Nucleic acid binding agents exert local toxic effects on neurites via a non-nuclear mechanism.

Abstract

The mechanism by which drugs that target nucleic acids cause neurotoxicity is not well described. We characterized the neurotoxicity of Hoechst 33342 (bis-benzimide), a common cell permeable nuclear dye, in primary neuronal cultures. The mechanism of cell death was not apoptotic, as death is rapid, not accompanied by typical nuclear morphological changes, and is insensitive to inhibitors of transcription, translation and caspase activity. In addition, free-radical scavenging agents failed to attenuate cell death, and damage was not accompanied by mitochondrial dysfunction. Neuronal processes of cells exposed to Hoechst 33342 display dramatic fragmentation prior to cell death. When this compound was applied selectively to the distal axons of sympathetic neurons grown in compartmented cultures, the distal axons were destroyed. However, the proximal processes present in the cell body compartment were spared, demonstrating direct axonal toxicity rather than a remote effect of nuclear dysfunction. Other cell-permeable nucleic acid binding dyes similarly caused rapid dendritic and axonal toxicity. The hypothesis that these nucleic acid binding dyes target RNA localized to dendrites and axons is supported by observations that RNaseV1 induced similar, rapid neurite fragmentation. We conclude that the neurotoxic effects of nucleic acid binding compounds are mediated, at least in part, by direct neurite injury, which does not require involvement of the cell body and nucleus.

Full text

Page 1

Nucleic acid binding agents exert local toxic effects on neurites via
a non-nuclear mechanism
Sokhon Pin,* Huiling Chen,? Pamela J. Lein? and Michael M. Wang§
*Department of Anesthesiology/Critical Care Medicine, Johns Hopkins University, Baltimore, Maryland, USA
?Children’s Research Institute, Washington DC, USA
?Center for Research on Occupational and Environmental Toxicology, Oregon Health and Science University, Portland, Oregon, USA
§Departments of Neurology and Molecular and Integrative Physiology, University of Michigan, Ann Arbor, Michigan, USA
Abstract
The mechanism by which drugs that target nucleic acids
cause neurotoxicity is not well described. We characterized
the neurotoxicity of Hoechst 33342 (bis-benzimide), a com-
mon cell permeable nuclear dye, in primary neuronal cultures.
The mechanism of cell death was not apoptotic, as death is
rapid, not accompanied by typical nuclear morphological
changes, and is insensitive to inhibitors of transcription,
translation and caspase activity. In addition, free-radical
scavenging agents failed to attenuate cell death, and damage
was not accompanied by mitochondrial dysfunction. Neuronal
processes of cells exposed to Hoechst 33342 display dra-
matic fragmentation prior to cell death. When this compound
was applied selectively to the distal axons of sympathetic
neurons grown in compartmented cultures, the distal axons
were destroyed. However, the proximal processes present in
the cell body compartment were spared, demonstrating direct
axonal toxicity rather than a remote effect of nuclear dys-
function. Other cell-permeable nucleic acid binding dyes
similarly caused rapid dendritic and axonal toxicity. The
hypothesis that these nucleic acid binding dyes target RNA
localized to dendrites and axons is supported by observations
that RNaseV1 induced similar, rapid neurite fragmentation.
We conclude that the neurotoxic effects of nucleic acid binding
compounds are mediated, at least in part, by direct neurite
injury, which does not require involvement of the cell body and
nucleus.
Keywords: cortical neurons, Hoechst, neurite, neurotoxicity,
nucleic acid dyes, RNase.
J. Neurochem. (2006) 96, 1253–1266.
A large variety of pharmaceuticals have neurological side-
effects and cause neuronal damage. Yet, in most cases, the
precise mechanisms of how these diverse compounds cause
neuronal injury remain elusive. Compounds which interact
with nucleic acids or nucleic acid metabolism frequently
affect the nervous system. For example, the chemotherapeu-
tic agent cytosine arabinoside causes confusion, seizures, and
cerebellar dysfunction (Resar et al. 1993). Several drugs that
are used for cancer are associated with peripheral neuro-
pathies, including chorambucil (Druker et al. 1989), isofosf-
amide (Patel et al. 1994; Frisk et al. 2001), cis-platin (Ozols
and Young 1985), and etoposide (Imrie et al. 1994). Finally,
dideoxy nucleoside analogues used for the treatment of HIV
commonly result in peripheral neuropathy (Dubinsky et al.
1989; Dalakas 2001).
These drugs are useful clinically because they target DNA
synthesis, which is crucial for the growth of tumor cells and
viral replication. Cytosine arabinoside is a nucleotide
analogue which inhibits DNA synthesis (Slapak et al.
1985). Chlorambucil and isofosfamide are alkylating agents
(Kundu et al. 1994; Povirk and Shuker 1994; Fleming 1997)
which are capable of modifying DNA and inhibiting DNA
replication and transcription. Cis-platin directly binds to
DNA and is thought to target DNA synthesis (Wang and
Lippard 2005). Etoposide inhibits DNA topoisomerase II,
which is crucial for DNA synthesis and repair (Meresse et al.
2004). Finally, HIV dideoxynucleosides impair reverse
transcriptase, and therefore inhibit replication of the virus
Received June 6, 2005; revised manuscript received October 3, 2005;
accepted October 19, 2005.
Address correspondence and reprint requests to Michael Wang,
7629 Medical Sci II Box 0622, 1137 Catherine St., Ann Arbor, MI
48109–0622, USA. E-mail: micwang@umich.edu
Abbreviations used: GFP, green fluorescent protein; PI, propidium
iodide; TMRM, tetramethylrhodamine methylester.
Journal of Neurochemistry, 2006, 96, 1253–1266 doi:10.1111/j.1471-4159.2006.03653.x
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
1253

