Available via license: CC BY
Content may be subject to copyright.
Toward Personalized Interventions for
Psoriasis Vulgaris: Molecular
Subtyping of Patients by Using a
Metabolomics Approach
Dan Dai
1
, Chunyan He
2
, Shuo Wang
3
, Mei Wang
4
,
5
*, Na Guo
6
* and Ping Song
1
*
1
Department of Dermatology, Guang’anmen Hospital, China Academy of Chinese Medical Sciences, Beijing, China,
2
Department
of Dermatology, Hubei Provincial Hospital of TCM, Wuhan, China,
3
Department of Oncology, Guang’anmen Hospital, China
Academy of Chinese Medical Sciences, Beijing, China,
4
Leiden University-European Center for Chinese Medicine and Natural
Compounds, Institute of Biology Leiden, Leiden University, Leiden, Netherlands,
5
SU BioMedicine, BioPartner Center 3, Leiden,
Netherlands,
6
Experimental Research Center, China Academy of Chinese Medical Sciences, Beijing, China
Aim: Psoriasis vulgaris (PV) is a complicated autoimmune disease characterized by
erythema of the skin and a lack of available cures. PV is associated with an increased
risk of metabolic syndrome and cardiovascular disease, which are both mediated by the
interaction between systemic inflammation and aberrant metabolism. However, whether
there are differences in the lipid metabolism between different levels of severity of PV
remains elusive. Hence, we explored the molecular evidence for the subtyping of PV
according to alterations in lipid metabolism using serum metabolomics, with the idea that
such subtyping may contribute to the development of personalized treatment.
Methods: Patients with PV were recruited at a dermatology clinic and classified based on
the presence of metabolic comorbidities and their Psoriasis Area and Severity Index (PASI)
from January 2019 to November 2019. Age- and sex-matched healthy controls were
recruited from the preventive health department of the same institution for comparison. We
performed targeted metabolomic analyses of serum samples and determined the
correlation between metabolite composition and PASI scores.
Results: A total of 123 participants, 88 patients with PV and 35 healthy subjects, were
enrolled in this study. The patients with PV were assigned to a “PVM group”(PV with
metabolic comorbidities) or a “PV group”(PV without metabolic comorbidities) and further
subdivided into a “mild PV”(MP, PASI <10) and a “severe PV”(SP, PASI ≥10) groups.
Compared with the matched healthy controls, levels of 27 metabolites in the MP subgroup
and 28 metabolites in the SP subgroup were found to be altered. Among these, SM (d16:
0/17:1) and SM (d19:1/20:0) were positively correlated with the PASI in the MP subgroup,
Edited by:
Christopher Staley,
University of Minnesota Health Twin
Cities, United States
Reviewed by:
Jie Zheng,
Shanghai Jiao Tong University, China
Claudio Luchinat,
University of Florence, Italy
*Correspondence:
Mei Wang
m.wang@biology.leidenuniv.nl
Na Guo
guona5246@126.com
Ping Song
songping_cacms@163.com
Specialty section:
This article was submitted to
Metabolomics,
a section of the journal
Frontiers in Molecular Biosciences
Received: 20 May 2022
Accepted: 15 June 2022
Published: 19 July 2022
Citation:
Dai D, He C, Wang S, Wang M, Guo N
and Song P (2022) Toward
Personalized Interventions for Psoriasis
Vulgaris: Molecular Subtyping of
Patients by Using a
Metabolomics Approach.
Front. Mol. Biosci. 9:945917.
doi: 10.3389/fmolb.2022.945917
Abbreviation: BMI, body mass index; Cer, ceramide; CV, coefficient of variation; FFA, free fatty acid; HC, healthy control; HC-
MP, healthy controls matched with mild psoriasis vulgaris; HC-SP, healthy controls matched with severe psoriasis vulgaris;
HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LPC, lysophosphatidylcholine; LPE,
lysophosphatidylethanolamine; MP, mild psoriasis vulgaris; PASI, Psoriasis Area and Severity Index; PC, phosphatidylcholine;
PE, phosphatidylethanolamine; PLS-DA, partial least squares discriminant analysis; PV, psoriasis vulgaris; PVM, psoriasis
vulgaris with metabolic diseases; QC, quality control; SM, sphingomyelin; SP, severe psoriasis vulgaris; UPLC-MS/MS, ultra-
performance liquid chromatography–tandem mass spectrometry; VIP, variables of importance in projection.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459171
ORIGINAL RESEARCH
published: 19 July 2022
doi: 10.3389/fmolb.2022.945917
while Cer (d18:1/18:0), PC (18:0/22:4), and PC (20:0/22:4) were positively correlated with
the PASI in the SP subgroup. In the PVM group, levels of 17 metabolites were increased,
especially ceramides and phosphatidylcholine, compared with matched patients from the
PV group. In addition, the correlation analysis indicated that Cer (d18:1/18:0) and SM (d16:
1/16:1) were not only correlated with PASI but also has strongly positive correlations with
biochemical indicators.
Conclusion: The results of this study indicate that patients with PV at different severity
levels have distinct metabolic profiles, and that metabolic disorders complicate the disease
development. These findings will help us understand the pathological progression and
establish strategies for the precision treatment of PV.
Keywords: psoriasis vulgaris, metabolomics, lipid metabolites, severity biomarkers, metabolic diseases, molecular
subtyping
BACKGROUND
Psoriasis is a chronic, relapsing, immunoinflammatory skin
disease that affects nearly 125 million of the global population
(Takeshita et al., 2017). Psoriasis vulgaris (PV), also known as
“plaque-type psoriasis,”accounts for approximately 90% of
all cases (Boehncke and Schön, 2015) and has a multifactorial
etiology, including polygenetic disorders, environmental
factors, inflammation, and mental health (e.g., depression)
(Kamiya et al., 2019). PV progression has been reported to be
closely associated with metabolic disorders, such as obesity,
diabetes, hypertension, and cardiovascular disease (Elmets
et al., 2019;Kim et al., 2019), which may be attributed to the
interplay between inflammation and metabolic dysfunction
(Dutkiewicz et al., 2016;Kang et al., 2017). Aberrations in
lipid expression and metabolism, as well as in receptors,
enzymes, and lipid transport proteins, are frequently
observed in patients with psoriasis. Such lipid
abnormalities exacerbate psoriatic lesions and increase the
risk of developing hyperlipidemia, metabolic syndrome, and
cardiovascular disorders (Jia et al., 2018;Caietal.,2021;
Nowowiejska et al., 2021). Moreover, the drugs used in
psoriasis therapy may have an impact on the patient’slipid
profile (Cao et al., 2021); therefore, monitoring the lipid
profile is of great importance not only to disease
development but also for possible adverse response caused
by treatments. A high-fat diet rich in saturated fatty acids not
only induces obesity but also exacerbates psoriasiform
dermatitis (Herbert et al., 2018). A large clinical
observation sample also highlighted the link between lipid
metabolites and cardiovascular events in patients with PV
(Colaco et al., 2021). The treatments of PV involve a variety of
medications such as interleukins, nonsteroidal anti-
inflammatory drugs, lithium, interferons, beta-blockers,
antimalarial medications, calcium channel blockers,
terbinafine (Jain, 2017), lipid-lowering medications (James,
2005),andcontentiousTNFinhibitorssuchasinfliximab or
adalimumab (Guerra and Gisbert, 2013). Untargeted
metabolomics data revealed that interleukin-17A
monoclonal antibody (ixekizumab) treatment improves
lipids metabolism and has the potential to reduce the
cardiovascular risk in patients with psoriasis (Cao et al.,
2021). Therefore, developing a personalized intervention
strategy using an advanced technology platform will be
cost-effective and increase the quality of life of patients.
For this purpose, we first aimed to stratify patients with
PV into subtypes. The inherent relationship between PV
and metabolic diseases sheds lights on new intervention
strategies.
Metabolomics approaches enable us to decode the complex
perturbations between individual genetic inheritance and
dynamic environments by combining high-throughput
analysis technology combined with pattern recognition and
expert systems (Donnelly et al., 2019). Such methods have
been applied to capture the overall metabolic trajectory of
patients with PV by profiling small molecular metabolites
(Kamleh et al., 2015;Alonso et al., 2016;Ottas et al., 2017). As
the molecular mechanisms underlying PV pathogenesis are
not fully understood, it is difficult to identify a conclusive
method for PV treatment (Raychaudhuri et al., 2014). The
application of metabolomicsisofgreatimportancein
exploring pathological features for successful clinical
management and individualized medicine (Tarentini et al.,
2021). Kang et al. (2017) identified hypoxanthine,ornithine,
azelaic acid, and crotonic acid as candidate biomarkers for
patients with psoriasis in an untargeted serum metabolomics
study. Kamleh et al. (2015) revealed that PV has specific
metabolic profiles at different degrees of severity, and that the
severity-associated metabolic perturbations may stem from
keratinocyte hyperproliferation, active collagen synthesis, or
the incidence of cachexia. However, whether differences in
lipid metabolism can indeed be observed in different
severities of PV remains largely unknown.
