Senescent fibroblasts induce moderate stress in lung epithelial cells in vitro

Experimental Gerontology 41 (2006) 532–539 www.elsevier.com/locate/expgero

Senescent fibroblasts induce moderate stress in lung epithelial cells in vitro Babett Bartling *, Grit Rehbein, Rolf-Edgar Silber, Andreas Simm Department of Cardio-thoracic Surgery, Martin Luther University Halle-Wittenberg, Ernst-Grube-Street 40, 06120 Halle (Saale), Germany Received 23 November 2005; received in revised form 9 February 2006; accepted 20 February 2006 Available online 4 April 2006

Abstract Epithelial–mesenchymal interactions contribute to functionality and integrity of the lung epithelium, which might change during ageing and associated cellular ageing. Therefore, we studied the effect of senescent versus pre-senescent lung fibroblasts (WI-38) on mitogenic and stressprotective factors in lung epithelial cells (H358). By use of conditioned medium, we found a growth promoting impact of fibroblasts compared with control medium from epithelial cells associated with activation of ERK1/2, Akt, p70S6K, and EGF receptor. Although senescent fibroblasts mediated similar growth stimulation compared with pre-senescent cells, we observed less protection against spontaneous mitochondrial dysfunction in vitro, higher production of reactive oxygen species and activation of copper/zinc superoxide dismutase. Moreover, senescent cells induced activation of caspase-3/7 in epithelial cells, which was associated with down-regulation of the caspase-inhibitory protein XIAP. In summary, senescent lung fibroblasts induce moderate stress in lung epithelial cells in vitro without affecting growth signaling. q 2006 Elsevier Inc. All rights reserved. Keywords: Fibroblast; Aging; Lung epithelial cell; Proliferation; Cell death; Reactive oxygen species

1. Introduction The temporal activation of defined signaling molecules, transcription factors, extracellular matrix proteins and their receptors contributes to differentiation and development of the lung epithelium. Direct cell–cell interactions of epithelial with mesenchymal cells and the secretion of a variety of molecules are essential in this complex process. The induction of tracheal buds by pulmonary mesenchyme, which was grafted onto tracheal epithelium, is one of the examples demonstrating Abbreviations: Ac-DEVD-AMC, acetyl Asp-Glu-Val-Asp 7-amino-4methylcoumarin; DCF, dichlorodihydrofluorescein; DMEM, Dulbecco’s modified Eagle’s medium; EGFR, epidermal growth factor (EGF) receptor; EGFP, enhanced green fluorescent protein; ERK, extracellular signal-regulated kinase; FCS, fetal bovine serum; GAPDH, glyceraldehydes-3-phosphate dehydrogenase; H2DCFDA, dichlorodihydrofluorescein diacetate; IAP, inhibitor of apoptosis; JNK, c-Jun N-terminal kinase; MAPK, mitogen-activated protein kinase; p70S6K, 70-kDa ribosomal protein S6 kinase; PD, population doubling; PBS, phosphate-buffered saline; PI, propidium iodide; ROS, reactive oxygen species; SAPK, stress-activated protein kinase; SOD, superoxide dismutase; TMRE, tetramethylrhodamine ethyl ester perchlorate; TBS, Trisbuffered saline; XIAP, X chromosome-linked inhibitor of apoptosis. * Corresponding author. Address: Klinik fu¨r Herz- und Thoraxchirurgie, Ernst-Grube-St. 40, D-61120 Halle/Saale, Germany. Tel.: C49 345 552 2878; fax: C49 345 552 2890. E-mail address: [email protected] (B. Bartling). 0531-5565/$ - see front matter q 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.exger.2006.02.006

the importance of epithelial–mesenchymal interactions in mammals (Alescio and Cassini, 1962). However, the multifactorial control of epithelial growth and differentiation does not only play a crucial role in developmental processes. In this regard, cell signaling mediated by the fibroblast growth factor (FGF) contributes to cellular function in fetal and adult lung as well (Demayo et al., 2002; McKeehan et al., 1998). In addition to FGF, fibroblasts are a relevant source of other growth promoting factors, cytokines and proteases that can act at a distance within tissues and considerably modify the local tissue microenvironment. The release of those factors strongly depends on tissue localization and functional status of the fibroblasts, which alters during cellular life-span, in response to chronic inflammation or environmental impacts, such as irradiation. In this regard, it has been shown that irradiated fibroblasts of the tumor stroma support the tumor-forming ability of transplanted cells in mice (Barcellos-Hoff and Ravani, 2000), whereas fibroblasts undergoing cellular senescence can alter the epithelial cell differentiation (Parrinello et al., 2005) and cause the oncogenic transformation of pre-neoplastic epithelial cells (Krtolica et al., 2001). One of the most frequently described effects of fibroblasts or fibroblast-related factors is the induction of growth processes and mediation of cell protection against cellular stress. This is of advantage for stabilizing tissue function and integrity in physiological and pathophysiological conditions (Demayo et al., 2002) but of disadvantage in anti-cancer therapy of

