Tumor Biol. DOI 10.1007/s13277-016-5350-1

ORIGINAL ARTICLE

FABP5 correlates with poor prognosis and promotes tumor cell growth and metastasis in cervical cancer Wei Wang 1 & Hong-juan Chu 2 & Yan-chun Liang 1 & Jia-ming Huang 1 & Chun-liang Shang 1 & Hao Tan 1 & Duo Liu 1 & Yun-he Zhao 1 & Tian-yu Liu 1 & Shu-zhong Yao 1

Received: 30 January 2016 / Accepted: 6 September 2016 # International Society of Oncology and BioMarkers (ISOBM) 2016

Abstract Fatty acid-binding protein 5 (FABP5) was found in our previous study to be a potential biomarker for lymph node metastasis of cervical cancer. However, the roles of FABP5 in cervical cancer remain unclear. In the present study, FABP5 expression was found to be significantly upregulated in cervical cancer tissues, and high FABP5 expression was significantly correlated with lymph node metastasis, lymphovascular space invasion, the International Federation of Gynecology and Obstetrics (FIGO) stage, and tumor size. Moreover, FABP5 was an independent factor for poor prognosis in cervical cancer patients. Silencing of FABP5 inhibited cell proliferation, colony formation, cell migration, and invasion in vitro. Furthermore, FABP5 silencing significantly reduced tumor growth and lung metastases in a murine allograft model in vivo. In addition, FABP5 silencing decreased the expression of matrix metalloproteinase-2 (MMP-2) and matrix metalloproteinase-9 (MMP-9) in vitro and in vivo. Collectively, these findings indicated that FABP5 plays an important role in the carcinogenesis and metastasis of cervical cancer, and FABP5 may be a novel predictor for prognostic assessment of cervical cancer patients.

Keywords Cervical cancer . Metastasis . Prognosis . FABP5 Wei Wang and Hong-juan Chu contributed equally to this work. * Shu-zhong Yao [email protected] 1

Department of Obstetrics and Gynecology, the First Affiliated Hospital, Sun Yat-sen University, Zhongshan Second Road 58, Guangzhou 510700, China

2

Department of Otorhinolaryngology Head and Neck Surgery, the Third Affiliated Hospital of Southern Medical University, Guangzhou, China

Introduction Cervical cancer is the second most frequently diagnosed cancer and third leading cause of cancer death among females in less developed countries [1]. In order to reduce its mortality, great efforts have been made to understand the molecular mechanisms underlying carcinogenesis and malignant dissemination of cervical cancer [2–4]. For elucidating these complicated molecular processes, it is crucial to identify the molecular involved in the malignant progression of cervical cancer. By using tissue-based proteomics, we have identified some differentially expressed proteins between primary cervical cancer tissues with and without lymph node metastasis (LNM) [5]. Through validation with an independent tissue cohort, three proteins were characterized as potential biomarkers for lymph node metastasis of cervical cancer. One such protein is the fatty acid-binding proteins 5 (FABP5). FABP5 is a small (15 kDa) cytoplasmic protein and is a member of the intracellular lipid-binding protein family. It is thought to be involved in the uptake and transport of longchain fatty acids. In addition, they play important roles in cell signaling, gene regulation, cell growth, and differentiation [6]. Several recent studies have reported that FABP5 was upregulated in some cancers [7–10]. In our previous work, FABP5 was one of the candidate biomarkers for predicting LNM in cervical cancer, which exhibited significantly expressed differences between cervical cancer with and without lymph node metastasis [5]. However, the distinctive molecular role of FABP5 in tumorigenesis and metastasis of cervical cancer and the prognostic significance of FABP5 in cervical cancer are elusive. The aims of this study were to examine the molecular function of FABP5 in the carcinogenesis and metastasis of cervical cancer and to investigate the prognostic significance of FABP5 in cervical cancer. In this study, we firstly investigated

Tumor Biol.

the expression of FABP5 in cervical cancer tissues and cell lines. Then, we examined its roles in cervical cancer cell proliferation, migration, invasion, and metastasis by regulating its expression levels in cancer cell lines in vitro and in vivo. Finally, the prognostic significance of FABP5 was investigated in a tissue cohort of cervical cancer.

