Genetic Reduction of Mitochondria Complex I Subunits is Protective against Cisplatin-Induced Neurotoxicity in Drosophila.

Chemotherapy-induced peripheral neuropathy (CIPN) is a prevalent side effect of widely used platinum-based anticancer agents. There are few predictable risk factors with which to identify susceptible patients. Effective preventive measures or treatments are not available. Here, we have used a model of CIPN in Drosophila melanogaster to identify genetic changes that confer resistance to cisplatin-induced neuronal damage but not in the rapidly dividing cells of the ovary. The Drosophila strain attP40, used as a genetic background for the creation of RNAi lines, is resistant to cisplatin damage compared with the similar attP2 background strain. attP40 flies have reduced mRNA expression of ND-13A, a component of the mitochondria electron transport chain complex I. Reduction of ND-13A via neuron-specific RNAi leads to resistance to the dose-dependent climbing deficiencies and neuronal apoptosis observed in control flies. These flies are also resistant to acute oxidative stress, suggesting a mechanism for resistance to cisplatin. The mitochondria of attP40 flies function similarly to control attP2 mitochondria under normal conditions. Mitochondria are damaged by cisplatin, leading to reduced activity, but attP40 mitochondria are able to retain function and even increase basal respiration rates in response to this stress. This retained mitochondrial activity is likely mediated by Sirt1 and peroxisome proliferator-activated receptor gamma coactivator-1α, and is key to cisplatin resistance. Our findings represent the potential for both identification of susceptible patients and prevention of CIPN through the targeting of mitochondria.SIGNIFICANCE STATEMENT Chemotherapy-induced peripheral neuropathy is a major, debilitating side effect of many platinum-based cancer drugs. There are few available screening tools to identify patients at risk, and there are no effective treatments. Here, we report a novel genetic change that confers resistance to cisplatin-induced neurotoxicity in a Drosophila model while preserving the toxic effect in rapidly dividing cells. This work has the potential to influence patient susceptibility testing and development of novel CIPN preventive treatments.

Cisplatin induces mitochondrial deficits in Drosophila larval segmental nerve.

Cisplatin is an effective chemotherapy drug that induces peripheral neuropathy in cancer patients. In rodent dorsal root ganglion neurons, cisplatin binds nuclear and mitochondrial DNA (mtDNA) inducing DNA damage and apoptosis. Platinum-mtDNA adducts inhibit mtDNA replication and transcription leading to mitochondrial degradation. Cisplatin also induces climbing deficiencies associated with neuronal apoptosis in adult Drosophila melanogaster. Here we used Drosophila larvae that express green fluorescent protein in the mitochondria of motor neurons to observe the effects of cisplatin on mitochondrial dynamics and function. Larvae treated with 10µg/ml cisplatin had normal survival with deficiencies in righting and heat sensing behavior. Behavior was abrogated by, the pan caspase inhibitor, p35. However, active caspase 3 was not detected by immunostaining. There was a 27% decrease in mitochondrial membrane potential and a 42% increase in reactive oxygen species (ROS) in mitochondria along the axon. Examination of mitochondrial axonal trafficking showed no changes in velocity, flux or mitochondrial length. However, cisplatin treatment resulted in a greater number of stationary organelles caused by extended pausing during axonal motility. These results demonstrate that cisplatin induces behavior deficiencies in Drosophila larvae, decreased mitochondrial activity, increased ROS production and mitochondrial pausing without killing the larvae. Thus, we identified particular aspects of mitochondrial dynamics and function that are affected in cisplatin-induced peripheral neuropathy and may represent key therapeutic targets.

Krüppel-like factor 11 differentially couples to histone acetyltransferase and histone methyltransferase chromatin remodeling pathways to transcriptionally regulate dopamine D2 receptor in neuronal cells.

The importance of Krüppel-like factor (KLF)-mediated transcriptional pathways in the biochemistry of neuronal differentiation has been recognized relatively recently. Elegant studies have revealed that KLF proteins are important regulators of two major molecular and cellular processes critical for neuronal cell differentiation: neurite formation and the expression of neurotransmitter-related genes. However, whether KLF proteins mediate these key processes in a separate or coordinated fashion remains unknown. Moreover, knowledge on the contribution of chromatin dynamics to the biochemical mechanisms utilized by these proteins to perform their function is absent. Here we report the characterization of two antagonistic, chromatin-mediated mechanisms by which KLF11, also known as TIEG2 (transforming growth factor-ß-inducible early gene 2) and MODY VII (maturity onset diabetes of the young VII), regulates transcription of the fopamine D2 receptor (Drd2) gene. First, KLF11 activates transcription by binding to a distinct Sp-KLF site within the Drd2 promoter (-98 to -94) and recruiting the p300 histone acetyltransferase. Second, Drd2 transcriptional activation is partially antagonized by heterochromatin protein 1 (HP1), the code reader for histone H3 lysine 9 methylation. Interestingly, KLF11 regulates neurotransmitter receptor gene expression in differentiating neuronal cell populations without affecting neurite formation. Overall, these studies highlight histone methylation and acetylation as key biochemical mechanisms modulating KLF-mediated neurotransmitter gene transcription. These data extend our knowledge of chromatin-mediated biochemical events that maintain key phenotypic features of differentiated neuronal cells.