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

solid tumors (Song et al., 2000). However, these data mainly result from experimental studies that used mitotic fibroblasts or isolated factors thereby neglecting age-related changes of the fibroblasts. Therefore, our study focused on the impact of senescent fibroblasts on the induction of mitogenic and cell death-protective signaling pathways in human lung epithelial cells. 2. Material and methods 2.1. Cell culture National Cancer Institute (NCI)-H358 lung epithelial cell line and primary embryonic WI-38 lung fibroblasts (ATCC cell bank; Manassas, VA) were used. WI-38 cells undergo replicative senescence after multiple cell passages, that was assessed by reduction of the population doubling and positive cell staining for acid b-galactosidase (Fig. 1A) as previously described (Bartling et al., 2006). H358 cells were stably transfected with the mammalian expression vector

A

533

pIRES2-enhanced green fluorescent protein (EGFP) (Clontech; Palo Alto, CA) allowing their identification by fluorescence in co-culture (Bartling et al., 2005). Moreover, the lung epithelial cell lines NCI-A549 and NCI-H322 were used. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin and 100 mg/ml streptomycin (Invitrogen; Karlsruhe, Germany) at 37 8C in a 10% CO2 atmosphere. Three days after culture, conditioned medium was taken from WI-38 and H358 cells that corresponds to 3.5 to 4!105 cells/ml. H358 cells were treated with conditioned medium from control H358 cells or WI-38 fibroblasts (mixed with one volume of fresh DMEM/10% FCS medium) for 2 days or by direct co-cultivation for 3 days as described recently (Bartling et al., 2006). Cell proliferation was spectrophotometrically determined by measuring the amount of reduced per oxidized alamarBluee reagent (Bioscource Europe; Nivelles, Belgium) according to the manufacturer’s instruction or by direct cell counting using a Multisizere3 Coulter Counterw (Beckman

propidium iodide uptake (%)

B

5.0

4.0

3.0

2.0

* *

*

1.0

0.0

-galactosidase staining of fibroblasts in co-culture

co-culture D 100 DCF fluorescence (median intensity)

loss of TMRE staining (%)

C 5.0

4.0

3.0

2.0

conditioned medium

(*) *

1.0

0.0

co-culture

*# 80

60

40

20

conditioned medium

control H358

pre-senescent WI-38

conditioned medium senescent WI-38

Fig. 1. Impact of pre-senescent and senescent WI-38 fibroblasts (A) on spontaneous death of H358 cells (EGFP-labeled) is indicated by membrane leakage (propidium iodide uptake, B), mitochondrial dysfunction (loss of mitochondrial TMRE uptake, C) for direct co-culture and the use of conditioned medium. Additionally, intracellular level of reactive oxygen species (DCFH oxidation, D) is shown in response to fibroblasts. *P!0.05, (*)P!0.1 versus co-culturing with/conditioned medium from H358 cells and #PZ0.05 versus pre-senescent WI-38 cells (nZ6).

534

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

Coulter; Krefeld, Germany). H358 cells were additionally treated with hydrogen peroxide (3 mM to 1.5 mM; Sigma, Deisenhofen, Germany) for 2 h after synchronizing them in serum-free medium—or they were cultured in conditioned medium supplemented either with hydrogen peroxide (3–300 mM) or with ascorbic acid (50 mM; Sigma) for 2 days. 2.2. Measurement of cell death, reactive oxygen species and caspase activity Cell death was assessed by uptake of propidium iodide (5 mg/ml PI in PBS) and subsequent cytometric analysis using the FACS Calibur flow cytometer with CellQuest Pro software (Becton Dickinson; San Jose, CA). In the case of direct co-culture with WI-38 fibroblasts, H358 cells were gated by the EGFP signal. Reduction of the mitochondrial membrane potential (DJm) was determined by staining of cells with tetramethylrhodamine ethyl ester perchlorate (25 nM TMRE; Molecular Probes Europe; Leiden, The Netherlands) and flow cytometry as described earlier (Bartling et al., 2004). Cytometric analysis of reactive oxygen species (ROS) based on the deacetylation and oxidation of dichlorodihydrofluorescein diacetate (5 mM H2DCFDA; Molecular Probes Europe). According to the manufacturer, H2DCFDA reacts with hydrogen peroxide, hydroxyl radicals, peroxyl radicals, and peroxynitrite anions. Increase in fluorescence of the DCF product was assessed for TMRE-positive cells only. Caspase activity was measured by cleavage of the fluorogenic substrate specific for caspase-3-like proteases (20 mM Ac-DEVD-AMC as a common substrate for active caspase-3 and -7; Calbiochem; Darmstadt, Germany). After culture of H358 cells in conditioned medium (1.5!105 cells were initially seeded), cells were lysed and subjected to enzyme reaction as described (Bartling et al., 2004). In positive controls, H358 cells were treated with 40 mM cisplatin (Medac; Hamburg, Germany) for 24 h. In negative controls, DEVD-CHO peptide (Calbiochem) has been added for competitive caspase inhibition. AMC liberation was monitored by a FLUOstar OPTIMA reader (BMG Labtechnologies; Offenburg, Germany) at 380 nm excitation and 460 nm emission, finally converted to fluorescence units per minute and normalized per milligram protein. 2.3. Immunoblot analysis Proteins were isolated from cells, quantified and subjected to standard immunoblot procedure as described earlier (Bartling et al., 2006). Phospho-specific rabbit polyclonal antibodies against p-Erk1/2/p-p42/44-MAPK (Thr202/Tyr204), p-Akt (Ser473), p-SAPK/cJNK (Thr183/Tyr185), p-Bad (Ser136), p-EGFR (Tyr845/992/1045/1068) (all are from Cell Signaling Technology; Beverly, MA), the rabbit polyclonal antibodies against XIAP (BD Transduction Laboratories; Lexington, KY), caspase-9 p35 (BD Pharmingen, San Jose, CA) and catalase (Calbiochem), the sheep polyclonal antibodies against manganese superoxide dismutase (SOD2; Calbiochem) and copper/zinc superoxide dismutase (SOD1; Calbiochem) or the

