Limits...
Melanoma brain colonization involves the emergence of a brain-adaptive phenotype.

Nygaard V, Prasmickaite L, Vasiliauskaite K, Clancy T, Hovig E - Oncoscience (2014)

Bottom Line: The brain-adaptive phenotype was found as more prominent in the early metastatic growth phases compared to a later phase, emphasizing a temporal requirement of critical events in the successful colonization of the brain.Combined experimental and computational approaches clearly highlighted genes and signaling pathways being shared with neurodegenerative diseases.Importantly, the identification of essential molecular networks that operate to promote the brain-adaptive phenotype is of clinical relevance, as they represent leads to urgently needed therapeutic targets.

View Article: PubMed Central - PubMed

Affiliation: Department of Tumor Biology, Institute for Cancer Research, The Norwegian Radium Hospital, Oslo University Hospital, Oslo, 0310, Norway.

ABSTRACT
The brain offers a unique microenvironment that plays an important role in the establishment and progression of metastasis. However, the molecular determinants that promote development of melanoma brain metastases are largely unknown. Utilizing two species of immune-compromised animals, with in vivo cultivated metastatic tissues along with their corresponding host tissues in a metastasis model, we here identify molecular events associated with melanoma brain metastases. We find that the transcriptional changes in the melanoma cells, as induced by the brain-microenvironment in both host species, reveal the opportunistic nature of melanoma in this biological context in rewiring the molecular framework of key molecular players with their associated biological processes. Specifically, we identify the existence of a neuron-like melanoma phenotype, which includes synaptic characteristics and a neurotransmission-like circuit involving glutamate. Regulation of gene transcription and neuron-like plasticity by Ca(2+)-dependent signaling appear to occur through glutamate receptor activation. The brain-adaptive phenotype was found as more prominent in the early metastatic growth phases compared to a later phase, emphasizing a temporal requirement of critical events in the successful colonization of the brain. Analysis of the host tissue uncovered a cooperative inflammatory microenvironment formed by activated host cells that permitted melanoma growth at the expense of the host organism. Combined experimental and computational approaches clearly highlighted genes and signaling pathways being shared with neurodegenerative diseases. Importantly, the identification of essential molecular networks that operate to promote the brain-adaptive phenotype is of clinical relevance, as they represent leads to urgently needed therapeutic targets.

No MeSH data available.


Related in: MedlinePlus

Signal intensity heatmaps of significant gene sets(A) Hierarchical clustering analysis of genes with significant variance across organ-specific metastasis specimens and in vitro cell cultures (marked with ∆) derived from MM1 and MM5 cell lines. Gene vectors with SD≥1 were selected for cluster analysis (844/20458 transcripts). The columns of the matrix represent samples and rows represent gene transcripts. The red/green color in each cell represents expression levels ranging from low (green) to high (red) levels on a log2-transformed scale. The invasive and proliferative signatures could not distinguish the cell lines in vivo. Organ-specific signatures were proposed for brain and lung. (B) Expression pattern of genes differentially expressed in brain metastasis. Signal intensity heatmap of genes with at least a 1.5-fold difference (FDR≤10%) between brain metastases samples vs. other in vivo sites (columna, tibia and lung) and in vitro samples.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4295759&req=5

Figure 1: Signal intensity heatmaps of significant gene sets(A) Hierarchical clustering analysis of genes with significant variance across organ-specific metastasis specimens and in vitro cell cultures (marked with ∆) derived from MM1 and MM5 cell lines. Gene vectors with SD≥1 were selected for cluster analysis (844/20458 transcripts). The columns of the matrix represent samples and rows represent gene transcripts. The red/green color in each cell represents expression levels ranging from low (green) to high (red) levels on a log2-transformed scale. The invasive and proliferative signatures could not distinguish the cell lines in vivo. Organ-specific signatures were proposed for brain and lung. (B) Expression pattern of genes differentially expressed in brain metastasis. Signal intensity heatmap of genes with at least a 1.5-fold difference (FDR≤10%) between brain metastases samples vs. other in vivo sites (columna, tibia and lung) and in vitro samples.

