SEARCH

SEARCH BY CITATION

Keywords:

  • Sprouty;
  • NSCLC;
  • cell cycle;
  • Spry1

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

Sprouty1 protein belongs to a family of receptor tyrosine kinase-mediated signaling inhibitors, whose members are usually regulated by growth factors to form a negative feedback loop. Correspondingly fluctuations of Sprouty1 mRNA in response to single growth factors have been observed.

In this report, we investigate Sprouty1 protein levels and show that in non-small cell lung carcinoma-derived cells, the expression levels are unaffected by the serum content in the cellular environment. Although cells harboring K-Ras mutations express insignificant higher Sprouty1 levels, ectopic expression of constitutive active Ras in normal human lung fibroblasts fails to augment Sprouty1 protein content. Furthermore, serum starvation for three days has no influence on Sprouty1 expression and addition of serum or of singular growth factors leaves Sprouty protein levels unchanged. Cell cycle analysis reveals that Sprouty1 levels remain constant throughout the whole cell cycle. These data demonstrate that Sprouty1 expression is not connected with mitogenic signaling and cell proliferation. Copyright © 2013 John Wiley & Sons, Ltd.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

The first member of the Sprouty (Spry) proteins was identified as an inhibitor of growth factor-mediated processes in Drosophila.[1, 2] Consecutively, Spry proteins were recognized as modulators of receptor tyrosine kinase (RTK) signaling in various organisms.[3-5] In mammalians, this protein family comprises four Spry family members, termed Spry1–Spry4.[6] Several studies demonstrated that Spry proteins function in dosing of growth factor-mediated processes. Their loss is usually associated with a phenotype comparable to excess RTK signaling and uncontrolled branching in different organs.[7-12]

Since during development, expression of Spry proteins is mainly observed at sites of known growth factor signaling, a negative feedback loop was postulated.[1, 3, 2, 4]

Consistently, an up-regulation of Spry2 mRNA levels upon stimulation with fibroblast growth factor (FGF) 2 was found in different reports and cells.[13-16] Furthermore, Spry2 expression is shown to be induced by FGF10 in the developing mouse lung[4] and by hepatocyte growth factor (HGF) in human hepatocytes.[17] Also, Spry4 expression can be induced when cells are stimulated by the activation of RTK-mediated signaling via addition of epidermal growth factor (EGF), FGF2, serum or RAS expression.[15, 16]

Although at mRNA level, two reports describe a reduction of Spry1 levels[13, 14], whereas in another experiment, Spry1 mRNA is induced in response to growth factor addition,[18] Spry1 protein in response to growth factor activation was not investigated. The goal of this study is to examine the influence of RTK-signaling on Spry1 protein expression in tumor and normal cells of the lung.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

Cell lines

Five of 15 non–small cell lung cancer (NSCLC) cell lines (A427, A549, Calu-3, Calu-6 and SK-LU-1) were purchased from the American Type Culture Collection (ATCC). The other ten were established at our institute from surgical samples. The characterization concerning K-Ras mutations and histological origin is described in [19]. Additionally, normal human embryonic lung fibroblasts WI-38 and immortal normal human bronchial epithelial cells BEAS2B from ATCC were used. All cell lines were grown and maintained in DMEM (Gibco) supplemented with 10% fetal calf serum (FCS) (Fisher Scientific), glutamine (2 mmol·l–1), penicillin (100U/ml) and streptomycin (100 µg/ml) (Sigma Aldrich). The growth factors used were applied in a final concentration of 10 ng/ml. FGFs were purchased from Strathmann Biotec and EGF from Sigma Aldrich.

Antibodies

Polyclonal antibodies against Spry1 were raised using glutathione S-transferase–tagged NH2-terminal 200 amino acids of human Spry1 as immunogen and immunopurified as described.[19] Antibody recognizing all three RAS isoforms (27H5) was purchased from Cell Signaling Technology and the one recognizing beta-actin (AC-15) from Novus Biological. Cyclin A (H-432), Cyclin D1 (M20) and K-Ras (F234) were detected with antibodies ordered from Santa Cruz.

