Abstract

   The import of proteins into the nucleus plays an important role in
   tumor development. In addition to the classical nuclear import proteins
   importin‐β and importin‐α, there are many nonclassical nuclear import
   proteins that include TNPO2. The role of TNPO2 as a nonclassical
   nuclear import protein in tumors is limited. Our previous studies have
   shown that DYNC1I1 is a poor prognostic factor for gastric cancer and
   can promote the proliferation and metastasis of gastric cancer cells.
   An expression profile chip showed that TNPO2 was its potential
   downstream. DYNC1I1 upregulated TNPO2 expression by upregulating SP1,
   following which, SP1 recruited and bound to the P300‐acetylated TNPO2
   promoter region histones, and thus promoted TNPO2 expression. At the
   same time, TNPO2 promoted gastric cancer cell proliferation and
   inhibited apoptosis by a mechanism that might be depending on the
   functional expression of P21.

   Keywords: DYNC1I1, gastric cancer, P300, proliferation, SP1, TNPO2
     __________________________________________________________________

   TNPO2 promotes gastric cancer cell proliferation and inhibits apoptosis
   by upregulating P21. DYNC1I1 upregulates TNPO2 expression by
   upregulating transcription factor SP1

   graphic file with name CAM4-8-7299-g008.jpg

1. INTRODUCTION

   Gastric carcinoma is one of the most common malignancies, with a
   relatively high rate of mortality across the world.[38]1 In addition,
   due to lack of effective and predictive capabilities for early
   screening, gastric cancer in most of the patients has been metastasized
   by the time of their initial diagnosis. Furthermore, there is currently
   no particular effective therapeutic target for gastric cancer.
   Therefore, it is urgent and imperative to explore the molecular
   mechanism of gastric cancer development.

   According to our previous experiments, cytoplasmic dynein intermediate
   chain 1 (DYNC1I1) is associated with poor prognosis of gastric cancer
   and promoted proliferation and migration of gastric cancer cells both
   in vivo and in vitro.[39]2 To study the specific underlying mechanism
   involved in the functional role of DYNC1I1 in gastric cancer, an
   affymetrix‐mediated scanning microarray of genome‐wide expression
   profiles after knockdown of DYNC1I1 was conducted. Transportin 2
   (TNPO2) is identified as the most prominent gene in downregulating
   transcriptional expression.

   TNPO2 belongs to β‐karyopherin family, which involves a family of
   factors that can be divided into importins and exportins.[40]3, [41]4
   The major family members included in it are importin‐β and importin‐α.
   The cargo proteins are mostly transported by importin‐β into the
   nucleus, and often use importin‐α as an adaptor. Importins bind to
   nuclear localization signal (NLS) of the cargo protein in the cytoplasm
   in many different ways. The complex then passes through the nuclear
   pore complex (NPC) by interaction with imported protein and
   nucleoporin.[42]5, [43]6 Importins are bound by Ran‐GTP, releasing the
   cargo proteins in the nucleus and recycling them in the cytoplasm.
   According to the recent research, karyopherin‐β family is regarded as a
   potential target for cancer therapy.[44]3, [45]7, [46]8, [47]9, [48]10
   The TNPO2‐mediated transport of the material into the nuclear process
   is similar to the above described system. But TNPO2 only transports
   limited nuclear‐destined cargoes, and participates only in the nuclear
   process, without participating in the cargo export process.[49]11

   TNPO2 is a nuclear accessory protein, in which its transport‐assisted
   protein belongs to the class of RNA‐binding proteins, including HuR and
   hnRNP. This in turn belongs to noncanonical nuclear pathway[50]11,
   [51]12 that promotes muscle maturation and differentiation. At the same
   time, it also plays an important role in tumor cells. During apoptosis,
   TNPO2 inhibits apoptosis in Hela cells.[52]13 Recent studies have
   demonstrated that TNPO2 is highly expressed in malignant pleural
   mesothelioma (MPM) tissues and cells, which might represent a new
   cancer gene.[53]14 However, the role played by TNPO2 in gastric cancer
   has not been previously elucidated or reported. As far as we know,
   there is no report till date on the regulation of TNPO2, and there is
   very little research on the mechanism of action of it in tumors
   currently.

   Hence, in this study, we initially found that the downregulation of
   TNPO2 mRNA expression was most significantly observed after silencing
   DYNC1I1 in gastric cancer cells by gene expression microarray analysis,
   which indicated that DYNC1I1 as a potential upstream signaling mediator
   of TNPO2. In addition, the putative mechanism by which DYNC1I1
   regulates TNPO2 was explored. The results showed that the TNPO2
   transcription factor SP1 plays an important role, as it not only serves
   as a transcription factor but also simultaneously recruits the
   P300‐acetylated TNPO2 promoter region histone. Furthermore, the role of
   TNPO2 in gastric cancer cells was explored, revealing TNPO2 in the
   promotion of proliferation of gastric cancer cells, and inhibition of
   apoptosis. It is suggested that the potential downstream mediator of
   TNPO2 might be P21 under these settings.

