Head and neck squamous cell carcinoma (HNSCC) is comprised of metabolically linked distinct compartments. Cancer-associated fibroblasts (CAF) and nonproliferative carcinoma cells display a glycolytic metabolism, while proliferative carcinoma cells rely on mitochondrial oxidative metabolism fueled by the catabolites provided by the adjacent CAFs. Metabolic coupling between these reprogrammed compartments contributes to HNSCC aggressiveness. In this study, we examined the effects of cigarette smoke–exposed CAFs on metabolic coupling and tumor aggressiveness of HNSCC. Cigarette smoke (CS) extract was generated by dissolving cigarette smoke in growth media. Fibroblasts were cultured in CS or control media. HNSCC cells were cocultured in vitro and coinjected in vivo with CS or control fibroblasts. We found that CS induced oxidative stress, glycolytic flux and MCT4 expression, and senescence in fibroblasts. MCT4 upregulation was critical for fibroblast viability under CS conditions. The effects of CS on fibroblasts were abrogated by antioxidant treatment. Coculture of carcinoma cells with CS fibroblasts induced metabolic coupling with upregulation of the marker of glycolysis MCT4 in fibroblasts and markers of mitochondrial metabolism MCT1 and TOMM20 in carcinoma cells. CS fibroblasts increased CCL2 expression and macrophage migration. Coculture with CS fibroblasts also increased two features of carcinoma cell aggressiveness: resistance to cell death and enhanced cell migration. Coinjection of carcinoma cells with CS fibroblasts generated larger tumors with reduced apoptosis than control coinjections, and upregulation of MCT4 by CS exposure was a driver of these effects. We demonstrate that a tumor microenvironment exposed to CS is sufficient to modulate metabolism and cancer aggressiveness in HNSCC.

Implications:

CS shifts cancer stroma toward glycolysis and induces head and neck cancer aggressiveness with a mitochondrial profile linked by catabolite transporters and oxidative stress.

Visual Overview:

http://mcr.aacrjournals.org/content/molcanres/17/9/1893/F1.large.jpg.

This article is featured in Highlights of This Issue, p. 1775

Head and neck cancer is the sixth most common type of cancer worldwide, with an incidence of 600,000 new cases every year (1). Head and neck squamous cell carcinoma (HNSCC) accounts for nearly 95% of head and neck malignancies. Cigarette smoke (CS) is the major causative agent of HNSCC. Smokers are at much higher risk to develop the disease than nonsmokers, as well as being more likely to have worse treatment outcomes and shorter disease survival (2, 3). CS contains over 70 known carcinogens (4). DNA damage and adduct formation is thought to be the common mechanism by which these compounds cause mutations and drive carcinogenic transformation of the epithelial cells in the head and neck region (4). However, the effects of CS on the stromal cells within the tumor microenvironment of HNSCC have not been explored in detail.

The tumor stroma plays an important role in HNSCC development and progression, and there is increasing interest in the metabolic interplay between cancer cells and the surrounding noncancerous cells (5–8). Two studies from Curry and colleagues show that at least two metabolically distinct compartments exist within the tumor microenvironment of HNSCC (9, 10). The tumor stroma, which contains abundant cancer-associated fibroblasts (CAF), is highly glycolytic and secretes high-energy catabolites such as lactate and pyruvate. The proliferating carcinoma cells take advantage of this metabolic compartmentalization because they are mitochondria-rich and utilize these catabolites to fuel their oxidative metabolism. Markers of metabolic compartmentalization have been described in HNSCC and are associated with aggressive disease (5, 9). The monocarboxylate transporter 4 (MCT4), which is an exporter of lactate and has a hypoxia response element regulated by HIF1α, is a marker of glycolysis in CAFs. The importer of monocarboxylates MCT1 and the translocase of the outer mitochondrial membrane 20 (TOMM20) are markers of lactate intracellular uptake and high mitochondrial oxidative phosphorylation (OXPHOS) in carcinoma cells. Studying the metabolic compartmentalization of tumors is important not only to understand the pathophysiology of cancer but also to develop therapeutic targets. For instance, it has been recently demonstrated that the antidiabetic drug metformin, a mitochondrial inhibitor, affects tumor metabolic compartmentalization and has anticancer effects in HNSCC (11, 12).

Research on the pathogenesis of smoking-related diseases such as pulmonary emphysema and lung cancer has prompted the study of the effects of CS on tissue fibroblasts. It has been demonstrated that exposure of lung fibroblasts to CS induces oxidative stress, cellular senescence and apoptosis, as well as inhibits proliferation, migration, and extracellular matrix deposition (13–15). Some of these effects have also been reported in human gingival and skin fibroblasts exposed to CS (16–19). Numerous studies have also shown that CS induces proinflammatory signaling cascades and chemokine secretion in fibroblasts (20–22), creating a chronic inflammatory state that may contribute to the development and progression of cancer. The mechanisms by which CS elicits its effects on fibroblasts include generation of intracellular reactive oxygen species (ROS) with alteration in the cellular redox state. In fact, treatment with antioxidants such as NAC or overexpression of endogenous antioxidant systems protects fibroblasts from CS-induced ROS and cellular damage (13, 23–25). Moreover, signaling through the aryl hydrocarbon receptor (AhR), a regulator of the inflammatory response, attenuates oxidative stress, and reduces apoptosis and inflammation triggered by CS in lung fibroblasts (26–29).

While the effect of CS on isolated fibroblasts has been studied, very little is known about how these altered fibroblasts affect the microenvironment and epithelial cells in proximity. To our knowledge, only two reports have shown the effects of CS on fibroblasts in the context of cancer (19, 30). Salem and colleagues demonstrated in breast cancer that CS-exposed fibroblasts have the ability to metabolically promote tumor growth in a paracrine fashion, thus highlighting the importance of the stromal compartment in tumors (19). Studies of the effect of cigarette smoke on HNSCC stroma, however, have not been published. Therefore, we aimed to study the effects of CS exposure to fibroblasts on HNSCC metabolic compartmentalization, oxidative stress, inflammation, and aggressiveness.

Preparation of CS extract

Cigarette smoke extract (CSE) was generated as described by attaching one end of a three-way-stopcock to a 60-mL syringe containing 10 mL of unsupplemented DMEM, and another end to a Marlboro Red cigarette (Philip Morris; Supplementary Fig. S1A; refs. 13, 19, 31–34). Briefly, 40 mL of cigarette smoke was pulled through the stopcock into the syringe and dissolved in the media by vigorous shaking. The cigarette was “smoked” down to the filter, approximately 7 to 8 puffs for one cigarette. The 10 mL of media with contents of one cigarette was treated as 100% CSE stock. This stock was diluted to 5% CSE in DMEM, supplemented with 10% FBS and penicillin–streptomycin, and sterile filtered through a 0.22-μm filter.

Cell culture

The human tongue and hypopharynx squamous cell carcinoma cell lines, CAL27 and FaDu, respectively, were purchased from ATCC. The human papillomavirus (HPV+)-transformed mouse tonsil epithelial cell, MTEC, from male C57Bl/6 mice was a kind gift from Dr. Ulrich Rodeck (Thomas Jefferson University, Philadelphia, PA). Human skin fibroblasts immortalized with human telomerase reverse transcriptase catalytic domain (BJ1) were purchased from Clontech, and clones were generated with GFP. The BALB/c mouse monocyte/macrophage cell line, RAW 264.7, was a kind gift from Dr. Douglas C. Hooper (Thomas Jefferson University, Philadelphia, PA). All cell lines were obtained from cell banks and authenticated by short tandem repeat (STR) DNA analyses prior to receipt by our laboratory. These cells were passaged in our laboratory for fewer than 8 weeks after resuscitation. All cells were tested for Mycoplasma using the Universal Mycoplasma Detection Kit (30-1012K; ATCC) and were negative for Myocoplasma within 2 months of cell thawing and experimental work. Cells were cultured in DMEM (10566016; Gibco) supplemented with 10% heat-inactivated FBS (161407-071; Gibco), 100 U/mL penicillin, and 100 U/mL streptomycin (penicillin–streptomycin; 30-002-CI; Corning). Cell cultures were maintained in a 37°C and 5% CO2 humidified incubator.

Isolation of mouse embryonic fibroblasts

Mouse embryonic fibroblasts (MEF) were isolated from genetically engineered C57BL/6 mice lacking MCT4 (MCT4−/−) and their wild-type control (MCT4+/+; provided by N. Philp, Thomas Jefferson University, Philadelphia, PA; ref. 35). Pregnant female mice were euthanized at E14.5 by CO2 inhalation. Under sterile conditions, the abdominal wall was exposed and the uterus was removed and transferred to a Petri dish containing PBS and penicillin–streptomycin. Embryos were separated from the placenta and placed into a new dish with PBS for individual manipulation. Head, arms, and legs were cut off, and internal red tissue (heart and liver) was removed. The remaining embryo was minced, transferred to a tube containing 0.25% trypsin-EDTA, incubated for 10 minutes at 37°C, and then MEF culture media (DMEM 10% FBS) was added to the tube. Cell suspension was left for 5 minutes to allow large fragments to precipitate and supernatant was transferred to 10-cm Petri dishes. Media was replaced once MEFs were seen attached to the plate. MEFs were maintained in DMEM 10% FBS penicillin–streptomycin for no longer than ten passages.

Cell treatments

BJ1 fibroblasts were plated in 12- or 6-well plates or 10-cm dishes with standard growth media (CTRL-BJ1) or 5% CSE media (CSE-BJ1), as indicated. The seeding densities were 1 × 105 cells/well (for 12-well), 2 × 105 cells/well (for 6-well), and 106 cells/dish. Control media or 5% CSE was replaced every 2 days, for a total of 4 or 6 days. Fresh CSE was generated for each use. When indicated, 30 μg/mL of liposomal glutathione, 50 or 100 μmol/L of l-kynurenine (L-KYN, K8625; Sigma), 5 or 10 μmol/L of α-naphthoflavone (α-NF, N5757; Sigma), or the vehicle DMSO, were added to CTRL- and CSE-BJ1 at time of seeding and every time media were replaced, for a total of 4 days. WT and MCT4-KO MEFs were plated in 6- or 12-well plates, or T-75 flasks with standard growth media, at a density of 1 × 105 or 2 × 105 cells/well, or 2 × 106 cells/flask, respectively. The following day, MEFs were treated with either 5% or 10% CSE or control media, with or without 30 μg/mL liposomal glutathione, as indicated, for 24 hours. RAW 264.7 cells were plated in 6-well plates with standard growth media at a density of 6 × 105 cells/well. The following day, RAW 264.7 were treated with 5% CSE or control media for 24 hours.

