Accepted Manuscript

Inhibition of miR-32 activity promoted EMT induced by PM2.5 exposure through

the modulation of the Smad1-mediated signaling pathways in lung cancer cells

Dan Yang, Mingyue Ma, Weiqiang Zhou, Biao Yang, Chunling Xiao

PII: S0045-6535(17)30867-6

DOI: 10.1016/j.chemosphere.2017.05.152

Reference: CHEM 19367

To appear in: Chemosphere

Received Date: 16 March 2017

Revised Date: 21 May 2017

Accepted Date: 27 May 2017

Please cite this article as: Dan Yang, Mingyue Ma, Weiqiang Zhou, Biao Yang, Chunling Xiao,
Inhibition of miR-32 activity promoted EMT induced by PM2.5 exposure through the modulation of
the Smad1-mediated signaling pathways in lung cancer cells, Chemosphere (2017), doi: 10.1016/j.
chemosphere.2017.05.152

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1 Inhibition of miR-32 activity promoted EMT induced by PM2.5 exposure

2 through the modulation of the Smad1-mediated signaling pathways in
3 lung cancer cells

4 Dan Yanga, b, 1, Mingyue Maa,c ,Weiqiang Zhoua, Biao Yanga, Chunling Xiaoa, *
5 aKey Laboratory of Environmental Pollution and Microecology of Liaoning
6 Province, Shenyang Medical college, No. 146, North Huanghe Street, Huanggu
7 District, Shenyang City 110034, P.R. China
8 bDepartment of Pharmacology, Shenyang Medical college, No. 146, North
9 Huanghe Street, Huanggu District, Shenyang City 110034, P.R. China
10 cDepartment of Toxicology, School of Public Health, Shenyang Medical
11 College, No. 146, North Huanghe Street, Huanggu District, Shenyang City
12 110034, P.R. China
13 * Corresponding author.
14 Chunling Xiao
15 Phone：+86 24 62215656.
16 E-mail address: E-mail: xiaochunling2000@163.com

1
 ACCEPTED MANUSCRIPT

18 Abstract
19 Epithelial mesenchymal transition (EMT) is a crucial morphological event
20 during tumor progression. The present study reported that EMT could be
21 triggered by airborne fine particulate matter (PM) with a mean diameter of less
22 than 2.5 μm (PM2.5) in human lung cancer cells. We also aimed to elucidate
23 the possible mechanisms of these processes. The results showed that
24 treatment with PM2.5 promoted the activity of the SMAD family member 1
25 (Smad1)-mediated signaling pathway and downregulated the expression of the
26 inhibitory Smad proteins Smad6 and Smad7 in lung cancer cells. Moreover, the
27 knockdown of Smad1 suppressed the EMT process induced by PM2.5
28 exposure. Our data further revealed that miR-32 has a negative effect on
29 PM2.5-induced EMT. The results showed that the expression level of miR-32
30 was significantly upregulated in the PM2.5-induced EMT process. The
31 knockdown of miR-32 enhances the activity of the Smad1-mediated signaling
32 pathway, which promotes the EMT process induced by PM2.5. Thus, these
33 findings indicate that PM2.5 can induce the EMT process through the Smad1-
34 mediated signaling pathway, and miR-32 may act as an EMT inhibitor in lung
35 cancer cells.

36 Capsule: PM2.5 exposure induced the EMT process in lung cancer cells, and
37 the inhibition of miR-32 activity promoted the EMT process through the
38 modulation of the Smad1-mediated signaling pathways.

39 Key words: PM2.5; miR-32; EMT; Smad1; lung cancer

40 Introduction
41 In recent years, a growing number of studies have focused on the impact
42 of ambient air particulate matter (PM) on the human body (Xiao X et al., 2016；
43 Colao A et al., 2016). PM has been classified as human carcinogen by the
44 International Agency for Research on Cancer (Loomis D et al., 2013).
45 Respiratory PM with a mean aerodynamic diameter less than 2.5 μm (PM2.5)
46 is a significant constituent of PM. PM2.5 has impacts the human respiratory
2
 ACCEPTED MANUSCRIPT

47 system, causing malignant pulmonary disorders, such as pulmonary fibrosis,

48 asthma, respiratory inflammation, chronic obstructive pulmonary disease
49 (COPD), and even lung cancer (Hystad P et al., 2013; Krall JR et al., 2016; Gu
50 LZ et al., 2017). A growing number of epidemiological and toxicological studies
51 have strengthened the evidence of the linear relationship between the elevation
52 in PM2.5 exposure concentration and the increase in lung cancer morbidity and
53 mortality (Hystad P et al., 2013; Turner MC et al., 2011; Gao Y et al., 2017;
54 Chen L et al., 2017). Although some biological mechanisms underlying PM2.5-
55 induced damaging effects on the respiratory system have been explored,
56 including oxidative damage of DNA, imbalanced intracellular calcium
57 homeostasis, inflammatory injury, and autophagy, the molecular mechanisms
58 involved in the toxicity of PM2.5 are still not completely understood, and
59 preventive methods are highly desirable (Borgie M et al., 2015; Duarte FV et
60 al., 2012; Oh SM et al., 2011; Su R et al., 2017; Deng X et al., 2017; Yan XD et
61 al., 2017).
62 Epithelial mesenchymal transition (EMT) is an evolutionarily conserved
63 and genetically controlled process during which epithelial cells acquire
64 mesenchymal features and increased motility (Tam WL et al., 2013). It is a
65 complex, multi-step program. The characteristics of EMT include reduced
66 intercellular adhesion, acquisition of mesenchymal markers (including vimentin
67 and N-cadherin) and loss of epithelial markers (such as E-cadherin) (Lamouille
68 S., 2014). EMT is an early and critical step in the metastasis of many epithelial
69 tumors, and emerging evidence has demonstrated that EMT plays an important
70 role in cell adhesion, tumor migration, invasion and metastasis (Yue X et al.,
71 2016; Sun K et al., 2015). Considering the close relationship between human
72 exposure to PM2.5 and elevated lung cancer mortality，the high risks of PM2.5
73 promoting tumor metastasis, including the initiation and progression of EMT in
74 lung cancer, should be studied. To our knowledge, there are only two studies
75 on the relationship between EMT and PM exposure, but the mechanisms of the
76 initiation and progression of PM2.5-induced EMT have not yet been elucidated
3
 ACCEPTED MANUSCRIPT

77 (Barker TH et al., 2014; Yue H et al., 2015).

