| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
NERVOUS SYSTEM CELL BIOLOGY
Department of Biochemistry and Molecular Biology and the Molecular, Cellular, and Integrative Neuroscience Program, Colorado State University, Fort Collins, Colorado
Submitted 9 February 2006 ; accepted in final form 21 May 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
-peptide oligomers also induce rods. Immunostaining of neurites indicates that these rods also contain the majority of the actin filament dynamizing proteins, actin-depolymerizing factor (ADF) and cofilin (AC). If the rods reappear within 24 h after the stress is removed, the neurite degenerates distal to the rod but with no increase in neuronal death. Here, rods were generated in cultured rat E18 hippocampal cells by overexpression of a green fluorescent protein chimera of AC. Surprisingly, we have found that, for a short period (
60 min) immediately after initial rod formation, the loss of mitochondrial membrane potential (
m) and ATP in neurites with rods is slower than in neurites without them. The m was monitored with the fluorescent dye tetramethylrhodamine methyl ester, and ATP was monitored with the fluorescent ion indicator mag-fura 2. Actin in rods is less dynamic than is filamentous actin in other cytoskeletal structures. Because m depends on cellular ATP and because ATP hydrolysis associated with actin filament turnover is responsible for a large fraction of neuronal energy consumption (50%), the formation of rods transiently protects neurites by slowing filament turnover and its associated ATP hydrolysis. actin dynamics; neurodegeneration; actin inclusions; neuroprotection; ischemia
-amyloid peptide1–42 (41). They vary from needle- to sausage-shaped inclusions. They may disappear when the insult is removed but return in a reduced number of cells within 24 h without affecting cell survival. Rod reappearance may contribute to synaptic dysfunction and/or loss (32, 48), characteristic of neurodegenerative disease because, distal to the rod site, mitochondria are inactive and the neurite degenerates. It is not difficult to understand why these rods are damaging when they reappear and persist for days. They commonly achieve a width great enough to supplant all microtubules and inhibit axonal transport (29, 41, 48). Moreover, the rods are shown here stably to sequester essentially all AC within the neurite and thus are likely to impair numerous neurite functions, ranging from vesicle transport (11, 59) to regulation of ionic homeostasis (14, 46). These functions depend on dynamically regulated actin assembly. The AC proteins are critical modulators of actin filament turnover (15, 57a, reviewed in Ref. 9). They accelerate turnover by enhancing depolymerization from the minus end of filaments and by severing them, thus creating plus ends for elongation and minus ends for preferential release of monomer actin (G-actin) complexed with AC. Regulation of these activities is complex. Acidification, binding to phosphatidylinositol 4,5-bisphosphate, tropomyosin binding to actin, and phosphorylation of AC all inhibit their activity. Inactivation by phosphorylation of the serine 3 residue of AC proteins occurs through TES (testicular) or LIM kinases (1, 4, 49, 68, 69, 73). Slingshot phosphatases (21, 53) and the haloacid dehalogenase phosphatase chronophin (22) reactivate AC proteins. Another level of regulation, involving the protein 14-3-3, coordinates phosphorylation/dephosphorylation with the presence of filamentous actin (23, 51, 63). The complexity of AC behavior extends to promoting assembly of AC-actin filaments, rather than disassembly, if the AC-to-actin ratio exceeds a critical level (16, 19, 31, 74). Finally, these proteins also inhibit monomer nucleotide exchange (27, 52), often a rate-limiting step for actin reassembly.
All cell treatments that generate rods have in common the effect of elevating the concentration of AC/G-actin complex. Microinjecting AC protein and overexpressing AC-green fluorescent protein (GFP) fusion proteins do this directly. ATP depletion for 20–30 min generates rods in >80% of cultured hippocampal neurons (48). In this case, ATP depletion elevates AC/G-actin complex by increasing the dephosphorylated (i.e., active) AC fraction and by increasing the pool of ADP bound to actin (actinADP) (6). AC proteins bind actinADP with higher affinity than actinATP (15, 16). In neurons, AC dephosphorylation is complete within 20 min after blocking ATP synthesis (48); similar AC activation has been reported in kidney cells (5) and in pancreatic endothelial cells (67). Three other rod inducers, glutamate, peroxide, and oligomers of A1–42, also likely activate AC by reducing ATP levels. In these cases, mitochondrial generation of ATP probably is impaired either by elevating mitochondrial Ca2+ to abnormal levels (42) or by generating more reactive oxygen species.
Persistence of rods probably contributes to the development of dementia and its hallmark, the amyloid plaque. Staining of brain tissue from patients with Alzheimer disease shows that >97% of dense core amyloid plaques have actin and/or AC rodlike inclusions nearby, but only 55% of the inclusions are near amyloid plaques (48). Furthermore, concentrations of amyloid -peptide as low as 10 nM will induce rods in cultured hippocampal neurons (41).
Even with the above findings, we still do not know whether rods are only an abnormal actin-AC aggregate and an early stage in the degeneration of neurites, eventually ending in cell death. The possibility exists that, because actin turnover can account for up to 50% of the ATP utilization in neurons (10), the initial sequestering of AC proteins into rods may be a temporary defense mechanism for stressed cells. The experimental approach we have taken to address this possibility (i.e., overexpression of a GFP chimera of Xenopus laevis XAC(wt) for rod generation) uniquely allows simultaneous real-time visualization of rods emerging in a fraction of neurites and monitoring mitochondrial potential and ATP level in all neurites. The resulting data support rods being more than merely a stress-induced protein aggregate: the short-term effects of rod generation are beneficial.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Cell culture and viral infection.