Page 2

(De Clercq 2002). It is widely accepted that these agents
impair normal stability and/or replication of DNA, resulting
in cell cycle disruption, apoptosis or inhibition of viral
replication. However, in post-mitotic neurons, the mechan-
ism of action of these agents in causing cellular damage is
unclear, as these cells do not depend on DNA replication for
survival. Several toxic mechanisms of action of have been
proposed, including direct inhibition of DNA stability
resulting in programmed cell death via multiple pathways
(Park et al. 1998; Stefanis et al. 1999). Yet other studies
have suggested neurotoxic compounds influence novel, non-
nuclear biochemical pathways (Wallace and Johnson 1989;
Martin et al. 1990). Nucleoside analogues have been shown
to inhibit mitochondrial DNA synthesis (Cui et al. 1997)
which may result in axonal energy failure and neurite
dysfunction. In addition, recent studies of DRG neurons
exposed to dideoxy drugs demonstrate neurite damage as a
result of mitochondrial dysfunction on a timescale much
faster than could be accounted for by mitochondrial DNA
damage (Keswani et al. 2003), implicating non-nucleic acid
mitochondrial toxicity. Overall, given the chemical diversity
of these compounds and the multifactorial susceptibility of
neurons, it is expected that the neurological side-effects of
the many drugs in use are a result of multiple, diverse
mechanisms.
This study demonstrates an additional mechanism by
which nucleic acid binding compounds induce neuronal
injury. We report the neurotoxic properties of the nucleic acid
binding agent Hoechst 33342, a cell-permeable dye used to
label live cell nuclei. Curiously, in contrast to the slower
actions of genotoxic compounds in non-neuronal cells,
Hoechst 33342 induces extremely rapid morphological
changes in neurites marked by fragmentation and blebbing
of axons and dendrites. We find that Hoechst 33342 mediates
toxicity through non-nuclear actions and further demonstrate
that a variety of other nucleic binding agents mediate similar
rapid neuronal damage. The morphological changes seen
with nucleic acid binding compounds is mimicked by
treatment with RNaseV1, suggesting that RNA may be an
essential component of neurite homeostasis and that neuronal
damage through nucleic acid binding drugs could be
mediated by actions on neurite RNA.
Materials and methods
Cell culture
All animals were handled in accordance with established national
and institutional guidelines under the auspices of the animal care
and use committee. HEK293A cells (Qbiogene, Irvine, CA, USA)
were grown as described in Dulbecco’s modified Eagle’s medium
containing 10% fetal bovine serum (Xu et al. 2003). Mouse cortical
cultures were prepared from E18 embryos and plated on poly
D-lysine-treated dishes in neurobasal/B27 media using established
protocols (Brewer 1995; Xu et al. 2003). Immunohistochemical
analysis has demonstrated that our cultures are more than 99%
neuronal. Experiments shown were performed on immature neurons
(days in vitro 3–5) prior to the onset of glutamate sensitivity. Cells
were transfected with pEGFP plasmids from Clontech (San Jose,
CA, USA) using Lipofectamine 2000 (Xu et al. 2003). For viability
studies, cells were usually pretreated with DNaseI (10–20 lg/mL for
2–6 h followed by media replacement) before addition of drugs to
decrease background counting of dead cells in the culture. Live–
dead assays were performed using calcein AM and propidium
iodide (PI) (Molecular Probes, Eugene, OR, USA); live cells
metabolize calcein AM and fluoresce green, while dead cells
fluoresce red because of PI access to the nucleus. Cells were
imaged and photographed using an inverted Nikon TE200 equipped
with a Spot RT digital camera. For time-course studies, cells were
grown in HEPES-buffered media in an enclosed chamber on the
microscope stage which was heated to 37?C. At least four different
high powered (40 ·) fields were photographed, and then manually
counted. Morphological changes in axons and dendrites (collec-
tively referred to as neurites) were scored in a blinded fashion. A
neuron was scored as exhibiting significant neuritic damage if more
than one axonal or dendritic process had at least three regions of
distinct blebbing or fragmentation along its length. In general, if a
neuron had one damaged neurite, almost all of the processes,
including dendrites and axons, were damaged. Nuclear morphology
was also analyzed blinded using established criteria for apoptosis
(Stefanis et al. 1999). Data from all groups represent at least three
wells of cultures and each experiment was performed three times to
validate consistency between cultures.
Culture of sympathetic neurons in compartmented Campenot
chambers
Compartmented Campenot chambers were set up as described by
Senger and Campenot (1997). Briefly, 35-mm dishes were pre-
coated with a layer of ammoniated rat tail collagen (0.75 mg/mL),
followed by a layer of air-dried rat tail collagen supplemented
with laminin (10 lg/mL). Three-compartmented Teflon dividers
(Camp10, Tyler Research Instruments, Edmonton, AB, Canada)
were seated on top of parallel tracks scratched in the collagen
substrate with a pin rake (Tyler Research Instruments) and secured
in place with silicone vacuum grease (Dow Corning, Huntington
Beach, CA, USA). Integrity of the grease seals was assessed by
placing culture medium into the side chambers only and incubating
the chambers overnight in a 37?C incubator. Only chambers that did
not leak were used in subsequent experiments. Medium was
removed from side compartments and dissociated sympathetic
neurons were plated in the center compartment in serum-free
medium (Dulbecco’s modified Eagle’s medium/F12 with N2
supplement) containing nerve growth factor (100 ng/mL, Harlan
Bioproducts, Madison, WI, USA). Sympathetic neurons were
dissociated from the superior cervical ganglia of embryonic
day 21 or post-natal day 1–2 Holtzmann rats (Harlan Sprague–
Dawley, Rockford, IL, USA) according to previously described
methods (Lein and Higgins 1989). The next day the integrity of the
seals between the compartments was reconfirmed by checking for
leakage from the compartment containing the cell suspension into
the adjacent compartments. Culture medium then was added to the
compartments not containing cells of those culture dishes exhibiting
1254
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 3

