Injury or infection can cause tissues in the body to become inflamed. The immune system triggers this inflammation to help repair the injury or fight the infection. A signal molecule known as TNF – which is produced by immune cells called macrophages – triggers inflammation. This protein is normally attached to the surface of the macrophage, and it only activates inflammation once it has been cut free.

An enzyme called TACE cuts and releases TNF from the surface of macrophages. This enzyme is made inside the cell and is then transported to the surface. On the way, TACE matures from an inactive form to a fully functional enzyme. Previous work revealed that a protein called iRhom2 controls TACE maturation, but it has been unclear whether iRhom2 affects TACE in any additional ways.

Grieve et al. studied the relationship between iRhom2 and TACE in more detail. The experiments show two new roles for iRhom2: in protecting TACE from being destroyed at the cell surface, and prompting TACE to release TNF to trigger inflammation. Injury or infection causes small molecules called phosphate groups to be attached to iRhom2 in macrophages, which causes TACE to release TNF.

The findings of Grieve et al. provide the first evidence that iRhom2 influences the activity of TACE throughout the enzyme’s lifetime. Excessive inflammation, often triggered by the uncontrolled release of TNF, can lead to rheumatoid arthritis, cancer and many other diseases. Therefore, iRhom2 could be a promising new target for anti-inflammatory drugs that may help to treat these conditions.

Signalling ligands are often synthesised as transmembrane domain (TMD) containing precursors. Upon their delivery to the plasma membrane, protease activity is required to shed the bioactive extracellular domain to allow signal release and subsequent binding to receptors on signal-receiving cells. TACE (also known as ADAM17) is a primary shedding enzyme and as such regulates multiple signalling pathways through its ability to cleave and release numerous membrane-tethered signalling ligands and receptors (Peschon et al., 1998). Of particular interest, it controls the shedding of TNFα, the principal inflammatory cytokine (Black et al., 1997; Moss et al., 1997), amphiregulin, TGFα and most other ligands of the epidermal growth factor (EGF) family (Sunnarborg et al., 2002; Sahin et al., 2004). Disturbances in these cytokine and growth factor signalling pathways are hallmarks of inflammation and cancer, respectively, as well as other diseases. This illustrates the potentially dangerous consequence of unregulated TACE activity and explains the tight post-translational regulation to which TACE is subject.

TACE is first synthesised in the endoplasmic reticulum (ER) as an immature form containing an inhibitory pro-domain that prevents its proteolytic activity. We and others identified the iRhom proteins, catalytically inactive members of the rhomboid-like superfamily, as essential regulators of TACE maturation (Adrain et al., 2012; McIlwain et al., 2012; Siggs et al., 2012; Issuree et al., 2013). We reported that iRhoms control transport of TACE from the ER to the Golgi apparatus, where removal of its pro-domain by pro-protein convertases such as furin occurs (Endres et al., 2003). In this way, iRhoms regulate the conversion of TACE from an inactive immature form to a mature proteolytically competent shedding enzyme (Adrain and Freeman, 2012; Freeman, 2014; Lemberg and Adrain, 2016). Without iRhoms there is no TACE maturation and therefore no TACE activity (Christova et al., 2013; Li et al., 2015). Of the two mammalian iRhoms, iRhom1 is broadly expressed, whereas macrophages express only iRhom2. Since macrophages are the major TNFα-releasing cell type (Parameswaran and Patial, 2010), this makes iRhom2 an important regulator of inflammation. Accordingly iRhom2 knock-out mice have profound inflammatory defects, are sensitive to bacterial infection and are resistant to LPS-induced toxic shock and the development of inflammatory arthritis (Adrain et al., 2012; McIlwain et al., 2012; Issuree et al., 2013). More recently iRhom2 was shown to modulate innate immunity to DNA viruses by regulating the ER-to-Golgi transport and the stability of the immune adaptor STING (Luo et al., 2016). Although there are also hints that iRhom2 might regulate TACE protein stability (Maney et al., 2015), this has not been explored in detail.

iRhoms comprise a seven transmembrane rhomboid-like domain, with a short luminal C-terminus, a well conserved loop between TMDs 1 and 2 (termed the iRhom homology domain, IRHD), and a long cytoplasmic N-terminus (Lemberg and Freeman, 2007, 2016). There is accumulating physiological and cell biological evidence that mutation or deletion of the cytoplasmic N-terminus of iRhoms leads to alterations in TACE function (Johnson et al., 2003; Blaydon et al., 2012; Saarinen et al., 2012; Brooke et al., 2014; Hosur et al., 2014; Siggs et al., 2014; Liu et al., 2015; Maney et al., 2015). Moreover, iRhoms act downstream of G-protein coupled receptor (GPCR) signalling in the transactivation of EGFR signalling (Christova et al., 2013; Li et al., 2015). Taken together, these observations indicate that the cytosolic N-terminal domain of iRhoms regulates TACE, but the underlying mechanisms are unclear. Importantly, TACE-induced ligand shedding is rapid: it is activated within ten minutes by physiological and pharmacological stimulation (Le Gall et al., 2010), and this is compatible with the timescale on which iRhoms have been reported to influence TACE substrate selection (Maretzky et al., 2013). However, this speed of TACE activation is not easily reconciled with its relatively slow ER-to-Golgi trafficking and maturation (3–6 hr, [Schlöndorff et al., 2000]), the process by which iRhoms are known to control TACE.

Here we resolve this apparent paradox by demonstrating that iRhom2-mediated TACE maturation represents just the first step in the relationship between them. Once mature TACE has been trafficked to the cell surface, it requires a further iRhom2-dependent activation step, mediated by phosphorylation of the iRhom2 cytoplasmic domain. We reveal two unanticipated molecular roles for iRhom2 beyond its described function in TACE maturation. First, iRhom2 interacts with mature TACE at the plasma membrane, simultaneously stabilising TACE from lysosomal degradation and limiting its proteolytic activity. Second, iRhom2 acts as a signalling hub whereby phosphorylation of its N-terminus and subsequent 14-3-3 protein binding controls stimulated TACE proteolytic activity and ligand shedding. We show that this regulation of mature TACE activity occurs not only in immortalised cell models but also regulates inflammatory TNFα release from primary macrophages. Overall, we conclude that iRhom2 binds to TACE and controls its function at multiple stages, indicating that iRhom2 can be treated as a multifunctional regulatory subunit of TACE.

We have discovered that iRhom2 regulates the metalloprotease TACE by multiple, distinct mechanisms throughout the secretory pathway. Because of its ability to trigger both cytokine and growth factor signalling, TACE activity is potentially dangerous, so it is unsurprising that it is itself subject to multiple layers of control. What is more unexpected is that iRhom2, a catalytically inactive relative of rhomboid intramembrane proteases, should be so intimately involved in TACE regulation throughout its lifecycle. We and others have already reported that iRhom2 is responsible for the first step of generating mature TACE, controlling its ER to Golgi trafficking (Adrain et al., 2012; McIlwain et al., 2012; Siggs et al., 2012; Issuree et al., 2013); we now show that this was only the first of several stages at which iRhom2 controls TACE. Once TACE has been exported from the ER under the influence of iRhom2 (Figure 7a), and its inhibitory pro-domain is removed in the trans-Golgi network, interaction with iRhom2 is essential to stabilise it at the plasma membrane, preventing its lysosomal degradation (Figure 7b). Stimulation of TACE proteolytic activity is then also mediated by iRhom2: the cytoplasmic N-terminus of iRhom2 is phosphorylated in response to, for example, GPCR signalling, leading to 14-3-3 binding. Our data support a model in which phosphorylation and 14-3-3 binding elicit a change in the interaction between iRhom2 and mature TACE that licenses TACE activity to shed substrates from the cell surface (Figure 7c).

