Delayed caffeine treatment prevents nigral dopamine neuron loss in a progressive rat
model of Parkinson's disease
Patricia K. Sonsallaa,⁎, Lai-Yoong Wonga, Suzan L. Harrisa, Jason R. Richardsonb, Ida Khobahyc,
Wenhao Lic, Bharathi S. Gadadc, Dwight C. Germanc
aDepartment of Neurology, UMDNJ Robert Wood Johnson Medical School, Piscataway NJ, USA
bDepartment of Environmental and Occupational Medicine, UMDNJ Robert Wood Johnson Medical School, Piscataway NJ, USA
cDepartment of Psychiatry, University of Texas Southwestern Medical Center, Dallas TX, USA
a b s t r a c ta r t i c l e i n f o
Received 14 December 2011
Revised 9 January 2012
Accepted 19 January 2012
Available online 28 January 2012
Parkinson's disease (PD) is characterized by a prominent degeneration of nigrostriatal dopamine (DA) neu-
rons with an accompanying neuroinflammation. Despite clinical and preclinical studies of neuroprotective
strategies for PD, there is no effective treatment for preventing or slowing the progression of neurodegenera-
tion. The inverse correlation between caffeine consumption and risk of PD suggests that caffeine may exert
neuroprotection. Whether caffeine is neuroprotective in a chronic progressive model of PD has not been eval-
uated nor is it known if delayed caffeine treatment can stop DA neuronal loss. We show that a chronic uni-
lateral intra-cerebroventricular infusion of 1-methyl-4-phenylpyridinium in the rat brain for 28 days
produces a progressive loss of DA and tyrosine hydroxylase in the ipsilateral striatum and a loss of DA cell
bodies and microglial activation in the ipsilateral substantia nigra. Chronic caffeine consumption prevented
the degeneration of DA cell bodies in the substantia nigra. Importantly, neuroprotection was still apparent
when caffeine was introduced after the onset of the neurodegenerative process. These results add to the clin-
ical relevance for adenosine receptors as a disease-modifying drug target for PD.
© 2012 Elsevier Inc. All rights reserved.
Parkinson's disease (PD) is a devastating neurodegenerative disor-
der characterized by extensive loss of the nigrostriatal dopamine
(DA) neurons and neuroinflammation (Appel et al., 2009; Hirsch
and Hunot, 2009). At present, the only therapy for PD is symptomatic
and such treatments eventually fail. Disease-modifying approaches
that can slow or stop the progression of neurodegeneration are des-
Neuropathology in PD occurs long before any substantive clinical
symptoms appear. It is estimated that at the time of symptom presen-
tation there may be 60–80% loss of striatal DA (Hornykiewicz, 1979).
Experimental PD animal models have provided a plethora of informa-
tion on mechanisms of DA neurodegeneration. However, most of
these models are based on acute neurotoxicant exposure and may
not accurately portray the chronic pathology that is seen in the
human PD brain. Moreover, the introduction of disease-modifying
drugs to PD patients will occur long after neurodegeneration has
been initiated and under conditions of an on-going pathological pro-
cess. To better predict the efficacy of potential disease-modifying
agents in PD, it is important that a progressive PD model be used
and that the drug is introduced during the pathological processes of
neurodegeneration and neuroinflammation. We have developed a
chronic progressive rat model of PD in which MPP+is infused into
the left cerebral ventricle for 28 days. In this model, there is a selec-
tive loss of nigrostriatal DA neurons accompanied by neuroinflamma-
tion in the nigrostriatal DA brain regions ipsilateral to the side of
infusion (Yazdani et al., 2006; Zeevalk et al., 2007). In the present
studies, we have used this model to examine whether DA neurons
can be rescued from neurodegeneration by concurrent or delayed caf-
In the past decade, interest in caffeine has emerged as a possible
neuroprotective compound. Epidemiological studies show an inverse
relationship between caffeine consumption and the risk of develop-
ing PD which suggests caffeine may exert neuroprotection in humans
(Morelli et al., 2011). In acute animal models of PD, caffeine has been
found to be neuroprotective. Treatment of mice with caffeine protects
DA neurons from the acute neurotoxicity of 1-methyl-4-phenyl-
1,2,3,6-tetrahydropyridine (MPTP) (Chen et al., 2001; Kalda et al.,
2006; Xu et al., 2010). In the acute 6-hydroxydopamine rat model
of PD, chronic caffeine treatment protects against the loss of striatal
DA neurochemistry and nigral DA cell bodies produced by a single
intrastriatal infusion of 6-hydroxydopamine (Aguiar et al., 2006;
Joghataie et al., 2004). In the one chronic study, caffeine was shown
Experimental Neurology 234 (2012) 482–487
⁎ Corresponding author at: 661 Hoes Lane, Staged Research Building, Rm 142,
RWJMS-UMDNJ, Piscataway, NJ 08854, USA. Fax: +1 732 235 5295.
E-mail address: email@example.com (P.K. Sonsalla).
