Brit. J. Pharmacol. (1953), 8, 366.

TISSUE DISTRIBUTION OF a-'4C-LABELLED ALKYLOXYa-PHENYLETHYLAMINES IN RELATION TO

ANALGESIC ACTIVITYBY

J. B. BRIERLEY AND A. McCOUBREYFrom the Departments of Neuropathology and Biochemistry, Institute of Psychiatry (British Postgraduate Medical

Federation, University of London), Maudsley Hospital, S.E.5

(RECEIVrD APRIL 18, 1953)

Minor changes in chemical structure couldinfluence the physiological response to a drug bycausing changes in those physical properties whichdetermine access to susceptible tissues. The tissuedistributions of '4C-labelled p-cyclohexyloxy-a-phenylethylamine (I) and p-isopropyloxy-a-phenyl-ethylamine (II) have been studied with this conceptin mind, since I has morphine-like properties notpossessed by II (McCoubrey, 1953). The a-C-atoms(marked with an asterisk in formulae I and Ii) werethe labelled atoms.

(I)

CH * NH2CH/ -DK CH.

(II)

METHODSThe amines were given as the hydrochlorides in water.

The doses (40 mg./kg. i.p. and 10 mg./kg. i.v.) weresufficient for I, but not for II, to induce well-markedanalgesia (M.E.D. 10-15 mg./kg. i.p.). Animals werekilled by excision of the heart under ether anaesthesia(3 minutes) unless stated otherwise. Tissues were alwaysdissected in the same order. There was little changein tissue concentration during thirty minutes afterexsanguination (Brierley, unpublished). Visceral tissueswere removed rapidly after intraperitoneal injection toavoid an observed post-mortem absorption of drug fromthe peritoneal cavity. To obtain sufficient tissue, sixmale albino rats (300-350 g.) were used for each timeinterval. Approximately equal amounts of any onetissue were taken from each animal and pooled. Rabbitsof one litter were used for each drug, two animals pertime interval. The analgesia due to I was sufficient toallow killing, without anaesthesia, by medullary sectionthrough the atlanto-occipital membrane, sometimes

after withdrawal of cerebrospinal fluid. Ether wasrequired for II. The monkey (Macaca mulatta, cf,3.5 years) was lightly anaesthetized (ethyl chloride, thenether) before giving I into the saphenous vein.

Defi'ition of Tissues.-Meninges and vessels wereremoved as completely as possible. Cerebral cortex-thin slices excluding visible white matter. Subcorticalwhite matter-substantial contamination with greymatter is inevitable in small animals. Monkey corpuscallosum is essentially pure white matter. Thalam is-isolated by coronal cuts at the interventricular foramenand habenula, laterally by an oblique cut at the junctionwith the internal capsule, inferiorly at the hypothalamicsulcus. Hypothalamus-from optic chiasma anteriorlyto the mammillary bodies posteriorly and extendingsuperiorly to the hypothalamic sulcus. Cerebellum-midline vermis with white core removed. Medqlla-fromauditory tubercle to obex with removal of area postrema.Spinal cord-cervical or lumbo-sacral enlargements. Incertain cases cervical cord grey matter was carefullydissected from the white matter. Dorsal root ganglia-lumbo-sacral and cervical ganglia were pooled from rats,but were treated as separate groups in rabbit and monkey.The ventral roots were removed. Peripheral nerve anddorsal root were removed as close to the ganglion aspossible. Sympathetic ganglia-superior cervical andstellate ganglia. Adrenals-rat adrenals were decap-sulated. Rabbit adrenals were cut in three. The twoouter thirds (" cortex ") were assayed separately fromthe inner third.Radioassay.-All tissues were dried at 600 C. 50 or

30 mg. of powdered tissue were assayed as " infinitethickness" samples (Popji!k, 1950) under a G.E.C.EHM2 mica window (1.9 mg./cm.2) counter. Thecounting efficiency by comparison with a standard14C-polythene disc was 1.9%. Where insufficient drytissue was available, kieselguhr was added and thoroughlymixed to make 30 mg. A linear relation between suchdilution and activity was found in trial assays withvarious tissues. 2,000 counts, including background(10-13 counts per minute), were recorded for mosttissues (error not greater than ni4.5y.). 1,000-1,500counts were recorded for tissues with low activity

TISSUE DISTRIBUTION OF ANALGESICS

1.40

1.20

1.00ai

J. B. BRIERLEY and A. McCOUBREY

0.7

0.6

0.5

0.4

0.3

0.2

0.1

2.5 5 10 5TIME (minutes) (a)

FIG. 2.-Rabbit tissue concentrations at various times after p-cyclohexyloxy-a-phenylethylamine hydrochloride (10 mg./kg. i v.).* Indicates where tissues were pooled from two animals. (a) Gives the tissue activities of p-isopropyloxy-a-phenylethyl-amine under comparable conditions at five minutes.

