GPR81 nuclear transportation is critical for cancer growth and progression in lung and other solid cancers

Abstract

BACKGROUND

The Warburg effect is common in cancers. Lactate and its receptor GPR81 play an important role in cancer progression. It is widely accepted that membrane receptor nuclear translocation plays some novel role in cancer pathology. The mechanism by which the lactate/GPR81 axis regulates cancer malignancy remains unclear.

AIM

To elucidate the mechanism of GPR81 nuclear transportation promoted by exogenous lactate.

METHODS

Lung cancer cells were stimulated with exogenous lactate and GPR81 levels were measured by immunofluoresence and western blot analysis in membrane, cytoplasmic, and nuclear fractions. Lung cancer cells were transduced with a mutant GPR81 nuclear localization signal (NLS) construct, wild type GPR81 or empty vector and used to examine how GPR81 nuclear transportation affects lung cancer cells malignancy in vitro and in vivo. Immunoprecipitation Proteomics analysis and Chromatin immunoprecipitation (ChIP) sequencing were used to determine GPR81 interacting proteins and genes.

RESULTS

In response to hypoxia/Lactate stimulation, GPR81 translocates and accumulates in the nucleus of lung cancer cells. Functionally, GPR81 nuclear translocation promotes cancer cell proliferation and motility. Depletion of the GPR81 NLS depletes GPR81 nuclear levels and decreases cancer cell growth and invasion in vitro, as well as cancer cell malignancy in vivo. Proteomics analysis revealed a set of proteins including SFPQ, that interact with GPR81 in the cancer cell nucleus. Notably, the interaction of GPR81 with SFPQ promotes cancer cell growth and motility. ChIP sequencing analysis discovered that there is a set of genes targeted by GPR81.

CONCLUSION

The interaction of GPR81 with SFPQ promotes cancer cell malignancy. GPR81 nuclear translocation is critical in conferring cancer progression and may be a potential therapeutic target for limiting cancer progression.

Key Words: Solid cancers; GPR81; Nuclear translocation; Proteomics; Chromatin immunoprecipitation sequencing; Ingenuity pathway analysis; Warburg effect; Self-renewal; Invasion

Core Tip: Lactate promotes GPR81 expression and nuclear transportation. where GPR81 interacts with nuclear proteins and regulates cancer cell function. Then targeting GPR81 and its nuclear transportation provides an opportunity to develop novel treatments targeting cancers.

INTRODUCTION

Solid tumor cancers are the most common and lethal cancers. Understanding the biological development of cancers remains critical to improving treatment efficacy. The progression of lung cancer is dependent on the interaction between tumor cells and the microenvironment composed of different cellular components and metabolites, including the products of the fermentation of glucose to lactate. Often known as the Warburg effect, this metabolic pathway is common in solid cancers, but the manner in which this series of reactions contributes to cancer progression is not fully understood[1-3]. Aerobic glycolysis, also a component of the Warburg effect, and anaerobic glycolysis, also known as hypoxia-induced glycolysis, are both metabolic processes that involve the conversion of glucose into energy. When the mechanism of hypoxia-induced energy conversion is intensively studied the Warburg effect remains unclear[4-6]. While lactate has long been known to play an essential role in cellular metabolism, more recently it has been discovered that it also serves as a signal molecule regulating a variety of cellular processes associated with cancer progression and cardiovascular disease[4,7-9].

The GPR81 is a lactate receptor, also known as the HCAR1. GPR81 is expressed at high levels in cancer, where it promotes cancer cell proliferation and prevents the presentation of tumor-specific antigens to immune cells[10-12]. G-protein coupled receptors belong to a large receptor family although these signal transduction pathways and ligand/receptor mechanisms are well described[13-16]. GPR81, one member of this family, and its mechanism in disease pathology are not well studied[2,17]. The manner in which lactate/GPR81 plays a role in cancers remains unclear, but in the current study, we identify that there is aberrant localization of GPR81 in cancer cells. Such inappropriate protein localization is common in cancers and is of interest for both the mechanism and its potential as a therapeutic target[18-20].

Lactate receptor GPR81 is highly expressed in cancer cells and appears important in the malignant phenotype[21,22]. We demonstrate that exposure to exogenous administration of lactate promotes GPR81 nuclear transportation and cancer cell self-renewal. GPR81 nuclear transportation may play a critical role in cancer malignancy and holds promise as a potential therapeutic target to arrest cancer progression.

MATERIALS AND METHODS

Cancer cell lines, primary cell lines and patient tissue sections

Cell lines from ATCC are used in this study. RWPE-1 and HBCE3-KT were cultured with Keratinocyte-SFM (Gibco 10724-011) and the 0.1% soybean trypsin inhibitor as ATCC instructed. PCS-201-020, MHCC97, HepG2, Caki-1, MCF, MDA468, PC3 and DU145 were cultured in DMEM with 5% FBS. Primary lung cell lines were established from patients fulfilling diagnostic criteria for lung cancer and other lung diseases including a pathological diagnosis of usual interstitial pneumonia. Patient controls were selected to be similar in age to lung cancer patients with non-cancer lung disorders. Control and lung cancer cell lines were derived from lungs, and cultivated as previously described[23]. Human patient tissue sections were collected and prepared through Bionet, University of Minnesota. We utilized 4 non-cancer primary control cell lines from lung tissue uninvolved by the primary process: Histologically normal lung tissue from a gunshot victim (n = 1) or chronic obstructive pulmonary disease (COPD) (n = 3); lung cancer cell lines are all non small cell lung cancer; 12 non-cancer patient tissue: Interstitial lung disease (n = 5) or COPD (n = 7) for control tissue sections. A pathologist verified all tissue to be tumor free.

Lactate treatment

For lactate stimulation, lactate sodium L-lactate (Sigma-Aldrich, 71718, St. Louis, MO, United States) was added to the media followed by pH adjustment to pH 7.6, the addition of pre–pH-adjusted lactate to the media, and the addition of serum to the media after pH adjustment. Cells were co-cultured as indicated in the figure legends for 24 hours for most experiments except where indicated.

Self-renewal assay

Single cell suspensions of lung cancer cells were incorporated into methylcellulose gels (Stemcell Technologies, Vancouver, Canada) and maintained in MSC SFM CTS (Thermo Scientific/Gibco, Rochford IL, United States) (37 °C, 5% CO₂; 1 week). Enumeration of colonies was performed microscopically and colony size was quantified by Image J. In some self-renewal assays, the cells were treated with the indicated concentrations of lactate (Sigma, United States).