Figure 7

Download asset Open asset

The cytoplasmic domain of iRhom2 maintains the plasma membrane stability of mature TACE. Endocytic clearance of mature TACE from the cell surface has been previously demonstrated after stimulation with PMA (Doedens and Black, 2000; Lorenzen et al., 2016), although more physiological stimuli do not alter the plasma membrane abundance of the protease (Lorenzen et al., 2016), suggesting that this is not a primary control point of TACE-dependent shedding. Instead, as with many cell surface proteins, there appears to be a constitutive endocytic recycling of mature TACE, mediated by PACS-2 (Dombernowsky et al., 2015). The loss of PACS-2 triggers activation-independent mature TACE lysosomal degradation, and it will be interesting to investigate further whether there is a mechanistic link between iRhom2 and PACS-2 in the stabilisation of mature TACE. Interestingly, this resonates with a recent report that iRhom2 maintains the stability of another client protein, the innate immune response adaptor protein, STING (Luo et al., 2016), although in this case, STING is degraded by the proteasome.

Signals that lead to the phosphorylation of the cytoplasmic N-terminal domain of iRhom2 cause 14-3-3 binding to iRhom2. The consequence of this signal-dependent 14-3-3 binding is a detectable alteration of the interaction between mature TACE and iRhom2 at the plasma membrane, and activation of TACE-dependent shedding of substrates (Figure 7c). Importantly, we have confirmed that this control of stimulated shedding occurs not only in embryonic fibroblasts, but is also an essential regulator of TNFα production in primary macrophages. Furthermore, we find that iRhom2 phosphorylation and 14-3-3 binding is dependent on ERK1/2 signalling, which is itself essential for TACE-dependent TNFα release in macrophages (Dumitru et al., 2000; Eliopoulos et al., 2003). Consistent with the physiological significance of iRhom2 phosphorylation as a signalling mechanism, Chanthaphavong et al. reported its phosphorylation in primary hepatocytes in response to nitric oxide signalling (Chanthaphavong et al., 2012). It is significant that stimulated TACE activity does not depend on its own cytoplasmic domain (Reddy et al., 2000; Horiuchi et al., 2007; Le Gall et al., 2010). Instead, the TACE TMD is required for its rapid activation (Le Gall et al., 2010), but the molecular mechanism underlying this observation has been mechanistically obscure. iRhoms have evolved from rhomboid intramembrane proteases, which are polytopic membrane proteins that recognise and cleave transmembrane domains, and it is therefore probable that TMD interaction remains a core function of iRhoms (Freeman, 2016). It is possible that the iRhom2-TACE interaction resembles a rhomboid-substrate complex, but that instead of then cleaving the TACE TMD, iRhom2 binding and phosphorylation alters the state of TACE, licensing its activation. Another possible mechanism that reflects the presumed TMD binding role of iRhom2 would be that the phosphorylation of the iRhom2 N-terminus may regulate its binding to TMD-containing TACE substrates, such as TNFα; in other words, that iRhom2 provides a substrate presentation or selection mechanism for TACE.

It will be important to understand how this newly identified phosphorylation-dependent regulatory mechanism relates to other previously identified modes of acute TACE regulation. Le Gall et al. (2010) reported that stimulation rapidly alters the accessibility of the TACE active site to a tight binding inhibitor, suggesting conformational changes in TACE itself. It has also been observed that iRhom2 controls substrate selectivity of mature TACE (Maretzky et al., 2013). Recently, Sommer et al. provided another model for acute TACE activation (Sommer et al., 2016). A variety of external signals cause the translocation of the lipid phosphatidylserine (PS) from the inner to outer leaflet of the plasma membrane. It has been shown that positive amino-acids in the extracellular membrane-proximal domain of TACE interact with the anionic head groups of exposed of PS, and it was proposed that this supports the positioning of TACE to cleave its substrates (Sommer et al., 2016). Another model for TACE activation is its conversion from dimers to monomers, which is associated with decreased TIMP3 association (Xu et al., 2012). Again, this model has significant overlap with our findings, as it is dependent on p38 MAPK and ERK signalling. The challenge is now to unify these earlier data with our discovery that phosphorylation and 14-3-3 binding of the iRhom2 cytoplasmic domain is necessary and sufficient to trigger TACE-dependent shedding of its substrates. A possible unifying model is that iRhom2 phosphorylation and subsequent conformational change in the iRhom2/TACE complex is necessary to allow mature TACE to engage with the above events of PS exposure, monomerisation, conformational change of the active site, and substrate selection. It must be stressed that this sequence of events is entirely speculative, but it is consistent with all published data, and does highlight the extraordinarily complex and tight regulation of TACE. Further study of the structure of iRhom2 and TACE should yield further insight into the enzymatic activation process.

Overall, we propose that iRhoms have evolved as long-term molecular partners of TACE: first escorting and controlling TACE secretion and maturation; then stabilising it at the cell surface; and subsequently acting as a signalling hub at the plasma membrane that licenses stimulated TACE activity upon receipt of appropriate signals. These results greatly strengthen the relationship between iRhom2 and TACE, leading us to propose that iRhom2 should perhaps be considered as a multifunctional cofactor or regulatory subunit of TACE.

The incidence of inflammatory diseases is increasing, and there is a need for new anti-inflammatory therapies (Myasoedova et al., 2010a, 2010b). TNFα blocking antibodies are used to treat rheumatoid arthritis, inflammatory bowel disease and psoriasis, and this therapy is one of the biggest-selling pharmaceuticals in the world. It is, however, unsuccessful for approximately a third of patients, requires injection, and is expensive (Palladino et al., 2003; Scott and Kingsley, 2006; Esposito and Cuzzocrea, 2009; Monaco et al., 2015). By expanding substantially the scope of the relationship between iRhom2 and TACE, the work we report here strengthens the case for the therapeutic potential of targeting iRhom2. Its previously reported role in regulating TNFα signalling by the ER-to-Golgi trafficking and maturation of TACE makes iRhom2 conceptually a valuable target, but that function will be hard to target pharmaceutically as it depends on a TMD-based interaction in the ER (our unpublished observation). In contrast, the regulatory mechanism between iRhom2 and mature TACE at the plasma membrane described here is cytoplasmic and thus perhaps more targetable. Moreover, the molecular components such as 14-3-3 proteins and the kinases that directly phosphorylate iRhom2 are more conventional, and therefore accessible, pharmaceutical targets. Finally, beyond targeting inflammatory cytokine signalling, pharmaceutical manipulation of the iRhoms-TACE interface could also influence growth factor signalling through the EGF receptor family, which is heavily implicated in a variety of diseases, most notably cancer (Kenny and Bissell, 2007; Ciardiello and Tortora, 2008).

1. 1. Adrain C
   2. Freeman M

   (2012) New lives for old: evolution of pseudoenzyme function illustrated by iRhoms

   Nature Reviews Molecular Cell Biology 13:489–498.

   https://doi.org/10.1038/nrm3392

   • PubMed
   • Google Scholar
2. 1. Adrain C
   2. Zettl M
   3. Christova Y
   4. Taylor N
   5. Freeman M

   (2012) Tumor necrosis factor signaling requires iRhom2 to promote trafficking and activation of TACE

   Science 335:225–228.

   https://doi.org/10.1126/science.1214400

   • PubMed
   • Google Scholar
3. 1. Black RA
   2. Rauch CT
   3. Kozlosky CJ
   4. Peschon JJ
   5. Slack JL
   6. Wolfson MF
   7. Castner BJ
   8. Stocking KL
   9. Reddy P
   10. Srinivasan S
   11. Nelson N
   12. Boiani N
   13. Schooley KA
   14. Gerhart M
   15. Davis R
   16. Fitzner JN
   17. Johnson RS
   18. Paxton RJ
   19. March CJ
   20. Cerretti DP

   (1997) A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells

   Nature 385:729–733.

   https://doi.org/10.1038/385729a0

   • PubMed
   • Google Scholar
4. 1. Blaydon DC
   2. Etheridge SL
   3. Risk JM
   4. Hennies HC
   5. Gay LJ
   6. Carroll R
   7. Plagnol V
   8. McRonald FE
   9. Stevens HP
   10. Spurr NK
   11. Bishop DT
   12. Ellis A
   13. Jankowski J
   14. Field JK
   15. Leigh IM
   16. South AP
   17. Kelsell DP

   (2012) RHBDF2 mutations are associated with tylosis, a familial esophageal Cancer syndrome

