Nucleotides released from palmitate-challenged muscle cells through pannexin-3 attract monocytes

Nicolas J. Pillon,1 Yujin E. Li,1 Lisbeth N. Fink,2 Joseph T. Brozinick,3 Alexander Nikolayev,3

Ming-Shang Kuo,3 Philip J. Bilan,1 and Amira Klip1

Nucleotides Released FromPalmitate-Challenged MuscleCells Through Pannexin-3Attract MonocytesDiabetes 2014;63:3815–3826 | DOI: 10.2337/db14-0150

Obesity-associated low-grade inflammation in metabol-ically relevant tissues contributes to insulin resistance.We recently reported monocyte/macrophage infiltrationin mouse and human skeletal muscles. However, themolecular triggers of this infiltration are unknown, andthe role of muscle cells in this context is poorly un-derstood. Animal studies are not amenable to thespecific investigation of this vectorial cellular commu-nication. Using cell cultures, we investigated the cross-talk between myotubes and monocytes exposed tophysiological levels of saturated and unsaturated fattyacids. Media from L6 myotubes treated with palmitate—but not palmitoleate—induced THP1 monocyte migra-tion across transwells. Palmitate activated the Toll-likereceptor 4 (TLR4)/nuclear factor-kB (NF-kB) pathway inmyotubes and elevated cytokine expression, but themonocyte chemoattracting agent was not a polypeptide.Instead, nucleotide degradation eliminated the chemo-attracting properties of the myotube-conditioned media.Moreover, palmitate-induced expression and activity ofpannexin-3 channels in myotubes were mediated byTLR4-NF-kB, and TLR4-NF-kB inhibition or pannexin-3knockdown prevented monocyte chemoattraction. Inmice, the expression of pannexin channels increasedin adipose tissue and skeletal muscle in response tohigh-fat feeding. These findings identify pannexins asnew targets of saturated fatty acid–induced inflamma-tion in myotubes, and point to nucleotides as possible

mediators of immune cell chemoattraction toward mus-cle in the context of obesity.

Nutrient excess is a major factor contributing to thealarming incidence of obesity worldwide (1). Obesity leadsto whole-body insulin resistance, a leading cause of type 2diabetes and cardiovascular complications (1). A paradigmshift in our understanding of the inflammation accompa-nying obesity was the discovery that a high-fat diet (HFD)increases the number of immune cells in adipose tissue,and a growing body of evidence now suggests that obesityand type 2 diabetes are inflammatory diseases (2). Surpris-ingly, equivalent studies in skeletal muscle had been rela-tively scant despite this tissue being responsible for themajority of postprandial glucose use (3).

We and others recently demonstrated an increase inmacrophage number and inflammatory phenotype withinskeletal muscle from HFD-fed mice and obese subjects(4,5), but, strikingly, the factors responsible for immunecell infiltration in obese muscle (or other metabolically rel-evant tissues) are largely unknown. Skeletal muscle secretesseveral cytokines, recently renamed “myokines” (6), but thespecific soluble mediators, channels, and receptors involvedin the crosstalk between skeletal muscle and immunecells are virtually undefined. Interestingly, in additionto myokines, selective stimuli induce the release of smallmolecules from muscle, such as prostanoids, lactate, and

1Program in Cell Biology, The Hospital for Sick Children, Toronto, Ontario, Canada2Diabetes Research Unit, Novo Nordisk A/S, Maaloev, Denmark3Eli Lilly and Company, Indianapolis, IN

Corresponding author: Amira Klip, [email protected].

Received 28 January 2014 and accepted 5 June 2014.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1.

© 2014 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

Diabetes Volume 63, November 2014 3815

IMMUNOLOGYAND

TRANSPLANTATIO

N

nucleotides (7–12). The mechanism of release of thesesmall molecules is unclear, but likely occurs throughchannels expressed at the plasma membrane. Pannexinsare recently discovered channel-forming proteins thatallow the release of cytoplasmic molecules to the ex-tracellular space (13,14). Whereas pannexin-1 and itsrole in the physiological release of small molecules havebeen widely studied (15–17), the functions of pannexin-2and pannexin-3 remain elusive. Moreover, the contributionof pannexin channels and small molecule release duringmetabolic inflammation remains unexplored.

Although the initial trigger of inflammation in vivo isdebated, excess lipids contribute to the induction ofinsulin resistance and proinflammatory genes in meta-bolic tissues (18). Compellingly, palmitic acid (also calledhexadecanoic acid, 16:0), a major dietary saturated fattyacid in blood, promotes a proinflammatory phenotype inseveral cell types in vitro (19–22). On the other hand,unsaturated fatty acids such as the monounsaturatedfatty acid palmitoleic acid ([Z]-9-hexadecenoic acid,16:1D9) are either innocuous or able to suppress inflam-mation (22,23).

Given the diversity of cells coexisting within metabolictissues (parenchymal, endothelial, and myeloid), it isdifficult to dissect the particular role of myocytes in theimmune cell infiltration of muscle tissue during HFDfeeding. Only cell culture paradigms allow for the estab-lishment of vectorial communication between distinct celltypes and their response to defined hyperlipidic environ-ments. Here we show that when myotubes are challengedwith palmitate—but not with palmitoleate—they releasenonpeptidic factors that attract monocytes. We provideevidence that nucleotides released through pannexin-3are major factors in this palmitate-induced crosstalk be-tween muscle and immune cells.

RESEARCH DESIGN AND METHODS

ReagentsMyeloid differentiation factor-88 (MYD88) inhibitorypeptides were from InvivoGen (San Diego, CA). Smallinterfering RNA (siRNA) oligonucleotides for connexin(Cx)-43, Cx45, and Toll-like receptor (TLR) 4 were fromGenePharma (Shanghai, People’s Republic of China), andfor pannexin-3 from Qiagen (Chatsworth, CA). Otherchemicals were from Sigma-Aldrich (St. Louis, MO).

Palmitate PreparationPalmitate or palmitoleate (P9767 and P9417; Sigma-Aldrich) stock solutions (200 mmol/L) were prepared in50% ethanol by heating at 50°C. Fatty acid–free, low-endotoxin BSA (A8806; Sigma-Aldrich) was dissolved inserum-free a-minimum essential medium (aMEM) to10.5%. Fatty acid stocks were diluted 253 in the BSAsolution and conjugated under agitation at 40°C for 2 h.This solution (lipid/BSA ratio 5:1) was further diluted inculture media. Palmitate and palmitoleate solutions thuscoupled to BSA are denoted as PA and PO, respectively.

Cell Culture, Viability, and TransfectionL6 muscle cells were grown and differentiated as de-scribed previously (24). THP1 monocytes were grown inRPMI 1640 medium containing 5% FBS. Cellular viabilitywas assessed from lactate dehydrogenase (LDH) activityand MTT reduction (Cytotoxicity Detection Kit and CellProliferation Kit I; Roche Applied Science, India-napolis, IN). Oligonucleotide siRNAs were transfectedwith JetPRIME (Polyplus Transfection, Illkirch, France).Myotubes were treated with 200 nmol/L siRNA for 24 hthen stabilized in fresh media for 24 h before performingexperiments.