Reverse transcriptase quantitative PCR

GPR81, CD44, MMP2, Ki-67, and SFPQ gene quantification was conducted by Quantitative PCR (Q-PCR) as previously described[9]. Total RNA was isolated and cDNA was transcribed using a TaqMan reverse transcriptase reagent kit (Roche, Indianapolis, IN, United States) and primed with random primers. Primer sequences were designed using NCBI Primer-BLAST. Q-PCR was performed using a LightCycler FastStart DNA Master^PLUS SYBR Green I Kit (Roche, Indianapolis, IN, United States). Primer sequences were as follows: GPR81 Forward: 5′- AATTTGGCCGTGGCTGATTTC-3’, GPR81 Reverse: 5′- ACCGTAAGGAACACGATGCTC-3’, Ki-67 Forward: 5′- TCCTTTGGTGGGCACCTAAGACCTG-3’, Ki-67 Reverse: 5′- TGATGGTTGAGGTCGTTCCTTGATG-3’, MMP2 Forward: 5′-CTCAGATCCGTGGTGAGATCT-3′, MMP2 Reverse: 5′-CTTTGGTTCTCCAGCTTCAGG-3′, CD44 Forward: 5′- GCTACCAGAGACCAAGACACA -3’, CD44 Reverse: 5′- GCTCCACCTTCTTGACTCC -3’, CD44v6 Forward: 5′- CCAGGCAACTCCTAGTAGTACAACG -3’, CD44v6 Reverse: 5′- CGAATGGGAGTCTTCTTTGGGT-3’, CTNNA1 Forward: 5′-GGGCAATGCTGGACGTAAAG-3′, CTNNA1 Reverse: 5′-TCTGAAACGTGGTCCATGACA-3′, ANXA7 Forward: 5′-GCTATCCCCCAACAGGCTAC-3′, ANXA7 Reverse: 5′-CCTGGTGGGACTCCAAATC-3′, SFPQ Forward: 5′-GATCTACAGGGAAAGGCATTGTTG-3′, SFPQ Reverse: 5′-GATACATTGGATTCTTCTGGGCA-3′, GAPDH Forward: 5′- TGTTGCCATCAATGACCCCTT-3′, GAPDH Reverse: 5′-CTCCACGACGTACTCAGCG-3′. Samples were quantified by using LightCycler analysis software and compared to an external calibration standard curve.

Plasmids/constructs

For altering gene function, GPR81 was knocked down using shRNA (pGIPZ-GPR81 shRNA; IDT and UMN Genomics Center). Scrambled shRNA served as a control. The GPR81 nuclear localization sequence (NLS) mutant was constructed by site-directed mutagenesis using wild-type GPR81 cDNA as a template. The putative NLS ²⁹⁴ICSLKPKQP³⁰² was mutated to ²⁹⁴ICAAAPAAA³⁰². The mutant constructs were verified by DNA-Seq. Cloning of the NLS GPR81, wild-type, or empty vector construct was done following standard cloning procedures. Lentiviral packaging was done following standard protocols[24].

Mass spectrometry

Nuclear fractions of cancer cells were isolated by using NE-PER Nuclear and Cytoplasmic Extraction reagents (Thermo Scientific). GPR81 was immunoprecipitated from the resulting nuclear fraction using a GPR81 antibody (NLS2095, R&D Systems). An isotype control antibody was used in control immunoprecipitation samples. The immunoprecipitates were then run SDS-PAGE followed by silver staining (Pierce Silver Stain Kit, catalog 24612, Thermo Fisher Scientific). Bands were processed for in-gel trypsin digestion. Peptide samples were resuspended in 98: 2 water/acetonitrile, 0.1% formic acid and run on an Orbitrap Velos (Thermo Fisher Scientific) mass spectrometer. The Thermo RAW files were analyzed with Peaks Studio 7.0 build 20140912 (Bioinformatics Solutions) software package for interpretation of tandem mass spectrometry and protein inference.

Chromatin immunoprecipitation sequencing analysis

Chromatin immunoprecipitation (ChIP)-sequencing analysis was performed as previously described with minor modifications[25,26]. The rabbit antibody GPR81 (PA5-114741, Thermo Fisher Scientific) was obtained from Thermo Scientific, and the rabbit polyclonal GPR81 (NLS2095, R&D Systems) was obtained from R&D systems. In brief, 15 million human cancer cells were harvested by trypsinizing into a single-cell suspension and crosslinked with formaldehyde (1%) for 10 minutes. at 37 °C. Cell pellets were subsequently sonicated, and extracts equivalent to 5 million cells were used for ChIP assays using 5 µL antibody. Additional steps are as by the published protocol[27]. Sequencing was conducted by the Genomics Center, University of Minnesota, using the Illumina system.

Sequencing handling. Sequence reads were screened for low-quality sequences and adaptor contamination with FastQC version 0.11.9 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Low-quality sequences, adaptor contaminants, and reads that were shorter than 18nt after trimming were removed with Trimmomatic version 0.39[28]. Cleaned reads were mapped to the GRCh38 repeat-masked assembly with BWA-MEM version 0.7.17[29]. Mapping parameters were kept at their default values. The resulting SAM files were sorted and converted to BAM files with SAMtools version 1.14[30]. The “NRF” (nonredundant fraction), “PBC1” (PCR bottleneck 1), and “PBC2” (PCR bottleneck 2) library quality summary statistics as defined by the ENCODE consortium (https://www.encodeproject.org/data-standards/terms/) were calculated with a custom script. BAM files were then post-processed to remove reads that mapped to alternate scaffolds, patches, and the mitochondrial genome, as well as to remove reads with MAPQ < 30 and those with SAM flag 780 (unmapped, mate unmapped, not primary alignment, or fail vendor checks). Cleaned BAM files were used for peak calling and binding analyses.