intact seals between compartments, and cytosine-arabinoside (1 lM)
was added to the medium in the compartment containing the cell
bodies for 48 h to eliminate non-neuronal cells. After axons had
extended through the grease seals into the adjacent compartments
(5–7 d after plating), nerve growth factor was withdrawn from
the compartment containing cell bodies but was maintained in the
adjacent compartments to encourage growth of axons into the
adjacent compartments and to eliminate neurons that had failed to
extend axons through the grease seals. Experiments were initiated
when there was significant neurite outgrowth in the adjacent
compartments (10–14 d after plating).
Reagents
All drugs and chemicals were obtained from Sigma (St Louis, MO,
USA), unless noted. Cell culture media were from Invitrogen
(Carlsbad, CA, USA). Fluorescent probes were obtained from
Molecular Probes. Biochemical grade RNaseV1 and purified tRNA
were purchased from Ambion (Austin, TX, USA).
Statistical analyses
Data is presented from representative experiments, which were
repeated at least three times. Standard error bars are shown which
represent cell counts from three wells of cells treated in identical
fashion.
Results
Cell death induced by nucleic acid binding dyes
Hoechst 33342 is a commonly used nucleic acid binding dye,
which is highly cell permeable. Because of these properties, it
has been used to label nuclei of live cells for cell sorting,
tracking, and for analysis of nuclear morphology. Although
Hoechst 33342 has been shown to induce apoptosis in HL-60
and BC3H-1 cells (Zhang and Kiechle 1997; Zhang et al.
1999),itseffectonprimaryneuronshasnotbeencharacterized.
In initial experiments, we found that Hoechst 33342 causes
rapid concentration-dependent toxicity in cultured murine
cortical neurons (Figs 1a and c). Neurons were significantly
moresensitivetothetoxiceffectsofHoechst 33342relativeto
non-neuronal cells (Figs 1b and d).
Non-apoptotic mechanism of Hoechst 33342 cell death
Several lines of evidence suggest that Hoechst 33342 may
cause neuronal cell death via apoptosis. First, Hoechst has
been shown to cause apoptosis in HL-60, BC3H-1 myocytes,
H4-II-E-C3 rat hepatoma cells, HT-114 melanoma cells, and
human skin fibroblasts (Zhang and Kiechle 1997; Zhang
et al. 1999; Kiechle 2002). Second, Hoechst 33342 not only
binds to minor grooves in DNA, but also inhibits topoisom-
erase I, and other agents that inhibit this enzyme, including
camptothecin, cause neuronal apoptosis (Morris and Geller
1996). To determine if Hoechst 33342 similarly induces
apoptosis in cultured cortical neurons, we compared the
effects of Hoechst 33342 and camptothecin on nuclear
morphology.
Nuclei of cortical neurons treated with camptothecin for
24 h appeared shrunken, blebbed and fragmented (Fig. 2a),
which is characteristic of cells undergoing apoptosis
(Stefanis 1999). In contrast, nuclei of cortical neurons treated
with Hoechst 33342 for 24 h appeared round and intact
(Fig. 2b) and did not exhibit the morphological characteris-
tics of nuclei undergoing programmed cell death. Actino-
mycin D and cycloheximide, inhibitors of RNA and protein
synthesis, respectively, failed to attenuate Hoechst 33342-
induced cell death (Figs 2c and d). In addition, DNA isolated
from cells treated with Hoechst 33342 did not demonstrate
fragmentation, as seen with camptothecin treatment (Fig. 2e),
although there was a slight apparent change in mobility in
Hoechst-treated DNA, which could have been the result
of infrequent nicking of DNA or of a change in mobility
induced by dye intercalation. In total, we found no evidence
that Hoechst 33342 acts by programmed cell death as
determined by a variety of established markers of apoptosis.
Morphological effects of nuclei acid binding dyes in
neurites
To investigate the morphological effects of Hoechst 33342
on cortical neurons, we transfected neurons with a green
fluorescent protein (GFP) expression plasmid prior to adding
the drug. This enabled us to visualize the effects of
Hoechst 33342 over time on individual cells. Nearly all
neurons displayed striking morphological changes marked by
progressive blebbing and fragmentation of the neurite
(Fig. 3a). Hoechst 33342 caused similar morphological
changes in neurons grown in low-density cultures that were
loaded with the fluorescent vital dye calcein AM (not
shown). Hoechst 33342 does not induce these rapid mor-
phological changes by inhibition of topoisomerase I, as
camptothecin failed to induce similar morphologic changes
(Fig. 3b). The response was rapid, as all cells displayed
abnormal morphology after 30 min of 30 lM Hoechst 33342
(Fig. 3c).
Next, we examined whether other cell permeable nucleic
acid binding compounds exhibited similar properties as
Hoechst 33342. Indeed, a large number of dyes were able to
induce visible neurite morphological changes in a rapid time
frame at micromolar concentrations (Fig. 4). Of note, non-
permeant dyes such as propidium iodide failed to elicit rapid
neurite damage (not shown).
Hoechst 33342 death mechanism is independent of free
radical formation and mitochondrial stress
Neurons are extremely sensitive to free radical-induced cell
death. We therefore tested whether free radical scavengers
could attenuate cell death induced by Hoechst 33342. Free
radical-mediated mechanisms appear to be unlikely as
scavengers such as phenyl-tert-butylnitrone (Gould and
Bickford 1994; Sturgess et al. 2000; Fig. 5) failed to
attenuate Hoechst 33342 toxicity.
Local toxic effects of nucleic acid binding agents on neurites
1255
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 4

Another important determinant of neurite integrity is
mitochondrial stability. We thus assessed whether the
morphological changes seen in response to Hoechst 33342
could be linked to mitochondrial toxicity (Fig. 6). To
determine whether mitochondrial stress could play a role in
Hoechst 33342 neurite toxicity, we loaded cells with tetra-
methylrhodamine methylester (TMRM) (Nieminen et al.
1995; Ward et al. 2000) to determine whether loss of
mitochondrial membrane potentials preceded neurite frag-
mentation. It has been previously reported that acute
exposures to Hoechst 33342 cause a transient mitochondrial
depolarization in HL-60 cells (Chen et al. 2004). Similarly,
we observed a 50% reduction in TMRM fluorescence in
cortical neurons 15 min after adding H33342 at 60 lM, but
TMRM fluorescence returned to control levels within 30 min
(unpublished data). However, mitochondrial potentials were
unaffected at early time points when exposed to Hoe-
chst 33342 at lower concentrations (15 lM) that cause
neurite fragmentation. In fact, the mitochondrial signal
persisted long after initial signs of neurite fragmentation
and blebbing, suggesting that mitochondrial injury does not
play a significant role in neurite instability. Therefore,
progressive changes to the mitochondria, at least at this
level, do not appear to correlate with neurite damage.
Analysis of the TMRM fluorescence in neurites was
compared with cell body fluorescence to discern potential
differences in neurite versus cell body mitochondria (not
shown). This analysis failed to demonstrate evidence of
neurite mitochondrial dysfunction prior to appearance of
morphological changes.
Role of excitotoxicity in Hoechst 33342-induced neurite
changes
Neurites can be damaged by excitotoxic injury and display
characteristic changes that are remarkably similar to those
induced by Hoechst 33342, with fragmentation, swelling
293 cells
Neurons
(a)
(b)
(c)
(d)
Fig. 1 Hoechst 33342 causes rapid, concentration-dependent neuro-
toxic effects in primary cultures of cortical neurons. (a) Untreated cor-
tical cultures show intense green fluorescence of calcein AM viability
dye and a modest uptake of propidium iodide (red fluorescence), which
marks dead cells. These cells were not pretreated with DNaseI. (b)
After 24 h of treatment with 15 lM Hoechst 33342, all cells are dead.
(c) Quantitative analysis of the ratio of viable to non-viable cells in
cortical cultures demonstrates concentration dependent toxicity of
Hoechst 33342. (d) 293A cell lines are much less sensitive to Hoechst
33342. Cells used in quantitative analysis were pretreated with DNaseI
(before Hoechst treatment) to reduce the background of cells which
had already died before addition of Hoechst 33342. Data are repre-
sented as the mean ± SEM from three identically treated wells.
Experiments were repeated with similar outcomes at least three times.
1256
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 5