Coculture system

CAL27 and FaDu human carcinoma cells were cocultured with either CTRL- or CSE-BJ1 (Supplementary Fig. S1B) in 12- or 6-well plates or 10-cm dishes, as specified. All cocultures were seeded at a 5:1 fibroblast-to-carcinoma ratio in standard DMEM. The total number of cells per well was 1 × 105 (12-well), 2.4 × 105 (6-well), and per dish was 1.2 × 106. MTEC mouse carcinoma cells were cocultured with WT MEFs, untreated or previously exposed to 5% CSE. A total of 6 × 104 MTEC cells were cultured at the bottom of a 6-well plate, and 2 × 105 MEFs were cultured on top of a polycarbonate cell culture insert with a 0.4-μm pore-size membrane (353090; Thermo Fisher Scientific). The next day, culture media were changed to DMEM with 10% Nu-serum, a low protein alternative to FBS (355100; BD Biosciences) containing penicillin–streptomycin. Cocultures were maintained in this media for 3 days.

Immunofluorescence

Immunofluorescence staining from cells in 12-well plates was performed as described previously (36). Primary antibodies used included anti-MCT4 (19-mer peptide sequence CKAEPEKNGEVVHTPETSV-cooh affinity purified rabbit antibody; YenZym Antibodies), anti-GLB1 (ab96239; Abcam), anti-HMGB1 (NB100-2322; Novus Biologicals), and anti-MCT1 (19-mer peptide sequence CSPDQKDTEGGPKEEESPV-cooh affinity purified rabbit antibody; YenZym). Anti-rabbit Alexa Fluor 568 (A11036; Invitrogen) secondary antibody was used. Nuclear counterstaining was performed with DAPI. Images were collected with a 40× objective and 1.5 or 2× zoom, when indicated, using a Nikon A1R confocal microscope.

Immunoblotting

Protein lysates were obtained from CTRL- and CSE-BJ1 cultured in 10-cm dishes. Protein extraction, quantification, and immunoblotting were performed as described previously (19). Primary antibodies used included anti-MCT4 (sc-50329; Santa Cruz Biotechnology), anti-GLB1 (ab96239; Abcam), anti-HMGB1 (NB100-2322; Novus Biologicals), anti-NF-κB (3033; Cell Signaling Technology), anti-pRB (9308; Cell Signaling Technology), anti-TIGAR (ab37910; Abcam), anti-β-actin (A5441; Sigma-Aldrich), anti-β-tubulin (T4026; Sigma-Aldrich), and anti-vinculin (#4650; Cell Signaling Technology).

Lactate assay

After 4 days in standard growth media or 5% CSE, with or without 30 μg/mL of glutathione, media from BJ1 fibroblasts cultured in 12-well plates were replaced by 400 μL of phenol-free DMEM 10% FBS. After 16 hours, cell media were collected and l-lactate levels were assessed according to the manufacturer's instructions using the EnzyChromTM L-Lactate Assay Kit (ECLC-100; BioAssay Systems). Results were normalized for total cell number.

ROS assay

After 4 days in standard growth media or 5% CSE, with or without 30 μg/mL of glutathione, media from BJ1 fibroblasts cultured in 12-well plates were replaced by DMEM 10% FBS containing 5 μmol/L CellROX Deep Red Reagent (C10422; Invitrogen). After 30-minute incubation at 37°C, BJ1 were harvested and CellROX fluorescence was detected by flow cytometry using an APC signal detector. Analysis was performed with FlowJo software.

Cell viability assessment

BJ1 fibroblasts were washed twice with PBS and fixed with 4% paraformaldehyde (PFA) for 10 minutes at room temperature. Fixed cells were then rinsed with PBS and stained with 0.5% crystal violet (CV, C0775; Sigma) in 10% ethanol. CV was removed and cells were washed with deionized water. Cells were left to dry and then photographed under an inverted phase microscope.

Flow cytometry assessment of metabolic markers

After 4 days of coculture in 6-well plates, cells were collected and subject to immunostaining and flow cytometry analysis, as described previously (11). Primary antibodies used were anti-MCT1 (YenZym), anti-TOMM20 (sc-17764; Santa Cruz Biotechnology), and anti-MCT4 (YenZym). Secondary antibodies used were anti-rabbit and anti-mouse conjugated to an APC or PE fluorophore, as indicated. Cell populations were separated by GFP detection when indicated.

Apoptosis and cell death assessment

After 4 days of coculture in 12-well plates, the culture media and cells were collected, centrifuged, and resuspended in Annexin-V (AnnV) binding buffer containing Ann-V-APC conjugate (550474; BD Biosciences) and propidium iodide (PI; KPL 71-04-01; SeraCare). Populations of cells were separated by GFP detection. AnnV and PI staining were assessed by flow cytometry and analysis was performed with FlowJo software. The same protocol was followed on monocultures of WT and MCT4-KO MEFs treated with or without CSE and GSH for 2 days. Quadruplicates were run for each condition.

Transwell migration assay

Twenty-four–well polycarbonate plates with 6.5-mm diameter inserts and 8-μm pore–sized membranes (3422; Costar) were used to study the migratory capabilities of carcinoma cells and macrophages. A total of 5 × 105 CAL27 or FaDu cells or 106 RAW 264.7 cells were added to the top of the insert, and 1 × 105 CTRL- and CSE-BJ1 or 15 × 104 MEFs were seeded in the bottom chamber, all in 10% Nu-serum DMEM. Cells were allowed to migrate for 24 hours and membranes were stained with crystal violet. Images were then taken under an inverted phase microscope and staining was quantified with ImageJ.

Analysis of CCL2 mRNA levels

Total RNA from WT and MCT4-KO MEFs exposed to CTRL- or 5% CSE-media was isolated using TRIzol reagent (15596026; Invitrogen) following manufacturer's instructions. Total RNA concentration and purity was determined using NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific). Reverse transcription PCR (RT-PCR) was performed using a High-Capacity cDNA Reverse Transcription Kit (4368814; Applied Biosystems) and carried out in a Bio-Rad PTC-100 thermal cycler. cDNA obtained was used to perform real-time PCR target amplification. qPCR master mixes were prepared using TaqMan Universal PCR Master Mix (4304437; Applied Biosystems) according to manufacturer's recommendations and loaded into MicroAmp Optical 384-Well reaction plates (4483319; Applied Biosystems). Inventoried TaqMan Gene expression assays CCL2 (Mm00441242_m1), β-actin (Mm00607939_s1), and GAPDH (Mm00607939_s1) were used. Each qPCR reaction was performed in triplicates. Plates were analyzed in a QuantStudio 12K Flex Real-Time PCR System (Thermo Fisher Scientific). mRNA expression levels were calculated using the relative comparison CT (2−ΔΔCt) method after normalization with GAPDH as endogenous control.

IHC

Excised tumors were frozen with liquid nitrogen in Tissue Tek O.C.T. (4583; Electron Microscopy Sciences) and sectioned at 6-μm thickness. Sections were fixed with 4% PFA in PBS for 10 minutes at room temperature, washed with PBS, and blocked with 5% BSA/PBS for 30 minutes. Endogenous biotin activity was blocked using the Avidin-Biotin Kit (AB972 L; Biocare Medical), and endogenous peroxidase activity with 0.3% H2O2 0.3% sodium azide in PBS for 15 minutes. Primary antibodies anti-CD45 (550539; BD Biosciences) and anti-CD68 (137001; BioLegend) were incubated for 1 hour at room temperature. Biotinylated anti-rat secondary antibody (BA-4001; Vector Laboratories), and avidin-horseradish peroxidase complex (Vectastain Elite ABC kit, PK-6100; Vector Laboratories) were each incubated for 30 minutes. Antibody reactivity was detected using liquid DAB substrate chromagen (K346711-2; Agilent) and counterstaining of nuclei was performed with Tacha's hematoxylin (NH-HEM M; Biocare Medical). Representative images were taken at a 20× magnification and quantified by ImageJ.

Quantification with ImageJ

CD45 and CD68 by IHC and crystal violet (CV) staining were quantified with ImageJ. RGB images were first split into single-color channels with the “RGB stack” tool. Staining was highlighted using the “threshold” tool. Thresholds were adjusted to select all stained areas and the same values were maintained across all images for comparison. The “measure” tool calculated the percentage of stained area relative to the total area of the region of interest (ROI), which was the same for all images.

TUNEL assay and quantification of apoptotic cells in tumors

Frozen sections were fixed with 4% paraformaldehyde in PBS for 10 minutes, washed with PBS, and permeabilized with 30% acetic acid in 100% ethanol at −20°C for 5 minutes. After washing, sections were incubated with equilibration buffer (S7106; Millipore-Sigma) for 15 minutes at room temperature and then with ApopTag TdT enzyme (S7107; Millipore-Sigma) and reaction buffer (S7105; Millipore-Sigma) at a ratio of 30:70 for 30 minutes at 37°C. Sections were washed in PBS and incubated with anti-digoxigenin-POD (11207733910; Sigma) for 30 minutes at room temperature. TUNEL-positive cells were visualized with liquid DAB substrate kit (Agilent).

Apoptotic nuclei were quantified using Aperio software (Aperio). Digital images were captured with Leica and Aperio slide scanners under 320 magnification with an average scan time of 120 seconds (compression quality 70). A nuclear algorithm was used to identify TUNEL-positive cells and generate values for percentage of 3+ nuclei. Four tumors from each group were evaluated. In the CAL27 xenografts, 9 to11 areas of tumor cells were analyzed for each xenograft and quantified for a total of 38 areas in the control and 40 areas in CSE treated. For the FaDu xenografts, 10 to 15 areas of tumor cells were analyzed for each xenograft and quantified for a total of 50 areas in the control and 51 areas in CSE treated.

Animal studies

All animal work was approved by the Institutional Animal Care and Use Committee (IACUC) at Thomas Jefferson University (Philadelphia, PA). Male athymic nude mice aged 4 to 6 weeks (Charles River Laboratories) were maintained under standard pathogen-free conditions in our animal facility. Animals were provided with sterilized chow and water. All animal handling and experiments were conducted under the standards of the IACUC.

