Phosphorylation of iRhom2 at the plasma membrane controls mammalian TACE-dependent inflammatory and growth factor signalling

7 figures

Figures

Figure 1 with 2 supplements
TACE activity, but not TACE maturation is regulated by iRhom2 phosphorylation.

(a) HA immunoprecipitates (IP) and lysates (lys) from HEK293 cells stably expressing GFP or mouse iRhom2WT-3xHA were immunoblotted for phosphorylated serine, HA and beta-actin after treatment with or without 200 nM PMA for 15 min. (b) Phosphorylation sites identified (in red) in human iRhom2 by iRhom2-3xHA immunoprecipitation and mass spectrometry. Ion scores derive from separate peptides from at least two independent datasets. For comparison, the corresponding conserved residue in mouse iRhom2 is stated (c) A model of mouse iRhom2 (in red), with the conserved N-terminal serine/threonine phosphorylation sites (yellow stars) identified in Figure 1b and from public source databases. All of these sites have been mutated to alanine in iRhom2pDEAD. The iRhom homology domain (IRHD) is indicated between the first two transmembrane domains. (d) Lysates from iRhom1/2 double knockout (DKO) mouse embryonic fibroblasts (MEFs) expressing GFP or different versions of HA-tagged iRhom2 enriched for glycoproteins with Concanavalin A (ConA) were immunoblotted for TACE, HA, and transferrin receptor (TfR) as a labelling control. Lysates were probed with beta-actin antibody. Black arrowhead: immature TACE. Grey arrowhead: mature TACE. (e) Immunofluorescence of DKO MEFs transduced with different versions of iRhom2-3xHA, labelled with HA (green), GM130 label for the cis-Golgi (red) and DAPI for DNA (blue). GM130-labelled regions (within white boxes) have been magnified. Scale bar = 10 µm. (f) Lysates and neutravidin-enriched preparations from MEFs transduced with GFP, or HA-tagged iRhom2WT and iRhom2pDEAD labelled with (biotin pull down) or without biotin (no biotin control), probed with HA, pan-cadherin (cell surface positive control; arrowhead indicates specific band) and p97 (cell surface negative control) antibodies. (g–j) DKO MEFs transduced with GFP or different versions of iRhom2-3xHA transfected with alkaline-phosphatase tagged amphiregulin (Areg), transforming growth factor (TGFα), tumour necrosis factor (TNFα), epidermal growth factor (EGF) or betacellulin (BTC), stimulated for indicated times with 200 nM PMA. Values represent the level of released alkaline phosphatase/total alkaline phosphatase multiplied by 100. Error bars representing standard deviations and values for two-tailed student t-tests are depicted.

https://doi.org/10.7554/eLife.23968.003
Figure 1—figure supplement 1
Human iRhom2 phosphorylation can be detected by mass spectrometry.

(a) Protein sequence coverage of human iRhom2 in mass spectrometry analysis. The peptides identified in mass spectrometry are indicated in red. The identified phosphorylation sites are marked in grey. (b) Spectra of the identified human iRhom2 phosphopeptides. Mascot software was used to identify peptides of the human proteome from mass spectrometry data.

https://doi.org/10.7554/eLife.23968.004
Figure 1—figure supplement 2
Characterisation of iRhom2KDEL.

(a) Lysates (lys) and neutravidin-enriched preparations (biotin PD) from MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2pDEAD and iRhom2KDEL labelled with biotin and probed with HA antibody. (b) Cell surface FACS staining. Left panel: cell surface labelling with anti-HA antibody of DKO MEFs transduced with iRhom2WT (red) and iRhom2KDEL (blue). Cells stained only with the secondary antibody (grey) served as control. One experiment of 5 replicates is shown. Right panel: quantification of cell surface staining of iRhom2WT (red) and iRhom2KDEL (blue). The percentage of positive stained cells (relative to control) from 5 replicates has been analysed with a Student’s t-test (**p value < 0.01).

https://doi.org/10.7554/eLife.23968.005
Figure 2 with 1 supplement
Mature TACE activity is dependent upon iRhom2 phosphorylation at three distinct sites, which co-ordinate binding to 14-3-3 family proteins.

