AMINO ACID ANALOG TOXICITY IN PRIMARY RAT NEURONAL
AND ASTROCYTE CULTURES: IMPLICATIONS FOR PROTEIN
MISFOLDING AND TDP-43 REGULATION
Kalavathi Dasuri
a
, Philip J. Ebenezer
a
, Romina M. Uranga
b
, Elena Gavilán
c
, Le Zhang
a
, Sun
OK Fernandez-Kim
a
, Annadora J. Bruce-Keller
a
, and Jeffrey N. Keller
a,‡,†
a
Pennington Biomedical Research Center, Louisiana State University System, Baton Rouge, LA
70808-4124, USA
b
Instituto de Investigaciones Bioquímicas de Bahía Blanca, Universidad Nacional del Sur and
Consejo Nacional de Investigaciones Científicas y Técnicas, Bahía Blanca, Argentina
c
Departamento de Bioquímica, Bromatología, Toxicología y Medicina Legal, Facultad de
Farmacia, Universidad de Sevilla, Seville, Spain
Abstract
Amino acid analogs promote translational errors that result in aberrant protein synthesis, and have
been used to understand the effects of protein misfolding in a variety of physiological and
pathological settings. TDP-43 is a protein that is linked to protein aggregation and toxicity in a
variety of neurodegenerative diseases. In this study we exposed primary rat neurons and astrocyte
cultures to established amino acid analogs (Canavanine and Azetidine-2-carboxylic acid), and
observed both cell types undergo a dose-dependent increase in toxicity, with neurons exhibiting a
greater degree of toxicity as compared to astrocytes. Neurons and astrocytes exhibited similar
increases in ubiquitinated and oxidized protein following analog treatment. Analog treatment
increased Heat shock protein (Hsp) levels in both neurons and astrocytes. In neurons, and to a
lesser extent astrocytes, the levels of TDP-43 increased in response to analog treatment. Taken
together, these data indicate that neurons exhibit preferential toxicity and alterations in TDP-43, in
response to increased protein misfolding, as compared to astrocytes.
Keywords
Aging; Alzheimer’s disease; cell death; neurodegeneration; neurotoxicity; protein aggregation;
ubiquitin
INTRODUCTION
Protein misfolding and protein aggregation are characteristic features of many
neurodegenerative diseases including Alzheimer’s disease (AD), Parkinson’s disease (PD),
frontotemporal lobar degeneration (FTLD), and amyotrophic lateral sclerosis (ALS). The
amino acid analogs canavanine (Can) and azetidine-2-carboxylic acid (AZC) mimic the
‡
Corresponding author: Dr Jeffrey N. Keller, Professor, Associate Executive Director of Basic Research, Director, Institute for
Dementia Research and Prevention, Pennington Biomedical Research Center, LSU System, 6400 Perkins Road, Baton Rouge, LA
70808-4124, (P) 225-763-3190; (F) 225-763-3193;, (E) jeffrey.keller@pbrc.edu.
†
This work was supported by grants from the NIA (AG029885, AG025771) and the Hibernia National Bank/Edward G Schlieder
Chair (J.N.K.).
The authors declare that they do not have any conflict of interests.
NIH Public Access
Author Manuscript
J Neurosci Res. Author manuscript; available in PMC 2011 September 19.
Published in final edited form as:
J Neurosci Res
. 2011 September ; 89(9): 1471–1477. doi:10.1002/jnr.22677.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
natural amino acids arginine and proline, and are incorporated into nascent polypeptides
promoting irreversible abnormal protein confirmation (Fowden and Richmond, 1963;
Fowden et al., 1967; Bessonov et al., 2010; Grant et al 1975; Zagari et al., 1990; Rosenthal
et al., 1989; Prouty et al., 1975; Rodgers and Shiozawa, 2008) and alterations in global
protein synthesis ( Qian et al., 2010; Kretz-Remy, et al., 1998). Can and AZC can therefore
be used to mimic the increased levels of abnormal proteins observed in aging cells, and
potentially model increased protein misfolding observed in a variety of neurodegenerative
conditions.
In order to prevent proteotoxicity from increased protein misfolding, cells rely on the
function of numerous heat shock proteins (Hsps) including Hsp70 and Hsp40 (Trottel et al.,
2002; Li et al., 1985; Watowich and Morimoto, 1988; Hightower 1991; Ananthan et al.,
1986; Barrett et al., 2004). Both Can and AZC have been shown to induce a variety of Hsps
(Trottel et al., 2002; Li et al., 1985; Watowich and Morimoto, 1988; Qian et al., 2010;
Kozutsumi et al., 1998; Thomas and Mathews, 1984), consistent with both analogs
promoting proteotoxic stress. Currently it is not known whether neurons and astrocytes
differ in regards to their sensitivity to toxicity or Hsp induction in response to amino acid
analogs such as Can and AZC.