Here, we therefore recruited patients with PV as well as healthy
controls (HCs) and further su bclassified the patients by the presence
of metabolic comorbidities and their Psoriasis Area and Severity
Index (PASI). A targeted lipidomics approach based on ultra-
performance liquid chromatography–tandem mass spectrometry
(UPLC-MS/MS) was applied to characterize lipid biomarkers for
PV subtyping and examine distinct metabolic signatures.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459172
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
MATERIALS AND METHODS
Participant Selection
Patients with PV and HC participants were recruited at
Guang’anmen Hospital, China Academy of Chinese Medical
Sciences (Beijing, China). We posted recruitment
advertisements on the hospital’s websites and notice boards,
providing a brief overview of the study’sgoals,themedical
assessments participants were to undergo, the eligibility criteria,
and instructions on how to get involved. Patients were enrolled
based on the following criteria: 1) aged 18–65 years, 2) diagnosis
of PV, and 3) provision of a signed informed consent form. The
following participants were excluded: 1) those who underwent
systemic treatment within 4 weeks; 2) those with a diagnosis of
pustular psoriasis, psoriatic arthritis, or erythrodermic psoriasis;
3) those with severe cardiovascularorcerebrovasculardisease,
abnormal liver or renal function, cancer, or psychosis disorders;
and 4) pregnant or lactating women. None of the participants
had dietary restrictions. Enrolled patients with PV were
assigned to either the “PVM group”(PV with metabolic
diseases) or the “PV group”(PV without metabolic diseases)
and further subdivided into the “mild PV”(MP, PASI <10) and
the “severe PV”(SP, PASI ≥10) groups. Healthy individuals
without a history of psoriasis, chronic inflammatory systemic
diseases, or obesity-related metabolic diseases served as the HC
group and were age- and sex-matched to the MP and SP
subgroups. Participants’eligibility was examined by two
dermatology physicians at the dermatological clinic of our
institution, both deemed them fit for this study. Written
informed consent was obtained from all participants. Serum
samples were collected and analyzed from each recruited
participant to obtain lipid biomarkers for PV subtyping. The
recruitment process began in January 2019 and was completed
in November 2019.
Chemicals and Materials
We purchased internal standards for ceramide (Cer (d18:1/17:0)),
phosphatidylethanolamine (PE [12:0/13:0]), sphingomyelin (SM
[d18:1/12:0]), and phosphatidylcholine (PC (19:0/19:0)) from
Avanti Polar Lipids (Alabaster, AL, United States). Free fatty
acid (FFA C19:0) was sourced from Larodan (Stockholm,
Sweden). MS-grade isopropanol, acetonitrile, methanol, and
formic acid were purchased from Fisher Scientific (Waltham,
MA, United States). Ultra-pure water was obtained from a Milli-
Q water purification system and used throughout the experiments
(Millipore, Bedford, MA, United States). All other reagents and
chemicals used were of analytical grade and commercially
available.
Sample Preparation
Following collection, whole blood samples were centrifuged for
10 min at 3,000 rpm in a refrigerated centrifuge (4°C), after which
the supernatants were aliquoted into 1.5-ml Eppendorf tubes and
stored at −80°C until the subsequent metabolomic analysis.
Standard stock solutions were prepared by dissolving the
compounds in methanol to a concentration of 1 mg/ml and
stored below −20°C.
For the targeted lipid and fatty acid analyses, 10 μl of serum
was added to 150 μl of cold methanol containing a mixture of the
following internal standards: FFA C19:0 (200 ng/ml), SM (d18:1/
12:0) (40 ng/ml), PC (19:0/19:0) (30 ng/ml), Cer (d18:1/17:0)
(200 ng/ml), and PE (12:0/13:0) (100 ng/ml). The samples were
mixed for 30 s, after which 500 µl of methyl tert-butyl ether was
added, and the samples were incubated under gentle agitation for
20 min at room temperature to extract the full lipids. After the
addition of 125 µl of water, the samples were shaken and
centrifuged for 10 min at 13,200 rpm at 4°C. The upper lipid
extracts of 100 µl were transferred into a new centrifuge tube,
vacuum-dried, and resuspended in a 200 µl solution of
acetonitrile/isopropanol/water (65:30:5, v/v/v). The samples
were again vortexed for 1 min and centrifuged as described
previously, and the supernatants were instantly analyzed using
UPLC-MS/MS.
To validate the stability of the LC-MS system, pooled quality
control (QC) samples were prepared by mixing 1 ml of each
serum sample prepared as described previously.
UPLC-MS/MS Conditions
The UPLC and MS systems were controlled using MassLynx
Mass Spectrometry Software (v4.1) (Waters Corp., MA,
United States). All chromatographic separations were
performed using an Acquity UPLC BEH C8 Column (1.7 μm,
100 × 2.1 mm
2
; Waters Corp.). The mobile phase consisted of
5 mM ammonium formate in acetonitrile/water (6:4, v/v; mobile
phase A) and 5 mM ammonium formate in isopropanol/
acetonitrile (9:1, v/v; mobile phase B). The column was
maintained at 55°C, and the flow rate was set at 0.26 ml/min.
The linear elution gradient settings are shown in Supplementary
Table S1. An injection volume of 1 μl was applied in the positive
ion mode and a 2 μl volume in the negative ion mode. During the
mass spectrometer (Xevo TQ-S; Waters Corp.) operation, we
applied multiple reaction monitoring in the positive and negative
ion modes. The TargetLynx Application Manager (Waters Corp.)
was used to process the data.
System Stability
To guarantee the stability of the LC-MS system throughout our
analyses, we evaluated the pooled QC samples in both the positive
and the negative ion modes. The retention time and intensity
measurements suggested that the stability and repeatability of the
measurements was satisfactory and high, respectively, throughout
for the experiment. The bulk of the coefficient of variation (CV)
values belonging to the internal standard peak area for the
targeted lipid and fatty acid tests fell below 15%.
Data Analysis
Raw data from the UPLC-MS/MS analysis were first imported
into Progenesis QI Software (v2.3; Waters Corp.). Baseline
filtering, peak identification, integration, retention time
correction, and peak alignment were then performed to
optimize the setting parameters. All chromatographic peaks
were confirmed manually by careful investigation to verify
the accuracy of the results, and a data matrix was obtained,
including information such as mass-to-charge ratio (m/z),
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459173
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
retention time, and peak area (intensity). The data matrix
obtained was exported to SIMCA software (v14.1; Umetrics,
Umeå, Sweden) for partial least squares discriminant analysis
(PLS-DA). The R
2
and Q
2
statistics were used to assess the
quality of the PLS-DA model, whereas its reliability was
determined using a permutation test, because of the model’s
potential to overestimate the separation performance. All
statistical tests were performed using SPSS (v25.0; SPSS Inc.,
TABLE 1 | Demographics of the study cohort.