0014-4886/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/yexnr
to protect mouse nigral DA cell bodies in the chronic pesticide (para-
quat/maneb) exposure model (Kachroo et al., 2010). However, in all
of these studies, caffeine was introduced prior to or concurrent with
In the present study, we have further characterized the progres-
sive nature of damage in the chronic PD rat model and have investi-
gated the ability of caffeine to prevent neurodegeneration as well as
to rescue DA neurons. Our findings demonstrate that caffeine protects
the nigral DA cell bodies even when treatment is initiated later in the
Male Sprague–Dawley rats (weighing ~300 g at the beginning of
the study) (Taconic Farms, Germantown, NY) were maintained on a
12-h light–dark cycle with food and water available ad libitum. Exper-
iments were performed in accordance with the National Institutes of
Health Guide for the Care and Use of Laboratory Animals and were
approved by the animal care committee of UMDNJ.
Surgeries and MPP+infusions
Vehicle (sodium-iodide) or MPP+-iodide (Sigma-Aldrich) dis-
solved in saline was infused into the left cerebral ventricle via tubing
linked to an Alzet osmotic minipump (model 2ML4) implanted sub-
cutaneously as previously described (Yazdani et al., 2006; Zeevalk et
al., 2007). Stereotaxic cannula placement was at coordinates relative
to bregma: anterior
−3.9 mm (Paxinos and Watson, 1986). Stereotaxic surgery and mini-
pump placement were performed under anesthesia. MPP+-iodide or
vehicle was administered at a dose of 75 μg/day for 28 days with a
drug delivery rate of 2.5 μl/h. With the exception of the time course
studies, animals were killed 27–28 days after cannula placement.
lateral left +1.4 mm,depth
was prepared fresh every 3–4 days. The dose chosen was to approxi-
mate caffeine intake in humans and equates to 60–80 mg/kg/day. In
humans with high caffeine intake, doses approximate 10–20 mg/kg/
day (Stavric et al., 1988). While the dose in our rats is higher than that
consumed by humans, thehalf-life of caffeine in the rat is ~1 h whereas
inhumansit is ~5 h(Xuetal., 2010). Thus, a higherdoseinrats is need-
ed to achieve blood or brain concentrations (~22 μM) similar to those
achieved in the serum of coffee drinkers (Bienvenu et al., 1990;
Costenla et al., 2010; Gandhi et al., 2010).
In rats used for neurochemistry and midbrain DA cell counts,
brains were rapidly removed and sectioned at mid-hypothalamus.
Left and right striata were dissected from the forebrain, weighed
and frozen at −80 °C until analyzed. The hind brain containing the
substantia nigra (SN) was immersion fixed for immunohistochemis-
try. Striatal tyrosine hydroxylase (TH) was determined by ELISA and
monoamines and metabolites determined by HPLC with electrochem-
ical detection, as previously described (Alfinito et al., 2003).
Immunohistochemistry for nigral DA cells and activated microglia
The hind brain was used for performing TH+cell counts using im-
munohistochemical methods similar to those described previously
(Yazdani et al., 2006). Briefly, 30-μm-thick coronal sections were cut
through the entire SN. Every fourth section through the rostral–
caudal extent of the SN was stained with an antibody against TH
(1:4000; Protos Biotech Corp.) to identify the DA neurons.
Some rats were transcardially perfused for evaluating microglial
response and TH staining in both the striatum and the SN. Animals
were deeply anesthetized and were transcardially perfused with
100 ml of 0.9% NaCl in 0.1 M sodium phosphate buffer, pH 7.3, con-
taining 50 units/ml heparin. This was followed by 500 ml of 4% para-
formaldehyde in 0.1 M sodium phosphate buffer, pH 7.3. The brains
were dissected out and taken through a series of 10, 20, and 30% su-
crose in 0.1 M sodium phosphate buffer, pH 7.3, for cryoprotection.
Cryostat sections were cut in the coronal plane at a thickness of
20 μm. Sections from the striatum and the SN were immunostained
with an antibody against ED1 (1:100; AbD Serotec) to identify acti-
vated microglia. Some sections were double labeled by immunofluo-
rescence for TH and ED1. Digital photographs were taken using a
Zeiss Axioplan microscope (Carl Zeiss, Inc.) equipped with an epi-
fluorescence illuminator and Axiovision software.
StereoInvestigator software (version 9.0. MicroBrightfield Inc.,
Williston, VT) was used to count TH-immunoreactive (TH-IR) SN
cells. Cells were counted with a 40× objective using a Leica DMRE mi-
croscope. The cell counting frame was 50×50×5 μm with a 2 μm
upper and lower guard zone. A cell was defined as a TH-IR somata
with a clearly visible unstained nucleus. The TH cell counts were
taken from 6 sections, spaced 4 apart (120 μm) and 200–250 cells
were counted in the SN on the left side of the brain. The SN region
was defined according to previous anatomical demarcation in the
rat (German and Manaye, 1993).