and blebbing seen within minutes of NMDA application
(Park et al. 1996). Although the neurons used in this study
were immature and resistant to glutamate toxicity, we
performed experiments to determine if blocking excitotoxic
pathways was sufficient to attenuate Hoechst 33342 toxic-
ity(Fig. 7). Excitotoxicitydepends on activation of
Fig. 2 Non-apoptotic mechanism of neuronal cell death in cortical
neurons treated with Hoechst 33342. (a) Nuclear morphology of cor-
tical neurons treated with camptothecin (800 nM for 24 h) then stained
with Hoechst 33342 for 30 min to visualize chromatin in nuclei. Note
the appearance of multiple fragmented and bright nuclei with evidence
of nuclear condensation (see arrows). (b) Cells treated with 15 lM
Hoechst 33342 overnight display round nuclei with no evidence of
fragmentation and condensation. A series of cultures were treated with
cycloheximide (c) or actinomycin D (d) at indicated concentrations
overnight. Control cells (white bar on the left) and Hoechst 33342
treated cells (four bars on the right) were scored for viability after 24 h
using calcein AM and PI fluorescence. Neither cycloheximide nor
actinomycin D blocked the ability of Hoechst 33342 to induce cell
death. (e) Cortical neurons were treated with medium alone (lane 1),
medium supplemented with Hoechst 33342 for 20 min (lane 2), 2 h
(lane 3), or 24 h (lane 4), or medium supplemented with camptothecin
(800 nM) for 2 h (lane 5) or 24 h (lane 6). Genomic DNA was analyzed
on an agarose gel to assess DNA fragmentation, which is an indicator
of programmed cell death. (c, d) Data are represented as the
mean ± SEM (three wells per group).
Local toxic effects of nucleic acid binding agents on neurites
1257
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 6

Fig. 3 Temporal analyses of neurite fragmentation in response to
Hoechst 33342. Sequential images of GFP-transfected neurons (in
min) after treatment with 15 lM Hoechst 33342 (a) or camptothecin
(800 nM) (b). (c) Percentage of GFP-transfected neurons with neurite
fragmentation after different times of exposure to 15 lM Hoe-
chst 33342.
1258
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 7

voltage-sensitive sodium channels and calcium influx
through both voltage-sensitive calcium changes and glu-
tamate receptors (Hasbani et al. 1998). As shown in
Fig. 7, blocking voltage-gated sodium channels with TTX
(1 lM), did not alter the toxic effects of Hoechst 33342 on
neurite morphology. Similarly, sequestration of extracellu-
lar calcium by EDTA, or blocking of calcium influx
through glutamate receptors by the glutamate receptor
antagonists MK801 and NBQX failed to block the
morphological changes seen with Hoechst 33342.
Toxicity of Hoechst 33342 on isolated axons
The rapid toxicity to neurites could involve direct action on
the neurite, and/or an indirect action on the cell body, which
then signals to the axonal and dendritic processes. To
distinguish between these mechanisms, we examined the
effect of Hoechst 33342 on sympathetic neurons grown in
compartmented cultures (Senger and Campenot 1997).
Addition of Hoechst 33342 to one of the two compartments
containing distal axons resulted in progressive degeneration
of axonal morphology within that compartment only (Figs 8a
and b). Axonal fragmentation was not observed in the
compartment containing cell bodies and proximal processes,
nor was it observed in the second compartment containing
distal axons that were not exposed directly to Hoe-
chst 33342. Similarly, degeneration of processes was also
seen when the drug was added to the chamber containing cell
bodies and proximal axons; however, distal axons not in
direct contact with drugs maintained normal morphology
(Fig. 8c). These results demonstrated that Hoechst 33342
causes local, direct toxic effects on axons in vitro, and that
this toxicity is not dependent on interactions with nuclear
DNA.
Fig. 4 Toxicity of cell-permeable nucleic acid dyes. (a) Exposure of
GFP-transfected cortical neurons to a selection of nucleic acid binding
dyes resulted in rapid morphological changes in neurites (1 h); dyes
were used at 15 lM. (b, c) Quantitative analysis of neurite toxicity of
cortical neurons treated with a set of cell permeable nucleic acid dyes;
criteria for neurite abnormality are described in the Materials and
methods.
Fig. 5 Hoechst 33342 causes neurotoxicity via a free radical-inde-
pendent mechanism. Cortical cultures were treated overnight with the
free radical scavenger phenyl-tert-butylnitrone at varying concentra-
tions. Control cells (five bars on the left) and Hoechst 33342 (five bars
on the right) -treated cells were scored for viability using calcein AM
and PI fluorescence after 24 h.
Local toxic effects of nucleic acid binding agents on neurites
1259
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 8

Toxicity of RNaseV1 on neurites
RNA is present in the neurites of neurons (Knowles et al.
1996; Job and Eberwine 2001). RNA frequently adopts
secondary structure involving inter- and intramolecular base
pairing, which confers it with common chemical properties
with DNA. Consequently, RNA and DNA share affinity for
most nucleic acid binding drugs, including Hoechst 33342
dyes (Wilson et al. 1993; Dassonneville et al. 1997) and the
SYTO compounds (Molecular Probes) (Knowles et al.
1996). Therefore, the neurotoxicity of nucleic acid binding
compounds could potentially be mediated by interactions
with RNA, rather than DNA. As most nucleic acid probes do
not distinguish between RNA and DNA, we assessed the
effect of RNaseV, an endonuclease that specifically targets
double-stranded RNA (Lockard and Kumar 1981; Lowman
and Draper 1986), on neurite morphology. As demonstrated
in Fig. 9(a and b), incubation of this enzyme with cortical
neurons causes rapid morphological changes in neurites that
are comparable with those observed with Hoechst 33342
over the same time period (results with RNaseV1 are
quantified in Fig. 9e). The enzymatic action of RNase can be
inhibited by heparin or tRNA (Mahalakshmi et al. 2000),
and pre-incubation of cortical cultures with either heparin or
tRNA (Mahalakshmi et al. 2000) completely blocked the
RNaseV1-induced changes in neurite morphology, suggest-
ing that the cytotoxic actions of RNase were mediated by its
RNA degradation function (Figs 9c and d; results are
quantified in Fig. 9f). In addition, heat treatment of
RNaseV1 strongly inhibited its toxic effects in cortical
neurons (Fig. 9g), suggesting that an enzymatic component
rather than a small chemical contaminant from the prepara-
tions, mediated RNaseV1 neurotoxicity.
We next attempted to test whether RNaseV1 was
crossing the cell membrane to disrupt RNA. When neurons
were treated with RNaseV1, there was an immediate
decrease in the amount of total RNA extracted from
treated cells at 30 min (Fig. 9h). There was no further
decrease in the amount of RNA for 4 h. However,
Fig. 6 Toxicity of Hoechst 33342 is not caused by persistent mito-
chondrial failure. Neurons were exposed to 15 lM Hoechst 33342 for
varying periods of time (top row). Thirty minutes before each time point,
cells were labeled with the mitotracker dye TMRM (200 nM for 30 prior
to end of time point) and photographed to assess the functional status
of mitochondria (maintenance of normal mitochondrial electrical
potential is indicated by the fluorescence of TMRM). Photographs were
obtained with constant exposure settings. To prevent possible photo-
bleaching of TMRM or light-induced damage to cultures, all cells were
processed in the dark and exposure to excitation light was limited to the
time required to focus and obtain the photograph. In parallel experi-
ments using sister cultures; representative GFP-transfected neurons
were photographed to document potency of Hoechst 33342 in neurite
fragmentation as a function of time (bottom row).
Fig. 7 Hoechst 33342 does not cause neurite injury via excitotoxic
mechanisms. GFP-transfected neurons were scored for neurite frag-
mentation. The first bar (white) shows the baseline fragmentation in
the absence of Hoechst 33342. The remaining bars (black) show the
effect of 1 h of 15 lM Hoechst 33342 treatment. Bars 3–6 show
neurite fragmentation by Hoechst 33342 in neurons pre-incubated
(60 min) with drugs that block glutamate receptors (MK801 20 lM,
NBQX 30 lM), calcium influx (EDTA 1 mM), or voltage-gated sodium
channels (TTX, 10 lM). Data are represented as the mean ± SEM
(n ¼ 3 per experimental condition).
1260
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 9

