Nitric Oxide–Releasing Nanoparticles PreventPropionibacterium acnes–Induced Inflammation byBoth Clearing the Organism and Inhibiting MicrobialStimulation of the Innate Immune ResponseMin Qin1, Angelo Landriscina2, Jamie M. Rosen2, Gabrielle Wei1, Stephanie Kao1, William Olcott1,George W. Agak1, Karin B. Paz2, Josephine Bonventre3, Alicea Clendaniel3, Stacey Harper3,4,Brandon L. Adler2, Aimee E. Krausz2, Joel M. Friedman5, Joshua D. Nosanchuk6,7, Jenny Kim1,8 andAdam J. Friedman2,5,9

Propionibacterium acnes induction of IL-1 cytokines through the NLRP3 (NLR, nucleotide oligomerizationdomain-like receptor) inflammasome was recently highlighted as a dominant etiological factor for acne vulgaris.Therefore, therapeutics targeting both the stimulus and the cascade would be ideal. Nitric oxide (NO), a potentbiological messenger, has documented broad-spectrum antimicrobial and immunomodulatory properties. Toharness these characteristics to target acne, we used an established nanotechnology capable of generating/releasing NO over time (NO-np). P. acnes was found to be highly sensitive to all concentrations of NO-np tested,although human keratinocyte, monocyte, and embryonic zebra fish assays revealed no cytotoxicity. NO-npsignificantly suppressed IL-1β, tumor necrosis factor-α (TNF-α), IL-8, and IL-6 from human monocytes, and IL-8and IL-6 from human keratinocytes, respectively. Importantly, silencing of NLRP3 expression by small interferingRNA did not limit NO-np inhibition of IL-1 β secretion from monocytes, and neither TNF-α nor IL-6 secretion, norinhibition by NO-np was found to be dependent on this pathway. The observed mechanism by which NO-npimpacts IL-1β secretion was through inhibition of caspase-1 and IL-1β gene expression. Together, these datasuggest that NO-np can effectively prevent P. acnes-induced inflammation by both clearing the organism andinhibiting microbial stimulation of the innate immune response.

Journal of Investigative Dermatology (2015) 135, 2723–2731; doi:10.1038/jid.2015.277; published online 6 August 2015

INTRODUCTIONAcne’s multifactorial etiology, resulting from a mix ofhormone-induced elevations in sebum production, abnormalfollicular epithelial desquamation and proliferation, hyperco-lonization of P. acnes, and host inflammatory reactions,makes treatment often times challenging (Zouboulis et al.,2005; Castro and Ferreira, 2008). This challenge has fueledthe development of innovative delivery platforms thatimprove delivery and efficacy of established andinvestigative agents. Various particulate delivery platformshave been evaluated for the use in acne, including liposomesand nanoparticles (nps). These carriers have the advantage ofgradual release (Manconi et al., 2002), minimizing irritancy(Manconi et al., 2002), follicular targeting/enhanced follicularpenetration (Jung et al., 2006), and cutaneous retention of theactive ingredient (Manconi et al., 2006). More importantly,the unique properties conferred by the vehicles existing at thenanoscale are allowing for the investigation of newtherapeutics, which in their bulk form can not be deliveredtopically because of stability and solubility issues, includinglauric acid (Nakatsuji et al., 2009) and chitosan (Friedmanet al., 2013).

ORIGINAL ARTICLE

1Division of Dermatology, Department of Medicine, David Geffen School ofMedicine, University of California, Los Angeles, California, USA; 2Division ofDermatology, Department of Medicine, Montefiore Medical Center, Bronx,New York, USA; 3Department of Environmental and Molecular Toxicology,Oregon State University, Corvallis, Oregon, USA; 4School of Chemical,Biological and Environmental Engineering, Oregon State University, Corvallis,Oregon, USA; 5Department of Physiology and Biophysics, Albert EinsteinCollege of Medicine, Bronx, New York, USA; 6Division of Infectious Diseases,Department of Medicine, Albert Einstein College of Medicine, Bronx, New York,USA; 7Department of Microbiology and Immunology, Albert Einstein College ofMedicine, Bronx, New York, USA; 8Department of Dermatology, Greater LosAngeles Healthcare Service Veterans Affairs, Los Angeles, California, USA and9Department of Dermatology, George Washington School of Medicine andHealth Sciences, Washington, DC, USA

Correspondence: Adam J. Friedman, Department of Dermatology, GeorgeWashington School of Medicine and Health Sciences, 2150 PennsylvaniaAvenue, NW, South Pavilion 2nd Floor, Washington, DC 20037, USA.E-mail: ajfriedman@mfa.gwu.edu

Received 6 January 2015; revised 15 June 2015; accepted 20 June 2015;accepted article preview online 14 July 2015; published online 6 August 2015

Abbreviations: MOI, multiplicity of infection; NO, nitric oxide; np,nanoparticle; NLR, nucleotide oligomerization domain-like receptor; PBMC,peripheral blood mononuclear cell; siRNA, small interfering RNA; TNF-α,tumor necrosis factor-α

© 2015 The Society for Investigative Dermatology www.jidonline.org 2723

Along this vein, np platforms can be a vehicle throughwhich unstable therapeutically relevant agents are harnessed.Nitric oxide (NO) represents one such molecule—a diatomic,lipophilic gaseous molecule with functions described rangingfrom vascular modulation to cell cycle regulation, pro- andanti-inflammatory properties, and microbicidal/static activity(Friedman and Friedman, 2009). The simplicity of NO mixedwith its biological potency and complexity makes it anextremely promising pharmacological agent, but utilizationhas been limited because of the lack of effective and safedelivery systems. We previously reported on a hybridhydrogel-sugar-based nanoparticulate system (np) capable ofgenerating physiologic concentrations of NO over time. Onexposure to an aqueous environment, the hydrogen bondingnetwork that comprises the np matrix loosens, allowing releaseof the NO (Friedman et al., 2008; Han et al., 2011). Thisplatform is unique in that the np matrix facilitates thegeneration of NO from encapsulated sodium nitrite througha dinitrogen trioxide intermediate, as the rich disulfidebonding network allows for long-range proton and electrontransfer/movement. As opposed to organic nitrates, the mostcommonly used NO donor in clinical practice, there is nopotential for decreased efficacy with prolonged and contin-uous use, what is known as ‘nitrite tolerance,’ which resultsfrom the prerequisite host reduction in the administered nitrateto active NO (Bennett et al., 1994; Gori and Parker, 2002).We have previously shown that NO-nps significantly

increase the healing of methicillin-resistant Staphylococcusaureus-infected wounds in a murine excisional injury model(Martinez et al., 2009). NO-np also accelerated woundhealing in non-obese, diabetic, severe combined immunode-ficiency mice, with significantly decreased neutrophil andincreased macrophage infiltration noted in the NO-np-treatedwound beds as compared with another topical NO-donatingplatform, diethylenetriamine (Blecher et al., 2012). Given theantimicrobial and immunological properties of NO, weinvestigated whether NO-np has antimicrobial activity againstPropionibacterium acnes and modifies the inflammatorycascade associated with the evolution of acne lesions.