Differences among means were analyzed using one-way analysis
of variance (ANOVA) or two-way ANOVA. Two-way ANOVA revealed
significant differences between treatment groups and sides (left le-
sioned and right non-lesioned sides). However, one-way ANOVA on
data from the right sides revealed no significant differences across
treatment groups allowing use of one-way ANOVA for comparisons
in the left lesioned side across treatment groups. When ANOVA
showed significant differences, comparisons between means were
tested by the Tukey–Kramer or Bonferroni multiple comparisons
post hoc test. In all analyses, the null hypothesis was rejected at the
0.05 level. All values are expressed as the mean±SEM.
Loss of striatal DA and TH is linear over time with continuous icv MPP+
Reductions in the content of DA and TH in the striatum ipsilateral
to the infusion progressed over time in rats infused with MPP+
(75 μg/day) into the left cerebral ventricle (Fig. 1). Because no signif-
icant reductions in these measures were seen in the contralateral stri-
ata in any of the groups, the data are plotted as left/right ratios to
simplify data presentation. Reductions in DA in the left striatum
(plotted as left/right ratios) were significant by 2 weeks with pro-
gressively greater reductions seen in subsequent weeks (Fig. 1). TH
loss occurred more slowly, but the loss of both TH and DA exhibited
significant correlation coefficients over time; r2values of 0.98 and
0.99, respectively (pb.0001). Thus, MPP+infusion produces a pro-
gressive loss of striatal DA and TH making this model ideal for evalu-
ating neuroprotective agents when administered before or during the
period of MPP+insult.
P.K. Sonsalla et al. / Experimental Neurology 234 (2012) 482–487
Neither MPP+nor caffeine treatment alters rat body weights
As shown in Table 1, there were no significant differences in the
end-of-study body weights in the different groups of rats (naïve,
vehicle-treated, MPP+or caffeine plus MPP+). These data indicate
that neither MPP+nor caffeine treatment had an adverse effect on
weight gain in the rats.
Caffeine treatment initiated simultaneously or during the course of
ongoing neurodegeneration reduces loss of nigral DA neurons
MPP+produced a significant reduction in the number of TH-
immunostained cells (49±9%) in the left SN of animals killed
27–28 days after starting infusion (Fig. 2). Oral caffeine (1 g/l in the
drinking water) from the onset of MPP+infusions prevented the
loss of the nigral TH-immunostained cell bodies. More importantly,
supplying caffeine at 1 or 3 weeks after initiating MPP+infusions
also reduced the loss of nigral TH-immunostained cells. These time
points (1 and 3 wks) were selected for starting caffeine treatment
as they represent early and later stages of loss of striatal DA based
upon results shown in Fig. 1. These data demonstrate that degenera-
tion of nigral DA neurons can be halted or slowed even after the neu-
rodegenerative process has begun.
Caffeine does not modify MPP+-induced decreases in striatal DA or TH
In contrast to the protection seen in the SN, caffeine did not signif-
icantly modify the MPP+-induced changes in the striatum. MPP+
produced significant reductions in TH (by 35%), DA (by 50%),
DOPAC (by 56%) and HVA (by 61%) in the left striatum as compared
to the left striatum in naive rats (see Fig. 3A). The administration of
chronic caffeine did not modify TH, DA or DOPAC in the right non-
lesioned striatum indicating that the caffeine treatment did not mod-
ify synthesis or turnover of DA (Fig. 3B). In rats treated with caffeine
and MPP+, the reductions in DA and TH in the left striatum were gen-
erally less than in the MPP+group but did not, however, differ signif-
icantly from the MPP+group. In vehicle-treated rats, there were no
significant neurochemical differences between left and right striata
(L/R ratios for TH: 1.12±0.19 and for DA: 1.05±0.18. mean±SD; 5
rats) indicating that cannula placement did not significantly damage
DA nerve terminals. Fig. 3 also illustrates that the striatal serotonin
system was not affected by MPP+administration (no significant
loss of 5HT or 5HIAA in the MPP+-treated rats), indicating selectivity
of this dose of MPP+towards DA neurons as has been previously
reported (Yazdani et al., 2006).
Caffeine attenuates microglia activation in the SN but not in the striatum
of MPP+-treated rats
To evaluate the microglial response, brain sections containing the
SN or striatum were immunostained with an antibody to ED1 which
detects activated microglia and were counterstained with a TH anti-
body. Immunohistochemical staining revealed increased numbers of
activated microglia in the left SN and left striatum of the MPP+-trea-
ted and the caffeine/MPP+-treated rats. Fig. 4 illustrates data from
representative animals in the two groups (n=2 rats/group). In the
photomicrograph in the upper panel, ED1 immunostained cells
were prevalent in the left SN whereas only very few were seen in
the right SN, as would be expected in non-lesioned or non-damaged
tissue. It can also be seen that there is a loss of TH-immunostained
cells, especially in the medial region of the SN. In rats treated with
MPP+and caffeine, there were fewer cells that immunostained for
ED1. Additionally, there is preservation of TH-immunostained cells
in the rats treated with caffeine and MPP+.