Mentions: Left ventricular injection of MM1 and MM5 (106) cells respectively, into immune-suppressed rats caused development of metastasis in multiple organs, based on positive findings using immunomagnetic bead selection of melanoma cells in host tissues. To compare gene expression profiles from organ-specific metastatic lesions, we isolated metastatic melanoma cells from the brain, columna, tibia and lungs. MM1 was mainly represented by brain samples as the selected number of MM1 metastatic cells from organs other than brain was low. To explore the relationship between the 21 profiled specimens (Illumina HumanWG-6v3 beadarrays) representing in vivo and in vitro melanoma samples, we performed a hierarchical clustering analysis of normalized and log2-transformed gene expression data (Fig. 1A). The MM1 in vivo samples clustered closer to the MM5 in vivo samples than to their in vitro counterpart, and expressed genes representing the highly proliferative phenotype, such as the melanocytic markers TYR and MLANA, indicating a switch in phenotype of the MM1 progeny grown in vivo. A subset of genes showed an apparent organ-specific expression in either brain or lung. To determine the brain-specific gene expression, we assessed the number of genes differentially expressed in the brain metastasis samples (n=7), compared to the other in vivo (n= 10) and in vitro (n=4) specimens by Significance Analysis of Microarray (SAM). We found 32 genes to be up/down-regulated (FC=±1.5, FDR≤10%) in brain metastasis specimens (Supporting Information Dataset S1 and Fig. 1B). Neuron-specific functions were associated with up-regulated genes in this brain signature. Specifically, we found genes involved in neurotransmission (GRIA2, GRM3, GRM4 and SCN2B), neuron excitation (CAMKV, JPH4) and synaptic active zones (BSN, SNAP91). The canonical pathways which were most significantly associated with the up-regulated genes included glutamate receptor signaling (p-value= 2.7x10−5), neuropathic pain signaling (p-value=1.5x10−4) and synaptic long term potentiation (p-value=1.6x10−4).


Melanoma brain colonization involves the emergence of a brain-adaptive phenotype.

Nygaard V, Prasmickaite L, Vasiliauskaite K, Clancy T, Hovig E - Oncoscience (2014)

Signal intensity heatmaps of significant gene sets(A) Hierarchical clustering analysis of genes with significant variance across organ-specific metastasis specimens and in vitro cell cultures (marked with ∆) derived from MM1 and MM5 cell lines. Gene vectors with SD≥1 were selected for cluster analysis (844/20458 transcripts). The columns of the matrix represent samples and rows represent gene transcripts. The red/green color in each cell represents expression levels ranging from low (green) to high (red) levels on a log2-transformed scale. The invasive and proliferative signatures could not distinguish the cell lines in vivo. Organ-specific signatures were proposed for brain and lung. (B) Expression pattern of genes differentially expressed in brain metastasis. Signal intensity heatmap of genes with at least a 1.5-fold difference (FDR≤10%) between brain metastases samples vs. other in vivo sites (columna, tibia and lung) and in vitro samples.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4295759&req=5