DiscussIoN

I and Il, like most basic substances, readilypenetrate nervous tissues (Friedcmann, 1942). TableII clearly shows that pharmacological differencesbetween these amines cannot be explained in termsof access to tissues. These amines have nomarkedly different preferences for nervous tissues,and II, which is not analgesic in the rat, penetratesall rat tissues examined more readily than I. Theconverse applies to the rabbit, but no pharmacolo-

gical data with respect to these substances areavailable for this species. The tissue concentration-time curve (Fig. 1) is similar to the analgesictime-action curve for I in the rat and does notcorrespond to the time course of respiratorydepression (McCoubrey, 1953). The ephemeralaction of I is probably due to rapid elimination bythe kidney. It was not possible to show anymarked plasma activity, even after intravenousdosage, and the amines appear to be rapidly con-

LU

#AUL

F

-

z

0

-

0UAz

368

TISSUE DISTRIBUTION OF ANALGESICS

TABLE IICOMPARATIVE TISSUE CONCENTRATIONS OF p-CYCLO-HEXYLOXY (I) AND p-ISOPROPYLOXY-a-PHENYLETHYL-

AMINE (II) IN RAT AND RABBITActivities are calculated from axn where a= counts per gramme wettissue, n= number of animals, p=counts per mil. in plasma. N=tota'injected counts for the group. The plasma activities (/. injectedcounts per animal per gramme wet tiisue) were respectively: rat:

I, 0.071, II, 0.075; rabbit: I, 0.010, Hi, 0.013

Rat Rabbit10 min. after 5 min. after

Tissue 40 mg./kg. i.p. 10 mg./kg. i.v.I II Ratio I II Ratio

Cerebral cortex .. 10-6 13-5 1 27 - -occipital . - - - 47-0 20-6 0 45frontal - -

-

43-4 20-4 0 47Hypothalamus .. 4-2 8-2 1 95 29-8 12 4 0-42Thalamus .. _ - - 34-6 15 6 0 45Cerebellum .. 61 80 1 31 33*8 15 6 0-46Medulla .. .. 35 7-3 2-10 30-0 13-8 0-46Olfactory bulb . - - - 25 0 13-9 0 56Dorsal root ganglia .. 2-8 44 158 - - -

lumbo-sacral. - - - 10-5 3-5 0-38cervical

-

-_ 15-6 6-6 0-42

Sympathetic gangia .. -_ 24-2 6-6 0-27Adrenal .. .. 124 17-0 1-37 23-2 12-4 0-54Thyroid .. 70 5 9 0-84 67-5 8-8 0-13Kidney .. 17-1 19-6 1-15 42-5 38-0 0-89Muscle .. .. 2-2 2-6 118 6-8 4.4 0 65

centrated in nervous tissue. With the exception ofrat and rabbit (but not monkey) cerebral cortex,there was no regional variation within grey matterwhen allowance was made for white matter con-tamination. The low activity in sympathetic anddorsal root ganglia was certainly due to low cellcontent, since in typical rat lumbo-sacral and cervicalganglia it was shown by projection of Nissl stainedserial sections on to paper that the proportion ofcells to fibres was 35 and 47% respectively (excludingventral roots). In spite of the solubility of theseamine hydrochlorides in non-polar solvents (McCou-brey and Iyengar, 1951) penetration into whitematter was quite low.

It is not possible at present to relate the selectiveconcentration in rat or rabbit cerebral cortex toanalgesic activity. Similar selective localizationdoes not occur in the monkey and in this animalthere is no preference for either sensory or frontalcortex. There is little evidence that pain sensationis normally projected on to the cerebral cortexeven in man. Penfield (1947) has shown thatelectrical stimulation of exposed human cerebralcortex in conscious subjects evokes no painfulsensation, and, while section of frontal lobe tracts(lobotomy) relieves the anxiety associated withpain, it apparently does not reduce pain sensation,and indeed the threshold for experimental painmay be reduced (Chapman, Rose, and Solomon,1948). In isolated cases excision of sensory cortexcan relieve pain of long standing (Lewin andPhillips, 1952), but in cases of intractable pain the

most effective procedure appears to be section ofspinothalamic tracts (Falconer, 1953).There are indications that some, but certainly

not all, aspects of analgesic mechanism are mediatedat a spinal level. Spinal structures were thereforeexamined in some detail, especially the dorsal rootganglia, since all afferent sensory impulses traversethese ganglia. Brierley (1951, 1952) has shownthat they concentrate a number ofions, and, althoughthere are no synapses here, a brief delay occurs inimpulse transmission across the ganglia (Dun,1951), which could indicate actual participation bythe cells in transmission. In the dorsal horn greymatter, Pearson (1952) finds histologically that slowpain fibres make their synapse in the dense neuropilof the substantia gelatinosa Rolandi, whereas fastpain fibres by-pass this region to synapse in thenucleus proprius. This is of interest in view ofCushny's remark (1941) that " sudden shocks "(fast pain ?) are distinguishable under morphinewhile pain (slow pain ?) is under control. Further,spinal animals react equally well in analgesic assayswhich measure depression of avoidance reflexes(Bonnycastle and Leonard, 1950; Houde andWikler, 1951). The appreciable concentration ofdrug, however (approx. 1O-7 moles per g. tissue),which appeared at all cell stations on the painpathway, viz., dorsal root ganglia, spinal cordgrey matter, medulla and thalamus, with little inthe interconnecting fibre tracts and peripheralnerve, suggests that the action of analgesics maybe an attenuation of pain impulses at the successivecell stations.The difference in pharmacological response

between I and II had suggested that some differencesin localization might occur in the nervous system.Marked divergence was only observed in theadrenal and thyroid glands. Mere concentrationdoes not necessarily imply a role for these glandsin analgesia; there is considerable evidence thattheir secretions markedly influence the action ofanalgesics, but discussion can only be speculativein the absence of experimental results.