Invasion assay

Invasion assay was conducted with the Millipore kit (8 µm pore Transwell inserts in 24-well tissue culture plates) (Millipore, United States). Cancer cells were cultured in serum-free DMEM for 24 hours, and seeded into the upper chamber at 2 × 10⁴ cells/well in 300 µL serum-free DMEM. The lower chamber contained 500 µL 10% FBS DMEM (positive control), conditioned DMEM, or serum-free DMEM (negative control). The invaded cancer cells and the cells remaining in the upper chamber were detected with CyQuant GR Dye at 16 hours and were quantified with a fluorescence reader.

Peak calling and binding quantification

Cleaned BAM files were used as input for ChIP-seq peak calling with MACS2[31]. Peaks were called on a per-sample basis, with the IPF+Lactate library considered a “treatment” condition and the IPF without lactate library considered a “control” condition. MACS2 was run with the “hs” (human genome) size and complexity preset. The per-sample narrow peak call sets were then merged with the MSPC package in R[32]. Peaks were merged and considered “stringent” with a per-sample P-value threshold of P < 10^-3 and a requirement that the peak be present in at least one sample. Merged peaks were converted to a SAF file, to be used for quantification, with a custom script. Consensus peaks were quantified in each library with the cleaned BAM files and featureCounts version 2.0.5[33]. Counting was done in a strand-nonspecific manner, and duplicate fragments, non-primary mapping locations, and fragments with MAPQ < 20 were excluded from counting. Functional Annotation: Peaks were assigned to genes by overlap with annotated gene features with 2kb upstream flanks, to account for gene promoters. The annotated Gene Ontology terms for the genes were then attached to the annotated peaks.

ChIP assay

ChIP assay was conducted with Abcam kit (Ab185913, Abcam, Cambridge, MA, United States) by following the manufacturer’s instructions. 6 × 10⁶ cells with treatment as indicated or vehicle control were used for chromatin fragment preparation. GPR81 or control IgG were immunoprecipitated from nuclear fractions using anti-GPR81 (NLS2095, Novus, United States) antibodies. ChIP-PCR was performed as delineated above.

DNA accessibility assay

CTNNA1 and ANXA7 nuclease accessibility was performed using a chromatin accessibility assay kit (Abcam, Cambridge, MA United States) and CTNNA1 and ANXA7 specific primer sets. The experiment was conducted following the manufacturer’s instructions.

Western blot and immunoprecipitation

Western blots were performed as previously described[4-6]. Briefly, Cell lysates were prepared with RIPA lysis buffer (150 mmol/L NaCl, 50 mmol/L Tris pH 8.0, 1 mmol/L EDTA, 1 mmol/L EGTA, 0.5% sodium deoxycholate, 0.1% SDS, and 1% Triton X-100) plus protease inhibitor cocktail (0.1 M phenylmethylsulfonyl fluoride, 5 μg/mL leupeptin, 2 μg/mL aprotinin, and 1 μg/mL pepstatin). Protein concentrations were measured using the BCA method and protein samples were separated on an 8%-14% SDS-PAGE at 80 V. Proteins were then transferred to a polyvinylidene difluoride membrane for 8 minutes on the Turbo transfer System (Invitrogen, United States). After blocking in 5% skim milk powder for 1 hour at RT, the membrane was incubated with primary antibody overnight at 4 °C. The membrane was washed three times and then incubated for 1 hour at RT with the horseradish peroxidase-conjugated secondary antibody. After 3 washes, protein bands were visualized by ECL Plus according to the manufacturer’s instructions (Cell Signaling, United States). For immunoprecipitation, nuclear fractions were precleared for 1 hour at 4 ºC with protein A/G beads and immunoprecipitated for 2 hours at 4 ºC with the appropriate primary antibody.

Immunohistochemistry of cancer tissues

Immunohistochemistry was conducted on 4 µm paraffin-embedded serial sectioned lung, liver, prostate, kidney and breast tissues. The sections were deparaffinized, quenched with 0.3% Hydrogen Peroxide, and boiled for 30 minutes in Citrate Buffer (pH 6.0) for antigen retrieval. Sections were blocked in 5% Normal Horse Serum (Jackson Immunoresearch, West Grove, PA, United States). The tissue specimens were incubated overnight at 4 ºC with the following primary antibodies: Anti-rabbit GPR81 monoclonal antibody (1:800) (Ab1888647, Abcam, United States), anti-human CTNNA1 antibody (1:500) (Ab15032, Abcam, United States), anti-human Ki67 antibody (1:500) (ab279653, Abcam, United States). Specimens were then visualized under the microscope.

Mouse xenograft model of cancer study

NOD/SCID/IL2rg/B2M (NSG) mouse model was used to assess the metastatic ability of NSC lung cancer cells in vivo[34]. All mouse studies followed the protocols reviewed and approved by the University of Minnesota Institutional Animal Care and Use Committee. 10-week-old age-matched NSG male and female mice (Jackson Laboratories) were used for intravenous injections in meglinancy studies. 3 days before IV cell injection, the mice were irradiated (225 cGy) to deplete immune cells. 5 × 10⁵ of lung cancer cells suspended in 50 μL PBS were IV injected into the mice tail with a 28-gauge needle after the mice were anesthetized with 5% isoflurane. All experimental mice were monitored until fully recovered from anesthesia, and were subsequently monitored for disease progression through measuring body weight and behavioral signs daily. Mice were euthanized by CO₂ and different organ tissues were harvested 4 weeks later. Histological (H&E and trichrome staining) and immunohistochemical analysis were performed on paraffin-embedded mouse tissues.

Quantification of human cancer cells in mouse lung tissue

Mice harvested above were also used for this quantification. The lungs were digested, and genomic DNA was prepared using a Genomic DNA Mini Kit by following the manufacturer’s instructions (Invitrogen). qPCR was used to quantify human cancer cells in the mouse per a previously published protocol[35]. The PCR assay was performed for 40 cycles using the human genomic DNA–specific primers (forward: 5’-ATGCTGATGTCTGGGTAGGGTG-3’; reverse: 5’-TGAGTCAGGAGCCAGCGTATG-3’). Genomic DNA from 1 × 10⁶ A549 was used as a reference control in qPCR. Image quantification with Image J was processed following the method described before[7].

Statistical analysis

Comparisons of data among experiments were performed using the 2-tailed Student’s t-test and one way ANOVA as indicated. P < 0.05 was considered significant. To present sample distribution, data normalized to a randomly chosen control sample (assigned as 1) in most of the experiments.

