Tải bản đầy đủ - 0 (trang)


Tải bản đầy đủ - 0trang

this compound did not produce PET images that showed any particular

regional pattern of brain localization when injected into a baboon. It

showed poor uptake but was widely distributed in the brain. Clearance of

radioactivity that did appear to enter the brain was rapid. Furthermore,

uptake of radioactivity in the skull was apparent, which suggested in vivo

decomposition of the radiotracer, leading to the production of labeled

fluoride ion, which then accumulated in bone. It is likely therefore that the

PET images represented only nonspecific uptake of the tracer with a

negligible component due to specific binding to cannabinoid receptors.

Studies with [18F] D8-THC supported the view that a successful

radiotracer must have adequate metabolic stability and a fairly high

affinity for the CB1 receptor to ensure that radioactivity is retained in

brain tissues long enough for tomographic measurement. Furthermore,

a good radiotracer should exhibit high uptake into the brain. This is

likely to be a difficulty for cannabinoid receptor radioligands, since

these molecules are extremely lipophilic. High log P values are generally

associated with poor blood – brain barrier penetrability, presumably

because they remain dissolved in lipid structures in the blood during

transit through the brain capillary bed. The nonclassical cannabinoid

CP55,940, the aminoalkylindole WIN55, 212-2, and THC are reported

to possess log P values of about 6, 5, and 7, respectively. Even the

lowest log P value in this series (5) has been associated with poor brain

penetration in other classes of molecules (see, e.g., Refs. 18, 19).

Furthermore, although [3H]WIN55,212-2 has an affinity about 10-fold

higher than THC, it does not exhibit preferential localization in CB1receptor-rich areas of the mouse brain when injected intravenously

(Gifford et al., unpublished).

Figure 2 Incorporation of fluorine-18 into 5V-[18F]fluoro-D8-THC.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Figure 3 Relationships between structures of SR141716A, AM251, and AM281.

The introduction of the diarylpyrazole CB1 receptor antagonist

SR141716A [12] immediately suggested that exploitation of this class of

molecules might lead to development of a successful CB1 receptor radioligand. Not only is the affinity of SR141716A of the order of 100-fold

higher than that of THC, but the structure suggested that the log P value

would be considerably lower. In addition, SR141716A is highly selective

for the brain cannabinoid receptor (CB1) relative to the CB2 receptor

found in cells of the immune system. It is also an antagonist, which is a

potential advantage because, in the binding of antagonists of G-proteincoupled receptors, there is no discrimination between receptors in different

affinity states. This is unlike the situation with agonists, which bind predominantly to a high-affinity state of the receptor. Finally, SR141716A

contains three chlorine atoms, suggesting that replacement of one of these

with a radioactive iodine atom might produce a compound with the desired


Our ‘‘mark I’’ pyrazole radioligand, code-named AM251, was synthesized in nonradioactive and radioactive forms [20,21]. Following

intravenous injection in mice and rats, the radioiodinated compound

concentrated preferentially in brain areas known to contain densities

of CB1 receptors [22]; however, it failed to enter the brain in SPECT

experiments conducted with baboons [23]. On the hypothesis that this

failure was associated with too high a log P value, we synthesized a

related, ‘‘mark II’’ radioligand with an additional structural modification. This was replacement of the piperidine ring of SR141716A and

AM251 with the more polar morpholino ring (Fig. 3). This compound

AM281, was able to visualize CB1 receptors in baboon (Fig. 4) and

rodent (Fig. 5) brains in vivo [23].

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Figure 4 Sagittal sections of the brains of two baboons injected intravenously

with [123I]AM251 (left) or [123I]AM281 (right). These experiments indicated that

AM281 is to penetrate the baboon brain much more readily than AM251.

Figure 5 Ex vivo autoradiography of [123I]AM281 in rat brain gave distribution

patterns that were essentially identical to in vitro autoradiographs obtained using

tritiated high affinity agonists.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Figure 6 Comparison of the behavior of radioidinated AM251 and AM281

in mice.