CAL27, FaDu, CTRL-BJ1, and CSE-BJ1 cells were cultured in parallel, then trypsinized and resuspended in PBS (Supplementary Fig. S1B). A total of 3 × 105 CAL27 or FaDu carcinoma cells were coinjected with 106 CTRL- or CSE-BJ1 fibroblasts bilaterally into the flanks of nude mice. Each group contained 5 mice and a total of 10 tumor xenografts. As controls, 5 mice were also injected with 106 CTRL- or CSE-BJ1 alone. Starting on day 14, tumors were measured twice weekly with an electronic caliper. Mouse weight was also monitored biweekly. Mice were sacrificed and tumors harvested at postinjection days 13 (CAL27) or 21 (FaDu) as an early timepoint, and day 24 as the endpoint. Excised tumor measurements were taken by electronic caliper. Tumor volume (mm3) was calculated by the formula V = (x2y/2), where v is the volume of the tumor, x is the length of the short axis, and y is the length of the long axis.

Male C57Bl/6 mice were purchased from Taconic Farms. A total of 105 MTEC cells were coinjected with 106 WT-MEF or MCT4-KO MEF bilaterally in the flank. Two different sets of coinjections were performed at different times: coinjections with CSE-unexposed WT and MCT4-KO MEFs (control MEFs) and coinjections with CSE-treated WT and MCT4-KO MEFs (CSE MEFs). Both sets were allowed to grow for a period of 28 days. Tumor volume and weight were measured at the time of collection.

Human study

This study was approved by the institutional review board (IRB) at Thomas Jefferson University (Philadelphia, PA). Samples of primary tumors from 36 patients with head and neck cancer were obtained from archived paraffin-embedded tissue blocks for histologic analysis. Patient data were collected including: age, sex, tobacco use, HPV status, tumor primary site, and subsite.

MCT4 IHC and analyses in head and neck cancer samples

Paraffin sections were deparaffinized, rehydrated, and antigen retrieval was performed in 0.01 mol/L Tris-EDTA, pH 9.0 with 0.1% Tween-20 for 10 minutes using a pressure cooker. The sections were blocked with 3% H2O2 and for endogenous biotin; then incubated with 10% goat serum (Vector Laboratories) at 4°C overnight. MCT4 antibody (sc-376140; Santa Cruz Biotechnology) was applied for 1 hour, followed by biotinylated secondary antibody and avidin–biotin–HRP complex. A DAB liquid substrate kit was used to detect binding and slides were stained with hematoxylin and mounted.

MCT4 staining was evaluated in a blinded fashion by two independent observers. Multiple fields for each slide were viewed at low and high power and an overall score of either ≤50% or >50% staining in cancer-associated stroma was determined. Samples with disparate scores were reviewed and a score was jointly determined. For immune-rich versus immune-poor status, MCT4 and hematoxylin and eosin–stained slides were evaluated in a similar fashion. The samples were labeled as having an immune-rich stroma if greater than 50% of the stromal cells were immune cells. Samples with disparate scores were reviewed jointly and in addition by an expert pathologist (M. Tuluc).

Statistical analysis

Statistical significance was examined using the Student t test and Fisher exact text. Differences were considered statistically significant at P < 0.05.

CSE induces senescence and glycolysis in fibroblasts

To examine the effects of CSE exposure on the metabolism and stress response mechanisms of BJ1 human skin fibroblasts, cells were exposed to 5% CSE (CSE-BJ1) or standard media (CTRL-BJ1) for 6 days. After CSE exposure, changes in BJ1 proliferation and cell morphology were observed. Compared with untreated cells, CSE-BJ1 were less confluent and cells were enlarged (Fig. 1A), which are two features of cellular senescence. Decreased proliferation rates of CSE-BJ1 were consistent with decreased expression of phosphorylated retinoblastoma protein (p-RB), an inducer of cell-cycle progression. CSE-BJ1 had decreased pRB, which prevents G1 progression through the cell cycle, compared with CTRL-BJ1 (Fig. 1B). CSE-BJ1 were also more senescent and showed increased β-galactosidase expression and loss of nuclear HMGB1 (note intranuclear dots) compared with CTRL-BJ1 by immunofluorescence (Fig. 1C) and Western blot (Fig. 1D). In terms of metabolism, CSE-BJ1 had increased MCT4 expression compared with CTRL-BJ1 (Fig. 1E and F), indicating a switch toward a glycolytic metabolism. To confirm the increased glycolytic flux in CSE-BJ1, we determined the levels of lactate secretion, which is the endproduct of glycolysis, and intracellular ROS levels, which is a driver of MCT4 expression. CSE-BJ1 increased lactate secretion compared with controls (Fig. 1G), consistent with an augmented glycolytic metabolism of fibroblasts upon CSE exposure. Intracellular ROS levels were also increased in CSE-BJ1 (Fig. 1H), correlating with the upregulation of MCT4 expression.

Figure 1.

CS induces senescence and a metabolic switch toward glycolysis in fibroblasts. BJ1 fibroblasts were cultured in standard or CSE media (CTRL-BJ1 and CSE-BJ1, respectively) for 6 days, refreshing the media every 2 days. A, Visual assessment of fibroblast density and morphology under the miscroscope at 10×. B, Analysis of the proliferation marker p-RB by Western blot. Analysis of the markers of senescence β-gal and HMGB1 by immunofluorescence (C) and western blot (D). Analysis of the marker of glycolysis MCT4 by immunofluorescence (E) and Western blot analysis (F). In all immunofluorescence staining, β-gal, HMGB1 and MCT4 are shown in red, and nuclei are shown in blue (DAPI). DAPI channel intensity was removed in HMGB1 zoom-ins to appreciate nuclear staining. Confocal microscopy images were acquired at the 40× magnification, with additional 2× zoom for HMGB1 and MCT4. Scale bar, 50 μm (β-gal) or 20 μm (HMGB1 and MCT4). G, Assessment of lactate levels in media of CTRL- and CSE-BJ1 by colorimetric assay. H, Assessment of intracellular ROS levels of CTRL- and CSE-BJ1 by flow cytometry.

Figure 1.

CS induces senescence and a metabolic switch toward glycolysis in fibroblasts. BJ1 fibroblasts were cultured in standard or CSE media (CTRL-BJ1 and CSE-BJ1, respectively) for 6 days, refreshing the media every 2 days. A, Visual assessment of fibroblast density and morphology under the miscroscope at 10×. B, Analysis of the proliferation marker p-RB by Western blot. Analysis of the markers of senescence β-gal and HMGB1 by immunofluorescence (C) and western blot (D). Analysis of the marker of glycolysis MCT4 by immunofluorescence (E) and Western blot analysis (F). In all immunofluorescence staining, β-gal, HMGB1 and MCT4 are shown in red, and nuclei are shown in blue (DAPI). DAPI channel intensity was removed in HMGB1 zoom-ins to appreciate nuclear staining. Confocal microscopy images were acquired at the 40× magnification, with additional 2× zoom for HMGB1 and MCT4. Scale bar, 50 μm (β-gal) or 20 μm (HMGB1 and MCT4). G, Assessment of lactate levels in media of CTRL- and CSE-BJ1 by colorimetric assay. H, Assessment of intracellular ROS levels of CTRL- and CSE-BJ1 by flow cytometry.

Close modal

α-naphthoflavone mimics the effects of CSE on fibroblasts

Because CSE induces ROS generation and increases MCT4 expression in our model, we investigated whether the aryl hydrocarbon receptor (AhR), which is involved in compensatory responses to inflammation and oxidative stress, was able to regulate MCT4 expression. We pharmacologically modulated the AhR in unexposed BJ1 fibroblasts and assessed MCT4 and NF-κB expression, to determine the inflammatory state of the cell. Treatment of BJ1 fibroblasts with the AhR agonist l-KYN decreased the expression of both p65 subunit of NF-κB and MCT4 (Fig. 2A and B). Conversely, the AhR antagonist α-naphthoflavone (α-NF) triggered a dose-dependent upregulation of both p65 subunit of NF-κB and MCT4 expression (Fig. 2C and D). These data in BJ1 fibroblasts are consistent with the known anti-inflammatory role of the AhR and suggest that the AhR signaling pathway regulates ROS scavenging and, consequently, the levels of MCT4. To determine whether these modulators of the AhR have effects on CSE-exposed fibroblasts, BJ1 cells were treated with 5% CSE or control media, containing increasing concentrations of l-KYN or α-NF, and apoptosis and cell death rates were quantified. As expected, CSE-cultured fibroblasts, regardless of drug treatment, presented higher apoptosis and cell death rates than fibroblasts cultured in control media (Fig. 2E and F). l-KYN had no effects on fibroblast viability in either CTRL- or CSE-BJ1 (Fig. 2E). α-NF did not mimic the effects of the CSE on CTRL-BJ1; however, it increased cell death and apoptosis in CSE-BJ1 in a dose-dependent manner (Fig. 2F). This suggests that the anti-inflammatory effects derived from the AhR agonist l-KYN are not sufficient to protect fibroblasts from the CSE, whereas antagonizing the AhR signaling with α-NF has an additive effect to the detrimental effect of the CSE on fibroblasts.

Figure 2.

Signaling through the aryl hydrocarbon receptor (AhR) modulates expression of MCT4 in fibroblasts. BJ1 fibroblasts were cultured in CTRL- and CSE media and treated with increasing concentrations of the AhR agonist l-KYN or the antagonist α-NF. A, Western blot assessment of MCT4 and NF-κB expression in CTRL-BJ1 untreated or treated with 50 and 100 μmol/L of l-KYN. B, Immunofluorescence staining of MCT4 (red) in CTRL-BJ1 untreated or treated with 100 μmol/L l-KYN. Merged images show GFP-expressing BJ1 in green and nuclei stained with DAPI in blue. Scale bar, 50 μm. C, Western blot assessment of MCT4 and NF-κB expression in CTRL-BJ1 treated with 5 and 10 μmol/L of α-NF or vehicle DMSO. D, Immunofluorescence staining of MCT4 (red) in CTRL-BJ1 treated with 10 μmol/L α-NF or vehicle DMSO. Merged images show GFP-expressing BJ1 in green and nuclei stained with DAPI in blue. Scale bar, 50 μm. E and F, Apoptosis and cell death percentages assessment of CTRL- and CSE-BJ1 treated with l-KYN (E) or α-NF (F). Apoptosis was detected by AnnV staining and cell death by propidium iodide (PI) staining by flow cytometry.

Figure 2.

