Dissection of a Novel Autocrine Signaling Pathway ... - ACS Publications

Jun 12, 2014 - We show here that the signal transducer/transcription factor FIZ1 (Flt3 ... human immortalized keratinocytes that is driven by more rap...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/jpr

Dissection of a Novel Autocrine Signaling Pathway via Quantitative Secretome and Interactome Mapping Nathalie Larivière,†,‡ Jennifer Law,† and Laura Trinkle-Mulcahy*,† †

Department of Cellular & Molecular Medicine and Ottawa Institute of Systems Biology, University of Ottawa, 451 Smyth Road, Ottawa, Ontario K1H 8M5, Canada ‡ The Ottawa Hospital Research Institute, 725 Parkdale Avenue, Ottawa, Ontario K1Y 4E9, Canada S Supporting Information *

ABSTRACT: Epidermal homeostasis is a balancing act governed by a multitude of underlying regulatory events, and several growth factors and signaling pathways have been implicated in regulation of the balance between proliferation and differentiation in keratinocytes. We show here that the signal transducer/transcription factor FIZ1 (Flt3 interacting zinc finger protein-1) is a previously unknown player in this regulatory axis, promoting an increase in proliferation of HaCaT human immortalized keratinocytes that is driven by more rapid G1/S progression and mediated by activation of the MAP/ERK kinase pathway. Utilizing quantitative SILAC-based secretome analysis, we identified the insulin growth factor binding protein IGFBP3 as the key mediating factor, demonstrating that elevated FIZ1 levels promote increased IGFBP3 expression and secretion and a concurrent increased sensitivity to IGF1 signaling, while antibodybased neutralization of IGFBP3 abrogates the FIZ1-induced growth advantage. To identify underlying protein−protein interactions likely to govern these events, we mapped the interactome of FIZ1 and found eight novel binding partners that form complexes with the protein in the cytoplasm and nucleus. These include signal transduction and transcription factors and the cell cycle regulatory NDR (Nuclear Dbf2-related) kinases. Our results provide further insight into the complex balance of epidermal homeostasis and identify FIZ1 as a novel therapeutic target. KEYWORDS: SILAC, proteomics, growth factors, keratinocytes, proliferation, FIZ1

■

FIZ1 (Flt3 interacting zinc finger protein-1), which was originally identified as a novel interactor for the FLT3 receptor tyrosine kinase.4 FLT3 (FMS-like receptor tyrosine kinase 3) is a key therapeutic target that is mutated in 30% of acute myeloid leukemia (AML) patients and associated with poor prognosis and has also been linked to autoimmune diseases. (See ref 5 for review.) FIZ1 has been implicated in FLT3 signaling6 and is widely expressed,4 but most studies have focused on its role as a transcriptional repressor on active photoreceptor-specific gene promoters in retinal cells.7−9 We set out to functionally assess the downstream effects of altered FIZ1 levels in the HaCaT human keratinocyte cell line, which is a widely used model system for the analysis of epidermal homeostasis. HaCaT cells retain all functional differentiation properties of normal keratinocytes, as when grafted onto severe combined immunodeficiency (SCID) mice or grown in organotypic cultures they undergo normal keratinization and stratification occurs.10 Using a range of cell biology and quantitative proteomic assays, we demonstrated a previously unknown regulatory role for this protein in the balance of proliferation and differentiation. FIZ1 is pro-proliferative in these cells,

INTRODUCTION The mammalian epidermis is a stratified squamous epithelium that forms the selective epidermal permeability barrier essential for mammalian survival. Tight regulation of the processes that underlie self-renewal of the mature epidermis and wound healing is critical, with dysfunctions in the homeostatic balance between epidermal proliferation and differentiation leading to diseases such as atopic dermatitis, psoriasis, and cancer.1 Development begins with commitment of a single layer of multipotent ectodermal cells to an epidermal cell fate and lineage. Keratinocytes in the proliferative basal compartment receive signals to differentiate and exit this layer, becoming irreversibly committed to terminal differentiation (for review, see ref 2). The cells exit the cell cycle and initiate expression of differentiation markers, moving suprabasally as they become part of the upper stratified layers. The end result is a barrier separating the external and internal environments of the organism, with selective permeability to solutes regulated by tight junction complexes.3 Although the contributions of several known signaling pathways have been evaluated, a comprehensive list of the key players and how all of the underlying events are coordinated remains elusive. A preliminary screen for proteins with the potential to mediate tight junction signaling in epidermal cells turned up © XXXX American Chemical Society

Received: April 17, 2014

A

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Aldrich) for 2 h or GFP-Trap_A (Chromotek) for 1 h. For endogenous FIZ1 pulldowns, α-FIZ1 (Abcam) or purified rabbit IgG (control) were covalently conjugated to protein G Sepharose beads (GE Healthcare) at 1 mg/mL. Western blots were imaged/quantified using an LAS4000 chemiluminescence system (GE Healthcare). See the Supporting Information for a detailed list of primary antibodies. HRP-conjugated secondary antibodies were obtained from Thermo Scientific. Total cellular RNA was extracted from cells and reversetranscribed, and transcripts were amplified as previously described.12 For qPCR, 4 μL of each cDNA reaction was analyzed in duplicate using iTaq Universal SYBR Green Supermix (BioRad) and a Rotor-Gene Q PCR Machine (Qiagen). The comparative CT method13 was used to compare relative levels of mRNA in HaCaTFIZ1 versus parental HaCaT cells, using GAPDH as an internal standard. See the Supporting Information for a full list of PCR primers.

inducing increased expression and secretion of autocrine growth factors, activation of MAP/ERK signaling, and accelerated G1/S progression. Furthermore, we identified novel interactions between FIZ1 and proteins involved in these pathways in both the cytoplasm and nucleus, demonstrating its potential to act as both a signal transducer and transcriptional regulator.

■

EXPERIMENTAL PROCEDURES

Plasmids and Cell Culture

The pCMV6-XL5-FIZ1-Myc-DDK construct encodes fulllength human FIZ1 fused to C-terminal Myc and FLAG tags. Coding sequences for FIZ1 interactors were amplified from donor constructs (Supporting Information) and subcloned into pEGFP-C1 (Clontech). Constructs were confirmed by DNA sequencing. HaCaT spontaneously immortalized human keratinocytes (ATCC) were maintained at 37 °C/5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS and 100 U/mL penicillin/streptomycin. HaCaTFIZ1 cell lines were established by stable incorporation of transiently transfected plasmids and clonal selection in G418-containing media. All cell counts were performed in triplicate using the Countess cell counter (Life Technologies), with Day 0 counts carried out 2 h after seeding to confirm the number of adhered cells. For FGF7/KGF stimulation, cells were plated at a lower density, and the media changed at Day 2 to include 10 ng/mL FGF7/KGF. Cell cycle progression was monitored by FACS as previously described.11 For RNAi experiments, cells were transfected 18 h after plating with 1 μg of either FIZ1-targeted siRNA duplexes or a nontargeted control duplex (see Table S2 in the Supporting Information for sequences), using DharmaFECT (Dharmacon).

