MUNI | RECETOX
SCI
MASS SPECTROMETRY BASED METABOLOMICS AND LIPIDOMICS
Darshak Gadara Supervisor - Dr. Zdenek Spacil Targeted Proteomics and Metabolomics Group RECETOX, MUNI, Brno
i
Flow of Presentation
• Systematic Data Processing of Untarqeted Metabolomics
• Application - Case/Control Fecal Sample of CVID
• High-Throughput Microbore LC-MS Lipidomics
• Application - APOE3 vs APOE 4 of iPSC Cerebral Organoid
• Application - 3D-Hepatospheroids to Study Hazardous Compounds
Why Metabolomics?
Phenotype (Metabolome)
• Metabolites are the endpoint of biological reactions
• Constantly synthesized, degraded, and interacting within the system and with the environment
• Represents dynamic relation between genotype and environment
• Extremely important in clinical diagnosis
Challenge of Data Processing
LC-MS (high resolution)
>85% Noise
m/z 100
m/z 1000
LC-MS Spectra
Data Processing
Full Scan Mode
Unbiased Measurement
Contaminants +
Background +
Metabolites
Major Challenge Thousands of peaks
No Consensus
Metabolomics Data Processing Workflow
Peak Peaking
Noise Filtering     Quanti. Filtration     Database Annotation
Blank
Contaminats Background -
Fecal Extract
Metabolites
Gadara, D..Systematic Feature Filtering in Exploratory Metabolomics: Application toward Biomarker Discovery. Anal. Chem. 2021, 93(26), 9103-9110. https://doi.Org/10.1021 /ACS.ANALCHEM.1 C00816.
5
RT vs ClogP Relationship
1224 Features Rl^C!°9P  839 Features Filtration
10
CUD 5
_o
U
s 0
Q. O
-10
Poor Retention
o o $Q>
8   °o °
_I_
12
Retention time (min)
Gadara, D..Systematic Feature Filtering in Exploratory Metabolomics: Application toward Biomarker Discovery. Anal. Chem. 2021, 93(26), 9103-9110. https://doi.Org/10.1021/ACS.ANALCHEM.1 C00816.
6
RTvsClogP Filtration on other Datasets
1 0
cl 5
o
ST00092 (79 Plasma Sample)
Poor Prediction]
O Features
O Identifications
Linear Regre. line 90% Pis line (both) O
is
752 Features
CL
o
.o
TO
447 Features
8       Retention Time [min]
10
ST001362 (78 Fecal Sample)
8
O Features Oldentifications ' O
Linear Regre. line O
90% Pis line (both)    O O
736 Features
O}
o
.o
CX3
447 Features
Retention Time [min] 20
Metabolomics Data Processing Workflow
RT Filtering Statistical Analysis      Structure Elucida. Interpretation
8
Application in Clinical Case/Control sample
Common Variable Immunodeficiency Disease (CVID) - inadequate antibody responses and low levels of immunoglobulins
Case/Control Fecal sample (n=6) subjected to developed metabolomlcs data processing workflow
Peak Peaking
Noise threshold-10,000 6940 Features
Biological Interpretation
„Puinrie Meta.
Tr^y-p^N&n^eta. Pathway Mapping
Blank Subtraction
Contaminants Background
Filtered if QC/Blank <3
4180 Features
MS2 & Manual Inspection
Peak Shape, l> = Fragment match 8 Features
Quantitative Filtration
x
Ulk
Injections
RSD in QC < 25
3205 Features
Statistical Analysis
FC>1.5 or <0.67, p-value <0,05 97 Features
Database Search MMD3
Error <3 ppm 1224 Features
Chemistry Filtration
RT
90% Prediction Interval
839 Features
9
Interpretation of metabolomics data for CVID vs healthy subject
Eight metabolites differ significantly between CVID vs Healthy subject (p<0.05)
Controls
rM C r- If! D 01
toooooo r Q ÜQÜÜÜ
0. Q.Q.Q.Q.O. M •
L Samples
ID Ol (N 5 ID CO
OOQOOQ □□□000
Metabolite (Pathway)
Adenosine (Purine metabolism) Inosine (Purine metabolism)
D-Glucosamine (Amino- and nucleotide sugar metabolism) Glycocholic acid (Bile acid biosynthesis) Glycoursodeoxycholic acid (Bile acid biosynthesis) Pyridoxine (Vitamin B6 metabolism) 4-Aminobenzenesulfonate (Aminobeizoate degradation) 8-Hydroxyguanine (Oxidative damage product)
0.0.0.0.0.0.
