Xyloglucan from Tropaeolum majus Seeds Induces Cellular

Article pubs.acs.org/Biomac

Xyloglucan from Tropaeolum majus Seeds Induces Cellular Differentiation of Human Keratinocytes by Inhibition of EGFR Phosphorylation and Decreased Activity of Transcription Factor CREB Dominika M. Zacharski,†,‡ Simone Brandt,†,‡ Stefan Esch,‡ Simone König,§ Michael Mormann,# Gudrun Ulrich-Merzenich,∥ and Andreas Hensel*,‡ ‡

University University Germany ∥ University # University §

of Münster, Institute of Pharmaceutical Biology and Phytochemistry, Corrensstrasse 48, D-48149 Münster, Germany of Münster, Interdisciplinary Centre for Clinical Research, Core Unit Proteomics, Röntgenstr. 21, D-48149 Münster, Clinic Centre Bonn, Medical Clinic III, Centre for Internal Medicine, Sigmund-Freud-Str. 25, D-53127 Bonn, Germany of Münster, Institute for Hygiene, Robert-Koch-Strasse 41, D-48149 Münster, Germany

S Supporting Information *

ABSTRACT: Xyloglucan XG (molecular weight 462 kDa, 1,4-/1,4,6-(pGlc) linked backbone, side chains of 1-pXyl, 1,2-pXyl, 1-p-Gal) was isolated from the seeds of Tropaeolum majus. XG (100 μg/mL) induced terminal cellular differentiation of human keratinocytes, as demonstrated by immunofluorescence staining and Western blot using cytokeratin 10 and involucrin as marker proteins. Differentiation was also induced by XG-derived oligosaccharides (degree of polymerization 7−9). Quantitative real-time polymerase chain reaction (qPCR) revealed the induction of gene expression of typical differentiation markers (cytokeratin, filaggrin, involucrin, loricrin, transglutaminase) in a time-dependent manner. Whole human genome microarray indicated that most of upregulated genes were related to differentiation processes. Microarray findings on selected genes were subsequently confirmed by qPCR. For identification of the molecular target of xyloglucan PAGE of keratinocyte membrane preparations was performed, followed by blotting with fluorescein isothiocyanate-labeled XG. XG interacting proteins were characterized by MS. Peptide fragments of epidermal growth factor receptor (EGFR) and integrin β4 were identified. Subsequent phospho-kinase array indicated that phosphorylation of EGFR and transcription factor cAMP response element-binding protein (CREB) was decreased in the XG-treated cells. Thus, the XG-induced differentiation of keratinocytes is proposed to be mediated by the inhibition of the phosphorylation of EGFR, leading to a dimished CREB activation, which is essential for the activation of cellular differentiation.

■

INTRODUCTION Wounding has a tremendous impact on healthcare economy. Especially chronic wounds represent a major health burden and drain on healthcare resources in most countries.1 Concerning the treatment of acute wounds significant differences and different paradigms consist between industrialized and developing countries: while acute wounding is cured in Europe and U.S. routinely by antiseptic and surgical care, most physicians in developing and emerging countries depend on local treatment of acute wounds using antiseptic, antiinflammatory, or cell-proliferation-inducing remedies. In contrast to that, treatment of chronic wounds is still a big problem all over the world. For example, leg ulcerations are suggested to occur at an incidence of about 1% of the population; in the U.S., only chronic wounds affect 3−6 million patients, with estimated annual costs of $5−10 billion each year. Therefore, the need for intensified development of new lead compounds for effective and evidence-based wound healing strategies is obvious. For © 2015 American Chemical Society

several years polysaccharide-containing wound dressings are increasingly used for chronic wounds with alginate or cellulose ethers as active ingredients.2 These create an occlusive and moist environment at the wound/dressing interface leading to the removal of exudate, reduced inflammation and microbial contamination and potentially increased migration of keratinocytes.3 Plant-derived polysaccharides have been described recently for wound treatment. Interestingly, these biopolymers did not only exert physical effects, but influenced specifically the physiology of the skin cells, for example, by inducing cellular proliferation,4 differentiation of keratinocytes,5,6 and stimulation of collagen and extracellular matrix formation from dermal fibroblasts.7 This would imply that such polymer-based Received: April 24, 2015 Revised: June 5, 2015 Published: June 12, 2015 2157

DOI: 10.1021/acs.biomac.5b00553 Biomacromolecules 2015, 16, 2157−2167

Article

Biomacromolecules

HaCaT keratinocytes (human adult low calcium high temperature) are spontaneously immortalized keratinocytes established and kindly provided by Prof. Dr. Fusenig from the German Cancer Research Center DKFZ (Heidelberg, Germany). NHEK were cultivated in NHEK growth medium (PCT epidermal keratinocyte medium, low BPE) at 37 °C (5% CO2), whereas HaCaT keratinocytes were cultured in HaCaT growth medium (D-MEM high glucose with glutamine, 10% FCS, 1% NEAA, 1% Pen/Strep) at 37 °C/8% CO2. Adherent cells were cultured until they reached a confluence of 70− 90%. Cell counting was performed using CASY cell counter. Assays that evaluate the impact of test compounds on NHEK or HaCaT were conducted in MCDB 153 complete medium as test medium. MCDB 153 complete medium is a defined medium supplemented with human recombinant EGF, human recombinant insulin, phosphoethalonamine, ethanolamine hydrocortisone, and Lglutamine. Use of this test medium results in a lower proliferation rate of the cells than in their respective growth medium. Isolation, Quantitation, and Reverse Transcription of Total RNA. NHEKs were seeded in six-well plates and cultured until 70− 80% confluence. After incubation with test solutions for specified time intervals, total RNA was isolated using NucleoSpin RNA II kit (Machery-Nagel) according to the manufacturer’s instructions. Medium was removed, and cells were briefly rinsed in PBS and directly lysed by adding 600 μL buffer RA1 (containing 1% freshly added β-mercaptoethanol) to the cell culture dish. Lysates were transferred to reaction tubes, immediately flash frozen in liquid N2, and stored at −80 °C until further processing. Samples were slowly thawed and passed five times through a 0.9 mm needle fitted to a syringe. Lysates were applied to NucleoSpin filters (violet ring) and centrifuged for 1 min at 11000 × g. Subsequently, lysates were mixed with 600 μL of 70% ethanol and loaded on NucleoSpin RNA/ Protein columns (blue ring) and centrifuged in two successive steps for 30 s at 11000 × g. After samples (RNA bond to the column membrane) were centrifuged with 350 μL of membrane desalting buffer (MDB) for 1 min at 11000 × g, 95 μL of rDNase reaction mixture was applied on the column and incubated for 15 min at room temperature. Following three washing steps (200 μL buffer RA2, 30 s, 11000 × g; 600 μL buffer RA3, 30 s, 11000 × g; 250 μL buffer RA3, 2 min, 11000 × g), RNA was eluted in 60 μL of RNase-free water by centrifugation for 1 min at 11000 × g and is ready to use for reverse transcription. RNA quality and concentration was examined at 260 and 280 nm, representing the maximum absorption wavelengths of nucleic acid and proteins/phenols respectively, by use of μCuvette G1.0 and the BioPhotometer plus (both Eppendorf). The A260/A280 ratio indicates purity of the RNA and should be in the range of 1.8−2.0.14 For measurement with the μCuvette G1.0 2 μL of RNA sample were directly applied to the measuring cell without any dilution. Depending on RNA concentration, 0.3−1 μg of total RNA were mixed with 1 μL of anchored-oligo(dT)18 primer and 2 μL of random hexamer primer. A variable amount of RNase-free water was added to make a total volume of 13 μL and the template-primer mixture was heated for 10 min at 65 °C in a thermal cycler. Samples were immediately cooled on ice and remaining components of the RT mix (transcriptor reverse transcriptase reaction buffer, 4 μL; protector RNase inhibitor, 0.5 μL; deoxynucleotide mix, 2 μL; transcriptor reverse transcriptase, 0.5 μL) were added. The RT reaction mixture was incubated at 25 °C for 10 min, followed by incubation at 50 °C for 60 min. Transcriptase was inactivated by heating to 85 °C for 5 min and reaction was further stopped by cooling down to 4 °C. cDNA was stored at −20 °C. As it is important to detect the presence of contaminating genomic DNA, so-called “no RT”-controls, have to be included when performing quantitative real-time polymerase chain reaction (qPCR). Therefore, each sample reaction was additionally carried out by replacing the reverse transcriptase with RNase-free water. Quantitative real time PCR (qPCR) was performed using TaqMan gene expression assays (Applied Biosystems), containing forward and reverse primers and probe, and the SensiMix II probe kit (Bioline) containing a hot-start DNA polymerase, dNTPs and stabilizers.

