The aim of our study was to estimate how well in vivo [123I]β-CIT SPECT/CT can discriminate the severity of dopaminergic dysfunction initiated by different intrastriatal 6-OHDA injections in rats. The unilateral partial 6-OHDA lesion model of PD has been widely used in pre-clinical studies of therapeutic interventions, but to our knowledge, the model has not yet been evaluated in both dose- and time-dependent manner with small-animal SPECT/CT using the DAT ligand [123I]β-CIT. Our results show that [123I]β-CIT SPECT/CT can reliably be used to distinguish between single- and two-site intrastriatal 6-OHDA injections and that the striatal binding of [123I]β-CIT correlates well with immunohistochemical findings.
Radioligands used for imaging in PD research focus on the dopaminergic system (for review, see ). Common tracers are [18F]l-3,4-dihydroxyphenylalanine (DOPA), that give an estimate of dopamine synthesis and radioligands with affinity for proteins important for the dopaminergic transmission (e.g., DAT, D2 receptor, vesicular monoamine transporter type 2). In a recent study, Kyono et al.  showed that PET detection of striatal uptake of [18F]FDOPA, after inhibition of aromatic l-amino acid decarboxylase and catechol-O-methyltransferase, is significantly correlated to striatal dopamine levels in the partially 6-OHDA-lesioned rat brain. They did four-site striatal 6-OHDA lesions of a total of 7, 14, or 28 μg 6-OHDA. The [18F]FDOPA uptake was significantly reduced in all three 6-OHDA groups as compared to vehicle-injected rats, but there was no significant difference in [18F]FDOPA uptake between the different 6-OHDA lesion groups using six animals per group. Uptake of DOPA (which includes transport, decarboxylation, and storage) can be affected by compensatory mechanisms in PD and PD models, leading to changes in the turnover [26, 27]. In a study by Forsback et al. , the measured uptake of [18F]FDOPA showed less sensitivity and weaker correlation to nigral cell loss than the DAT tracer [18F]2-β-carbomethoxy-3β-(4-fluorophe-nyl)tropane ([18F]CFT) in an ex vivo study of 6-OHDA-lesioned rats. Compared to β-CIT, CFT is more selective for DAT since β-CIT also has affinity for the serotonin transporter (SERT) [29, 30].
Most of the studies examining DAT binding in vivo have been conducted on rats that have received a unilateral injection of 6-OHDA to the MFB or SN [13–18, 20–22]. In these studies, information about correlations between DAT binding and immunohistochemical findings are largely lacking, and immunohistochemistry has mainly been used as a tool to confirm the dopaminergic lesion after image acquisition. Unilateral injection of 8 μg 6-OHDA into the MFB led to an approximately 59% decrease in ipsilateral binding of [123I]β-CIT in the striatum, and the imaging data were highly correlated to TH-reactive cell counts in the SN . When the same amount of 6-OHDA was injected into the SN, an approximately 29% decrease in striatal DAT binding of [123I]-N-ω-fluoropropyl-CIT ([123I]FP-CIT) was observed . Injection of a larger dose of 6-OHDA (24 μg) lead to a 74% decrease in the [123I]FP-CIT binding .
In studies using PET cameras and tracers, decreases in striatal DAT binding ranging from 33% to 85% have been reported following injection of 8 μg 6-OHDA to the MFB of rats [17, 20–22]. In addition, Inaji  and Pellegrino  and their co-workers showed that the decrease in binding was significantly correlated to drug-induced rotational behavior. In the SN 6-OHDA lesion model, ipsilateral DAT binding was decreased by 50% and 65% to 85% following injection of 4 μg  and 8 μg 6-OHDA , respectively. Moreover, Hume and co-workers  showed that the unilateral 6-OHDA injection did not affect the contralateral DAT binding despite the compensatory changes that are known to take place in the 6-OHDA model . However, Van Camp et al.
 reported a somewhat higher (although not significant) striatal activity of [123I]FP-CIT in the contralateral side of 6-OHDA-lesioned rats as compared to intact animals. When comparing the right (intact) side of 6-OHDA-treated and intact animals, we did not see any changes in striatal [123I]β-CIT binding, indicating that possible compensatory mechanisms did not affect the DAT binding activity in our assay.
Using the partial lesion model, with intrastriatal injection of 4 × 6 μg 6-OHDA, Cicchetti and co-workers  observed an approximately 65% decrease in striatal [11C]CFT uptake, while the DAT binding was not affected by a sham lesion. Sun et al.  reported a 24% decrease in [11C]CFT activity ex vivo in the ipsilateral striatum following intrastriatal injection of a total of 28 μg (4 × 7 μg) 6-OHDA. This is in line with our study in which intrastriatal injection of 2 × 10 μg resulted in an approximately 30% decrease in DAT density measured as the uptake of [123I]β-CIT. The loss of DAT (and TH)-immunoreactive fiber density in the striatum was much more pronounced showing a decrease of approximately 86% to 90% in the lesioned side as compared to the contralateral side. The difference in DAT density measured by the uptake of [123I]β-CIT and DAT-immunoreactive fiber density is probably due to methodological issues, such as the area analyzed (dorsal versus whole striatum).
