Glioblastoma is the most aggressive and the most common type of primary brain cancer, and despite treatment consisting of surgical tumor removal, radiotherapy and chemotherapy, glioblastoma patients have a poor prognosis with a median survival of 15 months and a five-year survival of 5-7 %. Identifying novel prognostic factors and therapeutic targets are of utmost importance in order to identify the fraction of patients with longest survival and in order to improve the treatment response. The glial cell line-derived neurotrophic factor (GDNF) is an important neuronal growth factor, which plays an essential role in neuronal development and maintenance. Four GDNF family receptors have been identified, namely GDNF family receptor alpha 1-4 (GFRA1-4). The rearranged during transfection (RET) receptor is the primary downstream co-receptor for the GDNF family in neuronal cells. Expression of GDNF and GFRA1 has previously been found in human glioblastoma tissue samples, and in experimental studies, GDNF and GFRA1 have been shown to increase proliferation, migration, angiogenesis and increase chemoresistance in glioblastoma cells.The overall aim of this PhD thesis was to investigate the expression of GDNF, the GDNF family of receptors and RET in human glioblastoma and the prognostic potential of GFRA1 and GFRA2 in glioblastoma. We also investigated the role of GDNF and GFRA1 in resistance to treatment with radiotherapy and chemotherapy in glioblastoma cells in order to investigate whether GDNF-GFRA1 signaling represents a therapeutic target for optimizing chemotherapy and radiotherapy for treatment of glioblastoma patients.In Study I, the expression of GDNF, the GDNF family receptors GFRA1-4 and the RET receptor were investigated in human glioblastoma tissue samples. Varying levels of GDNF protein expression was present in frozen tumor tissue from 22 of the 27 patients investigated, while varying levels of GDNF mRNA was present in all of the formalin-fixed paraffin embedded (FFPE) tumor tissue from 10 patients investigated. Double stainings revealed high GDNF mRNA expression in GFAP-positive tumor cells and limited mRNA expression in IBA1-positive microglia and macrophages and CD34-positive endothelial cells. Receptor immunostainings showed strong expression of GFRA1, limited expression of GFRA2 and very low expression of GFRA3, GFRA4 and RET. Immunofluorescence double stainings showed strong GFRA1 expression in GFAP positive tumor cells and in less than 20 % of stem-like tumor cells positive for either OLIG2 or SOX2. GFRA1 expression was very low in IBA1-positive microglia and macrophages and absent in CD34-positive endothelial cells. The results suggest that GDNF mainly signal in an autocrine and paracrine manner to GFRA1-expressing tumor cells, including tumor stem cells. Furthermore, as RET expression was mostly absent, our results suggest that GDNF-GFRA1 signaling is RET-independent in glioblastoma.In study II, the prognostic value of of GFRA1 and GFRA2 and co-expression with OLIG2 and IBA1, respectively, was further investigated. A fluorescent multiplex staining protocol was developed and stainings were performed on a glioblastoma patient cohort containing tumor tissue from 181 glioblastoma patients. The stained slides were scanned and spectrally unmixed digitally, and marker classifications and quantifications were performed. The results showed a wide range of GFRA1 expression. GFRA2 expression was only present in tumor tissue from a single patient. A small fraction of OLIG2 positive tumor stem cells also expressed GFRA1. Co-expression of GFRA1 and IBA1 was very low. There was no association between overall survival and expression of GFRA1 or co-expression of GFRA1 and OLIG2. Likewise, OLIG2 and IBA1 did not provide individual prognostic value. Even though GFRA1 was not associated with