The molecular pathomechanisms of major depressive disorder (MDD) are still not completely understood. Here, we follow the hypothesis, that mitochondria dysfunction which is inevitably associated with bioenergetic disbalance is a risk factor that contributes to the susceptibility of an individual to develop MDD. Thus, we investigated molecular mechanisms related to mitochondrial function in induced neuronal progenitor cells (NPCs) which were reprogrammed from fibroblasts of eight MDD patients and eight non-depressed controls. We found significantly lower maximal respiration rates, altered cytosolic basal calcium levels, and smaller soma size in NPCs derived from MDD patients. These findings are partially consistent with our earlier observations in MDD patient-derived fibroblasts. Furthermore, we differentiated MDD and control NPCs into iPS-neurons and analyzed their passive biophysical and active electrophysiological properties to investigate whether neuronal function can be related to altered mitochondrial activity and bioenergetics. Interestingly, MDD patient-derived iPS-neurons showed significantly lower membrane capacitance, a less hyperpolarized membrane potential, increased Na⁺ current density and increased spontaneous electrical activity. Our findings indicate that functional differences evident in fibroblasts derived from MDD patients are partially present after reprogramming to induced-NPCs, could relate to altered function of iPS-neurons and thus might be associated with the aetiology of major depressive disorder.

Conventional 2D cultures are commonly used in cancer research though they come with limitations such as the lack of microenvironment or reduced cell heterogeneity. In this study, we investigated in what respect a scaffold-based (Matrigel™) 3D culture technique can ameliorate the limitations of 2D cultures. NGS-based bulk and single-cell sequencing of matched pairs of 2D and 3D models showed an altered transcription of key immune regulatory genes in around 36% of 3D models, indicating the reoccurrence of an immune suppressive phenotype. Changes included the presentation of different HLA surface molecules as well as cellular stressors. We also investigated the 3D tumor organoids in a co-culture setting with tumor-infiltrating lymphocytes (TILs). Of note, lymphocyte-mediated cell killing appeared less effective in clearing 3D models than their 2D counterparts. IFN-γ release, as well as live cell staining and proliferation analysis, pointed toward an elevated resistance of 3D models. In conclusion, we found that the scaffold-based (Matrigel™) 3D culture technique affects the transcriptional profile in a subset of GBM models. Thus, these models allow for depicting clinically relevant aspects of tumor-immune interaction, with the potential to explore immunotherapeutic approaches in an easily accessible in vitro system.

Brain-tumor-initiating cells (BTICs) of proneural and mesenchymal origin contribute to the highly malignant phenotype of glioblastoma (GB) and resistance to current therapies. BTICs of different subtypes were challenged with oxidative phosphorylation (OXPHOS) inhibition with metformin to assess the differential effects of metabolic intervention on key resistance features. Whereas mesenchymal BTICs varied according to their invasiveness, they were in general more glycolytic and less responsive to metformin. Proneural BTICs were less invasive, catabolized glucose more via the pentose phosphate pathway, and responded better to metformin. Targeting glycolysis may be a promising approach to inhibit tumor cells of mesenchymal origin, whereas proneural cells are more responsive to OXPHOS inhibition. Future clinical trials exploring metabolic interventions should account for metabolic heterogeneity of brain tumors.

The phosphatase chronophin (CIN/PDXP) has been shown to be an important regulator of glioma cell migration and invasion. It has two known substrates: p-Ser3-cofilin, the phosphorylated form of the actin binding protein cofilin, and pyridoxal 5′-phosphate, the active form of vitamin B6. Phosphoregulation of cofilin, among other functions, plays an important role in cell migration, whereas active vitamin B6 is a cofactor for more than one hundred enzymatic reactions. The role of CIN has yet only been examined in glioblastoma cell line models derived under serum culture conditions. We found that CIN is highly expressed in cells cultured under non-adherent, serum-free conditions that are thought to better mimic the in vivo situation. Furthermore, the substrates of CIN, p-Ser3-cofilin and active vitamin B6, were significantly reduced as compared to cell lines cultured in serum-containing medium. To further examine its molecular role we stably knocked down the CIN protein with two different shRNA hairpins in the glioblastoma cell lines NCH421k and NCH644. Both cell lines did not show any significant alterations in proliferation but expression of differentiation markers (such as GFAP or TUBB3) was increased in the knockdown cell lines. In addition, colony formation was significantly impaired in NCH644. Of note, in both cell lines CIN knockdown increased active vitamin B6 levels with vitamin B6 being known to be important for S-adenosylmethionine biosynthesis. Nevertheless, global histone and DNA methylation remained unaltered as was chemoresistance towards temozolomide. To further elucidate the role of phosphocofilin in glioblastoma cells we applied inhibitors for ROCK1/2 and LIMK1/2 to our model. LIMK- and ROCK-inhibitor treatment alone was not toxic for glioblastoma cells. However, it had profound, but antagonistic effects in NCH421k and NCH644 under chemotherapy. In non-adherent glioblastoma cell lines cultured in serum-free medium, chronophin knockdown induces phenotypic changes, e.g. in colony formation and transcription, but these are highly dependent on the cellular background. The same is true for phenotypes observed after treatment with inhibitors for kinases regulating cofilin phosphorylation (ROCKs and LIMKs). Targeting the cofilin phosphorylation pathway might therefore not be a straightforward therapeutic option in glioblastoma.

We previously reported that DIRAS-3 is frequently inactivated in oligodendrogliomas due to promoter hypermethylation and loss of the chromosomal arm 1p. DIRAS-3 inactivation was associated with better overall survival. Consequently, we now investigated regulation and function of its family members DIRAS-1 and DIRAS-2. We found that DIRAS-1 was strongly downregulated in 65% and DIRAS-2 in 100% of analyzed glioma samples compared to non-neoplastic brain tissue (NNB). Moreover, a significant down-regulation of DIRAS-1 and -2 was detected in glioma data obtained from the TCGA database. Mutational analyses did not reveal any inactivating mutations in the DIRAS-1 and -2 coding regions. Analysis of the DIRAS-1 and -2 promoter methylation status showed significantly higher methylation in IDH-mutant astrocytic and IDH-mutant and 1p/19q-codeleted oligodendroglial tumors compared to NNB. Treatment of U251MG and Hs683 glioblastoma cells lines with 5-azacytidine led to significant re-expression of DIRAS-1 and -2. For IDH-wild-type primary gliomas, however, we did not observe significantly elevated DIRAS-1 and -2 promoter methylation levels, but still detected strong downregulation of both DIRAS family members. Additional analyses revealed that DIRAS-1 and -2 expression was also regulated by histone modifications. We observed a shift towards promoter heterochromatinization for DIRAS-1 and less promoter euchromatinization for DIRAS-2 in IDH-wild-type glioblastomas compared to controls. Treatment of the two glioblastoma cell lines with a histone deacetylase inhibitor led to significant re-expression of DIRAS-1 and -2. Functionally, overexpression of DIRAS-1 and -2 in glioblastoma cells translated into significantly higher sensitivity to lomustine treatment. Analyses of DNA damage markers revealed that DIRAS-1 and -2 may play a role in p53-dependent response to alkylating chemotherapy.