Comparison 1
Characteristic MP patient (n= 31) HC-MP subject (n= 31) p-value
Age (years) 19–62, 33.45 ± 11.38 19–58, 34.84 ± 10.74 0.667
Female 14 18 0.309
PASI 6.08 ± 2.52 NA NA
Duration of psoriasis (years) 9.81 ± 7.01 NA NA
Family history 11 NA NA
BMI (kg/m
2
) 23.23 ± 2.98 22.35 ± 2.66 0.383
Total cholesterol (mmol/L) 4.53 ± 0.63 4.77 ± 0.79 0.252
Triglyceride (mmol/L) 1.37 ± 0.77 0.91 ± 0.26 0.071
LDL cholesterol (mmol/L) 2.76 ± 0.43 2.94 ± 0.58 0.203
HDL cholesterol (mmol/L) 1.29 ± 0.26 1.43 ± 0.28 0.063
Comparison 2
Characteristic SP patient (n= 32) HC-SP subject (n=32) p-value
Age (years) 19–57, 36.03 ± 10.16 19–58, 36.25 ± 11.38 0.92
Female 13 19 0.134
PASI 17.69 ± 6.41 NA NA
Duration of psoriasis (years) 13.07 ± 8.26 NA NA
Family history 9 NA NA
BMI (kg/m
2
) 23.33 ± 3.99 22.34 ± 2.60 0.243
Total cholesterol (mmol/L) 4.68 ± 0.90 4.79 ± 0.82 0.665
Triglyceride (mmol/L) 1.23 ± 0.56 0.94 ± 0.26 0.059
LDL cholesterol (mmol/L) 2.92 ± 0.72 2.97 ± 0.58 0.812
HDL cholesterol (mmol/L) 1.27 ± 0.30 1.41 ± 0.26 0.066
Comparison 3
Characteristic MP patient (n= 31) SP patient (n=32) p-value
Age (years) 19–62, 33.45 ± 11.38 19–57, 36.03 ± 10.16 0.346
Female 14 13 0.716
PASI 6.08 ± 2.52 17.69 ± 6.41 0
Duration of psoriasis (years) 9.81 ± 7.01 13.07 ± 8.26 0.097
Family history 11 9 0.53
BMI (kg/m
2
) 23.23 ± 2.98 23.33 ± 3.99 0.91
Total cholesterol (mmol/L) 4.53 ± 0.63 4.68 ± 0.90 0.488
Triglyceride (mmol/L) 1.37 ± 0.77 1.23 ± 0.56 0.741
LDL cholesterol (mmol/L) 2.76 ± 0.43 2.92 ± 0.72 0.339
HDL cholesterol (mmol/L) 1.29 ± 0.26 1.27 ± 0.30 0.766
Comparison 4
Characteristic PVM patient (n= 25) PV patient (n= 25) p-value
Age (years) 26–65, 45.04 ± 9.34 24–57, 42.60 ± 8.22 0.332
Female 9 9 1
PASI 18.99 ± 14.62 15.62 ± 8.74 0.816
Duration of psoriasis (years) 15.83 ± 10.76 14.48 ± 8.23 0.621
Family history 7 7 1
BMI (kg/m
2
) 27.29 ± 3.75 23.91 ± 3.28 0.002
Total cholesterol (mmol/L) 5.49 ± 1.13 5.06 ± 0.77 0.165
Triglyceride (mmol/L) 2.85 ± 1.27 1.43 ± 0.64 0
LDL cholesterol (mmol/L) 3.56 ± 0.88 3.13 ± 0.64 0.084
HDL cholesterol (mmol/L) 1.15 ± 0.19 1.29 ± 0.34 0.107
BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; HC-MP, healthy controls matched with mild psoriasis vulgaris; HC-SP, healthy controls matched with severe psoriasis
vulgaris; LDL-C, low-density lipoprotein cholesterol; MP, mild psoriasis vulgaris (PASI <10); PASI, Psoriasis Area and Severity Index; PVM, psoriasis vulgaris with metabolic diseases; PV,
psoriasis vulgaris without metabolic diseases; SP, severe psoriasis vulgaris (PASI ≥10).
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459174
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
IL, United States). The non-parametric Student’st-test,
Kruskal–Wallis test, and chi-square test were used to
compare the data, as appropriate. The p-values of <0.05 were
considered significant, and no corrections were performed for
multiple testing as the study was exploratory. The variables of
importance in projection (VIP) analysis following PLS-DA
modeling and a Student’st-test analysis were applied to
identify biomarkers (VIP >1, p<0.05). Spearman’s
correlation coefficients were calculated to study the
associations between the expression of metabolites, PASI, and
biochemical indicators using the ggplot2 package (v3.3.5) in R
(https://www.R-project.org).
RESULTS
Participant Baseline Characteristics
In total, 123 participants, including 88 patients with PV and
35 HCs, were enrolled in this study. The PV group consisted of
63 patients, and the PVM group consisted of 25; the patients in
the PV group were further subdivided into 31 in the MP subgroup
and 32 in the SP subgroup. We performed four pairwise
comparisons, one between the MP subgroup and matched
HCs (MP vs. HC-MP; 31 vs. 31), one between the SP
subgroup and matched HCs (SP vs. HC-SP; 32 vs. 32), one
between both subgroups (MP vs. SP; 31 vs. 32), and one
between the PVM group and matched patients from the PV
group (PVM vs. PV, 25 vs. 25). The HC-MP and HC-SP
subgroups were based on the same 25 shared healthy subjects.
The characteristics of the study cohorts are shown in Table 1, and
the metabolic complications of the patients in the PVM group are
illustrated in Supplementary Table S2. Among the 38 patients
with PV whose PASI was <10, we identified 7 (18.42%) patients
with metabolic disorders, while of the 50 patients with PV whose
PASI was ≥10, 18 (36%) had metabolic disorders (Table 2). These
data demonstrated a significant increase in metabolic disorders in
patients with severe PV, suggesting that dysfunctions in lipid
metabolism may contribute to the progression of PV.
Targeted Metabolomic Analysis
In our targeted metabolomics analysis, we focused on the
expressions of specific lipids and fatty acids in the different
groups. During the UPLC-MS/MS analysis, QC samples were
regularly run to ensure the reproducibility of the results, and the
majority of the CV values corresponding to the internal standard
peak area were <15%, which is consistent with good stability and
reliability. In total, 25 FFAs and 131 lipid metabolites were
TABLE 2 | Metabolic disorder in sub-population of PV.
Patient PASI <10 PASI ≥10
No. of PV patient 38 50
No. of PVM patient 7 (18.42%) 18 (36%)
FIGURE 1 | Comparison of metabolites using PLS-DA models. (A) PLS-DA model of the MP and HC-MP subgroups. (B) PLS-DA model of the SP and HC-SP
subgroups. (C) PLS-DA model of the MP and SP subgroups. (D) PLS-DA model of the PVM and PV groups. HC-MP, healthy controls matched to the group with mild
psoriasis vulgaris; HC-SP, healthy controls matched to the group with severe psoriasis vulgaris; MP, mild psoriasis vulgaris; PLS-DA, partial least squares discriminant
analysis; PV, psoriasis vulgaris without metabolic diseases; PVM, psoriasis vulgaris with metabolic diseases; SP, severe psoriasis vulgaris.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459175
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
detected. Relevant information on precursor ions, product ions,
and collisions for all the lipid and fatty acid samples is provided in
Supplementary Table S3.
Serum Metabolic Profiling Identifies
Patients With PV at Different Severity Levels
The PLS-DA analysis was performed to investigate the metabolic
differences among the groups and subgroups (Figure 1).
Permutation tests (200 permutations) were performed to verify
the PLS-DA models. Both the permuted R
2
and Q
2
values were
significantly lower than the corresponding original values (see
Supplementary Figure S1), suggesting good compatibility of the
data and predictive ability of the model. Compared with the
matched HC-MP and HC-SP control groups, participants in the
MP and SP subgroups exhibited a significantly distinctive
metabolic signature, suggesting dysfunctional lipid metabolism
in patients with PV (Figures 1A,B). Although the MP subgroup
coincided with the SP group, PC, LPC, and SM showed
alterations between the two (Figure 1C). A clear distinction
was also observed between the PVM group and matched
patients in the PV group, with differences in Cer, FFA, LPC,
PC, and PE between the two groups (Figure 1D). This finding
suggests an overlap of PV and metabolic disorders (Table 3).
Altered Metabolites in Patients With PV at
Different Severity Levels
Our univariate statistical analysis revealed significant differences
in lipid and fatty acid levels among the groups and subgroups
(Table 4). The MP subgroup exhibited higher levels of three types
of PE, six types of PC, and FA 16:2, and lower levels of five types
of PC, four types of LPC, seven types of SM, and PE (18:2/16:1)
than the HC-MP subgroup (Figure 2A). In the comparison
between the SP and HC-SP subgroups, the levels of two types
of PC, two types of Cer, PE (18:2/16:0), and FA 16:2 were elevated
in the SP subgroup, while the levels of two types of LPC, ten types
of SM, nine types of PC, and lysophosphatidylethanolamine 20:2
(LPE 20:2) were reduced (Figure 2B). The lower levels of PC,
LPC, and SM observed in the SP subgroup distinguished it from
the MP subgroup (Figure 2C). As for the comparison between the
PVM and PV groups, levels of six types of Cer, two types of FFA,
seven types of PC, LPC 20:2, and PE (22:5/16:0) were all
significantly high in the PVM group (Figure 2D), indicating
that the disorder in lipid metabolism served a critical role in
metabolic complications of psoriasis.
