Abstract Acute respiratory distress syndrome is a lethal condition of acute bilateral lung disease associated with trauma, sepsis, and shock that occurs as a result of fluid build up in the alveoli. Prior work in the field suggests that ARDS and its pathophysiology may be mediated by epigenetically controlled factors, such as DNA methylation, and that the syndrome itself may incur genome-wide DNA methylation changes. In this investigation, we compare DNA methylation data derived from blood collected from 39 adults with ARDS, 75 ICU controls, and 30 healthy individuals made available by Szilagyi et al (1). Our analysis reveals significant demethylation in the promoter regions of 7 histone deacetylase, suggesting up-regulation of HDAC transcription. These results indicate that patients with ARDS have increased HDAC expression, potentially leading to closed chromatin structure and gene repression, suggesting a potential mechanism for increased HDAC expression in patients with ARDS.

Background

Materials and Methods

Results

Conclusion

1.StudyDesignandSamples

Pa#entMetric PercentageofSampleSize(numberofsamples)

ARDS 27.1%(39)

ICU(NoARDS) 52.1%(75)

Healthy 20.8%(30)

Male 54.9%(79)

Female 45.1%(65)

Black 50.0%(72)

Age>58yearsold 49.0%(71)

Age<58yearsold 51.0%(73)

•  A nested case–control design was used to select whole blood DNA

samples from 2 Chicago-based ARDS cohorts; data provided by Szilagyi et al. (Accession: GSE67530)

•  Exclusion Criteria: Hispanic heritage, diagnosis of cancer, history of organ transplant, or concurrent drug overdose to exclude known confounders of cytosine modification

•  Blood DNA was collected from healthy and sick patients, with and without ARDS

•  Available data for all patients included age, sex, and ethnicity which was used to inform linear modeling

•  DNA was treated with bisulfite, converting unmethylated cytosine bases to uracil

•  DNA methylation was measured using the Infinium 450k Methylation Array

•  Microchip targets > 450,000 CpG sites, measuring methylation via two color (type I probes) and one color (type II probes) read out

•  Intensity values for each color convey allelic ratio of a given locus •  The raw data contains summary intensities for each probe assayed in

the array; values must be converted to quantify methylation state •  Szilagyi et al published data via NCBIs GEO data base (GSE67530)

2.Infinium450kArray

3.NormalizaRonandModeling

4.GOAnalysis

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Beta

Density

Raw

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Beta

Density

Normalized

−0.1 0.0 0.1 0.2 0.3

−0.0

50.

000.

050.

