-
Categories
-
Pharmaceutical Intermediates
-
Active Pharmaceutical Ingredients
-
Food Additives
- Industrial Coatings
- Agrochemicals
- Dyes and Pigments
- Surfactant
- Flavors and Fragrances
- Chemical Reagents
- Catalyst and Auxiliary
- Natural Products
- Inorganic Chemistry
-
Organic Chemistry
-
Biochemical Engineering
- Analytical Chemistry
-
Cosmetic Ingredient
- Water Treatment Chemical
-
Pharmaceutical Intermediates
Promotion
ECHEMI Mall
Wholesale
Weekly Price
Exhibition
News
-
Trade Service
Today I would like to share with you an article published in Cancer Cell (IF:38.
585):
01 Research BackgroundSingle-cell RNA sequencing (scRNA-seq) studies of high- and low-grade gliomas have shown that intratumor heterogeneity and dynamic plasticity of cellular states are hallmarks of malignant brain tumors.
This dynamic adaptation occurs in four distinct states, namely mesenchymal (MES-like), neural progenitor-like (NPC-like), astrocyte-like (AC-like), and oligodendrocyte-like (PAC-like) states
.
Although brain tumors exhibit similar transcriptional adaptations and evolution to healthy brains, they have long been studied as separate entities, ignoring the role of
the local microenvironment in tumorigenesis.
However, the recent literature has reported local interactions between tumor cells and the neuronal environment, where neurons, glial cells, and immune cells contribute
to the complex and dynamically heterogeneous glioma network.
Due to the loss of spatial organization information, single-cell analysis provides only indirect inferences
of cellular interactions.
In the brain, spatial organization and function are closely related
.
Therefore, the investigators hypothesized that central nervous system malignancies are also organized
functionally and spatially.
Spatial transcriptomics (stRNA-seq) is a new technique that allows cell interactions and organization to be characterized in situ, thus deciphering the ecosystem
of malignant brain tumors.
02 Researchers characterize glioblastoma through spatially resolved transcriptomics, metabolomics, and proteomics, identify microenvironments marked by immune and metabolic stressors, and reveal local tumor-host interdependencies, leading to the generation of space-specific adaptive transcriptional programs.
In addition, they implanted glioblastoma stem cell models into human and rodent neocortical tissues to mimic various environments, confirming that transcriptional states stem from dynamic adaptation
to various environments.
PART
03 RESULTS OF THE STUDY SPATIALLY RESOLVED TRANSCRIPTOME MAP OF GLIOBLASTOMA (GBM).
of methods and cohorts Resolving spatial resolution of transcriptional heterogeneity investigators reproduce spatial transcriptional patterns by linking to existing transcriptional and histological classification data.
Through the analysis of spatial transcriptomics, five transcriptional patterns were identified, and these 5 patterns were present in 90% of tumors (Figure 2A).
The researchers then integrated the patient's spatially weighted correlation matrix, followed by hierarchical clustering, to confirm that five spatially distinct transcription patterns were spatially isolated (Figure 2A).
To determine whether the different spatial transcription patterns detected exist at all cell cycle stages, researchers classify all SPOS according to their transcription and cell cycle procedures.
Transcription patterns are evenly distributed across all cell cycle stages (Figure 2A).
To further integrate the findings with existing classification systems, spatially weighted regression was performed on different transcriptional pattern classifications on existing bulkRNA-seq, scRNA-seq, and stRNA-seq to map spatial relationships (Figure 2C).
Different spatial transcription patterns and altered copy number (CNA) of subclonal structures, including focal amplification of oncogenes or loss of tumor suppressors, are markers of malignant cells that have been reported to lead to treatment resistance and tumor recurrence
.
In known cell state classifications, unique CNAs are associated with specific states but are unevenly
distributed within individual tumor cells.
In Reactive Hypoxia-associated spots, the researchers found significant accumulation of CNA as an independent subclonal event (Figure 2D).
Metabolic changes associated with the Reactive Hypoxia model To further explore the spatially distinct Reactive Hypoxia patterns, the investigators performed a spatial metabolomics study (Figure 4A).
Further analysis identified three significant metabolic subgroups (Figures 4A, 4B).
Functional metabolic analysis showed that the pentose phosphate pathway was significantly enriched in the first metabolic module (M-G1
).