2. MATERIALS AND METHODS

2.1. Cell lines and culture condition

   HGC‐27, SGC‐7901, SNU‐216, and MGC‐803 gastric cancer cells were
   obtained from the Type Culture Collection of the Chinese Academy of
   Sciences (China). Cells were cultured in RPMI1640 (Gibco) supplemented
   with 10% fetal bovine serum, 100 U/ml penicillin, and 100 µg/mL
   streptomycin, and maintained in a 5% CO[2] incubator at 37°C.
   Experiments were carried out using logarithmic growth phase cells.

2.2. Total RNA isolation and RT‐qPCR

   The TRIzol kit was used to extract the total RNA, then the RNA
   reversely transcribed into cDNA according to the PrimeScript® RT
   reagent Kit with gDNA Eraser (Takara) kit protocol. After the
   complementary Deoxyribose Nucleic Acid (cDNA) was amplified, RT‐PCR was
   used to detect the relative expressions of related genes using SYBR
   Premix Ex Taq II kit. Primers used in the study were: TNPO2 F:
   5′‐CTTTAGGGTGTCTTCCGCGA‐3′, R: 5′‐GAAGACAGTAGGGCACCGTT‐3′; SP1 F:
   5′‐CCCTTGAGCTTGTCCCTCAG‐3′, R: 5′‐TGAAAAGGCACCACCACCAT‐3′; GAPDH F:
   5′‐GGTGTGAACCATGAGAAGTATGA‐3′, R: 5′‐GAGTCCTTCCACGATACCAAAG‐3′; the
   relative mRNA expression of TNPO2 and SP1 were estimated by ΔΔCt and
   normalized to GAPDH.

2.3. Western blot analysis

   The cells were treated according to different requirements of the
   experiments then cells were collected using lysis buffer, after lysed
   at 4°C for 40 minutes centrifuged at 13 000 rpm for 25 minutes, take
   the supernatant and the Coomassie Brilliant Blue method was used for
   protein quantification. Mix with 3× Loading Buffer and boil at 95°C for
   5 minutes. Then samples were subjected to electrophoresis in 5%, 8%, or
   15% SDS‐polypropylene gel at a concentration of 30‐50 µg/lane for
   3 hours and then transferred to a nitrocellulose membrane. DYNC1I1
   antibody (Abcam 23905 1:2000), β‐actin antibody (Santa sc‐47778
   1:1000), TNPO2 antibody (Abcam ab127165 1:1000), SP1 antibody (CST
   9389S 1:1000), P300 (CST 86377S 1:1000), HH3 (CST 4499P 1:2000),
   Ace‐H3K9 (CST 9649 1:1000), Ace‐H3K14 (CST 7627S 1:1000), Ace‐H3K27
   (CST 8173 1:1000), HuR (CST 12582 1:1000), CyclineD1 (Santa 753 1:500),
   CyclineE (Santa 481 1:500), P53 (CST 2527S 1:1000), P21 (Santa 6246
   1:500), and CDK2 (Santa 748 1:1000) were added and stored at 4°C
   overnight. Add horseradish peroxide‐labeled goat anti‐mouse (1:2000) or
   goat anti‐rabbit (1:2000) secondary antibody for 30 minutes, signal was
   visualized through a chemiluminescent detection system.

2.4. Cell transfection

   The gastric cancer cells were inoculated into a six‐well plate at
   1.5 × 10^5/well, after 24 hours, siRNA/shRNA transfection was then
   carried out by using Lipofectamine 2000 (Invitrogen).Transfection was
   done according to the manufacturer's instructions. Culture medium was
   replaced 6‐8 hours later. Cells were continued to be processed at
   different time points according to the different needs of the
   experiment.

2.5. Cell cycle assay

   The effect of TNPO2 on the cell cycle was assessed using propidium
   iodide staining. Cells were trypsinized after transfection with siTNPO2
   for 48 hours, centrifuged at 1000 rpm for 5 minutes, fixed overnight
   with 75% alcohol precooled to 4°C, then centrifuged at 1000 rpm for
   5 minutes, and suspended in PBS (precooled). Centrifugation was again
   carried out at 4°C, and then 400 μL of PBS containing 50 μg/mL PI and
   100 μg/mL RNase was added to the precipitate at 37°C for 30 minutes.
   After staining, the percentage change of G1, S, and G2 phase cells was
   evaluated by flow cytometry (BD AccuriTM C6 flow cytometry).

