Abstract
Myasthenia gravis (MG) is an autoimmune disorder resulting from
antibodies against the proteins at the neuromuscular junction. Emerging
evidence indicates that long non-coding RNAs (lncRNAs), acting as
competing endogenous RNAs (ceRNAs), are involved in various diseases.
However, the regulatory mechanisms of ceRNAs underlying MG remain
largely unknown. In this study, we constructed a lncRNA-mediated ceRNA
network involved in MG using a multi-step computational strategy.
Functional annotation analysis suggests that these lncRNAs may play
crucial roles in the immunological mechanism underlying MG.
Importantly, through manual literature mining, we found that lncRNA
SNHG16 (small nucleolar RNA host gene 16), acting as a ceRNA, plays
important roles in the immune processes. Further experiments showed
that SNHG16 expression was upregulated in peripheral blood mononuclear
cells (PBMCs) from MG patients compared to healthy controls. Luciferase
reporter assays confirmed that SNHG16 is a target of the microRNA
(miRNA) let-7c-5p. Subsequent experiments indicated that SNHG16
regulates the expression of the key MG gene interleukin (IL)-10 by
sponging let-7c-5p in a ceRNA manner. Furthermore, functional assays
showed that SNHG16 inhibits Jurkat cell apoptosis and promotes cell
proliferation by sponging let-7c-5p. Our study will contribute to a
deeper understanding of the regulatory mechanism of MG and will
potentially provide new therapeutic targets for MG patients.
Keywords: myasthenia gravis, ceRNA network, SNHG16, IL-10, let-7c-5p
Introduction
Myasthenia gravis (MG) is an autoimmune disorder caused by antibodies
that attack proteins of the postsynaptic membrane at the neuromuscular
junction, leading to muscle weakness and abnormal fatigability.[49]^1
Most MG patients have detectable antibodies against the acetylcholine
receptor (AChR), while a small group of patients have antibodies
against muscle-specific kinase (MuSK) or lipoprotein receptor-related
protein 4 (LRP4).[50]^2 The production of antibodies is a
T cell-dependent and B cell-mediated process. Cytokines produced by
immune cells are crucial regulators of the pathogenesis of MG. For
example, cytokines such as interleukin (IL)-2, interferon (IFN)-γ, and
tumor necrosis factor (TNF)-α secreted by T helper (Th) cells stimulate
the production of pathogenic antibodies, which are mediated by B
cells.[51]^3 Moreover, both the expression of cytokines, such as IL-2,
IFN-γ, and IL-10, and the number of peripheral blood mononuclear cells
(PBMCs) secreting these cytokines are higher in patients with MG.[52]4,
[53]5, [54]6 Increasing evidence shows that noncoding RNAs (ncRNAs) are
involved in the regulation of gene expression in the immune system, and
thus they provide novel insight into the pathogenesis, diagnosis, and
treatment of MG.[55]^7
Long ncRNAs (lncRNAs; more than 200 nt) interact with DNA, RNA, or
proteins to modulate the expression of protein-coding genes, which play
important roles in various biological processes, including the
regulation of immune responses.[56]^7 It has been reported that
aberrant expression of lncRNA IFNG-AS1 in PBMCs regulates CD4^+ T cell
activation in MG patients partly by influencing human leukocyte antigen
(HLA)-DRB1 expression.[57]^8 MicroRNAs (miRNAs; ∼22 nt) are small
functional ncRNA molecules that repress target gene expression and are
involved in a wide range of biological processes.[58]^9 Accumulating
evidence indicates that miRNAs contribute to the pathogenesis of MG by
regulating important genes. For example, downregulation of miR-320a in
MG patients induces inflammatory cytokine production by targeting
mitogen-activated protein kinase 1 (MAPK1).[59]^10 miR-15a is
downregulated in MG patients and causes abnormal activation of the
immune response by regulating IFN-γ-induced protein 10 (IP-10), which
is a highly inducible chemoattractant causing secretion of more IFN-γ
by activated Th1 cells.[60]^11 Although some miRNAs and lncRNAs have
been found to be implicated in MG, few studies have focused on
exploring the interactions among miRNAs, lncRNAs, and genes underlying
the pathogenesis of MG.