E18 rat hippocampal cells were acutely dissociated and plated on poly-D-lysine-coated no. 1 German glass coverslips (22 mm2) fixed to the bottom of drilled out 35-mm plastic dishes with aquarium sealant (Dow Corning). Sealant was cured for >48 h before cells were plated to minimize leeching of acid into the growth medium. Cells were grown in 2 ml of neurobasal medium with B27 supplement (Invitrogen) and 200 mM glutamine at a density of 1.5 x 104 cells/120 mm2 for 4 days. These were then infected at a multiplicity of infection of 50 with adenovirus for expression of GFP alone and XAC(wt)-GFP (45) or a monomeric red fluorescent protein chimera of actin (actin-mRFP); 1.7 ml of medium were removed and virus (7 µl) was added to the remaining medium in the well directly above the coverslip. Infected cells were incubated overnight before the 1.7 ml of medium were replaced. The actin-mRFP virus was made by excising human -actin cDNA from pEGFP-actin (Clontech, Mountain View, CA) and ligating it into pmRFP-C1 vector (13). The mRFP-actin cDNA was excised from this vector and ligated into the BglII site of pShuttle cytomegalovirus. Correctly oriented mRFP-actin-pShuttle cytomegalovirus was linearized and electroporated into BJ5183 Escherichia coli containing the pAdEasy-1 plasmid, and the virus was prepared according to He et al. (28) and Minamide et al. (47).
Electron microscopy. Hippocampal neurons were cultured for 3 days on polylysine-coated, formvar-covered gold grids. Some cultures were infected at a multiplicity of infection of 250 with adenovirus for expressing XAC(wt)-GFP, and both infected and uninfected cells were ATP depleted (10 mM NaN3/6 mM 2-deoxyglucose, 20 min) 3 days postinfection. The ATP depletion medium was washed out, and the cells were left overnight to recover. The cells were then fixed in 2% glutaraldehyde in PBS for 45 min at room temperature, washed quickly two times with PBS, and placed in an 8% sucrose-PBS solution. These cells were plunge frozen in liquid ethane, transferred to a vial containing 1% osmium tetroxide and 0.1% uranyl acetate in acetone, which was cooled to –196°C, and then warmed to –90°C in a freeze-substitution chamber. The cells were held at this temperature for 3 days, allowed to warm to room temperature over a 24-h period, rinsed three to five times in 100% acetone, infiltrated with Epox-araldite resin over 8–24 h, flat embedded, and polymerized at 60°C. Serial sections (150–200 nm) were cut on a Reichart ultramicrotome, collected on formvar-coated slot grids, and viewed on a JEOL 2000 electron microscope.
Monitoring mitochondrial membrane potential.
Cells, virally infected 2 days earlier, were examined for the initial appearance of rods before they were incubated for 15 min in 10 nM tetramethylrhodamine methyl ester (TMRM), a fluorescent dye whose uptake into mitochondria is a reversible function of mytochondrial membrane potential (m) (39, 56, 60). The TMRM-loaded culture was immediately transferred, without washing, to a 37°C, computer-controlled stage of an inverted Nikon Diaphot microscope fitted with Nikon x20 PlanApo objective, fluorescein filter cube, Texas red filter cube, a 0.5 neutral density filter, and computer-controlled phase and fluorescence shutters. A maximum of 30 min was spent recording positions of cells expressing XAC(wt)-GFP and capturing with a cooled charge-coupled device camera (PXL with 1400 Kodak chip; Roper Scientific, Tucson, AZ) three types of images: a phase image, a fluorescence image of the XAC(wt)-GFP, and a TMRM image. At 30-min intervals, this sequence of image acquisition was repeated two more times (see Data analysis section).
Monitoring ATP and Ca2+ levels.
Because ATP has 10-fold greater affinity for Mg2+ than does ADP or AMP (25, 38), the measurable increase in free intracellular Mg2+ concentration ([Mg2+]i) is a reliable indicator of ATP decline (38). We have demonstrated in embryonic rat cortical neurons that the time course of ATP decline, following blockage of ATP synthesis, is identical when monitored either with a luciferase assay that measures ATP in cell extracts (48) or with mag-fura 2 in live cells (10).
To determine whether intracellular free Ca2+ concentration ([Ca2+]i) exceeds 1 µM and thus contributes to the mag-fura 2 observations, the experiments were repeated with [Ca2+]i monitored through fura 2 and a fura 2 calcium imaging calibration kit. Cytoplasmic free Ca2+ does not contribute significantly to mag-fura 2 intensity unless the concentration exceeds 1 µM (24). The Ca2+-mag-fura 2 complex generates only 2% of the mag-fura 2 signal when ion concentrations are close to physiological (100 nM [Ca2+]i and 0.5 mM [Mg2+]i). This estimate is based on the affinity of Ca2+ for mag-fura 2 being 1,000-fold greater than that of Mg2+ and on the [Mg2+]i being 20,000-fold greater than that of [Ca2+]i. If [Ca2+]i reaches 1 µM and [Mg2+]i reaches 1.5 mM, as they may under pathological or stress conditions, then Ca2+ is estimated to generate 8% of the mag-fura 2 signal.