Generation of Muscle-Conditioned MediumL6 myotubes were treated in aMEM (2% FBS) for dosesand times indicated. Supernatants were centrifuged at10,000 rotations per minute for 5 min to pellet debris,aliquoted, and frozen immediately at 280°C. Superna-tants (conditioned media [CM]) from PA-, PO-, andBSA-treated myotubes are herein denoted as CM-PA,CM-PO, and CM-BSA, respectively.

Fatty Acid UptakeNonesterified fatty acids were quantified in the media offatty acid–incubated myotubes using the NEFA-HR(2) R2Set (Wako Chemicals USA, Richmond, VA). Uptake wasindirectly estimated from the fatty acid content in themyotube supernatant at the beginning and the end ofthe incubation.

Monocyte Chemoattraction AssayIn Boyden chambers (Transwell, 6.5 mm diameter, 5 mmpore diameter; Corning, Lowell, MA), 600 mL of attractantwas added to the lower chamber. THP1 monocytes(100 mL of 5 3 106cells/mL) in aMEM supplementedwith 2% FBS were placed in the upper chamber. After3 h at 37°C, cells were dislodged from the filter by gentleshaking, the upper chamber was discarded, and mono-cytes that transmigrated to the bottom well were countedusing a Z2 Coulter Counter (Beckman Coulter Canada,Mississauga, ON, Canada).

Nucleotide MeasurementsATP was specifically measured using the luciferase-basedENLITEN ATP Assay (Promega, Madison, WI). Othernucleotides (monophosphate, diphosphate, and triphos-phate) and nucleosides were measured by hydrophilicinteraction liquid chromatography coupled to mass spec-trometry, as previously reported (25). The method wasmodified by limiting the monitored metabolites to nucleo-sides and nucleotides. Supernatants were analyzed bya Nexera UPLC (Shimadzu Corporation, Kyoto, Japan) cou-pled to AB/SCIEX Triple Quad 5500 mass spectrometer(AB SCIEX, Framingham, MA) using an electrospray ioniza-tion technique operating in multiple reactions monitoringmode. Parent-to-product transitions used for each de-tected metabolite are presented in Supplementary Table 1.Calibration curves (12.5–500 ng/mL) were generated foreach detected metabolite. Uridine triphosphate (UTP),

3816 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

ATP, thiamine diphosphate, thiamine triphosphate (TTP),cytidine diphosphate, and guanosine triphosphate (GTP)were undetectable (limit of detection 1 ng/mL). The standardof uridine diphosphate (UDP) was not available at the time ofanalysis, and consequently results were reported in relativeunits by comparing the UDP peak areas.

RNA Isolation and Quantitative PCRAll reagents were from Life Technologies (Carlsbad, CA).RNA was isolated using Trizol, and cDNA was synthesizedusing the SuperScript VILO cDNA kit. Ten nanograms perreaction were used for quantitative RT-PCR using prede-signed Taqman probes for target genes and hprt1 or eef2(housekeeping references).

ImmunoblottingCells were scraped in lysis buffer, and protein content wasmeasured using the bicinchoninic acid–based assay. Sam-ples were boiled in Laemmli buffer, separated by SDS-PAGE, and transferred onto nitrocellulose. Membraneswere blotted using primary and peroxidase-coupled second-ary antibodies, then developed using chemiluminescence(ECL Kit; Bio-Rad, Hercules, CA), and analyzed using

ImageJ software (National Institutes of Health, Bethesda,MD).

YO-PRO UptakeYO-PRO (1 mmol/L) and Texas-red dextran (10 kDa, 0.1mg/mL) from Life Technologies were added to cells for 15min. Cells were then washed with PBS and fixed (3%paraformaldehyde for 10 min). Images were acquiredwith a Leica DMIRE2 fluorescent microscope with a103 air objective. Total green and red fluorescence weremeasured on 15 random fields per condition using ImageJsoftware.

Multiplex Cytokine AnalysisCytokines were determined with a rat Milliplex MAPMagnetic Bead Panel (Millipore, Hellerup, Denmark) ona Bio-Plex-200 System (Bio-Rad Laboratories, Copenha-gen, Denmark).

Animal StudiesThe study was approved by The Hospital for Sick ChildrenAnimal Care Committee. Male C57BL/6J mice (TheJackson Laboratory, Bar Harbor, ME), singly caged andmaintained at 21–22°C with light from 0600 to 1800 h,

Figure 1—CM from palmitate-treated muscle cells attract monocytes. A: Migration of THP1 monocytes toward CM-PA, CM-PO, or CM-BSA collected from L6 myotubes treated for 18 h with PA, PO, or BSA, respectively. B: Transmigration of THP1 monocytes toward CM fromL6 myotubes treated with 0.5 mmol/L fatty acids for 0, 3, 6, and 18 h. C: CCL2/MCP1 (100 ng/mL) was used as a positive control, and theeffect of PA and PO on migration was controlled by placing 0.5 mmol/L fatty acids in the bottom chamber and testing THP1 monocytetransmigration. D: Fatty acid uptake by L6 myotubes treated for 18 h was measured as described in RESEARCH DESIGN ANDMETHODS. All resultsare reported as the mean 6 SEM, n $ 4. *P < 0.05, **P < 0.01 vs. BSA control. AMEM, aMEM; conc., concentration; RM, regular mediumused to culture L6 myotubes.

diabetes.diabetesjournals.org Pillon and Associates 3817

were fed a standard chow diet (5P07 Prolab RMH 1000;LabDiet, St. Louis, MO) or an HFD (60% by kcal) (D12492;Research Diets, New Brunswick, NJ) for 18 weeks. Aftera 4-h fast, mice were killed via cervical dislocation, andtissue was collected, flash frozen in liquid nitrogen, andpreserved at 280°C.

Statistical AnalysisAnalyses were performed using Prism software (GraphPadSoftware, San Diego, CA). Results from dose responseswere compared by two-way ANOVA (dose, treatment)followed by Tukey post hoc tests. One-way ANOVA wasused to test differences between groups with equalvariances. Statistical significance was set at P , 0.05.

RESULTS

Palmitate-Treated Myotubes Attract MonocytesPalmitate is the most abundant saturated fatty acid inWestern diets, and it is widely used to challenge cells inculture. L6 myotubes were treated with BSA-conjugatedPA or with the equivalent 16-carbon chain–length mono-unsaturated PO. Thereafter, supernatants (i.e., CM) werecollected, and their chemoattracting activity toward THP1monocytes was determined (Fig. 1A and B). CM-PA fromPA-treated myotubes had a marked chemoattractingeffect that was not reproduced by CM-PO or CM-BSAfrom PO- or BSA-treated myotubes, respectively. Maximalchemoattraction was observed after treatment with 0.5mmol/L PA for 18 h. As the CM still contain fatty acids,monocyte migration was also measured toward regularmedium containing 0.5 mmol/L fatty acids (Fig. 1C). Un-like the chemoattracting chemokine (C-C motif) ligand 2(CCL2)/MCP1 (100 ng/mL), neither PA nor PO on its ownsignificantly affected monocyte migration. The uptake ofPA and PO by myotubes was similar (Fig. 1D), and anequivalent reduction in fatty acid and glucose content inthe CM was observed in all conditions (SupplementaryTables 1 and 3). These results indicate that monocytechemoattraction by CM-PA was not mediated by the fattyacid itself or by differences in myotube uptake of eitherfatty acids or glucose.