The degree of failure of AM251 to enter the baboon brain, as assessed

by SPECT scanning, was surprising because in rats and mice AM251 does

clearly enter the brain. Figure 6 presents a comparison of AM251 and

AM281 in mice. In absolute terms, the graphed data should be interpreted

cautiously because the experiments with the two radioligands were not

conducted simultaneously, and other experimental details were not identical [22,23]. However, apparently greater brain uptake of AM281 at early

times is consistent with its smaller log P value. Moreover, it is clear that

there was significant clearance of AM281 between 30 and 120 min, whereas

there was no significant difference between the 30, 60, and 120 min data

points for AM251. The in vivo brain uptake data, which show a more

prolonged retention of AM251, were thus consistent with the in vitro

binding data, which indicated that AM251 has an approximately threefold

higher affinity for the CB1 receptor than AM281. The mouse data shown

in Figure 6, however, did not predict the large difference seen in baboons

(Fig. 4). It may be that tight binding of AM251 to a specific blood protein,

rather than a greater distribution of AM251 into lipophilic blood components, is responsible for its low brain penetration in baboons.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Although our SPECT studies with [123I]AM281 provided a proof of

principle that in vivo imaging of CB1 receptors in primates is possible, the

behavior of this radioligand is far from ideal in that, relative to radiotracers

used to study other neurochemical systems, its uptake in the brain is low

and its clearance is rapid. These factors limit the count rate obtained in

radionuclide imaging studies, and thus the quality of the images. Because

PET is more sensitive than SPECT by an order of magnitude, a CB1

receptor radioligand labeled with fluorine-18 with pharmacokinetic properties to those of similar AM281 would probably be quite acceptable for

human use. On the other hand, a radioiodinated compound for use with

SPECT would be more useful if it had at least the initial brain uptake of

AM281, but a higher affinity, to ensure a longer clearance time. Since our

initial published work with AM281, several other candidate radioligands

have been prepared [24 – 26]. However, to our knowledge no reports of

human studies have appeared.


In our own laboratories, we have continued to synthesize and evaluate new

labeled cannabinoid receptor radioligands. One of these is [18F]AM284,

where the labeled atom is part of a fluoropentyl group on position 1 of the

pyrazole ring (Fig. 7; see also Table 1). Although this (unpublished) study

demonstrated in vivo binding of AM284 to CB1 receptors, as shown by

the fact that co-injection with SR141716A reduced brain binding, the

Figure 7 Structure of an 18F-labeled pyrazole ligand; for brain uptake (see Table


Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Table 1 Brain Uptake Data for the


F-labeled Pyrazole Ligand of Figure 7

Injected activity in whole brain (%)a


15 min

60 min (vehicle)

60 min (+ SR14176A)



0.80 F 0.05

0.66 F 0.05*

0.33 F 0.05***

0.07 F 0.008

0.028 F 0.004**

0.020 F 0.001****


p < 0.003; cf. 15 min time point.

p < 0.001; cf. 15 min time point.


p < 0.001; cf. vehicle.


p < 0.035; cf. vehicle.


Values are the mean FSD (n = 5).


binding was less than one-tenth that of AM281 measured simultaneously

in a dual-isotope experiment. It would therefore not be a practicable PET

radioligand. Other fluorine-18 and radioiodinated cannabinoid receptor

ligands are being synthesized and studied in our laboratories.

While we have not yet started human PET or SPECT studies, we have

used AM281 to conduct fundamental studies of the CB1 receptor system in

Figure 8 Comparison of ability of WIN 55, 212-2 to sedate mice (triangles) and

to block specific binding of [131I]AM281 (squares).

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

rodents. For example, during experiments in mice designed to measure the

occupancy of CB1 receptors associated with physiological effects of

exogenous cannabinoids, we found that doses of WIN 55,212-2 that

induced profound sedation did not reduce binding of AM281 to cerebellum and hippocampus (Fig. 8). This observation indicates that the

occupancy of the CB1 receptor necessary for physiological effects of

cannabinoids is very low [27].