survival time, it was often expressed, and may still represent a therapeutic target for improved outcome for glioblastoma patients.In study III, we investigated the role of GDNF and GFRA1 in resistance to treatment with either chemotherapy or irradiation in patient-derived glioblastoma cells. We saw an upregulation in mRNA expression of GDNF and GFRA1 after treatment with either temozolomide and lomustine, respectively, or irradiation. Using the CRISPR CAS9 technology, stable transfection models were created and validated for two of the cell lines with knockout of either GDNF or GFRA1. In the GDNF knock-out cell lines, the cells became sensitized to treatment with either temozolomide or lomustine, and exogenous addition of GDNF reversed the effect and desensitized the cells. GDNF knockout did not sensitize the cells to irradiation treatment. When GFRA1 was knocked out, the cells became sensitized to treatment with either temozolomide, lomustine or irradiation. The results suggest that GDNF and GFRA1 are mediators of resistance to chemotherapy, and that inhibition of GDNF-GFRA1 signaling may sensitize the tumor cells to chemotherapy treatment and irradiation treatment, and thus improve the efficacy of treatment in glioblastoma patients.In conclusion, the studies presented in this thesis show that GDNF and GFRA1 are widely expressed in human glioblastomas, mainly in tumor cells. GDNF is also expressed in a small number of micro-glia and macrophages and endothelial cells, and GFRA1 is expressed in a subset of stem-like tumor cells. Due to very limited RET expression, we suggest that GDNF-GFRA1 signaling is RET-independent. GFRA1 expression did not carry prognostic value in glioblastoma patients. The results from the experimental part of the study suggest that GDNF-GFRA1 play a role in resistance to anti-neo-plastic treatment in glioblastomas. Further elucidating the mechanisms of GDNF-GFRA1 mediated treatment resistance in glioblastoma tumors can pave the way for targeted treatment that may increase the efficacy of the current standard treatment with chemotherapy and radiotherapy.

Glioblastom er den mest aggressive og hyppigst forekommende type af primær hjernekræft, og på trods af behandling bestående af kirurgisk tumorfjernelse, strålebehandling og kemoterapi har glioblastompatienter en dårlig prognose med en medianoverlevelse på 15 måneder og en femårs over-levelse på 5-7 %. Det er derfor yderst vigtigt at identificere nye prognostiske faktorer og terapeutiske mål for at identificere andelen af patienter med den længste overlevelse og for at forbedre behandlingen. Den gliacelle-deriverede neurotrofiske vækstfaktor GDNF (glial cell line-derived neurotrophic factor) er en neuronal vækstfaktor, som spiller en væsentlig rolle i neuronal udvikling og vedligehold. Der er blevet identificeret fire GDNF-familiereceptorer, GDNF-familiereceptor alfa 1-4 (GFRA1-4). RET (rearranged during transfection) receptoren er den primære co-receptor for GDNF-familien i neuronale celler. Der er tidligere blevet fundet ekspression af hhv. GDNF og GFRA1 i glioblastom-væv fra patienter, og i eksperimentelle undersøgelser har man fundet at GDNF og GFRA1 kan øge celledeling, migration, karnydannelse og øge resistens mod kemoterapi i glioblastomceller.I studie I blev ekspressionen af GDNF, GDNF-familiereceptorerne GFRA1-4 og RET-receptoren undersøgt i glioblastomvæv fra patienter. Der blev fundet varierende niveauer af GDNF protein-ekspression i frossent tumorvæv fra 22 af de 27 undersøgte patienter, mens varierende niveauer af GDNF mRNA var til stede i alle vævene ved undersøgelse på formalin-fikseret paraffinindstøbt tumorvæv fra 10 patienter. Vi fandt høj GDNF mRNA-ekspression i GFAP-positive tumorceller og begrænset mRNA-ekspression i IBA1-positive mikroglia og makrofager og i CD34-positive endotel-celler. Receptor-immunfarvninger