Fluorescence Imaging

For direct detection of endogenous proteins, cells were grown on coverslips, PFA-fixed, permeabilized, stained, and mounted as previously described.12 Anti-Occludin (Life Technologies) and anti-FIZ1 (Abcam) primary antibodies were detected using rabbit DyLight488 secondary antibodies (Jackson Laboratories). Images were acquired using a DeltaVision CORE widefield fluorescence system fitted with a 60× NA 1.4 PlanApochromat objective (Olympus) and CoolSNAP coupledcharge device (CCD) camera (Roper Scientific). The microscope was controlled and images processed by SoftWorX acquisition and deconvolution software (GE Healthcare). All images are single, deconvolved optical sections. For fluorescence imaging of GFP-tagged interactors, HaCaT cells were grown on coverslips and transfected using Effectene (Qiagen), followed by PFA fixation and mounting 18 h later. For nuclear retention assays, proliferating (Day 2) and differentiating (Day 10) cells grown on coverslips were incubated for 5 min with 0.1% Triton X-100 in ASE buffer (20 mM Tris pH 7.5, 5 mM MgCl2, 0.5 mM EGTA) at 37 °C prior to PFA-fixation and staining. For nuclease digestion, benzonase was added at 125U/ coverslip. Mock permeabilized samples were incubated for 5 min in ASE buffer with no Triton X-100.

Western Blotting, Affinity Purification, and PCR

Whole cell extracts were prepared by sonication (6 × 10 s at 75% power) in ice-cold RIPA buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate, protease inhibitors) and cleared by centrifuging at 2800g for 10 min at 4 °C. Cytoplasmic and nuclear extracts were prepared from subcellular fractions, which were obtained as follows: harvested cells were resuspended in 4 mL of ice-cold mild detergent buffer (20 mM Tris pH7.4, 10 mM KCl, 3 mM MgCl2, 0.1% NP40, 10% glycerol) for 10 min, followed by centrifugation at 1350g for 10 min at 4 °C to break open cells and release and pellet nuclei and other fragments. The supernatant was retained as the cytoplasmic fraction, while the nuclear pellet was resuspended in 3 mL of 0.25 M sucrose/10 mM MgCl2, layered over a 3 mL cushion of 0.35 M sucrose/0.5 mM MgCl2 and centrifuged at 1430g for 5 min at 4 °C. Cytoplasmic extracts were prepared by adding 1 mL of ice-cold 5× RIPA buffer to the 4 mL cytoplasmic fraction, following by a brief sonication and centrifugation at 2800g for 10 min at 4 °C. Nuclear extracts were prepared by resuspending the purified nuclei in 5 mL of 1× RIPA buffer, sonicating, and clearing as previously described. For membrane isolation from cytoplasmic fractions, the fractions were subjected to centrifugation at 100 000 g for 1 h at 4 °C and membrane extracts prepared by solubilizing the membrane fractions in urea/SDS. Total protein concentrations were measured using the Pierce BCA Protein Assay Kit (Thermo Scientific). For affinity purification of FLAG- and GFP-tagged proteins, extracts were incubated with either FLAG-M2 agarose (Sigma-

Mass Spectrometry

For quantitative secretome and interactome analysis, cells were differentially labeled by growth in SILAC DMEM supplemented with 10% dialyzed FBS and 100 U/mL penicillin/ streptomycin and containing either “light” L-arginine and Llysine (Sigma-Aldrich; R0K0) or “heavy” L-arginine 13C and Llysine 4,4,5,5-D4 (Cambridge Isotope Laboratories; R6K4). To compare the heparin-binding secretomes of the two cell lines, we labeled 4 × 15 cm dishes of “light” HaCaT and 4 × 15 cm dishes of “heavy” HaCaTFIZ1 cells for 7 days in serumcontaining SILAC media, and then switched to serum-free SILAC media for 3 days. For each condition, 60 mL of conditioned serum-free SILAC media was collected, cleared (1000 rpm × 4 min; a higher speed preclearing spin is recommended for nonquantitative approaches, to more efficiently remove smaller cellular debris) and incubated with 250 μL of heparin-agarose for 1 h at 4 °C. Beads were washed and combined and proteins were eluted for gel separation and trypsin digestion, as previously described.11 An aliquot of each tryptic digest was analyzed by LC−NSI− MS/MS (liquid chromatography nanospray ionization tandem mass spectrometry) on an LTQ Orbitrap XL hybrid mass B

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 1. Altering FIZ1 levels affects keratinocyte proliferation. (A) Seven-day growth curves demonstrating the effects of sustained knockdown of endogenous FIZ1 by siRNA (dotted blue line vs solid blue line) and stable overexpression of exogenous FIZ1 (solid red line) in HaCaT cells. (B) Growth curves demonstrating the effect of reducing FIZ1 levels back to near endogenous levels in HaCaTFIZ1 cells by siRNA (dotted red line vs solid red line). (C) Western blot analysis confirming sustained siRNA-mediated knockdown of endogenous FIZ1 levels by >80% in parental HaCaT cells. (D) Western blot analysis confirming sustained siRNA-mediated knockdown of FIZ1 levels in HaCaTFIZ1 cells to near endogenous levels (dotted line). (E) Representative FACS analyses of proliferating HaCaT (blue) and HaCaTFIZ1 (red) cell populations. (F) Summary of FACS analyses, showing distribution of cells between cell cycle stages. (G) Quantitation of Western blot analysis of Cyclin E levels (normalized to GAPDH) in proliferating cell lysates. For all experiments, data are plotted as mean ± SE for at least three biological replicates. Symbols (*, #, ‡) indicate statistically significant differences for a threshold p < 0.01.

using data-dependent acquisition of the top five ions from each MS scan. Database searching (against the human Uniprot database) and quantitation were performed using MaxQuant software v1.2.7.4.14 The following criteria were used: peptide tolerance = 10 ppm, trypsin as the enzyme (two missed cleavages allowed), and carboxyamidomethylation of cysteine as a fixed modification. Variable modifications were oxidation of methionine and N-terminal acetylation. Heavy SILAC labels were Arg6 (R6) and Lys4 (K4). Minimum ratio count was 2, and quantitation was based on razor and unique peptides. Peptide

spectrometer with nanospray source (Thermo Scientific) and an UltiMate 3000 RSLC nano HPLC (Dionex). The system was controlled by Xcalibur software, version 2.0.7 (Thermo Scientific). Peptides were loaded onto a trap column (C18 CapTrap, Michrom USA) for 5 min at 15 μL/min, then eluted over a 60 min gradient of 3−45% acetonitrile with 0.1% formic acid at 0.3 μL/min onto a 10 cm long column with integrated emitter tip (Picofrit PF360-75-15-N-5; New Objective) packed with Zorbax SB-C18 5 μm (Agilent), and nanosprayed into the mass spectrometer. MS scans were acquired in the Orbitrap module, and MS2 scans were acquired in the ion trap module C

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 2. FIZ1 overexpression overrides Ca2+ but not contact-induced differentiation. (A) Western blot analysis of cell lysates collected on Day 2 (proliferating), Day 6 (confluent/early differentiation), and Day 10 (late differentiation) postseeding to assess levels of total (α-FIZ1) and tagged (αFLAG) FIZ1 and markers for proliferation (α-KI67) and differentiation (α-INV). GAPDH is included as a loading control. (B) Immunostaining with α-Occludin (green) at Days 2 and 10. DNA was stained using DAPI (blue). Images are single optical sections and scale bars are 5 μm. (C) Growth curves for HaCaT (blue) and HaCaTFIZ1 (red) cells plated in low Ca2+ (0.03 mM) media and at Day 2 either maintained in this (solid lines) or switched to high (2.8 mM) Ca2+ media (dashed lines). (D) Areas under the curve (AUCs) calculated for panel C. (E) Quantitation of Western blot analysis of Involucrin levels (normalized to GAPDH) at Day 5 with and without the Ca2+ switch. (F) Western blot assessing changes in global tyrosine phosphorylation in response to the Ca2+ switch. Red arrow indicates the phospho-p62 band. (G) Quantitation of phosTyr-p62 levels relative to Tubulin. For all experiments, data are plotted as mean ± SE for three biological replicates. Symbols (*, #) indicate statistically significant differences for a threshold p < 0.01.