A number of abnormal proteins have been shown to accumulate in neurodegenerative
diseases such as AD, PD, and FTLD ( Uversky 2008; Koo et al., 1999; Agorogiannis et al.,
2004; Reddy, 2006; Meridith 2005; Ross and Poirier, 2004; Zhu et al., 2005) suggesting the
genesis of proteotoxic stress. Recent studies have suggested an important role for TAR
DNA-binding protein of 43-kDa (TDP-43) in modulating proteotoxicity associated with
increased protein misfolding (Neumann 2006; Chen-plotkin et al., 2010). TDP-43 is
abundantly expressed in neurons and glia, and has been identified as a major component of
ubiquitinated neuronal cytoplasmic inclusions (NCI) and neuronal intranuclear inclusions.
Full length TDP-43, as well as cleavage products of ~25kDa and 35kDa, are observed in
ALS and FTLD (Zhang et al., 2009; Chanson et al., 2010; Ritson et al., 2010; Kabashi et al.,
2008; Barmada et al., 2010; Halawani and Latterich, 2006; Dalal et al., 2004). Currently, it
is not known whether analogs such as Can and AZC modulate TDP-43 homeostasis in
primary neuron and astrocyte cultures.
In the present paper, we demonstrate that treatment of primary rat neurons and astrocytes
results in a dose-dependent increase in cell death, with neurons being more vulnerable to the
toxicity of Can and AZC. The preferential increase in neuronal toxicity did not appear to be
linked to differences in ubiquitinated proteins, oxidized proteins, or Hsp induction. Amino
acid analogs induced increased levels of TDP-43 and its cleavage products. Taken together
these data have implications for understanding how increased levels of aberrant proteins
during aging and neurodegenerative disease contribute to neuronal death and dysfunction in
the brain.
MATERIALS AND METHODS
Materials
The antibodies to β-actin (SC-47778) and ubiquitin (SC - 8017) were purchased from Santa
Cruz Biotechnology Company (Santa Cruz, CA, USA). The antibodies to TDP-43 (3448S)
were purchased from Cell Signaling Technology, Inc.(Cambridge, MA, USA). The
antibodies to Hsp70 (SPA-810D) and Hsp40 (SPA-400D) were purchased from Enzo Life
Sciences International, Inc. (Plymouth Meeting, PA). Oxyblot kit was purchased from
Millipore Company (Billerica, MA, USA). All the chemicals including Hoechts 33342
(bisBenzamide trihydrochloride) staining, Triton X-100, protease inhibitor mix, EDTA,
DNase I, AZC (L-Azetidine-2-carboxylic acid) and L-Canavanine were purchased from
Dasuri et al. Page 2
J Neurosci Res. Author manuscript; available in PMC 2011 September 19.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Sigma-Aldrich, Corp. (St. Louis, MO, USA). All electrophoresis and immunoblot reagents
were purchased from Bio-Rad Laboratories (Hercules, CA, USA). All cell culture supplies
were obtained from GIBCO Life Sciences (Gaithersburg, MD, USA). The BCA reagent was
purchased from Thermo Scientific, Inc. (Pittsburgh, PA, USA).
Establishment and maintenance of primary neuron and astrocyte cultures: treatment with
analogs
Neuronal cultures were established as described previously by our laboratory (Ding et al.,
2006; Dasuri et al.,, 2010b, Ebenezer et al.,, 2010). Briefly, primary rat cortical neuronal
cells were cultured from E18 Sprague-Dawley rats and maintained in 5% CO2 at 37°C in
MEM or Neurobasal medium containing 5% fetal bovine serum (heat inactivated), N2
supplement, B27 supplement, and 1% antibiotic. Cells were used in experiments between
days 6–9 (Neurons) and 12–15 (Astrocytes) post plating. Rat astrocyte cultures were
established from E18 Sprague-Dawley rats as described previously by our laboratory (Ding
et al., 2006; Dasuri et al.,, 2010b, Ebenezer et al.,, 2010). Astrocytes were maintained in 5%
CO
2
at 37°C in MEM medium containing 5% fetal bovine serum (heat inactivated), N2
supplement and 1% antibiotic solution. All animals were utilized in accordance with IACUC
approved protocols at the Pennington Biomedical Research Center. For the analysis of
protein levels following analog treatment, the primary cultures of neurons and astrocytes
were treated with various concentrations of analogs (AZC or Can) and the pelleted cells
were frozen at −80°C until further use.
