Dopamine transporter relation to levodopa-derived synaptic dopamine in a rat model of Parkinson’s: an in vivo imaging study


Address correspondence and reprint requests to Vesna Sossi, PhD, Pacific Parkinson’s Research Centre, Room M37, Purdy Pavilion, 2221 Wesbrook Mall, Vancouver, BC, Canada V6T 2B5. E-mail:


Studies showed that the dopamine (DA) transporter (DAT) modulates changes in levodopa-derived synaptic dopamine levels (Δ(DA)) in Parkinson’s disease (PD). Here we evaluate the relationship between DAT and Δ(DA) in the 6-hydroxydopamine model of Parkinson’s disease to investigate these mechanisms as a function of dopaminergic denervation and in relation to other denervation-induced regulatory changes. 27 rats with a unilateral 6-hydroxydopamine lesion (denervation ∼20–97%) were imaged with 11C-dihydrotetrabenazine (VMAT2 marker), 11C-methylphenidate (DAT marker) and 11C-raclopride (D2-type receptor marker). For denervation <75%Δ(DA) was significantly correlated with a combination of relatively preserved terminal density and lower DAT. For denervation <90%, Δ(DA) was significantly negatively correlated with DAT with a weaker dependence on VMAT2. For the entire data set, no dependence on pre-synaptic markers was observed; Δ(DA) was significantly positively correlated with 11C-raclopride binding-derived estimates of DA loss. These findings parallel observations in humans, and show that (i) regulatory changes attempt to normalize synaptic DA levels (ii) a lesion-induced functional dependence of Δ(DA) on DAT occurs up to ∼ 90% denervation (iii) for denervation < 75% relative lower DAT levels may relate to effective compensation; for higher denervation, lower DAT levels likely contribute to oscillations in synaptic DA associated with dyskinesias.

Abbreviations used



dopamine transporter








Parkinson’s disease


positron emission tomography



Parkinson’s disease (PD), which affects approximately 300/100 000 of the general population (de Lau and Breteler 2006), is characterized by a progressive degeneration of the dopaminergic system leading to increasingly severe dopamine (DA) deficiency resulting in debilitating motor and often cognitive impairment. It is known that clinical symptoms generally occur when at least 50% of dopaminergic neurons have died; this implies the existence of a pre-clinical period during which disease-induced neurochemical changes take place (Adams et al. 2005). Increase in DA turnover in humans and animal models of PD (Zigmond et al. 1990; Sossi et al. 2002) and DA transporter (DAT) down-regulation (Lee et al. 2000; Adams et al. 2005) have been identified as possible early compensatory changes. It is also known that the majority of PD patients develop treatment-related motor complications within 5–10 years of treatment. While the neurochemical changes underlying motor complications are not completely understood, current knowledge favors a synergistic combination of altered DA release patterns and post-synaptic receptor sensitization in response to pulsatile DA stimulation (de la Fuente-Fernández et al. 2004a, Bibbiani et al. 2005; Cenci and Lundblad 2006) Experimental evidence suggests that an increase in both DA turnover and levodopa-induced changes in synaptic DA levels (Δ(DA)) is associated with increased risk of motor complications (de la Fuente-Fernández et al. 2001b, 2004a,b;Pavese et al. 2006). A link between an increase in DA turnover and/or Δ(DA) and lower DAT levels has also been recently suggested in human studies including symptomatic patients (Sossi et al. 2007a) as well as in 6-hydroxydopamine (6-OHDA) lesioned rats with dopaminergic denervation > 60% (Lee et al. 2008). These results suggest that once disease is present, the functional role of DAT is to preserve DA in the terminals and to reduce oscillations in synaptic DA levels that contribute to a pulsatile stimulation of the post-synaptic receptors. Indeed, in a recent clinical human study, down-regulated DAT levels have been directly linked to the occurrence of dyskinesia (Troiano et al. 2009).