RESULTSCharacterization of nps, detection of NO release, and tissuepenetrationNO-nps were formed using a previously described protocol(Friedman et al., 2008). In brief, the highly anhydrous interiorof the nps facilitates the formation of N2O3 from nitrite andprotons. N2O3 is both the source of NO and is a potentS-nitrosating agent that likely contributes to the biologicalefficacy of NO-np.Using scanning electron microscopy, NO-np aggregates

were found to average ~ 127 nm in diameter (Figure 1a). Wehave previously shown by transmission electron microscopythat individual particles measure ~ 10 nm in diameter(Friedman et al., 2008). Using dynamic light scattering, NO-np revealed an average hydrodynamic diameter of 216.9 nmbased on 40 acquisition attempts (Figure 1b). The SD was14.2 nm (3.3%), proving that NO-nps are homogenous insize. As NO-nps swell with moisture, the average diameter is

likely an overestimate as compared with the scanningelectron microscopy. NO release from the NO-np has beenpreviously reported using both amperometric detection andozone chemiluminescence (Friedman et al., 2008; Cabraleset al., 2010).

Antimicrobial activity of NO-npNO generated from the NO-np has demonstrated antimicro-bial activity against various Gram-positive and -negativepathogens, including Staphylococcus aureus (Martinez et al.,2009) and Pseudomonas aeruginosa (Friedman et al., 2011),and therefore we sought to evaluate the impact on P. acnes.P. acnes strains were incubated with various concentrations

of NO-np and np (0.625, 1.25, 2.5, or 5 mgml− 1) for 24 hours(Figure 2a). For all isolates tested, all concentrations of NO-npsignificantly inhibited bacterial growth compared with con-trols in a dose-dependent manner for up to 24 hours.Concentrations over 2.5 mgml− 1 completely inhibitedgrowth over 24 hours, whereas 1.25 and 2.5 mgml−1

significantly inhibited growth after 24 hours of co-incubation,respectively (Po0.0001), as compared with all controls, aswell as 0.625mgml− 1 NO-np. Statistical significance was notmet when comparing the two treatment groups (P=0.07). Todemonstrate the bactericidal activity against P. acnes, bacteriawere incubated with varying doses of NO-np, as well as npcontrol alone for 4 hours. Subsequently, the bacteria wereplated and the viability was determined by a colony-formingunit assay. The NO-nps were effective at significantly killingP. acnes, resulting in only 8.9 and 4.6% survival for the twoconcentrations illustrated (Figure 2b). Interestingly, the 2.5and 5mgml− 1 np control had a modest, albeit a significanteffect, compared with untreated cells, likely owing to thephysical presence of the nps that probably imparted a stericeffect, interfering with cell–cell interactions, noted in previousinvestigations (Friedman et al., 2011).

NO-nps inhibit P. acnes–stimulated inflammatory cytokinesGiven the importance of the inflammatory response in thepathophysiology of acne, we evaluated NO’s ability to inhibitinflammatory cytokine production from a P. acnes–chal-lenged human keratinocyte cell line, HaCaT, and peripheral

50

40

30

20

10

00 200 400 600

% In

tens

ity

Radius (nm)800 1,000

Figure 1. Nanoparticles (nps) characterization. (a) Scanning electronmicroscopy revealed distinct spherical nps with irregular surface structureindicative of the porous matrix lattice (bar = 100 nm). (b) Using dynamic lightscattering (DLS), nitric oxide–releasing np (NO-np) revealed an averagehydrodynamic radius of 108.47 nm based on 40 acquisition attempts. The SDwas 14.2 nm (3.3%), proving that NO-nps are homogenous in size.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

2724 Journal of Investigative Dermatology (2015), Volume 135

blood mononuclear cells (PBMCs). Previous studies showedthat P. acnes induces the inflammatory cytokines IL-1, IL-8,and IL-12 in human monocytes and IL-6 in humankeratinocytes (Friedman et al., 2013; Qin et al., 2014). Atthe same concentrations tested in the antimicrobial assays,NO-np significantly reduced P. acnes–stimulated humanmonocyte (PBMC) expression of IL-1 β (52.0% reduction)and tumor necrosis factor-α (TNF-α) (91.3% reduction) at thehighest dose tested (Figure 3a and b). Control np had noinhibitory effect on IL-1β secretion at all concentrations tested(S2). NO-np significantly inhibited P. acnes–stimulatedPMBCs and HaCaT IL-6 and IL-8 secretion in a dose-dependent manner as compared with controls (Figure 3c–f).For IL-6 and IL-8 secretion, respectively, a statisticallysignificant difference was noted when comparing controlwith 2.5 mgml−1 (33.4 and 54.84% reduction) and 5mg-ml− 1 (54.8 and 88.47% reduction) for the PBMC group andwith all treatment groups with the HaCaT cells (15.1, 51.5,and 65.0%, and 48.31, 90.76, and 98.93, respectively).With the recent attention paid to IL-1β as the initial

subclinical stimulus for follicular hyperkeratinization, itsplace in and importance of the NLRP3 (NLR, nucleotideoligomerization domain-like receptor) inflammasome withrespect to P. acnes induction of the host defense (Qin et al.,