Cells, cultured and virally infected as described above, were incubated for 25 min in either membrane-permeable mag-fura 2-AM (10 µM) or fura 2-AM (10 µM), the acetoxymethyl ester form of the ratiometric fluorescent dyes used to monitor [Mg2+]i and [Ca2+]i, respectively (26). The excess dye was removed and cells were washed three times with PBS before replacement of growth medium and the immediate transfer of the culture dish to a 37°C, computer-controlled stage of a microscope equipped as described above. A maximum of 30 min was spent recording positions of cells expressing XAC(wt)-GFP. For both mag-fura 2 and fura 2 experiments, four types of images were acquired: a phase image, a fluorescence image of the XAC(wt)-GFP, and two emission images. A fluorescein filter cube was used for XAC(wt)-GFP images. A computer-controlled excitation filter wheel (Ludl Electronics Products, Exton, PA) with 340 ± 7.5 nm and 380 ± 7.5 nm filters, a 400 nm dichroic mirror, and a 460 ± 25 nm bandpass emission filter were used for dye images. At 30-min intervals, this sequence of image acquisition was repeated four times (see Data analysis section).
For the calibration of total ATP depletion and relative sensitivity of hippocampal neurons to inhibition of glycolysis and oxidative phosphorylation, the mag-fura 2 experiments described above were repeated but without virally infecting the cells. Time-lapse images were taken before and after the addition of inhibitors or their buffers.
Actin-AC rod stability.
For all rod stability studies, hippocampal cells were cultured, virally infected, and transferred to the microscope stage as described above. For the latrunculin A studies, phase and XAC(wt)-GFP images of nonneuronal cells (probably glial cells) were taken before and 30 min after addition of the reagent (final concentration 10 µM) (see Data analysis section).
For rod-stability studies using fluorescence recovery after photobleaching (FRAP) of XAC(wt)-GFP and actin-mRFP, a Zeiss LSM 5 Live confocal microscope with 488 nm and 532 nm diode lasers, LP 505 and LP 550 emission filters, and Alpha Plan x100/1.45 oil objective was used. A 1-µm-wide stripe was photobleached with 80% laser power for 1 s in neurites with and without rods. Because the Zeiss LSM 5 Live is an x-direction slit-scanning system, it allows only rectangular areas orientated perpendicular to the microscope stage x-axis to be photobleached. Cells were selected that permitted a zone perpendicular to the long axis of areas of interest to be bleached. Fluorescence recovery images (exposure 100 ms, 15% laser intensity) were analyzed after background subtraction by measuring average pixel intensity of bleached area and correcting for unintended photobleaching by measuring an area of similar size immediately adjacent to the bleached area (MetaMorph version 6.1 software, Universal Imaging, Downingtown, PA; and Microsoft Excel 2003).
Cofilin translocation to mitochondria. To detect cofilin translocation to mitochondria, cells, overexpressing XAC(wt)-GFP, were loaded with 1 µM MitoTracker red CM-H2XRos for 20–30 min, washed three times with PBS, and returned to growth medium. This dye is concentrated in active mitochondria, but, unlike TMRM, its binding is irreversible. Time-lapse fluorescence images of GFP were acquired every 45 min for 3 h as described above; accompanying MitoTracker red images were acquired with a Texas red filter cube.
Nondenaturing PAGE. The amount of actin displaced by latrunculin A from the ADF-MgATP-actin 1:1 complex was measured by nondenaturing gel electrophoresis. To form the starting complex, 5 µM MgATP-actin and 10 µM ADF were incubated in 2 mM Tris, pH 8.4, 0.5 mM DTT, 0.2 mM ATP, 0.21 mM MgCl2, and 0.2 mM EGTA for 10 min on ice. A 2-to-1 molar ratio of ADF to actin was used to drive most of the actin into complex. Increasing concentrations of latrunculin A were incubated with the complex for 20 min to 2 h at room temperature or overnight at 4°C. Some experiments were performed starting with latrunculin A-actin mixtures that were then incubated for 1 h at room temperature with ADF. The proteins were then separated on a 7.5% nondenaturing polyacrylamide gel at 4°C with 50 mM bicine, 40 mM triethanolamine, 0.2 mM ATP, and 0.2 mM EGTA, pH 8.4 as the running buffer. The gel was stained with Coomassie blue R250. Digitized gel images were obtained with a chilled charge-coupled device camera (Roper Scientific, Tucson, AZ), and band densities were analyzed using 1D Phoretix gel analysis software (Nonlinear Dynamics, Newcastle upon Tyne, UK).
Data analysis.
For the analyses of all images, measurements were taken with MetaMorph version 6.1 software (Universal Imaging) and transferred to either Microsoft Excel 2002 or MINITAB version 14 for statistical analysis. To monitor m, the background was subtracted before measurement of average TMRM intensity of each mitochondrial region with a line scan, a two-pixel-wide line drawn along the long axis of the region. Time-lapse imaging strongly suggested that individual mitochondria were identified, but we cannot be certain that some were not clusters. Infrequently, a large TMRM region splits into two pieces between time points. The line scan for each mitochondrial region at time point 1 was copied to images of the same cells taken at subsequent time points and repositioned as needed to compensate for movement.
To monitor intracellular ATP and Ca2+, background was subtracted before ratioing the two mag-fura 2 or fura 2 images and measuring the average ratioed pixel intensity along each neurite with a 2-pixel-wide line drawn from the cell body distally. Intensities of the line scan along the neurite were used to generate average curves of [Mg2+]i increase (i.e., ATP loss) as a function of time and distance from the somata. Beyond 50 µm from the somata, neurites with old rods were too thin to measure. Analysis of covariance was performed with MINITAB version 14.1, using a general linear regression model (Minitab, Summerville, SC).
Lamellipodial and rod areas were measured by subtracting background and thresholding the structures of interest. MetaMorph software (Universal Imaging) was used to transfer measurements of the area of these structures to Microsoft Excel for statistical analysis.