In the striatum, there was intense ED1 immunostaining, particu-
larly in regions near the ventricle, in the MPP+-treated rats. Very
few ED1 immunostained cells are seen in the right striatum, consis-
tent with the lack of damage in the non-lesioned striatum. Also, in
the right striatum, there is intense TH immunostaining whereas in
the left striatum, TH immunostaining is markedly reduced. In rats
that were treated with MPP+and caffeine, the ED1 immunostaining
in cells in the left striatum appears similar to that seen in rats treated
with only MPP+. Likewise, the reduced striatal TH staining in rats
treated with MPP+and caffeine is similar to that seen in rats treated
with only MPP+. We also found that the expression of glial markers
(GFAP for astrocytes and MAC-1 for activated microglia) was elevated
by approximately 3-fold in the left striata of MPP+treated rats and
that caffeine treatment did not significantly modify these effects
(data not shown). We note that these are preliminary findings
obtained from a small number of animals and that the effect of caf-
feine on the microglia response requires further characterization
Arresting the progression of neurodegeneration in PD remains a
critical, unmet goal and is a genuine challenge to research in PD. We
have found that caffeine treatment protects against the loss of nigral
DA neurons in a chronic progressive rat model of PD. Most important-
ly, caffeine treatment was protective even when introduced late into
the neurodegenerative process suggesting that it may be capable of
arresting or slowing neurodegeneration. To our knowledge, this is
the first evidence that delayed pharmacological therapy can retard
degeneration of DA neurons.
The chronic MPP+rat model provides an excellent progressive
model of neurodegeneration. We have previously shown a progres-
sive loss over time of TH-immunostained neurons in the SN ipsilateral
to the infusion and that the reduction in the number of neurons is due
Fig. 1. Continuous icv MPP+administration produces a progressive and linear reduc-
tion in striatal DA and TH. Rats received continuous MPP+infusions (75 μg/day) into
the left cerebral ventricle and were killed at 7, 14, or 28 days after starting the MPP+
infusion. DA and TH protein were measured in both the left and right striata. Results
are plotted as the ratio of DA or TH content in the left striatum to the right striatum
(L/R ratio) as a function of days of MPP+exposure from the number of rats indicated
in parenthesis. Data at time 0 are L/R ratios from naive rats. DA and TH content in
the right striatum of the treatment groups did not differ significantly from right stria-
tum of naive rats nor did they differ across the treatment groups. Linear regression
analysis showed correlation coefficient r2values of 0.98 (pb0.01) for DA and of 0.99
(pb0.01) for TH. *pb0.01 vs L/R ratio in naive rats.
Treatment effects on body weight. Rats were weighed at the end of
the study (28 days after surgery). Results are the mean±SD of the
number of rats shown in parenthesis. Treatment did not significant-
ly alter weight gain over the 28-day course of treatment.
Treatment groupWeight (g)
Caffeine 1 wk/MPP+
Caffeine 3 wk/MPP+
P.K. Sonsalla et al. / Experimental Neurology 234 (2012) 482–487
to loss of TH-containing cell bodies and not just to loss of TH immu-
nostaining asconfirmed by counting
immunostained neurons (Yazdani et al., 2006). We now show that
the decline of DA content and TH protein in the ipsilateral striatum
is linear over the 4-week time period studied, with the loss of TH
being less than DA. There was a significant loss of striatal DA after
2 weeks of MPP+treatment, but striatal TH was not significantly re-
duced until 4 weeks of treatment. Interestingly, when caffeine was
given after 1 and 3 weeks of MPP+treatment, when striatal TH was
still minimally affected compared to DA, caffeine reduced the degen-
eration of nigral DA neurons seen after 4 weeks of MPP+treatment.
These data suggest that: (a) loss of striatal DA nerve terminal function
occurs before loss of the nigral DA neurons; (b) nigral neurodegen-
eration becomes apparent at about 3 weeks after MPP+treatment;
and (c) caffeine given at or before this time blocks the nigral neurode-
generative process without restoring the striatal nerve terminal neu-
rochemistry. Thus, after the neurodegenerative process has begun, as
indicated by the striatal DA neurochemical reductions, initiation of
caffeine treatment can still block the loss of nigral DA neurons.
Neuroprotection by caffeine in the MPP+rat model is seen pri-
marily at the level of the DA cell bodies in the SN. The reasons for
the regional differences in protection by caffeine are not known.
One possibility is that the icv route exposes the striatum to higher
MPP+concentrations than the SN and thus to a greater toxic insult.