CHEMICAL SECTIONThe route previously used for synthesis of alkyl-

oxy-a-phenylethylamines was not wholly suitablefor 14C-labelling. p-Cyclohexyloxyphenyl magne-sium bromide was carbonated to give the carboxylicacid, whence the lithium salt and lithium methyl(Gilman and Van Ess, 1933) gave p-cyclohexyl-oxyacetophenone. The ketoxime was reduced tothe amine as previously described (McCoubrey andIyengar, 1951). The isopropyl ether was preparedsimilarly.

369

370 J. B. BRIERLEY and A. McCOUBREY

p-Bromophenyl CyclohexylEther.-p-Bromophenol(20 g.) and cyclohexyl bromide (57 g.) by themethod described (McCoubrey and Iyengar, 1951)gave the ether (10.9 g.), b.p. 115-117/0.8 mm. Itcrystallized from light petroleum in prisms m.p.40-41. (Found: Br, 31.3. C12H41OBr requiresBr, 31.4%.)p-Bromophenyl Isopropyl Ether.-Prepared con-

ventionally. The ether was a fragrant oil, b.p.1 19-120/20 mm., n2D 1.5368. It could be crystal-lized from light petroleum m.p. 17.50. (Found:C, 51.4; H, 4.2; Br, 36.5. C9HLLOBr requiresC, 50.2; H, 5.1; Br, 37.2%.)

p-Cyclohexyloxy- and p-Isopropyloxybenzoic Acid.-The above bromides were converted to Grignardreagents by boiling with magnesium and a fewdrops of ethyl iodide in ether under nitrogen forfive hours. The filtered solutions were carbonatedin conventional manner with carbon dioxidegenerated from barium carbonate. The productswere isolated conventionally. The cyclohexyl ether(77%) crystallized from ethanol in plates m.p. 1760.(Found: C, 70.7; H, 7.2. C13H1603 requiresC, 70.1; H, 7.3%.) The isopropyl ether (72%l)crystallized from ethanol in plates m.p. 1590 (softensat 140). (Found: C, 66.1; H, 6.4. C10H1203requires C, 66.7; H, 6.7%.)The acids were dissolved in ether and were treated

cautiously with three molecular proportions oflithium methyl and boiled overnight. The aceto-phenones were isolated, converted to the ketoximesand reduced by sodium amalgam to the amines.

a-14C-laballedp-cyclohexyloxy-a-phenylethylaminewas obtained in 40% overall yield (1.28 g.) with anactivity of 8,296 counts per minute per mg. a-l4C_lab-l-ed p-isopropyloxy-a-phenylethylamine was

obtained in 40% overall yield (0.95 g.) with anactivity of 10,562 counts per minute per mg. Yieldsare calculated on barium carbonate used.

SUMMARYp-Cyclohexyloxy- and p-isopropyloxy-a-phenyl-

ethylamine hydrochlorides penetrated with equalfacility into rat, rabbit, or monkey cerebral greymatter. A selective localization in rat and rabbitcerebral cortex cannot at present be correlatedwith analgesic mechanism. There was no local-ization in peripheral nerve.

The authors thank the Governors of the BethlemRoyal and Maudsley Hospitals for a grant from theResearch Fund. Assistance with counting was givenby Miss N. Billery.

REFERENCESBonnycastle, D. D., and Leonard, C. S. (1950). J.

Pharmacol., 100, 141.Brierley, J. B. (1951). J. Physiol., 116, 24P.- (1952). Ibid., 117, 6P.

Chapman, W. P., Rose, A. S., and Solomon, H. C.(1948). Proc. Ass. Ros. nerv. Dis., 27, 754.

Cushny, A R. (1941). Pharmacology and Therapeutics,12th ed., p. 378. London: Churchill.

Dun, F. T. (1951). J. cell. comp. Physiol., 38, 131.Falconer, M. A. (1953). Brit. med. J., 1, 299.Friedemann, U. (1942). Physiol. Rev., 22, 125.Gilman, H., and Van Ess, P. R. (1933). J. Amer. chem.

Soc., 55 1258.Houde, R. W., and Wikler, A. (1951). J. Pharmacol.,

103, 236.Lewin, W., and Phillips, C. G. (1952). J. Neurol.

Psychiat., 15, 143.McCoubrey, A. (1953). Brit. J. Pharmacol., 8, 22.- and Iyengar, N. K. (1951). J. chem. Soc., 3430.

Pearson, A. A. (1952). Arch. Neurol. Psychiat., 68, 515.Penfield, W. (1947). Proc. roy. Soc., 134B, 329.PopjAk, G. (1950). Biochem. J., 46, 560.