In-vivo Measurement of LDOPA Uptake,Dopamine Reserve and Turnover in the Rat Brain Using [ 18 F]FDOPA PET

Abstract

Longitudinal measurements of dopamine (DA) uptake and turnover in transgenic rodents may be critical when developing disease-modifying therapies for Parkinson's disease (PD). We demonstrate methodology for such measurements using [¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine ([¹⁸F]FDOPA) positron emission tomography (PET). The method was applied to 6-hydroxydopamine lesioned rats, providing the first PET-derived estimates of DA turnover for this species. Control (n = 4) and unilaterally lesioned (n = 11) rats were imaged multiple times. Kinetic modeling was performed using extended Patlak, incorporating a k_loss term for metabolite washout, and modified Logan methods. Dopaminergic terminal loss was measured via [¹¹C]-(+)-dihydrotetrabenazine (DTBZ) PET. Clear striatal [¹⁸F]FDOPA uptake was observed. In the lesioned striatum the effective DA turnover increased, shown by a reduced effective distribution volume ratio (EDVR) for [¹⁸F]FDOPA. Effective distribution volume ratio correlated (r > 0.9) with the [¹¹C]DTBZ binding potential (BP_ND). The uptake and trapping rate (k_ref) decreased after lesioning, but relatively less so than [¹¹C]DTBZ BP_ND. For normal controls, striatal estimates were k_ref = 0.037 ± 0.005 per minute, EDVR = 1.07 ± 0.22 and k_loss = 0.024 ± 0.003 per minute (30 minutes turnover half-time), with repeatability (coefficient of variation) ≤11%. [¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine PET enables measurements of DA turnover in the rat, which is useful for developing novel therapies for PD.

Keywords

dopamine synthesis FDOPA 6-OHDA lesion Parkinson's disease PET

INTRODUCTION

In-vivo imaging using [¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine ([¹⁸F]FDOPA) positron emission tomography (PET) allows quantification of the brain kinetics of exogenous LDOPA. This includes the LDOPA uptake, including its decarboxylation to dopamine (DA) and subsequent sequestration in vesicles, and DA turnover. Such methodology has been successfully applied in humans^1–5 and nonhuman primates^6–8 but only limited studies have been reported in the rat.⁹ Given that rodent models are critical in medical research, it is important to develop imaging methodology in these species, and for specific biological readouts to be verified. A number of diseases are associated with changes or dysfunction in the dopaminergic system, most notably Parkinson's disease (PD). Several key features of this multifaceted disease remain poorly understood. The application of longitudinal in-vivo imaging to rodent models of PD may aid our understanding of genetic and environmental risk factors for PD, and facilitate the development of novel therapies aimed at preventing the disease or slowing its progression. Here, we describe findings from [¹⁸F]FDOPA PET imaging in control rats and in unilaterally 6-Hydroxydopamine (6-OHDA) lesioned rats. This work lays the foundations for [¹⁸F]FDOPA imaging in this and other rat models of PD, in which the tracer may provide unique and valuable information regarding DA synthesis, storage, and turnover.

Our aim was to quantify the rate constant describing FDOPA uptake and the DA turnover in the 6-OHDA rat model of PD. The effective fluorodopamine (FDA) turnover was previously defined as the ratio between the PET-derived DA clearance rate k_loss and the net-uptake rate K_i,⁸ and subsequently estimated as the direct inverse of the effective distribution volume (DV) of trapped FDA.³ Evidence from the PET imaging of asymptomatic humans that carry mutations in LRRK2⁵ suggests that measures of effective DA turnover may be more valuable than, and synergistic with, measures of VMAT2 or DAT in the study of PD and hence Parkinsonism in rodents. The accurate realization of such measurements is hindered by the complex metabolism of the radiotracer and the presence of radiolabeled metabolites that enter the brain.^1,2,4,10 The data-driven graphical methods, used in the current work, do not require a priori specification of a compartmental model and provide robust estimates when applied to the complex metabolic pathway traced by FDOPA.^3,11–13

In addition to the aforementioned findings in humans, we are motivated by reports from rodent-based models of PD that suggest changes in DA turnover take place before overt pathology¹⁴ and that excessive amounts of α-synuclein may cause dysfunction in the vesicular release of DA.¹⁵ Despite such motivation, there is a paucity of literature that describes [¹⁸F]FDOPA imaging in rats. The only such report is that of Kyono et al,⁹ who undertook [¹⁸F]FDOPA PET in 6-OHDA lesioned rats and found a negative relationship between striatal [¹⁸F]FDOPA uptake and methamphetamine-induced rotational behavior, and a positive relationship between the [¹⁸F]FDOPA uptake and the striatal levels of DA and DA metabolites in tissue homogenates. The results from ex vivo [¹⁸F]FDOPA and [¹⁴C]L-DOPA autoradiography in 6-OHDA lesioned rats have been reported.^16,17 In both cases, the tracer uptake was correlated with the number of remaining nigrostriatal neurons. The PET measured striatal [¹⁸F]FDOPA uptake was progressively reduced with age in a genetic mouse model of PD.¹⁸

Several reports examine the central and peripheral metabolism of [¹⁸F]FDOPA in rats^19–21 and the changes in metabolism caused by inhibitors of aromatic L-amino acid decarboxylase (AADC) and catechol-O-methyltransferase (COMT).^22,23 In the absence of AADC inhibition, [¹⁸F]FDOPA is rapidly metabolized to [¹⁸F]FDA in the periphery, thus decreasing the fraction of injected [¹⁸F]FDOPA that crosses the blood–brain barrier to become a substrate for the processes of interest. When AADC but not COMT is inhibited [¹⁸F]FDOPA is rapidly O-methylated in the periphery, with the radiolabeled metabolite 3-O-methyl-[¹⁸F]FDOPA ([¹⁸F]OMFD) comprising ~50% of the total radioactivity in plasma within 15 minutes in the rat.²² [¹⁸F]OMFD crosses the blood–brain barrier to produce a reasonably uniform background with a DV close to unity.²⁴ As the peripheral metabolism of FDOPA occurs faster in rodents compared with humans, these effects have limited the use of [¹⁸F]FDOPA imaging in the rat.

