|Home | About | Journals | Submit | Contact Us | Français|
Dexras1, a small G-protein localized predominantly to the brain, is transcriptionally upregulated by the synthetic glucocorticoid dexamethasone. It has close homology to the Ras subfamily, but differs in that Dexras1 contains an extended 7 kDa C-terminal tail. Previous studies in our laboratory showed that NMDA receptor activation, via NO and Dexras1, physiologically stimulates DMT1, the major iron importer. A membrane permeable iron chelator substantially reduces NMDA-excitotoxicity suggesting that Dexras1-mediated iron influx plays a crucial role in NMDA/NO-mediated cell death. We here report that iron influx is elicited by nitric oxide but not by other pro-apoptotic stimuli such as H2O2 or staurosporine. Deletion of Dexras1 in mice attenuates NO-mediated cell death in dissociated primary cortical neurons and retinal ganglion cells in vivo. Thus Dexras1 appears to mediate NMDA-elicited neurotoxicity via NO and iron influx.
The neurotransmitter glutamate acting via NMDA receptors elicits a variety of cellular alterations that are mediated by nitric oxide (NO). NO, in turn, can signal by activating guanylyl cyclase. Additionally, S-nitrosylation of cysteines in diverse proteins is increasingly appreciated as a major vehicle for NO actions (Foster et al., 2009;Hara and Snyder, 2007;Kim et al., 2005). One mode whereby NO is conveyed to its targets involves the binding of neuronal NO synthase (nNOS) to CAPON, a 55 kDa scaffold protein with a C-terminal domain that binds to the PDZ domain of nNOS (Jaffrey et al., 1998). CAPON then binds to Dexras1, a small GTPase that is a member of the Ras family and was discovered on the basis of selective induction by dexamethasone (Fang et al., 2000;Kemppainen and Behrend, 1998).
Dexras1 displays about 35% homology with the Ras family of proteins but differs in incorporating a 7 kDa C-terminal extension which it shares with Rhes (Ras Homologue Enriched in Striatum), a G protein highly enriched in the corpus striatum and involved in the neurotoxicity associated with Huntington’s Disease (Blumer et al., 2005;Subramaniam et al., 2009). Dexras1 plays a role in synchronizing circadian rhythms, as its deletion impairs circadian entrainment to light cycles and alters phase shifts to light (Cheng et al., 2004). A variety of influences upon adenylyl cyclase and G protein linked neurotransmitter influences have been reported for Dexras1. Also, Dexras1 can interact with FE65, an adaptor protein that occurs in a complex with the intracellular domain of the amyloid precursor protein (APP) (Cismowski et al., 2000;Lau et al., 2008;Nguyen and Watts, 2005).
NMDA receptor-mediated neurotransmission, via stimulation of nNOS, enhances Dexras1 activity. Thus, NMDA transmission leads to the binding of nNOS to CAPON, which in turn binds to Dexras1 with the ternary complex of proteins facilitating the S-nitrosylation of Dexras1 to activate its GTP binding activity (Fang et al., 2000).
Recently, we discovered a signaling cascade wherein Dexras1 binds to the peripheral benzodiazepine receptor-associated protein (PAP7), which in turn binds to the divalent metal transporter (DMT1), an iron import channel (Cheah et al., 2006). Stimulation of NMDA receptors activates nNOS leading to nitrosylation and activation of Dexras1, which through linkage to PAP7 and DMT1, physiologically enhances iron uptake. As iron is a potentially toxic substance, we wondered whether following cell stress, Dexras1 might mediate neurotoxicity via an excitotoxic pathway elicited by NMDA neurotransmission and iron entry. In the present study we have developed mice with targeted deletion of the gene for Dexras1. We demonstrate that deletion of Dexras1 markedly impairs iron uptake elicited by neurotoxic concentrations of NMDA and virtually abolishes NMDA neurotoxicity in cortical cultures. In intact mice NMDA destruction of retinal ganglion cells is abolished in Dexras1 knockout mice.
HEK 293T cells were maintained in DMEM with 10% fetal bovine serum (FBS), 2 mM L-Glutamine and 100U/ml penicillin-streptomycin (PS) at 37°C with 5% CO2 atmosphere in a humidified incubator. PC12 cells were maintained in DMEM with 10% FBS, 5% horse serum, 2 mM L-glutamine and 100U/ml PS in the same environment. All chemicals were purchased from Sigma, unless otherwise indicated.