Signaling through the aryl hydrocarbon receptor (AhR) modulates expression of MCT4 in fibroblasts. BJ1 fibroblasts were cultured in CTRL- and CSE media and treated with increasing concentrations of the AhR agonist l-KYN or the antagonist α-NF. A, Western blot assessment of MCT4 and NF-κB expression in CTRL-BJ1 untreated or treated with 50 and 100 μmol/L of l-KYN. B, Immunofluorescence staining of MCT4 (red) in CTRL-BJ1 untreated or treated with 100 μmol/L l-KYN. Merged images show GFP-expressing BJ1 in green and nuclei stained with DAPI in blue. Scale bar, 50 μm. C, Western blot assessment of MCT4 and NF-κB expression in CTRL-BJ1 treated with 5 and 10 μmol/L of α-NF or vehicle DMSO. D, Immunofluorescence staining of MCT4 (red) in CTRL-BJ1 treated with 10 μmol/L α-NF or vehicle DMSO. Merged images show GFP-expressing BJ1 in green and nuclei stained with DAPI in blue. Scale bar, 50 μm. E and F, Apoptosis and cell death percentages assessment of CTRL- and CSE-BJ1 treated with l-KYN (E) or α-NF (F). Apoptosis was detected by AnnV staining and cell death by propidium iodide (PI) staining by flow cytometry.

Close modal

The antioxidant glutathione abrogates the effects of CSE

As ROS seem to be driving the effects of CSE on fibroblasts, we next aimed to assess the effects of the antioxidant glutathione (GSH) on cell viability, ROS levels, MCT4 expression, and lactate efflux in our model. Treatment with GSH improved the viability of CSE-exposed fibroblasts, as seen by crystal violet staining (Fig. 3A). Regarding ROS levels, CSE-BJ1 have greatly increased intracellular ROS levels compared with CTRL-BJ1 (Fig. 3B, left). GSH treatment reduces ROS levels by 80% (Fig. 3B, right). MCT4 expression, quantified by flow cytometry, followed the same trend as ROS levels. CSE-BJ1 not treated with GSH increased MCT4 levels by almost 3-fold respect to CTRL-BJ1 (Fig. 3C, left), whereas addition of the antioxidant attenuated this response to 1.8-fold (Fig. 3C, right). The decrease in MCT4 expression observed in the GSH-treated cells was accompanied by a decrease in lactate levels in media of CSE-BJ1 (Fig. 3D). The antioxidant NAC, precursor of glutathione, elicits the same effects as GSH in terms of increasing cell viability (Supplementary Fig. S2A) and reversing MCT4 expression (Supplementary Fig. S2B) upon CSE exposure. Moreover, NAC reverses the effects of CSE on the expression of the TP53-induced glycolysis and apoptosis regulator (TIGAR) protein, which is involved in the antioxidant defense and apoptosis suppression (Supplementary Fig. S2C). This data suggests that fibroblasts increase MCT4 expression and lactate production in response to elevated ROS levels induced by the CSE treatment.

Figure 3.

The antioxidant glutathione abrogates the effects of CSE. Fibroblasts were cultured in CTRL or CSE media and treated with 30 μg/mL of liposomal glutathione (GSH). A, Crystal violet (CV) staining of CTRL- and CSE-BJ1 untreated or treated with GSH. B and C, Assessment of intracellular ROS levels (B), MCT4 expression (C), and secreted lactate levels (D) in BJ1 fibroblasts, by flow cytometry. E, Assessment of cell viability in WT and MCT4-KO MEF exposed to CSE in the presence or absence of GSH. Left, flow cytometry contour plots of representative samples. AnnV staining intensity is represented in the x-axis and PI staining on the y-axis. Dead cells are found in quadrant 1 (Q1), apoptotic cells in Q2+Q3, and live cells in Q4. Right, quantification of quadrants Q1–Q3 corresponding to apoptotic and dead cells (AnnV and/or PI positive).

Figure 3.

The antioxidant glutathione abrogates the effects of CSE. Fibroblasts were cultured in CTRL or CSE media and treated with 30 μg/mL of liposomal glutathione (GSH). A, Crystal violet (CV) staining of CTRL- and CSE-BJ1 untreated or treated with GSH. B and C, Assessment of intracellular ROS levels (B), MCT4 expression (C), and secreted lactate levels (D) in BJ1 fibroblasts, by flow cytometry. E, Assessment of cell viability in WT and MCT4-KO MEF exposed to CSE in the presence or absence of GSH. Left, flow cytometry contour plots of representative samples. AnnV staining intensity is represented in the x-axis and PI staining on the y-axis. Dead cells are found in quadrant 1 (Q1), apoptotic cells in Q2+Q3, and live cells in Q4. Right, quantification of quadrants Q1–Q3 corresponding to apoptotic and dead cells (AnnV and/or PI positive).

Close modal

To examine whether the increase in MCT4 expression was linked to viability of CSE-exposed fibroblasts, we exposed MEFs isolated from MCT4 knockout mouse embryonal fibroblasts (MCT4-KO MEF) and their wild-type (WT) counterparts (WT-MEF) to CSE, in the presence or absence of GSH, and assessed apoptosis and cell death rates (Fig. 3E). MCT4-KO MEFs under baseline conditions have slightly higher apoptosis and cell death levels than WT MEF (1.2 fold, P = 0.04). This difference is not significant when unexposed MEFs are treated with GSH (1.2 fold, n.s.). Both genotypes of MEFs have increased apoptosis and cell death rates when treated with CSE. However, MCT4-KO MEF are significantly more sensitive to CSE with 90% higher apoptosis and cell death rates than WT-MEFs. Treatment of CSE-exposed MEFs with GSH greatly reduces cell death and apoptosis in both genotypes. Interestingly, treatment with GSH completely reverses the detrimental effects of CSE on MCT4-KO MEF cell death and apoptosis. Altogether, these data suggests that CSE-induced ROS drives the glycolytic phenotype of fibroblasts, and that MCT4 upregulation is partially necessary for survival of fibroblasts under conditions of CSE exposure.

CSE fibroblasts reprogram cancer cell metabolism

Metabolic compartmentalization is a common feature in head and neck cancers. Therefore, we wanted to determine whether CSE-exposed fibroblasts with high MCT4 were able to reprogram the metabolism of carcinoma cells toward a more oxidative profile. CAL27 and FaDu cells were cocultured with CTRL- or CSE-BJ1 for 4 days and the markers of mitochondrial metabolism MCT1 and TOMM20 were assessed by flow cytometry. Coculture with CSE-BJ1 increased both MCT1 and TOMM20 expression in CAL27 and FaDu carcinoma cells (Fig. 4A and B). For CAL27, intensity of APC+ staining was quantified by gating the MCT1+ (Fig. 4A, top) and TOMM20+ (Fig. 4A, bottom) cells. For FaDu, percentage of MCT1+ (Fig. 4B, top) and TOMM20+ (Fig. 4B, bottom) cells was quantified. Upregulation of carcinoma cell MCT1 could be visually detected by immunofluorescence staining of CAL27 (Supplementary Fig. S3A) and FaDu (Supplementary Fig. S3B) cocultures with CSE-BJ1. Moreover, to ensure that the glycolytic phenotype of CSE-BJ1 was preserved after CSE exposure, we assessed the expression of MCT4 in fibroblasts after 4 days in coculture. CSE-BJ1 maintain MCT4 upregulation after 4 days in coculture with CAL27 (Fig. 4C) and FaDu (Fig. 4D) cells. We also assessed whether soluble factors secreted by BJ1 could be mediating these effects on carcinoma cells. We treated homotypic cultures of CAL27 and FaDu cells with conditioned media (CM) from CTRL- or CSE-BJ1. CSE-BJ1 CM elicited similar but lessened effects on carcinoma cell MCT1 and TOMM20 in both CAL27 (Supplementary Fig. S3C) and FaDu (Supplementary Fig. S3D) cells. The same effects were seen in transwell cocultures of MTEC cells with CTRL- or CSE- WT MEFs. WT MEFs exposed to CSE were able to induce MCT1 expression in MTEC cells (Fig. 4E) while MEFs maintained high MCT4 expression 4 days post-CSE-treatment (Fig. 4F).

Figure 4.

CSE fibroblasts induce markers of mitochondrial metabolism on carcinoma cells. A–D, CAL27 and FaDu cells were cocultured with CTRL- or CSE-BJ1 and cancer cell MCT1 and TOMM20, and fibroblast MCT4 expression were assessed by flow cytometry. A, Flow cytometry plots and quantification of MCT1 (top) and TOMM20 (bottom) staining intensity in CAL27 cocultures. B, Flow cytometry plots and quantification of percentage of cells expressing MCT1 (top) and TOMM20 (bottom) in FaDu cocultures. MCT4 expression in CTRL- and CSE-BJ1 after 4 days in co-culture with CAL27 (C) and FaDu (D) cells. E and F, MTEC cells were co-cultured with WT MEF previously exposed to CTRL or CSE media, and cancer cell MCT1 and fibroblast MCT4 expression were assessed by flow cytometry. E, Flow cytometry plots and quantification of MCT1 staining intensity in MTEC cells. F, Flow cytometry plots and quantification of MCT4 staining intensity in WT MEFs 4 days post-CTRL or CSE treatment. For all markers, staining intensity was gated into APC− and APC+ or PE− and PE+ based on the signal into the APC and PE channels, respectively, emitted by the unstained control. Only APC+ and PE+ populations were used for quantification.

Figure 4.

CSE fibroblasts induce markers of mitochondrial metabolism on carcinoma cells. A–D, CAL27 and FaDu cells were cocultured with CTRL- or CSE-BJ1 and cancer cell MCT1 and TOMM20, and fibroblast MCT4 expression were assessed by flow cytometry. A, Flow cytometry plots and quantification of MCT1 (top) and TOMM20 (bottom) staining intensity in CAL27 cocultures. B, Flow cytometry plots and quantification of percentage of cells expressing MCT1 (top) and TOMM20 (bottom) in FaDu cocultures. MCT4 expression in CTRL- and CSE-BJ1 after 4 days in co-culture with CAL27 (C) and FaDu (D) cells. E and F, MTEC cells were co-cultured with WT MEF previously exposed to CTRL or CSE media, and cancer cell MCT1 and fibroblast MCT4 expression were assessed by flow cytometry. E, Flow cytometry plots and quantification of MCT1 staining intensity in MTEC cells. F, Flow cytometry plots and quantification of MCT4 staining intensity in WT MEFs 4 days post-CTRL or CSE treatment. For all markers, staining intensity was gated into APC− and APC+ or PE− and PE+ based on the signal into the APC and PE channels, respectively, emitted by the unstained control. Only APC+ and PE+ populations were used for quantification.

Close modal

Altogether, our results show that the CSE-induced metabolic reprogramming of fibroblasts is sufficient to trigger a switch toward oxidative metabolism in carcinoma cells, and these effects are at least, in part, mediated by soluble factors.