membrane permeability to propidium iodide was seen in
parallel cultures only at 4 h and beyond (Fig. 9i). These
data suggest that RNaseV1 is capable of crossing the cell
membrane and digesting intracellular RNA, and that this
precedes the onset of cell death.
Discussion
Our studies demonstrate unique neurotoxic properties of
nucleic acid binding dyes. Although nucleic acid binding
dyes have been developed specifically for their nuclear
Fig. 8 Neurite toxicity of Hoechst 33342 is independent of nuclear
action. Sympathetic neurons labeled with calcein AM were grown in
compartmented chambers (left schematic) which allows physical
separation of the medium bathing cell bodies and proximal processes
(center compartment) from that bathing distal axons (left and right
compartments). In this experiment, Hoechst 33342 (30 lM) was added
only to the right chamber. (a) Adjacent microscopic fields demonstrate
the toxicity of Hoechst 33342 to axons exposed to Hoechst 33342 for
1 h. Note that the cell bodies are intact. (b) Adjacent microscopic fields
demonstrating that control axons and cell bodies in the left and center
compartment, which were not exposed to Hoechst 33342, are intact.
(c) Addition of Hoechst 33342 to the center well induced rapid mor-
phological changes in proximal axons within the center chambers.
However, the distal axons of the same cells (left and right compart-
ments) remained intact.
Local toxic effects of nucleic acid binding agents on neurites
1261
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 10

0 30 60 120
time (min)
Viability after RnaseV1 treatment
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 50100150 200250 300
(i)
Time (min)
% Viable
(a)
(b)
(d)
(c)
(e)
(f)
(g)
(h)
Fig. 9 RNaseV1 causes rapid neurite fragmentation. GFP-transfected
cortical neurons were incubated with RNaseV1 (4 units/mL) for 1 h, in
the absence or presence of inhibitors of the enzyme. (a) Control
neurons exhibit minimal neurite breakdown. (b) Treatment with
RNaseV1 causes rapid neurite blebbing and fragmentation. In con-
trast, incubation of cells with the RNase inhibitors (c) heparin (100 lg/
mL) or (d) tRNA (100 lg/mL) attenuated the neurotoxic effects of
RNaseV1-induced toxicity. (e) Quantitation of neurite damage induced
by RNaseV1 at 3 and 24 h demonstrated uniform cell injury at 1 day.
(f) Quantitative analysis of the ability of heparin and tRNA to inhibit the
neurite toxicity of RNaseV1. (g) Heat inactivation of RNaseV1 reduces
the neurite toxicity of the enzyme. (h) RNaseV1 treatment (5 units/mL)
reduced the amount of rRNA extracted from neurons. (i) This reduction
in rRNA preceded the onset of cell death, marked by propidium iodide
nuclear fluorescence. Cells were pretreated with DNaseI (as in Fig. 1)
and live cells remaining were counted using calcein AM fluorescence.
1262
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 11