Recently, the competing endogenous RNA (ceRNA) theory was proposed,
pointing out that RNA molecules containing miRNA response elements
(MREs) could compete with each other by binding to a common
miRNA.[61]^12 There is increasing evidence that lncRNAs function as
ceRNAs, competing with mRNAs by acting as sponges of miRNAs, which
relieve miRNA-mediated target repression.[62]^13 lncRNAs, acting as
ceRNAs, are implicated in various diseases. For example, the lncRNA
ZNFX1-AS1 has been shown to play an important role in the progression
and metastasis of colorectal cancer by acting as a ceRNA of miR-144,
thereby leading to the depression of its endogenous target gene
polycomb group protein enhancer of zeste homolog 2 (EZH2).[63]^14
Additionally, lncRNAs, acting as ceRNAs, also play crucial roles in the
regulation of the immune system and the development of autoimmune
diseases. lncRNA MEG3 regulates RORγt expression by sequestering
miR-17, which affects the regulatory T (Treg)/Th17 balance in
asthma.[64]^15 Despite advances in ceRNA regulation, their potential
roles in MG remain largely unknown. Thus, there is an urgent need to
explore the ceRNA regulatory mechanism of MG and to develop novel
biomarkers for the diagnosis and treatment of MG.
In the present study, we first constructed a global lncRNA-mediated
ceRNA network involved in MG using a multi-step computational approach.
Functional enrichment analysis was performed to reveal the potential
roles of these lncRNAs in the network. Through analyzing the network
and reviewing reliable publications, we designed biological experiments
to further confirm the presence of an SNHG16 (small nucleolar RNA host
gene 16)-mediated ceRNA regulatory mechanism through the
let-7c-5p/IL-10 axis in MG. Our study provides novel insights into the
ceRNA network and reveals potential roles of SNHG16 in MG.
Results
Construction of the lncRNA-Mediated MG-Associated ceRNA Network
It is well known that lncRNAs can act as ceRNAs by binding
miRNAs.[65]^13 To evaluate the lncRNA-mediated ceRNA regulation in MG,
an lncRNA-mediated MG-associated ceRNA network (LMGCN) was constructed
using a multi-step approach ([66]Figure 1A). As a result, the LMGCN
contained 9 miRNAs, 20 genes, 32 lncRNAs, and 147 edges
([67]Figure 1B). We referred to the previous study[68]^16 to count the
number of the primary relationship pairs of lncRNA-miRNA and the
secondary relationship pairs of miRNA-mRNA ([69]Table S1). For a given
lncRNA(i), the number of the primary relationship pairs of lncRNA-miRNA
represents the number of miRNAs linked with lncRNA(i), and the number
of the secondary relationship pairs of miRNA-mRNA represents the sum of
mRNAs linked with the above miRNAs. HELLPAR had the highest total
number of lncRNA-miRNA and miRNA-mRNA pairs, which may play important
roles in the network.
Figure 1.
[70]Figure 1
[71]Open in a new tab
The lncRNA-Mediated MG-Associated ceRNA Network (LMGCN)
(A) Schematic workflow to construct the LMGCN based on the ceRNA
hypothesis. (B) LMGCN. Red circles represent miRNAs, blue triangles
represent mRNAs, and yellow rhombi represent lncRNAs. Lines represent
their regulatory interactions. (C) Pathway enrichment analysis (left)
and GO annotation (right) of co-expressed mRNAs with lncRNAs.
To further explore the roles of lncRNAs, we first used RNALocate[72]^17
to investigate the subcellular localization of each lncRNA in the LMGCN
([73]Table S2). RNALocate is a comprehensive database that provides
experimentally supported high-quality RNA subcellular localizations.
Then, we performed functional annotation analysis using the
co-expressed mRNA. 58 pathways and 113 Gene Ontology (GO) terms were
identified (p < 0.05). Moreover, more than one third of the identified
terms and pathways are relevant to immune or inflammatory mechanism
([74]Table S3). We mainly listed several immune-related GO functions
and pathways that play important roles in the immunological mechanism
of MG ([75]Figure 1C). For example, the significant GO functions
included immune response, positive regulation of B cell proliferation,
thymus development, positive regulation of T cell proliferation, and
inflammatory response. The significant pathways included
cytokine-cytokine receptor interaction, T cell receptor signaling
pathway, and Toll-like receptor signaling pathway. These findings
highlighted the fundamental characteristics of these lncRNAs in the
pathogenesis of MG.
