70-kDa Heat Shock Protein Downregulates Dynamin in Experimental Stroke
A New Therapeutic Target?
Background and Purpose—The 70-kDa heat shock protein (Hsp70) protects brain cells in models of cerebral ischemia. Proteomic screening of mice subjected to middle cerebral artery occlusion identified dynamin as a major downregulated protein in Hsp70-overexpressing mice (Hsp70 transgenic mice). Dynamin-1 is expressed in neurons and participates in neurotransmission, but also transports the death receptor Fas to the cell surface, where it can be bound by its ligand and lead to apoptosis.
Methods—Mice were subjected to distal middle cerebral artery occlusion. Neuro-2a cells were subjected to oxygen glucose deprivation. Hsp70 transgenic and Hsp70-deficient (Hsp70 knockout) mice were compared with wild-type mice for histological and behavioral outcomes. Some mice and neuro-2a cell cultures were given dynasore, a dynamin inhibitor.
Results—Hsp70 transgenic mice had better outcomes, whereas Hsp70 knockout mice had worse outcomes compared with wild-type mice. This correlated with decreased and increased dynamin expression, respectively. Dynamin colocalized to neurons and Fas, with higher Fas levels and increased caspase-8 expression. Hsp70 induction in neuro-2a cells was protected from oxygen glucose deprivation, while downregulating dynamin and Fas expression. Further, dynamin inhibition was found to be neuroprotective.
Conclusions—Dynamin may facilitate Fas-mediated apoptotic death in the brain, and Hsp70 may protect by preventing this trafficking. Dynamin should be explored as a new therapeutic target for neuroprotection.
The 70-kDa inducible heat shock protein (Hsp70) is known to protect the brain from stressful stimuli, including stroke. Hsp70 also interferes with cell death pathways, such as apoptosis. Precise protective mechanisms are many, and protection by Hsp70 might lead to the identification of new therapeutic targets.1 To this end, we applied a proteomic approach to mice overexpressing Hsp70 (Hsp70 transgenic [Hsp70Tg]) subjected to experimental stroke and compared protein expression patterns to that of similarly injured wild-type (WT) mice. The result of this analysis showed marked reduction of dynamin in the brains of Hsp70Tg mice (Figure I and Table in the online-only Data Supplement).
Dynamin is a member of a family of guanine triphosphatase (GTPase) proteins largely known for their endocytic functions.2 Dynamin-1, the subject of this study, is exclusive to the brain.3 Recently, dynamin has been shown to translocate Fas protein from the Golgi apparatus to the cell surface, where it can be bound by its ligand, FasL,4 and subsequently trigger one of several extrinsic apoptosis pathways leading to caspase-dependent cell death.
Ischemic stroke leads to cell death via several pathways, including apoptosis. The intrinsic apoptotic pathway occurs within mitochondria,5 whereas the extrinsic pathway is receptor-mediated.6 Extrinsic apoptosis involves the engagement of death receptors located on the plasma membrane. Receptor ligation causes activation of caspase-8, leading to activation of effector caspase-3.7 After focal cerebral ischemia, the interaction between the Fas receptor and its ligand (FasL) initiates intracellular signaling cascades that ultimately terminate in caspase-dependent cell death after ischemic stroke.
Dynamin has not been studied extensively and has not been studied in brain ischemia. In the central nervous system, dynamin has been associated with endocytic processes and synaptic transmission.3 Prior related work has shown that a similar protein in the same family, dynamin-related protein-1 (DRP-1), may be important in brain ischemia because of its role in mitochondrial fission, and DRP-1 inhibition may be neuroprotective.8 However, to our knowledge, dynamin-1 has not been studied in brain ischemia, nor has its potential connection to Hsp70 been explored.
All experimental procedures in animals were approved by the local Institutional Animal Care and Use Committee (IACUC) and were in accordance with National Institutes of Health guidelines.
Male Hsp70Tg and Hsp70 knockout (Hsp70Ko) mice weighing 25 to 30 g were produced from breeder mice originally generated by the Dillmann (UCSD)9 and Pandita (Southwestern University) laboratories.10 Hemizygotic (Hsp70Tg) and homozygotic (Hsp70Ko) mice were compared with WT littermates.
Distal middle cerebral artery occlusion (dMCAO) was performed as described previously.11 Briefly, mice were anesthetized with isoflurane (5% for induction, 2% for maintenance via facemask) in a mixture of medical air/oxygen (3:1). The middle cerebral artery was permanently occluded by short coagulation proximal to the olfactory branch.
At the end of the experiments, mice were euthanized and transcardially perfused with normal saline and fixed in 4% paraformaldehyde (PFA) plus 20% sucrose, then frozen.