wound dressings could specifically interfere with the skin cell regulation and affect the complex wound healing process. In this context, a xyloglucan (XG) from Tamarind seeds promoting keratinocyte proliferation by interacting with the Erk signaling pathway was reported recently.8 This finding prompted us to investigate the interaction of a similar xyloglucan obtained from nasturtium seeds (Tropaeolum majus L.) with the skin cells, especially with respect to specific influence on keratinocyte physiology. Interestingly, we found this polysaccharide to be a strong inductor of keratinocyte terminal differentiation, which again is a crucial step during wound healing. The mode of action of this xyloglucan opens new perspectives, as this polysaccharide inhibits the activation of the epidermal growth factor receptor (EGFR). EGFR regulates the intracellular signaling toward the transcription factor cAMP response element-binding protein (CREB), which again triggers the cell to move from cellular proliferation into terminal differentiation. Furthermore, xyloglucans from plant seeds, serving physiologically as storage polymers for the embryo, are easily available from natural sources and can be produced very economically in good yields and high purity. From an economical point of view, xyloglucans could thus serve as innovative wound care products.

■

MATERIALS AND METHODS

General Experimentation Procedure. If not stated otherwise, all chemicals were purchased from VWR (Darmstadt, Germany). XG oligosaccharide mixture (degree of polymerization (DP) 7−9) was obtained from Megazyme (Bray, Ireland). Characterization by nanoESI-Q-TOF-MS mainly revealed the presence of heptasaccharides (e.g., m/z 1063.30 [4 Hex + 3 Pent + H]+), octasaccharides (e.g., m/z 1225.31 [5 Hex + 3 Pent + H]+) and nonasaccharides (e.g., m/z 1387.44 [6 Hex + 3 Pent + H]+); additionally, minor signals for tri- to hexasaccharides were detected. The following antibodies were used: β-Actin (AC-15), mouse monoclonal (1:4000) from Sigma-Aldrich, St. Louis, U.S.A., Alexa Fluor 488 rabbit anti-mouse IgG (1:300) from Life Technologies, Carlsbad, U.S.A., Alexa Fluor 594 goat anti-rabbit IgG (1:300) Life Technologies, Carlsbad, U.S.A., Cytokeratin 10, rabbit, both monoclonal and polyclonal (dilution: 1:15000 and 1:500, respectively) from Abcam, Cambridge, U.K., goat anti-rabbit IgG HRP (1:10000) from Jackson ImmunoResearch, West Grove, U.S.A., Involucrin (SY5), mouse monoclonal antibody (1:200) from Thermo Scientific, Waltham, U.S.A., and mouse TrueBlot ULTRA anti-mouse Ig HRP (1:1000) from Rockland Immunochemicals, Gilbertsville, U.S.A. Isolation of XG. T. majus L. seeds were obtained from the Medicinal Plant Garden of the University of Münster (Germany) and were identified by the comparison with reference herbal material. A voucher species is deposited in the archives of IPBP, University of Münster (designation IPBP391). XG from grounded nasturtium seeds was extracted9 in hot (100 °C) NaOH (2 M) containing sodium borohydride (0.05%, w/v) and was purified by precipitation with the 4-fold volume of glacial acetic acid and ethanol (1:10, v/v). XG was purified from the precipitate by Fehlings’s solution, as described by Edwards et al.10 Analytical characterization of the XG was performed according methods described in refs 6 and 11. Sequencing and identification of peptides was performed as described in ref 12. Labeling of XG with fluorescein isothiocyanate (FITC) was performed as described by ref 13. Cells and Cell Culture. Primary normal human epidermal keratinocytes (NHEK), isolated from juvenile human foreskin, were purchased from CELLnTEC Advanced Cell Systems AG, Bern, Switzerland (epidermal keratinocyte progenitors, pooled cells from three subjects). Several different batches have been used for this study. 2158