Statistical Analysis

and protein FDR was 0.01. All data sets (minus common environmental contaminants as per http://maxquant.org and proteins identified via the decoy database) are provided in the Supporting Information. For interactome mapping by quantitative AP/MS, 10 × 15 cm dishes of cells were encoded in SILAC media (“light” HaCaT and “heavy” HaCaTFIZ1) for 14 days, passaging at days 2, 6, 9, and 12 to maintain in a proliferative state. Equal protein amounts of whole cell extract were incubated with 100 μL of FLAG-M2 agarose for 2 h at 4 °C, the beads were washed and combined, and the proteins were eluted for gel separation, trypsin digestion, and LC−MS/MS analysis, as previously described.

Three independent experiments were carried out for all growth assays, FACS analyses, and quantified Western blots. AUCs were calculated and statistical analyses (ANOVA and Student’s t tests) performed using GraphPad Prism 5.0 (GraphPad Software).

■

RESULTS AND DISCUSSION

Elevated FIZ1 Levels Promote Keratinocyte Proliferation

While RNAi-mediated knockdown of endogenous FIZ1 levels (by ∼80%) in HaCaT cells did not affect the rate of proliferation at early time points in a growth curve, cells were D

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 3. FIZ1-induced proliferation is mediated by autocrine signaling. (A) Growth curves for HaCaT (blue) and HaCaTFIZ1 (red) cells in standard (10%, solid lines) and reduced (2%, dashed lines; 0%, dotted lines) serum conditions. (B) AUCs calculated for panel A) and expressed as a ratio of HaCaTFIZ1/HaCaT. (C) Two-day growth assays showing the significant (blue asterisk) increase in HaCaT proliferation when cells are grown in HaCaTFIZ1-conditioned media, and ablation of this growth advantage (pink asterisks) by addition of heparan sulfate (Hep; 20 μg/mL) or depletion of heparin-binding factors using heparin-agarose (Hep-Ag Depleted). (D) Two-day growth assays in serum-free media ± heparan sulfate. (E) Design of quantitative secretome experiment. (F) Plot of log H/L ratios for all proteins identified in heparin agarose eluants, highlighting those enriched in either HaCaTFIZ1 (red)- or HaCaT (blue)-conditioned media. For the growth curves, data are plotted as mean ± SE for three biological replicates. Symbols (*, #, ‡) indicate statistically significant differences for a threshold p < 0.01.

closure in scratch wound assays (Supporting Information), with ablation by treatment with the mitotic inhibitor mitomycin C, indicating that this is driven solely by enhanced proliferation and not migration. Analysis of HaCaTFIZ1 cell cycle distribution revealed a significant decrease in the percentage in G1 (Figure 1E,F), indicating faster progression through this stage. Consistent with this, HaCaTFIZ1 cells have higher levels of cyclin E (Figure 1G, Supporting Information), which, together with the cyclindependent kinase Cdk2, promotes progression to S phase.15

observed to exit the cell cycle prematurely (Figure 1A,C). Conversely, stable overexpression of FIZ1 (at levels 2.5 times that in parental cells; Figure 1C,D and the Supporting Information) promotes proliferation, with HaCaTFIZ1 cells reaching confluence a full day earlier than parental cells (Figure 1A). This was consistently observed for six separate clonal cell lines generated using two different FIZ1 expression plasmids. RNAi-mediated reduction of FIZ1 to near-endogenous levels in HaCaTFIZ1 cells (Figure 1D) leads to a concomitant drop in the proliferation rate (Figure 1B), confirming that it is excess FIZ1 that drives the enhanced proliferation. Despite their increased basal proliferation rate, HaCaTFIZ1 cells remain responsive to exogenous growth factors such as FGF7 (Supporting Information), indicating that mitogenic signaling is not saturated. Given that an inherent feature of keratinocytes is their ability to respond rapidly to epidermal injury with a surge in cell proliferation and migration to regenerate the damaged region, it is possible that FIZ1 participates in these signaling pathways. HaCaTFIZ1 cells do show more rapid

Excess FIZ1 Does Not Impede Contact-Induced Differentiation

Similar to parental cells, proliferative markers disappear and early differentiation markers appear in HaCaTFIZ1 cells shortly after reaching confluence (Figure 2A). Occludin, which accumulates at tight junctions, is observed at the periphery of both HaCaT and HaCaTFIZ1 cells in late differentiation (Figure 2B), confirming that excess FIZ1 does not inhibit the formation of these structures. FIZ1-mediated signaling can thus promote E

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Table 1. Top Signaling-Related Hits in the Secretome Experimenta Uniprot ID

gene name

protein name

increased secretion from HaCaTFIZ1 cells relative to parental cells: P17936 IGFBP3* insulin-like growth factor-binding protein 3 O94907 DKK1 Dickkopf-related protein 1 P23352 KAL1* anosmin-1 Q76M96 CCDC80* coiled-coil domain-containing protein 80 G5E971 MMP13* collagenase 3 decreased secretion from HaCaTFIZ1 cells relative to parental cells: P02751 FN1* fibronectin E9PLM6 MDK* midkine B4DIF2 SERPINE2 glia-derived nexin a

no. peptides

ratio H/L

normalized intensity

16 3 2 35 18

18.22 7.51 2.94 2.79 2.40

649.30 17.09 3.27 182.92 265.33

33 5 3

0.06 0.10 0.13

29.14 88.64 4.67

Asterisks indicate known heparin-binding proteins.

Figure 4. IGFBP3 mediates FIZ1-induced mitogenic signaling. (A) Western blot confirming increased levels of IGFBP3 in HaCaTFIZ1 cell extracts and captured on heparin-agarose beads from HaCaTFIZ1 cell-conditioned media. (B) Quantitative RT-PCR results for IGFBP3 and two other selected secretome hits and for the IGF signaling factors IGF1, IGF2, and IGF1R (IGF receptor). FIZ1 is shown for comparison. (C) Two-day growth assays in serum-free media comparing the relative response of HaCaTFIZ1 versus parental HaCaT cells to exogenous IGF1 (10 ng/mL). (D) Two-day growth assays in serum-free media testing the effects of IGFBP3 and IGF1 neutralizing antibodies (neut. ab). Data plotted as mean ± SE for three biological replicates. Asterisks indicate statistically significant differences for a threshold p < 0.01.

certain factors shown to be upregulated in primary cutaneous SCC cells (IGFBP3, MMP13, MATN218), and so it would be interesting to assess FIZ1 levels and function in those cells in future. The online integrated cancer database canSAR19 reports upregulated levels of FIZ1 in five cancers of epithelial origin and two leukemias, and it is thus worth considering that FIZ1 may possess oncogenic potential linked to its role as a signal transducer, transcription factor, or both.