Immunoblotting

Cells were harvested by scraping and pelleted by centrifugation (1000 x g, 3’). For protein lysis, the cell pellet was resuspended in chilled TGH buffer (50 m mol·l–1 of Hepes, pH 7·6; 150 mmol·l–1 of NaCl; 1.5 mmol·l–1 of MgCl2; 5 mmol·l–1 of NaF; 1 mmol·l–1 of EGTA; 1% TritonX-100; 10% glycerol, Complete® (Roche); 0·5 mmol·l–1 of Na-vanadate, 1 µmol·l–1 of PMSF) and incubated for at least 10 min on ice. After sonication for 2 min, cell debris was removed as a pellet after centrifugation. Protein concentration was determined using the Micro BCATM® Protein Assay Reagent Kit (Pierce). Equal amounts of protein were subjected to PAGE (using 37·5:1-acrylamide:bisacrylamide for polymerization). Separated proteins were transferred onto nitrocellulose membrane (Protran, Schülker und Schleich) by tank blotting. The membrane was probed with respective primary protein-specific, followed by species-specific secondary horseradish peroxidase (HRP)–conjugated antibodies. Western blot signals were visualized by HRP-mediated light emission onto X-ray film using a detection buffer (200 mmol·l–1 of p-Coumaric acid, 1.25 mmol·l–1 of Luminol, 0·009% H2O2 in 0·1 mol·l–1 of Tris pH 8.8).

Propidium iodide–based DNA content analysis (PI-staining)

Analysis of the DNA content was performed as previously described.[20]

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

Spry1 expression in NSCLC-derived cells is unchanged by serum removal

Owing to the importance of mitogenic signaling for the expression of the Spry2 and Spry4 family members, we analysed the influence of serum availability on Spry1 expression in NSCLC-derived cell lines originated from different tumor entities and different stages as well as in immortalized normal bronchial epithelial cells (BEAS2B). To this end, we depleted logarithmically growing cells of serum for 24 h and compared the Spry1 levels with those of cells further cultivated in medium supplemented with 10% FCS (Figure 1a and 1c). As shown in Figure 1b, some cell lines (6/15) show, on average, a slightly decreased Spry1 expression in absence of serum components (less than 1·5 fold), whereas others show an increased expression in the serum-deprived condition, but in all investigated cell lines, the Spry1 levels show no significant change between cells cultivated with or without serum. Furthermore, the presence of serum had no significant influence on Spry1 levels detected in BEAS2B (Figure 1c).

image

Figure 1. Spry1 protein levels in response to serum removal in lung-derived cell lines. a. Spry1 levels of NSCLC-derived cell lines growing for 24 h in absence or presence of serum were analysed using immunoblot. Cell lines harboring a mutated K-Ras allele are indicated by an asterisk (*). b. Spry1 levels of 15 NSCLC-derived cell lines cultivated with (■ boxes) or without serum (▲triangles) are depicted. large cell carcinoma (LCC) Densitometric analyses were performed using Image Quant 5.0 and the mean values obtained from at least two independent experiments were calculated. β-actin levels were used for normalization. c. BEAS2B were cultivated for 24 h in the presence (+) or absence (–) of 10% FCS. Spry1 and β-actin blots of a representative experiment are shown. d. Spry1 levels in logarithmically growing cell lines of squamous cell carcinoma (SCC) and adenocarcinoma (AC) are depicted. e. Spry1 levels from cell lines harboring a K-Ras mutation are compared with wt cell lines. Indicated p-values were calculated using the unpaired t-test in GraphPad Prism

Download figure to PowerPoint

Irrespective of the serum content, we observed that, on average, cells derived from squamous cell carcinoma express slightly less Spry1 than cells isolated from adenocarcinoma (3·3–4·2 fold compared with WI-38 cells) (Figure 1d). Because only adenocarcinoma usually are associated with K-Ras mutations, we analysed if such increases in Spry1 expressions are the consequence of induced signaling via oncogenic Ras. Although cell lines harboring a mutated K-Ras allele (SK-LU-1, A427, A549, Calu-6, VL-2 and VL-4) express, on average, higher Spry1 levels than cell lines expressing wild-type (wt) K-Ras (7·8–2·3 fold compared with WI-38 fibroblasts), the difference was not significant (Figure 1e). Owing to the low number of K-Ras mutated cell lines, we cannot exclude that this is the cause for the statistical insignificance.

