2. Materials and methods
Monitoring of CCHF diagnosed patients and collection of the
samples from the patients: Patients who were diagnosed with CCHF were
selected from Çorum Hitit University Hospital. The first sample
collection was conducted in April 2022, and blood samples were collected
daily using Ethylenediaminetetraacetic acid (EDTA) tubes. In current
metabolomics studies, opinions regarding the appropriate determination
of sample size can considerably vary. However, numerous statistical
analyses have underscored that a substantial sample size, for achieving
meaningful results, around 30 samples (14). In this study, serum
specimens were procured from 29 patients (n=29) diagnosed with CCHFV
infection, as well as from 10 healthy control group. CCHF patients also
admitted to the hospital with an infection diagnosis were subclassified
into two categories based on their blood test values and symptoms:
moderate (n=24) and severe (n=5). Four blood serum samples from patients
each consecutive day (n=116 samples) and a single sample from the
control group (n=10 samples) was taken. These blood samples were then
subjected to centrifugation for 5 minutes at 3000g to separate the
sample into plasma, white blood cell, and red blood cell phases. The
plasma phase was extracted and subjected to metabolite extraction using
multiple approaches: single methanol extraction, triple alcohol
extraction, Methanol-chloroform, Acetone, Acetonitrile, and
Ultrafiltration. Cold methanol-chloroform was chosen as the most
effective extraction and used for extraction of all samples. Only the
polar metabolites in the plasma were investigated, while proteins and
apolar compounds were removed from the plasma samples. Overall in this
study, we utilized a rigorous approach to sample collection,
preparation, and analysis to investigate polar metabolites in the plasma
of CCHF patients (15, 16, 17).
Ethical Statement: The study was approved by Koç University
Ethics Committee and all the procedures performed in this study
involving human participants were in accordance with the ethical
standards of the institutional research committee ethical standards. A
written informed consent was obtained from all patients.
NMR sample preparation: 4 mL ice-cold methanol-chloroform (1:1)
mixture was added to 2 mL serum for methanol-chloroform extraction. The
mixture was vortexed for 30 seconds and incubated for 10 minutes on ice.
After incubation the mixture was centrifuged at 4500g at 4°C for 30
minutes. The methanol phase was collected and dried using a vacuum
concentrator. The dried samples were dissolved in 550 uL
D2O based NMR sample solution (50 mM PBS (pH 7.4), 20mM
NaCl, 1 mM DSS) for standardized sample preparation.
NMR data collection, processing and statistical analysis: 500
MHz Bruker Ascend magnet with BBO paired resonance probe and Avance NEO
console was used for NMR data collection. 1D NOESY-presat (noesygppr1d)
pulse sequence was used for data collection. Each NMR data spectrum is
composed of 4K screening and 32K complex data points. Spectrum widths
were set to 9615.4 Hz. Bruker Topspin 4.2.0 software was used for NMR
data processing. Data was divided into 0.02 ppm data packages along with
their normalization coefficients. The dataset, which comprises data
packets with a resolution of 0.02 ppm, was analyzed using the
MetaboAnalyst 5.0 online metabolomics statistical analysis software.
Henceforth the data will be referred to as (Bin.x.xx [ppm]) data
packets and the day of when sample was taken. All data points were
normalized using the average centering normalization method. Following
this normalization, the dataset underwent statistical analysis using
Partial Least Squares Discriminant Analysis (PLS-DA). PLS-DA is a
classification and discrimination technique based on the Partial Least
Squares (PLS) regression method. This method is widely utilized to
determine the differences between classes, particularly in
high-dimensional and multivariate datasets. PLS-DA is a commonly
employed method in analyzing complex biological systems, such as
metabolomics studies. The VIP Projection variable importance score plot
is a graph that is utilized to assess the outcomes of PLS-DA and
determine the most significant variables in the analysis. VIP scores
quantify the importance of each variable (e.g., metabolites) in
classification and aid in identifying the most critical features. VIP
scores are computed based on the contribution of each variable to the
components (latent variables) in the PLS-DA model. The values begin at
1, and higher VIP scores indicate that the variable is more important
for classification. Variables with VIP scores greater than 1 are
generally deemed significant, although this threshold may vary in
practice. The VIP score plot displays the VIP scores of the variables on
the vertical axis, while the variables themselves or their indices are
shown on the horizontal axis. This graph facilitates the identification
of important variables visually and helps focus on the variables that
require prioritization in the analysis. In the VIP score graph, the
peaks at the relevant ppm values that make up the data packages
(Bin.x.xx) have been examined in more detail and the metabolites to
which they belong have been identified. For this operation, NMR spectra
have been reopened, and the metabolites to which the peak in the
relevant ppm region belongs have been determined using the Chenomx
software. It is thought that some peaks might belong to metabolites not
found in the database, and the molecules these peaks belong to have not
been identified. Further investigation and characterization may be
required to fully understand these unidentified peaks and their role in
the overall metabolic profile, ensuring that the final analysis provides
an accurate reflection of the biological system under investigation.
Metabolomic pathway visualization : All metabolic pathways
visualized using Metastate software Version BETA
(https://metastate.bio) (Metastate Bio
Inc.). Metastate algorithm employs the Kyoto Encyclopedia of Genes and
Genomes (KEGG) database as its foundational input source. Software
systematically retrieves details pertaining to biological pathways,
chemical compounds, and molecular reactions of interest. Software
curates and assembles a dynamic graphical representation of the data.