Dissection of Potential ceRNA Mechanisms through Manual Literature Mining
Based on the LMGCN, four miRNA-gene regulation pairs had reportedly
been verified in MG: let-7c-5p/IL-10,[76]^18 miR-145-5p/CD28,[77]^19
miR-15a-5p/CXCL10,[78]^11 and miR-181c-5p/IL7.[79]^20 We extracted the
sub-networks of these four miRNA-gene pairs and their linked lncRNAs in
the global triple network. We found eight lncRNAs that may act as
ceRNAs to regulate the miRNA-gene pairs mentioned above. Through
reviewing reliable publications, we summarized the functions of these
lncRNAs ([80]Figure 2; the detailed information is summarized in
[81]Table S4). These lncRNAs are mainly involved in the development of
various cancers,[82]21, [83]22, [84]23 with other functions including
immune processes,[85]^24 neuroprotective effects in
ischemia/reperfusion injury,[86]^25 and vascular calcification.[87]^26
More importantly, SNHG16 acting as a ceRNA was reported to be involved
in inflammatory and immune processes[88]^24 that may participate in the
pathogenesis of MG. We found that SNHG16 potentially competed with
IL-10 in the sub-network. IL-10 is an important growth factor for B
cells, augmenting B cells activation into antibody-producing cells, and
it plays important roles in the pathogenesis of MG.[89]^4 These
findings suggest that the IL-10/SNHG16 competing pair might be involved
in the immunological pathogenesis of MG. Therefore, we primarily
focused on SNHG16/let-7c-5p/IL-10 interaction for further experimental
verification.
Figure 2.
[90]Figure 2
[91]Open in a new tab
The Sub-Network of lncRNA Acting as a ceRNA to Regulate miRNA–Gene
Pairs Has Been Verified in MG.
(A) The sub-network of let-7c-5p/IL-10 pair. (B) The sub-network of
miR-145-5p/CD28 pair. (C) The sub-network of miR-15a-5p/CXCL10 pair.
(D)The sub-network of miR-181c-5p/IL-7 pair. The left pipeline are the
sub-networks of four experimentally-supported miRNA–gene pairs in MG
and their linked lncRNAs. The right pipeline are the summarized
functions of lncRNAs.
lncRNA SNHG16 Is Upregulated in MG Patients and Is a Target of let-7c-5p
Real-time PCR analysis was performed to examine lncRNA SNHG16 in PBMCs
from MG patients and control subjects. The expression of SNHG16 was
higher in patients with MG compared with controls (p = 0.004,
[92]Figure 3A). Bioinformatics analysis revealed that the SNHG16
sequence contains a putative let-7c-5p binding region. A previous study
had indicated that let-7c-5p was downregulated in PBMCs of MG
patients.[93]^18 To investigate the association between let-7c-5p and
SNHG16, let-7c-5p mimics were transfected into Jurkat cells. The
transfection efficiency of let-7c-5p mimics was confirmed by real-time
PCR (p = 0.0003, [94]Figure 3B). Overexpression of let-7c-5p
significantly inhibited the expression of SNHG16 in Jurkat cells (p <
0.01, [95]Figure 3C). To further confirm the direct interaction between
SNHG16 and let-7c-5p, we constructed luciferase reporter vectors of
SNHG16-wild type (WT) and SNHG16-mutated type (MUT) ([96]Figure 3D).
The SNHG16-WT or SNHG16-MUT was then co-transfected with let-7c-5p
mimics or negative control into HEK293T cells. The dual-luciferase
reporter assay showed that let-7c-5p mimics suppressed the luciferase
activity of SNHG16-WT but had no effect on the luciferase activity of
SNHG16-MUT ([97]Figure 3E). These results indicated that SNHG16 is the
target of let-7c-5p.
Figure 3.
[98]Figure 3
[99]Open in a new tab
Upregulation of SNHG16 Is a Target of let-7c-5p in MG
(A) SNHG16 expression was examined in 24 MG patients and 29 control
subjects by real-time PCR. (B) Transfection efficiency of let-7c-5p
mimics was measured by real-time PCR. (C) The relative expression level
of SNHG16 in Jurkat cells transfected with miRNA NC or let-7c-5p mimics
was measured using real-time PCR. (D) The putative let-7c-5p binding
sequence of the wild-type and mutation sequence of SNHG16. (E) The
luciferase reporter plasmid containing SNHG16-WT or SNHG16-MUT was
co-transfected with let-7c-5p mimics or miRNA NC into HEK293T cells.
Luciferase activities were calculated as the ratio of firefly/Renilla
activities. The experiment was repeated at least three times, and data
are presented as the mean ± SD. **p < 0.01.
SNHG16 Promotes IL-10 Expression by Sponging let-7c-5p
MG is a T cell-dependent and B cell-mediated autoimmune disease.
