GRK6 Depletion Induces HIF Activity in Lung Adenocarcinoma

G protein-coupled receptor kinase 6 (GRK6) is expressed in various tissues and is involved in the development of several diseases including lung cancer. We previously reported that GRK6 is down-regulated in lung adenocarcinoma patients, which induces cell invasion and metastasis. However, further understanding of the role of GRK6 in lung adenocarcinoma is required. Here we explored the functional consequence of GRK6 inhibition in lung epithelial cells. Analysis of TCGA data was coupled with RNA sequencing (RNA-seq) in alveolar epithelial type II (ATII) cells following depletion of GRK6 with RNA interference (RNAi). Findings were validated in ATII cells followed by tissue microarray analysis. Pathway analysis suggested that one of the Hallmark pathways enriched upon GRK6 inhibition is ‘Hallmark_Hypoxia’ (FDR = 0.014). We demonstrated that GRK6 depletion induces HIF1α (hypoxia-inducible factor 1 alpha) levels and activity in ATII cells. The findings were further confirmed in lung adenocarcinoma samples, in which GRK6 expression levels negatively and positively correlate with HIF1α expression (P = 0.015) and VHL expression (P < 0.0001), respectively. Mechanistically, we showed the impact of GRK6 on HIF activity could be achieved via regulation of VHL levels. Taken together, targeting the HIF pathway may provide new strategies for therapy in GRK6-depleted lung adenocarcinoma patients.


INTRODUCTION
G protein-coupled receptor kinases (GRKs) are a family of kinases that play a critical role in G protein-coupled receptors (GPCRs) homologous desensitization. GRKs phosphorylate specific serine and threonine residues of activated GPCRs which promote high affinity binding of arrestins and then suppress further G protein activation by interrupting receptor-G protein coupling (1)(2)(3). Desensitization of GPCRs has a critical role in maintaining homeostasis. As such, abnormal GPCRs desensitization can cause a variety of human diseases, including autoimmune diseases (4), asthma (5), heart failure (6), Parkinson's disease (7), inappropriate diuresis (8) and tumour progression and metastasis (9). Therefore, GRKs are important therapeutic targets for these diseases.
G protein-coupled receptor kinase 6 (GRK6) is a member of the GRK family, which is expressed in various tissues and involved in the development of several diseases (10)(11)(12). High expression of GRK6 has been reported in hepatocellular carcinoma (13), colorectal cancer (14); whilst lower expression was reported in hypopharyngeal squamous cell carcinoma (15) compared to normal tissues. Further, Grk6 knock out mice (Grk6 -/-) showed a significant increase in the growth and metastasis of Lewis lung cancer (LLC) compared to the control mice (Grk6 + / + ) (2). Our previous study suggested that GRK6 expression was significantly down-regulated in lung adenocarcinoma patients, and its level was an independent prognostic factor for overall survival (16). Moreover, we also showed that the promoter region of the GRK6 gene was hypermethylated in lung adenocarcinoma tissues compared to the normal tissue samples, leading to a down-regulation of GRK6 expression and in turn, inducing cell invasion and metastasis (17). However, further understanding of the role of GRK6 in lung adenocarcinoma is required.
In this study, we aimed to investigate the functional consequence of GRK6 depletion in lung epithelial cells. Analysis of TCGA data was coupled with RNA sequencing (RNA-seq) in alveolar epithelial type II (ATII) cells following the depletion of GRK6 with RNA interference (RNAi). Tissue microarrays were used to investigate the expression and function of GRK6 in lung adenocarcinoma. Our data suggests that GRK6 depletion induces HIF1a (hypoxia-inducible factor 1 alpha) activity. Targeting the HIF pathway may provide new strategies for therapy in GRK6depleted lung adenocarcinoma patients.
Short interfering RNA (siRNA) oligos against GRK6 or control siRNA were purchased from Biomics Biotechnologies Co., Ltd, China. Sequences are available from an earlier publication (17). Cells were transfected with the indicated siRNA oligos at a final concentration of 35 nM using Dharmafect 2 reagent (Dharmacon).

