Substrate stiffness modulates mRNA expression profiling in breast cancer cells

Abstract

Identifying effective targets induced by ECM stiffness is of critical importance for treating metastatic cancer diseases, which are followed by changes in the mechanical microenvironment in cancer cells. In this study, polyacrylamide hydrogel substrates with different stiffnesses were prepared and mRNA microarrays were performed to analyze the mRNA expression profiles in breast cancer cell line SK-BR-3 grown on different stiffness substrates. The results indicated that the expressions of 1831 genes were changed significantly in the SK-BR-3 cells on the different stiffness substrates. GO and KEGG pathway analyses of the differently expressed genes in five significant profiles annotated that the most significant pathways were cell cycle, ubiquitin mediated proteolysis RNA transport and pathways in cancer. Finally, the network of genes and gene interaction based on these differently expressed genes was established, and the phosphorylation of AKT and ERK, respectively the downstreams of the PI3K and Ras signal pathways, was further validated. The genes identified in this study may represent good therapeutic targets, and further study of these targets may help to increase our understanding of the mechanisms underlying the pathological processes and therapy for metastatic breast cancer disease.

Keywords

Stiffness breast cancer tumor microenvironment mRNA microarray bioinformatics

1 Introduction

Although great efforts have been made to understand the mechanism underlying the pathogenic process of breast cancer, breast cancer remains a leading cause of death amongst women worldwide [21]. In recent years the tumor microenvironment (TME) has become the focus of intense research, which may prove useful in prognosis and generate new therapeutic targets by understanding the alterations which occur in the stroma around a tumor [1].

In fact, cells contact with other cells as well as with the ECM, both of which provide not only a biochemical context, but also a mechanical context. Worth noting is that the changing composition, organization and biomechanical properties of the ECM surrounding tumors indicate to us the importance of biomechanical factors in cancer [15 , 22]. At present, it has become increasingly apparent that cells respond to physical cues, such as ECM stiffness. It has also been found that cells sensed the stiffness of their microenvironment by cell-mediated contraction, and responded with appropriate mechanosignaling events [16, 19]. In addition to the variances of chemical components, biological molecules or microstructures in the connective tissue surrounding carcinoma cells, matrix mechanics have been found to dramatically influence cell behavior. For example, Schrader et al. found that increasing the matrix stiffness promoted proliferation and chemotherapeutic resistance, while a soft environment induced the reversible cellular dormancy and stem cell characteristics in hepatocellular carcinoma (HCC) [20]. Another study indicated that microenvironmental stiffness enhanced glioma cell proliferation by stimulating epidermal growth factor receptor signaling [27]. However, the putative molecular mechanism underlying the biomechanical forces in the tumor microenvironment remains poorly understood. The emergence of the genomic approach has provided a way by which to identify the stiffness-associated genes and understand the mechanism of breast cancer development.

In the present study, polyacrylamide hydrogel substrates with different stiffness were prepared to mimic the mechanical environment of breast cancer cells, and used RNA microarrays to define the gene expression profile in breast cancer SK-BR-3 cells induced by different stiffness substrates. We further integrated the differently expressed mRNAs results and constructed a gene-gene interaction network map, which attempted to determine how the mechanical properties in the tumor microenvironment modulate functionally related genes involved in breast cancer progression.

2 Material and methods

2.1 Substrate preparation and characterization

Polyacrylamide hydrogels with different stiffnesses were generated according to the previously described method, with some modifications [25]. After 20 mm glass-bottom plates were washed clearly and dipped in 0.1 M NaOH overnight, 100 μl of 3-aminopropyltrimethoxysilane (APES, Sigma) was added to the surface of the plates to react for 10 min, followed by 0.5% glutaraldehyde in PBS for 3 h.. Then the glass-cover plates were pretreated with 20 μl of dichlorodimethylsilane (DCDMS) in the fume hood for 10 min, and different percentages of acrylamide, bis-acrylamide, APS, TEMED(all from Sigma) and distilled H₂O were mixed to obtain the desired stiffness gels. Next, the gels were polymerized, the glass-cover plates were discarded, and the gels were rinsed three times to remove the unpolymerized acrylamide. After the polyacrylamide hydrogel substrates were transferred into 24/96 well cell culture plates or culture dishes for UV irradiation for 3 hours, the gels were washed with 0.1 M HEPES three times, and sulfo-SANPAH (ProteoChem) solution was added to cover the gels. Then the gels were placed under a 365 nm UV light source for 15 min and rinsed with HEPES three times to remove the excess solution. After all of the plates were coated with type I collagen(gibco^® by life technologies^TM) at a concentration of 0.1 mg/mL for 3 h at 37°C and rinsed once with sterile water, the stiffness measurement was performed by Atomic Force Microscope (AFM), which combines an inverted microscope (Nikon,TE2000U) with a MAC mode AFM instrument (Agilent, 5500). Individual force curves contained 1000 data points, with a Z piezo-displacement between 0 and 6 μm. The AFM cantilever with a typical spring constant (k) of about 0.02 N m–1 was attached to a silica sphere and the diameter was about 10 μm. Optical images were recorded with a high resolution CCD (Roper, Cascas 512B). To calculate the Young’s modulus of the substrates, the force curves were converted into force–indentation curves and fitted with the spherical Hertzmodel [3, 23].