Study approval

Patient samples were provided by Bionet and collection was approved by the University of Minnesota Institutional Review Board (University of Minnesota IRB ID: 1504M68341). Animal studies were approved and conducted following the University of Minnesota Institutional Animal Care and Use Committee regulations (approval #1706-34890A).

RESULTS

Lactate promotes GPR81 expression and nuclear transportation in cancer cells but not in control lung cells

Cancer cells produce high levels of lactate associated with altered cell metabolism as has recently been described[36-38]. When we examined GPR81 expression in lung cancer cells we found that GPR81 is highly expressed in lung cancer cells (Figure 1A) compared to control cells both in reverse transcriptase-PCR (RT-PCR) and western blot analysis. When we treated cells with lactate the level of GPR81 appeared enhanced following lactate exposure (Figure 1B). Prior work indicates that cancer tissue contains high lactate levels and is widespread in cancer tissue. Lactate enhances cancer cell proliferation and regulates cancer cell invasion[36,37,39-41]. We also examined the effect of lactate on lung cancer cells in controlling self-renewal and invasion. Increased cancer cell self-renewal is seen in the presence of 5 mmol/L of lactate compared to vehicle (Figure 1C). When we explored both the cancer stemness marker CD44, the expression levels doubled following lactate treatment compared to controls (Figure 1D). The same treatment also leads to lung cancer cells demonstrating increased capacity for invasion (Figure 1E). Levels of the cell motion marker MMP2 demonstrated a more than 50% increase with those treatments (Figure 1F). These data indicate that cancer cells display increased self-renewal proliferation and higher levels of invasion capacity under lactate treatment, a result similar to published data in other cancers[22,42,43]. However, we also found that GPR81 accumulated in lung cancer cell nuclei after lactate treatment. We next examined the dynamics of GPR81 nuclear accumulation following lactate exposure in lung cancer cells. Lactate increased GPR81 in the nucleus of lung cancer cells within a few hours of exposure to lactate but not with a control vehicle (Figure 1G). Lactate also promotes GPR81 nuclear transportation in lung cancer cells within 30 minutes in a dose-dependent pattern (Figure 1G and H). When we observe if lactate also affects other cancer cells in a similar pattern we measured GPR81 expression in non cancer cell lines Lung cell PCS-201-020, bronchial cell HBEC3-KT, prostate cell RWPE-1, kidney cancer cell Caki-1, breast cancer cell MCF7 and MDA231, prostate cancer cell LNCaP and Du145, liver cancer cell HepG2 and MHCC97 with RT-PCR and Western blot (Figure 1I). Cancer cells have much higher GPR81 expression. We then examined whether lactate promotes GPR81 nuclear transportation in those cells. We found that lactate stimulates nuclear translocation in those cancer cells as well (Figure 1J). These data demonstrate that lactate both triggers and promotes GPR81 nuclear transportation in solid cancer cells.

Figure 1 Lactate promotes GPR81 expression and GPR81 nuclear translocation in cancer cells. A: GPR81 expression level in lung cancer and control cells were measured with reverse transcriptase-PCR (RT-PCR) (left panel) and western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control; B: GPR81 expression level in lung cancer and control cells cultured under vehicle/Lactate (5 mmol/L) conditions were measured with RT-PCR (left panel) and western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control; C: Lung cancer cells self-renewal was assessed in the colony forming assay; D: Cancer stemness marker CD44 and cell proliferation marker Ki67 were quantified with RT-PCR and western blot analysis; E: Lung cancer cell invasion was assessed; F: Cell motion marker MMP2 was quantified in RT-PCR and Western blot analyses. A-F: n = 4, each of control and lung cancer cells (Control 202, 205, 272, 392; lung cancer A549, 661, 858, 645); G: Lung cancer cells were cultured with varying concentration of lactate or vehicle and the cells were collected at 4 hours and nuclear fractions isolated for Western blot analysis. Lamin served as a loading control; H: Lung cancer cells were cultured with 5 mmol/L of lactate and the cells were collected at the indicated time and nuclear fractions were isolated for Western blot analysis. G and H: n = 4 Lung cancer cell lines (A549, 645, 661, 858); I: GPR81 expression level in cancer and control cells was measured with RT-PCR (left panel) and western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control; J: GPR81 Level in cancer and control cells cultured under vehicle/Lactate (5 mmol/L) conditions were measured with RT-PCR (left panel) and the nuclear fraction was analysed with western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. Lamin served as a loading control.

GPR81 positive cancer cells are present in the majority of cancer tissue sections from cancer patients

We next examined whether GPR81 exists in the nucleus of lung cancer cell lines, as well as cells in patient tissues. IHC was conducted in lung cancer patient tissue sections. We found that GPR81 exists in the nucleus of the majority of the lung cancer patient tissue sections (Figure 2A and B). When we quantify the GPR81 color area and density of cytoplasm and nucleus from the sections of 12 cases each of control and lung cancer tissues, cancer cell cytoplasmic staining was 5 times greater than control tissue. Nuclear staining was even more prominent as there was no staining in control nuclei (Figure 2C). Next, we checked if GPR81 also exists in lung cancer cell lines via an immunofluorescence stain. Many primary lung cancer cells had GPR81 nuclear staining (Figure 2D). Four cases of each control and lung cancer cell lines were quantified and a similar result to tissue was found (Figure 2E). GPR81 existed in lung cancer cell nuclei but not in control lung cells.

Figure 2 Nuclear GPR81 is present in non-small cell lung cancer cells and in non-small cell lung cancer tissue sections. Immunohistochemistry (IHC) stain was performed on human lung tissue. A and B: Control and lung cancer. Tissue sections were used in IHC immunostain. IHC was performed using antibodies to GPR81 to display GPR81 distribution in control and cancer tissue sections (red); Counterstain was Mayer’s hematoxylin to display the nucleus (blue). Scale bar = 100 µm; C: GPR81 stain was quantified with Image J and summarized in figure C. Twelve each of control and non-small cell lung cancer (NSCLC) patient specimens, 5 sections were imaged in each specimen; D: Immunoflurescent stain was performed on human NSCLC and control cells with GPR81 antibody. Scale bar = 20 μmol/L; E: GPR81 stain in figure D was quantified with Image J and summarized (4 cell lines each, 202, 205, 272, 392, A549, H661, H858, H645); F: Tissue sections were used in IHC immunostain. IHC was performed using antibodies to GPR81 to display GPR81 distribution in control and cancer tissue section (red); left panels are example images for non cancer liver, prostate, kidney and breast tissue; the middle two panels are example images for liver cancer, prostate cancer, kidney cancer and breast cancer (scale bar = 100 μmol/L). The right panel is a summary of the IHC stain for those cancers. Counterstain was Mayer’s hematoxylin to display the nucleus (blue). Scale bar = 50 µm.