The same mechanisms (i.e., an increase in Δ(DA) and/or DA turnover and regulatory DAT changes) seem to be involved in both delaying the onset of clinical symptoms in early disease as well as increasing the risk of treatment-induced motor complications as disease progresses. It is thus of interest to explore the relationship between DAT expression and Δ(DA) as a function of severity of dopaminergic denervation avoiding possible confounds of treatment, such as a reported differential responses of DAT to levodopa or DA agonists (Parkinson’s Study Group 2002, 2004), and to relate such changes to other denervation-induced effects, such as decrease of synaptic DA levels. In this study, we used a unilateral 6-OHDA-lesion rat model of PD and positron emission tomography (PET) imaging to investigate the relation between these mechanisms over a range of denervation corresponding to the pre-symptomatic to moderate/advanced symptomatic PD stage.

The same three PET tracers used in similar human studies (Sossi et al. 2007a) were used here: (+) 11C-dihydrotetrabenazine (DTBZ, a vesicular monoamine transporter type 2 (VMAT2) marker) to estimate the degree of dopaminergic denervation, 11C-methylphenidate (MP) to evaluate DAT expression and 11C-raclopride (RAC, a D2-type receptor marker) to evaluate Δ(DA) as well as denervation-induced changes in RAC binding, taken to be an estimate of synaptic DA loss. For a subset of rats data obtained from the pre-synaptic PET markers were also compared to post-mortem autoradiography with DTBZ (Strome et al. 2006) and [3H]WIN 35,428 (a DAT marker) to confirm the accuracy of the in vivo denervation assessment (= 25 for VMAT2 and for DAT). [3H]WIN 35 428 was used for DAT autoradiography because of its accepted use as DAT marker and ready availability.

Lesion-induced changes in baseline RAC binding were used for two purposes, the first of which was a relative estimation of baseline synaptic DA levels. RAC binding has been indeed shown to be sensitive to synaptic DA levels (de la Fuente-Fernández et al. 2001a,Piccini et al. 2003; Pavese et al. 2006). An increase in baseline RAC binding has been shown to occur in early symptomatic disease in untreated human subjects (Rinne et al. 1995) and in symptomatic MPTP lesioned non-human primates (Doudet et al. 2002), while no change has been observed in lesioned, but asymptomatic non-human primates. It is thus reasonable to assume that as long as synaptic DA levels are sufficient to maintain adequate motor function, RAC binding remains at control levels, while an increase in baseline RAC binding can be directly (by competition) or indirectly (by receptor regulation) related to lower endogenous synaptic DA levels (Doudet et al. 2002; Schiffer et al. 2005). The second purpose was to compare denervation-induced baseline RAC binding changes to those observed with the pre-synaptic markers to gain some insight into their respective dependence on the level of denervation.

Materials and methods

Subjects and animal lesioning

Twenty-seven 6-OHDA unilaterally lesioned male 400–450 g Sprague-Dawley rats were included in this study. The animals were anesthetized with isofluorane and placed in the stereotaxic headholder with the skull flat between Lambda and Bregma. A 2% solution of 6-OHDA hydrobromide (8 μg in 4 μL 0.05% ascorbic acid in saline; Sigma, St Louis, MO, USA) was infused at two sites along the medial forebrain bundle [site 1: AP −2.8 mm, ML −1.8 mm, DV −8.0 (all from Bregma); site 2: AP −4.7 mm (from Bregma), ML −1.5 mm (from midline), DV −7.9 mm (from skull) according to Paxinos and Watson (Paxinos and Watson 1998)]. Approximately 30 min prior to the 6-OHDA infusion the subjects were given desipramine (25 mg/kg i.p) to protect noradrenergic fibers (Breese and Traylor 1971; Kelly and Iversen 1976; Whishaw et al. 1994). The 6-OHDA solution was delivered at a rate of 1 μL/min using a 50 μL Hamilton syringe through a 26-gauge needle. Following the injection of 6-OHDA, the needle was left in place for an additional 3 min prior to withdrawal. After surgery, rats were monitored and placed in a recovery incubator for 2–4 h, and then returned to their home cage after full recovery from the anesthesia. The animals were allowed 3–4 weeks recovery before undergoing the PET studies. All studies were approved by the UBC ethics board.