Article

Biomacromolecules TaqMan gene expression assays used for gene expression studies are listed in Table 1. cDNA was diluted to 5 ng/μL (RNA equivalent)

membrane or stained with Coomassie. Blots were incubated with FITC-labeled XG for 45 min at room temperature. Fluorescence intensity was detected with a Typhoon fluorescence imager (Amersham Biosciences). SDS-PAGE.15 Precast gradient polyacrylamide gels (4−15%, BioRad), Mini-PROTEAN Tetra Cell system (BioRad), 150 V for 45 min, protein marker Precision Plus Protein Dual Color (BioRad). Native PAGE. Precast polyacrylamide gels (10%, BioRad), MiniPROTEAN Tetra Cell system (BioRad), 110 V for 60 min, gel filtration standards (BioRad), containing bovine thyroglobulin, bovine γ-globulin, chicken ovalbumin, horse myoglobulin, and vitamin B12 were used as molecular weight markers. 2D-PAGE. 2D-PAGE with 500 μg of total protein was performed using BioRad equipment as recommended by the manufacturer and nonlinear IPG strips pH 3−10 and 12% gels.16 Coomassie Staining. Gels were stained with Coomassie staining solution containing Coomassie Brilliant Blue R250 for 30 min and were destained with destaining solution for approximately 30 min until the background became nearly transparent again. Finally, gels were washed with water and documentation was carried out with the Intas Gel iX Imager (Intas Science Imaging Instruments GmbH). Western Blotting. Blotting was carried out using a semidry transfer system (BioRad). Transfer of the proteins was obtained in 30 min at 10 V, followed by blocking for 1 h. Blots were incubated with diluted primary antibodies overnight at 4 °C, afterwards washed three times for 5 min in TBS-T buffer, and incubated with diluted HRPconjugated secondary antibodies for 45 min. After washing for 3 × 10 min in TBS-T buffer, visualization was performed by addition of freshly prepared enhanced chemiluminescence (ECL) reagent (GE Healthcare). Digital images of the Western blots were captured using the Intas ChemoCam Imager (Intas Science Imaging Instruments GmbH). Immunofluorescence Analysis by Confocal Laser Scanning Microscopy. Cells were grown in imaging chambers (PAA) to 70% of confluence, followed by incubation with the test solutions for 7 days. After rinsing in PBS, cells were fixated in 4% formaldehyde for 15 min at RT. Following permeabilization with 100% ice-cold methanol at −20 °C for 10 min, cells were blocked with Ultra-V-Block (Thermo Scientific) for 10 min at RT and were incubated with the primary antibodies overnight at 4 °C. Cells were washed 3 × 5 min at RT with PBS and were incubated with fluorochrome-conjugated secondary antibodies for 45 min at RT. Subsequently, cells were washed 3 × 5 min with PBS, incubated with DAPI (1 μg/mL) for 3−5 min at room temperature to stain the nuclei and were washed again 3 × 5 min with PBS. Finally, the imaging chamber coverslip bottom was fixed to a slide using Fluoromount Mounting Medium (Sigma-Aldrich) and sealed with nail polish. Samples were immediately examined using the confocal laser scanning microscope (CLSM) Leica TCS SP2, equipped with a 63 × objective (HCX PL Apo 63x/1.4 oil), and the Leica LCS Lite software. For excitation, lasers with a wavelength of 405 nm (25% laser power), 488 nm (25% laser power), and 594 nm (19% laser power) were used. Human Phospho-Kinase Array. ProteomProfiler (RnDSystems, Minneapolis, MN, U.S.A.). NHEK were grown to 70% confluence, followed by a 6 h incubation with the test solutions. A total of 1 × 107 cells were trypsinized, lysed, and 214 μg protein was used per array. The array itself was performed as given in the manual of the supplier. Statistics. Results are expressed as interquartile mean ± interquartile range. The unpaired Student’s t test was used for the comparison against the untreated control with p < 0.05 considered as statistical significant (*) and p < 0.01 considered as statistically highly significant (**).

Table 1. TaqMan Gene Expression Assays Used for Gene Expression Studies of NHEK by qPCR gene

assay ID

gene

assay ID

PPIA UBC IVL KRT10 LOR FLG TGM1

Hs04194521_s1 Hs00824723_m1 Hs00846307_s1 Hs00166289_m1 Hs01894962_s1 Hs00856927_g1 Hs00165929_m1

ZBTB16 C10orf99 ELF3 HSPB3 IGFBP3 FOXN1 ADD2

Hs00957433_m1 Hs01379644_m1 Hs00963881_m1 Hs00937412_s1 Hs00426289_m1 Hs00186096_m1 Hs00242289_m1

with RNase-free water, whereof 2 μL were used for qPCR. Reaction mix: SensiMix II probe, 10 μL; TaqMan gene expression assay, 1 μL; ROX, 0.4 μL; H2O, 6.6 μL; template, 2 μL; 20 μL final volume. For detection of contaminating DNA, no template controls (NTC) were included when performing qPCR. Therefore, templates were replaced with RNase-free water for each TaqMan gene expression assay. Furthermore, “no RT” controls, as described above, were included. PCR master mixes containing SensiMix II probe, TaqMan gene expression assays, and ROX (passive reference dye to normalize wellto-well differences due to artifacts, e.g., pipetting errors) were prepared and pipetted under a vertical laminar air flow workbench for PCR (Telstar). Components were transferred into a 96-well optical reaction plate and after short centrifugation the plate was sealed with an optical adhesive cover and placed in a real-time PCR instrument. Cycling conditions: 1 cycle at 95 °C, 10 min for polymerase activation; 40 cycles at 95 °C, 10 s and 60 °C, 60 s with acquisition at the end of each cycle. PPIA and UBC were used as housekeeping genes and relative gene expression was calculated using the comparative CT method with the following equation:

RQ = 2−(ΔCTsample−ΔCTreference) Gene Expression Analysis by Agilent Whole Human Genome Oligo Microarrays. NHEK were cultivated in six-well plates to 70−80% confluence and incubated with test solutions for 12 and 24 h. Cells were harvested, frozen in liquid N2, and stored at −80 °C until further processing. Further steps were carried out by Miltenyi Biotech GmbH, Bergisch Gladbach. After quantification and quality control of total RNAs (all RNA samples revealed RIN values between 9.8 and 10.0), these were amplified and labeled with cyanine 3 (Cy-3), yielding in fluorescent labeled cRNA and hybridized to Agilent Whole Human Genome Oligo Microarrays (Agilent Technologies). Read out and processing of microarray image files was performed by use of the Agilent Feature Extraction Software, whereas the Rosetta Resolver gene expression data analysis system was used for comparison of two single intensity profiles in a ratio experiment. Analysis of ratio experiments was performed using the Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems). Preparation of Protein Lysates. (A) Differentiation markers by western blot: Cells were collected by trypsinization, washed with PBS, and lysed in freshly prepared RIPA lysis buffer. After incubation on ice for 20 min, samples were centrifuged at 20000 × g for 20 min at 4 °C, and supernatants were collected in fresh reaction tubes for subsequent SDS-PAGE and Western blotting. (B) Identification of molecular targets: Cells were scraped and subsequently homogenized using a Dounce homogenizer. After centrifugation of the lysates for 2 h at 100000 × g at 4 °C, the supernatant was used as cytosolic fraction, and the pellet was dissolved in homogenization buffer or lysis buffer for 2D PAGE and was used as membrane fraction. Fractions were separated by different gel electrophoresis methods and either transferred to a nitrocellulose