Palmitate-Dependent Myotube-InducedChemoattraction of Monocytes Requires the TLR-MYD88/Nuclear Factor-kB PathwayTLRs couple with adaptor MYD88 to activate the nuclearfactor-kB (NF-kB) transcription factor, a critical step inpalmitate-induced inflammatory response (18,21). In myo-tubes, PA evoked a dose-dependent decrease in inhibitor ofkB (IkB), the canonical NF-kB repressor (Fig. 2A). Prevent-ing IkB degradation in myotubes with parthenolide ora cell-permeant MYD88 inhibitory peptide (Fig. 2B andC) blunted the increase in monocyte migration inducedby CM-PA. The involvement of this pathway was furtherconfirmed by gene silencing. With an achieved reduction inthe expression of TLR4 and MYD88 of 65% and 80%, re-spectively, the CM-PA–evoked monocyte chemoattractionwas abolished (Fig. 2D). Together, these experiments

Figure 2—Inhibition of TLR-NF-kB signaling in muscle preventsATP release and monocyte migration. A: Activation of the NF-kBpathway was measured by the degradation of its repressor IkB in L6myotubes treated for 18 h with PA, PO, or BSA vehicle. Immuno-blotting was performed using specific antibodies to IkB. Blots werequantified, and densitometry results were expressed relative toactinin-1 as a loading control. Results are the average of sevenindependent experiments, and a representative gel is illustrated.B and C: L6 myotubes were pretreated with the NF-kB inhibitorparthenolide (PTN, 25 mmol/L for 1 h) or with an MYD88 inhibitorypeptide (Pep-MYD, 50 mmol/L for 3 h), and then treated with PA,PO, or BSA for 18 h in the presence of the same inhibitor. D: Myo-tubes were transfected with siRNA to target TLR4 or MYD88 or anon-targeting, non-related (NR) sequence, as described in RESEARCH

DESIGN AND METHODS. PA (0.5 mmol/L) or BSA were then added to themedia for 18 h. CM were collected, and THP1 monocyte migrationwas measured as described in RESEARCH DESIGN AND METHODS. Insert:Efficiency of gene silencing measured by qPCR. Results arereported as the mean 6 SEM, n $ 4. *P < 0.05, **P < 0.01 vs.BSA control. ns, not significant; Pep, peptide; RM; regular mediumused to culture L6 myotubes; si, siRNA.

3818 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

demonstrate that PA engages the TLR/MYD88/NF-kBpathway in myotubes to attract monocytes.

The Chemoattractant Released by Myotubes Is Nota Polypeptidic CytokineThe expression of chemokines and cytokines, which areclassic monocytes attractants, was analyzed using quan-titative PCR (qPCR) arrays (Fig. 3A). In myotubes exposedto PA, but not to PO, there was a significant rise in thegene expression of several chemokines (Cxcl2, Ccl2, andCxcl1) and cytokines (Il1a, Tnfa, and Il6), but, surpris-ingly, that did not translate into a corresponding elevatedsecretion for them into the medium, as measured byLuminex Multiplex Immunoassay (Fig. 3B). This was con-firmed using a membrane cytokine array and ELISAs for

tumor necrosis factor-a and CCL2 (Supplementary Fig. 1).To explore whether the chemoattracting factor would beaffected by conditions altering protein stability, CM-PAand CM-BSA were either heated or treated with protein-ase K (Fig. 3C). These manipulations did not prevent themonocyte migration induced by CM-PA, suggesting thatthe chemoattractant is unlikely to be a polypeptide. Fur-thermore, upon filtering the CM through a Vivaspin col-umn, only the fraction containing molecules of ,3,000 Dadisplayed monocyte-chemoattracting activity (Fig. 3C).

As TLR4 signaling in myotubes was involved in themonocyte-chemoattracting ability of CM-PA (Fig. 2), weexplored whether the activation of TLR4 signaling withlipopolysaccharide (LPS) could elicit similar effects. CMfrom LPS-treated myotubes (CM-LPS) also induced

Figure 3—The attractant released by muscle cells is not a chemokine. L6 myotubes were treated with 0.5 mmol/L PA or PO for 18 h. A:Cytokine and chemokine expression was analyzed using qPCR arrays. B: Cytokine and chemokine content in myotube CM were analyzedby Luminex multiplex immunoassay. C: CM from myotubes were heat inactivated (95°C, 20 min) or treated with proteinase K (PROK; 100mg/mL for 1.5 h at 37°C) to denature proteins. Untreated CM was also filtered through a 3-kDa cutoff membrane, and the included andexcluded fractions were tested separately for monocyte chemoattraction activity. D: Myotubes were treated for 18 h with 100 ng/mL LPS.CM-LPS was then collected and tested in a migration assay as described above. E and F: CM-PA and CM-LPS were treated with blockingantibodies against CCL2/MCP1 or IgG control. Results are reported as the mean 6 SEM, n $ 4. *P < 0.05, **P < 0.01 vs. BSA control.AMEM, aMEM; MW, molecular weight; N.D, not detectable; ns, not significant; RM, regular medium used to culture L6 myotubes.

diabetes.diabetesjournals.org Pillon and Associates 3819

monocyte migration, but, interestingly, and unlike CM-PA, the monocyte-chemoattracting activity of CM-LPSwas heat sensitive and proteinase K sensitive, and wasretained in the fraction containing molecules .3,000 Da(Fig. 3D). Blocking antibodies neutralizing the chemokineCCL2/MCP1 did not affect CM-PA–induced monocyte mi-gration (Fig. 3E) but completely prevented monocyte mi-gration toward CM-LPS (Fig. 3F). Together, these resultssuggest that in response to palmitate, the myotube-derived factors responsible for monocyte migration isa low-molecular-weight compound. In contrast, CCL2/MCP1is the monocyte chemoattractant released by myotubes inresponse to LPS.

Nucleotides Are Released by Myotubes and Are PotentMonocyte ChemoattractantsSmall molecules such as eicosanoids and nucleotides arebona fide regulators of the immune response, and canpotentially affect monocyte migration (26,27). We conse-quently tested the chemoattractant effect of a variety of

such small molecules. Eicosanoids (prostaglandins E2 andF2a, and arachidonic acid) and histamine were monocyterepellents, urate and lactate had no effect, and only CCL2/MCP1, formyl peptides (WKYMVdM), and ATP were ableto attract monocytes (Fig. 4).

Nucleotides were the only small molecules that inducedhigher monocyte transmigration than CM-PA, with thepotency order UTP . ATP . ADP . UDP . TTP (Fig. 5Aand B). The attracting ability of CM-PA was eliminatedwhen nucleotides were hydrolyzed with the nucleotidaseapyrase (Fig. 5C). Notably, using an ATP-specific luciferase–based assay, ATP levels were detected and found to behigher in CM-PA compared with CM-PO or CM-BSA,and, consistent with the blunted monocyte attraction,apyrase treatment eliminated any measurable amountsof ATP in the CM (Fig. 5D).