2.6. Immunoprecipitation

   35 ul Protein G‐sepharose beads (DYNC1I1) and 35 μL Protein A‐sepharose
   beads (P300, SP1) were added to the EP tube and the beads were washed
   twice with PBS and then lysis buffer, respectively. The total protein
   of the control and treatment groups was extracted with lysis buffer,
   applying Coomassie Bright. The total protein concentration was
   quantitatively extracted by the method. Forty microliters of protein
   was taken from each sample as input. The remaining protein lysate and
   10 µg of anti‐DYNC1I1 antibody, 10 µg of anti‐P300 antibody (CST
   86377S), and 10 µg of anti‐SP1 antibody (CST 9389S) in each sample were
   then added to Protein G‐sepharose beads and Protein A‐sepharose beads,
   respectively, and the solution containing beads, protein lysate, and
   antibody was slowly shaken overnight at 4°C. Then the beads were washed
   four times with lysis buffer, the supernatant was aspirated, and
   2× Buffer 40‐50 µL was added and boiled for 5 minutes, and then the
   samples were subjected to Western blot analysis.

2.7. MTT assay

   For MTT assay, gastric cancer cells transfected with TNPO2 siRNA or
   overexpressed TNPO2 were uniformly tiled in a 96‐well plates (3000
   cells per well), 48, 72, and 96 hours later, added MTT (concentration
   with 5 mg/mL), then continue to incubated for 4 hours. Removed the
   supernatants, 200 μL DMSO was added to each well and the absorbance was
   read at 570 nm. Cell viability: relative cell activity = (OD570
   measurement − OD570 blank)/(OD570 control − OD570 blank) × 100%.

2.8. Cellular apoptosis detection

   The Annexin V‐fluorescein isothiocyanate (FITC) Apoptosis Detection kit
   (BD Biosciences) was used to detect cell apoptosis. Cells were
   collected 48 hours after transfection with TNPO2 siRNA or over
   expressed TNPO2. The manufacturer's protocol was followed. Apoptosis
   changes were then detected using a flow cytometer (BD AccuriTM C6 flow
   cytometry).

2.9. Cytoplasmic and nuclear extraction

   Cytoplasmic and nuclear extraction kit for cells purchased from Active
   Motif. The samples were then subjected to Western blot analysis. Lamin
   A/C was considered as the internal reference of nuclear protein, and
   the internal reference of cytoplasmic was used with GAPDH. The cells
   were collected into a 1.5 mL microcentrifuge tube precooled to 4°C,
   centrifuged at 500 g for 1 minute, the supernatant was discarded, and
   an appropriate amount of cytoplasmic extraction buffer was added to the
   cell pellet, vortexed vigorously, and incubated on ice for 30 minutes.
   Procedures for cytoplasmic and nuclear protein extraction are described
   in the protocol.

2.10. Transcription factor and the binding site prediction

   The ALGGEN PROMO software program ([54]http://alggen.lsi.upc.es)[55]15
   and GeneCards ([56]https://www.genecards.org/)[57]16 were used to
   predict transcription factors. JASPAR
   ([58]http://jaspar.binf.ku.dk/)[59]17 and UCSC website
   ([60]https://genome.ucsc.edu/)[61]18 were used to predict transcription
   factor binding sites.

2.11. Bioinformatics

   Involvement of positive correlation genes of KEGG pathway and GO
   pathway enrichment analysis were evaluated using DAVID online software
   ([62]https://david.ncifcrf.gov/).[63]19 The GEPIA website[64]20 was
   used to predict gene correlation in gastric cancer.

2.12. The EdU incorporation assay

   The treated gastric cancer cells were seeded into 96‐well plates at a
   concentration of 2000‐5000 cells/200 µL. After 24 hours of incubation,
   50 μmol/L of 5‐ethynyl‐2'‐deoxyuridine (EdU; Ribobio) was added to each
   well, incubated at 37°C for 2 hours, and then incubated with 4%
   formaldehyde at room temperature. Fix the cells for 30 minutes.
   Incubate with 2 mg/mL glycine for 5 minutes. After washing five times
   with PBS, the cells were reacted with 100 μL of a 1× Apollo reaction
   mixture for 30 minutes. Then, the nuclei were stained with 1× Hoechst
   33342 (5 μg/mL).

2.13. Luciferase activity assay

   The binding sites on the promoter region of TNPO2 by SP1 were predicted
   by online data. We construted two plasmids, pGL4.10‐TNPO2 Promoter(Wt,
   wild type) and pGL4.10‐TNPO2 Promoter(Mut, mutant type). The plasmid
   was then cotransfected with the reporter plasmid into gastric cancer
   cells. Determination of luciferase activity was carried out on TECAN
   Infinite M200Pro reader according to the manufacturer recommendations
   (Promega); Renilla luciferase was used for normalization.

2.14. Statistical analysis

   All statistical analyses were carried out using SPSS version 22.0.
   Differences between groups were compared by using Student's t test.
   Each experiment was repeated three times and the data were expressed as
   mean + standard deviation. A value of P < 0.05 was considered to be
   statistically significant.