Materials and methods Patients and tissue specimens Informed consent was obtained from each patient. The study was approved by the Ethical Committee of the First Affiliated Hospital of Sun Yat-sen University (Guangzhou, China). For immunohistochemistry (IHC) assessment, a tissue cohort with 206 cervical squamous cell cancer samples was recruited. This cohort of paraffin-embedded tissue was collected from the archives of the pathology department at the First Affiliated Hospital of Sun Yat-sen University. All enrolled cervical cancer patients were stage IA2–IIA2 with available follow-up data. Surgeries were performed from January 2001 to January 2010. None of the patients received radiotherapy and/or chemotherapy before surgery. As a control group, normal cervical tissues (NCTs) were collected from 40 women who underwent simple hysterectomy for uterine leiomyoma at our hospital. All specimens were handled according to legal and ethical standards. Cell line culture All cervical cancer cell lines were purchased from American Type Culture Collection (ATCC, MD, USA). The cervical cancer cell lines HeLa, Siha, HCC94, and C33A were cultured in Eagle’s minimum essential medium (Gibco BRL, Rockville, MD), and Caski cells were cultured in RPMI1640 medium (Gibco BRL, Rockville, MD) in a 5 % CO2humidified atmosphere at 37 °C. Media were supplemented with 10 % fetal bovine serum (HyClone Laboratories, Logan, UT), 100 μg/ml streptomycin, and 100 U/ml penicillin. RNA extraction, reverse transcription, and quantitative PCR Total RNA from cultured cells and xenografts were extracted with RNAiso Plus reagent (Takara, Dalian, China) and transcribed into complementary DNA by using PrimeScript RT Master Mix (Takara, Dalian, China) according to the manufacturer’s instructions. Quantitative PCR (qPCR) was used to examine FABP5, matrix metalloproteinase-2 (MMP-2), and MMP-9 messenger RNA (mRNA) levels. All qPCRs were run in triplicate with a 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA). Expression data were

normalized to the geometric mean with reference to glycerol-3-phosphate dehydrogenase (GAPDH). Construction of the shRNA lentiviral expression vector and lentiviral infection Four RNA interference candidate target sequences for human FABP5 and negative control (NC) sequence were designed by GeneCopoeia, Co, Ltd. (Guangzhou, China). The short hairpin RNA (shRNA) sequences were as follows: shRNA#1, 5′AAGCCAGATTGTATCATCA-3′; shRNA#2, 5′-GTGC AT T G G T T C A G C AT C A - 3 ′ ; s h R N A # 3 , 5 ′ - T G A A G A A A C C A C A G C T G AT- 3 ′ ; s h R N A # 4 , 5 ′ - T G G A G T G T G T C AT G A A C A A - 3 ′ ; a n d N C , 5 ′ - G C C A GCTTAGCACTGATAC-3′. All oligonucleotides encoding shRNA sequences were synthesized and annealed into double strands. Then, double-stranded DNAs were inserted into pGLV3/H1/RFP lentiviral frame plasmids (GeneCopoeia Co, Ltd). All of the constructed plasmids were validated by using DNA sequencing. Lentiviral vector DNAs and packaging vectors were co-transfected into 293T cells. Lentiviral particles were harvested from supernatants 48 h later. We have performed subsequent purification by using ultracentrifugation, and the titer of lentivirus was examined. For transfection, 2 ml of the cell suspension was seeded into a 24-well plate at a density of 5 × 104 cells/well for 24 h. The cells were then transfected with lentivirus-mediated shRNA-targeting FABP5 or lentivirus-mediated negative shRNA (NC). The multiplicity of infection was 50. The interference efficiency was examined by qPCR after transfection for 72 h. Western blotting Protein lysates from xenografts and cell lines were obtained by using cold RIPA buffer. Proteins (50 μg) were fractionated with 10 % sodium dodecyl sulfate-polyacrylamide gel elect r o p h o r e s i s ( S D S - PA G E ) a n d t r a n s f e r r e d t o a polyvinylidenedifluoride membrane as previously described [5]. The signal was examined by using an enhanced chemiluminescence (ECL) plus kit (Millipore, Bedford, MA, USA). Rabbit anti-FABP5 (1:1000), anti-MMP-2 (1:800), antiMMP-9 (1:800), and anti-GAPDH (1:1000) were used. All of these antibodies were purchased from Abcam, Inc. (Cambridge, MA, USA). Immunohistochemistry Immunohistochemistry staining was performed by using antiFABP5 (1:600) as previously described [5]. The antibodies are the same as that used in Western blotting. Primary antibodies were omitted in negative controls. The results were scored by adding the staining area (0 = no, 1 < 30 %, 2 = between 30 and 60 %, 3 = between 60 and 100 % stained cells) and the