Associations Between the Altered
Metabolites and the Metabolic Signatures
of Severity
Spearman’s correlation was applied to explore the associations
between all the measured metabolites and the patient’s PASI
scores (Supplementary Table S4). SM (d16:1/16:1) and LPE 20:
4 were negatively correlated with PASI, whereas a positive
correlation was observed between Cer (d18:1/18:0) and the
PASI. The correlation analysis between all the measured
metabolites and biochemical indicators including cholesterol,
triglyceride, high-density lipoprotein cholesterol (HDL-C), and
low-density lipoprotein cholesterol (LDL-C) was also performed
to enrich the significance of the harvested lipid biomarkers
(Supplementary Table S5). Surprisingly, Cer (d18:1/18:0) and
SM (d16:1/16:1) were not only correlated with PASI but also
strongly correlated with biochemical indicators. Cer (d18:1/18:0)
was strongly positively correlated with cholesterol, triglyceride,
and LDL-C, whereas SM (d16:1/16:1) had the strongly positive
correlations with cholesterol, HDL-C, and LDL-C. Furthermore,
we observed positive correlations (all values of p<0.05) among
most of the altered metabolites (Figure 3). SM and PC, and PC
and PE showed a strong correlation within the altered metabolites
in the MP subgroup (Figure 3A). In the SP subgroup, a strong
correlation between SM and PC, and PC and LPC was observed in
the altered metabolites (Figure 3B). This observation indicates
that the progression of PV symptoms may be associated with an
alteration of lipid metabolism. Moreover, we also found a strong
correlation between PC and FFA, and PE and Cer in the PVM
group (Figure 3E). To further investigate the severity biomarkers
of PV, we conducted Spearman’s correlation analysis between the
PASI scores and the altered metabolites (MP vs. HC-MP and SP
vs. HC-SP). The results show that SM (d19:1/20:0) and SM (d16:
1/17:0) in the MP subgroup were positively correlated with the
PASI scores (Figure 4A). In addition, PC (18:0/22:4), PC (20:0/
22:4), and Cer (d18:1/18:0) in the SP subgroup were positively
correlated with the PASI scores (Figure 4B). There was no
metabolite that strongly was correlated with the PASI in both
the MP and the SP subgroups.
DISCUSSION
Based on the idea that subtyping of PV may contribute to the
development of personalized treatment, we explored the
molecular evidence for alterations in lipid metabolism at
different severity levels of the disease using serum
metabolomics. The identified metabolic disorders related to
the progress of PV are summarized in Figure 5 excessive
hydrolysis causes the accumulation of PC and subsequently
lipids droplet formates , which may be an important factor
that contributes to disease progression. Using the targeted
UPLC-MS/MS-based lipidomics platform to characterize the
TABLE 3 | Significantly altered lipids between different groups.
Lipid MP vs. HC-MP SP vs. HC-SP SP vs. MP PVM vs. PV
Cer –√–√
FFA √√–√
LPC √√√√
LPE –√––
PC √√√√
PE √√–√
SM √√√–
Cer, ceramide; FFA, free fatty acid; LPC, lysophosphatidylcholine; LPE,
lysophosphatidylethanolamine; PC, phosphatidylcholine; PE,
phosphatidylethanolamine; SM, sphingomyelin.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459176
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
TABLE 4 | Expression trends of metabolites altered between different groups.
Metabolite MP vs. HC-MP SP vs. HC-SP SP vs. MP PVM vs. PV
Cer (d18:1/16:0) ––––
Cer (d18:1/18:0) –△↑––
Cer (d18:1/24:1) –△↑–*↑
Cer (d18:1/25:0) –––△↑
Cer (d18:1/22:0) –––*↑
Cer (d18:1/22:2) –––*↑
Cer (d18:1/24:0) –––*↑
Cer (d18:2/22:0) –––△↑
FA 16:2 *↑△↑––
FA 19:0 –––△↑
FA 24:0 –––△↑
LPC 17:0 △↓–––
LPC 20:0 △↓–––
LPC 20:2 –––△↑
LPC 20:3 ––△↓–
LPC 20:5 –△↓△↓–
LPC 21:0 △↓–––
LPC 22:0 △↓–––
LPC 22:6 –△↓––
LPE 20:2 –△↓––
PC (10:0/19:1) △↓*↓––
PC (16:1/16:0) △↑–△↓–
PC (16:0/17:1) ––△↓–
PC (14:0/18:2) –––*↑
PC (18:0/16:0) –––△↑
PC (15:1/18:2) △↓△↓––
PC (16:1/18:2) ––△↓△↑
PC (18:3/16:0) ––△↓–
PC (20:4/14:0) ––△↓△↑
PC (18:1/17:0) –△↓––
PC (18:2/17:0) –△↓––
PC (18:2/19:1) –––△↑
PC (17:1/18:2) ––△↓–
PC (18:0/18:1) △↑–△↓–
PC (20:4/16:1) –△↓△↓–
PC (20:4/17:0) –*↓––
PC (15:0/22:6) △↓△↓––
PC (18:0/20:2) △↑––△↑
PC (18:0/20:5) ––△↓–
PC (18:0/22:4) △↑△↑––
PC (18:0/22:5) △↑–––
PC (20:4/20:4) △↓△↓––
PC (20:4/22:6) △↓△↓––
PC (20:0/18:0) –––*↑
PC (20:0/22:4) △↑△↑––
PE (12:0/13:0) ––––
PE (16:0/16:0) ––––
PE (16:0/18:1) *↑–––
PE (18:2/16:0) △↑△↑––
PE (18:2/16:1) △↓–––
PE (18:0/22:5) △↑–––
PE (22:5/16:0) –––△↑
SM (d16:0/15:1) –#↓––
SM (d16:1/16:0) –△↓△↓–
SM (d16:1/17:0) *↓*↓––
SM (d18:1/15:1) △↓*↓––
SM (d18:1/17:0) △↓△↓––
SM (d18:1/17:1) *↓*↓––
SM (d18:1/19:0) ––––
SM (d18:1/19:1) △↓△↓––
SM (d19:1/20:0) △↓*↓––
SM (d20:0/22:6) –△↓––
SM (d17:1/26:1) –△↓––
SM (d16:0/22:3) ––––
SM (d16:0/17:0) △↓–––
△,p<0.05; *, p<0.01; and #, p<0.001
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459177
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
serum samples of patients, we inspected the pathological
mechanism underlying PV that is closely associated with
complex lipid metabolism dysfunction and thereby
identified a set of PV’s lipid biomarkers. The correlation
analysis indicated that Cer (d18:1/18:0) and SM (d16:1/16:
1) were not only correlated with PASI but also has strongly
positive correlations with biochemical indicators in our
analysis. Positive correlations were observed among the
most altered metabolites. Among the altered metabolites in
the MP subgroup, SM and PC levels showed a strong
correlation. A strong correlation between SM and PC, and
PC and LPC was observed for the altered metabolites in the SP
subgroup. We focused on the biomarkers associated with the
severity of PV and found that in the MP subgroup, SM (d16:0/
17:1) and SM (d19:1/20:0) were positively correlated with the
PASI, whereas in the SP subgroup, PC (18:0/22:4), PC (20:0/
22:4), and Cer (d18:1/18:0) were positively correlated with
PASI scores. These results suggest that SM disorders
dominated the lipid abnormalities in the MP subgroup,
while Cer and PC disorders were predominant in the SP
subgroup. In addition, we found more patients with severe PV
complicated by metabolic diseases than patients with mild
PV, suggesting that dysfunction in lipid metabolism may
contribute to the progression of PV. Cer, FFA, PC, LPC,
and PE levels were significantly higher in patients with PV
complicated by metabolic diseases than in those without
metabolic diseases. Although the pathogenesis of metabolic
complications of PV is unclear, it has been linked to lipid
accumulation caused by dysfunctional lipid metabolism. Our
findings thus offer novel insights into the metabolic nature of
PV. The metabolic biomarkers we suggest for the subtyping of
PV may assist practitioners in identifying potentially
important risk factors in the future and support the
development of precision medicine treatment for PV.