2014), we sought to determine whether the aforementionedanti-inflammatory effects resulted from NO-np’s impact onthis multi-protein complex. To evaluate whether NO-npreduction of IL-1β, IL-6, and TNF-α secretion was mediatedspecifically through NO activity on NLRP1 and/or NLRP3,small interfering RNAs (siRNAs) were used to silence specificgene expression as previously described (Qin et al., 2014).Transfection of monocytes with specific siRNA targetingNLRP3 resulted in a significant reduction in IL-1 β expression(93.8%, Po0.01) (Figure 4a), as compared with stimulatedcontrols, respectively. A nonspecific control siRNA had noeffect in all cases. Although siNLRP3 transfection of PBMCsresulted in a 93.8% reduction in IL-1β secretion, NO-nptreatment provided further statistically significant reduction(499%) as compared with stimulated siNLRP3 control. Incongruence with past results, siNLRP1 transfection of PBMCsresulted in a minimal reduction in P. acnes–induced IL-1 βsecretion, whereas a 55.27% reduction in IL-1 β expressionfollowing NO-np treatment compared with untreated wasnoted. Interestingly, knockout of NLRP3 did not influence P.acnes–stimulated secretion of IL-6, IL-8, or TNF-α from PMBCs,and NO-np treatment yielded similar reductions in cytokinesecretion as noted above (S1a-c). However, knockdown ofNLRP1 had a significant effect in NO-np inhibition of IL-6secretion from PBMCs as compared with NLRP3 knockdown.

NO-nps inhibit P. acnes–induced caspase-1 and IL-1β, but notNLRP3, gene expressionGiven the additional reduction in IL-1β secretion fromsiNLRP3-transfected PBMCs beyond the effects of knockdownalone, we sought to evaluate NO-np activity on geneexpression of key inflammasome components. IL-1β andcaspase-1 mRNA levels were significantly reduced in PBMCsexposed to P. acnes and treated with NO-np as compared withcontrols (87.1 and 80.0% reduction, respectively; Figure 4aand b), with an associated decrease in caspase-1 productionand activity (Figure 5a and b). NO-np treatment had minimaleffect on both NLRP3 and apoptosis-associated speck-likeprotein containing a CARD mRNA expression as comparedwith controls (Figure 4d and e). NO-np alone did cause a slightbut a significant increase in IL-1β mRNA expression.

NO-nps are non-toxicAll concentrations of NO-np tested had minimal toxicity tokeratinocytes and PBMCs using the MTT (3-(4,5-dimethylthia-zol-2-yl)-2,5-diphenyltetrasodium bromide tetrazolium) assay(P=0.31 and 0.13, respectively, as compared with control)(Figure 6a and b), which was further substantiated usingembryonic zebra fish assays (Figure 6c). The embryoniczebrafish assay was used to assess the toxicological impact ofNO-np in an in vivo vertebrate model system. Embryosexposed to NO-nps did not exhibit differences in rates ofmalformations or mortality at any concentration testedcompared with controls. Any differences in morphology,mortality, development, larval morphology, or behavioralendpoints between treatment and control were attributedto laboratory background. Therefore, our data suggest thatNO-nps can inhibit P. acnes–induced cytokine production in

Per

cent

sur

viva

l (%

)

Contro

l

np (0

.625

mg/

ml)

np (1

.25

mg/

ml)

np (2

.5 m

g/m

l)

np (5

mg/

ml)

NO-np

(0.6

25m

g/m

L)

NO-np

(1.2

5 m

g/m

L)

NO-np

(2.5

mg/

mL)

NO-np

(5 m

g/m

L)

0

50

100

150

******

*

0 4 8 12 16 20 240.0

0.2

0.4

0.6

0.8

Hour

OD

600

Controlnp (0.625 mg/ml)np (1.25 mg/ml)np (2.5 mg/ml)np (5 mg/ml)NO-np (0.625 mg/ml)NO-np (1.25 mg/ml)NO-np (2.5 mg/ml)NO-np (5 mg/ml)

***

Figure 2. Antimicrobial effects of nitric oxide–releasing nanoparticles(NO-nps). (a) Susceptibility of P. acnes isolates to varying concentrations ofempty nps and NO-np (0.625, 1.25, 2.5, and 5mgml−1) was investigatedby real-time Bioscreen analysis and (b) the colony-forming unit assaydetermination (mean colony-forming unit per ml). Experiments were repeatedin triplicate and performed at least thrice on separate days. Asterisks denoteP-value significance calculated by unpaired two-tailed t-test analysis.*Po0.05; **Po0.01.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

www.jidonline.org 2725

human keratinocytes and PBMCs, and this is not a conse-quence of the release of cytokines at cell death.

DISCUSSIONIn this work, we demonstrated that the NO-nps effectivelydecreased P. acnes viability and inhibited IL-1β, TNF-α, IL-8,and IL-6 secretion from P. acnes–stimulated PMBCs, and IL-8and IL-6 from keratinocytes, while neither having in-vitro norin-vivo cytotoxicity. Although NO-np was found to have animpact on components of the NLRP3 inflammasome pathway,its anti-inflammatory effects were not specific to NLRP3 andNLRP3 knockout did not impair NO-np’s anti-inflammatoryproperties.Nearly every member of the skin cell population expresses

NO synthase and is thereby able to produce NO toaccomplish key physiologic processes. NO’s role in skin hostdefense is concentration dependent and is imparted in abimodal manner. At low concentrations it stimulates theimmune system by enhancing immune cell activity, cytokineproduction, expression of adhesion factors, and extracellular

matrix constituent synthesis. High output of endogenous NOis achieved through the stimulation of pathogen recognitionreceptors that promote induced NO synthase transcriptionand subsequent L-arginine-dependent NO release (Griffithand Stuehr, 1995). This calcium-independent enzymaticpathway along with the nonenzymatic reduction in naturallyoccurring nitrosothiols and nitrate (NO3

−) are the two keymethods of endogenous NO synthesis pertinent to microbialeradication (Weller, 2009). The coupling of NO and O2

−

produces potentially toxic reactive nitrogen species, such asthe peroxynitrite anion (ONOO−), and induces increased O2

−

production (Fang, 2011). Reactive nitrogen species caninterfere with normal protein interactions as protein thiolsbecome nitrosated (Han et al., 2011). Collectively, thesemolecules adversely affect microbial viability by traversingcellular membranes, interrupting DNA replication, instigatinglipid peroxidation, and inactivating metalloproteinasesessential for microbial metabolism (Wink and Mitchell, 1998).P. acnes, which resides in pilosebaceous follicles in both

acne and non-acne subjects, elicits inflammatory responses

1,400

1,200

1,000

800

600

400

200

16,000

14,000

12,000

6,000

4,000

0

2,000

2,000

1,500

1,000

500

0

2,500

3,000

8,000

10,000

0

P. acnes–– –

– 1.25 52.5++ + + +

P. acnes

–

– –

– 1.25 52.5+

+ + + +

P. acnes–– –

– 1.25 52.5++ + + +

2,0001,8001,6001,4001,2001,000

800600400200

0

2,000

1,500

1,000

500

0

2,500

2,000

1,500

1,000

500

0

2,500

TN

Fα

(pg

ml–1

)IL

-8 (

pg m

l–1)