properties, the neurotoxic properties of these compounds do
not appear to involve apoptosis and surprisingly cause direct
local neurotoxic effects in neurites via mechanisms that do
not involve the cell body or nucleus. Similar neurite damage
was observed for a large family of cell permeable nucleic
acid binding agents, suggesting a common molecular
pathway.
Park et al. (1996) have previously demonstrated the acute
morphological changes induced by glutamate in mature
cortical cultures. Glutamate-induced changes in neurite
morphology are strikingly similar to those we observe in
immature cultures exposed to nucleic acid binding dyes. In
contrast to these studies, we used immature cortical cultures
in which excitotoxicity is not a significant mechanism of
death. Therefore, it is not surprising that the mechanism by
which these neurons die in response to Hoechst 33342
appears not to involve glutamate toxicity, voltage-sensitive
sodium channels, or extracellular calcium, suggesting a
neurotoxic mechanism distinct from glutamate-induced tox-
icity that results in a grossly similar phenotype. Also, in
contrast to studies with glutamate, Hoechst 33342 toxicity
involves both axons and dendrites, whereas excitotoxin-
induced neuritic damage has been described only in dendrites
(Park et al. 1996). We have shown similar toxic effects of
several of these nucleic acid binding compounds in mature,
glutamate-sensitivecortical
(unpublished data). Combined with our findings that periph-
eral neurons also demonstrate sensitivity to these agents, we
conclude that our results appear to be generalized to multiple
neuronal types.
What could account for the exquisite sensitivity of neurites
to damage by nucleic acid binding compounds? Although
non-nucleic acid-dependent mechanisms are possible, the
common dramatic effect of numerous structurally diverse
nucleic acid binding agents suggests that these compounds
act on neurons through nucleic acid binding. We therefore
consider two classes of mechanisms: those involving DNA
versus those targeting RNA.
and hippocampal cultures
Nuclear DNA
The most direct evidence arguing against the hypothesis that
nucleic acid binding drugs cause neurotoxicity via interac-
tions with nuclear DNA are the experiments using sympa-
thetic neurons grown in compartmented Campenot chambers.
Rapid axonal damage was seen only with direct application
of nucleic acid binding compounds and did not require
application to the nucleus. While there are great similarities
in the temporal, quantitative and qualitative nature of the
neurite damage observed in response to Hoechst 33342
between sympathetic and cortical neurons in our study, one
cannot rule out the possibility that nuclear DNA is involved
in Hoechst 33342-mediated cortical neuronal injury. How-
ever, this seems unlikely as neurite blebbing and fragmen-
tation is obvious within 10 min of Hoechst application. This
amount of time seems insufficient to allow the sequence of
events that would have to occur if DNA were the primary
target, e.g. diffusion of the dye through the plasma
membrane and nuclear membrane into the nucleus, signaling
within the nucleus, and signal transduction from the nucleus
to the neurite several hundred microns away. Further
evidence arguing against the involvement of nuclear DNA
in Hoechst 33342 neurotoxicity comes from experiments
(not shown) demonstrating that staining of the nucleus with
these nuclear acid binding dyes is not detectable until at least
10 min after dye application.
Mitochondrial DNA
Another possible target of nucleic acid binding dyes is
mitochondrial DNA. Mitochondria play essential roles in
producing the energy required for normal neurite home-
ostasis. It seems logical that inhibition of mitochondrial DNA
could impair energy metabolism, leading to rapid neurite
dysfunction. However, the rapidity of Hoechst 33342-
induced changes in neurite morphology is inconsistent with
energy failure because of mitochondrial genomic inhibition,
which would require rapid turnover of mitochondrial gene
products. In addition, our time-course studies indicate that,
although acute exposure to high concentrations of Hoechst
(60 lM) causes a transient decrease in mitochondrial poten-
tial, there is no long-term progressive disruption which
parallels neurite instability. Furthermore, acute exposure to
lower Hoechst concentrations that cause neurite fragmenta-
tion (15 lM) do not decrease mitochondrial potential. This
strongly suggests that mitochondria are functional during the
time that nucleic acid binding compounds are disrupting
neurites, although we cannot rule out the possibility that a
transient depolarization of mitochondria plays a destabilizing
role.
RNA
We present evidence that RNA may be an important target of
nucleic acid binding compounds. RNA within neurites has
been described recently, and is thought to be transported
long distances based on cis-acting sequences encoded mostly
in the 3¢ untranslated region. Local translation of this RNA
has been demonstrated in dendrites (reviewed Steward 2003)
and in axons (reviewed Piper 2004), and plays a role in
growth cone collapse (Wu 2005). Ribosomal RNA is also
localized to neurites and is undoubtedly required for
translation of this RNA. While some of this RNA appears
to be localized in foci that are traceable with dyes such as
SYTO14 (Knowles et al. 1996), a large number of studies
have also demonstrated broadly distributed localization
depending on the RNA species (Mohr 2000; Wu 2005).
Therefore, current information suggests that as a whole,
RNA is diffusely localized throughout the neurite. This
matches well with the diffuse distribution of neurite
breakpoints we describe.
Local toxic effects of nucleic acid binding agents on neurites
1263
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 12

Most nucleic acid binding compounds are able to bind to
RNA as well as DNA, some with affinities greater for RNA
(Wilson et al. 1993; Knowles et al. 1996; Dassonneville
et al. 1997). Many of these agents selectively bind to double-
stranded RNA with significantly higher affinity compared
with single-stranded RNA. Experiments demonstrating that
RNaseV1 causes neurotoxic effects comparable with those
induced by Hoechst 33342 suggest that dsRNA is the
principal target in Hoechst 33342 neurotoxicity and indicate
that dsRNA may play an important role in neurite home-
ostasis. Two observations suggest that the neurotoxic effects
of RNaseV1 are mediated by its RNA degrading function: (i)
known inhibitors of RNaseV1 attenuate its neurotoxic effects
on cortical neurons; and (ii) RNaseV1 neurotoxicity is
sensitive to heat. We cannot yet exclude the possibility that
minor contaminants within the RNaseV1 preparation used in
these studies that are also sensititve to heparin, tRNA, and
heat inactivation may be responsible for the enzyme’s
neurotoxic activity. Phospholipase A2 from cobra venom
can contaminate preparation of RNaseV1; however, heparin
increases the activity of phospholipases (Patel et al. 1997)
and phospholipase A2 is not known to be affected by tRNA.
Heparin is known to inhibit complement-activating toxins
from cobra venom (Edens et al. 1994); however, our
experiments were performed in serum-free media, which
lack complement. We postulate that the RNaseV1 protein is
cell permeable or that the RNA target that mediates neurite
morphology is extracellular. The former possibility is more
likely, and not necessarily surprising as many proteins,
particularly in venom extracts (the source of the RNaseV1
used in our studies) are capable of crossing the plasma
membrane. This is supported in experiments showing that
RNaseV1 diminishes ribosomal RNA content within cells in
a time course that precedes membrane permeability to
propidium iodide, a molecule much smaller than RNaseV1
which is commonly used as a marker of cell death.
Conclusive demonstration that RNaseV1 is able to cross
the cytoplasmic membrane will require development of
specific antisera which are not yet available.
The selectivity of RNaseV1 for double-stranded RNA
in vitro suggests that the putative RNA target in neurons
has a significant secondary structure. The possible known
RNA targets include tRNA, rRNA, mRNA, ribonuclear
proteins and microRNAs. Although microRNAs are single
stranded, processing of microRNA precursors requires
double-stranded RNA (Hutvagner and Zamore 2002).
Targeting tRNA or rRNA would be predicted to impair
protein synthesis within neurites. Because transient inhibi-
tion of protein synthesis by cycloheximide fails to cause
neurite fragmentation, it is more likely that mRNA or
structural RNAs that are not integrally involved in protein
synthesis are targeted by RNaseV1 and the nucleic acid
binding dyes. The rapidity of morphological changes seen
in our experiments also strongly suggests that RNA may
play a structural role within the neurite, as opposed to a
role in protein synthesis. Recent data has demonstrated that
spreading initiation centers, novel cytoskeletal structures
present early in cell spreading, are highly enriched in RNA
binding proteins and ribosomal RNA (de Hoog et al.
2004). Taken together with our experiments, these studies
suggest a novel role for RNA in cytoskeletal maintenance
and dynamic modulation.
We note that this mechanism does not apply to all
classes of drugs that interfere with nucleic acid functions
(e.g. topoisomerase inhibitors or dideoxy nucleoside ana-
logues). However, considering the wide variety of potential
RNA structures that could exist in neurites, it is possible
that even drugs without a known affinity for nucleic acids,
could act on neurons through specific RNA binding. For
example,AZT,a specific
transcriptase which causes neuropathy has been shown to
bind to RNA (Ahmed Ouameur et al. 2004). In further
support that RNA could be a general target of neurotoxic
compounds, we note that several drugs, including strepto-
mycin (Wallace andSchroeder
(Thom and Prescott 1997), and tobramycin (Hermann
and Westhof 1999), have been shown to bind to specific
RNA sequences with high affinity.
In conclusion, nucleic acid binding agent-induced neuro-
toxicity through non-nuclear paths provides an alternative
mechanism for neurite damage which is independent of
excitotoxicity and free radicals and is possibly mediated by
interactions with RNA. The precise mechanism of damage to
neurites may have important implications regarding the cause
and treatment of neuropathies associated with nucleic acid
binding drugs. Finally, from a practical point of view, these
studies indicate that cell permeable nucleic acid binding
agents should be used with caution in neuronal culture
studies.
inhibitorof HIV reverse
1998), spectinomycin
Acknowledgements
This work was supported by grants from the National Institutes of
Health (NS041342 to MMW and NS046649 to PJL) and a
Burroughs Wellcome Fund Career Award in Biomedical Sciences
(MMW).
References
Ahmed Ouameur A., Marty R., Neault J. F. and Tajmir-Riahi H. A.
(2004) AZT binds RNA at multiple sites. DNA Cell Biol. 23, 783–
788.
Brewer G. J. (1995) Serum-free B27/neurobasal medium supports dif-
ferentiated growth of neurons from the striatum, substantia nigra,
septum, cerebral cortex, cerebellum, and dentate gyrus. J. Neuro-
sci. Res. 42, 674–683.
Chen J. C., Zhang X., Singleton T. P. and Kiechle F. L. (2004)
Mitochondrial membrane potential change induced by Hoechst
33342 in myelogenous leukemia cell line HL-60. Ann. Clin. Lab.
Sci. 34, 458–466.
1264
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors

Page 13

Cui L., Locatelli L., Xie M. Y. and Sommadossi J. P. (1997) Effect of
nucleoside analogs on neurite regeneration and mitochondrial
DNA synthesis in PC-12 cells. J. Pharmacol. Exp. Ther. 280,
1228–1234.
Dalakas M. C. (2001) Peripheral neuropathy and anti-retroviral drugs.
J. Peripher. Nerv. Syst. 6, 14–20.
Dassonneville L., Hamy F., Colson P., Houssier C. and Bailly C. (1997)
Binding of Hoechst 33258 to the TAR RNA of HIV-1. Recognition
of a pyrimidine bulge-dependent structure. Nucleic Acids Res. 25,
4487–4492.
De Clercq E. (2002) New developments in anti-HIV chemotherapy.
Biochim. Biophys. Acta 1587, 258–275.
Druker B. J., Rosenthal D. S. and Canellos G. P. (1989) Chlorambucil,
vinblastine, procarbazine, and prednisone. An effective but less
toxic regimen than MOPP for advanced-stage Hodgkin’s disease.
Cancer 63, 1060–1064.
Dubinsky R. M., Yarchoan R., Dalakas M. and Broder S. (1989)
Reversible axonal neuropathy from the treatment of AIDS and
related disorders with 2¢,3¢-dideoxycytidine (ddC). Muscle Nerve
12, 856–860.
Edens R. E., Linhardt R. J., Bell C. S. and Weiler J. M. (1994) Heparin
and derivatized heparin inhibit zymosan and cobra venom factor
activation of complement in serum. Immunopharmacology 27,
145–153.
Fleming R. A. (1997) An overview of cyclophosphamide and ifosfamide
pharmacology. Pharmacotherapy 17, 146S–154S.
Frisk P., Stalberg E., Stromberg B. and Jakobson A. (2001) Painful
peripheral neuropathy after treatment with high-dose ifosfamide.
Med. Pediatr. Oncol. 37, 379–382.
Gould T. J. and Bickford P. C. (1994) The effects of chronic treatment
with N-tert-butyl-a-phenylnitrone on cerebellar noradrenergic
receptor function in aged F344 rats. Brain Res. 660, 333–336.
Hasbani M. J., Hyrc K. L., Faddis B. T., Romano C. and Goldberg M. P.
(1998) Distinct roles for sodium, chloride, and calcium in excito-
toxic dendritic injury and recovery. Exp. Neurol. 154, 241–258.
Hermann T. and Westhof E. (1999) Docking of cationic antibiotics to
negatively charged pockets in RNA folds. J. Med. Chem. 42,
1250–1261.
de Hoog C. L., Foster L. J. and Mann M. (2004) RNA and RNA binding
proteins participate in early stages of cell spreading through
spreading initiation centers. Cell 117, 649–662.
Hutvagner G. and Zamore P. D. (2002) A microRNA in a multiple-
turnover RNAi enzyme complex. Science 297, 2056–2060.
Imrie K. R., Couture F., Turner C. C., Sutcliffe S. B. and Keating A.
(1994) Peripheral neuropathy following high-dose etoposide and
autologous bone marrow transplantation. Bone Marrow Transplant
13, 77–79.
Job C. and Eberwine J. (2001) Localization and translation of mRNA in
dendrites and axons. Nat. Rev. Neurosci. 2, 889–898.
Kiechle F. L. and Zhang X. (2002) Apoptosis: biochemical aspects and
clinical implications. Clin. Chim. Acta 326, 27–45.
Keswani S. C., Chander B., Hasan C., Griffin J. W., McArthur J. C. and
Hoke A. (2003) FK506 is neuroprotective in a model of anti-
retroviral toxic neuropathy. Ann. Neurol. 53, 57–64.
Knowles R. B., Sabry J. H., Martone M. E., Deerinck T. J., Ellisman
M. H., Bassell G. J. and Kosik K. S. (1996) Translocation of RNA
granules in living neurons. J. Neurosci. 16, 7812–7820.
Kundu G. C., Schullek J. R. and Wilson I. B. (1994) The alkylating
properties of chlorambucil. Pharmacol. Biochem. Behav. 49, 621–
624.
Lein P. J. and Higgins D. (1989) Laminin and a basement membrane
extract have different effects on axonal and dendritic outgrowth
from embryonic rat sympathetic neurons in vitro. Dev. Biol. 136,
330–345.
Lockard R. E. and Kumar A. (1981) Mapping tRNA structure in solution
using double-strand-specific ribonuclease V1 from cobra venom.
Nucleic Acids Res. 9, 5125–5140.
Lowman H. B. and Draper D. E. (1986) On the recognition of helical
RNA by cobra venom V1 nuclease. J. Biol. Chem. 261, 5396–
5403.
Mahalakshmi Y. V., Jagannadham M. V. and Pandit M. W. (2000)
Ribonuclease from cobra snake venom: purification by affinity
chromatography and further characterization. IUBMB Life 49, 309–
316.
Martin D. P., Wallace T. L. and Johnson E. M. Jr (1990) Cytosine
arabinoside kills post-mitotic neurons in a fashion resembling
trophic factor deprivation: evidence that a deoxycytidine-depend-
ent process may be required for nerve growth factor signal trans-
duction. J. Neurosci. 10, 184–193.
Meresse P., Dechaux E., Monneret C. and Bertounesque E. (2004)
Etoposide: discovery and medicinal chemistry. Curr. Med. Chem.
11, 2443–2466.
Mohr E. and Richter D. (2000) Axonal mRNAs: functional significance
in vertebrates and invertebrates. J. Neurocytol. 29, 783–791.
Morris E. J. and Geller H. M. (1996) Induction of neuronal apoptosis by
camptothecin, an inhibitor of DNA topoisomerase-I: evidence for
cell cycle-independent toxicity. J. Cell Biol. 134, 757–770.
Nieminen A. L., Saylor A. K., Tesfai S. A., Herman B. and Lemasters
J. J. (1995) Contribution of the mitochondrial permeability trans-
ition to lethal injury after exposure of hepatocytes to t-butyl-
hydroperoxide. Biochem. J. 307, 99–106.
Ozols R. F. and Young R. C. (1985) High-dose cisplatin therapy in
ovarian cancer. Semin. Oncol. 12, 21–30.
Park J. S., Bateman M. C. and Goldberg M. P. (1996) Rapid alterations
in dendrite morphology during sublethal hypoxia or glutamate
receptor activation. Neurobiol. Dis. 3, 215–227.
Park D. S., Morris E. J., Stefanis L., Troy C. M., Shelanski M. L., Geller
H. M. and Greene L. A. (1998) Multiple pathways of neuronal
death induced by DNA-damaging agents, NGF deprivation, and
oxidative stress. J. Neurosci. 18, 830–840.
Patel S. R., Forman A. D. and Benjamin R. S. (1994) High-dose ifosf-
amide-induced exacerbation of peripheral neuropathy. J. Natl
Cancer Inst. 86, 305–306.
Patel H. V., Vyas A. A., Vyas K. A., Liu Y. S., Chiang C. M., Chi L. M.
and Wu W. (1997) Heparin and heparan sulfate bind to snake
cardiotoxin. Sulfated oligosaccharides as a potential target for
cardiotoxin action. J. Biol. Chem. 272, 1484–1492.
Piper M. and Holt C. (2004) RNA translation in axons. Annu. Rev. Cell.
Dev. Biol. 20, 502–503.
Povirk L. F. and Shuker D. E. (1994) DNA damage and mutagenesis
induced by nitrogen mustards. Mutat. Res. 318, 205–226.
Resar L. M., Phillips P. C., Kastan M. B., Leventhal B. G., Bowman
P. W. and Civin C. I. (1993) Acute neurotoxicity after intrathecal
cytosine arabinoside in two adolescents with acute lymphoblastic
leukemia of B-cell type. Cancer 71, 117–123.
Senger D. L. and Campenot R. B. (1997) Rapid retrograde tyrosine
phosphorylation of trkA and other proteins in rat sympathetic
neurons in compartmented cultures. J. Cell Biol. 138, 411–
421.
Slapak C. A., Fine R. L. and Richman C. M. (1985) Differential pro-
tection of normal and malignant human myeloid progenitors
(CFU-GM) from Ara-C toxicity using cycloheximide. Blood 66,
830–834.
Stefanis L., Park D. S., Friedman W. J. and Greene L. A. (1999)
Caspase-dependent and -independent death of camptothecin-trea-
ted embryonic cortical neurons. J. Neurosci. 19, 6235–6247.
Steward O. and Schuman E. M. (2003) Compartmentalized synthesis
and degradation of proteins in neurons. Neuron 40, 347–359.
Local toxic effects of nucleic acid binding agents on neurites
1265
? 2006 The Authors
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266

