biological disposition of morphine andits surrogates 1

Bull. Org. mond. Sante 1961, 25, 227-262Bull. Wld Hlth Org.

The Biological Disposition of Morphineand its Surrogates 1 *

E. LEONG WAY, Ph.D.' & T. K. ADLER, Ph.D.'

CONTENTS

Page

INTRODUCTION ..... . . . . . . . . . . . . . . . 227

MORPHINE ...................... 228Methods of estimation ............... 229Absorption .................... 234Distribution ..... . . . . . . . . . . . . . . . 236

Metabolism ..... . . . . . . . . . . . . . . . 243Excretion ..... . . . . . . . . . . . . . . . . 252

REFERENCES ..... . . . . . . . . . . . . . . . . . 259

INTRODUCTION

The last decade has seen the elucidation of thestereochemistry (Rapoport & Payne, 1950, 1952;Rapoport & Lavigne, 1953; Ginsburg, 1953) andstructural configuration (Lindsay & Barnes, 1955)of the morphine molecule, as well as the verificationof the structure by total synthesis (Gates & Tschudi,1952); this has supplemented previous chemicalknowledge of the opium alkaloids (Small, 1932;Bentley, 1954). Moreover, the relationship betweenchemical structure and biological activity has beenexplored not only for morphine but for the syntheticsubstances that can act as surrogates for morphinein vivo (Bockmuhl & Ehrhart, 1949; Eddy, 1950;Schwartzman, 1950; Beckett & Casy, 1954; Braen-den, Eddy & Halbach, 1955; Reynolds & Randall,1957). From these studies it is apparent that muchprogress has been made towards delineating those

* This study on the biological disposition of morphineand its surrogates is being published in the Bulletin of theWorld Health Organization in four instalments. This-thefirst-is devoted to morphine per se; the second and thirdinstalments will deal, respectively, with derivatives of mor-phine and synthetic surrogates of morphine, and the finalinstalment will discuss general considerations. The fourinstalments will eventually be available as a joint reprint.

1 Department of Pharmacology, University of CaliforniaMedical Center, San Francisco, Calif., USA. The authorswere aided in their preparation of this report respectively byGrant RG-1839 from the National Institutes of Health andby Senior Research Fellowship SF 271 from the PublicHealth Service, US Department of Health, Education, andWelfare.

structural features that endow these molecules withtheir biological potential. This potential can berealized only when the molecule has reached itstarget organ, for the very nature of the pharmacolo-gical response indicates a highly selective action atspecific receptor sites within the central nervoussystem. Accordingly, the onset, intensity, and dura-tion of action will be governed by the relative easewith which the molecules can reach the effective site;this, in turn, depends upon a series of physico-chemical processes that regulate the biological dis-position of the compound. The present study con-siders in detail the absorption, distribution, bio-transformation and elimination of each compound,as well as evaluating the analytical methods used inthe investigation of these processes. Wheneverpossible, the significance of the findings in relationto the over-all pharmacology of the compound isdiscussed. The study is divided into four parts. Thefirst three parts, which deal, successively, withmorphine, the partially synthetic derivatives ofmorphine and the wholly synthetic substances withmorphine-like action, are presented in the form ofindividual monographs. The structure of eachagent is shown in Fig. 1. The final part is devoted toa discussion of general considerations pertinent tothis group of drugs as a whole.The material for this study has been selected

mainly from papers published within the past twentyyears, attention being directed primarily to those

1042 -227-

E. LEONG WAY & T. K. ADLER

FIG. I

MORPHINE AND MORPHINE-MIMETIC COMPOUNDS a

NATURALLY N-CM5 N-CM

OCCURRING CHI cm

OPIUMALKALOIDS 4"o ON c-

MOAHIM CODEINE

N-CM5 N-CM5 N-CM3NCH-HUHPARTIALLY ICMSYNTHETIC

DERIVATIVES OFMORPHINE O%G14- -O 0 0-C-CM5 Mo 0 0 04g-0 0 0 No 0 ONMORPHINE q__ -,,csH.ooozo

MUA0OIN DIMYDROMORPHMINNONE AYDROCODEINONE NALOAHINI

SYNTHETIC N-MHNC NC-*M

COMPOUNDS 0 -HMORPHINANS MO C,5-o Mo

LEVORAPNOL DEXTPOMETMORPMAN LEVALLOAPHN

CM,^-C-O Q~CM5 ,cH} 0 iCMC /CHM/CM-N CM- CMCCM CCQM WsMMCN~tF N-M0CC-M-METHADONES oC CCHS_CHMN CH3-CH-CH-C-CHZ-CH-N\ 0

METHADONE ArETYLMETHADOL U AONI POPOXYPHENtE

-C-0-cma-CM5 O-C-CaCN-CM5aCH L -X0C-CMI-CH5PHENYL- M 4M CHS It H C H

PIPERIDINES HaN Ha Ha HSMgNgMaMEPERIDINE AUM%APODINL KETOSEMIDONE ANIUDIN

-°-CH-04 NI-CH-^MISCELLANEOUS HSM CMa

N3 CHI

ETMOMEPTAZINE PM4NAZOCINE

a The compound " meperidine " is henceforth referred to in these studies by its proposed international non-proprietary name,pethidine.

papers appearing since the publication of the mono- more recent findings. Certain aspects of this materialgraph on the pharmacology of the opium alkaloids have been reviewed elsewhere in a more limitedby Krueger, Eddy & Sumwalt (1941, 1943) to the fashion (Krueger, 1955; Peterson, 1955; Perrine &time of writing (December 1960). Occasionally, Eddy, 1956; Reynolds & Randall, 1957; Schaumann,references to the older literature are discussed at 1957; Stolman & Stewart, 1960; Way & Adler,length when this is pertinent to the interpretation of 1960).

MORPHINE

In discussing morphine and its derivatives thepresent authors have adhered to the Gulland &Robinson formula as presented in 1925. Accord-ingly, morphine is presented as a partially saturatedphenanthrene derivative containing an ethane-amine side chain. The fourteen carbon atoms of thephenanthrene part of the formula are drawn andnumbered in the classical manner. The remaining

carbon atoms of the side chain, which, together withnitrogen and carbons 13 and 14 of the phenanthreneskeleton, form a piperidine ring, are numbered 15and 16. No number is assigned to either the ringnitrogen or the ring oxygen. It should be noted,however, that inasmuch as morphine is a complexmolecule, the graphic formula, the numbering of theatoms, and the nomenclature of morphine deriva-

228

BIOLOGICAL DISPOSION OF MORPHINE AND ITS SURROGATES. 1

FIG. 2NUMBERING SYSTEMS USED FOR THE GRAPHIC FORMULA OF MORPHINE

"CLASSICAL"PHENANTHREN E

tives will depend on the relative emphasis placed onparticular structures in the molecule. Schaumann(1940) considers morphine to be primarily a piperi-dine derivative, but this proposal has not gained wideacceptance among chemists or pharmacologists.For the most part, the phenanthrene structure hasbeen emphasized during the thirty-seven years sinceGulland & Robinson (1923) pointed out the strikingtendency of the morphine alkaloids to give phenan-threne derivatives during chemical degradation.Even here several methods of numbering the atomsare in current use. The Ring Index (Patterson &Capell, 1940) regards morphine as a derivative ofimino-ethano-phenanthrene-furan, and both thefuran oxygen and the imino nitrogen are included inthe numbering system. Rasmussen & Berger (1955)have suggested that the resemblance between mor-phine and the steroids be stressed, and in theirgraphic formula the phenanthrene part of themolecule is drawn and numbered in the manner usedin steroid chemistry. Fig. 2 shows the various waysin which the formula for morphine may be drawn.None of these conveys a feeling for the gnarled andtwisted shape of the molecule revealed by X-rayexamination (Lindsay & Barnes, 1955).

METHODS OF ESTIMATION

The major problem encountered in the estimationof morphine in body fluids and tissues is usuallyconcerned with the processes involved in the isola-

PHENANTHROFURAN "STEROID"PHENANTHREN E

tion and separation of the morphine from theextraneous biological phase. The morphine-con-taining moiety finally obtained is rarely free frominterfering substances and often the yield is far fromquantitative. In addition, the small size of the doseof morphine required to elicit the pharmacologicalor toxicological response has often necessitated theuse of large amounts of the biological material forthe analysis.

In speaking of morphine recovered from biologicalmedia, the terms frequently encountered are " freemorphine ", " bound, combined or conjugatedmorphine" and " total morphine ". Free morphineis generally assumed to be the unchanged parentcompound. The terms bound, combined and con-jugated have been used interchangeably to indicatethe morphine released after acid hydrolysis, and thequantity of bound morphine is estimated by takingthe difference between the total and free morphinefound in a given sample. Recent evidence suggeststhat bound morphine is chiefly, if not entirely,morphine-3-monoglucuronide dihydrate, but sincethis has not been established with absolute certaintyin all cases, and since the metabolic pathway ofmorphine has not been completely delineated, wehave, in general, preferred to retain the morefamiliar term, bound morphine. It is presumed thatwhen morphine is referred to in papers publishedbefore the existence of bound morphine was estab-lished that the various authors mean the free orunchanged alkaloid. However, it is probable that in

229


some of these studies, especially when direct tests foralkaloid reactants were performed on the biologicalspecimens without prior purification by adsorption orextraction, some bound morphine was also measured.Many methods have been proposed for the estima-

tion of morphine in biological fluids, but no satis-factory procedure was available prior to 1928. Themethods employed since then are summarized inTable 1, together with an evaluation of their use-fulness. A critique of the earlier methods will befound in the publications by Pierce & Plant (1932)and Wolff, Riegel & Fry (1933). Only those methodswhich have direct bearing on data concerning thebiological disposition of morphine have been in-cluded. Some recent methods for estimatingmorphine for forensic purposes have been cited inthe text, but are not discussed because the applicabi-lity of these methods has been studied only to theextent of noting recovery values after adding mor-phine to tissue; the methods have not been actuallyemployed for investigating the fate of morphine.Most of these forensic methods are usually onlysemi-quantitative but some of them appear to besufficiently sensitive for estimating small amounts ofmorphine in biological fluids. Biological methods forestimating morphine have also been reported, butsuch procedures lack the precision and sensitivity ofchemical methods (Forst & Deininger, 1949;Fichtenberg, 1951; Janssen & Jagenau, 1956). Otherreferences with respect to biological procedures maybe obtained from Schaumann's monograph (1957),which also includes methods for estimating morphinein opium. A comprehensive list of micro-chemicaltests for morphine and its surrogates has been com-piled by Clarke (1959).Most of the methods for estimating morphine

follow a general pattern in that the morphine is re-covered from the biological fluid by processes thatselectively separate the compound from its con-taminants. The purified morphine is then measuredby some general reaction for a phenol or alkaloidalsubstance. The reliability of the procedure is,therefore, chiefly dependent upon the adequacy ofthe purification processes, which usually involveextraction, adsorption or precipitation of morphinewithout actual isolation of the crystalline product.The isolation and separation techniques used are of

two general types: (a) the extraction procedures and(b) the adsorption methods. The extraction proce-dures involve the partition of morphine betweendifferent solvent phases and have, in general, beenmodifications of the classic Stas Otto method. The

method utilizes multiple extraction techniques undervarying pH conditions, the major separation of theimpurities from the morphine being effected by themore or less differential distribution of the compo-nents between various solvent pairs as the pH issystematically changed. Unfortunately, morphine isdifficult to extract into the organic solvent phase,owing to the concomitant presence of the phenolichydroxyl and the tertiary nitrogen groups, whichconfer amphoteric properties on the compound. Ingeneral, optimal extraction of morphine has beenobtained at about pH 9, which is near its isoelectricpoint, using a variety of solvents such as ethylacetate and mixtures of alcohols with chloroform.More recently, Fujimoto, Way & Hine (1954)proposed that the extraction of morphine be effectedwith n-butanol from a highly concentrated alkalinesolution. Application of the procedure to morphinepresent in biological media yielded good recoveriesalong with only negligible quantities of naturallyoccurring interfering substances.

Ion-exchange and adsorption chromatographyhave been developed for the assay of the crude drugand its galenical preparations (Brochmann-Hanssen,1955), but their use in the separation of morphinefrom animal material has been limited (Achor &Geiling, 1954; Tompsett, 1960). In a preliminaryreport, Stewart, Chatterji & Smith (1937) gave pro-mising indications of using adsorption on kaolin, butthe details have not been forthcoming. Stolman &Stewart (1949) subsequently developed the methodof adsorbing the morphine on to a column ofmagnesium trisilicate (Florisil), and fractionallyeluting the morphine with methanol. Other studies(Oberst, 1938-39; Fischer & Chalupa, 1950; Fischer& Goll, 1950) demonstrate the purifying power ofadsorbents. Paper chromatography methods for theseparation of morphine from its contaminants havebeen reported by many groups (Goldbaum &Kazyak, 1952; Breinlich, 1953; Jatzkewitz, 1954;Kaiser & Jori, 1954; Mannering, Dixon, Carrol &Cope, 1954; Curry & Powell, 1954; Siebert, Williams& Huggins, 1954; Woods, 1954; Schultz & Strauss,1955; Kariyone & Hashimoto, 1957; Vidic, 1955,1957; Way, Kemp, Young & Grassetti, 1960). Asuitable solvent system for application of thecounter-current distribution technique has also beendeveloped; this system consists of phosphate buffer,pH 6.6, and 10% butanol in chloroform (Young &Way, unpublished data).Once the morphine has been obtained in a form

more or less free from impurities, a variety of chemi-

230

231BIOLOGICAL DISPOSMON OF MORPHINE AND ITS SURROGATES.

TABLE I

METHODS FOR THE ESTIMATION OF MORPHINE IN BIOLOGICAL FLUIDS

ISample Apia lnReference Purification Measurement Sample size Alica- Blank Remarks

(g or ml) bility (Mg) (Mg/mI)

RADIOACTIVE METHODS

Adler, Elliott & Addition of Isotopic dilution Plasma 0.4 0.05 Nil Sensitivity can be enhancedGeorge (1957) carrier morphine, of crystalline deri- Urine 0.002-0.01 0.1 by Increasing sample size

extractions and vative Tissues 1.0-1.5 0.1conversion todinitrophenylether

Mule &Woods Extractions Radioactivity of Brain 0.1-1.0 0.05 Nil Probably has high degree(1960) extracted alkaloid of specificity, but assess-

ment not given

Achor & Geiling None Radioactivity of Tissue ? Nil Method non-specific, direct(1953) plated tissue Urine plating sublect to greater

homogenate or errors, insufficient detailurine

Miller & Elliott None Radioactivity of Tissues 0.025-0.05 0.1-0.2 Nil Non-specific(1955) plated pepsin

digest of tissue

Elliott, Tolbert, None Radioactivity after Urine 15 0.01-0.1 Nil Method non-specific forAdler & combustion to Faeces 10-15 morphine but useful forAnderson (1954) C02 and precipi- Breath 3-6 min. tracing metabolites

tation as BaCO3 sample

PHOTOMETRIC METHODS

Way, Kemp, Extractions With Folin- Tissues 1 10 0.3 RapidYoung & CiocalteuGrassetti (1960)

Young & Way Extractions With Folin- Tissues 1 2 0.2 Modification of earlier(unpublished Clocalteu methoddata)

Siminoff & Extractions and With methyl Blood 10 2 0.02 Combined method adoptingSaunders (1958)a conversion to orange Tissues 10 2 0.02 extraction procedures of

nitrobenzoyl ester Fujimoto, Way & Hine (1954)and Woods, Cochin,Fornefeld &Seevers (1954)

Szerb, MacLeod, Precipitation, With Folin- Plasma 5 5 0.2 Other phenolic narcoticsMoya & McCurdy extraction, Ciocalteu Tissues 1 10 0-1.2 can interfere(1957) a adsorption on

resins

Feldstein & Extractions and As nitroso- Urine 50 250 2 Bound blanks considerablyKlendshoj (1956)a nitrosation morphine higher; other phenolic

narcotics interfere

Fujimoto, Way & Extractions With silico- Urine 15 45 0.3 Fairly rapid, adaptable forHine (1954) a molybdic acid ultraviolet analysis

Woods, Cochin, Extractions and With methyl Plasma 5 15 0.04 Concentration and opticalFornefeld & conversion to orange Urine 5 20 0.3 density non-linear. BoundSeevers (1954) a nitrobenzoyl ester Faeces 4 80 2 morphine blanks not stated

Tissues 1.5 10 0.4-2

Guarino (1946) Extraction Oxidation withiodic acid andcomplexing withferric chloride

a Adapted for measuring bound morphine.

232 E. LEONG WAY & T. K. ADLER

TABLE I

METHODS FOR THE ESTIMATION OF MORPHINE IN BIOLOGICAL FLUIDS (concluded)

I ~~~~~~~~~Sample Applica- BlankReakReference Purification Measurement Sample size bilitl(ttg) Aglml) Remarks

(g or ml) y(g (gmI

PHOTOMETRIC METHODS (continued)

Oberst (1938-39) Extraction, With molybdo- Urine 50 150 0.8 Amino-phenols wouldadsorption on phosphotungstic interferepermutit acid

Mull (1937) Precipitation With phospho- Blood 1 5 1 Reliability not sufflcientlyfrom deprotein- molybdic acid establishedIzed blood

Pierce & Plant Diazotization Urine 60 2 000 25 Low sensitivity but(1932) Extractions with sulfanilic Faeces 10 (dried) reasonably accurate withlglauer (1949) acid fresh specimens

Fleischmann Deproteinatlon With phospho- Blood 5 20-2 000 ? Not reliable(1929) with uranyl nitrate molybdic acid

OTHER METHODS

Paerregaard Extractions, paper Polarographic Urine 5 5 Nil Dihydromorphinone(1957)a chromatography, after nitrosation Blood 5 or 10 10 Nil InterferesMllthers (1958) elution

Achor & Geiling Extraction, ad- Ultraviolet Reported to be applicable(1954) sorption on lon- spectrophoto- for tissue but no data

exchange resins, metric presentedelution

Deckert(1936a, 1936b)Endo & Kato(1937,1938) Extraction, pre- Nephelometric Urine 25-100 30-100 SemlquantitativeOberst cipitation as(1938-39) molybdate-Schirm (1940) vanadate complexlglauer (1949)Oettel (1950)

Balls & Wolff Extraction, pre- Gravimetric Urine 100-200 8 000 1 000- Low sensitivity, but reason-(1928) cipitation with 2 000 ably accurateWolff, Riegel & silicotungsticFry (1933) acid

Ikeshima (1935) Tissue digestion lodometric Blood 5 600 ? Low sensitivity, reliabilityKabasawa with papain ex- Brain 500 not sufficiently established(1933-35) traction

To & RI (1938) Extractions, pre- Titration of Urine 1 000 10 000 ? Not reliable - reportedcipitation with HI bound to values too highiodine-potassium morphineIodide (Wagner'sreagent)

a Adapted for measuring bound morphine.