The second metabolic module (M-G2) is characterized by enrichment
of adenylate phosphate metabolism.
The third metabolic module (M-G3) is significantly enriched in glycolysis and aminoglycan metabolism (Figure 4C).
The Reactive Hypixia pattern accumulates chromosomal alterations due to hypoxia has been shown to cause DNA damage responses and altered expression of repair genes, thereby inhibiting recombination-mediated DNA double-strand break repair, increasing mutation rates and CNA
.
To investigate the hypothesis, regions where the Reactive Hypoxia pathway is enriched, which is also enriched with glycolytic pathways (Figure 4D)
was explored.
And the CNA map showed that chromosomes 15p, 14q, and 7p/q were significantly missing
in the hypoxic core region.
Individual samples also showed loss and amplification on multiple chromosomes (8p, 9p, 13q, 19q, and 21q), most likely caused by a single subclone within the hypoxic core region, demonstrating that hypoxia-related metabolism is a driver of potential genomic instability (Figure 4E).
Fig.
4 Integrating spatial metabolomics data
The "go or grow" potential in hypoxic metabolism regulation GBM has been documented in the literature to suggest that hypoxic stress triggers cell cycle arrest, specifically S-phase arrest (Figure 6A).
According to the study model, an inverse relationship between hypoxic metabolism and cell migration was observed (Figures 6B, 6C, and 6D).
Together, their findings provide evidence that metabolic changes and oxidative stress are potential interdrivers of genomic diversity, leading to clonal evolution
of GBM.
When hypoxia occurs, this hypoxia-induced glycolytic transition leads to the induction of go programs, which help cells escape/migrate to the normoxic zone
.
Due to hypoxic conditions, phase S stagnation occurs, which subsequently leads to the accumulation
of de novo CNA.
The authors hypothesize that a large proportion of these cells affected by hypoxia will go into an apoptotic state, leading to the characteristic necrosis
seen in GBM.
Only a small percentage of cells successfully escaped through upregulation of migration-related transcription patterns
.
In addition to hypoxia, other stressors, such as radiation or chemotherapy, have been reported to cause this pressure-related perturbation in GBM (Figure 6A).
According to this model, an inverse relationship
between hypoxic metabolism and cell migration can be seen.
To explore the enrichment of migratory gene expression traits in regions of metabolic alteration, the researchers determined the spatial direction of the directional gradient between low and high enrichment of specific gene expression traits.
The direction vector of each point is a hierarchical enrichment based on the gene expression characteristics studied in its local neighborhood
.
These vector field calculations enable authors to approximate spatial gene expression trajectories and thus be able to identify spatially oppositely opposed transcriptional pathways (Figure 6B).
Based on these vector field calculations, the researchers reported that the hypoxic response and migration characteristics showed reverse spatial trajectories (Figures 6C and 6D).
Exploring tumor-host interdependence in the Reactive Immune region Spatial proteomics analysis found a significant increase in myeloid cells and lymphocytes in the region of Reactive Immune (Figures 7A, 7B, and 7C).
。 To study tumor cell differentiation in the Reactive Immune region, the researchers subdivided the cells into Radial Glia (EGFR+HOPX+), Reactive Immune (EGFR+CHI3L1+VIM+), Spatial OPC (EGFR+OLIG1+), and Neuronal Development (EGFR+SNAP25+CALM2+).
and Reactive Hypoxia (EGFR+).
7 Spatial proteome data integration
of environmental conditions facilitates bidirectional subtype transition to simulate tissue infiltration through a previously established GBM model based on human organotype neocortical tissue, which has no metabolic and immune stress
.
To assess the effects of various microenvironments, the researchers used human tissue of different ages and tissue from 2-week-old mice and 2-year-old rats (Figure 8A).
Tissue culture for 4 days and then seed a similar MES-like primary patient-derived cell line into all cultured tissue sections (Figure 8A).
After 7 days of culture, the tissue is digested and tumor cells are isolated using FACS for scRNA-seq analysis
.
Using inferred CNAs analysis (Figures 8C and 8D), computational identification
of tumor cells based on characteristic expansion of tumor cells in chromosome 7 is performed.
04 SUMMARY In summary, this work uses spatially resolved multiomics in glioblastoma samples and identifies microenvironments
characterized by immune and metabolic stress factors.