Analysis of cell viability
Cell survival was determined by quantification of apoptotic and necrotic nuclei using
Hoechts 344 staining as described previously (Keller et al., 1998; Ding et al., 2006; Dasuri
et al.,, 2010b, Ebenezer et al.,, 2010). Briefly, Neuronal and astrocyte cells were treated with
increasing concentrations of analogs (AZC or Can) for 3 hours or with 5mM of analog for
the indicated time intervals. The treated cells were stained with the fluorescent DNA-
binding dye Hoechts 344 at a concentration of 1μg/μl, and the percentage of viable cells
were determined by counting the number of dead cells (condensed and fragmented nuclei)
using a fluorescence microscope equipped with a 32X objective. Additional confirmation of
cell viability was determined using MTT reduction as a measure of cell viability as reported
previously (Kruman et al., 1997; Ding et al., 2006; Dasuri et al.,, 2010b, Ebenezer et al.,
2010).
Protein estimation
Protein concentration of the cell lysates made in cell lysis buffer was estimated using BCA
(Thermo Scientific) reagent as described by the manufacturer.
Western blotting
The protein samples were analyzed by SDS-PAGE and immunoblotted with specified
antibodies as described previously by our laboratory (Dasuri et al., 2010a).
Analysis of protein oxidation levels
Protein carbonyl levels were analyzed using Oxyblot kit (Millipore) as described by the
manufacturer. Briefly, 10 μg of protein lysate was derivatized with 2, 4-
dinitrophenylhydrazone, (DNPH) and then the derivatized products were detected by the
Western blot analysis as described by the manufacturer.
Dasuri et al. Page 3
J Neurosci Res. Author manuscript; available in PMC 2011 September 19.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Triton X-100 fractionation of cell lysates and analysis
Triton X-100 fractionation was done as described by our laboratory (Dasuri et al., 2010b).
Briefly, Cells were suspended in extraction buffer (150μl) containing 50mM Tris- HCl,
pH-7.4, 150mM NaCl, and 1mM EDTA, 1% triton X-100, protease inhibitor cocktail and
DNase I (buffer A). The protein lysates were incubated for 30 minutes with shaking at 4°C.
Then the triton X-100 soluble and insoluble fractions were separated by centrifugation for
20 minutes at 14,000 X g. The protein levels in triton X-100 soluble and insoluble fractions
were estimated and equal amounts of protein were used for the analysis of protein levels
using the western blotting.
Statistics
Statistical analyses were conducted using the Prism 3.0 software (GraphPad Software, San
Diego, CA). Student’s two-tailed t-test was used to determine whether observed differences
were statistically significant (p<0.05).
RESULTS
Neuronal cells exhibit increased sensitivity to amino acid analog toxicity
Canavanine and AZC are amino acid analogs which incorporate into newly synthesized
proteins resulting in protein misfolding and the accumulation of abnormal proteins (Kelley
and Schlesinger, 1978; Ananthan et al., 1986; Hightower 1980). In order to study the
relative susceptibilities of primary rat neuronal and astrocyte cultures to the toxicity of
misfolded and abnormal proteins, we conducted studies analyzing the acute toxicity of Can
and AZC. We have observed that AZC and Can promoted a dose-dependent increase in cell
death in both neurons and astrocytes (Figures 1 & 2). The toxicity of AZC and Can was
nearly identical, with both agents observed to induce more pronounced cell death in neurons
as compared to (Figures 1 & 2) at every dose analyzed. Treatment with analogs induced a
time dependent increase in cell death, with neurons observed to undergo more severe loss of
viability than astrocytes (Figure 1 & 2), at every dose analyzed.
Neuronal and astrocyte cells exhibit higher level of ubiquitinated and oxidized proteins
following amino acid analog treatment
To overcome the accumulation of abnormal or misfolded proteins cells rely on the function
of multiple proteolytic pathways, including ubiquitin-dependent pathways (Hershko and
Ciechanover, 1992; Hochstrasser 1992; Navon and Ciechanover, 2009). In our analysis, we
observed that Can and AZC increased the levels of ubiquitinated protein in both neurons and
astrocytes (Figure 3). Although much more variable, amino acid analog treatment also
increased the levels of oxidized proteins in both neurons and astrocytes (Figure 3).
Amino acid analog treatment increases the levels of Heat shock protein (Hsp) in neurons
and astrocytes
Next, we sought to elucidate whether Can and AZC treatment altered the levels of Hsps in
neurons and astrocytes. In our studies we observed that both neurons and astrocytes
exhibited increases in Hsp70 and Hsp40 levels following Can and AZC treatment (Figure 4).