   The American Journal of Human Genetics 90:340–346.

   https://doi.org/10.1016/j.ajhg.2011.12.008

   • PubMed
   • Google Scholar
5. 1. Brooke MA
   2. Etheridge SL
   3. Kaplan N
   4. Simpson C
   5. O'Toole EA
   6. Ishida-Yamamoto A
   7. Marches O
   8. Getsios S
   9. Kelsell DP

   (2014) iRHOM2-dependent regulation of ADAM17 in cutaneous disease and epidermal barrier function

   Human Molecular Genetics 23:4064–4076.

   https://doi.org/10.1093/hmg/ddu120

   • PubMed
   • Google Scholar
6. 1. Chanthaphavong RS
   2. Loughran PA
   3. Lee TY
   4. Scott MJ
   5. Billiar TR

   (2012) A role for cGMP in inducible nitric-oxide synthase (iNOS)-induced tumor necrosis factor (TNF) α-converting enzyme (TACE/ADAM17) activation, translocation, and TNF receptor 1 (TNFR1) shedding in hepatocytes

   Journal of Biological Chemistry 287:35887–35898.

   https://doi.org/10.1074/jbc.M112.365171

   • PubMed
   • Google Scholar
7. 1. Christova Y
   2. Adrain C
   3. Bambrough P
   4. Ibrahim A
   5. Freeman M

   (2013) Mammalian iRhoms have distinct physiological functions including an essential role in TACE regulation

   EMBO Reports 14:884–890.

   https://doi.org/10.1038/embor.2013.128

   • PubMed
   • Google Scholar
8. 1. Ciardiello F
   2. Tortora G

   (2008) EGFR antagonists in Cancer treatment

   New England Journal of Medicine 358:1160–1174.

   https://doi.org/10.1056/NEJMra0707704

   • PubMed
   • Google Scholar
9. 1. Doedens JR
   2. Black RA

   (2000) Stimulation-induced down-regulation of tumor necrosis factor-alpha converting enzyme

   Journal of Biological Chemistry 275:14598–14607.

   https://doi.org/10.1074/jbc.275.19.14598

   • PubMed
   • Google Scholar
10. 1. Dombernowsky SL
    2. Samsøe-Petersen J
    3. Petersen CH
    4. Instrell R
    5. Hedegaard AM
    6. Thomas L
    7. Atkins KM
    8. Auclair S
    9. Albrechtsen R
    10. Mygind KJ
    11. Fröhlich C
    12. Howell M
    13. Parker P
    14. Thomas G
    15. Kveiborg M

    (2015) The sorting protein PACS-2 promotes ErbB signalling by regulating recycling of the metalloproteinase ADAM17

    Nature Communications 6:7518.

    https://doi.org/10.1038/ncomms8518

    • PubMed
    • Google Scholar
11. 1. Dumitru CD
    2. Ceci JD
    3. Tsatsanis C
    4. Kontoyiannis D
    5. Stamatakis K
    6. Lin JH
    7. Patriotis C
    8. Jenkins NA
    9. Copeland NG
    10. Kollias G
    11. Tsichlis PN

    (2000) TNF-alpha induction by LPS is regulated posttranscriptionally via a Tpl2/ERK-dependent pathway

    Cell 103:1071–1083.

    https://doi.org/10.1016/S0092-8674(00)00210-5

    • PubMed
    • Google Scholar
12. 1. Eliopoulos AG
    2. Wang CC
    3. Dumitru CD
    4. Tsichlis PN

    (2003) Tpl2 transduces CD40 and TNF signals that activate ERK and regulates IgE induction by CD40

    The EMBO Journal 22:3855–3864.

    https://doi.org/10.1093/emboj/cdg386

    • PubMed
    • Google Scholar
13. 1. Endres K
    2. Anders A
    3. Kojro E
    4. Gilbert S
    5. Fahrenholz F
    6. Postina R

    (2003) Tumor necrosis factor-alpha converting enzyme is processed by proprotein-convertases to its mature form which is degraded upon phorbol ester stimulation

    European Journal of Biochemistry 270:2386–2393.

    https://doi.org/10.1046/j.1432-1033.2003.03606.x

    • PubMed
    • Google Scholar
14. 1. Esposito E
    2. Cuzzocrea S

    (2009) TNF-alpha as a therapeutic target in inflammatory diseases, ischemia-reperfusion injury and trauma

    Current Medicinal Chemistry 16:3152–3167.

    https://doi.org/10.2174/092986709788803024

    • PubMed
    • Google Scholar
15. 1. Favata MF
    2. Horiuchi KY
    3. Manos EJ
    4. Daulerio AJ
    5. Stradley DA
    6. Feeser WS
    7. Van Dyk DE
    8. Pitts WJ
    9. Earl RA
    10. Hobbs F
    11. Copeland RA
    12. Magolda RL
    13. Scherle PA
    14. Trzaskos JM

    (1998) Identification of a novel inhibitor of mitogen-activated protein kinase kinase

    Journal of Biological Chemistry 273:18623–18632.

    https://doi.org/10.1074/jbc.273.29.18623

    • PubMed
    • Google Scholar
16. 1. Freeman M

    (2014) The rhomboid-like superfamily: molecular mechanisms and biological roles

    Annual Review of Cell and Developmental Biology 30:235–254.

    https://doi.org/10.1146/annurev-cellbio-100913-012944

    • PubMed
    • Google Scholar
17. 1. Freeman M

    (2016) Rhomboids, signalling and cell biology

    Biochemical Society Transactions 44:945–950.

    https://doi.org/10.1042/BST20160035

    • PubMed
    • Google Scholar
18. 1. Fu H
    2. Subramanian RR
    3. Masters SC

    (2000) 14-3-3 proteins: structure, function, and regulation

    Annual Review of Pharmacology and Toxicology 40:617–647.

    https://doi.org/10.1146/annurev.pharmtox.40.1.617

    • PubMed
    • Google Scholar
19. 1. Gschwendt M
    2. Dieterich S
    3. Rennecke J
    4. Kittstein W
    5. Mueller HJ
    6. Johannes FJ

    (1996) Inhibition of protein kinase C mu by various inhibitors. differentiation from protein kinase c isoenzymes

    FEBS Letters 392:77–80.

    https://doi.org/10.1016/0014-5793(96)00785-5

    • PubMed
    • Google Scholar
20. 1. Horiuchi K
    2. Le Gall S
    3. Schulte M
    4. Yamaguchi T
    5. Reiss K
    6. Murphy G
    7. Toyama Y
    8. Hartmann D
    9. Saftig P
    10. Blobel CP

    (2007) Substrate selectivity of epidermal growth factor-receptor ligand sheddases and their regulation by phorbol esters and calcium influx

    Molecular Biology of the Cell 18:6–188.

    https://doi.org/10.1091/mbc.E06-01-0014

    • PubMed
    • Google Scholar
21. 1. Hosur V
    2. Johnson KR
    3. Burzenski LM
    4. Stearns TM
    5. Maser RS
    6. Shultz LD

    (2014) Rhbdf2 mutations increase its protein stability and drive EGFR hyperactivation through enhanced secretion of amphiregulin

    PNAS 111:E2200–E2209.

    https://doi.org/10.1073/pnas.1323908111

    • PubMed
    • Google Scholar
22. 1. Issuree PD
    2. Maretzky T
    3. McIlwain DR
    4. Monette S
    5. Qing X
    6. Lang PA
    7. Swendeman SL
    8. Park-Min KH
    9. Binder N
    10. Kalliolias GD
    11. Yarilina A
    12. Horiuchi K
    13. Ivashkiv LB
    14. Mak TW
    15. Salmon JE
    16. Blobel CP

    (2013) iRHOM2 is a critical pathogenic mediator of inflammatory arthritis

    Journal of Clinical Investigation 123:928–932.

    https://doi.org/10.1172/JCI66168

    • PubMed
    • Google Scholar
23. 1. Johnson KR
    2. Lane PW
    3. Cook SA
    4. Harris BS
    5. Ward-Bailey PF
    6. Bronson RT
    7. Lyons BL
    8. Shultz LD
    9. Davisson MT