These spatial niches are influenced by the tumor microenvironment, reflect transcriptional adaptations to inflammatory or metabolic stimuli, and reproduce neurodevelopmental stages
.
This article has a lot to learn, such as experimental techniques, using many advanced spatial omics techniques, spatial transcriptome, spatial metabolome and spatial proteome
.
In terms of mechanism, in vivo and in vitro models are combined to gradually verify the underlying mechanism
.
▎Orange typesetting ▎XX
585):
Spatial multiomics to resolve tumor-host interdependence in glioblastoma
01 Research BackgroundSingle-cell RNA sequencing (scRNA-seq) studies of high- and low-grade gliomas have shown that intratumor heterogeneity and dynamic plasticity of cellular states are hallmarks of malignant brain tumors.
This dynamic adaptation occurs in four distinct states, namely mesenchymal (MES-like), neural progenitor-like (NPC-like), astrocyte-like (AC-like), and oligodendrocyte-like (PAC-like) states
.
Although brain tumors exhibit similar transcriptional adaptations and evolution to healthy brains, they have long been studied as separate entities, ignoring the role of
the local microenvironment in tumorigenesis.
However, the recent literature has reported local interactions between tumor cells and the neuronal environment, where neurons, glial cells, and immune cells contribute
to the complex and dynamically heterogeneous glioma network.
Due to the loss of spatial organization information, single-cell analysis provides only indirect inferences
of cellular interactions.
In the brain, spatial organization and function are closely related
.
Therefore, the investigators hypothesized that central nervous system malignancies are also organized
functionally and spatially.
Spatial transcriptomics (stRNA-seq) is a new technique that allows cell interactions and organization to be characterized in situ, thus deciphering the ecosystem
of malignant brain tumors.
Given the complementary nature of scRNA-seq and stRNA-seq, the integration of the two technologies is critical
.
Cell communication and metabolism are two key factors that have a decisive impact on the dynamic adaptation of brain cancer, promoting growth, invasion and treatment resistance
.
These metabolic changes can be attributed to the microenvironment or regional heterogeneity
of tumor metabolism.
For example, forced metabolic deterioration due to hypoxia has been shown to significantly drive transcriptional adaptation and genomic instability
.
Cellular interactions between tumor cells and the immune system are also receiving increasing attention, with a recent study demonstrating that epigenetic immune editing drives acquired immune invasion programs, leading to a spatially heterogeneous landscape
in glioblastoma.
These studies further highlight the need for a comprehensive study of the various transcriptional adaptations
of gliomas due to the microenvironment in the context of spatial resolution.
This study aims to provide a spatially resolved transcriptional program and cellular interaction map
of glioblastoma and its local microenvironment.
02 Researchers characterize glioblastoma through spatially resolved transcriptomics, metabolomics, and proteomics, identify microenvironments marked by immune and metabolic stressors, and reveal local tumor-host interdependencies, leading to the generation of space-specific adaptive transcriptional programs.
In addition, they implanted glioblastoma stem cell models into human and rodent neocortical tissues to mimic various environments, confirming that transcriptional states stem from dynamic adaptation
to various environments.
PART
03 RESULTS OF THE STUDY SPATIALLY RESOLVED TRANSCRIPTOME MAP OF GLIOBLASTOMA (GBM).
To characterize the spatial structure of GBM, the researchers generated spatial transcriptomics (stRNA-seq) profiles of 28 samples, resulting in a total of 88,793 spots
.
The investigators also supplemented spatial metabolomics and proteomics from tissues for stRNA-seq-assisted integration analysis (Figure 1A).
Horizontal integration based on mutual nearest neighbor (MNN) and shared nearest neighbor (SNN) clustering showed that non-malignant samples showed similarity
between patients.
Malignant samples are characterized by unique gene expression profiles, further confirmed by the high Shannon entropy of sample diversity between clusters (Figure 1B).
If the malignant origin sample shows significantly lower entropy, it is indicated that the cluster consists of spots from individual patients (Figures 1B and 1C).
These findings are consistent with recent single-cell studies in which a large number of individual copy number changes and mutation profiles contribute to heterogeneity between patients, leading to changes
in individual transcription profiles.
Of the 88,793 spots, 63,121 were from malignant samples, of which 46,459 spots, with a proportion of tumor cell content above 95% in each spot (Figure 1C).