mouse monoclonal anti-Bcl-2 antibody (BD Pharmingen) were used for immunoblot detection. Equal protein loading was controlled by using the polyclonal antibody against GAPDH (Abcam, Cambridge, UK). Protein lysate of K-562 cells (human erythromyeloblastoid leukemia, BD Pharmingen) were used as positive control for Bcl-2 expression. Bound antibodies were detected by calf intestinal alkaline phosphatase-conjugated secondary antibodies (Dianova, Hamburg, Germany) and using the 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium system (Sigma). The intensity of visualized signals was densitometrically estimated by use of the LAS 3000 computer-based imaging system (FUJIFilm; Tokyo, Japan) equipped with AIDA 3.5 software (Raytest; Straubenhardt, Germany). 2.4. Statistics The ANOVA procedure was used for multiple comparisons followed by Dunnett’s or Dunn’s method as appropriate (SigmaStat and SigmaPlot software; Jandel Corp., San Rafael, CA). Data are reported as meanGSEM (nR3 in all cases). In the case of median presentation, the boundary of the Box plots closest to zero indicates the 25th percentile centered about the mean, the line within the box marks the 50th percentile (median), and the boundary of the box farthest from zero indicates the 75th percentile. Whiskers above and below the box represent the 5th and 95th percentiles. P values less than 0.05 were accepted as indicating a significant difference. 3. Results 3.1. Impact of fibroblasts on spontaneous epithelial cell death Direct co-culture with WI-38 lung fibroblast as well as the use of conditioned medium from WI-38 fibroblasts with lung epithelial H358 cells resulted in a diminished amount of spontaneous cell death of H358 cells in vitro. This has been demonstrated by reduced plasma membrane leakage and mitochondrial dysfunction of H358 cells (Fig. 1B and C). However, the protection against mitochondrial dysfunction was less pronounced in response to senescent compared with pre-senescent fibroblasts (Fig. 1C). Subsequent analysis of reactive oxygen species (ROS) by oxidation of the H2DCFDA reagent revealed higher intracellular ROS levels in H358 cells in the presence of conditioned medium from senescent WI-38 fibroblasts (Fig. 1D). This impact has not been detected for presenescent WI-38 fibroblasts (Fig. 1D). However, the higher production of ROS mediated by senescent WI-38 fibroblasts did not significantly affect the fibroblast-induced proliferation of H358 cells (Fig. 2A). This has been demonstrated for conditioned medium from lung WI-38 fibroblasts (Fig. 2A) as well as for direct co-culturing in two- and three-dimensional cell culture models (Bartling et al., 2006). Moreover, we investigated the fibroblast-induced cell proliferation for other lung epithelial cell lines (A549 and H322) and confirmed a better cell propagation in response to fibroblasts (data not shown).