Infarct size was determined from cryosections (50 μm) of brains of animals survived 14 days and stained with cresyl violet. Infarct volume was determined as previously described.11
Neurological assessments were performed as previously described.11 All studies were recorded using a video camera, and scoring was performed by 2 different investigators.
The Bederson score was modified for use in mice, as previously reported11 (grade 0=no observable neurological deficit; grade 1=unable to extend the contralateral forelimb; grade 2=flexion of the contralateral forelimb; grade 3=mild circling to the contralateral side; grade 4=severe circling; and grade 5=falling to the contralateral side).
Mice were coaxed onto a 40-rung ladder. The number of forelimb faults were counted. Fewer foot falls indicated improved sensorimotor function.12
Adhesive Removal (Sticky Tape) Test
50-mm2 (4 mm diameter) adhesives were attached to the palm of each forepaw, and mice were observed for 2 minutes and scored for identifying the presence of and removal of the adhesive. Shorter removal times indicated improved neurological recovery.
Elevated Body Swing Test
To measure motor deficits, mice were suspended vertically by the tail with their heads elevated 3 inches above the test bench. A lateral swing was counted each time the animal moved its head >10 degrees away from the vertical axis.13
In Vitro Models
Neuro-2a (N2a) cells were purchased from the American-Type Culture Collection. Cultures were grown and maintained in DMEM (Cellgro) supplemented with 10% fetal bovine serum defined (Hyclone). Under humidified 5%, CO2 95% air atmosphere and at 37°C, cells were plated in 25-cm2 cell culture flasks (Corning). Media was changed 3 days after seeding and split twice a week. For experiments, cells were plated on 12-well dishes (1×106 cells/well).
Cells were exposed to oxygen glucose deprivation (OGD) by placing in an anoxia chamber (O2 tension <0.001%; Coy Laboratories) for 2 hours at 37°C. Media was removed and replaced with balanced salt solution lacking serum or glucose or oxygen. Control cultures were incubated at 37°C with balanced salt solution containing 5.5 mmol/L glucose. After OGD, glucose was added to each well to a final concentration of 5.5 mmol/L, and plates were returned to normoxia.
Cell death and viability were assessed by morphological assessment by light microscopy, vital staining (trypan blue; Sigma), and the tetrazolium dye MTT 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT; Sigma) assay.
17-N-Allylamino-17-demethoxygeldanamycin and Dynasore Administration
After control or OGD exposure, N2a cell cultures were treated with 60 μM 17-N-allylamino-17-demethoxygeldanamycin (Sigma) to induce Hsp70 or 30 μM dynasore (Dyna; Sigma), a dynamin inhibitor. Each compound was solubilized in 0.1% dimethyl sulfoxide (Sigma) and dimethylformamide (Sigma) and diluted in PBS.
Intracerebroventricular injections took place 30 minutes after dMCAO or sham surgeries. Anesthetized mice were placed in a stereotaxic frame and a total volume 2 μL of Dyna (0.3 mg/mL dose) or vehicle (0.1% dimethylformamide in PBS) was injected into the right lateral cerebral ventricle (stereotaxic coordinates: 1 mm caudal to bregma, 1.3 mm lateral to sagittal suture, and 2 mm in depth) at a speed of 0.5 μL/min via a burr hole. The needle was left in place for 5 minutes to allow drug diffusion into tissue before it was removed and the burr hole filled with bone wax.
Coimmunoprecipitation, Membrane Fractionation, and Immunoblotting
For immunoblots, ipsilateral hemispheres or N2a cells were homogenized and solubilized in radioimmunoprecipitation assay (RIPA) lysis buffer (Sigma) with protease inhibitor mixture (ROCHE). Coimmunoprecipitation was performed as previously described by our laboratory.14 Brain lysates were incubated with 2.5 mg of mouse anti-Hsp70 (Stressgen) or an IgG isotype control (2.5 mg normal mouse IgG; Santa Cruz), and the protein A/G PLUS-Agarose was collected. To detect Fas in the cell membrane, a Subcellular Protein Fractionation Kit (Thermo) was used according to the manufacturer’s instructions to separate and isolate membrane from tissue samples. Twenty micrograms protein samples were subjected to 10% SDS-polyacrylamide gel electrophoresis, then transferred to polyvinylidinene fluoride membranes (Millipore), and probed for the protein of interest by incubation with mouse anti-Hsp70 (1:1000; Stressgen) or dynamin-1 (1:1000; Santa Cruz) and rabbit anti-caspase-8 (1:1000; Santa Cruz) or Fas (1:1000; Santa Cruz) antibodies, followed by a horseradish peroxidase–conjugated secondary antibody. Blots were visualized using the enhanced chemoluminescence (ECL) system (Amersham) according to the manufacturer’s directions and imaged using LAS-4000 (Fuji).