(a–c) DKO MEFs stably transduced with GFP or different versions of iRhom2-3xHA transfected with alkaline-phosphatase tagged TGFα or TNFα stimulated for 30 min with 200 nM PMA. The contributing residues of Sites 1–3 in iRhom2 are highlighted in red text. (a–c). (d) As stated for (a–c), except DKO MEFs were co-transfected with Areg, TGF or EGF and the histamine HA1 receptor, then stimulated for 60 min with 300 µM histamine a–d. Values represent the level of released alkaline phosphatase/total alkaline phosphatase multiplied by 100. Error bars representing standard deviations and values for two-tailed student t-tests are depicted. ns = p>0.05, *p<0.05, **p<0.01, ***p<0.001. (e) HA immunoprecipitates from HEK293 cells stably expressing GFP or HA-tagged mouse iRhom2WT or iRhom2Site1-3 were immunoblotted for phosphorylated serine and HA after treatment with or without 200 nM PMA for 15 min. Lysate was probed with HA antibody. (f) Lysates from HEK293T cells overexpressing HA-tagged human iRhom2WT, iRhom2ΔN (lacking aa 1–382) and unc93B1 were subjected to anti-HA immunoprecipitation and analysed by semi-quantitative label-free mass spectrometry. The ratio of the spectral index of iRhom2WT is compared to a control pull-down with the polytopic membrane protein unc93b1 to compensate for unspecific binding to hydrophobic TMDs. The ratio of the spectral index of iRhom2WT and iRhom2ΔN shows the decreased binding of 14-3-3 proteins to iRhom2ΔN. (g) HA immunoprecipitations and lysates from DKO MEFs transduced with GFP, HA-tagged mouse iRhom2WT or stated phosphorylation site-to-alanine mutants, treated with or without 200 nM PMA for 30 min, immunoblotted for HA, TACE, pan-14-3-3 and beta-actin.

https://doi.org/10.7554/eLife.23968.006
Figure 2—figure supplement 1
Analysis of the contribution to TGF shedding of each iRhom2 residue within each phosphorylation site.

(a–c) DKO MEFs transduced with GFP or different versions of HA-tagged iRhom2 transfected with alkaline-phosphatase tagged TGFα stimulated for 30 min with 200 nM PMA. Site 1 is depicted in (a), Site 2 in (b), and Site 3 in c. (d-g) HA immunoprecipitations and lysates from DKO MEFs transduced with GFP, mouse iRhom2WT-3xHA, stated phosphorylation site mutants or iRhom2KDEL, treated with or without 200 nM PMA for 15 min, immunoblotted for HA, indicated 14-3-3 isoforms (epsilon, eta in d,g, alpha/beta, gamma in e, tau, zeta in f) and beta-actin.

https://doi.org/10.7554/eLife.23968.007
iRhom2 phosphorylation and 14-3-3 binding is dependent on active ERK signalling.

(a) Lysates from cells overexpressing HA tagged human iRhom2WT and iRhom2∆N (lacking aa 1–382) and unc93B1 were subjected to HA immunoprecipitation and analysed by semi-quantitative label-free mass spectrometry. The ratio of the spectral index of iRhom2WT is compared to the control pull-down with unc93b1 to compensate for unspecific binding to hydrophobic TMDs. The ratio of the spectral index of iRhom2WT and iRhom2∆N shows the decreased binding of the identified kinases upon deletion of the N-terminus. (b) DKO MEFs stably transduced with HA tagged murine iRhom2WT transfected with alkaline-phosphatase tagged amphiregulin or EGF, pretreated for 1 hr with the indicated kinase inhibitors (20 µM) followed by stimulation for 1 hr with 20 nM PMA. (c) DKO MEFs stably transduced with HA tagged murine iRhom2WT transfected with alkaline-phosphatase tagged amphiregulin and the histamine HA1 receptor were pretreated with U0126 (20 µM) for 1 hr followed by stimulation for 1 hr with 300 µM histamine. In b–c, ***p<0.001. (d-e) HA immunoprecipitates (IP) and lysates (lys) from DKO MEFs transduced with GFP and HA-tagged iRhom2WT (and iRhom2Site1-3, in e) pre-treated with U0126 (20 µM) for 1 hr and stimulated with 20 nM PMA for 15 min, immunoblotted for 14-3-3, HA, ERK1/2 and phosphorylated ERK1/2. Arrowheads indicate the iRhom2 phosphoserine positive bands. (f) HA immunoprecipitates (IP) and lysates (lys) from DKO MEFs transduced with GFP and HA-tagged iRhom2WT and iRhom2Site1-3, transfected with 4 µg histamine HA1 receptor (per 10 cm plate), pre-treated with U0126 (20 µM) for 1 hr and stimulated with 300 µM histamine for 15 min, immunoblotted for 14-3-3, HA, ERK1/2 and phosphorylated ERK1/2.