Tumor Biol.

staining intensity (0 = no, 1 = weak, 2 = moderate, 3 = strong staining). The immunostaining scores (ranging from 0 to 6) were evaluated by two experienced pathologists in a blinded fashion. A staining score of 3 was defined as the cutoff. The samples were divided accordingly into low and high staining groups.

plate were trypsinized, and the numbers of cells were detected with cell counting. Assays were done by using MMP-2 and MMP-9 ELISA kits (R&D Systems, UK) following the manufacturer’s instructions. Results were normalized to the cell number.

Cell proliferation assay Enzyme-linked immunosorbent assay Cells were seeded into 6-well plates. The culture media were collected and centrifuged after 24 h of seeding. The cells in the

Fig. 1 FABP5 protein and mRNA expression determined by immunohistochemical assay, Western blotting, and real-time PCR. a Representative results of immunohistochemical staining of FABP5 in early stage cervical cancer and normal cervical tissue. (NCT normal cervical tissues, LNM lymph node metastasis, NLNM no lymph node

Tumor cells were seeded into 96-well plates at a density of 5000 cells per well. Then, CCK-8 solution (Dojindo, Japan) was carefully added to plates (10 μl per well). The plates were

metastasis). b Western blotting assay of FABP5 expression in cervical cancer cell lines and normal cervical tissues (NCT). c Real-time PCR assay of FABP5 expression in cervical cancer cell lines and normal cervical tissues (NCT). *P < 0.05

Tumor Biol.

incubated for 1 to 4 h at 37 °C. The absorbance was periodically detected at 450 nm by a microplate reader. Colony formation assay Cells were plated into 6-cm culture dishes (100 cells per dish). Then, medium containing 10 % FBS was added to culture dishes (4 ml per well). After 2 weeks, the cells were fixed for 10 min with methanol and stained for 30 min with 0.1 % crystal violet. Visible colony and the rate of colony formation were counted. Invasion assay Invasion assays were done by using 24-well transwells (Costar, Cambridge, MA, USA) with polycarbonate filters (8-μm pore size). The transwell filters were coated with Matrigel (BD Biosciences, Bedford, MA, USA). Cells (2 × 104) were seeded onto the top side of transwells and incubated for 24 h at 37 °C. Invaded cells on the lower membrane surface were fixed with 4 % paraformaldehyde, stained with 0.1 % crystal violet. The invaded cells were counted in ten random fields at ×100 magnification under a light microscope.

artificial wound was created with a sterile 200-μl pipette tip. After scratching, the cells were washed with serum-free medium. Images of cell migration into the scratched region were obtained at 0, 24, and 48 h with an inverted microscope. Tumor growth analysis in mice All animal experiment was approved by the Animal Ethical and Welfare Committee of Sun Yat-sen University. Twelve female athymic nude mice (4–6 weeks of age) were purchased from Guangdong Medical Laboratory Animal Center (Guangzhou, China) and randomly divided to three groups. Mice were maintained and bred under SPF circumstances in the Department of Animal Center, the First Affiliated Hospital, Sun Yat-sen University. Aliquots of 1 × 107 cells were inoculated subcutaneously into the dorsal rear flanks of nude mice. After injection, tumor size was measured by external caliper, and tumor volumes were calculated according to the formula as follows: V = (length / 2) × (width2) [11]. Mice were sacrificed on 35 days after the cell injection. Xenograft tumors were dissected from each mouse, weighed. Total protein and RNA were extracted from the primary xenografts to examine FABP5, MMP-2, and MMP-9 expression levels with real-time PCR and Western blotting.

