ADFP promotes cell proliferation in lung adenocarcinoma via Akt phosphorylation

Previously, we identified differentially expressed proteins, including ADFP, between lung adenocarcinoma (LAC) tissue and paired normal bronchioloalveolar epithelium. In this study, we investigated the role of ADFP in LAC. ADFP levels in the serum of patients with lung cancer and benign diseases were measured by enzyme‐linked immunosorbent assays (ELISA). shRNA was used to knock‐down or overexpress ADFP in A549 and NCI‐H1299 cells. The biological function of ADFP and its underlying mechanisms was evaluated in vivo and in vitro. ADFP was highly expressed in the serum of lung cancer patients, especially those with LAC. ADFP promoted cell proliferation and up‐regulated the p‐Akt/Akt ratio in A549 and NCI‐H1299 cells in vitro. Furthermore, in nude mice, ADFP promoted tumour formation with high levels of p‐Akt/Akt, Ki67 and proliferating cell nuclear antigen (PCNA). Similar to the effect of ADFP knock‐down, MK‐2206 (a phosphorylation inhibitor of Akt) reduced A549 and NCI‐H1299 cell proliferation. In ADFP‐overexpressing A549 and NCI‐H1299 cells, proliferation was suppressed by MK‐2206 and returned to the control level. ADFP did not regulate invasion, migration or adhesion in LAC cells. Together, these results suggest that ADFP promotes LAC cell proliferation in vitro and in vivo by increasing Akt phosphorylation level.

fairly prevalent, occurring both as early and late events during tumour evolution, including amplifications in EGFR, MET, KRAS, ERBB2 and MDM2, and deletions in LRP1B, PTPRD and CDKN2A. Detection of these genomic alterations has substantially changed the setting of lung cancer diagnosis and treatment.
The transformation from normal tissues to malignant tumours is strictly regulated by the type, location and quantity of proteins that perform biological functions. Comparative proteomics is a powerful method for identifying a range of proteins that are differentially expressed between normal and tumour tissues. We previously purified LAC cells and normal bronchial alveolar epithelial cells from surgical LAC tissues and matched normal lung tissues, respectively, by laser-capture microdissection. The isobaric tags for relative and absolute quantification (iTRAQ) and two-dimensional liquid chromatography-tandem mass spectrometry (2D-LC-MS/MS) were used to screen the differentially expressed membrane proteins. A total of 568 proteins were identified, including ADFP (adipophilin, ADRP, PLIN2), which was up-regulated by 2.46-fold in LAC. 3 Further, we measured ADFP levels in 62 cases of primary lung cancer tissues and 24 cases of normal lung tissue by immunohistochemical (IHC) staining and found that ADFP was significantly highly expressed in LAC. 4 ADFP is called adipose differentiation-related protein and has a molecular weight of 50 kDa. The human ADFP gene is located in 9 p22.1 encoding 437 amino acids. ADFP is a membrane-associated protein belonging to the ancient PAT protein family and was originally considered to be a unique protein on the surface of fat droplets in fat cells. 5 Since then, ADFP has been detected in most tissue cells, such as breast, macrophages, liver, heart, intestinal mucosa, skeletal muscle, pancreas and others. ADFP is also involved in cellular fatty acid uptake, lipid droplet formation and lipid storage. Several conserved regions in the protein structure have the potential to bind lipids. [6][7][8][9][10][11][12] There is a dynamic balance between the decomposition and storage of fat in lipid droplets. Perilipin 5 (PLIN5) distributed on the surface of lipid droplets allows triglyceride lipase and hormone-sensitive lipase to enter the lipid droplets, promoting lipid degradation and free fatty acid consumption. ADFP hinders the entry of lipase into lipid droplets, resulting in the accumulation of lipids in lipid droplets.
PLIN5 and ADFP help to maintain equilibrium on the surface of lipid droplets. 13,14 In the absence of lipids to bind, ADFP is easily degraded by the proteasome; thus, ADFP is positively correlated with intracellular lipid content. Therefore, ADFP is used as an index for evaluating lipid accumulation under physiological and pathological conditions. Abnormal ADFP function and expression are also associated with a variety of diseases related to disordered lipid metabolism. 14,15 Recent studies showed that ADFP is highly expressed in Burkitt lymphoma, 16 20,22,23 Bioplasmonic paper-based assays that can quickly detect the content of ADFP in the urine of patients are expected to be widely used in clinical practice. 24 Cao et al knocked down ADFP in the human kidney cancer cell line A498 using an oligonucleotide sequence and observed enhanced proliferation, invasion and migration abilities. 23 Therefore, ADFP may also play a role as a tumour suppressor gene. Significantly, the role of ADFP in LAC remains unclear. In our previous research, the sample number of early-stage lung cancer patients was insufficient. Limitations of it included the heterogeneity of the population used and the small sample size. Thus, ADFP level in the peripheral blood of lung cancer patients, especially LAC patients, is still to be observed and more samples are needed to be included in the study.