Cytokines play an important role in the regulation of autoantibody
production and cell-mediated immunity in MG.[100]^2 It has been
reported that there is a negative correlation between let-7c-5p and
IL-10 mRNA levels in MG patients, and that let-7c-5p mediates
regulation of IL-10 by directly targeting IL-10 in Jurkat
cells.[101]^18 Therefore, we designed experiments to explore whether
SNHG16 could regulate IL-10 expression by targeting let-7c-5p. First,
we measured the expression of IL-10 mRNA and protein in Jurkat cells
after transfection with let-7c-5p mimics or negative control. The
results indicated that let-7c-5p overexpression decreased IL-10
expression at both the mRNA and protein level (p < 0.01, [102]Figures
4A and 4B), which was consistent with the results of a previous
study.[103]^18
Figure 4.
[104]Figure 4
[105]Open in a new tab
SNHG16 Regulates IL-10 Expression by Binding let-7c-5p in a ceRNA
Manner
(A) Relative mRNA levels of IL-10 were determined by real-time PCR
after transfection with negative control or let-7c-5p mimics in Jurkat
cells. (B) Relative protein expression levels of IL-10 were determined
by western blotting after transfection with negative control or
let-7c-5p mimics in Jurkat cells. (C) Relative mRNA levels of IL-10
were determined by real-time PCR analysis after transfection with
negative control, siSNHG16, and siSNHG16 + let-7c-5p inhibitor in
Jurkat cells. (D) Relative protein expression levels of IL-10 were
determined by western blotting after transfection with negative
control, siSNHG16, and siSNHG16 + let-7c-5p inhibitor in Jurkat cells.
The experiment was repeated at least three times, and data are
presented as the mean ± SD. **p < 0.01.
To verify whether SNHG16 regulates IL-10 expression by targeting
let-7c-5p, Jurkat cells were transfected with negative control,
siSNHG16, and siSNHG16 in combination with let-7c-5p inhibitor. Then,
the IL-10 mRNA and protein levels were determined. The results showed
that knockdown of SNHG16 suppressed the IL-10 mRNA and protein
expression levels in Jurkat cells, whereas let-7c-5p inhibitor blocked
the reduction in IL-10 expression induced by SNHG16 suppression (p <
0.01, [106]Figures 4C and 4D). These findings suggest that SNHG16
regulates the expression of IL-10 by sponging let-7c-5p in a ceRNA
manner.
SNHG16 Inhibits Apoptosis and Promotes Proliferation by Sponging let-7c-5p in
Jurkat Cells
MG is a T cell-dependent autoimmune disease, and the proliferation and
activation of T cells play an important role in the pathogenesis of MG.
Therefore, we used Jurkat T cells for functional verification of MG,
referring to previous studies.[107]^10^,[108]^18 To determine whether
SNHG16 affects apoptosis and proliferation of Jurkat cells by targeting
let-7c-5p, Jurkat cells were transfected with negative control,
siSNHG16, and siSNHG16 along with let-7c-5p inhibitor. Then, flow
cytometry and a Cell Counting Kit-8 (CCK-8) assay were used to assess
the rates of apoptosis and proliferation. The rate of apoptosis
increased following transfection with siSNHG16, whereas the addition of
let-7c-5p inhibitor abrogated the above trend (p < 0.01, [109]Figures
5A and 5B). Moreover, the CCK-8 assay revealed that knockdown of SNHG16
inhibited proliferation of Jurkat cells compared with transfection with
negative control, and this effect was reversed by transfecting with
let-7c-5p inhibitor (p < 0.01, [110]Figure 5C). These results suggest
that SNHG16 inhibits cell apoptosis and promotes proliferation by
sponging let-7c-5p in Jurkat cells. Taken together, our results suggest
that SNHG16 regulates T cell apoptosis and proliferation by sponging
let-7c-5p, which is involved in the immunological pathogenesis of MG
([111]Figure 5D).
Figure 5.
[112]Figure 5
[113]Open in a new tab
SNHG16 Inhibits Apoptosis and Promotes Cell Proliferation by Sponging
let-7c-5p
(A) After transfecting negative control, siSNHG16 or siSNHG16+let-7c-5p
inhibitor, Jurkat cells were stained with Annexin-V-FITC/PI, and the
apoptosis was detected by flow cytometric analysis. (B) The apoptosis
rate of Jurkat cells after transfecting negative control, siSNHG16 or
siSNHG16+let-7c-5p inhibitor. (C) Cell proliferation was analyzed using
CCK-8 assays by transfecting negative control, siSNHG16, or siSNHG16 +
let-7c-5p inhibitor into Jurkat cells. (D) Schematic diagram of SNHG16
involved in the immunological pathogenesis of MG. The experiment was
repeated at least three times, and data are presented as the mean ± SD.