3. RESULTS

3.1. DYNC1I1 upregulated TNPO2 in gastric cancer cells

   Previous studies have confirmed the effect of DYNC1I1 on the biological
   function of gastric cancer cells. To further explore the mechanism of
   DYNC1I1 in gastric cancer, affymetrix scanner microarray genome‐wide
   expression analysis was used after knockdown (KD) of DYNC1I1 in HGC‐27
   cells. As shown in Figure [65]1A, TNPO2 was considered as the primary
   gene among the downregulated genes. Therefore, the main focus is on
   TNPO2 in this study. First, to verify the accuracy of the expression
   profile chip results, the correlation between DYNC1I1 and TNPO2 in
   gastric cancer was predicted by using the Gene Expression Profiling
   Interactive Analysis (GEPIA) website. As shown in Figure [66]1B,
   DYNC1I1 showed positive correlation with TNPO2 in gastric cancer with a
   correlation coefficient of 0.69 (P < 0.05). Furthermore, RT‐qPCR
   analysis was performed to detect the changes in TNPO2 mRNA levels after
   DYNC1I1 knockdown. The results were consistent with those of microarray
   analysis in HGC‐27 cells (Figure [67]1C). Western blotting analysis was
   then performed to detect changes in the relative protein expression
   levels of TNPO2 after knockdown of DYNC1I1, and this was consistent
   with mRNA data. The proteins levels of TNPO2 were downregulated after
   knockdown of DYNC1I1 (Figure [68]1D). These results indicated that
   knockdown of DYNC1I1 in gastric cancer cells also downregulated TNPO2,
   suggesting that DYNC1I1 might be a potential upstream signaling
   molecule for the activity of TNPO2.

Figure 1.

   Figure 1
   [69]Open in a new tab

   DYNC1I1 upregulated TNPO2 in gastric cancer cell. HGC‐27, SGC‐7901, and
   SNU‐216 cell lines were transfected with siDYNC1I1 for 48 h. A, After
   knockdown of DYNC1I1 in HGC‐27 cells, the top 10 upregulated genes
   reported by microarray genome‐wide expression analysis. B, Correlation
   between DYNC1I1 and TNPO2 in gastric cancer was performed using GEPIA
   online platform; differences with P < 0.05 were considered
   statistically significant. C, YNC1I1 and TNPO2 mRNA expression levels
   were tested by qRT‐PCR in HGC‐27, SGC‐7901, and SNU‐216 cells. D,
   DYNC1I1 and TNPO2 protein expressions were detected by Western blot
   analysis in HGC‐27, SGC‐7901, and SNU‐216 cells (*P < 0.05, **P < 0.01,
   ***P < 0.001, ****P < 0.0001; n = 3, student s t test, means ± 95% CI)

3.2. DYNC1I1 upregulated TNPO2 expression by increasing SP1 in gastric cancer
cells

   To further investigate the mechanism by which DYNC1I1 upregulates TNPO2
   expression, TNPO2 transcription factor was first predicted by exploring
   the ALGGEN PROMO website (Figure [70]2A). At the same time, the TNPO2
   transcription factor was predicted on the genecard website. The major
   four transcription factors were as follows: Arnt, Nkx2‐5, Pax‐6, and
   SP1. The common transcription factor in both the sites was SP1. It was
   speculated that DYNC1I1 might regulate the expression of TNPO2 by
   modulating its transcription factor SP1. Furthermore, the correlation
   between DYNC1I1 and SP1, as well as TNPO2 and SP1 in gastric cancer was
   verified by the GEPIA website. As predicted, DYNC1I1 showed positive
   association with SP1 (correlation coefficient 0.48; P < 0.05), and
   TNPO2 showed positive association with SP1 (correlation coefficient
   0.59; P < 0.05; Figure [71]2B). RT‐qPCR and Western blotting analysis
   were performed to detect SP1 mRNA and protein expression levels after
   knockdown of DYNC1I1. As shown in Figure [72]2C, SP1 was downregulated
   regardless of mRNA or protein levels after knockdown of DYNC1I1. This
   was consistent with the relationship between DYNC1I1 and SP1 as
   verified previously by exploring the website. Next, SP1 was knocked
   down in gastric cancer cells with the aim to detect altered TNPO2 mRNA
   and protein expression levels. As shown in Figure [73]2E, it was found
   that TNPO2 was downregulated in gastric cancer cells at both mRNA and
   protein expression levels after SP1 knockdown. Dual luciferase assay
   was then performed to detect whether SP1 binds to TNPO2 promoter site.
   The results of the assay indicated that luciferase activity was
   significantly higher when cotransfected with pGL4.10‐TNPO2 promoter
   (Wt) and SP1 (P < 0.05). The luciferase activity in pGL4.10‐TNPO2
   promoter (Mu) and pcDNA4.1‐SP1 cotransfection groups was not
   significantly higher when compared to normal control cells (Figure
   [74]2G).These observations demonstrated that DYNC1I1 decreased TNPO2
   expression by downregulating SP1 expression in gastric cancer cells,
   further confirming that SP1 serves as a transcription factor of TNPO2.