Wound healing assay Experimental metastasis assay Cells (5 × 105 cells per well) were seeded into 6-well culture plates. When cell density reached approximately 90 %, the serum was depleted from the cell culture medium. Cells were starved for 24 h. When the cell confluence reached 100 %, an

Fig. 2 Stably transfected Siha cell lines (×100). a Siha negative control (white light). b Siha negative control (fluorescence). c Siha-FABP5-RNAi (white light). d Siha-FABP5-RNAi (fluorescence)

Fifteen female athymic nude mice (4–6 weeks of age) were purchased from Guangdong Medical Laboratory Animal Center (Guangzhou, China) and randomly divided to three

Tumor Biol.

groups. Aliquots of 1 × 106 (0.1 ml) tumor cells were injected through the lateral tail vein. At 30 days after injection, all mice were sacrificed and their lungs were removed. The number of surface metastases per lung was counted. For further validation, the lung tissue samples were trimmed, processed, embedded, sectioned, and subjected to hematoxylin and eosin (H&E) staining to detect the lung metastasis. Follow-up After treatment completion, cervical cancer patients were followed up every 3 months in the first 2 years, every 6 months in the next 3 years, and annually thereafter. The enrolled cervical cancer patients in the present study were followed up between January 2004 and January 2015. Overall survival (OS) time was measured from the date of surgery to the date of death or censoring. Recurrence-free survival (RFS) time was calculated from the date of surgery to the date of recurrence or censoring.

in the epithelial cell layers. However, basal cell layers were more immunoreactive (Fig. 1a).

Elevated expression of FABP5 in cervical cancer cell lines Real-time PCR (RT-PCR) and Western blotting analyses showed elevated expression levels of FABP5 mRNA and protein in all five cervical cancer cell lines compared to NCTs (Fig. 1b, c). Because Siha expressed the highest level of FABP5 among these cervical cancer cell lines, we chose Siha cells to perform the following experiments.

Statistical analysis Statistical analyses were performed by using SPSS 13.0 (SPSS Inc., Chicago, IL, USA) and MedCalc statistical software (MedCalc Software, Mariakerke, Belgium). Independent-sample t test was used to compare the difference between two groups. The relationships between FABP5 expression level and clinicopathological factors were evaluated by the Pearson χ2 test or Fisher’s exact test. Multivariate Cox’s proportional hazard model analysis was used to determine the independent prognostic factors of cervical cancer patients. The Kaplan–Meier method was used to construct relapse-free probability curves and overall survival curves, and the curves were compared by the log-rank test. P < 0.05 was regarded as statistically significant.

Results Elevated expression of FABP5 in cervical cancer tissue We detected FABP5 protein expression in 206 samples of cervical cancer tissues (41 LNM, 165 no LNM) and 40 samples of NCTs. IHC analysis showed that FABP5 expression was significantly higher in cervical cancer tissues compared with NCTs. FABP5 expression occurred in primary cervical cancer tissues. Strong immunoreactivity was concentrated primarily in the nuclei and cytoplasm of cervical cancer cells. Semi-quantitative analysis indicated that the LNM group had significantly higher expression of FABP5 than the no lymph node metastasis (NLNM) group. In NCT, FABP5 was weakly expressed

Fig. 3 Western blotting and RT-PCR assay of FABP5 protein and mRNA expression in cells. a The knockdown effect of lentivirus-mediated shRNA vectors (HSH005077-1-LVRH1MP (1), HSH005077-2LVRH1MP (2), HSH005077-3-LVRH1MP (3), HSH005077-4LVRH1MP (4)) compared with untransfected Siha (Siha) and negative control (NC) group. Error bars correspond to means (SD) for each experiment in three replicates. b Western blotting of FABP5 expression in untransfected SiHa (Siha), Siha-NC (NC), and Siha-FABP5-RNAi (RNAi) cell. c Real-time PCR assay of FABP5 expression in untransfected Siha (Siha), SiHa-NC (NC), and SiHa-FABP5-RNAi (RNAi) cells. **P < 0.01 and *P < 0.05

Tumor Biol. Fig. 4 Downregulation of FABP5 inhibits cell proliferation, colony formation, and tumorigenicity of cervical cancer cells. a A clear inhibitory effect of cell proliferation was observed from the second day after FABP5 knockdown. b, c Siha-FABP5RANi (RNAi) cells had lower rate of colony formation than Siha and Siha-NC (NC) groups. d Tumor growth curves in nude mouse xenograft models. e The tumor weights of xenograft tumor. f The photograph of primary xenograft tumor. *P < 0.05