78 Several oncogenic pathways, including Notch, Wnt/β-catenin, transforming
79 growth factor β (TGF-β)/Smad, Smad1-mediated, and PI3K/GSK-3β/β-catenin
80 signaling pathways have been reported to induce EMT (Shan S et al., 2015;
81 Shen M et al., 2016; Zhu Y et al., 2016; Qi F et al., 2015; Richter A et al., 2014).
82 Additionally, a number of miRNAs also play important roles in the EMT process
83 (Yue X et al., 2016).
84 MicroRNAs (miRNAs) are an abundant class of endogenously expressed
85 small non-coding RNAs that are 19 to 24 nucleotides in length (Maitri Y. Shah
86 et al., 2016). They are known to regulate gene products at the post-
87 transcriptional level by binding to multiple target mRNAs (Tang J et al., 2016;
88 Lin CW et al., 2014). The dysregulation of miRNAs is involved in various
89 pathological processes, including EMT and cancer metastasis (Li Q et al., 2016;
90 Brivio S et al., 2015; Chandra Mangalhara K et al., 2017). Hence, the altered
91 expression of miRNAs and their target genes might represent a vital
92 mechanism attributable to the cytotoxicity of PM2.5 exposure. Several recent
93 studies have reported that changes of miRNA levels are involved in the
94 response to PM2.5 exposure in vitro or in vivo (Fossati S et al., 2014; Liu CW
95 et al., 2014). Li et al reported that miR-802 was downregulated, while miR-4516
96 was upregulated in A549 cells and serum of individuals from moderately
97 polluted cities (Li X et al., 2016; Li X et al., 2016). Other studies have focused
98 on the association between miRNA expression and PM2.5 exposure, and the
99 results showed that miR-21,182, 185, and 222 were induced by PM2.5
100 exposure (Bollati V et al., 2010; Liu C et al., 2015). However, until now, the
101 important roles of miRNAs as regulators or biomarkers for PM2.5 exposure
102 have remained largely unknown.
103 In the present work, we have investigated the effects of PM2.5 exposure
104 on the induction of EMT, the regulatory effects of the Smad1-mediated signaling
105 pathways, and the regulatory roles of miR-32 in the process of EMT induced by
106 PM2.5 in lung cancer cells. The results provide novel insights into the
4
 ACCEPTED MANUSCRIPT

107 mechanism of increasing morbidity and mortality of lung cancer caused by

108 PM2.5 exposure.
109 2. Materials and Methods
110 2.1. Reagents and antibodies
111 Anti-Smad6, anti-Smad7, anti-E-cadherin, anti-vimentin and anti-actin
112 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA,
113 USA). Anti-Smad 1 and anti-phospho-Smad1 antibodies were purchased from
114 Cell Signaling Technology (Danvers, MA, USA). Lipofectamine 2000, TRIzol,
115 and primers for Smad1, Smad6, Smad7, and Glyceraldehyde 3-phosphate
116 dehydrogenase (GAPDH) were obtained from Invitrogen (CA, USA). miR-
117 shNC, miR-32 sponge inhibitor, pGPH1/GFP/Neo-shNC, and
118 pGPH1/GFP/Neo-Smad1 were purchased from GenePharma Technology
119 (Shanghai, China). Promega GoScript, Promega GoTaq® qPCR Master Mix
120 and GoTaq®Green Master Mix were purchased from the Promega Corporation
121 (Madison, Wisconsin, USA). MiR-32 miScript Primer Assays and U6 miScript
122 Primer Assays were purchased from RIBOBIO CO., LTD (Guangzhou, China),
123 and the miScript Ⅱ RT Kit and miScript SYBR® Green PCR Kit were purchased
124 from QIAGEN (Hilden, Germany).
125 2.2. Cell culture
126 A549 cells and H292 cells were cultured in RPMI 1640 medium (GIBCO,
127 Grand Island, NY, USA) supplemented with 10% (v/v) heat-inactivated fetal
128 bovine serum (FBS), penicillin (100 U/ml), and streptomycin (100 mg/ml) at
129 37°C in an atmosphere of 5% CO2 and 95% air. The cells were routinely sub-
130 cultured every 2-3 days, and all of the cell samples were verified to be in the
131 logarithmic growth phase.
132 2.3. Sampling and preparation
133 The PM2.5 sample collection and preparation were conducted according
134 to our previously reported methods (Ma, M. et al., 2015). Briefly, samples
135 were collected on nitrocellulose filters using a high-volume sampler particle
136 collector in the center of Shenyang City in winter. PM2.5 was extracted from
5
 ACCEPTED MANUSCRIPT

137 the nitrocellulose filters by immersing in deionized water and then sonicated
138 for 20 minutes (Ma, M. et al., 2015). The extracted samples were then stored
139 at -80°C until they were exposed to A549 cells. The collected PM2.5 samples
140 were chemically analyzed for their composition. The main chemical
141 compositions of PM2.5 are toxic heavy metals (Mn, Zr, Cu, etc.) and
142 polycyclic aromatic hydrocarbon substances (anthracene, chrysene,
143 benzo(a)pyrene, etc.) (Ma, M. et al., 2015).
144 2.4. RNA extraction and quantitative real-time polymerase chain reaction
145 (qRT-PCR)

146 Cells seeded in 10-cm culture dishes were treated with PM2.5 (0, 5, 10, or

147 20 μg/cm2) or TGF-β1 (10 ng/ml) for 72 hours. Total RNA was extracted from
148 cells using TRIzol reagent, according to the manufacturer’s instructions, and
149 cDNA synthesis was performed using Promega reverse transcription reagent
150 or miScript Ⅱ RT Kit.
151 qRT-PCR reactions were performed in an ABI PRISM 7500 system. The
152 amplification of specific PCR products was detected using SYBR Green PCR
153 Master Mix (Applied Biosystems) or miScript SYBR® Green PCR Kit.
154 The primer sequences used were as follows: human GAPDH – forward 5’-
155 GGAGCCAAAAGGGTCATCAT-3’ and reverse 5’-
156 GTGATGGCATGGACTGTGGT-3’;
157 human Smad1 – forward 5’-GCTCAACAATCGTGTGGGTG-3’ and reverse 5’-
158 CCAATATGCCGCCTGGTGTT-3’;
159 human Smad6 – forward 5’-CCTACTCTCGGCTGTCTCCT-3’ and reverse 5’-
160 GGAGCAGTGATGAGGGAGTTG-3’;
161 human Smad7 – forward 5’-GTGGCCGAACCAGAACCAA-3’ and reverse 5-
162 AAAGACGCTACAATGGCAGG-3.
163 For Smad1, Smad6, and Smad7, the data were normalized to the
164 housekeeping gene GAPDH. U6 primers were used for the normalization of
165 miR-32. Data were analyzed using the comparative threshold cycle (CT)

6
 ACCEPTED MANUSCRIPT

166 method, and each sample was analyzed in duplicate.