After OGD, 1×106 N2a cells were washed once in PBS and blocked with flow cytometry staining buffer (eBioscience) for 10 minutes. Samples were stained with mouse anti-Fas (1:1000; Santa Cruz) followed by Alexa Fluor 488-conjugated goat anti-mouse IgG (1:1000; Invitrogen). Mouse IgG-PE (Santa Cruz) was used as an isotype control. N2a cells were sorted on FACSCalibur (BD Biosceinces), and data were analyzed using CellQuest (BD Biosciences).
Three days after dMCAO, brain sections (10 μm thick) were incubated with primary antibodies against mouse anti-dynamin-1 (1:1000; Santa Cruz) and rabbit anti-caspase-8 (1:500; Santa Cruz) or Fas (1:200; Santa Cruz), followed by secondary biotinylated antibodies and 3,3′-diaminobenzidine (Vector Laboratories) and counterstained with cresyl violet.
Brain sections were colabeled with antibodies against dynamin-1 plus cell markers for neurons (MAP-2), astrocytes (GFAP), microglia (CD11b), or Fas plus dynamin-1 to determine which cell populations expressed dynamin-1, followed by a fluorescent secondary antibody.
Terminal deoxynucleotidyl transferase dUTP nick end-labeling (TUNEL, ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit; Millpore) was used according to the manufacturer’s instructions to assess DNA fragmentation in neuron and described previously.15 Number of TUNEL-positive cells in the ipsilateral hemisphere were normalized and the numbers of cells counted from brains of sham animals.
All studies were randomized and assessments performed by investigators blinded to experimental conditions. All data were analyzed with standard statistical methods (t test; Systat Software, Inc). Data were presented as the mean±SE. Differences between the 2 groups were compared using an unpaired t test, and multiple comparisons were performed using 1-way analysis of variance followed by Bonferroni’s post hoc test. P≤0.05 was considered significant.
Hsp70 Protects Against Experimental Stroke
Consistent with previously published reports,16 Hsp70 overexpression led to improved neurological outcomes, whereas Hsp70 deficiency led to worsened outcomes compared with WT mice. Infarct size 14 days after dMCAO from WT, Hsp70Ko, and Hsp70Tg mice was compared. Infarct volumes quantified from cresyl violet–stained sections were significantly smaller in the Hsp70Tg mice compared with WT and Hsp70Ko mice (Figure 1A). Motor function was also improved among Hsp70Tg mice and worsened in Hsp70Ko mice (Figure 1B). Behavioral indices among sham controls did not reveal any baseline differences because of gene manipulation (data not shown).
Hsp70 Overexpression Decreases and Interacts With Dynamin-1
Using our in vitro model, we induced Hsp70 in N2a cells using 17-N-allylamino-17-demethoxygeldanamycin as we previously reported.17 17-N-Allylamino-17-demethoxygeldanamycin treatment led to decreased cell death after OGD (Figure IIA and IIB in the online-only Data Supplement) and 2-fold higher Hsp70 induction and 2.3-fold lower dynamin-1 expression relative to controls (Figure IIC in the online-only Data Supplement).
Consistent with observations from the proteomic analysis, Hsp70Tg mice showed reduced dynamin-1 levels compared with knockout and WT mice after brain ischemia, whereas Hsp70Ko mice showed the highest dynamin-1 levels (Figure 2A and 2B). This was seen especially within cells of the ischemic border zone. Double labeling showed that dynamin-1 was present in neurons but not astrocytes or microglia under noninjury conditions. In ischemic brains, dynamin-1, Fas, and Hsp70 were expressed in neurons (Figure III in the online-only Data Supplement).
Coimmunoprecipitation studies showed that Hsp70 was associated with dynamin-1 (Figure 2C). The highest Hsp70–dynamin-1 associations were seen in the transgenic mice and the least in the knockout mice. Colabeling for dynamin-1 and Fas showed that both colocalized to neuron-like cells and were the highest in Hsp70Ko brains and lowest in the transgenic brains (Figure 2D). Further, several dynamin-positive cells were found to be TUNEL-positive (Figure 2D).