proliferation without impeding normal contact-induced differentiation. HaCaTFIZ1 cells do, however, show resistance to the differentiation program induced in HaCaT cells by a switch from low to high Ca2+.16 While parental cells exited the cell cycle and showed upregulation of transcript and protein levels for early differentiation markers following the Ca2+ switch, HaCaTFIZ1 cells continued to proliferate, albeit with a lag that delayed reaching confluency (Figure 2C−E). A hallmark of this pathway is the rapid and transient tyrosine phosphorylation of the ras-GTPase activating protein (ras-GAP)-associated protein p62/SAM68.17 This phospho-band was observed in HaCaT but not HaCaTFIZ1 cells, 5 min after the Ca2+ switch (Figure 2F,G), indicating that FIZ1 overexpression abrogates activation of the relevant signaling pathway. Interestingly, a similar resistance to Ca2+- but not contact-induced differentiation was observed for the murine squamous cell carcinoma (SCC) cell line PAM212.17 HaCaTFIZ1 cells also have increased levels of

Increased HaCaTFIZ1 Proliferation Involves Autocrine Signaling

The growth advantage of HaCaTFIZ1 over parental HaCaT cells is further increased when serum levels in the growth media are reduced, with a persistence of proliferation observed even in serum-free media lacking any exogenous growth factors (Figure 3A,B). This suggests that the mitogenic signaling is mediated, at least in part, by secreted factors, which was confirmed by the demonstration that parental HaCaT cells show increased F

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 5. FIZ1-induced proliferation requires ERK1/2 activation. (A) Western blot analysis of Phos-ERK1/2 levels (normalized to total ERK1/2 levels) in HaCaT and HaCaTFIZ1 cell lysates during proliferation (day 2 of growth curve, 0% or 10% serum) or early differentiation (day 10 of growth curve, 10% serum). Tubulin is shown as a loading control. (B) Summary of the Western blot results in panel A, showing relative levels of phospho/activated ERK1/2 in HaCaTFIZ1 versus parental HaCaT cells. (C) Western blot analysis of cell lysates collected at day 2 of serum-free or 10% serum growth curves, showing that treatment with 1 μM PD98059 reduces the level of phosphoERK1/2 in HaCaTFIZ1 cells to that observed in parental HaCaT cells. (D) Two-day growth assays in serum-free (solid bars) or 10% serum (hashed bars) media in the presence or absence of 1 μM PD98059 (or DMSO carrier alone). Data plotted as mean ± SE for three biological replicates. Asterisks indicate statistically significant differences for a threshold p < 0.01.

proliferation when grown in HaCaTFIZ1 cell-conditioned media (Figure 3C). Pretreatment with heparan sulfate or depletion of heparin-binding factors removed the stimulatory effect of the conditioned media (Figure 3C), while direct addition of heparan sulfate to HaCaTFIZ1 cells removed their growth advantage (Figure 3D), suggesting the involvement of one or more heparin-binding growth factors. To identify this factor(s), we utilized the quantitative mass spectrometry technique SILAC (stable isotope labeling by amino acids in culture20). Parental HaCaT cells were labeled by growth in “light” media containing environmental isotopes of arginine and lysine (R0K0), while HaCaTFIZ1 cells were labeled by growth in “heavy” media containing 13C-arginine and D4-lysine (R6K4; Figure 3E). The isotopes shift the mass of Arg- and Lyscontaining peptides, with relative levels of heavy and light peptides reflecting the amount of protein under each condition. Cells were passaged in 10% serum SILAC media for 7 days and then transferred to serum-free SILAC media for 3 days. The conditioned media was collected and incubated with heparinagarose, the beads washed and combined, and bound proteins were eluted for gel separation and MS analysis. We identified/ quantified 353 proteins, the majority of which were detected in equal amounts (Figure 3F). Known heparin-binding growth factors such as FGF7/KGF, HB-EGF, and AREG were not present at detectable levels, although we did observe upregulation of HB-EGF transcript levels in HaCaTFIZ1 cells (data not shown). Of the subset of heparin-binding factors

enriched in HaCaTFIZ1-conditioned media, the top hit was IGFBP3, an insulin growth factor binding protein (Table 1). Both Western blot (Figure 4A) and qPCR (Figure 4B) analysis confirmed higher levels of IGFBP3 expression/secretion. Although we do not yet know the mechanism by which elevated FIZ1 results in increased IGFBP3 secretion, the dual roles that FIZ1 plays in signal transduction and transcriptional regulation suggest that it could regulate transcription of IGFBP3 either indirectly (e.g., by influencing an upstream signaling pathway such as JAK/STAT) or directly (e.g., via repression of a transcription factor). Increased FIZ1 Expression Sensitizes Cells to IGF Signaling

The role of IGFBP3 is controversial, given that its high affinity binding to IGFs has been shown in various cell lines to either prevent the growth factors’ pro-proliferative signaling or to prolong the lifespan of active IGF molecules without interfering with receptor binding. (For review, see ref 21.) IGFBP3 is also known to have IGF-independent roles.22−24 We first confirmed the essential role of IGBP3 in FIZ1-mediated mitogenic signaling by demonstrating that addition of IGFBP3 neutralizing antibodies completely abolishes the growth advantage of HaCaTFIZ1 cells in serum-free media (Figure 4D) and significantly reduces it in the more complicated background of 10% serum media (Supporting Information). The more robust response of HaCaTFIZ1 cells to exogenous IGF1 (Figure 4C) is consistent with sensitization to this growth factor, as is the fact that HaCaTFIZ1 cells are more resistant than G

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 6. Quantitative interactome mapping identifies numerous novel FIZ1 interactors. (A) Design of AP/MS experiment. (B) Plot of log ratio H/ L versus relative abundance (summed peptide intensities normalized by molecular weight) for all identified/quantified proteins. The bait protein FIZ1 (red circle) and interactors that validated by co-IP/Western blot analysis (green diamonds) are highlighted, along with common FLAG matrix contaminants that were manually filtered out of the data set (purple diamonds).

hits in the secretome, are contributing players in the signaling pathway(s) activated by elevated FIZ1 activity. Lastly, we also detected factors in the secretome that were secreted to a greater extent by parental cells (Figure 3F, Table 1). Although the significance is not yet clear, it is interesting to note that the glycoprotein fibronectin, growth factor midkine, and serine peptidase inhibitor SERPINE2 have all been linked to promotion of cell migration, a property that our wound healing assays specifically suggest is not associated with elevated FIZ1 (Supporting Information). The full secretome data set is provided as Supporting Information.

parental cells to the growth inhibitory effect of IGF1 neutralizing antibodies in serum-free media (Figure 4D). Levels of IGF1/IGF2 transcripts in HaCaTFIZ1 cells are similar to those in parental cells; however, there is an upregulation of transcript levels for the IGF1R receptor, through which both can signal (Figure 4B). Thus, although we cannot rule out an IGF-independent role for IGFBP3 in FIZ1-mediated signaling, our data suggest that increased IGFBP3, perhaps in combination with increased IGF1R levels, mediates hypersensitivity of HaCaTFIZ1 cells to IGF signaling. IGF1R expression has already been linked to regulation of the proliferation/differentiation axis in keratinocytes, as it is necessary for proliferation but has also been shown to actively suppress differentiation.25 The interplay between these signaling factors, together with our knockdown results suggesting that FIZ1 may function in part to prevent premature differentiation, provides novel insight into the complex balance of epidermal homeostasis.