2.2 Cell culture on different substrates

The breast cancer cell line SK-BR-3 cells were cultured in an RPMI1640 medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 mg/ml streptomycin in a humidified 37°C chamber with 5% CO2. For cell viability analysis, a total of 5×10³ cells were seeded per well in a 96-well plate with different substrates for 72 h. Next, the supernatant was discarded, and 100 μL CCK-8 working solution was added into each well and incubated at 37°C for 3 h. The absorbance at 450 nm was read on a spectrophotometer. The cell viability on plastic plate was considered as 100%.

2.3 Total mRNA isolation and mRNA microarray analysis

The total RNA from each sample was isolated using a TRIZOL Reagent, and the RNA integrity was checked by SMA3000 spectrophotometry and 1.5% formaldehyde denaturant gel electrophoresis. Then, the total RNA was amplified and hybridized to GeneChip Human Gene 1.0 ST Array (Affymetrix), according to the protocols. After washing with the holding buffer, the GeneChip images were scanned using a GeneChip Scanner 3000 7 G. Three microarrays were performed for each condition.

2.4 Multi-class dif analysis

The random-variance model F-test was used to filter the differentially expressed genes for both the control and experiment groups. The differentially expressed genes were identified according to the p-value threshold (P < 0.05).

2.5 Series test of cluster (STC) analysis

The series test of cluster algorithm of gene expression dynamics was used to profile the gene expression time series and to select the most probable set of clusters generating the observed time series [29, 30]. After the differential expression genes were identified in a logical sequence according to RVM corrective ANOVA, a set of unique model expression profiles was selected in accordance with the different signal density change profiles of genes under different conditions. Next, the raw expression values were converted into log2 ratios, and the expression model profiles were related to the actual or expected number of genes assigned to each model profile. After Fisher’s exact test and multiple comparison tests were performed, the significant profiles were shown to have higher probability than expected.

2.6 Gene ontology (GO) and pathway analysis

GO Analysis is applied to determine the differently expressed genes belonging to the main function according to the Gene Ontology, which is the key functional classification of NCBI. In general, Fisher’s exact test and χ² test are applied to classify the GO category, and the false discovery rate (FDR) is calculated to correct the P-value.

Similarly, pathway analysis was used to analyze the significant pathways of the differential genes according to the Kyoto encyclopedia of genes and genomes (KEGG) database. Two-side Fisher’s exact test and χ2-test were applied to select the significant pathway category, and the FDR was used to correct the p-value.

2.7 Signal-net analysis

The gene-gene interaction network map was built based on the data of differentially expressed genes. To investigate the global network, we turn to the connectivity (also known as degree), defined as the sum of connection strengths with the other network genes. For a gene in the network, the number of source genes of a gene is known as the indegree of the gene, and the number of target genes of a gene is its outdegree. The character of genes is described by betweenness centrality measures reflecting the importance of a node in a graph relative to other nodes.

2.8 Western blot

After the SK-BR-3 cells were harvested in a lysis buffer containing inhibitors for phosphatase and protease (Na2P2O4 and PMSF), the concentration of the protein was determined with an Enhanced BCA Protein Assay Kit (Beyotime, China). A total of 20 μg of proteins in each sample were loaded for SDS-PAGE (10% SDS gel). Then the proteins were electrophoresed and transferred onto a polyvinylidene difluoride (PVDF) membrane (GE Healthcare, UK) at 4°C. Following several rinses with TTBS (20 mM Tris–Cl, pH 7.5, 0.15 M NaCl and 0.05% Tween-20), the transferred PVDF membrane was then blocked with 10% non-fat milk in TTBS for 1 h and incubated with primary antibodies overnight at 4°C and corresponding secondary antibodies (Cell Signaling Technology, USA) for 1 h. Next, an Enhanced Chemiluminescence (ECL) kit (GE Healthcare, UK) was used to detect the signals. In order to verify the equal loading of the proteins, the blots were reprobed with primary monoclonal antibody against GAPDH (Cell Signaling Technology, USA).