viste kraftig ekspression af GFRA1, begrænset ekspression af GFRA2 og meget lav ekspression af GFRA3, GFRA4 og RET. Immunfluorescens-dobbeltfarvninger viste kraftig GFRA1-ekspression i GFAP-positive tumorceller samt GFRA1 ekspression i mindre end 20 % af tumorceller positive for tumorstamcelle-markørerne OLIG2 eller SOX2. Ekspression af GFRA1 var meget lav i IBA1-positive mikroglia og makrofager og helt fraværende i CD34-positive endotelceller. Resultaterne tyder på, at GDNF hovedsageligt signalerer på en autokrin og parakrin måde målrettet GFRA1-udtrykkende tumorceller, herunder tumorstamceller, i glioblastomer. Endvidere tyder vores resultater på at GDNF-GFRA1-signalering er RET-uafhængig i glioblastomer, i det RET-ekspression var meget begrænset.I studie II blev den prognostiske værdi af GFRA1 og GFRA2 og samtidige ekspression af henholdsvis OLIG2 og IBA1 undersøgt yderligere. En fluorescens-baseret multiplex-farvningsprotokol blev udviklet, og farvninger blev udført på en glioblastom-patientkohorte indeholdende tumorvæv fra 181 glioblastom-patienter. De fluorescens-farvede glas blev scannet og spektralt justeret digitalt, hvorefter markørerne blev klassificeret og kvantificeret. Resultaterne viste en stor variation i GFRA1-ek-spression. GFRA2-ekspression var kun til stede i tumorvæv fra en enkelt patient. En lille del af OLIG2-positive tumorstamceller udtrykte også GFRA1. Co-ekspression af GFRA1 og IBA1 var meget lav. Der var ingen sammenhæng mellem samlet overlevelse og ekspression af GFRA1 eller samtidig ekspression af GFRA1 og OLIG2. Ligeledes havde henholdsvis OLIG2 og IBA1 ikke individuel prognostisk værdi. Selvom GFRA1 ikke var associeret med overlevelsestid, blev det ofte udtrykt og kan muligvis være et mål for behandling hos glioblastom-patienter.I studie III undersøgte vi om henholdsvis GDNF og GFRA1 har betydning for resistens mod behandling med enten kemoterapi eller stråleterapi i patient-deriverede glioblastomceller. Vi så en opregulering i mRNA-ekspressionen af GDNF og GFRA1 efter behandling med henholdsvis kemoterapi-medikamenterne temozolomid og lomustin eller bestråling. Ved hjælp af CRISPR CAS9-teknologien blev der lavet stabile transfektionsmodeller for to af cellelinjerne med knockout af enten GDNF eller GFRA1. I GDNF knockout-cellelinjerne blev cellerne sensibiliserede over for behandling med enten temozolomid eller lomustin, og eksogen tilsætning af GDNF ophævede effekten og desensibiliserede cellerne. GDNF knockout sensibiliserede ikke cellerne over for strålebehandling. Ved GFRA1 knock-out blev cellerne sensibiliserede over for behandling med enten temozolomid, lomustin eller bestråling. Resultaterne tyder på, at GDNF og GFRA1 er mediatorer af resistens over for kemoterapi, og at blokering af GDNF-GFRA1-signalering kan sensibilisere tumorcellerne over for kemoterapibehandling og strålebehandling og dermed gøre behandlingen af glioblastom-patienter mere effektiv.Baseret på de studier, som præsenteres i denne afhandling, kan vi konkludere at GDNF og GFRA1 er bredt udtrykt i humane glioblastomer, primært i tumorceller. GDNF udtrykkes også i et lille antal mikroglia og makrofager og endotelceller, og GFRA1 udtrykkes i en undergruppe af tumorstamceller. Resultaterne peger endvidere på at GDNF-GFRA1 signaling er uafhængig af RET, da RET-ekspressionen var meget begrænset. GFRA1-ekspression var ikke associeret med overlevelse hos glioblastompatienter. Resultaterne fra det eksperimentelle studie peger i retning af at GDNF-GFRA1 spiller en rolle i resistens mod antineoplastisk behandling i glioblastomer. Yderligere belysning af mekanismerne for GDNF-GFRA1-medieret behandlingsresistens i glioblastom-tumorer kan bidrage til at udvikle mere målrettet behandling, der kan øge effektiviteten af den nuværende standardbehandling med kemoterapi og strålebehandling.