    (2003) Curly bare (cub), a new mouse mutation on chromosome 11 causing skin and hair abnormalities, and a modifier gene (mcub) on chromosome 5

    Genomics 81:6–14.

    https://doi.org/10.1016/S0888-7543(02)00013-7

    • PubMed
    • Google Scholar
24. 1. Kenny PA
    2. Bissell MJ

    (2007) Targeting TACE-dependent EGFR ligand shedding in breast Cancer

    Journal of Clinical Investigation 117:337–345.

    https://doi.org/10.1172/JCI29518

    • PubMed
    • Google Scholar
25. 1. Le Gall SM
    2. Maretzky T
    3. Issuree PD
    4. Niu XD
    5. Reiss K
    6. Saftig P
    7. Khokha R
    8. Lundell D
    9. Blobel CP

    (2010) ADAM17 is regulated by a rapid and reversible mechanism that controls access to its catalytic site

    Journal of Cell Science 123:3913–3922.

    https://doi.org/10.1242/jcs.069997

    • PubMed
    • Google Scholar
26. 1. Lemberg MK
    2. Freeman M

    (2007) Functional and evolutionary implications of enhanced genomic analysis of rhomboid intramembrane proteases

    Genome Research 17:1634–1646.

    https://doi.org/10.1101/gr.6425307

    • PubMed
    • Google Scholar
27. 1. Lemberg MK
    2. Adrain C

    (2016) Inactive rhomboid proteins: new mechanisms with implications in health and disease

    Seminars in Cell & Developmental Biology 60:29–37.

    https://doi.org/10.1016/j.semcdb.2016.06.022

    • PubMed
    • Google Scholar
28. 1. Li X
    2. Maretzky T
    3. Weskamp G
    4. Monette S
    5. Qing X
    6. Issuree PD
    7. Crawford HC
    8. McIlwain DR
    9. Mak TW
    10. Salmon JE
    11. Blobel CP

    (2015) iRhoms 1 and 2 are essential upstream regulators of ADAM17-dependent EGFR signaling

    PNAS 112:6080–6085.

    https://doi.org/10.1073/pnas.1505649112

    • PubMed
    • Google Scholar
29. 1. Liu B
    2. Xu Y
    3. Li WL
    4. Zeng L

    (2015) Proteomic analysis of differentially expressed skin proteins in iRhom2(Uncv) mice

    BMB Reports 48:19–24.

    https://doi.org/10.5483/BMBRep.2015.48.1.022

    • PubMed
    • Google Scholar
30. 1. Lorenzen I
    2. Lokau J
    3. Korpys Y
    4. Oldefest M
    5. Flynn CM
    6. Künzel U
    7. Garbers C
    8. Freeman M
    9. Grötzinger J
    10. Düsterhöft S

    (2016) Control of ADAM17 activity by regulation of its cellular localisation

    Scientific Reports 6:35067.

    https://doi.org/10.1038/srep35067

    • PubMed
    • Google Scholar
31. 1. Luo WW
    2. Li S
    3. Li C
    4. Lian H
    5. Yang Q
    6. Zhong B
    7. Shu HB

    (2016) iRhom2 is essential for innate immunity to DNA viruses by mediating trafficking and stability of the adaptor STING

    Nature Immunology 17:1057–1066.

    https://doi.org/10.1038/ni.3510

    • PubMed
    • Google Scholar
32. 1. Maney SK
    2. McIlwain DR
    3. Polz R
    4. Pandyra AA
    5. Sundaram B
    6. Wolff D
    7. Ohishi K
    8. Maretzky T
    9. Brooke MA
    10. Evers A
    11. Vasudevan AA
    12. Aghaeepour N
    13. Scheller J
    14. Münk C
    15. Häussinger D
    16. Mak TW
    17. Nolan GP
    18. Kelsell DP
    19. Blobel CP
    20. Lang KS
    21. Lang PA

    (2015) Deletions in the cytoplasmic domain of iRhom1 and iRhom2 promote shedding of the TNF receptor by the protease ADAM17

    Science Signaling 8:ra109.

    https://doi.org/10.1126/scisignal.aac5356

    • PubMed
    • Google Scholar
33. 1. Maretzky T
    2. McIlwain DR
    3. Issuree PD
    4. Li X
    5. Malapeira J
    6. Amin S
    7. Lang PA
    8. Mak TW
    9. Blobel CP

    (2013) iRhom2 controls the substrate selectivity of stimulated ADAM17-dependent ectodomain shedding

    PNAS 110:11433–11438.

    https://doi.org/10.1073/pnas.1302553110

    • PubMed
    • Google Scholar
34. 1. Martiny-Baron G
    2. Kazanietz MG
    3. Mischak H
    4. Blumberg PM
    5. Kochs G
    6. Hug H
    7. Marmé D
    8. Schächtele C

    (1993)

    Selective inhibition of protein kinase C isozymes by the indolocarbazole gö 6976

    The Journal of Biological Chemistry 268:9194–9197.

    • PubMed
    • Google Scholar
35. 1. McIlwain DR
    2. Lang PA
    3. Maretzky T
    4. Hamada K
    5. Ohishi K
    6. Maney SK
    7. Berger T
    8. Murthy A
    9. Duncan G
    10. Xu HC
    11. Lang KS
    12. Häussinger D
    13. Wakeham A
    14. Itie-Youten A
    15. Khokha R
    16. Ohashi PS
    17. Blobel CP
    18. Mak TW

    (2012) iRhom2 regulation of TACE controls TNF-mediated protection against Listeria and responses to LPS

    Science 335:229–232.

    https://doi.org/10.1126/science.1214448

    • PubMed
    • Google Scholar
36. 1. Monaco C
    2. Nanchahal J
    3. Taylor P
    4. Feldmann M

    (2015) Anti-TNF therapy: past, present and future

    International Immunology 27:55–62.

    https://doi.org/10.1093/intimm/dxu102

    • PubMed
    • Google Scholar
37. 1. Moss ML
    2. Jin SL
    3. Milla ME
    4. Bickett DM
    5. Burkhart W
    6. Carter HL
    7. Chen WJ
    8. Clay WC
    9. Didsbury JR
    10. Hassler D
    11. Hoffman CR
    12. Kost TA
    13. Lambert MH
    14. Leesnitzer MA
    15. McCauley P
    16. McGeehan G
    17. Mitchell J
    18. Moyer M
    19. Pahel G
    20. Rocque W
    21. Overton LK
    22. Schoenen F
    23. Seaton T
    24. Su JL
    25. Becherer JD

    (1997) Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-alpha

    Nature 385:733–736.

    https://doi.org/10.1038/385733a0

    • PubMed
    • Google Scholar
38. 1. Munro S
    2. Pelham HR

    (1987) A C-terminal signal prevents secretion of luminal ER proteins

    Cell 48:899–907.

    https://doi.org/10.1016/0092-8674(87)90086-9

    • PubMed
    • Google Scholar
39. 1. Myasoedova E
    2. Crowson CS
    3. Kremers HM
    4. Therneau TM
    5. Gabriel SE

    (2010a) Is the incidence of rheumatoid arthritis rising?: results from Olmsted County, Minnesota, 1955-2007

    Arthritis & Rheumatism 62:1576–1582.

    https://doi.org/10.1002/art.27425

    • PubMed
    • Google Scholar
40. 1. Myasoedova E
    2. Davis JM
    3. Crowson CS
    4. Gabriel SE

    (2010b) Epidemiology of rheumatoid arthritis: rheumatoid arthritis and mortality

    Current Rheumatology Reports 12:379–385.

    https://doi.org/10.1007/s11926-010-0117-y

    • PubMed
    • Google Scholar
41. 1. Palladino MA
    2. Bahjat FR
    3. Theodorakis EA
    4. Moldawer LL

    (2003) Anti-TNF-alpha therapies: the next generation

    Nature Reviews Drug Discovery 2:736–746.

    https://doi.org/10.1038/nrd1175

    • PubMed
    • Google Scholar
42. 1. Parameswaran N
    2. Patial S

    (2010) Tumor necrosis factor-α signaling in macrophages

    Critical Reviews™ in Eukaryotic Gene Expression 20:87–103.