BIOLOGICAL DISPOSMON OF MORPHINE AND ITS SURROGATES. 1

cal, physical and biological methods is availablefor the identification, estimation and characteriza-tion of the morphine. The chemical methods consistof qualitative and quantitative tests of crystallinesalt formation, precipitation reactions, and variouscolour reactions. The formation of crystalline saltsis highly desirable but biological media generallyyield neither the quantity nor the degree of purity ofthe morphine residue required to accomplish thistask. Nephelometric methods of measuring theamount of precipitation of morphine as an insolublecomplex have been used, but such methods have notbeen perfected to the degree attained with colori-metric procedures. The latter have been mostpopular because many colour reactions are possibleinvolving the functional groups present in morphine.The aromatic hydroxyl on the 3-position can be de-tected by a variety of phenolic reagents with varyingdegrees of sensitivity and precision. A colouredderivative of morphine can be formed by couplingthe phenol group with diazotized sulfanilic acid.The basic nitrogen group offers the possibility ofutilizing the general coupling reaction of certainacidic dyes with organic bases to form an extractablecomplex of morphine. These colour reactions aregenerally quite sensitive and easily performed, butare not specific for morphine alone; hence, theextracted morphine must be in a highly purified statebefore the reaction is carried out. Most of themethods summarized in Table 1 are based on theadaptation of these colour reactions to the quantifi-cation of morphine.

Physical methods have the inherent advantage ofnot requiring the chemical alteration of the morphinemolecule for its estimation, characterization or

identification. The subject has been recentlyreviewed (Farmilo & Levi, 1953; Farmilo, Oest-reicher & Levi, 1954). A wide selection of methods isavailable for the purified material, utilizing polaro-graphic (Paerregaard, 1957a; Milthers, 1958),electrophoretic (Spengler, 1958), fluorescence(Nadeau & Sobolewski, 1958), infra-red (Seagers,Neuss & Mader, 1952), ultraviolet (Elvidge, 1940;Bradford & Brackett, 1958), optical rotation(Soehring & Frahm, 1949) and X-ray diffraction(Gross & Oberst, 1947) techniques. Unfortunately,most of these methods are not suitable at the presenttime for measuring morphine in biological fluids.The ultraviolet absorption spectrum has been usedwith some success for estimating and identifyingmorphine in biological media (Biggs, 1952; Gold-baum & Kazyak, 1952; Fujimoto, Way & Hine,

1954), but it has a relatively low order of sensitivity.A polarographic procedure has been developed whichappears to be sufficiently sensitive for morphine inconcentrations as low as 1 ,ug/ml (Paerregaard,1957a; Milthers, 1958).Tracer techniques have been applied to 14C mor-

phine labelled either in the N-methyl position(Rapoport, Lovell & Tolbert, 1951; Andersen &Woods, 1959) or randomly by cultivation of thepoppy in an atmosphere of 14CO2 (Achor & Geiling,1954). Earlier measurements using isotopic mor-phine were largely confined to determination ofradioactivity (Achor& Geiling, 1953; Elliott,Tolbert,Adler & Anderson, 1954; March & Elliott, 1954;Miller & Elliott, 1955). While this is extremelyuseful in providing information on the routes ofexcretion and, possibly, on the metabolic pathwaysof morphine, measurements of radioactivity do notgive specific information as to concentrations ofmorphine in various tissues. More recently Adler,Elliott & George (1957) were able to detect concen-trations of morphine as low as 0.028,g/ml in plasma,using an isotope dilution technique. MorphineN-14CH3 in biological media was determined byadding carrier morphine to the sample to be analysed.After recovery and purification of the isotopicallydiluted morphine N-14CH$ by conventional pro-cedures, it was converted to crystalline dinitrophenyl-morphine-N-l"CH3 and the specific activity deter-mined.The purity of the crystalline derivative was ascer-

tained by its powder X-ray diffraction pattern, thusmaking this method one of the most sensitive as wellas specific for determination of morphine.The chief deterrent to the development of a suit-

able method for the estimation of morphine inbiological media may be attributed to the fact thatmicrogram quantities of morphine need to be mea-sured in the presence of a large amount of naturallyoccurring biological substances, which may react asmorphine. Most of the methods we have citedgenerally measure some blank material which gives amorphine test. This greatly limits the precision ofthe method, especially at the lower limits of sen-sitivity. With isotopic procedures and apparentlywith a recent polarographic method (Paerregaard,1957a; Milthers, 1958) blank contributions need notbe considered. The sensitivity of the methods whichwe have listed in Table 1 does not necessarily agreewith what is claimed by the various authors,inasmuch as we are reporting a concentration whichwe consider will give a precision of approximately

233


i 15% after making allowances for the blank con-tribution.The absolute specificity of the methods for deter-

mining morphine listed in Table 1 has not beenestablished beyond doubt. However, there can be noquestion that morphine is excreted in the urine sinceStevenson & Rapoport (1955) reported the isolationand crystallization of morphine from the 24-hoururine of six patients on morphine. It is quite prob-able that most of the methods listed show a highdegree of selectivity for morphine. Those procedureswhich we considered non-specific are so indicatedunder the column " Remarks " in Table 1. Althoughthe specificity of some of the procedures has notalways been fully evaluated, the uniformity of theresults on the excretion of morphine obtained by thevarious investigators with divergent techniques givesa strong indication that these methods are essentiallyvalid for measuring morphine. While it is conceiv-able that some biotransformation products ofmorphine might retain the characteristics of a phenolas well as those of an alkaloid and thus react asmorphine (e.g., as normorphine), the available evi-dence indicates that any contribution of normorphineor other metabolic products to the readings ofmorphine by the above methods is very minor.Nevertheless, one should bear in mind that minorcontributions from alkaloidal phenols may exist.

It should be noted that while many of thesemethods are applicable for studying the metabolismof the morphine, this does not necessarily mean thatthe method is applicable for forensic purposes,especially when the substance ingested is unknown.In such instances some closely related narcoticanalgesic congeners or aminophenol may be mea-sured as morphine. Also, with the advent of the useof nalorphine for the treatment of acute morphinism,one must be cognizant of the fact that nalorphinegives many of the reactions of morphine and is noteasily separated from the latter by general purifica-tion procedures.The more recent methods, as noted in Table 1,

have been applied also to the estimation of boundmorphine. This is obtained by difference afterestimation of the total (free plus bound) morphinepresent in a given sample after subjecting it to heatand strong acid. It has tacitly been assumed thatany morphine that is bound, combined or con-jugated is completely released without decompositionunder conditions used for hydrolysis of the boundmorphine. This assumption was based on the studiesof Gross & Thompson (1940) and of Oberst (1940).

However, Rapoport and his colleagues (personalcommunication) have recently found that conditionsnecessary for complete hydrolysis of bound mor-phine often need to be much more rigorous thanthose proposed in these earlier investigations.Furthermore, if more than one form of boundmorphine exists, as some authors claim (Thompson& Gross, 1941; Woods, 1954), until the postulatedproducts are isolated in purified form for properstudy it cannot be stated with certainty that theabsolute amount of morphine measured after hydro-lysis represents the total amount of bound morphinepresent. It should be noted, however, that recentstudies indicate that only one form of bound mor-phine-namely, morphine monoglucuronide-is pre-sent in the urine in significant quantities (Fujimoto &Way, 1957) and that hydrolysis of this compoundcan be effected under conditions which do not causedecomposition of the liberated morphine (Gross &Thompson, 1940; Woods, 1954; Fujimoto & Way,1957).

ABSORPTION

Considering the amount of work that has beencarried out on morphine, there is a surprising dearthof quantitative studies on the absorption of mor-phine. Numerous citations from an earlier review(Krueger, Eddy & Sumwalt, 1941) attest thatmorphine can enter the mammalian body by anynumber of routes. In addition to the usual means forobtaining morphine effects by parenteral administra-tion, morphine is reported to have produced sys-temic effects after being placed in contact with thelining or walls of the alimentary canal, respiratorytract, vagina, urethra, prepuce, bladder, abradedskin, ear, conjunctival sac and brain. However, acomparative study of the efficiency of morphineabsorption by diverse routes has not been exten-sively investigated. From the vast amount ofpharmacological, toxicological and clinical datapublished, it is possible to conclude that morphine iswell absorbed by all species when injected paren-terally, but is erratically absorbed when given bymouth.

In man the absorption of morphine appears to befairly prompt after parenteral administration. It iswell known from clinical usage that the effects ofmorphine can appear within a few minutes afterhypodermic administration, but judging from animalstudies about one hour is probably required toattain peak levels of the compound. There is a needfor reliable quantitative studies on man concerning

234

BIOLOGICAL DISPOSITION OF MORPHINE AND ITS SURROGATES. 1

the rate of appearance of morphine in the bloodafter its administration. Only recently have someobjective data been reported concerning the rate ofabsorption of morphine. Elliott, Tolbert, Adler &Anderson (1954) administered 10-15 mg of morphine-N-methyl-l4C intramuscularly and found measurableamounts of 14CO2 in the breath within 10-15 minutes,with the peak rate of excretion occurring between30 and 90 minutes.Absorption of morphine by mouth appears to be

poor. Isbell & Fraser (1953) found that orallyadministered morphine produced little, if any,pupillary constriction and respiratory depression informer addicts. Beecher, Keats, Mosteller &Lasagna (1953), in assessing analgesia, were unableto distinguish between a placebo and 10 mg ofmorphine orally, but found that 0.6 g of aspirinproduced analgesia greater than that of the placebo.Comroe & Dripps (1948) reported minimal side-effects in normal volunteers given 15-30 mg ofmorphine by mouth. Walton & Lacey (1935a,1935b) administered morphine sublingually in dosesranging between 20 and 120 mg and noted negligiblepharmacological effects.

In contrast to the dearth of objective evidence onthe absorption of morphine in man, ample data tothis effect exist for animals. In monkeys, a level of6 ,ug/ml of the free alkaloid was attained in plasmawithin 30 minutes after subcutaneous administrationof 30 mg/kg and the maximum level of 8 ,tg/mloccurred after 120 minutes. The biological half-lifewas reported to be about 3 Y2-4 hours (Mellett &Woods, 1956).As to dogs, 25% and 42% of the alkaloid were

recovered after 30 minutes from the injection site intwo dogs given 100 mg/kg morphine sulfate intra-muscularly (Hatcher & Gold, 1929). A peak plasmalevel of 7 ,tg/ml was attained about 45 minutes aftersubcutaneous administration of 30 mg/kg morphine,but the same dose given orally yielded barely de-tectable or no plasma levels of morphine (Cochin,Haggart, Woods & Seevers, 1954). In the same com-munication, it was reported that absorption ofmorphine by the intramuscular route does not differgreatly from that found with subcutaneous adminis-tration, but the former route yielded an earlier andgreater peak.

Rabbits given doses of morphine varying from4 to 10 mg/kg exhibited high levels in the bloodwithin 35 minutes after intramuscular injection(Mull, 1937); maximum blood concentrations wereobtained 30 minutes after subcutaneous injection

(Fleischmann, 1929). Peak blood levels were noted1 hour after intraperitoneal administration of20 mg/kg; 30-minute levels were almost as high(Siminoff& Saunders, 1958).

In the rat, detectable levels of 14C in various partsof the central nervous system were found within15 minutes after hypodermic injection of 5 mg/kg 14C-labelled morphine and peak levels of radioactivity,equivalent to between 0.2 and 0.6 ,ug/g tissue, occur-red at 60 minutes (Miller & Elliott, 1955). Over 90%of the radioactivity disappeared from the subcuta-neous injection site within 1 hour after injectingmorphine-N-methyl-14C. Absorption was equallyrapid following injection into the popliteal space,and was not markedly altered by vasopressin, ACTHor adrenalectomy (Adler, Elliott & George, 1957).

Pretreatment with an antihistamine (mepyramine)or depletion of tissue histamine with compound48/80 (a condensation product of p-methoxyphenyl-ethylmethylamine and formaldehyde) enhanced pro-nouncedly the subcutaneous absorption of morphineas measured by blood levels of morphine 10 minutesafter injection (Milthers & Schou, 1958). The expla-nation was offered that morphine, being a histaminereleaser, causes self-depression of absorption. Thisconclusion was based on the belief that any substancereleasing tissue histamine would tend to inhibit itsown absorption after subcutaneous administration,since histamine given locally was found to decreasethe absorption of sulfacetamide (Schou, 1958). Theconcept of self-depression of absorption was exten-ded to include 5-hydroxytryptamine liberators(Milthers, 1959), since enhanced absorption of mor-phine was noted also after pretreatment with theserotonin antagonist, BOL 148, and with depletion oftissue 5-hydroxytryptamine. It was pointed outfurther that newborn animals, being resistant to theaction of histamine and 5-hydroxytryptamine, shouldabsorb morphine more rapidly than adult animals.As evidence it was noted that blood levels of mor-phine 10 minutes after subcutaneous administrationof 100 mg/kg morphine hydrochloride were nearly6 times higher in 12-day-old rats than in adultanimals (Milthers, 1960) and these findings wererelated to the increased susceptibility of youngeranimals to morphine. With respect to the latterpoint, Kupferberg & Way (unpublished data)maintain, on the basis of their disposition studies onmorphine, that a decreased blood-brain barrier tomorphine rather than enhanced absorption is theprimary cause for the decreased resistance of new-born animals to morphine.

235


In the guinea-pig, morphine was found in tissues15 minutes after subcutaneous administration(Keeser, Oelkers & Raetz, 1933).

In the mouse, after subcutaneous administrationin the hind leg, approximately 50% of the injecteddose disappeared after 15 minutes and over 90%after 60 minutes (Way, Young, Kemp & Johnson,unpublished data). Within 15 minutes after intra-peritoneal injection of approximately 10 mg 14C-labelled morphine per kg, mouse blood contained,per millilitre, over 5% of the dose of 14C andappreciable levels of radioactivity were present inother tissues at that time (Achor & Geiling, 1954).

DISTRIBUTION

Information on the distribution of morphine hasbeen obtained in several species. The earlier data areconcerned only with the free alkaloid; more recentwork includes results on bound morphine. Mostextensive information is available on the dog and rat,but considerable data are available as well onmonkeys, cats, rabbits, mice, and guinea-pigs.

In general, free morphine, like most basic amines,rapidly leaves the blood and concentrates in tissues,particularly parenchymatous tissues. Kidney, lung,liver, and spleen show a predilection for the drug,with by far the highest concentration being found inthe kidney. Certain endocrine organs, such as theadrenal, thyroid and pancreas, also appear to con-centrate the drug. While skeletal muscle may show asomewhat lower level of morphine, it generallyaccounts for the major fraction of the drug left in theanimal body, except for what may be present inbladder urine and gall-bladder bile. Extensivecumulation of morphine in tissues does not occur;tissue levels fall to quite low levels within 24 hoursafter the last administration. Relative to the doseadministered the amount of the drug found in thebrain is extremely minute and the concentration inthe central nervous system is but a fraction of thatattained in most other organs. It may well be thatthe wide range in toxicity of morphine found amongthe different animal species is related to the ability ofthe compound to gain access to the central nervoussystem in the various species.Bound morphine, which has relatively low pharma-

cological activity, is found in highest concentra-tions in organs concerned with its excretion-namely,the kidney and gall-bladder. The bound morphinefound in tissues is very probably identical with themorphine-3-monoglucuronide isolated from urineand bile.

ManThe store of knowledge on the tissue distribution of

morphine in man is meagre. Forensic data offersome qualitative information but lack of preciseknowledge of dosage levels, time relationships,routes of administration, and complications owingto possible simultaneous ingestion of other com-pounds, etc., makes it difficult to view the informationin proper perspective. A satisfactory method ofsufficient sensitivity for the measurement ofmorphineconcentrations in blood well below 0.5 ,mg/ml stillneeds to be developed. Even if one neglects animaldata indicating that morphine rapidly leaves theblood and concentrates in certain tissues, the dis-tribution of a 20 mg/70 kg dose of morphine in bodywater would yield a blood level of only 0.4 ,ug/ml. Itis known that the blood levels of morphine at anytime after an intravenous dose of 20 mg are less than1 ,Ag/ml (Milthers, 1958). Measurable levels ofmorphine are found in the urine, bile, faeces, salivaand sweat and these are discussed under the sectionon excretion.

MonkeyIn monkeys after subcutaneous administration of

30 mg/kg of morphine peak plasma levels of freemorphine (about 8 ug/ml) were obtained 1 /2-2 hoursafter injection, the biological half-life being approxi-mately 3 Y2-4 hours; bound morphine reached apeak in about 2 hours (26 ,ug/ml) and had a half-lifeof about 6 hours. The distribution of free morphine90 minutes after injection was found to be highest inthe adrenal, lung, pancreas, and kidney, the con-centration range being between 11 and 16 ,ug/g.Brain, spinal cord, and cerebrospinal fluid failed toshow a detectable level of free morphine with amethod sensitive to 5 ,tg/g. After 4 hours only theadrenal (9 ,tg/g) and bile (35 ,tg/ml) showed ap-preciable concentrations of free morphine. Boundmorphine was found in high concentrations in thebile (4120 ,ug/ml) and the kidneys (16 pg/g) at4 hours (Mellett & Woods, 1956).A comparison of the distribution of morphine in

four non-tolerant monkeys and in five monkeysmaintained on 15 mg/kg morphine injected sub-cutaneously twice a day for nearly a year indicatedthat tolerant monkeys did not differ greatly fromnon-tolerant animals. With both tolerant and non-tolerant monkeys, high concentrations were foundin the adrenals and in the bile at 90 minutes and4 hours and no morphine was detected in the brain,spinal cord, cerebrospinal fluid, skeletal muscle, or

236

BIOLOGICAL DISPOSITION OF MORPHINE AND ITS SURROGATES. 1 237

TABLE 2TISSUE DISTRIBUTION OF MORPHINE IN DOGS AFTER ADMINISTRATION OF VARIOUS DOSES

1111IDose infused 375 SC a 37.5 SC 37.5 SC 37.5 SC 30 IV b 30 IV 30 SC 30 SC 30 SC 30 SC

(mg/kg) IV for5 hours

Time afterratiois- 0.6 4 4 24 24 1.5 4 1.5 1.5 4 4

(hours)

No. of 1 6 6d 6 61 1 1 1d 1animals

g of drug per g of wet tissue

Brain 230 9.8 7.8 7.5 3.2 5 3 6 5

Blood 50 5.3 5.9 2.2 3.6 5 5 c 2 4 2

Kidney 810 0.67 c 0.77 c 0.27 c 0.29 c 68 9 c 25 55 12 15

Liver 160 18.5 14.8 4.7 8.2 5 1 c 4 11 3 8

Spleen 77 7 c 20 33 9 10

Lung 230 16.9 13.9 4.7 5.0 30 5 c 12 20 5 12

Heart 110 11.3 8.2 3.4 6.4 14 3 c 6 8 6

Muscle 360 11.3 6.4 2.9 4.8 27 7 c 5 10 2

Skin 90

Bone 370

Pancreas 25 32 7 6

Adrenals 10 19 6 8

7 e 14 e6GI tract 480 4.7c 74c 2.3c 46c 8f 32f 9fGI contents 205 c 1.7 c 8.1 1.8C 7.6 c

Bile 7 1Oc 340 52 27 29Urine 896 c 16.1 17.1 20.9 c 32.0 c

Percentagerecovery 39

Wolff et al. Plant & Pierce (1933) Woods (1954)(1933)

a SC = subcutaneous administration.b IV = intravenous administration.

spleen. The small intestine of tolerant animals hadhigher levels of morphine than did that of non-tolerant ones at 4 hours (Mellett & Woods, 1956).