Correlation coefficients, calculated with Microsoft Excel, were used to determine the colocalization of XAC(wt)-GFP and mitochondria stained with Mito Tracker red in neurites. The intensities of the two fluorophors in single-pixel line scans tracing the neurites were drawn in MetaMorph and exported to Excel for analysis. The correlation coefficient is shown in Eq. 1
 | (1) |
 | (2) |
x,y) are mean and standard deviation, respectively, of each fluorophor over the entire line scan. |
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
m decline.
In noninfected cells, ATP depletion, after metabolic inhibition, induces rods presumably by dephosphorylating AC, thus increasing AC-actin complex (48). However, throughout this study, rods were generated by increasing the AC-actin complex through overexpression of XAC(wt)-GFP. Only fluorescently tagged AC allows the identification of emergent rods in live cell neurites. This identification is required for comparing in an individual cell the m and ATP level in neurites with and without emerging rods. Rods first appeared in hippocampal neurons 60 h after adenoviral infection with constructs for XAC(wt)-GFP; adenoviral-mediated expression of GFP alone produced only diffuse fluorescence. Figure 1a shows a typical cell at time point 1 with a few small rods generated through overexpression of an AC protein that elevates AC-actin complex levels. By time point 2 (30 min later; Fig. 1b), many more rods have appeared.
|
Figure 2 shows a typical hippocampal cell with fluorescent XAC-actin rods (a), a phase image of the same cell (b), and TMRM staining (c) of its active mitochondria at time point 1. Figure 2c shows line scans, marking all mitochondrial regions, in neurites with rods at time point 1. These line scans were transferred to an image of the same cell at time point 2 (Fig. 2d) and were used to determine the decline in m. The same procedure was followed for mitochondria in neurites with no rods from the same cell.
|
1 rod at time point 2 (i.e., neurites with newly formed or "young" rods), and 3) those with 1 rod at both time points 1 and 2 (i.e., neurites with more mature or "old" rods). Figure 3A shows that m is retained best in neurites that form rods between time points 1 and 2 (young rods). By time point 2, neurites with old rods show less retention of m than neurites with young rods. However, even neurites with old rods retain m more effectively than neurites with no rods. The greater decline in potential of mitochondria in neurites with no rods than neurites with young rods is not because of a difference in the m at time point 1; there is no difference (see Fig. 3B). However, between time points 2 and 3, the m drops faster in neurites with old rods than in neurites with no rods (see Fig. 3C). Hence, rod formation initially slows mitochondrial decline but with time accelerates it. Photobleaching is not responsible for the decline in TMRM intensity because we observed that the same number of exposures at 10-s intervals caused <10% intensity decrease. Neurites with old rods show a higher m at time point 1 than do neurites with young rods (see Fig. 3B). This finding is consistent with rods initially slowing the m drop because experiments were initiated soon after the first appearance of rods.
|
m would accompany preservation of ATP levels, we monitored ATP levels in neurites with rods over a period of 90 min. Overexpression of XAC(wt)-GFP leads to rod formation 60 h after viral infection. The cells were then loaded with mag-fura 2. Figure 4A shows the GFP fluorescence image of neurons with young rods merged with a phase image. Time-lapse phase, GFP fluorescence, and mag-fura 2 dual excitation (340 and 380 nm) images were taken at 30-min intervals. The increase in [Mg2+]i, i.e., ATP decline, was derived from the ratioed mag-fura 2 images generated for each time point (Fig. 4b). Values of ratioed fluorescence intensity in a 2-pixel-wide line, drawn along the neurite from cell body distally (Fig. 4b), were transferred to an electronic spreadsheet.
|
Analysis of the mag-fura 2 time-lapse images at 0- and 90-min time points indicates that ATP levels drop faster in neurites with no rods than in neurites with young rods. However, the ATP levels drop even faster in neurites with old rods than in neurites with no rods (see Fig. 5A). This decline becomes progressively more severe with distance from the somata. ATP levels in neurites with old rods are actually high near the somata, suggesting that rod impedance of axonal transport may preserve ATP levels proximal to their location and deplete it in regions distal to them. No ATP decline in any neurite class is seen at 30- and 60-min time points. Analysis of covariance with a general linear regression model yielded the estimate of mean, SE, and significance of difference shown in Fig. 5B. Data from Fig. 5A are replotted in Fig. 5C to show the ATP percentage remaining 90 min after the initiation of observation, assuming that a fourfold increase in [Mg2+]i (the maximum change observed at 50 min) corresponds to total ATP depletion (Fig. 5D). Previously, our group (10) demonstrated that this mag-fura 2 assay accurately indicates the time course of total ATP depletion by combined NaN3-2-deoxyglucose inhibition of ATP synthesis. The inhibition of only glycolysis (2-deoxyglucose treatment) causes complete reduction of ATP (Fig. 5D) because the addition of NaN3 to glycolysis-inhibited cells does not further reduce ATP. Moreover, glycolysis inhibition reduces ATP faster than does inhibition of oxidative phosphorylation (Fig. 5D). Thus embryonic rat hippocampal neurons, like chick sympathetic neurons (71), energetically rely primarily on glycolysis. The glycolytic generation of ATP may explain why neurite ATP levels do not follow more closely the decline in m.
|
Surprisingly, in neurites with old rods, [Ca2+]i nearest the somata is high enough to generate 8% of the mag-fura 2 signal in this region (Fig. 5E; see MATERIALS AND METHODS). Although mag-fura 2 intensity is much lower here than in distal regions, the high [Ca2+]i suggests that the true Mg2+-generated mag-fura 2 signal is even lower. The high [Ca2+]i and low [Mg2+]i near the cell body may reflect the high concentration of somal mitochondria. A fraction of them could be capable of retaining ATP levels but not capable of sequestering all the Ca2+ released by failing ones.
