Download HIV-1 Transactivator of Transcription Protein Induces Mitochondrial

This information is current as
of June 18, 2017.
HIV-1 Transactivator of Transcription
Protein Induces Mitochondrial
Hyperpolarization and Synaptic Stress
Leading to Apoptosis
Seth W. Perry, John P. Norman, Angela Litzburg, Dabao
Zhang, Stephen Dewhurst and Harris A. Gelbard
J Immunol 2005; 174:4333-4344; ;
doi: 10.4049/jimmunol.174.7.4333
http://www.jimmunol.org/content/174/7/4333
Subscription
Permissions
Email Alerts
This article cites 59 articles, 18 of which you can access for free at:
http://www.jimmunol.org/content/174/7/4333.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2005 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
References
The Journal of Immunology
HIV-1 Transactivator of Transcription Protein Induces
Mitochondrial Hyperpolarization and Synaptic Stress Leading
to Apoptosis1
Seth W. Perry,2*†‡ John P. Norman,# Angela Litzburg,*‡ Dabao Zhang,§ Stephen Dewhurst,储
and Harris A. Gelbard2*‡¶储
T
ransactivator of transcription protein (Tat)3 is a small,
nonstructural transcriptional regulator essential for the
replication of the HIV type 1 (HIV-1). Tat has been detected in the brain of patients with HIV-1-associated dementia
(HAD) by immunoblot analysis (1), and is unique among nonEnvelope proteins in that it is actively secreted by infected glial
*Center for Aging and Developmental Biology, †Interdepartmental Program in Neuroscience, ‡Department of Neurology, §Department of Biostatistics and Computational Biology, ¶Department of Pediatrics, 储Department of Microbiology and Immunology, and #Department of Environmental Medicine, University of Rochester
Medical Center, Rochester, NY 14642
Received for publication October 7, 2004. Accepted for publication January 4, 2005.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by National Institutes of Mental Health P01MH64570 (to
H.A.G. and S.D.); National Institutes of Mental Health MH56838 (to H.A.G.); National Institute on Environmental Health Sciences Training Grant ES07026 (to
J.P.N.); and National Institutes of Allergy and Immunology T32 AI49815 (to S.W.P.
and S.D.).
2
Address correspondence and reprint requests to Dr. Seth W. Perry or Dr. Harris A.
Gelbard, Center for Aging and Developmental Biology, University of Rochester Medical Center, Rochester, NY 14642. E-mail addresses: [email protected]
or [email protected]
Abbreviations used in this paper: Tat, transactivator of transcription protein; ⌬␺m, mitochondrial membrane potential; ⌬␺p, plasma membrane potential; AP, action potential;
CCD, charge-coupled device; CM-H2DCFDA, 5 (and 6)-chloromethyl-2⬘,7⬘-dichlorodihydrofluorescein diacetate, acetyl ester; DiBAC4(3), bis-(1,3-dibutylbarbituric acid) trimethine oxonol; DIV, day in vitro; FCCP, carbonyl cyanide 4-(trifluoromethyoxy)phenylhydrazone; FM1-43, N-(3-triethylammoniumpropyl)-4-(4-(dibutylamino)styryl)
pyridinium dibromide; HAART, highly active antiretroviral therapy; HAD, HIV-1associated dementia; KATP, ATP-sensitive potassium; MCS, mean cell signal; NAC,
N-acetylcysteine; NwNR, with neurite retraction; NwoNR, without neurite retraction;
PAF, platelet-activating factor; RFU, relative fluorescent unit; ROS, reactive oxygen
species; TMRE, tetramethylrhodamine ethyl ester; TMRM, tetramethylrhodamine methyl
ester; TUDCA, tauroursodeoxycholic acid; TUS, total uptake signal.
3
Copyright © 2005 by The American Association of Immunologists, Inc.
cells (2), thus increasing the probability that it will be found at
higher concentrations in the extracellular space in brain parenchyma relative to other HIV gene products.
Tat is often described as pleiotropic because of its diverse effects
in the periphery and the CNS. In keeping with this, Tat exerts
transcriptional control over numerous cellular genes in a variety of
cell types (1, 3), and it increases TNF-␣ expression in neuronal
cells at the transcriptional level (3); this in turn leads to an increase
in platelet-activating factor (PAF) production by mononuclear
phagocytes (4). Furthermore, Tat has been shown to have a number of direct effects on neurons, including excessive depolarization
and calcium influx (5) as well as apoptosis (5, 6).
Oxidative stress has been implicated as a potential cause of neuronal dysfunction in a number of neurological diseases, and may
represent a final common mediator of both synaptic and somal
stress in neurons afflicted during HAD and other neurodegenerative disorders (7). Tat has been reported to induce oxidative stress
in the periphery and the CNS by a variety of mechanisms, including down-regulation of manganese-dependent superoxide dismutase and increased protein oxidation (8, 9). A particularly intriguing aspect of oxidative stress is the activation of proapoptotic
pathways in the synapse that may be involved in plasticity and
remodeling (10, 11). This has led to the concept of synaptic apoptosis, or deconstruction of established synapses. Because HIV-1associated neurologic disease does not appear to correlate with
neuronal apoptosis (12), but rather diminution of the dendritic arbor, this is a compelling phenomenon to study in the context of
HIV-1-associated neurologic disease. Furthermore, it is particularly germane to investigate how Tat and other proinflammatory
cellular metabolites may disrupt cellular energy metabolism under
0022-1767/05/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Despite the efficacy of highly active antiretroviral therapy in reducing viral burden, neurologic disease associated with HIV-1
infection of the CNS has not decreased in prevalence. HIV-1 does not induce disease by direct infection of neurons, although
extensive data suggest that intra-CNS viral burden correlates with both the severity of virally induced neurologic disease, and with
the generation of neurotoxic metabolites. Many of these molecules are capable of inducing neuronal apoptosis in vitro, but
neuronal apoptosis in vivo does not correlate with CNS dysfunction, thus prompting us to investigate cellular and synaptic events
occurring before cell death that may contribute to HIV-1-associated neurologic disease. We now report that the HIV-1 regulatory
protein transactivator of transcription protein (Tat) increased oxidative stress, ATP levels, and mitochondrial membrane potential
in primary rodent cortical neurons. Additionally, a proinflammatory cellular metabolite up-regulated by Tat, platelet-activating
factor, also induced oxidative stress and mitochondrial hyperpolarization in neurons, suggesting that this type of metabolic
dysfunction may occur on a chronic basis during HIV-1 infection of the CNS. Tat-induced mitochondrial hyperpolarization could
be blocked with a low dose of the protonophore FCCP, or the mitochondrial KATP channel antagonist, tolbutamide. Importantly,
blocking the mitochondrial hyperpolarization attenuated Tat-induced neuronal apoptosis, suggesting that increased mitochondrial
membrane potential may be a causal event in precipitating neuronal apoptosis in cell culture. Finally, Tat and platelet-activating
factor also increased neuronal vesicular release, which may be related to increased mitochondrial bioenergetics and serve as a
biomarker for early damage to neurons. The Journal of Immunology, 2005, 174: 4333– 4344.
4334
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
conditions of synaptic stress. We now report that Tat and the proinflammatory metabolite it up-regulates, PAF, are able to paradoxically increase mitochondrial membrane potential (⌬␺m), ATP levels, and vesicle recycling before neuronal demise.
Materials and Methods
Reagents
Carbamyl-PAF (c-PAF or simply PAF in this work), a nonhydrolyzable
form of PAF, was obtained from BIOMOL. rHIV Tat1–72 protein and the
biologically inactive mutant Tat ⌬31– 61 were received as generous gifts
from the laboratory of A. Nath (Johns Hopkins University, Baltimore,
MD), prepared as previously described (13).
Cell culture
Assessment of neurite damage
冋冘
No. time-matched field pairs
After experimental treatment of cortical cultures in 24-well plates, for each
condition, neurons in 30 random ⫻30 fields from three replicate wells/
condition (10 random fields/well) were scored as either with (NwNR) or
without (NwoNR) neurite retraction by a trained observer blind to the
experimental conditions. Neurons classified as with neurite retraction exhibited either truncated neurites or complete absence of neurites (see arrows, Fig. 1, B and C, for examples). All cells included in the analysis were
still viable, as cells positive for cell death by TUNEL or trypan blue dye
exclusion assays were excluded from analysis (37). Percentage of NwNR
per field values was determined by the equation: % NwNR/field ⫽ (number
of NwNR/(NwNR ⫹ NwoNR)) ⫻ 100. These values were averaged for all
fields per condition, and for each condition (⫾Tat treatment), data were
expressed as mean percentage of neurons with neurite retraction per field ⫾
SEM (% NwNR/field ⫾ SEM). Comparisons were made by unpaired t tests
with significance level of p ⱕ 0.01.
Monitoring alterations in ⌬␺m
Changes in ⌬␺m in response to Tat treatment were assessed by the lipophilic cationic dyes tetramethylrhodamine ethyl and methyl ester
(TMRE and TMRM, respectively). TMRE and TMRM are positively
charged dyes that equilibrate across membranes in a Nernstian fashion, and
therefore, will accumulate across the mitochondrial membrane and into the
matrix space in inverse proportion to the ⌬␺m (16). Due to the quenching
properties of aggregated membrane-bound probe at higher dye concentrations, sensitivity to fluctuations in ⌬␺m is maximized by using the lowest
possible (i.e., subquenching) dye concentrations, because this ensures that
the dye signal remains directly proportional to dye concentration and therefore ⌬␺m. Moreover, because TMRE and similar dyes must equilibrate
across the plasma membrane before entering the mitochondria, in whole
cell applications dye concentration in the mitochondria will be dependent
on both plasma membrane potential (⌬␺p) and ⌬␺m. At sufficiently low dye
concentrations (i.e., high cell/dye ratio) and normal membrane potentials,
ⱖ95% of the dye will be in the mitochondria, and ⱕ1% in the medium,
which results in a ⬃100⫻ sensitivity of the dye to ⌬␺m over ⌬␺p, given
equivalent changes in plasma or ⌬␺m (17). In other words, a 33% drop in
⌬␺p would result in a ⬍1% change in cell fluorescence, but an equivalent
change in ⌬␺m would result in a ⬎90% change in cell fluorescence (17).
共共共fMCStxT /fMCSctlT兲 ⫻ 100兲
T⫽1
册
⫺ 100)T /no. time matched field pairs ⫾ SEM
Significance between conditions was determined by Student’s t test at a
level of p ⱕ 0.01.
For some experiments, mitochondrial TMRE or TMRM signal was assessed by fluorometry, as described below; both methods produced equivalent fluorescence values for mitochondrial TMRE/M uptake (see Results
and Fig. 3, G and H).