The gene encoding mouse Dexras1, Rasd1, is located on chromosome 17 and consists of two exons. Rasd1+/− mice were generated at Ozgene. The targeting construct was based on the sequence of the C57BL/6 strain Rasd1 gene (GenBank accession number AF239157). The PGK-neo selection cassette was inserted downstream of exon 2. The PGK-neo cassette was flanked by flippase recognition target (FRT) sites and can be deleted with enhanced flippase recombinase. All the exons were flanked by loxP sites and can be deleted with Cre recombinase. All mice were maintained on a C57BL/6 background. Mice were housed in a 12 h light/dark cycle at an ambient temperature of 22°C and fed standard rodent chow. Animal protocols, approved by the Institutional Animal Care and Use Committee of University of Pennsylvania, were used in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.
Mice were genotyped by PCR analysis of genomic DNA from tail biopsies. Primer sets P1 (CGATCCGCGGCGAAGTCTAC) and P2 (GCGGTGCAAGTCGGGGCTCATCT) yielded a 579 bp product from the wild-type (RASD1+) allele. Reverse transcription-PCR (RT-PCR) analysis was used to assess whether the RASD1 transcript was missing in the knock-out mice. RNA was prepared from brain tissues obtained from RASD1 knock-out mice and their wild-type littermates by using TRIZOL reagent (Invitrogen). cDNA was prepared by using Oligo(dT) primers (Invitrogen) and the Omniscript reverse transcriptase kit (Qiagen). PCR was performed by using cDNA templates and primers P1 and P2 as above.
qPCR was performed as we have published previously (Hadziahmetovic et al., 2011), as was retinal iron quantification using bathophenanthroline sulfate.
Primary cortical neurons were dissected out of E16–E18 wild-type or Dexras knockout mice and plated in 6 well plates at 3 × 106 cells per well. Cells were maintained in Primary Neuron Media (Neurobasal media supplemented with B27 serum, 2 mM L-glutamine and 100U/ml PS) at 37°C with 5% CO2 atmosphere in a humidified incubator. Neurons were aged 14–20 days after plating before being used for iron uptake assays. Cells were treated with various concentration of NMDA for 10 minutes. Cells were then washed once with warm PBS. Iron uptake was measured as described before (Cheah et al., 2006).
A 5 mg/ml stock of MTT (Thiazolyl Blue Tetrazolium Bromide, Sigma) was diluted to a final concentration of 0.25 mg/ml in Hanks' Balanced Salt Solution (HBSS) buffer and added to cells after various treatments. Cells were incubated at 37 °C for 2–4 h, then the MTT reagent was removed and the cells were washed one time in HBSS. The samples were read in a spectrophotometer at OD 580 nm and OD 630 nm. The OD 580 nm – OD 630 nm reading was normalized to control and expressed as a percentage of cell viability.
The media was removed and the cells were rinsed once with PBS. PI was diluted into full media at a final concentration of 1 µg/ml, then incubated with cells at 37 °C for 10 min. Cells were washed once with PBS, then fixed with 4% paraformaldehyde in PBS for 30 min. To identify cells, the nuclei were stained with Hoescht stain at 1:15000 for 5 min. PI selectively stains dying cells nuclei (red), while Hoescht stains all cells (blue).
Retrograde labeling of RGCs was performed as described previously (Shindler et al., 2006). Briefly, Dexras KO male mice and wild type male mice were anesthetized by intraperitoneal injection with 2 mg ketamine (Sigma, St. Louis, MO) and 0.2 mg xylazine (Sigma). Holes were drilled through the skull above each superior colliculus through a midsagittal skin incision. 2.5 µL of 1.25% hydroxystilbamidine (Fluorogold; Molecular Probes, Eugene, OR) in sterile water was injected stereotactically into each superior colliculus 1 week before NMDA injection.
The method for intravitreal injections was adapted from previous studies (Liang et al., 2001). Dexras KO and wild type mice were anesthetized with ketamine/xylazine, and eyes were visualized under a dissecting microscope. The conjunctiva was lifted with forceps and cut from the sclera with Vannas scissors along the corneal limbus. Sclera was penetrated with a 30-gauge needle passed into the vitreous just posterior to the lens. NMDA (0.8 µL, 1.2 mM) (Sigma) in PBS was then injected into the vitreous using a 10 µL Hamilton syringe (Hamilton, Reno, NV) with a 32-gauge blunt tip needle. PBS alone was injected in the contralateral eye of each mouse. After injection, antibioitic ointment (Polysporin; Pfizer, New York, NY) was applied to each eye. The final NMDA concentration in the eye is estimated to be 200 nM, one sixth the concentration of the solution injected based on the volume injected and the average size of the vitreal space.