Actin-AC rods are stable structures. Hydrolysis of ATP associated with actin filament turnover consumes a significant portion of the total ATP used by unstimulated neurons (10). We probed the idea that immediately after rod formation these inclusions may delay ATP loss by sequestering actin and the actin dynamizing proteins ADF and cofilin (AC). To test this idea, we determined the stability of rods compared with other actin-based structures. Rods were generated as before by overexpression of XAC(wt)-GFP, and cells were treated with either latrunculin A or subjected to FRAP analysis (30). Latrunculin A is a membrane-permeable natural product that does not actively depolymerize filaments but rather passively sequesters actin monomers released by filaments undergoing turnover (18). It prevents monomer reassembly into F-actin (64). Analysis of rod area in images of nonneuronal cells, before and after latrunculin A addition (Fig. 6A), indicates that rods are more stable than the actin filaments within lamellipodia (Fig. 6B). With latrunculin A, rod area does not decline as does lamellipodial area (Fig. 6A and B), and rod intensity does not decline as does the intensity of rod-free neurites (Fig. 6C).
|
|
|
AC translocation to mitochondria.
Translocation of dephosphorylated cofilin, capable of binding actin, was demonstrated to be both necessary and sufficient for triggering cytochrome c release and initiating apoptosis in staurosporine-treated neuroblastoma cells (17). Therefore, another potential mechanism for the observed protection of m is the reduction of AC translocation to mitochondria in neurites with rods. To determine whether rods affect AC translocation to mitochondria, cells with rods, generated as above, were loaded with MitoTracker red, another fluorescent dye that is taken up only by active mitochondria. However, unlike TMRM, MitoTracker red is not released as the membrane potential drops. GFP and MitoTracker red images were taken at 30-min intervals (Fig. 9). Discrete regions of concentrated XAC(wt)-GFP and rods were circled, and these markings were transferred from the GFP to MitoTracker red fluorescence images. When GFP fluorescence completely overlapped mitochondrial fluorescence, XAC(wt)-GFP was judged to be colocalized with a mitochondrion. Only 1% of the mitochondria in neurites with rods had colocalized regions of concentrated XAC(wt)-GFP. In contrast, 12% of the mitochondria in neurites without rods had colocalized XAC(wt)-GFP; the two experiments performed yielded the same percentages. The mere sequestering of AC in rods may be enough to reduce its translocation to mitochondria and help preserve m. To determine the extent of increased colocalization of XAC(wt)-GFP and MitoTracker red with time in nonrod neurites, we calculated the correlation coefficients at each time point. These coefficients support the idea that the two species are colocalized rather than the possibility that XAC(wt)-GFP increases uniformly throughout the neurite and randomly overlaps mitochondria. A random overlapping would also produce yellow mitochondria in merged images, but the correlation coefficient would remain close to zero with time due to the (xi – µx) term in Eq. 2 (see Data analysis above) rather than increase as observed. The (yi – µy) term, representing irreversible MitoTracker red staining, remains relatively constant over time. Neurites without rods, judged to show XAC translocation to mitochondria, consistently showed an increase in correlation coefficient with time, whereas neurites with rods showed a decrease (Fig. 9). One would expect this correlation coefficient decrease if XAC is increasingly concentrated in rods over time.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
m rundown; however, after 60 min, neurites with rods lose m more quickly. In this study, rods were initially generated through overexpression of an AC chimera protein, XAC(wt)-GFP. Subsequent light microscopy probably accelerated rod generation because it produces reactive oxygen species, particularly in the presence of intrinsically fluorescent protein (20, 37, 42). Reactive oxygen species damage membranes, probably contributing to the mitochondrial potential decline observed and likely converting mitochondria from ATP producers, under normal conditions, into powerful ATP consumers in stressed cells. F1F0-ATPase is activated to prevent the collapse of the proton gradient (65). The resulting ATP drop exacerbates rod formation, generated initially through AC overexpression, but secondarily through dephosphorylation of AC. AC dephosphorylation contributes to abnormally high levels of AC-actin complex (48). However, those neurites with newly formed rods transiently show a reduced decline in ATP and m. Possible mechanisms for this short-term protective effect are discussed below. Immunostaining shows that only a fraction of the total actin in neurites is confined to rods, but the majority of the AC protein appears so sequestered (48). This protein sequestering first appears energetically beneficial. However, prolonged sequestering probably accelerates the decline because many cellular processes, including organelle transport and ionic homeostases, depend on actin dynamics (11, 36, 59). For example, the following ion regulatory proteins depend on actin dynamics for positioning or modulation of their activities: Na+/K+-ATPase, NHE1, Na+/Ca2+ exchanger, nonvoltage-dependent Na+ channels (14, 46), voltage-dependent Ca2+ channels (33), and NMDA receptors (58). Clearly, extended interruption in their function would produce conditions damaging to mitochondria, for example, elevated [Ca2+]i. Although reduction in actin dynamics initially slows ATP decline by limiting actin-associated ATP hydrolysis (10), prolonged reduction in actin dynamics would perturb Na+/K+-ATPase activity, leading to intracellular accumulation of Na+ and reversal of the normal Ca2+ extrusion by the Na+/Ca2+ exchanger (75). Thus it is not surprising that, compared with neurites with no rods, neurites with young rods lose ATP more slowly and neurites with old rods lose ATP more rapidly.