By taking average pixel intensity over a neuronal cell body, MCS values
could be impacted by changes in cell size or morphology. To ensure that
fluorescence signal changes were not caused by changes in cell size or
morphology, total TMRE uptake in individual culture wells was determined by loading with 1 nM TMRE for 20 min, followed by removal of the
dye solution, readdition of fresh dye-free HBSS⫹, and assessment of the
total accumulated dye signal minus background fluorescence (total uptake
signal (TUS)) from the cells using a Bio-Rad Fluoromark plate reader
(excitation, 544; emission, 605). For these experiments, TUS for each condition ⫽
冋冘
No. Replicates
(well RFU valueR
R⫽1
册
⫺ mean background RFU)R /no. replicates ⫾ SEM
TUS values were normalized to cell number and expressed as percentage
of control ⫾ SEM, and significance between conditions was determined by
Student’s t test at a level of p ⱕ 0.01. Compared with digital image analysis
methods, by design fluorometers can read a TUS from a monolayer culture,
solution, or both, without significant specimen depth or focal plane limitations (i.e., the fluorescence signal collected is more three dimensional and
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Primary neuronal cell cultures. Primary neuronal cortical cultures were
prepared from embryonic day 18 rats by modification of the protocol by
Brewer et al. (14). In brief, cortices were dissected from a litter of E18
embryonic rats, dissected free of meninges and other tissue, and incubated
in 2.0 ml of Ca⫹/Mg⫹-free HBSS (with 10 mM HEPES, pH 7.3) with PSN
antibiotics (50 mg/L penicillin, 50 mg/L streptomycin, 100 mg/L neomycin) plus 0.5 ml of 2.5% trypsin (for 0.25% final) for 15 min at 37°C per
brain. After the 15-min incubation, trypsin was removed, cells were
washed twice with HBSS (with Ca⫹/Mg⫹), then dissociated in growth
medium (below) by Pasteur pipette trituration by 8–10 passages through a
0.9-mm bore 1000-␮l blue pipette tip. Dissociated cells were counted by
trypan blue viability assay and plated in cell culture plates at 0.5–0.6 ⫻ 105
cells/cm2, on poly(D-lysine)-coated cell culture plastic. The plating and
maintenance medium used consisted of Neurobasal with B27 supplement
(Invitrogen Life Technologies), as described by Brewer et al. (14), and as
modified for antioxidant-free culture (14, 15). This medium formulation
inhibits the outgrowth of glia, resulting in a neuronal population that is
98% pure (14); thus, glial inhibitors are unnecessary. Cells were cultured
for 10–21 days at 37°C in a humidified atmosphere of 5% CO2/95% air,
changing medium every 4 days. Cells were used for experiments at days in
vitro (DIV) 14–21, unless otherwise indicated.
Thus, we used 1 nM TMRE or TMRM for these experiments, a dose that
we confirmed was exponentially more sensitive to ⌬␺m over ⌬␺p, and did
not exhibit quenching effects (17–19). At this concentration, dye signal
emanated almost exclusively from mitochondria (see Results). TMRM has
been reported to exhibit lower mitochondrial binding and toxicity (16, 20);
thus, we used this probe exclusively in later experiments. However, our
results did not vary whether TMRE or TMRM was used.
For these studies, primary rat cortical neurons were treated with reagent
under normal culture conditions for the indicated time period, followed by
removal of the medium, 1 ⫻ 2-min wash in prewarmed 37°C HBSS plus
10 mM glucose and 10 mM HEPES (HBSS⫹), then incubation in HBSS⫹
with 1 nM TMRE/M. Following equilibration of the dye for 20 min to
ensure distribution across the mitochondrial membrane, the cells were imaged while remaining in the 1 nM TMRE/M solution, as is necessary for
a continued dye equilibrium state. Random field images were taken using
an Olympus IX-70 microscope and ⫻40 objective (fluorescent excitation,
545; emission, 610) and Sony DCX-9000 color charge-coupled device
(CCD) camera, and the mean relative fluorescent unit (RFU) value (red
channel only) of the neuronal soma and processes were quantified using
Scanalytics IPLab software. This total neuronal (i.e., cytoplasmic, including soma and neurites) area was determined by thresholding to exclude
cell- or mitochondria-deficient regions of the field. A mean neuronal fluorescence value (mean cell signal (MCS)) was acquired for this total mitochondria-containing neuronal area per field, equal to the sum of pixel
values over the total included neuronal area, divided by the total number of
pixels; the equation for this is as follows: for each field, the average pixel
intensity value over the total cytoplasmic area, or field MCS (fMCS) ⫽
sum of all cytoplasmic pixel values/total number of cytoplasmic pixels. By
equally weighting all pixels over the total mitochondrial-containing neuronal area per field, this method is preferable to the more common method
of taking the mean of the mean pixel value per cell (which is undesirable
because it gives more weight to pixel values that occur more frequently).
To control for variations in signal intensity over time during the course of
image acquisition, the integrated MCS from each treatment field (fMCStx)
was expressed as percentage of difference from the MCS of a time-matched
control field (fMCSctl), and percentage of differences for each pair was
averaged over the total number of paired fields (at least 15 fields per condition) and expressed as mean percentage of difference from control ⫾
SEM. The equation for this is: mean %⌬ in TMRM/E signal vs Ctl ⫽
The Journal of Immunology
less subject to blockade effects or signal interference than can be achieved
by the two-dimensional image limitations of microscope optics and CCD
camera). As such, readings taken with and without cell lysis (in HBSS⫹
with 2% SDS) were comparable.
Mitotracker green (Molecular Probes), a largely potential independent
mitochondrial marker (21), was also quantitated to ensure that changes in
TMRE fluorescence were not caused by changes in total mitochondrial
mass or volume. After relevant experimental treatment, Mitotracker green
(100 nM) was used in exactly the same fashion as TMRE and TMRM
above. To ensure a valid comparison against the TMRE probe, the fluorescent Mitotracker green signal was also quantified by both MCS and TUS
(with cell lysis) methods above, producing similar results regardless of the
quantification method.
The bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC4(3))
control for monitoring ⌬␺p was performed and analyzed in exactly the
same way as for the TMRE and TMRM experiments, except that 1 ␮M
DiBAC4(3) was used in place of TMRE/M. Contrary to TMRE and
TMRM, DiBAC4(3) is an anionic dye, thus showing an inverse relationship
with membrane potential.
Measuring production of intracellular reactive oxygen species
(ROS)
Measurement of adenosine levels and ATP/ADP ratio
Adenosine di- and triphosphate (ADP and ATP, respectively) levels in
cortical cultures were measured after experimental treatment using a kit
from Cambrex. Briefly, cortical neurons were plated in white poly(D-lysine)-coated 96-well plates at a density of 32,000 cells/well. Each treatment
group contained five replicate wells. After treatment, ATP and ADP levels
were then measured using an ATP/ADP assay kit (Cambrex), as per the
manufacturer’s directions. This kit assays ATP levels via the luciferase
reaction, then assays ADP levels by a proprietary method of ADP to ATP
conversion; the difference in luminescence signal pre- and post-ADP conversion represents the ADP levels in the culture. Luminescent signal was
integrated over 1 s and read on a Packard Instrument LumiCount microplate luminometer. For each replicate well, data were expressed as both
ATP level/ADP level (i.e., the ATP/ADP ratio, the chief indicator of cellular energy status) (see Fig. 6A), as well as individual ADP and ATP
levels (see Fig. 6B). Thus, for each condition, data represent the mean
ATP/ADP ratio ⫾ SE of 5–10 replicates per condition, expressed relative
to 1-h control (1-h control ⫽ 1.0). Statistical comparisons of ATP/ADP
ratios for all main effects and interaction effects of time (1, 24, and 48 h)
and dose (0, 0.1, and 2.5 ␮g/ml Tat) were made by SAS statistical software,
using the PROC GLM procedure for unbalanced ANOVA analysis, with
significance determined at p ⬍ 0.01. Actual ATP and ADP levels were
analyzed in the same fashion, except presented as raw (RFU) values, and
significance was analyzed by Student’s t test at p ⬍ 0.01. Background
luminescence was negligible, but nonetheless was subtracted from all readings before data analysis. Experimental conditions were run two or more
times with similar results.
FM1-43 uptake assay
Total spontaneous activity-dependent vesicular uptake was assessed using
the lipophilic styryl dye N-(3-triethylammoniumpropyl)-4-(4-(dibutylamino)styryl) pyridinium dibromide (FM1-43). FM1-43 is an amphipathic molecule with a ⫹2 charge that prevents it from passively crossing
membranes (23). It is nonfluorescent in aqueous solution, but becomes
fluorescent upon reversibly binding to exposed membrane by partitioning
into the lipid bilayer. Therefore, extracellular membranes bind FM1-43,
which then becomes encapsulated into a vesicle during endocytosis, gen-
erating a fluorescent signal from the endocytosed vesicle. Upon subsequent
exocytosis, the vesicle fuses with the membrane, releasing its contents, and
the FM1-43 dissociates back into the extracellular solution and loses its
fluorescence.
These features of this dye have made it a useful tool for studying endocytosis and exocytosis, notably synaptic vesicle recycling and size estimation of various synaptic vesicle pools. Because we wished to study the
effects of Tat and PAF on spontaneous activity-dependent FM1-43 uptake
of cortical neuronal cultures, we loaded FM1-43 in the absence of a depolarizing stimulus to neurons. This method loads FM1-43 into vesicles
undergoing both spontaneous miniature synaptic release and spontaneous
action potential (AP)-dependent release. Spontaneous miniature synaptic
activity consists of AP-independent presynaptic release of one or more
neurotransmitter quanta; exists throughout the nervous system; is considered an integral participant in the maintenance, modulation, and plasticity
of synaptic connections; and can be identified with FM1-43 (24, 25). Spontaneous AP activity also results in uptake of FM1-43, and has been shown
to occur in vitro under control conditions. Under control conditions, our
neuronal cultures fire spontaneous APs at a rate of at least 1.67 ⫻ 10⫺2 Hz
per cell, which translates to a minimal rate of 2.1 ⫻ 103 Hz for the culture
as a whole, although the rate may be up to 2.3 Hz for some neuronal
subtypes (26). Moreover, Tat has been shown to induce rapid, repetitive,
and continuous chains of APs in neuronal cultures (5), and PAF also increases neuronal transmission (24, 27).