RNA Isolation, Library Construction, and Sequencing
To identify global transcriptomic changes in ATII cells upon GRK6 depletion, RNA sequencing (RNA-seq) was performed. In brief, ATII cells were transfected with either control siRNA or siRNA against GRK6 for 3 days. Total RNA was isolated using an RNeasy mini kit (Qiagen) according to the manufacturer's instructions and quantified using a Nanodrop Spectrophotometer 2000c (Thermo Fisher Scientific). A total amount of 3 µg RNA per sample was used as input material for library construction. Sequencing libraries were generated using NEBNext ® UltraTM RNA Library Prep Kit for Illumina ® (NEB, Ipswich, Massachusetts, USA) following the manufacturer's instructions. Libraries were pooled in equimolar and sequenced using the paired-end strategy (2 × 150) on the Illumina NovaSeq 6000 platform following the standard protocols (Novogene, UK). RNA-seq data have been deposited in the Gene Expression Omnibus (GEO) database (accession code GSE164921).

Data Mining GRK6 Related Data From the Cancer Genome Atlas (TCGA)
The expression of mRNAs in the TCGA lung adenocarcinoma (LUAD) (IlluminaHiSeq) dataset was obtained from the UCSC Xena Browser (https://xenabrowser.net/). To separate the low and high GRK6 group in the TCGA dataset, hierarchical cluster was performed on the high correlated genes with GRK6 via Pearson analysis in RStudio (version 3.4.4). According to the correlation analysis, there were 17 samples in the high GRK6 group and 26 samples in the low GRK6 group. Then, an unpaired t-test was performed to identify significantly expressed mRNAs (FDR < 0.05) between the high and low GRK6 groups in RStudio (version 3.4.4). Codes are available upon request.

Identification of Top Hit Genes and Pathway Analysis
The statistically significant (FDR < 0.05) differentially expressed mRNAs in the TCGA (IlluminaHiseq) dataset that were highly expressed in the low GRK6 lung adenocarcinoma group were merged with statistically different genes in the RNA-sequencing dataset, which showed higher gene expression in siGRK6 samples compared to the control samples by using RStudio (version 3.4.4) to identify the top hit candidate gene(s) (Figure 2).
For pathway analysis, Metascape (https://metascape.org/gp/ index.html#/main/step1) was used to detect functional enrichment of the identified top hit genes. The pathways were sorted from the lowest q-value and pathways with a q-value of less than 0.05 were chosen to create a histogram plot in GraphPad Prism 8.

Clinical Data and Tissue Samples
The study population comprised of 174 lung adenocarcinoma (LUAD) patients who were examined and treated at the Thoracic Surgery Department of the Affiliated Hospital of Nantong University and Thoracic Surgery Department of Second Affiliated Hospital of Nantong University between January 1, 2015, and December 31, 2016. The median age of patients at the time of diagnosis was 63 years (range 41-83 years). Study protocol was approved by the Ethics Committee of the Affiliated Hospital of Nantong University (No. 2018-L068), and all experiments were performed in accordance with approved guidelines of the Affiliated Hospital of Nantong University. Written informed consent was obtained from the patients for publication of this study and any accompanying images. Details of the clinical and demographic information were collected retrospectively. All patients underwent standard surgery aiming for maximal tumour resection. Patient clinical data were recorded in detail, and the diagnoses were confirmed by at least two pathologists. Tumour histological grades and clinical stages were evaluated according to the pathological results after surgery. All tumours were staged according to the pathological tumour/node/metastasis (pTNM) classification (7th edition) of the International Union against Cancer.
Tissue Microarray (TMA) Construction and Immunohistochemistry Analysis (IHC) Tissue microarray system (Quick-Ray, UT06, UNITMA, Korea) in the Department of Clinical Pathology, Nantong University Hospital, Jiangsu, China, was used to generate TMA. Specifically, core tissue biopsies (2 mm in diameter) were taken from individual FFPE blocks and arranged in recipient paraffin blocks. TMA specimens were cut into 4 µm sections and placed on super frost-charged glass microscope slides. TMA analysis was used as a quality control for hematoxylin and eosin staining. Tissue sections were deparaffinized and rehydrated through graded ethanol. Antigen retrieval was performed with 0.01 M citrate buffer pH 6.0 and microwave heat induction. Endogenous peroxidase activity was blocked with 3% H 2 O 2 for 30 min. Sections were then incubated with a rabbit polyclonal antibody specific to GRK6 (1:100; Proteintech, 11439-1-AP), HIF1a (1:100; Proteintech, 20960-1-AP) and VHL (1:100; Abcam, ab140989) at 4°C overnight, followed by incubation with a biotinylated antirabbit secondary antibody at 37°C for 30 min. Slides were then processed using horseradish peroxidase and 3,3-diaminobenzidine chromogen solution and counterstained with hematoxylin. The staining intensity of GRK6, HIF1a or VHL for each slide was evaluated and scored by two independent pathologists. Staining intensity was scored as follows: 0 (negative), 1+ (weak staining), 2+ (moderate staining), and 3+ (intense staining). For each of the four staining intensity scores, the percentage of cells stained at each intensity were determined. The intensity percentage score was the product of staining intensity and percentage of stained cells. The final staining scores were then evaluated from the sum of the four intensity percentage scores; thus, the staining score had a range from the minimum value of 0 (no staining) to a maximum of 300 (100% of cells with 3+ staining intensity), as described previously (27). The cut-off of 140 was selected to evaluate expression: score 0-140 was considered low expression, while 141-300 was considered high expression. For all subsequent analyses, GRK6, HIF1a and VHL protein expression levels were considered either as "low" or "high" according to these cut-off values.