The FRAP experiments herein support the possibility that rods can briefly preserve ATP levels by reducing the rate of ATP hydrolysis associated with filament turnover: actin-mRFP recovery in a neurite region containing a rod is one-third that of a neurite without rods. The stability of the XAC(wt)-GFP bleached zone (Fig. 8) in a rod region supports the idea that AC proteins, key drivers of filament turnover, are stably sequestered in rods. AC recovery being slower than actin recovery is consistent with AC being more effectively sequestered in rods (48). The AC sequestration by rods eliminates a major player in actin dynamics, blocking its normal ability to recycle continuously to filaments and promote turnover.
Latrunculin A efficiently disperses F-actin-supported structures in lamellipodia of these cells (Fig. 6, A and B) despite the fact that it competes only weakly with an AC protein for actin (Fig. 7) and these cells have an excess of exogenous AC protein. The observed lamellipodial collapse could be predicted from the well-documented ability of latrunculin A to sequester released actin subunits in other cells and prevent reassembly of those subunits. However, to the best of our knowledge, the XAC(wt)-GFP increase into rods is unique. Because rods do not recover XAC(wt)-GFP in the absence of latrunculin A (FRAP data), latrunculin A seems responsible for the addition. We speculate that a pool of latrunculin A-actin-AC complexes is built up from rapidly turning over F-actin. Because it cannot add to normal F-actin structures (hence the lamellipodial dispersion), it is free to add to rods. Thus less stable actin structures are supplying actin-AC-latrunculin A complex to the highly stable rods.
A somewhat similar result was reported with latrunculin B addition to mast cells. It dramatically increased the amount of actin translocated to the nucleus. This translocation is cofilin-dependent, and, as with latrunculin A, resulted in the accumulation of concentrated actin regions (55). However, it is also possible that latrunculin A enhances rod fluorescence through an induced conformational change of the actin-bound XAC(wt)-GFP or an alteration in the GFP environment.
The fact that, over 60 min of observation, rods do not move (Fig. 4, f–h) is consistent with their being stable structures because often, if actin disassembly is followed by reassembly, the reassembled structure appears in a somewhat different location. The process of disassembly-reassembly is a major component of cell locomotion (66).
Another possible mechanism for the protective effects of rod formation reported here may derive from the inhibition of cofilin translocation to mitochondria observed in neurites with rods. The preservation of m and ATP is only short-term after rod formation in cultured cells with overexpressed fusion protein because the rods in this experimental model are not reversible. However, it is conceivable that, if their presence were transient in whole animals, as they can be in culture (48), they could have long-term beneficial effects. Microischemia, generating only transient rods in whole animals, could, by sequestering AC, reduce the chances of triggering cytochrome c release and apoptosis.
Several other factors must be considered as potentially contributing to the observed transient reduction in mitochondrial decline associated with rod formation. They include the direct effects of actin, gelsolin, and amyloid precursor protein on mitochondria. Translocation of the amyloid precursor protein to mitochondria is deleterious and depends on fast axonal transport of vesicles containing it (2). Rods probably inhibit this transport and may initially slow the damaging delivery to mitochondria. However, eventually, the impedance of axonal transport by rods spanning the neurite is likely to result in a build up of vesicles carrying amyloid precursor protein and its proteases, -secretase and presenilin (57, 61). Thus, with time, rod blockage of axonal transport may promote the accumulation of amyloid precursor protein and its proteolytic fragment A-amyloid1–42 peptide (29, 40, 41).
In addition, the rod sequestering of AC may preserve ATP levels and m by increasing the amount of gelsolin free to bind to mitochondria. Gelsolin is a well-characterized actin filament severing and plus-end capping protein with known AC interdependencies (70). Its binding to mitochondria inhibits the voltage-dependent anion channel of the mitochondrial permeability transition pore and protects against apoptosis (35). Some part of the transient beneficial effect of overexpressed AC rod formation may be through enhancement of gelsolin binding to mitochondria.
Transient neuronal stress, such as occurs with microischemia, is a common medical event with serious consequences, including increasing the likelihood of long-term neurodegeneration seen in cerebrovascular dementia and Alzheimer disease (34, 72). In this study, we have tried to determine whether rods, when occurring only transiently, might play a beneficial role in medical conditions like microischemia. Here, we have found that rod appearance has an initial beneficial energetic effect on neurites compared with neurites of the same cell in which rods do not appear. Is this beneficial effect on neurites sustained? Because rod generation through AC overexpression is not reversible and because overexpression is presently the only way to follow rod formation in live cells while monitoring mitochondrial potential and ATP level, it is unfortunately not possible to test for the long-term benefit. However, brief ischemia in whole animals does protect against the damage of longer ischemia-reperfusion, a phenomenon termed "ischemic preconditioning" (7, 54). Short-term rod appearance may be a contributing factor in this preconditioning. It would be interesting to know whether such preconditioning generates transient rods. The findings of this study do suggest that the brief appearance of actin-AC rods may have a pivotal protective long-term role because the extent of cell death in a neuronal population depends on the degree of m decline experienced (3). Beyond a critical point, mitochondrial recovery is no longer possible. Transient rod formation may reduce the number of mitochondria that cross the line and are irreversibly impaired.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Present address of H. Chen: Department of Biological Chemistry, Johns Hopkins University School of Medicine, Baltimore, MD 21205.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
2. Anandatheerthavarada HK, Biswas G, Robin MA, and Avadhani NG. Mitochondrial targeting and a novel transmembrane arrest of Alzheimer's amyloid precursor protein impairs mitochondrial function in neuronal cells. J Cell Biol 161: 41–54, 2003.