IL-8

(pg

ml–1

)

IL-8

(pg

ml–1

)

IL-8

(pg

ml–1

)IL

-8 (

pg m

l–1)

P. acnes

–

– –

– 1.25 52.5+

+ + + +

NO-np (mg ml–1)NO-np (mg ml–1)

NO-np (mg ml–1)NO-np (mg ml–1)

NO-np (mg ml–1)NO-np (mg ml–1)

P. acnes–– –

– 1.25 52.5++ + + +

P. acnes

–

– –

– 1.25 52.5+

+ + + +

*

***

*

**

*

**

**

***

*

**

**

*

Figure 3. Nitric oxide–releasing nanoparticle (NO-np) significantly inhibits P. acnes–induced pro-inflammatory cytokines in peripheral blood mononuclearcell (PMBC) and keratinocytes. P. acnes–stimulated PBMC expression of IL-1β (a), tumor necrosis factor-α (TNF-α) (b), IL-6 (c), and IL-8 (e), and HaCaT cellexpression of IL-6 (d) and IL-8 (f) with and without various concentrations of NO-np. NO-np at 2.5 mgml−1 was used as a control. IL-1β and TNF-α wereundetectable in the HaCaT supernatant via ELISA assay (data not shown). The data are presented as the mean of triplicate wells± SD and are representative ofthree individual experiments. *Po0.05; **Po0.01.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

2726 Journal of Investigative Dermatology (2015), Volume 135

essential for the pathogenesis and clinical phenotype of acnevulgaris (Bellew et al., 2011). P. acnes was found to be highlysensitive to NO derived from the NO-np. Perhaps the mostsignificant aspect to be considered of NO’s microbicidalactivity is the lack of demonstrated resistance to date. This isthought to be a direct result of the aforementioned multiplemechanisms of NO’s action—to overcome would require thesimultaneous development of distinct resistance mutations.Some minor compensatory mechanisms against the actions ofNO have been elucidated; however, no protection against thehigh concentrations of NO released by donating materials isevident. Isolates of S. aureus, Staphylococcus epidermidis,Escherichia coli O157:H7, and P. aeruginosa that survivedexposure to high NO concentrations showed no increase inminimum inhibitory concentration levels afterward (Privettet al., 2012). Given the ongoing concern regarding antibioticresistance in P. acnes, NO-np represents a much neededmulti-mechanistic antimicrobial approach (Eady, 1998; Rosset al., 2003).Although P. acnes is implicated in the pathophysiology of

acne vulgaris, it serves more as a pro-inflammatory stimulusrather than a pathogenic organism, and therefore muting thisresponse is central to clinical treatment. We found thatNO-np inhibited IL-8 and IL-6 production from P. acnes–

stimulated PBMCs and keratinocytes in a dose-dependentmanner. IL-8 is a CXC chemokine that serves as a potentchemoattractant for neutrophils, a predominant cell type inacne-related lesions. IL-8 gene expression has been shown tobe significantly elevated in biopsy specimens of acne lesions(Trivedi et al., 2006). IL-6 is a pleiotropic cytokine having animportant role in acute and chronic inflammation. IL-6 hasalso been shown to contribute to persistent neutrophiltrafficking in wounds, resulting in delayed wound healing inthe setting of sustained stress. Interestingly, we found thatNLRP1 may have a role in NO-np downregulation of P.acnes–induced IL-6 expression, as compared with NLRP3,warranting further investigation.NO-np had a profound effect on the secretion of TNF-α

from PBMCs. TNF-α is central to stimulating the acute phasereaction of the immune response. Implicated in multipledermatologic diseases including acne, several studies havehighlighted that single-nucleotide polymorphisms of theTNF-α gene (such as the −308 G/A polymorphism) areassociated with an increased risk to develop chronic inflamma-tory disease (Yang et al., 2014). NO-np significantly inhibitedPBMC induction of TNF-α triggered by P. acnes, almostcompletely at the highest concentration tested (5mgml−1).

**

*

**

*

350

300

250

200

150

100

50

0

6

5

4

3

2

1Cas

pase

- 1

rela

tive

mR

NA

exp

ress

ion

NLR

P3

rela

tive

mR

NA

ex

pres

sion

AS

C r

elat

ive

mR

NA

expr

essi

on

0

IL-1β

rela

tive

mR

NA

ex

pres

sion

400

**

**

**

*

*

IL-1

β (p

g m

l–1)

1,200

1,000

800

600

400

150

100

50

0

P. acnes

NO-np (mg ml–1)

NO-np (mg ml–1)

NO-np (mg ml–1)NO-np (mg ml–1)

NO-np (mg ml–1)

– –

––

52.51.25

P. acnes

Cells alone Scramble siRNA siNLRP3 siNLRP1

– + ++ +– –– – + +

+ +– ––– + +

+ +– ––– + +

+ +– ––

+

++++

– –––

52.51.25P. acnes

+++++

– –

––

52.51.25

P. acnes

+

++++

– –

––

52.51.25

P. acnes

+

++++

2

1.5

1

0.5

0

3

2.5

2

1.5

1

0.5

0

Figure 4. Nitric oxide–releasing nanoparticle (NO-np) inhibits peripheral blood mononuclear cell (PBMC) IL-1β secretion by directly downregulating IL-1βgene and caspase-1 gene expression via the inflammasome pathway. Cells were infected by P. acnes at a multiplicity of infection (MOI) 0.5 with and withoutNO-np and with scramble small interfering RNA (siRNA), siNLRP3, or siNLRP1, and IL-1β secretion in culture supernatants was tabulated via ELISA assay (NO-np2.5 mgml− 1) (a). The expressions of caspase-1, IL-1β, NLRP3, and apoptosis-associated speck-like protein containing a CARD (ASC) were measured by real-timePCR (b–e). Data are shown as the mean of triplicate wells± SD and are representative of three individual experiments. *Po0.05; **Po0.01.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