The regulation of apoptosis by the downstream regulatory element antagonist modulator/potassium channel interacting protein 3 (DREAM/KChIP3) through interactions with hexokinase I.

The EF-hand protein, DREAM/KChIP3 (henceforth referred to as DREAM), regulates apoptosis by incompletely understood mechanisms. We demonstrate that in the presence of Ca2+, DREAM interacts with hexokinase I, a protein known to bind mitochondria and regulate apoptosis. A mutant DREAM protein construct incapable of binding Ca2+ does not associate with hexokinase I. The amino-terminal portion of DREAM is required for binding to hexokinase I, as a DREAM construct lacking the first 94 amino terminal residues fails to bind hexokinase I. Expression of DREAM in neuroblastoma cells enhances cisplatin mediated caspase-3 activity. Simultaneous expression of hexokinase I in such cells reduces DREAM-stimulated apoptosis. DREAM overexpression in neuroblastoma cells reduces hexokinase I localization on isolated mitochondria. The interaction of DREAM with hexokinase I may be important in the regulation of neuronal apoptosis.

Drosophila melanogaster: a new model to study cisplatin-induced neurotoxicity.

Platinum-based compounds are widely used and effective chemotherapeutic agents; however, sensory peripheral neuropathy is a dose-limiting and long term side effect for 20-30% of patients. A critical question is whether the mechanisms of cell death underlying clinical efficacy can be separated from the effects on neurons in order to develop strategies that prevent platinum-induced neuropathy. In rodent dorsal root ganglion neurons (DRG), cisplatin has been shown to bind and damage neuronal DNA, inducing apoptosis; however genetic manipulation in order to study mechanisms of this phenomenon in the rodent model system is costly and time-consuming. Drosophila melanogaster are commonly used to study neurological disorders, have DNA damage-apoptosis mechanisms homologous to mammalian systems, and have readily-available, inexpensive tools for rapid genetic manipulation. We therefore sought to develop adult Drosophila as a new model to study cisplatin-induced neurotoxicity. Adult Drosophila were exposed to 10, 25, 50, 100, 200 and 400 µg/ml cisplatin for 3 days and observed for fly survival and geotactic climbing behavior, cisplatin-DNA binding and cellular apoptosis. On day 3, 50 µg/ml cisplatin reduced the number of flies able to climb above 2 cm to 43% while fly survival was maintained at 92%. 100% lethality was observed at 400 µg/ml cisplatin. Whole fly platinum-genomic DNA adducts were measured and found to be comparable to adduct levels previously measured in rat DRG neurons. Brain, ovaries, kidney and heart harvested from cisplatin treated flies were stained for active caspase 3. Apoptosis was found in ovaries and brain but not in heart and kidney. Brain apoptosis was confirmed by transmission electron microscopy. Expression of the anti-apoptotic baculoviral protein, p35, in neurons using the GAL4-UAS system prevented cisplatin-induced apoptosis in the brain and restored climbing behavior. In conclusion, cisplatin-induced behavioral and apoptotic changes in Drosophila resemble those seen in mammals. Furthermore, the use of lethality and climbing assays combined with powerful gene manipulation, make Drosophila a suitable model to study mechanisms of cisplatin neurotoxicity.

Cisplatin induced mitochondrial DNA damage in dorsal root ganglion neurons.

Cisplatin is a platinum-based chemotherapeutic agent that induces peripheral neuropathy in 30% of patients. Peripheral neuropathy is the dose limiting side effect, which has no preventative therapy. We have previously shown that cisplatin induces apoptosis in dorsal root ganglion (DRG) sensory neurons by covalently binding to nuclear DNA (nDNA), resulting in DNA damage, subsequent p53 activation and Bax-mediated apoptosis via the mitochondria. We now demonstrate that cisplatin also directly binds to mitochondrial DNA (mtDNA) with the same binding affinity as nDNA. Cisplatin binds 1 platinum molecule per 2166 mtDNA base pairs and 1 platinum molecule per 3800 nDNA base pairs. Furthermore, cisplatin treatment inhibits mtDNA replication as detected by 5-bromo-2'-deoxy-uridine (BrdU) incorporation and inhibits transcription of mitochondrial genes. The relative reduction in mtDNA transcription is directly related to the distance the gene is located from the transcription initiation point, which implies that randomly formed platinum adducts block transcription. Cisplatin treated DRG neurons exhibit mitochondrial vacuolization and degradation in vitro and in vivo. Taken together, this data suggests that direct mtDNA damage may provide a novel, distinct mechanism for cisplatin-induced neurotoxicity separate from the established nDNA damage pathway.