2.9 Statistics

The experiments were repeated at least three times. All data are expressed as the means±SE, and were acquired using the software Origin 8.5.

3 Results

3.1 Viability of SK-BR-3 cells grown on different stiffness substrates

The development of breast cancer involves a series of intermediate processes from ductal hyperproliferation, to invasive carcinoma, and finally to metastatic disease. Young’s Modulus of 0.45 kPa, 4.04 kPa and 23.51 kPa, equivalent to the stiffness of the corresponding matrices of frequent organ metastasis, such as brain, lymph node and mammary tissue, mammary tumor and lung tissue, spinal cord and collagenous bone tissue, are composed of different ratios of acrylamide and bis-acrylamide [5, 8]. After the SK-BR-3 cells were cultured on three different substrates for 72 h, the cell viability was assessed by using the CCK-8 kit. The results indicated that the cell viability was elevated with the increase in substrate stiffness (Fig. 1).

3.2 Differential expression of mRNAs in SK-BR-3 cells grown on different stiffness substrates

The mRNA microarray, which contained probes for 28,869 genes, was used to investigate the mRNA expression profiles induced by the substrate stiffness. The expression of 1831 genes was changed significantly in the SK-BR-3 cells on different stiffness by the RVM F-test (P < 0.05). The hierarchical clustering analysis of these changed mRNAs is shown in Fig. 2. In order to further define the tendency of gene expression induced by stiffness, the series test of cluster (STC) algorithm was analyzed to profile the gene expression series, and to identify the most probable set of clusters generating the observed stiffness series. These differential genes were divided into 15 types of expression profiles according to the alignment of the 0.45 kPa, 4.04 kPa and 23.51 kPa substrates (Fig. 3). Of the total, the top five profiles contained genes with significant tendency changes in expression following the observed stiffness series (P < 0.05).

3.3 GO and pathway analysis of differently expressed genes from five significant profiles

In order to interpret the biological meaning of the mechanical microenvironment in breast cancer development, we performed the GO and KEGG pathway analyses of these differently expressed genes from five significant profiles in SK-BR-3 cells grown on different stiffness substrates. First, we interpreted the significant molecular function of the differential expression of mRNAs using Gene Ontology (GO). All genes revealed a total of 590 significant GO categories. The top 40 significant GO categories are shown in Fig. 4A. We found that the most prominent terms were the mitotic cell cycle, regulation of transcription, DNA-dependent, cell division, protein phosphorylation, response to DNA damage stimulus, apoptotic process, and cell cycle. Furthermore, from the KEGG pathway analysis it was shown that there were 96 significant signaling pathways in total. Figure 4B lists 40 of the most significant signaling pathways. Among these, cell cycle, ubiquitin mediated proteolysis, pathways in cancer, proteoglycans in cancer, and MAPK signaling pathway were ranked the highest.

3.4 Signal-net construction of the differentially expressed genes from the top five profiles

After we analyzed the functions of the differentially expressed genes in the top five profiles, we further constructed the gene-gene interaction network according to the relationship between gene and gene (Fig. 5). From the Signal-net, we can see that some genes such as PIK3CA, PIK3CB, NRAS and KRAS are in the key position of the map. Genes with degrees greater than 10 are listed in Table 1. We then tested whether the PI3K and Ras family signalling pathways were activated in the different stiffness substrates. As shown in Fig. 6, the Western blot analyses results indicated that the AKT activity downstream of PI3K was enhanced with the increase of modulus. Similarly, the phosphorylated level of the Ras downstream signal ERK was increased on the stiff substrates. Therefore, the results indicated that endogeneous PI3K/AKT and Ras/ERK activity is correlated with changes in the substrate modulus. Furthermore, the signal-net results not only confirmed that the activity of PI3K/AKT and Ras/ERK signalling pathways were induced in the SK-BR-3 cells grown on stiff substrates, but also implied that other genes with high degrees, such as PLCB4, IMPAD1 and PAPSS2, may account for the gene expressions in different mechanical microenvironments.