SM is primarily produced in the Golgi apparatus and
transferred to all other cellular membranes (Campelo et al.,
FIGURE 2 | Volcano plots illustrating differential metabolites. (A) Volcano plot of differential metabolites between the MP and HC-MP subgroups. (B) Volcano plot of
differential metabolites between the SP and HC-SP subgroups. (C) Volcano plot of differential metabolites between the MP and SP subgroups. (D) Volcano plot of
differential metabolites between the PVM and PV groups. HC-MP, healthy controls matched to the group with mild psoriasis vulgaris; HC-SP, healthy controls matched
to the group with severe psoriasis vulgaris; MP, mild psoriasis vulgaris; SP, severe psoriasis vulgaris; PV, psoriasis vulgaris without metabolic diseases; PVM,
psoriasis vulgaris with metabolic diseases.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459178
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
2017). Sphingomyelinase catalyzes the hydrolysis of the
phosphodiester bond of sphingomyelin and yields Cer and PC
after stimulation by TNF-αor IL-1β(Sindhu et al., 2021). In our
sample, SM disorders dominated the lipid abnormalities observed
in the MP subgroup and thus provide possible treatment targets
for MP. Our results indicate that SM was reduced in patients in
the MP subgroup compared to HCs. Among all the measured
lipid metabolites, SM (d16:1/16:1) is negatively correlated with
PASI and had strongly positive correlations with cholesterol,
HDL-C, and LDL-C. SMs are the most abundant surface
components of plasma lipoproteins including LDL and HDL,
and abnormal SM metabolism dysfunction impairs the structure
FIGURE 3 | Heatmaps illustrating Spearman’s correlations between the altered metabol ites. (A) Correlation heatmap of altered metabolites (MP vs. HC-MP) in the
MP subgroup. (B) Correlation heatmap of altered metabolites (SP vs. HC-SP) in the SP subgroup. (C) Correla tion heatmap of altered metabolites (MP vs. SP) in the MP
subgroup. (D) Correlation heatmap of altered metabolites (MP vs. SP) in the SP subgroup. (E) Correlation heatmap of altered metabolites (PVM vs. PV) in the PVM
group. Correlation analysis performed with Spearman’s correlation coefficient, *p<0.05, **p<0.01, and ***p<0.001. HC-MP, healthy controls matched to the
group with mild psoriasis vulgaris; HC-SP, healthy controls matched to the group with severe psoriasis vulgaris; MP, mild psoriasis vulgaris; SP, severe psoriasis vulgaris;
PV, psoriasis vulgaris without metabolic diseases; PVM, psoriasis vulgaris with metabolic diseases.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 9459179
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
FIGURE 4 | Spearman’s correlations between altered metabolites and the PASI. (A) Correlation analyses between PASI and altered metabolites (MP vs. HC-MP) in
the MP subgroup: SM (d16:1/17:0): R= 0.37, p= 0.041; SM (d19:1/20:0): R= 0.39, p= 0.032. (B) Positive correlations between PASI and altered metabolites (SP vs.
HC-SP) in the SP subgroup: Cer (d18:1/18:0): R= 0.42, p= 0.015; PC (18:0/22:4): R= 0.45, p= 0.01; and PC (20:0/22:4): R= 0.39, p= 0.029. PASI, Psoriasis Area and
Severity Index; HC-MP, healthy controls matched to the group with mild psoriasis vulgaris; HC-SP, healthy controls matched to the group with severe psoriasis
vulgaris; MP, mild psoriasis vulgaris; SP, severe psoriasis vulgaris.
FIGURE 5 | Graphical abstract showing the major findings of the study.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 94591710
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
and biofunctions of LDL and HDL (Iqbal et al., 2017;Tsuji et al.,
2021). Reduced SM levels may contribute to disease progression,
as it has been reported that SM supplements appear to protect
against aberrant lipid metabolism, intestinal dysbiosis, and
inflammation (Norris et al., 2017;Norris et al., 2019). The SM
derivative sphingosine-1-phosphate (S1P) activates
differentiation and inhibits keratinocyte proliferation (Jeon
et al., 2020). S1P receptor agonists have been found to
alleviate psoriasis-like dermatitis in mice (Liu et al., 2021).
Moreover, controversially, a high-fat diet can induce
sphingomyelin accumulation in the liver (Chocian et al.,
2010). The lack of sphingomyelinase promotes sphingomyelin
accumulation and causes defective cholesterol trafficking and
efflux, which may occur in diabetes and atherosclerotic heart
disease (Leventhal et al., 2001;Summers, 2006). The seemingly
opposite functions of SM may thus explain why SM is deficient in
MP but positively correlated with the PASI.
Cer and PC were found to be related to lipid metabolism
dysfunction in the SP subgroup. Cer is the key precursors for the
synthesis of other sphingolipids, and the decreased SM and
elevated Cer values suggest excessive hydrolysis or impaired
synthesis of SM in this group. Different Cer species have
specific and sometimes opposing biological functions (Gomez-
Larrauri et al., 2020). Cer promotes the NLRP3 inflammasome,
which increases IL-1βlevels (Vandanmagsar et al., 2011) and
enhances TNF-α, MCP-1, and IL-6 production in adipocytes
(Samad et al., 2006). High concentrations of Cer in the white
adipose tissue exacerbate chronic inflammation and insulin
resistance (Kolak et al., 2007;Chaurasia et al., 2016), which is
an independent risk factor for cardiovascular death in patients
with stable coronary artery disease (Laaksonen et al., 2016).
However, the overall serum content of Cer was reported to be
lower in psoriatic patients (Myśliwiec et al., 2017), which is not in
line with our results. Cer deficiency in the stratum corneum has
also been found to lead to the dry desquamation of the skin
(Bocheńska and Gabig-Cimińska, 2020). The correlation analysis
highlighted that the expression of Cer (18:1/18:0) was strongly
positively correlated with PASI not only in an early stage of PV
such as SP group but also in all enrolled patients with PV, which
can be considered a severity biomarker for clinical reference.
PC plays a role in regulating blood lipoprotein homeostasis.
Impaired hepatic PC biosynthesis significantly reduces the
concentrations of circulating very low-density lipoproteins as
well as high-density lipoproteins (HDLs) (da Silva et al., 2020;
Sprenger et al., 2021). HDL not only eliminates excess cholesterol
but also has anti-inflammatory and antioxidant properties (Kuai
et al., 2016). The phospholipid composition determines HDL
function and secretion. The composition of HDL
phosphatidylcholine affects the hepatic absorption of HDL
lipid cargo (Kadowaki et al., 1993). In our sample, some kinds
of PC were downregulated, while some were upregulated in
patients with PV serum, which is consistent with the results of
previous studies (Zeng et al., 2017;Li et al., 2020). Downregulated
PC was also found in patients with PV (Łuczaj et al., 2020),
potentially due to the excessive proliferation of keratinocytes
during the progression of PV. Dietary PC supplementation
substantially ameliorates lipid transportation dysfunction and
atherosclerosis (Aldana-Hernández et al., 2021). Excess PC
may be catabolized and promote triglyceride synthesis and
triglyceride-mediated steatosis (Martínez-Uña et al., 2013),
which increases the risk of obesity-related diseases in patients
with PV. The positive correlation between triglyceride and PC in
our experiment also supported this conclusion. The significantly
high triglyceride levels observed in our comparison of the PV and
PVM groups may have been a result of PC accumulation. PC
disorders dominated the lipid abnormalities observed in the SP
subgroup, which suggests the need for attention to lipoprotein
functions when it comes to the development of targeted
preventive treatment for PV.
LPC is generated from PC through the hydrolysis of
phospholipase A2 and is the core component of oxidatively
damaged low-density lipoprotein (Shoda et al., 1997;Ohigashi
et al., 2019). Thus, the altered PC and LPC levels we observed in
the SP subgroup were strongly correlated. Increasing evidence
indicates that LPC levels are elevated in inflammation-related
diseases, including psoriasis (Zeng et al., 2017;Li et al., 2020).
Nevertheless, in the current study, the entire PC composition was
decreased. LPC usually has a bidirectional effect on inflammation
regulation, resulting in various functions in the progression of
inflammatory diseases (Liu et al., 2020). LPC with
polyunsaturated acyls can curtail inflammation induced by
saturated LPC by acting as an anti-inflammatory lipid
mediator (Hung et al., 2012). Moreover, LPC can reinforce the
immunosuppressive function of regulatory T cells by increasing
the production of TGF-βand Foxp3 via the G2A-JNK pathway
(Hasegawa et al., 2011).