Figure 1: Signal intensity heatmaps of significant gene sets(A) Hierarchical clustering analysis of genes with significant variance across organ-specific metastasis specimens and in vitro cell cultures (marked with ∆) derived from MM1 and MM5 cell lines. Gene vectors with SD≥1 were selected for cluster analysis (844/20458 transcripts). The columns of the matrix represent samples and rows represent gene transcripts. The red/green color in each cell represents expression levels ranging from low (green) to high (red) levels on a log2-transformed scale. The invasive and proliferative signatures could not distinguish the cell lines in vivo. Organ-specific signatures were proposed for brain and lung. (B) Expression pattern of genes differentially expressed in brain metastasis. Signal intensity heatmap of genes with at least a 1.5-fold difference (FDR≤10%) between brain metastases samples vs. other in vivo sites (columna, tibia and lung) and in vitro samples.
Mentions: Left ventricular injection of MM1 and MM5 (106) cells respectively, into immune-suppressed rats caused development of metastasis in multiple organs, based on positive findings using immunomagnetic bead selection of melanoma cells in host tissues. To compare gene expression profiles from organ-specific metastatic lesions, we isolated metastatic melanoma cells from the brain, columna, tibia and lungs. MM1 was mainly represented by brain samples as the selected number of MM1 metastatic cells from organs other than brain was low. To explore the relationship between the 21 profiled specimens (Illumina HumanWG-6v3 beadarrays) representing in vivo and in vitro melanoma samples, we performed a hierarchical clustering analysis of normalized and log2-transformed gene expression data (Fig. 1A). The MM1 in vivo samples clustered closer to the MM5 in vivo samples than to their in vitro counterpart, and expressed genes representing the highly proliferative phenotype, such as the melanocytic markers TYR and MLANA, indicating a switch in phenotype of the MM1 progeny grown in vivo. A subset of genes showed an apparent organ-specific expression in either brain or lung. To determine the brain-specific gene expression, we assessed the number of genes differentially expressed in the brain metastasis samples (n=7), compared to the other in vivo (n= 10) and in vitro (n=4) specimens by Significance Analysis of Microarray (SAM). We found 32 genes to be up/down-regulated (FC=±1.5, FDR≤10%) in brain metastasis specimens (Supporting Information Dataset S1 and Fig. 1B). Neuron-specific functions were associated with up-regulated genes in this brain signature. Specifically, we found genes involved in neurotransmission (GRIA2, GRM3, GRM4 and SCN2B), neuron excitation (CAMKV, JPH4) and synaptic active zones (BSN, SNAP91). The canonical pathways which were most significantly associated with the up-regulated genes included glutamate receptor signaling (p-value= 2.7x10−5), neuropathic pain signaling (p-value=1.5x10−4) and synaptic long term potentiation (p-value=1.6x10−4).

Bottom Line: The brain-adaptive phenotype was found as more prominent in the early metastatic growth phases compared to a later phase, emphasizing a temporal requirement of critical events in the successful colonization of the brain.Combined experimental and computational approaches clearly highlighted genes and signaling pathways being shared with neurodegenerative diseases.Importantly, the identification of essential molecular networks that operate to promote the brain-adaptive phenotype is of clinical relevance, as they represent leads to urgently needed therapeutic targets.

View Article: PubMed Central - PubMed

Affiliation: Department of Tumor Biology, Institute for Cancer Research, The Norwegian Radium Hospital, Oslo University Hospital, Oslo, 0310, Norway.

ABSTRACT
The brain offers a unique microenvironment that plays an important role in the establishment and progression of metastasis. However, the molecular determinants that promote development of melanoma brain metastases are largely unknown. Utilizing two species of immune-compromised animals, with in vivo cultivated metastatic tissues along with their corresponding host tissues in a metastasis model, we here identify molecular events associated with melanoma brain metastases. We find that the transcriptional changes in the melanoma cells, as induced by the brain-microenvironment in both host species, reveal the opportunistic nature of melanoma in this biological context in rewiring the molecular framework of key molecular players with their associated biological processes. Specifically, we identify the existence of a neuron-like melanoma phenotype, which includes synaptic characteristics and a neurotransmission-like circuit involving glutamate. Regulation of gene transcription and neuron-like plasticity by Ca(2+)-dependent signaling appear to occur through glutamate receptor activation. The brain-adaptive phenotype was found as more prominent in the early metastatic growth phases compared to a later phase, emphasizing a temporal requirement of critical events in the successful colonization of the brain. Analysis of the host tissue uncovered a cooperative inflammatory microenvironment formed by activated host cells that permitted melanoma growth at the expense of the host organism. Combined experimental and computational approaches clearly highlighted genes and signaling pathways being shared with neurodegenerative diseases. Importantly, the identification of essential molecular networks that operate to promote the brain-adaptive phenotype is of clinical relevance, as they represent leads to urgently needed therapeutic targets.

No MeSH data available.


Related in: MedlinePlus