When we examined GPR81 expression in other solid cancers. We found GPR81 present in most of the cancer cells including liver cancer, kidney cancer, breast cancer and prostate cancer tissue sections. The majority of the cancer cases had GPR81 staining in the nucleus (Figure 2F).

Lactate increases levels of a nuclear GPR81/transportin1 complex

Prior work showed that in response to the ligation of a receptor by its cognate ligand, receptors are commonly internalized and transported to the nucleus by transportin[44,45]. As lactate markedly increased nuclear GPR81 Levels, this suggested that lactate may promote the association of GPR81 with transportins. We analyzed the GPR81 sequence with a web-based program (https://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi) to predict NLS and found there is a potential NLS segment in GPR81 (²⁹⁴ICSLKPKQPGHSKTQRPEEMPISNLGRRSCIS³²³). The presence of an NLS suggests GPR81 is transported into the lung cancer nucleus with importin components via this sequence. In support of this, we found markedly increased levels of a nuclear GPR81/transportins complex in lung cancer cells exposed to lactate compared to the control medium (Figure 3). When we immunoprecipitated importin beta from both lactate treated and control lung cells, importin beta and GPR81 were both higher in lactate treated cell nuclei (Figure 3A). Following treatment of lung cancer cells with the nuclear importin beta inhibitor ivermectin (10 μmol/L), GPR81 decreased in the lung cancer cell nucleus (Figure 3B). GPR81 was increased by 187% in the cytoplasmic fraction but decreased by 81% in the corresponding nuclear fraction cells were treated with ivermectin. In order to validate these data, we constructed an NLS deleted mutant with the GPR81 NLS ²⁹⁴ICSLKPKQP³⁰² replaced with a mutant GPR81 NLS construct (²⁹²AAAAAQPAAA³⁰¹). When ligated, the GPR81 NLS mutant failed to enter the nucleus. Following the over-expression of the GPR81 mutant NLS construct we analyzed GPR81 Levels in cell fractions from GPR81 wild-type vector control, lactate treated, or vehicle treated GPR81 NLS mutant transduced cells. GPR81 Levels were high in the cytoplasm (Figure 3C) and cell membrane fractions (Figure 3D) but not in the nuclear fraction in GPR81 NLS mutant expressing cells (Figure 3E and F). Nuclear GPR81 in NLS mutant expressing cells is decreased by 70% compared to wild-type GPR81 cells. Similar observations were found in other solid cancer cells (Supplementary Figure 1). There was not much GPR81 nuclear accumulation in kidney cancer cell Caki-1, prostate cancer cell Du145, breast cancer cell MDA231 and liver cancer cell MHCC97 in GPR81 NLS mutant cells but there was GPR81 accumulation in GPR81 wild type cancer cells. These data support the concept that lactate triggers GPR81 nuclear import via importin beta and NLS in essential for the transportation.

Figure 3 GPR81 is translocated to cell nucleus with importin beta. A: Anti-importin beta was used to immunoprecipitate proteins from lung cancer cell with lactate or vehicle overnight treatment and the precipitated material was used in Western blot analysis. Densitometry values summarizing western blot data are shown in the right panel. Ig served as a loading control. Cell H661 was used; B: GPR81 expression levels were quantified in lung cancer cell exposed to 5 mmol/L lactate, with or without ivermectin (10 μmol/L) vs vehicle control in western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control. A549, H661, H858, H645 were used; C-E: GPR81 expression level in cell fractions of GPR81 NLS mutant transduced lung cancer cells, GPR81 wild-type transduced and empty vector transduced cells cultured under lactate or vehicle conditions were measured with western blot; D and E: Loading markers were GAPDH for cytoplasm, cadherin for cell membrane and lamin for nucleus. Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control. n = 4 Lung cancer cells (A549, H661, H858, H645); F: Immunoflurescent stain was performed on human lung cancer A549 cells (transduced with Empty vector, GPR81 wild-type and GPR81 NLSmutant as marked) with GPR81 antibody. Scale bar = 20 μmol/L.

Inhibition of GPR81 nuclear translocation reduces lung cancer cell self-renewal and invasion

To describe nuclear GPR81 effects on lung cancer cell function, we examined the effect of GPR81 nuclear accumulation on lung cancer cell self-renewal and cell invasion. Lactate increased cell proliferation and cell invasion in lung cancer cells (Figure 4). When we compared GPR81 NLS mutant transduced lung cancer cells with wild-type GPR81 gene transduced lung cancer cells, it was found that lung cancer cells with impaired GPR81 nuclear translocation have much less self-renewal in a colony assay than the lung cancer cells with wild-type GPR81 transduction. In addition, cancer stemness marker CD44 expression was decreased in lung cancer cells. Likewise, a similar result was obtained in a cell invasion assay (Figure 4C and D). Cell invasiveness decreased in lactate treated GPR81 NLS mutant cells compared to wild-type GPR81 cells. The cell motion marker MMP2 Level was more than 50% decrease in lactate treated GPR81 NLS mutant cells as compared to wild-type GPR81 cells. When we examined other solid cancer cells, NLS mutation transduced cells had reduced self renewal and invasion (Supplementary Figure 2). Although there was variation between the cancer cells, all but kidney cancer cell Caki-1, prostate cancer cell Du145, breast cancer cell MDA231 and liver cancer cell MHCC97 with GPR81 NLS mutant had less cell renewal and invasion compared to cancer cells with GPR81 wild type. These findings support a model wherein GPR81 nuclear translocation plays a role in lung cancer cell proliferation and cell invasion.