**p < 0.01.
Discussion
MG is a neurological autoantibody-mediated disease, but the triggering
autoimmune processes involved are not clearly defined. Immunomodulatory
therapies have been widely used to improve the prognosis for MG
patients. Given the complex pathogenesis and heterogeneity of MG, no
one treatment therapy is best for all MG patients.[114]^27 Extensive
evidence indicates that lncRNAs play critical roles in regulation of
the immune system and in autoimmune disease.[115]^7 The recently
uncovered lncRNA-mediated ceRNA regulatory theory and networks improve
our understanding of the precise molecular mechanism of MG.
During the past few years, ceRNA regulatory mechanisms have been
validated in various diseases. To date, several databases have been
developed to curate ceRNA interactions based on experimentally
supported evidence or computationally predicted methods, such as
LncACTdb,[116]^28 starBase,[117]^29 and PceRBase[118]^30. MNDR
v2.0[119]^31 and RAID v2.0[120]^32 are also useful resources for
analyzing RNA-disease associations. These databases are valuable
resources for studying ceRNA regulation underlying complex diseases.
However, so far, most studies concerning ceRNA network construction and
mechanisms have focused on the cancer field. For example, Wang
et al.[121]^33 constructed a lncRNA-associated ceRNA network to reveal
global patterns and prognostic markers across 12 types of human cancer.
However, only a few ceRNA interactions have been reported in autoimmune
diseases.[122]^16 Accordingly, there is a need to explore the
regulatory mechanisms of ceRNAs in MG. In the present study, we first
constructed an LMGCN based on the ceRNA theory using a comprehensive
approach. The LMGCN was composed of 9 miRNA nodes, 20 mRNA nodes, and
32 lncRNA nodes. The results of functional annotation analysis suggest
that these lncRNAs may play crucial roles in the development of MG.
Next, we summarized the functions of lncRNA as a ceRNA to regulate
miRNA-gene interactions that have been verified in MG. Of note, a
recent study showed that SNHG16 affected the LPS-induced inflammatory
and immune processes. Subsequently, mechanistic investigations revealed
that SHNG16 acts as a ceRNA to positively regulate Toll-like receptor 4
(TLR4) by competitively binding miR-15a/16.[123]^24 Indeed, growing
evidence shows that SNHG16 serving as a ceRNA is involved in several
diseases.[124]^34^,[125]^35 However, the underlying mechanism of the
involvement of SNHG16 in MG remains unclear. The present study found
that SNHG16 expression is increased in PBMCs from MG patients.
Functionally, we found that knockdown of SNHG16 suppresses cell
proliferation and promotes apoptosis in Jurkat cells. These results
suggest that SNHG16 might be involved in the immunological pathogenesis
of MG.
Bioinformatics analysis predicted that SNHG16 was a direct target of
let-7c-5p. A previous study has reported that let-7c-5p is
downregulated in PBMCs from MG patients, and that IL-10 is a target for
let-7c-5p. Meanwhile, let-7c regulates IL-10 secretion in Jurkat
cells.[126]^18 We found that let-7c-5p overexpression decreased IL-10
mRNA and protein expression levels in Jurkat cells, findings that were
consistent with the results of previous study. Hence, we hypothesized
that SHNG16 acted as a ceRNA to regulate the let-7c-5p/IL-10 axis in MG
([127]Figure 5D). We found that transfection of let-7c-5p significantly
decreased SNHG16 expression in Jurkat cells. Then, we confirmed that
SNHG16 is a direct target of let-7c-5p by using luciferase reporter
assays. Furthermore, we found that SNHG16 knockdown suppressed the
IL-10 mRNA and protein levels, but these effects were reversed by
co-transfecting siSHNG16 and let-7c-5p inhibitor into Jurkat cells.
Moreover, cell proliferation and apoptosis assays showed that knockdown
of SNHG16 inhibited cell proliferation and promoted cell apoptosis, but
these were reversed by co-transfecting siSHNG16 and let-7c-5p inhibitor
into Jurkat cells. Taken together, our findings suggest that lncRNA
SNHG16 functions as a ceRNA to regulate IL-10 by competitively binding
let-7c-5p.