167 2.5. Western blotting analysis
168 Western blotting analysis was conducted using sodium dodecyl sulfate
169 polyacrylamide gel electrophoresis (SDS-PAGE). The procedures and
170 quantification methods were the same as those described previously (Yang,
171 D., et al., 2015). Briefly, the protein concentration was determined using the
172 Lowry method. Total proteins were electrophoresed in SDS-PAGE system and
173 were electrophoretically transferred to polyvinylidene fluoride (PVDF)
174 membranes. Membranes were incubated with the primary antibodies [anti-
175 Smad1, anti-phospho-Smad1, anti- E-cadherin, anti-vimentin, anti-Smad6, anti-
176 Smad7, anti-actin]. Immune complexes were finally detected with enhanced
177 chemiluminescence (ECL) reagent (SuperSignal Western Pico
178 Chemiluminescent Substrate, Pierce, USA).
179 2.6. Transfection
180 Cells were seeded into 24-well plates (5×104 cells/well) without the addition
181 of antibiotics. After 24 h, the cells were transfected with miR-shNC, miR-32
182 sponge inhibitor, pGPH1/GFP/Neo-shNC, or pGPH1/GFP/Neo-Smad1 using
183 Lipofectamine 2000 reagent according to the manufacturer’s instructions. After
184 transfection of the cells with shRNA for 48 h, the cells were selected in medium
185 containing kanamycin (10 μg/cm2) or spectinomycin (20 μg/cm2), and a stable
186 cell pool was obtained for further study.
187 2.7. Immunofluorescence

188 After treatment with PM2.5 or TGF-β1 for 72 hours, the cells were washed

189 and permeabilized before being blocked with 5% bovine serum albumin (BSA)
190 in phosphate-buffered saline (PBS) for 1 h. Next, the cells were incubated with
191 anti-E-cadherin or anti-vimentin antibodies at a dilution of 1:100 overnight at
192 4°C, followed by washes and the green fluorescent protein (GFP)- or
193 rhodamine-conjugated secondary antibodies at a 1:50 dilution (CWBIO, Beijing,
194 China). Finally, the cells were washed 3 times and incubated with 1 μg/ml 4',6-

7
 ACCEPTED MANUSCRIPT

195 diamidino-2-phenylindole (DAPI). All images were obtained from an

196 immunofluorescence microscope (Leica CTR 4000, Germany).
197 2.8 Wound healing assay
198 Cells were seeded into 6-well plates after treatment with PM2.5 for 72 h at
199 a concentration of 1×106/ml. After 24 h of growth, the cells reached
200 approximately 80% confluence. The cells were gently and slowly scratched
201 across the center of the well with a 200 μl pipette. After scratching, the detached
202 cells were removed by washing the well gently with PBS. The recording sites
203 of each well was signed, and photos were taken on a microscope. The well was
204 replenished with serum-free medium for 48 h; then, photos were taken at the
205 signed sites. The gap distance was quantitatively evaluated using ImageJ
206 software (National Institutes of Health, Bethesda, MD USA). The experiments
207 were performed in triplicate.
208 2.9. Transwell invasion assay
209 Transwell invasion assay was performed using Transwell chambers
210 (Corning Life Sciences, Corning, NY, USA) with 8-μm pores. The upper
211 chamber of the Transwell inserts was coated with 50 μl of 5 mg/ml Matrigel.
212 The cells were treated with PM2.5 (0, 5, 10, or 20 μg/cm2) for 72 h before
213 starting the experiment. Then, 2×104 cells with 200 μl serum-free medium were
214 seeded into the top chamber of the Transwell insert. The lower chamber was
215 loaded with 600 μl medium containing 10% FBS. Following a 24 h incubation
216 period, the non-invading cells on the upper surface of the membrane were
217 gently removed with a cotton swab. The cells that invaded to the lower surface
218 of the membrane were stained with 0.1% crystal violet and were imaged and
219 counted in 6 randomly selected fields. The experiments were performed in
220 triplicate.
221 2.10. Statistical analysis
222 Data are presented as the mean ± the standard deviation (SD) from 3
223 independent experiments unless otherwise indicated. Differences between the
224 groups were evaluated for significance by a Mann-Whitney U test using
8
 ACCEPTED MANUSCRIPT

225 GraphPad Prism 5 software. P-values of < 0.05 were considered statistically
226 significant.
227 3. Results
228 3.1. PM2.5 induced EMT in human lung cancer cells
229 A549 cells and H292 cells were seeded in 6-well plates and pretreated with
230 RPMI 1640 without FBS and antibiotics for 24 h before the addition of PM2.5
231 or TGF-β1for 72 h. Cells treated with PM2.5 or TGF-β1 presented obvious
232 morphological changes, including a transition from cobblestone phenotype
233 (epithelial phenotype) to a spindle phenotype (mesenchymal phenotype)
234 (Figure 1A). The expression of protein markers associated with EMT was also
235 altered, which accompanied the observed morphological changes. The western
236 blotting results showed that the expression level of the epithelial marker E-
237 cadherin was significantly decreased, while the expression level of the
238 mesenchymal marker vimentin was markedly increased (Figure 1B). Moreover,
239 the immunofluorescence assays in A549 cells and H292 cells confirmed that
240 exposure to PM2.5 led to a significant decrease in E-cadherin protein
241 expression and an increase in vimentin protein expression (Figure 1C).
242 Taken together, our data showed that exposure to certain concentrations
243 of PM2.5 could lead to EMT in lung cancer cells.
244 Fig. 1. EMT induced by PM2.5 exposure in human lung cancer cells.
245 (A) The morphological changes induced by exposure of A549 cells and H292
246 cells to PM2.5. TGF-β1 was used as a positive control. The cell shape
247 changed from cobblestone-like to a long spindle-shaped phenotype after
248 A549 cells and H292 cells were treated with PM2.5 or TGF-β1. Scale bars
249 in (A) = 50 μm.
250 (B) The expression levels of E-cadherin and vimentin proteins were analyzed
251 by western blotting. The final results are summarized in the bar graphs. The
252 data are presented as the means ± SD of 3 independent experiments. Actin
253 was used as the loading control.
254 * P < 0.05 compared with the cells in the PM2.5 untreated group.
9
 ACCEPTED MANUSCRIPT