    https://doi.org/10.1615/CritRevEukarGeneExpr.v20.i2.10

    • PubMed
    • Google Scholar
43. 1. Peschon JJ
    2. Slack JL
    3. Reddy P
    4. Stocking KL
    5. Sunnarborg SW
    6. Lee DC
    7. Russell WE
    8. Castner BJ
    9. Johnson RS
    10. Fitzner JN
    11. Boyce RW
    12. Nelson N
    13. Kozlosky CJ
    14. Wolfson MF
    15. Rauch CT
    16. Cerretti DP
    17. Paxton RJ
    18. March CJ
    19. Black RA

    (1998) An essential role for ectodomain shedding in mammalian development

    Science 282:1281–1284.

    https://doi.org/10.1126/science.282.5392.1281

    • PubMed
    • Google Scholar
44. 1. Reddy P
    2. Slack JL
    3. Davis R
    4. Cerretti DP
    5. Kozlosky CJ
    6. Blanton RA
    7. Shows D
    8. Peschon JJ
    9. Black RA

    (2000) Functional analysis of the domain structure of tumor necrosis factor-alpha converting enzyme

    Journal of Biological Chemistry 275:14608–14614.

    https://doi.org/10.1074/jbc.275.19.14608

    • PubMed
    • Google Scholar
45. 1. Ryves WJ
    2. Evans AT
    3. Olivier AR
    4. Parker PJ
    5. Evans FJ

    (1991) Activation of the PKC-isotypes alpha, beta 1, gamma, delta and epsilon by phorbol esters of different biological activities

    FEBS Letters 288:5–9.

    https://doi.org/10.1016/0014-5793(91)80989-G

    • PubMed
    • Google Scholar
46. 1. Saarinen S
    2. Vahteristo P
    3. Lehtonen R
    4. Aittomäki K
    5. Launonen V
    6. Kiviluoto T
    7. Aaltonen LA

    (2012) Analysis of a finnish family confirms RHBDF2 mutations as the underlying factor in tylosis with esophageal Cancer

    Familial Cancer 11:525–528.

    https://doi.org/10.1007/s10689-012-9532-8

    • PubMed
    • Google Scholar
47. 1. Sahin U
    2. Weskamp G
    3. Kelly K
    4. Zhou HM
    5. Higashiyama S
    6. Peschon J
    7. Hartmann D
    8. Saftig P
    9. Blobel CP

    (2004) Distinct roles for ADAM10 and ADAM17 in ectodomain shedding of six EGFR ligands

    The Journal of Cell Biology 164:769–779.

    https://doi.org/10.1083/jcb.200307137

    • PubMed
    • Google Scholar
48. 1. Sapkota GP
    2. Cummings L
    3. Newell FS
    4. Armstrong C
    5. Bain J
    6. Frodin M
    7. Grauert M
    8. Hoffmann M
    9. Schnapp G
    10. Steegmaier M
    11. Cohen P
    12. Alessi DR

    (2007) BI-D1870 is a specific inhibitor of the p90 RSK (ribosomal S6 kinase) isoforms in vitro and in vivo

    Biochemical Journal 401:29–38.

    https://doi.org/10.1042/BJ20061088

    • PubMed
    • Google Scholar
49. 1. Schlöndorff J
    2. Becherer JD
    3. Blobel CP

    (2000) Intracellular maturation and localization of the tumour necrosis factor alpha convertase (TACE)

    Biochemical Journal 347 Pt 1:131–138.

    https://doi.org/10.1042/bj3470131

    • PubMed
    • Google Scholar
50. 1. Scott DL
    2. Kingsley GH

    (2006) Tumor necrosis factor inhibitors for rheumatoid arthritis

    New England Journal of Medicine 355:704–712.

    https://doi.org/10.1056/NEJMct055183

    • PubMed
    • Google Scholar
51. 1. Siggs OM
    2. Xiao N
    3. Wang Y
    4. Shi H
    5. Tomisato W
    6. Li X
    7. Xia Y
    8. Beutler B

    (2012) iRhom2 is required for the secretion of mouse TNFα

    Blood 119:5769–5771.

    https://doi.org/10.1182/blood-2012-03-417949

    • PubMed
    • Google Scholar
52. 1. Siggs OM
    2. Grieve A
    3. Xu H
    4. Bambrough P
    5. Christova Y
    6. Freeman M

    (2014) Genetic interaction implicates iRhom2 in the regulation of EGF receptor signalling in mice

    Biology Open 3:1151–1157.

    https://doi.org/10.1242/bio.201410116

    • PubMed
    • Google Scholar
53. 1. Sommer A
    2. Kordowski F
    3. Büch J
    4. Maretzky T
    5. Evers A
    6. Andrä J
    7. Düsterhöft S
    8. Michalek M
    9. Lorenzen I
    10. Somasundaram P
    11. Tholey A
    12. Sönnichsen FD
    13. Kunzelmann K
    14. Heinbockel L
    15. Nehls C
    16. Gutsmann T
    17. Grötzinger J
    18. Bhakdi S
    19. Reiss K

    (2016) Phosphatidylserine exposure is required for ADAM17 sheddase function

    Nature Communications 7:11523.

    https://doi.org/10.1038/ncomms11523

    • PubMed
    • Google Scholar
54. 1. Soond SM
    2. Everson B
    3. Riches DW
    4. Murphy G

    (2005) ERK-mediated phosphorylation of Thr735 in TNFalpha-converting enzyme and its potential role in TACE protein trafficking

    Journal of Cell Science 118:2371–2380.

    https://doi.org/10.1242/jcs.02357

    • PubMed
    • Google Scholar
55. 1. Sunnarborg SW
    2. Hinkle CL
    3. Stevenson M
    4. Russell WE
    5. Raska CS
    6. Peschon JJ
    7. Castner BJ
    8. Gerhart MJ
    9. Paxton RJ
    10. Black RA
    11. Lee DC

    (2002) Tumor necrosis factor-alpha converting enzyme (TACE) regulates epidermal growth factor receptor ligand availability

    Journal of Biological Chemistry 277:12838–12845.

    https://doi.org/10.1074/jbc.M112050200

    • PubMed
    • Google Scholar
56. 1. Trudgian DC
    2. Thomas B
    3. McGowan SJ
    4. Kessler BM
    5. Salek M
    6. Acuto O

    (2010) CPFP: a central proteomics facilities pipeline

    Bioinformatics 26:1131–1132.

    https://doi.org/10.1093/bioinformatics/btq081

    • PubMed
    • Google Scholar
57. 1. Trudgian DC
    2. Ridlova G
    3. Fischer R
    4. Mackeen MM
    5. Ternette N
    6. Acuto O
    7. Kessler BM
    8. Thomas B

    (2011) Comparative evaluation of label-free SINQ normalized spectral index quantitation in the central proteomics facilities pipeline

    PROTEOMICS 11:2790–2797.

    https://doi.org/10.1002/pmic.201000800

    • PubMed
    • Google Scholar
58. 1. Trudgian DC
    2. Mirzaei H

    (2012) Cloud CPFP: a shotgun proteomics data analysis pipeline using cloud and high performance computing

    Journal of Proteome Research 11:6282–6290.

    https://doi.org/10.1021/pr300694b

    • PubMed
    • Google Scholar
59. 1. Wang B
    2. Yang H
    3. Liu YC
    4. Jelinek T
    5. Zhang L
    6. Ruoslahti E
    7. Fu H

    (1999) Isolation of high-affinity peptide antagonists of 14-3-3 proteins by phage display

    Biochemistry 38:12499–12504.

    https://doi.org/10.1021/bi991353h

    • PubMed
    • Google Scholar
60. 1. Wiśniewski JR
    2. Zielinska DF
    3. Mann M

    (2011) Comparison of ultrafiltration units for proteomic and N-glycoproteomic analysis by the filter-aided sample preparation method

    Analytical Biochemistry 410:307–309.

