LIVER ENLARGEMENT AND THE MODIFICATION OF HEPATIC MICROSOMAL DRUG METABOLISM BY PYRETHRUM

Liver enlargement in rats resulting from oral administration of pyrethrum has been observed in various laboratories. This study was undertaken to investigate the nature of this enlargem en t as well as resulting changes in hepatic drug metabolism. Oral administration of pyrethrum at 200 mg/kg for 13 or 23 days resulted in decreases in h~patic DNA concentrations~ While total lipid concentrations were increased significantly, tha increases did not account for the enlargement. Protein concentrations of whole liver homogenates, 9,000 x g supernatants and the 105,000 x g pellet were not different from controls, No significa nt c han ges occurred in hepatic water concentrations. Accompanying histologica l observations of the enlarged livers indicated cellular hypertrophy~ Significant decreases in hexobarbital induced hypnosis without conco~itant changes ir1 ba rbital induced hypnosis suggested a pyrethrum ca used alteration in hepatic drug metabolism. The activities of hepatic microsomal enzymes responsible for EPN detoxification, p-nitroanisole de~ethylation and hexobarbital oxidation were increased to 150, 173 and 241 percent of control, respectively. No significant increases were noted in the demethylation of aminopyrina or oxidation of L-tryptophan. Incre ases in liver weight, the detoxification of EPN and derneth ylation of p-nitroa n is~le were fou~d to be ~ose related.