Figure 4 Inhibition of GPR81 nuclear translocation reduces cancer cell self-renewal and invasion. Lung cancer cells were transduced with empty vector, GPR81 wild-type or NLS mutant and cultured under vehicle or 5 mmol/L lactate conditions. A: Lung cancer cell self-renewal was assessed in the colony forming assay; B: Cancer stemness marker CD44 and proliferation marker Ki67 were quantified with reverse transcriptase-PCR (RT-PCR) and western blot analysis; C: Lung cancer cells invasion was assessed; D: Cell motion marker MMP2 was quantified in RT-PCR and western blot analyses using cell lines H549, H661, H848 and H448. Data are shown as mean ± SE. 3 technical replicates were performed for all experiments. GPR81 translocation affects lung cancer progression in vivo. NSG mice implanted with lung cancer cells (A549, H848 and H661) stably transduced with either empty vector, GPR81 NLS mutant, or wild-type GPR81 via IV (3 × 10⁵ cells/50µl); 6 mice/group; E-P: Serial 4 µm sections of right lung tissue from lung cancer tumors transduced with GPR81NLS mutant or wild-type GPR81. Representative H&E (E-G; scale bar: 500 µm) and immunohistochemistry (IHC) stains assessing GPR81, Ki67 and MMP2 positivity. IHC identifies GPR81, Ki67 and MMP2 and determines the distribution of GPR81 expression cells (H-J), Ki67 expressing cells (K-M) and MMP2 expressing cells (N-P). IHC for GPR81 and Ki67 was conducted to assess the distribution of human cells expressing GPR81, Ki67 and MMP2 protein expressing cells from mice from tumors of lung cancer cells transduced with empty vector, GPR81 wild-type (GPR81wt) or GPR81 NLS mutant (GPR81NLSmu). Inner frame shows an enlarged image from the indicated area. Scale bar = 200 μmol/L; Q-S: GPR81, Ki67 and positive cells in lung cancer cell explanted mice lung tissue sections; T: Cell quantification was conducted with quantitative PCR.

We tested if blocking GPR81 nuclear translocation affects cancer growth in vivo. GPR81 wild-type and GPR81 NLS mutant stably transduced cells were used. GPR81 wild-type and GPR81 NLS mutant cells were injected intravenously into NSG mice and 3 weeks later the lung tumors were harvested. GPR81 and Ki-67 expression in the resulting tumors were analyzed (Figure 4E-M). GPR81 staining identified that there were more GPR81 positive cells in tumors derived from GPR81 wild-type cells compared to stably transduced GPR81 NLS mutant cell tumors (Figure 4H-J). Stain density quantification also found that GPR81 positivity in GPR81 wild-type tumors was more than doubled compared to the density of GPR81 staining identified in tumors arising from GPR81 NLS mutant cells (Figure 4Q). When these tissue sections were stained for the proliferation marker Ki-67, it was discovered that the lung cancer xenografts with wild-type GPR81 cells had much greater proliferation than tumors from the contrasting NLS mutant cells. In addition, the cell number in GPR81 NLS mutant tumors was 37% lower than GPR81 wild-type tumors using Image J quantification following IHC stain (Figure 4K-M, R). Invasion marker MMP2 was stained with those sections too, it was discovered that the lung cancer xenografts with wild-type GPR81 cells were much greater than tumors from the contrasting NLS mutant cells. In addition, the cell number in GPR81 NLS mutant tumors was 63% lower than GPR81 wild-type tumors using Image J quantification following IHC stain (Figure 4N-P, S). To correlate with the IHC findings demonstrating the decreased expression of human nuclear protein IHC in tumors derived from GPR81 NLS mutant cells we quantified human cell numbers in the mice at the time of tissue harvest to assess the expansion of the human cells during the duration of the experiment. The lung tumors arising from mice receiving GPR81 NLS mutant cells contained decreased (62% less than GPR81 wild-type) levels of human DNA compared to mice that received GPR81 wild-type cells (Figure 4T). Together, these data indicate that inhibition of GPR81 nuclear translocation attenuates the ability of lung cancer to grow and spread in vivo.

GPR81 interactome in proteomics analysis in lung cancer cells

Once GPR81 is transported into the nucleus, it may interact with nuclear proteins potentially altering cell function. To define the nuclear GPR81 interactome, we performed a GPR81 immunoprecipitation proteomics analysis. Proteomics analysis identified a number of proteins bound to GPR81 (Supplementary Table 1). When we analyze their possible function and pathway importance with ingenuity pathway analysis (IPA) (https://digitalinsights.qiagen.com/products/qiagen-ipa), these proteins are found to include potential pathways of signal transduction impacting cell stemness, metabolism, senescence, migration and proliferation (Figure 5A and B). Some genes of interest include SFPQ, LDHa, PARP1, SRSFs and CD44. To confirm these protein interact with GPR81, GPR81 antibody immunoprecipitation and western blot were performed with PARP1, SRSF2, and SFPQ antibodies in lung cancer cells (Figure 5C). PARP1, SRSF2, and SFPQ bands were present in the GPR81 pull down material. A similar immunoprecipitation result was observed in other solid tumor cancer cells (Supplementary Figure 3). All cancer cells tested, including kidney cancer cell Caki-1, prostate cancer cell Du145, breast cancer cell MDA231 and liver cancer cell MHCC97 with GPR81 wild type had more GPR81/SFPQ complex than the cells with GPR81 NLS mutant. These data demonstrate that GPR81 interacts with nuclear proteins and suggest these interactions may impact cell function.