The top three metabolic comorbidities in the PVM group were
hyperlipidemia (48%), hypercholesterolemia (28%), and
hyperlipidemia and hypertension (12%), which were presented
with serum accumulation of Cer, PC, LPC, FFA, and PE. Elevated
PC levels promote triglyceride synthesis and triglyceride-
mediated steatosis (Łuczaj et al., 2020). In addition to the
routine synthesis pathway (CDP-choline pathway) (Kennedy
and Weiss, 1956), the liver possesses a unique PC synthesis
pathway that involves three consecutive methylations of the
ethanolamine moiety of PE catalyzed by PE methyltransferase
and accounts for approximately 30% of hepatic PC synthesis
(Sundler and Akesson, 1975). Abnormal cellular PC/PE molar
ratios in diverse tissues have been associated with disease
progression and alterations in energy metabolism (van der
Veen et al., 2017). Ottas et al. (2017) found that PE levels are
high in the serum of patients with psoriasis. This finding is
consistent with our observations. Increased PE fraction
breakdown could release arachidonic acid, which would be
accompanied by lower quantities of bile acids in the serum of
patients with psoriasis, as well as poorer antioxidant indicators,
such as glutathione (Sorokin et al., 2018;Miura et al., 2021). We
also found higher Cer levels in patients with only PV than HCs,
but their levels were still lower than those of patients with PV and
metabolic diseases, suggesting that Cer may contribute to
metabolic and inflammatory responses. Circulating FFAs are
the primary energy source for almost all tissues and are
obtained mostly from adipose tissue lipolysis, and elevated
saturated fatty acids levels exacerbate early psoriatic skin
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 94591711
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
inflammation (Herbert et al., 2018). Fatty acid synthesis appears
to drive the formation of Th17 cells (Young et al., 2017;Nicholas
et al., 2019). Polyunsaturated fatty acids can induce thrombosis
and proinflammation, resulting in an increased prevalence of
atherosclerosis, obesity, and diabetes (Simopoulos, 2013;
Kromhout and de Goede, 2014;Simopoulos, 2016).
This study has some limitations. The reported lipids could not be
identified with high certainty using MS/MS because of poor lipid
abundance, ion suppression, and co-elution. Moreover, the correlations
obtained in the current study were retrieved from cross-sectional data;
thus, the identified biomarkers need to be verified in a large
independent cohort before being applied in clinical practice. Future
work should focus on a detailed understanding of the intricate
mechanisms governing the interaction between systemic
inflammation and abnormal lipid metabolic responses.
CONCLUSION
Classification into molecular subtypes enables the identification of
potential therapeutic targets that are specifictocomplexand
refractory diseases, such as PV. We validated the subtypes of PV
with different lipid metabolic profiles and identified their molecular
biomarkers related to disease progression. Our findings therefore
contribute to the understanding of the pathogenesis and
development of novel strategies for precision treatment of PV.
DATA AVAILABILITY STATEMENT
The raw data supporting the conclusion of this article will be
made available by the authors, without undue reservation.
ETHICS STATEMENT
The studies involving human participants were reviewed and
approved by the Ethics Committee of the Guang’anmen Hospital,
China Academy of Chinese Medical Sciences. The patients/
participants provided their written informed consent to
participate in this study.
AUTHOR CONTRIBUTIONS
PS, NG, and MW conceived the study, designed the experiments,
interpreted the results, and reviewed the manuscript. DD wrote
the manuscript. DD and CH collected clinical samples and
information. CH performed the targeted metabolomics
experiment. SW performed the targeted metabolomics analysis.
All authors read and approved the final manuscript.
FUNDING
The laboratory of the authors benefits from ongoing support from
the Scientific and Technological Innovation Project of China
Academy of Chinese Medical Sciences (CI2021A02301), National
Natural Science Foundation of China (82074448), and the
Fundamental Research Funds for the Central Public Welfare
Research Institutes (JBGS2021002).
ACKNOWLEDGMENTS
The authors thank the patients, their families, and the
investigators who participated in this study for their support
and valuable input. Also, the authors thank the Figdraw (https://
www.figdraw.com/) for assistance regarding illustrations.
SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at:
https://www.frontiersin.org/articles/10.3389/fmolb.2022.945917/
full#supplementary-material
REFERENCES
Aldana-Hernández, P., Azarcoya-Barrera, J., van der Veen, J. N., Leonard, K.-A.,
Zhao, Y.-Y., Nelson, R., et al. (2021). Dietary Phosphatidylcholine
Supplementation Reduces Atherosclerosis in Ldlr Male Mice2. J. Nutr.
Biochem. 92, 108617. doi:10.1016/j.jnutbio.2021.108617
Alonso, A., Julià, A., Julià, A., Vinaixa, M., Domènech, E., Fernández-Nebro, A.,
et al. (2016). Urine Metabolome Profiling of Immune-Mediated Inflammatory
Diseases. BMC Med. 14 (1), 133. doi:10.1186/s12916-016-0681-8
Bocheńska, K., and Gabig-Cimińska, M. (2020). Unbalanced Sphingolipid
Metabolism and its Implications for the Pathogenesis of Psoriasis. Molecules
25 (5). doi:10.3390/molecules25051130
Boehncke, W.-H., and Schön, M. P. (2015). Psoriasis. Lancet 386 (9997), 983–994.
doi:10.1016/s0140-6736(14)61909-7
Cai, J., Cui, L., Wang, Y., Li, Y., Zhang, X., and Shi, Y. (2021). Cardiometabolic
Comorbidities in Patients with Psoriasis: Focusing on Risk, Biological Therapy,
and Pathogenesis. Front. Pharmacol. 12, 774808. doi:10.3389/fphar.2021.
774808
Campelo, F., van Galen, J., Turacchio, G., Parashuraman, S., Kozlov, M. M., García-
Parajo, M. F., et al. (2017). Sphingomyelin Metabolism Controls the Shape and
Function of the Golgi Cisternae. Elife 6. doi:10.7554/eLife.24603
Cao, H., Su, S., Yang, Q., Le, Y., Chen, L., Hu, M., et al. (2021). Metabolic Profiling
Reveals interleukin-17A Monoclonal Antibody Treatment Ameliorate Lipids
Metabolism with the Potentiality to Reduce Cardiovascular Risk in Psoriasis
Patients. Lipids Health Dis. 20 (1), 16. doi:10.1186/s12944-021-01441-9
Chaurasia, B., Kaddai, V. A., Lancaster, G. I., Henstridge, D. C., Sriram, S., Galam,
D. L. A., et al. (2016). Adipocyte Ceramides Regulate Subcutaneous Adipose
Browning, Inflammation, and Metabolism. Cell. Metab. 24 (6), 820–834. doi:10.
1016/j.cmet.2016.10.002
Chocian, G., Chabowski, A., Zendzian-Piotrowska, M., Harasim, E., Łukaszuk, B.,
and Górski, J. (2010). High Fat Diet Induces Ceramide and Sphingomyelin
Formation in Rat’s Liver Nuclei. Mol. Cell. Biochem. 340 (1-2), 125–131. doi:10.
1007/s11010-010-0409-6
Colaco,K.,Lee,K.A.,Akhtari,S.,Winer,R.,Welsh,P.,Sattar,N.,etal.
(2021). Targeted Metabolomic Profiling and Prediction of Cardiovascular
Events: a Prospective Study of Patients with Psoriatic Arthritis and
Psoriasis. Ann. Rheum. Dis. doi:10.1136/annrheumdis-2021-220168
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 94591712
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
da Silva, M. T., Mujica-Coopman, M. F., Figueiredo, A. C. C., Hampel, D., Vieira, L.
S., Farias, D. R., et al. (2020). Maternal Plasma Folate Concentration Is
Positively Associated with Serum Total Cholesterol and Low-Density
Lipoprotein across the Three Trimesters of Pregnancy. Sci. Rep. 10 (1),
20141. doi:10.1038/s41598-020-77231-7
Donnelly,D.,3rd,Aung,P.P.,andJour,G.(2019).The"-OMICS"FacetofMelanoma:
Heterogeneity of Genomic, Proteomic and Metabolomic Biomarkers. Seminars
Cancer Biol. 59, 165–174. doi:10.1016/j.semcancer.2019.06.014
Dutkiewicz, E. P., Hsieh, K.-T., Wang, Y.-S., Chiu, H.-Y., and Urban, P. L. (2016).
Hydrogel Micropatch and Mass Spectrometry-Assisted Screening for Psoriasis-
Related Skin Metabolites. Clin. Chem. 62 (8), 1120–1128. doi:10.1373/clinchem.
2016.256396
Elmets, C. A., Leonardi, C. L., Davis, D. M. R., Gelfand, J. M., Lichten, J., Mehta, N.
N., et al. (2019). Joint AAD-NPF Guidelines of Care for the Management and
Treatment of Psoriasis with Awareness and Attention to Comorbidities. J. Am.