Downregulation of FABP5 expression by lentiviral vector Four lentiviral vectors (HSH005077-1-LVRH1MP, HSH005077-2-LVRH1MP, HSH005077-3-LVRH1MP, HSH005077-4-LVRH1MP) were transfected into Siha cells. Stably transfected Siha cell lines were constructed (Fig. 2). After transfection for 72 h, RT-PCR was done to assess the knockdown efficiency of lentivirus-mediated FABP5-specific shRNA. As shown in Fig. 3a, compared with negative control (Siha-NC) cells, FABP5 expression was significantly decreased

Fig. 5 Cell migration, invasion, and metastasis were inhibited by FABP5„ knockdown, respectively. a Cell migration was measured by wound healing assay at 0, 24, and 48 h (×200). b Cell invasion was measured by transwell invasion assay (×200). c Bar graphs represent the cell numbers adherent in the lower membrane. d Data of metastatic nodules per lung from five mice in each group are presented as the mean ± SD. e Representative pictures of lung tissues taken at the time of sacrifice after 30 days of injection through the tail vein. f Representative H&E-stained lung sections. (Siha negative control (NC), Siha-FABP5-RNAi (RNAi), untransfected Siha (Siha))

Tumor Biol.

Tumor Biol.

(72.5 % decrease) in HSH005077-2-LVRH1MP transfected cells (P < 0.01). We selected HSH005077-2-LVRH1MP to do the following experiments. There was no significant difference between untransfected Siha and Siha-NC cells. Western blotting and RT-PCR confirmed that both FABP5 mRNA and protein expression levels were significantly downregulated in SihaFABP5-RNAi cells compared to Siha-NC cells and untransfected Siha cells (Fig. 3b, c). Downregulation of FABP5 inhibits cell proliferation and colony formation To investigate the effect of FABP5 on cervical cancer cell proliferation, CCK-8 proliferation assay and colony formation assay were performed. The results indicated that the stably transfected Siha-FABP5-RNAi cells had lower proliferative ability, compared with untransfected Siha or Siha-NC cells (Fig. 4a). Consistent with the CCK-8 proliferation assay, the ability of colony formation in Siha-FABP5-RNAi cells was

Fig. 6 Downregulated FABP5 inhibits MMP-2 and MMP-9 expression in vivo and in vitro. a, b MMP-2 and MMP-9 expression in cell culture by ELISA. c, d Real-time PCR assays of MMP-2 and MMP-9 expression in primary xenografts. e Western blotting assays of FABP5, MMP2, and MMP-9 expression in primary xenografts. (SihaFABP5-RNAi groups (RNAi), Siha-NC (NC), untransfected Siha (Siha)). f Column graph shows the relative gray level of each band compared with GAPDH. *P < 0.05

significantly lower than Siha or Siha-NC cells (Fig. 4b, c). These results suggest that FABP5 is essential for cervical cancer cell proliferation.

Downregulation of FABP5 inhibits cell migration and invasion To assess the potential role of FABP5 in regulating cell migration and invasion, we performed a wound healing assay and transwell invasion assay. Wound healing assay indicated a slow-healing effect in Siha-FABP5-RNAi cells compared with Siha and Siha-NC cells (Fig. 5a). A transwell invasion assay showed that the number of cells in the lower chamber significantly decreased in SihaFABP5-RNAi group compared with Siha and Siha-NC groups (Fig. 5b, c). The above results showed that the downregulation of FABP5 inhibits the cervical cancer cell migration and invasion in vitro.

Tumor Biol.

Downregulation of FABP5 inhibits tumorigenicity of cervical cancer cells

Correlation of FABP5 expression with postoperative relapse and survival

We further investigated the effects of FABP5 on the tumorigenesis of cervical cancer cells. Tumors in nude mice formed from Siha-FABP5-RNAi cells grew more slowly than those from Siha-NC and Siha cells (Fig. 4d). After 35 days, the weights of tumors induced by Siha-FABP5-RNAi cells were dramatically reduced compared with those from Siha and Siha-NC cells (Fig. 4e, f). Taken together, these results suggested that FABP5 plays an important role in the tumorigenicity of cervical cancer cells in vivo.