■

RESULTS AND DISCUSSION Isolation and Characterization of XG. XG was isolated in yields of 4.9% from dried nasturtium seeds (Tropaeolum majus). The mean molecular weight was determined by HP-SEC on Suprema 3000/100 Å stationary phase with 462 kDa after standard calibration with pullulan standards.11 The polysac-

2159


Article

Biomacromolecules

XG (147% related to UC). Dextran was used as an additional negative control and showed similar fluorescence intensities as the untreated control (91% related to UC), confirming that the protein expression of the differentiation marker was not affected by any polysaccharide in an unspecific manner. XG also revealed an increased expression of involucrin (219% related to UC) compared to the untreated control cells and the negative control dextran (88% related to UC), as displayed in Figure 2.

charide was composed of glucose, xylose, and galactose in the ratio 3:2:1, as determined by GC-quantification of the respective alditol acetates obtained after acid hydrolysis of the polymer and subsequent derivatization.6,11 Linkage analysis11 revealed the presence of the characteristic structural motifs of xyloglucans with a 1,4- /1,4,6-(pGlc) linked backbone (20 respectively 44%), and side chains of 1-pXyl (11%), 1,2-pXyl (16%), and 1-p-Gal (9%). These findings are comparable to published structural data for XG from T. majus.17 Functional Investigation of XG on Skin Cells. XG (100 μg/mL) was investigated concerning its influence on cellular proliferation by BrdU incorporation ELISA18 and viability and metabolic activity (MTT assay19) of HaCaT keratinocyte cell line. No significant changes of the XG-treated cells compared to the untreated control groups were observed (data not shown). However, during the cultivation typical morphological changes of the cells were observed, which indicated that the keratinocytes underwent terminal cellular differentiation. For a more detailed investigation of the potential XGinduced differentiation of the skin cells, protein expression of the differentiation-specific markers cytokeratin 10 (CK) and involucrin was analyzed in primary normal human epidermal keratinocytes (NHEK) after incubation with XG (100 μg/mL) for 7 days using immunofluorescence staining and confocal laser scanning microscopy (CLSM). Fluorescence intensities were calculated using the software ImageJ and differences in the expression of differentiation-specific marker proteins were related to the respective untreated controls (UC). CLSM revealed strong accumulation of CK in the cytoplasm (Figure 1) after treatment with Ca2+ as positive control (fluorescence intensity of 192% related to UC) as well as after treatment with

Figure 2. Confocal laser scanning microscopy after immunostaining of involucrin (A−D) in NHEK incubated with test compounds for 7 days: (A) untreated control; (B) positive control Ca2+ (2 mM); (C) XG from T. majus (100 μg/mL); (D) dextran (100 μg/mL). Nuclei were stained with DAPI (blue). Fluorescence intensities (%) are related to the respective untreated control. Images are from one representative experiment from n = 3 independent experiments, performed with different cells at different times.

To confirm this finding, protein expression of CK and involucrin was monitored by Western blot analysis using XGtreated NHEK. In this experiment cells grown at a higher cell density and undergoing contact inhibition (leading to induction of cellular differentiation20) were used as positive controls (Figure 3A). Again, CK as well as involucrin levels were upregulated by XG, similar to the cell groups undergoing contact inhibition. The differentiation-inducing influence on CK and involucrin was concentration-dependent from 0.1 to 100 μg/mL (Figure 3B) and was also observed for a commercially available mixture of XG oligosaccharides mainly containing hepta- to nonasaccharides (Figure 3C). This finding clearly indicated that the effect on cell physiology was not dependent on the high molecular weight of the polysaccharide but was due to the specific structural elements of the oligomers. The methods used for pinpointing the cellular differentiation visualize directly the increased formation of specific early and late differentiation markers on protein level by two independent protocols, namely, by immune microscopy and blotting. These methods are much more sensitive and specific as monitoring cellular differentiation by radioactive transglutaminase monitoring as often used in older literature.

Figure 1. Confocal laser scanning microscopy after immunostaining of cytokeratin 10 (A−D) in NHEK incubated with test compounds for 7 days: (A) untreated control; (B) positive control Ca2+ (2 mM); (C) XG from T. majus (100 μg/mL); (D) dextran (100 μg/mL). Nuclei were stained with DAPI (blue). Fluorescence intensities (%) are related to the respective untreated control. Images are from one representative experiment from n = 3 independent experiments, performed with different cells at different times. 2160


Article

Biomacromolecules

differentiation. However, gene expression progression was comparable in all experiments. Furthermore, background expression of differentiation markers was observed. More precisely, an extension of the time interval itself in the absence of any test compound caused an increase in the gene expression of differentiation markers. Representative results are displayed in Figure 4 and diagrams of biological replicates are depicted in the Supporting Information, Figures S1 and S2. The gene expression of CK (Figure 4A) was clearly affected in a time-dependent manner by XG and transcript levels were higher for XG-treated groups compared to the respective untreated control groups between 48 and 144 h. Transcript levels of involucrin (Figure 4B) were higher for XG-treated cells compared to the respective untreated controls between 48 and 72 h. XG only induced a slight increase in gene expression of transglutaminase between 24 and 72 h (Figure 4C). As transglutaminase catalyzes the cross-linking of involucrin to membranous proteins,21 its gene expression should evolve simultaneously with involucrin gene expression, as presented in Figure 4B. Gene expression of loricrin (Figure 4D) was highly upregulated by XG, while filaggrin was not influenced (Figure 4E). However, an induction of loricrin and filaggrin gene expression was not detected until 144 h of incubation as loricrin and filaggrin are typical late differentiation markers. Thus, XG served as an inductor for early differentiation markers and only to some extent for late differentiation markers. Concerning qPCR as method for monitoring cellular differentiation, it might be stated that the reproducibility between independent experiments can be problematic: the tendency was always the same, but problems occur often in terms of absolute comparison of the different data sets, especially due to variability between different donors for the primary cells and background expression. From this point of view, qPCR is assessed not as the method of choice for these kinds of investigations. Whole Human Genom Oligo Microarray. To obtain a more detailed insight into the effects induced by XG on keratinocytes Agilent Whole Human Genome Oligo Microarray was used for screening of the gene expression. Microarray experiments were conducted for three different samples: two untreated control samples (UC) at two different time intervals (12 and 24 h) and one group of NHEK treated with XG (100 μg/mL) for 24 h (XG 24 h). An incubation interval of 24 h for XG-treated cells was of interest as typical differentiation markers have been shown to be upregulated after 48 h. UC 12 h was chosen as changes in gene expression analyzed by qPCR were also related to UC 12 h. The additional untreated control group UC 24 h was used to differentiate between the influence of XG and the timedependent process of cellular differentiation itself. Two single intensity profiles were compared (UC 12 h vs UC 24 h and UC 12 h vs XG 24 h). Results were evaluated using Qiagen’s Ingenuity Pathway Analysis (IPA, Qiagen Redwood City, fold factor ≥2, p-value ≤1.00 × 10−02). Initially, downstream biological processes were elucidated (Supporting Information, Figure S3). Table 2 summarizes the results for 24 h treatment of NHEK with XG. Genes grouped to cell cycle processes, cellular assembly, movement, growth, and death, as well as DNA replication, were affected. Again, these results provide further evidence for the differentiation-stimulating

