Ninety-eight tissue samples including 49 primary RCC tissues and 49 paired tumor adjacent renal tissues from the same patients were collected between May 2009 and April 2013 at The First Bethune Hospital of Jilin University, Jilin, China. All the samples were immediately stored in liquid nitrogen vapor at −196°C. All tumor specimens contained more than 80% cancer tissue and few necrosis in some. Tumor adjacent renal tissues were obtained from the region 5 cm distant from the tumor location, which ensured the absence of any kidney lesions. The present study was approved by the Ethics Committee of The First Bethune Hospital of Jilin University, and consent forms were obtained from all participating patients.

Human clear cell renal cell carcinoma (ccRCC) cell line 786-O (TCHu3), human RCC cell line OSRC-2 (TCHu40) and human non-tumorous embryonal kidney cell line HEK293T were purchased from the Cell Resource Center of Shanghai Life Science Institute (Shanghai, China). HEK293T and OSRC-2 cells were cultured and maintained in high glucose Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Carlsbad, CA, USA), while 786-O cells were cultured in RPMI-1640 medium (Gibco), both media being supplemented with 10% fetal bovine serum (Gibco), 100 U/ml penicillin, 100 mg/ml streptomycin in a 5×7 cm flask in a moist atmosphere containing 5% CO₂ at 37°C (Thermo Scientific, Waltham, MA, USA). Cultured cells used for extracting total RNA or proteins were harvested from 1 well of a 6-well plate and lysed using TRIzol reagent (Life Invitrogen, Carlsbad, CA, USA) or precooling RIPA lysis buffer [150 mM NaCl, 50 mM Tris-HCl (pH 7.4)], 0.25% Na-deoxycholate, 1% Nonidet P-40, 1 mM EDTA and protease inhibitor cocktail tablets (Roche, Penzberg, Germany). The methods of the cell culture were previously described (9).

Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed to cDNA using PrimeScript™ reverse transcriptase (Takara, Dalian, China). The resulting cDNA was analyzed by RT-qPCR with a real-time PCR detector (Ilumina Eco; Ilumina, San Diego, CA, USA). RT-qPCR reactions were completed under the following conditions: the initial denaturation step consisted of 95°C for 30 sec, followed by 40 cycles of further denaturation at 95°C for 5 sec, and annealing and extension at 60°C (59°C for TPM2, TPM3 and TPM4) for 30 sec. The following steps were used to obtain the melting curve: 95°C for 15 sec, then 55°C for 15 sec and finally 95°C for 15 sec again. The final concentration of all reagents in the reaction were as follows: SYBR^® Premix Ex Taq 1X, forward and reverse primers 0.2 μM each, DNA templates ≤5 ng/μl. Unless otherwise indicated, the relative mRNA expression levels were calculated as −ΔΔCt where ΔΔCt = ΔCt_tumor − ΔCt_normal and ΔCt = Ct_target − Ct_control. TPM1-4 were the target genes, and GAPDH was the control gene in the present study. Table I lists the primer sequences (synthesized by BGI Gene Company, Beijing, China) used for RT-qPCR.

Methods for the protein extraction and western blotting were previously described (9). Whole cell lysate from the cultured cells was mixed with 4X SDS loading buffer (8% SDS, 30% glycerol, 0.25 M Tris-HCl pH 6.8, 0.02% bromophenol blue with 10% BME). Kidney tissue samples (100 mg) were homogenized with tissue grinders (XinZhi, Ningbo, China) and solubilized in 100 μl precooling RIPA lysis buffer. Whole cell lysate was prepared by sonication (Scientz-IID, Ningbo, China) for 12 sec with 30% work cycle and following centrifugation at 15,000 × g for 10 min at 4°C. Total proteins in the supernatant after centrifugation were quantified using the Bradford method according to the handling manual (Bio-Rad, Hercules, CA, USA). Equal total amounts of boiled protein samples with SDS loading buffer were separated by SDS polyacrylamide gel (SDS-PAGE) electrophoresis, and specific proteins were detected by immunoblotting using TPM1 (1:500, #AV41392, rabbit anti-human polyclonal; Sigma-Aldrich, St. Louis, MO, USA) antibodies and α-tubulin (1:800, #bs-0159R, rabbit anti-human polyclonal; Bioss, Beijing, China) antibodies. Protein bands were analyzed by Quantity One^® (Bio-Rad) and the TPM1 protein level was defined as the grey scale value ratio of TPM1 to α-tubulin.