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

3.2. Influence of fibroblasts on growth signaling In order to the fibroblast-mediated proliferation of H358 cells, intracellular factors of the growth signaling cascade are induced. This has been demonstrated for activation of the protein kinase Akt, 70-kDa ribosomal protein S6 kinase (p70S6K) and extracellular signal-regulated kinase (ERK) 1/2 (Fig. 2B–D). Moreover, the use of conditioned medium from pre-senescent versus senescent WI-38 fibroblasts revealed an even higher induction of p70S6K in response to senescent fibroblasts (Fig. 2C). This tendency can be observed for activation of ERK1/2 as well (Fig. 2D). With respect to elevated ROS levels in H358 cells through senescent fibroblasts, we proved the mitogenic induction of H358 cells after treatment with hydrogen peroxide in comparison with conditioned medium derived from fibroblasts. As demonstrated in Fig. 3A, hydrogen peroxide induced the EGF receptor (EGFR) tyrosine kinase as well as the intracellular protein kinase ERK1/2, which corresponds to the induction observed for conditioned medium from senescent WI-38 fibroblasts. However, the hydrogen peroxide-mediated stimulation of ERK1/2 follows a distinct concentration dependency (Fig. 3B) that has not been observed for EGFR activation. Because ERK1/2 plays a major role in mitogenic

p-p70S6 kinase signal (fold changes)

C

cell signaling, we supplemented the conditioned medium from senescent WI-38 cells with ascorbic acid (50 mM) as a hydrophilic ROS scavenger. Adding of ascorbic acid resulted in a slightly reduced stimulation of ERK1/2. In contrast, this was conversely regulated for conditioned medium from control H358 cells (Fig. 3C). However, ascorbic acid had no influence on the significant activation of ERK1/2 (Fig. 3C) and improved proliferation of H358 cells (data not shown) in response to WI-38 fibroblasts. In accordance with this observation, we demonstrated that low-dose hydrogen peroxide (%10 mM) did not stimulate the proliferation of H358 cells cultured in conditioned medium from WI-38 fibroblasts but only of H358 cells cultured in control conditioned medium (Fig. 3D). 3.3. Influence of fibroblasts on stress-associated factors Several kinases of the survival/growth signaling cascade have been reported to impair the induction of mitochondrial cell death by phosphorylation of pro-apoptotic Bad (Datta et al., 1997; Harada et al., 1999, 2001), a member of the bcl-2 gene family. As graphed in Fig. 4A, we indeed found a better phosphorylation of Bad in response to conditioned medium from WI-38 fibroblasts. This has been observed for pre-senescent and senescent fibroblasts as well (Fig. 4A B 2.5 p-Akt kinase signal (fold changes)

1.3

1.2

**

**

1.1

1.0

3.0

D ** #

2.5

2.0

* 1.5

1.0

control H358

p-ERK1/2 signal (fold changes)

alamarBlue reduction (fold changes)

A

535

2.0

1.5

**

**

1.0

2.5

** 2.0

**

1.5

1.0

pre-senescent WI-38

senescent WI-38

Fig. 2. Influence of conditioned medium from pre-senescent and senescent WI-38 fibroblasts on the proliferation of H358 cells (reduction of alamarBluee reagent, A) and activation (phosphorylation, p-) of intracellular protein kinases (Akt, B; p70S6K, C; ERK1/2, D) as determined by respective antibodies. *P!0.05, **P!0.01 versus conditioned medium from H358 cells and #P!0.05 versus pre-senescent WI-38 cells (nZ5).

536

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

WI-38a

conditioned medium 0.6 mM

0.3 mM

0 mM

hydrogen peroxide

H358

A

p-ERK1/2 p-EGFR

p-ERK1/2 signal (fold changes)

B

4

3

2

1

0.3

0

0.75

1.5 mM

hydrogen peroxide C

2.0

D 1.5 proliferation (alamarBlue reduction)

p-ERK1/2 signal (fold changes)

control H358 senescent WI-38

1.8 1.6 1.4 1.2 1.0

ascorbic acid:

–

+

–

+

control H358 pre-senescent WI-38

1.4 1.3 1.2

*

1.1 # 1.0

0.003

0.009

0.03

0.1

0.3 mM

hydrogen peroxide

Fig. 3. Treatment with hydrogen peroxide for 2 h resulted in the activation (phosphorylation, p-) of ERK1/2 and EGFR, which is shown in comparison with the activation induced by conditioned medium from asenescent WI-38 fibroblasts after 2 days (A). ERK1/2 stimulation depends on the concentration of hydrogen peroxide used (nZ3, B). Moreover, activation of ERK1/2 is demonstrated in the presence of conditioned medium from senescent WI-38 cells supplemented with ascorbic acid (50 mM, nZ5, C). Proliferation of H358 cells (reduction of alamarBluee reagent) is shown after cultivation in conditioned medium from control H358 cells and pre-senescent WI-38 fibroblasts supplemented with different concentrations of hydrogen peroxide (nZ6, D). *PZ0.01 versus hydrogen peroxide-free, conditioned medium from WI-38 cells and #P!0.01 versus hydrogen peroxide-supplemented, conditioned medium from H358 cells.