Statistical Analysis
Two-tailed, unpaired Student's t-test for the TCGA data were performed in RStudio (version 3.4.4). For multiple t-test, P-values were adjusted by using Benjamini-Hochberg (BH) method. Codes are available upon request. Fisher's exact test was used to evaluate the relationship of GRK6 and HIF1a expression in lung adenocarcinoma patient samples in IHC using GraphPad Prism 8 software. P < 0.05 was considered statistically significant.

Global Transcriptomic Changes in ATII Cells Upon GRK6 Depletion
We previously reported that GRK6 knockdown promotes cell migration and invasion in lung epithelial cells (17). To determine if, and how, lung epithelial cells responded to GRK6 inhibition, we characterised the global transcriptomic changes in alveolar epithelial type II (ATII) cells transfected with either siRNAs against GRK6 (siGRK6) or control siRNA (Control) by performing RNA sequencing (RNA-seq). Principal component analysis (PCA) showed good separation between Control compared to siGRK6 samples (n = 3 in each group) (Supplementary Figure 1).
Genes with a false discovery rate (FDR) adjusted P value (or qvalue) of less than 0.05 were considered as differentially expressed genes (DEGs). In total, 7,116 DEGs were identified, including 3,430 up-regulated (Supplementary Table 1) and 3,686 downregulated (Supplementary Table 2

Candidate Pathways Enriched Upon GRK6 Inhibition Are Identified by TCGA Analysis Coupled With RNA-seq
To understand the role of GRK6 in lung adenocarcinoma, we performed TCGA analysis coupled with the RNA-seq data described above. As shown in Figure 2A, correlation analysis was performed in the TCGA lung adenocarcinoma (LUAD) (IlluminaHiseq) dataset; samples were separated into high vs. low GRK6 expression based on an unsupervised hierarchical clustering (Supplementary Figure 2). We identified 2,345 genes as differentially expressed in the high vs. low GRK6 samples in the TCGA dataset ( Figure 2B). A total of 7,116 genes were differentially expressed in ATII cells transfected with control siRNA or siRNA against GRK6 (siGRK6) in RNA-seq, among which 3,430 up-regulated ( Figure 2C). By cross-referencing the results from the TCGA analysis with the RNA-seq analysis, we identified 274 candidate genes, which were highly expressed in low GRK6 samples in the TCGA dataset ( Figure 3A; Supplementary Table 5) and in siGRK6 samples in the RNA-seq analysis ( Figure 3B; Supplementary Table 6).
To further validate the in vitro findings, the correlation between GRK6 expression and HIF1a levels or GRK6 expression and VHL levels were analysed in lung adenocarcinoma samples using tissue microarrays ( Figure 5). Representative images of low and high expression of GRK6, HIF1a or VHL in lung adenocarcinoma samples are shown in Figures 5A-C, respectively. Importantly, the percentage of patients with high HIF1a expression (61%) in the low GRK6 group was significantly higher than in the high GRK6 group (41%) ( Figure 5D; P < 0.05), while patients with low GRK6 tended to have a low level of VHL compared to those within high GRK6 group ( Figure 5D; P < 0.0001).