MAP/ERK Kinase Activation Is Involved in FIZ1-Induced Proliferation

The major pathways through which IGFs signal are the phosphotidylinositol-4,5-bisphosphate 3-kinase (PI3K)/Akt kinase and mitogen-activated protein (MAP) kinase/extracellular signal-regulated kinase (ERK) 1/2 signaling pathways. Using phosphospecific antibodies, we detected higher levels of phosphorylated/active ERK1/2 (also known as p44/p42 MAPK) in proliferating HaCaTFIZ1 versus parental cells that correlate with their growth advantage in both 10% serum and serum-free media (Figure 5A,B) and suggest increased signaling through this pathway. PD98059 is a potent and selective inhibitor of the upstream MAPK/ERK kinase (MEK) that phosphorylates/activates ERK1/2. We noted that at concentrations typically used in cell culture (5 and 10 μM), treatment with PD98059 inhibited proliferation of both cell lines (data not shown), whereas at a lower concentration (1 μM), it had a selective effect on HaCaTFIZ1 cells, reducing levels of active ERK1/2 to those in parental cells (Figure 5C) and abolishing their growth advantage in both serum-free and 10% serum media (Figure 5D). Thus, while multiple pathways likely contribute to basal HaCaT proliferation, inhibition of MAP/ERK signaling removes the selective growth advantage conferred by excess FIZ1. Given that IGFs can signal through this pathway, we propose that the predominant pro-proliferative mechanism activated by excess FIZ1 in HaCaT cells is sensitization to IGF with increased downstream signaling via ERK1/2.

Other Factors Implicated in FIZ1-Mediated Signaling

While IGFBP3 was by far the most highly enriched protein in the HaCaTFIZ1 cell secretome, other heparin-binding signaling factors were also detected at elevated levels (Table 1). They include CCDC80/Equarin and DKK1, which, like IGFBP3, have been reported to have conflicting and system-dependent effects on proliferation.26−28 We confirmed upregulation of CCDC80 mRNA levels in HaCaTFIZ1 cells (Figure 4B), and the role of this protein as an FGF signaling modulator in the regulation of lens cell differentiation29,30 makes it an attractive candidate for future studies. DKK1 is a Wnt antagonist, and Wnt signaling has been shown to inhibit growth and promote terminal differentiation in keratinocytes.31 Consistent with this, DKK1 has been shown to promote proliferation of keratinocytes and regulate skin thickness.32 Although we did not see upregulation of DKK1 at the mRNA level (Figure 4B), the addition of DKK1 neutralizing antibodies to HaCaTFIZ1 cells had a small but significant negative effect on their growth advantage in serum-free media (data not shown). Thus, the potential remains that CCDC80 and DKK1, and possibly other H

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Table 2. Subset of Top Hits in the FIZ1 Interactomea pathway MOB/NDR complexes

nuclear import transcription JAK/STAT signaling receptor recycling a

Uniprot ID

gene name

protein name

no. peptides

ratio H/L

Q96SL8 Q70IA6 Q9Y2H1 Q15208 O15397 Q9BUZ4 O00268 P23458 P40763 Q6P3W7

FIZ1 MOB2 STK38L STK38 IPO8 TRAF4 TAF4 JAK1 STAT3 SCYL2

Flt3-interacting zinc finger protein 1 MOB kinase activator 2 NDR2; serine/threonine-protein kinase 38-like NDR1; serine/threonine-protein kinase 38 importin-8 TNF receptor-associated factor 4 transcription initiation factor TFIID subunit 4 Tyrosine-protein kinase JAK1 signal transducer and activator of transcription 3 SCY1-like protein 2

5 3 13 21 26 2 11 4 2 34

11.17 4.41 4.14 2.47 5.75 4.18 2.47 2.47 3.15 2.69

Of the nine proteins chosen for follow-up validation, all but STAT3 were demonstrated by Western blot analysis to reproducibly co-IP with FIZ1.

FIZ1 Interacts with Signaling and Transcription Factors

The cytoplasmic-specific interactor SCYL2/CVAK104 kinase is involved in receptor internalization and regulates the Wnt/βcatenin pathway,36 while the JAK1 tyrosine kinase is involved in numerous signaling pathways, coupling binding of cytokine ligands (including IGF1) to phosphorylation of signaling proteins and activation of STAT transcription factors. Although we could not reliably validate a FIZ1/STAT3 interaction by coIP/Western blot, which may be due to it being indirect/ transient. JAK1 signaling via STAT1 and STAT3 has been shown to promote myoblast proliferation while preventing premature differentiation.37 With regard to nuclear interactors, TAF4 is a subunit of the core TFIID RNA Polymerase II complex that mediates promoter responses to various activators/repressors. It has been linked to autocrine signaling pathways and negative regulation of proliferation38 and inactivation of Taf4 in mouse embryonic fibroblasts-induced serum-independent autocrine growth accompanied by deregulation of >1000 genes.39 Furthermore, conditional Taf4 inactivation in mouse basal keratinocytes promotes epidermal hyperplasia, affects the development of the epidermal permeability barrier, and provokes enhanced formation of chemically induced tumors.40 It is likely that more than one signaling pathway feeds into TAF4 regulation in keratinocytes, and we suggest that its interaction with FIZ1, a known transcriptional repressor that we have shown here to promote keratinocyte proliferation, raises the possibility that FIZ1 is a negative regulator of TAF4 in this system. This will be assessed directly in future studies by determining the effect of altered FIZ1 levels on expression of both known TAF4 target genes and on genes that we have shown here to be positively regulated by FIZ1, including IGFBP3, CCDC80, and IGF1R. We will also carry out a more comprehensive analysis of the global impact of altered FIZ1 levels on transcription and ChIP-seq experiments to identify docking sites on chromatin. Other nuclear FIZ1 interactors include the kinases NDR1/2 and their associated regulatory protein MOB2. Interestingly, these kinases have been linked to regulation of G1/S progression,41 and NDR1 opposes the cell cycle inhibition mediated by TGFβ signaling in HaCaT and other cell lines.42 Their association with FIZ1 may therefore be related to the increased G1/S progression in HaCaTFIZ1 cells, and this warrants further investigation. Given that the role of FIZ1 may change following the switch from proliferation to differentiation, even though FIZ1 levels remain the same (Figure 2A), we also tested whether the interactors remained associated with FIZ1 under these conditions. As shown in Figure 7E, all but IPO8 remained associated with FIZ1 in differentiating cells. While IPO8 may