255 (C) The expression levels of the E-cadherin (green) and vimentin (red) proteins
256 were analyzed by immunofluorescence assays. Scale bars in (C) = 50 μm.
257 3.2. Induction of EMT by PM2.5 was correlated with up-regulated activity
258 of the Smad1-mediated signaling pathway in A549 cells
259 To investigate the mechanisms of EMT induced by PM2.5, the activity of the
260 Smad1-mediated signaling pathway, which is one of the most important
261 regulators of EMT, was evaluated. Thus, we conducted quantitative RT-PCR
262 and western blotting assays to examine the expression of Smad1 in A549 cells
263 after exposure to PM2.5. The results indicated that the treatment of A549 cells
264 with PM2.5 led to a dose-dependent upregulation of the levels of Smad1 mRNA
265 as well as the levels of p-Smad1 and Smad1 proteins (Fig. 2A, 2B). We also
266 determined that the mRNA and protein expression levels of the inhibitory Smad
267 proteins (Smad6 and Smad7), which can inhibit the expression of Smad1, were
268 downregulated as a result of PM2.5 exposure (Fig. 2A, 2B).
269 To prove that Smad1 was involved in PM2.5-induced EMT in the lung
270 cancer cell line A549, we used Smad1-specific shRNA to inhibit the expression
271 of Smad1, and then analyzed E-cadherin and vimentin expression. As shown
272 in Figure 2C, the expression levels of E-cadherin and vimentin were not
273 significantly changed in the shRNA-Smad1 group compared with the shRNA-
274 NC (negative control) group. The interference experiment indicated that the
275 knockdown of Smad1 could effectively suppress the induction of EMT by PM2.5
276 exposure in A549 cells. Notably, Smad1 depletion inhibited the ability of A549
277 cells to migrate as detected by wound healing assay. As shown in Figure 2D,
278 the wound healing rates which indicated the migration ability in the shRNA-
279 Smad1 group were lower compared with the shRNA-NC group regardless of
280 PM2.5 treatment.
281 All these findings suggested that PM2.5 could induce EMT via the
282 upregulation of the activity of the Smad1-mediated pathway, which might be
283 partially due to the downregulation of the expression of Smad6 and Smad7 in
284 A549 cells.
10
 ACCEPTED MANUSCRIPT

285 Fig. 2. Induction of EMT by PM2.5 was correlated with upregulated Smad1
286 activity in A549 cell lines.
287 (A) Quantitative RT-PCR analysis of the mRNA expression of Smad1, Smad6,
288 and Smad7. The final results are summarized in the bar graphs. The data
289 are presented as the means ± SD of 3 independent experiments. The
290 expression of every target gene was quantified using GAPDH as a
291 normalization control. * P < 0.05 compared with the cells in the PM2.5
292 untreated group.
293 (B) Western blot analysis of the expression levels of p-Smad1, Smad1, Smad6,
294 and Smad7 proteins.
295 (C) Western blot analysis of the expression levels of p-Smad1, Smad1, E-
296 cadherin, and vimentin proteins. Smad1 was inhibited by shRNAs.
297 (D) A lower wound healing rate was found in A549 cells treated with shRNA-
298 Smad1 by wound healing assay.
299 (B)(C)(D) The final results are summarized in the bar graphs. The data are
300 presented as the means ± SD of three independent experiments. Actin was
301 used as the loading control. * means P < 0.05.
302 3.3. miR-32 exerted an inhibitory effect on PM2.5-induced EMT
303 MiR-32 has recently gained widespread attention for its roles in the
304 regulation of many critical processes in various types of human cancer,

305 including non-small cell lung cancer (NSCLC) (Bai Y et al., 2015). To determine

306 the expression levels of miR-32 during EMT induced by PM2.5, we performed
307 a quantitative RT-PCR analysis of miR-32 expression in A549 cells treated with
308 increasing doses of PM2.5. The results showed that the treatment of A549 cells
309 with PM2.5 led to a dose-dependent upregulation of miR-32 (Fig. 3A, 3B). To
310 determine the effect of miR-32 on EMT induced by PM2.5, miR-32 sponge
311 inhibitor and miR-shNC were transfected into A549 cells. The results showed
312 that the knockdown of miR-32 resulted in the acquisition of a spindle-shaped
313 cell morphology, and these changes were enhanced by exposure to PM2.5

11
 ACCEPTED MANUSCRIPT

314 (Fig. 3C). Western blotting indicated that the expression of E-cadherin in the
315 miR-32 sponge inhibitor group was lower than that in the miR-shNC group,
316 either with or without exposure to PM2.5 (Fig. 3D). The expression of vimentin
317 in the group treated with the miR-32 sponge inhibitor was higher than that in
318 the group treated with miR-shNC either with or without exposure to PM2.5 (Fig.
319 3D). The miR-32 sponge inhibitor enhanced the abilities of A549 cells to migrate

320 and invade detected by wound healing assay and transwell assay (Fig. 3E，

321 3F). The wound healing rates and the number of invaded cells in the miR-32
322 sponge inhibitor group were higher than those in the miR-shNC group
323 regardless of PM2.5 treatment.
324 The results indicated that the knockdown of miR-32 could promote EMT
325 induced by PM2.5.
326 Fig. 3. miR-32 exerted an inhibitory effect on EMT induced by PM2.5
327 (A) The relative expression of miR-32 was higher in A549 cells exposed to
328 PM2.5 compared with the PM2.5 untreated group.
329 qRT-PCR analysis of miR-32 expression, which was normalized to U6 RNA.
330 The results are expressed as the mean ± SD of 3 replicate experiments. *
331 P < 0.05 compared with the cells in the PM2.5 untreated group.
332 (B) (C) The morphological changes in the miR-32 sponge inhibitor group and in
333 the miR-shNC group whether they were exposed to PM2.5 or not. The cell
334 shape changed from cobblestone-like to long, spindle-shaped cells after
335 the A549 cells were treated with the miR-32 sponge inhibitor. Scale bars
336 in (B) (C) = 50 μm.
337 (D) Western blotting analysis of the expression of E-cadherin and vimentin
338 proteins. The final results are summarized in the bar graphs. The data are
339 presented as the means ± SD of 3 independent experiments. Actin was
340 used as the loading control.
341 (E) A higher wound healing rate was found in A549 cells treated with miR-32
342 sponge inhibitor by wound healing assay.