For this assay, FM1-43 was loaded into cells in the absence of any
additional depolarizing stimulus for 15 min, a length of time that has been
determined to label spontaneous activity without saturating the vesicle pool
(25). Following washout of excess dye, release of the spontaneously endocytosed FM1-43 in a high KCl-depolarizing bath confirmed the activitydependent nature of the staining (28), and ensured that all FM1-43 vesicular uptake was released. As would be expected for cultures undergoing
spontaneous AP-dependent and AP-independent synaptic activity, some
nonlinear (i.e., vesicular or quantal) spontaneous FM1-43 release could be
achieved in the absence of KCl depolarization over the same monitored
release period. However, FM1-43 release in the presence of KCl (⫹KCl)
was much more rapid and many magnitudes greater than FM1-43 release
in the absence of KCl (⫺KCl). In one representative comparison, release
rate (t1/2) was 1.4 times faster ( p ⬍ 0.001), and relative release magnitude
3.9 times greater ( p ⬍ 0.001), in the ⫹KCl vs the ⫺KCl condition, respectively, further emphasizing the activity-dependent nature of the
FM1-43 uptake (data not shown).
For these assays, primary rat cortical neurons were plated in 24-well
plates and cultured, as described above, then treated for 24 h with appropriate experimental agents, as indicated in Results. For the last 15 min of
treatment, FM1-43 was added directly to the medium to a final concentration of 10 ␮M, and incubated for 15 min at 37°C. Following loading,
excess FM1-43 was washed out in HBSS⫹ for 15 min at room temperature.
Wash solution was prechilled to 4°C to limit exocytosis of dye from spontaneous activity. Following washout, cells were incubated at 37°C in a
depolarizing solution of HBSS⫹ with 100 mM KCl added. Immediately
after addition of the depolarizing solution, cells were transferred to a BioRad fluoromark plate reader, and release of FM1-43 was monitored at 20-s
intervals using 438 excitation and 605 emission filters and Bio-Rad microplate reader III software. Because some endocytosed vesicles are recycled into the endosomal pathway and thus take longer to release (29), and
some FM1-43 may remain trapped in the membrane of rapidly recycling
vesicles due to the dissociation kinetics of the dye (30), we monitored
release over 1 h to ensure that all vesicular FM1-43 signal was accounted
for. However, the majority of activity-dependent FM1-43 signal was lost
well before 15 min, as would be expected for exocytotic processes occurring at the synapse. Because background nonvesicular staining by FM1-43
is not released by KCl (28), then for each condition, the total loss of
FM1-43 signal during the release period equals the total spontaneous activity-dependent vesicular uptake of FM1-43 in each well over the loading
period. This absolute FM1-43 uptake value reflects the total activity of the
culture, and is equivalent to the synaptic or vesicular release probability of
the culture (25).
These values from at least three replicates per condition, expressed in
RFU, were averaged and expressed as mean FM1-43 uptake ⫾ SEM.
(Some data were expressed as percentage of control to allow direct comparisons across different experimental runs and conditions.) To ensure that
changes in total FM1-43 uptake could not be accounted for by differing
well densities, neurons (which are postmitotic and do not divide after plating) were plated at equal densities, and preassay visual screening for uniform cell number and density was performed. Within the central reading
area of the plate reader, wells with greater than 10% variation in cell number (as assessed by cell counts) or total cell area (as estimated visually), or
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Generation of ROS in response to Tat and PAF was assessed with the
oxidizable dye indicator 5 (and 6)-chloromethyl-2⬘,7⬘-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA, abbreviated as
DCF) (Molecular Probes). CM-H2DCFDA is an improved analog of dichlorodihydrofluorescein diacetate that enables better retention in live cells
by virtue of the chloromethyl group. Upon entering the cell by passive
diffusion, the acetate groups are cleaved by intracellular esterases, and the
thiol-reactive chloromethyl group binds to intracellular thiols, rendering
the CM-H2DCF product trapped within the cell. Oxidation by ROS, including hydrogen peroxide (H2O2), hydroxyl radical (HO⫺), peroxyl radical (HOO⫺), or peroxynitrite anion (ONOO⫺) (22) (manufacturer’s data),
leaves the final fluorescent product dichlorofluorescein diacetate, which is
then imaged (excitation, 485; emission, 538) and quantitated in the same
fashion as TMRE (below) as an indicator of relative levels of ROS in the
culture. Comparisons were made by unpaired t tests with significance level
of p ⱕ 0.01.
4335
4336
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
FIGURE 1. Tat induces neurite retraction in rat cortical neurons. A, Treatment with 3.5 ␮g/ml Tat for 24 h resulted in a 3-fold (300%) increase in the
percentage of neurons exhibiting retracted neuritic processes (ⴱ, p ⬍ 0.0001; unpaired t test). A neuron was judged to have retracted neurites based upon
criteria described in Materials and Methods, and was only counted as having retracted processes if a clear judgment could be made. Fifteen ⫻30 fields
(average field ⬇75 neurons ⫻ 15⬇1125 neurons total) per condition were scored by a blinded observer trained to recognize neurite retraction. Data were
expressed as mean number of neurons with retracted processes ⫾ SEM. Data are from one representative experiment replicated at least three times. B,
Representative vehicle control subfield showing neurons with an extensive neurite network, including numerous long, largely intact processes forming a
dense neurite mat (see boxed area). C, Representative 3.5 ␮g/ml Tat-treated subfield showing several neurons with truncated or retracted neurites (filled
arrows), and only a sparse neurite network (compare boxed areas in C and B, respectively).
from analysis). Nonetheless, this finding raises two intriguing
questions: 1) does cell death begin at the synapse, and 2) do compensatory metabolic changes occur because of loss of synaptic
connectivity?
Quantifying cell death
Tat increases ⌬␺m in a biphasic fashion
Cell death was assessed by visualizing fragmented DNA per the TUNEL
method, as described previously (15). Briefly, after experimental treatment
in 24-well plates, cells were fixed with Histochoice MB Tissue Fixative
(AMRESCO), then TUNEL labeled with the ApopTag kit (Chemicon International), according to kit instructions. Cells were visualized under
Hoffman modulation contrast optics using a ⫻40 objective, and images
were taken of 10 random fields per well from 3 replicate wells per condition. Data were expressed as percentage of Apoptag-positive neurons
(((number of Apoptag-positive neurons)/(total neurons)) ⫻ 100) per field,
and then field values were averaged for a mean cell death value per well.
Mean values from 3 wells were averaged for a final mean cell death value
for each condition ⫾ SEM; the analyses were made without knowledge of
the treatment group. Significance of differences between conditions was
determined by Student’s t test at p ⱕ 0.01.
Mitochondrial function is essential for normal synaptic activity
(33–35). Because ⌬␺m is a sensitive indicator of mitochondrial
function, and because collapse (i.e., depolarization) of the resting
⌬␺m is commonly associated with apoptosis (36), we used the
ethyl and methyl esters of the cationic dye tetramethylrhodamine
(TMRE and TMRM, respectively) to quantitatively define changes
in ⌬␺m in response to Tat1–72. At low nM concentrations, TMRE
and TMRM localize almost exclusively to the mitochondrial membrane, and are exponentially more sensitive to ⌬␺m over ⌬␺p.
Thus, increased TMRE/M uptake reflects increased ⌬␺m (i.e., hyperpolarization) and decreased dye uptake represents depolarization of ⌬␺m.
Using this method, our data indicated that treatment of cortical
neurons with Tat caused a biphasic increase in ⌬␺m (i.e., hyperpolarization) that was both concentration and time dependent (Fig.
2). A low dose of Tat (100 ng/ml) caused a gradual increase in
Statistical considerations
Statistical analyses were made by either unpaired Student’s t tests, or unbalanced ANOVA analysis using the PROC GLM procedure in the SAS
statistical software package, as described under each Materials and Methods subsection. Differences between results were determined to be statistically significant if p ⬍ 0.01.
Results
Tat induces loss of neurites
We used primary rodent monolayer cortical cultures to investigate
early effects of Tat on the integrity of synapses. In Fig. 1, we
present data to demonstrate that 24-h application of 3.5 ␮g/ml Tat
to cortical neuronal cultures that have already established synapses
can alter synaptic connectivity, such that there is a 300% increase
(23.1% Tat vs 7.7% control, p ⬍ 0.0001) in the percentage of
neurons exhibiting neurite retraction (see Materials and Methods
for description of analysis). The validity of the assessment criteria
used to determine neurite retraction is further supported by noting
the significant simplification of the neural network in Tat-treated
vs control cultures (compare boxed regions in Fig. 1, C (Tat) and
B (control), respectively). Although Tat is by no means the sole
arbiter of HIV-1-associated neurotoxicity, this observation is in
agreement with previously published neuropathologic data demonstrating simplification of the dendritic arbor in patients with
HAD (12, 31). It is also obvious that, at this dose of Tat, neurite
retraction may simply be part of the morphologic progression of
Tat-induced apoptosis (32), with the presumption that cells undergoing neurite retraction would undergo cell death at a later time
(because cells concurrently positive for cell death were excluded
FIGURE 2. Tat causes a dose-dependent biphasic mitochondrial hyperpolarization in cortical neurons. Treatment of rat cortical neurons with 100
ng/ml or 2.5 ␮g/ml Tat for 1, 4, 10, 24, 26, 36, or 48 h resulted in a
dose-dependent biphasic increase in ⌬␺m over the time course. As described in Results, initial peaks for both doses were followed by periods of
apparent ⌬␺m stabilization, followed by increased ⌬␺m again at later time
points. For both doses, the later increase in ⌬␺m persisted until the end of
the analysis (48 h). A biologically inactive mutated Tat peptide (green)
produced no effect. Curve fits are nonformulaic software interpolations of
the data points. TMRE assay and analysis were performed, as described in
Materials and Methods and Results.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
wells with considerable cell clumping were not used for the assay. Averaging multiple replicates (ⱖ3 for each condition) also helped to eliminate
this potential source of error. Figures reflect representative data from multiple experiments. Significance was determined by Student’s t test at
p ⱕ 0.01.
The Journal of Immunology
faint uptake of TMRE in rodent cortical neurons under control
conditions, while B shows a marked increase in TMRE uptake in
response to 1 h of 2.5 ␮g/ml Tat treatment. In C, the uptake of
TMRE in cortical neurons exposed for 1 h to an equimolar concentration of the biologically inactive Tat mutant (⌬31– 61) is indistinguishable from A, establishing the specificity of Tat’s effect
on TMRE uptake. D–F show ⫻40, ⫻60, and ⫻90 (respectively)
magnifications of sample cultures to illustrate the subcellular distribution of TMRE in rodent cortical cultures, with pronounced
uptake of TMRE into mitochondrial membranes in the perinuclear
and process-bearing areas. Because of the very high density of
mitochondria typically present in perinuclear regions of neurons,
some portions of these images appear to have indiscriminate
TMRE labeling within the perinuclear cytoplasm; however, the
lack of cytoplasmic staining and distinctly punctate labeling of
TMRE in cell body and neurite regions further from the nucleus
emphasizes the specificity of dye uptake and its mitochondrial localization. Distribution, localization, and morphology of mitochondria and cells were not significantly affected by Tat treatment
(and see additional controls below).
