Document mBvoavq7QeV97MgJ84N57xZdb

NATURE. VOL. 214. APRIL 20. 1967 nature. 1 tit it lit ov*---. *f tho-,,- list'll, lu filit-nil,i.'l culture- i*i i><! 1* <1, ; Iti-r mriitj.iti<iii For 1 day mid tli'*iiif*zr-iti**ii "t cells. in) medium uluiie, (k) suporimfnut n-d111;11 htou intact macrophugcs alone and From intact colla maintained in medium containing PNG, and (e) supernatant medium from initially intact macrophages incubated with quartz (O'l rag/2x 10* cells/ml.) mid from intact macrophages incubated with quartz in the same concentrations together with PNO. In one experiment. 1 mg/ml. I'NO was added to tho niaerophago cultures 1 h before adding quartz. Controls (a) and (<>) showed no rise of HOP concentration, A significant increase of HOP occurred at 2 and 4 days in quartz cultures whether treated with PNO or not. there being no significant dif ference between the increases in the two cases. In a second experiment PNO was applied to macrophage cultures at 2 mg/ml. for l day before adding quartz. The experiment lasted only 2 days, but HOP production was raised with quartz alone (Pc 0-001) but not at all with quartz in the presence of PXO. The proximate mechanism of fibrogenesis by silicon dioxide is still obscure but the present experiments provide, apparently for the first time, tissue culture evidence of a quantitative nature that implicates the reaction between silica nd macrophages as an essential initial step in the process. The evidence permits the conclusion that this is a vital phenomenon, which does not necessarily involve a direct effect of dissolved or particulate silicate on fibroblasts. Moreover, the comparative experiments with titanium dioxide and PNO (higher concentration, longer action) suggest that the silica-macrophage reaction is specific. The nature of the active factor has still to be elucidated, but another experiment suggests that the debris from quartz-treated, disintegrated macrophages lacks the factor. The idea that the fibrogenio agent is lipid in nature1-1* now faces preliminary evidence to the IJcoDntrr.arHy1a*n. s Leiteritz kindly provided the Dfirentrup quartz. Dr. IV. Twist the titanium dioxide and Professor Dr. H.-1V. Schlipkoter the PNO. This work was sup ported by a grant from the National Coal Board. A. G. Hepfleston J. A. Styles Department of Pathology, University of Newcastle upon Tyne, Royal Victoria Infirmary, Newcastle upon Tyne. * Fallon, J. T.. Canod. JM. Arne. J,, 33.223 (1937). * Marks, J . anil XagelschmiiU. G.. Amtr. Med. A not. Arch, Indml. Health. 20. 3S3 (1350). * Curran, K. C..and Agsr. J. A. IT.. J, Path. Kiel; 33.1 (1032). * Vigilant. E. C,, and Persia, B., A dr. Tubtre'doni Ret.. IS. 230 (1083). * Marks, J , Mason, it. A,, and Naselsr-hmlUt, G.. Brit. J. tnduit. Hid., 13. 187 (19.M). * VicIlani.E. C., Vernis.B., and Monaco. L.. In Inhnitd Pnrticiei and Vapovrt (die. by Davits, C. X.). 3(8 (Pergamon Press. London, 1081). ' SchKpkbter. H.-W., and Beck. E. O., 3/nf. Aurora. 36. (85 (1985), * Trovveil, O. A.. Exp. Celt. Am., 18.118 (1050). * Cerlottl, 0.. J. Bid. CAm..l9S, 297 (1052). " Woesaner. J. F., Areh. Biaehtm. Biaphpi., 93. ((0(1081). " Schlipk.-:r. H..\V,, and Brorkhaua, A.. Ditch. Mid. Wvht., 38. 9i0 (1080). 11 Schlipkoter, H.-W.. and Brockhana. A., Elin. 1Ttche., 30.11S2 (1001). " Beck. E. (?., limch. J.. and Brnckhau*. A.. Z. ZeUfoneh., 69,563(1063). 14 Harinston. J. 8.. .s'. African Med. J,, 37. (51 (1063). 14 Weh*t-r, I,, avi .liarnsns. I... X. African Council for .Sri. and fndnA. Ree., .ViiAh Am. t;,p., (1065). Haemolytic Activity of Asbestos and other Mineral Dusts Silica in the particulate form has been shown to have "vtic action ou intracellular structures1"1 and on crythri evtes1-*. Tin- present investigation has shown that ehrysi tile asbestos i' as potent a haemolytic agent as sili< powder. Other forms of asbestos lysctl erythrocytes on after priiliiiii't-rl iuc-iib.ilinti u ith tlii-m. and .i v. id>- rang*- of mineral dost.