In an attempt to circumvent this problem, we chose to pursue [¹⁸F]FDOPA PET imaging in rats that had received both AADC and COMT inhibitors. This enabled tracer kinetic modeling to be performed using a reference region (cerebellum), which was assumed devoid of AADC.²⁵ The model neglects [¹⁸F]OMFD, as we assume the concentration of [¹⁸F]OMFD to be low compared with [¹⁸F]FDOPA and [¹⁸F]FDA. If the inhibitors also act centrally they can alter the tracer kinetics. In particular, central inhibition of COMT is expected to lower the FDA turnover and increase the accumulation of radioactivity in the striatum.¹³

In the 6-OHDA lesion rat model of PD,²⁶ we acquired 3-hour duration [¹⁸F]FDOPA PET data with the aim of separately quantifying: (1) the FDOPA uptake rate (i.e., its decarboxylation and subsequent incorporation into vesicles) and (2) the effective FDA turnover, which was quantified by the inversely related effective DV ratio (EDVR). These different aspects of neuronal function might change by different amounts following 6-OHDA lesioning, as a result of differences in compensatory capabilities. It is known that DA synthesis (and presumably the rate of vesicular trapping) is relatively increased in surviving neurons in this model,^27,28 and we thus predicted that measurements of FDOPA uptake (decarboxylation and FDA vesicular trapping) would not fall at the same rate as neuronal terminal density. We also wished to examine the relationship between DA turnover (via EDVR) and terminal density. This was achieved by correlating results from FDOPA PET with estimates of denervation severity, derived from [¹¹C]-(+)-dihydrotetrabenazine (DTBZ) PET. [¹¹C]-Dihydrotetrabenazine binds to VMAT2 and its binding potential (BP_ND) is a validated marker of DA terminal integrity.^29,30

[¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine PET was performed multiple times in individual rats to assess repeatability. Arterial blood was collected and the concentrations of [¹⁸F]FDOPA, [¹⁸F]OMFD and other metabolites were determined. Two different COMT inhibitors (tolcapone and entacapone) were tested and their known differences in pharmacological action were used to verify our interpretation of the PET data. Entacapone acts mainly in the periphery, while tolcapone also acts centrally. The central potency of tolcapone is about 10 times that of entacapone.³¹ Peripheral inhibition is similar for both compounds.

MATERIALS AND METHODS

Animals and 6-Hydroxydopamine Lesioning

Procedures involving animals were approved by the University of British Columbia's ethics committee and performed in accordance with the Canadian Council on Animal Care guidelines. Four normal control and 11 unilaterally 6-OHDA lesioned²⁶ male Sprague Dawley rats were used. The animals had free access to standard diet and tap water. They were housed at 21°C with a 12-hour light cycle (light from 0700 to 1900 hours). In lesioned rats, a naturally variable degree of dopaminergic denervation was produced by injection of 10 μg of 6-OHDA hydrobromide (Sigma-Aldrich, St Louis, MO, USA) dissolved in 4 μL of 0.05% ascorbic acid solution into the substantia nigra pars compacta at coordinates: AP −4.7 mm (from bregma), ML −1.5 mm (from the midline), and DV −7.9 mm (from skull surface).³² Desipramine (Sigma-Aldrich) was given at least 30 minutes before surgery (25 mg/kg intraperitoneally) to protect noradrenergic nerve terminals, ensuring selectivity for dopaminergic neurons.³³

Radiotracer Synthesis

[¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine was synthesized as described in Namavari et al³⁴; the specific activity at injection was 16 ± 7 MBq/μmol (mean ± s.d.). [¹¹C]-Dihydrotetrabenazine was synthesized as described in Adam et al³⁵; the specific activity at injection was 224 ± 7 GBq/μmol. The injected activity for both compounds was 0.045 ± 0.002 MBq/g.

Positron Emission Tomography Scanning

Each animal underwent 4 to 8 [¹⁸F]FDOPA PET scans and one [¹¹C]DTBZ scan, with no more than one scan per week. An AADC inhibitor (benserazide, 10 mg/kg) was given intraperitoneally 30 minutes before the [¹⁸F]FDOPA injection. Ninety minutes prior to the [¹⁸F]FDOPA injection, a COMT inhibitor (40 mg/kg tolcapone, 40 mg/kg entacapone) was given intraperitoneally; limited testing using 10 mg/kg entacapone was also performed. The COMT inhibitors were obtained from a local pharmacy in pill form, ground to fine powders, and then mixed in distilled water to a concentration of 40 mg/mL (or 10 mg/mL) for doses of 40 mg/kg (or 10 mg/kg). Benserazide (Sigma-Aldrich) was mixed in distilled water (10 mg/mL). The animals were anesthetized with 2.5% isoflurane gas during scanning. Their heart rate and blood oxygen saturation level were continuously monitored using a pulse oximeter; temperature was measured using a digital thermometer, and respiration was monitored visually at 15 minutes intervals. Temperature was maintained at 35°C to 36°C using a heat lamp. Ear-bars immobilized the head and provided accurate positioning of the brain at the center of the scanner's field of view.

