Abstract
The brain is the main oxygen-consuming organ and is vulnerable to
ischemic shock or insufficient blood perfusion. Brain hypoxia has a
persistent and detrimental effect on resident neurons. Previous studies
have identified alterations in genes and metabolites in ischemic brain
shock by single omics, but the adaptive systems that neurons use to
cope with hypoxia remain uncovered. In the present study, we
constructed an acute hypoxia model and performed a multi-omics analysis
from RNA-sequencing and liquid chromatography-mass spectrometry
(LC-MS)-based metabolomics on exploring potentially differentially
expressed genes (DEGs) and metabolites (DEMs) in primary cortical
neurons under severe acute hypoxic conditions. The TUNEL assay showed
acute hypoxia-induced apoptosis in cortical neurons. Omics analysis
identified 564 DEGs and 46 DEMs categorized in the Kyoto encyclopedia
of genes and genomes (KEGG) database. Integrative pathway analysis
highlighted that dysregulated lipid metabolism, enhanced glycolysis,
and activated HIF-1 signaling pathways could regulate neuron physiology
and pathophysiology under hypoxia. These findings may help us
understand the transcriptional and metabolic mechanisms by which
cortical neurons respond to hypoxia and identify potential targets for
neuron protection.
Keywords: Hypoxia, Neuron, RNA sequencing, Metabolomics, Multi-omics
analysis
Highlights
* •
Acute hypoxia induces apoptosis in primary cortical neurons.
* •
Acute hypoxia alters the metabolomics and transcriptomics profiles
of cortical neurons.
* •
Multi-omics analysis shows dysregulated lipid metabolism, enhanced
glycolysis and activated HIF-1 signaling.
1. Introduction
The brain is one of the most metabolically active tissues and organs,
consuming ∼20% of the total basal oxygen (O[2]) to meet the high energy
demands of neuronal activity [[33][1], [34][2], [35][3], [36][4]]. As a
highly oxygen-consuming organ, the brain is inherently vulnerable to
hypoxia [[37]3,[38]5]. In humans and most mammals, substantially
reducing oxygen supply severely impairs brain function [[39][6],
[40][7], [41][8], [42][9]], and prolonging hypoxia irreversibly damages
the brain structure and function. For example, acute ischemic shock
leads to massive neuron death and blood-brain barrier disruption, while
inadequate cerebral perfusion induces various neuronal disorders
[[43]7,[44][10], [45][11], [46][12]]. The cortex is the most affected
brain region during ischemia and neonatal hypoxic-ischemic
encephalopathy [[47]13,[48]14]. Cerebral hypoxia can lead to cerebral
infarcts that are challenging to recover from or even life-threatening.
Although various studies have documented possible mechanisms involving
differentially expressed genes (DEGs) and metabolites (DEMs), there is
still a lack of effective interventions for ischemia based on current
knowledge. In particular, neurons, the primary cell types resident in
the brain tissue, including cortical neurons, bear the brunt of oxygen
shortage because they depend on oxygen to continuously produce
metabolic energy [[49]4,[50]7,[51]15]. Neuron death is an apparent
pathological change in the ischemia cortical region [[52]16];
therefore, further studies should examine how cortical neurons perceive
oxygen and identify the adaptive systems that cortical neurons use to
cope with hypoxia.
Omics approaches are powerful tools for clarifying pathogenic
mechanisms. Previous studies have shown that hypoxia induces adaptive
changes in many cellular events, including metabolic adaptations
[[53]16,[54]17]. Moreover, metabolites were identified as regulators or
indicators in ischemic reperfusion and other hypoxia-associated
conditions [[55][18], [56][19], [57][20], [58][21], [59][22]]. Single
omics (e.g., microarray and transcriptomics) approaches enable us to
dissect the underlying mechanisms of cellular adaptations under hypoxic
conditions [[60]23]. Yet, despite considerable advances in our
understanding of cellular adaptations to hypoxia promoted by single
omics, a more detailed assessment of the hypoxic response, particularly
from an integrative analysis perspective, has not been performed yet.
Multi-omics analysis provides insights into pathogenic mechanisms,
allowing us to interrogate diseases from multiple perspectives. Here,
we combined metabolic and transcriptomic analysis to probe the
molecular basis of acute hypoxia in cortical neurons. Our integrated
analysis revealed that hypoxia in primary cortical neurons
significantly alters metabolic and hypoxia-inducible factor 1 (HIF-1)
transcriptional signaling pathways. Therefore, targeting these pathways
is a potential therapeutic option for treating hypoxia-related ischemic
shock and other diseases.
2. Materials and methods
2.1. Chemicals
As previously described [[61]24], all chemicals and solvent reagents
used were HPLC or analytical grade.
2.2. Primary cortical neurons culture and hypoxia treatment
Primary cortical neurons (PCNs) were cultured as previously described
[[62]25,[63]26]. The PCNs cultures were maintained for 11 days in vitro
(DIV) and incubated in a hypoxia chamber (Smartor 118pro, Hua Yue
Enterprise Holdings Ltd., Guangzhou, China) containing 0.1%
Oxygen/5%CO2/94.9N2) for 2 h. Then, cells were collected for subsequent
mRNA extraction (n = 3), RNA-sequencing (RNA-seq) (n = 3), and
metabolomics analysis (n = 6). The protocol used to prepare primary
cortical neuron cultures from mice was approved by the research ethics
committee of the Tongren Hospital, Shanghai Jiao Tong University School
of Medicine.
2.3. Apoptosis detection by TUNEL assay
PCNs subjected to normoxia or acute hypoxia were fixed with 4%
paraformaldehyde, and the apoptotic cells were detected using the
One-step TUNEL Apoptosis Assay kit (#C1090, Shanghai, China). Prior to
mounting, the cells were counterstained with the nuclear dye
Hoechst33342 (DOJINDO, Japan). The apoptosis rate was calculated as the
number of TUNEL-positive cells divided by the number of
Hoechst33342-stained cells. Images were acquired under an A1-scope
fluorescence microscopy (Zeiss, Germany).