To obtain homogeneous samples for immunohistochemical analysis, we collected 25 formalin-fixed, paraffin-embedded tissue blocks from the 49 patients mentioned above. The method for immunohistochemical staining was previously described (9). Tissue blocks were sectioned and then heated in a constant temperature oven at 60°C for 3 h. The sections were deparaffinized in xylene and rehydrated through a series of graded ethanol. The sections were washed by distilled water once then with phosphate-buffered saline (PBS) buffer twice every 3 min. Each section was treated with 3% hydrogen peroxide solution at room temperature for 10 min to eliminate endogenous peroxidase. Then they were washed 3 times with PBS buffer, 5 min each time. The tissue sections were blocked with normal goat serum, incubated at room temperature for 15 min, the serum was decanted, but not washed, and 1:100 dilution of TPM1 polyclonal antibodies (#AV41392, rabbit anti-human; Sigma-Aldrich), was added and maintained at 4°C overnight. Tissue slides were washed by PBS buffer again and biotinylated secondary antibody was added at 37°C for 10 min. Tissue sections were incubated in horseradish peroxidase-labeled streptavidin working solution at 37°C for 10 min. Finally, the immunoreaction was detected using the DAB detection kit (Maixin, Fuzhou, China).

The staining intensity (SI) and the percentage of positive cells (PP) were semi-quantitatively assessed using the following criteria: SI was scored as 0, no staining; 1, weak; 2, moderate; and 3, strong; and PP: 0, no positive cells; 1, <10%; 2, 10–50%; 3, 51–80%; and 4, >80%. For each section, the immunoreactive score (IRS) was calculated as SI × PP with a maximum possible value of 12 (10). Observations were recorded separately by two different individuals and the average score was used.

The full-length coding sequence of the TPM1 gene transcription variant 5 was amplified from the HepG2 cell line cDNA library. Then the TPM1 coding sequence was linked into the T vector and finally constructed into the eukaryotic expression vector pcDNA3.1 to obtain the pcDNA3.1-TPM1 plasmid for cell transfection (Comate Bioscience, Changchun, China). HEK293T, 786-O and OSRC-2 cells were cultured in 6-well plates (1–2×10⁵ cells/well) in specific complete culture medium as described above. These cells were transiently transfected with 0.2–1 μg of pcDNA3.1-TPM1 plasmid using Lipofectamine^® 2000 reagent (Life Invitrogen) in the Opti-MEM^® diluent condition (Gibco) following the manufacturer’s instructions. Cells were harvested for further analysis 48 h after transfection.

OSRC-2 and 786-O cells were seeded in a 6-well plate and incubated to form cell monolayers. Subsequently, cross-shaped scratches were made gently with sterile pipette tips across the diameter of the wells and washed with PBS to wipe off residuals. The scratched monolayers of cells were then transfected with different plasmid DNA. At least 5 microscopic (Olympus, Tokyo, Japan) fields were analyzed for each well at a magnification of ×100 at 0, 24 and 48 h after scratching (11). The ratio of the recovered area at 24 or 48 h after scratching to the initial wound area was identified as the wound healing rate.

OSRC-2 and 786-O cells were transfected with pcDNA3.1-NC or pcDNA3.1-TPM1 and starved for 24 h. An in vitro Transwell^® membrane (Corning Costar, Corning, NY, USA) precoated with BD Matrigel matrix (BD Biosiences, Franklin Lakes, NJ, USA) and medium mixture (matrix:medium, 1:1) was prepared for the overnight invasion assay and the uncoated membrane was prepped for the migration assay. The cell seeding amounts for this 24-well plate were 1×10⁴ and 6×10⁴ cells/well for the migration and invasion assay, respectively. Then culturing under the indicated condition for 24 h was carried out as described in the product instructions for users. Finally, the bottom filters were fixed in methanol and stained with 1% crystal violet solution. Images were captured for 5 random optical fields at ×200 magnification on each filter.