Figure 2.

   Figure 2
   [75]Open in a new tab

   DYNC1I1 upregulated TNPO2 expression by upregulating SP1 in gastric
   cancer cells. A, Transcription factors (TF) prediction in the promoter
   region of TNPO2 on GeneCards (only the top four) and ALGGEN PROMO
   website, respectively, intersect the results obtained by the two
   websites. B, Correlation between DYNC1I1 and SP1, TNPO2, and SP1 in
   gastric cancer were performed using GEPIA online platform; differences
   with P < .05 were considered statistically significant. C, SP1 mRNA
   expression levels were tested by qRT‐PCR in HGC‐27, SGC‐7901, and
   SNU‐216 cells after transfected with siDYNC1I1 for 48 h. D, SP1 protein
   expression was detected by Western blot analysis in HGC‐27, SGC‐7901,
   and SNU‐216 cells after transfected with siDYNC1I1 for 48 h. E, SP1 and
   TNPO2 mRNA expression levels were tested by qRT‐PCR in HGC‐27,
   SGC‐7901, and SNU‐216 cells after transfected with siSP1 for 48 h. F,
   SP1 and TNPO2 protein expressions were detected by Western blot
   analysis in HGC‐27, SGC‐7901, and SNU‐216 cells after transfected with
   siSP1 for 48 h. G, The luciferase activities in HGC‐27 cotransfected
   with SP1 overexpression or NC and luciferase reporters containing TNPO2
   promotor WT or TNPO2 promotor MUT. Student's t tests were used for
   statistical analyses (**P < 0.01, ***P < 0.001, ****P < 0.0001; n = 3,
   student s t test, means ± 95% CI)

3.3. SP1 enhanced histone acetylation levels in TNPO2 promoter regions by
binding to P300

   Acetylation of H3K27 in TNPO2 promoter region was found by exploring
   the UCSC website ([76]https://genome.ucsc.edu/) (Figure [77]3A).
   Previous studies have shown that SP1 can bind to the acetylation
   coactivator P300. Coregulation of acetylated target gene promoter
   region then promotes transcription, and whether similar mechanism
   exists in this study requires further validation. The level of
   acetylation in different parts of histone 3 was detected after
   knockdown of SP1 in HGC‐27 cell. As shown in Figure [78]3B, H3K9 and
   H3K27 acetylation levels showed significant downregulation after SP1
   knockdown, and both these sites were present at TNPO2 promoter. It was
   speculated that SP1 affected the levels of TNPO2 promoter acetylation,
   thus affecting its transcription. Changes in acetylation levels were
   also found after knockdown of DYNC1I1 in the same cell line (Figure
   [79]3C). Next, coimmunoprecipitation assay was performed in HGC‐27 cell
   to determine whether SP1 binds to P300 or to determine whether DYNC1I1
   binds to P300. The results revealed that SP1 can bind to P300 instead
   of DYNC1I1 (Figure [80]3D). These results showed that DYNC1I1 regulates
   SP1 expression in gastric cancer cells, and SP1 not only binds to TNPO2
   promoter region but also recruits acetylated coactivator P300 to
   increase TNPO2 promoter region acetylation, thus driving TNPO2
   transcription.

Figure 3.

   Figure 3
   [81]Open in a new tab

   SP1 enhanced the histone acetylation levels in TNPO2 promoter regions
   by binding to P300. A, The UCSC Genome Bioinformatics Site
   ([82]http://genome.ucsc.edu/)showed high enrichment of H3K27Ac at the
   promoter of TNPO2. B and C, Protein expression level of H3, H3[K9],
   H3[K14], and H3[K27] were detected by Western blot analysis after
   transfected with siSP1 or transfected with siDYNC1I1 for 48 h. D,
   Immunoprecipitation assay was performed to test the complex formation
   of SP1 and P300 physically in HGC‐27 cells (*P < 0.05,**P < 0.01;
   n = 3, student s t test, means ± 95% CI)

3.4. High TNPO2 expression defines a poor prognosis in gastric cancer
patients

   The role of TNPO2 in gastric cancer has not been studied till date. The
   expression level of TNPO2 in gastric cancer and adjacent tissues was
   predicted by using the GEPIA website. As shown in Figure [83]4A,
   although no statistical significance was observed, the expression of
   TNPO2 in gastric cancer was higher than that found in normal tissues.
   Moreover, the Kaplan‐Meier plotter program
   ([84]http://kmplot.com/analysis/index.php?p=service%26cancer=gastric)[8
   5]21 predicted the effect of TNPO2 on gastric cancer prognosis and the
   results revealed that high TNPO2 expression showed a lower survival
   rate in gastric cancer cells (Figure [86]4B).