However, pharmacological protection by several diverse compounds
is greater in the SN than in the striatum in mice treated systemically
with acute or subacute doses of MPTP. For example, pharmacological
intervention with adenosine A2Areceptor antagonists, rosglitazone
(an agonist at peroxisome proliferators-activated receptor-gamma),
or an inhibitor of monoacylglycerol lipase (which reduces brain pros-
taglandin synthesis) completely protect against loss of nigral TH im-
munostained neurons but only minimally protect the striatal DA
nerve terminals from MPTP (Dehmer et al., 2000; Nomura et al.,
2011; Pierri et al., 2005; Schintu et al., 2009; Yu et al., 2008). These
findings may indicate a region-selective effect of the drugs or alterna-
tively that the striatal DA nerve terminals are much more sensitive to
MPTP/MPP+. In mice with targeted mitochondrial damage to DA neu-
rons, loss of striatal DA markers occurs long before any loss of nigral
DA cell bodies, suggesting that the DA nerve terminals are particularly
sensitive to mitochondrial dysfunction (Pickrell et al., 2011). Inhibi-
tion of complex I of the mitochondrial electron transport chain and
hence of mitochondrial function by MPP+is a principal mechanism
underlying neurodegeneration in DA neurons (Vyas et al., 1986). Re-
cent data indicate that MPP+also damages mitochondrial transport
in DA axons (Kim-Han et al., 2011), providing another possible expla-
nation for why the DA nerve terminals are more vulnerable to MPP+-
induced damage than are the cell bodies in the SN.
Caffeine treatment provides neuroprotection to DA neurons in
the SN. Based on our preliminary data, we think that this neuroprotec-
tion is due to a diminished immune response in the SN although addi-
tional studies are required to further characterize and quantify the
glial responses. We propose this because adenosine, which is promi-
flammatory actions through activation of A2Areceptors on microglia
(Fiebich et al., 1996; Hasko et al., 2008; Orr et al., 2009; Saura et al.,
2005; Trincavelliet al., 2008). Caffeine or selective A2Areceptor antag-
onists prevent or attenuate microglia responses such as proliferation,
Fig. 2. Caffeine protects nigral DA neurons when administered prior to or during the course of neurodegeneration initiated by MPP+. Rats were infused with MPP+(75 μg/day) for
4 wks. Caffeine in the drinking water (1 g/l) was provided from the start of the infusion, or beginning 1 or 3 weeks later. (A) Results are the mean±SEM of TH+cell counts in the
left SN, with the number of rats indicated in the columns. Controls are naive untreated rats. When caffeine was given along with MPP+, or delayed by 1 or 3 weeks after beginning
MPP+infusions, there was less neurodegeneration of nigral DA neurons compared to MPP+given alone.⁎pb0.01 control vs. group treated with only MPP+. (B) Data are plotted
from individual animals, and also show the median cell count (horizontal line) for each of the 5 treatment groups.
Fig. 3. Caffeine does not alter the MPP+-induced reductions in striatal TH, DA or DA me-
tabolites. Rats were treated as described in Fig. 2. (A) Results are from the left striatum
and are presented as the % of control±SEM (naives, n=3; MPP+, n=7; simultaneous
MPP+and caffeine, n=6). Control values are (mean±SEM in ng/mg tissue): TH, 174±
6; DA, 16.7±0.3; DOPAC, 1.7±0.1; HVA, 1.6±0.1; 5HT, 0.8±0.1; 5HIAA, 1.6±0.1.
⁎pb0.05 vs. respective side of controls. (B) Results are presented as % control from
right sides of naive rats or rats treated with MPP+or MPP+and caffeine.
P.K. Sonsalla et al. / Experimental Neurology 234 (2012) 482–487
retard their recruitment to sites of injury and reduce the productionof
Moreover, chronic caffeine partially reverses the microglia activation
seen in the brains of older rats (Brothers et al., 2010). That caffeine's
protection may be mediated by its blockade of A2Areceptors is sup-
ported by findings that A2Areceptor antagonists or genetic ablation
of A2Areceptors protect nigral DA neurons and reduce neuroinflam-
mation in MPTP-treated mice (Carta et al., 2009; Chen et al., 2001;
Pierri et al., 2005; Yu et al., 2008). While the mechanism(s) by
which caffeine treatment modifies the microglia response is not fully
established, blockade of microglia A2Areceptors is a candidate target
site. Alternatively, it may be that the microglial response is not as
pronounced in the caffeine-MPP+treated rats because damage in
the SN is considerably less than in the rats treated only with MPP+,
thus leading to a much lower recruitment of microglial cells to the
SN. Additional studies are needed to sort out this “chicken and egg”
In addition to A2Areceptors located on microglia, neuronal A2Are-
ceptors may also participate in the neuroprotection afforded by caf-
feine. Stimulation of pre- and/or post-synaptic A2Areceptors in the
striatum drives activity in the striato–pallidal–subthalamic–nigral
pathway (the indirect pathway) causing nigral glutamate release
(Morelli et al., 2011). Nigral glutamate is elevated in animal PD
models due to the loss of the inhibitory actions of DA on D2 receptors
which then leave unopposed the excitatory drive by A2Areceptors on
the indirect pathway (ibid). Excessive stimulation of the nigral gluta-
matergic N-methyl-D-aspartate (NMDA) receptors located on DA
neurons can be excitotoxic, especially to metabolically compromised
DA neurons as would occur in the MPP+-treated rats. Selective
blockade of nigral NMDA receptors protects DA neurons from meta-
bolic stress, indicating the importance of nigral glutamate and excito-
toxicity to DA neurons (Zeevalk et al., 2000). In the striatum, blockade
of striatal post-synaptic A2A receptors reduces nigral glutamate
release (Morelli et al., 2011). Indeed, in mice lacking neuronal A2Are-
ceptors in the forebrain, neurotoxicity by sub-acute MPTP is attenuated
(Carta et al., 2009). Selective blockade of nigral A2Areceptors also pro-
tects DA neurons against metabolic stress, an action thought to impact
on nigral glutamate release (Alfinito et al., 2003). Thus, blocking striatal
mate release and excitotoxic damage.