12
 ACCEPTED MANUSCRIPT

343 (F) A higher number of invaded cells was found in miR-32 sponge inhibitor
344 group by transwell assay.
345 (B)(C) (D) (E) (F) * means P<0.05.
346 3.4. Knockdown miR-32 enhanced the activity of the Smad1-mediated
347 pathway
348 We demonstrated that PM2.5-induced EMT was correlated with
349 upregulated activity of the Smad1-mediated pathways, which occurred partially
350 via the downregulation of the expression of Smad6 and Smad7 in A549 cells.
351 Moreover, miR-32 negatively regulated EMT induced by PM2.5. Our next
352 objective was to evaluate the role of miR-32 in the regulation of the Smad1-
353 mediated pathways. The A549 cells treated respectively with miR-shNC and
354 miR-32 sponge inhibitor were then exposed to PM2.5 for 72 h. The expression
355 levels of p-Smad1, Smad1, Smad6, and Smad7 were determined by
356 quantitative RT-PCR and western blotting assays. The results showed that
357 knockdown of miR-32 significantly increased the mRNA and protein levels of p-
358 Smad1 and Smad1, but decreased the mRNA and protein levels of Smad6 and
359 Smad7 compared with those of the miR-shNC group either with or without
360 exposure to PM2.5 (Fig. 4A, 4B). These data indicated that knockdown of miR-
361 32 could significantly enhance the activity of the Smad1-mediated signaling
362 pathway.
363 Fig. 4. Knockdown of miR-32 enhanced the activity of the Smad1 mediated
364 pathways
365 (A) Quantitative RT-PCR analysis of the mRNA expressions of Smad1, Smad6,
366 and Smad7. The final results are summarized in the bar graphs. Data are
367 presented as the means ± SD of 3 independent experiments. The
368 expression of every target gene was quantified using GAPDH as a
369 normalization control.
370 (B) Western blotting analysis of the protein expressions of p-Smad1, Smad1,
371 Smad6, and Smad7. The final results are summarized in the bar graphs.
372 Data are presented as the means ± SD of 3 independent experiments.
13
 ACCEPTED MANUSCRIPT

373 Actin was used as a loading control.

374 (A) (B) *, means P<0.05.
375 4. Discussion
376 The activation of EMT in cancer involves a variety of cancer-related
377 features, including increased stem cell properties, invasion, metastasis, and
378 drug resistance (Siebzehnrubl FA et al., 2013). Although several studies have
379 focused on the impact of PM2.5 on lung cancer, few studies have evaluated the
380 EMT process of lung cancer cells induced by PM2.5 exposure (Eckel SP et al.,
381 2016; Yue H et al., 2015; Li X et al., 2016). In the present study, we provide a
382 novel mechanism of EMT induced by PM2.5 exposure in human lung cancer
383 cells. Our data showed that PM2.5 exposure can induce a morphological
384 change (from an epithelial phenotype to a mesenchymal phenotype) in A549
385 cells. According to western blot results, A549 cells acquired mesenchymal
386 features (vimentin) and lost the expression of the epithelial marker (E-cadherin)
387 after exposure to PM2.5. The expression levels of vimentin and E-cadherin
388 were also evaluated by the immunofluoresence test. These results indicated
389 that A549 cells had undergone an EMT process after exposure to PM2.5. These
390 findings were consistent with the reports of other researchers (Yue H et al.,
391 2015; Barker T H et al., 2014). Moreover, we also certified that the mechanisms
392 of this EMT process induced by PM2.5 exposure were related to Smad1-
393 mediated signaling pathways and miR-32. This represents a novel mechanism
394 of toxicity of PM2.5 on the body, which may be associated with metastasis and
395 drug resistance in human lung cancer.
396 Smad1 is a member of receptor-regulated Smads (R-Smads) which is one
397 subgroup of the Smad family (Wang Y et al., 2016). Smad1 is a key mediator
398 present at the crossroads of multiple EMT-related cell signaling pathways, such
399 as the bone morphogenetic protein (BMP), TGF-β, and Wnt/β-catenin signaling
400 pathways (Yong Wang et al., 2016; Pieniazek M et al., 2012; Massague J
401 et al., 2005; Lagna G et al., 2007; Virtanen S et al., 2011; Richter A et al., 2014
402 ). For example, Smad1 can be activated by BMP signaling (Ahn BN et al., 2016).
14
 ACCEPTED MANUSCRIPT

403 BMP ligands bind to BMP receptors on the cell surface, which leads to the
404 phosphorylation of intracellular Smad1. Then, phosphorylated R-Smad1 forms
405 a complex with Smad4, which then translocates to the nucleus, where these
406 proteins function as transcriptional co-modulators that alter the transcription of
407 specific target genes (McCormack N et al., 2013). Smad1 also acts as an
408 upstream factor, regulating EMT-associated transcription factors, such as
409 Snail, Slug and Twist1 expression (Yong Wang et al., 2016; Anne Richter et al.,
410 2016; Liu CW et al., 2014). All these studies indicate that Smad1 participates
411 in the EMT process. Several studies have demonstrated that knockdown of
412 Smad1 expression suppresses EMT, and the upregulation of Smad1 can
413 promote the EMT phenotype (Yong Wang et al., 2016; Xu T et al., 2011; Jiang
414 B et al., 2016). The regulation of the Smad1-mediated signaling pathway
415 includes a negative feedback control mechanism, which occurs via the direct
416 upregulation of the inhibitory Smad proteins (Smad6 and Smad7) (von Bubnoff
417 A et al., 2001; Xu P et al., 2012; Yu J et al., 2016). In this study, our data showed
418 that exposure to PM2.5 leads to a dose-dependent upregulation of p-Smad1
419 and Smad1 expression and downregulation of Smad6 and Smad7 expression
420 in A549 cells. The results showed that PM2.5 exposure might overwhelm the
421 negative feedback system by the downregulation of Smad6 and Smad7
422 expression, which leads to the activation of the Smad1-mediated signaling
423 pathway. To determine whether Smad1-mediated signaling is required for the
424 induction of EMT by PM2.5, shRNA-Smad1 was used to knock down Smad1
425 expression. The results showed that the shRNA-Smad1 could successfully
426 knock down Smad1 and block the process of EMT induced by PM2.5. These
427 results also proved that Smad1-mediated signaling pathways played an
428 important role in the mechanism by which PM2.5 promoted the EMT process.
429 Further, the migration ability in the shRNA-Smad1 group was lower compared
430 with the shRNA-NC group whatever regardless of PM2.5 treatment.
431 MiRNAs modulate cellular pathways involved in cell growth, apoptosis, and
432 migration mainly through control stability and translation of mRNAs (Gozuacik
15
 ACCEPTED MANUSCRIPT