2.4. RNA isolation and RNA-seq library construction
The RNA extraction and sequencing library construction were performed
as previously described [[64]24].
2.5. RNA-seq and differentially expressed genes (DEGs) analysis
An Illumina NovaSeq 5000 platform sequenced the RNA libraries and
generated 150 bp paired-end reads. The raw reads ([65]BioProject
accession number PRJNA827649) were technically processed as previously
described [[66]24], and then gene expression was analyzed using the
DESeq (2012) R package [[67]27]. A p-value <0.05 and an |FC| >1.2 were
applied to screen for DEGs and shown by hierarchical cluster analysis.
Then the DEGs were used in the Kyoto encyclopedia of genes and genomes
(KEGG) [[68]28] pathway enrichment analysis with OEcloud
([69]https://cloud.oebiotech.com/task/). Gene ontology (GO) enrichment
was performed in Metscape ([70]https://metascape.org/gp/index.html).
Gene set enrichment analysis (GSEA) was employed to screen for pathways
related to biological events, using the following cut-offs: KEGG
pathway terms with enrichment score (ES) > 0.40, normalized enrichment
score (NES) > 1.4, and p < 0.01 were considered significantly enriched.
2.6. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
RNA extraction and RT-qPCR were performed as previously described
[[71]24,[72]25,[73]29]. All primers used in validated genes are listed
in [74]Supplementary Table 1.
2.7. Non-targeting metabolomics analysis
As reported in our recent studies [[75]24,[76]30], we performed
LC-MS-based metabolomics analysis using the OE Biotechnology Co. Ltd
(Shanghai, China) platform. Briefly, ∼1 × 10^7 11-DIV cultured PCNs in
acute hypoxia or normoxia (n = 6) were used to prepare LC-MS samples by
sonication with chloroform and dissociation with methanol: water (1:4)
solution. Subsequently, these samples were separated by
ultra-performance liquid chromatography (UPLC) on a Nexera UPLC System
(Shimadzu Corporation, Japan). Then, a Q Exactive mass spectrometer was
employed to acquire MS/MS spectra with information-dependent
acquisition (IDA) mode under the Xcalibur 4.0.27 software (Thermo
Fisher). Lastly, the acquired LC-MS raw data were processed in
Progenesis QI v2.3 (Waters Corporation, Milford, USA) to identify
metabolites based on public databases, including the human metabolome
database (HMDB) ([77]http://www.hmdb.ca/), METLIN and Lipid Maps (v2.3)
([78]http://www.lipidmaps.org/). The resulting metabolites were further
analyzed by the principal component analysis (PCA) and the orthogonal
projections to latent structures discriminant analysis (OPLS-DA) to
determine metabolic alterations in the hypoxia (H) and normoxia (N)
groups. The criteria for differentially expressed metabolites (DEMs)
were variable importance in the projection (VIP) > 1.0 as determined by
OPLS-DA) and p < 0.05, as shown by a two-tailed Student's t-test on the
normalized peak areas.
2.8. Integrated pathway analysis
DEGs (
[MATH: |FC|
:MATH]
>1.2, p < 0.05) and DEMs (VIP>1, p < 0.05) were used for integrated
analysis by Metabolyst 5.0 [[79]31]. We selected all pathway
(integrated) databases, the hypergeometric test enrichment, and the
combined p-values (pathway-level) integration method to compare results
from integrated pathway analysis.
2.9. Statistical analysis
The raw data for each group were statistically analyzed by unpaired
Student's t-test tested with GraphPad Prism 8 software. Data
represented as mean ± SD. Differences with p < 0.05 were considered
significant, and differences with p ≥ 0.05 were considered
non-significant.
3. Results
3.1. Acute hypoxia increases vascular endothelial growth factor A (Vegfa)
mRNA expression and causes apoptosis in primary cortical neurons
We first constructed the hypoxia model with the cultured mouse PCNs,
followed by a severe hypoxic treatment (0.1% O[2]) for 2 h. RT-qPCR
analysis showed that hypoxia marker Vegfa mRNA increased ∼2.8 fold in
the hypoxia group when compared to the normoxia group ([80]Fig. 1A), in
line with previous studies [[81]32,[82]33]. In addition, acute hypoxia
led to cortical neuron apoptosis, in contrast to normoxia ([83]Fig. 1B
and C). These results indicated that our model of primary cortical
neurons reflected the hallmarks of hypoxia and therefore could be used
for subsequent studies.
Fig. 1.
[84]Fig. 1
[85]Open in a new tab
Acute hypoxia increases Vegfa mRNA expression and causes apoptosis of
primary cortical neurons. (A) Vegfa mRNA upregulated in primary
cortical neurons undergoing acute severe hypoxia. Cultured mouse
cortical neurons (11-DIV) were incubated in a hypoxia chamber
containing 0.1% O[2] for 2 h and collected for Vegfa and Hif1a RT-qPCR
analysis. Mean ± SD, n = 4, unpaired student's t-test, ***p < 0.001,
ns, no significant. (B) Representative TUNEL staining of PCNs in Acute
hypoxia. Red, Cy3-labeled TUNEL-positive cells; Blue, Hoechst
counterstaining. Bar 50 μm. (C) The apoptosis rate in (B) was
quantified by calculating the ratio of TUNEL-positive cells to
Hoechst-stained cells. Data are expressed as mean ± SD (n = 5–6),
unpaired Student's t-test, **p < 0.001. (For interpretation of the
references to color in this figure legend, the reader is referred to