The MicroPET Focus120 small animal scanner (Concorde/Siemens, Knoxville, TN, USA) was used.³⁶ A 10-minute ⁵⁷Co transmission scan was acquired, following which the radiotracer was injected intravenously at the tail using a 25-gauge, 19 mm long catheter (BD Insyte). Emission data were collected for 3 hours ([¹⁸F]FDOPA) or 1 hour ([¹¹C]DTBZ), and split into 26 or 17 frames, respectively. The frames had increasing durations, ranging from 30 to 900 seconds ([¹⁸F]FDOPA) or 30 to 480 seconds ([¹¹C]DTBZ). Fourier rebinning was used to collapse the 3D data, which were fully corrected for attenuation, scatter, normalization, and dead-time. Images were reconstructed using filtered backprojection, calibrated, and decay corrected. The spatial resolution of this system is < 1.5 mm full width at half maximum.³⁶

To measure the efficacy of the AADC and COMT inhibitors, arterial blood samples were collected from the tail during [¹⁸F]FDOPA scanning. A 25-gauge, 19 mm long catheter (BD Insyte) was inserted into the ventral artery before the start of the scan. Three samples each of 200 μL (plus 100 μL dead-space removal) were drawn at 20, 100, and 160 minutes post tracer administration. They were analyzed for [¹⁸F]FDOPA and its metabolites following the method of McLellan et al.³⁷ The method separates compounds in plasma into four compartments: those trapped by the anionic (F-HVA, F-DOPAC, FDA-sulf) and cationic (FDA) exchange columns, and the nonretained supernatant (OMFD) and alumina (FDOPA). Due to the difficulty of sampling from the ventral artery, blood was sampled from the tail vein (opposite side to tracer injection) in instances where arterial catheterization failed. This enabled comparison of the metabolic profile between venous and arterial blood. Blood data were not used to generate an arterial input function since only a small number of samples were taken per scan, limited by the blood volume of the rat for this longitudinal study.

Image Analysis

Time-activity curves (TACs) were extracted from the images using regions-of-interest (ROIs) that had predefined shape and size using the ASIPro software (CTI Concorde Microsystems, Knoxville, TN, USA). For each scan, three ROIs were placed on summation radioactivity images, defining the cerebellum and left and right striatum. Each ROI covered three consecutive transverse planes; in each of these planes, the ROI was rectangular and placed manually on the given region so that the ROI tracked changes in anatomy between-planes; this simple procedure ensured similar treatment of the contralateral striatum and the ipsilateral striatum, capturing a high proportion of striatal uptake with minimal contamination from adjacent tissue. The ROI volumes were 0.043 cm³ (cerebellum) and 0.022 cm³ (striatal). Kinetic modeling and analysis was performed using Matlab (The Mathworks, Natick, MA, USA) with in-house software. The average standard uptake value (SUV; kBq/cm³/MBq/kg) for [¹⁸F]FDOPA was calculated between 75 and 150 minutes. The SUV enables a simple comparison of the tracer uptake between different PET images, after accounting for the animal's weight and the injected activity. It was used here for quality assurance of the data.

[¹¹C]-Dihydrotetrabenazine was analyzed using Logan graphical analysis with a cerebellum reference region.¹² The time after which data are fitted, t*, was set to 30 minutes and the k₂ term was omitted.³⁸ The ratio of the total DV in the striatum to that for the reference region was thus estimated. From this DVR, the BP of DTBZ (BP_ND) was calculated as BP_ND = DVR–1. In the 6-OHDA lesion model, changes in DTBZ BP_ND reflect changes in VMAT2 density (B_max),³⁰ which is linearly correlated with tyrosine hydroxylase activity.²⁹ The denervation severity (DS_DTBZ) caused by the lesioning was calculated from the asymmetry in BP_ND via:

D S_{DTBZ} = \frac{B P_{N D}^{C o n t r a} - B P_{N D}^{l p s i}}{B P_{N D}^{C o n t r a}}

(1)

where the superscripts contra and ipsi represent the striatum on the contralateral side and the ipsilateral side to the lesioning. The method was previously verified against autoradiography.³⁸

[¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine was similarly analyzed using the Logan plot, with the modification that the TAC from the reference region was subtracted from the striatal TAC before running the analysis.³ Following this subtraction, the estimated slope from the linear fit equals the EDVR, which informs on the DV of the striatal FDA compartment as compared with the reference region. The asymmetry in the EDVR estimate was calculated as:

A S Y M_{FDOPA} = \frac{E D V R^{Contra} - E D V R^{l p s i}}{E D V R^{Contra}}

(2)

The inverse of the EDVR is interpreted as a measured of the effective DA turnover.

In addition to this Logan-type analysis, we applied extended Patlak graphical analysis.^11,13 The extended formulation includes allowance for the loss of radiolabeled metabolites from the sequestered compartment (described by the rate constant k_loss). The k_loss estimate was that which provided the optimal least-squares fit in the linear regression used to estimate k_ref. Note that k_ref reflects a multiplicity of processes, namely the decarboxylation of FDOPA to FDA and the sequestration of FDA in vesicles.

Reliability and Repeatability

The reliability and repeatability of measurements of EDVR, k_ref, and k_loss were calculated by applying a single factor analysis of variance to the results obtained from repeated (two to three) measurements, considering each striatum separately for an individual rat. Calculations were performed separately for each COMT inhibitor and for each condition (normal control, lesioned contralateral, lesioned ipsilateral). We use the definitions:

Reliability = \frac{σ_{b}^{2}}{σ_{b}^{2} + σ_{w}^{2}}

(3)

and,

Reliability (%) = 100 \frac{σ_{w}}{\overset{ˉ}{x}}

(4)

where σ _b ² is the between-subject variance, σ _w ² is the within-subject variance, and x is the mean value for the group.