DogThe tissue uptake of morphine in dogs has been

studied over a wide dosage range. Since a direct com-parison of the various studies is not always possiblebecause of the many differences in experimental con-ditions, the results obtained in three different labo-ratories have been grouped in Table 2. General

c Total quantity recovered in mg.d Tolerant.

e Jejunum.f Colon.

characteristics of the distribution of morphine in thedog are: (a) its rapid disappearance from the blood;(b) the high levels attained in the kidneys; (c) therelatively large amounts recovered in skeletalmuscle; and (d) the low levels attained in thebrain.

Evidence that morphine disappears rapidly fromblood and can be recovered to a large extent inskeletal muscle was furnished some thirty years agoby Hatcher & Gold (1929). Using a procedure forestimating morphine which was rather insensitive


but apparently reliable for the high doses used, theywere able to demonstrate that less than 2% of aninjected intravenous dose of 50-100 mg/kg ofmorphine sulfate was present in the blood of animalsexsanguinated after 30 minutes. Analyses of muscle30 minutes after intravenous administration of50 mg/kg and 100 mg/kg in two animals yieldedrecoveries of morphine which accounted for 47%and 22%, respectively, of the injected dose. Mor-phine leaves the muscle rapidly, only 2% of thedose being present in an animal 2 hours afterreceipt of 100 mg/kg infused intravenously for onehour. In the same animal the morphine in the liver,heart, kidneys, lungs, spleen, stomach, intestines andpancreas accounted for less than 1.0% of theinjected dose.

Wolff, Riegel & Fry (1933) injected morphineintravenously (1111 mg/kg base equivalent) into adog under ether anaesthesia over a 5-hour period.Thirty-five minutes after the injection the variousorgans were analysed for the free morphine content.The estimated recovery of the dose administered was39%, and more than half of this quantity could beaccounted for by the morphine in skeletal musclealone. As indicated in Table 2, the kidney had thehighest concentration, but appreciable levels werealso present in the gastro-intestinal tract, skeletalmuscle and bone. It is of interest to note the rela-tively high brain levels of morphine which wereattained at this high dosage level.

Plant & Pierce (1933) studied the distribution intolerant and non-tolerant dogs 4 and 24 hours aftersubcutaneous administration of 50 mg/kg of mor-phine sulfate. Their data, recalculated in terms ofthe free base, are summarized in Table 2. Highconcentrations were found in the kidney. Althoughthe data on the kidney were reported in terms of totalmilligrams recovered rather than on a concentrationbasis as with most other organs a reasonable appro-ximation of kidney concentrations can be obtainedby assuming the organ to weigh 30 grams. The brainlevels appear high when compared with the distribu-tion characteristics of morphine in other species andwith more recent data on the dog (Woods, 1954,1957). Since Plant & Pierce (1933) claimed goodreproducibility only down to 15 ,ug/g (actually0.3 mg/20 g tissue), the brain values they reported donot fall within the limits of adequate precision. Wefeel, therefore, that these values are, at best, accept-able only as maximum values. Tolerant and non-tolerant animals showed some difference in tissuelevels, but the differences were not striking and

variation was such that it is difficult to attach anyspecial significance to the findings (Plant & Pierce,1933).Woods (1954), using a dose of morphine (30 mg/kg

subcutaneously) close to that of Plant & Pierce,found high concentrations of morphine in the kid-neys, pancreas, and spleen at 90 minutes. Somewhatlower levels were found in the lung and adrenalsand even lower levels in cardiac, skeletal, andsmooth intestinal muscle. The concentrations in thebrain and blood were below the sensitivity of themethod (5 ,g/g). The kidneys also had extremelyhigh concentrations of bound morphine; the liverhad somewhat lower but still high concentrations ofthe bound metabolite. However, the concentrationof bound morphine in both organs was still farbelow that in gall-bladder bile. After 4 hours thetotal morphine values in all tissues had decreasedconsiderably and after 12 hours the concentration oftotal morphine was less than 5 ,ug/g. After adminis-tration of 30 mg/kg morphine intravenously, brainlevels of free morphine were barely detectable or notmeasurable although the kidney, spleen, lung, andskeletal muscle levels (named in order of decreasingconcentrations) were roughly twofold to fivefoldhigher than after subcutaneous administration. At1 hour the concentration of free morphine in theplasma of these animals was threefold greater andthat of bound morphine 30-fold greater than thecorresponding concentrations in erythrocytes.The distribution of N-methyl-4C-labelled mor-

phine in the white and grey matter of the cerebralcortex was determined after administration of2 mg/kg subcutaneously. At 4 hours the morphinelevels were 0.23 ,tg/g in grey areas and 0.13 ,ug/g inwhite areas. At 16 hours the level in both areas wasabout 0.05 ,tg/g (Mule & Woods, 1960). The grossconcentration of morphine in dog brain appears to belittle affected by the simultaneous subcutaneousadministration of nalorphine (3 mg/kg) with mor-phine (30 mg/kg) (Woods, 1957).Woods (1954) also studied the distribution of the

drug in animals made tolerant to morphine, 30 mg/kg.Although not specifically stated, it is presumed thatthe degree of tolerance and its duration in theseanimals were comparable to that achieved in thebitches used in a previous study (Cochin, Haggart,Woods & Seevers, 1954). The distribution of freeand bound morphine was compared in three non-tolerant and three tolerant animals, using one animalfrom each group at three different time intervals after30 mg/kg of morphine injected subcutaneously. At

238


12 hours tissue levels of morphine were too low fordetection, but at 90 minutes and 4 hours the concen-trations of free and bound morphine were found, ingeneral, to be higher for most tissues in tolerant thanin non-tolerant animals; brain levels in both groupswere either below or barely exceeded the sensitivity ofthe method (5 ,g/g). Woods concluded that thesedata do not provide acceptable evidence that altereddistribution is responsible for the development oftolerance to morphine. To this we readily agree, butwe would add that inasmuch as some possible altera-tions in the distribution of morphine within thecentral nervous system could not have been detectedby the methods used, the study also does not refutesuch an idea.Woods (1954) minimized the differences noted

between the tolerant and the non-tolerant dogs andoffered the explanation that " such differences arenot of such magnitude that they cannot be explainedon the basis of quantitative differences in the vascularresponses of tolerant as compared with non-tolerantanimals resulting in altered mobilization from thesite of administration and differences in the vascularsupply to the several tissues." The meaning of thelatter part of the explanation is not quite clear to ussince, in an earlier experiment, after intravenousadministration of morphine, he postulated that thehigh levels of morphine found in certain tissues werethe consequence of a marked vasodilatory effect andpooling of blood in these organs. If we assume thatthe vasodilatory effect of morphine is greater in non-tolerant than in tolerant dogs, then this should resultin higher rather than lower tissue morphine levels inthe non-tolerant animals, but such is not the case.We would like to offer an alternative explanation

for the higher organ levels of morphine found intolerant than in non-tolerant dogs. Since thetolerant animals received daily injections of mor-phine, it appears more plausible to suppose thatresidual morphine may be present in this group.That such may be the case is apparent from thebiliary excretion data of Woods (1954), wheresignificant concentrations of morphine in the freeand especially in the bound form were found inthe gall-bladder bile even 72 hours after injection.While this finding was obtained in a non-tolerantanimal, measurements in tolerant animals (studiedonly up to 12 hours) yielded biliary levels of totalmorphine comparable to, if not higher than, thoseobserved in the non-tolerant animals. In thetolerant animals there would be a residual freemorphine pool which receives sizeable contributions

not only from biliary free morphine secreted in theintestine and reabsorbed, but also from biliarybound morphine, since the latter is readily hydro-lysed in vivo (Blaney, Bloom & Woods, 1955). Thus,with chronic administration, mobilization of thebiiary stores of free and bound morphine wouldaugment the effect of injected morphine and lead tohigher tissue levels.

Rabbit

Concentrations of morphine in rabbits after ad-ministration of 20 mg/kg intraperitoneally weredetermined by Siminoff & Saunders (1958). Ap-preciable amounts of both free and bound morphinewere detected within 30 minutes in brain, blood, liver,and kidney, and peak levels occurred at 1 hour. Thekidney had high concentrations of both free andbound morphine. Brain levels at 1 hour were1.8 ,tg/g. Animals rendered tolerant to the respira-tory effects of 20 mg/kg of morphine by dailyinjection of increasing doses for 100 days yieldedtissue levels comparable to those of non-tolerantanimals 1/2, 1, 2, 4, and 8 hours after intraperitonealadministration of 20 mg/kg of the drug. Half of theintravenous dose of morphine hydrochloride(1-2 mg/kg) disappeared from the blood within afew minutes after injection; the levels at 20 minuteswere approximately 1 ,ug/ml or less (Milthers, 1958).

Guinea-pigIn the guinea-pig studies were carried out after

injecting 400 mg/kg of morphine subcutaneously.Morphine was found in all organs examined exceptthe muscles 16-24 hours after injection. The highestlevels were found in the kidney and liver. The levelsin the blood and skeletal muscle were lower. Thebrain was the organ with the lowest concentration ofmorphine. In animals which had received morphinedaily for 3-6 weeks, morphine was still detectable inthe kidney and blood 24 hours after the last injection(Keeser, Oelkers & Raetz, 1933).

Rat

Depending on the route of administration bloodlevels of morphine in rats reach a peak value soonafter administration which is followed by a more orless rapid fall. With the exception of the centralnervous system, the partition between blood andtissues is largely in favour of tissues and, as equili-brium is approached, this is reflected by high tissueand low blood levels.

239


After intravenous administration of 75 mg/kg ofmorphine sulfate, blood levels fell rapidly from alevel of 27 ,ug/ml at 10 minutes to 16 ,ug/ml at60 minutes and to 5 ,ug/ml at 180 minutes (Szerb &McCurdy, 1956). In rats given 150 mg/kg morphinesubcutaneously blood levels of free morphine fellfrom 19 ,ug/ml at 90 minutes to less than 3 ,ug/ml at240 minutes. During the same interval, levels offree morphine in most other organs were generallyseveral-fold higher than the blood levels (Woods,1954). Plasma levels of free morphine 60 minutesafter injection of 2 mg/kg of labelled morphine, sub-cutaneously or intrapopliteally, averaged 0.08 ,ug/mlin Sprague-Dawley rats and 0.10 u.g/ml in Long-Evans rats (Adler, Elliott & George, 1957).

Levels of bound morphine in the blood fluctuatemore widely than free morphine levels and are de-tectable in tissue after a longer period than freemorphine. The variations and the lag are to be ex-pected since the appearance of bound morphine isdependent on conjugation processes, which in turnare subject to influences by dosage, physiologicalstate of the animals, prolonged morphine adminis-tration, etc. No bound morphine was detected in theblood 10 minutes after intravenous administration of75 mg/kg morphine sulfate intravenously, but at60 minutes the level was 3.5 ,ug/ml and at 180 minutesit was 7.8 ,ug/ml (Szerb & McCurdy, 1956). Aftersubcutaneous administration of 150 mg/kg, nobound morphine was detected at 90 minutes but at4 hours, when no free morphine was found, the levelof bound morphine was 23 ,ug/ml (Woods, 1954).After injection of 2.0 mg/kg of labelled morphinesubcutaneously or intrapopliteally, over 95% of theradioactivity in the plasma of Sprague-Dawley rats(1.9 ,ug/ml) and about 85% of that of Long-Evansrats (0.67 ,ag/ml) at 60 minutes represented radio-activity contributed by bound morphine. Wholeblood levels of 14C ranged between 66% and 78% ofthe 14C concentration of plasma, the lower valuesbeing associated with a high haematocrit (Adler,Elliott & George, 1957). Szerb & McCurdy (1956)reported that free morphine was equally distributedbetween plasma and erythrocytes while there wasabout 4 times more bound morphine in plasma thanin formed element.

There appears to be a considerable blood-brainbarrier to morphine, although the uptake of smallquantities of the drug by the organ is relativelyrapid. The concentration of free morphine attainedin the brain on a mg/kg basis is generally but a verysmall fraction of that to be expected assuming

uniform distribution of the drug in the animal,whereas other organs show a selective preference formorphine.Age greatly influences the uptake of morphine by

the central nervous system. Sixteen-day-old ratsshow brain levels that are roughly threefold higherthan those of 33-day-old animals after either intra-peritoneal or intravenous injection of comparabledoses (Kupferberg & Way, unpublished data). Asmight be surmised, higher levels of free morphine inthe brain are attainable by intravenous administra-tion of high doses of the compound than by otherroutes. With an intravenous dose of 75 mg/kg,which produced almost immediate catalepsis inSprague-Dawley rats, the brain level at 15 minuteswas 13 ,ug/g, and at 40 minutes it had decreased to7 ,ug/g (Young & Way, unpublished data). On theother hand, in male Holtzman rats receiving 75mg/kg morphine intravenously, Szerb & McCurdy(1956) found a lower and later peak, with a morphinebrain level of 6.2 ,ug/g at 10 minutes rising to 9.3,ug/g at 60 minutes. The latter result is rather sur-prising, since organic bases as a general rule leave theblood rapidly owing to rapid tissue uptake andordinarily one would expect peak organ levels ofmorphine to occur about 10 minutes after intra-venous administration.

Administration of morphine by parenteral routesother than intravenous yields considerably lowerlevels of the compound in the brain. With a dose ashigh as 150 mg/kg injected subcutaneously, Woods(1954) found no morphine in the brain, using amethod with a sensitivity of 5 ,ug/g. Hosoya (1956)reported values of 10± 1.9 ,tg/g in brain 40 minutesafter intraperitoneal injection of 150 mg/kg mor-phine and of 5.66 ± 2.0 ,ug/g after 4.0 mg/kg. Thelatter value is inordinately high for such a low doseof morphine and is difficult to explain. The centralnervous system was reported to contain negligibleamounts of radioactivity 1 hour after subcutaneousadministration of 5 mg/kg of morphine-N-methyl-14C (March & Elliott, 1954). However, subsequentstudies by the same laboratory using improvedtechniques resulted in measurement of morphinelevels in the brain even with doses as low as 2 mg/kg(Adler, Elliott & George, 1957) and 5 mg/kg (Miller& Elliott, 1955). During the interval between 30 and60 minutes (when the pharmacological effects weremaximal) after a 5 mg/kg subcutaneous dose ofmorphine-N-methyl-14C in Long-Evans rats, themorphine levels in various parts of the centralnervous system were between 0.2 and 0.7 ,ug/g.

240


These values represent maximum values for mor-phine since only 14C levels were measured and thebrain was not perfused free of blood containingappreciable radioactivity. The spinal cord showedthe highest radioactivity, yielding levels representingbetween 0.6 and 0.7 ,ug of morphine per g of wettissue; the levels in the hypothalamic area were onlyslightly lower. Concentrations in the cerebellum,medulla, mid-brain and cerebrum were between0.2 and 0.4 ,ug/ g (Miller & Elliott, 1955). If calcula-tions of the morphine concentration are based onthe lipid content of the tissue rather than on theweight of wet tissue, the cerebrum levels actuallyexceed those of the spinal cord (Adler, Elliott &George, 1957). Sixty minutes after subcutaneous orintrapopliteal injection of a 2 mg/kg dose of labelledmorphine in Sprague-Dawley or Long-Evans rats,cerebrum levels ranged between 0.04 and 0.09,ug/g in all but one of eight animals (Adler, Elliott& George, 1957). Despite the fact that these levelsrepresent maximum values, they are the lowestmeasurable level of morphine recorded.The minute amounts of morphine needed in the

central nervous system to elicit pharmacologicaleffects can be further emphasized by the fact that themaximal values given above may be several-fold toohigh. Adler, Elliott & George (1957) have shownthat specific activity studies after isotope dilutionindicated that only 10-20% of the 14C present in thetissue could be recovered as free morphine. Withhigher doses of morphine the free morphine-N-14CH3recovered represented 51 % of the 14C in the cereb-rum; the recovery was not substantially increasedafter hydrolysis, indicating that there is practicallyno bound morphine in the brain after injection ofsmall doses. Moreover, since blood levels of radio-activity were more than 20 times higher than cereb-rum levels in Sprague-Dawley rats and about6.5 times higher in Long-Evans animals, any bloodtrapped in the central nervous system would tend toelevate the radioactivity of the organ and a correc-tion for this should be applied. Analysis of theresidual blood content in the brain indicated that,on the basis of the blood concentration, the cereb-rum values could be reduced roughly by one-thirdin the Sprague-Dawley rats and by one-ninth in theLong-Evans rats. Finally, if one considers that mostof the "4C after injection of N-"4CH3 morphine is notin the neuronal elements of the brain but rather in thehighly vascularized choroid plexus and ventriclesas judged by radioautographs of brain sections(Miller & Elliott, 1955), it becomes quite apparent

that an extremely sensitive and selective responseto morphine is exhibited by the central nervoussystem.