To identify protein−protein interactions that may mediate the role of FIZ1, we carried out a quantitative affinity purification/ mass spectrometry (AP/MS) experiment (Figure 6A), directly comparing a pulldown of FLAG-tagged FIZ1 from HaCaTFIZ1 whole cell extracts to a control pulldown (incubation of parental HaCaT whole cell extracts with the same anti-FLAG affinity matrix). We identified/quantified 113 proteins, of which a subset were enriched specifically with FIZ1 (Figure 6B). Eight of the top hits (Table 2; Supporting Information) were further validated by co-IP/Western blot analysis with tagged (Figure 7D,E) and endogenous (Supporting Information) FIZ1 and by reciprocal co-IP (Figure 7F). The full data set is provided as Supporting Information. The eight novel FIZ1 interactors identified here range from signaling molecules and import factors to proteins involved in receptor recycling and gene transcription. Interestingly, half are protein kinases: the NDR1 and 2 kinases (and their associated regulatory protein MOB2), JAK1, and SCYL2/CVAK104, highlighting the potential for FIZ1 to feed into phosphorylation-mediated signaling cascades. Given that FIZ1 has been shown to function both as a signal transducer and a transcription factor, it was not surprising to identify both cytoplasmic and nuclear interactors. Similar to its localization in other cell lines,4,7 we show here that FIZ1 is predominantly nuclear in HaCaT cells, with a small pool detected in the cytoplasm (Figure 7B,C). We did not detect FIZ1 in membrane fractions, suggesting that interactions with membrane-associated proteins are likely to be transient. Assessment of compartment-specific interactions with FIZ1 (Figure 7D) revealed that some are exclusively cytoplasmic (the import factor IPO8, the signaling protein JAK1 and the receptor recycling factor SCYL2) and some exclusively nuclear (the NDR1 and 2 kinases and their regulatory protein MOB2, the transcription factor TAF4), whereas the TNF receptorassociated scaffold protein TRAF4 is found in complex with FIZ1 in both compartments. TRAF4 is a signaling scaffold that links receptors to kinases and their regulators in signaling cascades.33 Like FIZ1, TRAF4 is found in both the cytoplasm and in the nucleus (Figure 7D34). Interestingly, TRAF4 has been shown to associate with tight junctions in a dynamic fashion and to potentiate ERK1/2 phosphorylation/activation in proliferating but not quiescent cells.35 Given the novel association that we report here between TRAF4 and FIZ1, it is tempting to speculate that these proteins may work together to transduce extracellular signals from tight junctions that affect downstream regulation of the balance of keratinocyte proliferation/differentiation. I

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 7. FIZ1 is found in protein complexes in both the cytoplasm and the nucleus. (A) In addition to its 11 C2H2-type Zn finger domains, FIZ1 contains a monopartite nuclear localization signal (NLS) and 2 predicted nuclear export signals (NESs). (B) Western blot analysis of HaCaT cell fractions confirms the distribution of FIZ1 between the nucleus and cytoplasm in a 70:30 ratio. FIZ1 was not present at detectable levels in plasma membrane fractions. Lamin A/C, alpha-tubulin, and E-cadherin are markers for nuclear, cytoplasmic, and plasma membrane fractions, respectively. (C) Nuclear retention assay demonstrating the colocalization of endogenous FIZ1 (green) with chromatin (Hoechst 33342; blue) in proliferating HaCaT cells both before and after mild permeabilization with 0.1% Triton X-100. When the chromatin was subsequently digested by nuclease treatment (125 U benzonase/coverslip; position of nuclei outlined), the nuclear FIZ1 signal was lost. Scale bars are 5 μM. (D) Co-IP/Western blot analysis of the association of endogenous interactors with FLAG-tagged FIZ1 in nuclear and cytoplasmic extracts from HaCaTFIZ1 cells. For the control IP, parental HaCaT cells were incubated with the anti-FLAG matrix. (E) Co-IP/Western blot analysis of the association of endogenous interactors with FLAG-tagged FIZ1 in HaCaTFIZ1 whole cell extracts collected under proliferating versus differentiating conditions. For the control IP, parental HaCaT cells were incubated with the anti-FLAG matrix. (F) Reciprocal validation of novel interactions by co-IP of endogenous FIZ1 with GFP-tagged interactors transiently overexpressed in HaCaT cells. For the control IPs, the transiently overexpressed GFP tag alone was pulled down.

suggests an additional level of regulation to mediate any changes in signaling, such as post-translational modification or changes in multiprotein complex composition. FIZ1 is a putative phosphoprotein, although to date only two phosphosites have been mapped (in a Cell Signaling Technology curation data set in PhostphoSitePlus based on enrichment of putative Akt substrates46), and the relevance is not yet known. We did not detect any FIZ1 phosphopeptides in pulldowns from proliferating versus differentiating cells, nor did we observe any difference in migration of FIZ1 when lysates were separated on Phos-tag gels (which enhance the slower migration of phosphoproteins), although we did observe a

mediate nuclear import of FIZ1, which contains a predicted monopartite NLS (Figure 7A), FIZ1 is predominantly nuclear even in differentiating cells (Supporting Information). A stable and proliferation-specific complex with IPO8 suggests an alternate or parallel function. IPO8 mediates nuclear import of Smad4,43 which is involved in TGFβ-induced cell cycle arrest in HaCaT cells.44 It is also a gene-silencing factor that targets Argonaute (Ago) proteins to specific mRNAs.45 Interestingly, although FIZ1 did not enrich Ago proteins, we did find overlap with Ago3/4 interactors (IPO8, NDR1, JAK1, spindlin). For the remaining seven interactors identified here, persistence of interaction with FIZ1 in differentiating cells J

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

and fluorescence assays demonstrating the lack of nuclear retention of FIZ1 in response to mild permeabilization in differentiating cells. Full datasets are also provided for the secretome and interactome experiments. This material is available free of charge via the Internet at http://pubs.acs.org.

band shift in nocodazole-arrested cell extracts that suggests cellcycle-dependent phosphorylation of FIZ1 (data not shown). One key difference that we did note is the loss of association of FIZ1 with chromatin following the switch from proliferation to differentiation. Using a nuclear retention assay, we demonstrated that in proliferating cells the nuclear pool of FIZ1 shows that a nucleoplasmic localization similar to that of chromatin is retained following mild permeabilization of HaCaT cells and lost upon treatment with the endonuclease benzonase, which degrades DNA and RNA (Figure 7C). Although we cannot rule out RNA-mediated associations, interaction with chromatin is more likely to mediate the nuclear retention of FIZ1, given that it is a demonstrated transcription factor already shown to associate with chromatin.7 In differentiating cells, the steady-state localization of FIZ1 remains predominantly nuclear; however, the protein freely diffuses out of the nucleus following mild permeabilization (Supporting Information), indicating loss of the underlying interaction(s) with nuclear structures that mediate nuclear retention under proliferating conditions. In summary, we show here that the transcription factor/ signal transducer FIZ1 is expressed in human keratinocytes as a predominantly nuclear protein and that excess FIZ1 drives increased proliferation via autocrine signaling pathways. HaCaTFIZ1 cells remain responsive to exogenous growth factors and are subject to contact-induced terminal differentiation, and we have shown that the autocrine mitogenic signaling induced by excess FIZ1 involves increased secretion of the heparinbinding factor IGFBP3 and sensitization to IGF signaling and is mediated via activation of MAP/ERK kinases. Taken together, our results provide valuable functional clues to the role(s) of FIZ1 in epidermal homeostasis and open up new avenues of research into its therapeutic potential, both in promoting wound healing and in treating disorders in which epidermal homeostasis is dysregulated. An important next step will be to extend these analyses to keratinocyte homeostasis in vivo by assessing both the distribution of FIZ1 in the stratified epidermis and the downstream effects of altering its expression levels. We have also identified underlying molecular mechanisms by which FIZ1 can act as both a signaling and a transcription factor, which is consistent with previously published roles in regulating homeostasis in retinal and hematopoietic cells. The latter study also demonstrated increased proliferation of mouse B-cell lymphoblastic lymphomas in response to co-overexpression of Stat5a and Fiz1 (via dual MLV integration).6 It will thus be interesting to determine whether the proproliferative function of this ubiquitously expressed protein is conserved in other cell types.