REFERENCES ••• ••••••••••••••••••••• • ··~···•••••••
Eff e ct of pyrethrum, 500 mg/kg/day orally, on the deto xif ic ation of EPN by 9,000 x g super- Composite graph: Effect of pyrethrum, 500 mg/kg/day or ally , on the liver to body weight ratio, deto xi fic ation of EPN and demeth yl a tion of p-nitro anisole by 9,000 x g supernatants and NADPH-cytochrome c reducta se act ivity in In recent years increa3ing interest has developed in tha toxicity of the various classes of pesticides. The organochlorine insecti~ides, and DDT in particular, are of growing ecologic al concern becauss of their widespread use and extreme persistence and the difficulty in evaluating the harmful effects of th ese low residues upon such parameters as the fertility of animals and fragility of bird eggs (O'Brien, 1967). The unrestricted use of these halogenated hydrocarbons in libor atory animal quarters has also been of concern (Fouts, 1963). Since enzyme induction can seriously affect the results of tests meant to evaluate the metabolism or the acute or chronic toxicity of drugs, care should be taken to exclud3 such insecticides from quarters of animals used in such studies (Conney, 1967).
Another type~ of insecticide, the organophosphates, fer example parathion and malathion, are inhibitors of cholineste~ase.
Although they are rapidly hydrolyzed and thus are not persistent, the acute mammalian toxicity is high.
Pyrethrum, an extract of the pyrethrum flo wer Chrysant hemum cinararia efolium , contains four esters of complex alcohols and acids called pyrethrin I and II and cinerin I and II (S hepard , 1939).
As an insec t icid e , it has found wide use in agriculture, forestry, the home, and laboratory animal quarters because of its favorable properties. It demonstrates a low mammalian toxicity (Negherbon 1 1959), but at the same time has a high knockdown and kill rate on many classes of insects (Metcalf,195 5 ). Instability towards mild acid or alkali, and direct sunlight limits its persistence in the environment (Chen and Casid a , 1969). It has been generally assumed that pyrethrum 7 is safe for use .i n situations requiring a non-~nducing inssctici de, such as in quarters housing animals undergoing ph3rmacological t esting (Fouts, 1963;Conn ey , 1967). Although no data was presented , the use of pyrethrum and piperonyl butoxide was reco mmen ded since there was no detected effect of these materials on hexobarbital induced narcosis or microsomal drug metabolizing enzyme activity .
In examining the combin3d effe cts or DDT , pyrethr um , and pi pe ro ny l butoxide , Kimbrough et al (1968) found that -pyrethr um , especially synergized pyrethrum, produced liver enlargement and cellular changes including margination and cyto plasmic inclusions . These changes were proportional to dosa ge and similar in ch arac ter to changes caus Gd by DDT.
However, these studies were concerned only with morph ologic al chang es and did not include biochemical observations. Resu!ts from our laboratory of 90 day toxicit y studies cond u cted in rats e~ploying 25 percent of the oral LD50 also showed that pyrethrum, pipgronyl butoxide , tropital and sulfoxide, alone and in co mbination , caused liver enlargement ( Bond andDefeo, 1967-1969).
Liv er enlargement which may be indicative of abnormal structur al changes or other pathology often accompanies the stimulation or inhibition of drug biotransformation (Conney, 1967). Since the rate of hepatic drug biotransformation influences the resultant pharmacodynamic activity of a compound (Conney, 1967), it seemed important to examine the nature of the liver enlargement caused by pyrethrum, as well as possible Early work dealing with the toxicity of pyrethrum centered on its effects on mammalso The reported results varied with route of administration, species and vehicle used (Shimkin and Anderson, 1936;Ambro3e and Rob~ins, 1951;Negherbon, 1959;Bond andDeFeo, 1967-1969) •. The acute oral LD50 in rats for an 86~2 percent solution of pyrethrins was greater than 2.6 g/kg. No mortality was observed in rats administered 1.6 g/kg subcutaneously (Ambrose and Robbins, 1951). Negherbon (1959) reported an oral LD50 for rats of 820 mg/kg and 1500 mg/kg for guinea pigs. A dose of 6-8 mg/kg was toxic to do~s when administered intravenously, suggesting that pyrethrum possessed a potent intrinsic toxicity. The acute oral LD50 in rats for a commercial solution of pyrethrum (20 percent pyrethrins, 80 percent petroleum distillates) was 7200 mg/kg or the equivalent of 1440 mg/kg of pyrethrum alone (Bond andDeFeo, 1967-1969)0 The acute oral LOSO in rats for pyreth~in I dissolved in dimethylsulfo x ida was reported to be 260-420 mg/kg while that of pyrethrin II was greater than 600 Tig/kg.