https://doi.org/10.7554/eLife.23968.008
Figure 4 with 1 supplement
The amino-terminus of iRhom2 is required for binding to, and stabilisation of, mature TACE.

(a–b) HA immunoprecipitates (IP) and lysates (lys) from WT MEFs (in a) and DKO MEFs (in b) transduced with GFP and HA-tagged iRhom2WT, iRhom2ΔIRHD or iRhom2ΔNT, probed for HA, TACE and beta-actin. Lysates from WT MEFs, enriched for glycoproteins with Concanavalin A (ConA), show equal levels of mature TACE and transferrin receptor (TfR) (in a). (c) Lysates from DKO MEFs expressing GFP or different versions of HA-tagged iRhom2 constructs were enriched for glycoproteins with ConA and were immunoblotted for TACE, HA, ADAM10 and TfR as a labelling control. Lysates were probed with HA and beta-actin antibody. Where indicated, cells had been treated with the lysosomal degradation inhibitor bafilomycin A1 (Baf A, 100 nM) or the proteosomal degradation inhibitor MG132 (10 µM) for 4 hr. Lane numbers have been added for clarity. All panels: black arrowhead: immature TACE; grey arrowhead: mature TACE.

https://doi.org/10.7554/eLife.23968.009
Figure 4—figure supplement 1
iRhom2 interaction with TACE, and further characterisation of the deletion mutants of iRhom2.

(a) HA immunoprecipitates (IP) and lysates (lys) from DKO MEFs transduced with GFP or HA-tagged iRhom2WT, immunoblotted for HA, TACE and beta-actin. Empty beads (without conjugated HA antibody) are used to show the specificity of the co-immunoprecipitation. (b) Lysates (lys) and neutravidin-enriched preparations (biotin PD) from MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2∆NT and iRhom2∆IRHD labelled with biotin and probed with HA and transferrin receptor antibodies. (c) Lysates from DKO MEFs expressing HA-tagged iRhom2WT and iRhom2∆NT enriched for glycoproteins with ConA were immunoblotted for TACE, HA, and transferrin receptor (TfR) as a labelling control. Where indicated, cells had been treated with the lysosomal degradation inhibitor bafilomycin A1 (Baf A, 100 nM) or the proteosomal degradation inhibitor MG132 (10 µM) for 4 hr. Lysates were probed for beta-actin, and immunoblotted for ubiquitin to demonstrate the efficacy of MG132 treatment.

https://doi.org/10.7554/eLife.23968.010
Figure 5 with 1 supplement
Phosphorylation and 14-3-3 protein recruitment uncouples the interaction between iRhom2 and mature TACE to drive its activity at the cell surface.