PET scanning

Each rat underwent four scans: one DTBZ, one MP and two RAC scans, the first at baseline (RAC0) and the second 45 min after I.P. administration of levodopa/benserazide (50 mg/kg, 15 mg/kg prepared) (RAC1).

The scanning procedure was identical for all three tracers. The rats were anesthetized with a 2% isofluorane gas mixture. All studies were performed on a Siemens/Concorde Focus 120 (Kim et al. 2007). Following a 6 min transmission scan with a 57Co source, a 1 h emission scan was performed starting at tracer injection. Data were histogrammed into 6 × 30, 2 × 60, 5 × 300, 2 × 450 and 2 × 480 s frames and reconstructed using Fourier rebinning and filtered backprojection after applying normalization, scatter, attenuation and sensitivity corrections. The specific activity for each tracer was kept high enough to ensure negligible transporter or receptor occupancy (SA > 1100 nCi/pmole) at an injection dose of 100 μCi/100 g of animal weight (Schiffer et al. 2005; Sossi et al. 2007b).

Image and data analysis

Regions of interest (ROI) were placed on both striata and the cerebellum. Since the lesioned striatum and the cerebellum were in most cases not clearly identifiable, ROI placement was guided indirectly by a rat brain atlas, with striatal and cerebellar regions clearly delineated (Sossi et al. 2007b). Using the aligned atlas, an ROI (area 9.00 mm2) was placed on each striatum on three consecutive axial planes (for a total of three ROIs per striatum). In addition, a reference ROI (area 13.50 mm2) was placed on the cerebellum on three consecutive axial planes. All ROIs were then transferred to the PET image and adjusted when necessary, while keeping the relative locations of the ROIs constant, to capture the maximum activity in the striatum or the most uniform area of the cerebellum. Time activity curves were generated for each ROI and averaged across all three planes. Tissue input binding potential BPND (Innis et al. 2007) values were calculated from the average time activity curves by applying the simplified reference tissue method (Lammertsma and Hume 1996; Gunn et al. 1997) to each striatum separately.

Dopaminergic denervation

Data from the DTBZ scan were used to determine the level of dopaminergic denervation. A good correlation has previously been found between DTBZ binding and tyrosine hydroxylase activity in the striatum, indicating that DTBZ binding is a good measure of striatal terminal density (Masuo et al. 1990). Dopaminergic denervation was estimated as (1-BPND_DTBZ_L/BPND_DTBZ_C), with the subscripts L and C referring to the lesioned and control side respectively. A difference in DTBZ BPND between the control and lesioned side was previously demonstrated to reflect a change in the maximum free VMAT2 density Bmax and not ligand-transporter affinity KD, justifying this definition of denervation (Sossi et al. 2007b).

A possible confound might arise from the fact that DTBZ is not an exclusive marker for dopaminergic activity, but is also sensitive to serotonergic terminals, which are unaffected by 6-OHDA lesion (Zigmond et al. 1987). The contribution of such neurons has been found to be approximately 5% of total DTBZ binding (Masuo et al.1990). While entirely negligible for an intact dopaminergic system, it potentially affects the estimate of a more severe denervation by only a small amount. The highest level of denervation estimated by PET in this rat population was 96%, with a corresponding value of 98% estimated from autoradiography; such high denervation estimates confirm that the possible confound introduced by DTBZ binding to serotonergic terminals is indeed very small. Furthermore, since DAT is not present on serotonergic terminals and MP has been found to selectively bind to DAT (Deutsch et al.1996, Meltzer et al. 2003), we also visually examined the ratio between BPND_MP and BPND_DTBZ for the lesioned and control side with the reasoning that large decreases from normal values would potentially be indicative of an overestimation of dopaminergic innervation by DTBZ binding.