Figure 3. Western blot analysis of differentiation markers cytokeratin 10 and involucrin in NHEK incubated with test compounds for 7 days. (A) Influence of XG on cytokeratin 10 and involucrin expression; UC: untreated control, CI: positive control cells undergoing contact inhibition; XG: xyloglucan from T. majus (100 μg/mL). β-Actin confirms equivalent protein loading (7.5 μg). (B) Concentrationdependent stimulation of cytokeratin 10 and involucrin by XG; UC: untreated control. β-Actin confirms equivalent protein loading (15 μg). (C) Influence of XG polysaccharide (100 μg/mL) and xyloglucan oligosaccharide mixture DP 7−9 (XG7−9; 100 μg/mL) on expression of cytokeratin 10; CI: positive control cells undergoing contact inhibition. β-Actin confirms equivalent protein loading (15 μg).

For a detailed investigation of this differentiation inducing effect of XG gene expression analysis of various differentiationspecific genes (CK, filaggrin, involucrin, loricrin, transglutaminase) was monitored in a time-dependent manner to reflect changes during the differentiation process; therefore, quantitative real-time PCR of NHEKs, cultivated over a 2.5 respectively 6 days period (four independent experiments over 12, 24, 48, and 60 h, and two independent experiments with additional 72 and 144 h check points) was performed. Relative gene expression was calculated using the comparative CT method for quantification; the genes for the peptidylprolyl isomerase A and ubiquitin C served as reference. Changes in gene expression were related to the first untreated kinetic point of 12 h, which was used as basal level. Besides a Ca2+-treated cell group, also, cells cultivated at higher density and undergoing contact inhibition, triggering the cells into terminal differentiation, were used as positive controls. Biological replicates showed high variation in qPCR analysis due to the use of primary keratinocytes from different donors taken at different days. The relatively long incubation time also contributed to the variability in the absolute amount of 2161


Article

Biomacromolecules

Figure 4. qPCR analysis of expression of the differentiation markers: (A) cytokeratin 10, (B) involucrin, (C) transglutaminase, (D) loricrin, (E) filaggrin, related to UC 12 h (x-fold/control) in NHEK incubated with test compounds over 12, 24, 48, 60, 72, and 144 h. UC: untreated control, XG: xyloglucan from T. majus (100 μg/mL), Ca: Ca2+ 2 mM, CI: contact inhibition. n = 1 experiment.

effect of XG based on the correlation of the affected biological processes to differentiation processes. In terms of statistically over-representation of focus genes in preformulated canonical pathways, especially “p38 MAPK signaling” and “granulocyte adhesion and diapedesis”, were detected, the first of which is known to be involved in regulation of keratinocyte differentiation.22,23 The latter pathway is typical for inflammatory responses (Ingenuity Target Explorer, http://targetexplorer. ingenuity.com). This pathway includes integrins and selectins, which were of interest with respect to cell morphology and tissue formation. TGF-β, TNF, KRT14, ErbB2, and INFG were identified as the main upstream regulators influenced by XG (Table 2). These findings were in correlation with the cellular differentiation of keratinocytes induced by XG. Three of the detected top five transcriptional regulators were cytokines, which are known to regulate maturation, growth, and responsiveness of cell populations. Transforming growth factor β (TGF-β) regulates cellular growth (TGF-β2, TGF-β3) and differentiation (TGF-β1) and stimulates synthesis and secretion of protein constituents of the extracellular matrix. TGF-β1 is associated with a differentiated state and is present as an extracellular network surrounding the keratinocytes.24

Tumor necrosis factor (TNF) is upregulated during the inflammatory phase of wound healing and indirectly promotes reepithelialization by induction of fibroblast growth factor FGF7 production.25 Cytokeratin 14 (KRT14) is an intermediate filament typical for proliferating cells.26 Interferon-γ (INFG) modulates keratinocyte differentiation, cell shedding, and the expression of cell surface molecules such as desquamin, a glycoprotein with a critical role in desquamation as it degrades amino sugar.27 ErbB2 is a member of the EGFR family. Whereas EGFR is located in proliferating cells, ErbB2 is found in differentiated cells.28 Additionally, gene expression of single genes was analyzed, and an overview on fold changes of the top up- and downregulated genes of the ratio experiments UC 12 versus 24 h and UC 12 versus XG 24 h are presented in the Supporting Information, Table S1. Most of the top 10 upregulated genes after XG treatment were related to differentiation processes (Table S1): upregulation of transcription factors regulating the expression of differentiation related genes (e.g., ZBTB16, ELF3), modulation of the timing of differentiation stages (e.g., FOXN1), increase of proteins that modulate differentiation2162