Administered orally in 6 doses over 12-54 hour periods, rats survived 450-2000 mg/kg of pyrethrin I and greater than 2900 mg/kg of pyrethrin II (Cassida tl ~' 1971). Carpenter tl al . ( 1950) found tha-: rats were able to ingest over a 2L~ hour period a quantity which was lethal if administered in a single dose and were able to maintain the intake for long periods without apparent injury. Bond andDereo (1967-1969) noted that rats were able to tolerate a dcse of i' the oral LD50 for periods of 90 days without significant increases in mortality over that of controls.
The chronic dermal toxicity of pyrethrum (20 percent pyrethrum, 80 percent petroleum distillates) to rabbits and the chronic vapor toxicity to mice, rats, guinea pigs and rabbits, as well as the associated histopathology of these animals is reported by the Office of the Surgeon General (1951,1952). Slight to mild skin irritation was observed in rabbits upon application of 1 ml/kg/day for approximately 30 days, with further applications progressing in severity. However, these same effects were produced by the solvent which was tested alone, suggesting that the effects were not due to pyrethrum but to the solve~t. Continuation of exposure to saturated vapors of pyrethrum extract for 5 weeks, 8 hours daily, 5 days per week caused a 40 percent mortality in mice; preening, eye irritation and unthrifty appearance in rats; unthrifty appearance in guinea pigs; 3nd no apparent effects in rabbits.
On a comparable basis, vapors from the solvent appeared to be more toxic than the extract and accounted for the toxicity of extract_ The histopathology of the tissues from these animals showed primarily vascular congestion which was more pronounced ih the lungs and kidneys than other tissues. These effects 1 1 were more severe in solvent treated anim a ls.
Early studies by Feinberg (1934), Garratt and Bigger(1923), and I Y1 c C o rd tl ~ ( 1 9 2 1 ) s h ow e d t hat s k i n con ta ct a n d i n ha 1 a ti on may cause allerg3nic attacks in sensitive people. Patients sensitive to ragweed developed allergic reactions to extracts of ground pyret~rum flowers; severe dermatitis and anaphylactoid reactions were also reported, IYlore recent studies using refined pyrethrum suggested th a t there was no potential allergy to insecticides containing pyrethrins, as opposed to the unrefined material (Zucker, 1965).
Structurally related new synthetic derivatives of pyrethrum.
many reviewers of this subject conclude that the effective and rapid knockdown action of pyrethrum in insects results from an initial impairment of neuromotor systems (Shepard, 1951;Roeder, 1953;Metcalf, 1955;Negherbon, 1959;O'Brien, 1967 The results from isolated nreve preparations showed that spontaneously active ganglia , or the electrically stimulated peripheral nerve of the cr a yfish, wa s pa rtic u larly sensitive to the blocking effects of pyr e thr um, but ins en sitive tu lido-ca~ne.
In contrast: both the fro g ner ve a n d r a t nerve we re resistant to a cute block by pyrethrum but se ns itiv e to lidocaine.
These apparent differences could not be attributed to the vertebrate nerve sheath, as shown in de sheathe d nerve preparations, and suggested a possible basic difference in the mechanisms of action.
In whole animal studies, ~ow do ses of pyrethrum, 1-10 ug/kg, disrupted and blocked electrical activity of the heart in .crayfish Luhile doses c:.s high as. 1000 mg/kg failed to stop the frog heart.
The earliest work on the metabolism of pyrethrum in insects led to the sugg es tion and virtual assumption that hydrolysis of the ester group was a major metabolic route (Bridges, 1957;Chamberlain, 1950;Chang and Kearns, 1964;O'Brien, 1967; W i n t er i n g ham At ~ , 1 9 5 5 ; and Z e id ~t .51J:. , 1 9 5 3 ) • T he s e s tu d i e s had limited success because the radiolabeled compounds used had either low specific activity, were mixtures of isomers -.m were otherwise impure.
Sever a l c 14 -labeled pyrethroids, stereochemical ly pure isomers of high activity, were recently synthesized (Nishizwa and Casida, 1965;Yamamoto and Casida, 1968). With these com- The physiological significance of enhanced liver growth and function in animals treated with various drugs has not yet been determined and it is not clRar whet her these ~ffects are harmful or beneficial (Conney, 1967). While erlargement of the liver is generally regarded as a pathological phenomenon (Kunz, 1966), the concern lies with the increase in relative liver weight in the absence of histopatholoical effects indicative of liver dam a ge (Golber g , 1964).
Organ weight is related to body weight according to the principle of allrn , 1 etry, so th a t a log-log pl o t of body wE i ght against liver weight yields a str a ight lin e { Addis and Gray,