To study the tumor invasion region, the authors predicted the spots' histological phenotype according to the Ivy GAP histological classification system (Figure 1D).
Compared to histopathological features, samples with low tumor cell frequencies mainly contain areas of invasion in which tumor cells are rarely present within the healthy cortex (Figure 1E).
Figure 1: Overview
of methods and cohorts Resolving spatial resolution of transcriptional heterogeneity investigators reproduce spatial transcriptional patterns by linking to existing transcriptional and histological classification data.
Through the analysis of spatial transcriptomics, five transcriptional patterns were identified, and these 5 patterns were present in 90% of tumors (Figure 2A).
The researchers then integrated the patient's spatially weighted correlation matrix, followed by hierarchical clustering, to confirm that five spatially distinct transcription patterns were spatially isolated (Figure 2A).
By exploring these different procedures, the researchers identified two spatially distinct transcriptional patterns
associated with high expression of glia-related genes (e.
g.
, GFAP, AQP4, VIM, CD44).
While both models showed glial lineages, one was associated with increased expression of radial glia-related genes (e.
g.
, HOPX, PTPRZ1) called "Radial Glia"; The other shows functional enrichment of inflammation-related genes (e.
g.
, HLA-DRA, C3) and INF-γ pathway, called "Reactive Immune" (Figure 2A).
The remaining transcription patterns showed consistency with the glial lineage and were named "Neuronal Development" according to their nerves or oligodendrocyte origin named "Spatial OPC" (Figure 2C).
。 A fifth transcription pattern, further known as "Reactive Hypoxia," is associated with hypoxia (e.
g.
, VEGFR, HMOX1, GAPDH) and glycolysis (e.
g.
, LDHA, PGK1) genes, indicating metabolic alterations bound to low concentrations of oxygen driving transcriptional different states in certain regions (Figures 2A, 2B).
To determine whether the different spatial transcription patterns detected exist at all cell cycle stages, researchers classify all SPOS according to their transcription and cell cycle procedures.
Transcription patterns are evenly distributed across all cell cycle stages (Figure 2A).
To further integrate the findings with existing classification systems, spatially weighted regression was performed on different transcriptional pattern classifications on existing bulkRNA-seq, scRNA-seq, and stRNA-seq to map spatial relationships (Figure 2C).
The scRNAseq dataset of Neftel et al.
was used to integrate the gene features with the highest scores in different spatial transcription patterns, and it was confirmed that the three transcription modes of Radial Glia, Spatial OPC and Neuronal Development had great overlap
with AC-like, OPC-like and NPC-like, respectively.
Compared to data from Richards et al.
, all of the above patterns have most of the overlap with neurodevelopmental phenotypes (Figure 2C).
The Reactive Hypoxia model overlaps substantially with the MES-like subtype, particularly the hypoxia-dependent "MES2" state (Figure 2C).
Fig.
2 Exploring spatially resolved transcriptional heterogeneity
Different spatial transcription patterns and altered copy number (CNA) of subclonal structures, including focal amplification of oncogenes or loss of tumor suppressors, are markers of malignant cells that have been reported to lead to treatment resistance and tumor recurrence
.
In known cell state classifications, unique CNAs are associated with specific states but are unevenly
distributed within individual tumor cells.
In Reactive Hypoxia-associated spots, the researchers found significant accumulation of CNA as an independent subclonal event (Figure 2D).
Next, the authors investigated whether spatially different transcriptional diversity could directly reflect intratumor subclonals
.
They reconstructed the clonal structure
through hierarchical clustering of patient-specific CNA.
A total of 57 subclones were identified, each containing between 2 and 6 subclones (Figure 3A).
Then, the distribution of samples, different spatial transcription patterns in each subclone was determined (Figures 3B and 3C).
The analysis showed that in 26.
32% of subclons, a single transcription pattern predominated (more than 75% of the spots, per subclonal).
These results suggest that clonal structures have little effect on the occurrence of different spatial transcription patterns, although subcloning occasionally favors Spatial OPC (8.
77%) or Reactive Hypoxia (10.
52%) (Figure 3D).
Fig.
3 Transcription patterns are independent of subclonal structures
Metabolic changes associated with the Reactive Hypoxia model To further explore the spatially distinct Reactive Hypoxia patterns, the investigators performed a spatial metabolomics study (Figure 4A).