Article

Biomacromolecules

Concerning the use of microarray methodology for monitoring of cellular differentiation, it can be stated that the data sets obtained must be evaluated with advanced bioinformatics tools and programs in order to select the relevant pathways influenced by the test compounds. The problem during microarray interpretation is the need for intensive evaluation of the complete literature for every gene influenced by the test compounds concerning published data on functionality concerning cellular differentiation. This is an extremely time-consuming process, but is absolutely essential for correct interpretation of the data set. Target Fishing for the Cellular Binding Partner of XG. In order to investigate the primary target of XG for starting the differentiation process XG was labeled with fluorescein isothiocyanate (FITC) under dibutyltindilaurate catalysis.29 Soluble cytosolic and membrane protein fractions were prepared from NHEK. Dot blot assays indicated a strong association of FITC-XG with the cellular membrane and no interaction with the cytosolic fraction (data not shown); thus, XG seemed to interact with cell surface rather than with intracellular or subcellular compartment structures. Conclusively, the membrane protein fraction was further investigated. Proteins were separated by SDS-PAGE. One lane was stained with Coomassie Brilliant Blue R250, while proteins from the second lane were blotted onto a nitrocellulose membrane, followed by floating the membrane with FITC-XG. Fluorescence imaging revealed nine bands (Figure 5A), which were related to the respective protein bands on the Coomassiestained gel. Those were excised and subjected to MS-based protein identification using LC-MS/MS (nanoAcquity coupled to Q-TOF Premier, Waters Corp.). Peptide gas phase fragmentation spectra were assigned to proteins, as shown in the Supporting Information, Table S2. As expected, keratin peptides were identified as this protein is part of the intermediate cell filaments of differentiated keratinocytes. Additionally, tryptic peptides for epidermal growth factor receptor (EGFR), integrin α2 (ITA2), integrin β4 (ITB4), lysosomal-associated membrane protein 1 (LAMP1), and galectin-7 (LEG7) were detected, all of which are potential binding partners of FITC-XG (Table 3). Especially EGFR was highly interesting due to the fact that this receptor is known as an essential keratinocyte regulating protein closley related to integrins for cellular signaling: α6β4, α2β1, α3β1, and αvβ5 integrins expressed in human epidermis.30 β4 integrin is known to contribute strongly to proliferation.31 If integrins are cross-linked, proliferation is maintained. On the contrary, reduction of ligand binding of integrins is a stimulus for keratinocyte terminal differentiation.32,33 The pro-proliferative signaling occurs in close cooperation with EGFR, which is again controlled by the activated β4 integrin,34 leading to phosphorylation of EGFR on Tyr877 by a Src-family kinase.35 Three different downstream signaling pathways are triggered by β4/EGFR activation: i. The JNK1-mediated activation of the pro-proliferative proto-oncogene c-Myc.36,37 ii. Induction of signal transducer and activator of transcription 3 (STAT3), which enables the cross talk to intercellular junctions and prevents contact-induced growth inhibition by reversibly disassembling intercellular junctions.38

Table 2. Evaluation of Whole Human Genome Oligo Microarray with NHEK, Treated with Xyloglucan (100 μg/ mL) for 24 h (XG 24 h) Against Untreated NHEK Control at 12 and 24 h Incubation Timea UC 12 h vs UC 24 h molecular and cellular functions cell cycle cellular assembly and organization DNA replication, recombination and repair cellular movement cellular growth and proliferation cell death and survival canonical pathways p-value ratio p38 MAPK signaling

2.81E-05

role of IL-17 in psoriasis

1.54E-03

LXR/RXR activation

1.40E-04

mitotic roles of Polo-like 4.03E-04 kinase signaling by Rho family 5.90E-04 GTPases granulocyte adhesion and diapedesis upstream regulators CSF2 FOXM1 CCND1 YY1 KRT14 TGFB1 TNF ERBB2 IFNG

2.06E-11 1.31E-10 1.69E-08 1.74E-08 2.73E-08

16/118 (0.136) 4/13 (0.308) 15/126 (0.119) 10/70 (0.143) 21/250 (0.084)

UC 12 h vs XG 24 h

118 129 52 164 229

208 208

p-value

315 431 438 ratio

5.02E-08 2.11E-04 3.15E-04

1.86E-05

28/118 (0.237) 6/13 (0.462) 21/126 (0.167)

30/175 (0.171)

4.61E-12 4.93E-14 1.69E-12 7.18E-11 5.54E-10

Data represent the top five gene clusters within “Molecular and Cellular Functions” and the number of affected genes; additionally the top canonical pathways are given, with known relation to skin physiology with respective p-values and ratios; the top upstream regulators with p-value of overlap are displayed. a

specific enzymes (e.g., CALML5), upregulation of heat shock proteins regulating protein processing and trafficking (e.g., HSPB3), formation of the cornified envelope (e.g., SBSN), and formation of the filamentous network (e.g., KRT1). Top downregulated genes could mostly not be assigned to differentiation processes with the exception of proliferationpromoting cytokines (e.g., TNFSF15, CXCL5, IL21R) and cell surface receptors (e.g., SIRPG). Microarray findings on selected genes significantly involved in keratinocyte differentiation (marked in Table S1), were subsequently validated by qPCR. The expression levels of a subset of seven genes (chromosome 10 open reading frame 99 (C10orf99), E74-like factor 3 (ELF3), zinc finger and BTB containing 16 (ZBTB16), forkhead box N1 (FOXN1), heat shock 27 kDa protein 3 (HSPB3), insulin-like growth factor binding protein 3 (IGFBP3), and adducin 2 (ADD2)) were examined. Microarray results could be confirmed (data not shown). This indicates that the main information on changes in gene expression of selected differentiation-specific genes after XG treatment obtained from the single microarray experiment have been highlighted by a further, independent methodology. 2163


Article

Biomacromolecules

Figure 5. (A) SDS-PAGE of membrane protein fraction from NHEK after Coomassie staining (middle) and after blotting on nitrocellulose membrane and staining with FITC-labeled xyloglucan (right); protein marker ladder (left). Arrows indicate the isolated bands subjected to further MS analysis. (B) Native PAGE of subcellular fractions, separated into soluble (cytoplasmic, S) and insoluble pellet (membranous, P) fraction from NHEK after Coomassie staining (left) and after blotting and FITC-XG staining (right). Arrows indicate isolated bands subjected to further MS analysis. (C) 2D-PAGE of membrane fraction from NHEK after Coomassie staining (left) and after blotting and FITC-XG staining (right). Marked areas indicate isolated spots subjected to further MS analysis.

triggering cell differentiation. Peus et al.41 already demonstrated that EGFR tyrosine kinase inhibition induces growth arrest and terminal differentiation. Electrophoretic separation under native conditions and FITC-XG-staining of the blotted proteins (Figure 5B) led to the identification of protein disulfide isomerase A3 (PDIA3) and α-tubulin beside actin and heat shock proteins as binding

iii. PI3K/Akt activation, which ensures cell survival in keratinocytes and inhibits several components of the cell-death machinery like BAD.39,40 Following these considerations, the binding of FITC-XG to integrin β4 and EGFR is reasonable, and we hypothesized that this interaction leads to a diminished activation/phosphorylation of EGFR followed by diminished cellular proliferation 2164


Article

Biomacromolecules Table 3. Synopsis of Membrane Proteins from NHEK with Binding to FITC-Labeled Xyloglucan after Separation by SDS-PAGE, Native PAGE under Nondenaturing Conditions, and 2D-PAGE and Identification by nanoLC-Q-TOF MS

actin cytoplasmic 1 epidermal growth factor receptor galectin-7 hydroxymethylglutaryl CoA synthase heat shock proteins HSP7C/71/74/76/ HS71L/HS90A/B integrin α2 integrin β4 keratin lysosomal-associated membrane protein 1 myosin 9 nucleobindin 1 protein disulfide isomerase A1/A3 α-tubulin 1A/1B vimentin