Most cells express surface extracellular ecto-nucleotidasesthat cleave nucleotides from the interstitial space (28).Accordingly, we explored whether endogenous ecto-nucleotidases would tonically reduce the amount of che-moattractant released by the myotubes. The inhibitionof myotube ecto-nucleotidases with ARL67156 aug-mented the levels of ATP (Fig. 5E) and concomitantlypotentiated the monocyte-chemoattracting activity ofCM-PA (Fig. 5F), without affecting either parameter inCM-BSA or CM-PO. Compellingly, the ATP concentra-tion in CM-PA correlated strongly with monocyte che-moattraction (r = 0.747, P , 0.0001) across treatmentsand conditions (Fig. 5G), buttressing the propositionthat nucleotides are responsible for the enhanced mono-cyte transmigration toward CM-PA.

Blocking the P2 family of receptors on monocytesusing the broad-spectrum nonselective antagonistssuramin and pyridoxal-phosphate-6-azophenyl-29,49-disulfonate (PPADS) also prevented CM-PA–inducedmonocyte chemoattraction (Fig. 5H). However, since P2receptors can be activated by several nucleotides (29) andapyrase can cleave all nucleotide triphosphates and diphos-phates (Fig. 5I), the results suggested that, in addition toATP, other nucleotides may also be involved. Indeed, usinghydrophilic interaction liquid chromatography coupled tomass spectrometry, we found elevated concentrations ofADP, UDP, and several monophosphate nucleotides in CM-PA compared with CM-BSA and CM-PO (Fig. 5J and K).

Since cytotoxic effects of fatty acids have been de-scribed in several cell types, and cells can release ATP aftermembrane damage, we assayed myotube viability. MTTreduction, LDH release, and caspase-3 cleavage werenot significantly affected in myotubes treated with PAcompared with BSA or PO (Supplementary Table 1 andSupplementary Fig. 2). Moreover, after mechanically dam-aging myotubes by scraping and vortexing to induce ne-crosis, LDH release increased 15-fold, indicating that.95% of the cells were undergoing necrosis. However,contrary to palmitate treatment, the supernatant fromthese scraped cells did not attract monocytes. In addition,no correlation was found between LDH release and

Figure 4—Compounds affecting monocyte migration across trans-wells. Chemokines (CCL2/MCP1 and CXCL1), formylated peptides(WKYMVdM and formyl-methionyl-leucyl-phenylalanine [fMLP]),ceramides (C2, C8, and C8–1-phosphosphate), eicosanoids (arach-idonic acid [ARA], prostaglandin E2 [PGE2], and prostaglandin F2a[PGF2a]), lactic and uric acids, histamine, and ATP were tested indose response for their ability to attract THP1 monocytes. Resultsare the average of at least four independent experiments, and datawere fit to a nonlinear sigmoidal dose-response curve.

3820 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

Figure 5—Palmitate induces nucleotide release by muscle cells to attract monocytes. A and B: Migration of THP1 monocytes towardgraded doses of purine and pyrimidine nucleotides. C and D: CM-PA and CM-PO collected from muscle cells were treated with apyrase(0.2 international units [IU]/mL final) for 1 h. E and F: Myotubes were treated with 0.5 mmol/L BSA/PA in the presence of the ecto-nucleotidase inhibitor ARL67156 (100 mmol/L) for 18 h. The ATP content and chemoattraction activity of the corresponding CM werethen measured. G: Correlation between ATP concentration in CM and monocyte migration. H: CM from myotubes were supplemented withthe P2 receptor antagonists suramin (100 mmol/L) and PPADS (200 mmol/L) before testing monocyte migration. I: Apyrase (0.2 IU/mL finalfor 1 h) was added to solutions containing nucleotide triphosphates at the concentration that caused half-maximal stimulation of monocytechemoattraction for each nucleotide before testing monocyte migration. J and K: Nucleotides were measured using liquid chromatographycoupled to mass spectrometry. Monocyte migration and ATP concentration were measured as described in RESEARCH DESIGN AND METHODS.Results are reported as the mean 6 SEM. n $ 4. *P < 0.05, **P < 0.01, ***P < 0.001 vs. BSA control. AMEM, aMEM; CMP, cytidinemonophosphate; CTP, cytidine triphosphate; GMP, guanosine monophosphate; IMP, inosine monophosphate; ns, not significant; RM,regular medium used to culture L6 myotubes; UMP, uridine monophosphate.

diabetes.diabetesjournals.org Pillon and Associates 3821

monocyte chemoattraction (Supplementary Fig. 2). Thus,it is unlikely that the monocyte-chemoattracting activityof CM-PA was due to sporadic myotube cell death.

Nucleotides Are Released From Myotubes ThroughPannexin-3 ChannelsThe route of ATP release from myotubes was nextexamined. Pannexins are channels that allow the releaseof cytoplasmic molecules into the extracellular space (14).In addition, gap junction molecules (i.e., Cxs) can formhemichannels, also allowing the release of small molecules(13). In L6 myotubes, Cx43 and Cx45 were the mostabundant Cxs (Fig. 6A), while Cx40 and pannexin-2 wereundetectable. Notably, only the expression of pannexin-3significantly rose with PA treatment (Fig. 6B), and thiseffect was blocked by inhibiting NF-kB with parthenolide(Fig. 6C).

To mediate ATP release, channels must be open at theplasma membrane. The small dye YO-PRO-1 has a molec-ular weight (630 g/mol) close to that of ATP (507 g/mol),readily diffuses through open Cx/pannexin channels, andfluoresces upon binding to nucleic acids; hence, it hasbeen used to ascertain the presence of open channels atthe cell surface (15). YO-PRO was administered to myo-tubes treated with PA, PO, or BSA, along with a largeTexas-Red-dextran (10 kDa) polysaccharide that cannot

go through Cx/pannexin channels but enters cells withdamaged membranes (Supplementary Fig. 3). In myo-tubes treated with PA, the YO-PRO/Texas-Red-dextranfluorescence ratio was significantly higher than that inPO-treated or BSA-treated myotubes (Fig. 6D), indicatingthat PA increases the number of open channels at themyotube plasma membrane.

Next, we reduced the expression of channels usingcognate siRNA sequences. Only the knockdown ofpannexin-3 eliminated the CM-PA–induced monocytetransmigration (Fig. 6E). Consistently, the ATP content sig-nificantly diminished in CM-PA derived from pannexin-3–depleted myotubes (Fig. 6F). These results demonstratethat pannexin-3 is required for nucleotide release by PA-treated myotubes into the media and for the consequentmonocyte chemoattraction.