and C). Anti-apoptotic Bcl-2 is relatively low expressed in H358 cells compared with human leukemia (K-562) cells and remained unaffected by conditioned medium from fibroblasts (data not shown). In contrast, we found a reduced protein expression of the caspase-inhibitory protein XIAP in response to senescent WI-38 fibroblasts (Fig. 4B and C). The downregulation of XIAP was associated with higher caspase-3-like proteolytic activities (enzymatic cleavage of the DEVD peptide substrate) in conditioned medium from senescent fibroblasts (Fig. 4D). This is definitely no strong stimulation because the simultaneous treatment with cisplatin (40 mM) resulted in about 10-fold higher enzymatic activities of caspase-3/7 (data not shown). In addition, a slight proteolytic processing of caspase-9 into active p35 fragment has been detected after treating H358 cells with conditioned medium—but this was unaffected by senescent compared with pre-senescent fibroblasts (Fig. 4E). On the basis of ROS production and activation of caspase3-like proteases, we additionally determined whether senescent cells might induce other stress-related factors in H358 cells. However, the stress-associated protein kinase (SAPK) and catalase cannot be identified by immunoblot analysis at detectable level. The copper/zinc superoxide dismutase

(SOD1) and manganese superoxide dismutase (SOD2) are mostly down-regulated in response to pre-senescent WI-38 fibroblasts (Fig. 5A and B). However, this was abolished when culturing H358 cells in conditioned medium from senescent WI-38 cells (Fig. 5A and B). Senescent fibroblasts even significantly induced the protein expression of SOD1 in H358 cells (Fig. 5A). 4. Discussion Replicative senescence limits the proliferation of normal somatic cells in culture causing an irreversible growth arrest and, in contrast to post-mitotic mature cells, multiple changes in their cell function. Those functional changes include the secretion of a variety of soluble molecules, including metalloproteinases, pro-inflammatory cytokines and growth factors (reviewed in Krtolica and Campisi, 2002) that can strictly impair the local tissue homeostasis. On the basis of epithelial–mesenchymal interactions in the organism, the constitutive secretion of these molecules from senescent cells may contribute to the age-related decline in tissue function and integrity. Therefore, we established conditioned cell culture medium enriched with cellular factors that are released from

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

*

2.0 1.5 1.0

1.0

*

P-Bad

0.5

XIAP GAPDH

D 10.0

E 600 caspase-3/-7 activity (DEVD-AMC release/min/mg)

caspase-9 p35 (% cleavage)

sen. WI-38

2.5

C pre-sen. WI-38

*

control H358

B 1.5 XIAP expression (fold changes)

p-Bad signal (fold changes)

A 3.0

537

8.0 6.0 4.0 2.0

* # 500

400

300 control H358

pre-senescent WI-38

senescent WI-38

Fig. 4. Effect of conditioned medium from pre-senescent and senescent WI-38 fibroblasts on the phosphorylation (p-) of pro-apoptotic Bad in H358 cells (A, C) and expression of the caspase-inhibitory protein XIAP (B, C) (each nZ3) as determined by antibody staining per GAPDH. Moreover, activity of caspases-3-like proteases (Ac-DEVD-AMC release, D) and proteolytic processing of caspase-9 into active p35 fragment (E) was determined (each nZ6). *P!0.05 versus conditioned medium from H358 cells and #P!0.05 versus pre-senescent WI-38 cells.

senescent human lung fibroblasts (WI-38) in vitro. By use of this conditioned medium, we found that secreted factors of senescent fibroblasts had the capacity to induce mild cellular stress in epithelial cells from human lung (H358) associated with a less pronounced protection against spontaneous mitochondrial dysfunction of H358 cells in vitro as compared to pre-senescent fibroblasts. Although mitochondrial dysfunction proceeds the final cell death in apoptosis and necrosis as well (Kroemer et al., 1998), we did not observe a significant impact of senescent fibroblasts on the spontaneous cell death

B 2.00

1.75

#

1.50 1.25 1.00 * 0.75

control H358

SOD2 expression (fold changes)

SOD1 expression (fold changes)

A 2.00

in vitro. Nevertheless, we found a reduced expression of the inhibitor of apoptosis (IAP) family protein XIAP. XIAP inhibits the proteolytic activation of procaspase-9 into active caspase-9, the proximal caspase of mitochondria-induced apoptosis, as well as the activation of caspase-3/7 terminating the apoptotic caspase cascade (Deveraux et al., 1998). Although we did not detect fibroblast-induced differences in the proteolytic cleavage of procaspase-9, down-regulation of XIAP correlates with an increased enzymatic activity of caspase-3/7 in response to senescent fibroblasts. Moreover, the

1.75 1.50 1.25 1.00

(*)

0.75

pre-senescent WI-38

senescent WI-38

Fig. 5. Impact of conditioned medium from pre-senescent and senescent WI-38 fibroblasts on the expression of SOD1 (A) and SOD2 (B) as determined by immunoblot per GAPDH (nZ6). *PZ0.05, (*)P!0.1 versus conditioned medium from H358 cells and #P!0.05 versus pre-senescent WI-38 cells.