FIGURE 3. Increased TMRE signal represents a verifiable increase in ⌬␺m. After 1-h treatment, mitochondrial TMRE uptake was visibly lower in
vehicle control cultures (A) as compared with 2.5 ␮g/ml Tat-treated cultures (B). Cultures treated with an equimolar amount of biologically inactive mutated
Tat peptide, ⌬31– 61 (C), were no different from control. Furthermore, ⫻40 field of cortical neurons equilibrated with 1 nM TMRE (D) demonstrates that
the TMRE signal is localized primarily to the mitochondria, as evidenced by the punctate signal in the cytoplasm and neuronal processes, but excluded from
nuclear regions, and the relative absence of staining in the intermitochondrial cytoplasm. E, ⫻60 close-up of cultures in D further demonstrates mitochondrial localization of TRME signal. Triangular arrows indicate mitochondrial TMRE labeling along clusters of neurites; square arrow indicates dense
mitochondrial TMRE labeling of a cluster of neuronal cell bodies. F, ⫻90 high magnification field indicates punctate mitochondrial labeling (excluding
the mitochondria-deficient nucleus) in a single neuronal cell body (square arrow) and along its neurite (triangular arrow). G, Comparing total cellular TMRE
uptake with mitochondrial TMRE signal in control- vs Tat-treated cultures indicated that the increased signal seen with Tat treatment could not be accounted
for by changes in cell morphology, nor changes in total mitochondrial mass or volume, as evidenced by Mitotracker green labeling analysis (H). I, Nor was
the Tat-induced increase in mitochondrial TMRE staining an artifact of Tat-induced increases in ⌬␺p, because Tat actually caused a slight depolarization
of ⌬␺m. Images were taken on an Olympus IX-70 microscope with ⫻40 and ⫻60 objectives (with ⫻1.5 inline slider lens) and 545 excitation and 610
emission filters, with a Sony DCX-9000 CCD camera in grayscale mode. Statistics were done by t test at p ⬍ 0.01.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
⌬␺m, peaking with a 17% increase in mitochondrial TMRE uptake
at 4 h ( p ⬍ 0.004), then declining to baseline by 14 h, before rising
again (6% increase vs control vehicle) 26 h after application, followed by a plateau at 36 h that persists until the end of the analysis
(48 h) (19% increase vs control vehicle) (Fig. 2, blue squares). In
contrast, a higher dose of Tat (2.5 ␮g/ml) caused a sharp increase
in ⌬␺m vs control, peaking at a 39% increase over control by 1 h
( p ⬍ 0.0003) before declining to control level by 10 h, then increasing again to a second peak of 27% vs control at 36 h, followed
by a plateau of 23% vs control at 44 h that also persisted until the
end of the analysis (48 h) (Fig. 2, red circles). Incubation of cortical cultures with the biologically inactive Tat mutant (⌬31– 61) at
either 1 or 26 h resulted in TMRE uptake values that were indistinguishable from control vehicle (green diamonds), demonstrating
specificity of Tat’s effect on mitochondrial hyperpolarization. The
mitochondria-depolarizing protonophore trifluoromethoxycarbonylcyanide phenylhydrazone FCCP (5 ␮M), in contrast, substantially diminished mitochondrial TMRE/M signal (see Fig. 5C).
Fig. 3 provides visual representations of the increased TMRE
signal seen with Tat treatment, and shows that the TMRE signal
emanates primarily from mitochondria. Fig. 3A shows relatively
4337
4338
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
Changes in cell morphology or mitochondrial mass do not
account for increased TMRE signal
Increased mitochondrial TMRE is not due to increased ⌬␺p
By the Nernst equation, TMRE/M concentration in the mitochondria is dependent upon both ⌬␺p and ⌬␺m. By increasing dye
concentration in the cytoplasm, an increased ⌬␺p will increase
TMRE/M concentration in the mitochondria independent of any
changes in ⌬␺m. As mentioned previously, this effect is greater at
higher TMRE/M concentrations. Nonetheless, to further verify that
the observed increased TMRE uptake following Tat treatment reflects hyperpolarization of the mitochondrial, but not the ⌬␺p, we
performed additional experiments using the voltage-sensitive anionic fluorescent dye DiBAC4(3) after a 1-h treatment with 2.5
Tat and PAF increase ROS in cortical neuronal cultures
independently of ⌬␺m
We then wondered whether mitochondrial membrane hyperpolarization represented a compensatory response to potential increases
in reactive oxygen species (ROS) that may occur after exposure to
Tat. Because Tat can exert some of its neurotoxic effects through
increased production of the phospholipid mediator PAF (4), we
also measured ROS production in neuronal cultures exposed to
PAF, and mitochondrial TMRM uptake in rodent cortical neuronal
cultures treated for 1 h with increasing doses of PAF. Data in Fig.
4A demonstrate a 50% increase in ROS after a 1-h exposure to 464
nM dose of cPAF, while Tat, at a dose of 2.5 ␮g/ml, produced a
110% increase in ROS relative to control vehicle. As a positive
control, H2O2, at a dose of 500 ␮M, sufficient to induce neuronal
apoptosis (37), induced a 90% increase in ROS relative to control
(Fig. 4A). Data in B demonstrate that PAF also induced a dosedependent increase in mitochondrial TMRM uptake, reflecting a
dose-dependent mitochondrial hyperpolarization. However, in
these same rodent cortical neuronal cultures, coincubation of 2.5
␮g/ml Tat with a congener of the endogenous antioxidant tauroursodeoxycholate, derived from bile salts (38, 39), or the potent antioxidant N-acetylcysteine (NAC), both failed to normalize ⌬␺m to
control values (C). Taken together, these data suggest that mitochondrial hyperpolarization is a generalized response to both the
HIV-1 virotoxin Tat and its proinflammatory cellular mediator
PAF, but that cellular ROS production does not directly cause the
mitochondrial hyperpolarization.
FIGURE 4. Tat and PAF increase ROS in cortical neuronal cultures independently of ⌬␺m. A, Treatment of rat cortical neurons for 24 h with Tat or
PAF induced elevated levels of ROS vs control, as measured with oxidizable dye indicator CM-H2DCFDA (DCF). This effect was greatest with 2.5 ␮g/ml
Tat, resulting in a 110% increase in ROS production (ⴱ, p ⬍ 0.0001 vs control; unpaired t test), whereas 464 nM PAF resulted in a 50% increase in ROS
production, and the H2O2 positive control induced a 90% increase. B, Like Tat, its downstream mediator PAF also induced a dose-dependent rise in ⌬␺m
(ⴱ, p ⬍ 0.002 or p ⬍ 0.0001 vs control, respectively; unpaired t test). C, However, the antioxidants NAC (150 ␮M) and TUDCA (300 ␮M) failed to block
the 2.5 ␮g/ml Tat-induced increase in ⌬␺m. NAC (150 ␮M) itself induced a rise in ⌬␺m, and augmented the increase in ⌬␺m caused by 2.5 ␮g/ml Tat (ⴱ,
p ⬍ 0.01; unpaired t test). These results suggest that ROS do not mediate Tat’s increase of ⌬␺m. Data acquisition and analysis were performed, as described
in Materials and Methods, and comparable results were obtained over several experiments.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
By taking average pixel intensity over a neuronal cell body, MCS
values could be impacted by changes in cell size or morphology.
To ensure that fluorescence signal changes were not caused by
changes in cell size or morphology, total TMRE uptake in individual culture wells was also determined at matching time points
by loading sample culture wells with 1 nM TMRE, and assessment
of the total accumulated dye signal from the cells using a fluorescent plate reader (see Materials and Methods). Total dye uptake
was normalized to cell number, expressed as TUS, and compared
with MCS values. Quantification of TMRE uptake by both methods yielded equivalent values for TMRE signal intensity (Fig. 3G),
indicating that increases in fluorescent TMRE signal intensity were
not artifacts of changes in cell size, morphology, or three-dimensional mitochondrial localization.
To ensure that the increased TMRE signal we observed in neuronal mitochondria after Tat treatment was not due to an increase
in mitochondrial mass or volume, Mitotracker green, a largely potential independent mitochondrial marker (21), was also quantitated by both MCS and TUS methods (Fig. 3H). Contrary to results
with TMRE, Tat treatment did not increase Mitotracker green signal by either method, suggesting that mitochondrial mass and volume remained constant in Tat-treated cultures.
Together, these data show that the Tat-induced increase in mitochondrial TMRE signal was not an artifact of alterations in cell
size, morphology, or total mitochondrial surface area, but rather
reflects a verifiable increase in ⌬␺m.
␮g/ml Tat or 100 mM KCl (a concentration known to induce
strong depolarization of the ⌬␺p). In Fig. 3I, Tat induces a significant 11% increase in the relative intensity of the signal from
DiBAC4(3) fluorescence ( p ⬍ 0.001), compared with a 22% increase after KCl ( p ⬍ 0.001). These findings demonstrate that
despite the ability of 2.5 ␮g/ml Tat to depolarize the ⌬␺p, thus
decreasing the influx of cationic TMRE into the intracellular space
relative to control vehicle, the mitochondrial TMRE signal is still
significantly higher in Tat-treated vs control cells (Fig. 3, compare
B and A, respectively).
The Journal of Immunology
4339
Possible mitochondrial loci for Tat-induced hyperpolarization
of ⌬␺m
We next sought to identify mitochondrial agents that could block
the Tat-induced mitochondrial hyperpolarization. As expected, our
data demonstrate that coincubation of Tat with a low dose (100
nM) application of the mitochondrial uncoupling protonophore
FCCP completely abolishes low and high dose Tat-mediated increases in ⌬␺m at both 1 and 24 h (Fig. 5A). FCCP decreases ⌬␺m
by increasing proton transport inward across the mitochondrial
membrane, down the proton concentration gradient. However, it is
important to note that at the FCCP concentration used in this study
(100 nM), FCCP coincubated with Tat did not result in a complete
loss of ⌬␺m, but rather simply returned ⌬␺m to approximately
resting (i.e., baseline or control) levels (Fig. 5A). Importantly, at 1
and 24 h, 100 nM FCCP by itself had little effect on ⌬␺m vs control
(Fig. 5A). In stark contrast, however, 1 h of 5 ␮M FCCP alone
resulted in a strong depolarization of ⌬␺m, reflected by significant
loss of the TMRM signal (Fig. 5C).