10

Principal Component 1

Prin

cipa

l Com

pone

nt 2

GSM1648896

GSM1648897

GSM1648898

GSM1648899

GSM1648900

GSM1648901

GSM1648902

GSM1648903

GSM1648904

GSM1648905

GSM1648906

GSM1648907

GSM1648908

GSM1648909

GSM1648910

GSM1648911

GSM1648912

GSM1648913

GSM1648914

GSM1648915

GSM1648916

GSM1648917

GSM1648918GSM1648919

GSM1648920

GSM1648921

GSM1648922

GSM1648923

GSM1648924

GSM1648925

GSM1648926

GSM1648927

GSM1648928

GSM1648929

GSM1648930

GSM1648931

GSM1648932

GSM1648933

GSM1648934

GSM1648935GSM1648936

GSM1648937

GSM1648938

GSM1648939

GSM1648940

GSM1648941GSM1648942

GSM1648943

GSM1648944

GSM1648945

GSM1648946

GSM1648947

GSM1648948

GSM1648949GSM1648950

GSM1648951

GSM1648952

GSM1648953

GSM1648954

GSM1648955GSM1648956

GSM1648957

GSM1648958

GSM1648959

GSM1648960

GSM1648961

GSM1648962

GSM1648963

GSM1648964

GSM1648965

GSM1648966

GSM1648967

GSM1648968

GSM1648969

GSM1648970

GSM1648971

GSM1648972

GSM1648973

GSM1648974

GSM1648975

GSM1648976

GSM1648977GSM1648978

GSM1648979

GSM1648980

GSM1648981

GSM1648982

GSM1648983

GSM1648984

GSM1648985

GSM1648986

GSM1648987

GSM1648988

GSM1648989

GSM1648990

GSM1648991

GSM1648992

GSM1648993GSM1648994

GSM1648995

GSM1648996

GSM1648997

GSM1648998

GSM1648999

GSM1649000

GSM1649001

GSM1649002

GSM1649003

GSM1649004

GSM1649005GSM1649006

GSM1649007

GSM1649008

GSM1649009

GSM1649010

GSM1649011

GSM1649012

GSM1649013

GSM1649014GSM1649015

GSM1649016

GSM1649017

GSM1649018

GSM1649019

GSM1649020

GSM1649021

GSM1649022

GSM1649023

GSM1649024

GSM1649025

GSM1649026

GSM1649027

GSM1649028

GSM1649029GSM1649030

GSM1649031

GSM1649032

GSM1649033

GSM1649034

GSM1649035

GSM1649036

GSM1649037

GSM1649038

GSM1649039

Female

Male

−0.1 0.0 0.1 0.2 0.3

−0.0

4−0

.02

0.00

0.02

0.04

0.06

Principal Component 1

Prin

cipa

l Com

pone

nt 3

GSM1648896

GSM1648897

GSM1648898

GSM1648899

GSM1648900

GSM1648901

GSM1648902

GSM1648903

GSM1648904

GSM1648905

GSM1648906

GSM1648907

GSM1648908

GSM1648909

GSM1648910

GSM1648911

GSM1648912

GSM1648913

GSM1648914

GSM1648915

GSM1648916

GSM1648917

GSM1648918

GSM1648919

GSM1648920

GSM1648921

GSM1648922

GSM1648923

GSM1648924

GSM1648925

GSM1648926

GSM1648927

GSM1648928

GSM1648929

GSM1648930

GSM1648931

GSM1648932

GSM1648933

GSM1648934

GSM1648935

GSM1648936

GSM1648937

GSM1648938

GSM1648939

GSM1648940

GSM1648941

GSM1648942

GSM1648943

GSM1648944

GSM1648945

GSM1648946

GSM1648947

GSM1648948

GSM1648949

GSM1648950

GSM1648951

GSM1648952

GSM1648953GSM1648954

GSM1648955

GSM1648956

GSM1648957

GSM1648958

GSM1648959

GSM1648960

GSM1648961

GSM1648962

GSM1648963

GSM1648964

GSM1648965

GSM1648966

GSM1648967GSM1648968

GSM1648969

GSM1648970

GSM1648971

GSM1648972

GSM1648973

GSM1648974

GSM1648975

GSM1648976

GSM1648977

GSM1648978

GSM1648979

GSM1648980

GSM1648981

GSM1648982

GSM1648983

GSM1648984

GSM1648985

GSM1648986

GSM1648987

GSM1648988

GSM1648989

GSM1648990GSM1648991

GSM1648992

GSM1648993

GSM1648994

GSM1648995

GSM1648996

GSM1648997

GSM1648998

GSM1648999

GSM1649000

GSM1649001GSM1649002

GSM1649003

GSM1649004

GSM1649005

GSM1649006

GSM1649007

GSM1649008

GSM1649009

GSM1649010

GSM1649011

GSM1649012

GSM1649013

GSM1649014

GSM1649015

GSM1649016

GSM1649017

GSM1649018

GSM1649019

GSM1649020GSM1649021

GSM1649022

GSM1649023

GSM1649024

GSM1649025

GSM1649026

GSM1649027

GSM1649028

GSM1649029

GSM1649030

GSM1649031

GSM1649032

GSM1649033

GSM1649034

GSM1649035

GSM1649036

GSM1649037

GSM1649038

GSM1649039

Black

White

−0.1 0.0 0.1 0.2 0.3

−0.0

8−0

.06

−0.0

4−0

.02

0.00

0.02

0.04

Principal Component 1

Prin

cipa

l Com

pone

nt 4

GSM1648896

GSM1648897

GSM1648898

GSM1648899

GSM1648900

GSM1648901

GSM1648902

GSM1648903

GSM1648904

GSM1648905

GSM1648906

GSM1648907

GSM1648908

GSM1648909

GSM1648910

GSM1648911

GSM1648912GSM1648913

GSM1648914GSM1648915

GSM1648916

GSM1648917