    https://doi.org/10.1016/j.ab.2010.12.004

    • PubMed
    • Google Scholar
61. 1. Xu P
    2. Liu J
    3. Sakaki-Yumoto M
    4. Derynck R

    (2012) TACE activation by MAPK-mediated regulation of cell surface dimerization and TIMP3 association

    Science Signaling 5:ra34.

    https://doi.org/10.1126/scisignal.2002689

    • PubMed
    • Google Scholar

Thank you for submitting your article "Phosphorylation of iRhom2 at the plasma membrane controls TACE-dependent inflammatory and growth factor signalling" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Vivek Malhotra as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The manuscript by Grieve et al. investigates how phosphorylation of the rhomboid pseudoprotease iRhom2 controls the activity of the sheddase TACE, which serves as the key activator of TNFalpha and growth factor signaling. The authors add to a flurry of papers published in 2012 revealing an unanticipated role of iRhom2 in regulating ER-to-Golgi-trafficking and subsequent maturation of TACE. In this manuscript, Grieve et al. further elaborate their previous trafficking model, now providing strong evidence that iRhom2 and TACE remain associated throughout the entire secretion route to the plasma membrane. Interestingly, phosphorylation of the N-terminal cytoplasmic domain and subsequent binding of 14-3-3 proteins appears to weaken this interaction thereby increasing TACE shedding activity. Although the molecular mechanism of how TACE activity is controlled in this complex still remains unclear, data presented in this manuscript support an interesting new model that iRhom2 serves as multifunctional subunit of TACE, which simultaneously limits shedding activity and protects TACE from lysosomal degradation. The paper is well written and the experiments are sufficiently explained, but two major conceptual issues and several technical concerns need to be resolved in a revision:

1) An important conceptual gap is the identity of the kinase(s) and how they are controlled. At least the identity of the kinase(s) need to be clarified or narrowed down using well described kinase inhibitors or/and sets of siRNA after physiological (not PMA) stimulation. Without this information the underlying mechanism is not sufficiently developed to warrant publication in eLife.

2) The manuscript falls short in defining the role of iRhom2 phosphorylation on complex formation with TACE. From the data presented in the current manuscript (Figure 4F), the claim that upon PMA stimulation "the complex is more stable with the non-phosphorylatable mutant" (referred to as Site 1-3 mutant) is not convincingly supported, and serious concerns exist regarding the reliability of this data set. Since x-ray films used for ECL detection of western blots hardly provide a linear readout, the authors need an independent experimental approach to support this biotinylation/co-IP experiment (e.g. proximity ligation assays, or split GFP approach to demonstrate direct physical interaction between TACE and the iRhom2 N-terminal domain). The surface interaction of TACE and iRhom2 should be determined after physiological stimulation (not after PMA treatment). The authors also need to determine whether phosphorylation can also occur in the secretory route or only at the cell surface.

Technical comments:

Introduction:

In order to prevent confusion, the authors need to indicate that TACE also is known as ADAM17.

"[…] with the relatively slow process of ER-to-Golgi trafficking[…]" Define what is meant by relatively slow. Can the authors provide any numbers here? Why do the authors describe this as an "apparent paradox"? Trans-Golgi derived vesicles could also rapidly fuse with the plasma membrane.

The recent paper by Luo et al. describing iRhom2 as an essential factor for STING activity should be mentioned (not only in the Discussion section) suggesting role(s) of iRhom2 next to the regulation of TACE.

Results:

Do the authors also detect other post translational modifications next to phosphorylation? Mono-ubiquitinylation would be very interesting in the context of the suggested role of modulation of iRhom2´s endocytosis.

Subsection “TACE activity, but not TACE maturation, is regulated by iRhom2 phosphorylation”, second paragraph: iRhom´s only function is not only the regulation of ER-to-Golgi transport (see above).

Subsection “TACE activity, but not TACE maturation, is regulated by iRhom2 phosphorylation”, second paragraph: Steady state analysis of mature TACE level (Figure 1d and 5a) and cellular localization (Figure 1e and F) do not allow precisely measuring trafficking rates and therefore the conclusion that "any phosphorylation-dependent function must be restricted to an unknown post-TACE maturation role of iRhom2" and "trafficking is phosphorylation independent" need to be toned down.

How do the authors explain the higher molecular weight in the iR2Pdead mutant? It is very surprising that a significant fraction of the ER-retention mutant is also found in GM130 positive compartments similar to the WT and pDead mutants? The ER-retention mutant should also be included in the biotinylation assay shown in Figure 1F. How far do these over-expressed iRhom2 variants reflect the physiological intracellular steady-state distribution? Are the authors able to show an effect of the iR2Pdead mutant in comparison with the wildtype protein on an endogenous (not overexpressed alkaline phosphatase tagged) substrate of TACE? Is there any effect of the mutant without induction using PMA? It should be noted (and labelled) that these experiments are performed in mouse fibroblasts using PMA as a non-physiological kinase stimulant.

In Figure 2—figure supplement 1 (and in part also in Figure 2a) there is no major differences in TACE activation towards the TGF-α substrate and the authors´ conclusion that this revealed the three distinct sites that contribute to PMA-induced stimulation is unclear. Is the mass spectroscopy analysis performed earlier not a more reliable methodology here?

Subsection “Mature TACE activity is dependent upon iRhom2 phosphorylation at three distinct sites, which co-ordinate binding to 14-3-3 family proteins”, second paragraph: The authors report a proteomic screen, and therefore need to present data showing the entire list of interacting proteins found in this screen. This is an important point.

Do 14-3-3 proteins also bind to iRhom2 upon physiological (non PMA) stimulation?

Figure 3a, c: Only bead and only antibody controls are required for the indicated IP experiments. The multiple bands after HAIP are not explained. Is the (co)transport of the iRhom2 to the cell surface mutants affected? No controls are provided that the used inhibitors (BafA and MG132) worked. Bafilomycin not only affects lysosomal degradation. It would be interesting to analyse the effects of an inhibition of endocytosis or and lysosomal degradation by cathepsins.

Figure 4: improve the labelling of all subpanels.

Subsection “Phosphorylation and 14-3-3 protein recruitment uncouples the interaction between iRhom2 and mature TACE to drive its activity at the cell surface”, first paragraph: Does the R18 peptide sequence inserted in the N-terminus of iRhom2 affect its subcellular trafficking?

Why is the complex of iRhom2 and TACE less stable after expression of iR2Site1-3 mutant and 5 min PMA treatment?

[…] The paper is well written and the experiments are sufficiently explained, but two major conceptual issues and several technical concerns need to be resolved in a revision:

1) An important conceptual gap is the identity of the kinase(s) and how they are controlled. At least the identity of the kinase(s) need to be clarified or narrowed down using well described kinase inhibitors or/and sets of siRNA after physiological (not PMA) stimulation. Without this information the underlying mechanism is not sufficiently developed to warrant publication in eLife.

We have addressed this important point extensively, leading to the identification of ERK1/2 as being essential kinases in this process. This has led to the addition of a new Figure 3. Specifically:

Upon re-examination of our mass spectrometry hits for iRhom2 interactors, we found that both ERK1 and ERK2, and RSK3 were significant and reproducible hits. Therefore, we screened their activity with their respective inhibitors.

We found that inhibition of ERK1/2 signalling with U0126, a widely used inhibitor of their upstream kinases MEK1/2, reduced TACE-dependent shedding after stimulation with both the non-physiological stimulant, PMA, and a physiological trigger for TACE activation, histamine.

We analysed 14-3-3 recruitment after TACE activation with PMA and histamine -/+ U0126, and found that inhibition of ERK1/2 phosphorylation strongly and reproducibly inhibited 14-3-3 recruitment to iRhom2.

Analysis of these signalling pathways with U0126 confirmed that both PMA and histamine stimulate ERK1/2 signalling – and that U0126 blocks the phosphorylation of ERK1/2.

Consistent with ERK1/2 functioning upstream of iRhom2, we found that ERK phosphorylation was unaltered in iRhom2 site 1-3 mutants.

We also found that serine phosphorylation of iRhom2 was lost upon treatment with U0126.