1950)~
Published data for the rat varies somewhat in this respect (Zumoff and Pachter, 1964;Setnik ar and Magistretti, 1965), and attention has been called to the error introduced  ' 1960;Conney and Gilman, 1963) ..
In contrast, 3-methylcholanthrene, a polrcyclic hydrocarbon 1 stimulated liver growth and the synthes;s of total liver protein with little or no increase in the concentration of microsomal protein (Conney and Gilman, 1963;Arcos . §.l al, 1961). The~e studies suggested that the ability of polycyclic hydr~carbons to stimulate liver growth was .related to their ability to induce microsomal enzyme formation (Arcos et al, . . -1961) . Some drugs, such as chloroform, halothane and penthrane, caused liver enlargement without enhanced liver microsomal activity and the enlargement was showr. biochemically not to be due to · fat or ede ma (Kunz, 1964;Kunz . §.l tl, 1966). In a structure activity relationship study of p heno l derivatives, no corr e lation was found between the enhance me nt of liver growth and stimulation of aminopyrine demethylat ion or hexobarbital oxidation (Golber g , 1964;Gilbert et tl' 1969).

-
In view of the diversified effects of various drugs on the liver , it was of interest to determinG what biochemic a l effects would be associated with liver enlargement due to pyrethrum .

III. EXPERIMENTAL
A. Animals c, Me asurement Ef Dr ug -Induced Narcosis Duration of hexobarbit a l induced narcosis was taken as the interval between injection of the drug and the return of righting reflex, the endpoint of which was determined according to Wenzel and Lal (1959)  Portions of this fraction were then stored at -20 C until assayed.
The 105,000 x g pellet was obtained by centrifuging 0 aliquots of the 9,000 x g supernatant at 0 C for 2 hours in an IEC preparative ultracentrifuge, #A-170 rotor, 37,000 rpm.
This pellet was dispersed in 0.1 M phosphate buffer (pH 7.4) and stored at -20°c until assayed. In instances when only the protein concentration of this pellet was de t erw. ined, the pellet was digested in 0.5 M NaOH and assayed.

E. Analytical Procedures
.
All incub ations were carr ied out in 2 Dubnoff metabolic shaker under ai.r. All absorbance re a dings were determined in a Beckman DB-G spectrophctometer.

Determin ation of Hepatic ~ater Concentrations
Four ml aliquots of whole liver homog enate equal to 1.0 g of liver were placed in 50 ml tared beakers and frozen in liquid nitrogen~ The samples were lyophilized for 40 hours at 0.5 mm pressure with no external heat in a Virtis LyophilizDr, Research Equipment, Gardner, N.Y •• Water content was determined as the percent loss in ~eight.

Detern:2-nation of Hepatic Total Lipids
The method of isolation of total .lipids was essentially that of Folch ~ ~l (1957) with modifications by Sperry and Brand (1955) and further modified by using lyophilized tissue. The extraction of DNA was carried out according to the procedure of Schneider (1945), while the method of ~easurement was that of Burton (1956).
One ml of whole liver homogenate was mixed with 2r5 ml of ice-cold 10 percent trichloroacetic acid (TCA) 1 and centri- The TCA extracts were combined to form a nucleic acid extract (7.5 ml).

21
In the development of a chromophore, equal volumes of TCA extract and 1o0 N perchloric acid were mixed. To 2 ml of this mixture was added 2.0 ml of diphenylamine reagent containing acetaldehyde. This reagent was prepared by dissolving 1.5 g of diphenylamine in 100 ml of glacial acetic acid and adding 1.5 ml of concentrated sulfuric acid. On the day of use, D.1 ml of aqueous acetaldehyde (16 mg/ml) was added for each 20 ml of reagent required. A reagent blank was prepared containing an equal valuma of 5 percent TCA and 1.0 N perchloric acid.
The above mixtures were incubated after capping .at 30°C for 16-20 hours and read at 6QO nanometers verses the reagent blank.

Determination of Hexobarbital Oxid3tion
The oxidation of hexobarbital was estimated by measuring the disappearance of substrate using the method of Cooper and Brodie (1955). One ml of 9,000 x g supernatant was added to an incubation mixture containing 2.12 umoles of hexobarbital, The barbiturate was extracted and quantitatively me asured accordi n g to a procedure described by Brodie~~ (1953).
The above incubation mixtures were tran s ferred to 50 ~l centrifuge tubes containing 1 g of sodium chlorid e and 30 ml of petroleum ether (technical grade, B.P. 37-47°c and purified by washing with 1 N NaOH, 1 N HCl and 2 washings with water) containing 1.5 pe r cent isoamyl alcohol. Phosphate buffer, 1.5 ml, The r eaction was started by adding 0.1 ml of microsomal suspension containing 1-2 mg protein/ml to the sample cuvette.
The contents were mixed and recording was started immediately.
The reaction rate was linear for 3-5 mino The amo'...lnt of cytochrorie reduced was calculated from the 3 molar extinction coefficient of 18.5 x 10 for the reduced minus the oxidized cytoch~ome c.

Determination of £Q.-Binding Pigmen.t (£:1.-450) Concentration~
The determination of P-450 concentrations was based on the recording of a differen:e spectrum generated by microsomes treated with carbon monoxide as specifiad by Dahlner (1g63).
A reaction mixture of 6.D ml was prepared containing 10- Relative liver weights of the pyrethrum treated animals increased 23 percent as compared to controls, while body weights and hepatic water concentratio,s were unaffected by the tre atment . Although the increase in total lipids is statistically significant at the P: .05 level, it does not account for the total increase in liver weight.
In ,000 x g supernatants anj 10 5 ,000 x g pellets were not  In earlier studies (Table 4) pyrethrum administerad in isoparaffinic oil verses a corn oil treated group, caused a similar decrease in the duration of hexobarbital narcosis.
Hexobarbital serum levels at return of righting reflex were measured and were not sign±ficantly different (P=0.05) from control lev e ls.