433 D et al., 2017). Abnormalities of miRNAs often correlate with cancer formation
434 and progression. Studies have shown that the ribosome is a general site for the
435 miRNA-mediated mRNA degradation pathway (Antic S et al., 2015). The
436 pathogenic mechanisms of lung damage caused by PM2.5 exposure are
437 complex, and recent data has demonstrated that molecular pathways involving
438 miRNAs and ribosomal proteins play a key role in the response to PM 2.5
439 exposure. Some target genes of microRNAs are involved in ribosome
440 biogenesis or function and could be deregulated by exposure to PM2.5.
441 Perturbation of ribosome biogenesis is able to activate nucleolus stress
442 signaling pathways involving p53 activity, and the expression of p53 has been
443 found to be dramatically increased after exposure to PM2.5 (Russo A and
444 Russo G., 2017; Li X et al., 2016; Zhao H et al., 2017).

445 In NSCLC, miRNAs were found to influence EMT via the targeting of genes
446 that play central roles in EMT (Fischer KR et al., 2015; Chen QY et al., 2016; Li Y
447 et al., 2016). MiR-32 has recently been shown to inhibit EMT induced by

448 targeting TWIST1 in NSCLC cells (Li L et al., 2016).

449 Additionally, miR-32 may be a potential therapeutic target or molecular
450 biomarker for the prediction of the prognosis of NSCLC patients (Bai Y et al.,
451 2015; Zhan B et al., 2016), but there is still no consensus on the mechanisms
452 of miR-32 involved in carcinogenesis and the progression of NSCLC or other
453 types of carcinoma. For example, Ma et al demonstrated that miR-32 functioned
454 as a tumor suppressor, and the upregulation of miR-32 expression could inhibit
455 NSCLC (Zhu D et al., 2015; Ma ZL et al., 2015). Different functions of miR-32
456 were also reported in other types of cancers. It was shown to function in
457 migration by targeting phosphatase and tensin homolog (PTEN) and as a tumor
458 suppressor by directly targeting enhancer of zeste homolog 2 (EZH2) in
459 gastric carcinoma, breast cancer, hepatocellular carcinoma, or oral squamous
460 cell carcinoma (Yan C et al., 2015; Xia H et al., 2015; Yan SY et al., 2015;
461 Zhang D et al., 2014; Xia W et al., 2017; Zhao L et al., 2017). In this paper, we
462 confirmed that the expression levels of miR-32 were significantly higher in the
463 PM2.5-treated group compared with the PM2.5-untreated group by qRT-PCR.
16
 ACCEPTED MANUSCRIPT

464 The higher levels of miR-32 might function as a tumor suppressor to weaken or
465 suppress EMT induced by PM2.5 exposure. To further evaluate the roles of
466 miR-32 during EMT in A549 cells, miR-32 sponge inhibitor was used to knock
467 down miR-32. The results showed that the knockdown of miR-32 could
468 augment the migration ability and EMT process induced by PM2.5 exposure in
469 A549 cells. These findings were consisted with those presented in Li’s report
470 (Li L et al., 2016). The results also showed that the knockdown of miR-32 could
471 augment the downregulated expression of inhibitor Smad proteins (Smad6 and
472 Smad7) but enhance the activity of Smad1 induced by PM2.5 exposure in A549
473 cells. All of the results indicated that inhibition of miR-32 activity could promote
474 EMT induced by PM2.5 exposure through the Smad1-mediated signaling
475 pathways.
476 5. Conclusions
477 The present study showed that PM2.5 exposure could induce EMT and
478 upregulate the activity of the Smad1-mediated signaling pathway in human lung
479 cancer cells. Furthermore, we found that knockdown of miR-32 could promote
480 PM2.5-induced EMT through the upregulation of Smad1-mediated signaling
481 pathway activity. These findings provide a novel mechanism of PM2.5-induced
482 cytotoxicity and suggest that patients with lung cancer should avoid PM2.5
483 exposure, which can aggravate disease progression and increase the mortality
484 rate.
485 Acknowledgments: This study was supported by the Chinese National
486 Foundation of National Sciences grants (No.30471425) and the Science and
487 Technology Plan Project of Shenyang City (No.F14-181-1-00).
488 Conflict of interest: The authors declare no conflicts of interest.
489 References
490 Antic, S., Wolfinger, M.T., Skucha, A., Hosiner, S., Dorner, S., 2015. General and MicroRNA-Mediated
491 mRNA Degradation Occurs on Ribosome Complexes in Drosophila Cells. Molecular and cellular
492 biology 35, 2309-2320.
493 Bai, Y., Wang, Y.L., Yao, W.J., Guo, L., Xi, H.F., Li, S.Y., Zhao, B.S., 2015. Expression of miR-32 in human non-
494 small cell lung cancer and its correlation with tumor progression and patient survival. International

17
 ACCEPTED MANUSCRIPT

495 journal of clinical and experimental pathology 8, 824-829.