Denervation-induced changes in baseline RAC binding – estimate of DA loss

Denervation-induced increase in baseline RAC binding was estimated as the ratio between BPND_RAC_0_L and BPND_RAC_0_C As previously discussed, these changes are either a direct or indirect consequence of lower synaptic DA levels and can thus be considered as a surrogate measure of synaptic DA loss. In order to satisfy the statistical requirements of the analysis we defined DA loss as:


Change in synaptic DA levels (Δ[DA])

An estimate of the levodopa-derived change in synaptic DA levels was obtained by comparing the BPND values obtained in the first (baseline) RAC scan (RAC_0) (BPND_RAC_0) and those obtained in the second RAC scan (RAC_1) performed after administration of levodopa/benserazide (BPND_RAC_1). Such estimate can be performed according to the accepted definition (Innis et al. 2007):


In order to account for the statistical characteristics of the mathematical expression and the non-Gaussian distribution of the data, the statistical analysis was however performed on the variable:


and log(BPND_RAC_0) was used as an explanatory variable in the regression. Δ[DA] was estimated separately for the control and lesioned side.


After PET scans were complete, rats were killed by decapitation, their brains were extracted and frozen in isopentane, and stored at −80°C. Brains were cut into 16 μm coronal slices (equivalent orientation to PET transverse) at five anterior-to-posterior positions, three redundant slices per position, to representatively sample the whole striatum as seen in PET. For DTBZ autoradiography separate sets of slices were incubated in 5 nM 11C-DTBZ solution or 1 μM unlabeled tetrabenazine and 5 nM 11C-DTBZ combined solution, in order to measure total and non-specific binding, respectively. After drying, slides and standards were placed against radiosensitive phosphor screens for several hours, and the screens were then read with a Cyclone storage phosphor system, producing autoradiographic images of DTBZ binding. Activity standard curves were created by serially diluting an initial tracer solution of a known radioactivity concentration so as to cover the expected tissue radioactivity concentration range and dropping a fixed volume of each solution on a test plate. The test plate was placed on the phosphor imaging system and the measured light intensity was calibrated against the radioactivity of the solutions. The standard curves and the measured tracer specific activity values were used for the conversion of the light intensity as measured by the phosphor imaging system to pmol/mL (Strome et al. 2006). Care was taken to select brain slices to be consistent with PET image planes included in ROI averages. A similar procedure was used for [3H]WIN35,428 binding, where the standard curves were obtained from commercially available samples available [3H]microscales (Amersham, Buckinghamshire, UK). Denervation severity (for DTBZ) and denervation-induced reduction in[3H]WIN35,428 binding to DAT were estimated as (1-bindingL/bindingC)*100, analogous to the PET studies.

Statistical analysis

DA loss

The baseline RAC binding-derived estimate of synaptic DA loss (see Materials and methods) was first stepwise regressed on log(BPND_MP) and log(BPND_DTBZ) to identify effects most significantly associated with the loss of synaptic DA.


In order to account for the intrinsic correlation between DA loss and Δ(DA) because of the common log(BPND_RAC_0) term in their definition, DA loss and Δ(DA) were separately regressed on log(BPND_RAC_0) and residuals were calculated. Stepwise multiple regression was then performed between residuals from the regression between Δ[DA] and log(BPND_RAC_0) (dependent variable),and log(BPND_DTBZ), log (BPND_MP), log (BPND_RAC_) and residuals of the regression between DA loss and (BPND_RAC_0) (independent variables) to identify the mechanisms most significantly associated with changes in synaptic DA levels. Although the statistical analysis was performed using residuals, the quantities DA loss and Δ(DA) will be referred to in the results sections, with the understanding that the significance levels reported are derived from the statistical analysis using residuals.

As visual inspection of the data showed a relationship between increased baseline RAC binding in the lesioned side compared with control side and denervation (see Results), the same analysis was performed on a subset of rats with denervation levels lower than 75% (= 17). Likewise, after a visual inspection of the ratio between BPND_MP_L and BPND_DTBZ_L, the analysis was performed on the subset of rats with denervation levels lower than 90% (see Results) (= 22).

Analysis was performed for the lesioned and the control striatum separately.