Pannexin Channel Expression Rises in HFD-Fed Miceand Palmitate-Treated Primary Human MyotubesFinally, in a pilot experiment we measured the expressionof pannexins in epididymal white adipose tissue (eWAT)and quadriceps muscle from a small cohort of mice fed anHFD for 18 weeks. Pannexin-1 was highly expressed inboth tissues and increased in response to HFD feeding ineWAT but not in quadriceps muscle (Fig. 7A and B). Theexpression of pannexin-2 and pannexin-3 was lower, but

Figure 6—Nucleotides are released through pannexin-3 channels. L6 myotubes were challenged with 0.5 mmol/L PA, PO, or BSA for 18 h.A and B: The expression of several pannexins and Cxs was measured by qPCR. C: Myotubes were treated with NF-kB inhibitor parthe-nolide (PTN, 25 mmol/L) during the fatty acid challenges, and then pannexin-3 expression was determined. D: Uptake of YO-PRO was usedas an index of channel opening as described in RESEARCH DESIGN AND METHODS. E and F: Silencing Cx43, Cx45, and Pan3 in myotubes wasperformed using siRNA oligonucleotides (200 nmol/L), CM were collected, and their THP1 monocyte-chemoattracting activity or ATP concen-tration was determined. Insert: Efficiency of transfection measured by qPCR. Results are reported as the mean6 SEM. n $ 4. *P < 0.05, **P <0.01. ND, not detectable; ns, not significant; Panx and Pan, pannexin; RM, regular medium used to culture L6 myotubes; si, siRNA.

3822 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

HFD feeding induced a significantly higher expression ineWAT. In particular, pannexin-3 expression in this tissuewas undetectable in chow-fed mice and was noticeablyinduced by HFD feeding (Fig. 7B). In quadriceps muscle,the trend was very limited, and more mice will be neededfor a proper statistical analysis of the results. Of note, inhuman primary myotubes, pannexin-2 expression rosesignificantly in response to palmitate (Fig. 7C). Theseresults demonstrate that, although pannexin isoformsmay differ, palmitate-induced expression of pannexin

channels is common across species and relevant duringHFD feeding in mice.

DISCUSSION

Recent studies document a rise in macrophages withinskeletal muscle of HFD-fed mice and obese individuals(4,5,30–33), but a key unresolved issue is whether andhow skeletal muscle is capable of attracting monocytes.Here, we present evidence that CM from myotubes ex-posed to the saturated fatty acid PA, but not the unsat-urated PO, attract monocytes in culture. Myotubeschallenged in this manner activate endogenous inflamma-tory programs leading to the expression of pannexin-3channels, and we identify nucleotides as new potentialfactors in muscle-to-monocyte crosstalk during metabolicinflammation.

In vivo, downregulation of CCL2/MCP1 (34) or its re-ceptor CCR2 (35) only partially prevented the gain in in-flammatory macrophages in adipose tissue and skeletalmuscle of HFD-fed mice (5), suggesting that additionalfactors contribute to macrophage infiltration of tissues.In addition, two important studies (36,37) recently as-cribed the gain in adipose tissue macrophages to in situproliferation, which was dependent on CCL2/MCP1.Hence, the chemoattracting function of CCL2/MCP1 inobesity is debatable. Even when exerting a chemoattract-ing effect, CCL2/MCP1 might have been produced in re-sponse to endotoxin-like stimuli. Finally and importantly,it is not possible to ascertain from the above in vivostudies whether CCL2/MCP1 arose from myocytes/adipo-cytes, or whether instead it was contributed by endothe-lial cells or myeloid cells inside the tissues. It is preciselyto approach these questions that we here explored theability of a reconstituted cellular system to deconstructthe potential crosstalk between muscle and immune cellsin the presence of fatty acids. We show that L6 myotubesselectively challenged with PA can attract monocytes, andthat this does not rely on any chemokine. Instead, thechemoattracting factors are nucleotides released throughpannexin-3 channels. By contrast, LPS-treated myotubesprovoked a CCL2/MCP1-dependent monocyte chemoat-traction. This suggests that both nucleotides and CCL2/MCP1 might contribute to the immune cell infiltration ofskeletal muscle in vivo that occurs during HFD feeding.Finally, primary human myotubes also showed increasedpannexin expression in response to palmitate, as did tis-sues from high fat-fed mice.

The Inflammatory Response of Myotubes to SaturatedFat Leads to Expression of Pannexin-3Dietary fats, in particular saturated fats, confer a state oflow-grade inflammation to skeletal muscle. Either TLRactivation or intracellular lipid intermediates can triggeran inflammatory response through activation of stresskinases (e.g., Jun NH2-terminal kinase, extracellular signal–related kinase), generation of reactive oxygen species, andstimulation of NF-kB signaling, classically enhancing the

Figure 7—Pannexin (Panx) channel expression rises with HFDfeeding in mice and upon palmitate exposure in human myotubes.A and B: Male C57BL/6J mice were fed an HFD for 18 weeks. Theexpression of pannexins in quadriceps muscle and eWAT was mea-sured by quantitative RT-PCR. Results are reported as the mean,n $ 4. Dotted line represents the threshold of detection. C: Primaryhuman myotubes were treated with 0.5 mmol/L palmitate, palmito-leate, or the BSA control for 24 h. Expression of pannexins wasmeasured using quantitative RT-PCR. Results are reported asthe mean 6 SEM, n $ 4. *P < 0.05, **P < 0.01. eEF2, eukaryotictranslation elongation factor 2; N.D., not detectable; ns, not signif-icant vs. control. Since chow samples were undetectable for Panx3,the Wilcoxon test was used, setting the hypothetical value at thedetection threshold.

diabetes.diabetesjournals.org Pillon and Associates 3823

expression of proinflammatory cytokines (38,39). How-ever, there is no previous evidence of inflammatory cuesregulating the expression of channels for small molecules.We report here that activation of the TLR4/MYD88/NF-kB pathway in myotubes challenged with palmitate sig-nificantly elevates the expression of pannexin-3. Thisresponse was not observed when exposing myotubes tothe unsaturated fatty acid palmitoleate, and thereforerepresents a nutrient-specific response. In particular, weobserved a significant rise in several nucleotides in theCM from palmitate-challenged myotubes, and silencingthe expression of pannexin-3 abolished this gain.

Nucleotide Release From Myotubes: A “Find-Me”Signal for MonocytesThe importance of extracellular nucleotides in cell-to-cellcommunication is evident in the immune system, wherereleased ATP acts as a “find-me” signal for immune cells topromote phagocytic clearance of damaged or apoptoticcells (27). Specific release of ATP also occurs in skeletalmuscle during physiological exercise as well as duringpathological situations such as sepsis or myopathies (forreview, see Pillon et al. [40]). Attempts to measure in-terstitial nucleotides using microdialysis have been scant,but have estimated ATP concentrations at ;1 mmol/L inskeletal muscle from anesthetized cats (41) and ; 0.1mmol/L in humans (42). Studies in other tissues reportedinterstitial ATP in the nanomolar range (43,44), but noneof these experiments could take into account the ATP con-centration in the unstirred layer covering the cell surface(45). Overall, the basal interstitial concentration of ATPcan be estimated in the range of 1–100 nmol/L, but inpathological situations, extracellular ATP concentration canrise markedly, reaching up to 10 mmol/L, as was recentlyobserved in a tumor microenvironment (46). In contrast, itis unknown whether extracellular/interstitial nucleotide lev-els change during obesity-associated inflammation.