538

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

activity of caspase-9 might be impaired as well, because IAP proteins inhibit both, the proteolysis of procaspases and the enzymatic activity of processed caspases (reviewed in Shi, 2002). Independent of its function as caspase-inhibitory protein, XIAP acts as an activator of the nuclear factor (NF)-kB-mediated cell survival (Hofer-Warbinek et al., 2000), suggesting a possible impact of senescent fibroblasts on the NF-kB signaling via XIAP. In addition to altered compounds of the cell death machinery, we observed a more frequently induced production of reactive oxygen species (ROS) in H358 cells. This increased level of ROS and the limited protection against spontaneous mitochondrial dysfunctions in response to senescent fibroblasts suggests the simultaneous induction of anti-oxidative defense mechanisms and other stress-protecting factors in H358 cells. However, both mitochondria-associated proteins manganese superoxide dismutase (SOD2) and anti-apoptotic Bcl-2 remained unchanged, whereas the expression of catalase has not been found at detectable level. The presence of fibroblasts even reduced the expression of copper/zinc superoxide dismutase (SOD1), which is in contrast to mitochondrial SOD2 located in the in intracellular cytoplasmic spaces (Zelko et al., 2002). Nonetheless, the down-regulation of SOD1 was abolished when H358 cells had been cultured in conditioned medium from the senescent fibroblasts, suggesting partial activation of anti-oxidative defense mechanisms. This observation is of high interest because the continuous up-regulation of SOD1 might contribute to a long-term stabilization of adjacent epithelial cells and cellular protection against additional/stronger stress signals (i.e. pre-conditioning). In this context, it has additionally to point out that even active caspase-3 is not only involved in cell death but also cell death-independent mechanisms and cell proliferation (Hueber et al., 2000). The interaction of senescent fibroblasts with epithelial cells seems rather to be associated with a mild than an intense stress situation because we did not find stimulation of the stress-activated protein kinase (SAPK). Additionally, we did not reveal a significant impact of senescent fibroblasts on the fibroblast-induced proliferation in comparison with pre-senescent cells and activation of growth signaling pathways confirming our assumption. This has been demonstrated by detection of active protein kinases that are commonly induced in growth factor response (Schaeffer and Weber, 1999; Thomas, 2002; Vanhaesebroeck and Alessi, 2000), including extracellular signal-regulated kinase (ERK) 1/2, Akt kinase and 70-kDa ribosomal protein S6 kinase (p70S6K). Ribosomal p70S6K, known to regulate cell growth by inducing protein synthesis components (reviewed in Thomas, 2002), was even more induced in the presence of conditioned medium from senescent cells. Moreover, p70S6K mediates site-specific phosphorylation of Bad, which consequently inactivates this proapoptotic molecule. Enhanced levels of inactive Bad have also been observed in H358 cells due to conditioned medium from fibroblasts. Although there was, in contrast to p70S6K, no difference between conditioned medium from pre-senescent and

senescent fibroblasts, Bad phosphorylation might additionally be mediated by active Akt kinase (Datta et al., 1997) or the cAMP-dependent protein kinase (PKA) (Harada et al., 1999). This suggests that senescent fibroblasts seem not to impair direct coupling of growth factor/survival signals to the cellintrinsic death machinery via Bad pathway. Presently, we do not know what factors released from senescent WI-38 fibroblasts are responsible for an increased generation of ROS in H358 epithelial cells. Moreover, a number of mechanisms can generate oxygen radicals within cells (reviewed in Simm and Broemme, 2005). Among the senescence-related proteins (reviewed in Krtolica and Campisi, 2002), proinflammatory interleukin-1a might be one candidate protein able to induce ROS (Bohler et al., 2000), whereas the up-regulation of extracellular matrix metalloproteinases in senescent fibroblasts (Millis et al., 1992) seems rather to be a consequence of the oxidative stress (Nelson and Melendez, 2004) than an inductor. Moreover, changes in the redox homeostasis are not necessarily implicated in cellular injury. There is growing literature suggesting transient increase in ROS levels as an important mediator of cell proliferation associated with activation of various signaling molecules and pathways (reviewed in Esposito et al., 2004). In this context it has been demonstrated, that intra- or extracellularly produced ROS stimulate ERK and Akt pathways through direct activation of growth factor receptors themselves, thereby mimicking ligand induced signaling by growth factors, such as platelet-derived growth factor (PDGF) (Gonzalez-Rubio et al., 1996) and epidermal growth factor (EGF) (Gamou and Shimizu, 1995). Interestingly, the EGF-like growth factor heregulin is also described as a candidate molecule upregulated in senescent fibroblasts (Linskens et al., 1995). Therefore, we analyzed the impact of hydrogen peroxide on the mitogenic signaling in H358 cells showing that hydrogen peroxide stimulates ERK1/2 as did conditioned medium from WI-38 fibroblasts. The hydrogen peroxide-induced activation of ERK1/2 occurred in a biphasic manner, which has also been described for other types of cells (Meloche et al., 1992). Moreover, we demonstrated a receptor activation of the EGF receptor tyrosine kinase in response to hydrogen peroxide and conditioned medium from senescent WI-38 fibroblasts. Adding of hydrogen peroxide at moderate concentration even enhanced the proliferation of H358 cells. However, this effect was abolished when culturing H358 cells in conditioned medium from pre-senescent fibroblasts suggesting that fibroblast-related factors are primarily sufficient to improve the proliferation of their neighboring epithelial cells. Nevertheless, this potency might be impaired by senescent fibroblasts but compensated by transient induction of ROS. Adding of the hydrophilic antioxidant ascorbic acid at least showed a slightly reduced activation of ERK1/2 in response to conditioned medium from senescent fibroblasts. Nonetheless, it has to keep in mind that increased levels of ROS in response to senescent fibroblasts might also be directly associated with cellular stress as suggested by an increased activation of caspase-3 and activation of SOD1 in H358 epithelial cells.