RGC numbers were counted as described previously (Shindler et al., 2006). Briefly, following sacrifice each eye was removed and fixed in 4% paraformaldehyde. Dissected retinas were flat mounted on glass slides, viewed by fluorescence microscopy (Eclipse 80i; Nikon, Tokyo, Japan), and photographed at 20× magnification in 12 standard fields: 1/6, 3/6, and 5/6 of the retinal radius from the center of the retina in each quadrant. RGC numbers shown in each experiment represent the total number of RGCs counted in 12 fields per eye. RGCs were counted by a blinded investigator using image analysis software (Image-Pro Plus 5.0; Media Cybernetics, Silver Spring, MD). Statistical comparisons of RGC numbers were performed by ANOVA.
PC12 cells display many neuronal properties and possess endogenous Dexras1 as well as PAP7 and DMT1 (Cheah et al., 2006). Accordingly, we selected these cells for studies of a possible role of Dexras1 in mediating the effects of neurotoxic levels of NO upon iron uptake and cell viability (Figure 1). We exposed cells to the NO donor GSNO (1 mM), H202 (200 µM) or staurosporine (1 µM). The concentrations at which we administered these agents are cytotoxic with 50% cell death 24 h after treatment (data not shown). At 2 h, when cell viability is normal, GSNO treatment doubles iron uptake, while H2O2 and staurosporine did not change uptake (Figure 1A). To assess a role for Dexras1 in the neurotoxic actions of these substances, we developed an shRNA construct which provides almost complete elimination of Dexras1 protein in PC12 cells (Figure 1B and C). Depletion of Dexras1 abolishes the cytotoxic actions of GSNO. Strikingly, loss of Dexras1 does not impair the cytotoxic actions of staurosporine or H2O2.
To evaluate the impact of Dexras1 on neurotoxicity in intact rodents, we developed mice with targeted deletion of Dexras1 (Figure 2). We employed a targeting vector to delete the complete open reading frame of Dexras1 (Figure 2A). PCR genotyping and RT-PCR confirm the complete genomic deletion of Dexras and the absence of Dexras mRNA in brain and liver (Figure 2B and C). The mutant mice appear grossly normal. There was no difference from wild-type in body size, weight, or locomotor activity at the age of 8 weeks. Mice lived up to 16–18 months and gross anatomic dissection of 16 month old mice reveals no apparent abnormalities in organs of the adult mutant mice. The absence of major aberrations in the mice corresponds to results from another group that also developed Dexras1 knockouts (Cheng et al., 2004).
To determine whether Dexras1 influences iron uptake associated with neurotoxicity, we exposed cortical cultures to various concentrations of NMDA, including cytotoxic levels (100 – 300 µM) for 10 min and measured iron uptake. At the cytotoxic concentrations of NMDA, we observe substantial increases in iron uptake with 300 µM NMDA eliciting a doubling of iron uptake (Figure 3A). No increase of iron uptake is apparent in Dexras1 deleted brain cultures. On the other hand, iron efflux was not affected in primary neurons from Dexras−/− mice (Fig3B). Moreover, we found that the expression levels of proteins involved in iron homeostasis such as APP, TfR, DMT1 and ferritin are similar in neuronal cultures between WT and Dexras −/− mice (Fig 3C). Thus, these findings establish a role for Dexras1 in mediating iron uptake in NMDA-mediated neurotoxic insults. These findings confirm and extend our earlier observations in PC12 cells that the stimulation of iron uptake by GSNO (100 µM) is prevented by depletion of Dexras1 utilizing RNA interference (Cheah et al., 2006).
We also examined cytotoxic actions of NMDA in the same cortical cultures (Figure 3D and E). We employed the well characterized regimen of NMDA treatment that elicits delayed neurotoxicity, thought to mimic events in vascular strokes (Dawson et al., 1991;Koh and Choi, 1988). Cortical cultures were exposed to 300 µM NMDA for ten min and then examined after 8 and 24 h. As has been reported by numerous investigators, no loss of cell viability is evident at 8 h, whereas at 24 h viability is decreased by more than 90%. Dexras1 deletion provides dramatic protection against this toxicity, with no evident loss of viability.
To extend these findings to intact animals, we evaluated retinal toxicity elicited by NMDA injection into the eyes of mice, a procedure known to selectively destroy retinal ganglion cells (Sucher et al., 1997). First we examined various genes involved in iron trafficking as well as markers for retinal cells. The qPCR cycle threshhold for Dexras1 in wild-type neural retinas was 29 (mid-range, out of 40 cycles), but was undetectable in Dexras1 knockouts. There were no significant differences for wild-type versus knockout for transferrin receptor, rhodopsin (rod photoreceptor specific), Thy1 (ganglion cell specific), or any of the DMT1 isoforms. These results are further confirmed by Western blotting examining the levels of protein expression for Ferritin and DMT1 (Fig 4B and C). Interestingly, the protein level of TfR is slightly higher in Dexras−/− and it may be a reflection of slightly lower levels of iron. We measured retinal iron levels directly, but the levels were variable and there was no statistically significant (Fig 4D) difference between the genotypes. Thus Dexras1 KO does not notably affect baseline iron levels or ganglion cell or rod photoreceptor numbers in the normal retina.