Ras-mediated signaling fails to increase Sprouty1 levels

To further evaluate the importance of Ras-mediated signal transduction on Spry1 expression, we introduced adenoviruses expressing mutated Ras isoforms. We used serum starved normal human lung fibroblasts (WI-38) because BEAS2B cells were hardly susceptible to viral infection. Two days post-infection, cells were harvested to determine Spry1 levels. All three oncogenic Ras isoforms are clearly expressed but fail to modulate the endogenous Spry1 levels (Figure 2). Corroborating a dominant negative Ras variant (K-RasS17N) leaves the level of Spry1 expression unchanged (Figure 2). These data indicate that Ras-activated signaling pathways alone are not sufficient to increase Spry1 expression.

image

Figure 2. Spry1 protein levels after activation of Ras signaling through the expression of mutated Ras isoforms. Serum-starved normal human lung fibroblasts (WI-38) were infected with adenoviruses expressing the indicated Ras proteins. Two days post infection, protein extracts were prepared, and Spry 1 levels were determined. Immunoblots of a representative experiment using specific antibodies for Spry1, pan-Ras, K-Ras and β-actin are shown. Spry1 levels from two experiments were calculated after densitometric analysis with Image Quant 5.0. The values obtained for control cells were arbitrarily set as 1. Values measured for β-actin were used for normalization. Using GraphPad Prism software p-values were calculated in an unpaired t-test

Download figure to PowerPoint

Growth factor availability is not a determinant of Sprouty1 expression in normal human lung fibroblasts

To exclude that other, unidentified alterations of signal transduction in cancer-derived cells are responsible for the insensitivity of Spry1 to mitogen availability, we examined the importance of serum for Spry1 expression in primary human cells. First, Spry1 levels in WI-38 cells were analysed at different time points after serum removal. The time curve shows that within a 3-day interval, growth factors present in the serum are not necessary for expression of Spry1 (Figure 3a). In addition to serum, stimulation of the cells with 10 ng·ml–1 of EGF or members of the fibroblast growth factor family (FGF2 and FGF9) fails to increase Spry1 levels at least within a 4-h time frame (Figure 3b). Therefore, we conclude that in primary cells, mitogen-activated signaling is not sufficient to modulate Spry1 levels.

image

Figure 3. Growth factor influence on Spry1 levels in normal human lung fibroblasts. a. WI-38 cells were serum-starved and harvested at the indicated times. Spry1 and β-actin protein levels were monitored. Spry1 levels from two experiments were calculated after densitometric analysis. The values obtained for cells grown in presence of serum were arbitrarily set as 1. Values measured for β-actin were used for normalization. The box blot was drawn using Graph Pad Prism; p-values were determined by an unpaired t-test. b. Serum-starved WI-38 cells were incubated for 4 h with 20% serum (FCS), EGF, FGF2 or FGF9, respectively. All growth factors were added at a concentration of 10 ng/ml. An unspecific band was used as a loading control. The ratio of Spry1 protein levels normalized to the value obtained in unstimulated cells (⊥) is given below the figure

Download figure to PowerPoint

Sprouty1 is constantly expressed throughout the cell cycle

Because pathways governing the transition between quiescence (G0) and proliferation are not essentially involved in regulating Spry1 expression, we next examined if Spry1 levels fluctuate in other phases of the cell cycle. To synchronize cells, WI-38 cells were serum deprived for 3 days. As verified by densitometric DNA analysis, more than 90% of cells accumulate with a 2N DNA content (Figure 4 upper panel). Additionally, low levels of cyclin D1 indicate that the majority of cells are in G0 phase. After addition of serum, more than 80% of the cells are released from quiescence. At 5 h, when cells have entered early G1 phase, cyclin D1 levels increased. Fourteen hours later, a broader G1 peak indicates that cells progress through the G1/S transition. After 26 h, the majority of cells are in S phase, and cyclin A1 expression is detectable. As cells continue to replicate DNA, cells with 4N DNA content accumulate. As an additional feature of G2 phase, Cyclin A1 is further increased 30 h after induction of the proliferating cycle. Thirty-four hours after cell cycle entry, the synchronized cells reach mitosis as indicated by a DNA profile with a strong G2 peak and the reappearance of the G1 peak. Although the collected samples clearly accumulate cells in different cell cycle phases, a fluctuation of Spry1 during the cell cycle could not be observed. These data indicate that the proliferation status of cells is not important for determining Spry1 protein expression.