RESULTS

In almost one-fifth of scans, the AADC inhibition appeared ineffective; this occurred in individual rats repeatedly, suggesting that the 10 mg/kg dose of benserazide used was insufficient for some animals. The effect was clearly seen in the blood data, by an FDOPA fraction of <30% in the first (20 minutes) sample, and in the PET images by a low tracer uptake in the brain. The SUV in both the striatum and cerebellum was about 0.15 in these cases, far lower than the normal SUVs of around 1.5 in the striatum and 0.64 in the cerebellum. Catechol-O-methyltransferase inhibition was ineffective in one scan (out of 55), likely explained by an ineffective intraperitoneal injection. These clear outliers were not useful for quantitative analysis and were discarded from all further analyses, tables, and figures.

Results from the measurement of radiolabeled metabolites in arterial blood are shown in Figure 1; in all cases, the majority of radioactivity is derived from the parent compound. The data confirm that COMT was effectively inhibited by 40 mg/kg tolcapone or 40 mg/kg entacapone, since the measured level of [¹⁸F]OMFD in plasma remains close to zero. No differences in radiolabeled metabolites were observed between the two COMT inhibitors at this dose. The data show that AADC is reasonably inhibited since anionic metabolites remain low. 10 mg/kg entacapone did not fully inhibit COMT, seen by an OMFD fraction of 0.13 at 160 minutes. Venous blood yielded a similar but delayed profile compared with arterial blood. The arterial-to-venous ratio of the metabolite fractions for the main (anionic) metabolites were 2.3, 1.4, and 1.1 for samples at 20, 100, and 160 minutes. The average ratio of the activity concentrations between plasma and whole blood was 0.77 at 100 minutes.

Figure 1.

Radiolabeled metabolites in arterial blood from the rat, as found after administration of different catechol-O-methyltransferase (COMT) inhibitors (t = −90) paired with an aromatic L-amino acid decarboxylase (AADC) inhibitor (10 mg/kg benserazide, t = −30). Data are for 40 mg/kg tolcapone (–o–, lines 1, 5, and 9), 40 mg/kg entacapone (•Δ•, lines 2, 6, and 8), and 10 mg/kg entacapone (•□•, lines 3, 4, and 7). Data from 40 mg/kg tolcapone and 40 mg/kg entacapone are overlapping in all cases. The upper three lines (1 to 3, light gray) are results from the alumina (parent compound, fluoro-3,4-dihydroxyphenyl-L-alanine (FDOPA)). The middle three lines (numbered 4 to 6, black) are from the anion exchange column (fluorodopamine (FDA)-sulfate, F-HVA, and F-DOPAC). The lowest three lines (numbered 7 to 9, dark gray) are from the supernatant (OMFD). Error bars represent ± 1 s.d. from the three to four measurements used to calculate the mean (i.e., three to four rats per COMT inhibitor). The activity retained by the cation exchange column (e.g., FDA) was negligible.

Table 1 lists the number of reported rats and scans at the 40 mg/kg COMT inhibitor dosing. Two subgroups of lesioned rats were identified on the basis of their DS_DTBZ values: (1) three rats with ~30% denervation (DS_DTBZ range = 0.25 to 0.38, mean = 0.30); (2) three rats with 90% denervation (DS_DTBZ range = 0.83 to 0.92, mean = 0.88). The average parameter estimates (k_ref, k_loss, and EDVR) obtained from these groups are listed in Table 2, along with the repeatability and reliability for the normal control and lesioned groups.

Table 1

Numbers of animals and scans (40 mg/kg COMT inhibitor dosing)

	Normal control	Lesioned (all)	Lesioned (30%)	Lesioned (90%)
n, Rats (tolcapone)	4	9	3	3
n, Rats (entacapone)	4	8	3	3
n, Rats (total)	4	9	3	3
n, Scans (tolcapone)	8	16	7	5
n, Scans (entacapone)	8	14	5	5
n, Scans (total)	16	30	12	10

COMT, catechol-O-methyltransferase.

Table 2

Repeatability and reliability of parameter estimates for the striatum, with values from normal controls

Parameter	Group	k _ref (tolc.)	k _ref (entac.)	k _loss (tolc.)	k _loss (entac.)	EDVR (tolc.)	EDVR (entac.)
Repeatability (%)	NC	10	8	9	11	6	9
	LC	10	11	18	11	17	6
	LI	12	29	17	27	18	12
Reliability	NC	0.75	0.93	0.59	0.81	0.98	0.95
	LC	0.6	0.74	0.32	0.68	0.44	0.89
	LI	0.91	0.65	0.94	0.58	0.96	0.98
Value (mean±s.d.)	NC	0.034±0.004	0.037±0.005	0.020±0.002	0.024±0.003	1.40±0.24	1.07±0.22
	LC 30%	0.035±0.003	0.043±0.007	0.020±0.001	0.029±0.003	1.29±0.15	0.99±0.13
	LI 30%	0.032±0.005	0.038±0.006	0.024±0.001	0.033±0.003	0.92±0.17	0.72±0.14
	LC 90%	0.032±0.004	0.039±0.003	0.019±0.003	0.024±0.003	1.37±0.06	1.11±0.15
	LI 90%	0.021±0.009	0.023±0.004	0.049±0.016	0.051±0.005	0.18±0.07	0.14±0.01

k_ref and k_loss have units of 1/min. Tolc=tolcapone, entac.=entacapone. Groups are coded as: NC=normal control, LC=lesioned contralateral (all), LI=lesioned ipsilateral (all), LC 30%=lesioned contralateral (30%), LI 30%=lesioned ipsilateral (30%).