Denervation induced changes in baseline RAC binding

The ratio of lesioned to control side baseline RAC binding as a function of degree of denervation is shown in Fig. 1. The BPND_RAC_0_L/BPND_RAC_0_C ratio remained close to 1 (indicating no substantial difference between the two sides) until approximately 75% denervation, at which point a sharp increase was observed. These results suggest that compensatory changes are effective at maintaining adequate synaptic DA levels until approximately 75% of denervation is reached. In the regression analysis, we thus considered the subset of rats with denervation < 75% in an additional separate analysis.

Figure 1.

 Top: Lesion-induced changes in RAC binding, defined as the ratio between BPND_RAC_0_L and BPND_RAC0_C as a function of dopaminergic denervation determined from DTBZ data. Bottom: BPND_MP/BPND_DTBZ ratio for the control (o) and lesions side (*). On the lesioned side a sharp increase is observed for denervation levels greater than approximately 90%.

Ratio between BPND_MP and BPND_DTBZ

Figure 1 shows the BPND_MP to BPND_DTBZ ratio for the control and lesioned side. The ratio is similar between the two sides and fairly constant until greater than 90% denervation, after which it increases sharply on the lesioned side. While this does not completely eliminate the potential contribution of serotonergic terminals to DTBZ binding, it shows that significant overestimation of VMAT2 is highly unlikely. Since a sharp increase of this ratio might be indicative of other mechanisms not investigated in this study, we additionally performed the statistical analysis on the rats with denervation levels lower than <90% separately.

Dopamine loss

The results of the stepwise multiple regression between DA loss and log(BPND_DTBZ_L) and log(BPND_MP_L) on the entire data set identified log(BPND_DTBZ_L) as the only significant explanatory variable (< 0.0001, negative correlation). When the entire data set is considered, the degree of DA loss is thus directly proportional to the degree of denervation (Figure 2). Almost identical results are observed when only those animals with denervation <90% were included in the analysis (= 0.0003). As expected (see Fig. 1), when the analysis was limited to those rats with denervation lower than 75%, no correlation between DA loss and log(BPND_DTBZ_L) was found.

Figure 2.

 Top: Dopamine loss as estimated from the baseline RAC scan as a function of log(BPND_DTBZ_L) for the entire data set. The line is a result of the linear regression. Bottom: Levodopa-derived change in synaptic DA levels Δ(DA) as a function of DA loss for the entire data set. While the significance level used for the interpretation of the results was that obtained by using residuals (see text, Statistical analysis) here we are showing the relationship between the variables themselves for more intuitive illustrative purposes. The line is a result of the linear regression performed on the variables and the corresponding correlation coefficients and significance levels are shown.

Change in synaptic DA levels (Δ[DA])

No significant relation between Δ[DA] and any other variables was observed for the control side. No further results for this side will be reported. Since all quantities refer to the lesioned side, the subscript L will be omitted.

When the entire data set was considered the stepwise multiple regression identified DA loss as the only significant explanatory variable (positive correlation, = 0.004, when using residuals) and Figure 2 shows the corresponding relation between the variables. This observation supports the notion that levodopa-induced changes primarily serve to normalize synaptic DA levels, attempting in that way to reach normal dopaminergic neurotransmission (de la Fuente-Fernández et al. 2004a).

When the analysis was limited to those animals with denervation lower than 90% the stepwise multiple regression analysis identified a significant negative dependence of Δ(DA)on log(BPND_MP) (< 0.036) and an almost significant positive dependence on log(BPND_DTBZ) (= 0.073). The latter result, which may appear paradoxical, could imply that for the same degree of DA loss, levodopa-induced changes in synaptic DA levels are dependent on free VMAT2 site availability (see Fig. 2 in de la Fuente-Fernández et al. 2004a.).

Finally, limiting the analysis to rats with denervation < 75%, the analysis identified a significant negative correlation with log(BPND_MP) (= 0.003) (Fig. 3) and positive with log(BPND_DTBZ) (= 0.01).

Figure 3.