Interestingly, some similar effects were observed when our cortical neuronal cultures were treated with both low and high dose
Tat coincubated with 100 ␮M tolbutamide (Fig. 5B). Tolbutamide
is a potent antagonist for the ATP-sensitive K⫹ channels in the
inner mitochondrial membrane (40). As expected, 1-h incubation
with 100 ␮M tolbutamide resulted in an apparent trend toward
mitochondrial hyperpolarization, as would be expected by acutely
blocking mitochondrial K⫹ influx, although this trend was lost by
24 h (Fig. 5B). At 1 h, coincubation of 100 ␮M tolbutamide with
2.5 ␮g/ml Tat resulted in a 75– 80% attenuation of the Tat-mediated increase in ⌬␺m. At 24 h, tolbutamide blocked the hyperpolarizing effect of high-dose Tat by ⬃50%.
Neuronal ATP/ADP ratios and adenosine levels increase with
prolonged exposure to Tat
Because mitochondrial membrane hyperpolarization might affect
mitochondrial bioenergetics by numerous mechanisms, including
alterations in the ability of cytochrome c oxidase to regulate mitochondrial energy metabolism (41, 42), we measured intracellular
ATP and ADP content in response to Tat treatment. Changes in the
ratio of ATP to ADP content is a key indicator of cells’ bioenergetic status, with rising ATP/ADP ratios indicating increased energy reserves, and declining ATP/ADP ratios indicating lower energy supplies (or increased ATP use). As shown in Fig. 6A, ATP/
ADP ratios trended higher at most doses and time points analyzed,
although this effect was only significant for 2.5 ␮g/ml Tat at 24 h
(unbalanced ANOVA analysis, p ⫽ 0.0053). Equally important,
however, adenosine (ATP and/or ADP) levels were significantly
elevated under the majority of treatment conditions analyzed. We
found that both low and high dose Tat were able to increase ATP
and ADP levels relative to time-matched control vehicle after 24 h
of exposure (Fig. 6B). After 48 h of either low or high dose Tat
treatment, both ATP and ADP levels were still significantly increased vs control, but less so (Fig. 6B). Tat’s effects on adenosine
levels at 1 h trended toward increases, but were not significant.
As additional controls, Tat’s downstream inflammatory mediator, PAF, which, like Tat, hyperpolarizes mitochondrial membranes (see Fig. 4B), also increased ATP and ATP levels in cortical
neurons, while a ⌬␺m-depolarizing dose of the protonophore
FCCP (5 ␮M; see Fig. 5C) abolished ATP production (data not
shown). These results suggest that the Tat-induced rises in adenosine levels are not isolated or aberrant effects of Tat or the assay,
FIGURE 6. Changes in neuronal ATP/ADP ratios and adenosine levels
with chronic Tat exposure. Primary cortical neurons were treated for 1, 24,
or 48 h with 0 (control), 0.1, or 2.5 ␮g/ml Tat. A, The ATP/ADP ratio was
significantly increased at time ⫽ 24 h when the Tat dosage was 2.5 ␮g/ml
(ⴱ, p ⫽ 0.0053 for 0 ␮g/ml vs 2.5 ␮g/ml Tat at 24 h; SAS PROC GLM
procedure for unbalanced ANOVA), but there were no other significant
changes, despite a general upward trend. All main effects and interaction
effects of time (i.e., 1, 24, or 48 h) and dosage (i.e., 0, 0.1, or 2.5 ␮g/ml Tat)
were included in the unbalanced ANOVA analysis. B, Importantly, both
ATP and ADP levels were significantly increased by both doses of Tat at
24 and 48 h, even where ATP/ADP ratios were not significantly affected (at
24 and 48 h; ⴱ, p ⬍ 0.0002 for all changes in ATP and ADP levels vs
time-matched 0 ␮g/ml control; ANOVA and unpaired t tests).
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 5. Tat-induced mitochondrial hyperpolarization is reversible with low-dose protonophore or KATP channel antagonist. A, One- or 24-h
coincubation of Tat (0.1 or 2.5 ␮g/ml) with a low dose (100 nM) of the protonophore FCCP normalized ⌬␺m to approximately baseline values. A total
of 100 nM FCCP by itself had minimal effect on ⌬␺m. High dose FCCP (5 ␮M) for 1 h, in contrast, induced a strong mitochondrial depolarization (see
C) (ⴱ#, p ⬍ 0.001; unpaired t tests). B, Similarly, coincubation of 100 ␮M tolbutamide with the same doses of Tat completely abrogated 2.5 ␮g/ml Tat’s
increase of ⌬␺m at 1 h, and partially abrogated (by 50%) this increase at 24 h (ⴱ, p ⬍ 0.01; #, p ⬍ 0.01 at 1 h, p ⬍ 0.1 at 24 h; unpaired t tests). C, High
dose protonophore FCCP (5 ␮M) decreases mitochondrial TMRM uptake at 1 h (ⴱ, p ⬍ 0.01; unpaired t test). This effect of 5 ␮M FCCP (which would
induce strong mitochondrial depolarization) on TMRM uptake, when contrasted to the increased TMRM signal seen with 2.5 ␮g/ml Tat treatment,
confirmed that Tat was increasing ⌬␺m (and see other controls described in Materials and Methods).
4340
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
but rather may more broadly represent a pathologic response seen
in HAD.
Taken together, these results suggest that although cultures may
not experience general energy deficits (as would traditionally be
associated with a decline in ATP/ADP ratio, although these data do
not exclude the possibility of local intracellular energy deficits),
cultures appear to respond to Tat by increasing ATP levels, most
likely by increased production rather than decreased consumption
of ATP. Moreover, increases in ATP production may be a result of
the Tat-induced mitochondrial hyperpolarization seen over a similar time scale (see Fig. 2), because increased ⌬␺m will increase
protonmotive driving force through the F1F0-ATPase, thus increasing
ATP production. Together, these results may suggest that neurons
exposed to varying concentrations of Tat respond by increasing
cellular energy production (with a corresponding mitochondrial
hyperpolarization), and this increased metabolic activity may contribute to the increased oxidative stress induced by Tat or its proinflammatory mediator PAF (see Fig. 4).
To test the functional correlates of our bioenergetic data, and because both Tat and PAF have been shown to increase excitatory
neuronal activity (5, 24, 27), we investigated changes in FM1-43
uptake in response to Tat and PAF. These experiments revealed
punctate FM1-43 staining along neuronal processes (characteristic
of synaptic vesicle labeling), with heavier punctate staining in areas of high neurite density, and preferential loss of punctate staining relative to background signal upon release in KCl (Fig. 7,
A–C).
FIGURE 7. Tat and PAF cause an
antioxidant-sensitive rise in vesicular
activity. Images of Brightfield (A)
and fluorescent FM1-43 (B) labeling
show that FM1-43 uptake results in
labeling that is punctate and oriented
primarily along neuronal processes.
C, Moreover, activity-dependent
FM1-43 release by KCl depolarization preferentially reduces the punctate vesicular labeling vs background
signal. D, In ⱕ14- and ⱖ14-day-old
primary neurons, 2.5 ␮g/ml Tat and
4.25 ␮g/ml Tat induce an increase in
spontaneous activity-dependent vesicular uptake, an effect that is both
dose and culture age dependent. E,
However, no effect is seen with mutated Tat protein, suggesting that
Tat’s effect is due to biologically sp.
act. of its functional region. F, A total
of 464 nM PAF had an even greater
effect on FM1-43 uptake at 24 h. G,
The antioxidant TUDCA completely
eliminated Tat’s effects on FM1-43
uptake for all doses of Tat, whereas
the antioxidant (H) NAC partially reduced 2.5 ␮g/ml Tat’s effect on
FM1-43 in an NAC dose-dependent
fashion. (ⴱ, #, p ⬍ 0.01 for comparisons vs relevant control; unpaired t
tests).
Stabilizing the mitochondrial electrochemical gradient or
blocking ATP-sensitive K⫹ channels attenuates Tat-induced
neuronal apoptosis
A central question that remained unanswered in these studies was
what perturbations in mitochondrial bioenergetics ultimately
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Vesicle recycling increases with exposure to Tat or PAF
Twenty-four-hour application of Tat to primary rat cortical neurons, a time when cellular production of ATP is maximal (Fig. 6),
caused a dose- and age-dependent increase in vesicular activity, as
measured by FM1-43 uptake (Fig. 7D). In neurons aged greater
than 14 DIV, Tat resulted in a 50% ( p ⬍ 0.02) and 75% ( p ⬍
0.002) increase in FM1-43 uptake at doses of 2.5 and 4.25 ␮g/ml,
respectively. Moreover, at each dose of Tat tested, relative to control, neurons aged greater than 14 DIV showed 40 –50% more
Tat-induced vesicular activity than neurons aged ⬍14 days DIV
(Fig. 7D). Tat’s effects on FM1-43 uptake were specific, as they
could be blocked by both polyclonal Tat Abs (data not shown) and
a mutated functionally deficient (functional region-deleted) Tat
peptide (Fig. 7E).
Not surprisingly, PAF, the proinflammatory mediator of Tat,
also increased FM1-43 uptake, to an even greater extent than Tat
(6-fold vs 1.75-fold) (Fig. 7F). Interestingly, the antioxidant tauroursodeoxycholic acid (TUDCA) (Fig. 7G) was able to completely reverse Tat’s effects on FM1-43 uptake. The antioxidant
NAC had a partial dose-dependent effect on reducing the Tat-induced increase in FM1-43 uptake (Fig. 7H). Thus, increased metabolic activity of nerve terminals, reflected by increased vesicle
recycling, can be returned to basal or near-basal levels by treatment with antioxidants.
The Journal of Immunology
FIGURE 8. Stabilization of the mitochondrial electrochemical gradient
or blockade of ATP-sensitive K⫹ channels attenuates Tat-induced neuronal
apoptosis. Cotreatment of 0.1 ␮g/ml or 2.5 ␮g/ml Tat-treated cortical cultures with A, a low (100 nM) dose of the protonophore FCCP, capable of
returning ⌬␺m to baseline (see Fig. 5A), or B, the mitochondrial KATP
channel antagonist tolbutamide (100 ␮M), partially attenuated apoptotic
cell death in response to Tat. FCCP was protective only against the high
dose of Tat, which was not unexpected, whereas tolbutamide was protective against both doses of Tat and showed no toxicity when applied alone.
(ⴱ, ⫹, #, p ⬍ 0.01 for all comparisons vs relevant control, and as indicated;
unpaired t tests).
As a whole, these results suggest that preventing the Tat-induced increase in ⌬␺m may serve to ameliorate a subsequent neuronal apoptosis. This finding may have critical implications for
protecting against neuronal dysfunction and death in HAD and
other neurodegenerative diseases.
Discussion
Our data bring to the forefront issues relevant to HIV-1-associated
neurologic disease. Because highly active antiretroviral therapy
(HAART) can decrease viral burden to very low or undetectable
levels, florid HIV-1 encephalitis has become an uncommon neuropathologic outcome. Rather, HAART may have altered the phenotype of CNS disease by indirectly decreasing inflammatory burden in the CNS.