B. Bartling et al. / Experimental Gerontology 41 (2006) 532–539

Taken together, from our data we conclude that the integrity and function of the lung epithelium can be affected by senescent fibroblasts due to induction of mild stress in neighboring epithelial cells. Although this does not impair the fibroblast-induced activation of various signaling molecules and pathways involved in epithelial cell growth and survival, it might still be of disadvantage in that case, when oxidatively damaged cells do not undergo but overcome cell arrest (i.e. become abnormal). The simultaneous up-regulation of anti-oxidative mechanisms including SOD1 might stabilize this fragile condition.

Acknowledgements This project was supported by Wilhelm Roux grant of the Bundesministerium fu¨r Bildung und Forschung, BMBF (FKZ7/04) and by the Deutsche Forschungsgemeinschaft, DFG (SFB 598/TPA5). References Alescio, T., Cassini, A., 1962. Induction in vitro of tracheal buds by pulmonary mesenchyme grafted on tracheal epithelium. J. Exp. Zool. 150, 83–94. Barcellos-Hoff, M.H., Ravani, S.A., 2000. Irradiated mammary gland stroma promotes the expression of tumorigenic potential by unirradiated epithelial cells. Cancer Res. 60, 1254–1260. Bartling, B., Lewensohn, R., Zhivotovsky, B., 2004. Endogenously released Smac is insufficient to mediate cell death of human lung carcinoma in response to etoposide. Exp. Cell. Res. 298, 83–95. Bartling, B., Hofmann, H., Weigle, B., Silber, R., Simm, A., 2005. Down-regulation of the receptor for advanced glycation end-products (RAGE) supports non-small cell lung carcinoma. Carcinogenesis 26, 293–301. Bartling, B., Demling, N., Silber, R.E., Simm, A., 2006. Proliferative stimulus of lung fibroblasts on lung cancer cells is impaired by RAGE. Am. J. Respir. Cell. Mol. Biol. 34, 83–91. Bohler, T., Waiser, J., Hepburn, H., Gaedeke, J., Lehmann, C., Hambach, P., Budde, K., Neumayer, H.H., 2000. TNF-alpha and IL-1alpha induce apoptosis in subconfluent rat mesangial cells. Evidence for the involvement of hydrogen peroxide and lipid peroxidation as second messengers. Cytokine 12, 986–991. Datta, S.R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., Greenberg, M.E., 1997. Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91, 231–241. Demayo, F., Minoo, P., Plopper, C.G., Schuger, L., Shannon, J., Torday, J.S., 2002. Mesenchymal–epithelial interactions in lung development and repair: are modeling and remodeling the same process? Am. J. Physiol. Lung. Cell. Mol. Physiol. 283, L510–L517. Deveraux, Q.L., Roy, N., Stennicke, H.R., Van Arsdale, T., Zhou, Q., Srinivasula, S.M., Alnemri, E.S., Salvesen, G.S., Reed, J.C., 1998. IAPs block apoptotic events induced by caspase-8 and cytochrome c by direct inhibition of distinct caspases. EMBO J. 17, 2215–2223. Esposito, F., Ammendola, R., Faraonio, R., Russo, T., Cimino, F., 2004. Redox control of signal transduction, gene expression and cellular senescence. Neurochem. Res. 29, 617–628. Gamou, S., Shimizu, N., 1995. Hydrogen peroxide preferentially enhances the tyrosine phosphorylation of epidermal growth factor receptor. FEBS Lett. 357, 161–164.