In contrast to the central nervous system, othertissues usually attain morphine levels higher thanthose found in blood, and those organs concernedwith the excretion of morphine, particularly thekidneys, show a considerable capacity to concen-trate the drug. With a 5 mg/kg subcutaneous dose oflabelled morphine, the renal level in terms of radio-activity at 60 minutes was more than three timeshigher than that to be expected assuming uniformdistribution of the drug and was roughly about 80times higher than cerebral levels. Radioautographstaken of the kidney of the animals given labelledmorphine indicated that the majority of the radio-activity present was localized in the cortex and calyx,suggesting concentration in the region rich in glome-ruli and tubules (Miller & Elliott, 1955). Undoubt-edly, a considerable portion of the radioactivity isdue to bound morphine, since with a dose of 150mg/kg injected subcutaneously the total morphinelevel in the kidney at 90 minutes represented a freemorphine concentration of 133 ,ug/g and a boundmorphine concentration of 158 ,tg/g (Woods, 1954).At 4 hours the free morphine level had decreased to65 ,tg/g and the bound morphine level had in-creased to 213 ,ug/g. By 12 hours the free morphinelevel was less than 10 ,ug/g and the bound morphinelevel was less than 15 ug/g. In contrast to in vivoconditions, morphine is readily taken up in vitro byrat cerebral cortex slices, so that at 30 minutes theconcentration in the slices (240 micromoles/kg) wasmore than twice the initial concentration of mor-phine in the incubating bath (10-4 M) (Bell, 1958).Morphine uptake by the liver is generally con-

siderably less than that by the kidney, but is stillmuch greater than that by the brain. With a 2 mg/kgdose of labelled morphine in Long-Evans rats, theradioactivity was approximately 20-25 times higherin the liver than in the cerebrum (Miller & Elliott,1955; Adler, Elliott & George, 1957). Radio-autographs of the liver from an animal receivinglabelled morphine showed distribution of activitythroughout the tissue, with points of concentrationprobably at portal areas (Miller & Elliott, 1955).Appreciable free morphine levels appeared in theliver 90 minutes after subcutaneous administration of150 mg/kg, but by 4 hours, although the levels offree morphine in other tissues were still quite high,there was scarcely any free morphine left in the liver(Woods, 1954). However, as might be expected from

241


the role of the liver in conjugating morphine, therewere substantial amounts of bound morphinepresent at this time as well as at 12 hours (Wood1954).The skeletal muscle appears to be an important

depot for morphine and its metabolites. In Sprague-Dawley rats receiving 2.0 mg/kg of labelled mor-phine by subcutaneous or intrapopliteal injection, theaverage maximum morphine concentration in muscleone hour after injection was about 0.7 ,ug/g andwas about tenfold higher than the cerebrum level.Approximately 15% of the injected dose was re-covered from the muscle as 14C. About one-thirdof this radioactivity was represented by free mor-phine, another third by the bound form and theremainder by a fraction which was unidentified(Adler, Elliott & George, 1957). In Long-Evansrats receiving 5.0 mg/kg of labelled morphineapproximately 12% of the injected 14C was presentin muscle (Adler, Elliott & George, 1957), but withhigher doses of morphine a higher fraction of thedose can be accounted for in muscle. Thus, after adose of 150 mg/kg injected subcutaneously, the freemorphine level at 90 minutes was reported to be72 ,tg/g (Woods, 1954). On the assumption thatskeletal muscle makes up 40% of the total body-weight, this would mean that roughly one-fifthof the total dose of morphine would be accountedfor in skeletal muscle as free morphine alone. Fromthe data at 4 hours, indicating free and boundmorphine to be present in approximately the sameconcentrations, it may be similarly calculated thatone-tenth of the total dose can still be recoveredas morphine in skeletal muscle at this time.Other tissues also show considerable ability to

localize morphine. However, the ratio of bound tofree morphine in all tissues is generally much lowerthan that in the kidneys and liver, especially duringthe early stages after drug administration. Organssuch as the lung, spleen, thyroid, adrenals, and heartappear to take up high concentrations of morphine.Adrenal levels of radioactivity between 30 and 60minutes after injection of 5 mg/kg of labelled mor-phine subcutaneously reached levels that wereroughly 25-fold higher than those in the cerebrum,and considerable radioactivity was still present in theadrenal at 150 minutes (Miller & Elliott, 1955).After a dose of 150 mg/kg of morphine injected sub-cutaneously, the levels of free morphine in thethyroid, spleen, lung, heart, and skeletal muscle at90 minutes were roughly one-fourth to one-half thelevel of 133 ,tg/g found in the kidney. At 12 hours

the total morphine levels in these organs other thanthe kidney were not detectable, being lower than5 ,ug/g (Woods, 1954).

In rats made gradually tolerant to 150 mg/kg mor-phine in approximately a month, the blood levels offree morphine were found to decrease more rapidlythan they did in non-tolerant animals. The differencewas attributed to impaired conjugation of morphinein the non-tolerant animals resulting from inhibitionof glycogen synthesis by the injected morphine(Szerb & McCurdy, 1956). The same authors (op.cit.) also reported that tolerant animals were foundto be hyperactive with a free morphine brain level of6.5 ,ug/g 60 minutes after injection, while non-tolerant ones were still immobile 180 minutes afterinjection with an average brain concentration of4.3 ,ug/g, and suggested that the reaction of thecentral nervous system to morphine was altered bychronic morphine administration. A comparison ofthe tissue distribution of morphine in non-tolerantrats and in rats made tolerant to 150 mg/kg over anunspecified period of time did not reveal strikingdifferences beyond a higher concentration of mor-phine in the thyroid and spleen of the tolerantanimals (Woods, 1954).Although vasopressin increases and ACTH de-

creases sensitivity to morphine, beyond decreasingthe bound morphine levels in plasma, treatment withneither substance altered the gross distributioncharacteristics of morphine markedly (Adler, Elliott& George, 1957), nor did pretreatment with neo-stigmine (Szerb & McCurdy, 1956), SKF525A/3-ethylaminoethyldiphenylpropyl acetate hydro-chloride) (Hosoya, 1956), or nalorphine (Woods,1957). On the other hand, adrenalectomy resultedin increased tissue levels of morphine without im-pairing the animals' ability to conjugate morphine(Adler, Elliott & George, 1957).

Mouse

The tissue distribution of 14C in the mouse afterintraperitoneal injection of morphine has beenstudied by Achor & Geiling (1956), using 14C-labelled morphine obtained from radioactive pop-pies. In general, all tissues studied showed maximallevels of radioactivity 15 minutes after injection of10 mg/kg. The tissue levels of 14C declined rapidlywith time except in the case of the liver and gastro-intestinal tract where, after an initial decline at60 minutes, the 14C levels rose to a secondary andhigher peak at 120 minutes after injection. Somechanges in the tissue distribution pattern of radio-

242


FIG. 3

KNOWN AND POSTULATED METABOLIC PATHWAYS OF MORPHINE

H CH3N

0~~=O 0 0 OH

/6H\ MORPHINE-5-MONOGLUCURO-,CH-CH/ NIDE

HOH H<TWNN/

4TION

HO 0 OHNORMORPH I N E

CH3 PSEUDOMORPHINE

CODEINE

activity were noted as a result of treatment of themice with nalorphine or 5-monoamino-acridine, butthe meaning of this is not clear.

METABOLISM

A schematic representation of known and postu-lated pathways of morphine metabolism in vivo ispresented in Fig. 3. The metabolism of morphinehas been of particular interest because, notwithstand-ing the fact that the disposition of an appreciablefraction of the administered dose is still unknown,important roles in producing analgesia and physicaldependence have been assigned at various times toknown and unknown biotransformation products.However, of the several metabolic pathways in-dicated above, only one-the conjugation of mor-

phine with glucuronic acid-has been conclusivelyestablished in vivo by isolation of the metabolite incrystalline form. No such firm foundation supportsthe other pathways, although the equivocal nature ofthe evidence of their existence in vivo has been no

barrier to speculation concerning their presumed rolein producing the pharmacological effects of mor-

phine. In past years the withdrawal syndrome hasbeen attributed to the vainly sought oxidationproduct of morphine, pseudomorphine. Morerecently N-demethylation has been the focus ofinterest and the spur to much experimental work inattempts to correlate this metabolic pathway withanalgesia and with the development of tolerance. Adiscussion of the recent hypotheses concerningN-demethylation and pharmacological effects will befound in a later instalment.

Conjugation

It is now known that conjugation with glucuronicacid is a major pathway for the detoxication ofmorphine. Development of the evidence leading tothis conclusion can be traced from early experimentsshowing that morphine is excreted in a bound formto the present knowledge of both the chemicalstructure of bound morphine and the sequences ofenzyme reactions leading to its formation.

MORPHINE

,CH3N

Ho 0

N- DEMETYLA

243


Before the turn of the century, earlier workers(Stolnikow, 1884; Marquis, 1896) suggested thepossibility that morphine might be bound or con-jugated in the body, but the evidence was not con-clusive. In 1938 Endo reported that when the urinefrom morphinized rabbits was allowed to stand indilute sulfuric acid, a larger amount of morphine wasrecovered than from untreated urine. The existenceof bound morphine as a morphine metabolite inurine was demonstrated in a convincing fashion in1940 by Gross & Thompson using dogs, and shortlyafter by Oberst (1940) in human addicts. Theisolation of crystalline bound morphine from theurine of the dog and its identification as the glucu-ronide was reported by Woods in 1954, and sub-sequently from the urine of human addicts byFujimoto & Way (1954, 1957, 1958). It has beensuggested (Thompson & Gross, 1941; Woods, 1954)that more than one form of bound morphine exists,but this view has not been universally accepted. It isof interest, therefore, to consider in some detail themore recent studies with respect to the nature ofbound morphine.

Gross & Thompson (1940) hydrolysed samples ofurine from morphinized dogs with one-tenth thevolume of concentrated hydrochloric acid in anautoclave for 30 minutes at a pressure of 15 poundsper square inch (I atm.). They found an increase inphenolic substances and reported that this was dueprimarily to morphine. The morphine liberated fromthe conjugated or bound morphine was isolated byrepeated extraction and reprecipitation to obtain ayellow-white residue. This residue gave the usualcolour test for morphine and, when injected intodogs, produced effects similar to equivalent doses ofauthentic morphine. The excretion pattern of thebiologically obtained morphine was similar to thatof pure morphine in dogs (Gross & Thompson,1940). In a follow-up study (Thompson & Gross,1941) further evidence was obtained. Specific rota-tion measurements of the isolated morphine matchedthose of an authentic sample. The diacetyl derivativewas also made and the mixed melting-point withpure heroin showed no change.

Oberst (1940) demonstrated the presence of boundmorphine in the urine ofhuman addicts. He acidifiedthe urine with one-fifth its volume of acid and afterrefluxing for three hours he found an increasedamount of morphine. The identity of the lattercompound was established by mixed melting-pointdeterminations and by colorimetric and nephelo-metric tests for morphine. The specific rotation of

the derived morphine and the mixed melting-pointafter conversion to heroin were also determined.Most early workers were of the opinion that

bound morphine was a glucuronide. Ashdown(1890) and Mayer (1899) reported that a glucuronidewas excreted in the urine after morphine administra-tion and acknowledged that the original observationwas reported by Mering in 1874. Mayer postulatedthat morphine was conjugated as glucuronide on thebasis of optical rotation measurements made onhydrolysed urine from patients receiving morphine.Endo (1938) felt that morphine was conjugated withglucuronic acid, but the English abstract of hisoriginal manuscript (in Japanese) cited no experi-mental observations to support his conclusion.Oberst (1941) observed a correlation between theamount of glucuronic acid excreted in the urine andthe dose of morphine. Direct proof was lacking,however, that morphine and glucuronic acid werepaired with each other. Oberst & Gross (1944)prepared a bound morphine, morphine sulfuricether, and studied its actions and fate, but no evidenceexists that such a derivative may be biosynthesizedfrom morphine.More recently three groups of workers reported

almost simultaneously that they had obtainedevidence that conjugated morphine was a glucuronide(Fujimoto & Way, 1954; Seibert, Williams &Huggins, 1954; Woods, 1954). Convincing proofwasfurnished by Woods (1954), who was the first toreport the isolation of bound or conjugated mor-phine in crystalline form from the urine and bile ofdogs. Subsequently, a crystalline product wasisolated from the urine of addicts by Fujimoto &Way (1957) which yielded an infra-red curve identicalwith that obtained from the dog (Fujimoto & Way,1958).The presence of morphine in bound morphine was

established by powder X-ray diffraction analysis ofthe dinitrophenyl derivatives of the hydrolysedmorphine conjugate (Fujimoto & Way, 1958).Liberated morphine after hydrolysis was alsoidentified by paper chromatography (Fujimoto &Way, 1954; Seibert, Williams & Huggins, 1954;Woods, 1954) and by mixed melting-point de-terminations of the free base, the picrate (Seibert,Williams & Huggins, 1954) and the diacetyl (Woods,1954) derivatives.The product conjugated with morphine was

identified as glucuronic acid by various chemical andphysical tests. The morphine conjugate aftercounter-current distribution was hydrolysed and

'2AA


analysed for morphine and glucuronic content. Themorphine and glucuronic curves were found to bealmost identical within the limits of experimentalerror (Fujimoto & Way, 1957). The infra-red curveof the conjugate was found to give a very strongband in the 3 ,u region, which is in agreement with theOH stretch frequency broadened by association to beexpected from a polyhydroxy glucuronide com-pound. A strong absorption in the 6-7 ,u region wasinterpreted as being in agreement with the presenceof a carboxylate ion (Fujimoto & Way, 1958). Thebound morphine, separated by paper chromato-graphy, on hydrolysis with acid or /-glucuronidaseyielded a positive test for glucuronide as well as formorphine (Seibert, Williams & Huggins, 1954).Elemental analyses were found to be consistent withthe assay procedures for a conjugate of morphinewith glucuronic acid being combined in the ratio ofone to one, and associated with two molecules ofwater (Woods, 1954).The molecular site of conjugation of bound

morphine with glucuronic acid was established to beat the 3-phenolic position on the basis that thephenol reagent used in the morphine determinationprocedure did not give the *olour with the con-jugated material until after acid pressure hydrolysis(Fujimoto & Way, 1957). The ultraviolet absorptioncharacteristics of the conjugate in acid and base gaveno evidence as to the presence of a free phenol.Morphine, like other phenols, shows a bathochromicshift with an increase in pH, whereas the ultravioletcurve for morphine conjugate was more similar tothat for codeine, in which the phenol group ismasked (Fujimoto & Way, 1958).The experimental evidence indicated that the

morphine conjugate is a zwitterion (Fujimoto &Way, 1958). The infra-red curve of bound morphineshowed a maximum at 6.2 , with no band between5.6 and 6.2 ,t. This was interpreted as meaning thatthe carboxyl group of the glucuronic acid moiety ispresent in an ionized form. Such an interpretationwould necessitate the presence of a positive chargeon the piperdine nitrogen. The titration curve of themorphine conjugate gave two pK values which wereconsistent with the values predicted by Kumler(1955), who based his calculations on the assumptionthat morphine glucuronide exists as an ampholyte.The question whether there is more than one

form of bound morphine is of considerable interest.Thompson & Gross (1941) reported that two formsof bound morphine existed in dog urine, one "easily"hydrolysable and the other " difficultly " hydro-

lysable. Woods (1954) also suggests that two formsof bound morphine may be excreted in dog urinethe monoglucuronide, which is crystalline andpossesses low water solubility, and another com-pound, possibly a di-conjugated morphine, which isamorphous and possesses high water solubility.Fujimoto & Way (1957), however, concluded that3-morphine-monoglucuronide is the only boundmorphine present in appreciable quantity in theurine of addicts. While they did not completelyexclude the possibility that other bound morphinescould be formed, they felt that any quantitiesformed would be of a low order of magnitude.Thompson & Gross (1941) concluded that there

were two forms of bound morphine from followingthe rate of hydrolysis of the bound morphine in dogurine. Upon adjusting the pH of the urine tobetween 1 and 2 and heating at 100°C, morphineappeared to be set free at a fairly uniform rate for thefirst 60 minutes, but thereafter the reaction proceed-ed at a very slow rate. The morphine fraction liber-ated by two hours of hydrolysis under these condi-tions was designated the " easily" hydrolysablefraction. The remainder of the bound morphine,which yielded morphine only after 30 minutes in theautoclave in the presence of 5 % hydrochloric acid,was called the " difficultly" hydrolysable fraction.Of the two forms, the latter was present in muchlarger quantities in both tolerant and non-tolerantdogs, but significant amounts of the " easily " hydro-lysable fraction were also found. Although itappears plausible to accept the conclusions ofThompson & Gross, kinetic studies on a system ascomplicated as urine place restrictions on the inter-pretations. Any substance in urine which mightcatalyse the hydrolysis of bound morphine and isslowly destroyed by heat would show a drop in theabsolute rate of hydrolysis of bound morphineover a given time. Thus, the findings of Thompson& Gross are suggestive but far from conclusiveevidence of the existence of two bound morphines.Woods (1954) also suggested that there must be at

least two bound forms of morphine excreted by thedog, although his paper presents chromatographicevidence to the contrary. Woods based his con-clusions instead on the results he obtained with hisexperiments designed to isolate bound morphine.He was successful in isolating a crystalline substancefrom the urine which he demonstrated to be mor-phine-3-monoglucuronide. However, he found thata far larger fraction of bound material was stillpresent which was amorphous, highly water soluble,

245


and contained 30% glucuronic acid and 40% mor-phine. He suggested that the compound was adi-conjugated morphine, the alcoholic group beingconjugated as a glucuronide and the phenolic groupas an ethereal sulfate. No data on sulfur analysiswere given.

Fujimoto & Way (1957), in their studies on urineobtained from addicts, found an amorphous,highly water-soluble morphine conjugate whichapparently resembled the amorphous bound mor-phine Woods found in dog urine. However, afterfurther purification and especially after removal ofninhydrin-reacting material, a crystalline substancewas isolated which was identical with the morphine-3-monoglucuronide isolated by Woods (1954).Since this crystalline monoglucuronide was entirelyderived from the amorphous material, it was con-cluded that the crystalline and amorphous boundmorphine actually represented simply different statesof a single substance. Moreover, since the amor-phous conjugate was found in large amounts in theurine and was the only bound morphine noted, itwas concluded that only one form of bound mor-phine is excreted in the urine in any appreciablequantity. Further evidence supporting this con-clusion that only one major bound form was presentin urine was derived from experiments using counter-current distribution, paper chromatography andinfra-red analysis. The paper chromatographystudies of Woods (1954) and of Seibert, Williams &Huggins (1954) were also cited in support of theargument for a single main form of bound morphine.Subsequently, Woods (personal communication)found that the water-soluble amorphous boundmcrphine in dog urine could also be changed intothe crystalline poorly water-soluble substance bymanipulative techniques. Thus it would appear thatmorphine-3-monoglucuronide may be strongly asso-ciated with substances in urine which enhance itswater solubility and prevent it from crystallizing.It is quite possible that the difference in phy-sical properties between the " easily " hydrolysableand the " difficultly " hydrolysable forms of boundmorphine found in dog urine may be relatedto these factors of asseciation with extraneoussubstances.