Hsp70 Attenuates Fas-Mediated Apoptotic Death
To explore a link between Hsp70- and Fas-mediated cell death, dMCAO-exposed brains from WT, Hsp70Ko, and Hsp70Tg mice were stained for Fas. Fas-positive cells were found to be increased by brain ischemia, and brains of Hsp70Ko mice expressed more Fas compared with those of either WT or Hsp70Tg mice. Because Fas must be expressed on the cell’s surface to trigger cell death, immunoblots of membrane fractions showed that membrane-bound Fas was significantly decreased in the brain of Hsp70Tg mice, whereas Hsp70Ko mice had 4-fold higher Fas expression (Figure 3A). Further, N2a cells exposed to OGD showed increased Fas-positive cells, and treatment with 17-N-allylamino-17-demethoxygeldanamycin decreased these numbers by ≈42% (Figure 3B). The percentage of Fas+ cells were calculated as follows. R2 demarcated the fluorescein isothiucyanate (FITC) signal threshold for stained cells, excluding baseline FITC levels in unstained controls. R3 excludes FITC levels from nonspecific staining in control samples (secondary antibody only control). R3 thereby conservatively estimates the population of Fas-expressing cells within a given sample.
Because caspase-8 is activated downstream of Fas, but not during intrinsic (mitochondria) apoptosis, we assayed for caspase-8. Caspase-8 expression was the most decreased among ischemic Hsp70Tg brain samples compared with either WT or Hsp70Ko mice, whereas caspase-8 expression was highest among Hsp70Ko (Figure 3C). Increased dynamin-1, Fas, and caspase-8 protein also corresponded to an increased number of TUNEL-positive injured neurons, particularly in Hsp70Ko mice. Hsp70Tg mice showed significantly reduced numbers of TUNEL-positive cells by ≈20% to 25% (WT versus transgenic) and ≈35% to 40% (knockout versus transgenic), respectively (Figure 3D). These data suggest that dynamin contributes to Fas-mediated cell death in experimental stroke and that Hsp70 may protect by inhibiting dynamin expression and preventing surface expression of Fas.
Dynamin Inhibition Is Protective
To determine functional significance of dynamin, OGD-exposed N2a cells were treated with Dyna, a dynamin inhibitor which interferes with its GTPase activity.18 This led to improved cell viability and reduced cell death (Figure 4A and 4B) and 3.4-fold lower dynamin-1 expression relative to Veh groups (Figure 4C) and decreased numbers of Fas-positive N2a cells as demonstrated by flow cytometry (Figure 4D).
At the in vivo level, we first established that Dyna could engage target tissues in the brain. Texas red-conjugated Dyna (courtesy of Dr Nick Cairns, Combinix, Inc Mountain View, CA) was injected intracerebroventricular into uninjured animals. Fluorescent signals were observed in brain tissues showing that Dyna did in fact travel to and bind to brain cells (Figure IV in the online-only Data Supplement). Dyna administration did not seem to cause any seizures or obvious neurological impairment in uninjured mice.
Dyna treatment in dMCAO mice led to smaller strokes compared with vehicle (Figure 5A)-treated controls, with improved motor function (Figure 5B). Dynamin-1 expression was also decreased with Dyna treatment. To determine whether dynamin inhibition might prevent Fas translocation to the cell surface, immunoblots of cytoplasmic and membrane fractions showed that dMCAO led to increased expression of Fas in the membrane fraction, and Dyna treatment decreased this (Figure 5C).
To explore any synergistic actions of Hsp70 and dynamin inhibition, we administered Dyna to Hsp70Tg mice exposed to dMCAO. However, Dyna did not lead to any further lesion size reduction or improvement in neurological function (Figure V in the online-only Data Supplement).
Hsp70Tg mice exposed to stroke led to the identification of dynamin as a markedly downregulated protein. To our knowledge, this is the first report of dynamin in brain ischemia. We show that dynamin increases after stroke, along with Fas and caspase-8, and these proteins are decreased in mice overexpressing Hsp70, whereas the opposite was observed in Hsp70 deficiency. Further, we show that the proportion of membrane-bound Fas is increased after stroke, but is reduced by Hsp70Tg overexpression. This is consistent with prior work that showed that dynamin trafficks Fas to the cell’s surface4 and is in line with our hypothesis that Hsp70 prevents this trafficking (Figure 6). We also demonstrate for the first time that inhibition of dynamin improves outcome from experimental stroke.