Caffeine metabolites may also contribute to neuroprotection.
Paraxanthine (1,7-dimethylxanthine), which is the major metabolite,
protects DA neurons in vitro and in vivo by non-adenosine receptor
dependent mechanisms (Geraets et al., 2006; Guerreiro et al., 2008;
Xu et al., 2010). In vitro studies demonstrate paraxanthine is a potent
inhibitor of poly (ADP-ribose) polymerase-1 (PARP-1; at an IC50 of
15 μM) and a less potent activator of ryanodine receptors (Geraets
et al., 2006; Guerreiro et al., 2008).
Based on the data presented here, we propose that caffeine's neu-
roprotection in the MPP+rat model of PD is mediated, at least in part,
by blockade of A2Areceptors and an attenuation of neuroinflamma-
tion in the SN. If the effects of caffeine in our model are mediated
by A2Areceptors, then it is possible that the administration of A2Aan-
tagonists to PD patients would modify the neuropathology and slow
disease progression. Selective A2Aantagonists may be superior to caf-
feine as they would not block the neuroprotective effects of adeno-
sine as mediated through A1 receptors (Alfinito et al., 2003).
Clinical studies are currently underway to evaluate several A2A
Fig. 4. Caffeine appears to attenuate the microglia response in the SN but not in the striatum. Rats were treated as described in Fig. 2 except that caffeine treatment was started
1 week after onset of MPP+infusion. ED1 immunostaining for microglia is red. TH+immunostaining for DA neurons is green. (Upper panels) Representative photomicrograph
from one of two rats treated with only MPP+. Left side is ipsilateral to the icv infusion. Top photomicrographs are at lower magnification. ED1 immunostained cells are prominently
noted in the left side whereas few are seen in the right side. Also, note the reduction of TH+cells in the medial region of the SN on the left side as compared to the right SN. The
findings in the second rat were similar. Representative photomicrographs from one of two rats that received both caffeine and MPP+. Note that there are fewer ED1 immunostained
cells in the left SN region in the MPP+and caffeine treated rat than in the MPP+rat. Also note the preservation of TH+cells in the medial region of the SN as compared to MPP+rats.
Similar findings were observed in the second rat. (Lower panels) Caffeine does not attenuate the microglia response in the striatum. Areas shown are from sections at two different
rostral–caudal levels from one rat; the upper photomicrographs are from near the icv cannula placement region whereas the lower photomicrographs are from sections rostral to
cannula placement. Note intense ED1 staining in the left striatum of the MPP+-treated rat near the cannula placement site with much less intense staining at the more rostral site.
Note absence of ED1 staining in the right striatum. Also note loss of TH staining in the left striatum vs. the right striatum. The rat treated with MPP+and caffeine shows similar ED1
staining as in the rat treated with only MPP+, but the TH staining is somewhat more intense in this animal vs. MPP+alone. As in the MPP+treated rat, there is a paucity of ED1
staining in the right striatum.
P.K. Sonsalla et al. / Experimental Neurology 234 (2012) 482–487
receptor antagonists for their effects both on symptom relief and in
arresting disease progression (reviewed in Morelli et al., 2011).
Based on our findings with caffeine, we would predict that the A2Aan-
tagonists should slow disease progression.
This work was supported by NIH grants NS052733, NS058329 and
ES005022 and in part by the James Webb Fund of the Dallas Founda-
tion, the Dallas Area Parkinsonism Society. The authors thank Evelyn
Ho for validation of the nigral DA cell count data.
Aguiar, L.M., Nobre Jr., H.V., Macedo, D.S., Oliveira, A.A., Freitas, R.M., Vasconcelos, S.M.,
Cunha, G.M., Sousa, F.C., Viana, G.S., 2006. Neuroprotective effects of caffeine in the
model of 6-hydroxydopamine lesion in rats. Pharmacol. Biochem. Behav. 84,
Alfinito, P.D., Wang, S.P., Manzino, L., Rijhsinghani, S., Zeevalk, G.D., Sonsalla, P.K., 2003.
Adenosinergic protection of dopaminergic and GABAergic neurons against mito-
chondrial inhibition through receptors located in the substantia nigra and stria-
tum, respectively. J. Neurosci. 23, 10982–10987.