-, was inactive. The haemolytic action of clirysotilo was completely prevented by disudimu ser- scuate and simple phosphates Imt only partially by polyvinylpyridino-iV-uxide and aluminium, both of wliich protect cells Against lysis by silica1. The following finely ground material* were used: crystalline and amorphous silica, silicic acid powder, silica gel, clirysotilo, amosite, anthophvllitc and crocidoiite asbestos, serpentine, aluminium oxide, `Sephadox G25', high and low rank coal dust, mine dust, colloidal carbon, talc, knolin. glass powder, carborundum and diamond dust. For the majority of tests n standardized 2 per cent suspension of washed sheep erythrocytes in a veronal buffered isotonic saline buffer (pH 7-4) was used. Four millilitres of this suspension were completely lysed by 50 mg of clirysotilo when this was incubated at 37 C with a fixed volume (4 ml.) of suspension fur u period of 50 miu. This was taken as the standard time for all subsequent quantitative tests. The effects of various agents in pre venting haemolysis by silica and chrysotile were in vestigated after the minerals hod been coated with suitable concentrations of the protective materials, followed by repeated washing with saline. Significant hoemulytic activity was found for chrysotile, serpentine and all forms of silica tested. (Tho particle size ranges of the aotive dusts were the following: chryso tile, 11 per cent <0-5p; 37 per cent 0-5-2p; 33 per cent 2-5|X; 15 per cent 5-10|i.; 4 per cent 10-50(i length; serpentine, very fibrous, no size distribution obtained; crystalline silica, 15 per cent <2-5p; 49 per cent 2-5-Sp; 30 per cent 5-7-5p; 5 per cent 7-5-10p; amorphous silica, 250-400 A; silicic acid powder, not determined; silica gel, 10-~40p.) The remaining powders were either completely inactive or only weakly lytie. Quantitative values for the powders found to be lytic are shown in Table 1. Table 1. haemolytic actiyttt or sauxs srsrtxsioxs or cmrsomz JASZSTOS AND UUCA POSTS Hlaersl Fully hnemolysed erythrocyte euepenelon Aeroeil (colloidal silica) Crystalline silica Crystalline silica (aeld-srsahed)* Silicic acid powder Silica gel Chrysotile asbestos Serpentine Per cent haemoglobin in supernatant 100 1s0s0ostt si all me*taSlsil.ica repeatedly waehed with boiling hydrochloric acid to remora surface t No adsorption or haemoglobin--some erythrocytes left intact. hnemolyaed.t Lower valuti caused by adsorption nf haemoglobin on tbasa AUS*-- erythrocytes completely QQ ^)) ) j The rate of haemolysis by chrysotile was greater than that shown by silica, the apparent lower percentage haemolysis by the asbestos (Table 1) being the result of the adsorption of released haemoglobin during the early stages of lysis. Haemolysis by chrysotile took place rapidly, with tho immediate adsorption of haemoglobin on the asbestos fibres. Because of this, the degree of haemo lysis in the supernatant (as measured colorimetricallv) never reached 100 per cent. This was also found in the case of the adsorbents, silicic acid and silica gel. Several samples of atnosito, anthophyllite and crocidolite asbestos gave consistently negative results when incubated for 50 min at 37 C. In tho case of tho first two, haemolysis started only after 5 h at 37s C and reached its maximum aftt-r 24 h. Crociclolite had a very mild haemolytic effect after 24 h. Haemolysis by chrysotile was inhibited by serum and bylow concentrations of disodium ethylenediamine tetraacetic acid (EDTA) and phosphate ions (35 mg FO'") in the standard test medium. In tho case of EDTA it was found that 20 (imoles of chelating agent in 4 ml. oferythro cyte suspension at 37s C completely prevented the haemo- 1 I , lie ilCtHMi j [u-i- vi-nti - L per c rot PC-live ]vj|- by cli it complete The pot< silica has mechanical -urf-ice cai du-ts, for e powder, gi- i-xpcriment pensions ar. The resul suggest the their haem .icid intern. ,md that h> acids and Recent inv phagosoma. that side e with the t preventive mbcellular to the preftho silica si acts more e i-rythrocytt protection with the p The marl (and, presu phagosoma formation < ionic form surface. In the c40 per cent given by indicates ti this must b the moleci. offer not c interaction eo-ordinati: 100 T 80 t 20 0 r-- Fiji. 1* Itihi lvsia by chr> vlimcvtt su* il-e curve. j BB 0015499 j V tl967 igeof action of ilium ver'.y by polyof which ere used: i powder, uid crociphadex Gt, colloidal udtim and 2 per cent a veronal iscd. Four .* lysed by 37 C with I of 50 min. subsequent uta in pre; were inith suitable ollowed by chrysotile, he particle ig: chryso33 per cent Op length; l obtained; ?nt 2-5"5p; 'liodkilica, 1 ^Bgel. Cv^^Ietely lues for the 1. ir emmonu -moglobln .mant t 't t it remove turtle# t.ict. i III** dilute-- router than percentage he result of .'z the early look place uoglobin on e of liaemoi metrically) mid in the ;ol. md croci-.nits when i nmand l>v mine tetrn- i mg po;-) I )TA it wh- . oforythro- die liiiemo- NATURE. VOL. 214. APRIL 29. 