DISCUSSION
Lung cancer is the most prevalent and the leading cause of cancer death (34). Adenocarcinoma is the most common type of lung cancer, in both smokers and non-smokers, in females and males, and represents 40% of the lung cancer cases (35). Lung adenocarcinoma progresses from the small airway; one of the most abundant cell types present here are alveolar type II epithelial cells, which secrete mucus and other substances (36). Lung adenocarcinoma is one of the most aggressive cancers and the survival rate of patients is short after diagnosis with an overall survival rate of less than 5 years (35). The major challenge for lung adenocarcinoma is its resistance to conventional radiotherapies and chemotherapies (35).
Hypoxia is one of the typical features of the tumour microenvironment that increases the aggressiveness of different tumours such as lung cancer (37), colorectal cancer (38), hepatocellular carcinoma (39) and oesophageal squamous cell carcinoma (40). Hypoxic conditions lead to the activation of various transcription factors, such as HIF1; and the activation of downstream signalling pathways that regulate cell death, motility and proliferation (41). HIF1 is a heterodimeric transcription factor, capable of controlling the cellular adaptive response to hypoxia and has two subunits; HIF1a and HIF1b (42,43). Cellular oxygen concentration regulates the protein expression of HIF1a so is a key factor for cellular adaptive response to hypoxia (43). HIF activities can also be up-regulated by other mechanisms (44,45).
G protein-coupled receptor kinases (GRKs) are a family of kinases which can desensitize G protein-coupled receptors (GPCRs) homologous (1). GRK6 is of the members of  theGRK family (10-12) and we previously showed that GRK6 is down-regulated in lung adenocarcinoma, which is associated with malignant tumour progression (16,17), by an unknown mechanism.
To identify global transcriptomic changes in ATII cells upon GRK6 depletion, RNA-seq coupled with siRNA-mediated depletion of GRK6 was performed in ATII cells. We identified 3,430 up-regulated and 3,686 down-regulated DEGs. GO functional analysis with DEGs demonstrated that DEGs are mainly enriched in mRNA metabolism, ribonucleoprotein complex biogenesis, and regulation of cellular response to stress. To understand the role of GRK6 in lung adenocarcinoma, analysis of TCGA data was coupled with the RNA-seq data, described above. Pathway analysis suggested that one of the Hallmark pathways enriched upon GRK6 inhibition is 'Hallmark_Hypoxia'. We demonstrated that GRK6 depletion induces HIF1a expression and activity in ATII cells. The findings were further confirmed in lung adenocarcinoma samples, in which GRK6 expressions negatively correlate with HIF1a protein levels. Mechanistically, the impact of GRK6 on HIF activity could be achieved via regulation of VHL levels, which is a master regulator of HIF activity by targeting the prolyl-hydroxylated HIF1a subunit for ubiquitylation and rapid proteasomal degradation (30)(31)(32)(33). This study provides evidence that GRK6 inhibition causes a decrease in VHL expression, leading to HIFa stabilisation with increased activity in lung adenocarcinoma, although the underlying mechanism merits further investigation.  Earlier reports suggest that hypoxia regulates mRNA translation (46). RNA-binding proteins (heterogeneous nuclear ribonucleoproteins) have a role in post-transcriptional gene regulation under hypoxic conditions and are associated with hypoxia-induced transcripts that regulate encoded protein levels (47). Hypoxia can affect tumour cells; by acting as a stressor and inhibiting cell growth or inducing cell death. Alternatively, it can act by contributing to cancer progression and resistance to treatments, leading to hypoxia-induced genomic and proteomic changes in the cancer cells (48,49).
We previously demonstrated that cell migration and invasion in lung epithelial cells is induced upon GRK6 knockdown (17). In addition to the hypoxia, this analysis showed EMT is also enriched upon GRK6 inhibition, which can explain our previous findings (17). The hypoxic tumour microenvironment can regulate EMT (50,51). EMT is a biological process and the cell polarity and cell-cell adhesion of epithelial cells are lost and in turn become mesenchymal cells, which have migratory and invasive features (52). In a similar manner to our findings (17), previous studies in medulloblastoma (53) and Lewis lung carcinoma (2) show that when GRK6 was downregulated, migration and metastasis were increased. Consistently, it has been found that hypoxia-related genes CA9, NDRG1, SLC2A1, P4HA1 and ENO1 induced EMT in hepatocellular carcinoma (54), bladder cancer (55), laryngeal cancer (56) and gastric cancer (57,58), respectively. Our study showed an increase of hypoxiainduced gene expression and HIF1a expression in GRK6  knockdown cells, this suggests that GRK6 knockdown may induce EMT in lung adenocarcinoma.
In summary, this study shows that GRK6 is involved in different disease-related pathological features; mRNA metabolism, ribonucleoprotein complex biogenesis, regulation of cellular response to stress, as well as EMT and hypoxia. Targeting the HIF pathway may provide new strategies for therapy in GRK6-depleted lung adenocarcinoma patients.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi. nlm.nih.gov/geo/, GSE164921.

ETHICS STATEMENT
This study was approved by the Ethics Committee of the Affiliated Hospital of Nantong University (No. 2018-L068). The patients/participants provided their written informed consent to participate in this study.