| Cells culture
Human LAC cell lines A549 and NCI-H1299 were obtained from the Cell Bank of the Chinese Academy of Science (China). Both cell lines were cultured in RPMI1640 (Gibco) medium with 10% foetal bovine serum (Biological Industries, Israel) and maintained in a cell incubator (Thermo) at 5% CO 2 and 37°C.

| MK-2206
MK-2206, a phosphorylation inhibitor of Akt (Akt1, Akt2 and Akt3), was obtained from Selleck Chemicals. MK-2206 powders were dissolved by dimethyl sulfoxide (DMSO) and diluted to 1 mM (5% DMSO) by RPMI1640. The solution was stored at − 80°C. Where indicated, cells were exposed to MK-2206 after being cultured for 24 h, which was recorded as the start of the experiment.

| Real-time quantitative polymerase chain reaction (RT-qPCR)
Total RNA was extracted using an RNA Extraction Kit (Takara).

| CCK-8 assay
Cells were seeded in a 96-well plate (Corning) at a density of 6 × 10 3 cells/100 μL complete medium per well and cultured for 24-96 h.
Next, 10 μL of the CCK-8 reagent was added to each well and incubated for 1 h at 37°C. The optical density was measured using an absorbance microplate reader (Thermo, USA) at a wavelength of 450 nm.

| 5-Ethynyl-2'-deoxyuridine (EdU)
Cells were seeded in a 96-well plate (Corning) at a density of 3 × 10 3 cells/100 μL complete medium per well and cultured for 24 h, followed by serum starvation in serum-free medium for 24 h. Then, cells were labelled with EdU and cultured for 24 h. Apollo staining and DNA staining were performed according to the protocol described by the manufacturer of the EdU kit (Ribobio). Finally, cells were observed under a fluorescence microscope (Leica). The Hoechst33342 nuclear staining signal was excited by blue light, representing the viable cells containing DNA (total cells), and the Apollo staining signal was excited by red light, representing the cells with replicated DNA (proliferating cells).

| Plate colony formation
Cells were seeded in a 6-well plate (Corning) at a density of 200 cells with 3 mL complete medium per well and cultured for 10 days. Then, cells were fixed in methanol and stained with 0.1% crystal violet, and the number of colonies was counted.

| Flow cytometry
Cells were seeded in a 6-well plate (Corning) at a density of 1 × 10 5 cells with 3 mL complete medium per well and cultured for 24 h, followed by serum starvation in serum-free medium for 24 h. Then, cells were cultured in complete medium for 24 h and processed according to the protocol described by the manufacturer of the PI Cell Cycle Analysis Kit (7sea Biotech). Samples were analysed by a flow cytometer (BD).

| Transwell migration and invasion assays
Cells were seeded in a 24-well transwell insert (Millipore) that was either uncoated (migration assay) or coated (invasion assay) with Matrigel at a density of 1.5 × 10 5 cells (migration assay) or 5 × 10 5 cells (invasion assay) with 200 μL serum-free medium per transwell insert. Complete medium (800 μL) was added into each lower chamber of the 24-well plate (Corning) as an attractant. Cells were incubated for 24-48 h. Afterwards, cells in the upper chamber were removed, and cells that had migrated/invaded through the membrane were fixed in methanol and stained with 0.1% crystal violet.
The number of migrated/invaded cells was counted under a microscope (Nikon, Japan).

| Wound healing assay
Cells were seeded in a 6-well plate at a density of 1 × 10 5 cells with 3 mL complete medium per well and cultured until confluent. Then, cells were scratched using a 100 μL tip and cultured in serum-free medium for 24-72 h. The wound area was calculated and analysed under a microscope (Nikon, Japan).