Figure 5 The interaction of GPR81 with other nuclear proteins affects lung cancer cell function. Proteins identified from GPR81 immunoprecipitated proteins in lung cancer cells in Proteomics analysis were applied to Ingenuity Pathway Analysis (IPA). A: Top functions associated with the lung cancer cell dataset as shown by IPA pathway analysis; B: Top canonical pathways associated with proteins from the lung cancer cell dataset as shown by IPA pathway analysis. Cell functions, or pathways identified are represented on the y-axis. The x-axis corresponds to the –log of the P-value (Fisher’s exact test) and the orange points on each pathway bar represent the ratio of the number of proteins in a given pathway that meet the cutoff criteria, divided by the total number of proteins that map to that pathway; C: Immunoprecipitation and western blot analysis with GPR81 bound proteins. The nuclear fraction from A549 was used in immunoprecipitation with the GPR81 antibody. Then SFPQ, PARP1 and SRSF2 antibodies were used in western blot analysis with the GPR81 pull down portion of the lung cancer nuclear fraction. Lactate promotes SFPQ expression and formation of GPR81/SFPQ complex in lung cancer cells; D: GPR81 reverse transcriptase-PCR (RT-PCR) was conducted with lung cancer cells cultured under vehicle/Lactate (5 mmol/L) conditions. (left panel). GPR81 antibody was used in immunoprecipitation with lung cancer cells cultured under vehicle/Lactate (5 mmol/L) conditions and the SFPQ levels were measured with western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control; E: GPR81 immunoprecipitation was conducted with the cell nuclear fraction from lung cancer cell cultured under vehicle/Lactate (5 mmol/L) conditions. Then the protein bands were detected with anti-GPR81 and SFPQ antibodies. Western blot was shown in the left panel. Densitometry values summarizing western blot data are shown in the right panel. Lamin served as a loading control; F: Lung cancer cells transduced with Lenti virus with empty vector, GPR81 wild-type or GPR81 NLS mutant, antibody was used in immunoprecipitation with lung cancer cell nuclear fraction; G: The same set of cells was used to analyze CD44 and CD44v6 expression. The CD44 and CD44v6 Levels were measured with RT-PCR (left panel) and western blot (middle panel). Densitometry values summarizing western blot data are shown in the right panel. GAPDH served as a loading control. Lung cancer cell A549, H661, H858, H645 were used in figure D-G; H: Anti-CD44v6 was used to assess the distribution of human cells expressing CD44v6 protein expressing cells from mice lung regions from mice receiving lung cancer cells transduced with empty vector, GPR81 wild-type (GPR81wt) (middle panel) or GPR81 NLS mutant (GPR81NLSmu) (right panel). Scale bar = 200 μmol/L. Quantification was summed up in the bottom figure. Image J was used to quantify CD44v6 protein positive cells in lung cancer cell explanted mice lung tissue sections.

Lactate affects SFPQ function via a GPR81/SFPQ nuclear complex

Based on the above data, we next examined the effect of lactate on the expression and function of the GPR81 binding partner SFPQ in lung cancer cells. Lactate promoted SFPQ and GPR81 Levels and the GPR81/SFPQ complex in lung cancer cells as compared to control cells (Figure 5D and E). There was not much GPR81/SFPQ complex in GPR81 NLS mutant cancer cells (Figure 5F). GPR81 accumulated in lung cancer cell nuclei after lactate treatment. We previously reported that SFPQ promotes CD44 isoform expression[46]. Our previous analysis indicates that SFPQ interacts with GPR81 in lung cancer cell nuclei and changes CD44v6 expression. Therefore, we examined the effect of GPR81 nuclear transportation on CD44v6 expression. The CD44v6 expression of lung cancer cells with empty vector and wild type GPR81 is greater than in GPR81 NLS mutant transduced cells. We also observed the CD44v6 expression in mouse lungs from our in vivo experiment. The tumors derived from GPR81 NLS mutant cells have much fewer CD44v6 positive cells; 80% less than the mice with GPR81 wild-type (Figure 5G and H). This suggests that lactate may promote lung cancer cell function via GPR81 nuclear transportation and GPR81/SFPQ interaction.

GPR81 targets a set of genes in ChIP sequencing analysis in lung cancer cells

To define the nuclear GPR81 targetome, we performed a GPR81 immunoprecipitation ChIP sequencing analysis (GSE253282). The analysis identified a number of genes associated with GPR81 (Supplementary Table 2). The genes were applied to IPA. Among these were genes involving cell stemness, metabolism, senescence, motility and proliferation. The top conical pathways include cell cycle control, oxidative phosphorylation, and cell metabolism. The top functions involve cell apoptosis, proliferation and motion (Figure 6A and B). We further screened these genes against the protein atlas (https://www.proteinatlas.org). Many of the identified proteins are cancer related (Supplementary Table 3). Genes BCR, ADK, ANXA7, ATG2B and CTNNA1 are among those identified. When we examined the ChIP binding plot, they all are present at different patterns (Figure 6C). We selected two of them with different binding peak patterns in the ChIPseq plot and validated if GPR81 binds with these genes using ChIP PCR. To confirm that GPR81 targets CTNNA1 (ENSG00000044115) and ANXA7 (ENSG00000138279), GPR81 antibody immunoprecipitation and ChIP PCR were performed with GPR81 antibody with either control or lung cancer cell lines (Figure 6C). CTNNA1 and ANXA7 were present in the GPR81 pull down chromatin. We further investigated if the GPR81 nuclear transportation alters the CTNNA1 gene and its expression in lung cancer cells. First, we found that GPR81 interacts with CTNNA1 using ChIP PCR and DNA accessibility assay (Figure 6D). Lung cancer cells with the GPR81 NLS mutant had much less GPR81 binding than lung cancer cells with the empty vector and GPR81 wild-type. We then examined whether GPR81 nuclear transportation alters CTNNA1 expression in lung cancer cells with RT-PCR and western blot analysis. We found that CTNNA1 Levels were much lower in GPR81 NLS mutant cells than in GPR81 wild-type cells. When nuclear GPR81 transportation is blocked, the CTNNA1 expression level is reduced by more than 50% in both RT-PCR and western blot analysis (Figure 6D and E). When we examined other solid cancer cells, CTNNA1 ChIP PCR was positive in Caki-1, MHCC97, MDA-231 and Du145 with GPR81 wild type while there was much less bound in cancer cells with GPR81c NLS mutant (Supplementary Figure 4). These data indicate that nuclear GPR81 targets genes at the level of chromatin potentially altering their expression.