Conjugation of morphine appears to be a meta-bolic pathway common to many species. Increasedamounts of morphine were found to have beenliberated after acid hydrolysis from the excreta ortissues of monkeys (Mellett & Woods, 1956),rabbits (Hosoya, 1959), rats (Zauder, 1952; Way,

Sung & Fujimoto, 1954), and mice (Kokka, Elliott& Way, unpublished data) given morphine. Therate of conjugation of morphine in vivo did not differmaterially among rats ranging in age from 2 days to32 days. However, newborn rats less than 8 hours ofage showed a somewhat lesser ability to conjugatemorphine (Kupferberg & Way, unpublished data).Although the crystalline compound was not isolatedin each instance, the conjugated product is verylikely morphine-3-monoglucuronide.A great deal of information on the sequences

leading to the conjugation of morphine has beengained from studies using isolated tissue preparations.The most extensive studies have involved the liverand have resulted in considerable clarification ofearly observations that morphine " is altered " or" disappears " when perfused through the liver(Hatcher & Gold, 1929; Rink, Gray & Rueckert,1956) or incubated with liver mince or slices (Ko,1937; Kuwahara, 1938; Inoue, 1940; Bernheim &Bemheim, 1944, 1945; Fawaz, 1948; Deneau,Woods & Seevers, 1953; Zauder, 1952).Inoue (1940) studied the ability of liver tissues of

various animal species to metabolize morphine in anattempt to arrive at some conclusion concerning theetiology of natural tolerance. He reported that whena 10-mg quantity of morphine hydrochloride wasadded to 10 g of liver from various animals andincubated at room temperature for 30 minutes, theproportion recovered was as follows: dogs, 52-82 %;cats, 51-73 %; rabbits, 48-61 %; guinea-pigs, 48-57 %;pigeons, 43-60 %; hens, 41-58 %. He concluded thatthe difference in the ability of liver to alter morphinemay be a cause (not necessarily the chief one) of thedifference in the natural tolerance to morphine ofvarious animal species. Inasmuch as a full descrip-tion of the experimental methods is not presented,it is difficult to evaluate the findings. It appears,however, that even if the concept turns out to becorrect, the experiments as presented are inadequateto establish the point.Bernheim & Bernheim (1944) found that when

morphine was added to rat liver slices, the compounddisappeared under aerobic conditions. Kidney andbrain slices were ineffectual. Liver cell suspensionsshowed reduced activity. They concluded thatoxidation of morphine had resulted and ruled outconjugation as a possible mechanism. In a secondstudy (Bernheim & Bernheim, 1945), however, theyreported that their method for determining con-jugated morphine was inadequate. After applyingthe method of Gross & Thompson (1940) for bound

246


morphine, they concluded that conjugation and notoxidation accounts for the disappearance of mor-phine when the compound is incubated with liverslices from the dog, cat, rat, and guinea-pig. Thereaction is completely inhibited by M/1500 iodo-acetic acid, M/500 sodium cyanide, or M/50 sodiumfluoride. With respect to the cat, the disappearanceof morphine in the presence of liver slices ought tobe explored more thoroughly, inasmuch as it hasbeen reported that cats do not form glucuronidesin vivo (Robinson & Williams, 1958), presumablybecause of the lack of glucuronyl transferase(Dutton & Grieg, 1957).A fairly extensive store of knowledge has been

attained in recent years which emphasizes theimportance of carbohydrate metabolism in theformation of glucuronides by liver preparations. Itappears that the key substance in glucuronide syn-thesis is uridine diphosphate glucose (UDPG) andthat this compound is derived from uridine triphos-phate and glucose-l-phosphate (Mills, Lockhead& Smith, 1958). Since adenosine triphosphate(ATP) is required for both hexose phosphate forma-tion and resynthesis of uridine triphosphate, theavailability of carbohydrate substrates and theregeneration of ATP will greatly influence the abilityof liver slices or liver mince to conjugate morphine.Thus, with rat liver slices respiring in Krebs-Ringersolution the replacement of sodium by potassiumresulted in the concomitant delay of glycogenolysisand of morphine conjugation, although the potas-sium had no effect on the conjugation of morphinewhen glucose was added to the medium (Marks &Huggins, 1959). It is quite possible that depletionof carbohydrate stores, which, in turn, limits theamount of UDPG formed, is primarily responsiblefor the impaired ability to conjugate morphineobserved in liver mince obtained from traumatizedrats (Goldbaum, Gray, Rink, Rueckert & Ostash-ever, 1956) or from tolerant rats during withdrawal(Deneau, Woods & Seevers, 1953).The importance of UDPG in the conjugation of

morphine was first suggested by the studies ofStrominger, Kalckar, Axelrod & Maxwell (1954)following the demonstration by Dutton & Storey(1954) that uridine diphosphate glucuronic acid(UDPGA) participates in glucuronide synthesis.Strominger, Kalckar, Axelrod & Maxwell (1954)have shown that when the supernatant fluid fromhomogenates of guinea-pig or calf liver containingboth microsomes and soluble enzymes was incubatedwith UDPG, diphosphopyridine nucleotide (DPN+),

MgCl2, and morphine, a marked reduction in freephenol concentration occurred. It may be inferredthat synthesis of morphine glucuronide took placeunder these conditions, although the product wasnot isolated and identified. Further experimentsindicated that UDPG is first oxidized to UDPGAby a dehydrogenase system present in the particle-free supernatant and that for every mole oxidizedtwo moles of DPN+ were reduced. Strominger,Maxwell, Axelrod & Kalckar (1957) have purifiedthe enzyme from calf liver and Strominger & Map-son (1957) have obtained a purified enzyme from peaseedlings and report that it closely resembles theone from the calf liver. The enzyme-catalysedoxidation of UDPG is inhibited by sodium fluoridebut not by iodoacetate. Recently, Takemori (1960)has found that the UDPG dehydrogenase activityin male rat liver is increased often after only a singleinjection of morphine.The product of oxidation is not a substrate for

/-glucuronidase, indicating an a-linkage of glucu-ronic acid in UDPGA. An enzyme or enzymespresent in the microsomes catalyse the transfer ofthe glucuronic acid moiety of UDPGA to morphine.Several steps may be involved in this reaction sincethe transferase has no 3-glucuronidase activity(Isselbacher, 1956). Thus, an inversion of thea-linkage of glucuronic acid probably takes placejust prior to or during the final coupling with mor-phine.

Inscoe & Axelrod (1960), using o-aminophenol asthe substrate, have shown that liver microsomes ofnewborn rats and guinea-pigs have increasedglucuronyl transferase activity 24 hours afterinjection of 3,4-benzpyrene. They have found alsothat in adult rats the in vitro activity of the glu-curonyl transferase system of liver microsomes issex-dependent and can be markedly enhanced byprevious in vivo treatment of the rat with androgensor the carcinogenic polycyclic hydrocarbon, 2,3-benzpyrene, or the activity can be as dramaticallyreduced by pretreatment of rats with oestrogens.During cold stress the reduction in glucuronyltransferase activity is probably due to a reductionin total liver microsomal content as evidenced by thereduction in microsomal nitrogen. However, inchronically morphinized rats a reduction in micro-somal glucuronyl transferase activity occurs in theabsence of any decrease in nitrogen content of themicrosomes (Takemori, 1960).

These interesting changes in the in vitro activityof the glucuronyl transferase system that result

247


from pretreatment of the animal in vivo raise thequestion of whether one can predict the in vivoability to conjugate morphine on the basis of thein vitro liver microsomal activity. For example, onewould expect to find a marked increase in urinaryconjugated morphine in male rats or androgen-treated female rats as compared with normal femalesor oestrogen-treated males, and in benzpyrene-treated rats as compared with untreated ones.Conversely, a decreased urinary excretion of con-jugated morphine should occur during cold stress,or during chronic morphine administration. Wehave found no published studies in which suchcomparisons have been made except in respect ofchronically morphinized rats and here the in vivoreduction in urinary conjugated morphine appearsto be correlated with the reduced glucuronyl trans-ferase activity of the microsomes. However, theresults obtained with liver slices suggest that theparallelism may be merely fortuitous, since severalexperiments have shown that with liver slices frommorphinized rats there is no reduction in the abilityto conjugate morphine (Fawaz, 1948; Way, Sung &Fujimoto, 1954) and possibly even an increase(Zauder, 1952). Moreover, the conjugating abilityof liver slices obtained from rats receiving otherkinds of treatment, such as adrenalectomy or ACTHinjections, provides no basis for predicting thecomparable conjugating ability in vivo. Thus, liverslices from adrenalectomized rats are either unableto conjugate morphine at all (Zauder, 1952) or showno change from normal (Way, Sung & Fujimoto,1954), whereas adrenalectomized rats show higherconcentrations of both free and conjugated morphinein the plasma than are found in normal rats (Adler,Elliott & George, 1957). The ability of liver slicesfrom ACTH-treated rats to conjugate morphine isincreased (Zauder, 1952), but plasma levels of con-

jugated morphine in ACTH-treated rats are lowerthan those in normal rats (Adler, Elliott & George,1957).In general, extrapolation from liver slice in vitro

experiments to in vivo predictions may be meaning-less since, on the one hand, under in vitro conditionsthe residual carbohydrate stores can profoundlyaffect the conjugation of morphine (see above) and,on the other hand, under in vivo conditions tissuesother than the liver may contribute to the totalamount of conjugated morphine excreted. Finally,as will be discussed in a later section, although manyconditions profoundly affect the conjugation mecha-nism in vitro, there is no concomitant effect on the

over-all pharmacological action of morphine.Despite this lack of correlation, conjugation ofmorphine must be viewed as a detoxication processin so far as the product of conjugation is far lessactive than the parent substance.

N-demethylationThe second of the metabolic pathways outlined in

Fig. 3-namely, N-demethylation-has thus farproved difficult to categorize as either a detoxicationor an enhancement process. Moreover, in neitherin vitro nor in vivo experiments has the product ofdemethylation, normorphine, been isolated in crys-talline form. While the in vitro experiments have hadsome measure of success in establishing normorphineas a morphine metabolite by providing evidence ofa product with the solubility characteristics of nor-morphine, similar success in in vivo experiments hasbeen obtained only in respect of the rat. The bulk ofthe experimental evidence available at present sug-gesting normorphine formation in vivo are theobservations that 14CO2 appears in the breathshortly after injection of morphine-N-14CH3 inseveral species, including man.

In rats (Wistar type) the pulmonary excretion of14CO2 after subcutaneous administration of 5 mg/kgof morphine-N-methyl-14C hydrochloride was foundto be greater in males than in females (March &Elliott, 1954). The excretion rate for both groups wasfound to be more rapid during the first two hours.Four male rats excreted nearly 5% of the dose asradioactive CO2 within six hours, whereas the per-centage excreted by seven females over the same timeinterval was less than 0.5 %. Four female rats treatedwith a total of 45 mg of cyclopentyltestosterone pro-pionate in subcutaneous doses over a 38-dayperiod prior to morphine administration exhibited apulmonary excretion curve almost identical with thatof the male rats. In male Sprague-Dawley rats aftersubcutaneous injection of 10 mg/kg tritiated mor-phine the 24-hour urine contained some 5% of thedose as a free and conjugated radioactive metabolitethat was chromatographically homogeneous withnormorphine in two separate solvent systems(Misra, Mule & Woods, 1961). It has been reportedin a preliminary communication that paper chroma-tographic evidence was obtained for the presenceof two normorphine derivatives in the liver andbrain of rats given morphine (Penna, Arevalo, Fer-nandez, Navia & Mardones, 1959). An assessmentof these findings must await the publication of theexperiments in greater detail.

248


In the dog and the monkey, 24 hours after sub-cutaneous injection of 2 mg of morphine-N-14CH3per kg, the exhaled labelled carbon dioxide repre-sented, respectively, 0.2% and 1.2% of the dose(Mellett & Woods, 1961).

In man, the pulmonary excretion of 14CO2 in fivesubjects given 10-15 mg of morphine-N-methyl-14Csulfate intramuscularly ranged from 3.5 % to 6% ofthe injected dose in 24 hours (Elliott, Tolbert, Adler& Anderson, 1954). No sex difference in the amountof expired 14CO2 was observed between the two maleand three female subjects. The peak rate of 14CO2excretion in the breath was found to be between 30and 90 minutes after drug administration. A plateauin the rate of excretion occurred after six hours.Measurable amounts of radioactivity were found tobe present in two subjects 4-5 days after drug admini-stration. It was suggested as a consequence thatsome transfer of 14CH3 groups from the morphinemolecule may occur from which 14CO2 is slowlyliberated by catabolic processes.

In none of these early '4C studies was there anattempt to identify any of the morphine metabolitesother than carbon dioxide. However, since N-de-methylation with the release of the correspondingnor-compound has been established as an in vivometabolic pathway for such closely related con-geners as codeine and pethidine (Adler, Fujimoto,Way & Baker, 1955; Plotnikoff, Elliott & Way, 1952;Plotnikoff, Way & Elliott, 1956) the inference that asimilar process occurs with morphine is easy to make.It can be noted that the amount of 14CO2 obtained inthe breath in man after injection of morphine-N-14CH3 is about the same as that obtained in the breathin man after injection of codeine-N-"4CH3 (Adler,Fujimoto, Way & Baker, 1955) and accordingly, ifthe N-demethylation processes are analogous, some10% of a therapeutic dose of morphine ought to befound in a 24-hour urine sample as free and boundnormorphine. However, attempts to show that anynormorphine is present in urine have met with nosuccess, and the recent work of Rapoport (personalcommunication) indicates that the total amount ofnormorphine excreted in 72 hours in the urine ofman must be less than 0.5% of the dose of morphine.The absence of normorphine from the urine after

morphine administration is, of course, no proof thatthe metabolite is not formed in vivo. Any one ofseveral explanations could account for a failure todetect normorphine even if small amounts werereleased in the body. Although it has been shownthat after very large doses (75-150 mg) of normor-

phine in man about 75% of the dose appears in urineas an extremely labile conjugate plus a large propor-tion of free alkaloid (Sloan, Eisenman, Fraser &Isbell, 1958), a different fate may characterize therelatively small amounts which might be expected tobe released from a therapeutic dose of morphine invivo. On purely speculative grounds the reviewerssuggest the following possibilities: (a) biosyntheticnormorphine may be excreted preferentially in thefaeces; (b) biosynthetic normorphine may be furthermetabolized in an unknown manner; (c) biosyn-thetic normorphine may be retained by tissues for afairly prolonged period of time. Although the lastsuggestion is compatible with the pharmacology ofnormorphine in that marked cumulative effects ofthe drug are seen after multiple doses (Fraser,Wikler, Van Horn, Eisenman & Isbell, 1958), thesesuggested possibilities can be placed in proper per-spective only when data are furnished regarding theexcretion of normorphine after injection of smalldoses of this compound.That morphine may be demethylated in vitro was

first suggested by the work of March & Elliott (1954).Using rat liver slices and morphine-N-14CH3 assubstrate, these authors showed that under aerobicconditions there is a release of 14CO2. Subsequently,Axelrod (1955a, 1956a) found that an oxidativemechanism located in liver microsome preparationsis responsible for the oxidation of the methyl groupto formaldehyde. Qualitative evidence for normor-phine formation was obtained by paper chromato-graphy and it is assumed that the amount offormaldehyde formed during the reaction bears astoichiometric relationship to the amount of nor-morphine released.The demethylating system described by Axelrod

resembles the system responsible for oxidativedemethylation of methylated aminazo dyes, firstobserved by Mueller & Miller (1953) in liver homo-genates and later (Conney, Brown, Miller & Miller,1957) located in the microsomes. Although the twodemethylating systems are not identical (Takemori& Mannering, 1958), both enzyme systems requirereduced triphosphopyridine nucleotide (TPNH) andoxygen. The work of Gillette, Brodie & LaDu(1957) has shown that oxidative dealkylation ofmono-methyl-4-aminoantipyrine occurs when thereis a concomitant oxidation of TPNH by a specificTPNH-oxidase and not when TPNH is oxidized bythe cytochrome system. The same authors havefurther shown that the reaction of TPNH withTPNH-oxidase yields " organic peroxides " even in

249


the absence of drug substrates. This suggests thatthe initial oxidative attack at the N-methyl positionof morphine may depend on the formation of aspecific oxidant, possibly an " activated " peroxide.Apparently H202 per se does not function in themicrosome system but can function in an iron-containing model system since generation of H202by the glucose/glucose-oxidase system was effectivein promoting oxidative dealkylation of N-alkyl ami-nes in the model system but not in the microsomesystem (Gillette, Dingell & Brodie, 1958). If theinitial reaction between an " activated peroxide "and morphine results in the formation of an N-oxide, the subsequent steps leading to formaldehyderelease need not necessarily depend on enzymaticreactions. It has been shown that rearrangement oftertiary amine N-oxides to a carbinol compoundfollowed by hydrolysis to formaldehyde and thesecondary amine can occur under mild chemicalconditions (pH 5-7, 38°C) when catalysed by aferric-ion-tartrate complex (Fish, Johnson, Law-rence & Homing, 1955; Fish, Sweeley, Johnson,Lawrence & Horning, 1956). Under such conditionsone would expect a secondary reaction to occur-namely, reduction of some of the N-oxide by form-aldehyde leading to regeneration of the tertiaryamine and oxidation of formaldehyde to formic acid.This secondary reaction would be minimized in asystem containing a formaldehyde-trapping re-agent. If morphine-N-oxide is, indeed, formedduring the demethylation process it cannot be ignoredas a possible contributor to some of the morphineeffects, since after injection it exerts certain centraldepressant effects such as respiratory depressionand antitussive action, although it apparently hasno analgesic properties (Kelentey, Stenszky, Czollner,Szlavik & Meszaros, 1957).Axelrod (1956a) has found that microsomes

prepared from livers of the rat, rabbit, and guinea-pig, but not the mouse, are capable of catalysing theformation of formaldehyde from morphine. Theinability of mouse microsomes to catalyse the reac-tion is apparently restricted to certain strains, sinceTakemori & Mannering (1958) have found consider-able N-demethylating activity in mouse liver micro-some preparations. Microsomes prepared from thekidney, brain, muscle, or spleen of male rats areinactive in .this respect (Axelrod, 1956a). Speciesdifferences in activity of the enzyme system occur,but these may be related to the presence or absenceof co-factors. For example, rabbit livers appear tocontain a dialysable co-factor necessary for N-

demethylation. This inference is based on Axelrod'sfinding that the only reaction catalysed by a dialysedpreparation of rabbit liver microsomes when in-cubated with codeine is the formation of morphineand formaldehyde in equimolar amounts (Axelrod,1955b). It appears to the reviewers that if such adialysed preparation were capable of effectingN-demethylation an additional amount of form-aldehyde would be formed from the N-methyl ofthe released morphine as well as from the N-methylof the otherwise unmetabolized codeine. Anotherexample of the importance of co-factors in speciesdifference is the finding that rabbit livers do not, but ratlivers do, contain a heat-labile inhibitory factor local-ized in the nuclei and mitochondria (Axelrod, 1956a).The early observations made by March & Elliott

(1954) of the large sex difference in the ability of ratsto form 14CO2 from morphine-N-'4CH3 both in vivoand with liver slices, and the marked effect of pre-treatment of the animal with either androgens oroestrogens, were confirmed and extended by Axelrod(1956a) to rat liver microsomal preparations.Furthermore, Axelrod & Cochin (1957) report amarked (non-competitive) inhibition of microsomalenzymatic N-demethylation of morphine by addednalorphine. Although March & Elliott (1954) foundno such inhibition in their liver slice experiments, wefeel that this apparent discrepancy may possibly berelated to the fact that nalorphine is rapidly conju-gated in liver slices (Seibert & Huggins, 1953) and,hence, may not reach the demethylating enzymeswithin the microsomes.

In 1956 Axelrod found that in chronically mor-phinized male rats there was a marked reduction inthe ability of the liver microsomes to demethylatemorphine, dilaudid or meperidine; cocaine, on theother hand, was N-demethylated exceedingly wellby microsomes from either morphine-treated oruntreated rats. This interesting finding has served asthe basis for a provocative hypothesis concerning themechanism of tolerance which assumes that the liverN-demethylating enzyme and the receptors in thecentral nervous system are closely related (Axelrod,1956b, Cochin & Axelrod, 1959). A detailed dis-cussion of the evidence for, and the implications of,this hypothesis will be found in a later instalment.An equally provocative hypothesis relating N-

demethylation to analgesia has been advanced byBeckett and his colleagues (Beckett, Casy & Harper,1956; Beckett, Casy, Harper & Phillips, 1956). Thishypothesis, which would assign a prominent role ininitiating analgesia to the N-demethylation process,

250


will also be discussed in detail in the final instalmentof this series of papers. Both of these hypotheseshave stimulated much excellent experimental workon the examination of factors affecting N-de-methylation of morphine in both tolerant and non-tolerant animals (Axelrod & Cochin, 1957; Lockett& Davis, 1958; Mannering & Takemori, 1958;Takemori & Mannering, 1958; Chernov, Miller &Mannering, 1959; Cochin & Axelrod, 1959; Cochin& Economon, 1959; Herken, Neubert & Timm-ler, 1959; Horlington & Lockett, 1959; Cochin &Sokoloff, 1960; Elison, Rapoport & Elliott, to bepublished). Suffice it to say at this time that attemptsto classify N-demethylation of morphine either asan activation or as a detoxication process are thwartedby the complex nature of the pharmacologicalproperties of normorphine.