image

Figure 4. Spry1 levels throughout the cell cycle. WI-38 cells were serum-deprived for 3 days (0) and stimulated with 20% serum. At the indicated times, cells were harvested to perform densitometric analysis of the DNA content (upper panel) and immunoblot using the indicated antibodies. The ratios of Spry1 protein levels normalized to the loading control β-actin and compared with Spry1 prior to serum addition are given

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

Spry proteins are supposed to function in a negative feedback loop of various growth factors as the expression of these molecules is induced in the presence of mitogenic signaling to deactivate the induced signaling cascades. In accordance to this model, Spry2 and, to a lesser extent, also Spry4 protein levels were shown to be strictly dependent on serum and growth factor availability.[13-16] In this study, we demonstrate that Spry1 fails to respond to serum removal in NSCLC-derived, as well as, in normal lung fibroblasts. Corroborating addition of growth factors to serum-deprived cells had no effect on Spry1 protein expression. In contrast, earlier studies investigating RNA level show that Spry1 expression is modulated as a consequence of growth factor availability. Ozaki et al. observed that Spry1 mRNA levels were temporarily increased in response to different growth factors already after 30 min.[21] In two other reports examining mouse fibroblasts and endothelial cells, respectively, Spry1 levels dropped in response to FGF2 addition,[13, 14] but these studies examined Spry1 expression at a time point where the Spry1 levels measured by Ozaki et al. had already dropped again.[21] These observations indicate that growth factor–induced Spry1 fluctuations at mRNA level are not necessarily expressed at protein level. A similar observation is already documented in response to 16 K prolactin, where a strong increase at RNA level is hardly pronounced at the protein level.[22] Although mitogen-induced signaling as well as expression of oncogenic Ras isoforms had no influence on Spry1 levels in serum-starved cells, tumor cells harboring a mutated version of K-Ras express on average more Spry1 protein. In agreement with our observation, it was shown that in tumor-derived cells, increased Spry1 expressions are accompanied by higher levels of phosphorylated ERK1 and 2 (pERK1/2).[21] Corroborating in rhabdomyosarcoma, Schaaf et al. observed that Spry1 levels are elevated in cell lines with higher pERK1/2 levels mostly caused by mutations of a Ras isoform.[23] These observations indicate that signaling downstream of Ras is involved in Spry1 regulation, although our data demonstrate that Ras signaling alone is not sufficient to elevate Spry1 expression. Most likely, other determinants are necessary to modulate Spry1 levels. Because our cell cycle analysis reveals that Spry1 does not fluctuate throughout the cell cycle, we conclude that Spry1 expression is not modulated during proliferation processes. Earlier reports describe that Spry1 levels are influenced by WT-1, a transcription factor known to play an important role in differentiation.[18] Accordingly, Spry1 levels are increased when neuronal cells are induced to perform axonal growth by treatment with either FGF2 or NGF for 4 days,[24] and Spry1 levels are especially high in non proliferating muscle stem cells indicating, that Spry1 is more likely modulated by differentiation processes.[25]

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

Our data demonstrate that Spry1 levels are insensitive to serum and growth factor availability in NSCLC-derived cell lines as well as in normal human lung fibroblasts. Therefore, we conclude that mitogenic signaling is not sufficient to modulate Spry1 expression.

CONFLICT OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

This work was supported by Hertzfelder'sche Familienstiftung and Hochschuljubiläumsstiftung H-01984-2007 and H-3097-2011. The authors report no declarations of interest.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES

We thank Barbara Haigl, Angelina Doriguzzi and Elsa Mühlbacher for helpful discussion and proofreading of the manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. ACKNOWLEDGEMENTS
  10. REFERENCES