539

Gonzalez-Rubio, M., Voit, S., Rodriguez-Puyol, D., Weber, M., Marx, M., 1996. Oxidative stress induces tyrosine phosphorylation of PDGF alpha-and beta-receptors and pp60c-src in mesangial cells. Kidney Int. 50, 164–173. Harada, H., Becknell, B., Wilm, M., Mann, M., Huang, L.J., Taylor, S.S., Scott, J.D., Korsmeyer, S.J., 1999. Phosphorylation and inactivation of BAD by mitochondria-anchored protein kinase A. Mol. Cell 3, 413–422. Harada, H., Andersen, J.S., Mann, M., Terada, N., Korsmeyer, S.J., 2001. p70S6 kinase signals cell survival as well as growth, inactivating the pro-apoptotic molecule BAD. Proc. Natl Acad. Sci. USA 98, 9666–9670. Hofer-Warbinek, R., Schmid, J.A., Stehlik, C., Binder, B.R., Lipp, J., de Martin, R., 2000. Activation of NF-kappa B by XIAP, the X chromosomelinked inhibitor of apoptosis, in endothelial cells involves TAK1. J. Biol. Chem. 275, 22064–22068. Hueber, A.O., Zornig, M., Bernard, A.M., Chautan, M., Evan, G., 2000. A dominant negative Fas-associated death domain protein mutant inhibits proliferation and leads to impaired calcium mobilization in both T-cells and fibroblasts. J. Biol. Chem. 275, 10453–10462. Kroemer, G., Dallaporta, B., Resche-Rigon, M., 1998. The mitochondrial death/life regulator in apoptosis and necrosis. Annu. Rev. Physiol. 60, 619–642. Krtolica, A., Campisi, J., 2002. Cancer and aging: a model for the cancer promoting effects of the aging stroma. Int. J. Biochem. Cell Biol. 34, 1401–1414. Krtolica, A., Parrinello, S., Lockert, S., Desprez, P.-Y., Campisi, J., 2001. Senescent fibroblasts promote epithelial cell growth and tumorigenesis: a link between cancer and aging. Proc. Natl Acad. Sci. USA 98, 12072–12077. Linskens, M.H.K., Feng, J., Andrews, W.H., Enlow, B.E., Saati, S.M., Tonkin, L.A., Funk, W.D., Villeponteau, B., 1995. Cataloging altered gene expression in young and senescent cells using enhanced differential display. Nucleic Acids Res. 23, 3244–3251. McKeehan, W.L., Wang, F., Kan, M., 1998. The heparan sulfate-fibroblast growth factor family: diversity of structure and function. Prog. Nucleic Acid Res. Mol. Biol. 59, 135–176. Meloche, S., Seuwen, K., Pages, G., Pouyssegur, J., 1992. Biphasic and synergistic activation of p44mapk (ERK1) by growth factors: correlation between late phase activation and mitogenicity. Mol. Endocrinol. 6, 845–854. Millis, A.J., Hoyle, M., McCue, H.M., Martini, H., 1992. Differential expression of metalloproteinase and tissue inhibitor of metalloproteinase genes in aged human fibroblasts. Exp. Cell Res. 201, 373–379. Nelson, K.K., Melendez, J.A., 2004. Mitochondrial redox control of matrix metalloproteinases. Free Radic. Biol. Med. 37, 768–784. Parrinello, S., Coppe, J.P., Krtolica, A., Campisi, J., 2005. Stromal–epithelial interactions in aging and cancer: senescent fibroblasts alter epithelial cell differentiation. J. Cell Sci. 118, 485–496. Schaeffer, H.J., Weber, M.J., 1999. Mitogen-activated protein kinases: specific messages from ubiquitous messengers. Mol. Cell Biol. 19, 2435–2444. Shi, Y., 2002. Mechanisms of caspase activation and inhibition during apoptosis. Mol. Cell 9, 459–470. Simm, A., Broemme, H.-J., 2005. Reactive oxygen species (ROS) and aging: Do we need them, can we measure them, should we block them? Signal Transduction 3, 115–125. Song, S., Wientjes, M.G., Gan, Y., Au, J.L., 2000. Fibroblast growth factors: an epigenetic mechanism of broad spectrum resistance to anticancer drugs. Proc. Natl Acad. Sci. USA 97, 8658–8663. Thomas, G., 2002. The S6 kinase signaling pathway in the control of development and growth. Biol. Res. 35, 305–313. Vanhaesebroeck, B., Alessi, D.R., 2000. The PI3K–PDK1 connection: more than just a road to PKB. Biochem. J. 346 (Pt 3), 561–576. Zelko, I.N., Mariani, T.J., Folz, R.J., 2002. Superoxide dismutase multigene family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and EC-SOD (SOD3) gene structures, evolution, and expression. Free Radic. Biol. Med. 33, 337–349.