SDSPAGE

native PAGE

2DPAGE

+ + +

+

+

+ + + + + +

+

+ +

+ + + + +

partners of the polysaccharide (Table 3). 2D-PAGE revealed similar results (Figure 5C). PDIs are ubiquitously expressed multifunctional proteins found in the endoplasmatic reticulum and at the plasma membrane, besides a crucial role in protein folding and in glycosylation of proteins.42,43 PDI is discussed to either alter TGF-β1 activity or to modify the TGF-β1 receptor so that cells lose their responsiveness to TGF-β1 and become resistant to the growth-inhibitory effects of TGF-β1.43 TGF-β1 regulates cellular differentiation and is strongly associated with a differentiated keratinocyte state.44 Summarizing the gel electrophoresis experiments integrin/ EGFR and PDI seemed to be reasonable protein targets to explain the differentiation-inducing properties of XG. Especially EGFR-mediated signaling seemed promising as XG-binding might prevent EGFR phosphorylation. Therefore, a Human Phosho-Kinase Array (ProteomeProfiler) was used for parallel determination of the relative levels of phosphorylation of 43 different target proteins. NHEK (107 cells) were incubated for 6 h either with MCBD medium alone (untreated control, UC) or XG (100 μg/mL), followed by trypsinization, lysis, and determination of phosphorylation status. The experiment was performed in two independent assays with two different batches of NHEK at two different dates about six months apart. Both experiments confirmed that phosphorylation of EGFR was decreased in the XG-treated cell groups (Figure 6). Other phosphorylated proteins were reduced in the XG-treated group, for example, cAMP response binding element (CREB), a transcription factor influencing keratinocyte proliferation45 was downregulated by XG (Figure 6). Inhibition of the activation of CREB is known to decrease cell cycle progression46 which, in turn, will trigger the cell into a state of terminal differentiation. The role of CREB in skin cell differentiation is still unclear.47 Some reports show that the CREB protein level is induced during keratinocyte differentiation,48 and others say that it is suppressed.49 It seems that CREB levels are strictly related to the promotor region of loricrin, a late differentiation marker in NHEK.50 As XG apparently inhibits the formation of the phosphorylated CREB (Figure 6) increased gene expression for loricrin would follow.

Figure 6. Influence of 6 h incubation of NHEK with XG (100 μg/mL) on phosphorylation profiles of kinases and protein substrates by Human Phosho-Kinase Array ProteomeProfiler. #: Epidermal growth factor receptor EGFR; *: cAMP response binding element (CREB).

This increase in loricrin gene expression has been observed in our study in the initial experiments by qPCR (Figure 4). Besides the influence of XG on the phosphorylation of EGFR and CREB minor changes were observed in the kinase array for the XG-treated groups, namely STAT2 and HSP60, targets for which the connection to the XG-induced differentiation remains unclear. From these data XG-induced differentiation of keratinocytes can be explained by the mechanism depicted in Figure 7: Inhibition of phosphorylation of EGFR, either by direct interaction with the receptor protein or by interaction with integrins, initiates reduced intracellular signaling by the MAPK/ ERK pathway via RAS/RAF/MEK51 which in turn results in diminished CREB activation leading to a stop of the cell cycle and exit into terminal cellular differentiation. The binding area of XG to EGFR is now of interest. So far, this interaction has only been detected in SDS-PAGE under denaturing conditions, but not in native PAGE, which may have method-inherent reasons. However, it cannot be excluded that the interaction takes place at the intracellular domain of the receptor, although, due to the high molecular weight of XG, we hypothesized an interaction of the extracellular domain. Further studies on the potential internalization of XG into the cell, respectively, binding studies of XG to the receptor by computational and biophysical investigations will shed more light on the underlying mechanisms.

■

CONCLUSION We have described, to our knowledge, for the first time, a polysaccharide that triggers by a specific mechanism skin cells 2165


Article

Biomacromolecules Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Financial support from the German Research Council DFG (Project GRK 1549, International Research Training Group “Molecular and Cellular Glyco Sciences) is acknowledged.

■

(1) Harding, K. G.; Morris, H. L.; Patel, G. K. Brit. Med. J. 2000, 324, 160−163. (2) Thomas, S. J. Wound Care 2000, 9, 56−60. (3) Mosti, G. Phlebology 2013, 28 (Suppl 1), 79−85. (4) Deters, A. M.; Schröder, K. R.; Smiatek, T.; Hensel, A. Planta Med. 2005, 71, 33−39. (5) Deters, A.; Petereit, F.; Schmidgall, J.; Hensel, A. J. Pharm. Pharmacol. 2008, 60, 197−204. (6) Zippel, J.; Deters, A.; Pappai, D.; Hensel, A. Carbohydr. Res. 2009, 344, 1001−1008. (7) Deters, A.; Schröder, K. R.; Hensel, A. J. Cell Physiol. 2005, 202, 717−722. (8) Nie, W.; Deters, A. M. Dermatol. Res. Pract. 2013, 359756. (9) Hensel, A.; Brummell, D. A.; Hanna, R.; Maclachlan, G. Planta 1991, 183, 321−326. (10) Edwards, M.; Dea, I. C. M.; Bupin, P. V.; Reid, J. S. G. Planta 1985, 163, 133−140. (11) Sehlbach, M.; König, S.; Mormann, M.; Sendker, J. D.; Hensel, A. Carbohydr. Polym. 2013, 98, 522−531. (12) Hermann, A.; König, S.; Lechtenberg, M.; Sehlbach, M.; Vakrushev, S.; Peter-Katalinic, J.; Hensel, A. Glycobiology 2012, 22, 1424−1439. (13) Schmidgall, J.; Hensel, A. Int. J. Biol. Macromol. 2002, 30, 217− 225. (14) Manchester, K. L. Biotechniques 1995, 19, 208−209. (15) Lämmli, U. K. Nature 1970, 227, 680−685. (16) Pakkianathan, B. C.; Singh, N. K.; Krishnan, M.; König, S. BCM Biochem. 2012, 13, 5. (17) Le Dizet, P. Carbohydr. Res. 1972, 24, 505−509. (18) Porstmann, T.; Ternynck, T.; Avrameas, S. J. Immunol. Methods. 1985, 82, 169−179. (19) Mosmann, T. J. Immunol. Methods 1983, 65, 55−63. (20) Poumay, Y.; Pittelkow, M. R. J. Invest. Dermatol. 1995, 104, 271−276. (21) Rice, R. H.; Green, H. Cell 1979, 18, 681−694. (22) Efimova, T.; Broome, A.-M.; Eckert, R. L. J. Biol. Chem. 2003, 278, 34277−34285. (23) Eckert, R. L.; Efimova, T.; Dashti, S. R.; Balasubramanian, S.; Deucher, A.; Crish, J. F.; Sturniolo, M.; Bone, F. J. Invest. Dermatol. 2002, 1, 36−40. (24) Levine, J. H.; Moses, H. L.; Gold, L. I.; Nanney, L. B. Am. J. Pathol. 1993, 143, 368−380. (25) Barrientos, S.; Stojadinovic, O.; Golinko, M. S.; Brem, H.; Tomic-Canic, M. Wound Rep. Regener. 2008, 16, 585−601. (26) Eichner, R.; Sun, T.-T.; Aebi, U. J. Cell Biol. 1986, 102, 1767− 1777. (27) Brysk, M. M.; Bell, T.; Hoida, C.; Tyring, S. K.; Rajaraman, S. Exp. Cell Res. 1991, 197, 140−147. (28) Säaf̈ , A.; Pivarcsi, A.; Winge, M. C. G.; Wahlgren, C.-F.; Homey, B.; Nordenskjö ld, M.; Tengvall-Linder, M.; Bradley, M. Arch. Dermatol. Res. 2012, 304, 773−780. (29) Schmidgall, J.; Hensel, A. Int. J. Biol. Macromol. 2002, 30, 217− 225. (30) Adams, J. C.; Watt, F. M. J. Cell Biol. 1991, 115, 829−841. (31) Nikolopoulos, S. N.; Blaikie, P.; Yoshioka, T.; Guo, W.; Puri, C.; Tacchetti, C.; Giancotti, F. G. Mol. Cell. Biol. 2005, 25, 6090−6102. (32) Adams, J. C.; Watt, F. M. Nature 1989, 340, 307−309. (33) Adams, J. C.; Watt, F. M. Cell 1990, 63, 425−435.