Recently, cytokines that drive autoantibody secretion have received
widespread attention as specific immunotherapy for MG. Various
cytokines have been reported to play a critical role in the
immunopathogenic mechanisms of MG.[128]^36 Some anti-cytokine agents
relevant to MG immunopathogenesis have provided new targeted
immunotherapies.[129]^37 For example, etanercept (an antagonist of
TNF-α) has been proposed to treat patients with MG.[130]^38 IL-10 plays
a crucial role in B cell activation and autoantibody
production.[131]^39 It has been found that MG patients had increased
numbers of AChR-reactive IL-10 mRNA-expressing PBMCs.[132]^4^,[133]^40
High IL-10 levels were associated with AChR antibody production and
treatment response in juvenile MG.[134]^5 These findings support the
hypothesis that IL-10 plays important roles in the pathogenesis of MG.
The symptoms of MG are caused by the autoantibodies that attack the
receptors in the postsynaptic muscle membrane, leading to muscle
fatigue and weakness. The production of autoantibodies is a
T cell-dependent and B cell-mediated process. It has been reported that
IL-10 secreted by Th2 cells promotes B cell activation into
antibody-producing cells[135]^4 ([136]Figure 5D). Thus, the present
study showed that SNHG16 regulates the expression of IL-10 by sponging
let-7c-5p, and thus improved our understanding of the precise molecular
mechanism of MG.
In summary, in the present study, we for the first time constructed a
lncRNA-mediated ceRNA network involved in MG. Further experiments
revealed a novel regulatory mechanism of lncRNA SNHG16 involved in MG,
which regulates IL-10 expression by sponging let-7c-5p. Our findings
provide a global view of the mechanisms of ceRNA regulation and
candidate lncRNAs as biomarkers for diagnosis and therapies in MG.
Materials and Methods
Human MG Risk miRNA Collection
We mainly applied the following two steps in the data collection
process: searching for relevant articles, and extracting useful
information from the selected articles. To ensure a high quality, human
MG risk miRNA collection referred to our previous studies published
online.[137]^41 We first obtained all publications from the PubMed
database using the keyword combinations “myasthenia gravis” and
“microRNA” or “miRNA” or “miR” to collect all experimentally supported
MG-associated miRNAs. Then, we manually curated the MG-associated
miRNAs that met the following criteria: (1) the species was human; (2)
MG patients and healthy controls were included in the study, and at
least 10 samples were included in each group; (3) the expression levels
of miRNA were analyzed in PBMCs and showed significant differences
between the MG patients and controls; (4) selected miRNAs were
experimentally confirmed by reliable low-throughput experiments, such
as real-time PCR and northern blot; and (5) the collected miRNAs were
agreed to by at least two researchers. Finally, detailed information of
MG risk miRNAs that we collected is summarized in [138]Table S5.
Identification of miRNA-lncRNA and miRNA-mRNA Interactions
First, miRNA-lncRNA interactions were predicted using TargetScan
(release 7.2)[139]^42 and miRanda (2010 version)[140]^43 computational
methods. The miRNA and lncRNA sequences were obtained from miRBase
(release 21)[141]^44 and GENCODE (v26)[142]^45, respectively. We
predicted miRNA target binding sites on the whole lncRNA sequences
using the intersection of the TargetScan and miRanda methods. Then,
predicted miRNA-lncRNA interactions were further filtered by starBase
(v3.0)[143]^29, which is a database containing the Argonaute (AGO)
crosslinking immunoprecipitation sequencing (CLIP-seq)
experiment-supported miRNA-lncRNA interactions. The intersections with
starBase (v3.0) were selected as candidate miRNA-lncRNA interactions.
Then, miRNA-mRNA interactions were obtained from miRTarBase (release
7.0),[144]^46 which is a manually curated database that provides
experimentally supported miRNA-gene interactions. We only retained
high-confidence functional miRNA-gene interactions supported by
reporter assay and/or western blot data. The miRNA target genes were
further filtered by human MG risk genes, which were manually curated
using strict criteria from our previous study.[145]^47 We ultimately
retained the intersections of experimentally supported genes of miRNA
and human MG risk genes as the targets of miRNAs.
Identification of Potential mRNA-lncRNA Competing Interactions
To identify competing mRNA-lncRNA interactions, hypergeometric tests
and the Pearson correlation coefficient (PCC) were employed to identify
the competing pairs.[146]^48 First, the hypergeometric test was used to
evaluate the significance of the shared miRNAs between each mRNA and
lncRNA. The formula used was as follows:
[MATH: P=1−∑k=0x(mk
)(N−mn−k)(
Nn). :MATH]
For each mRNA-lncRNA pair, N denoted the total number of miRNAs in the
genome, n represented the number of miRNAs that were associated with
one mRNA, m represented the number of miRNAs that were associated with
one lncRNA, and x represented the number of common miRNAs that shared
the mRNA and lncRNA. The mRNA-lncRNA pairs with a p value <0.05 were
considered significant interactions.
