or BRAF mutants

Fig. 7

Ubiquitination of BRAF serves to maintain BRAF oncogenic function in melanoma cells. a Immunoblot (IB) analysis of whole-cell lysates (WCL) derived from WM1346 cells stably expressing EV, WT-BRAF, or the 5KR-BRAF mutant. The stable cell lines were further infected with the indicated lentiviral vectors to deplete endogenous BRAF. *The BRAF cDNA used in this figure was mutated to confer resistance to the shBRAF lentiviral vector. b IB analysis of anti-BRAF immunoprecipitates (IP) BRAF-depleted WM1346 cells stably expressing WT-BRAF or the 5KR-BRAF mutant. c WM1346 cells generated in (a) were subjected to cell proliferation assays. Cell viability was calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test. d WM1346 cells generated in (a) were seeded (300 cells per well) for clonogenic survival assays. e, f WM1346 cells generated in (a) were subjected to Matrigel invasion assays (e). The invaded distance was calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test (f). g, h WM1346 cells generated in (a) were subjected to endothelial cell invasion assays (g). The invaded distance was calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test (h). Scale bar: 50 μm. i, j WM3918 cells generated in Supplementary Fig. 11a were subjected to clonogenic survival assays without or with M2-differentiated THP1 (i). The colony numbers were counted from three independent experiments. The colony numbers were calculated as mean ± SD (n = 3), *P < 0.05; Student’s t test (j). k, l WM1346 cells generated in (a) were subjected to soft agar colony- formation assays (k). The colony numbers were counted and calculated as mean ± SD (n = 3) from three independent experiments. *P < 0.05; Student’s t test (l). m, n Tumor pictures (m) and the growth curves (n) for the xenograft experiments with the WM1346 cells generated in (a) were inoculated subcutaneously. The visible tumors were measured at the indicated days. Error bars represent ± SEM (n = 6). *P < 0.05; Student’s t test