4 Discussion

Aside from chemical components and biological molecules, previous studies have indicated that the changes of mechanical factors in ECM also play an important role in directing cell behaviors [9 , 14]. Our previous study prepared polydimethylsiloxane (PDMS) of different stiffnesses to mimic the mechanical microenvironments of different organs, and found that breast cancer MCF-7 cells on rigid substrates entered the logarithmic growth phase earlier than cells on soft substrates [6]. Breast tumorigenesis has been reported to be accompanied by collagen crosslinking, which modulated the ECM stiffening to force focal adhesions, growth factor signaling and breast malignancy. Moreover, the development of breast cancer involves a series of intermediate processes, beginning with in situ carcinoma, invasive carcinoma and finally metastatic disease. Due to the fact that the matrix elasticity in various organs is disparate, it is important to investigate the manner by which the changing force which breast cancer cells undergo regulates their behaviors. Therefore, the changing force which cells undergo should to be considered when trying to understand the complex nature of tumorigenesis [4].

Therefore, identifying the genes induced by stiffness can increase our understanding of the mechanism of breast tumorigenesis, and provide an optional therapy for metastatic breast cancer. In the present article, we applied polyacrylamide hydrogel to fabricate different stiffness (0.45 Kpa, 4.04 Kpa and 23.51 Kpa) to mimic different tissues, including brain, lymph node and mammary tissue, mammary tumor and lung tissue, spinal cord and collagenous bone tissue, and verified the effects of stiffness on breast cancer line SK-BR-3 cell viability. Our results indicated that rigid substrates promoted breast cancer SK-BR-3 cell growth more effectively than that on soft substrates (Fig. 1). Then we performed mRNA expression profiling to identify the differently expressed genes on 0.45 Kpa, 4.04 Kpa and 23.51 Kpa substrates using the human genome microarrays. 1831 differently expressed genes were obtained in SK-BR-3 cells grown on three substrates. In order to better understand the role of differently expressed genes in this process, it was crucial to interpret the functions of these genes by bioinformatics analyses. Based on differentially regulated genes in five significant profiles from microarrays results, the GO database was first used to interpret the significant molecular function of these genes induced by stiffness. Then, the KEGG database was also applied to analyze the significant pathways of the differential expression of mRNAs. Among the 40 most significant signaling pathways, cell cycle, ubiquitin mediated proteolysis, pathways in cancer, proteoglycans in cancer, and MAPK signaling pathway were ranked the highest.

Furthermore, the gene-gene interaction network of the above differentially expressed mRNAs was built according to the relationship between gene and gene based on the KEGG database. From the results, we can see that some genes such as PIK3CA, PIK3CB, NRAS and KRAS are in the key position of the map. The genes of PIK3CA and PIK3CB encode the catalytic subunit of the phosphatidylinositol 3-kinases (PI3K) protein, while NRAS and KRAS produce Ras family protein. At present, the PI3K and Ras signal pathways have been found to control gene expression, proliferation and apoptosis [2, 24]. In order to further investigate the role of PI3K and Ras signal pathways in this process, endogeneous AKT and ERK activity, which are downstream of PI3K and Ras, respectively, was detected by Western blotting. As shown in Fig. 6, the activation of both AKT and ERK is up-regulated as the substrate stiffness increases. The enhancement of AKT and ERK activity in SK-BR-3 on the rigid substrate indicated that these signalings are critical for the aggressive breast tumor growth in a stiff microenvironment. This work will broaden our understanding of how extracellular mechanical cues regulate breast cell function at the molecular level. Furthermore, the question of whether or not other key genes with high degrees in the signal-network participate in this process is worthy of further study. For example, one group indicated that PLCβ4 may play an important role in the maintenance of the status epilepticus [12]. Additionally, phospholipase C epsilon 1 (PLCɛ1) has recently been found to play a critical role in inflammation and tumorigenesis to been considered as a novel potential biomarker for gastric cancer [31]. We believe that the primary network should be the foundation for a more advanced and complicated understanding of the role of mechanical factors in breast cancer development.

Presently, tissue engineering and regenerative medicine have been implemented in the clinic. Although the scaffold does not need to completely recapitulate the native tissue structures and functionalities before implantation, it is instructive to consider the mechanical factor, as cells and scaffolds are plastic and mutually responsive. Until present, the important properties of these cell-culture scaffolds can be informed in cardiovascular disease, nerve damage and stem cell regeneration [7 , 28]. Moreover, the less effectiveness of drug therapies in vivo on tumors, may highly reflect the importance of the mechanical microenvironment, which directly regulates drug susceptibility [17]. However, little attention has been paid to the pharmacological issues under different mechanical factors in tumors and other stiffness related diseases. In fact, we introduced the concept of “biomechanopharmacology” into the field of pharmacology at the 4th International Conference on Clinical Hemorheology in early 2002. As a new borderline discipline, it emphasizes equivalently pharmacological effects of biomechanical factors while highlighting the role of drugs by changing or interacting with biomechanical events [11].