At the end of the study, 54 patients (26.21 %) experienced disease recurrence, 49 patients (23.79 %) had died of cervical cancer, and 157 patients (76.21 %) were still alive. In multivariate Cox’s proportional hazard model analysis, age, LNM, tumor size, and upregulation of FABP5 were found to be the independent prognostic Table 1 Associations of FABP5 expression levels with clinicopathological factors in cervical cancer

Downregulated FABP5 suppresses distant metastasis of cervical cancer cells To further assess the effects of FABP5 on metastasis, different group cells were injected into the lateral tail vein of nude mice. After 30 days, the Siha and Siha-NC groups formed 10 to 18 metastatic nodules per lung in all ten mice. In contrast, mice injected with the same amount of Siah-FABP5-RNAi cells formed zero to four metastatic nodules per lung (Fig. 5d, e, f). These results indicated that downregulated FABP5 expression reduced the metastasis ability of cervical cancer cells in vivo. Downregulated FABP5 inhibits MMP-2 and MMP-9 expression in vivo and in vitro The levels of MMP-2 and MMP-9 in the culture medium of Siha-FABP5-RNAi cells were significantly low compared with Siha and Siha-NC cells (Fig. 6a, b). Moreover, the MMP-2 and MMP-9 expression levels of the primary SiHa-FABP5-RNAi xenografts were much lower than Siha and Siha-NC xenografts (Fig. 6c, d, e). Collectively, these data showed that downregulation of FABP5 reduced MMP-2 and MMP-9 expression in vivo and in vitro. Correlation of FABP5 with clinicopathological factors in cervical cancer To assess the association between FABP5 protein expression and clinicopathological factors in cervical cancer, the expression of FABP5 was tested by IHC in 206 paraffin-embedded cervical squamous cell cancer tissues. FABP5 was significantly correlated with LNM (P < 0.001), LVSI (P < 0.001), FIGO stage (P < 0.001), and tumor size (P = 0.011). No significant correlations were identified between the expression of FABP5 and other clinicopathological characteristics, including age, tumor differentiation, parametrial invasion, and vaginal invasion (Table 1).

Number

FABP5

206

Low (0–3)a 136

High (4–6)a 70

46 160

35 101

11 59

25 65

25 39

0 26

50 30 36

33 24 15

17 6 21

Age (years) ≤35 >35 FIGO stage IA2 IB1 IB2 IIA1 IIA2 Tumor size (cm) ≤4 >4 LNM Positive Negative LVSI Positive Negative Differentiation G1 G2 G3 Stromal invasion ≤1/2 >1/2 Parametrial invasion Positive Negative Vaginal invasion Positive Negative

P

0.102

0

0.026 122 84

88 48

34 36

41 165

13 123

28 42

0

0.001 25

9

16

181

127

54 0.940b

12 72 122

8 49 79

4 23 43

98 108

70 66

28 42

13 193

8 128

5 65

31 175

18 118

13 57

0.118

0.766

0.411

FIGO the International Federation of Gynecology and Obstetrics, LVSI lymphovascular space invasion, LNM lymph node metastasis a

The immunostaining scores of FABP5

b

P value from Fisher’s exact test

Tumor Biol. Table 2 Multivariate Cox’s proportional hazard model analysis of relapse-free and overall survivals

Variables

Overall survival

Relapse-free probability

HR (95 % CI)

P

HR (95 % CI)

P

Age (year)

1.084 (1.049–1.122)

0.000

1.089 (1.055–1.124)

0.000

Tumor size (cm)

1.977 (1.031–3.790)

0.000

1.356 (1.108–1.660)

LNM Negative (reference)

0.000 1

1

Positive

5.438 (2.824–10.473)

6.122 (3.232–11.596)

0.003 0.000

FABP5 Low (reference)

0.040

0.030

1

1

High

1.977 (1.031–3.790)

2.012 (1.068–3.790)

For the stepwise multivariate analysis, backward LR method was used to drop insignificant variables. Variables entered for analysis were the following: age, FIGO stage, tumor size, LNM, LVSI, differentiation, stromal invasion, vaginal invasion, parametrial invasion, and FABP5 HR hazard ratio, 95 % CI 95 % confidence interval, LNM lymph node metastasis

factors (Table 2). The survival curves showed that the overall survival rate decreased significantly with upregulation of FABP5 (Fig. 7b). The recurrence-free probability curves showed that the recurrence rate increased significantly with upregulation of FABP5 (Fig. 7a).