In Vitro Metabolic Studies
The 9,000 x g supernatant fr act ions of liver ho mogenates  Further investigation of the in vitro metabolism of aminopyrin e showed the production of formaldehyde by this fractio n was not linear for 60 min (Fig ure 1). However, when portions of the above prepar ed samples were combined and assayed in duplicate determir1ations for the production of formaldehyde at 10, 20, and 30 Tiin there were no significan~ differences between pyrethrum treated and control animals.
The production of formaldehyd9 was found to be linear for at least 30 min.

The investigation for alterations in in vitro drug
metabolism was continued with examin3tion of EPN detoxification, and p-nitroanisol 0-demethylation. In Table 7, rats that had been treated with pyrethrum, 200 mg/ kg/day orally for 13 days, showed significant increases in the detoxification of EPN, and demethylation of p-nitroanisol, 148 and 176 percent of co n trol levels, respectively. Hexobarbital oxidation was increa sed to 173 percent of control le vels.
The in vitro metabolism of EPN and p-nitroanisole by 9,000 x g supernatants were chosen as indicators of microsomal drug metabolism in studies d3signad to establish a dose-effect relations hip as well as the time course and reversibility of the pyrethrum caused liver changes.
In a dose-effect study, groups of 4 rats were administered pyreth rum , 85,200 or 500 mg/kg/day for 15 days. An additional gro~p of animals was administ8red corn oil. Figure 2 graphically relat es the informatio~ from the tables and shows the results of this study.      FI GU RE 2. C omp~sit e Graph : Effec t of pyr ethrum , three do se l e v e ls or a lly for 1 5 days , on the liv er to body weight r at io , detoxific~tion of EPN and d emeth ylatio n of p-nitro a nisol e b y 9 , 000 x g supernatants .

As shown in
Close d symbols repr ese nt signif ic ance f rom control at Pf0. 05 or Pf0 . 01 . 174 percent of controls while the metabolism of p-nitroanisole was increased to 136 7 187 or 21j perc ent of controls. Figure 2 sho~s that with increasing dosage, these parameters increased in an app rox imately logarithmic fashion.
In order to examine the time cours e and reversibility of these changes, groups of 4 rats were administered. pyrethrum, 500 mg/kg/day f.or 4, 7 and 17 days and sacrificed 24 hours subsequent to the last administered dose. An additional group was treated for 17 days and sacrificed on the 24th da y.
Tables 11 to 13 and Figure 3 show the ~esults of this study As seen in Table 11, with increasing duration of treatment, relative liver weights of the pyrethrum treat ed ani mals were increased ~o 148, 15e or 160 percent of control levels after 4, 7 or 17 days of treatment, respectively. However, within 7 days following cessatio n of the 17 day treatment, the relative li0er weights of these animals returned t o control levels.
The in vitro metabolism of Epn and p-nitroanisole was ----~~ significantly increased in these animals as compared to controls. As illustrated in Table 12 and 13, the greatest increases in enzyme activity occurred within the first four days of treatment. EPN metabolism was maximum on the 8th day, where it was 185 percent of control, and had decreased somewhat by the 17th day to 169 percent of control. Return to control levels occurred within 7 days of cessation of tre atment. p-Nitroanisole metabolism increased co nti n~ally throughout the period of treat ment (Table 13), and did not   appear to reach maximum levels on the 17th day where it was 179 percent of controls. Return to control levels occurred within 7 days of cess atio n of treatmgnt.

49
In order to deter~ine if these changes were part of a generalized increase in cellular metabolism or were more specific, the in vitro oxidation of tryptophan was examined (Table 14). The tryptophan pyrrolase activities of 9,000 x g supgrnatants from livers of the 7 and 17 day treated animals were assayed and found to be not significantly different from controls (P =0.05).