In conclusion, our data demonstrate that a significant number of mRNAs can be affected by the stiffness in breast cancer cells. The commonly dysregulated genes identified in this study may represent strong candidates for future therapeutic targets for metastatic breast cancer, and the study of these targets may increase our understanding of the mechanism underlying stiffness related diseases. Our results may also provide the suggestion that the relevant mechanical microenvironment in the cell matrix should be taken into account in tissue engineering, drug delivery in vitro, as well as drug screening system in vitro.

Footnotes

Acknowledgments

This work is supported by the project of 973 in Ministry of Science and Technology of China (2012CB933800), National Natural Science Foundation of China (31470905) and Beijing Natural Science Foundation (7132025).

References

Allen

and Louise

, Jones, Jekyll and Hyde: The role of the microenvironment on the progression of cancer, The Journal of Pathology 223(2) (2011), 162–176.

Awada

, Bozovic-Spasojevic

and Chow

, New therapies in HER2-positive breast cancer: A major step towards a cure of the disease? Cancer Treatment Reviews 38(5) (2012), 494–504.

Braune

, Honow

, Mrowietz

, Cui

, Kratz

, Hellwig

, Uzum

, Klitzing

R.V.

, Lendlein

and Jung

, Hemocompatibility of soft hydrophobic poly(n-butyl acrylate) networks with elastic moduli adapted to the elasticity of human arteries, Clinical Hemorheology and Microcirculation 49(1-4) (2011), 375–390.

Butcher

D.T.

, Alliston

and Weaver

V.M.

, A tense situation: Forcing tumour progression, Nature Reviews Cancer 9(2) (2009), 108–122.

Engler

A.J.

, Sen

, Sweeney

H.L.

and Discher

D.E.

, Matrix elasticity directs stem cell lineage specification, Cell 126(4) (2006), 677–689.

Feng

, Tang

, Xu

, Sun

, Liao

and Han

, Substrate stiffness influences the outcome of antitumor drug screening in vitro, Clinical hemorheology and microcirculation 55(1) (2013), 121–131.

Keung

A.J.

, Asuri

, Kumar

and Schaffer

D.V.

, Soft microenvironments promote the early neurogenic differentiation but not self-renewal of human pluripotent stem cells, Integrative Biology: Quantitative Biosciences from Nano to Macro 4(9) (2012), 1049–1058.

Levental

, Georges

P.C.

and Janmey

P.A.

, Soft biological materials and their impact on cell function, Soft Matter 3(3) (2007), 299–306.

, Han

and Zhao

Y.P.

, Kinetic behaviour of the cells touching substrate: The interfacial stiffness guides cell spreading, Scientific Reports 4 (2014), 3910.

10.

, Dranoff

J.A.

, Chan

E.P.

, Uemura

, Sevigny

and Wells

R.G.

, Transforming growth factor-beta and substrate stiffness regulate portal fibroblast activation in culture, Hepatology (Baltimore, Md) 46(4) (2007), 1246–1256.

11.

Liao

, Li

, Han

, Cao

and Chen

, Biomechanopharmacology: A new borderline discipline, Trends in Pharmacological Sciences 27(6) (2006), 287–289.

12.

Liu

J.X.

, Hu

, Chen

X.L.

, Xu

J.H.

, Yang

P.B.

, Zhang

J.S.

and Liu

, Reduced expression of Phospholipase C beta in hippocampal interneuron during pilocarpine induced status epilepticus in mice, Neurochemistry International 68 (2014), 10–17.

13.

C.M.

, Wang

H.B.

, Dembo

and Wang

Y.L.

, Cell movement is guided by the rigidity of the substrate, Biophysical Journal 79(1) (2000), 144–152.

14.

Pelham Jr.

R.J.

and Wang

, Cell locomotion and focal adhesions are regulated by substrate flexibility, Proceedings of the National Academy of Sciences of the United States of America 94(25) (1997), 13661–13665.

15.

Prantl

, Englbrecht

M.A.