www.jidonline.org 2727

NO is known to modulate several innate immune path-ways, enhancing the host’s own immune response; forexample, NO is involved in lymphocyte activation andproliferation, and numerous cytokine pathways, includingTNF-α, transforming growth factor-β, p56, IL-6, and NF-κB(Fang, 1997). However, although an important recruiter of hostdefense against invading pathogens, NO’s immunoregulatoryfunction is clearly multi-faceted—NO can function as an anti-inflammatory agent when appropriate, as demonstrated in thisstudy. For example, investigators have shown that NO inhibitskey members of the inflammatory cascade, such as NF-κB andSTAT-1 in keratinocytes (Giustizieri et al., 2002), anddownregulates neutrophil aggregation, secretion, anddiapedesis (Dal Secco et al., 2003). We too demonstratedthat NO-np inhibits secretion of various pro-inflammatorycytokines from different cell types stimulated with P. acnes.In light of the recent data highlighting the role of caspase-

dependent IL-1β generation through the NLRP3 inflamma-some pathway in the pathophysiology of acne vulgaris, (Qinet al., 2014), we showed that NO-np inhibits P. acnes–stimulated PBMC IL-1β and caspase-1 gene expression, likely

explaining the additional reduction in IL-1β secretion from P.acnes–stimulated siNLRP3-transfected monocytes treatedwith NO-np. The NLRP3 inflammasome is a multi-proteincomplex that triggers the maturation of the pro-inflammatorycytokines IL-1β and IL-18 (Hernandez-Cuellar et al., 2012).This complex can be activated by pathogen-associatedmolecular patterns or endogenous danger-associated molecu-lar patterns generated in the setting of cellular injury or tissuedamage (Dostert et al., 2008; Schroder and Tschopp, 2010;Qin et al., 2014). Interestingly, one of NO’s roles may be as anegative regulator of inflammasome activity. It was previouslyshown that inflammasome activation was inversely correlatedto induced NO synthase expression, and that NO inhibitedIL-1β production and caspase-1 activation (Mao et al., 2013).The mechanism by which NO is believed to inhibit the variouselements of the inflammasome is via S-nitrosylation of severalproteins involved in cellular regulation, ultimately impactingtheir activity and function (Stamler et al., 2001). Wedemonstrated that NO-np did not reduce the gene expressionof NLRP3; however, this finding does not provide insight intoNLRP3 activity or the degree of nitrosylation. We havepreviously shown that the NO-nps are effective trans-nitrosat-ing agents (Friedman et al., 2011), likely owing to the npsforming a dinitrogen trioxide intermediate, capable of S-nitrosation under aerobic and anaerobic conditions, whichtherefore offers a mechanism for the observed activity.Regardless of NLRP3 function, we demonstrated NLRP3-

independent anti-inflammatory effects in a NLRP3 knock-down model, likely owing but not limited to NO-np’s directeffect on caspase-1 and IL-1β gene expression. This findinghas far reaching implications in many disease states given thebroad role of both in all inflammasome pathways. Interest-ingly, knockdown of NLRP3 did not impact NO-np suppres-sion of P. acnes–stimulated monocyte-derived IL-6, IL-8, andTNF-α, once again supporting NO’s multi-mechanistic anti-inflammatory activity and highlighting NO’s target diversity.Not surprisingly, we found that NO-np alone mildly

induces PBMC IL-1β expression in the absence of pathogenstimulus, highlighting the duality of NO’s functions inimmune regulation. We and other investigators have shownthat at low concentrations, NO stimulates PMBC generationof IL-6, but as NO levels increased using NO-donormolecules both IL-6 expression and NF-κB activity decreased.A greater degree of nitrosative stress increases the likelihoodof protein S-nitrosylation, in this case, of key NF-κB-relatedproteins including IκB kinase b and p50, which was shown todirectly inhibit pro-inflammatory activity (Marshall et al.,2004). This further supports any NO-releasing technology thatfacilitates trans-nitrosylation. Therefore, the current theory isthat in the early stage of infection, lower, initial concentra-tions of NO may accelerate the immune response bystimulating the synthesis of proinflammatory cytokines,whereas at a later stage, when NO concentrations havepeaked and cells are activated, it rather counteracts theundesired development of inflammation.Together, the presented and past data, demonstrating that

NO-nps have multi-mechanistic antimicrobial and anti-inflammatory properties, suggest that NO-np can potentially

350

*300

250

200

150

100Cas

pase

-1 a

ctiv

ity(%

of c

ontr

ol c

ells

)

50

0P. acnesControl np

No-np

––

– –

––

–

–

–

––+

+

+

+ ++

+

P. acnes

Caspase-120 kDa active form

β-Actin

Control npNo-np

––– –

––

––

–––+

++

+ ++

+

Figure 5. Nitric oxide–releasing nanoparticle (NO-np) downregulates P.acnes–induced caspase-1 protein expression and activity in peripheral bloodmononuclear cells (PBMCs). PBMCs were isolated from whole blood ofnormal healthy donors. Transfected cells were plated into six-well cell cultureplates, cultured at 37 °C overnight, and co-cultured with or without2.5 mgml− 1 control-np and 2.5 mgml− 1 NO-np for an hour. Cells were theninfected by P. acnes at a multiplicity of infection (MOI) 0.5 for 24 hours. Cellswere lysed and caspase-1 protein expression was measured with western blot(a) and activity assay (b). (a) Western blotting using anti-caspase-1: cells alone(lane 1); P. acnes (lane 2); control np (lane 3); NO-np (lane 4); P. acnes in thepresence of NO-np (lane 5); P. acnes in the presence of control np (lane 6). (b)Caspase-1 activity assay: cells were normalized as a percentage of controlcells. NO-np significantly decreases PBMC caspase-1 protein activity in thepresence of P. acnes compared with control-np. *Po0.05.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

2728 Journal of Investigative Dermatology (2015), Volume 135

target multiple pathophysiologic elements of acne vulgaris.Importantly, in the face of rising of antimicrobial resistance inthe setting of prolonged antibiotic use, NO’s bactericidal andstatic properties limit the risk of the emergence of resistantspecies. Therefore, our findings suggest that the NO-np offersa new approach to harness the therapeutic potential of NO,and that this platform should be translated to the bedside.