Acad. Dermatology 80 (4), 1073–1113. doi:10.1016/j.jaad.2018.11.058
Gomez-Larrauri, A., Presa, N., Dominguez-Herrera, A., Ouro, A., Trueba,
M.,andGomez-Muñoz,A.(2020).RoleofBioactiveSphingolipidsin
Physiology and Pathology. Essays Biochem. 64 (3), 579–589. doi:10.1042/
ebc20190091
Guerra, I., and Gisbert, J. P. (2013). Onset of Psoriasis in Patients with
Inflammatory Bowel Disease Treated with Anti-TNF Agents. Expert Rev.
Gastroenterology Hepatology 7 (1), 41–48. doi:10.1586/egh.12.64
Hasegawa, H., Lei, J., Matsumoto, T., Onishi, S., Suemori, K., and Yasukawa, M.
(2011). Lysophosphatidylcholine Enhances the Suppressive Function of
Human Naturally Occurring Regulatory T Cells through TGF-βProduction.
Biochem. Biophysical Res. Commun. 415 (3), 526–531. doi:10.1016/j.bbrc.2011.
10.119
Herbert, D., Franz, S., Popkova, Y., Anderegg, U., Schiller, J., Schwede, K., et al.
(2018). High-Fat Diet Exacerbates Early Psoriatic Skin Inflammation
Independent of Obesity: Saturated Fatty Acids as Key Players.
J. Investigative Dermatology 138 (9), 1999–2009. doi:10.1016/j.jid.2018.03.1522
Hung, N. D., Sok, D.-E., and Kim, M. R. (2012). Prevention of 1-palmitoyl
Lysophosphatidylcholine-Induced Inflammation by Polyunsaturated Acyl
Lysophosphatidylcholine. Inflamm. Res. 61 (5), 473–483. doi:10.1007/
s00011-012-0434-x
Iqbal, J., Walsh, M. T., Hammad, S. M., and Hussain, M. M. (2017). Sphingolipids
and Lipoproteins in Health and Metabolic Disorders. Trends Endocrinol.
Metabolism 28 (7), 506–518. doi:10.1016/j.tem.2017.03.005
Jain, S. (2017). Dermatology: Illustrated Study Guide and Comprehensive Board
Review. 2nd ed. Switzerland: Springer Cham, 83–87.
James, W. D. (2005). Andrews’Diseases of the Skin: Clinical Dermatology. 10th ed.
USA: Saunders, 191–197.
Jeon, S., Song, J., Lee, D., Kim, G.-T., Park, S.-H., Shin, D.-Y., et al. (2020).
Inhibition of Sphingosine 1-phosphate Lyase Activates Human Keratinocyte
Differentiation and Attenuates Psoriasis in Mice. J. Lipid Res. 61 (1), 20–32.
doi:10.1194/jlr.ra119000254
Jia, Y., Gan, Y., He, C., Chen, Z., and Zhou, C. (2018). The Mechanism of Skin
Lipids Influencing Skin Status. J. Dermatological Sci. 89 (2), 112–119. doi:10.
1016/j.jdermsci.2017.11.006
Kadowaki, H., Patton, G., and Robins, S. (1993). Effect of Phosphatidylcholine
Molecular Species on the Uptake of HDL Triglycerides and Cholesteryl Esters
by the Liver. J. Lipid Res. 34 (2), 180–189. doi:10.1016/s0022-2275(20)40745-x
Kamiya, K., Kishimoto, M., Sugai, J., Komine, M., and Ohtsuki, M. (2019). Risk
Factors for the Development of Psoriasis. Int. J. Mol. Sci. 20 (18). doi:10.3390/
ijms20184347
Kamleh, M. A., Snowden, S. G., Grapov, D., Blackburn, G. J., Watson, D. G., Xu, N.,
et al. (2015). LC-MS Metabolomics of Psoriasis Patients Reveals Disease
Severity-dependent Increases in Circulating Amino Acids that Are
Ameliorated by Anti-TNFαTreatment. J. Proteome Res. 14 (1), 557–566.
doi:10.1021/pr500782g
Kang, H., Li, X., Zhou, Q., Quan, C., Xue, F., Zheng, J., et al. (2017). Exploration of
Candidate Biomarkers for Human Psoriasis Based on Gas Chromatography-
Mass Spectrometry Serum Metabolomics. Br. J. Dermatol 176 (3), 713–722.
doi:10.1111/bjd.15008
Kennedy, E. P., and Weiss, S. B. (1956). The Function of Cytidine Coenzymes in the
Biosynthesis of Phospholipides. J. Biol. Chem. 222 (1), 193–214. doi:10.1016/
s0021-9258(19)50785-2
Kim, H.-N., Han, K., Park, Y.-G., and Lee, J. H. (2019). Metabolic Syndrome Is
Associated with an Increased Risk of Psoriasis: A Nationwide Population-Based
Study. Metabolism 99, 19–24. doi:10.1016/j.metabol.2019.07.001
Kolak, M., Westerbacka, J., Velagapudi, V. R., Wa˚gsa€ter, D., Yetukuri, L.,
Makkonen, J., et al. (2007). Adipose Tissue Inflammation and Increased
Ceramide Content Characterize Subjects with High Liver Fat Content
Independent of Obesity. Diabetes 56 (8), 1960–1968. doi:10.2337/db07-0111
Kromhout, D., and de Goede, J. (2014). Update on Cardiometabolic Health Effects
of ω-3 Fatty Acids. Curr. Opin. Lipidol. 25 (1), 85–90. doi:10.1097/mol.
0000000000000041
Kuai, R., Li, D., Chen, Y. E., Moon, J. J., and Schwendeman, A. (2016). High-
Density Lipoproteins: Nature’s Multifunctional Nanoparticles. ACS Nano 10
(3), 3015–3041. doi:10.1021/acsnano.5b07522
Laaksonen, R., Ekroos, K., Sysi-Aho, M., Hilvo, M., Vihervaara, T., Kauhanen, D.,
et al. (2016). Plasma Ceramides Predict Cardiovascular Death in Patients with
Stable Coronary Artery Disease and Acute Coronary Syndromes beyond LDL-
Cholesterol. Eur. Heart J. 37 (25), 1967–1976. doi:10.1093/eurheartj/ehw148
Leventhal, A. R., Chen, W., Tall, A. R., and Tabas, I. (2001). Acid
Sphingomyelinase-Deficient Macrophages Have Defective Cholesterol
Trafficking and Efflux. J. Biol. Chem. 276 (48), 44976–44983. doi:10.1074/
jbc.m106455200
Li, L., Yao, D.-n., Lu, Y., Deng, J.-w., Wei, J.-a., Yan, Y.-h., et al. (2020).
Metabonomics Study on Serum Characteristic Metabolites of Psoriasis
Vulgaris Patients with Blood-Stasis Syndrome. Front. Pharmacol. 11,
558731. doi:10.3389/fphar.2020.558731
Liu, L., Wang, J., Li, H.-j., Zhang, S., Jin, M.-z., Chen, S.-t., et al. (2021).
Sphingosine-1-Phosphate and its Signal Modulators Alleviate Psoriasis-like
Dermatitis: Preclinical and Clinical Evidence and Possible Mechanisms. Front.
Immunol. 12, 759276. doi:10.3389/fimmu.2021.759276
Liu, P., Zhu, W., Chen, C., Yan, B., Zhu, L., Chen, X., et al. (2020). The Mechanisms
of Lysophosphatidylcholine in the Development of Diseases. Life Sci. 247,
117443. doi:10.1016/j.lfs.2020.117443
Łuczaj, W., Dobrzyńska, I., Wroński, A., Domingues, M. R., Domingues, P., and
Skrzydlewska, E. (2020). Cannabidiol-Mediated Changes to the Phospholipid
Profile of UVB-Irradiated Keratinocytes from Psoriatic Patients. Int. J. Mol. Sci.
21 (18)6592. doi:10.3390/ijms21186592
Martínez-Uña, M., Varela-Rey, M., Cano, A., Fernández-Ares, L., Beraza, N.,
Aurrekoetxea, I., et al. (2013). Excess S-Adenosylmethionine Reroutes
Phosphatidylethanolamine towards Phosphatidylcholine and Triglyceride
Synthesis. Hepatology 58 (4), 1296–1305. doi:10.1002/hep.26399
Miura, H., Mizuguchi, H., Amano-Iwashita, M., Maeda-Kogure, R., Negishi, A.,
Sakai, A., et al. (2021). Clofibric Acid Increases Molecular Species of
Phosphatidylethanolamine Containing Arachidonic Acid for Biogenesis of
Peroxisomal Membranes in Peroxisome Proliferation in the Liver.