 Δ(DA) adjusted for the explanatory variables using the coefficients obtained from the stepwise multiple regression as a function of log(BPND_MP_L) for the subset of rats with dopaminergic denervation levels lower than 75%. The line is a result of the multiple linear regression performed directly on the variables for illustrative purposes, with overall r2 = 0.52 and < 0.008 for the dependence of the adjusted Δ(DA) on log(BPND_MP_L). As explained in the text, the significance level used for the interpretation of the results was that obtained using residuals.

Comparison of PET results with autoradiographic binding

The comparison of the denervation levels estimated with PET and autoradiography is shown in Fig. 4, indicating that in vivo PET measures of denervation agree well with post-mortem assessments. In particular, a dopaminergic denervation assessed by PET to be 96% was determined by autoradiography to be 98%. In addition to confirming the in-vivo measurement with post-mortem binding, these results also indicate that potential instrumentation-related limitations, such as partial volume effects, do not greatly impact the accuracy of the imaging results. Likewise, the autoradiographic ([3H]WIN35,428) and PET (MP) measures of denervation-induced reduction of binding to DAT are very well correlated (Fig. 4). Figure 5 shows an example of PET and autradiographic images obtained for the same animal.

Figure 4.

 Top: Comparison of dopaminergic denervation as estimated by PET and autoradiographic binding. Bottom: Comparison of binding reduction to DAT reduction as estimated by PET (MP) and autoradiography ([3H]WIN35,428). Values are expressed as percentages.

Figure 5.

 PET and autoradiographic transverse images of a medial slice through a lesioned rat striatum with an estimated denervation level of approximately 50%. (a) [11C]DTBZ PET; (b) [11C]MP PET; (c) [11C]DTBZ autoradiography; (d) [3H]WIN autoradiography.


This study represents, to our knowledge, the first in vivo multi-tracer study evaluating levodopa-derived changes in synaptic DA levels and comparing them to changes in pre-synaptic and synaptic markers for a wide range of denervation, ranging from 20% to almost complete denervation. A most useful approach to the interpretation of these results is to consider the difference between BPND_RAC0_L and BPND_RAC0_C as a reflection of the lesion-induced loss of endogenous synaptic DA levels, while Δ(DA) as a measure of the capacity of the lesioned system to release DA in response to exogenous levodopa. Likewise, it is useful to examine how the nature of the relationships between the variables may be affected by the degree of denervation, with a threshold effect on Δ(DA) seen at denervation levels of 75%, based on an examination of the raw data. The examination of the entire data set provides an opportunity to examine the consistent overall effects of the dopaminergic denervation; on the other hand, the analysis of isolated subsets provides an opportunity to identify mechanisms that are most relevant at specific ranges of denervation, and thus, by extrapolation, at particular stages of human PD.

Keeping this approach in mind, the analysis of the entire data set indicates that DA loss was significantly correlated with log(BPND_DTBZ) and thus degree of denervation. This finding indicates that most of the endogenously derived synaptic DA originates from the dopaminergic terminals even in the case of severe denervation. On the other hand, the relationship between Δ(DA) and dopaminergic denervation was not significant, despite a significant positive correlation of Δ(DA) with the degree of synaptic DA loss. Interpretation of this finding suggests that (i) throughout the entire range of denervation, the primary response of the system to exogenously administered levodopa is an attempt to normalize synaptic DA levels (de la Fuente-Fernández et al. 2004a) and (ii) at high levels of denervation, the engagement of other DA synthesis and release mechanisms might become relevant, as suggested by others (Carta et al. 2007; Lee et al. 2008).

No substantial loss of synaptic DA was observed for denervation levels lower than 75% indicating that pre-synaptic compensatory changes are effective at maintaining adequate synaptic DA levels up to this level of denervation. Increase in Δ(DA), presumably reflecting elevated release of exogenously derived DA (Zigmond et al. 1990), was found to be positively correlated with log(BPND_DTBZ) and negatively with log(BPND_MP). This indicates a more effective compensation for relatively milder degrees of denervation [higher log(BPND_DTBZ)] and shows that, once denervation is present, DAT assumes a functional role in the regulation of synaptic DA levels, which was found to be maintained until at least 90% denervation, such that lower DAT was consistently associated with higher Δ(DA).