Discussion In this study, our results indicate that FABP5 plays an important role in cervical cancer carcinogenesis and metastasis. First, FABP5 expression was significantly higher in cervical cancer cells and clinical cervical cancer specimens than in NCT. Downregulation of FABP5 expression in Siha cells significantly inhibits its proliferation in vitro and tumor growth in nude mice. Second, FABP5 expression was significantly higher in cervical cancer with LNM than in cervical cancer without LNM. The metastasis-promoting activity is shown by downregulation expression of FABP5 that inhibits Siha cell migration and metastasis shown in both cell mobility assay in vitro and experimental metastatic assay in vivo. Furthermore, downregulation of FABP5 reduced MMP-2 Fig. 7 Kaplan–Meier survival plots for early stage cervical cancer according to the expression levels of FABP5. a FABP5 expression and recurrence-free probability (logrank test, P < 0.0001). b FABP5 expression and overall survival (log-rank test, P = 0.0003)

and MMP-9 expression in vitro and in vivo. Third, high FABP5 expression was significantly associated with LNM, LVSI, tumor size, and FIGO stage through a cohort of cervical cancer specimen (n = 206) assay. Moreover, high FABP5 expression was an independent prognostic factor for poor prognosis. These data indicate that FABP5 plays a crucial role in cervical cancer progression, including metastasis. Upregulation of FABP5 mRNA and protein expression in cervical cancer tissues and cell lines provides a clue suggesting that FABP5 may behave as a tumor promoter (Fig. 1). Furthermore, our results showed that downregulation of FABP5 expression inhibited cellular proliferation and tumorigenesis in vitro and in vivo (Fig. 4). In addition, the high level of FABP5 expression was significantly correlated with tumor size and FIGO stage (Table 1). These results indicated that FABP5 was associated with malignant progression in cervical cancer and might play an important role in the tumorigenesis of cervical cancer. These results are consistent with previous reports for prostate cancer [8], oral cancer [12], and breast cancer [13]. Intriguingly, the expression levels of FABP5 were significantly higher in cervical cancer tissues with LNM than without LNM (Fig. 1a). Moreover, high expression

Tumor Biol.

of FABP5 was an independent prognosis factor for poor prognosis (Table 2). To further investigate the biological significance of FABP5 in cervical cancer, we downregulated the expression of FABP5 in Siha cells by lentivirusmediated shRNA. Downregulation of FABP5 has the potential to inhibit cervical cancer cell migration and suppress metastasis activity (Fig. 5). It was reported that transfection of FABP5 gene into non-metastatic rat mammary epithelial cell line induced metastases in vivo [14, 15]. A recent study reported that FABP5 plays an important role in triple-negative breast cancer progression. FABP5 mediated its enhanced effect on triple-negative breast cancer metastasis by inhibiting EGFR proteasomal degradation [16]. Therefore, FABP5 may be a potential therapeutic target by downregulation of FABP5 expression to inhibit metastasis of tumor cells. MMP-2 and MMP-9 are members of metalloproteases family that can degrade extracellular matrix proteins. Previous studies demonstrated that the expression of MMP-2 and MMP-9 is crucial in the invasion and metastasis of cervical cancer [17]. In the present study, we found that downregulation of FABP5 inhibited expression of MMP-9 and MMP-2 in vitro and in vivo. These data suggest that FABP5 increase the aggressiveness of cervical cancer cells, possibly by regulation of MMP-2 and MMP-9. However, the exact molecular mechanisms underlying the role of FABP5 in carcinogenesis and metastasis need further research. In summary, our preliminary results suggest that FABP5 is correlated with poor prognosis in cervical cancer. Moreover, FABP5 plays an important role in cervical cancer cell proliferation, migration, invasion, and metastasis. Future studies should be done to elucidate the molecular mechanisms underlying the role of FABP5 in the carcinogenesis and metastasis of cervical carcinoma, as well as its utility as a potential marker for predicting poor prognosis in cervical cancer. Acknowledgments This work is supported by the Sun Yat-sen University Clinical Research 5010 Program (grant number 2007010), the Guangdong Natural Science Fund (grant number 2016A030310147, S2013010015448), and the Science and Technology Program of Guangzhou (grant number 201510010289).