Biochimica Biophysica Acta (BBA) - Mol. Cell. Biol. Lipids 1866 (8), 158963.
doi:10.1016/j.bbalip.2021.158963
Myśliwiec, H., Baran, A., Harasim-Symbor, E., Choromańska, B., Myśliwiec, P.,
Milewska, A. J., et al. (2017). Increase in Circulating Sphingosine-1-Phosphate
and Decrease in Ceramide Levels in Psoriatic Patients. Arch. Dermatol Res. 309
(2), 79–86. doi:10.1007/s00403-016-1709-9
Nicholas, D. A., Proctor, E. A., Agrawal, M., Belkina, A. C., Van Nostrand, S. C.,
Panneerseelan-Bharath, L., et al. (2019). Fatty Acid Metabolites Combine with
Reduced βOxidation to Activate Th17 Inflammation in Human Type
2 Diabetes. Cell. Metab. 30 (3), 447–461.e5. doi:10.1016/j.cmet.2019.07.004
Norris, G. H., Milard, M., Michalski, M.-C., and Blesso, C. N. (2019). Protective
Properties of Milk Sphingomyelin against Dysfunctional Lipid Metabolism, Gut
Dysbiosis, and Inflammation. J. Nutr. Biochem. 73, 108224. doi:10.1016/j.
jnutbio.2019.108224
Norris, G. H., Porter, C. M., Jiang, C., Millar, C. L., and Blesso, C. N. (2017). Dietary
Sphingomyelin Attenuates Hepatic Steatosis and Adipose Tissue Inflammation
in High-Fat-Diet-Induced Obese Mice. J. Nutr. Biochem. 40, 36–43. doi:10.
1016/j.jnutbio.2016.09.017
Nowowiejska, J., Baran, A., and Flisiak, I. (2021). Aberrations in Lipid Expression
and Metabolism in Psoriasis. Int. J. Mol. Sci. 22 (12). doi:10.3390/ijms22126561
Ohigashi, T., Kanno, K., Sugiyama, A., Nguyen, P. T., Kishikawa, N., Otani, Y., et al.
(2019). Protective Effect of Phosphatidylcholine on Lysophosphatidylcholine-
induced Cellular Senescence in Cholangiocyte. J. Hepatobiliary Pancreat. Sci. 26
(12), 568–577. doi:10.1002/jhbp.684
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 94591713
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
Ottas, A., Fishman, D., Okas, T.-L., Kingo, K., and Soomets, U. (2017). The
Metabolic Analysis of Psoriasis Identifies the Associated Metabolites while
Providing Computational Models for the Monitoring of the Disease. Arch.
Dermatol Res. 309 (7), 519–528. doi:10.1007/s00403-017-1760-1
Raychaudhuri, S. K., Maverakis, E., and Raychaudhuri, S. P. (2014). Diagnosis and
Classification of Psoriasis. Autoimmun. Rev. 13 (4-5), 490–495. doi:10.1016/j.
autrev.2014.01.008
Samad, F., Hester, K. D., Yang, G., Hannun, Y. A., and Bielawski, J. (2006). Altered
Adipose and Plasma Sphingolipid Metabolism in Obesity. Diabetes 55 (9),
2579–2587. doi:10.2337/db06-0330
Shoda, J., Ueda, T., Ikegami, T., Matsuzaki, Y., Satoh, S., Kano, M., et al. (1997).
Increased Biliary Group II Phospholipase A2 and Altered Gallbladder Bile in
Patients with Multiple Cholesterol Stones. Gastroenterology 112 (6), 2036–2047.
doi:10.1053/gast.1997.v112.pm9178697
Simopoulos, A. (2016). An Increase in the Omega-6/Omega-3 Fatty Acid
Ratio Increases the Risk for Obesity. Nutrients 8 (3), 128. doi:10.3390/
nu8030128
Simopoulos, A. (2013). Dietary Omega-3 Fatty Acid Deficiency and High Fructose
Intake in the Development of Metabolic Syndrome, Brain Metabolic
Abnormalities, and Non-alcoholic Fatty Liver Disease. Nutrients 5 (8),
2901–2923. doi:10.3390/nu5082901
Sindhu, S., Leung, Y. H., Arefanian, H., Madiraju, S. R. M., Al-Mulla, F., Ahmad, R.,
et al. (2021). Neutral Sphingomyelinase-2 and Cardiometabolic Diseases. Obes.
Rev. 22 (8), e13248. doi:10.1111/obr.13248
Sorokin,A.V.,Domenichiello,A.F.,Dey,A.K.,Yuan,Z.-X.,Goyal,A.,Rose,S.M.,
et al. (2018). Bioactive Lipid Mediator ProfilesinHumanPsoriasisSkinandBlood.
J. Investigative Dermatology 138 (7), 1518–1528. doi:10.1016/j.jid.2018.02.003
Sprenger, R. R., Hermansson, M., Neess, D., Becciolini, L. S., Sørensen, S. B.,
Fagerberg, R., et al. (2021). Lipid Molecular Timeline Profiling Reveals Diurnal
Crosstalk between the Liver and Circulation. Cell. Rep. 34 (5), 108710. doi:10.
1016/j.celrep.2021.108710
Summers, S. (2006). Ceramides in Insulin Resistance and Lipotoxicity. Prog. Lipid
Res. 45 (1), 42–72. doi:10.1016/j.plipres.2005.11.002
Sundler, R., and Akesson, B. (1975). Regulation of Phospholipid Biosynthesis in
Isolated Rat Hepatocytes. Effect of Different Substrates. J. Biol. Chem. 250 (9),
3359–3367. doi:10.1016/s0021-9258(19)41523-8
Takeshita, J., Grewal, S., Langan, S. M., Mehta, N. N., Ogdie, A., Van Voorhees, A.
S., et al. (2017). Psoriasis and Comorbid Diseases. J. Am. Acad. Dermatology 76
(3), 377–390. doi:10.1016/j.jaad.2016.07.064
Tarentini, E., Odorici, G., Righi, V., Paganelli, A., Giacomelli, L., Mirisola, V., et al.
(2021). Integrated Metabolomic Analysis and Cytokine Profiling Define
Clusters of Immuno-Metabolic Correlation in New-Onset Psoriasis. Sci. Rep.
11 (1), 10472. doi:10.1038/s41598-021-89925-7
Tsuji,T.,Yuri,T.,Terada,T.,andMorita,S.-y.(2021).ApplicationofEnzymatic
Fluorometric Assays to Quantify Phosphatidylcholine, Phosphatidylethanolamine
and Sphingomyelin in Human Plasma Lipoproteins. Chem. Phys. Lipids 238,
105102. doi:10.1016/j.chemphyslip.2021.105102
van der Veen,J. N., Kennelly, J. P., Wan,S., Vance, J. E., Vance,D. E., and Jacobs, R. L.
(2017). The Critical Role of Phosphatidylcholine and Phosphatidylethanolamine
Metabolism in Health and Disease. Biochim. Biophys. Acta Biomembr. 1859 (9 Pt
B), 1558–1572. doi:10.1016/j.bbamem.2017.04.006
Vandanmagsar, B., Youm, Y.-H., Ravussin, A., Galgani, J. E., Stadler, K., Mynatt, R.
L., et al. (2011). The NLRP3 Inflammasome Instigates Obesity-Induced
Inflammation and Insulin Resistance. Nat. Med. 17 (2), 179–188. doi:10.
1038/nm.2279
Young, K. E., Flaherty, S., Woodman, K. M., Sharma-Walia, N., and Reynolds, J. M.
(2017). Fatty Acid Synthase Regulates the Pathogenicity of Th17 Cells.
J. Leukoc. Biol. 102 (5), 1229–1235. doi:10.1189/jlb.3ab0417-159rr
Zeng, C., Wen, B., Hou, G., Lei, L., Mei, Z., Jia, X., et al. (2017). Lipidomics Profiling
Reveals the Role of Glycerophospholipid Metabolism in Psoriasis. Gigascience 6
(10), 1–11. doi:10.1093/gigascience/gix087
Conflict of Interest: MW was employed by SU BioMedicine, BioPartner Center 3,
Leiden, Netherlands.
The remaining authors declare that the research was conducted in the absence of
any commercial or financial relationships that could be construed as a potential
conflict of interest.
Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations or those of
the publisher, the editors, and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.
Copyright © 2022 Dai, He, Wang, Wang, Guo and Song. This is an open-access
article distributed under the terms of the Creative Commons Attribution License (CC
BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice.
No use, distribution or reproduction is permitted which does not comply with these
terms.
Frontiers in Molecular Biosciences | www.frontiersin.org July 2022 | Volume 9 | Article 94591714
Dai et al. Molecular Subtyping of Psoriasis Vulgaris