Coculture with CSE fibroblasts increases features of tumor aggressiveness

Tumor aggressiveness is defined both by the intrinsic hallmarks of cancer cells and by the cancer-promoting effects of the surrounding stroma. To determine whether the reprogramming of fibroblasts following CSE exposure would produce changes in carcinoma cell aggressiveness, early and late apoptosis and cell death rates were quantified in CAL27 and FaDu cells after 4 days in coculture with CTRL- or CSE-BJ1. Flow cytometry assessment of AnnV and PI staining showed that coculture with CSE-BJ1 led to less carcinoma cell death and apoptosis in both CAL27 and FaDu cells (Fig. 5A), with a 50% and 60% decrease, respectively, from cocultures with CTRL-BJ1. Figure 5B shows the contour plots from three representative samples of each group. Note how in the cocultures of CAL27 and FaDu cells with CSE-BJ1 there is a reduction in the cell population in the first (Q1), second (Q2), and third quadrants (Q3), corresponding to the dead cells (Q1) and late- and early-apoptotic (Q2+Q3) carcinoma cells.

Figure 5.

CSE-exposed fibroblasts increase features of tumor aggressiveness. A and B, CAL27 and FaDu cells were cocultured with CTRL- or CSE-BJ1 and apoptosis and cell death rates were assessed by flow cytometry and cell migration by a transwell assay. A, Quantification of quadrants Q1–Q3 (from B) corresponding to apoptotic and dead CAL27 and FaDu carcinoma cells (AnnV and/or PI positive). B, Contour plots of AnnV (x-axis) and PI (y-axis) staining in CAL27 and FaDu cells cocultured with fibroblasts. Dead cells are found in quadrant 1 (Q1), apoptotic cells in Q2+Q3, and live cells in Q4. C, Assessment of the migratory ability of CAL27 and FaDu carcinoma cells stimulated by CTRL- or CSE-BJ1. On the left, crystal violet (CV) staining of the cells that migrated through the pores of the membrane; on the right, quantification of the percentage area of the membrane stained by CV. D–F, Assessment of the effects of CSE-fibroblasts on immune cell migration. D, Quantification of CCL2 mRNA levels in WT and MCT4-KO MEF exposed to CTRL or CSE media. E, Transwell migration assay with CV staining of RAW 264.7 cells that migrated through the membrane in response to WT MEFs previously exposed to CTRL or CSE media. F, Flow cytometry plots and quantification of MCT4 staining intensity in RAW 264.7 exposed to CTRL or CSE media.

Figure 5.

CSE-exposed fibroblasts increase features of tumor aggressiveness. A and B, CAL27 and FaDu cells were cocultured with CTRL- or CSE-BJ1 and apoptosis and cell death rates were assessed by flow cytometry and cell migration by a transwell assay. A, Quantification of quadrants Q1–Q3 (from B) corresponding to apoptotic and dead CAL27 and FaDu carcinoma cells (AnnV and/or PI positive). B, Contour plots of AnnV (x-axis) and PI (y-axis) staining in CAL27 and FaDu cells cocultured with fibroblasts. Dead cells are found in quadrant 1 (Q1), apoptotic cells in Q2+Q3, and live cells in Q4. C, Assessment of the migratory ability of CAL27 and FaDu carcinoma cells stimulated by CTRL- or CSE-BJ1. On the left, crystal violet (CV) staining of the cells that migrated through the pores of the membrane; on the right, quantification of the percentage area of the membrane stained by CV. D–F, Assessment of the effects of CSE-fibroblasts on immune cell migration. D, Quantification of CCL2 mRNA levels in WT and MCT4-KO MEF exposed to CTRL or CSE media. E, Transwell migration assay with CV staining of RAW 264.7 cells that migrated through the membrane in response to WT MEFs previously exposed to CTRL or CSE media. F, Flow cytometry plots and quantification of MCT4 staining intensity in RAW 264.7 exposed to CTRL or CSE media.

Close modal

Cell migration is another common feature of cancer aggressiveness. We assessed the migratory abilities of CAL27 and FaDu cells in response to CTRL- and CSE-BJ1 stimuli with a transwell assay. Carcinoma cells were seeded on top of the porous membrane in the top chamber of the insert, and CTRL- and CSE-BJ1 were seeded on the bottom chamber. Crystal violet staining of the insert membranes revealed that CSE-BJ1 increased carcinoma cell migration through the pores by 20% in CAL27 and 50% in FaDu cells (Fig. 5C), compared with CTRL-BJ1.

Enhanced replicative capacity is another hallmark of cancer aggressiveness. We assessed proliferation rates in carcinoma cells in coculture with CTRL- and CSE-BJ1. No differences were seen in cancer cell proliferation between the two types of coculture in either CAL27 or FaDu cells (Supplementary Fig. S4).

Macrophages constitute a large proportion of the tumor stroma and are associated with aggressive disease and poor prognosis in a variety of human cancers. We studied whether CSE fibroblasts were able to induce the recruitment of macrophages. We assessed the mRNA levels of the cytokine CCL2, involved in recruitment of monocytes, in WT and MCT4-KO MEFs unexposed or exposed to CSE. Both types of MEFs upregulated CCL2 expression in response to CSE; however, WT-MEF did it to a greater extent compared with MCT4-KO MEFs (Fig. 5D). Next, we determined the effects of CTRL- and CSE- WT MEF on the migration capabilities of the mouse macrophage cell line RAW 264.7 by a transwell assay. CSE-exposed MEFs increased the migration of RAW 264.7 cells compared with CTRL-MEFs (Fig. 5E). We were also interested in determining the effects of direct CS exposure to macrophages on MCT4 expression. MCT4 protein level was assessed by flow cytometry in RAW 267.4 after a 24-hour treatment with 5% CSE or CTRL media. MCT4 was highly upregulated in the macrophages exposed to CSE (Fig. 5F), indicating that the metabolic effects of CSE are not limited to fibroblasts and could be broadened to other cells within the stroma.

Coinjection of carcinoma cells with CSE fibroblasts increases tumor growth in vivo

To assess whether CSE-fibroblasts would have any effect on carcinoma cell growth in vivo, either CAL27 or FaDu cells were coinjected with CTRL- or CSE-BJ1 into the flanks of nude mice. Tumor size was followed and at day 24 mice were sacrificed and tumors harvested. Tumor growth curves show that coinjection with CSE-BJ1 increased tumor growth in both CAL27 (Supplementary Fig. S5A) and FaDu (Supplementary Fig. S5B) xenografts. Growth differences were significant as early as day 14, which was the first time point when tumor volumes could be measured. We collected tumors at two different timepoints for purposes of performing different types of studies, as specified below. At the early timepoint, tumor weights were already significantly different between CTRL- and CSE-BJ1 coinjections in both CAL27 (Supplementary Fig. S5C) and FaDu (Supplementary Fig. S5D) tumors. At the endpoint, xenografts generated from coinjections of carcinoma cells with CSE-BJ1 were larger than coinjections with CTRL-BJ1 (Fig. 6A). The volume of CAL27 + CSE-BJ1 xenografts were 2.1 fold (Fig. 6B), and the FaDu + CSE-BJ1 xenografts were 4.7 fold (Fig. 6C) larger than their respective controls.

Figure 6.

CSE exposure in fibroblasts promote tumor growth and inflammation. CAL27 and FaDu cells were coinjected with CTRL- and CSE-BJ1 into the flanks of immunocompromised nude mice. A, Images of harvested tumor xenografts for gross volume comparison. Volumes of harvested tumor xenografts generated from CAL27 (B) and FaDu (C) and fibroblast coinjections. D and E, Apoptosis rates were measured by staining (D) and quantification (E) of TUNEL in CAL27 and FaDu tumor xenografts. F and G, IHC assessment of the pan-lymphocyte marker, CD45, and macrophage marker, CD68, on CAL27 tumor samples. Representative areas of CD45 and CD68 infiltration (F) and quantification of percentage of stained area (G) in CAL27 xenografts.

Figure 6.

CSE exposure in fibroblasts promote tumor growth and inflammation. CAL27 and FaDu cells were coinjected with CTRL- and CSE-BJ1 into the flanks of immunocompromised nude mice. A, Images of harvested tumor xenografts for gross volume comparison. Volumes of harvested tumor xenografts generated from CAL27 (B) and FaDu (C) and fibroblast coinjections. D and E, Apoptosis rates were measured by staining (D) and quantification (E) of TUNEL in CAL27 and FaDu tumor xenografts. F and G, IHC assessment of the pan-lymphocyte marker, CD45, and macrophage marker, CD68, on CAL27 tumor samples. Representative areas of CD45 and CD68 infiltration (F) and quantification of percentage of stained area (G) in CAL27 xenografts.

Close modal

As we had seen differences in cancer apoptosis and cell death in coculture studies, we evaluated apoptosis rates in the tumors at the early timepoint, when both groups of tumors were actively growing. TUNEL staining revealed decreased cancer cell apoptosis rates in the CSE-BJ1 coinjections in both CAL27 and FaDu xenografts (Fig. 6D and E).

As studies show that CSE induces a proinflammatory state in fibroblasts, and our in vitro work demonstrates that CSE fibroblasts promote macrophage recruitment, we assessed the inflammatory infiltrates in the tumors at the endpoint. We performed IHC staining for the leucocyte marker CD45, and CD68, a marker of macrophages and other mononuclear phagocytes on frozen tumor sections. Xenografts from CAL27 coinjected with CSE fibroblasts had more CD45+ and CD68+ tumor infiltration than the control coinjection xenografts (Fig. 6F). Quantification of the IHC staining revealed a 4- and 2.1-fold increase in the percentage area invaded by CD45+ and CD68+ cells, respectively (Fig. 6G). FaDu xenografts had high intratumoral and intertumoral variability in macrophage infiltration, and quantification of the IHC staining showed no significant differences in the percentage of area invaded by CD45+ or CD68+ cells between CSE and CTRL groups (Supplementary Fig. S5E and S5F).

CSE-induced MCT4 in fibroblasts drives tumor growth

Finally, we wanted to investigate whether stromal MCT4 expression is a driver of the more aggressive growth in tumors observed when cancer cells were coinjected with CSE fibroblasts. We generated an allograft model by coinjecting MTEC cells with WT MEFs or MCT4-KO MEFs into the flanks of C57BL6 mice. No significant differences in tumor volume and weight were observed between coinjections of MTEC with control WT MEFs and control MCT4-KO MEFs at baseline (Fig. 7A). However, when using MEFs that had been treated with CSE, allografts generated from the coinjection with CSE WT MEF had larger volumes and weight than coinjections with CSE MCT4-KO MEFs (Fig. 7B).