Further analysis identified three significant metabolic subgroups (Figures 4A, 4B).
Functional metabolic analysis showed that the pentose phosphate pathway was significantly enriched in the first metabolic module (M-G1
).
The second metabolic module (M-G2) is characterized by enrichment
of adenylate phosphate metabolism.
The third metabolic module (M-G3) is significantly enriched in glycolysis and aminoglycan metabolism (Figure 4C).
The Reactive Hypixia pattern accumulates chromosomal alterations due to hypoxia has been shown to cause DNA damage responses and altered expression of repair genes, thereby inhibiting recombination-mediated DNA double-strand break repair, increasing mutation rates and CNA
.
To investigate the hypothesis, regions where the Reactive Hypoxia pathway is enriched, which is also enriched with glycolytic pathways (Figure 4D)
was explored.
And the CNA map showed that chromosomes 15p, 14q, and 7p/q were significantly missing
in the hypoxic core region.
Individual samples also showed loss and amplification on multiple chromosomes (8p, 9p, 13q, 19q, and 21q), most likely caused by a single subclone within the hypoxic core region, demonstrating that hypoxia-related metabolism is a driver of potential genomic instability (Figure 4E).
Fig.
4 Integrating spatial metabolomics data
To further validate the findings, the investigators analyzed TCGA-GBM samples (GBM IDH1/2 wild-type n=357) and classified them according to the patients' hypoxic gene expression scores (Figures 5A and 5B).
Hypoxia-driven tumors show a significant increase in chromosomal alterations, confirming the relationship between metabolism and genomic instability (Figures 5C and 5D).
Further experimental validation was that primary patient-derived GBM cell lines were cultured under normoxic and hypoxic conditions for 2-6 weeks (Figure 5E).
The significant accumulation of CNA events under chronic hypoxic conditions confirms the hypothesis of stress-induced CNA variation (Figure 4E).
In addition, the researchers studied the effect
of hypoxic metabolism on DNA methylation by analyzing the O-6 methylguanine-DNA methyltransferase (MGMT) promoter methylation at the CpG sites cg12434587 and cg12981137 under normoxic and hypoxic conditions.
Cell lines with non-methylated MGMT promoters under normobic conditions undergo hypermethylation under hypoxic conditions (Figure 5F).
Fig.
5 TCGA and cell culture validation of hypoxia-related CNA changes
The "go or grow" potential in hypoxic metabolism regulation GBM has been documented in the literature to suggest that hypoxic stress triggers cell cycle arrest, specifically S-phase arrest (Figure 6A).
According to the study model, an inverse relationship between hypoxic metabolism and cell migration was observed (Figures 6B, 6C, and 6D).
Together, their findings provide evidence that metabolic changes and oxidative stress are potential interdrivers of genomic diversity, leading to clonal evolution
of GBM.
When hypoxia occurs, this hypoxia-induced glycolytic transition leads to the induction of go programs, which help cells escape/migrate to the normoxic zone
.
Due to hypoxic conditions, phase S stagnation occurs, which subsequently leads to the accumulation
of de novo CNA.
The authors hypothesize that a large proportion of these cells affected by hypoxia will go into an apoptotic state, leading to the characteristic necrosis
seen in GBM.
Only a small percentage of cells successfully escaped through upregulation of migration-related transcription patterns
.
In addition to hypoxia, other stressors, such as radiation or chemotherapy, have been reported to cause this pressure-related perturbation in GBM (Figure 6A).
According to this model, an inverse relationship
between hypoxic metabolism and cell migration can be seen.
To explore the enrichment of migratory gene expression traits in regions of metabolic alteration, the researchers determined the spatial direction of the directional gradient between low and high enrichment of specific gene expression traits.
The direction vector of each point is a hierarchical enrichment based on the gene expression characteristics studied in its local neighborhood
.
These vector field calculations enable authors to approximate spatial gene expression trajectories and thus be able to identify spatially oppositely opposed transcriptional pathways (Figure 6B).
Based on these vector field calculations, the researchers reported that the hypoxic response and migration characteristics showed reverse spatial trajectories (Figures 6C and 6D).
Fig.
