We used multimodal neuroimaging to study how individual differences in nigrostriatal degeneration, as quantified by DAT scan, influenced BOLD responses to apomorphine, a potent and fast-acting dopamine agonist.
We found that DAT-BPND levels guided the striatal and PFC responses to apomorphine in PD patients during all working-memory loads. In particular, the apomorphine effect in PD patients with intermediate dopaminergic depletion was opposite to that found in patients with higher and lower dopaminergic depletion (i.e., patients with longer and shorter disease duration, respectively).
Consistent with some previous data, apomorphine tended to impair behavioral performance during working memory in all PD patients, regardless of the residual dopamine level (Ruzicka et al. 1994
; Costa et al. 2003
). However, only a trend effect of treatment was found for accuracy (P
= 0.08), and this may depend on our smaller sample size (n
= 12) compared with those commonly used to assess the behavioral effects of dopaminergic drugs (e.g., n
~20) (Costa et al. 2009
). Further studies with larger sample sizes are necessary to identify the precise amount of apomorphine stimulation that leads to cognitive dysfunctions. We also acknowledge that additional research is necessary to investigate how apomorphine influences cognition in PD patients with greater disease severity and longer disease duration that those reported here.
Nonetheless, it is important to point out that this study was designed to explore how apomorphine influenced working memory in PD at a neural rather than behavioral level. To this end, fMRI is a sensitive tool which can reveal subtle effects of drugs on brain responses, even before the occurrence of noticeable behavioral findings. In fact, apomorphine modulated neural responses independently from its behavioral effects, and this was demonstrated by the stability of the results when fMRI analyses assessing the main effect of treatment were repeated including RT and accuracy as variables of no interest. Overall, our data extend the knowledge about the neural mechanisms of apomorphine in PD by showing that this potent dopamine agonist increased striatal response and reduced SFG activation during working memory.
The enhanced striatal response to apomorphine might depend on the super-sensitivity of postsynaptic D2 receptors. There is evidence, in animal models of PD, that lesioning dopaminergic neurons causes reduced DAT-BPND
values, increased D2 receptor binding, and increased BOLD response to apomorphine in the striatum (Nguyen et al. 2000
). Comparative research has also suggested that this enhanced striatal BOLD response to apomorphine may indirectly reflect the state of postsynaptic D2 receptors (i.e., sensitivity and/or number) (Zhang et al. 2000
). Although the sensitivity and/or number of D2 receptors were not measured in this study, we speculate that the progressive nigrostriatal degeneration in PD induced a D2 receptor super-sensitivity state which, in turn, guided the abnormal striatal responses to apomorphine during all working-memory loads. However, it remains to be explained why we found an inverted-U-shaped relation between DAT-BPND
values and the brain responses to apomorphine. We hypothesize at least two, not mutually exclusive, explanations for this finding.
First, there is clear in vitro evidence that the number of striatal D2 receptors follows an inverted-U curve after lesioning dopaminergic neurons (i.e., the number of receptors continue to rise until the ~100th day after the dopaminergic damage; next, it gradually reverts to normal levels, which are reached after ~500 days in total) (Todd et al. 1996
Second, an inverted-U-shaped relation between D2 receptors number and/or sensitivity and disease progression has been also observed in vivo, in PD patients at different stages (Antonini et al. 1994
; Ichise et al. 1999
). In particular, patients with initial or advanced PD display normal D2 receptor number and/or sensitivity, while patients with intermediate disease progression show increased D2 receptors number and/or sensitivity (Antonini et al. 1994
; Ichise et al. 1999
). Other studies have also demonstrated that D2 receptors in PD return to normal levels after treatment and remain stable over the course of the disease (Guttman and Seeman 1985
; Guttman et al. 1986
). This variability may depend on differential levels of exposure to chronic dopaminergic therapy that might normalize the receptor number and/or sensitivity (i.e., patients with advanced PD would have been more exposed to chronic dopaminergic therapy when compared with patients with intermediate PD stages) (Alexander et al. 1993
). However, it is also possible that molecular mechanisms independent from drug therapy intervene to reduce the D2 receptor number and/or sensitivity over time. There is indeed evidence that the number and/or sensitivity of D2 receptors decreases in Parkinsonian monkeys with chronic nigrostriatal lesion even if they did not receive dopaminergic therapy (Decamp et al. 1999
). Nonetheless, differences in treatment duration in our PD patients may have played a role in determining the sensitivity of D2 receptors and thus the heterogeneity of their brain responses to apomorphine.