3. Ankarcrona M, Dypbukt JM, Bonfoco E, Zhivotovsky B, Orrenius S, Lipton SA, and Nicotera P. Glutamate-induced neuronal death: a succession of necrosis or apoptosis depending on mitochondrial function. Neuron 15: 961–973, 1995.[CrossRef][ISI][Medline]
4. Arber S, Barbayannis FA, Hanser H, Schneider C, Stanyon CA, Bernard O, and Caroni P. Regulation of actin dynamics through phosphorylation of cofilin by LIM-kinase. Nature 393: 805–809, 1998.[CrossRef][Medline]
5. Ashworth SL, Southgate EL, Sandoval RM, Meberg PJ, Bamburg JR, and Molitoris BA. ADF/cofilin mediates actin cytoskeletal alterations in LLC-PK cells during ATP depletion. Am J Physiol Renal Physiol 284: F852–F862, 2003.
6. Atkinson SJ, Hosford MA, and Molitoris BA. Mechanism of actin polymerization in cellular ATP depletion. J Biol Chem 279: 5194–5199, 2004.
7. Baines CP, Liu GS, Birincioglu M, Critz SD, Cohen MV, and Downey JM. Ischemic preconditioning depends on interaction between mitochondrial KATP channels and actin cytoskeleton. Am J Physiol Heart Circ Physiol 276: H1361–H1368, 1999.
8. Bamburg JR. Proteins of the ADF/cofilin family: essential regulators of actin dynamics. Annu Rev Cell Dev Biol 15: 185–230, 1999.[CrossRef][ISI][Medline]
9. Bamburg JR and Wiggan OP. ADF/cofilin and actin dynamics in disease. Trends Cell Biol 12: 598–605, 2002.[CrossRef][ISI][Medline]
10. Bernstein BW and Bamburg JR. Actin-ATP hydrolysis is a major energy drain for neurons. J Neurosci 23: 1–6, 2003.
11. Bernstein BW, DeWit M, and Bamburg JR. Actin disassembles reversibly during electrically induced recycling of synaptic vesicles in cultured neurons. Mol Brain Res 53: 236–250, 1998.[Medline]
12. Bershadsky AD, Gelfand VI, Svitkina TM, and Tint IS. Destruction of microfilament bundles in mouse embryo fibroblasts treated with inhibitors of energy metabolism. Exp Cell Res 127: 421–429, 1980.[CrossRef][ISI][Medline]
13. Campbell RE, Tour O, Palmer AE, Steinbach PA, Baird GS, Zacharias DA, and Tsien RY. A monomeric red fluorescent protein. Proc Natl Acad Sci USA 99: 7877–7882, 2002.
14. Cantiello HF. Changes in actin filament organization regulate Na+,K+-ATPase activity—role of actin phosphorylation. Ann NY Acad Sci 834: 559–561, 1997.
15. Carlier MF, Laurent V, Santolini J, Melki R, Didry D, Xia GX, Hong Y, Chua NH, and Pantaloni D. Actin depolymerizing factor (ADF/cofilin) enhances the rate of filament turnover: implication in actin-based motility. J Cell Biol 136: 1307–1322, 1997.
16. Chen H, Bernstein BW, Sneider JM, Boyle JA, Minamide LS, and Bamburg JR. In vitro activity differences between proteins of the ADF/cofilin family define two distinct subgroups. Biochemistry 43: 7127–7142, 2004.[CrossRef][Medline]
17. Chua BT, Volbracht C, Tan KO, Li R, Yu VC, and Li P. Mitochondrial translocation of cofilin is an early step in apoptosis induction. Nat Cell Biol 5: 1083–1089, 2003.[CrossRef][ISI][Medline]
18. Coue M, Brenner SL, Spector I, and Korn ED. Inhibition of actin polymerization by latrunculin A. FEBS Lett 213: 316–318, 1987.[Medline]
19. Didry D, Carlier MF, and Pantaloni D. Synergy between actin depolymerizing factor/cofilin and profilin in increasing actin filament turnover. J Biol Chem 273: 25602–25611, 1998.
20. Dixit R and Cyr R. Cell damage and reactive oxygen species production induced by fluorescence microscopy: effect on mitosis and guidelines for non-invasive fluorescence microscopy. Plant J 36: 280–290, 2003.[CrossRef][ISI][Medline]
21. Endo M, Ohashi K, Sasaki Y, Goshima Y, Niwa R, Uemura T, and Mizuno K. Control of growth cone motility and morphology by LIM kinase and slingshot via phosphorylation and dephosphorylation of cofilin. J Neurosci 23: 2527–2537, 2003.
22. Gohla A, Birkenfeld J, and Bokoch GM. Chronophin, a novel HAD-type serine protein phosphatase, regulates cofilin-dependent actin dynamics. Nat Cell Biol 7: 21–29, 2005.[CrossRef][ISI][Medline]
23. Gohla A and Bokoch GM. 14-3-3 Regulates actin dynamics by stabilizing phosphorylated cofilin. Curr Biol 12: 1704–1710, 2002.[CrossRef][ISI][Medline]
24. Grubbs RD, Beltz PA, and Koss KL. Practical considerations for using mag-fura-2 to measure cytosolic free magnesium. Magnesium Trace Elements 10: 142–1501992.
25. Gupta RK, Gupta P, Yushok WD, and Rose ZB. Measurement of the dissociation constant of MgATP at physiological nucleotide levels by a combination of 31P NMR and optical absorbance spectroscopy. Biochem Biophys Res Commun 117: 210–216, 1983.[CrossRef][ISI][Medline]
26. Haugland RP. Handbook of Fluorescent Probes and Research Chemicals. Eugene, OR: Molecular Probes, 1996.
27. Hayden SM, Miller PS, Brauweiler A, and Bamburg JR. Analysis of the interactions of actin depolymerizing factor with G- and F-actin. Biochemistry 32: 9994–10004, 1993.[CrossRef][Medline]
28. He TC, Zhou S, da Costa LT, Yu J, Kinzler KW, and Vogelstein B. A simplified system for generating recombinant adenoviruses. Proc Natl Acad Sci USA 95: 2509–2514, 1998.