Components .9.f. microsomal Electron Transport System
In order to account for the increa se in microsomal drug metabolizing activity, components of the microsomal electron transport system were examined for possible incre ase in activity. Significant increases were found in the NADPH dependant cytochrome c reductase activity of microsomes from livers of animals employed in the time-effect study (Table 15 and Figur8 3). With increasing d~ration of treatment, this activity was increased by 38, 54 and 27 percent on day 5 1 7 and 17, respectively, but returned to control levels within 7 days aft er cessation of treatment.
Cytochrome P-450 levels of pyrethrum and corn oil treated animals were examined. Rats were treated with pyrethrum, 500 mg/kg/day or corn oil for 4 days (Table 16). As correlates, liver to body weight ratios and microsomal drug metabolizing activities were measured. In co mpa ri son to controls, relative liver w3ights were increased to 141 percent while the in vitro   Composite Gr a ph : Effect of pyrethru m, 5 0 0 mg / kg / d a y orally , on the l iver to body we ight r a ti o , detoxific a tion of EPN and demethy l ation of p-n itroani s ol e by 9 , 000 x g sup e rnatants a nd NAD~H c ytochro me c reductase activity by microso mes . Cl osed sym bols r eprese nt si gni fic a nc e at Pf0.05 or PfD . 01 .  In th e absence of irr evers i ble pathological damage , the toxicological significance of this pheno m~non cannot be assessed until more is known of the essential natur e of the enlargement, as well as the ini t ial r e actions, mechanisms, and the effect the drug has on liver met a bo lis m.
Analyses were carri e~ out to determine if the pyrethrum induced liver weight increases were due to fat deposition , edema, proliferation of specific subcellul ar particl es , a coordinated growth of the whole cell (true hypertrophy) or an incre ase in the num ber of cells (hyperplasia).
Control animals received a volume . of corn oil vehicle equal to the total volume of pyr~thrum solution administered in ord e r to avoid liver weight or body we ight ch anges due to differences in diet.
Measurem en t of water and lipid concentrati ons (Ta ble 1) sho wed that th e liver weight incr eases we re nffit due to fat deposition er edema .
If As seen in Table 2 Duration of hexo ba rbital narcosis, which is limited by its metabolism in hepatic microsomes (Cooper and Brodie , 1955) was decreased in pyrethrum treated animals (Table 3). · The onset and duration of narcosis due to barbital, which is ex creted unmetabolized (Dorfman and Goldbaum, 1947;Mayn~rt andVan Dyke, 1958: Ebert . §l ~' 1964), was unaffe cted by pyrethrum pretre a t~ent. These da ta suggested that th 0 decreases in hexobarbital narcosis resulted from an increas e in microsomal drug metabolism r at her than alterations in central nervo us system sensitivit yo This idea was further sub sta ntiated by the l ack of differences in hexobarbital serum levels at return of ri ghting r eflex in the two groups (T ab le 4).
The study was c ont in ued wit h an examination of in vitro mic r os oma l drug metabolism in order to determine if the decrea ses in he~obarb ital narc osis resulted only from an incre ased metabolic cap 2cit y of a larger liv er or wa s co mb ined with incre ases in spec i fic enzymatic activity. Measurement of aminopyrin e N-demethylase activity was chosen as a possible indic ator of altered wicrosomal drug metabolism since its meta bolism is increas ed by two prototypes of inducing agents , 3-w.ethylch olanthr ene end phenob arb ital~ The N-demethylation of aminopyr ine was found to be unaltered by pyrethrum treatment (Table 5) . The N-demethylation of p-chloro-N-methylaniline was also found to be unaltered by pyrethrum treatment (Table 6) .
Since the decrease in hexobarbital narcosis ~ithout a concomitant alteration in barbital narcosis sugges te d an increase in microsomal drug metabolism, even though there were no incre a ses in the N-demethylation of p-chlorom~thyl aniline or aminopyrine, the in vitro metabolism of hexobarbital was examined. Tables 5 and 7 sho w that in vitro hexobarbital oxidas e activity was elevated by pyrethrum treatment.
In his revi e w, Conney (1967) drew a di st inction between the selective ind u ction effects of the poly cyclic hydrocarbons, which ind uce only a few activities such as zoxazolamine hydroxyl ase and azo dye demethylase bu t not hexobarbital oxid Gse or aminopyrine demethylase, and the generalized inductive effect of ph e nob arbita l which induces all fo ur a ctivities.
Becau se of the se otservations, the lack of incr ease in aminopyrine dem et hyl ase WES subjected to a more detailed study.
Samples that showed no incr ease in aminopyrine demethylase activity when assayed for formaldehyde production at 60 min were pooled in pairs and in duplicate determin at ions, assayed for formaldehyde production at 10, 20 and 30 min. As shewn in Figure 3, there was no significant difference between pyr ethrum treat e d and control animals in the prodcction of formald eh yd e at 10, 20 and 30 min , although the rate was approximately line ar for only 30 min.
This work confir~ed that of Soyka (1969), who noted that even if linearity for 15 min was accom plished, the use of too small a qcantity of glucose-6-phosph ate limited the production of formaldehyde. Thus the activity of adult liver erroneously appeared to be less than that from 20 day old animals if studied after 30 min of incubation.
The increase in hExobarbital oxidase activity withoct an increase in aminopyrine demethylase activity suggested that pyrethrum did not resemble in effects produced either prototype of enzyme induc er . It is interestirig to note that in studying th e structure activity relationships of substituted phenols on liver we ights and liver enzymes in the rat, Gilbert ~ ~ (1969) found there was no dirEct relationship between liver weight increases and incr eases in hexobarbital oxidase and aminopyrine demethyl ase acti viti es , Cert a in co mpc unds consi stently showed a greater effect on aminopyrine demethylase than on hexobarbital oxidase while the reverse also occurred.