Next, we evaluated co-expression correlation of mRNA-lncRNA
pairs identified above using the PCC. The lncRNA and mRNA expression
data were downloaded from the dbGaP database (the Genotype-Tissue
Expression Project, released in 2016;
[147]https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_
id=phs000424.v7.p2), which contains 32 healthy tissues in 7,862 samples
from 552 donors.[148]^49 The p values of co-expression analysis were
adjusted according to the false discovery rate (FDR). The mRNA-lncRNA
pairs with a PCC of ≥0.2 and a FDR of <0.01 were identified as
co-expressed pairs. The above analyses were performed by R software.
Construction of the LMGCN
We constructed the LMGCN based on the theory that lncRNAs share common
miRNA-binding sites with mRNAs and function as miRNA sponges to
regulate mRNAs. For a given lncRNA-miRNA-mRNA interaction, both mRNA
and lncRNA shared common miRNAs and were co-expressed for merging into
a competing triplet. After assembling all lncRNA-miRNA-mRNA competing
triplets, we constructed the LMGCN. The network was visualized using
Cytoscape software, in which nodes represent miRNAs, genes, and
lncRNAs, and edges represent their interactions.
Functional Enrichment Analysis
To further confirm the roles of lncRNAs, we performed Kyoto
Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis
and investigated biological processes in GO annotation of co-expressed
mRNAs in the LMGCN using Database for Annotation, Visualization and
Integrated Discovery (DAVID),[149]^50 which is an online functional
annotation tool. Pathways and GO terms with p < 0.05 were considered to
be significantly enriched function annotations.
Clinical Samples
A total of 24 MG patients who were followed at The Second Affiliated
Hospital of Harbin Medical University were included in this study. All
the patients were diagnosed initially and met the diagnostic criteria
for MG.[150]^51 A total of 29 sex- and age-matched healthy donors with
no history of autoimmune disease were included as the control. This
study was approved by the Ethics Committee of The Second Affiliated
Hospital of Harbin Medical University. Written informed consent was
obtained from all subjects. The study was carried out according to the
World Medical Association Declaration of Helsinki. Peripheral blood
samples were collected from each participant in tubes containing
ethylenediaminetetraacetic acid, and PBMCs were isolated using
lymphocyte separation medium.
Cell Culture
The T cell leukemia line (Jurkat cells) and human embryonic kidney 293T
(HEK293T) cells were purchased from the American Type Culture
Collection (Manassas, VA, USA). Jurkat cells, which had been used for
functional verification of MG according to previous
studies,[151]^10^,[152]^18 were cultured in RPMI 1640 medium (Gibco,
Carlsbad, CA, USA). HEK293T cells were cultured in Dulbecco’s modified
Eagle’s medium (DMEM; Gibco). All media were supplemented with 10%
fetal bovine serum (Gibco), together with 100 IU/mL penicillin and
100 μg/mL streptomycin (KeyGen Biotech, Nanjing, China). All cells were
incubated at 37°C in a humidified atmosphere of 5% CO[2].
Cell Transfection
lncRNA Smart Silencer for human SNHG16, let-7c-5p mimics, let-7c-5p
inhibitor, and negative control (NC), designed and synthesized by
Ribobio (Guangzhou, China), was transfected into Jurkat cells using
Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). SNHG16 Smart
Silencer (siSNHG16) was a pool containing three siRNAs and three
antisense oligonucleotides, which was applied to knock down the
expression of SNHG16. The sequences of Smart Silencer were as follows:
siSNHG16-1 target sequence, 5′-CCTTCCAGGAACAGATCTT-3′; siSNHG16-2
target sequence, 5′-TGTTGACTCACCAAGGCAA-3′; siSNHG16-3 target sequence,
5′-GGAACAGATCTTTGCATAG-3′; antisense oligonucleotides target
sequence-1, 5′-CCCAGCTATTTTTTCTTTCG-3′; antisense oligonucleotides
target sequence-2, 5′-CGTGCCCTAAATTGACCAAC-3′; and antisense
oligonucleotides target sequence-3, 5′-CCACTTACAATAAACTTGGG-3′.