496 Barker, T.H., Dysart, M.M., Brown, A.C., Douglas, A.M., Fiore, V.F., Russell, A.G., Committee, H.E.I.H.R., 2014.
497 Synergistic effects of particulate matter and substrate stiffness on epithelial-to-mesenchymal
498 transition. Research report, 3-41.
499 Boeri, M., Sestini, S., Fortunato, O., Verri, C., Suatoni, P., Pastorino, U., Sozzi, G., 2015. Recent advances of
500 microRNA-based molecular diagnostics to reduce false-positive lung cancer imaging. Expert review
501 of molecular diagnostics 15, 801-813.
502 Chandra Mangalhara, K., Manvati, S., Saini, S.K., Ponnusamy, K., Agarwal, G., Abraham, S.K., Bamezai, R.N.,
503 2017. ERK2-ZEB1-miR-101-1 axis contributes to epithelial-mesenchymal transition and cell
504 migration in cancer. Cancer letters 391, 59-73.
505 Chen, L., Shi, M., Gao, S., Li, S., Mao, J., Zhang, H., Sun, Y., Bai, Z., Wang, Z., 2017. Assessment of
506 population exposure to PM2.5 for mortality in China and its public health benefit based on BenMAP.
507 Environmental pollution 221, 311-317.
508 Deng, X., Feng, N., Zheng, M., Ye, X., Lin, H., Yu, X., Gan, Z., Fang, Z., Zhang, H., Gao, M., Zheng, Z.J., Yu, H.,
509 Ding, W., Qian, B., 2017. PM2.5 exposure-induced autophagy is mediated by lncRNA loc146880
510 which also promotes the migration and invasion of lung cancer cells. Biochimica et biophysica acta
511 1861, 112-125.
512 Farraj, A.K., Hazari, M.S., Haykal-Coates, N., Lamb, C., Winsett, D.W., Ge, Y., Ledbetter, A.D., Carll, A.P.,
513 Bruno, M., Ghio, A., Costa, D.L., 2011. ST depression, arrhythmia, vagal dominance, and reduced
514 cardiac micro-RNA in particulate-exposed rats. American journal of respiratory cell and molecular
515 biology 44, 185-196.
516 Guo, Y., Zeng, H., Zheng, R., Li, S., Pereira, G., Liu, Q., Chen, W., Huxley, R., 2017. The burden of lung
517 cancer mortality attributable to fine particles in China. The Science of the total environment 579,
518 1460-1466.
519 Gu, L.Z., Sun, H., Chen, J.H., 2017. Histone deacetylases 3 deletion restrains PM2.5-induced mice lung
520 injury by regulating NF-kappaB and TGF-beta/Smad2/3 signaling pathways. Biomedicine &
521 pharmacotherapy = Biomedecine & pharmacotherapie 85, 756-762.
522 Heldin, C.H., Miyazono, K., ten Dijke, P., 1997. TGF-beta signalling from cell membrane to nucleus
523 through SMAD proteins. Nature 390, 465-471.
524 Hystad, P., Demers, P.A., Johnson, K.C., Carpiano, R.M., Brauer, M., 2013. Long-term residential exposure
525 to air pollution and lung cancer risk. Epidemiology 24, 762-772.
526 Itoh, F., Asao, H., Sugamura, K., Heldin, C.H., ten Dijke, P., Itoh, S., 2001. Promoting bone morphogenetic
527 protein signaling through negative regulation of inhibitory Smads. The EMBO journal 20, 4132-
528 4142.
529 Jiang, B., Zhang, T., Liu, F., Sun, Z., Shi, H., Hua, D., Yang, C., 2016. The co-stimulatory molecule B7-H3
530 promotes the epithelial-mesenchymal transition in colorectal cancer. Oncotarget 7, 31755-31771.
531 Krall, J.R., Mulholland, J.A., Russell, A.G., Balachandran, S., Winquist, A., Tolbert, P.E., Waller, L.A., Sarnat,
532 S.E., 2017. Associations between Source-Specific Fine Particulate Matter and Emergency Department
533 Visits for Respiratory Disease in Four U.S. Cities. Environmental health perspectives 125, 97-103.
534 Lagna, G., Ku, M.M., Nguyen, P.H., Neuman, N.A., Davis, B.N., Hata, A., 2007. Control of phenotypic
535 plasticity of smooth muscle cells by bone morphogenetic protein signaling through the myocardin-
536 related transcription factors. The Journal of biological chemistry 282, 37244-37255.
537 Li, L., Wu, D., 2016. miR-32 inhibits proliferation, epithelial-mesenchymal transition, and metastasis by
538 targeting TWIST1 in non-small-cell lung cancer cells. OncoTargets and therapy 9, 1489-1498.

18
 ACCEPTED MANUSCRIPT

539 Liu, C.W., Li, C.H., Peng, Y.J., Cheng, Y.W., Chen, H.W., Liao, P.L., Kang, J.J., Yeng, M.H., 2014. Snail
540 regulates Nanog status during the epithelial-mesenchymal transition via the Smad1/Akt/GSK3beta
541 signaling pathway in non-small-cell lung cancer. Oncotarget 5, 3880-3894.
542 Li, X., Lv, Y., Hao, J., Sun, H., Gao, N., Zhang, C., Lu, R., Wang, S., Yin, L., Pu, Y., Chen, R., 2016. Role of
543 microRNA-4516 involved autophagy associated with exposure to fine particulate matter. Oncotarget
544 7, 45385-45397.
545 Ma, M., Li, S., Jin, H., Zhang, Y., Xu, J., Chen, D., Kuimin, C., Yuan, Z., Xiao, C., 2015. Characteristics and
546 oxidative stress on rats and traffic policemen of ambient fine particulate matter from Shenyang. The
547 Science of the total environment 526, 110-115.
548 Massague, J., Seoane, J., Wotton, D., 2005. Smad transcription factors. Genes & development 19, 2783-
549 2810.
550 Ma, Z.L., Zhang, B.J., Wang, D.T., Li, X., Wei, J.L., Zhao, B.T., Jin, Y., Li, Y.L., Jin, Y.X., 2015. Tanshinones
551 suppress AURKA through up-regulation of miR-32 expression in non-small cell lung cancer.
552 Oncotarget 6, 20111-20120.
553 McCormack, N., Molloy, E.L., O'Dea, S., 2013. Bone morphogenetic proteins enhance an epithelial-
554 mesenchymal transition in normal airway epithelial cells during restitution of a disrupted epithelium.
555 Respiratory research 14, 36.
556 Pieniazek, M., Donizy, P., Zietek, M., Szynglarewicz, B., Matkowski, R., 2012. [The role of TGF-beta-related
557 signal transduction pathways in pathogenesis of epithelial-mesenchymal transition as a key element
558 in cancer development and progression]. Postepy higieny i medycyny doswiadczalnej 66, 583-591.
559 Richter, A., Valdimarsdottir, L., Hrafnkelsdottir, H.E., Runarsson, J.F., Omarsdottir, A.R., Ward-van
560 Oostwaard, D., Mummery, C., Valdimarsdottir, G., 2014. BMP4 promotes EMT and mesodermal
561 commitment in human embryonic stem cells via SLUG and MSX2. Stem cells 32, 636-648.
562 Russo, A., Russo, G., 2017. Ribosomal Proteins Control or Bypass p53 during Nucleolar Stress.
563 International journal of molecular sciences 18.
564 Siebzehnrubl, F.A., Silver, D.J., Tugertimur, B., Deleyrolle, L.P., Siebzehnrubl, D., Sarkisian, M.R., Devers,
565 K.G., Yachnis, A.T., Kupper, M.D., Neal, D., Nabilsi, N.H., Kladde, M.P., Suslov, O., Brabletz, S., Brabletz,
566 T., Reynolds, B.A., Steindler, D.A., 2013. The ZEB1 pathway links glioblastoma initiation, invasion and
567 chemoresistance. EMBO molecular medicine 5, 1196-1212.
568 Skrzypski, M., Dziadziuszko, R., Jassem, J., 2011. MicroRNA in lung cancer diagnostics and treatment.
569 Mutation research 717, 25-31.
570 Su, R., Jin, X., Zhang, W., Li, Z., Liu, X., Ren, J., 2017. Particulate matter exposure induces the autophagy of
571 macrophages via oxidative stress-mediated PI3K/AKT/mTOR pathway. Chemosphere 167, 444-453.
572 Turner, M.C., Krewski, D., Pope, C.A., 3rd, Chen, Y., Gapstur, S.M., Thun, M.J., 2011. Long-term ambient
573 fine particulate matter air pollution and lung cancer in a large cohort of never-smokers. American
574 journal of respiratory and critical care medicine 184, 1374-1381.
575 Virtanen, S., Alarmo, E.L., Sandstrom, S., Ampuja, M., Kallioniemi, A., 2011. Bone morphogenetic protein -
576 4 and -5 in pancreatic cancer--novel bidirectional players. Experimental cell research 317, 2136-
577 2146.
578 von Bubnoff, A., Cho, K.W., 2001. Intracellular BMP signaling regulation in vertebrates: pathway or
579 network? Developmental biology 239, 1-14.
580 Wang, Y., Sun, B., Zhao, X., Zhao, N., Sun, R., Zhu, D., Zhang, Y., Li, Y., Gu, Q., Dong, X., Wang, M., An, J.,
581 2016. Twist1-related miR-26b-5p suppresses epithelial-mesenchymal transition, migration and
582 invasion by targeting SMAD1 in hepatocellular carcinoma. Oncotarget 7, 24383-24401.