Although significantly higher than the level of ATPdetected in CM-PO and CM-BSA, the concentration ofATP in CM-PA reached 10 nmol/L, a concentration thatwas insufficient to induce monocyte migration. However,the monocyte-chemoattracting activity of CM-PA wasabolished by apyrase, and we corroborated that apyrasecleaves all nucleotide triphosphates. While the vastmajority of studies on physiologically released nucleotideshave focused on ATP, UTP is also released during cardiacischemia (47) and is a potent inducer of cell migration(48). Since cytidine triphosphate, GTP, and TTP could notbe detected in CM-PA (assayed with a detection limit of 2nmol/L), and .1 mmol/L each is required for effectivechemoattraction (Fig. 5), these nucleotides are unlikely toparticipate in the CM-PA–induced chemoattraction. Hence,we surmise that a combination of ATP, ADP, UDP, and/orUTP constitute the chemoattracting find-me signal formonocytes that is broadcast toward palmitate-challengedmuscle cells. Consistent with this assertion, monocytesexpress several P2 receptors that selectively recognize

nucleotides. Even though all of them respond to ATP,P2Y2 and P2Y4 have high affinity for UTP; P2Y6, andP2Y8 have high affinity for UDP; and P2Y1, P1Y12, andP2Y13 have high affinity for ADP (29). As two broad-spectrum nonselective antagonists of the P2 receptors, sur-amin and PPADS, reduced the CM-PA–induced monocyteattraction, it is conceivable that nucleotides released byPA-challenged myotubes enact chemoattracting activity byactivating one or more P2X/P2Y receptors on monocytes.

Implications for Inflammation Associated WithMetabolic DiseaseThe results described in this study bring a new un-derstanding to the lipotoxic inflammatory response ofmyotubes. Along with our previously reported inflamma-tory polarization of macrophages conferred by palmitate-treated muscle cells (22), these results illustrate thebidirectional crosstalk that occurs between muscle andimmune cells in the context of hyperlipidic environments.Such bidirectional communication could be ascertainedonly through the described use of defined cell cultureparadigms.

In addition to the palmitate-induced increase inpannexin-3 in L6 myotubes and pannexin-2 in humanmyotubes, our pilot in vivo results suggest that pannexin-2and pannexin-3 are upregulated in eWAT from obese mice(and there might be a trend toward an increase in the

Figure 8—Schematic representation of how nucleotides arereleased through pannexin (Panx) channels from fatty acid–challenged muscle cells and attract monocytes. Activation of theTLR4/NF-kB pathways leads to an increase in pannexin expressionand opening at the plasma membrane. The subsequent nucleotiderelease attracts monocytes. The figure was created using ServierMedical Art (http://www.servier.com).

3824 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

quadriceps muscle). The response in tissues is more com-plex than in cells, as muscle and adipose tissue arecomposed of various cell types and express differentpannexin isoforms. Moreover, increasing the expressionof pannexin is not the only way to induce nucleotide re-lease, because opening of the channels can also be re-gulated. Animal models will be needed in the future toprovide deeper insight into the role of nucleotides andpannexin channels in diet-induced obesity.

Irrespective of whether changes in pannexin expressionoccur in muscle with obesity, our results show a previouslyunrealized principle, that treatment of myotubes with PAresults in the expression of pannexin-3 through the TLR/NF-kB pathway and a consequent release of nucleotides tothe medium to chemoattract monocytes (Fig. 8). Whetherincreases in pannexin expression and/or pannexin openingare required needs to be explored, but our siRNA resultsshow that the existence of pannexin channels is requiredfor nucleotide release and monocyte chemoattraction.

In conclusion, these findings constitute a proof ofconcept of muscle-to-monocyte communication in hyper-lipidic environments, and raise the possibility that, invivo, muscle fibers might also release nucleotides throughpannexins to promote macrophage infiltration of skeletalmuscle. Our findings would also predict that targeting thepathways responsible for chemokine production may beinsufficient to reduce macrophage infiltration of muscle,because other factors such as nucleotides may playa significant role in immune cell chemoattraction. Finally,our studies point to pannexins as interesting targets fortapering the recruitment of tissue inflammatory macro-phages during metabolic disease.

Acknowledgments. The authors thank Dr. Sheila Costford (The Hospitalfor Sick Children, Toronto, Ontario), for providing the RNA from obese animals,and Dr. Michael Salter (The Hospital for Sick Children), for helpful discussionaround nucleotides.Funding. This project was supported by grants from the Canadian DiabetesAssociation and the Canadian Institutes of Health Research (grant MT12601) toA.K. N.J.P. was supported by postdoctoral awards from the Research TrainingCentre at The Hospital for Sick Children and from the Banting and Best DiabetesCentre of the University of Toronto. Y.E.L. was supported by a summer student-ship from the Banting and Best Diabetes Centre of the University of Toronto.Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. N.J.P. participated in the design of the study,coordinated and carried out the majority of the experiments, performed thestatistical analysis, wrote the manuscript, and approved the final manuscript.Y.E.L. helped with experiments and analysis concerning muscle signaling andmonocyte migration, and read and approved the final manuscript. L.N.F. per-formed the Luminex multiplex analysis, and read and approved the final manu-script. J.T.B., A.N., and M.-S.K. performed the liquid chromatography massspectrometry analysis, and read and approved the final manuscript. P.J.B. par-ticipated in the design of the study, helped to perform the experiments, helped towrite the manuscript, and read and approved the final manuscript. A.K. con-ceived the study, participated in its design and coordination, participated in thewriting of the manuscript, and read and approved the final manuscript. A.K. is theguarantor of this work and, as such, had full access to all the data in the study

and takes responsibility for the integrity of the data and the accuracy of the dataanalysis.