Figure 4.

   Figure 4
   [87]Open in a new tab

   TNPO2 is overexpressed in GC tissues and patients with high TNPO2
   expression have a poor prognosis. A, Analysis of TNPO2 expression using
   GEPIA Website. B, Kaplan‐Meier Plotter analysis of the effect of TNPO2
   on GC patients' survival

3.5. TNPO2 promoted gastric cancer cell proliferation and inhibited apoptosis

   To assess the biological function of TNPO2 in gastric cancer cells,
   siRNA was used to transiently knockdown the expression of TNPO2, and
   Western blotting assay was performed to detect the knockdown
   efficiency. As shown in Figure [88]5A, siRNA reduced TNPO2 protein
   expression. MTT and Edu assays were used to quantify the effect of
   knockdown of TNPO2 on the ability of cancer cells to proliferate. As
   shown in Figure [89]5B, knockdown of TNPO2 reduced the ability of
   gastric cancer cells to proliferate and increased the inhibition of
   gastric cancer cell proliferation over time when compared with negative
   controls. The cell viability of HGC‐27 cells was decreased by about 25%
   (P < 0.05) after 48 hours of knockdown of TNPO2, and by approximately
   30% after 72 hours of knockdown of TNPO2 (P < 0.05). In addition, the
   cell viability of HGC‐27 cells was decreased by about 40% after
   96 hours of knockdown of TNPO2 (P < 0.05). In contrast, SGC‐7901 cell
   viability was decreased by about 20% (P < 0.05) after 48 hours of TNPO2
   knockdown, by 30% after 72 hours (P < 0.05), and then by 40% after
   96 hours (P < 0.05). Similar results were obtained after TNPO2
   silencing in SNU‐216 cells. As shown in Figure [90]5C, silencing of
   TNPO2 inhibited DNA replication in gastric cancer cells. In addition,
   flow cytometry was performed to detect the changes in the levels of
   gastric cancer cell apoptosis after 48 hours of transient knockdown of
   TNPO2 in HGC‐27, SGC‐7901, and SNU‐216 cells. The results revealed that
   the levels of apoptosis were increased after 48 hours of knockdown of
   TNPO2 when compared with control group (Figure [91]5D).
   Apoptosis‐related proteins were then detected by Western blotting assay
   (Figure [92]5E), which showed that the apoptotic proteins caspases 3
   and 9 and PARP showed significant increase. For further analyses,
   overexpression of TNPO2 in MGC‐803 cells and overexpression efficiency
   were detected by Western blotting (Figure [93]6A). MTT assay was
   performed and the results indicated that overexpression of TNPO2 in
   MGC‐803 cells enhanced the viability of MGC‐803 cells in a
   time‐dependent manner (Figure [94]6B). By 48 to 96 hours after
   overexpression of TNPO2 in MGC‐803 cells, proliferation was increased
   by approximately 20 to 50% as observed in those without TNPO2
   overexpression (P < 0.05). Similarly, the EdU results showed that
   overexpression of TNPO2 in gastric cancer cells increased DNA
   replication (Figure [95]6C). Furthermore, the apoptosis levels were
   decreased after overexpression of TNPO2at 48 hours (Figure [96]6D).
   These results showed that TNPO2 promoted gastric cancer cell
   proliferation and inhibited their apoptosis.

Figure 5.

   Figure 5
   [97]Open in a new tab

   Silencing TNPO2 in gastric cancer cells inhibited proliferation of
   gastric cancer cells and promoted apoptosis. A, TNPO2 protein
   expression was detected by Western blot analysis in HGC‐27, SGC‐7901,
   and SNU‐216 cells after transfected with siTNPO2 for 48 h. B, Analysis
   of cell proliferation following TNPO2 knockdown in HGC‐27, SGC‐7901,
   and SNU‐216 cells by MTT assay. C, The apoptotic rates of HGC‐27 and
   SGC‐7901 cells transfected with siNC or siTNPO2 for 48 h were
   visualized by flow cytometry. D, EdU incorporation assay. Seventy‐two
   hours after gastric cancer cells were transfected with siTNPO2 or siNC,
   followed by incubation with EdU and Hoechst in sequence. Hoechst 33342
   (blue) and EdU (green) represent cell nuclei and nuclei of
   proliferative cells, respectively. The percentages of the EdU‐positive
   cells are presented (right). E, Apoptotic proteins were detected by
   Western blot analysis in HGC‐27 cells after transfected with siTNPO2
   for 48 h (*P < 0.05, **P < 0.01, ***P < 0.001; n = 3, student s t test,
   means ± 95% CI)

Figure 6.