Five days following NMDA administration we observed a 60% reduction in numbers of retinal ganglion cells (Figure 5A and B). The Dexras1 deleted mice are completely protected from this neurotoxicity. Detailed histologic examination reveals no difference in the morphology of retinal ganglion cells or any other cell type of the retina between wild-type mice and Dexras1 deleted mice in the absence of NMDA administration (data not shown).
In the present study we have established a major role for Dexras1 in mediating both iron uptake and cell viability under NMDA-excitotoxic condition. The linkage between cell viability and iron uptake is selective for NO, as cytotoxic concentrations of GSNO increase iron uptake, whereas comparably toxic levels of H2O2 and staurosporine fail to do so. Moreover, the neurotoxic actions of GSNO, but not those of other agents, are prevented by Dexras1 depletion. Experiments using Dexras1 knockout mice provide compelling evidence for its importance in iron uptake and neurotoxicity. In cortical cultures of Dexras1 knockouts, stimulation by NMDA both of iron uptake and of neurotoxicity is abolished. In intact mice the marked loss of retinal ganglion cells elicited by NMDA is completely prevented in the Dexras1 knockouts. Our findings suggest that Dexras1 mediates NMDA neurotoxicity via its enhancement of iron uptake, as Dexras1 deletion prevents both processes. Moreover, previously we observed that NMDA neurotoxicity in cortical cultures is prevented by iron chelator treatment (Cheah et al., 2006).
The use of Dexras1 mutant mice substantially strengthens evidence for a signaling cascade wherein glutamate, acting via NMDA receptors, activates nNOS to form NO which nitrosylates and activates Dexras1 which, through a link to PAP7, increases iron uptake via DMT1. As reported earlier (Cheah et al., 2006), Dexras1 stimulation of iron uptake stems from its GTPase activity, as constitutively active Dexras1 is associated with enhanced stimulation of iron uptake.
Rhes (Dexras2) is the only Ras homologue that closely resembles Dexras1, with about 62 % amino acid homology (Falk et al., 1999). Whereas Dexras1 is induced by glucocorticoids, Rhes is selectively stimulated by thyroid hormone (Vargiu et al., 2001). Most strikingly, Rhes is uniquely concentrated in the corpus striatum, where its binding to mutant huntingtin is thought to underlie the selective damage to the corpus striatum in Huntington’s Disease (Subramaniam et al., 2009). Like Dexras1, Rhes does bind to PAP7 (Cheah et al., 2006) and so might regulate striatal iron deposition, which might participate in Huntington’s disease pathophysiology. Consistent with this possibility, iron levels in the striatum are increased in Huntington’s Disease patients (Dexter et al., 1991).
In our earlier study, we addressed physiologic regulation of iron transport by Dexras1 in response to NMDA-NO activation. Our present study focuses on the pathogenic actions of iron. Iron is well known to be toxic in excess and, in the brain, iron accumulation has been linked to numerous neurodegenerative diseases (McCord, 1998;Thomas and Jankovic, 2004). In particular, Salazar et al showed that DMT1, which is modulated by Dexras1, plays a critical role in iron-mediated neurodegeneration of Parkinson’s disease (Salazar et al., 2008;Snyder and Connor, 2009).
The dramatic protection from NMDA neurotoxicity elicited by deletion of Dexras1 may have therapeutic implications. Glutamate excitotoxicity has been implicated in the retinal ganglion cell loss of numerous optic neuropathies, including glaucoma (Dreyer, 1998;McCord, 1998) and dominant optic atrophy (Nguyen et al., 2011), as well as neuronal loss in multiple sclerosis models (Pitt et al., 2000), where significant ganglion cell loss occurs secondary to optic neuritis (Quinn et al., 2011;Shindler et al., 2008). Conceivably, drugs that selectively block Dexras1 may be neuroprotective in these optic neuropathies, vascular stroke and other neurodegenerative diseases. Because Dexras1 differs markedly in structure from other members of the Ras family and other small G proteins, it may be feasible to develop highly selective and safe inhibitors of Dexras1 function with therapeutic potential.
This work was funded by US Public Health Service Grant DA-00266 and Research Scientist Award DA-00074 (to S.H.S.), and National Institutes of Health Grant HD026979 and MH079614 (to S.F.K.)and EY015245 (to J.L.D.)