References1. World Health Organization (WHO). Obesity and overweight (fact sheet No.311) [article online], 2011. Available from http://www.who.int/mediacentre/factsheets/fs311/en/index.html. Accessed 12 June 20142. Donath MY, Shoelson SE. Type 2 diabetes as an inflammatory disease. NatRev Immunol 2011;11:98–1073. DeFronzo RA, Tripathy D. Skeletal muscle insulin resistance is the primarydefect in type 2 diabetes. Diabetes Care 2009;32(Suppl. 2):S157–S1634. Fink LN, Oberbach A, Costford SR, et al. Expression of anti-inflammatorymacrophage genes within skeletal muscle correlates with insulin sensitivity inhuman obesity and type 2 diabetes. Diabetologia 2013;56:1623–16285. Fink LN, Costford SR, Lee YS, et al. Pro-Inflammatory macrophages in-crease in skeletal muscle of high fat-Fed mice and correlate with metabolic riskmarkers in humans. Obesity (Silver Spring) 2014;22:747–7576. Pedersen BK. Exercise-induced myokines and their role in chronic diseases.Brain Behav Immun 2011;25:811–8167. Prisk V, Huard J. Muscle injuries and repair: the role of prostaglandins andinflammation. Histol Histopathol 2003;18:1243–12568. Mortensen SP, Thaning P, Nyberg M, Saltin B, Hellsten Y. Local release ofATP into the arterial inflow and venous drainage of human skeletal muscle: in-sight from ATP determination with the intravascular microdialysis technique.J Physiol 2011;589:1847–18579. Spriet LL, Howlett RA, Heigenhauser GJ. An enzymatic approach to lactateproduction in human skeletal muscle during exercise. Med Sci Sports Exerc2000;32:756–76310. Coll T, Palomer X, Blanco-Vaca F, et al. Cyclooxygenase 2 inhibition ex-acerbates palmitate-induced inflammation and insulin resistance in skeletalmuscle cells. Endocrinology 2010;151:537–54811. Osorio-Fuentealba C, Contreras-Ferrat AE, Altamirano F, et al. Electricalstimuli release ATP to increase GLUT4 translocation and glucose uptake viaPI3Kg-Akt-AS160 in skeletal muscle cells. Diabetes 2013;62:1519–152612. Riquelme MA, Cea LA, Vega JL, et al. The ATP required for potentiation ofskeletal muscle contraction is released via pannexin hemichannels. Neurophar-macology 2013;75:594–60313. Spray DC, Ye Z-C, Ransom BR. Functional connexin “hemichannels”:a critical appraisal. Glia 2006;54:758–77314. D’hondt C, Ponsaerts R, De Smedt H, et al. Pannexin channels in ATP re-lease and beyond: an unexpected rendezvous at the endoplasmic reticulum. CellSignal 2011;23:305–31615. Xiao F, Waldrop SL, Khimji A-K, Kilic G. Pannexin1 contributes to patho-physiological ATP release in lipoapoptosis induced by saturated free fatty acids inliver cells. Am J Physiol Cell Physiol 2012;303:C1034–C104416. Qu Y, Misaghi S, Newton K, et al. Pannexin-1 is required for ATP releaseduring apoptosis but not for inflammasome activation. J Immunol 2011;186:6553–656117. Sandilos JK, Chiu YH, Chekeni FB, et al. Pannexin 1, an ATP releasechannel, is activated by caspase cleavage of its pore-associated C-terminalautoinhibitory region. J Biol Chem 2012;287:11303–1131118. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innateimmunity and fatty acid-induced insulin resistance. J Clin Invest 2006;116:3015–302519. Lee JS, Pinnamaneni SK, Eo SJ, et al. Saturated, but not n-6 poly-unsaturated, fatty acids induce insulin resistance: role of intramuscular accu-mulation of lipid metabolites. J Appl Physiol (1985) 2006;100:1467–147420. Samokhvalov V, Bilan PJ, Schertzer JD, Antonescu CN, Klip A. Palmitate-and lipopolysaccharide-activated macrophages evoke contrasting insulin re-sponses in muscle cells. Am J Physiol Endocrinol Metab 2009;296:E37–E4621. Suganami T, Tanimoto-Koyama K, Nishida J, et al. Role of the Toll-likereceptor 4/NF-kappaB pathway in saturated fatty acid-induced inflammatory

diabetes.diabetesjournals.org Pillon and Associates 3825

changes in the interaction between adipocytes and macrophages. ArteriosclerThromb Vasc Biol 2007;27:84–9122. Pillon NJ, Arane K, Bilan PJ, Chiu TT, Klip A. Muscle cells challenged withsaturated fatty acids mount an autonomous inflammatory response that activatesmacrophages. Cell Commun Signal 2012;10:3023. Dimopoulos N, Watson M, Sakamoto K, Hundal HS. Differential effects ofpalmitate and palmitoleate on insulin action and glucose utilization in rat L6skeletal muscle cells. Biochem J 2006;399:473–48124. Tamrakar AK, Schertzer JD, Chiu TT, et al. NOD2 activation induces musclecell-autonomous innate immune responses and insulin resistance. Endocrinology2010;151:5624–563725. Zou W, Tolstikov VV. Pattern recognition and pathway analysis with geneticalgorithms in mass spectrometry based metabolomics. Algorithms 2009;2:638–66626. Alvarez Y, Valera I, Municio C, et al. Eicosanoids in the innate immuneresponse: TLR and non-TLR routes. Mediators Inflamm 2010;2010:20192927. Elliott MR, Chekeni FB, Trampont PC, et al. Nucleotides released by apo-ptotic cells act as a find-me signal to promote phagocytic clearance. Nature2009;461:282–28628. Miloševi�c M, Petrovi�c S, Veli�ckovi�c N, Grkovi�c I, Ignjatovi�c M, Horvat A. ATPand ADP hydrolysis in cell membranes from rat myometrium. Mol Cell Biochem2012;371:199–20829. Erlinge D. P2Y receptors in health and disease. Adv Pharmacol 2011;61:417–43930. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AWJr. Obesity is associated with macrophage accumulation in adipose tissue. J ClinInvest 2003;112:1796–180831. Patsouris D, Li PP, Thapar D, Chapman J, Olefsky JM, Neels JG. Ablation ofCD11c-positive cells normalizes insulin sensitivity in obese insulin resistantanimals. Cell Metab 2008;8:301–30932. Varma V, Yao-Borengasser A, Rasouli N, et al. Muscle inflammatory re-sponse and insulin resistance: synergistic interaction between macrophages andfatty acids leads to impaired insulin action. Am J Physiol Endocrinol Metab 2009;296:E1300–E131033. Hevener AL, Olefsky JM, Reichart D, et al. Macrophage PPAR gamma isrequired for normal skeletal muscle and hepatic insulin sensitivity and full an-tidiabetic effects of thiazolidinediones. J Clin Invest 2007;117:1658–166934. Inouye KE, Shi H, Howard JK, et al. Absence of CC chemokine ligand 2 doesnot limit obesity-associated infiltration of macrophages into adipose tissue. Di-abetes 2007;56:2242–2250

