Separate gene and protein expression profiling of the malignant and stromal cell compartments in tumors

"Separate gene and protein expression profiling of the malignant and stromal celI hjernesvulster skjer det interaksjoner mellom kreftcellene og omgivende normale celler. Stromale vertsceller endrer egenskaper pga disse interaksjonene som vi forsøker å kartlegge.GFP-positive scid mus graftes med humane hjernesvulster. Når musene har utviklet xenograftsvulster tas svulstene ut, dissossieres og FACS-sortes på fluorescnes i en tumor og stromal cellepopulasjon. Etter å ha sjekket renhet av celepopulasjonene vha kontrollmikroskopi og humanspesifikke antistoffer isolerer vi RNA fra disse prøvene samt stromale (gliale) celler fra normal musehjerne. Dette er nå gjort på 6 prøver og genekspresjonsanalysen fullføres våren 2010. Så langt er en prøve analysert og vi har der funnet 1800 gener som er differensielt uttrykt mellom gliale celler fra hjernesvulster og normal hjerne. Av disse er ca 1200 gener nedregulert og 600 gener oppregulert i tumoraktiverte gliale celler i forhold til normal hjerne. I tillegg ser vi at gliale celler fra svulster uttrykker et høyere nivå av den nevrale stamcellemarkøren nestin og at de dessuten prolifererer. I ko-implantasjonsstudier ser vi også at tumorceller implantert sammen med tumoraktiverte gliale celler gir tumor fortere enn om vi bare implanterer tumorceller. Differensielt uttrykte gener vil valideres i et humant biopsimateriale. Vi ønsker slik å finne hvilke signalveier som virker i tumoraktiverte gliale celler. Vi vil gjøre ytterligere funksjonelle studier og bla kartlegge migratoriske egenskaper. Deretter vil de valideres som terapeutiske targets. Gjennom siRNA knockout og bruk av mus med forskjellig genetisk bakgrunn av de differensielt uttrykte genene, ønsker vi å finne mekanismene som regulerer interaksjonene mellom kreftceller og tumor-aktiverte gliale celler i hjernesvulster.

Tumor stroma interactions in brain tumoursMalignant tumors are composed of cancerous cells as well as non-transformed stromal cells. Interactions between these two cells compartments are regulating tumor progression. We have developed a model to analyse these interactions using fluorescent immunodeficient animals that are engrafted with human brain tumour biopsies.While a substantial amount of data have accumulated to describe the properties of glioma and other cancer cells, it has become increasingly appearant that the stromal compartment within tumours can modulate disease progression to a significant degree. The “seed and soil hypothesis” dating back more than a century, emphasised the role of tumor-stroma interactions in organ-specific metastasis, and proliferation of host vessels (angiogenesis) was early recognised as a requirement for tumor growth. In addition, prolonged survival amongst cancer patients following anti-angiogenic therapy is a powerful demonstration of the therapeutic potential that lies within targeting tumor-stroma interactions. Finally, since mutations and genetic instability is mainly restricted to malignant cells, stromal cell may be less succeptible to treatment resistance. For the same reason, they may be attractive targets for therapy. The stroma in most solid tumours is composed of fibroblasts, immune cells and endothelial cells as well as extracellular matrix (ECM) components. The brain however, exhibit unique tissue structures with other ECM components and astrocytes, oligodendrocytes, microglia and neurons as its main cellular constituents, in addition to endothelial cells. Brain tumour progression is characterised by a diffuse invasion of glioma cells into these structures, accompanied by extensive remodelling of the host tissue. In addition, these tumours typically display florid angiogenesis due to microvascular proliferation of nearby host vessels. Thus, like with other malignancies, it seems clear that tumor-stroma interactions are critical to brain tumor progression as well. Over the last decade, tumor angiogenesis and the involvement of endothelial cells has received much attention in brain tumor research. More recently, the role of fibroblasts associated with other tumour subtypes have been investigated and produced some intriguing findings: While implantation of cancer cells in animals produced tumors at reproducible rates, co-implantation of cancer-associated fibroblasts (CAFs) along with cancer cells increased take rate and accentuated disease course4. The presence of these fibroblasts may even promote formation of immortalised, otherwise non-tumorogenic, prostatic cells1. Furthermore both Tumor Growth Factor beta (TGF-ß) and platelet derived growth factor (PDGF) signalling seem to be involved in the activation of CAFs. Although the brain is largely devoid of fibroblasts, it is conceivable that other cell types, such as reactive astrocytes, can sustain tumor growth in the CNS. However, due to the infiltrative growth of these tumours any biopsy will contain a mixture of cancer cells and stromal cells, including reactive astrocytes. Therfore, the lack of specific markers for these cells has made it difficult to investigate their precise role during tumour growth. In this project, we have developed a fluorescent and immuno-deficient mice strain that provide a tool for separating stromal and malignant cells. Using this tool, we intend to characterise the stromal cells and their potential cytogenetic abnormalities, methylation patterns, gene expression profiles and protein signatures. This information will subsequently be applied to tailor drug regimens that target the brain tumor-stroma interactions.