It is also noteworthy that apomorphine decreased activation of the SFG, a specific PFC region linked to stimulus manipulation during working memory (du Boisgueheneuc et al. 2006
). SFG is linked to basal-ganglia circuits involved in filtering irrelevant information during working memory (Moustafa et al. 2008
; Baier et al. 2010
); hence, apomorphine might indirectly alter the SFG function via dopaminergic receptors in the striatum. Our finding that DAT striatal levels modulated BOLD responses to apomorphine in SFG during all working-memory loads support this hypothesis. Alternatively, apomorphine might influence dopaminergic receptors on cortical neurons within the SFG itself. This possibility is supported by previous research in behaving monkeys showing that excessive levels of D1 receptor stimulation reduce delay-related firing of PFC neurons and erode the tuning of their responses during working memory (Vijayraghavan et al. 2007
). In line with a recent staging model of executive dysfunctions and mental fatigue in PD (de la Fuente-Fernandez 2012
), it is also possible that apomorphine stimulation “overdosed” the direct VTA-PFC dopaminergic pathways via D4 receptors, a D2 receptor family expressed in the neocortex and implicated in the pathophysiology of a range of neuropsychiatric disorders (Oak et al. 2000
; Wang et al. 2002
Two other PFC areas (i.e., the inferior frontal gyrus, IFG, and the dACC) showed a significant modulation by the striatal DAT levels and apomorphine therapy. The IFG has been consistently associated with response inhibition, a key neuropsychological function during working-memory tasks that require response inhibition (Aron and Poldrack 2005
; Aron 2011
). In contrast, the dACC has been linked to error and conflict monitoring, two other fundamental processes to execute a wide range of cognitive paradigms (van Veen and Carter 2002
; Hester et al. 2004
; Nee et al. 2011
). It is also important to note that apomorphine effects related to DAT levels were consistently observed for all working-memory loads (i.e., high-, medium-, and low-working-memory loads). Overall, this suggests a broad influence of apomorphine on PFC physiology that is independent from cognitive demands.
Finally, two additional points should be discussed. First, we did not assess the brain effects of apomorphine in our HCs because of the emetic of the high doses of drug employed in PD patients (Bowron 2004
; LeWitt 2004
). Furthermore, even if we had used apomorphine at low doses in HCs, the interpretation of the group by treatment interaction would have been limited by the fact that apomorphine at those doses activates presynaptic rather than postsynaptic D2 receptors (the latter is the case at the high doses employed in PD). Nonetheless, we enrolled HCs to examine the main effect of group (PD-Off, HCs). This analysis showed that PD patients Off-medication, relative to controls, displayed greater activations in the cuneus, precuneus, and thalamus, a group of regions previously implicated in attentional and working-memory processes (LaBar et al. 1999
). BOLD hyperactivations associated with normal behavioral performances have been previously described in PD and other neurological patients and may represent functional compensations and brain plasticity effects (Passamonti et al. 2009
; Hughes et al. 2010
). Alternatively, they may result from reduced “focusing” within working memory and attentional circuits (Mattay et al. 2002
; Helmich et al. 2009
Second, it could be argued that our results were partially driven by apomorphine effects on other neurochemical systems involved in arousal (i.e., apomorphine also modulates noradrenergic receptors, although to a lesser extent than dopaminergic ones) (LeWitt 2004
). Although this possibility cannot be completely ruled out, our patients did not report significant changes in the arousal state after apomorphine injection. Furthermore, a medication by task interaction was found in the dACC, demonstrating that at least part of the neural effects induced by apomorphine were specific for high-load working-memory trials.
In conclusion, our research offers strong support in demonstrating that the combination of fMRI and quantitative DAT imaging predicted individual differences in brain responses to a dopaminergic challenge. In the future, additional studies with larger sample sizes may open new possibilities for developing multiparametric brain markers (e.g., diffusional kurtosis imaging [Giannelli et al. 2012
]) that can be used to personalize pharmacological therapies according to the specific needs of PD patients.