As a result, HAART has changed the presentation of HIV-associated neurologic disease from a rapidly progressive process to
a more indolent pattern, characterized by minor cognitive and minor motor deficits (43– 47). Thus, we now face the challenge of
modeling a disease process that is characterized more by reversible, phenotypic changes in neuronal function and synaptic networks, rather than frank neuronal loss.
For these reasons, we have used the HIV-1 Tat and PAF in our
experiments, because they induce multiple biologic effects important to HAD, and they represent physiologically relevant reagents
that can induce both inflammatory and excitotoxic events at dose
ranges that elicit both sublethal and lethal effects for vulnerable
neurons. Moreover, several variants of Tat (Tat1–72, Tat1– 86, Tat1–
101) were assessed in our experiments, with similar results in the
TMRE/M, FM1-43, ROS, ATP, and cell death assays (data for
Tat1– 86, Tat1–101 are not shown) (see Refs. 6 and 32).
Given this changing face of HAD, and because HIV-1-associated neurologic disease does not appear to correlate with neuronal
apoptosis (12), but rather diminution of the dendritic arbor, we first
studied neurite retraction as a suitable starting point to investigate
how Tat might alter neurotransmission in a potential biologic
model of HAD. We found that Tat did indeed induce simplification
of the neuronal architecture, and have suggestive evidence that
these changes might be linked to either increased synaptic activity
and/or mitochondrial hyperpolarization. Studies are currently underway to more conclusively link aberrant synaptic activity and
mitochondrial hyperpolarization with recoverable diminution of
the dendritic arbor.
The ability of Tat and its cellular phospholipid metabolite PAF
to hyperpolarize neuronal ⌬␺m was more unexpected, given that
collapse of the ⌬␺m is frequently considered one of the penultimate events before neuronal apoptosis. Our findings in this study
are not without support, however, as only in recent years have
others begun to report mitochondrial hyperpolarizations associated
with neuronal apoptosis (19, 21, 48).
A potential clue to the biologic outcome of increased ⌬␺m, particularly as it relates to HAD, comes from a study of PBMCs in
HIV-1-infected patients. In this study, the investigators reported
that HIV-1-infected PBMCs have a hyperpolarized ⌬␺m, and behave like activated T cells, prone to apoptotic stimuli, while
HAART protease inhibitors can inhibit this effect (49). To our
knowledge, and contrary to the traditional dogma associating depolarization of ⌬␺m with cell death, this was the first report showing that pharmacologically inhibiting a pathology-induced rise in
⌬␺m can protect cells from death. We have extended that work in
this study by showing similar results in a neuronal culture model.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
meant for cell fate of neurons vulnerable to Tat. We speculated that
prolonged exposure to even low levels of Tat, with a concomitant
increase in ⌬␺m, energy production, and metabolic activity (reflected by increased vesicular uptake, recycling, and release),
would ultimately result in increased levels of neuronal apoptosis.
Thus, we used the mitochondrial uncoupling reagent FCCP, at
doses known to reverse hyperpolarization of ⌬␺m (Fig. 5A), to
determine whether it was neuroprotective against increasing doses
of Tat (Fig. 8A). A total of 100 nM FCCP was ineffective against
low dose Tat, but partially reversed cell death from high dose Tat
(Fig. 8A). Furthermore, the ATP-sensitive K⫹ channel blocker,
tolbutamide, at a concentration that prevented Tat-induced mitochondrial hyperpolarization, also attenuated both low and high
dose Tat-mediated neuronal apoptosis (Fig. 8B).
Parenthetically, low (i.e., 0.1 ␮g/ml) dose Tat exhibited moderately variable levels of neuronal apoptosis between replicate experiments (Fig. 8, compare A with B), which is a common occurrence for any cytotoxicity study, and most likely reflects intrinsic
variability of culture systems. However, the dose-response relationship between low and high dose (2.5 ␮g/ml) Tat was always
maintained. Although it is possible that this variability also accounts for the limited neuroprotective effect of FCCP observed at
this dose of Tat, we believe it more likely that FCCP’s failure to
provide neuroprotection under these conditions stems from the inherent toxicity of mitochondrial uncouplers. Indeed, while the
FCCP paradigm demonstrated in this study suggests a controlled
uncoupling effect (note in Fig. 5A, 100 mM FCCP returns ⌬␺m to
baseline, but not significantly below), and controlled mitochondrial uncoupling has been shown to have neuroprotective effects
(reviewed in Ref. 33), mitochondrial uncouplers (at low, controlled
doses), e.g., as weight loss agents, have shown only limited therapeutic potential due to toxic side effects.
In addition, it is important to note that FCCP and tolbutamide
may be exerting neuroprotective effects through different mechanisms or subcellular sites of action. Both can act as mitochondrial
uncouplers, but tolbutamide has additional effects as a KATPchannel antagonist; studies are currently underway to determine
which mechanism(s) of action accounts for tolbutamide’s ⌬␺mstabilizing and neuroprotective effects.
4341
4342
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
Mitochondrial bioenergetics during Tat-induced inflammation:
effectors or the affected?
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
HAART, despite achieving effective concentrations in cerebrospinal fluid, may have relatively limited ability to achieve high concentrations in CNS parenchyma. This is evidenced in part by the
fact that HIV-1 drug resistance mutations continue to occur in the
brain of patients on HAART, a finding that has been attributed to
the presence of ineffective or partially effective drug concentrations within CNS tissue (50). As a consequence, HAART may be
unable to rescue all HIV-1-associated effects on vulnerable neurons, thereby allowing mitochondrial hyperpolarization, oxidative
stress, and changes in energy metabolism to occur in brain neurons. This underscores the importance of finding alternative agents
that may limit the mitochondrial membrane hyperpolarization in
the nervous system, and our data show that a currently approved
pharmacologic, tolbutamide, may be one agent capable of achieving this effect with consequent neuronal protection. Importantly,
because we can achieve some neuronal protection by directly diminishing the mitochondrial proton gradient (the chief component
of ⌬␺m) with FCCP, this suggests that hyperpolarization of ⌬␺m
may be directly causal to neuronal apoptosis, and would be consistent with a paradigm of neuronal protection by controlled mitochondrial uncoupling (see Ref. 33 for review).
Mitochondrial hyperpolarization in neurons has been observed
in other model systems, including 2-amino-5-hydroxy-5-methyl4-isoxazolepropion acid-induced increases in ⌬␺m in cerebellar
Purkinje neurons with an acute time course (i.e., 30 –90 min) similar to our findings after Tat treatment (Fig. 2) (51). Exposure of
hippocampal neurons to staurosporine also induced elevations in
⌬␺m, followed by release of cytochrome c, ultimately leading to
apoptosis (19). However, neither of these studies addressed the
mitochondrial response to increasing doses of neurotoxic stimuli,
including the ability of mitochondria to normalize their electrochemical gradient to more homeostatic values (i.e., a ⌬␺m similar
to control values) over time (see Fig. 2).
Despite this transient stabilization of ⌬␺m after Tat treatment, it
would appear that mitochondria are ultimately affected by a cellular milieu altered during Tat (and PAF)-induced inflammation.
The increased ATP and ADP levels (and in some cases ATP/ADP
ratio) we observed are consistent with, and most likely correspond
to, the observed increases in ⌬␺m, and would appear to reflect an
increased activity of the metabolic pathways leading to ATP production, notably glycolysis and/or the TCA cycle. Although our
current data do not support the hypothesis that neurons are dying
from wholesale energy deprivation (because ATP/ADP ratios are
maintained), nor do they eliminate the possibility that local energy
deficits exists, for example, perhaps at hyperactive synapses. In
addition, our data also strongly indicate that excessive ATP production may be equally detrimental, resulting in neuronal cell
death by excess ROS levels as byproducts of oxidative phosphorylation. These concepts are discussed in greater detail below, and
studies are currently underway to test these hypotheses. Either
way, our data demonstrating that tolbutamide can normalize ⌬␺m
at 24 h (see Fig. 5), as well as ameliorate neuronal apoptosis (see
Fig. 8), suggest a temporal window of potentially reversible dysfunction, and buttress the concept of reversible neuronal dysfunction in HAD.
Ultimately, it is difficult to determine precisely what intracellular pathways or mechanisms cause the mitochondrial hyperpolarization. In this study, we have shown correlative, although not
conclusive, data that mitochondrial KATP channels may be involved in this effect, because blocking these channels with tolbutamide attenuates the mitochondrial hyperpolarization. This would
seem contrary to well-documented evidence that, in the absence of
other effects, blocking mitochondrial KATP channels frequently
results in a slight increase in ⌬␺m due to a decreased inward K⫹
gradient and consequent decreased K⫹ (out)/H⫹ (in) exchange, as
has been shown for tolbutamide and other mitochondrial KATP
antagonists (Ref. 40 and see Fig. 5); thus, other actions of tolbutamide, such as regulation of calcium or glycolytic pathways, may
explain its ability to ameliorate Tat-induced rises in ⌬␺m. For example, tolbutamide reduces glucose availability to the glycolytic
pathway, which in turn reduces the supply of pyruvate to the
Kreb’s cycle, resulting in decreased production of reduced NADH,
and consequent decline in ⌬␺m. Notably, one other report has
demonstrated a neurotoxic effect that is induced via activation of
neuronal KATP channels, and can be blocked by tolbutamide (52),
which may be consistent with results shown in this work. Tat may
also contribute to increased ⌬␺m by other means, including inhibition of the mitochondrial F1F0-ATP-ase, or increasing electrochemical potential through increased production of (NADH)
substrate.
Although it might be tempting to conclude that increased ⌬␺m
and ATP production represents a compensatory response to the
increased energetic demands of Tat or PAF treatment, such as
would be required for increased vesicle recycling or synaptic transmission, this may not be the interpretation most consistent with the
data we have shown in this work. Were this a true compensatory
response, i.e., neurons were increasing ⌬␺m and ATP as a protective measure, or in response to energy demands, then contrary to
what we found, we would expect that blocking the rise in ⌬␺m
would exacerbate cell death, and further, that at best ATP/ADP
ratios would remain constant (if ATP supply met demand), or
would decline (if ATP demand outstripped neurons’ production
capability). To the contrary, cell death induced by 2.5 ␮g/ml Tat
for 24 h was associated with increased ADP, ATP, and ATP/ADP
ratio, suggesting that Tat-treated neurons were not ATP starved.
Therefore, overall, it would appear that the condition of heightened
⌬␺m and ATP production was not favorable for cell survival.