See supplementary Figure 1 for quantification of blots from Figure 4.
Effect of amino acid analog treatment on the levels of TDP-43 in neurons and astrocytes
In order to determine if amino acid analog treatment alters the levels of TDP-43 we
conducted studies looking at the total levels of TDP-43, in neurons and astrocytes. In this
analysis we observed that the total levels of TDP-43 were robustly increased in neurons
(Figure 5). In astrocytes increases were transient or occurred to a lesser degree (Figure 5),
Dasuri et al. Page 4
J Neurosci Res. Author manuscript; available in PMC 2011 September 19.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
than was observed in neuronal cultures. Analysis revealed that the levels of the known ~35
kDa cleavage product of TDP-43 were selectively observed in neurons treated with the
highest concentrations of amino acid analogs, and not observed in astrocyte cultures (Figure
5). Please see supplementary Figure 2 for quantification of TDP-43.
DISCUSSION
Amino acid analogs AZC and Can are used to induce protein misfolding (Goldberg and
Dice, 1974; Trotter et al., 2002) It is known that AZC causes changes in the confirmation of
the protein back bone leading to functional alterations in multiple proteins (Bessonov et al.,
2010; Trotter et al., 2001;Trotter et al., 2002; Hoshikawa et al., 2003). In our study we show
for the first time that AZC and Can both cause a loss of cell viability in primary CNS
cultures. These post-mitotic neurons in the current study exhibited dramatically elevated
sensitivity to the toxicity associated with increased protein misfolding, when compared to
mitotic astrocyte cultures. Neurons are known to be more vulnerable than astrocytes to a
variety of stressors relevant to aging including oxidative stressors and proteasome inhibitors
(Keller et al., 1999, 2000; Ding et al., 2006; Dasuri et al., 2010b; Schmuck et al., 2002;
Watts et al., 2005). This increased vulnerability may be due in part to the increased
propensity of neurons to undergo increases in protein hydrophobicity following proteasome
inhibitor treatment (Dasuri et al., 2010a), as compared to astrocyte cultures. The key to
understanding the basis for this toxicity likely resides in defining if toxicity is mediated by
gross abnormalities in the proteome, or conversely mediated by selective or key
perturbations within the proteome, during periods of increased protein misfolding.
In the current study we observed that the levels of ubiquitinated and oxidized proteins are
increased in both neurons and astrocytes following amino acid analog treatment. The
increase in both ubiquitinated and oxidized proteins were similar in both neurons and
astrocytes, suggesting that the increased neuron death in neurons as compared to astrocyte
cultures, is not likely mediated by gross increases in either ubiquitinated or oxidized
proteins. Presumably, the misfolding of proteins by AZC and Can treatment results in their
rapid ubiquitination, with the increase in ubiquitinated proteins in the present study
consistent with the proteolytic pathways responsible for ubiquitinated protein degradation
being overwhelmed or inhibited following amino acid analog treatment. In our search of the
literature, this study appears to be the first to demonstrate amino acid analogs are capable of
increasing the levels of oxidized protein in any cell type. Neither AZC or Can would be
expected to be capable of producing reactive oxygen species, suggesting that the increased
levels of oxidized protein in the current study are due to the misfolded proteins themselves
being more vulnerable to the endogenous oxidative stressors in both neurons and astrocytes
in culture. Alternatively, the increased protein misfolding induced by both AZC and Can
may indirectly increase oxidative stress within both neurons and astrocytes. For example,
misfolding (and presumably loss of function) of antioxidant enzymes and key cell signaling
components, results in amino acid analogs promoting increases in oxidized proteins by
shifting the intracellular environment to a pro-oxidant environment. Interestingly, previous
studies have implied a role for increased reactive oxygen species as a mechanism for amino
acid analog-induced increases in NFkB activation (Kretz-Remy et al., 1998).
Our results suggest that increased protein misfolding in the brain during aging and in a
variety of neurodegenerative conditions promotes neurocognitive abnormalities through
direct effects on neurons, as well as indirect mechanisms. The direct mechanisms relate to
the acute neurotoxicity and neuron death, as observed in the present study. Alternatively, it
is likely the protein misfolding in the astrocytes in the current study is sufficient to promote
the induction of chronic inflammatory signaling, which indirectly is sufficient to promote
neurocognitive abnormalities based on their effects on neurons.
Dasuri et al. Page 5
J Neurosci Res. Author manuscript; available in PMC 2011 September 19.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