MATERIALS AND METHODSProduction of NO-npThe production of the NO-np has been previously reported(Friedman et al., 2008). Briefly, the NO-nps are prepared using thefollowing sequence of steps: (1) hydrolyzing tetramethylorthosilicate:stock of 5 ml of tetramethylorthosilicate, 600 μl of deioinized water,and 560 μl of 2 mM hydrochloric acid are added to a small vial. Thecontents of the vial are sonicated for ~ 20–30minutes, yielding aclear solution that is then placed on ice. (2) Mixing the sol-gelcomponents: 1.49 g of sodium nitrite is dissolved in 4ml of PBSbuffer at pH 7.5 followed by sequential addition and mixing of 0.5 mlof polyethylene glycol-200 and 500 μl of chitosan (1 mgml−1). Theresulting mixture is then vortexed thoroughly. Next, 2 ml ofpreviously hydrolyzed tetramethylorthosilicate is added to thesolution followed by vigorous vortexing until complete gelation. (3)Lyophilizing the sol-gel: The resulting gelled material is thenlyophilized for 24–48 hours, which removes all volatile components.(4) Ball milling the lyophilized sol-gel: following lyophilization, thedry material is ball milled at 150 r.p.m. for 8 hours, resulting in a veryfine white powder.

Nanoparticle imagingFor nanoparticle imaging, see Supplementary Methods online.

Dynamic light scatteringFor dynamic light scattering, see Supplementary Methods online.

Nanoparticle skin penetrationRetired breeder male Sprague–Dawley rats (650 g, Charles RiverBreeding Laboratories (Wilmington, MA)) were used in these studies.All animal protocols were approved by the Animal Use Committee atthe Albert Einstein College of Medicine. Rats (n= 3) were firstanesthetized via intraperitoneal injection of sodium pentobarbital(35 mg kg−1; Abbott Laboratory, Chicago, IL). Four aliquots of 50 μlof Alexa Fluor 594 nm–labeled NO-np (5 mgml−1) suspended incoconut oil were applied to the four abdominal quadrants of theabdomen. Coconut oil alone was used as a control (n= 1). After10minutes, quadrants were cleaned with alcohol wipes and 4-mmpunch biopsies were taken from each application site. Specimenswere subjected to frozen sectioning and vertical sections (4-mm-thick) were fixed to glass slides. Images were captured with aconfocal microscope (PerkinElmer, Waltham, MA) with a × 10, 1.4numericl aperture objective and a digital camera (Orca ER,Hamamatsu, Japan).

Colony-forming unit assay and bioscreen analysisP. acnes strain ATCC 6919 was obtained from American TypeCulture Collections and isolates grown in the miniMACS anaerobicworkstation (Don Whitley Scientific, West Yorkshire, UK). Briefly,Brucella agar plates supplemented with blood, hemin, and vitamin K(Remel Fisher Scientific, Lenexa, KS) were streaked with P. acnesovernight and single colonies isolated.The cultures were grown in ananaerobic chamber for 4–5 days. A spectrophotometer OD600 wasused to determine the bacterial log phase.Various concentrations of control and NO-nps (0.625, 1.25, 2.5,

and 5mgml− 1) were incubated with 3.75× 105 bacteria in a finalvolume of 30 ml at 37 °C for 4 hours. After incubation, 10-folddilutions were prepared and incubated for 4 days at 37 °C underanaerobic conditions. Individual colonies were counted and thenumber of colony-forming unit was tabulated.For growth kinetics, reinforced Clostridium medium was inocu-

lated with a fresh colony grown on the Brucella agar plates andsuspended in 1ml of medium. A suspension of 100 μl of bacteria wastransferred to a 200-well plate with 100 μl of reinforced Clostridiummedium per well containing NO-np or np (0.625, 1.25, 2.5, and5mgml−1). Bacteria and nps were incubated for 24 hours at 37 °C.Controls included wells containing bacteria with reinforced Clos-tridium medium alone. Growth was assessed at OD600 every 30minutes using a microplate reader (Bioscreen C, Growth Curves USA,Piscataway, NJ).

P. acnes culture and cell stimulationPBMCs were stimulated by P. acnes strain 6919 at multiplicity ofinfection (MOI) 0.5, whereas HaCaT keratinocytes were infected byP. acnes at MOI 20 (Qin et al., 2014). We used live P. acnes asopposed to sonicated-killed preparations of P. acnes due to thepathophysiological relevance of our model and consistency on theobserved responses.

120* *

100

80

60

40

20

0

100

% m

orta

lity

at 1

20 h

pf 0 p.p.m.

250 p.p.m.

80

60

40

20

0

0

0.01

6

0.08 0.

4

Concentration (p.p.m.)

2 10 50 250

% M

etab

olic

act

ivity

– 1.25 2.5 5.0

120

100

80

60

40

20

0

% M

etab

olic

act

ivity

– 1.25 2.5 5.0No-npmg/ml

Figure 6. Nitric oxide–releasing nanoparticles (NO-nps) are non-toxic. NO-nps at various concentrations (1.25, 2.5, and 5mgml−1, respectively) wereincubated with HaCaT cells (a) or peripheral blood mononuclear cells(PBMCs) (b). Cells were collected and evaluated for viability using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrasodium bromide tetrazolium)assay. (c) Representative images of zebrafish embryos at 120 hpf: control (top)and exposed to NO-np (bottom). No significant differences were observed inlarval morphology or behavioral endpoints (P40.05 for each endpointevaluated, Fisher’s exact test). Error bars denote SEM.