Drosophila strain specific response to cisplatin neurotoxicity.

Drosophila melanogaster has recently been developed as a simple, in vivo, genetic model of chemotherapy-induced peripheral neuropathy. Flies treated with the chemotherapy agent cisplatin display both a neurodegenerative phenotype and cell death in rapidly dividing follicles, mimicking the cell specific responses seen in humans. Cisplatin induces climbing deficiencies and loss of fertility in a dose dependent manner. Drosophila sensitivity to cisplatin in both cell types is affected by genetic background. We show that mutation or RNAi-based knockdown of genes known to be associated with CIPN incidence in humans affect sensitivity of flies to CIPN. Drosophila is a promising model with which to study the effect of genetics on sensitivity to CIPN.

Mechanisms Underlying the Regulation of HP1γ by the NGF-PKA Signaling Pathway.

Heterochromatin protein 1 γ (HP1γ) is a well-known chromatin protein, which regulates gene silencing during the execution of processes associated with embryogenesis, organ maturation, and cell differentiation. We find that, in vivo, the levels of HP1γ are downregulated during nervous system development. Similar results are recapitulated in vitro during nerve growth factor (NGF)-induced neuronal cell differentiation in PC12 cells. Mechanistically, our experiments demonstrate that in differentiating PC12 cells, NGF treatment decreases the association of HP1γ to silent heterochromatin, leads to phosphorylation of this protein at S83 via protein kinase A (PKA), and ultimately results in its degradation. Genome-wide experiments, using gain-of-function (overexpression) and loss-of-function (RNAi) paradigms, demonstrate that changing the level of HP1γ impacts on PC12 differentiation, at least in part, through gene networks involved in this process. Hence, inactivation of HP1γ by different post-translational mechanisms, including reduced heterochromatin association, phosphorylation, and degradation, is necessary for neuronal cell differentiation to occur. Indeed, we show that the increase of HP1γ levels has the reverse effect, namely antagonizing neuronal cell differentiation, supporting that this protein acts as a barrier for this process. Thus, these results describe the regulation and participation of HP1γ in a novel membrane-to-nucleus pathway, through NGF-PKA signaling, which is involved in NGF-induced neuronal cell differentiation.

Mechano growth factor, a splice variant of IGF-1, promotes neurogenesis in the aging mouse brain.

Mechano growth factor (MGF) is a splice variant of IGF-1 first described in skeletal muscle. MGF induces muscle cell proliferation in response to muscle stress and injury. In control mice we found endogenous expression of MGF in neurogenic areas of the brain and these levels declined with age. To better understand the role of MGF in the brain, we used transgenic mice that constitutively overexpressed MGF from birth. MGF overexpression significantly increased the number of BrdU+ proliferative cells in the dentate gyrus (DG) of the hippocampus and subventricular zone (SVG). Although MGF overexpression increased the overall rate of adult hippocampal neurogenesis at the proliferation stage it did not alter the distribution of neurons at post-mitotic maturation stages. We then used the lac-operon system to conditionally overexpress MGF in the mouse brain beginning at 1, 3 and 12 months with histological and behavioral observation at 24 months of age. With conditional overexpression there was an increase of BrdU+ proliferating cells and BrdU+ differentiated mature neurons in the olfactory bulbs at 24 months when overexpression was induced from 1 and 3 months of age but not when started at 12 months. This was associated with preserved olfactory function. In vitro, MGF increased the size and number of neurospheres harvested from SVZ-derived neural stem cells (NSCs). These findings indicate that MGF overexpression increases the number of neural progenitor cells and promotes neurogenesis but does not alter the distribution of adult newborn neurons at post-mitotic stages. Maintaining youthful levels of MGF may be important in reversing age-related neuronal loss and brain dysfunction.

Neurotoxicity of oxaliplatin and cisplatin for dorsal root ganglion neurons correlates with platinum-DNA binding.