These data have led to the creation of a new figure, Figure 3, and an updated Abstract.

In summary, we have found that iRhom2 phosphorylation is dependent on ERK1/2 (though it would take work beyond the scope of this project to prove that iRhom2 is a direct target). We show that PMA and histamine, both of which stimulate iRhom2 phosphorylation-mediated TACE activation, activate ERK1/2. Interestingly, our findings are consistent with previous reports that LPS stimulates ERK1/2, which is required for TNF release in macrophages (Dumitru CD et al., 2000, Cell; Eliopoulos AG et al., 2003, EMBO J). This suggests that ERK1/2 dependent iRhom2 phosphorylation may be a conserved mechanism for TACE activation in macrophages, now mentioned in the discussion.

There is also overlap with another study of TACE activation (Xu P, et al. 2012, Science Signalling), where it was shown that ERK and p38 MAPK signalling lead to TACE activation that coincide with TACE monomerisation and changes in surface presentation. The overlap between this study and ours is covered in the Discussion.

2) The manuscript falls short in defining the role of iRhom2 phosphorylation on complex formation with TACE. From the data presented in the current manuscript (Figure 4F), the claim that upon PMA stimulation "the complex is more stable with the non-phosphorylatable mutant" (referred to as Site 1-3 mutant) is not convincingly supported, and serious concerns exist regarding the reliability of this data set. Since x-ray films used for ECL detection of western blots hardly provide a linear readout, the authors need an independent experimental approach to support this biotinylation/co-IP experiment (e.g. proximity ligation assays, or split GFP approach to demonstrate direct physical interaction between TACE and the iRhom2 N-terminal domain). The surface interaction of TACE and iRhom2 should be determined after physiological stimulation (not after PMA treatment). The authors also need to determine whether phosphorylation can also occur in the secretory route or only at the cell surface.

Actually, our work does not demonstrate a direct interaction between the iRhom2 N-terminus and the cytoplasmic domain of TACE. Instead, we favour a model in which N-terminal phosphorylation and 14-3-3 recruitment to iRhom2 leads to a change in the TMD-based interaction between iRhom2 and TACE at the cell surface (see Discussion). As iRhom2 and both the pro and mature forms of TACE interact throughout the entire secretory pathway, a selective approach is required to assess the interaction at the cell surface. This is quite difficult, as the cell surface pool is only a small fraction of the total, but it is what we showed in the original Figure 4F.

We do accept the reviewers’ point that it would be good to provide a different type of experimental evidence to support the conclusion that there is a change in the interaction upon 14-3-3 binding and we have worked hard to address it.

A split GFP approach will not work because the lifetime interaction between iRhom2 and TACE would mean that the fluorescent GFP signal would be detected at all subcellular sites and is not specific to the small proportion at the plasma membrane.

PLA with unpermeabilised cells, using antibodies that recognise extracellular epitopes of iRhom2 and TACE, is a good idea and should be able to detect only the plasma membrane interaction. Following this suggestion, we have attempted PLA with iRhom2 and TACE, using the extracellular epitope of iRhom2-HA and an antibody that detects the TACE extracellular epitope. However, these experiments were unsuccessful, even in permeabilised cells. We then generated a cell line that expresses iRhom2-HA and TACE-FLAG, with a FLAG epitope placed just before the furin cleavage site- thus allowing us to detect the FLAG epitope at the cell surface. PLA using antibodies against HA and FLAG with this also failed. We therefore conclude that this approach does not work, and suggest that, as in some other cases, steric issues in the complex between iRhom2 and TACE prevent PLA from working – at least with these epitopes.

It is worth restating here that the quantification in the previous Figure 4F (now Figure 5) represent three separate experiments, giving us confidence that the reduction of interaction between iRhom2 and TACE that we detect at the cell surface is robust and reproducible.

Further supporting our interpretation, constitutive 14-3-3 recruitment leads to a loss of interaction between mature TACE and iRhom2. This leads to elevated TACE activity even without PMA treatment. This is fully consistent with the idea that 14-3-3 binding alters the interaction between iRhom2 and mature TACE, thereby stimulating TACE activity.

Nevertheless, we fully accept that without a completely independent experimental approach to this difficult problem, we need to word this conclusion more cautiously, and we have now done this in both the Results and Discussion.

We have additionally addressed the final sentence of this point, about the cellular location of iRhom2 phosphorylation. We performed an experiment with iRhom2-KDEL, to assess whether this mutant of iRhom2, which is restricted to the early secretory pathway and does not support TACE-dependent shedding, could recruit 14-3-3 upon PMA stimulation. We found that iRhom2-KDEL recruited 14-3-3 in a similar fashion to iRhom2-WT (Figure 2—figure supplement 1g). We therefore believe that iRhom2 is phosphorylated throughout the secretory pathway, but the only effect of iRhom2 phosphorylation on TACE that we can detect is at the cell surface.

Technical comments:

Introduction:

In order to prevent confusion, the authors need to indicate that TACE also is known as ADAM17.

In the earliest description of TACE in the Introduction, we now state that it is also known as ADAM17 (first paragraph).

"[…] with the relatively slow process of ER-to-Golgi trafficking[…]" Define what is meant by relatively slow. Can the authors provide any numbers here? Why do the authors describe this as an "apparent paradox"? Trans-Golgi derived vesicles could also rapidly fuse with the plasma membrane.

Until now, the only known function of iRhom2 in TACE regulation is the ability to promote ER to Golgi transport and maturation of TACE. Although it is formally possible that the TACE activation mechanism could be through gating of trans Golgi derived vesicles, this is shown not to be the case, as phosphorylation-defective mutants of iRhom2 do not affect its own surface presentation, nor the ability of mature TACE to reach the cell surface. The maturation of TACE is already well studied and occurs within a time frame of 3-6 hours (Schlondorff et al., 2000), which is not on the timescale of TACE activation (within 10 mins). We have made a comment to this effect in the manuscript (Introduction, third paragraph).

The recent paper by Luo et al. describing iRhom2 as an essential factor for STING activity should be mentioned (not only in the Discussion section) suggesting role(s) of iRhom2 next to the regulation of TACE.

The Luo et al. publication has now been mentioned in the Introduction, in the context of iRhom2 regulation of protein stability (second paragraph).

Results:

Do the authors also detect other post translational modifications next to phosphorylation? Mono-ubiquitinylation would be very interesting in the context of the suggested role of modulation of iRhom2´s endocytosis.

We do have some preliminary mass spec data potentially detecting mono-ubiquitination, but this has not been confirmed or validated so is not appropriate for publication without further work.

Subsection “TACE activity, but not TACE maturation, is regulated by iRhom2 phosphorylation”, second paragraph: iRhom´s only function is not only the regulation of ER-to-Golgi transport (see above).

Thank you for bringing this error to light. We have correctly reworded this sentence, to make clear the statement is with reference to TACE (subsection “TACE activity, but not TACE maturation, is regulated by iRhom2 phosphorylation”, second paragraph).

Subsection “TACE activity, but not TACE maturation, is regulated by iRhom2 phosphorylation”, second paragraph: Steady state analysis of mature TACE level (Figure 1d and 5a) and cellular localization (Figure 1e and F) do not allow precisely measuring trafficking rates and therefore the conclusion that "any phosphorylation-dependent function must be restricted to an unknown post-TACE maturation role of iRhom2" and "trafficking is phosphorylation independent" need to be toned down.

We have toned down the first statement to a hypothesis. We have rephrased the final conclusion to make a more specific statement that the ER-to-Golgi traffic of iRhom2 and TACE is independent of phosphorylation, which is supported by our data (TACE still undergoes ER-to-Golgi transport and maturation in iRhom2 phosphorylation mutant cell lines; iRhom2 pDEAD reaches the plasma membrane to a similar level as its wild type counterpart, therefore there is no block in ER-to-Golgi traffic).

How do the authors explain the higher molecular weight in the iR2Pdead mutant?