M Qin et al.Efficacy of NO Against P. acnes Inflammation

www.jidonline.org 2729

Isolation of PBMCs, HaCaT cell culture, and cytokine ELISAPBMCs were isolated from whole blood of normal healthy donors asapproved by the Institutional Review Board at UCLA using Ficoll-Paque gradients (GE Healthcare, Piscataway, NJ) and plated onto24-well tissue culture plates (5× 106 per well) in RPMI 1640 mediacontaining 10% fetal bovine serum (HyClone, South Logan, UT).Cells were co-cultured 1 hour with 0, 1.25, 2.5, and 5mgml− 1 ofNO-nps, respectively. Then, cells were infected by P. acnes at MOI0.5 for 24 hours at 37 °C.HaCaT cells, an immortalized human keratinocyte cell line, plated

into six-well tissue culture plates (5× 105 per well) and cultured at37 °C with 5% CO2 in DMEM (Gibco11995-060, Invitrogen,Carlsbad, CA) supplemented with 10% heat-inactivated fetal bovineserum (HyClone) and 1% penicillin/streptomycin (Gibco 15140-122,Invitrogen) for 18 hours. HaCat cells retain their ability to differ-entiate, to express Toll-like receptors involved in pathogen recogni-tion, and to produce a number of cytokines in response to bacterialstimulation in a similar manner to primary human keratinocytes(Grange et al., 2009). HaCaT cells were treated 1 hour with thedifferent doses of NO-nps as above. Then, cells were infected by P.acnes at MOI 20 for 24 hours at 37 °C.IL-6, IL-8, IL-1β, and TNFα levels in culture supernatants were

measured by ELISA assay following the manufacturer’s recommenda-tions (R&D ELISA Development System, Minneapolis, MN). Sampleswere assayed in triplicates. Results are expressed as mean± SD of atleast three independent experiments, with PBMCs obtained fromthree independent donors.

RNA isolation, cDNA synthesis, and real-time PCRFor RNA isolation, cDNA synthesis, and real-time PCR, seeSupplementary Methods online.

siRNA transfectionsiRNA transfection into PBMCs was accomplished by using the AmaxaNucleofector system and the Nucleofector Kit (Lonza, Allendale, NJ)according to the manufacturer’s recommendations, with programU-001 for high viability. siRNA constructs (scramble, siNLRP3, andsiNLRP1) were used at 1 μg per transfection. Transfection efficiency ofexpression constructs was assessed using the pMAX-GFP constructand it yielded an average of 70% transfection rate (Lonza, Allendale,NJ). Transfected cells were plated into six-well cell culture plates andcultured at 37 °C for overnight. Transfected cells were co-cultured withNO-np 2.5mgml−1 for an hour and then infected by P. acnes atMOI 0.5.

Western blot analysisDetection of caspase-1 was performed by western blot analysis.Isolated PBMCs were treated with 2.5 mgml− 1 NO-np or control npfor an hour. Cells were then infected by P. acnes with MOI 0.5. After24 hours, cells were collected and lysed in 100 μl of lysis buffer(M-PER Mammalian Protein Extraction Reagents, Thermo Scientific,Waltham, MA) with protease inhibitors. Protein levels were normal-ized with the Bradford assay. Protein profiles were separated byelectrophoresis in BIO-RAD (Hercules, CA) mini PROTEAN TGXpolyacrylamide precast gels and transferred onto 0.45 μm Millipore(Billerica, MA) Immobilon P-transfer membranes, and caspase-1 wasdetected by caspase-1 monoclonal antibody (D7F10, Cell Signaling,

Danvers, MA). The bands from western blotting were visualized byusing VersaDOC MP5000 (BIO-RAD).

Activity assayIsolated PBMCs were treated with 2.5 mgml− 1 NO-np or control npfor an hour. Cells were then infected by P. acnes with MOI 0.5. After24 hours, cells were collected and lysed with 100 μl of cold lysisbuffer. Caspase-1 protein activity was then measured using theCaspase-1 Colorimetric Assay (R&D Systems, Minneapolis, MN) asper the manufacturer’s instructions and the activity was determinedvia percentage of the control (cells alone). After 16 hours ofincubation at 37 °C, activity was read using BioTek Synergy 2(Winooski, VT) plate reader using a wavelength of 405 nm.

MTT toxicity assayThe MTT assay (Sigma Aldrich) was performed as previouslydescribed (Friedman et al., 2013) (see Supplementary Methodsonline for complete methods).

Statistical analysisGraphPad Prism 5.0 (GraphPad Software, La Jolla, CA) was used forstatistical analyses. P-values were determined by analysis of varianceand adjusted by the use of Bonferroni correction. P-values of o0.05were considered significant

In vivo toxicity assayZebrafish embryos (Danio rerio, wild type, 5D-Tropical strain) wereobtained from Sinnhuber Aquatic Research Laboratory, Oregon StateUniversity, and exposures and evaluations were conducted accordingto Truong et al. (2011). Exposures were conducted over 5 days ofdevelopment. All organ systems begin functioning during this timeperiod and all of the molecular signaling pathways are active andnecessary for normal development to occur. At 120 hpf, behavioralendpoints (motility and tactile response) and larval morphology (bodyaxis, eye, snout, jaw, otic vesicle, notochord, heart, brain, somite, fin,yolk sac, trunk, circulation, pigment, and swim bladder) wereevaluated in vivo and scored in a binary manner (present orabsent). Untreated control and exposed groups were comparedusing Fisher’s exact test for each endpoint and P-value o0.05 forsignificance.

CONFLICT OF INTERESTAJF and JMF are co-inventors of the NO-np, a technology that has beenlicensed to Nano Biomed Inc for commercialization. All other authors state noconflicts of interest.

ACKNOWLEDGMENTSSLH acknowledges NIH grants ES017552-01A2, ES016896-01, P30 ES03850,T32 ES0007060, and AFRL FA8650-05-1-5041. JDN acknowledges theNational Institutes of Health/National Institute of Allergy and InfectiousDiseases grant 1RC2A1087612-01. AJF acknowledges Dermatology Founda-tion Career Development Award. JK acknowledges support from NIAMS R01AR053542.

SUPPLEMENTARY MATERIAL

Supplementary material is linked to the online version of the paper at http://www.nature.com/jid

M Qin et al.Efficacy of NO Against P. acnes Inflammation

2730 Journal of Investigative Dermatology (2015), Volume 135

REFERENCES

Bellew S, Thiboutot D, Del Rosso JQ (2011) Pathogenesis of acne vulgaris:what’s new, what’s interesting and what may be clinically relevant. J DrugsDermatol 10:582–5

Bennett BM, McDonald BJ, Nigam R et al. (1994) Biotransformation of organicnitrates and vascular smooth muscle cell function. Trends Pharmacol Sci15:245–9

Blecher K, Martinez LR, Tuckman-Vernon C et al. (2012) Nitric oxide-releasingnanoparticles accelerate wound healing in NOD-SCID mice. Nanomedi-cine 8:1364–71

Cabrales P, Han G, Roche C et al. (2010) Sustained release nitric oxide fromlong-lived circulating nanoparticles. Free Radic Biol Med 49:530–8