Oxidation

Since oxidation of morphine to pseudomorphine(also known as dehydromorphine, oxydimorphineor 2,2'-bimorphine} occurs very easily under mildchemical conditions (Bentley & Dyke, 1959), it is notsurprising that this has been presumed to occurunder biological conditions as well. Althoughpseudomorphine exhibits some striking pharmaco-logical properties following intravenous injection,especially on the cardiovascular system where itproduces intense depressor effects (Travell, 1932;Schmidt & Livingston, 1933), it is of special interestbecause it has long been considered to be responsiblefor some of the effects ascribed to morphine, particu-larly the withdrawal syndrome. This hypothesis,however, has received practically no support fromexperimental evidence in the past, as can be judgedfrom Eddy's review of the conflicting and incon-clusive findings (Krueger, Eddy & Sumwalt, 1941).More recent attempts to establish the presence of

pseudomorphine in tissues after morphine admini-stration have been unsuccessful. Fichtenberg (1951)has reported that pseudomorphine is absent fromblood or muscle of normal rats or from blood,muscle, or liver of habituated rats after injection oflarge doses of morphine. Her conclusions are basedon the results of a bio-assay procedure reported to besensitive to quantities of pseudomorphine of theorder of 20-50 [kg/ml of extracted solution. Hosoya& Brody (1954) have demonstrated the formation ofa compound in vitro, indistinguishable by chromato-graphic analysis from authentic pseudomorphine, byrat liver homogenates when fortified with cyto-chrome c and incubated with morphine under aero-

bic conditions. Under these conditions there was aconcomitant reduction in the formation of morphineglucuronide which normally takes place in vitro inthe absence of added cytochrome c. These authors,however, were unsuccessful in their attempts todemonstrate the presence of the compound in ratliver after morphine administration in vivo. Inas-much as chromatographic homogeneity in any onesolvent system does not constitute proof of identity,it is possible that the compound obtained by Hosoya& Brody is something other than pseudomorphine.It is interesting to note that the chromatographicbehaviour of this compound resembles that of anunknown morphine metabolite present in the urineof a strain of rats showing low urine and plasmaconcentrations of bound morphine after injection ofmorphine-N-14CH3. The metabolite was absent fromthe urine of another strain of rats showing twofold tothreefold higher urine and plasma values of boundmorphine (Adler, Elliott & George, 1957). Theunknown metabolite, although chromatographicallyhomogeneous with pseudomorphine, differed mar-kedly from pseudomorphine in its spectrophoto-metric properties (Adler, unpublished data). It isthus apparent that a small part of the dose of mor-phine is metabolized in an unknown manner pro-bably not by oxidation to pseudomorphine.An " oxidized morphine " has recently been pro-

duced by partial chemical oxidation of morphine;this compound is different from pseudomorphineand is pharmacologically more active than morphinein many respects (Woods, Daly, Haggart & Seevers,1952). However, a full account of the work has notbeen published and the preliminary announcementsgive no indication that the compound can be derivedfrom morphine by biotransformation.

3-0-methylationThe most recently discovered metabolic pathway

for the biotransformation of morphine indicates thatcodeine is a metabolite of morphine. Qualitativeevidence of the presence of a small amount ofcodeine in pooled urine samples of male rats given5 mg/kg morphine-N-14CH3 was obtained by paperpartition chromatography. Evidence of codeine for-mation from morphine-N-14CH3 in in vitro experi-ments using liver homogenates or the liver solublefraction was obtained by counter-current distribu-tion analysis of radioactivity and basic amines. Theexperiments showed that methionine is not themethyl donor for the methoxy group of codeine(Elison, 1961).

251


EXCRETION

With certain quantitative differences, variousspecies excrete morphine more or less in a similarfashion. Most of a given dose of morphine can beaccounted for, largely in the urine and to a lesser butsignificant extent in the bile and faeces. Otherfluids, such as saliva, milk, and perspiration, appearto play only a very minor role in the excretion ofmorphine. The forms in which administered mor-

phine have been isolated in the crystalline state fromurine are free morphine and morphine-3-mono-glucuronide (bound morphine). Free morphine isexcreted in the urine by all species in small butsignificant amounts, the average percentage urinaryexcretion of free morphine being lowest in man(about 5 %) and highest in rats (about 20 %). Boundmorphine represents the main form in which mor-

phine appears in urine in the various species whichhave been studied, the amount excreted usually beingtwofold to sevenfold greater than that of free mor-phine. Normorphine and codeine may be excretedin minor amounts as biotransformation products ofmorphine, but the evidence is preliminary and hasnot been established with certainty in higher species.

Renal

Man. Morphine has been accounted for in theurine largely as a glucuronide conjugate and in lesseramounts as free morphine. The available evidencesuggests that unknown metabclic products are alsomostly excreted by way of the kidneys.

Free morphine constitutes only a minor fraction ofthe total dose of morphine which can be accountedfor in the urine. While most of the studies have beencarried out on addicts and post-addicts, a recentstudy by Paerregaard (1957b) on non-addicts indi-cates no striking differences in the excretory pattern.The excretion of free morphine is quite prompt.

Nearly all the excreted free morphine can be foundin urine within 8 hours after intravenous or sub-cutaneous administration (Paerregaard, 1957b), butonly negligible amounts are excreted at 3 hours(Eisenman, 1945). The excretion of free morphine,however, may persist in trace amounts for severaldays (Fry, Light, Torrance & Wolff, 1929; Oberst,1940; Paerregaard, 1957b).The average amount of free morphine excreted in

the urine within a 24-hour period is approximately7 %, the range being 1-14%. These mean values werecalculated from those data in Table 3 which were

considered to be reliable. In general, the values

reported by different workers are in good agreement.The variation in the amount of morphine excreted inthe urine can be attributed chiefly to variation amongindividuals, but even the same individual varies con-siderably in his ability to excrete morphine from dayto day (Fry, Light, Torrance & Wolff, 1929; Deckert,1936a, 1936b). The amount of free morphine ap-pearing in the urine is dependent in part on thedosage of morphine administered. In studies on non-addicts, less than 1 % of the dose can be recovered inthe urine 24 hours after subcutaneous administrationof 5 mg of morphine (Paerregaard, 1957b). With adose of 10 mg, the relative amount of morphineexcreted is increased, but further increases in thedose of morphine generally result only in increasedamounts of free morphine commensurate with thedosage administered. Oberst (1940) found littledifference in the relative amounts of free morphineexcreted in the urine using doses of morphineranging from 30 to 3317 mg.A few studies of factors which may influence the

excretion of morphine have been reported. It isclaimed that administration of lecithin and glucoseaccelerates the excretion of free morphine (Chopra,Chopra & Roy, 1941). Neostigmine does notproduce any marked change in the excretory patternof morphine (Himmelsbach, Oberst, Brown &Williams, 1942).Morphine monoglucuronide constitutes the chief

metabolic product of morphine and is excretedalmost exclusively in the urine. The amounts ofurinary bound morphine reported by various in-vestigators are summarized in Table 3. The valuesshow a range of 11-45% of the dose.

Recent work by Rapoport (personal communica-tion) on non-addicts indicates that conditions forhydrolysing bound morphine in the earlier studiesmay not always have been optimal. Using morevigorous hydrolytic conditions he and his associatesfound that more than half the administered dose ofmorphine-N-14CH3 could be attributed to a morphineconjugate. Their data on the 24-hour excretion oftotal morphine, that is, free and bound morphine,indicate that approximately 70% of the dose ofmorphine can be accounted for by these two com-pounds. The urinary excretion of 14C continues forseveral days beyond the initial 24-hour period andvirtually all of the dose of radioactivity can berecovered from urine within 72 hours after injectionof morphine-N-14CH3. No evidence has been foundfor the excretion of other forms of bound morphinenor for the excretion of free or bound normorphine.

252


TABLE 3

URINARY EXCPETION OF MORPHINE IN MAN

Percentage ofNo. morphine excreted

Subjects exam- Dose Route a in 24 hours Remarks Referenceined (mg)__

Free |Bound

Addicts stabilized Rate not influenced by age, Fry, Light,on morphine 5 972 SC 9(7-10) weight, urine volume, length or Torrance & Wolff

degree of addiction. (1929)

Addicts 3 20-120 SC 14(2-24) Considerable variation foundin same individual from day to Deckert (1936a)day.

Opium addicts 11 40-300 Oral 27(23-31)Opium addicts 10 10-100 Oral 29(18-37) Method apparently not reliable To & Ri (1938)-values too high.Morphine addicts 3 21-50 SC 57(45-72)

Addicts stabilized 12 30-90 SC 7(3-12) 18(11-20)on morphine Percentage excreted after oral

22 101-371 SC 6(4-12) 22(14-30) doses about half that after SC Oberst (1940)doses

28 524-3317 SC 3(1-9) 32(27-45)

Addicts stabilized 26 120-600 SC 6 33 Collection time not given. Oberst (1941)on morphine

Addicts stabilized 13 8-2072 SC 6 30 Data on 5 non-tolerant subjects Oberst (1942)on morphine included.

Patient 1 25-30 SC 10 30 Excretion also studied in opium Oettel (1950)smokers.

Addicts stabilized 6 260 SC 4(2-6) 20(16-37) Fujimoto & Wayon morphine (1957)

Volunteers 6 5 SC 1 38(37-42)Two-thirds of total excretion Paerregaard

Psychiatric patients 6 10 SC 8(4-10) 36(28-42) occurs within 8 hours. Traces (1957b)still detectable after 40 hours.

Surgical patients 3 20 IV 9 39

a SC = subcutaneous; IV = intravenous.

Following its formation, the excretion of bound mor-phine appears to be rapid. Paerregaard (1957b)reported that over 50%. of the total amount to beexcreted in the urine appears within 8 hours, and by24 hours the value is roughly 90%. Traces of boundmorphine are still detectable, however, after 48 hours.There is considerable variation in the amount ofbound morphine excreted by different individualsand Paaeregaard's results indicate that the sameindividual on a constant dosage regime of morphinevaries considerably from day to day in the amountof bound morphine he excretes.

Normorphine and its conjugate have not beenreported to be present in urine after morphineadministration, and, according to Rapoport (per-sonal communication), if they are present the total ex-cretion would be less than 0.5% of the injected dose.Monkey. The excretion of morphine in the urine

by monkeys (Macaca mulatta) was reported byMellett & Woods (1956). Excretion of free mor-phine is rapid, with detectable amounts appearingin bladder urine within 30 minutes after subcuta-neous administration of a single 30 mg/kg dose ofmorphine. Roughly 10% of the dose was accounted

253


for in the 24-hour urine of 4 non-tolerant animalsreceiving 15 mg/kg morphine twice daily, but therewas considerable variation in the amounts excretedfrom day to day even by the same animal, thestandard deviation being 15%. Tolerant animalsexcreted an average of 7% ± 3% of the morphinedose as the free alkaloid. Conjugated morphine,presumably the glucuronide, was also rapidly ex-creted, appearing in bladder urine within 30 minutes.The 24-hour excretion averaged 67% ± 25%. Intolerant animals the 24-hour excretion of boundmorphine averaged 59% ± 11 %. The total mor-phine recovered in the 24-hour urine exceeds thatfound in dogs, suggesting a more rapid rate ofelimination of morphine by monkeys.

Dog. In the dog the major fraction of the mor-phine dose can be accounted for within 24 hours inthe urine. About 15% appears as free morphine andgenerally over 50% is excreted as a conjugated pro-duct of glucuronic acid.The urinary excretion of free morphine begins

promptly. The compound appears in bladder urinewithin 15-30 minutes after parenteral administra-tion (Cochin, Haggart, Woods & Seevers, 1954).After subcutaneous injection over 80% of the totalamount to be excreted in the urine can be accountedfor within 8 hours (Wolff, Riegel & Fry, 1933;Paerregaard & Poulsen, 1958). The rate of elimina-tion of free morphine is greatest within the first twohours, when blood levels are elevated (Pierce &Plant, 1932). The renal clearance of morphine wasfound to be about the same as that of inulin. On theassumption that protein binding of morphine by theplasma was 25 %, it was concluded that morphine iscleared primarily by glomerular filtration and only toa very minor extent by tubular secretion (Baker &Woods, 1957). Traces of the morphine continue tobe excreted for several days (Pierce & Plant, 1932).This residual morphine probably has its origin in themorphine stored in the gall-bladder (Woods, 1957).The delayed appearance of morphine in the urinemay be due to re-absorption of free morphine eithersecreted into the gastro-intestinal tract via the bileor formed by hydrolysis of biliary bound morphinein the gut.Approximately 15% of the dose of morphine is

excreted as free morphine in the urine within24 hours. The data reported by various investigatorson urinary free morphine after subcutaneous ad-ministration indicate general agreement, virtuallyall values falling within the range 9-20% (Pierce &Plant, 1932; Wolff, Riegel & Fry, 1933; Cochin,

Haggart, Woods & Seevers, 1954; Paerregaard &Poulsen, 1958). This range also defines the variationwhich an individual animal may exhibit from day today (Pierce & Plant, 1932).

Various factors which may influence the excretionof free morphine have been studied. The dosage ofmorphine does not appear to be an important-factorin affecting the proportion excreted as free morphine,since animals receiving a subcutaneous dose ofmorphine as low as 0.2 mg/kg (Paerregaard &Poulsen, 1958) or as high as 200 mg/kg (Fry, Light,Torrance & Wolff, 1929) excrete free morphinewithin the percentage range given. A questionableslight increase in the total amount of free morphineexcreted in the urine may be induced by increasingthe daily water intake of dogs by 200 ml (Pierce &Plant, 1932). The excretion of free and boundmorphine is not greatly influenced by neostigminemethylsulfate, 0.1 mg/kg (Slaughter, Treadwell &Gales, 1941). Free urinary morphine excretion isincreased appreciably by liver damage produced withcarbon tetrachloride (Gross, 1942) or chloroform(Gross, Plant & Thompson, 1938).

Several studies comparing the excretion of freemorphine in tolerant and non-tolerant dogs havebeen carried out. Despite varying conditions for theproduction of tolerance, there is general accord thatthe excretion of free morphine is not altered by thedevelopment of tolerance to morphine (Pierce &Plant, 1932; Wolff, Riegel & Fry, 1933; Gross &Thompson, 1940; Woods, 1954).The increase in cellular metabolism that followed

administration of small doses of dinitrophenol orthyroid feeding decreased the excretion of freemorphine in non-tolerant dogs but not in tolerantanimals (Plant & Slaughter, 1936, 1938). It wasconcluded that there is a marked difference betweentolerant and non-tolerant dogs in the manner inwhich morphine is handled in the tissue. Thesestudies were conducted prior to the establishment ofbound morphine as a major morphine metabolite.Since free morphine excretion represents but a minorfraction of the morphine dose and the consequencesof dinitrophenol or thyroid feeding are complicated,it is difficult to evaluate the significance of thesefindings.The major urinary product of morphine is mor-

phine-3-monoglucuronide. Slightly more than one-half of the morphine dosage can be accounted forwithin 24 hours as morphine monoglucuronide,most values reported for the morphine conjugatefalling between 38% and 75 %. The fluctuation of

254

BIOLOGICAL DISPOSITlON OF MORPHINE AND ITS SURROGATES. 1

values is due to variation among dogs, daily in-dividual variations, dosage of morphine adminis-tered, and perhaps to conditions which may not havebeen optimal for complete hydrolysis of boundmorphine. With regard to variations due to dosage,it was shown that the proportion of conjugatedmorphine appearing in the urine of dogs after a doseof 2 mg/kg morphine was approximately 36% of thedose, whereas after a dose of 10 mg/kg the propor-tion of conjugated morphine in the urine of the sameanimals was increased to 60% of the dose. Duringthis same interval, the percentage excretion of freemorphine did not change (Paerregaard & Poulsen,1958).The formation and excretion of bound morphine

in the dog is rapid. Its appearance in bladder urinewas detected within 15-30 minutes after parenteraladministration (Cochin, Haggart, Woods & Seevers,1954). Approximately one-fifth of the morphinedosage after a subcutaneous injection of 20 mg/kgcan be accounted for in the urine as conjugatedproduct(s) at the end of two hours (Thompson& Gross, 1941). It can easily be calculated from thedata reported by Paerregaard & Poulsen (1958) forthe total and free morphine excreted over an 8-hourperiod that between 57% and 77% of the totalamount of conjugated morphine to be excretedappears within this interval. The clearance of boundmorphine was reported to be the same as that offree morphine, the values given being 66 ml/min.(Baker & Woods, 1957) and 80 ml/min. (Cochin,Haggart, Woods & Seevers, 1954). Since thesevalues were found to be similar to that for inulin,and since binding of morphine monoglucuronideby plasma was considered to be less than a few percent. (Blaney & Woods, 1956), it was concludedthat conjugated morphine is cleared essentiallyby glomerular filtration (Baker & Woods, 1957).Despite the initial rapid rate of excretion of conju-gated morphine, its presence in the urine may bedetected for several days, the prolonged excretion ofconjugated morphine persisting to a greater degreethe higher the dose of morphine (Paerregaard& Poulsen, 1958). The bound morphine whichappears in the urine of dogs later than 24 hoursafter injection stems from the compound stored inthe gall-bladder (Woods, 1954).

Studies comparing the excretion of bound mor-phine in tolerant and non-tolerant dogs are con-flicting. Gross & Thompson (1940) reported thatthe total (free and bound) morphine excreted in theurine of animals made tolerant by daily subcutaneous

injection of 20 mg/kg morphine for one or moreyears was 35-66%, whereas, in non-tolerant dogs,the total morphine excretion was 80-92%. Sincethe free morphine excretion in both groups wasapproximately the same (between 10% and 20 %), thedifference between the two groups was attributedto a decrease in bound morphine excretion by thetolerant animals. On the other hand, Cochin,Haggart, Woods & Seevers (1954) reported that theurinary excretion of bound morphine in dogs madetolerant to 30 mg/kg morphine by daily hypodermicinjections of gradually increasing doses of morphineover a period of 2-3 months was the same as that fornon-tolerant animals; the mean value for bothgroups was found to be between 50% and 60% ofthe injected dose. Woods (1954), on the basis of hisfindings that biliary excretion of morphine is animportant pathway for the disposal of morphine,offered the explanation that the reduced urinaryexcretion of bound morphine found in tolerant dogsby Gross & Thompson (1940) was most likelyparalleled by an increase in faecal excretion. Thiswould mean that between 34% and 65% of the doseought to be found in the faeces. Although such anexplanation is not without plausibility, the questionarises why there should be a greatly increased faecalexcretion in this group of tolerant dogs over thatfound by Woods in his tolerant animals. Woodssuggests that the difference may be due to a differencein the amount of exercise allowed the two groups ofdogs from the two laboratories.