GSM1648918

GSM1648919

GSM1648920

GSM1648921

GSM1648922

GSM1648923

GSM1648924

GSM1648925

GSM1648926

GSM1648927

GSM1648928

GSM1648929

GSM1648930

GSM1648931GSM1648932

GSM1648933

GSM1648934

GSM1648935

GSM1648936 GSM1648937

GSM1648938

GSM1648939

GSM1648940GSM1648941

GSM1648942

GSM1648943

GSM1648944

GSM1648945

GSM1648946

GSM1648947

GSM1648948

GSM1648949

GSM1648950

GSM1648951

GSM1648952

GSM1648953

GSM1648954

GSM1648955

GSM1648956

GSM1648957GSM1648958

GSM1648959

GSM1648960

GSM1648961

GSM1648962

GSM1648963

GSM1648964

GSM1648965GSM1648966

GSM1648967

GSM1648968

GSM1648969

GSM1648970

GSM1648971

GSM1648972

GSM1648973

GSM1648974

GSM1648975

GSM1648976

GSM1648977

GSM1648978

GSM1648979

GSM1648980

GSM1648981

GSM1648982

GSM1648983

GSM1648984

GSM1648985

GSM1648986

GSM1648987

GSM1648988

GSM1648989

GSM1648990

GSM1648991

GSM1648992

GSM1648993

GSM1648994

GSM1648995

GSM1648996

GSM1648997

GSM1648998

GSM1648999

GSM1649000GSM1649001

GSM1649002

GSM1649003

GSM1649004

GSM1649005

GSM1649006

GSM1649007

GSM1649008

GSM1649009

GSM1649010

GSM1649011

GSM1649012

GSM1649013

GSM1649014

GSM1649015

GSM1649016GSM1649017

GSM1649018

GSM1649019

GSM1649020

GSM1649021

GSM1649022

GSM1649023

GSM1649024

GSM1649025

GSM1649026GSM1649027

GSM1649028

GSM1649029

GSM1649030

GSM1649031

GSM1649032

GSM1649033

GSM1649034

GSM1649035GSM1649036

GSM1649037

GSM1649038

GSM1649039

0

1

2

Figure3:DistribuRonsforRawandNormalizedbetavalues.

Histone deacetylases (HDACs) are known to regulate gene expression. In general, HDAC inhibitors cause an overall increase in gene expression. HDAC inhibitors repress expression of inflammatory cytokines (5) and have been shown to aZenuatelipopolysaccharide-inducedacutelunginjuryinmice(6)

ARDS is a lethal condition of acute bilateral lung disease affecting 5% of adult patients on ventilators with mortality estimated at 40% Previous studies have demonstrated that ARDS pathophysiology is likely mediated, in part, by epigenetic mechanisms (2-4)

•  178,918CpGswerefoundtohavesignificantdifferenRalmethylaRon(q-value<0.05)

•  18,388oftheseprobesfellwithinaCpGislandthatwasalsointhepromoterregionofaknowngene,referredtoasisland-promoter(IP)CpGs.

•  GOanalysisrevealedsignificantenrichmentinthe“RepressingtranscripRonfactorbinding”GOfamily(Table2)

•  CrossvalidaRnggenesunderthisGOtermwithourlistofIPCpGsrevealedmanyHDACgenestobesignificantlyalteredinpaRentswithARDS(Table3)

•  Demethylation of CpG islands in the promoter regions of HDACs provide a potential mechanism for the increase in HDAC expression seen in patients with ARDS (4).

•  Results also support continued investigation into the application of HDAC inhibitors for treating ARDS (6)

Acknowledgements This project was made possible by Princeton’s Center for Health and Well Being, Office of the Dean of College, MOL/QCB SURP, and Illumina Adult sample data provided by Szilagyi et al. through NCBI GEO database (Accession: GSE67530) Notterman Lab

References 1.  Szilagyi K, Garcia JGN, Zhang W (2013) Exploring DNA Methylation of MYLK

as a Contributor to Acute Respiratory Distress Syndrome Disparities. J Pulm Respir Med 3:e127. doi:10.4172/2161-105X.1000e127

2.  Zhang, X., Lv, C., Liu, X., Hao, D., Qin, J., Tian, H., . . . Wang, X. (2013). Genome‑wide analysis of DNA methylation in rat lungs with lipopolysaccharide‑induced acute lung injury. Spandidos publications: Molecular Medicine Reports, 7(15), 1417-1424. doi:10.3892/mmr.2013.140

3.  Elangovan, V. R., Camp, S. M., Kelly, G. T., Desai, A. A., Adyshev, D., Sun, X., … Garcia, J. G. N. (2016). Endotoxin- and mechanical stress–induced epigenetic changes in the regulation of the nicotinamide phosphoribosyltransferase promoter. Pulmonary Circulation, 6(4), 539–544. http://doi.org/10.1086/688761

4.  El Gazzar M, Yoza BK, Hu JY, Cousart SL and McCall CE: Epigenetic silencing of tumor necrosis factor alpha during endotoxin tolerance. J Biol Chem. 282:26857–26864. 2007.