(a) A model of mouse iRhom2 (in red), with the two R18 sequence insertions in the N-terminus to facilitate the recruitment of 14-3-3 proteins (in green). Underneath, the sequence of the R18 insertion is indicated. (b) HA immunoprecipitates (IP) and lysates (lys) from DKO MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2Site1-3 or iRhom2R18, treated with or without 200 nM PMA for 30 min, immunoblotted for HA, 14-3-3 and beta-actin. (c) DKO MEFs transduced with GFP or different versions of HA-tagged iRhom2 transfected with alkaline-phosphatase tagged Areg or TGF. Values represent the level of released alkaline phosphatase overnight/total alkaline phosphatase multiplied by 100. Error bars representing standard deviations and values for two-tailed student t-tests, compared to iRhom2Site1-3 are depicted. *p<0.05. (d) Lysates from DKO MEFs expressing GFP or different versions of HA-tagged iRhom2 enriched for glycoproteins with Concanavalin A (Con A) were immunoblotted for HA, TACE and transferrin receptor (TfR) as a labelling control. Grey arrowhead: immature TACE. White arrowhead: mature TACE. (e) HA immunoprecipitates (IP) and lysates (lys) from DKO MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2Site1-2, iRhom2Site1-3 or iRhom2R18, treated with or without 200 nM PMA for 30 min, immunoblotted for HA, TACE, 14-3-3 and beta-actin. Grey arrowhead: immature TACE. White arrowhead: mature TACE. (f) HA immunoprecipitates (left) and neutravidin-enriched cell surface biotinylated protein preparations (right) with corresponding lysates (lys) from DKO MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2Site1-3, treated with or without 200 nM PMA for 5 min, immunoblotted for HA, TACE and beta-actin. As a molecular marker for immature and mature TACE, in the last lane, a lysate from DKO MEFs expressing HA-tagged iRhom2WT enriched for glycoproteins with Concanavalin A (ConA) was loaded. Lane numbers have been added for clarity. On the right-hand side, a quantification of the interaction between iRhom2 and mature TACE from three independent experiments has been plotted. This was achieved by dividing densitometry values of mature TACE by levels of iRhom2WT or iRhom2Site1-3, and normalised to the values before addition of PMA (a.u).

https://doi.org/10.7554/eLife.23968.011
Figure 5—figure supplement 1
Cell surface presentation of iRhom2R18.

Lysates (lys) and neutravidin-enriched preparations (biotin PD) from MEFs transduced with GFP, or HA-tagged iRhom2WT, iRhom2Site1-3 and iRhom2R18 labelled with biotin and probed with HA, transferrin receptor and beta-actin antibodies.

https://doi.org/10.7554/eLife.23968.012
Acute regulation of TACE by iRhom2 phosphorylation controls TNFα release in primary macrophages.

(a) Lysates from WT or iRhom2 KO primary macrophages transduced with GFP, or HA-tagged iRhom2WT, iRhom2Site1-3, enriched for glycoproteins with Concanavalin A (ConA) and probed for TACE, HA, transferrin receptor (TfR) and ADAM10. NB: The separation of immature and mature TACE is not always apparent in macrophage preparations (Adrain et al., 2012). (b–c) HA immunoprecipitations (IP) and lysates (lys) from iRhom2 KO primary macrophages transduced with GFP, or HA-tagged mouse iRhom2WTor iRhom2Site1-3, treated with or without 200 nM PMA (c) or 100 ng/ml LPS (d) for 15 min, immunoblotted for HA, phosphorylated serine, pan-14-3-3, TACE and beta-actin. (d) Levels of secreted TNFα by ELISA from primary macrophages described in a treated with 100 ng/ml LPS for indicated times. Error bars representing standard deviations and values for two-tailed student t-tests are depicted.

https://doi.org/10.7554/eLife.23968.013
iRhom2 regulates TACE at multiple steps in its lifecycle to control growth factor and cytokine signalling.

(a) iRhom2 and immature TACE must move together from the ER through the Golgi apparatus (Adrain et al., 2012) also shown by iRhom2KDEL (Figure 1d), the site of furin cleavage of the TACE prodomain (in yellow), and therefore production of mature TACE. This event requires a contribution from the iRhom2 IRHD. Without iRhom2, there is no mature TACE. (b) iRhom2 and mature TACE are in complex with one another at the plasma membrane. The iRhom2 N-terminus stabilises mature TACE levels at the cell surface; in its absence mature TACE is degraded in lysosomes. (c) Upon delivery and maintenance of mature TACE at the plasma membrane, GPCR stimulation or macrophage activation with LPS leads to iRhom2 phosphorylation and recruitment of 14-3-3 proteins. The recruitment of 14-3-3 proteins is sufficient to alter the inhibitory interaction between iRhom2 and mature TACE, allowing TACE to cleave its cytokine and growth factor substrates.

https://doi.org/10.7554/eLife.23968.014

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Adam Graham Grieve
  2. Hongmei Xu
  3. Ulrike Künzel
  4. Paul Bambrough
  5. Boris Sieber
  6. Matthew Freeman
(2017)
Phosphorylation of iRhom2 at the plasma membrane controls mammalian TACE-dependent inflammatory and growth factor signalling
eLife 6:e23968.
https://doi.org/10.7554/eLife.23968