Figure 7.

MCT4 expression in fibroblasts induced by CSE drives tumor growth. MTEC carcinoma cells were coinjected with unexposed or CSE-exposed WT and MCT4-KO MEFs into the flanks of C57BL/6 mice. A, Tumor volume (top) and weight (bottom) of harvested allografts generated from the coinjection of MTEC with control MEFs. B, Tumor volumes (top) and weights (bottom) of harvested allografts generated from the coinjection of MTEC with CSE-exposed MEFs.

Figure 7.

MCT4 expression in fibroblasts induced by CSE drives tumor growth. MTEC carcinoma cells were coinjected with unexposed or CSE-exposed WT and MCT4-KO MEFs into the flanks of C57BL/6 mice. A, Tumor volume (top) and weight (bottom) of harvested allografts generated from the coinjection of MTEC with control MEFs. B, Tumor volumes (top) and weights (bottom) of harvested allografts generated from the coinjection of MTEC with CSE-exposed MEFs.

Close modal

High MCT4 expression is associated with an immune-poor or fibroblast-rich tumor stroma

HNSCC samples from 36 subjects were stained for MCT4 to better understand its relationship with type of stromal infiltrate (immune-rich versus immune-poor), tobacco use, and HPV status. The mean age of the subjects was 64 with a range of 41 to 85 and 28 subjects were male and 8 were female. Supplementary Table S1 summarizes the subject's baseline characteristics. High stromal MCT4 expression was associated with the presence of immune-poor stroma (P < 0.015; Supplementary Table S2). The predominant stromal cell type in all cases of immune-poor stroma were fibroblasts. HPV+ disease was associated with an immune-rich stroma (P < 0.045), which is consistent with HPV+ disease arising in the lymphoid tissue of the base of tongue and tonsil. However, tobacco use was not associated with an immune-poor stroma. Also, no statistically significant differences were found in stromal MCT4 expression based on tobacco use or HPV status.

CS causes many changes in the microenvironment of head and neck carcinomas. Here, we show that CS induces phenotypic changes, including metabolic reprogramming, in fibroblasts, and that these altered fibroblasts are sufficient to reprogram the overall metabolism of HNSCC and induce its aggressiveness in vitro and in vivo. Moreover, we demonstrate that stromal monocarboxylate transporter MCT4 drives cancer aggressiveness in the context of CS and describe the aryl hydrocarbon receptor (AhR) pathway as a novel pathway regulating the expression of MCT4.

The tumor stroma supports cancer cells and, for example, CAFs promote proliferation, migration, invasion, and resistance to anticancer drugs in a wide array of cancers, including HNSCC (37). CAFs phenotypically differ from normal fibroblasts because they have increased remodeling of the extracellular matrix and increased levels of soluble factors and cytokines, which drive a host of effects on carcinoma cells (38). CAFs also have a reprogrammed metabolism with high glycolytic flux, autophagy, and senescence, which are cellular processes that provide a nutrient-rich environment for cancer cells (39). Metabolically reprogrammed CAFs have been identified in the tumor stroma of HNSCC and have been shown to promote malignant progression (9, 40, 41). Here, we demonstrate that CS metabolically reprograms fibroblasts. CSE exposure increases glycolysis rates in fibroblasts, as seen by upregulation of MCT4 and increases secretion of lactate, as well as inducing senescence. Generation of ROS drives glycolysis and senescence (35, 42) and this study demonstrates that oxidative stress is a mechanism by which CSE exerts these effects on fibroblasts.

MCT4 is a marker of tumor-associated fibroblasts in a number of cancers, including HNSCC, and is not present in the normal fibroblasts within the adjacent tissue (9). An MCT4-high stroma may be able to promote transformation by oxidative stress and lactate production. Increased MCT4 expression in the tumor stroma has been associated with decreased overall survival or decreased disease-free survival in several human malignancies (43). In HNSCC, MCT4 expression in CAFs is associated with higher tumor stage (9) and development of invasive HNSCC in a chemical carcinogenesis model (36). Our cohort of patients revealed that tumors with a stroma rich in CAFs, as opposed to a stroma rich in immune cells, is associated with high stromal MCT4 expression. In addition, high stromal expression of the carbonic anhydrase IX (CA IX), another marker of oxidative stress and glycolysis, is associated with a poor prognosis in HNSCC (44). There is an interest in developing therapies against CAFs or the elements participating in the cross-talk between stromal and cancer cells (37). Here we report ways to modulate fibroblast MCT4 to revert the CAF phenotype induced by CS. The AhR pathway regulates inflammatory responses (29) and protects fibroblasts from CSE-induced oxidative stress (26–28). Stimulation of the AhR with the agonist l-KYN decreased the proinflammatory transcription factor NF-κB, consistent with the anti-inflammatory role of AhR, at the same time that it downregulated MCT4 expression. However, l-KYN was not able to promote viability of fibroblasts exposed to smoke. Conversely, the antagonist α-NF increased both NF-κB and MCT4 expression, and had an additive effect to CS. This is the first publication demonstrating that signaling through the AhR modulates MCT4 expression. In addition, we studied the effects of antioxidants in our model. Treatment with glutathione abrogated the effects of CSE on ROS levels, MCT4 expression, and lactate secretion in fibroblasts. Also, fibroblasts lacking MCT4 had increased rates of CSE-induced apoptosis and cell death compared with MCT4-expressing fibroblasts, an effect abolished by glutathione treatment. Altogether, our data demonstrate that oxidative stress generated by CSE exposure is a main driver of MCT4 upregulation in fibroblasts, and that MCT4 expression is necessary for fibroblast survival under these conditions.

Cancer cells reprogram stromal cells to derive nutrients, establishing a metabolic cooperation that favors tumor progression (45). The coexistence of reprogrammed CAFs with high glycolytic metabolism and proliferative cancer cells with high mitochondrial oxidative metabolism is known as “two-compartment tumor metabolism.” Metabolic compartmentalization with high stromal MCT4 and/or high cancer cell MCT1 and TOMM20, markers of mitochondrial metabolism, occurs in HNSCC (9). Coculture models in vitro and coinjection models in vivo are useful to study the influence of the stroma on carcinoma cells (7, 11, 41). In breast cancer, a coculture model of fibroblasts and human breast adenocarcinoma cells showed that the lactate released by fibroblasts induces mitochondrial biogenesis in adjacent carcinoma cells (46). Here we saw that carcinoma cells cocultured with CSE fibroblasts upregulated MCT1 and TOMM20 expression, compared with carcinoma cells cocultured with unexposed fibroblasts. Moreover, culture of carcinoma cells with conditioned media (CM) from CSE fibroblasts also induced MCT1 and TOMM20 expression, compared to CM from CTRL fibroblasts. These data suggests that increased stromal MCT4 and lactate production, together with soluble factors, are sufficient to induce the metabolic reprograming of carcinoma cells.

When we studied the effects of reprogrammed fibroblasts on carcinoma cell aggressiveness, we found that cocultures of carcinoma cells with CSE-exposed fibroblasts greatly reduced apoptosis and cell death rates in the carcinoma compartment. Coinjection of carcinoma cells with CSE-exposed fibroblasts generated larger tumors than coinjection with unexposed fibroblasts. It is well described that tumor-associated fibroblasts enhance features of aggressiveness in HNSCC cells (38); however, we are the first to show that a microenvironment exposed to CS may be further permissive for HNSCC progression. We also demonstrated that MCT4 upregulation in CS fibroblasts is an important driver of this effect. Coinjection of syngeneic carcinoma cells with MCT4-KO MEFs exposed to CS generated smaller tumors than coinjections with CS-exposed WT MEFs. No significant difference was observed in the coinjections with unexposed WT and MCT4-KO MEFs, which can be explained by the low basal MCT4 levels in WT MEFs, resembling those of an MCT4-KO MEF. When WT MEFs are stimulated with CS, MCT4 levels rise, and a differential effect versus the MCT4-KO MEFs can be seen. Future studies will need to confirm that the difference in growth is due to low baseline levels of MCT4. We demonstrate that CS-reprogrammed fibroblasts are capable of switching cancer cell metabolism toward a more oxidative metabolism and this metabolic reprogramming in carcinoma cells may be also involved in the promotion of cancer cell aggressiveness. Further studies will need to be performed to assess the effects of mitochondrial-inhibitory drugs on tumor growth. HMGB1 arises as another candidate to mediate the cancer-promoting effects of CSE-exposed stroma. HMGB1 is a nuclear factor that can be released from the cell in response to oxidative stress (47). In our model, CS exposure decreases nuclear HMGB1, presumably releasing it to the extracellular media. Extracellular HMGB1 plays signaling functions through its RAGE receptor that have been shown to promote features of aggressiveness in HNSCC among other cancers (48–50). Moreover, HMGB1 acts also as a proinflammatory cytokine. Altogether, these results demonstrate that CS-induced reprogramming of fibroblasts creates a more favorable tumor microenvironment that induces a metabolic switch in carcinoma cells and promotes their aggressiveness.

HNSCC has abundant immune infiltration (51). CD68, a macrophage marker, is abundantly expressed in HNSCC tumor tissue, as opposed to normal tissue, and correlates with aggressive behavior, lymph node metastasis, increased incidence of extracapsular spread, and overall worse survival (52–54). We determined the effects of fibroblast CSE exposure on macrophage recruitment. CSE-exposed fibroblasts have increased levels of the chemokine CCL2 and are capable of inducing migration of macrophages in vitro. Moreover, we assessed the expression of the inflammatory markers CD45, which is a pan-leukocyte marker, and CD68 in our tumors. Despite being and immunocompromised host, nude mice have an intact innate immune system with functional macrophages. Tumors generated from the coinjection with CS fibroblasts had higher expression of CD45 and CD68, predominantly within the stroma, compared with coinjections with CTRL fibroblasts. Our data fits with other studies that show that exposure of mice to CSE induces macrophage recruitment in lung tissues and lung tumors, and that these macrophages are polarized to a protumorigenic phenotype (55, 56). In HNSCC, macrophages promote cell migration, invasion and lymph node metastasis in in vitro and in vivo models, and are associated with worse outcome in patients (53, 57, 58). Moreover, studies show the correlation between the density of CAFs and TAMs with poor outcome in HNSCC (59, 60). Furthermore, one study reported that protumorigenic macrophage polarization is promoted in HNSCC tumors with high lactic acid concentrations (61). Interestingly, CS fibroblasts in our model secrete much higher lactate than CTRL fibroblasts. Our data supports that CS-exposed CAFs with high MCT4 and lactate secretion are capable of recruiting tumor-associated macrophages (TAM) and promote tumor progression. Future studies will determine the role of MCT4 in the crosstalk between CAFs and TAMs, as well as the contribution of TAMs to tumor growth in the context of a CS-exposed stroma.