5.  The antitumor histone deacetylase inhibitor suberoylanilide hydroxamic acid exhibits antiinflammatory properties via suppression of cytokines. Leoni F, Zaliani A, Bertolini G, Porro G, Pagani P, Pozzi P, Donà G, Fossati G, Sozzani S, Azam T, Bufler P, Fantuzzi G, Goncharov I, Kim SH, Pomerantz BJ, Reznikov LL, Siegmund B, Dinarello CA, Mascagni P Proc Natl Acad Sci U S A. 2002 Mar 5; 99(5):2995-3000.

6.  Ni, Yun-Feng et al. “Histone Deacetylase Inhibitor, Butyrate, Attenuates Lipopolysaccharide-Induced Acute Lung Injury in Mice.” Respiratory Research 11.1 (2010): 33. PMC. Web. 3 Aug. 2017.

Next Steps

Methylation Data from Case-Control Study Reveals Potential Mechanism for HDAC Up-Regulation in ARDS Sam Chiacchia, Lisa Schneper, Daniel Notterman Molecular Biology Department, Princeton University, Princeton NJ, USA

•  MethylaRonpaZernsfromARDSpaRentsandICUcontrolsclustertogetherwhilehealthypaRentsclusterontheirown(Fig4)

•  HealthypaRentstendtohavemoremethylatedCpGsthansickpaRents(Fig4)

Figure4:MDSplotsdepicRngrelaRonshipbetweendifferentprincipalcomponents

Table1:SampledemographicinformaRon

•  Significant CpGs that fell within CpG islands were mapped to the genome, those that fell within the promoter regions of genes were used to conduct GO analysis

•  Analysis was conducted using the online GO Enrichment Analysis tool (http://www.geneontology.org/page/go-enrichment-analysis)

•  Raw data was normalized using quantile normalization where methylation values are transformed to maintain a common distribution of intensities (Fig 3)

•  In order to assess what types of variables may confound methylation, we used multidimensional scaling (MDS) to visualize the degree of similarity between individuals in various patient groups. MDS converts a set of possibly correlated variables into a set of linearly uncorrelated variables called principal components.

•  MDS confirms previously understood correlations between DNA methylation, sex, and ethnicity (Fig 4)

Figure1:Cartooncomparingahealthyalveolus(led)toonewithARDS(right).DepictsdeathoftypeIIepithelialcellsinresponsetoproteinrichedemafluidcausedbythebody’simmuneresponsetoaprimarymorbidityPhotocitaRon:hZp://www.daviddarling.info/encyclopedia/A/acute_respiratory_distress_syndrome.html

Figure4:HeatmapdepicRngbetavaluesforIPCpGsassociatedwithHDACs.ThecolumnsrepresentindividualpaRents,whiletherowsrepresentCpGloci.OrangerepresentspaRentswithARDS,purplerepresentsICUcontrols,andgreyrepresentshealthypaRents.Themoregreenthemoremethylatedtheloci.ClusteringofpaRentsandlociwasdoneusinghclustalgorithm.

ColumnColor Pa#entHealthStatus

Orange ARDS

Purple ICUControl

Grey Healthy

Table3:NumberofIPCpGsbyHDAC

Table2:GOEnrichmentAnalysisResultsGOmolecularfunc#on

FoldEnrichment P-value

RepressingtranscripRonfactorbinding 7.95 2.54E-03RNAbinding 1.77 2.91E-02DNAbinding 1.73 3.48E-04Nucleicacidbinding 1.69 3.84E-08Heterocycliccompoundbinding 1.55 3.73E-09Organiccycliccompoundbinding 1.54 7.60E-09

Gene SignificantIPCpGsHDAC4 14HDAC5 5HDAC11 4HDAC1 3HDAC10 3HDAC7 3HDAC2 2

•  Pediatric acute respiratory distress syndrome (PARDS) is similar to adult ARDS in pathophysiology, incidence, and mortality.

•  Given the growing evidence that ARDS is mediated by epigenetic factors, we are interested in investigating similar phenomena in a pediatric cohort since genome wide methylation is known to vary with age

•  Through collaboration with Dr. Peter Mourani of Denver Children’s Hospital and Dr. Judie Howrylak of Pennsylvania State College of Medicine, we have designed a longitudinal case-control study designed to investigate variable DNA methylation in children with ARDS.

AcuteStress

DemethylaRonofCpGsinHDACPromoters

HDACexpression

InflammatoryResponse

HDACInhibitor

Figure 5: Proposed model for increased HDAC expression

Totalsamplesizeis144paRents

FemaleMale

BlackNotBlack

ICUControlARDSHealthy