Example [¹⁸F]FDOPA PET images corresponding to both COMT and AADC inhibition are displayed in Figures 2A–C. The ROI definition for the striatum is shown in part A. The images show that the striatum can be visualized and that there is a large reduction in radioactivity concentration in the lesioned stratum, seen in part C. Time-activity curves for the cerebellum and striatum, averaged from four control rats and from the three rats with 30% denervation, are presented in Figure 2D.

Figure 2.

(A–C) [¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine ([¹⁸F]FDOPA) positron emission tomography (PET) images (3-hour summation) showing slices through the striatum. (A) Striatal region-of-interest (ROI) definition shown on a normal control rat. (B) Normal control rat. (C) A unilaterally lesioned rat (dihydrotetrabenazine (DTBZ) _DS = 0.83). The arrow locates the lesioned striatum. (D) [¹⁸F]FDOPA TACs from normal control rats and from rats with ~30% denervation. Mean values are plotted (three to four rats); error bars represent ± 1 s.d. Ipsi. = ipsilateral to lesioning. Contr. = contralateral to lesioning. Cer. = cerebellum. Str. = striatum.

The estimates of k_ref and k_loss are compared with the corresponding estimates of DTBZ BP_ND in Figure 3 (left: k_ref versus DTBZ BP_ND, right: k_loss versus DTBZ BP_ND). We found a complex relationship between k_ref and DTBZ BP_ND, which demonstrated upregulation in k_ref (per terminal), and a significant linear relationship between k_loss and DTBZ BP_ND.

Figure 3.

(A) Relationship between k_ref and dihydrotetrabenazine (DTBZ) binding potential (BP_ND). (B) Relationship between k_loss and DTBZ BP_ND. A localized polynomial regression curve is shown in (A). The linear regression shown in (B) has a correlation coefficient of −0.82. Data are from 40 mg/kg entacapone. Similar trends were found for tolcapone.

The relationship between striatal EDVR and DTBZ BP_ND is shown in Figure 4, for both COMT inhibitors (left: entacapone; right: tolcapone). Significant linear correlations were found (correlation coefficients of 0.90 and 0.91, respectively). Comparison of Figures 4A and 4B shows that tolcapone increased EDVR as compared with the same dose of entacapone. The median increase in EDVR was 22%. Student's t-test (paired, single tailed) showed this increase to be statistically significant (P < 0.01). Bland–Altman analysis (not shown) found the increase in EDVR to be proportional to EDVR.

Figure 4.

Comparison of fluoro-3,4-dihydroxyphenyl-L-alanine (FDOPA)-derived effective distribution volume ratio (EDVR) with dihydrotetrabenazine (DTBZ) binding potential (BP_ND) for (A) 40 mg/kg entacapone and (B) 40 mg/kg tolcapone. Linear regressions are shown as lines; the correlation coefficients are 0.90 and 0.91.

Figure 5 shows the asymmetry in the FDOPA PET estimates when using entacapone (parts A and C for EDVR and k_ref), compared with the asymmetry in DTBZ BP_ND. The asymmetry in EDVR matched that in DTBZ BP_ND; the linear regression is not significantly different from the line of identity and the correlation coefficient is 0.99. This was not the case for k_ref, where a weaker correlation (correlation coefficient of 0.75) was found, and where the asymmetry in k_ref was less than the asymmetry in BP_ND. Similar results were obtained using tolcapone, as seen in Figures 5B and 5D, for EDVR and k_ref, respectively.

Figure 5.

Comparisons of asymmetry in fluoro-3,4-dihydroxyphenyl-L-alanine (FDOPA)-derived estimates of effective distribution volume ratio (EDVR) (A, B) and k_ref (C, D), as compared with the asymmetry in the dihydrotetrabenazine (DTBZ)-derived binding potential (BP_ND). Data in (A, C) are for 40 mg/kg entacapone; (B, D) are for 40 mg/kg tolcapone. The correlation coefficients for the linear regressions are 0.99, 0.98, 0.75, and 0.66 for parts (A–D), respectively.

DISCUSSION

The presented data confirm that [¹⁸F]FDOPA PET imaging in rats can provide a quantitative measure of FDA trapping in the striatum. In the 6-OHDA unilateral lesion model of PD, the DV of the striatal FDA compartment, quantified by the EDVR, was linearly related to the degree of DA denervation caused by the lesioning. The EDVR measure was sensitive to the pharmacological action of tolcapone, which decreased central FDA metabolism and thus increased the EDVR (or lowered the effective DA turnover). The quantitative analysis afforded estimates of k_ref and k_loss, the latter being the first such PET-derived estimates reported for the rat striatum. The repeatability for these estimates was quantified. The structure of the striatum could be seen in the PET images with high contrast, and the two sides were easily resolved by the microPET.