29. Hiruma H, Katakura T, Takahashi S, Ichikawa T, and Kawakami T. Glutamate and amyloid -protein rapidly inhibit fast axonal transport in cultured rat hippocampal neurons by different mechanisms. J Neurosci 23: 8967–8977, 2003.
30. Hotulainen P, Paunola E, Vartiainen MK, and Lappalainen P. Actin-depolymerizing factor and cofilin-1 play overlapping roles in promoting rapid F-actin depolymerization in mammalian nonmuscle cells. Mol Biol Cell 16: 649–664, 2005.
31. Idrissi FZ, Wolf BL, and Geli MI. Cofilin, but not profilin, is required for myosin-I-induced actin polymerization and the endocytic uptake in yeast. Mol Biol Cell 13: 4074–4087, 2002.
32. Jang DH, Han JH, Lee SH, Lee YS, Park H, Lee SH, Kim H, and Kaang BK. Cofilin expression induces cofilin-actin rod formation and disrupts synaptic structure and function in Aplysia synapses. Proc Natl Acad Sci USA 102: 16072–16077, 2005.
33. Johnson BD and Byerly L. Ca2+ channel Ca(2+)-dependent inactivation in a mammalian central neuron involves the cytoskeleton. Pflügers Arch 429: 14–21, 1994.[ISI][Medline]
34. Kalaria RN. The role of cerebral ischemia in Alzheimer's disease. Neurobiol Aging 21: 321–330, 2000.[CrossRef][ISI][Medline]
35. Kusano H, Shimizu S, Koya RC, Fujita H, Kamada S, Kuzumaki N, and Tsujimoto Y. Human gelsolin prevents apoptosis by inhibiting apoptotic mitochondrial changes via closing VDAC. Oncogene 19: 4807–4814, 2000.[CrossRef][ISI][Medline]
36. Langford GM. Myosin-V, a versatile motor for short-range vesicle transport. Traffic 3: 859–865, 2002.[CrossRef][ISI][Medline]
37. Levraut J, Iwase H, Shao ZH, Vanden Hoek TL, and Schumacker PT. Cell death during ischemia: relationship to mitochondrial depolarization and ROS generation. Am J Physiol Heart Circ Physiol 284: H549–H558, 2003.
38. Leyssens A, Nowicky AV, Patterson L, Crompton M, and Duchen MR. The relationship between mitochondrial state, ATP hydrolysis, [Mg2+]i and [Ca2+]i studied in isolated rat cardiomyocytes. J Physiol 496: 111–128, 1996.[ISI][Medline]
39. Loew LM, Tuft RA, Carrington W, and Fay FS. Imaging in five dimensions: time-dependent membrane potentials in individual mitochondria. Biophys J 65: 2396–2407, 1993.
40. Lustbader JW, Cirilli M, Lin C, Xu HW, Takuma K, Wang N, Caspersen C, Chen X, Pollak S, Chaney M, Trinchese F, Liu S, Gunn-Moore F, Lue LF, Walker DG, Kuppusamy P, Zewier ZL, Arancio O, Stern D, Yan SS, and Wu H. ABAD directly links A to mitochondrial toxicity in Alzheimer's disease. Science 304: 448–452, 2004.
41. Maloney MT, Minamide LS, Kinley AW, Boyle JA, and Bamburg JR. -secretase-cleaved amyloid precursor protein accumulates at actin inclusions induced in neurons by stress or amyloid : a feedforward mechanism for Alzheimer's disease. J Neurosci 25: 11313–11321, 2005.
42. Mattson MP and Kroemer G. Mitochondria in cell death: novel targets for neuroprotection and cardioprotection. Trends Mol Med 9: 196–205, 2003.[CrossRef][ISI][Medline]
43. Maupin P and Pollard TD. Improved preservation and staining of HeLa cell actin filaments, clathrin-coated membranes, and other cytoplasmic structures by tannic acid-glutaraldehyde-saponin fixation. J Cell Biol 96: 51–62, 1983.
44. McIntosh JR. Electron microscopy of cells: a new beginning for a new century. J Cell Biol 153: 25F-32, 2001.
45. Meberg PJ and Bamburg JR. Increase in neurite outgrowth mediated by overexpression of actin depolymerizing factor. J Neurosci 20: 2459–2469, 2000.
46. Mills JW, Schwiebert EM, and Stanton BA. The cytoskeleton and membrane transport. Curr Opin Nephrol Hypertens 3: 529–534, 1994.[Medline]
47. Minamide LS, Shaw AE, Sarmiere PD, Wiggan O, Maloney MT, Bernstein BW, Sneider JM, Gonzalez JA, and Bamburg JR. Production and use of replication-deficient adenovirus for transgene expression in neurons. Methods Cell Biol 71: 387–416, 2003.[ISI][Medline]
48. Minamide LS, Striegl AM, Boyle JA, Meberg PJ, and Bamburg JR. Neurodegenerative stimuli induce persistent ADF/cofilin-actin rods that disrupt distal neurite function. Nat Cell Biol 2: 628–636, 2000.
) and increase in actin (or latrunculin A-actin; 
) with increasing latrunculin A. Data point (
) was included from an experiment starting with latrunculin A-actin complex and adding ADF to show that steady state was reached. The solid line for the ADF-actin curve is a fit to the binding equations using a ratio of the Kd values for ADF-actin to latrunculin A-actin of 0.095.