Castro G, Ferreira L (2008) Novel vesicular and particulate drug deliverysystems for topical treatment of acne. Expert Opin Drug Deliv 5:665–79

Dal Secco D, Paron JA, de Oliveira SH et al. (2003) Neutrophil migration ininflammation: nitric oxide inhibits rolling, adhesion and induces apoptosis.Nitric Oxide 9:153–64

Dostert C, Petrilli V, Van Bruggen R et al. (2008) Innate immune activationthrough Nalp3 inflammasome sensing of asbestos and silica. Science 320:674–7

Eady EA (1998) Bacterial resistance in acne. Dermatology 196:59–66

Fang FC (2011) Antimicrobial actions of reactive oxygen species. mBio 2:e00141–11

Fang FC (1997) Mechanisms of nitric oxide –reated antimicrobial activities.J Clin Invest 99:2818–25

Friedman A, Blecher K, Sanchez D et al. (2011) Susceptibility of Gram-positiveand -negative bacteria to novel nitric oxide-releasing nanoparticletechnology. Virulence 2:217–21

Friedman A, Friedman J (2009) New biomaterials for the sustained release ofnitric oxide: past, present and future. Expert Opin Drug Deliv 6:1113–22

Friedman A, Han G, Navati M et al. (2008) Sustained release nitric oxidenanoparticles: characterization of a novel delivery platform based on nitritecontaining hydrogel/glass composites. Nitric Oxide 19:12–20

Friedman A, Phan J, Schairer D et al. (2013) Antimicrobial and anti-inflammatory activity of chitosan-alginate nanoparticles: a targeted therapyfor cutaneous pathogens. J Invest Dermatol 133:1231–9

Friedman AJ, Blecher K, Schairer D et al. (2011) Improved antimicrobial efficacywith nitric oxide releasing nanoparticle generated S-nitrosoglutathione.Nitric Oxide 25:381–6

Giustizieri ML, Albanesi C, Scarponi C et al. (2002) Nitric oxide donors suppresschemokine production by keratinocytes in vitro and in vivo. Am J Pathol161:1409–18

Gori T, Parker JD (2002b) The puzzle of nitrate tolerance: pieces smaller thanwe thought? Circulation 106:2404–8

Grange PA, Raingeaud J, Calvez V et al. (2009) Nicotinamide inhibitsPropionibacterium acnes-induced IL-8 production in keratinocytes throughthe NF-kappaB and MAPK pathways. J Dermatol Sci 56:106–12

Griffith OW, Stuehr DJ (1995) Nitric oxide synthases: properties and catalyticmechanism. Annu Rev Physiol 57:707–36

Han G, Friedman AJ, Friedman JM (2011) Nitric oxide releasing nanoparticlesynthesis and characterization. Methods Mol Biol 704:187–95

Hernandez-Cuellar E, Tsuchiya K, Hara H et al. (2012) Cutting edge: nitric oxideinhibits the NLRP3 inflammasome. J Immunol 189:5113–7

Jung S, Otberg N, Thiede G et al. (2006) Innovative liposomes as atransfollicular drug delivery system: penetration into porcine hair follicles.J Invest Dermatol 126:1728–32

Manconi M, Sinico C, Valenti D et al. (2006) Niosomes as carriers for tretinoin.III. A study into the in vitro cutaneous delivery of vesicle-incorporatedtretinoin. Int J Pharm 311:11–9

Manconi M, Sinico C, Valenti D et al. (2002) Niosomes as carriers for tretinoin.I. Preparation and properties. Int J Pharm 234:237–48

Mao K, Chen S, Chen M et al. (2013) Nitric oxide suppresses NLRP3inflammasome activation and protects against LPS-induced septic shock.Cell Res 23:201–12

Marshall HE, Hess DT, Stamler JS (2004) S-nitrosylation: physiologicalregulation of NF-kappaB. Proc Natl Acad Sci USA 101:8841–2

Martinez LR, Han G, Chacko M et al. (2009) Antimicrobial and healing efficacyof sustained release nitric oxide nanoparticles against Staphylococcusaureus skin infection. J Invest Dermatol 129:2463–9

Nakatsuji T, Kao MC, Fang JY et al. (2009) Antimicrobial property of lauric acidagainst Propionibacterium acnes: its therapeutic potential for inflammatoryacne vulgaris. J Invest Dermatol 129:2480–8

Privett BJ, Broadnax AD, Bauman SJ et al. (2012) Examination of bacterialresistance to exogenous nitric oxide. Nitric Oxide 26:169–73

Qin M, Pirouz A, Kim MH et al. (2014) Propionibacterium acnes inducesIL-1beta secretion via the NLRP3 inflammasome in human monocytes. JInvest Dermatol 134:381–8

Ross JI, Snelling AM, Carnegie E et al. (2003) Antibiotic-resistant acne: lessonsfrom Europe. Br J Dermatol 148:467–78

Schroder K, Tschopp J (2010) The inflammasomes. Cell 140:821–32

Stamler JS, Lamas S, Fang FC (2001) Nitrosylation: the prototypic redox-basedsignaling mechanism. Cell 106:675–83

Trivedi NR, Gilliland KL, Zhao W et al. (2006) Gene array expression profilingin acne lesions reveals marked upregulation of genes involved ininflammation and matrix remodeling. J Invest Dermatol 126:1071–9

Truong L, Harper SL, Tanguay RL (2011) Evaluation of embryotoxicity using thezebrafish model. In: Drug Safety Evaluation: Methods and ProtocolsGautier J-C (ed). Vol. 691. Humana Press: New York, NY, pp 271–9

Weller RB (2009) Nitric oxide-containing nanoparticles as an antimicrobialagent and enhancer of wound healing. J Invest Dermatol 129:2335–7

Wink DA, Mitchell JB (1998) Chemical biology of nitric oxide: insights intoregulatory, cytotoxic, and cytoprotective mechanisms of nitric oxide. FreeRadic Biol Med 25:434–56

Yang JK, WuWJ, Qi J et al. (2014) TNF-308 G/A polymorphism and risk of acnevulgaris: a meta-analysis. PLoS ONE 9:e87806

Zouboulis C, Eady A, Philpott M et al. (2005) What is the pathogenesis of acne?Exp Dermatol 14:143–52

M Qin et al.Efficacy of NO Against P. acnes Inflammation

www.jidonline.org 2731