U. Martinez-Outschoorn reports receiving a commercial research grant from Otsuka Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M. Domingo-Vidal, D. Whitaker-Menezes, P. Tassone, N. Philp, J. Curry, U. Martinez-Outschoorn

Development of methodology: M. Domingo-Vidal, D. Whitaker-Menezes, C. Martos-Rus, P. Tassone, U. Martinez-Outschoorn

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Domingo-Vidal, D. Whitaker-Menezes, C. Martos-Rus, P. Tassone, M. Tuluc

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Domingo-Vidal, D. Whitaker-Menezes, C. Martos-Rus, P. Tassone, M. Tuluc, N. Philp, U. Martinez-Outschoorn

Writing, review, and/or revision of the manuscript: M. Domingo-Vidal, D. Whitaker-Menezes, C. Martos-Rus, P. Tassone, C.M. Snyder, N. Philp, J. Curry, U. Martinez-Outschoorn

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Domingo-Vidal, D. Whitaker-Menezes, J. Curry

Study supervision: M. Domingo-Vidal, J. Curry, U. Martinez-Outschoorn

The NCI of the NIH under award numbers K08 CA175193 (to U. Martinez-Outschoorn) and 5P30CA056036-17 supported this work. The following Sidney Kimmel Cancer Center Shared Resources supported by 5P30CA056036-17 were used to conduct this research: Bioimaging Shared Resource, Flow Cytometry Shared Resource, Lab Animal Shared Resource and Translational Pathology Shared Resource. AAO-HNSF Research Award number 411726 (to P. Tassone) also supported this work.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2016
.
CA Cancer J Clin
2016
;
66
:
7
30
.
2.
Hatcher
JL
,
Sterba
KR
,
Tooze
JA
,
Day
TA
,
Carpenter
MJ
,
Alberg
AJ
, et al
Tobacco use and surgical outcomes in patients with head and neck cancer
.
Head Neck
2016
;
38
:
700
6
.
3.
Chen
AM
,
Chen
LM
,
Vaughan
A
,
Sreeraman
R
,
Farwell
DG
,
Luu
Q
, et al
Tobacco smoking during radiation therapy for head-and-neck cancer is associated with unfavorable outcome
.
Int J Radiat Oncol Biol Phys
2011
;
79
:
414
9
.
4.
Jethwa
AR
,
Khariwala
SS
. 
Tobacco-related carcinogenesis in head and neck cancer
.
Cancer Metastasis Rev
2017
;
36
:
411
23
.
5.
Curry
JM
,
Sprandio
J
,
Cognetti
D
,
Luginbuhl
A
,
Bar-ad
V
,
Pribitkin
E
, et al
Tumor microenvironment in head and neck squamous cell carcinoma
.
Semin Oncol
2014
;
41
:
217
34
.
6.
Kansy
BA
,
Dissmann
PA
,
Hemeda
H
,
Bruderek
K
,
Westerkamp
AM
,
Jagalski
V
, et al
The bidirectional tumor–mesenchymal stromal cell interaction promotes the progression of head and neck cancer
.
Stem Cell Res Ther
2014
;
5
:
95
.
7.
Wheeler
SE
,
Shi
H
,
Lin
F
,
Dasari
S
,
Bednash
J
,
Thorne
S
, et al
Enhancement of head and neck squamous cell carcinoma proliferation, invasion, and metastasis by tumor-associated fibroblasts in preclinical models
.
Head Neck
2014
;
36
:
385
92
.
8.
Puram
SV
,
Tirosh
I
,
Parikh
AS
,
Patel
AP
,
Yizhak
K
,
Gillespie
S
, et al
Single-cell transcriptomic analysis of primary and metastatic tumor ecosystems in head and neck cancer
.
Cell
2017
;
171
:
1611
24
.
9.
Curry
JM
,
Tuluc
M
,
Whitaker-Menezes
D
,
Ames
JA
,
Anantharaman
A
,
Butera
A
, et al
Cancer metabolism, stemness and tumor recurrence: MCT1 and MCT4 are functional biomarkers of metabolic symbiosis in head and neck cancer
.
Cell Cycle
2013
;
12
:
1371
84
.
10.
Curry
J
,
Johnson
J
,
Tassone
P
,
Vidal
MD
,
Menezes
DW
,
Sprandio
J
, et al
Metformin effects on head and neck squamous carcinoma microenvironment: window of opportunity trial
.
Laryngoscope
2017
;
127
:
1808
15
.
11.
Tassone
P
,
Domingo-Vidal
M
,
Whitaker-Menezes
D
,
Lin
Z
,
Roche
M
,
Tuluc
M
, et al
Metformin effects on metabolic coupling and tumor growth in oral cavity squamous cell carcinoma coinjection xenografts
.
Otolaryngol Head Neck Surg
2018
;
158
:
867
77
.
12.
Verma
A
,
Rich
LJ
,
Vincent-Chong
VK
,
Seshadri
M
. 
Visualizing the effects of metformin on tumor growth, vascularity, and metabolism in head and neck cancer
.
J Oral Pathol Med
2018
;
47
:
484
91
.
13.
Nyunoya
T
,
Monick
MM
,
Klingelhutz
A
,
Yarovinsky
TO
,
Cagley
JR
,
Hunninghake
GW
. 
Cigarette smoke induces cellular senescence
.
Am J Respir Cell Mol Biol
2006
;
35
:
681
8
.
14.
Carnevali
S
,
Petruzzelli
S
,
Longoni
B
,
Vanacore
R
,
Barale
R
,
Cipollini
M
, et al
Cigarette smoke extract induces oxidative stress and apoptosis in human lung fibroblasts
.
Am J Physiol Lung Cell Mol Physiol
2003
;
284
:
L955
63
.
15.
Rennard
SI
,
Togo
S
,
Holz
O
. 
Cigarette smoke inhibits alveolar repair: a mechanism for the development of emphysema
.
Proc Am Thorac Soc
2006
;
3
:
703
8
.
16.
Semlali
A
,
Chakir
J
,
Rouabhia
M
. 
Effects of whole cigarette smoke on human gingival fibroblast adhesion, growth, and migration
.
J Toxicol Environ Health A
2011
;
74
:
848
62
.
17.
Colombo
G
,
Dalle-Donne
I
,
Orioli
M
,
Giustarini
D
,
Rossi
R
,
Clerici
M
, et al
Oxidative damage in human gingival fibroblasts exposed to cigarette smoke
.
Free Radic Biol Med
2012
;
52
:
1584
96
.
18.
Yang
GY
,
Zhang
CL
,
Liu
XC
,
Qian
G
,
Deng
DQ
. 
Effects of cigarette smoke extracts on the growth and senescence of skin fibroblasts in vitro
.
Int J Biol Sci
2013
;
9
:
613
23
.
19.
Salem
AF
,
Al-Zoubi
MS
,
Whitaker-Menezes
D
,
Martinez-Outschoorn
UE
,
Lamb
R
,
Hulit
J
, et al
Cigarette smoke metabolically promotes cancer, via autophagy and premature aging in the host stromal microenvironment
.
Cell Cycle
2013
;
12
:
818
25
.
20.
Sato
E
,
Koyama
S
,
Takamizawa
A
,
Masubuchi
T
,
Kubo
K
,
Robbins
RA
, et al
Smoke extract stimulates lung fibroblasts to release neutrophil and monocyte chemotactic activities
.
Am J Physiol
1999
;
277
:
L1149
57
.
21.
Martey
CA
,
Pollock
SJ
,
Turner
CK
,
O'Reilly
KM
,
Baglole
CJ
,
Phipps
RP
, et al
Cigarette smoke induces cyclooxygenase-2 and microsomal prostaglandin E2 synthase in human lung fibroblasts: implications for lung inflammation and cancer
.
Am J Physiol Lung Cell Mol Physiol
2004
;
287
:
L981
91
.
22.
D'Anna
C
,
Cigna
D
,
Costanzo
G
,
Ferraro
M
,
Siena
L
,
Vitulo
P
, et al
Cigarette smoke alters cell cycle and induces inflammation in lung fibroblasts
.
Life Sci
2015
;
126
:
10
8
.
23.
Moldeus
P
,
Berggren
M
,
Grafstrom
R
. 
N-acetylcysteine protection against the toxicity of cigarette smoke and cigarette smoke condensates in various tissues and cells in vitro
.
Eur J Respir Dis Suppl
1985
;
139
:
123
9
.
24.
Kim
H
,
Liu
X
,
Kobayashi
T
,
Conner
H
,
Kohyama
T
,
Wen
FQ
, et al
Reversible cigarette smoke extract-induced DNA damage in human lung fibroblasts
.
Am J Respir Cell Mol Biol
2004
;
31
:
483
90
.
25.
Ishii
T
,
Matsuse
T
,
Igarashi
H
,
Masuda
M
,
Teramoto
S
,
Ouchi
Y
. 
Tobacco smoke reduces viability in human lung fibroblasts: protective effect of glutathione S-transferase P1
.
Am J Physiol Lung Cell Mol Physiol
2001
;
280
:
L1189
95
.
26.
Sarill
M
,
Zago
M
,
Sheridan
JA
,
Nair
P
,
Matthews
J
,
Gomez
A
, et al
The aryl hydrocarbon receptor suppresses cigarette-smoke-induced oxidative stress in association with dioxin response element (DRE)-independent regulation of sulfiredoxin 1
.
Free Radic Biol Med
2015
;
89
:
342
57
.
27.
Iu
M
,
Zago
M
,
Rico de Souza
A
,
Bouttier
M
,
Pareek
S
,
White
JH
, et al
RelB attenuates cigarette smoke extract-induced apoptosis in association with transcriptional regulation of the aryl hydrocarbon receptor
.
Free Radic Biol Med
2017
;
108
:
19
31
.
28.
Rico de Souza
A
,
Zago
M
,
Pollock
SJ
,
Sime
PJ
,
Phipps
RP
,
Baglole
CJ
. 
Genetic ablation of the aryl hydrocarbon receptor causes cigarette smoke-induced mitochondrial dysfunction and apoptosis
.
J Biol Chem
2011
;
286
:
43214
28
.
29.
Baglole
CJ
,
Maggirwar
SB
,
Gasiewicz
TA