   Figure 6
   [98]Open in a new tab

   Overexpression of TNPO2 promoted gastric cancer cell proliferation and
   inhibited apoptosis. A, TNPO2 protein expression was detected by
   Western blot analysis in MGC‐803 cells after transfected with
   overexpressing TNPO2 for 48 h. B, MTT shows cell viability of gastric
   cancer cells after overexpressing TNPO2 for 48, 72, and 96 h. C, The
   percentages of the EdU‐positive cells after overexpressing TNPO2 for
   72 h. D, The apoptotic rates of MGC‐803 cells after overexpressing
   TNPO2 for 48 h (*P < 0.05, **P < 0.01, ****P < 0.0001; n = 3, student s
   t test, means ± 95% CI)

3.6. TNPO2 promoted cell proliferation and inhibited apoptosis by
upregulating P21

   To further explore the specific mechanism by which TNPO2 modulates the
   biological functions in gastric cancer cells, some additional
   experiments were conducted. Previously it has been reported that TNPO2
   promotes Hela cell apoptosis by inhibiting the entry of HuR protein
   into the nucleus in response to severe stress conditions.[99]13 Next,
   TNPO2 was knocked down in HGC‐27 cells for undergoing nucleoplasmic
   separation. However, the results showed that the HuR protein showed no
   significant concentration in the cytoplasm (Figure [100]7A), suggesting
   that TNPO2 does not promote apoptosis through this mechanism in gastric
   cancer cells.

Figure 7.

   Figure 7
   [101]Open in a new tab

   TNPO2 promoted apoptosis and inhibited cell proliferation by
   upregulated P21. A, HuR protein expression were detected by Western
   blot analysis in HGC‐27 in cytoplasmic and nuclear extracts,
   respectively, after transfected with siTNPO2 for 48 h. B, The DAVID
   website enriches the pathways involved in TNPO2 positively related
   genes. C, Flow cytometry was performed using PI staining to determine
   the cell cycle of HGC‐27 cell lines after transfected with siTNPO2 for
   48 h. D, P53, CyclinD1, CyclinE, P21, and CKD2 proteins were detected
   by Western blot analysis in HGC‐27 cells after transfected with siTNPO2
   for 48 h. E, The apoptosis rate of cells was detected by flow cytometry
   after transfection with siNC, siTNPO2, and siP21 for 48 h in HGC‐27
   cells (*P < 0.05, **P < 0.01; n = 3, student s t test, means ± 95% CI)

   Furthermore, based on the DAVID website, the positive and negative
   TNPO2 gene pathways were enriched by GO method. The results found that
   “DNA replication” was the primary pathway observed from both methods,
   indicating that TNPO2 might regulate the biological behavior of gastric
   cancer cells through this pathway (Figure [102]7B). At the same time,
   the effect of transient TNPO2 knockdown in the cell cycle was detected
   (Figure [103]7C). The results showed that the cell cycle was arrested
   in G1 phase at 48 hours after knockdown of TNPO2 in gastric cancer cell
   line HGC‐27 [the frequency of G1 phase in the NC interference group was
   approximately 32%, and the frequency of G1 in TNPO2 siRNA interference
   group was approximately 56% (P < 0.01)]. The G1 phase‐detecting protein
   was assayed by Western blotting analysis (Figure [104]7D). The P21
   protein was upregulated, and this indicated that P21 might act as a
   potential downstream mediator of TNPO2. Next, the P21 was silenced in
   gastric cancer cells and found that P21 silencing in gastric cancer
   cells partially reversed the increased apoptosis (Figure [105]7E).
   These results showed that the potential downstream of TNPO2 might be
   P21.

4. DISCUSSION

   TNPO2 is an accessory protein that belongs to the class of RNA‐binding
   proteins including HuR and hnRNP in the nucleus and belongs to the
   noncanonical nuclear pathway.[106]11, [107]12 It plays an important
   role as a nuclear accessory protein in normal cells, and at the same
   time it also plays an important role in tumor cells. In Hela cells,
   knockdown of TNPO2 can lead to cytoplasmic accumulation of HuR protein,
   aggravating apoptosis.[108]13 Recent studies have shown that TNPO2 is
   highly expressed in malignant pleural mesothelioma (MPM) tissues as
   well as cells, and this might formally represent a novel
   cancer‐specific gene. However, its precise role in MPM has not been
   studied till now.[109]14 In the present study, we found that knockdown
   of TNPO2 inhibited gastric cancer cell proliferation and promoted their
   apoptosis.

   At present, most of the functional mechanisms of TNPO2 occur due to its
   role in nuclear import of HuR. TNPO2 promotes muscle fiber formation by
   influencing the distribution of HuR in cells.[110]22 However, in
   response to severe stress conditions, apoptotic cell death occurs
   subsequently in tumor cells. The activated caspase cleaves HuR into two
   fragments, namely HuR‐CP1 and HuR‐CP2. HuR‐CP1 competes with HuR in
   binding to TNPO2 for HuR accumulation in the cytoplasm, thereby
   promoting further apoptosis. This suggests that the effect of HuR is
   accumulated in the cytoplasm after silencing TNPO2 under normal
   conditions.