Cisplatin has been in use for 40 years, primarily for treatment of ovarian and testicular cancer. Oxaliplatin is the only effective treatment for metastatic colorectal cancer. Neurotoxicity occurs in up to 30% of patients and is dose-limiting for both drugs. The neuropathy is characterized by selective sensory loss in the extremities. Cisplatin treatment is associated with high levels of Pt-DNA binding and apoptosis of dorsal root ganglion (DRG) neurons. In this study, we directly compared the effects of oxaliplatin on DRG in vitro. Compared with cisplatin, oxaliplatin formed fewer Pt-DNA adducts following 6, 12, 24, and 48h (0.007ng Pt/mug DNA, 0.012ng/microg, 0.011ng/microg, 0.011ng/microg versus 0.014ng/microg, 0.022ng/microg, 0.041ng/microg, 0.030ng/microg), respectively. These findings closely correlated with data on cell survival where equimolar concentrations of oxaliplatin induced less cell death than cisplatin. Oxaliplatin-induced DRG death was associated with the morphological characteristics of apoptosis defined by 4'-6-diamidino-2-phenylindole and annexin/propidium iodide staining. Death was completely inhibited by the caspase inhibitor z-VAD-fmk. Our results demonstrate that both compounds cause apoptosis of DRG neurons but compared to cisplatin, oxaliplatin forms fewer Pt-DNA adducts and is less neurotoxic to DRG neurons in vitro.

Neurotoxicity to DRG neurons varies between rodent strains treated with cisplatin and bortezomib.

Chemotherapy-induced peripheral neuropathy (CIPN) is a major dose limiting side effect that can lead to long-term morbidity. Approximately one-third of patients receiving chemotherapy with taxanes, vinca alkaloids, platinum compounds or proteasome inhibitors develop this toxic side effect. It is not possible to predict who will get CIPN, however, genetic susceptibility may play a role. We explored this hypothesis using an established in vitro dorsal root ganglia neurite outgrowth (DRG-NOG) assay to assess possible genetic influences for cisplatin- and bortezomib-induced neurotoxicity. Almost all previous in vitro studies have used rats or mice. We compared DRG-NOG between four genetically defined, inbred mouse strains (C57BL/6J, DBA/2J, BALB/cJ, and C3H/HeJ) and one rat strain (Sprague Dawley). Our studies found differences in cisplatin and bortezomib-induced neurotoxicity between mouse and rat strains and between the different mouse strains. C57BL/6J and Balb/cJ DRG-NOG was more sensitive to cisplatin than DBA/2J and C3H/HeJ DRG-NOG, and all mouse strains were more sensitive to cisplatin than rat. Bortezomib induced a biphasic dose response in DBA/2J and C3H/H3J mice. C57BL/6J DRG-NOG was most sensitive and Balb/cJ DRG-NOG was least sensitive to bortezomib. Our animal data supports the hypothesis that genetic background may play a role in CIPN and care must be taken when rodent models are used to better understand the contribution of genetics in patient susceptibility to CIPN.

NGF rescues DRG neurons in vitro from oxidative damage produced by hemodialyzers.

Using dorsal root ganglion neurons (DRG), in vitro, we studied the effects of nerve growth factor (NGF) on a toxin extracted from ethylene oxide (EO) sterilized hemodialyzers. Tissue culture medium passed through dialyzers produced beading of DRG axons that was inhibited by increasing the concentration of NGF from 3.5 to 10 ng/ml. The antioxidant enzymes, catalase and glutathione peroxidase (GPx), prevented neurite beading while superoxide dismutase (SOD) alone did not. 3-amino-1,2,4-triazole (Az), an inhibitor of catalase blocked the protective effects of catalase and NGF. 1,3 bis[chloromethyl]-1-nitrosourea (BCNU) inhibits glutathione reductase, and reduces intracellular glutathione levels; it blocked the protective effects of NGF. Dialyzer treated medium was found to have increased peroxide content. In parallel experiments, NGF protected DRG neurons from hydrogen peroxide (H(2)O(2)) toxicity that was inhibited by Az and BCNU. NGF was also shown to upregulate glutathione in DRG neurons. We propose that EO gas used in the sterilization of hemodialyzers is responsible for the neurotoxicity and is most likely due to oxidative damage in DRG neurons. NGF protects DRG from this toxin by upregulating antioxidants such as catalase, GPx and GSH.

Bortezomib alters microtubule polymerization and axonal transport in rat dorsal root ganglion neurons.