Methodological issues may also explain why the line plots from the correlation analysis did not intercept with zero in the xy-axis. Another reason for this could be striatal binding of [123I]β-CIT (specific or non-specific) not related to DAT, e.g., [123I]β-CIT has also affinity to SERT [29, 30]. In the rat striatum, DAT concentration is several folds higher than the concentration of SERT, and in normal healthy rat, striatal tissue binding of [123I]β-CIT can be considered to be mediated mainly by DAT . Even though SERT binding activity is decreased in the rat 6-OHDA model , there may still be a change in the proportion of striatal DAT versus SERT binding of [123I]β-CIT. In a recent PET study, Sossi et al.  observed changes in the non-specific background signal from [11C]-methylphenidate (DAT tracer) in the striatum of unilaterally 6-OHDA-lesioned rats. When compared to the reference region, additional non-specific background signal was detected from the striatum that was almost fully denervated by 6-OHDA injection. This increase in the background signal in the lesioned striatum could interfere with the evaluation of the degree of denervation and may cause over-estimation of striatal DAT density and function.
No progression of the lesion could be seen during 4 weeks with any of the used measurement methods. This is in line with previous studies that have shown that degeneration of dopaminergic terminals happens mainly during the first week post-lesion and is complete within 3 weeks after intrastriatal injection of 6-OHDA [8, 34]. In the partial 6-OHDA lesion model, the onset of cell death in the SNpc starts between 1 and 2 weeks post-lesion [8, 35]. When nigral dopaminergic cells are labeled pre-lesioning (e.g., retrograde labeling with fluoro-gold), the cell loss following intrastriatal injection of 6-OHDA continues to progress for an extensive period of time (>4 weeks) . However, the expression of TH is highly regulated, and there is an acute repression of TH phenotype following 6-OHDA injection resulting in maximum loss of nigral TH-immunoreactive neurons at already 2 weeks post-injection . TH immunohistochemistry can therefore be considered to be a reliable marker for DA cell loss starting from approximately 4 weeks post-lesion.
A great advantage with in vivo [123I]β-CIT SPECT/CT to more conventional methods is the possibility to follow repeatedly the condition of the midbrain dopaminergic system in living animals. To our knowledge, the present study is the first to assess DAT binding in the partially 6-OHDA-lesioned rat, using small-animal SPECT/CT imaging. With our nanoSPECT/CT system, the test-retest variability was in a recent study of striatal [123I]β-CIT binding in mice defined to be 9% . Thus, this methodology is suggested to be capable of accurate and repeatable measurement of DAT binding using [123I]-β-CIT. In PD research, continuous evaluation of the animals is usually done by monitoring motor function in different behavioral test settings. The most sensitive behavioral test used for unilaterally lesioned rats is the detection of amphetamine-induced rotational asymmetry that can react to a 40% to 50% decrease in striatal dopamine levels and 30% to 50% loss of nigral cells in the SNpc . Criticism of the test concerns the irrelevance to motor symptoms seen in PD and the fact that the test requires administration of a drug (dopamine agonist) to induce the behavior (for review, see ). Even if it is a sensitive test, it has also been observed that the degree of rotation asymmetry does not necessarily correlate with the degree of striatal dopamine depletion and dopaminergic denervation [6, 37]. This is also seen in our study in which, at 2 weeks post-lesion, rats lesioned with 8 μg 6-OHDA showed higher amount of drug-induced rotations than rats lesioned with 2 × 10 μg 6-OHDA, even though other analyses indicated that 2 × 10 μg 6-OHDA resulted in a significantly bigger lesion.
Compared to the amphetamine-induced rotational test, many other tests used to evaluate motor function in 6-OHDA-lesioned rats require more extensive dopaminergic denervation  and are therefore not useful for the detection of very limited loss of dopamine neurons. In our hands, an intrastriatal injection of 8 μg 6-OHDA resulted in a 30% reduction in TH-positive cells in the SNpc. This is approximately the same extent of cell loss that has been estimated to be found at the time of onset of motor symptoms in Parkinson's disease (discussed in ). As diagnostic procedures improve, it would be important to be able to study also lesions of smaller magnitude to evaluate the effect of new neuroprotective treatments aimed at this stage of the disease. Our work shows that in vivo [123I]β-CIT SPECT/CT could be useful for this purpose.
Neuronal compensation mechanisms [5, 7, 27] and spontaneous recovery of motor function despite no improvement or re-innervation of the dopaminergic pathway are other problems that can affect the outcome of some behavioral test (discussed in ). This may explain the inconsistency in the amount of ipsilateral amphetamine-induced rotations in our study showing a decrease in the extent of rotational asymmetry in the 8-μg lesion group between 2 and 4 weeks post-lesion. Obviously, the over-interpretation of results could be partly avoided by the use of in vivo [123I]β-CIT SPECT/CT as a complement to behavioral tests in animal PD models. Compensatory mechanisms should, of course, also be taken into account in SPECT/CT analyses of dopaminergic neurons since the results can be affected by changes in DAT expression, dopamine turnover, and axonal sprouting. Last but not least, behavioral tests tend to result in rather big inter-individual variations and require therefore often numerous animals before differences between treatment groups can reliably be detected. Our work shows that with groups of only four to five animals, [123I]β-CIT SPECT/CT can give statistically significant differences in striatal [123I]β-CIT uptake between two different partial lesions of the dopaminergic system. The only other measurement that could discriminate between the two different lesions was loss of nigral TH-reactive cells.