It seems to the reviewers that more importantdifferences have been overlooked and that the twogroups of tolerant dogs are not strictly comparablesince one group had been treated with morphinefor one or more years while the other had beentreated for only 2-3 months prior to the study. Itmay well be that, in contrast to the tolerance pro-duced to the central effects, the long-term administra-tion of morphine has progressively disruptive effectson the mechanisms responsible for re-absorptionof bound morphine from the gut. This would be ofconsiderable importance in the dog and the rat, inwhich there is extensive biliary excretion of morphinemetabolites. Another possibility with the long-termadministration of morphine is that there is a depres-sion of the processes involved in the conjugation ofmorphine, with the resultant shunting of morphineto a metabolic pathway which is still unknown.These hypotheses, as well as the one implicatingexercise as an important factor in the dispositionof morphine in tolerant dogs, cannot be other than

255


tentative in the face of the inadequate data availableat present.

In a second report, Thompson & Gross (1941)carried out kinetic studies on the hydrolysis ofcombined morphine excreted by tolerant and non-tolerant dogs. On the basis of these studies theyfurther classified the combined morphine into twofractions-(a) an " easily " hydrolysable fraction,excreted in larger amounts in the non-tolerant dog,and (b) a "difficultly" hydrolysable fraction, excretedin greater quantities in the tolerant dog. We havealready pointed out in the section on metabolismour objection to interpretations based on kineticevidence obtained on systems as complicated asurine and we have also cited arguments for theexistence of only one form of bound morphine.If another bound product is excreted in urine, theamount present would be considerably less than the8-16% cited by Thompson & Gross for the form ofbound morphine which was excreted in the lesserquantity-namely, the " easily " hydrolysable frac-tion. Moreover, assuming the validity of theirexperiments, the difference between tolerant andnon-tolerant animals in respect of the excretion ofthe " easily " hydrolysable fraction does not appearto be at a level of statistical significance.

Rabbit. There have been numerous studies madeon this species, but most of the reports antedate thedevelopment of reliable methods for the estimationof morphine and the establishment of morphineglucuronide as a major biotransformation productof morphine in the urine. More recent work indi-cates that the urinary excretion pattern ofmorphine inrabbits is more or less similar to that in other species.

Keeser, Oelkers & Raetz (1933), after givingdoses ranging from 77 to 186 mg/kg, recoveredbetween 3% and 12% of the dose as morphine(free) in the 48-hour urine. Similarly, Yoshikawa(1940) found that, after subcutaneous administrationof 0.05-0.2 mg/kg morphine hydrochloride, 6-10% ofthe dose was excreted within one or two days in theurine and faeces, " but far more in urine than infeces". He compared the excretion of morphinein the cat and the hen under similar conditions andconcluded that natural tolerance cannot be satis-factorily explained on a basis of different capacityto destroy morphine. To the reviewers, the excretiondata alone appear to be insufficient to support or

refute this hypothesis.Hosoya (1959) and Otobe (1960) found consider-

able free and bound morphine appearing in the urineof a rabbit within 30 minutes after intravenous

administration. Pretreatment of rabbits with sodiumglucuronate or glucuronlactone before morphineinjection resulted in the earlier appearance and inlarger amounts of bound morphine in the urine;free morphine excretion was increased sometimesby the same treatment.

Rat. Free morphine excretion in the urine of therat appears to be higher than in most species. Thevalues cited by various authors, using different dosesof morphine and strains of rats, indicate a 24-hoururinary excretion of between 10% and 30% of thedose of morphine injected, with an average close to20%. Bound morphine excretion is more variableand appears to be related in part to strain differences.The amount excreted within 24 hours usually fallsbetween 20% and 50% of the dose injected.

Free and bound morphine appear in the urineshortly after parenteral administration of morphine.Within 30 minutes after subcutaneous injection of8 mg/kg of labelled morphine in a Long-Evans rat,both free and bound morphine were detected in theurine (Adler, Elliott & George, 1957). In Sprague-Dawley or Long-Evans rats receiving 2 mg/kg oflabelled morphine subcutaneously or via the popli-teal space, some 14% of the dose appeared as freemorphine in the urine at 1 hour in both strains;at the same time, the bound morphine in the urinewas 18-40% of the dose in the Sprague-Dawleyrats and 10-12% of the dose in the Long-Evans rats(Adler, Elliott & George, 1957). In rats given5 mg/kg of labelled morphine subcutaneously, 18%of the dose appeared as 14C in the urine at 1 hour,64% at 6 hours and 52% at 24 hours (March &Elliott, 1954).

Zauder (1952) studied the urinary excretion ofmorphine over a 9-week period in 12 Wistarrats. For the first 6 days the animals were injectedeach day with 12 mg/kg. During this period the24-hour excretion of free morphine averaged 28 %,and the bound morphine 36%. As the dose ofmorphine was increased by one-half over each 10-dayperiod and the animals developed tolerance, theaverage excretion offree morphine for the subsequent8 weeks remained fairly constant at about 19% ofthe dose administered. In contrast to this the per-centage of conjugated morphine in urine at firstincreased from 36% to 60% and remained elevatedfor 3 weeks, but with each succeeding week there-after it gradually decreased until after 9 weeks thebound morphine excretion was 38 %. In terms ofabsolute amounts there was a gradual increase in theamount of conjugated morphine excreted, although

256


this increase was not commensurate with the increasein morphine dosage.The excretion of free and conjugated morphine

was studied in 12 Long-Evans rats receiving 30or 45 mg/kg subcutaneously in two divided doses(Way, Sung & Fujimoto, 1954). The 24-hour urineshowed wide variations in the amount of free andconjugated morphine excreted. The excretion of freemorphine ranged from 8% to 29% on the first day,with a mean of 18 %, and the excretion of conju-gated morphine ranged from 3% to 31 %, with amean of 14 %. Continued administration of the samedose of morphine for 3 consecutive days resulted inan increase in free morphine and some decrease inconjugated morphine excretion.

In rats receiving 200 mg/kg morphine hydro-chloride subcutaneously, the average urinary excre-tion after 72 hours was found to be 48% ofthe morphine injected; 12% of the dose was freemorphine and 36% bound morphine. In ratsreceiving three successive subcutaneous injectionsof 100 mg/kg morphine hydrochloride daily and200 mg/kg on the fourth day, the 72-hour urinaryexcretion was 18% in the free form and 35% inthe bound form. Over a period of from 3 to 12weeks, the rats were gradually made tolerant withdoses of morphine hydrochloride increasing from100 to 500 mg/kg. The average excretion of freemorphine did not change materially with this treat-ment, fluctuating between 10% and 20% of the doseadministered. The bound morphine excretion didnot vary much for 3 weeks, remaining at levelsbetween 35% and 40%, but after 7 weeks of dailyinjection of morphine hydrochloride, the percentageexcretion of bound morphine was reduced to 4-16%of the dose (Fichtenberg, 1951).Mouse. Fairly rapid urinary excretion of 14C

occurred in the mouse after injection of 3.5 mg/kgbiosynthetically labelled morphine. Between one-third and three-quarters of the dose was found in theurine within 6 hours after injection, most of thisexcretion occurring from the fourth to the sixthhour. Intraperitoneal injection of 25 mg/kg nalor-phine following the morphine injection resulted ina somewhat enhanced urinary excretion during thefirst 6 hours, the greater part of the radioactivitybeing excreted during the second hour. These timedifferences observed are very probably related to thenalorphine antagonism of the morphine depressionof micturition (Achor & Geiling, 1953).

Other species. Yoshikawa (1940) reported thatthe cat excreted 11-27% of the dose in the urine as

free morphine within 1-2 days and the hen 10-13%of the dose.

Gastro-intestinalThe gastro-intestinal tract has been generally

considered to be a minor route for elimination ofmorphine, because only small amounts have beenfound in faeces and because the major fraction ofthe morphine dose is recovered from the urine. Onlyrecently has the importance of the biliary pathwayfor the disposal of morphine been established. Itnow appears that a significant fraction of morphinecan be accounted for in the bile, primarily as conju-gated morphine, and that biliary excretion is chieflyresponsible for the morphine appearing in the faeces.

Man. Surprisingly few studies have been carriedout on the gastro-intestinal excretion of morphinein man. The recovery of free morphine from thefaeces of addicts on a total daily dose ranging from972 to 3888 mg of morphine was less than 5%(Fry, Light, Torrance & Wolff, 1929). The studyof Oberst (1942) is of interest. In 9 samples exa-mined from an addict receiving 283-489 mg daily, hefound that the faecal excretion of free morphinedid not exceed 1 %. In another addict, the amountof free morphine appearing in faecal material wasless than 3% after a daily dose of 2455 mg. In bothindividuals the amount of conjugated morphineexcreted was even lower, the concentration beingbelow the sensitivity of the method. No free mor-phine and only a low concentration (0.7 jg/ml) ofconjugated morphine was detected in the bilecollected in a test-tube in the common duct of anon-tolerant patient who had had a cholecystectomyand who had been receiving 15 mg of morphinesulfate every 4 hours for 3 days. In four specimenswhich were examined, the gastric contents showedfree morphine concentrations ranging from 0.2 to0.7 ,ug/ml and bound morphine concentrationsranging from 0 to 1.3 ,ug/ml. No morphine wasfound in the vomitus of two morphine addicts on thesecond day of withdrawal. Unfortunately, noattempts were made to measure bound morphinein this study.

After intramuscular administration of 10-15 mgof labelled morphine sulfate in 5 human subjects,the faecal excretion of radioactivity over 72 hoursaccounted for 7-10 %. of the dose. In one of the sub-jects a duodenal tube was passed, and 14C wasmeasured at various time intervals. The radio-activity in the duodenal secretions was highestbetween 1l/2 and 6 hours after administration of the

257


dose. On the assumption that the radioactivity inthe duodenal samples originated in the bile and thatthe 24-hour bile output was 500 ml, it was estimatedthat 7.4% of the injected dose was potentiallyavailable for excretion by the intestinal route. Sincethis figure approximated that obtained in thefaecal excretion studies, it was concluded that thetotal faecal output arises from biliary excretion. Itwas also concluded that relatively little re-absorptionof biliary morphine occurs, since 24-hour retentionof faeces by one subject did not result in a de-creased amount of radioactivity excreted in thefaeces as compared with other subjects (Elliott,Tolbert, Adler & Anderson, 1954). This single studyis not consistent with the findings on other speciesdiscussed below. While biliary excretion may notbe an important route for the disposal of morphinein man, it should be pointed out that the duodenalsamples may not reflect accurately the total amountof morphine capable of being excreted via the bile.

Monkey. The bile of non-tolerant and tolerantmonkeys receiving 30 mg/kg morphine subcutane-ously was found to contain significant amounts ofmorphine; over 99% of the amount present in bothgroups was in the conjugated form. The totalmorphine excretion in the bile of a non-tolerantmonkey at 90 minutes was equivalent to 19.7% ofthe dose. In another non-tolerant monkey the totalbiliary morphine at 4 hours was equivalent to 11.5 %of the dose. A bound morphine concentration of300 ,ig/ml was noted in the bile of an animal sacri-ficed at 24 hours. The percentage excretion of onetolerant monkey at 90 minutes was 5.7 % and ofanother tolerant monkey at 4 hours was 5.6% of thedose. Apparently re-absorption and partial hydro-lysis of biliary bound morphine occurs, since onlysmall quantities of morphine appear in the faeces,chiefly in the free form (Mellett & Woods, 1956).Four non-tolerant monkeys receiving hypodermic

injections of 15 mg/kg morphine twice daily excretedonly 2% ± 1.5% free morphine in the faeces;5 tolerant monkeys excreted 4% ± 3 %. Theaverage concentration of bound morphine excretedin tolerant and non-tolerant animals was found to beless than 1% (Mellett & Woods, 1956).

Dog. In the dog the faecal excretion of freemorphine is considerably less than the urinaryexcretion. In studies on 11 tolerant and 7 non-tolerant dogs on daily doses of 10-50 mg/kg mor-phine sulfate, the average amount of morphineexcreted daily in the faeces over a period of 6-14 days

for both groups was between 2% and 3%. Therewas less variation in the daily excretion of an indivi-dual animal than between animals, but in neithercase did the amount of free morphine appearing inthe faeces on any given day exceed 10% of theadministered dose (Plant & Pierce, 1932). Wolff,Riegel & Fry (1933) reported no significant differencebetween tolerant and non-tolerant animals in respectof the amount of free morphine appearing in thefaeces. They gave figures averaging between 5 % and6 %, but felt that their values were on the high sidebecause their analytical procedure for estimatingmorphine in faeces was deemed to be inaccurate.Cochin, Haggart, Woods & Seevers (1954) foundabout 6% free morphine and 1 % bound morphinein 4 non-tolerant dogs receiving 30 mg/kg sub-cutaneously. Significantly higher amounts of freemorphine but not of bound morphine were observedin 4 tolerant dogs (stabilized on 30 mg/kg mor-phine for 60-90 days) than in non-tolerant animals,the average free morphine excretion for the tolerantanimals being 14%. They found a greater variationin the latter group and felt that the low faecalmorphine excretion in one of the 4 animals in thisseries was attributable to decreased faecal bulkresulting from a meat diet. The faecal excretion ofconjugated morphine in tolerant and non-tolerantanimals did not exceed 2% of the dose.

Studies on the biliary excretion of morphine in10 non-tolerant dogs at various time intervals afteradministration of the drug indicated that a largefraction of the morphine dose (30 mg/kg subcuta-neously) was excreted in the gall-bladder bile within24 hours, almost entirely as conjugated morphine(Woods, 1954). In general, the conjugated morphineaccounted for more than 99% of the total morphineexcreted in the bile, although low concentrations offree morphine (11-65 ug/ml) were detectable at alltime intervals studied. At 90 minutes the total mor-phine recovered was in excess of 10% of the dose.At various time intervals between 4 and 24 hours,approximately one-quarter to one-third of the dosewas recovered in the bile. The amount declined toabout 15% at 48 hours and to less than 1% at72 hours. Three tolerant dogs similarly excretedlarge amounts of conjugated morphine in the bile.Indeed, at 12 hours the gall-bladder of the tolerantanimals was found to contain 48% of the injecteddose and practically all of this was present in theconjugated form (Woods, 1954).

Evidence was also furnished that the morphinepresent in faeces arises almost exclusively from the

258


bile. In an animal with a ligated cystic duct and acommon bile duct fistula, no free or conjugatedmorphine was detectable in the faeces up to 96 hoursand yet 98% of the dose was recovered within24 hours in the excreta as follows: urinary freemorphine, 20 %; urinary conjugated morphine, 43 %;biliary free morphine, 1 %; and biliary conjugatedmorphine, 34% (Woods, 1954).

Rat. Morphine elimination via the gastro-intes-tinal tract is more important in the rat than in mostspecies. Although the amount of morphine appear-ing in the faeces is considerably less than that ap-pearing in the urine, a rather large fraction of themorphine dose appears initially in the bile, chiefly inthe bound form. However, only a fraction of this isexcreted in the faeces owing, undoubtedly, to re-absorption of the compound. Generally, less than2% of the injected dose is found in the faeces as freemorphine and less than 20% as bound morphine.

In rats given 200 mg/kg morphine hydrochloridesubcutaneously, about 1-2% of the dose can berecovered in the faeces within 72 hours as freemorphine and about 11% as bound morphine. Inanimals made tolerant to increasing doses of mor-phine administered over a period of 12 weeks, thefree morphine faecal excretion remained fairlyconstant and represented generally less than 5%of the injected morphine. Bound morphine excre-tion was more variable but after an initial increase,as with the urinary excretion, the bound morphinelevel fell from a peak of 28% at the end of 3 weeksto 8 % after 7 weeks (Fichtenberg, 1951).March & Elliott (1954), using morphine-N-14CH3,

found 25% of the radioactivity in the faeces within24-48 hours after subcutaneous injection of 5 mg/kg.In one rat they recovered 63% of the dose in bilecollected during the first 6 hours after injection, ascompared with 18% in the urine during this period.The material secreted in the bile was stated to be freeand bound morphine, with the latter being the major

fraction. Since the intact animal ex9retes a muchgreater percentage of the dose in the urine over thesame time interval, it was suggested that there isevidence for an entero-hepatic circulation of mor-phinie or its metabolites following initial excretion byliver in the bile.

Other routesWhile renal and alimentary excretion constitute

the major pathways for the elimination of morphine,other body fluids, such as saliva, tears, sweat andmilk, may be expected to carry at least traces of thedrug. That the presence of morphine has not alwaysbeen detected in these fluids by chemical tests in-dicates that the amount of morphine excreted bysuch routes is very low indeed. There has beenvery little information to supplement that found inthe review by Krueger, Eddy & Sumwalt (1941).Unequivocal chemical evidence of the presence ofmorphine in milk has still not been furnished.

In contrast to claims made in two reports cited byKrueger, Eddy & Sumwalt (1941), Oberst (1942) wasunable to detect free or bound morphine in the salivaof addicts on a daily dosage of between 105 and4200 mg of morphine, using a technique sensitive to0.2 ,ug/ml. However, according to Peterson (1955),the presence of morphine in the saliva and urine ofhorses after parenteral administration has beennoted in several laboratories.

Traces of free morphine (0.2-5 ,ug/ml) were foundin the perspiration of addicts after a daily 300-mgdose of morphine (Oberst, 1942).

Other animalsComparatively little work has been devoted to

studies on the excretion of morphine in animalspecies other than those already mentioned. Krueger,Eddy & Sumwalt (1941) cite earlier studies whichhave been reported on the cat, guinea-pig, goat, cow,chicken, dove, frog and toad. We have not encoun-tered any recent studies on these species.

REFERENCES

Achor, L. B. & Geiling, E. M. K. (1953) Proc. Soc. exp.Biol. (N. Y.), 84, 688

Achor, L. B. & Geiling, E. M. K. (1954) Analyt. Chem.,26, 1061

Achor, L. B. & Geiling, E. M. K. (1956) J. Pharmacol.,117, 16

Adler, T. K., Elliott, H. W. & George, R. (1957)J. Pharmacol., 120, 475

Adler, T. K., Fujimoto, J. M., Way, E. Leong & Baker,E. M. (1955) J. Pharmacol., 114, 251

Andersen, K. S. & Woods, L. A. (1959) J. org. Chem.,24, 274

Ashdown, H. (1890) Brit. med. J., 1, 169Axelrod, J. (1955a) J. PharmacoL, 114, 430Axelrod, J. (1955b) J. Pharmacol., 115, 259Axelrod, J. (1956a) J. Pharmacol., 117, 322


Axelrod, J. (1956b) Science, 124, 263Axelrod, J. & Cochin, J. (1957) J. Pharmacol., 121, 107Baker, W. P. & Woods, L. A. (1957) J. Pharmacol., 120,

371Balls, A. K. & Wolff, W. A. (1928) J. biol. Chem., 80, 379Beckett, A. H. & Casy, A. F. (1954) J. Pharm. (Lond.),

6, 986Beckett, A. H., Casy, A. F. & Harper, N. J. (1956)

J. Pharm. (Lond.), 8, 874Beckett, A. H., Casy, A. F., Harper, N. J. & Phillips,

P. M. (1956) J. Pharm. (Lond.), 8, 860Beecher, H. K., Keats, A. S., Mosteller, F. & Lasagna, L.

(1953) J. Pharmacol., 109, 393Bell, J. L. (1958) J. Neurochem., 2, 2, 265Bentley, K. W. (1954) The chemistry of the morphine

alkaloids, Oxford, Clarendon Press, p. ixBentley, K. W. & Dyke, S. F. (1959) J. chem. Soc., 3, 2574Bernheim, F. & Bemheim, M. L. C. (1944) J. Pharmacol.,

81, 374Bemheim, F. & Bernheim, M. L. C. (1945) J. Pharmacol.,

83, 85Biggs, A. I. (1952) J. Pharm. (Lond.), 4, 547Blaney, D. J., Bloom, V. & Woods, L. A. (1955)

J. Pharmacol., 113, 5Blaney, D. J. & Woods, L. A. (1956) J. Pharmacol., 116, 7Bockmuhl, M. & Ehrhart, G. (1949) Justus Liebigs Ann.