Bortezomib is part of a newer class of chemotherapeutic agents whose mechanism of action is inhibition of the proteasome-ubiquitination system. Primarily used in multiple myeloma, bortezomib causes a sensory-predominant axonal peripheral neuropathy in approximately 30% of patients. There are no established useful preventative agents for bortezomib-induced peripheral neuropathy (BIPN), and the molecular mechanisms of BIPN are unknown. We have developed an in vitro model of BIPN using rat dorsal root ganglia neuronal cultures. At clinically-relevant dosages, bortezomib produces a sensory axonopathy as evidenced by whole explant outgrowth and cell survival assays. This sensory axonopathy is associated with alterations in tubulin and results in accumulation of somatic tubulin without changes in microtubule ultrastructure. Furthermore, we observed an increased proportion of polymerized tubulin, but not total or acetylated tubulin, in bortezomib-treated DRG neurons. Similar findings are observed with lactacystin, an unrelated proteasome-inhibitor, which argues for a class effect of proteasome inhibition on dorsal root ganglion neurons. Finally, there is a change in axonal transport of mitochondria induced by bortezomib in a time-dependent fashion. In summary, we have developed an in vitro model of BIPN that recapitulates the clinical sensory axonopathy; this model demonstrates that bortezomib induces an alteration in microtubules and axonal transport. This robust model will be used in future mechanistic studies of BIPN and its prevention.

An automated climbing apparatus to measure chemotherapy-induced neurotoxicity in Drosophila melanogaster.

We have developed a novel model system in Drosophila melanogaster to study chemotherapy-induced neurotoxicity in adult flies. Neurological deficits were measured using a manual geotactic climbing assay. The manual assay is commonly used; however, it is laborious, time-consuming, subject to human error and limited to observing one sample at a time. We have designed and built a new automated fly-counting apparatus that uses a "video capture-particle counting technology" to automatically measure 10 samples at a time, with 20 flies per sample. Climbing behavior was assessed manually, as in our previous studies, and with the automated apparatus within the same experiment yielding statistically similar results. Both climbing endpoints as well as the climbing rate can be measured in the apparatus, giving the assay more versatility than the manual assay. Automation of our climbing assay reduces variability, increases productivity and enables high throughput drug screens for neurotoxicity.

Antioxidants are necessary for myelination of dorsal root ganglion neurons, in vitro.

We have demonstrated that myelination of dorsal root ganglion (DRG) axons occurs in a fully defined, serum-free medium (B27). This implies that there may be components in B27 medium that support myelination. To determine which of the components in B27 were essential for myelination, we systematically removed components from B27 until myelination was lost. We added these components to a fully defined minimal medium (N2) that supports neuron survival but not myelination. When antioxidants were removed from B27, myelination was lost. However, the individual antioxidants did not induce myelination when added to N2 medium. Addition of ascorbic acid along with the B27 antioxidants was sufficient to induce myelination in N2 medium, which was enhanced by retinyl acetate. Removal of vitamin E from B27 caused a partial loss of myelination, and addition of vitamin E to N2 medium containing ascorbic acid induced partial myelination. Addition of serum to the B27 myelinating medium inhibited myelination completely. These results indicate that antioxidants are important for myelination, in vitro. Vitamin E may play an important role. Use of a serum-free medium may be beneficial for in vitro myelination studies because serum has unknown inhibitory effects.

Acute glucose deprivation leads to apoptosis in a cell model of acute diabetic neuropathy.

OBJECTIVE: Our aims were to better understand the mechanisms underlying peripheral neuropathy with diabetes mellitus and to test the hypothesis that acute lowering of glucose levels induces apoptosis in hypoxic neurons. METHODS: We used rat dissociated dorsal root ganglion (DRG) neurons incubated in a medium high in glucose concentration (700 mg%) and room air (PO2 150 torr). After 5 days, DRG neurons were placed in hypoxic conditions (PO2 7.6 torr) with a normal-glucose (100 mg%) or high-glucose (700 mg%) medium containing 3 or 100 ng/mL of nerve growth factor. Acute lowering of glucose levels under hypoxic conditions led to apoptosis of DRG neurons. Apoptosis was demonstrated by bis-benzimide staining for nuclear fragmentation, electron microscopy, DNA laddering, and TUNEL staining. Caspase 3 immunocytochemistry and inhibition of neuronal death by the caspase inhibitor z-VAD-fmk (100 microM) confirmed that death was apoptotic. RESULTS: Hypoxia-induced death was decreased when DRG neurons were maintained in high-glucose medium, suggesting that high levels of substrate protected against hypoxia. Apoptosis was completely prevented by increasing the concentration of nerve growth factor from 3 to 100 ng/mL and was partially prevented by the addition of the antioxidant alpha-lipoic acid (500 microM). CONCLUSIONS: This model provides a novel means for studying the pathogenesis and treatment of early stages of diabetic neuropathy.