| IHC staining
Tissues were dehydrated in a graded series of alcohol (50%, 70%, 85%, 95% and 100%) and dimethylbenzene, followed by paraffin embedding at 65°C. Blocks of tissues were cut into slices and dried on glass slides at 60°C for 12 h. Then, the slides were dewaxed with dimethylbenzene and gradient alcohol. Antigen retrieval was performed by microwaving sections in citrate buffer for 2 min. Slides The slides were stained with 3,3'-diaminobenzidine and haematoxylin and then washed with flowing water. Afterwards, the slides were dehydrated with gradient alcohol (70%, 80%, 90% and 100%) and dimethylbenzene. Finally, the slides were sealed with neutral balsam and coverslips.

| Statistical analysis
SPSS Statistics 23 (SPSS Inc) was used for all statistical analyses. Wilcoxon signed-rank, and Kruskal-Wallis tests were used to determine differences in serum ADFP levels among cancer and control groups. Data were presented as Median (Q1, Q3). The comparisons between two groups or among three groups were performed using a t test or ANOVA. The differences in the CCK-8 assay and tumour volume analyses were determined by repeated measures ANOVA. Data were presented as the mean ± SD. All statistical tests were two-sided, and values of P < .05 were considered significantly different.

| ADFP is up-regulated in the peripheral blood serum of patients with LAC
We first collected peripheral venous blood samples from 177 patients with lung cancer (including 79 LAC, 55 lung squamous carcinoma and 43 small-cell lung cancer) and 22 patients with benign lung diseases. The ADFP protein levels in serum were detected by ELISA. The results showed that the ADFP level was 91.84 ng/mL (77.38 ng/mL, 107.74 ng/mL) in the peripheral blood serum of patients with lung cancer and 80.63 ng/mL (48.44 ng/mL, 96.18 ng/ mL) in patients with benign diseases. The difference was statistically significant (P < .05). Compared with the benign disease group, the ADFP levels in the LAC and small-cell lung cancer groups were significantly higher (P < .05). There was no statistical difference between the lung squamous cell carcinoma group and the benign disease group (Table 1). We further compared ADFP expression among the three common pathological types of lung cancer. The ADFP level in small-cell lung cancer was higher than in LAC (Bonferroni corrected P < .05) and squamous cell carcinoma (Bonferroni corrected P < .05). There was no difference between LAC and lung squamous cell carcinoma.
The relationship between serum ADFP levels and the clinical characteristics of patients with LAC was analysed. Patients with regional lymph node invasion had higher ADFP serum levels than those without (P < .05). There was also no significant difference

| ADFP promotes LAC cell proliferation
The up-regulation of ADFP in the serum of clinical LAC patients suggested the potential tumour-promoting ability of ADFP in the progression of LAC. To evaluate the role of ADFP in LAC progression, we selected A549 and NCI-H1299 LAC cell lines for further research.
The role of ADFP was further confirmed by the increased colony formation ability of ADFP-overexpressing A549 and NCI-H1299 cells and the reduced colony formation ability of ADFP knockdown A549 and NCI-H1299 cells ( Figure 1D). The results of EdU experiments showed that the proportions of proliferating ADFP knock-down A549 and NCI-H1299 cells were decreased, whereas the proportions of proliferating ADFP-overexpressing A549 and NCI-H1299 cells were increased ( Figure 1E). Furthermore, ADFP induced S phase arrest in A549 and NCI-H1299 cells ( Figure 1F).
These data collectively suggested that ADFP played a tumour-promoting role in LAC cells.

| ADFP does not regulate the invasion, migration or adhesion of LAC cells
The roles of ADFP in invasion, migration and adhesion in LAC were further evaluated. In transwell assays, the migration of ADFP knock-down A549 cells was significantly suppressed. However, ADFP knock-down had no effect on the invasion of A549 cells, and ADFP overexpression did not affect the migration or invasion of A549 cells (Figure 2A,B). Similarly, there was no change in the migration or invasion of NCI-H1299 cells (Figure 2A,B). In addition, the results of wound healing assays showed no significant differences between any of the groups ( Figure 2C). The number of adherent ADFP knock-down A549 cells was decreased, but there was no significant change among other groups in adhesion assays ( Figure 2D).