35. Chen A, Mumick S, Zhang C, et al. Diet induction of monocyte chemo-attractant protein-1 and its impact on obesity. Obes Res 2005;13:1311–132036. Amano SU, Cohen JL, Vangala P, et al. Local proliferation of macrophagescontributes to obesity-associated adipose tissue inflammation. Cell Metab 2014;19:162–17137. Haase J, Weyer U, Immig K, et al. Local proliferation of macrophages inadipose tissue during obesity-induced inflammation. Diabetologia 2014;57:562–57138. Green CJ, Macrae K, Fogarty S, Hardie DG, Sakamoto K, Hundal HS.Counter-modulation of fatty acid-induced pro-inflammatory nuclear factor kBsignalling in rat skeletal muscle cells by AMP-activated protein kinase. Biochem J2011;435:463–47439. Weigert C, Brodbeck K, Staiger H, et al. Palmitate, but not unsaturated fattyacids, induces the expression of interleukin-6 in human myotubes throughproteasome-dependent activation of nuclear factor-kappaB. J Biol Chem 2004;279:23942–2395240. Pillon NJ, Bilan PJ, Fink LN, Klip A. Cross-talk between skeletal muscle andimmune cells: muscle-derived mediators and metabolic implications. Am JPhysiol Endocrinol Metab 2013;304:E453–E46541. Li J, Gao Z, Kehoe V, Xing J, King N, Sinoway L. Interstitial adenosinetriphosphate modulates muscle afferent nerve-mediated pressor reflex. MuscleNerve 2008;38:972–97742. Mortensen SP, González-Alonso J, Nielsen J-J, Saltin B, Hellsten Y. Muscleinterstitial ATP and norepinephrine concentrations in the human leg during ex-ercise and ATP infusion. J Appl Physiol (1985) 2009;107:1757–176243. Gifford JR, Heal C, Bridges J, Goldthorpe S, Mack GW. Changes in dermalinterstitial ATP levels during local heating of human skin. J Physiol 2012;590:6403–641144. Nishiyama A, Majid DS, Walker M 3rd, Miyatake A, Navar LG. Renal in-terstitial atp responses to changes in arterial pressure during alterations in tu-buloglomerular feedback activity. Hypertension 2001;37:753–75945. Seminario-Vidal L, Lazarowski ER, Okada SF. Assessment of extracellularATP concentrations. Methods Mol Biol 2009;574:25–3646. Falzoni S, Donvito G, Di Virgilio F. Detecting adenosine triphosphate in thepericellular space. Interface Focus 2013;3:2012010147. Erlinge D, Harnek J, van Heusden C, Olivecrona G, Jern S, Lazarowski E.Uridine triphosphate (UTP) is released during cardiac ischemia. Int J Cardiol2005;100:427–43348. Rossi L, Manfredini R, Bertolini F, et al. The extracellular nucleotide UTP isa potent inducer of hematopoietic stem cell migration. Blood 2007;109:533–542

3826 Muscle Cell–Derived Nucleotides Attract Monocytes Diabetes Volume 63, November 2014

SUPPLEMENTARY DATA

©2014 American Diabetes Association. Published online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1

Supplementary Figure 1. Palmitate does not affect the release of cytokines and chemokines from L6 myotubes. L6 myotubes were treated with 0.5mM PA or PO for 18h. Cytokine and chemokine content in myotube conditioned-media was analyzed using a cytokine array (Rat Cytokine Array C2, RayBiotech) and concentration of TNFα and MCP1/CCL2 in supernatant was confirmed using ELISA immunoassays (Rat quantikine, R&D systems), following manufacturer’s instructions

SUPPLEMENTARY DATA

©2014 American Diabetes Association. Published online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1

Supplementary Figure 2. Necrosis and apoptosis do not correlate with monocyte migration. A) L6 myotubes were treated with PA, PO or BSA control for 18h. Caspase-3 was then measured in cell lysate using specific antibody recognizing both the proand clived forms of caspase-3. A representative blot is show. B) L6 myotubes were mechanically damaged by scraping and vortexing to induce necrosis. Conditioned media was then collected, centrifuged to pellet debris and tested for monocyte attraction as described in methods. Mean ± SEM, n=4. C) LDH release was plotted against CM-induced monocyte migration. Results are the mean from 18 independent experiments.

SUPPLEMENTARY DATA

©2014 American Diabetes Association. Published online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1

Supplementary Figure 3. YoPro uptake into myotubes in response to fatty acids. L6 myotubes differentiated in 6-well plates were treated with 0.5 mM BSA, PA or PO for 18 h as described in Methods. YoPro dye (1 μM) and Texas-red dextran (10 kDa, 0.1 mg/mL) were added for 15 min, then cells were washed 4x in ice-cold PBS, fixed with 3% PFA for 10 min and washed again with PBS. Images were taken with a Leica DMIRE2 fluorescence microscope using the 10X air objective. Total green and red fluorescence were measured using ImageJ software in 15 random fields per condition.

SUPPLEMENTARY DATA

©2014 American Diabetes Association. Published online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1

Supplementary Table 1. MRM transitions and ESI polarity mode used for detection of AMP, ADP, UDP, GMP, CMP, UMP, IMP and adenosine.

Component Precursor ion (m/z) Product ion (m/z) Polarity mode

AMP 348 136 positive

ADP 426 159 negative

UDP 403 159 negative

CMP 324 112 positive

GMP 364 152 positive

UMP 325 97 positive

IMP 349 137 positive

Adenosine 268 136 positive

Supplementary Table 2. Measured NEFA concentration.

Initial Supernatant Change Filtrate RM 25 ± 25 29 ± 34 +16 % < 10 0.2 mM BSA 15 ± 14 24 ± 26 +57 % < 10 PA 197 ± 58 90 ± 21 -54 % < 10 PO 252 ± 80 112 ± 24 -56 % < 10 0.5mM BSA 18 ± 17 28 ± 26 +56 % < 10 PA 478 ± 73 242 ± 36 -49 % < 10 PO 442 ± 119 299 ± 44 -32 % < 10 0.8 mM BSA 34 ± 35 49 ± 34 +45 % < 10 PA 723 ± 136 438 ± 58 -40 % < 10 PO 737 ± 222 467 ± 72 -37 % < 10

NEFA concentrations in the initial media, myotube conditioned-media and filtrates (<3000 Da) were measured using the enzymatic method based on Acyl-CoA oxidase described in Methods. Results are means ± SD expressed in micromoles/litre from at least 4 independent experiments (n≥4).

SUPPLEMENTARY DATA

©2014 American Diabetes Association. Published online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0150/-/DC1

Supplementary Table 3. Measured glucose concentration.

Initial Supernatant Change Filtrate

RM 6.7 ± 0.4 3.9 ± 0.6 -46 % 3.5 ± 0.2

BSA 6.6 ± 0.2 4.2 ± 0.2 -39 % 4.0 ± 0.2

PA 7.0 ± 0.1 3.5 ± 0.4 -48 % 3.8 ± 0.2

PO 6.4 ± 0.6 3.5 ± 0.7 -43 % 4.0 ± 0.1 Glucose concentration in the initial media, conditioned-media and filtrates from L6 myotubes treated for 18h with 0.5 mM palmitate, palmitoleate or BSA control. Concentrations were measured using a glucometer. Results are means ± SD, expressed in mM from at least 4 independent experiments (n≥4). Supplementary Table 4. Viability of L6 myotubes.

LDH release 0.2 mM 0.5 mM 0.8 mM BSA 0.97 ± 0.08 0.75 ± 0.05 0.90 ± 0.07 PA 0.95 ± 0.03 0.99 ± 0.08 1.22 ± 0.04* PO 0.85 ± 0.10 0.79 ± 0.04 1.05 ± 0.10 MTT reduction 0.2 mM 0.5 mM 0.8 mM

BSA 0.79 ± 0.02 0.75 ± 0.05 0.73 ± 0.05 PA 0.73 ± 0.09 0.54 ± 0.04 0.45 ± 0.07* PO 0.79 ± 0.07 0.70 ± 0.05 0.64 ± 0.04

The potential cytotoxicity caused by fatty acid treatment was estimated through LDH release and MTT reduction after exposing L6 myotubes to BSA, PA and PO for 18h. Results are normalized to the regular media control and expressed as mean ± SD, n ≥4. * p < 0.05 vs BSA control.