Real-Time PCR Analysis
Total RNA was extracted from PBMCs or Jurkat cells using TRIzol reagent
(Invitrogen) following the manufacturer’s instructions. Reverse
transcription of total RNA into cDNA was performed using Primescript RT
reagent kit (Takara, Dalian, China) according to the manufacturer’s
instructions. The sequences of the primers used are listed in
[153]Table S6. Quantitative real-time PCR was performed using the SYBR
Premix Ex Taq kit (Takara, Dalian, China) in a Roche LightCycler 480
instrument. The relative expression level was calculated using the
2^−ΔΔCt method. RPL13A was used as the reference for SNHG16 and IL-10,
and U6 was used as the reference for let-7c-5p.
Western Blot Analysis
Total protein was extracted from Jurkat cells using a
radioimmunoprecipitation assay lysis buffer (Beyotime Biotechnology,
Shanghai, China). The protein concentration was determined using a
bicinchoninic acid protein assay kit (Beyotime). Proteins were
separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) and electrophoretically transferred onto polyvinylidene
fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). The membrane
was incubated with the following antibodies at 4°C overnight: IL-10
(1:1,000, ABclonal, Wuhan, China) and β-actin (1:10,000, Abways,
Shanghai, China). The membranes were then washed and incubated with
anti-rabbit horseradish peroxidase (HRP)-linked secondary antibodies
(Cell Signaling Technology, Danvers, MA, USA) for 2 h at room
temperature. The PVDF membranes were washed with Tris-buffered saline
with Tween 20 (TBST), and the blots were visualized using enhanced
chemiluminescence (ECL) (Carestream, Wuxi, China). β-Actin was used as
the internal control. The experiments were independently repeated in
triplicate.
Luciferase Reporter Assay
A fragment from SNHG16 containing the predicted let-7c-5p binding site
was cloned into a PHY-811 vector (Hanyin Biotechnology, Shanghai,
China) to construct the luciferase reporter vector, named SNHG16-WT. To
mutate the putative binding site of let-7c-5p in SNHG16, the sequence
of the putative binding site was replaced, and the vector was named
SNHG16-MUT. The SNHG16-WT or SNHG16-MUT plasmids were co-transfected
together with let-7c-5p mimics or negative control into HEK293T cells
using Lipofectamine 3000 (Invitrogen). A dual-luciferase reporter assay
system (Promega, Madison, WI, USA) was used to assess luciferase
activity after 48 h of transfection according to the manufacturer’s
protocol. Relative luciferase activity was measured and normalized to
Renilla luciferase activity.
Cell Proliferation Assay
Cell proliferation was evaluated using a CCK-8 assay (Dojindo, Tokyo,
Japan). Cells from different groups were seeded onto 96-well plates at
a density of 1,500 cells/well and incubated at 37°C in a humidified
atmosphere of 5% CO[2]. According to the instructions, 10 μL of CCK-8
reagent was added to the wells at time points of 24, 48, 72, 96, and
120 h. The absorbance at 450 nm was measured using a microplate reader
(BioTek Instruments, Winooski, VT, USA) after the plates were incubated
at 37°C for 2 h. The experiments were independently repeated in
triplicate.
Apoptosis Analysis
Apoptosis was analyzed using an annexin V-fluorescein isothiocyanate
(FITC)/propidium iodide (PI) apoptosis detection kit (BD Biosciences,
San Jose, CA, USA) according to the manufacturer’s protocol. Jurkat
cells were transfected and cultured in a six-well plate. After a 48-h
incubation, cells were digested using trypsin (Gibco) and then stained
with annexin V-FITC and PI. Then, stained cells were analyzed by flow
cytometry using CellQuest software (BD Biosciences, Franklin Lakes, NJ,
USA). The experiments were independently repeated in triplicate.
Statistical Analysis
SPSS 17.0 software was used to perform the statistical analyses. Data
are expressed as mean ± standard deviation (SD). The differences
between the two groups were analyzed using Student’s t test,
while the differences among multiple groups were analyzed using ANOVA.
p < 0.01 was considered statistically significant.
Author Contributions
H.Z., L.W., and S.N. designed the study. J.W. and Y.C. wrote the
manuscript. J.W. and X.L. performed the bioinformatics analysis. X.W.,
X.K., C.B., and S.L. conducted the experiments. J.W. performed the
statistical analyses. M.B., Y.J., H.G., and X.Y. provided valuable
opinions in the process of writing the manuscript. All authors read and
approved the final manuscript.
Conflicts of Interest
The authors declare no competing interests.
Acknowledgments