Experiments in superfused hippocampal slices prepared from rats

were then conducted to compare inhibition of acetylcholine release by the

cannabinoid receptor agonist WIN55,212-2 with inhibition of AM281

binding by this agonist. The results (Fig. 9) show that half-maximal

response is achieved at less than 1% occupancy, confirming that the

agonist occupancy necessary to produce a physiological response is very

low in the CB1 system [27].

A consequence of a very large receptor reserve for CB1 receptors

would be that PET or SPECT could not be used to image the occupancy of

the CB1 receptor by biologically significant doses of agonist drugs. This is

Figure 9 Graph of inhibition of acetylcholine release in superfused hippocampal slices versus CB1 receptor occupancy estimated from inhibition of

[131I]AM281 binding.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

because even high doses of an agonist would not displace an appreciable

fraction of the radiotracer binding. On the other hand, these experiments

indicated that the binding of exogenous CB1 receptor antagonist radioligands should not be affected by changes in the levels of endogenous

ligands such as anandamide. This would thus remove a possible confounding factor in imaging experiments designed to detect changes in

cannabinoid receptor densities.

In the light of these experiments with WIN55,212-2, it is of interest to

speculate on the degree of occupancy of brain CB1 receptors that is

achieved by doses of THC that induce desired effects in human during

the smoking of a marijuana cigarette. THC is a partial agonist with an

efficacy of 20 to 25% [28,29], so that it would act at a higher receptor

occupany than a full agonist like WIN55,212-2. Intravenous doses of 0.5

mg/kg THC are effective in humans [30], and if 1% of the injected dose is

distributed in the brain [31], this would correspond to a concentration of

about 15 nmol/L. Applying the mass action equation and assuming that an

in vitro Kd value for the CB1 receptor of 100 nM is appropriate, and a Bmax

value of 100 nmol/L, an occupancy of 7% is estimated. However, it is likely

that the fraction of THC available for binding to the receptor in vivo is

quite small, since as an extremely lipophilic molecule, it will be distributed

in brain membranes. This is expected to increase the effective Kd value in

vivo and so lower the estimate of occupancy, possibly by more than one

order of magnitude. These considerations, therefore, suggest that only a

very small proportion of the brain CB1 receptors need be activated to

induce psychoactive effects in humans, consistent with our results in mice,

and that PET studies will not be able to measure the degree of occupancy

achieved by marijuana smokers.

Similar studies were done to evaluate the relationship between level

of occupancy of the CB1 receptor by nonradioactive AM281 and the

degree to which the antagonist AM281 was able to reverse the sedative

effect of the agonist WIN55,212-2 [32]. The AM281 effectively restored

the activity to normal levels (Fig. 10). In addition, AM281 alone was

found to significantly stimulate locomotor activity between 1 and 2 h

after its administration (Fig. 11). Both the antagonism of the effect of

WIN55,212-2 and the effect of AM281 alone increased progressively with

doses up to 0.3 mg/kg AM281, but did not further increase at 1 mg/kg.

A 50% occupancy of the CB1 receptor, as assessed by inhibition of

[131I]AM281 binding, was achieved at a dose of 0.45 mg/kg. These data

are consistent with prior in vitro indications that AM281 is a CB1 receptor antagonist or inverse agonist [33] and that AM281 inhibits an

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Figure 10 Effect of increasing doses of AM281 on binding of [131I]AM281 in

cerebellum and hippocampus, using brain stem as reference tissue.

endogenous cannabinoid tone. This baseline activity of the CB1 system

might be maintained either by constitutive activity of the receptor [34], or

by endogenous agonists such as anandamide [11]. These experiments [32]

indicate that PET could be used to measure the degree of occupancy of

CB1 receptors by antagonist or inverse agonist drugs in the human brain,

if these drugs turn out to have useful therapeutic effects, such as reducing

memory loss in the elderly [35 –37].