Th ey further found that 2,6-di-tert-butyl-4-hydroxymethylphenol induced hexobarbital oxidase strongly with no corresponding increase in the activity of aminopyrine demethylase .
This new catego~y appears td exemplify the effects of pyrethrum.
In order to further categorize the effects of pyrethrum, other microsomal enzymes were examined for alteration in activity, including p-nitroanisole 0-demethylase and EPN detoxification. Table 7 shows that the activities of both of these enzyme s were increasedo At this point it was of interest to determine if these incr eases in drug metabolizing activity were part of a generalized increase in cellular metabolic activity or were mere specific as suggested by the lack of increase in aminopyrine N-demethylase and p-chloromethylaniline N-demethylase activitieso Lack of increase in tryptophan pyrrolase activity (Table 14), an enzyme localized in the 105,000 x g supernatant fraction and therefore a probable cytoplasmic enzyme, further supported the hypothesis that the enzymatic increases were specific in nature, indicative of true enzyme induction.
Investig ation of liver enlargement and increases in drug metabolizing activities due to pyrethrum was continued with establishment of a dose-effect relationship (Figure 2). The results of this study indicated that increases in enzyme activity ware evident before liver enlargement. Increases in enzyme activity paralleled the liver enlargement . These parameters increased in a logarithmic fashion with increasing dos ages indic at ing a dose-effect rel at io nship . Th e data also sug~ested that tha maximum re sp ons e of the parameters had not been re ached , although higher dos es are within th a lethal range of pyrethru m.
An investiga tion was also undertaken to s~udy the time course a nd reversibility of these changes (Fig~re 3). Liver enlarge men t and p-nitroanisole metabolism increased with increasing duration of treat men t. The greatest increase occurred during the first four days of treatment, with smaller increases at the next consecutive measurements, suggesting the attainment of a platea u of peak activity after ab~ut 2 weeks of treatment. EPN de toxification and NADPH-cytochrome c reductase activity paralleled each other in activity. The greatest increas e in activity occurred during th e first four days, b~t unlike the time course of p-nitroanisole ~etabolism and liver enlarg ement , it was maximum on the 8th day and had declined so~ewhat by the 18th day. Martin and Gilbert (1968) and Gilbert et~ (1969) recog , 1ized that continued administration of a compounj could lead to regression of the initial effect . These studies suggest that increasas in liver size cannot necessarily be equated with increases in enzymatic activity .
All parameters measured, returned to control levels within 7 days after ces sat ion of treatment showing that the changes are spontaneously reversible upon discontinuation of exposure to pyrethrum.
True enzyme inducers, stimulants of hepatic microsomal drug metabolism, have the abilit y to increase components of the microsomal electron transport system. For example, pheno barbital and 3-~ethylcholanthrene cause increases in P-450 conc e ntrations, and phenobarbital increases th 3 activity of NADPH-cytochrome c reductase (Conney, 1967). These observatio ns suggested that pyrethrum may have a similar inductive effect on these components of the microsomal electro~ transport, and consequently be a true enzyme inducer. As shown in Tables 15 and 16 and Figure 3 , both of these components are elevated by pyrethrum treatment .
It sh ou ld be noted that the magnitude of the doses employed in this study would not be encountered in normal usage of pyrethrum. However, this study does not indicate a no-effect level for py:ethrum.
VI. SUMMARY AND CONCLUSIONS 1, Liver enlargement due to pyrethrum is associated with proportional increases in total water and total protein of whole liJer, 9,000 x g supernatant fraction and 105,000 x g pellets. St atistically significant (Pf0.0 5 ) increases in total lipid .per gram of liver did not account for th3 total incr ease in liver weight. The en lar gement was further associated with a decrease in DNA per gram of / liver, suggestive of hypertrophy. Microscopic examination of liver sections supported this hypothesis.
2. Liver enlargement due to pyrethrum is associated with a decrease in hexobarbital induced narcosis. Hexobarbital serum levels at awakening were not significantly different in control and pyrethrum treated animals . The onset and duration or barbital narcosis was not significantly al t ered by the pyrethrum pretreatmento These results suggested that the decrease in hexabarbital induced narcosis was due to increas ed microsomal metabolism of the drug rather than a ch ange in CNS sensitivity to the drugo 3. In vitro increases in specific microsomal oxidation of hexobarbital, 0-demethylation of p-nitroanisole and detoxification of EPN were associated with ~he liver enlargement.
Th e observed lack or increase in specific activity of aminopyrine demethylase did not result from the duration of incubation employed in the assa y, although the production of form a ldehyd e was not linear for 60 min jut lin ear for at l east 30 min. The lack or i ncrease in this enzyme , as well as th e l ack of increases in aminopyrine N-d eme thylase and p-chlorornethylaniline N-demethylase, suggested that the inc reases in microsomal drug metabolism were not part of a generalized increase in c e ll u lar metabolism .