6 Description of hypoxic stress concept and escape mechanism
Exploring tumor-host interdependence in the Reactive Immune region Spatial proteomics analysis found a significant increase in myeloid cells and lymphocytes in the region of Reactive Immune (Figures 7A, 7B, and 7C).
。 To study tumor cell differentiation in the Reactive Immune region, the researchers subdivided the cells into Radial Glia (EGFR+HOPX+), Reactive Immune (EGFR+CHI3L1+VIM+), Spatial OPC (EGFR+OLIG1+), and Neuronal Development (EGFR+SNAP25+CALM2+).
and Reactive Hypoxia (EGFR+).
The investigators quantified cellular connectivity between tumor cells and lymphoid or myeloid cells based on their distance, confirming enhanced cellular interactions between tumor cells and immune compartments in the transcriptionally defined Reactive Immune region (Figure 7D).
Next, the differences (regions
of interest) distribution
of cell types in ROI.
The researchers confirmed that AC-like and MES-like cells are enriched in Reactive Immune and Reactive Hypoxia
.
OPC-like cells are mainly enriched in Neuronal Development (Figures 7E and 7F).
In addition, both the Reactive Immune and Reactive Hypoxia regions showed significant enrichment of tumor-associated myeloid cells (TAM) and T cells (Figure 7G).
Given that T cells are enriched in both the Reactive Immune and Reactive Hypoxia regions, the authors looked at mean PD-1 protein levels on T cells (CD3+), which were significantly increased in the Reactive Immune region compared to Reactive Hypoxia and Neuronal Development, indicating enhanced local immunosuppression in Reactive Immune
。 In addition, the scRNA-seq dataset from GBM confirmed the enrichment of memory and depleted T cells in the tumor Reactive Immune region (Figure 7H).
Fig.
7 Spatial proteome data integration
of environmental conditions facilitates bidirectional subtype transition to simulate tissue infiltration through a previously established GBM model based on human organotype neocortical tissue, which has no metabolic and immune stress
.
To assess the effects of various microenvironments, the researchers used human tissue of different ages and tissue from 2-week-old mice and 2-year-old rats (Figure 8A).
Tissue culture for 4 days and then seed a similar MES-like primary patient-derived cell line into all cultured tissue sections (Figure 8A).
After 7 days of culture, the tissue is digested and tumor cells are isolated using FACS for scRNA-seq analysis
.
Using inferred CNAs analysis (Figures 8C and 8D), computational identification
of tumor cells based on characteristic expansion of tumor cells in chromosome 7 is performed.
To explore dynamic adaptations to different host environments, all malignant cells are enriched according to their corresponding cell states (Figures 8E and 8F).
The baseline state shows a MES-1/2-like phenotype, with low transcriptional diversity but mostly in the cell cycle cycle (Figure 8E).
GBM cells cultured in the rodent neural environment exhibit enrichment of MES-1/2-like and AC-like transcriptional features
.
To explore dynamic adaptation, the authors performed RNA rate analysis (Figure 8G).
In general, from initial in vitro cell culture (Figures 8G and 8H) to MES-like and AC-like, directional adaptation
driven by the neural environment is indicated.
PTPRZ1 is a common marker gene for radial glial differentiation of GBM cells, and based on its high-rate phase map in the MES-AC-hybrid state (Figures 8G, 8H, and 8I), the investigators' analysis confirmed bidirectional fate
within the OPC-like and NPC-like branches.
Fig.
8 In vitro human cortical culture shows a bidirectional transition through environmental influences
04 SUMMARY In summary, this work uses spatially resolved multiomics in glioblastoma samples and identifies microenvironments
characterized by immune and metabolic stress factors.
These spatial niches are influenced by the tumor microenvironment, reflect transcriptional adaptations to inflammatory or metabolic stimuli, and reproduce neurodevelopmental stages
.
This article has a lot to learn, such as experimental techniques, using many advanced spatial omics techniques, spatial transcriptome, spatial metabolome and spatial proteome
.
In terms of mechanism, in vivo and in vitro models are combined to gradually verify the underlying mechanism
.
END
Don't want to miss the daily hot spots and technologies welcome everyone to add a letter to write for the star recommendation▎Orange typesetting ▎XX
Latest ideas recommended
Cell death |
|
Liquid biopsy |
|
Single cellular |
|
m6A Feature |
|
Clinical topics |
|