■

■

AUTHOR INFORMATION

Corresponding Author

*Tel: 613-562-5800 x8068. Fax: 613-562-5434. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Delphine Chamousset for assistance with FACS analysis and colleagues in the Trinkle lab for helpful discussions and suggestions. We also thank Drs. Kursad Turksen, Brian Hemmings, Harald Wajant, and Nadine Wiper-Bergeron for advice and reagents and Lawrence Puente at the Ottawa Hospital Research Institute Proteomics Core Facility for technical support. This work was supported by the Terry Fox Research Institute (ref: 20148) and NSERC (ref: 372370). N.L. holds a CIHR Banting & Best Scholarship and L.T.-M. holds a CIHR New Investigator Award.

■

REFERENCES

(1) Ratushny, V.; Gober, M. D.; Hick, R.; Ridky, T. W.; Seykora, J. T. From keratinocyte to cancer: the pathogenesis and modeling of cutaneous squamous cell carcinoma. J. Clin. Invest. 2012, 122, 464− 472. (2) Fuchs, E. Scratching the surface of skin development. Nature 2007, 445, 834−842. (3) Furuse, M. Molecular Basis of the Core Structure of Tight Junctions. http://cshperspectives.cshlp.org/content/2/1/a002907. long (accessed Jan 10, 2014). (4) Wolf, I. Fiz1, a Novel Zinc Finger Protein Interacting with the Receptor Tyrosine Kinase Flt3. J. Biol. Chem. 1999, 274, 21478− 21484. (5) Whartenby, K. A.; Small, D.; Calabresi, P. A. FLT3 inhibitors for the treatment of autoimmune disease. Expert Opin. Invest. Drugs 2008, 17, 1685−1692. (6) Tsuruyama, T.; Imai, Y.; Takeuchi, H.; Hiratsuka, T.; Maruyama, Y.; Kanaya, K.; Kaszynski, R.; Jin, G.; Okuno, T.; Ozeki, M.; et al. Dual retrovirus integration tagging: identification of new signaling molecules Fiz1 and Hipk2 that are involved in the IL-7 signaling pathway in B lymphoblastic lymphomas. J. Leukocyte Biol. 2010, 88, 107−116. (7) Mali, R. S.; Peng, G.-H.; Zhang, X.; Dang, L.; Chen, S.; Mitton, K. P. FIZ1 is part of the regulatory protein complex on active photoreceptor-specific gene promoters in vivo. BMC Mol. Biol. 2008, 9, 87. (8) Mali, R. S.; Zhang, X.; Hoerauf, W.; Doyle, D.; Devitt, J.; Loffreda-Wren, J.; Mitton, K. P. FIZ1 is expressed during photoreceptor maturation, and synergizes with NRL and CRX at rod-specific promoters in vitro. Exp. Eye Res. 2007, 84, 349−360. (9) Mitton, K. P. Interaction of retinal bZIP transcription factor NRL with Flt3-interacting zinc-finger protein Fiz1: possible role of Fiz1 as a transcriptional repressor. Hum. Mol. Genet. 2003, 12, 365−373. (10) Boukamp, P.; Petrussevska, R. T.; Breitkreutz, D.; Hornung, J.; Markham, A.; Fusenig, N. E. Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line. J. Cell Biol. 1988, 106, 761−771. (11) Prevost, M.; Chamousset, D.; Nasa, I.; Freele, E.; Morrice, N.; Moorhead, G.; Trinkle-Mulcahy, L. Quantitative fragmentome mapping reveals novel, domain-specific partners for the modular protein RepoMan. Mol. Cell. Proteomics 2013, 12, 1468−1486.

ASSOCIATED CONTENT

S Supporting Information *

Details for all plasmids and antibodies and sequences for all RT-PCR and qPCR primers and siRNA duplexes are provided. Supplemental data include the following: Western blot analysis of additional clonal cell lines, response of cell lines to exogenous FGF7/KGF, representative Western blot demonstrating relative cyclin E levels, scratch wound assays, Western blot and RT-PCR analysis of the onset of differentiation marker expression in the Ca2+ switch experiments, effect of IGFBP3 neutralizing antibodies on growth in 10% serum, validation of co-purification of novel binding partners with endogenous FIZ1 K

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

(12) Chamousset, D.; De Wever, V.; Moorhead, G. B.; Chen, Y.; Boisvert, F.-M.; Lamond, A. I.; Trinkle-Mulcahy, L. RRP1B targets PP1 to mammalian cell nucleoli and is associated with Pre-60S ribosomal subunits. Mol. Biol. Cell 2010, 21, 4212−4226. (13) Schmittgen, T. D.; Livak, K. J. Analyzing real-time PCR data by the comparative CT method. Nat. Protoc. 2008, 3, 1101−1108. (14) Cox, J.; Matic, I.; Hilger, M.; Nagaraj, N.; Selbach, M.; Olsen, J. V.; Mann, M. A practical guide to the MaxQuant computational platform for SILAC-based quantitative proteomics. Nat. Protoc. 2009, 4, 698−705. (15) Siu, K. T.; Rosner, M. R.; Minella, A. C. An integrated view of cyclin E function and regulation. Cell Cycle 2012, 11, 57−64. (16) Deyrieux, A. F.; Wilson, V. G. In vitro culture conditions to study keratinocyte differentiation using the HaCaT cell line. Cytotechnology 2007, 54, 77−83. (17) Filvaroff, E.; Stern, D. F.; Dotto, G. P. Tyrosine phosphorylation is an early and specific event involved in primary keratinocyte differentiation. Mol. Cell Biol. 1990, 10, 1164−1173. (18) Toriseva, M.; Ala-aho, R.; Peltonen, S.; Peltonen, J.; Grénman, R.; Kähäri, V.-M. Keratinocyte Growth Factor Induces Gene Expression Signature Associated with Suppression of Malignant Phenotype of Cutaneous Squamous Carcinoma Cells. PLoS One 2012, 7, e33041. (19) Halling-Brown, M. D.; Bulusu, K. C.; Patel, M.; Tym, J. E.; AlLazikani, B. canSAR: an integrated cancer public translational research and drug discovery resource. Nucleic Acids Res. 2012, 40, D947−D956. (20) Ong, S.-E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell. Proteomics 2002, 1, 376−386. (21) Yamada, P. M.; Lee, K. W. Perspectives in mammalian IGFBP-3 biology: local vs. systemic action. Am. J. Physiol.: Cell Physiol. 2009, 296, C954−C976. (22) Ingermann, A. R.; Yang, Y.-F.; Han, J.; Mikami, A.; Garza, A. E.; Mohanraj, L.; Fan, L.; Idowu, M.; Ware, J. L.; Kim, H.-S.; et al. Identification of a novel cell death receptor mediating IGFBP-3induced anti-tumor effects in breast and prostate cancer. J. Biol. Chem. 2010, 285, 30233−30246. (23) Oh, Y.; Muller, H. L.; Lamson, G.; Rosenfeld, R. G. Insulin-like growth factor (IGF)-independent action of IGF-binding protein-3 in Hs578T human breast cancer cells. Cell surface binding and growth inhibition. J. Biol. Chem. 1993, 268, 14964−14971. (24) Rajah, R.; Valentinis, B.; Cohen, P. Insulin-like growth factor (IGF)-binding protein-3 induces apoptosis and mediates the effects of transforming growth factor-beta1 on programmed cell death through a p53- and IGF-independent mechanism. J. Biol. Chem. 1997, 272, 12181−12188. (25) Sadagurski, M.; Yakar, S.; Weingarten, G.; Holzenberger, M.; Rhodes, C. J.; Breitkreutz, D.; Leroith, D.; Wertheimer, E. Insulin-like growth factor 1 receptor signaling regulates skin development and inhibits skin keratinocyte differentiation. Mol. Cell. Biol. 2006, 26, 2675−2687. (26) Menezes, M. E.; Devine, D. J.; Shevde, L. A.; Samant, R. S. Dickkopf1: a tumor suppressor or metastasis promoter? Int. J. Cancer 2012, 130, 1477−1483. (27) Tremblay, F.; Revett, T.; Huard, C.; Zhang, Y.; Tobin, J. F.; Martinez, R. V.; Gimeno, R. E. Bidirectional modulation of adipogenesis by the secreted protein Ccdc80/DRO1/URB. J. Biol. Chem. 2009, 284, 8136−8147. (28) Yamaguchi, Y.; Passeron, T.; Watabe, H.; Yasumoto, K.-I.; Rouzaud, F.; Hoashi, T.; Hearing, V. J. The effects of dickkopf 1 on gene expression and Wnt signaling by melanocytes: mechanisms underlying its suppression of melanocyte function and proliferation. J. Invest. Dermatol. 2007, 127, 1217−1225. (29) Song, X.; Sato, Y.; Felemban, A.; Ito, A.; Hossain, M.; Ochiai, H.; Yamamoto, T.; Sekiguchi, K.; Tanaka, H.; Ohta, K. Equarin is involved as an FGF signaling modulator in chick lens differentiation. Dev. Biol. 2012, 368, 109−117.