   In this study, we found that TNPO2 knockdown promoted apoptosis of
   gastric cancer cells, while this effect was not achieved by abnormal
   cellular distribution of HuR. This was not completely consistent with
   the results of the previously published studies, which might be due to
   the use of different cell lines in experiments. However, it is because
   of the affect of cycle‐associated protein P21.

   The initial function of P21 is that it acts as a mediator in cellular
   senescence via p53‐dependent and independent pathways.[111]23, [112]24
   On one hand, senescence can also be mediated through reactive oxygen
   species (ROS)‐based mechanisms,[113]25 and on the other hand, p21 is
   currently considered as a potent universal cyclin‐dependent kinase
   (CDK) inhibitor,[114]26, [115]27 serving as an important regulator of
   cell cycle progression in G1 as well as S phases.[116]28, [117]29 In
   the current study, the cell cycle was arrested in G1 phase following
   TNPO2 knockdown in gastric cancer cells, while the expression of P21
   was upregulated. However, CyclinD1, CyclinE, CKD2, and P53 showed no
   significant changes.

   Unfortunately, the precise mechanism of action of P21 has not been
   clarified yet. Whether to weaken the binding stability of CDK2 and
   cyclinE or CDK4/6 and CyclinD1 by inhibiting its catalytic activity or
   by other unknown mechanisms remains to be clarified. In the aspect of
   P21 and its role in apoptosis, the current study found that P21 has two
   sides. First, p21 plays a role in apoptosis inhibition, and by
   inhibiting the initiation of apoptotic caspase cleavage cascade by
   tumor necrosis factor‐related apoptosis‐inducing ligand (DR4/TRAIL)
   receptor. This in turn inhibits apoptosis and provides survival
   advantages of human tumor cells.[118]30 Second, it is known that the
   functional expression of p21 has protective effects on tumor cell
   apoptosis by inducing cell growth arrest or cellular senescence.[119]31
   In contrast, the activated caspase cleaves P21 to P15 to promote
   cellular apoptosis. We further found that TNPO2 knockdown promoted
   gastric cancer cell apoptosis, and enhanced the functional expression
   of caspases 3 and 9, and P21 in this study, which indicated that
   activated P21 does not inhibit the caspase activity. It is speculated
   that activated caspase activity cleaves P21 in order to promote
   apoptosis.

   The specific mechanism of DYNC1I1 regulation of TNPO2 was further
   explored and the results revealed that SP1 plays an important role in
   it. The transcription factor SP1 is a basal transcription factor that
   regulates the so‐called housekeeping genes.[120]32 SP1 also showed
   overexpression in many cancer types and its high expression showed
   association with poor prognosis.[121]33, [122]34, [123]35, [124]36 In
   addition to direct binding to DNA, SP1 can simultaneously recruit P300.
   Functional activation of P300 serves as a transcriptional coactivator
   that does not directly bind to DNA, catalyzing the promoter‐bound
   histone acetylation, and leading to gene activation.[125]37, [126]38,
   [127]39 For the first time we found that DYNC1I1 decreased TNPO2
   expression by downregulating SP1. In addition to its role as a TNPO2
   transcription factor, SP1 acetylates TNPO2 promoter region histone 3,
   which in turn is mediated by recruiting P300 by SP1. However, DYNC1I1
   cannot be combined with P300.

5. CONCLUSIONS

   In conclusion, we found that DYNC1I1 upregulates TNPO2 expression, and
   increases the expression of transcription factor SP1. Activated SP1 not
   only binds to TNPO2 promoter site but also recruits P300 for
   acetylation of TNPO2 promoter site histones to promote TNPO2
   transcription. The role of TNPO2 as a nuclear accessory protein in
   tumors remained unclear. TNPO2 promoted gastric cancer cell
   proliferation and inhibited apoptosis, and this effect might be
   achieved by downstream engagement of P21. Whether TNPO2 exerts this
   effect on P21 is unclear and requires research regarding the
   transportation of substances into the nucleus and on to which
   substances. Hence, further research is warranted to clarify the
   specific mechanism. By investigating the underlying mechanism of
   DYNC1I1 and TNPO2 in gastric cancer, the molecular markers that help in
   predicting the prognosis of gastric cancer were screened. For more
   precise treatment of gastric cancer, these potential molecular targets
   are intended.

CONFLICT OF INTEREST

   The authors declare that they have no competing interests.

AUTHOR CONTRIBUTIONS

   Xiujuan Qu, Zhi Li, and Yunpeng Liu conceived and designed the
   experiments. Libao Gong, Yizhe Wang, and Jin Wang performed the
   experiments. Libao Gong and Yizhe Wang analyzed the data. Libao Gong,
   Xiujuan Qu, and Yunpeng Liu wrote the paper.

ACKNOWLEDGMENTS