OSRC-2 and 786-O cells were transfected with pcDNA3.1-NC or pcDNA3.1-TPM1, or mock treated for 72 h. Cells were collected and stained with Annexin V-FLUOS staining kit (Roche) following the instructions for users. Sample cells were then analyzed by a FACS flow cytometer (Becton-Dickinson, Franklin Lakes, NJ, USA). We recorded Annexin V-positive and propidium iodide (PI)-negative cells as early stage apoptotic cells.

Statistical analysis was achieved using IBM SPSS^® Statistics 21.0 (IBM, Armonk, NY, USA). Unless otherwise indicated, values are presented as the means ± SEM. The paired t-test was used to determine significant differences between the treatment and the control groups. The Chi-square test was used to analyze the associations between TPM1 and clinical features. The log-rank test was used to analyze the survival results. Due to the marked dispersion of the cancer specimens, the Mann-Whitney U test was performed to analyze RT-qPCR data instead of the t-test. Significance was defined at p<0.05.

The expression levels of four major TPM genes (TPM1, TPM2, TPM3 and TPM4) in the RCC tumor tissues and paired adjacent normal renal tissues from the same patients were detected by RT-qPCR. The mRNA levels of TPM2-4 had no significant differences in expression in the 33 RCC tumor tissues compared to these levels in the adjacent normal renal tissues (p=0.591, p=0.104 and p=0.191, respectively, Fig. 1A and B). However, compared with the normal kidney tissues, the RCC tumor tissues exhibited significantly lower TPM1 expression at the mRNA level. Even after the sample number was enlarged to 44 pairs, the statistical difference remained significant (p<0.05, Fig. 1A and B).

TPM1 protein levels in the primary RCC tumor tissues were further confirmed by western blotting. The results of the western blotting of the 40 paired RCC samples were consistent with the results of the RT-qPCR. RCC tumor tissue had significantly lower TPM1 expression at the protein level when compared with levels in the normal kidney tissues (p<0.01, Fig. 1C).

We then evaluated TPM1 protein expression at the tissue level using immunohistochemical staining of 25 tissue sections, including 25 RCC tissues and 24 paired tumor adjacent tissues (one section only had tumor tissue). In the tumor adjacent tissues, TPM1 was mainly expressed in the cytoplasm of renal tubular cells rather than the nucleus or the renal corpuscle. TPM1 was mainly expressed in the tumor adjacent normal kidney tissues, yet not in the RCC tissues. TPM1 expression was even lost in many highly developed RCC sections. Since ccRCC dominated the majority of RCC cases, we selected one microscopic sample of ccRCC as representative (Fig. 1D). IRS of TPM1 in the RCC tissues was significantly different than that in the adjacent normal renal tissues (p<0.001, Fig. 1D).

There was no significant difference between age, gender, pathologic type, T and N stage, metastasis, recurrence and alcoholic consumption and TPM1 expression levels among all the patients involved in the RT-qPCR analyses (Table II). TPM1 mRNA expression was significantly decreased in the various types of RCC cases; in other words, there was no significant difference in decreased TPM1 levels among the various RCC subgroups. However, tumor size was significantly associated with the TPM1 mRNA level (p=0.012, Table II). RCC patients with tumor size >4.5 cm had a higher chance to have a downregulated TPM1 level. In addition to tumor size, the association between smoking incidence and TPM1 was also observed (p=0.031, Table II). The markedly decreasing (>2-fold) rate of TPM1 mRNA in the primary RCC cases was 59.1% (26/44).