(30) Jarrin, M.; Pandit, T.; Gunhaga, L. A balance of FGF and BMP signals regulates cell cycle exit and Equarin expression in lens cells. Mol. Biol. Cell 2012, 23, 3266−3274. (31) Slavik, M. A.; Allen-Hoffmann, B. L.; Liu, B. Y.; Alexander, C. M. Wnt signaling induces differentiation of progenitor cells in organotypic keratinocyte cultures. BMC Dev. Biol. 2007, 7, 9. (32) Yamaguchi, Y.; Passeron, T.; Hoashi, T.; Watabe, H.; Rouzaud, F.; Yasumoto, K.-I.; Hara, T.; Tohyama, C.; Katayama, I.; Miki, T.; et al. Dickkopf 1 (DKK1) regulates skin pigmentation and thickness by affecting Wnt/beta-catenin signaling in keratinocytes. FASEB J. 2008, 22, 1009−1020. (33) Wajant, H.; Henkler, F.; Scheurich, P. The TNF-receptorassociated factor family: scaffold molecules for cytokine receptors, kinases and their regulators. Cell. Signalling 2001, 13, 389−400. (34) Rousseau, A.; McEwen, A. G.; Poussin-Courmontagne, P.; Rognan, D.; Nomine, Y.; Rio, M.-C.; Tomasetto, C.; Alpy, F. TRAF4 is a novel phosphoinositide-binding protein modulating tight junctions and favoring cell migration. PLoS Biol. 2013, 11, e1001726. (35) Kedinger, V.; Alpy, F.; Baguet, A.; Polette, M.; Stoll, I.; Chenard, M.-P.; Tomasetto, C.; Rio, M.-C. Tumor necrosis factor receptorassociated factor 4 is a dynamic tight junction-related shuttle protein involved in epithelium homeostasis. PLoS One 2008, 3, e3518. (36) Terabayashi, T.; Funato, Y.; Fukuda, M.; Miki, H. A coated vesicle-associated kinase of 104 kDa (CVAK104) induces lysosomal degradation of frizzled 5 (Fzd5). J. Biol. Chem. 2009, 284, 26716− 26724. (37) Sun, L.; Ma, K.; Wang, H.; Xiao, F.; Gao, Y.; Zhang, W.; Wang, K.; Gao, X.; Ip, N.; Wu, Z. JAK1-STAT1-STAT3, a key pathway promoting proliferation and preventing premature differentiation of myoblasts. J. Cell Biol. 2007, 179, 129−138. (38) Davidson, I.; Kobi, D.; Fadloun, A.; Mengus, G. New insights into TAFs as regulators of cell cycle and signaling pathways. Cell Cycle 2005, 4, 1486−1490. (39) Mengus, G.; Fadloun, A.; Kobi, D.; Thibault, C.; Perletti, L.; Michel, I.; Davidson, I. TAF4 inactivation in embryonic fibroblasts activates TGF beta signalling and autocrine growth. EMBO J. 2005, 24, 2753−2767. (40) Fadloun, A.; Kobi, D.; Pointud, J.-C.; Indra, A. K.; Teletin, M.; Bole-Feysot, C.; Testoni, B.; Mantovani, R.; Metzger, D.; Mengus, G.; et al. The TFIID subunit TAF4 regulates keratinocyte proliferation and has cell-autonomous and non-cell-autonomous tumour suppressor activity in mouse epidermis. Development 2007, 134, 2947−2958. (41) Cornils, H.; Kohler, R. S.; Hergovich, A.; Hemmings, B. A. Human NDR kinases control G(1)/S cell cycle transition by directly regulating p21 stability. Mol. Cell. Biol. 2011, 31, 1382−1395. (42) Pot, I.; Patel, S.; Deng, L.; Chandhoke, A. S.; Zhang, C.; Bonni, A.; Bonni, S. Identification of a Novel Link between the Protein Kinase NDR1 and TGFbeta Signaling in Epithelial Cells. PLoS One 2013, 8, e67178. (43) Yao, X.; Chen, X.; Cottonham, C.; Xu, L. Preferential utilization of Imp7/8 in nuclear import of Smads. J. Biol. Chem. 2008, 283, 22867−22874. (44) Levy, L.; Hill, C. S. Smad4 dependency defines two classes of transforming growth factor {beta} (TGF-{beta}) target genes and distinguishes TGF-{beta}-induced epithelial-mesenchymal transition from its antiproliferative and migratory responses. Mol. Cell. Biol. 2005, 25, 8108−8125. (45) Weinmann, L.; Hock, J.; Ivacevic, T.; Ohrt, T.; Mutze, J.; Schwille, P.; Kremmer, E.; Benes, V.; Urlaub, H.; Meister, G. Importin 8 is a gene silencing factor that targets argonaute proteins to distinct mRNAs. Cell 2009, 136, 496−507. (46) Hornbeck, P. V.; Kornhauser, J. M.; Tkachev, S.; Zhang, B.; Skrzypek, E.; Murray, B.; Latham, V.; Sullivan, M. PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res. 2012, 40, D261−D270.

L

dx.doi.org/10.1021/pr500392m | J. Proteome Res. XXXX, XXX, XXX−XXX