Our development of [123I]AM281, an antagonist radioligand for brain

cannabinoid receptors, has allowed us to image this receptor for the first

time in vivo. Ex vivo autoradiographic experiments have been conducted

in rodents, and SPECT studies have been conducted in baboons. Research

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Figure 11 Stimulation of locomotor activity at CB1 receptor occupancy levels

calculated from the effect of increasing doses of AM281 on binding of

[131I]AM281; squares, antagonism of the sedative effect of WIN55,212-2 at 0

to 15 min; triangles, induction of hyperactivity by AM281 alone at 61 to 120 min.

continues to develop superior radioligands for SPECT research and also

to develop CB1 receptor radioligands that can be labeled with positronemitting nuclides for PET. The results of the animal work to date will

provide the foundation for using AM281 and other cannabinoid receptor

radioligands in human imaging experiments. It is predicted from our

animal data that psychoactive or medicinal doses of agonists such as THC

will not alter CB1 receptor radioligand binding in the human brain. On

the other hand, PET or SPECT is likely to be useful in determining the

degree of CB1 receptor occupancy necessary for therapeutic effects of

antagonist drugs, as well as in evaluating CB1 receptor changes in

addiction and in other diseases.


This research was carried out at the Brookhaven National Laboratory

under contract DE-AC02-98CH10886 with the U.S. Department of

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Energy and supported by its Office of Health and Environmental

Research. The research was also supported by awards from the National

Institute on Drug Abuse to AM (DA 07515, DA 09158) and to ANG

(DA 12412).


1. Wolf AP, Fowler JS. Positron emission tomography. Biomedical research

and clinical application. Neuroimaging Clin N Am 1995; 5: 87 – 101.

2. Fowler JS, Wolf AP. The heritage of radiotracers for positron emission

tomography. Acta Radiol Suppl 1990; suppl 374: 13 – 16.

3. Gatley SJ, Volkow ND. Addiction and imaging of the living human brain.

Drug Alcohol Depend 1998; 51: 97 – 108.

4. Volkow ND, Fowler JS, Gatley SJ, Logan J, Wang GJ, Ding YS, Dewey

SL. PET evaluation of the dopamine system of the human brain. J Nucl

Med 1996; 37: 1242 – 1256.

5. Volkow ND, Gillespie H, Mullani N, Tancredi L, Grant C, Ivanovic M,

Hollister L. Cerebellar metabolic activation by delta-9-tetrahydrocannabinol in human brains. A study with positron emission tomography and 18F-2deoxy-2-fluoro-D-glucose. Psychiat Res 1991; 40: 69 – 80.

6. Mathew RJ, Wilson WH, Coleman RE, Turkington TG, DeGrado TR.

Marijuana intoxication and brain activation in marijuana smokers. Life Sci

1997; 60: 2075 – 2089.

7. Mathew RJ, Wilson WH, Turkington TG, Coleman RE. Cerebellar activity

and disturbed time sense after THC. Brain Res 1998; 797: 183 – 189.

8. Volkow ND, Gillespie H, Mullani N, Tancredi L, Grant C, Valentine A,

Hollister L. Brain glucose-metabolism in chronic marijuana users at baseline and during marijuana intoxication. Psychiatr Res Neuroimaging 1996;

67: 29 – 38.

9. Jansen EM, Haycock DA, Ward SJ, Seybold VS. Distribution of

cannabinoid receptors in rat brain determined with aminoalkylindoles.

Brain Res 1992; 575: 93 – 102.

10. Herkenham M, Lynn AB, Little MD, Johnson MR, Melvin LS, DeCosta

BR, Rice KC. Cannabinoid receptor localization in brain. Proc Natl Acad

Sci USA 190; 87: 1932 – 1936.

11. Felder CC, Briley EM, Axelrod J, Simpson JT, Mackie K, Devane WA.

Anandamide, an endogenous cannabimimetic eicosanoid, binds to the

cloned human cannabinoid receptor and stimulates receptor-mediated signal

transduction. Proc Natl Acad Sci USA 1993; 90: 7656 – 7660.

12. Rinaldi-Carmona M, Barth F, Heaulme M, Shire D, Calandra B, Congy C,

Martinez S, Maruani J, Neliat G, Caput D, Ferrara P, Soubrie P, Breliere

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về


Tải bản đầy đủ ngay(0 tr)