The arterial blood data verified the actions of the AADC and COMT inhibitors. For routine imaging, blood collection is not mandatory, since ineffective AADC inhibition presented itself clearly in the PET image and ineffective COMT inhibition was rare. The cautious investigator is likely satisfied by a single venous blood sample. The method of metabolite analysis did not determine the form of the main (anionic) metabolite, but we assume it to be FDA-sulfate.²²

The EDVR was calculated using Logan graphical analysis following subtraction of the reference region. This subtraction is designed to remove the precursor compartment from the target region, which is assumed to approximately equal the reference region.³ Estimates of k_ref and k_loss allowed the changes in EDVR (or the inversely related effective DA turnover) to be explained in terms of reduced FDA synthesis and/or vesicular trapping, and increases in the rate constant for the loss of labeled compounds from the target region. These two parameters were estimated using the extended Patlak plot, which includes a term describing the loss of radiolabeled metabolites from the ‘trapped’ compartment.¹¹ The method was found to be robust for control rats, but less reliable for the lesioned striatum (e.g., DS_DTBZ > 0.5) where the striatal TAC is marginally greater than the reference TAC. In these cases, the lesioning has destroyed a substantial fraction (but not all) of the dopaminergic terminals in the ipsilateral striatum, and hence the AADC-specific ‘trapping’ of the radiotracer is diminished (but not eliminated). This leads to a reduction in the ratio of the striatal TAC to the reference region TAC, and hence the signal-to-noise ratio for AADC-specific radiotracer accumulation is reduced. As may be expected, there is a corresponding reduction in the repeatability for estimates of k_ref and k_loss. The situation is common to many radiotracers where specific uptake of the tracer is reduced by disease or in a disease model. The different repeatability of the parameter estimates at different levels of denervation should be considered when designing a study that makes use of this method, as a given effect (e.g., a change in one of the parameters following some intervention) may be more difficult to observe in animal models that start with a high degree of denervation.

There is much literature that describes the characteristics of 6-OHDA unilateral lesions.^27,28 It is known that when 20% or more of the striatal DA neurons survive the lesioning, the concentration of DA in dialysates collected from striatal extracellular fluid (ECF) is unchanged. This remarkable compensation is further evidenced by an increase in DA synthesis, metabolism and fractional release in the surviving neurons, coupled with a reduction in DA reuptake.²⁷ In the present study, it was found that for partial (mild) striatal denervation (DS_DTBZ < 0.4, corresponding to DTBZ BP_ND > 2.5), there was an increase in k_loss with relatively little change in k_ref from extended Patlak analysis (see Figure 3). Such a finding is in good agreement with the known compensation that occurs in this model. If the concentration of DA in the ECF (and presumably the synapse) is normal, then there is no change in the total DA production per unit time (assuming no change in the rate at which DA is removed from the ECF). This would agree with the relative preservation of k_ref found here. Despite being diminished in number, regulation allows the remaining vesicles to sequester a near-normal total amount of radioactivity per unit time. On the other hand, the total DA content of striatal tissue (e.g., from tissue homogenates) is reduced, that is, the distribution of DA between tissue and the ECF is altered.²⁷ The implication is for a reduction in the total amount of DA sequestered at any one time, which for a normal level of DA production can be achieved by a reduced vesicular residence time. This is in keeping with the increase in effective DA turnover (reduced EDVR) and increase in k_loss that was found here. In more severe lesions, the compensation was insufficient to maintain k_ref at normal values.

The compensatory increase in k_ref per terminal in mild lesions in the 6-OHDA lesioned rat is similar to what is observed in early and preclinical PD using the same imaging technique.⁵ Clinical symptoms appear when DTBZ BP_ND is decreased by about 50%, and it is only after this stage that k_ref substantially decreases. The reduction in DA storage capacity (or increase in turnover) and the increase in k_loss found here are in agreement with prior reports from other species.^4–6,8

Although an unchanged k_ref in mild lesions is in keeping with predictions based on the characterization of the 6-OHDA model via non-PET methods, it disagrees with some PET-based reports of a reduced k_ref (or K_i) in lesioned nonhuman primates⁸ and 6-OHDA lesioned rats.⁹ This discrepancy may be explained by the disease-dependent negative bias introduced into k_ref when calculated by the traditional (nonextended) Patlak graphical analysis which neglects k_loss. This bias depends on k_loss, being most apparent in species where k_loss is large (such as the rat), and/or in the diseased state.³⁹ When calculating k_ref by the traditional Patlak method, we obtained a value of 0.0168 ± 0.003/min for normal control rats (using data over the period 10 to 60 minutes) in agreement with Kyono et al.⁹ The bias reduced the k_ref for control rats by 55%, and masked the compensatory effects of relatively preserved DA synthesis and trapping in the mildly lesioned striatum.

Further refinement of the method, particularly regarding the dose and timing of the inhibitors may be possible. The reliable inhibition of peripheral AADC and COMT is desired, normally without central inhibition. There may however be some central effects of both entacapone³¹ and benserazide⁴⁰ when administered intraperitoneally at doses of 40 mg/kg and 10 mg/kg, respectively. In the current study, these doses were chosen to ensure long lasting and near complete peripheral inhibition. A dose of 10 mg/kg entacapone (t = −90 minutes, intraperitoneally) did not provide COMT inhibition throughout the 3-hour acquisition. The use of inhibitors provides a mechanism for bias in the results that is dependent upon their efficacy in the particular rat during the scan. Substantial bias may arise from peripherally generated [¹⁸F]OMFD due to its contamination of the reference region, whereas central inhibition of COMT can increase EDVR and reduce k_loss (parameters of interest). Although we cannot determine the extent to which the central effects of 40 mg/kg entacapone changed EDVR and k_loss from their basal values, the observation that EDVR increased when changing to 40 mg/kg tolcapone confirmed that 40 mg/kg entacapone does not saturate central COMT in the rat. In theory, the target and reference regions could also be contaminated by [¹⁸F]OMFD made centrally; this is however not the case since the DV of OMFD in the rat brain is ~1,²⁴ and the concentration of [¹⁸F]OMFD in arterial plasma was measured and found to be near-zero. Note also that partial inhibition of central AADC would probably be unobservable in our data, since k_ref for the rat striatum is likely limited by FDA sequestration into vesicles as opposed to AADC activity.²⁰ Despite the above concerns coupled with the simplifications applied to the complex metabolic pathway of [¹⁸F]FDOPA, the results demonstrated that k_ref and EDVR retain distinct information. Interestingly, these measures were differentially affected by 6-OHDA lesioning in the rat, with the relative preservation of k_ref being consistent with our predictions based on the literature.