Hsp70 has previously been shown to interfere with many aspects of the intrinsic apoptotic pathway by inhibiting caspase activation, preventing mitochondrial release of cytochrome c or increasing the anti-apoptotic protein Bcl-2.19 Less has been studied with respect to the extrinsic or receptor-mediated apoptotic pathways. Death receptors include Fas, which initiates cell death with the binding of Fas by its ligand FasL and leads to caspase-8 activation and apoptosis.20 Fas activation has been documented in ischemic stroke and related pathologies.21 Fas and its ligand have been documented in the brain after ischemia,22 and several studies, including some from our laboratory, have shown that interrupting this pathway is protective.22,23
In neurological disease, dynamin-1 deficiency has been linked to defects in γ-amino butyric acid transmission and epilepsy.2 However, intact dynamin has also been shown to have negative consequences. Dynamin has been linked to Alzheimer’s disease pathology,24 whereas its inhibition or deficiency led to decreased neuronal degeneration.25 Dynamin has been shown to traffick Fas protein from the Golgi apparatus through the Trans Golgi network to the cell surface where it can be bound by FasL4 and may suggest an additional role in brain cell death and degeneration. There are scant reports of dynamin in the ischemia literature. The most widely studied seems to be DRP-1 and its role in mitochondrial fission. In cardiac and renal ischemia, DRP-1 inhibition has been shown to be cytoprotective by preventing mitochondrial apoptosis.26,27 There are a few reports characterizing DRP-1 in experimental stroke,28 and DRP-1 inhibition decreased apoptosis.29
The relationship of dynamin to Hsp70 is unknown, but the present data indicate an inverse relationship between Hsp70 and dynamin in the ischemic brain. Hsp70 overexpression attenuated increases in dynamin after experimental stroke, whereas its deficiency did the opposite. Transgenic mice overexpressing Hsp70 had decreased Fas, caspase-8, and membrane-associated Fas. Based on our own experiments and reports in the literature, we postulate 2 potential mechanisms for this relationship. Hsp70 may regulate dynamin at the transcriptional level. Dynamin’s promoter region contains a sequence similar to those recognized by nuclear factor kappa B (NF-kB),30 as does Fas.31 Hsp70 is known to inhibit NF-kB’s transcriptional activity16 and may disrupt dynamin at the transcriptional level. However, it is also possible that Hsp70 may directly inhibit dynamin-dependent Fas translocation by containing dynamin in the cytosol through specific chaperone interactions.
Pharmacological inhibition of dynamin can also protect against stroke. Dynamin inhibition led to protection in our stroke models and reduced Fas expression on the cell surface. However, it should be noted that we administered dynasore intracerebroventricularly shortly after stroke onset. Although this paradigm was designed as a proof of concept study, future studies should further address the optimal timing and dosing of dynamin inhibitors.
We reveal a previously unknown mechanism of protection by Hsp70 in the ischemic brain and identify dynamin as a potential therapeutic target. Future studies may focus on more precise interactions between Hsp70 and dynamin and the development of dynamin inhibitors.
Sources of Funding
This study was funded by grants from the National Institutes of Health (NS40516), Department of Defense and the Veteran’s Merit Award (I01 BX000589) to Dr Yenari, American Heart Association Western States Affiliate Postdoctoral Fellowship (13POST14810019) to Dr Kim, and National Research Foundation of Korea grant from the Korean government (NRF-2014R1A2A2A01006556) to Dr Lee. Grants to Drs Yenari and Kim were administered by the Northern California Institute for Research and Education and supported by resources of the Veterans Affairs Medical Center, San Francisco, CA.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.012763/-/DC1.
- Received February 1, 2016.
- Revision received May 22, 2016.
- Accepted June 8, 2016.
- © 2016 American Heart Association, Inc.
- Lou X,
- Fan F,
- Messa M,
- Raimondi A,
- Wu Y,
- Looger LL,
- et al
- Ivanov VN,
- Ronai Z,
- Hei TK
- Broughton BR,
- Reutens DC,
- Sobey CG
- Hunt CR,
- Dix DJ,
- Sharma GG,
- Pandita RK,
- Gupta A,
- Funk M,
- et al
- Kawabori M,
- Kacimi R,
- Kauppinen T,
- Calosing C,
- Kim JY,
- Hsieh CL,
- et al
- Hartmann J,
- Blum R,
- Kovalchuk Y,
- Adelsberger H,
- Kuner R,
- Durand GM,
- et al
- Borlongan CV,
- Sanberg PR
- Zheng Z,
- Kim JY,
- Ma H,
- Lee JE,
- Yenari MA
- Ong SB,
- Subrayan S,
- Lim SY,
- Yellon DM,
- Davidson SM,
- Hausenloy DJ
- Qi X,
- Qvit N,
- Su YC,
- Mochly-Rosen D
- Yoo J,
- Lee SS,
- Jeong MJ,
- Lee KI,
- Kwon BM,
- Kim SH,
- et al
- Liu F,
- Bardhan K,
- Yang D,
- Thangaraju M,
- Ganapathy V,
- Waller JL,
- et al