An alternate interpretation consistent with our findings would be
that Tat or PAF treatment increases activity of the Kreb’s cycle,
hence also elevating ⌬␺m by increased production of NADH, leading to increased electron transport chain activity and consequent
proton extrusion, and this enzymatic hyperactivity is detrimental to
the cell. By this model, the increased ATP production that we
observed, rather than being a compensatory response, would simply be a byproduct of the increased ⌬␺m; this would also explain
why ATP/ADP ratios did not decline (and in some cases even rose)
with Tat treatment, suggesting stable or increasing ATP reserves.
Studies are currently underway to pinpoint more specifically the
mechanisms that lead to mitochondrial hyperpolarization and increased ATP production in Tat- and PAF-treated cells.
Regardless of whether neurons were responding to Tat (or PAF)
by actively or by inadvertently increasing ⌬␺m and ATP production, and regardless of cellular bioenergetic status, either scenario
could be equally detrimental to the cell, because ROS production
is directly linked to mitochondrial membrane polarization and
ATP production. This relationship, i.e., limiting ROS byproducts
of oxidative phosphorylation, is thought to underlie the neuroprotective properties of controlled mitochondrial uncoupling. These
concepts are also consistent with our data indicating that Tat-mediated mitochondrial membrane hyperpolarization may be upstream of ROS production, because cotreatment with two different
antioxidants failed to reverse Tat-mediated increases in ⌬␺m (see
Fig. 4C). This is in agreement with data indicating that increased
⌬␺m will lead to increased ROS production from mitochondrial
complex III, the cell’s chief superoxide (䡠O2)-generating source,
The Journal of Immunology
Synaptic dysfunction during inflammation
Data in Fig. 7 demonstrate that neuronal vesicular uptake actually
increases after exposure to Tat or PAF, during the time that mitochondrial ⌬␺m is hyperpolarized, and ATP production is maximal. This is consistent with proven excitatory activity of Tat and
PAF (5, 24, 27), and also begs the question of whether the changes
in mitochondrial function (i.e., increased ⌬␺m and increased ATP
production) are causal to, a result of, or independent of changes in
vesicular activity. On the one hand, increased ATP production will
signal adequate energy reserves to the ATP-sensitive plasma membrane KATP channels, thus closing these channels, blocking K⫹
efflux, and resulting in membrane depolarization and increased
synaptic activity. This explanation may not be the one most consistent with the known excitatory actions of Tat and PAF, however. Rather, increased energetic demands from synaptic excitation
by these molecules may stimulate increased ATP production, made
possible by increased ⌬␺m. Alternatively, the potential failure in
the efficiency of synaptic transmission (from excitotoxic stress)
might elicit increased presynaptic vesicular supply of neurotransmitter as a compensatory mechanism. Ultimately, any of these
mechanisms would most likely fail due to continued energy demands on the nerve terminal, and the eventual failure to maintain
energy production sufficient for cellular functions relevant to neurotransmission, as would be consistent in the falloff in ATP/ADP
ratio at 48 h shown in this work. Finally, we cannot exclude the
possibility that increased vesicular activity is unrelated to changes
in mitochondrial status and ATP levels; thus, future studies will
attempt to dissect this relationship.
Synaptic apoptosis: a pathophysiologic substrate for HIV-1associated neurologic disease in the era of HAART
The concept of synaptic apoptosis has been advanced in other neurodegenerative diseases, particularly Alzheimer’s disease, by Gilman and Mattson (10), based on observations that caspases, the
proapoptotic Bcl-2 protein Bax, mitochondrial factors cytochrome
c, and apoptosis-inhibiting factor are present in neurites and synaptic terminals (56 –58). Studies with synaptosomal preparations
from hippocampal and cortical neurons have demonstrated that
treatment with a caspase inhibitor prevents collapse of the ⌬␺m
and release of proapoptotic factors into the cytoplasm (53). Additional support for the phenomenon of synaptic apoptosis in neurodegenerative disease comes from a report that demonstrates
amyloid ␤-mediated degeneration of primary hippocampal neurites by a caspase-mediated mechanism in compartmentalized cultures (11). Thus, our findings of Tat-mediated neurite retraction
(Fig. 1), as well as our findings of an accompanying synaptic and
metabolic stress, are not only consonant with the neuropathology
of HIV-1-associated neurologic disease (31, 59), but also with the
concept of synaptic apoptosis, and therefore, may represent a neurodegenerative mechanism similar to the type of neurite dissolution seen after exposure to A␤ (11). As reported in this work, our
findings suggest numerous potential mechanisms by which the effects we describe might participate directly in synaptic apoptosis
and neurite retraction, including toxic activation of neuronal
plasma membrane ATP receptors; activation of proapoptotic pathways at the synapse by ROS or other factors; either intracellularly
or extracellularly induced synaptic hyperactivity resulting in dying
back of neurites and loss of synaptic function; or excitotoxic cell
death, to name a few. Future studies will address these and other
relevant in situ events that occur in nerve terminals and dendrites
exposed to HIV-1 virotoxins such as Tat, or cellular metabolites
such as PAF, including activation of caspases and other effector
proteins and mechanisms potentially involved in mediating reversible synaptic apoptosis or dysfunction.
Acknowledgments
We are grateful to Dr. Sanjay Maggirwar for his critical reading of this
manuscript.
Disclosures
The authors have no financial conflict of interest.
References
1. Nath, A. 2002. Human immunodeficiency virus (HIV) proteins in neuropathogenesis of HIV dementia. J. Infect. Dis. 186(Suppl. 2):S193.
2. Tardieu, M., C. Hery, S. Peudenier, O. Boespflug, and L. Montagnier. 1992.
Human immunodeficiency virus type 1-infected monocytic cells can destroy human neural cells after cell-to-cell adhesion. Ann. Neurol. 32:11.
3. Sawaya, B. E., P. Thatikunta, L. Denisova, J. Brady, K. Khalili, and S. Amini.
1998. Regulation of TNF␣ and TGF␤-1 gene transcription by HIV-1 Tat in CNS
cells. J. Neuroimmunol. 87:33.
4. Perry, S. W., J. A. Hamilton, L. W. Tjoelker, G. Dbaibo, K. A. Dzenko,
L. G. Epstein, Y. Hannun, J. S. Whittaker, S. Dewhurst, and H. A. Gelbard. 1998.
Platelet-activating factor receptor activation: an initiator step in HIV-1 neuropathogenesis. J. Biol. Chem. 273:17660.
5. Haughey, N. J., C. P. Holden, A. Nath, and J. D. Geiger. 1999. Involvement of
inositol 1,4,5-trisphosphate-regulated stores of intracellular calcium in calcium
dysregulation and neuron cell death caused by HIV-1 protein tat. J. Neurochem.
73:1363.
6. New, D. R., S. B. Maggirwar, L. G. Epstein, S. Dewhurst, and H. A. Gelbard.
1998. HIV-1 Tat induces neuronal death via tumor necrosis factor-␣ and activation of non-N-methyl-D-aspartate receptors by a NF␬B-independent mechanism.
J. Biol. Chem. 273:17852.
7. Mattson, M. P., and D. Liu. 2002. Energetics and oxidative stress in synaptic
plasticity and neurodegenerative disorders. Neuromolecular Med. 2:215.
8. Flores, S. C., J. C. Marecki, K. P. Harper, S. K. Bose, S. K. Nelson, and
J. M. McCord. 1993. Tat protein of human immunodeficiency virus type 1 represses expression of manganese superoxide dismutase in HeLa cells. Proc. Natl.
Acad. Sci. USA 90:7632.
9. Aksenov, M. Y., U. Hasselrot, A. K. Bansal, G. Wu, A. Nath, C. Anderson,
C. F. Mactutus, and R. M. Booze. 2001. Oxidative damage induced by the injection of HIV-1 Tat protein in the rat striatum. Neurosci. Lett. 305:5.
10. Gilman, C. P., and M. P. Mattson. 2002. Do apoptotic mechanisms regulate
synaptic plasticity and growth-cone motility? Neuromolecular Med. 2:197.
11. Ivins, K. J., E. T. Bui, and C. W. Cotman. 1998. ␤-Amyloid induces local neurite
degeneration in cultured hippocampal neurons: evidence for neuritic apoptosis.
Neurobiol. Dis. 5:365.
12. Everall, I. P. 2000. Neuronal damage: recent issues and implications for therapy.
J. Neurovirol. 6(Suppl. 1):S103.
13. Ma, M., and A. Nath. 1997. Molecular determinants for cellular uptake of Tat
protein of human immunodeficiency virus type 1 in brain cells. J. Virol. 71:2495.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
and data that 䡠O2 generation by complex 1 is proportional to ATP
production (53). In summary, increased ROS production may be a
result of, rather than causal to, increased ⌬␺m, and may link Tat’s
neurotoxicity to its ability to enhance ⌬␺m and ATP production.
Calcium’s contributions to Tat’s effects on ⌬␺m also cannot be
ignored, because Tat increases intracellular calcium levels in neurons (5), resulting in excitotoxic cell death (54, 55). A role for
calcium would also be consistent with the possible model of mitochondrial hyperpolarization just described, as calcium can increase production of NADH by stimulating dehydrogenases. This
may be a principal mechanism by which in our model, increased
Ca2⫹ levels or altered Ca2⫹ signaling may lead to an increase in
⌬␺m, rather than the more frequently implicated model whereby
excessive Ca2⫹ influx leads to mitochondrial Ca2⫹ uptake, invariably followed by a consequent decline in ⌬␺m, release of proapoptotic factors, and ultimately cell death. Alternatively, increased
⌬␺m could be directly caused by diminished mitochondrial Ca2⫹
uptake, which in turn might result from either increased mitochondrial Ca2⫹ release or increased Ca2⫹ uptake by the endoplasmic
reticulum. Studies are underway to determine the significance of
Ca2⫹ altered handling, as well as Ca2⫹-mediated downstream signaling events including pertussin-sensitive G protein-coupled
events, protein kinase C, tyrosine kinases, and events modulated
by zinc binding sites on the N-methyl-D-asparate and other receptors, in our model of mitochondrial membrane hyperpolarization.
4343
4344
HIV-1 TAT APOPTOSIS VIA MITOCHONDRIAL HYPERPOLARIZATION
38. Keene, C. D., C. M. Rodrigues, T. Eich, M. S. Chhabra, C. J. Steer, and
W. C. Low. 2002. Tauroursodeoxycholic acid, a bile acid, is neuroprotective in
a transgenic animal model of Huntington’s disease. Proc. Natl. Acad. Sci. USA
99:10671.
39. Sola, S., R. E. Castro, P. A. Laires, C. J. Steer, and C. M. Rodrigues. 2003.
Tauroursodeoxycholic acid prevents amyloid-␤ peptide-induced neuronal death
via a phosphatidylinositol 3-kinase-dependent signaling pathway. Mol. Med.
9:226.
40. Inoue, I., H. Nagase, K. Kishi, and T. Higuti. 1991. ATP-sensitive K⫹ channel in
the mitochondrial inner membrane. Nature 352:244.