Page 14

Sturgess N. C., Rustad A., Fonnum F. and Lock E. A. (2000) Neurotoxic
effect of L-2-chloropropionic acid on primary cultures of rat
cerebellar granule cells. Arch. Toxicol. 74, 153–160.
Thom G. and Prescott C. D. (1997) The selection in vivo and charac-
terization of an RNA recognition motif for spectinomycin. Bioorg.
Med. Chem. 5, 1081–1086.
Wallace T. L. and Johnson E. M. Jr (1989) Cytosine arabinoside kills
post-mitotic neurons: evidence that deoxycytidine may have a role
in neuronal survival that is independent of DNA synthesis.
J. Neurosci. 9, 115–124.
Wallace S. T. and Schroeder R. (1998) In vitro selection and character-
ization of streptomycin-binding RNAs: recognition discrimination
between antibiotics. RNA 4, 112–123.
Wang D. and Lippard S. J. (2005) Cellular processing of platinum
anticancer drugs. Nat. Rev. Drug Discov. 4, 307–320.
Ward M. W., Rego A. C., Frenguelli B. G. and Nicholls D. G. (2000)
Mitochondrial membrane potential and glutamate excitotoxicity
in cultured cerebellar granule cells. J. Neurosci. 20, 7208–
7219.
Wilson W. D., Ratmeyer L., Zhao M., Strekowski L. and Boykin D.
(1993) The search for structure-specific nucleic acid-interactive
drugs: effects of compound structure on RNA versus DNA inter-
action strength. Biochemistry 32, 4098–4104.
Wu K. Y., Hengst U., Cox L. J., Macosko E. Z., Jeromin A., Urquhart
E. R. and Jaffrey S. R. (2005) Local translation of RhoA regulates
growth cone collapse. Nature 436, 1020–1024.
Xu Y., Traystman R. J., Hurn P. D. and Wang M. M. (2003) Neurite-
localized estrogen receptor-a mediates rapid signaling by estrogen.
J. Neurosci. Res. 74, 1–11.
Zhang X. and Kiechle F. L. (1997) Hoechst 33342-induced apoptosis in
BC3H-1 myocytes. Ann. Clin. Lab. Sci. 27, 260–275.
Zhang X., Chen J., Davis B. and Kiechle F. (1999) Hoechst 33342
induces apoptosis in HL-60 cells and inhibits topoisomerase I
in vivo. Arch. Pathol. Lab. Med. 123, 921–927.
1266
S. Pin et al.
Journal Compilation ? 2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 1253–1266
? 2006 The Authors