There are a large number of mutations in iRhom2 pDEAD that are likely to affect its overall charge in a manner likely unrelated to phosphorylation. For this reason we performed the screen in the manuscript to create iRhom2 with the site1-3 mutant, which is more minimal that the full pDEAD mutant, and in which this upward shift in molecular weight is not apparent. Depending on the duration of the SDS-PAGE gel that is run, an expected faster migration in the iRhom2 site 1-3 phosphorylation mutants can be observed e.g. HA lysates Figure 3e/3F, Figure 5f, Figure 6c.

It is very surprising that a significant fraction of the ER-retention mutant is also found in GM130 positive compartments similar to the WT and pDead mutants? The ER-retention mutant should also be included in the biotinylation assay shown in Figure 1F.

The KDEL insertion does not prevent Golgi localisation, it ensures retrieval from the cis-Golgi. We have reworded the text to make this clearer. Both biotinylation experiments and cell surface FACS analysis show a significant reduction in surface levels of iRhom2-KDEL. These data are included in Figure 1—figure supplement 2.

How far do these over-expressed iRhom2 variants reflect the physiological intracellular steady-state distribution? Are the authors able to show an effect of the iR2Pdead mutant in comparison with the wildtype protein on an endogenous (not overexpressed alkaline phosphatase tagged) substrate of TACE?

Without an antibody to iRhom2 (despite many attempts), it is difficult to answer the first question. We have avoided gross over-expression to the best of our ability, by ensuring low expression by infection with low titres of virus (aiming for an M.O.I. of 1 virus particle per cell). We provide evidence of an effect of iRhom2 phosphorylation on TACE-dependent release of endogenous TNF in macrophages, by ELISA (Figure 6d).

Is there any effect of the mutant without induction using PMA? It should be noted (and labelled) that these experiments are performed in mouse fibroblasts using PMA as a non-physiological kinase stimulant.

We had already addressed this important issue in Figure 2 but have now substantially strengthened the evidence that physiological GPCR stimulation of TACE activity is dependent on iRhom2 phosphorylation downstream of ERK1/2 (new Figure 3). This is also demonstrated with LPS-stimulated TACE activity in primary macrophages (Figure 5). In the Results section where we first introduce the use of histamine, we have stated that “PMA is not a physiological trigger for TACE activity”.

In Figure 2—figure supplement 1 (and in part also in Figure 2a) there is no major differences in TACE activation towards the TGF-α substrate and the authors´ conclusion that this revealed the three distinct sites that contribute to PMA-induced stimulation is unclear. Is the mass spectroscopy analysis performed earlier not a more reliable methodology here?

We demonstrate that sites 1, 2 and 3 independently and significantly affect TACE-dependent shedding in an additive fashion. Within sites 1, 2 and 3, some residues are more critical than others, which is reflected by the lack of major differences in TACE activation in some cases.

Subsection “Mature TACE activity is dependent upon iRhom2 phosphorylation at three distinct sites, which co-ordinate binding to 14-3-3 family proteins”, second paragraph: The authors report a proteomic screen, and therefore need to present data showing the entire list of interacting proteins found in this screen. This is an important point.

We now include data from the mass spec screen identifying the kinases involved in iRhom2 phosphorylation (Figure 3). We are reluctant to provide the whole list that was obtained from a different proteomic project, which it is not directly relevant to this project, mainly on the grounds that most of the hits have not been validated. Those that have been are being worked on separately.

Do 14-3-3 proteins also bind to iRhom2 upon physiological (non PMA) stimulation?

We include new data demonstrating that 14-3-3 proteins bind to iRhom2 upon stimulation with the physiological stimulus, histamine, in an ERK1/2 signalling dependent manner (Figure 3). This is in addition to the data that were already included, which showed that 14-3-3 proteins are also recruited to iRhom2 upon physiological stimulation of macrophages with LPS, in a phosphorylation-dependent manner (Figure 5).

Figure 3a, c: Only bead and only antibody controls are required for the indicated IP experiments. The multiple bands after HAIP are not explained. Is the (co)transport of the iRhom2 to the cell surface mutants affected? No controls are provided that the used inhibitors (BafA and MG132) worked. Bafilomycin not only affects lysosomal degradation. It would be interesting to analyse the effects of an inhibition of endocytosis or and lysosomal degradation by cathepsins.

The interaction between iRhom2 and TACE has already been well documented and shown to be specific. For this reason, we typically use commercially available pre-conjugated HA beads for all immunoprecipitation experiments, that have undergone rigorous in-house testing. Nonetheless, first, we performed a repeat IP with control beads without HA antibody, to demonstrate the specificity of our co-immunoprecipitation of TACE with iRhom2 (Figure 4—figure supplement 1a). Second, the beads used in the new Figure 4c are for enrichment of glycoproteins, not immunoprecipitations. All lysates are treated equally and, given the new control provided in Figure 4—figure supplement 1a with beads only, we do not believe there is a need for bead only controls in this ConA experiment (empty beads do not concentrate glycoproteins, such as TACE). Third, we include the whole western blot in these immunoprecipitation experiments so that the respective size of each mutant is clearly demonstrated. The lower molecular weight bands have not been well characterized, but are reproducible and presumably represent degradation products. Fourth, we provide a cell surface biotinylation assay of these mutants to demonstrate their ability to reach the cell surface (Figure 4—figure supplement 1b). Fifth, to demonstrate that MG132 has worked, we include a new figure that shows increased ubiquitin levels specifically upon treatment with MG132, demonstrating that the rescue of mature TACE levels is dependent on lysosomal degradation, with no influence from the ubiquitin-proteasome pathway (Figure 4—figure supplement 1c). Sixth, in the new Figure 4, we include a panel for ADAM10, which is known to undergo endocytosis and degradation from the cell surface (Marcello E et al., 2013, JCI), showing that bafilomycin treatment leads to an increase in ADAM10 levels. We agree that bafilomycin does not only inhibit endocytic degradation but since it is a specific inhibitor of the vacuolar-type H⁺-ATPase on lysosomes, we prefer to keep to our current wording that mature TACE is degraded in lysosomes.

Figure 4: improve the labelling of all subpanels.

We have changed the western blot subpanel labelling from white to black, for clarity.

Subsection “Phosphorylation and 14-3-3 protein recruitment uncouples the interaction between iRhom2 and mature TACE to drive its activity at the cell surface”, first paragraph: Does the R18 peptide sequence inserted in the N-terminus of iRhom2 affect its subcellular trafficking?

We now provide a blot from a surface biotinylation assay that demonstrates R18-iRhom2 reaches the plasma membrane in a similar fashion to both iRhom2 WT and Site 1-3, relative to their expression level (Figure 5—figure supplement 1).

Why is the complex of iRhom2 and TACE less stable after expression of iR2Site1-3 mutant and 5 min PMA treatment?

Quantification from three independent experiments shows that the iRhom2-mTACE complex is less stable after 5 min PMA treatment; however, there is no statistically significant difference in the stability of the iRhom2-mTACE complex in the iR2 site 1-3 mutant (Figure 5f).

Author details

1. Adam Graham Grieve

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Contribution

   AGG, Conceptualization, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Hongmei Xu

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0001-6420-5724

2. Hongmei Xu

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Present address

   National Key Laboratory of Medical Immunology, Institute of Immunology, Second Military Medical University, Shanghai, China

   Contribution

   HX, Conceptualization, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Adam Graham Grieve

   Competing interests

   No competing interests declared.

3. Ulrike Künzel

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Contribution

   UK, Conceptualization, Investigation, Methodology, Writing—original draft

   Competing interests

   No competing interests declared.

4. Paul Bambrough

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Contribution

   PB, Conceptualization, Investigation, Methodology

   Competing interests

   No competing interests declared.

5. Boris Sieber

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Contribution

   BS, Investigation, Methodology, Writing—original draft

   Competing interests

   No competing interests declared.

6. Matthew Freeman

   Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

   Contribution

   MF, Conceptualization, Supervision, Funding acquisition, Writing—original draft, Writing—review and editing

   For correspondence

   matthew.freeman@path.ox.ac.uk

   Competing interests

   MF: Reviewing editor, eLife

   "This ORCID iD identifies the author of this article:" 0000-0003-0410-5451

• 2,844

  Page views

• 601

  Downloads

• 70

  Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.