Appel, S.H., Beers, D.R., Henkel, J.S., 2009. T cell-microglial dialogue in Parkinson's dis-
ease and amyotrophic lateral sclerosis: are we listening? Trends Immunol. 31,
Bienvenu, T., Pons, G., Rey, E., Richard, M.O., d'Athis, P., Olive, G., 1990. Effect of hy-
pophysectomy on caffeine elimination in rats. Fundam. Clin. Pharmacol. 4,
Brothers, H.M.,Marchalant,Y., Wenk,
lipopolysaccharide-induced neuroinflammation. Neurosci. Lett. 480, 97–100.
Carta, A.R., Kachroo, A., Schintu, N., Xu, K., Schwarzschild, M.A., Wardas, J., Morelli, M.,
2009. Inactivation of neuronal forebrain A receptors protects dopaminergic neu-
rons in a mouse model of Parkinson's disease. J. Neurochem. 111, 1478–1489.
Chen, J.F., Xu, K., Petzer, J.P., Staal, R., Xu, Y.H., Beilstein, M., Sonsalla, P.K., Castagnoli, K.,
Castagnoli Jr., N., Schwarzschild, M.A., 2001. Neuroprotection by caffeine and
A(2A) adenosine receptor inactivation in a model of Parkinson's disease. J. Neu-
rosci. 21, RC143.
Costenla, A.R., Cunha, R.A., de Mendonca, A., 2010. Caffeine, adenosine receptors, and
synaptic plasticity. J. Alzheimers Dis. 20 (Suppl. 1), S25–S34.
Dehmer, T., Lindenau, J., Haid, S., Dichgans, J., Schulz, J.B., 2000. Deficiency of inducible
nitric oxide synthase protects against MPTP toxicity in vivo. J. Neurochem. 74,
Fiebich, B.L., Biber, K., Lieb, K., van Calker, D., Berger, M., Bauer, J., Gebicke-Haerter, P.J.,
1996. Cyclooxygenase-2 expression in rat microglia is induced by adenosine A2a-
receptors. Glia 18, 152–160.
Gandhi, K.K., Williams, J.M., Menza, M., Galazyn, M., Benowitz, N.L., 2010. Higher serum
caffeine in smokers with schizophrenia compared to smoking controls. Drug Alcohol
Depend. 110, 151–155.
Geraets, L., Moonen, H.J.J., Wouters, E.F.M., Bast, A., Hageman, G.J., 2006. Caffeine me-
tabolites are inhibitors of the nuclear enzyme poly(ADP-ribose)polymerase-1 at
physiological concentrations. Biochem. Pharmacol. 72, 902–910.
German, D.C., Manaye, K.F., 1993. Midbrain dopaminergic neurons (nuclei A8, A9, and
A10): three-dimensional reconstruction in the rat. J. Comp. Neurol. 331, 297–309.
Guerreiro, S., Toulorge, D., Hirsch, E., Marien, M., Sokoloff, P., Michel, P.P., 2008. Para-
xanthine, the primary metabolite of caffeine, provides protection against dopami-
nergic cell death via stimulation of ryanodine receptor channels. Mol. Pharmacol.
Hasko, G., Linden, J., Cronstein, B., Pacher, P., Hasko, G., Linden, J., Cronstein, B., Pacher,
P., 2008. Adenosine receptors: therapeutic aspects for inflammatory and immune
diseases. Nat. Rev. Drug Discovery 7, 759–770.
Hirsch, E.C., Hunot, S., 2009. Neuroinflammation in Parkinson's disease: a target for
neuroprotection? Lancet Neurol. 8, 382–397.
Hornykiewicz, O., 1979. The neurobiology of dopamine. In: Horn, A.S., Korf, J.,
Westerink, B.H.C. (Eds.), Brain Dopamine in Parkinson's Disease and Other Neuro-
logical Disturbances. Academic Press, pp. 633–654.
Joghataie, M.T., Roghani, M., Negahdar, F., Hashemi, L., 2004. Protective effect of caf-
feine against neurodegeneration in a model of Parkinson's disease in rat: behavior-
al and histochemical evidence. Parkinsonism Relat. Disord. 10, 465–468.
Kachroo, A., Irizarry, M.C., Schwarzschild, M.A., 2010. Caffeine protects against com-
bined paraquatand maneb-induced dopaminergicneurondegeneration. Exp.Neurol.
Kalda, A., Yu, L., Oztas, E., Chen, J.-F., 2006. Novel neuroprotection by caffeine and aden-
osine A(2A) receptor antagonists in animal models of Parkinson's disease. J. Neu-
rol. Sci. 248, 9–15.
Kim-Han, J.S., Antenor-Dorsey, J.A., O'Malley, K.L., 2011. The Parkinsonian mimetic,
MPP+, specifically impairs mitochondrial transport in dopamine axons. J. Neurosci.
31 (19), 7212–7221.
Morelli, M., Carta, A.R., Kachroo, A., Schwarzschild, M.A., 2011. Pathophysiological roles
for purines: adenosine, caffeine and urate. Prog. Brain Res. 183, 183–208.