Chem., 561, 52Bradford, L. W. & Brackett, J. W. (1958) Mikrochem.

Acta, 3, 353Braenden, 0. J., Eddy, N. B. & Halbach, H. (1955) Bull.

WldHlth Org., 13, 937Breinlich, J. (1953) Arzneimittel-Forsch., 3, 212Brochmann-Hanssen, E. (1955) Medd. norsk farm. Selsk.,

17, 76Chemov, H. I., Miller, J. W. & Mannering, G. J. (1959)

Fed. Proc., 18, 376Chopra, R. N., Chopra, G. S. & Roy, A. C. (1941)

Indian J. med. Res., 29, 195Clarke, E. G. C. (1959) Bull. Narcot., 11, No. 1, p. 27Cochin, J. & Axelrod, J. (1959) J. Pharmacol., 125, 105Cochin, J. & Economon, S. (1959) Fed. Proc., 18, 377Cochin, J., Haggart, J., Woods, L. A. & Seevers, M. H.

(1954) J. Pharmacol., 111, 74Cochin, J. & Sokoloff, L. (1960) Fed. Proc., 19, 270Comroe, J. H., jr & Dripps, R. D. (1948) Surg. Gynec.

Obstet., 87, 221Conney, A. H., Brown, R. R., Miller, J. A. & Miller, E. C.

(1957) Cancer Res., 17, 628Curry, A. S. & Powell, H. (1954) Nature (Lond.), 173,

1143Deckert, W. (1936a) Arch. exp. Path. Pharmak., 180, 656Deckert, W. (1936b) Klin. Wschr., 15, 714Deneau, G. A., Woods, L. A. & Seevers, M. H. (1953)

Fed. Proc., 12, 315Dutton, G. J. & Greig, C. G. (1957) Biochem. J., 66, 52PDutton, G. J. & Storey, I. D. E. (1954) Biochem. J.,

57, 275Eddy, N. B. (1950) J. Amer. pharm. Ass., sci. Ed., 39, 245

Eisenman, A. J. (1945) Fed. Proc., 4, 88Elison, C. (1961) Some aspects of the fate and relationshipofthe N-methylgroup ofmorphine to itspharmacologicalactivity (Dissertation, University of California)

Elliott, H. W., Tolbert, B. M., Adler, T. K. & Anderson,H. H. (1954) Proc. Soc. exp. Biol. (N. Y.), 85, 77

Elvidge, W. F. (1940) Quart. J. Pharm., 13, 219Endo, M. (1938) Jap. J. med. Sci., 11, 140Endo, M. & Kato, T. (1937) Jap. J. med. Sci., 10, 192Endo, M. & Kato, T. (1938) Jap. J. med. Sci. 11, 136;Hokkaido Ig. Z., 16, 1130

Farmilo, C. G. & Levi, L. (1953) Bull. Narcot., 5, No. 4,p. 20

Farmilo, C. G., Oestreicher, P. M. & Levi, L. (1954)Bull. Narcot., 6, No. 1, p. 7

Fawaz, G. (1948) Proc. Soc. exp. Biol. (N.Y.), 68, 262Feldstein, M. & Klendshoj, N. C. (1956) J. forensic Sci.,

1, 47Fichtenberg, D. G. (1951) Bull. Narcot., 3, No. 3, p. 19;No.4,p. 16

Fischer, R. & Chalupa, L. (1949) Mikrochem. Acta, 3, 257Fischer, R. & Goll, H. (1950) Mikrochem. Acta, 1, 63Fish, M. S., Johnson, N. M., Lawrence, E. P. & Horning,

E. C. (1955) Biochim. biophys. Acta, 18, 564Fish, M. S., Sweeley, C. C., Johnson, N. M., Lawrence,

E. P. & Horning, E. C. (1956) Biochim. biophys. Acta,21, 196

Fleischmann, P. (1929) Biochem. Z., 208, 392Forst, A. W. & Deininger, R. (1949) Arch. exp. Path.Pharmak., 206, 416

Fraser, H. F., Wikler, A., Van Horn, G. D., Eisenman,A. J. & Isbell, H. (1958) J. Pharmacol., 112, 359

Fry, E. G., Light, A. B., Torrance, E. G. & Wolff, W. A.(1929) Arch. intern. Med., 44, 862

Fujimoto, J. M. & Way, E. Leong (1954) Fed. Proc., 13, 58Fujimoto, J. M. & Way, E. Leong (1957) J. Pharmacol.,

121, 340Fujimoto, J. M. & Way, E. Leong (1958) J. Amer. pharm.

Ass., sci. Ed., 47, 273Fujimoto, J. M., Way, E. Leong, & Hine, C. H. (1954)

J. Lab. clin. Med., 44, 627Gates, M. & Tschudi, G. (1952) J. Amer. chem. Soc.,

74, 1109Gillette, J. R., Brodie, B. B. & LaDu, B. N. (1957)

J. Pharmacol., 119, 532Gillette, J. R., Dingell, J. V. & Brodie, B. B. (1958)

Nature (Lond.), 181, 898Ginsburg, D. (1953) Bull. Narcot., 5, No. 4, p. 32Goldbaum, L. R., Gray, I., Rink, R. A., Rueckert, R. R.& Ostashever, A. S. (1956) Fed. Proc., 15, 429

Goldbaum, L. R. & Kazyak, L. (1952) J. Pharmacol., 106,388

Gross, E. G. (1942) Proc. Soc. exp. Biol. (N. Y.), 51, 61Gross, E. G., Plant, 0. H. & Thompson, V. (1938)

J. Pharmacol., 63, 13Gross, E. G. & Thompson, V. (1940) J. Pharmacol., 68,413Gross, S. T. & Oberst, F. W. (1947) J. Lab. clin. Med.,

32, 94


Guarino, S. (1946) Boll. Soc. ital. Biol. sper., 22, 1226Gulland, J. M. & Robinson, R. (1923) J. chem. Soc.,

123, 980Gulland, J. M. & Robinson, R. (1924-25) Proc. Man-

chester Lit. & Phil. Soc., 69, 79Hatcher, R. A. & Gold, H. (1929) J. Pharmacol., 35, 257Herken, H., Neubert, D. & Timmler, R. (1959) Arch.

exp. Path. Pharmak., 237, 319Himmelsbach, C. K., Oberst, F. W., Brown, R. R. &

Williams, E. G. (1942) J. Pharmacol., 76, 50Horlington, M. & Lockett M. F. (1959) J. Pharm.

(Lond.), 11, 415Hosoya, E. (1956) Fed. Proc., 15, 440Hosoya, E. (1959) Conjugation ofmorphine with glucuronic

acid. In: Proceedings of the 5th Conference, GlucuronicAcid Symposium of Europe, Zurich, p. 17

Hosoya, E. & Brody, T. M. (1954) J. Pharmacol., 110, 26Iglauer, H. (1949) Angew. Chem., 61, 412Ikeshima, S. (1935) Jikken Yakubute Z., 9, 331Inoue, T. (1940) Jap. J. med. Sci., 12, 75Inscoe, J. K. & Axelrod, J. (1960) J. Pharmacol., 129,

128Isbell, H. & Fraser, H. F. (1953) J. Pharmacol., 107, 524Isselbacher, K. J. (1956) Recent Progr. Hormone Res., 12,

134Janssen, P. A. J. & Jagenau, A. (1956) Experientia, 12, 293Jatzkewitz, H. (1954) Dtsch. med. Wschr., 79, 541Kabasawa, I. (1933-35) Jap. J. med. Sci., 8, 97Kaiser, H. & Jori, H. (1954) Arch. Pharm. (Berl.),

287, 224Kariyone, T. & Hashimoto, Y. (1957) J. pharm. Soc.

Japan., 71, 436Keeser, E., Oelkers, H. A. & Raetz, W. (1933) Arch. exp.

Path. Pharmak., 173, 622Kelentey, B., Stenszky, E., Czollner, F., SzlAvik, I.& Mszdro's, Z. (1957) Arzneimittel-Forsch., 7, 594

Ko, K. (1937) Jap. J. med. Sci., 10, 199Krueger, H. (1955) Narcotics and analgesics. In: Manske,

R. H. F., ed., The alkaloids, New York, AcademicPress, vol. 5

Krueger, H., Eddy, N. B. & Sumwalt, M. (1941) Thepharmacology of the opium alkaloids. Part 1, Washing-ton, D.C., US Government Printing Office (US PublicHealth Reports, Suppl. 165)

Krueger, H., Eddy, N. B. & Sumwalt, M. (1943) Thepharmacology of the opium alkaloids. Part 2, Washing-ton, D.C., US Government Printing Office (US PublicHealth Reports, Suppl. 165)

Kumler, W. D. (1955) J. org. Chem., 20, 700Kuwahara, H. (1937) Jap. J. med. Sci., 10, 196Lindsay, J. M. & Barnes, W. H. (1955) Acta cryst.

(Camb.), 8, 227Lockett, M. F. & Davis, M. M. (1958) J. Pharm.

(Lond.), 10, 80Mannering, G. J., Dixon, A. C., Carrol, N. & Cope, 0.

(1954) J. Lab. clin. Med., 44, 292Mannering, G. J., & Takemori, A. E. (1958) Fed. Proc.,

17, 391

March, C. H. & Elliott, H. W. (1954) Proc. Soc. exp.Biol. (N. Y.), 86, 494

Marks, A. F. & Huggins, S. F. (1959) Tex. Rep. Biol.Med., 17, 114

Marquis, E. (1896) Pharm. Z. f. Russland, 35, 549Mayer, P. (1899) Berl. klin. Wschr., 36, 591Mellett, L. B. & Woods, L. A. (1956) J. Pharmacol.,

116, 77Mellett, L. B. & Woods, L. A. (1961) Proc. Soc. exp.

Biol. (N.Y.), 106, 221Miller, J. W. & Elliott, H. W. (1955) J. Pharmacol.,

113, 283Mills, G. T., Lockhead, A. C. & Smith, E. E. B. (1958)

Biochim. biophys. Acta, 27, 103Milthers, K. (1958) Acta pharmacol. (Kbh.), 15, 21Milthers, K. (1959) Acta pharmacol. (Kbh.), 16, 144Milthers, K. (1960) Acta pharmacol. (Kbh.), 17, 200Milthers, K. & Schou, J. (1958) Acta pharmacol. (Kbh.),

15, 175Misra, A. L., Mule, S. J. & Woods, L. A. (1961) Nature

(Lond.), 190, 82Mueller, G. C. & Miller, J. A. (1953) J. biol. Chem.,

202, 579MuIe, S. J. & Woods, L. A. (1960) Fed. Proc., 19, 271Mull, J. (1937) Proc. Soc. exp. Biol. (N. Y.), 35, 551Nadeau, G. & Sobolewski, G. (1958) Canad. J. Biochem.,

36, 625Oberst, F. W. (1938-39) J. Lab. clin. Med., 24, 318Oberst, F. W. (1940) J. Pharmacol., 69, 240Oberst, F. W. (1941) J. Pharmacol., 73, 401Oberst, F. W. (1942) J. Pharmacol., 74, 37Oberst, F. W. & Gross, E. G. (1944) J. Pharmacol., 80, 188Oettel, H. (1950) Arch. exp. Path. Pharmak., 212, 163Otobe, S. (1960) Jap. J. Pharmacol., 9, 105Paerregaard, P. (1957a) Acta pharmacol. (Kbh.), 14, 38Paerregaard, P. (1957b) Acta pharmacol. (Kbh. ), 14, 53Paerregaard, P. & Poulsen, E. (1958) Acta pharmacol.

(Kbh.), 14, 390Patterson, A. M. & Capell, L. T. (1940) The Ring Index,New York, Reinhold Publishing Corp., No. 3094,p. 425

Penna, A., Arevalo, L., Fernandez, L., Navia, E.& Mardones, J. (1959) Normorphine derivatives foundin rats intoxicated with morphine. In: Abstracts ofcommunications; XXI International Congress ofPhysiological Sciences, Buenos Aires, p. 209

Perrine, T. D. & Eddy, N. B. (1956) J. org. Chem., 21, 125Peterson, H. (1955) Bull. Narcot., 7, No. 3-4, p. 23Pierce, I. H. & Plant, 0. H. (1932) J. Pharmacol., 46, 201Plant, 0. H. & Pierce, I. H. (1933) J. Pharmacol., 49, 432Plant, 0. H. & Slaughter, D. (1936) J. Pharmacol., 58, 417Plant, 0. H. & Slaughter, D. (1938) J. Pharmacol., 62, 106Plotnikoff, N. P., Elliott, H. W. & Way, E. Leong (1952)

J. Pharmacol., 104, 377Plotnikoff, N. P., Way, E. Leong & Elliott, H. W.

(1956) J. Pharmacol., 117, 414Rapoport, H. & Lavigne, J. B. (1953) J. Amer. chem.

Soc., 75, 5329


Rapoport, H., Lovell, C. H. & Tolbert, B. M. (1951)J. Amer. chem. Soc., 73, 5900

Rapoport, H. & Payne, G. B. (1950) J. org. Chem.,15, 1093

Rapoport, H. & Payne, G. B. (1952) J. Amer. chem. Soc.,74, 2630

Rasmussen, H. B. & Berger, J. (1955) Bull. Narcot., 7,No. 3-4, p. 30

Reynolds, A. K. & Randall, L. 0. (1957) Morphine andallied drugs, Toronto, University Press

Rink, R. A., Gray, I. & Rueckert, R. R. (1956)Anesthesiology, 17, 377

Robinson, D. & Williams, R. T. (1958) Biochem. J.,68, 23

Schaumann, 0. (1940) Arch. exp. Path. Pharmak.,196, 109

Schaumann, 0. (1957) Morphin und Morphin&hnlichwirkende Verbinduingen. In: Handbuch der experimen-tellen Pharmakologie, Berlin, Springer-Verlag, vol. 12

Schirm, M. (1940) Apothekerztg. (Berl.), 55, 106 (Quotedin Quart. J. Pharm., 1946, 19, 200)

Schmidt, C. F. & Livingston, A. E. (1933) J. Pharmacol.,47, 473

Schou, J. (1958) Acta pharmacol. (Kbh.), 15, 43Schultz, 0. E. & Strauss, D. (1955) Arzneimittel-Forsch.,

5, 342Schwartzman, L. H. (1950) J. org. Chem., 15, 517Seagers, W. J., Neuss, J. D. & Mader, W. J. (1952)

J. Amer. pharm. Ass., sci. Ed., 41, 640Seibert, R. A. & Huggins, R. A. (1953) Proc. Soc. exp.

Biol. (N. Y.), 82, 518Seibert, R. A., Williams, C. E. & Huggins, R. A. (1954)

Science, 120, 222Siminoff, R. & Saunders, P. R. (1958) J. Pharmacol.,

124, 252Slaughter, D., Treadwell, C. R. & Gales, J. W. (1941)

J. Pharmacol., 72, 37Sloan, J., Eisenman, A., Fraser, H. F. & Isbell, H.

(1958) Preliminary observations of the urinary excretionof normorphine and morphine in man. In: Abstracts ofFall Meeting of the American Society of Pharma-cologists, Ann Arbor, Mich., p. 31

Small, L. F. (1932) PubI. Hlth Rep. (Wash.), Suppl. 103Soehring, K. & Frahm, M. (1949) Klin. Wschr., 27, 513Spengler, G. A. (1958) Helv. med. Acta., 25, 430Stevenson, G. W. & Rapoport, H. (1955) Fed. Proc.,

14, 389

Stewart, C. P., Chatterji, S. K. & Smith, S. (1937)Brit. med. J., 2, 790

Stolman, A. & Stewart, C. P. (1949) Analyst, 74, 536Stolman, A. & Stewart, C. P. (1960) Drugs extractable by

organic solvents from aqueous alklaline solution, NewYork & London, Academic Press, vol. 1, chapter 5

Stolnikow (1884) Hoppe-Seylers Z. physiol. Chem., 8, 235Strominger, J. L., Kalckar, H. M., Axelrod, J. & Maxwell,

E. S. (1954) J. Amer. chem. Soc., 76, 6411Strominger, J. L. & Mapson, L. W. (1957) Biochem. J.,

66, 567Strominger, J. L., Maxwell, E. S., Axelrod, J. & Kalckar,H. M. (1957) J. biol. Chem., 224, 79

Szerb, J. C., MacLeod, D. P., Moya, F. & McCurdy,D. H. (1957) Arch. int. Pharmacodyn., 109, 99

Szerb, J. C. & McCurdy, D. H. (1956) J. Pharmacol.,118, 446

Takemori, A. E. (1960) J. Pharmacol., 130, 370Takemori, A. E. & Mannering, G. J. (1958) J. Pharmacol.,

123, 171Thompson, V. & Gross, E. G. (1941) J. Pharmacol.,

72, 138To, S. & Ri, C. (1938) Jap. J. med. Sci., 11, 47Tompsett, S. L. (1960) Acta pharmacol. (Kbh.), 17, 295Travell, J. (1932) J. Pharmacol., 44, 123Vidic, E. (1955) Arzneimittel-Forsch., 5, 291Vidic, E. (1957) Arzneimittel-Forsch., 7, 314Walton, R. P. & Lacey, C. F. (1935a) J. Pharmacol.,

54, 53Walton, R. P. & Lacey, C. F. (1935b) J. Pharmacol.,

54, 61Way, E. Leong & Adler, T. K. (1960) Pharmacol. Rev.,

12, 383Way, E. Leong, Kemp, J. W., Young, J. M. & Grassetti,D. R. (1960) J. Pharmacol., 129, 144

Way, E. Leong, Sung, C-Y. & Fujimoto, J. M. (1954)J. Pharmacol., 110, 51

Wolff, W. A., Riegel, C. & Fry, E. G. (1933) J. Phar-macol., 47, 391

Woods, L. A. (1954) J. Pharmacol., 112, 158Woods, L. A. (1957) J. Pharmacol., 120, 58Woods, L. A., Cochin, J., Fornefeld, E. G. & Seevers,M. H. (1954) J. Pharmacol., 111, 64

Woods, L. A., Daly, J., Haggart, J. & Seevers, M. H.(1952) J. Pharmacol., 106, 426

Yoshikawa, N. (1940) Jap. J. med. Sci., 12, 74Zauder, H. L. (1952) J. Pharmacol., 104, 11

biological disposition of morphine andits surrogates 1

Documents