, Schoeneich

, Kuehlmann

, Jung

E.M.

and Kubale

, Semiquantitative measurements of capsular contracture with elastography–first results in correlation to Baker Score, Clinical Hemorheology and Microcirculation 58(4) (2014), 521–528.

16.

Raab

, Shin

J.W.

and Discher

D.E.

, Matrix elasticity in vitro controls muscle stem cell fate in vivo, Stem Cell Research & Therapy 1(5) (2010), 38.

17.

Rehfeldt

, Engler

A.J.

, Eckhardt

, Ahmed

and Discher

D.E.

, Cell responses to the mechanochemical microenvironment–implications for regenerative medicine and drug delivery, Advanced Drug Delivery Reviews 59(13) (2007), 1329–1339.

18.

Rjosk-Dendorfer

, Gurtler

V.M.

, Sommer

W.H.

, Reiser

and Clevert

D.A.

, Value of high resolution compression elastography and color doppler sonography in characterisation of breast lesions: Comparison of different high-frequency transducers, Clinical Hemorheology and Microcirculation 57(2) (2014), 129–135.

19.

Schedin

and Keely

P.J.

, Mammary gland ECM remodeling, stiffness, and mechanosignaling in normal development and tumor progression, Cold Spring Harbor Perspectives in Biology 3(1) (2011), a003228.

20.

Schrader

, Gordon-Walker

T.T.

, Aucott

R.L.

and van Deemter

, Quaas

, Walsh

, Benten

, Forbes

S.J.

, Wells

R.G.

and Iredale

J.P.

, Matrix stiffness modulates proliferation, chemotherapeutic response, and dormancy in hepatocellular carcinoma cells, Hepatology (Baltimore, Md) 53(4) (2011), 1192–1205.

21.

Sharma

, Beck

A.H.

, Webster

J.A.

, Espinosa

, Montgomery

, Varma

and van de Rijn

, Jensen

K.C.

and West

R.B.

, Analysis of stromal signatures in the tumor microenvironment of ductal carcinoma in situ, Breast Cancer Research and Treatment 123(2) (2010), 397–404.

22.

Shieh

A.C.

, Biomechanical forces shape the tumor microenvironment, Annals of Biomedical Engineering 39(5) (2011), 1379–1389.

23.

Teixeira

A.I.

, Ilkhanizadeh

, Wigenius

J.A.

, Duckworth

J.K.

, Inganas

and Hermanson

, The promotion of neuronal maturation on soft substrates, Biomaterials 30(27) (2009), 4567–4572.

24.

Thery

J.C.

, Spano

J.P.

, Azria

, Raymond

and Penault Llorca

, Resistance to human epidermal growth factor receptor type 2-targeted therapies, European Journal of Cancer (Oxford, England: 1990) 50(5) (2014), 892–901.

25.

Tse

J.R.

and Engler

A.J.

, Preparation of hydrogel substrates with tunable mechanical properties, Current protocols in cell biology / editorial board, Juan S Bonifacino [et a.l] 2010, Chapter 10:Unit 10.16.

26.

Tuft

B.W.

, Zhang

, Xu

, Hangartner

, Leigh

, Hansen

M.R.

and Guymon

C.A.

, Material stiffness effects on neurite alignment to photopolymerized micropatterns, Biomacromolecules 15(10) (2014), 3717–3727.

27.

Umesh

, Rape

A.D.

, Ulrich

T.A.

and Kumar

, Microenvironmental stiffness enhances glioma cell proliferation by stimulating epidermal growth factor receptor signaling, PloS One 9(7) (2014), e101771.

28.

Wong

J.Y.

, Velasco

, Rajagopalan

and Pham

, Directed movement of vascular smooth muscle cells on gradient-compliant hydrogels, Langmuir 19(5) (2003), 1908–1913.

29.

Xiao

, Jia

, Mo

, Wang

, Qin

, He

, Zhao

, Huang

, Li

and Yu

, et al., Understanding PRRSV infection in porcine lung based on genome-wide transcriptome response identified by deep sequencing, PloS One 5(6) (2010), e11377.

30.

Xiao

, Mo

, Wang

, Jia

, Qin

, Yu

, Niu

, Zhao

, Liu

and Chen

, Aberrant host immune response induced by highly virulent PRRSV identified by digital gene expression tag profiling, BMC genomics 11 (2010), 544.

31.

Yan

and Fu

, PLCepsilon1: A potential target of RNA interference therapy for gastric cancer, Biochemical and Biophysical Research Communications 448(4) (2014), 409–413.