In summary, [¹⁸F]FDOPA PET imaging of the rat brain enables quantification of the capacity for synthesis and storage of DA. Using a cerebellum reference region, we were able to calculate striatal estimates of: (1) EDVR, which is approximately the ratio of DVs between the specifically trapped compartment, assumed to represent striatal FDA levels, and the precursor compartment represented by the reference region.³ 1/EDVR is an index of the effective DA turnover; (2) k_ref, which is the uptake rate constant for FDOPA (and so includes both FDA synthesis and vesicular trapping); (3) k_loss, which is the rate constant for the washout of radioactivity from the striatum, that is, the metabolism of FDA and elimination of metabolic products. The EDVR estimate was found to be linearly related to the degree of DA denervation caused by 6-OHDA lesioning. The k_ref estimate remained at normal levels in mildly lesioned rats (up to about 50% DS_DTBZ) but was reduced in more heavily lesioned rats.

Changes in DA synthesis and turnover in animal models of disease can be related to the human condition using this technique. The repeated, longitudinal quantification of effective DA turnover (via EDVR) and k_ref may aid in the understanding of disease mechanisms and compensatory effects at the molecular level. The relevance of new rodent models of PD, such as transgenic models, may be directly assessed using the method. [¹⁸F]fluoro-3,4-dihydroxyphenyl-L-alanine PET can assist in the development and testing of novel therapies aimed at delaying or reversing neurodegeneration in PD.

DISCLOSURE/CONFLICT OF INTEREST

The authors declare no conflict of interest.

Footnotes

ACKNOWLEDGEMENTS

The authors thank the staff at TRIUMF for tracer synthesis and the UBC PET group for their assistance. DJ Doudet is thanked for constructive discussions.

References

Martin

WRW

Palmer

Patlak

Calne

. Nigrostriatal function in humans studied with positron emission tomography. Ann Neurol 1989; 26: 535–542.

Gjedde

Reith

Dyve

Leger

Guttman

Diksic

Dopa decarboxylase activity of the living human brain. Proc Natl Acad Sci USA 1991; 88: 2721–2725.

Sossi

Doudet

Holden

. A reversible tracer analysis approach to the study of effective dopamine turnover. J Cereb Blood Flow Metab 2001; 21: 469–476.

Kumakura

Cumming

. PET studies of cerebral levodopa metabolism: a review of clinical findings and modeling approaches. Neuroscientist 2009; 15: 635–650.

Sossi

Fuente-Fernandez

Nandhagopal

Schulzer

McKenzie

Ruth

Dopamine turnover increases in asymptomatic LRRK2 mutations carriers. Mov Disord 2010; 25: 2717–2723.

Barrio

Huang

Melega

Hoffman

Schneider

6-[F-18]Fluoro-L-Dopa probes dopamine turnover rates in central dopaminergic structures. J Neurosci Res 1990; 27: 487–493.

Doudet

Chan

GLY

Holden

Morrison

Wyatt

Ruth

. Effects of catechol-O-methyltransferase inhibition on the rates of uptake and reversibility of 6-fluoro-L-dopa trapping in MPTP-induced parkinsonism in monkeys. Neuropharmacology 1997; 36: 363–371.

Doudet

Chan

GLY

Holden

Mcgeer

Aigner

Wyatt

6-[F-18]fluoro-L-DOPA PET studies of the turnover of dopamine in MPTP-induced parkinsonism in monkeys. Synapse 1998; 29: 225–232.

Kyono

Takashima

Katayama

Kawasaki

Zochi

Gouda

Use of [18F]FDOPA-PET for in vivo evaluation of dopaminergic dysfunction in unilaterally 6-OHDA-lesioned rats. EJNMMI Res 2011; 1: 25.

10.

Huang

Barrio

Grafton

Melega

Hoffman

Kinetics and modeling of L-6-[F-18] fluoro-dopa in human positron emission tomographic studies. J Cereb Blood Flow Metab 1991; 11: 898–913.

11.

Patlak

Blasberg

. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data - generalizations. J Cereb Blood Flow Metab 1985; 5: 584–590.

12.

Logan

Fowler

Volkow

Wang

Ding

Alexoff

. Distribution volume ratios without blood sampling from graphical analysis of PET data. J Cereb Blood Flow Metab 1996; 16: 834–840.

13.

Holden

Doudet

Endres

Chan

GLY

Morrison

Vingerhoets

FJG

Graphical analysis of 6-fluoro-L-dopa trapping: effect of inhibition of catechol-O-methyltransferase. J Nucl Med 1997; 38: 1568–1574.

14.

Koprich

Johnston

Huot

Reyes

Espinosa

Brotchie

. Progressive neurodegeneration or endogenous compensation in an animal model of parkinson's disease produced by decreasing doses of alpha-synuclein. Plos One 2011; 6.

15.

Scott

Tabarean

Tang

Cartier

Masliah

Roy

. A pathologic cascade leading to synaptic dysfunction in alpha-synuclein-induced neurodegeneration. J Neurosci 2010; 30: 8083–8095.

16.

Ito

Fujita

Shimada

Watanabe

Okada

Kusuoka

Comparison between the decrease of dopamine transporter and that of L-DOPA uptake for detection of early to advanced stage of Parkinson's disease in animal models. Synapse 1999; 31: 178–185.

17.

Forsback

Niemi

Marjamaki

Eskola

Bergman

Gronroos

Uptake of 6-[F-18]fluoro-L-dopa and [F-18]CFT reflect nigral neuronal loss in a rat model of Parkinson's disease. Synapse 2004; 51: 119–127.