Figure 7. Xyloglucan inhibits phosphorylation of EGFR, which leads to decreased intracellular signaling by MAPK/ERK pathway and reduced CREB activity, which again triggers keratinocytes into cellular differentiation.

into differentiation. This result indicates potential use of this polysaccharide for wound healing. Further studies have to be conducted to evaluate if this cell cycle stop will also be possible in other cells, for example, key player cells of the immune cells, cancer cells, or hyperproliferating keratinocytes with impaired differentiation properties, as observed during psoriatic diseases.

■

ASSOCIATED CONTENT

S Supporting Information *

Table S1 gives an overview on fold changes and p-values of the top up- and downregulated genes after Agilent Whole Human Genome Oligo Microarray and evaluation of comparative experiments UC 12 h vs UC 24 h and UC 12 h vs XG 24 h. Details on functional properties are given for the “Top-ten” upregulated genes ZBTB16, CALML5, ELF3, C10orf99, ADD2, IGFBP3, HSPB3, SBSN, FOXN1. Table S2 displays peptide ions assigned to sequences of the different proteins by MS analysis. Figure S1 shows qPCR analysis from four independent experiments of expression of the differentiation markers cytokeratin 10 and involucrin related over 60 h. Figure S2 displays qPCR analysis of expression of the differentiation markers cytokeratin 10, involucrin, transglutaminase, loricrin, and filaggrin related over 144 h. Figure S3 gives a summary on up-/downregulated and unregulated downstream biological processes obtained by Agilent Whole Human Genome Oligo Microarray and evaluation of ratio experiments. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b00553.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Ph.: ++49 251 8333380. Fax: ++49 251 8338341. E-mail: [email protected]. Author Contributions †

These authors contributed equally, sharing first authorship (D.M.Z. and S.B.). 2166


Article

Biomacromolecules (34) Moro, L.; Venturino, M.; Bozzo, C.; Silengo, L.; Altruda, F.; Beguinot, L.; Tarone, G.; Defilippi, P. EMBO J. 1998, 17, 6622−6632. (35) Ishizawar, R.; Parson, S. J. Cancer Cell 2004, 6, 209−214. (36) Guo, W.; Pylayeva, Y.; Pepe, A.; Yoshioka, T.; Muller, W. J.; Inghirami, G.; Giancotti, F. G. Cell 2006, 126, 489−502. (37) Zenz, R.; Wagner, E. F. Int. J. Biochem. Cell Biol. 2006, 38, 1043−1049. (38) Guo, W.; Pylayeva, Y.; Pepe, A.; Yoshioka, T.; Muller, W. J.; Inghirami, G.; Giancotti, F. G. Cell 2006, 126, 489−502. (39) Toker, A.; Yoeli-Lerner, M. Cancer Res. 2006, 66, 3963−3966. (40) Vivanco, I.; Sawyers, C. L. Natl. Rev. Cancer 2002, 2, 489−501. (41) Peus, D.; Hamacher, L.; Pittelkow, M. R. J. Invest. Dermatol. 1997, 109, 751−756. (42) Sullivan, D. C.; Huminiecki, L.; Moore, J. W.; Boyle, J. J.; Poulsom, R.; Creamer, D.; Barker, J.; Bicknell, R. J. Biol. Chem. 2003, 278, 47079−47088. (43) Sipes, N. J.; Miller, D. A.; Bascom, C. C.; Winkler, J. K.; Matrisian, L. M.; Moses, H. L. Cell Growth Differ. 1990, 1, 241−246. (44) Levine, J. H.; Moses, H. L.; Gold, L. I.; Nanney, L. B. Am. J. Pathol. 1993, 143, 368−380. (45) Jang, S. I.; Steinert, P. M. J. Biol. Chem. 2002, 277, 42268− 42279. (46) Rozenberg, J.; Rishi, V.; Orosz, A.; Moitra, J.; Glick, A.; Vinson, C. Mol. Cancer Res. 2009, 7, 654−664. (47) Rozenberg, J. M.; Bhattacharya, P.; Chatterjee, R.; Glass, K.; Vinson, C. PLoS One 2013, 8 (11), e78179. (48) Rutberg, S. E.; Adams, T. L.; Olive, M.; Alexander, N.; Vinson, C.; Yuspa, S.H. Oncogene 1999, 18, 1569−1579. (49) Jang, S. I.; Steinert, P. M. J. Biol. Chem. 2002, 277, 42268− 42279. (50) Jang, S. I.; Steinert, P. M. J. Biol. Chem. 2002, 44, 42268−42279. (51) Orton, R. J.; Sturm, O. E.; Vyshemirsky, V.; Calder, M.; Gilbert, D. R.; Kolch, W. Biochem. J. 2005, 392, 249−261.

2167


Xyloglucan from Tropaeolum majus Seeds Induces Cellular

Recommend Documents