41. Kadenbach, B., M. Huttemann, S. Arnold, I. Lee, and E. Bender. 2000. Mitochondrial energy metabolism is regulated via nuclear coded subunit of cytochrome c oxidase. Free Radical Biol. Med. 29:211.
42. Murphey, M. P., and M. D. Brand. 1988. The stoichiometry of charge translocation by cytochrome oxidase and the cytochrome bc1 complex of mitochondria
at high membrane potential. Eur. J. Biochem. 173:645.
43. Brew, B. J. 2004. Evidence for a change in AIDS dementia complex in the era of
highly active antiretroviral therapy and the possibility of new forms of AIDS
dementia complex. AIDS 18(Suppl. 1):S75.
44. McArthur, J. C., N. Haughey, S. Gartner, K. Conant, C. Pardo, A. Nath, and
N. Sacktor. 2003. Human immunodeficiency virus-associated dementia: an
evolving disease. J. Neurovirol. 9:205.
45. Langford, T. D., S. L. Letendre, G. J. Larrea, and E. Masliah. 2003. Changing
patterns in the neuropathogenesis of HIV during the HAART era. Brain Pathol.
13:195.
46. Gray, F., F. Chretien, A. V. Vallat-Decouvelaere, and F. Scaravilli. 2003. The
changing pattern of HIV neuropathology in the HAART era. J. Neuropathol. Exp.
Neurol. 62:429.
47. Sperber, K., and L. Shao. 2003. Neurologic consequences of HIV infection in the
era of HAART. AIDS Patient Care STDS 17:509.
48. Piacentini, M., M. G. Farrace, L. Piredda, P. Matarrese, F. Ciccosanti, L. Falasca,
C. Rodolfo, A. M. Giammarioli, E. Verderio, M. Griffin, and W. Malorni. 2002.
Transglutaminase overexpression sensitizes neuronal cell lines to apoptosis by
increasing mitochondrial membrane potential and cellular oxidative stress.
J. Neurochem. 81:1061.
49. Matarrese, P., L. Gambardella, A. Cassone, S. Vella, R. Cauda, and W. Malorni.
2003. Mitochondrial membrane hyperpolarization hijacks activated T lymphocytes toward the apoptotic-prone phenotype: homeostatic mechanisms of HIV
protease inhibitors. J. Immunol. 170:6006.
50. Smit, T., B. J. Brew, W. Tourtellotte, S. Morgello, B. B. Gelman, and
N. K. Saksena. 2004. Independent evolution of human immunodeficiency virus
(HIV) drug resistance mutations in diverse areas of the brain in HIV-infected
patients, with and without dementia, on antiretroviral treatment. J. Virol.
78:10133.
51. Strahlendorf, J., C. Boxa, J. Attridgea, J. Diertienb, V. Finckbonea, W. M. Henne,
M. S. Medinaa, R. Milesa, S. Oommana, M. Schneider, et al. 2003. AMPAinduced dark cell degeneration of cerebellar Purkinje neurons involves activation
of caspases and apparent mitochondrial dysfunction. Brain Res. 994:146.
52. Dugravot, S., F. Grolleau, D. Macherel, A. Rochetaing, B. Hue, M. Stankiewicz,
J. Huignard, and B. Lapied. 2003. Dimethyl disulfide exerts insecticidal neurotoxicity through mitochondrial dysfunction and activation of insect KATP channels. J. Neurophysiol. 90:259.
53. Nicholls, D. G. 2002. Mitochondrial function and dysfunction in the cell: its
relevance to aging and aging-related disease. Int. J. Biochem. Cell Biol. 34:1372.
54. Kruman, I. I., A. Nath, and M. P. Mattson. 1998. HIV-1 protein Tat induces
apoptosis of hippocampal neurons by a mechanism involving caspase activation,
calcium overload, and oxidative stress. Exp. Neurol. 154:276.
55. Haughey, N. J., A. Nath, M. P. Mattson, J. T. Slevin, and J. D. Geiger. 2001.
HIV-1 Tat through phosphorylation of NMDA receptors potentiates glutamate
excitotoxicity. J. Neurochem. 78:457.
56. Bahr, M. 2000. Live or let die: retinal ganglion cell death and survival during
development and in the lesioned adult CNS. Trends Neurosci. 23:483.
57. Mattson, M. P., J. N. Keller, and J. G. Begley. 1998. Evidence for synaptic
apoptosis. Exp. Neurol. 153:35.
58. Sugimoto, T., C. Xiao, Y. F. He, and H. Ichikawa. 1996. Bax protein-like immunoreactivity in primary sensory and hypothalamic neurons of adult rats. Neurosci. Lett. 215:37.
59. Everall, I. P., R. K. Heaton, T. D. Marcotte, R. J. Ellis, J. A. McCutchan,
J. H. Atkinson, I. Grant, M. Mallory, and E. Masliah. 1999. Cortical synaptic
density is reduced in mild to moderate human immunodeficiency virus neurocognitive disorder: HNRC Group: HIV Neurobehavioral Research Center. Brain
Pathol. 9:209.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
14. Brewer, G. J., J. R. Torricelli, E. K. Evege, and P. J. Price. 1993. Optimized
survival of hippocampal neurons in B27-supplemented Neurobasal, a new serumfree medium combination. J. Neurosci. Res. 35:567.
15. Perry, S. W., J. P. Norman, A. Litzburg, and H. A. Gelbard. 2004. Antioxidants
are required during the early critical period, but not later, for neuronal survival.
J. Neurosci. Res. 78:482.
16. Nicholls, D. G., and M. W. Ward. 2000. Mitochondrial membrane potential and
neuronal glutamate excitotoxicity: mortality and millivolts. Trends Neurosci.
23:166.
17. Rottenberg, H., and S. Wu. 1998. Quantitative assay by flow cytometry of the
mitochondrial membrane potential in intact cells. Biochim. Biophys. Acta
1404:393.
18. Ward, M. W., A. C. Rego, B. G. Frenguelli, and D. G. Nicholls. 2000. Mitochondrial membrane potential and glutamate excitotoxicity in cultured cerebellar
granule cells. J. Neurosci. 20:7208.
19. Poppe, M., C. Reimertz, H. Dussmann, A. J. Krohn, C. M. Luetjens,
D. Bockelmann, A. L. Nieminen, D. Kogel, and J. H. Prehn. 2001. Dissipation of
potassium and proton gradients inhibits mitochondrial hyperpolarization and cytochrome c release during neural apoptosis. J. Neurosci. 21:4551.
20. Scaduto, R. C., Jr., and L. W. Grotyohann. 1999. Measurement of mitochondrial
membrane potential using fluorescent rhodamine derivatives. Biophys. J. 76:469.
21. Krohn, A. J., T. Wahlbrink, and J. H. Prehn. 1999. Mitochondrial depolarization
is not required for neuronal apoptosis. J. Neurosci. 19:7394.
22. Carter, W. O., P. K. Narayanan, and J. P. Robinson. 1994. Intracellular hydrogen
peroxide and superoxide anion detection in endothelial cells. J. Leukocyte Biol.
55:253.
23. Ryan, L. A., R. L. Cotter, W. E. Zink II, H. E. Gendelman, and J. Zheng. 2002.
Macrophages, chemokines and neuronal injury in HIV-1-associated dementia.
Cell. Mol. Biol. 48:137.
24. Bouron, A. 2001. Modulation of spontaneous quantal release of neurotransmitters
in the hippocampus. Prog. Neurobiol. 63:613.
25. Prange, O., and T. H. Murphy. 1999. Correlation of miniature synaptic activity
and evoked release probability in cultures of cortical neurons. J. Neurosci.
19:6427.
26. Prasad, B. M., and S. G. Amara. 2001. The dopamine transporter in mesencephalic cultures is refractory to physiological changes in membrane voltage.
J. Neurosci. 21:7561.
27. Bazan, N. G., C. F. Zorumski, and G. D. Clark. 1993. The activation of phospholipase A2 and release of arachidonic acid and other lipid mediators at the
synapse: the role of platelet-activating factor. J. Lipid Mediators 6:421.
28. Pyle, J. L., E. T. Kavalali, S. Choi, and R. W. Tsien. 1999. Visualization of
synaptic activity in hippocampal slices with FM1-43 enabled by fluorescence
quenching. Neuron 24:803.
29. Buckley, K. M., H. E. Melikian, C. J. Provoda, and M. T. Waring. 2000. Regulation of neuronal function by protein trafficking: a role for the endosomal pathway. J. Physiol. 525:11.
30. Klingauf, J., E. T. Kavalali, and R. W. Tsien. 1998. Kinetics and regulation of fast
endocytosis at hippocampal synapses. Nature 394:581.
31. Masliah, E., R. K. Heaton, T. D. Marcotte, R. J. Ellis, C. A. Wiley, M. Mallory,
C. L. Achim, J. A. McCutchan, J. A. Nelson, J. H. Atkinson, and I. Grant. 1997.
Dendritic injury is a pathological substrate for human immunodeficiency virusrelated cognitive disorders: HNRC Group: The HIV Neurobehavioral Research
Center. Ann. Neurol. 42:963.
32. New, D. R., M. Ma, L. G. Epstein, A. Nath, and H. A. Gelbard. 1997. Human
immunodeficiency virus type 1 Tat protein induces death by apoptosis in primary
human neuron cultures. J. Neurovirol. 3:168.
33. Mattson, M. P., and D. Liu. 2003. Mitochondrial potassium channels and uncoupling proteins in synaptic plasticity and neuronal cell death. Biochem. Biophys.
Res. Commun. 304:539.
34. Stanton, P. K., and F. A. Schanne. 1986. Hippocampal long-term potentiation
increases mitochondrial calcium pump activity in rat. Brain Res. 382:185.
35. Calabresi, P., P. Gubellini, B. Picconi, D. Centonze, A. Pisani, P. Bonsi,
P. Greengard, R. A. Hipskind, E. Borrelli, and G. Bernardi. 2001. Inhibition of
mitochondrial complex II induces a long-term potentiation of NMDA-mediated
synaptic excitation in the striatum requiring endogenous dopamine. J. Neurosci.
21:5110.
36. Narita, M., S. Shimizu, T. Ito, T. Chittenden, R. J. Lutz, H. Matsuda, and
Y. Tsujimoto. 1998. Bax interacts with the permeability transition pore to induce
permeability transition and cytochrome c release in isolated mitochondria. Proc.
Natl. Acad. Sci. USA 95:14681.
37. Perry, S. W., L. G. Epstein, and H. A. Gelbard. 1997. Simultaneous in situ
detection of apoptosis and necrosis in monolayer cultures by TUNEL and trypan
blue staining. BioTechniques 22:1102.