19
 ACCEPTED MANUSCRIPT

583 Xia, H., Long, J., Zhang, R., Yang, X., Ma, Z., 2015. MiR-32 contributed to cell proliferation of human
584 breast cancer cells by suppressing of PHLPP2 expression. Biomedicine & pharmacotherapy =
585 Biomedecine & pharmacotherapie 75, 105-110.
586 Xia, W., Zhou, J., Luo, H., Liu, Y., Peng, C., Zheng, W., Ma, W., 2017. MicroRNA-32 promotes cell
587 proliferation, migration and suppresses apoptosis in breast cancer cells by targeting FBXW7. Cancer
588 cell international 17, 14.
589 Xu, T., Yu, C.Y., Sun, J.J., Liu, Y., Wang, X.W., Pi, L.M., Tian, Y.Q., Zhang, X., 2011. Bone morphogenetic
590 protein-4-induced epithelial-mesenchymal transition and invasiveness through Smad1-mediated
591 signal pathway in squamous cell carcinoma of the head and neck. Archives of medical research 42,
592 128-137.
593 Yan, C., Yu, J., Liu, Y., Kang, W., Ma, Z., Zhou, L., 2015. MiR-32 promotes gastric carcinoma tumorigenesis
594 by targeting Kruppel-like factor 4. Biochemical and biophysical research communications 467, 913-
595 920.
596 Yan, S.Y., Chen, M.M., Li, G.M., Wang, Y.Q., Fan, J.G., 2015. MiR-32 induces cell proliferation, migration,
597 and invasion in hepatocellular carcinoma by targeting PTEN. Tumour biology : the journal of the
598 International Society for Oncodevelopmental Biology and Medicine 36, 4747-4755.
599 Yan, X.D., Wang, Q.M., Tie, C., Jin, H.T., Han, Y.X., Zhang, J.L., Yu, X.M., Hou, Q., Zhang, P.P., Wang, A.P.,
600 Zhang, P.C., Gao, Z., Jiang, J.D., 2017. Polydatin protects the respiratory system from PM2.5
601 exposure. Scientific reports 7, 40030.
602 Yu, J., Lei, R., Zhuang, X., Li, X., Li, G., Lev, S., Segura, M.F., Zhang, X., Hu, G., 2016. MicroRNA-182 targets
603 SMAD7 to potentiate TGFbeta-induced epithelial-mesenchymal transition and metastasis of cancer
604 cells. Nature communications 7, 13884.
605 Yue, H., Yun, Y., Gao, R., Li, G., Sang, N., 2015. Winter Polycyclic Aromatic Hydrocarbon-Bound Particulate
606 Matter from Peri-urban North China Promotes Lung Cancer Cell Metastasis. Environmental science
607 & technology 49, 14484-14493.
608 Zhan, B., Lu, D., Luo, P., Wang, B., 2016. Prognostic Value of Expression of MicroRNAs in Non-Small Cell
609 Lung Cancer: A Systematic Review and Meta-Analysis. Clinical laboratory 62, 2203-2211.
610 Zhang, D., Ni, Z., Xu, X., Xiao, J., 2014. MiR-32 functions as a tumor suppressor and directly targets EZH2
611 in human oral squamous cell carcinoma. Medical science monitor : international medical journal of
612 experimental and clinical research 20, 2527-2535.
613 Zhao, H., Yang, B., Xu, J., Chen, D.M., Xiao, C.L., 2017. PM2.5-induced alterations of cell cycle associated
614 gene expression in lung cancer cells and rat lung tissues. Environmental toxicology and
615 pharmacology 52, 77-82.
616 Zhao, L., Han, T., Li, Y., Sun, J., Zhang, S., Liu, Y., Shan, B., Zheng, D., Shi, J., 2017. The lncRNA SNHG5/miR-
617 32 axis regulates gastric cancer cell proliferation and migration by targeting KLF4. FASEB journal :
618 official publication of the Federation of American Societies for Experimental Biology 31, 893-903.
619 Zhu, D., Chen, H., Yang, X., Chen, W., Wang, L., Xu, J., Yu, L., 2015. MiR-32 functions as a tumor
620 suppressor and directly targets SOX9 in human non-small cell lung cancer. OncoTargets and
621 therapy 8, 1773-1783.
622
623
624
625

20
 ACCEPTED MANUSCRIPT

Highlights

 PM2.5 could induce epithelial-to-mesenchymal transition (EMT) in A549 cells and H292
cells.
 Smad1 was up-regualted and I-Smads（Smad6 and Smad7） were down-regulated
in the EMT process induced by PM2.5.
 miR-32 was up-regulated in the EMT process induced by PM2.5.
 Knockdown Smad1 can inhibit the EMT process induced by PM2.5.
 Knockdown miR-32 can improve the activation of Smad1 and the EMT process, but
reduced the levels of I-Smads