Figure 6 GPR81 interacts with genes and change gene expression. Genes identified from GPR81 immunoprecipitated chromatin in lung cancer cells in Chromatin immunoprecipitation (ChIP) sequencing analysis were applied to Ingenuity Pathway Analysis (IPA). A: Top cell functions associated with the lung cancer cell dataset as shown by IPA pathway analysis; B: Top canonical pathways associated with proteins from the lung cancer cell dataset as shown by IPA pathway analysis. Cell functions, or pathways identified are represented on the Y-axis. The X-axis corresponds to the –log of the P-value (Fisher’s exact test) and the orange points on each pathway bar represent the ratio of the number of proteins in a given pathway that meet the cutoff criteria, divided by the total number of proteins that map to that pathway; C: The read depth plots for the GPR81 ChIP peaks of gene interest. Each plot shows the gene body. The red line shows the mean read depth across the +lactate samples and the black line shows the mean across the -lactate samples. The vertical blue lines show the boundaries of the gene body. Right panel: ChIP assays were performed on cross linked chromatin from nuclear fractions from control and lung cancer cells with CTNNA1 and ANXA7 primers; D: ChIP assays were performed on cross linked chromatin from nuclear fractions from GPR81 wt and GPR81 NLSmutant cells (n = 4 cell lines each) using immune GPR81 antibody (Isotype IgG as blank control) (left panel). Right panel: DNA Accessibility Assays were performed on cross linked chromatin from nuclear fractions from GPR81 wt and GPR81 NLSmutant cells (n = 4 cell lines A549, H661, H858, H645); E: CTNNA1 expression level was quantified with reverse transcriptase-PCR and western blot analysis in above set of cells; F: Anti-CTNNA1 was used to assess the distribution of human cells expressing CTNNA1 protein expressing cells from mice lung regions from mice receiving lung cancer cells transduced with empty vector, GPR81 wild-type (GPR81wt) (left panel) or GPR81 NLS mutant (GPR81NLSmu) (right panel). Scale bar = 200 μmol/L. Quantification was summed up in the bottom figure. Image J was used to quantify CTNNA1 protein positive cells in lung cancer cell explanted mice lung tissue sections.

We also examined CTNNA1 expression in mouse lung tissue from our in vivo study. CTNNA1 staining shows there were more CTNNA1 positive cells in CTNNA1 wild-type cell implanted tissue (Figure 6F left top) than in GPR81 NLS mutant cells implanted tissue (Figure 6F left bottom). Stain density quantification found that GPR81 positive cells GPR81 wild-type cell implanted tissue more than doubled than GPR81 positive cells in GPR81 NLS mutant cells explanted tissue (Figure 6F). These data indicate GPR81 nuclear transportation plays an important role in CTNNA1 expression.

DISCUSSION

Protein mislocalization is common in cancers and is often identified as having potential as a target for cancer therapy[18-20,47,48]. In the current study, we have identified that GPR81 nuclear translocation is common in cancer cells, and the translocation of GPR81 may affect cancer cell functions particularly in response to hypoxia and lactate accumulation. This translocation may offer a unique insight into the biology and potential treatment of cancers.

The Warburg effect, a phenomenon whereby cancers consume large amounts of glucose which is fermented to lactate even with oxygen present, is commonly seen in most solid tumors. The identification of GPR81 nuclear transportation in cancer may lead to new opportunities for investigating this mechanism. For example, GPR81 is upregulated in cancer where it is critical for cancer growth[22]. Although there are no specific models to demonstrate this hypoxia-lactate effect in cancer, we found there were positive hypoxia markers, carbonic anhydrase 9 and increased lactate concentration in the regions of cancer cell clusters in lung cancer cell xenografted NSG mice. In the current study, we show that lactate promotes GPR81 nuclear transportation in cancer cells but not in control cells. To further explore potential pathological mechanisms and therapeutic targets, we examined the associated proteins impacted by this translocation. Proteomics analysis identified a cohort of proteins bound to nuclear GPR81. These involve cell proliferation and invasion in IPA signal pathway analysis. Among the proteins identified as associating with nuclear GPR81 is SFPQ. Among its functions, SFPQ may affect RNA splicing[49], regulate chemotherapy sensitivity in ovarian cancer[50], or change DNA repair in breast cancer[51]. It was also recently reported that SFPQ expression is regulated by epigenetic methylation in non-small cell lung cancer[52]. In the current report, it was demonstrated that SFPQ binds to GPR81 and this interaction may affect SFPQ function and further influence cancer cell functions. CD44 isoform expression was also changed by GPR81 binding to SFPQ. In the presence of less GPR81 in the cancer cell nucleus there was less SFPQ interacting with GPR81 and less CD44v6 expressed.

We also investigated the effect on nuclear genes by GPR81 by ChIP sequencing GPR81 bound chromatin. These studies revealed a set of genes with cancer-relevant functions including CTNNA1 and ANXA7. CTNNA1 is involved in cell adhesion among other functions related to malignancy[53-55]. We demonstrated that inhibition of GPR81 nuclear translocation inhibited ChIP PCR value with GPR81 Ab precipitation and inhibition of GPR81 nuclear translocation reduced CTNNA1 expression in RT-PCR and western blot analysis, suggesting that GPR81 nuclear translocation affects CTNNA1 expression by bound chromatin. GPR81/chromatin bound force may not high as other chromatin binding proteins but the result demonstrates it could affect gene expression in cancer cells. It is not a chromatin bound protein normally but it interacts with chromatin and affects gene expression. The mechanism of these interactions may be worth further investigation.

In order to better understand whether targeting the alteration in GPR81 Localization may change cancer cell growth or function, we chose to examine the effects of nuclear GPR81 transport on cell function in vivo. Interestingly, we observed that blocking GPR81 nuclear transportation reduces lung cancer cell growth. CD44v6 expression level was inhibited in GPR81 nuclear translocation inhibited mice, which might be influenced by GPR81/SFPQ interaction. Expression of GPR81 targeted gene CTNNA1 was reduced in GPR81 nuclear translocation inhibited mice. CD44v6 and CTNNA1 are both involved in cancer malignancy[51,52,56]. These findings indicate that GPR81 nuclear transportation is critical in cancer cell proliferation and invasion. A potential therapy disrupting these interactions, or the localization of GPR81 to the nucleus, would therefore alter lung cancer cell growth, invasion, and self-renewal. There was some effort put forward to target protein nuclear translocation to disrupt the cell function change for cancer therapy[57,58]. Those will be other potential treatments for GPR81-expressing cancers.

CONCLUSION

In summary, GPR81 nuclear transportation in cancer cells interacts with nuclear proteins and regulates cancer cell function. In addition, it may bind to chromatin and influence gene expression changing cell function. Targeting GPR81 and its nuclear transportation provides an opportunity to develop novel treatments targeting cancers.

ACKNOWLEDGEMENTS

Human sample collection and screen were conducted by Bionet at the University of Minnesota. The gene molecular work was performed at the viral vector and cloning Core at the University of Minnesota; supporting agencies are listed here: https://vvcc.umn.edu/. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health’s National Center for Advancing Translational Sciences.