,
Thatcher
TH
,
Phipps
RP
,
Sime
PJ
. 
The aryl hydrocarbon receptor attenuates tobacco smoke-induced cyclooxygenase-2 and prostaglandin production in lung fibroblasts through regulation of the NF-kappaB family member RelB
.
J Biol Chem
2008
;
283
:
28944
57
.
30.
Pal
A
,
Melling
G
,
Hinsley
EE
,
Kabir
TD
,
Colley
HE
,
Murdoch
C
, et al
Cigarette smoke condensate promotes pro-tumourigenic stromal-epithelial interactions by suppressing miR-145
.
J Oral Pathol Med
2013
;
42
:
309
14
.
31.
Nyunoya
T
,
Monick
MM
,
Klingelhutz
AL
,
Glaser
H
,
Cagley
JR
,
Brown
CO
, et al
Cigarette smoke induces cellular senescence via Werner's syndrome protein down-regulation
.
Am J Respir Crit Care Med
2009
;
179
:
279
87
.
32.
Monick
MM
,
Powers
LS
,
Walters
K
,
Lovan
N
,
Zhang
M
,
Gerke
A
, et al
Identification of an autophagy defect in smokers' alveolar macrophages
.
J Immunol
2010
;
185
:
5425
35
.
33.
Kim
SY
,
Lee
JH
,
Huh
JW
,
Ro
JY
,
Oh
YM
,
Lee
SD
, et al
Cigarette smoke induces Akt protein degradation by the ubiquitin-proteasome system
.
J Biol Chem
2011
;
286
:
31932
43
.
34.
Gross
TJ
,
Powers
LS
,
Boudreau
RL
,
Brink
B
,
Reisetter
A
,
Goel
K
, et al
A microRNA processing defect in smokers' macrophages is linked to SUMOylation of the endonuclease DICER
.
J Biol Chem
2014
;
289
:
12823
34
.
35.
Bisetto
S
,
Whitaker-Menezes
D
,
Wilski
NA
,
Tuluc
M
,
Curry
J
,
Zhan
T
, et al
Monocarboxylate transporter 4 (MCT4) knockout mice have attenuated 4NQO induced carcinogenesis; a role for MCT4 in driving oral squamous cell cancer
.
Front Oncol
2018
;
8
:
324
.
36.
Martinez-Outschoorn
UE
,
Curry
JM
,
Ko
YH
,
Lin
Z
,
Tuluc
M
,
Cognetti
D
, et al
Oncogenes and inflammation rewire host energy metabolism in the tumor microenvironment: RAS and NFkappaB target stromal MCT4
.
Cell Cycle
2013
;
12
:
2580
97
.
37.
Schmitz
S
,
Machiels
JP
. 
Targeting the tumor environment in squamous cell carcinoma of the head and neck
.
Curr Treat Options Oncol
2016
;
17
:
37
.
38.
Liao
Z
,
Tan
ZW
,
Zhu
P
,
Tan
NS
. 
Cancer-associated fibroblasts in tumor microenvironment - Accomplices in tumor malignancy
.
Cell Immunol
2018
Jan 31. [Epub ahead of print]
.
39.
Avagliano
A
,
Granato
G
,
Ruocco
MR
,
Romano
V
,
Belviso
I
,
Carfora
A
, et al
Metabolic reprogramming of cancer associated fibroblasts: the slavery of stromal fibroblasts
.
Biomed Res Int
2018
;
2018
:
6075403
.
40.
Curry
J
,
Tassone
P
,
Gill
K
,
Tuluc
M
,
BarAd
V
,
Mollaee
M
, et al
Tumor metabolism in the microenvironment of nodal metastasis in oral squamous cell carcinoma
.
Otolaryngol Head Neck Surg
2017
;
157
:
798
807
.
41.
New
J
,
Arnold
L
,
Ananth
M
,
Alvi
S
,
Thornton
M
,
Werner
L
, et al
Secretory autophagy in cancer-associated fibroblasts promotes head and neck cancer progression and offers a novel therapeutic target
.
Cancer Res
2017
;
77
:
6679
91
.
42.
Martinez-Outschoorn
UE
,
Balliet
RM
,
Rivadeneira
DB
,
Chiavarina
B
,
Pavlides
S
,
Wang
C
, et al
Oxidative stress in cancer associated fibroblasts drives tumor-stroma co-evolution: A new paradigm for understanding tumor metabolism, the field effect and genomic instability in cancer cells
.
Cell Cycle
2010
;
9
:
3256
76
.
43.
Bovenzi
CD
,
Hamilton
J
,
Tassone
P
,
Johnson
J
,
Cognetti
DM
,
Luginbuhl
A
, et al
Prognostic indications of elevated MCT4 and CD147 across cancer types: a meta-analysis
.
Biomed Res Int
2015
;
2015
:
242437
.
44.
Brockton
N
,
Dort
J
,
Lau
H
,
Hao
D
,
Brar
S
,
Klimowicz
A
, et al
High stromal carbonic anhydrase IX expression is associated with decreased survival in P16-negative head-and-neck tumors
.
Int J Radiat Oncol Biol Phys
2011
;
80
:
249
57
.
45.
Gouirand
V
,
Guillaumond
F
,
Vasseur
S
. 
Influence of the tumor microenvironment on cancer cells metabolic reprogramming
.
Front Oncol
2018
;
8
:
117
.
46.
Martinez-Outschoorn
UE
,
Pavlides
S
,
Howell
A
,
Pestell
RG
,
Tanowitz
HB
,
Sotgia
F
, et al
Stromal-epithelial metabolic coupling in cancer: integrating autophagy and metabolism in the tumor microenvironment
.
Int J Biochem Cell Biol
2011
;
43
:
1045
51
.
47.
Yu
Y
,
Tang
D
,
Kang
R
. 
Oxidative stress-mediated HMGB1 biology
.
Front Physiol
2015
;
6
:
93
.
48.
Nguyen
A
,
Bhavsar
S
,
Riley
E
,
Caponetti
G
,
Agrawal
D
. 
Clinical value of high mobility group box 1 and the receptor for advanced glycation end-products in head and neck cancer: a systematic review
.
Int Arch Otorhinolaryngol
2016
;
20
:
382
9
.
49.
Li
Y
,
Wang
P
,
Zhao
J
,
Li
H
,
Liu
D
,
Zhu
W
. 
HMGB1 attenuates TGF-beta-induced epithelial-mesenchymal transition of FaDu hypopharyngeal carcinoma cells through regulation of RAGE expression
.
Mol Cell Biochem
2017
;
431
:
1
10
.
50.
Di
X
,
He
G
,
Chen
H
,
Zhu
C
,
Qin
Q
,
Yan
J
, et al
High-mobility group box 1 protein modulated proliferation and radioresistance in esophageal squamous cell carcinoma
.
J Gastroenterol Hepatol
2019
;
34
:
728
35
.
51.
Mandal
R
,
Senbabaoglu
Y
,
Desrichard
A
,
Havel
JJ
,
Dalin
MG
,
Riaz
N
, et al
The head and neck cancer immune landscape and its immunotherapeutic implications
.
JCI Insight
2016
;
1
:
e89829
.
52.
Balermpas
P
,
Rodel
F
,
Liberz
R
,
Oppermann
J
,
Wagenblast
J
,
Ghanaati
S
, et al
Head and neck cancer relapse after chemoradiotherapy correlates with CD163+ macrophages in primary tumour and CD11b+ myeloid cells in recurrences
.
Br J Cancer
2014
;
111
:
1509
18
.
53.
Seminerio
I
,
Kindt
N
,
Descamps
G
,
Bellier
J
,
Lechien
JR
,
Mat
Q
, et al
High infiltration of CD68+ macrophages is associated with poor prognoses of head and neck squamous cell carcinoma patients and is influenced by human papillomavirus
.
Oncotarget
2018
;
9
:
11046
59
.
54.
Marcus
B
,
Arenberg
D
,
Lee
J
,
Kleer
C
,
Chepeha
DB
,
Schmalbach
CE
, et al
Prognostic factors in oral cavity and oropharyngeal squamous cell carcinoma
.
Cancer
2004
;
101
:
2779
87
.
55.
Oliveira da Silva
C
,
Monte-Alto-Costa
A
,
Renovato-Martins
M
,
Viana Nascimento
FJ
,
Dos Santos Valença
S
,
Lagente
V
, et al
Time course of the phenotype of blood and bone marrow monocytes and macrophages in the lung after cigarette smoke exposure in vivo
.
Int J Mol Sci
2017
;
18
:
E1940
.
56.
Kameyama
N
,
Chubachi
S
,
Hegab
AE
,
Yasuda
H
,
Kagawa
S
,
Tsutsumi
A
, et al
Intermittent exposure to cigarette smoke increases lung tumors and the severity of emphysema more than continuous exposure
.
Am J Respir Cell Mol Biol
2018
;
59
:
179
88
.
57.
Tsujikawa
T
,
Yaguchi
T
,
Ohmura
G
,
Ohta
S
,
Kobayashi
A
,
Kawamura
N
, et al
Autocrine and paracrine loops between cancer cells and macrophages promote lymph node metastasis via CCR4/CCL22 in head and neck squamous cell carcinoma
.
Int J Cancer
2013
;
132
:
2755
66
.
58.
Gao
L
,
Wang
FQ
,
Li
HM
,
Yang
JG
,
Ren
JG
,
He
KF
, et al
CCL2/EGF positive feedback loop between cancer cells and macrophages promotes cell migration and invasion in head and neck squamous cell carcinoma
.
Oncotarget
2016
;
7
:
87037
51
.
59.
Yu
Y
,
Ke
L
,
Lv
X
,
Ling
YH
,
Lu
J
,
Liang
H
, et al
The prognostic significance of carcinoma-associated fibroblasts and tumor-associated macrophages in nasopharyngeal carcinoma
.
Cancer Manag Res
2018
;
10
:
1935
46
.
60.
Fujii
N
,
Shomori
K
,
Shiomi
T
,
Nakabayashi
M
,
Takeda
C
,
Ryoke
K
, et al
Cancer-associated fibroblasts and CD163-positive macrophages in oral squamous cell carcinoma: their clinicopathological and prognostic significance
.
J Oral Pathol Med
2012
;
41
:
444
51
.
61.
Ohashi
T
,
Aoki
M
,
Tomita
H
,
Akazawa
T
,
Sato
K
,
Kuze
B
, et al
M2-like macrophage polarization in high lactic acid-producing head and neck cancer
.
Cancer Sci
2017
;
108
:
1128
34
.