AAA AA4i
Household -I
In collaboration with
Prof. David Smajs group, Faculty of medicine
Purine metabolites are important anti- inflammatory signal molecules modulating inflammatory response
Welihinda AA et al. Cell Signal (2016) 28:552-60 He B et al. J Exp Med (2017) 214:107-23
A decreased level of "purine-like" molecules was previously detected in blood of CVID patients
Callery EL et, al, Sci Rep (2019) 9:e7239)
Dysregulation of bile acid metabolism related to bacterial dysbiosis has been reported for several conditions
Zheng X et al. BMC Biol. 2017 Dec 14; 15(1): 120 Wang Yetal. mSystems. 2019 Dec 17; 4(6)
Decreased plasma levels of pyridoxine (B6) observed CVID patients
Bierwirth J et al. Eur J Clin Nutr (2008) 62:332-5
10
Summary of Metabolomics
We developed a systematic metabolomics data processing workflow for clinical biomarker discovery, explaining individual feature filtering stages
Identified eight metabolites in stool samples which clearly distinguished between CVID patients and controls
RETURN TO ISSUE
< PREV   ARTICLE    NEXT >
analytical. A
'chemistry
Systematic Feature Filtering in Exploratory Metabi Application toward Biomarker Discovery
Darshak Gadara, Katerina Coufalikova, Juraj Bosak, David Smajs, and Zdenek
0 Cite this: Anal. Chem. 2021,93,26,9103-9110 Publication Date: June 22,2021 v https://doi.Org/10.1021 /acs.analchem.1 c00816 Copyright © 2021 American Chemical Society rights & permissions ✓'Subscribed
Article Views
Altmetric
Citations
1085 2
ORIGINAL RESEARCH article
Front Immunol., 14 May 20211 https://doi.org/10.3389/fimmu.2021.671239
learn about these metrics
Patients With Common Variable Immunodeficiency (CVID) Show Higher Gut Bacterial Diversity and Levels of Low-Abundance Genes Than the Healthy Housemates
Juraj Bosák1,      Matěj Lexa2, _;\ Kristýna Fiedorová34,      Darshak C. Gadara5, Lenka Micenkováa,     Zdenek Spacil5,     Jiří Litzman45,     Tomáš Freiberger34 and      David Šmajs1"
department of Biology, Faculty of Medicine, Masaryk University, Brno, Czechia
2Faculty of Informatics, Masaryk University, Brno, Czechia
^Centre for Cardiovascular Surgery and Transplantation, Brno, Czechia
4Department of Clinical Immunology and Allergology, Faculty of Medicine, Masaryk University, Brno, Czechia 5RECETQX Center, Faculty of Science, Masaryk University, Brno, Czechia_
li
Flow of Presentation
• Systematic Data Processing of Untarqeted Metabolomics
• Application - Case/Control Fecal Sample of CVID
• High-Throughput Microbore LC-MS Lipidomics
• Application - APOE3 vs APOE 4 of iPSC Cerebral Organoid
• Application - 3D-Hepatospheroids to Study Hazardous Compounds
LC Gradient and Flow Optimization
Microbore Column 1 mm I.D.
nanoViper Tube
		
		i
		
Four flow rate - 50, 75,100,120 u L/min Three different gradient length -11,22, 33 min
75
Flow rate (|il/min)
UHPLC System
50       75      100
Flow rate (nl/min)
150
13
Single Cerebral Organoid Lipidomics
Reverse Phase Separation
~ 350 lipid species from single crebral organoid
1200
900
i 600
300-
T-r
5 10 Retention time (Min)
Hex3Cer 3 Sph 3
LPC
14
AP0E3 vs AP0E4 Lipidmics
Cerebral Organoids APOE Phenotypes (n=10)
E3
E4
Q Q
Tramiprosate
-1 0 1 PC 1 (68.8%)
_^
CO >
Q.
CD O
2^.
APOE E4 vs E3
OCE
#HexCer Hex2Cer*
PC-
PC-Hex3Cer <
In collaboration with DG - Loschmidt Laboratories
- Faculty of medicine
Cholesterol PE-O LPE
O      O    OO J
c2°PS  LPC LPC-0
dhCer
_SM__
PC-O
• TG
g-ya\ue_<_0.05
IIpi
• PG
Car
-0.5
0.0
log2(FC)
0.5
15
Single 3D-Hepatospheroid Lipidomics
, 200+ lipid species i Charatcerized
Control Spheroid Spheroid     + 1 rj|jM Amiodarone
In collaboration with - Marina Grossi Pavel Babica's group
Control Spheroid
10p M Amiodarone Treated Spheroid
°
o
1
°
o
°
b-d
1»
Wo
H 1
T
i
T
& ^     ^ ^ JP ^ <f> fp #>
V
16
Other Hazardous Chemicals?
Cynotoxins _CYN
Flam Retardant - EHDPP
Several chemicals
Scores Plot
1 U.M
o o o
Control
o GYN SC H20
o o
o o o
o o
im l3
Q_
Solwnt Control   ■= LjiM EHDPP    LOjiM EHDPP
Control
1 U.M
o1
° %
o o
o o
10 U.M
-i
0 1 PCI (55.6%)
TM3&TM4 Cell lines (2D)
10      11 12
)0
€)•#•
^(t^ ^(b^
Oí
Data Visulization is Ongoing
• Riju Roy Chowdhury Dr. Pavel Babica's group
■ Ghanaer Negi Prof. Ludek Bláha's group
Eliška Sychrova
Dr. Iva Sovadinova's group
17
Summary of Lipidomics
• |j LC-MS/MS lipidomics workflow allows sensitive, high-throughput and robust measurement of lipidome from small volume in vitro samples such as single hepatospheroid or single cerebral organoid.
• Application demonstrated on 5 different projects (1 first author + several co-authors manuscript under preparation)
• This work paves the way for a more routine application of \i LC-MS/MS lipidomics in high-throughput in vitro toxicity screening.
18
ACKNOWLEDGEMENT
• Sincerely Thankful to
• Supervisor-Dr.Zdenek Spacil
• Lab-members
• Collaborators
Thank you all for listening