| ADFP promotes tumour growth in vivo
The  NCI-H1299 cells. The expression of p-Akt protein was decreased in ADFP knock-down cells and increased in ADFP-overexpressing cells ( Figure 4A). There were no significant changes in Akt

| D ISCUSS I ON
Recent studies reported high levels of ADFP in the tissues and bodily fluids of patients with several malignant tumours. 20,21,25,26 ADFP was also an independent predictor of poor prognosis in patients with pancreatic ductal adenocarcinoma. 27 In clear cell renal cell carcinoma tissues, the transcript levels and protein expression of ADFP were higher than in normal renal tissues, 25 and its increased expression could be used as a diagnostic and prognostic factor for clear cell renal cell carcinoma. This study found significantly higher levels of ADFP in the serum of lung cancer patients, which was consistent with the results of tumour tissue analyses we previously conducted. 4 We further analysed three common pathological types of lung cancer. Compared with patients with benign lung diseases, serum ADFP levels in patients with LAC and small-cell lung cancer were significantly increased, but no statistical difference was found in lung squamous cell carcinoma. The ADFP level in small-cell lung cancer was higher than in LAC. The measurement of serum ADFP levels might be helpful for the diagnosis of LAC and small-cell lung cancer. Peripheral blood indicators have the advantages of quicker, more accessible and less trauma.
However, as a membrane protein, its level is not strictly linear in peripheral blood serum and intracellular. We should be cautious in judging whether an indicator can be used for diagnosis by its peripheral blood levels. The relationship between expression at cellular level and the concentration in the serum level needs to be further explored and better experimentally demonstrated.
Self-control specimens of primary LAC tissue and peripheral blood are the most anticipated, but difficult to be obtained. In LAC, the serum ADFP levels were not statistically different among patient groups with different genders, ages and smoking indexes, which is consistent with the results in our previous study. In LAC, ADFP was associated with regional lymph node invasion. There was difference of ADFP serum level between patients with or without lymph node metastases but no difference between patients with or without distal metastases. There is evidence that lymph node and distant metastases develop through fundamentally different evolutionary mechanisms. Lymph node metastasis and distant metastasis are not completely synchronized. 28 We consider that both of the arguments above need to be taken one stage further, for example, larger sample size, more detailed classification of clinicopathological characteristics and smaller population heterogeneity.
Previous studies found that the high ADFP level in urine and renal tissues was associated with a good prognosis in clear cell renal cell carcinoma patients, 20,22,23  F I G U R E 1 ADFP showed tumour-promoting ability in LAC cells. A and B, A549 and NCI-H1299 cells were stably transduced with ADFP knock-down (KD) and ADFP overexpression (OE) lentiviral vectors. Control groups (NC-KD and NC-OE) were transduced with the corresponding empty vector. Overexpression and knock-down of ADFP were confirmed by RT-qPCR and Western blotting. C, ADFP knock-down significantly reduced cell viabilities in the CCK-8 assay compared with those of control A549 and NCI-H1299 cells. ADFP overexpression significantly promoted cell viabilities in the CCK-8 assay compared with that of the control A549 and NCI-H1299 groups. D, ADFP knock-down significantly decreased colony numbers compared with those of control A549 and NCI-H1299 cells. ADFP overexpression significantly increased colony numbers compared with those of the control A549 and NCI-H1299 groups. E, ADFP knockdown significantly decreased the proportion of EdU-positive cells compared with that of control A549 and NCI-H1299 cells. ADFP overexpression significantly increased the proportions of EdU-positive cells compared with those of the control A549 and NCI-H1299 groups. Blue represents the cell nucleus, and red represents DNA replication. F, Cell cycle was analysed by flow cytometry. ADFP reduced the proportions of cells in the G1 phase and increased the fraction of cells in the S phase. Bars represent the SE of the mean ± SD from three independent experiments. *P < .05 and **P < .01 and ***P < .001 These results indicated that ADFP could promote LAC cell prolifera- p-Akt is the functional form of Akt, and p-Akt/Akt is often used to evaluate Akt phosphorylation levels. [30][31][32] Akt hyperphosphorylation can lead to the instability of the genome and even cancer formation. 33 Studies have shown that PI3K/Akt-mediated phosphorylation and lipid-mediated glycogen synthase kinase-3 (GSK3)/ADFP interactions for GSK3 regulation provided a link between energy homeostasis and cellular functionality. 34 Therefore, there might be a cross-talk between ADFP and PI3K/Akt pathways. PI3K activities are modulated by a variety of factors, including lipids. ADFP associated with HIF-2α promotes lipid storage and endoplasmic reticulum homeostasis under anaerobic conditions. 35 We consider that ADFP might help tumour cells survive hypoxia by enhancing anaerobic glycolysis.
The present study is the first comprehensive study to establish the functional role of ADFP in LAC cell proliferation and to identify a possible underlying mechanism. Our findings provide new mechanistic insight into the basic theory of LAC progression and identify a potential therapeutic strategy for LAC treatment.