TPM1 IHC staining status in the 25 tissue sections was systematically evaluated. We analyzed the statistical significance of the IRS variances compared to age, gender, grade and stage of the RCC cases. The results indicated that the average IRS in the normal kidney tissues in male and female patients were significantly different from each other (p=0.034, Table III), yet the difference in the IRS for TPM1 in the RCC tumor tissues according to gender was not significant. Notably, the IRS in the normal kidney tissues was significantly decreased with increasing Fuhrman nuclear grade while the IRS in the RCC tissues was significantly increased according to increasing Fuhrman nuclear grade (p=0.019 and p=0.009, Table IV). However, a similar phenomenon was not observed at different clinical stage or age for IRS.

We followed up 42 of these patients for >4 years after surgical intervention in the present study. One patient from the TPM1 increase group passed away from an unrelated cause (myocardial infarction) 629 days after surgery. Thus, this case was treated as a censored observation in the disease-specific survival rate analysis while as a death observation in the overall survival rate analysis. During the follow-up, 7 patients (7/27, 25.9%) passed away in the TPM1 decrease group while one unrelated death (1/15, 6.7%) occurred in the TPM1 increase group. No significant difference was observed in regards to the overall survival rate (p=0.137, Fig. 2A), yet a significant difference in the disease-specific survival rate was noted between the TPM1 decrease group and TPM1 increase group (p=0.039, Fig. 2B).

TPM1 also played a role in RCC cell motility. In the scratch assay, the TPM1-transfected OSRC-2 and 786-O cells exhibited significantly slower wound healing compared with the empty plasmid transfected and the mock treatment group. In regards to the OSRC-2 cells, the wound healing status between the experimental and the negative control group showed a significant difference after 48 h (p=0.0292, Fig. 3B and C). The same observation was made in the 786-O cells until 24 h after scratching (p=0.0016, Fig. 4B and C); a near complete wound closure was observed after 48 h. This was probably due to the weaker migratory ability of the OSRC-2 compared to the 786-O cells.

Correspondingly, the Transwell migration assay demonstrated that overexpression of TPM1 was associated with markedly inhibited migration in both the OSRC-2 and 786-O cells (p=0.0013 and p=0.0039 separately, Figs. 3E and 4E). Cell migration was detected by crystal violet staining and microscopic graphs 24 h after seeding into the Transwell inserts. In regards to the OSRC-2 cells, the number of migrated cells/microscopic field (x200) for the pc3.1-NC and pc3.1-TPM1 groups was 124.6±2.112 and 110.8±1.934, respectively. In regards to the 786-O cells, the two values for each group were 125.0±2.168 and 108.2±3.597, respectively.

For the Transwell invasion assay, the cells needed to invade the Matrigel matrix and then reach the filter bottom which was stained and then images could be captured. Compared with the negative control group, the invasive abilities of the OSRC-2 and 786-O cells in the TPM1-transfected group were significantly reduced (p=0.0155 and p=0.0411 for each cell type, Figs. 3F and 4F). In regards to the OSRC-2 cells, the numbers of invasive cells/microscopic field (x200) in the pc3.1-NC and pc3.1-TPM1 groups were 119.0±6.928 and 91±5.958, respectively. In regards to the 786-O cells, the two values for each of the pc3.1 groups were 122.4±2.502 and 112.2±3.367, respectively.

We then detected apoptosis in the OSRC-2 and 786-O cells following TPM1 transfection. For these two cell lines, we used flow cytometry based on Annexin V and PI staining. The results demonstrated that TPM1 transfection promoted cell apoptosis in these two RCC cell lines (Figs. 5 and 6). At 72 h after transfection, the early stage apoptosis rate (Annexin V⁺/PI⁻) following TPM1 transfection was 4.1±0.2 and 3.8±0.2% for the OSRC-2 and 786-O cells, respectively. This was significantly higher than the negative control: 1.7±0.4 and 2.4±0.4%, respectively (p=0.0051 and p=0.0361, Figs. 5B and 6B). In addition, the total apoptosis rate (Annexin V⁺/PI^±) following TPM1 transfection was 6.0±0.2 and 5.8±0.3% for the OSRC-2 and 786-O cells, respectively. This was also significantly higher than the empty vector transfection: 2.3±0.5 and 4.2±0.4%, respectively (p=0.0022 and p=0.0318, Figs. 5C and 6C).