5.
Incr eases in liver to body weight ratios and incre a3e s in met abolism of p-nitroanisole and of EPN by 9,000 x g supernatants were found to be dose rel ated . At the lo wast dose us ed of 85 mg /kg, inc reases in microsomal drug metabolism wer e evident before liver enlarge ment . Peak c c tivity of th ese parameters was not apparently r eached at the highest dose used of 500 mg/kg, although higher doses were within the leth a l range of pyrethrum.

6.
Incr eases i n liv ar to body weight ratios and incre a ses in metabol ism of p-nitroanisole and of EPN by 9,000 x g s upernatant s were found to be time rel ated . The g~eatest incre ases in the3e p arameter s occurred within the first four days of treatment, but did not appear to re a ch plateau l evels simultaneously. The incre ase in NADPH-cytochrome c r eductase , a component of the microsomal electron transport system , was also time rel ated . It is import ant to no te th a t all parameters measured return ed to control levels within 7 days after cessation of treatment, suggesting that irreversible changes had not occurred. 65

7.
Increases in P-450, a component of the microsomal electron transport system, were associated with increases in liver to body weight ratios and increases in EPN detoxification, p-nitroanisqle demethylase and NADPH-cytochrome c reductase.
8. These findings are consistent with and support the hypothesis that pyrethrum caused liver enlargement results from true hypertrophy. It is proposed that increases in specific microsomal drug metabolizing activity , as well as an incre ase in metabolic capacity resulting from the larger liver, account for the decrease in hexobarbital induced narcosis. It is proposed that the increases in specific enzyme activity result from the synthesis of new protein and that in fact, pyrethrum is a true enzyme inducer.
These data suggest the need for further investigation to reevaluate the use of pyrethrum as a non-inducing insecticide.