The analysis of the subset of animals with denervation < 90% indicated that as more severe denervation is included, a similar dependence on the pre-synaptic markers is still observed, but now the only significant correlate is DAT; there is a only trend towards significance for the Δ(DA) dependence on VMAT2. This finding likely indicates a decreasing ability of the pre-synaptic mechanism to provide adequate compensation for the denervation-induced DA loss and a potential gradual engagement of other sources of DA synthesis and release (Carta et al. 2007; Lee et al. 2008). In this situation of DA deficiency, larger changes in Δ(DA), lead to oscillatory behavior of synaptic DA levels. Thus, levodopa treatment in this scenario leads to a pulsatile stimulation of post-synaptic receptors and a heightened risk of dyskinesia (de la Fuente-Fernández et al. 2004a; Cenci and Lundblad 2006). The consistent negative correlation between Δ(DA) and DAT levels explains how, at this stage lower DAT levels become a risk factor.

Interestingly, our current results obtained in rodents agree very well with human studies showing a significant negative correlation between Δ(DA) and DAT levels (Sossi et al. 2007a) and a relative DAT down-regulation has been observed to be linked to a higher occurrence of dyskinesia in both humans (Troiano et al. 2009) and rodents (Lee et al. 2008).

Furthermore, since no significant relation was observed for the unlesioned side and no treatment had been administered to the animals, the correlation between DAT expression and DA release must clearly be determined by the dopaminergic deficit and not by treatment.

A potentially interesting finding is the increase of the BPND_MP to BPND_DTBZ ratio for very high levels of denervation (Fig. 1). Obviously more complex studies involving animals with very high degrees of denervation are needed to investigate this behavior and no conclusions can be drawn from this data set, especially, since at such high level of denervation the measured parameters are numerically very small. However, it can be stated that no obvious imaging-related methodological problems have been observed: as mentioned, partial volume effects are unlikely to explain this trend and neither do these animals appear as obvious outliers when PET data are compared with post-mortem binding (Fig. 4). Abnormal behavior of DAT in the presence of severe VMAT2 deficiency have been however suggested previously (Patel et al. 2003).

There are clearly limitations to this study. First, it was performed on a rat model of PD that differs from human disease. Nonetheless, the virtually identical relationships among the PET data measured in the current study and in a similar human study argue towards strong similarities between this model and human disease. Second, PD-like symptoms have been induced by a well-defined acute mechanism in rats, while the etiology of PD remains unknown and the disease itself progresses; in this study, we are not following progressive adaptive changes, but rather taking a snapshot of the relationships between regulatory changes at different, fixed levels of denervation. There are also data showing that a unilateral lesion might affect the contralateral side as well (Nikolaus et al. 2002), while in this study the contralateral side is used for control values. While the results presented here do not eliminate the possibility of some contralateral effect, such effect must be limited, since no significant relations between the observed variables were found for the unlesioned side.

Despite these caveats, the present study provides a plausible scenario that might be applicable to at least a subset of PD patients. Similar studies performed on different animal models of PD are required to ascertain whether this is a common mechanism or limited to specific disease origins.

In summary, this study shows, using multi-tracer in vivo imaging and the 6-OHDA rodent model of PD, that regulatory changes strive to normalize synaptic DA levels throughout the range of denervation examined and that DAT is consistently functionally related to levodopa-derived changes in synaptic DA levels for levels of dopaminergic denervation ranging from 20% to 90% with lower DAT levels being associated with higher changes. This may explain why relatively low DAT levels may provide an effective compensatory mechanism in a situation of mild dopaminergic deficit, when the system is still able to maintain sufficient endogenous DA levels in the synapse, while increasing the propensity to treatment-induced motor complication at more advanced levels of dopaminergic denervation.


This work was supporter by a CIHR grant, a CIHR Team grant, NSERC, MSFHR and CRC Awards, the James A. Moore Chair in Parkinson’s Research and a TRIUMF Life Science grant.