References 1.

2.

3.

4.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Compliance with ethical standards Conflicts of interest None

17.

Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA. A Cancer Journal for Clinicians. 2015;65:87–108. Lin LF, Chuang CH, Li CF, Liao CC, Cheng CP, Cheng TL, et al. Zbrk1 acts as a metastatic suppressor by directly regulating mmp9 in cervical cancer. Cancer Res. 2010;70:192–201. Chen YF, Chiu WT, Chen YT, Lin PY, Huang HJ, Chou CY, et al. Calcium store sensor stromal-interaction molecule 1-dependent signaling plays an important role in cervical cancer growth, migration, and angiogenesis. Proc Natl Acad Sci U S A. 2011;108:15225–30. He SY, Shen HW, Xu L, Zhao XH, Yuan L, Niu G, et al. Foxm1 promotes tumor cell invasion and correlates with poor prognosis in early-stage cervical cancer. Gynecol Oncol. 2012;127:601–10. Wang W, Jia HL, Huang JM, Liang YC, Tan H, Geng HZ, et al. Identification of biomarkers for lymph node metastasis in earlystage cervical cancer by tissue-based proteomics. Br J Cancer. 2014;110:1748–58. Storch J, Corsico B. The emerging functions and mechanisms of mammalian fatty acid-binding proteins. Annu Rev Nutr. 2008;28:73–95. Jeong CY, Hah YS, Cho BI, Lee SM, Joo YT, Jung EJ, et al. Fatty acidbinding protein 5 promotes cell proliferation and invasion in human intrahepatic cholangiocarcinoma. Oncol Rep. 2012;28:1283–92. Adamson J, Morgan EA, Beesley C, Mei Y, Foster CS, Fujii H, et al. High-level expression of cutaneous fatty acid-binding protein in prostatic carcinomas and its effect on tumorigenicity. Oncogene. 2003;22:2739–49. Campos B, Centner FS, Bermejo JL, Ali R, Dorsch K, Wan F, et al. Aberrant expression of retinoic acid signaling molecules influences patient survival in astrocytic gliomas. Am J Pathol. 2011;178:1953–64. Liu RZ, Graham K, Glubrecht DD, Germain DR, Mackey JR, Godbout R. Association of fabp5 expression with poor survival in triple-negative breast cancer: implication for retinoic acid therapy. Am J Pathol. 2011;178:997–1008. Chen Y, Lin C, Liu Y, Jiang Y. Hmgb1 promotes hcc progression partly by downregulating p21 via erk/c-myc pathway and upregulating mmp-2. Tumour Biol. 2015. Fang LY, Wong TY, Chiang WF, Chen YL. Fatty-acid-binding protein 5 promotes cell proliferation and invasion in oral squamous cell carcinoma. J Oral Pathol Med. 2010;39:342–8. Levi L, Lobo G, Doud MK, von Lintig J, Seachrist D, Tochtrop GP, et al. Genetic ablation of the fatty acid-binding protein fabp5 suppresses her2-induced mammary tumorigenesis. Cancer Res. 2013;73:4770–80. Jing C, Beesley C, Foster CS, Rudland PS, Fujii H, Ono T, et al. Identification of the messenger rna for human cutaneous fatty acidbinding protein as a metastasis inducer. Cancer Res. 2000;60:2390–8. Jing C, Beesley C, Foster CS, Chen H, Rudland PS, West DC, et al. Human cutaneous fatty acid-binding protein induces metastasis by up-regulating the expression of vascular endothelial growth factor gene in rat rama 37 model cells. Cancer Res. 2001;61:4357–64. Powell CA, Nasser MW, Zhao H, Wochna JC, Zhang X, Shapiro C, et al. Fatty acid binding protein 5 promotes metastatic potential of triple negative breast cancer cells through enhancing epidermal growth factor receptor stability. Oncotarget. 2015;6:6373–85. Libra M, Scalisi A, Vella N, Clementi S, Sorio R, Stivala F, et al. Uterine cervical carcinoma: role of matrix metalloproteinases. Int J Oncol. 2009;34:897–903 .Review