18.

Sharma

El ReFaey

Ebadi

. Complex-1 activity and F-18-DOPA uptake in genetically engineered mouse model of Parkinson's disease and the neuroprotective role of coenzyme Q(10). Brain Res Bull 2006; 70: 22–32.

19.

Cumming

Boyes

Martin

WRW

Adam

Grierson

Ruth

The metabolism of [F-18] 6-Fluoro-L-3,4-dihydroxyphenylalanine in the hooded rat. J Neurochem 1987; 48: 601–608.

20.

Reith

Dyve

Kuwabara

Guttman

Diksic

Gjedde

. Blood-brain transfer and metabolism of 6-[F-18]Fluoro-L-Dopa in rat. J Cereb Blood Flow Metab 1990; 10: 707–719.

21.

DeJesus

Haaparanta

Solin

Nickles

. 6-FluoroDOPA metabolism in rat striatum: time course of extracellular metabolites. Brain Res 2000; 877: 31–36.

22.

Pauwels

Dethy

Goldman

Monclus

Luxen

. Effect of catechol-O-methyl transferase inhibition on peripheral and central metabolism of 6-[F-18]Fluoro-L-Dopa. Eur J Pharmacol 1994; 257: 53–58.

23.

Okada

Nakao

Hosoi

Zhang

Fukumura

Suzuki

Microdialysis with radiometric monitoring of L-[beta-(11)C]DOPA to assess dopaminergic metabolism: effect of inhibitors of L-amino acid decarboxylase, monoamine oxidase, and catechol-O-methyltransferase on rat striatal dialysate. J Cereb Blood Flow Metab 2011; 31: 124–131.

24.

Stout

Huang

Namavari

Satyamurthy

Yee

Barrio

. Distribution volume of 3-O-methyl-6-[F-18]DOPA in rat brain. J Cereb Blood Flow Metab 2000; 20: 1717–1724.

25.

Baker

Abate

Szabo

Joh

. Species-specific distribution of aromatic L-amino-acid decarboxylase in the rodent adrenal-gland, cerebellum, and olfactory-bulb. J Comp Neurol 1991; 305: 119–129.

26.

Ungerste

. 6-hydroxy-dopamine induced degeneration of central monoamine neurons. Eur J Pharmacol 1968; 5: 107–110.

27.

Robinson

Whishaw

. Normalization of extracellular dopamine in striatum following recovery from a partial unilateral 6-Ohda lesion of the substantia nigra -a microdialysis study in freely moving rats. Brain Res 1988; 450: 209–224.

28.

Zigmond

Abercrombie

Berger

Grace

Stricker

. Compensations after lesions of central dopaminergic-neurons - some clinical and basic implications. Trends Neurosci 1990; 13: 290–296.

29.

Masuo

Pelaprat

Scherman

Rostene

. [H-3] dihydrotetrabenazine, a new marker for the visualization of dopaminergic denervation in the rat striatum. Neurosci Lett 1990; 114: 45–50.

30.

Sossi

Holden

Topping

Camborde

Kornelsen

McCormick

In vivo measurement of density and affinity of the monoamine vesicular transporter in a unilateral 6-hydroxydopamine rat model of PD. J Cereb Blood Flow Metab 2007; 27: 1407–1415.

31.

Kaakkola

Wurtman

. Effects of COMT inhibitors on striatal dopamine metabolism - a microdialysis study. Brain Res 1992; 587: 241–249.

32.

Paxinos

Watson

. The Rat Brain in Stereotaxic Coordinates. Academic Press: Orlando, FL, USA, 1997.

33.

Kelly

Iversen

. Selective 6OHDA-induced destruction of mesolimbic dopamine neurons - abolition of psychostimulant-induced locomotor-activity in rats. Eur J Pharmacol (1976); 40: 45–56.

34.

Namavari

Bishop

Satyamurthy

Bida

Barrio

. Regioselective radiofluorodestannylation with [F-18] F2 and [F-18] Ch3Coof - a high-yield synthesis of 6-[F-18]Fluoro-L-Dopa. Appl Radiat Isot 1992; 43: 989–996.

35.

Adam

Jivan

Huser

. C-11-methylations using C-11-methyl iodide and tetrabutylammonium fluoride. Radiochim Acta 2000; 88: 207–209.

36.

Kim

Lee

Kim

Lee

Performance measurement of the microPET focus 120 scanner. J Nucl Med 2007; 48: 1527–1535.

37.

McLellan

Doudet

Brucke

Aigner

Cohen

. New rapid analysis method demonstrates differences in 6-(F-18)Fluoro-L-Dopa plasma input curves with and without carbidopa and in hemi-Mptp lesioned monkeys. Appl Radiat Isot 1991; 42: 847–854.

38.

Topping

Dinelle

Kornelsen

McCormick

Holden

Sossi

. Positron emission tomography kinetic modeling algorithms for small animal dopaminergic system imaging. Synapse 2010; 64: 200–208.

39.

Sossi

Holden

Fuente-Fernandez

Ruth

Stoessl

. Effect of dopamine loss and the metabolite 3-O-methyl-[F-18]fluoro-dopa on the relation between the F-18-fluorodopa tissue input uptake rate constant K-occ and the [F-18]fluorodopa plasma input uptake rate constant K-i. J Cereb Blood Flow Metab 2003; 23: 301–309.

40.

Jonkers

Sarre

Ebinger

Michotte

. Benserazide decreases central AADC activity, extracellular dopamine levels and levodopa decarboxylation in striatum of the rat. J Neural Transm 2001; 108: 559–570.