Nomura, D.K., Morrison, B.E., Blankman, J.L., Long, J.Z., Kinsey, S.G., Marcondes, M.C.G.,
Ward, A.M., Hahn, Y.K., Lichtman, A.H., Conti, B., Cravatt, B.F., 2011. Endocannabi-
noid hydrolysis generates brain prostaglandins that promote neuroinflammation.
Science 334, 809–813.
Orr, A.G., Orr, A.L., Li, X.-J., Gross, R.E., Traynelis, S.F., 2009. Adenosine A(2A) receptor
mediates microglial process retraction. Nat. Neurosci. 12, 872–878.
Paxinos, G., Watson, C., 1986. The rat brain in stereotaxic coordinates-Second Edition.
Academinc Press, Inc., Orlando, FL.
Pickrell, A.M., Pinto, M., Hida, A., Moraes, C.T., 2011. Striatal dysfunctions associated
with mitochondrial DNA damage in dopaminergic neurons in a mouse model of
Parkinson's disease. J. Neurosci. 31, 17649–17658.
Pierri, M., Vaudano, E., Sager, T., Englund, U., 2005. KW-6002 protects from MPTP in-
duced dopaminergic toxicity in the mouse. Neuropharmacology 48, 517–524.
Rebola,N.,Simões, A.P., Canas,P.M.,Tomé, A.R., Andrade,G.M.,Barry,C.E., Agostinho,P.M.,
Lynch, M.A., Cunha, R.A., 2011. Adenosine A2A receptors control neuroinflammation
and consequent hippocampal neuronal dysfunction. J. Neurochem. 117, 100–111.
Saura, J., Angulo, E., Ejarque, A., Casado, V., Tusell, J.M., Moratalla, R., Chen, J.F.,
Schwarzschild, M.A., Lluis, C., Franco, R., Serratosa, J., 2005. Adenosine A2A receptor
stimulation potentiates nitric oxide release by activated microglia. J. Neurochem.
Schintu, N., Frau, L., Ibba, M., Caboni, P., Garau, A., Carboni, E., Carta, A.R., Schintu, N.,
Frau, L., Ibba, M., Caboni, P., Garau, A., Carboni, E., Carta, A.R., 2009. PPAR-
gamma-mediated neuroprotection in a chronic mouse model of Parkinson's disease.
Eur. J. Neurosci. 29, 954–963.
Stavric, B., Klassen, R., Watkinson, B., Karpinski, K., Stapley, R., Fried, P., 1988. Variabil-
ity in caffeine consumption from coffee and tea: possible significance for epidemi-
ological studies. Food Chem. Toxicol. 26, 111–118.
Trincavelli, M.L., Melani, A., Guidi, S., Cuboni, S., Cipriani, S., Pedata, F., Martini, C., 2008.
Regulation of A(2A) adenosine receptor expression and functioning following per-
manent focal ischemia in rat brain. J. Neurochem. 104, 479–490.
Vyas, I., Heikkila, R.E., Nicklas, W.J., 1986. Studies on the neurotoxicity of 1-methyl-4-
phenyl-1,2,3,6-tetrahydropyridine: inhibition of NAD-linked substrate oxidation
by its metabolite, 1-methyl-4-phenylpyridinium. J. Neurochem. 46, 1501–1507.
Xu, K., Xu, Y.H., Chen, J.F., Schwarzschild, M.A., 2010. Neuroprotection by caffeine:
time course and role of its metabolites in the MPTP model of Parkinson's disease.
Neuroscience 167, 475–481.
Yazdani, U., German, D.C., Liang, C.L., Manzino, L., Sonsalla, P.K., Zeevalk, G.D., 2006. Rat
model of Parkinson's disease: chronic central delivery of 1-methyl-4-phenylpyridi-
nium (MPP+). Exp. Neurol. 200, 172–183.
Yu, L., Shen, H.-Y., Coelho, J.E., Araujo, I.M., Huang, Q.-Y., Day, Y.-J., Rebola, N., Canas, P.M.,
Rapp, E.K., Ferrara, J., Taylor, D., Muller, C.E., Linden, J., Cunha, R.A., Chen, J.-F., 2008.
Adenosine A2A receptor antagonists exert motor and neuroprotective effects by dis-
tinct cellular mechanisms. Ann. Neurol. 63, 338–346.
Zeevalk, G.D., Manzino, L., Sonsalla, P.K., 2000. NMDA receptors modulate dopamine
loss due to energy impairment in the substantia nigra but not striatum. Exp. Neurol.
Zeevalk, G.D., Manzino, L., Sonsalla, P.K., Bernard, L.P., 2007. Characterization of intra-
cellular elevation of glutathione (GSH) with glutathione monoethyl ester and
GSH in brain and neuronal cultures: relevance to Parkinson's disease. Exp. Neurol.
P.K. Sonsalla et al. / Experimental Neurology 234 (2012) 482–487