1967 923 lytic action of 50 mg chrysotile (Fig. 1). On the other hand, with cations such as magnesium (II) ions only when 2 per cent aluminium chloride. 1-25 per cent gelatine and suitable anions are available to maintain electrical 0-9 per cent polyvinylpyridine-A'-oxide (all effective neutrality (T. Nash, personal communication).. protective agents against haemolysis by silica) retarded That magnesium in the ionic form is the principal lysis by chrysotile for short periods but failed to prevent cationic reactant on tho surface of chrysotile in haemo > it completely. lysis seems to bo indicated by the effective inhibitory The potent haemolytic action of both chrysotile and action of EDTA,- a material which also prevents lysis by silica has been demonstrated. The influence of any macromoleculcs containing ionized iron*. This belief is 1 mechanical action of these materials on the erythrocyte supported by the inhibitory action of simple phosphates ' surface can be largely excluded because other abrasive in the medium. Nash has suggested (personal communi dusts, for example, carborundum, diamond dust and gloss cation) that non-ionic compounds (for example, poly- . powder, gave little or no haemolysis under the same vinylpyridino-N-oxide) or betaine-like compounds would experimental conditions. Recent work4 on silica sua- be expected to protect cells against haemolysis by silica * pensions and inert dusts confirms this observation. because of the nonionic character of the latter. If mag The results of our investigations with protective agentB nesium ions are the lytic agents, however, aa appears to suggest that silica and chrysotile differ in the nature of be the case with chrysotile, polyacids would seem to be their haemolytic action. There is evidence that silicic suitablo inhibitors. Good co-ordinators like phosphate acid interacts effectively with lipoprotein monolayers*-' and polyphosphate should also be effective. The iron in and that hydrogen bonds are formed between keto-imino- chrysotile (up to 2 per cent) appears to play no part in lysis acids and the hydroxyl groups of the silicic acid4-'. by this form of asbestos, because no inhibitory effect was Recent investigations of the damage by silica to the obtained in tho presence of a.a'-dipyridyl in the medium. phagosomal membrane of macrophages1-* have indicated On the other hand, the presence of magnesium and that silica exerts its lytic effect by forming hydrogen bonds calcium ions had no enhancing influence on the weakly j with the biological membrane of the phagosome. The lytic effect of crocidolito and amosite. , preventive effect of polyvinylpyririine-N-oxide at the The haemolytic effects of chrysotile are of interest in , subcellular level was ascribed after experimental evidence* view of the relative inactivity of the amphiboie asbestos to the preferential formation of hydrogen bonds between forms, crocidolito and amosite, and the pronounced the silica surface and the polymer. That the latter inter activity of certain polymerized silicates which, unlike acts more extensively with the silica surface than with tho monomeric forms, bind firmly to the erythrocyte mem erythrocyte membrane is shouit by the low degree of brane*. The active fibrogenic properties common to the i protection obtained when blood cells are preincubated types of asbestos investigated here cannot yet be associated t with the polymer only*. with their haemolytic characteristics, a relationsliip The marked depression of silica solubility by aluminium suggested for different forms of silica4. The inactivity of t (and, presumably, the protective effect of this ion sgainst crocidolite, amosite and anthophyllite after incubation phagosomal membrane lysis by silica)1 is caused by the for 50 min suggested that if haemolytic activity- is to be 1 formation of aluminium hydroxide by aluminium in tho related to fibrogenicity, contact of these forms of asbestos \ ionic form reacting with the hydroxyl groups of the silica with erythrocytes would have to be prolonged; the weak . I surface. haemolytic activity observed by us for crocidolite, amosite In the caso of chrysotile, a fibrous silicate containing and anthophyllite after extended incubation with ery 40 per cent magnesium (aa oxide), the partial protection throcytes supports this possibility. given by aluminium and polyvinyipyridine-iV-oxide Finally, it also seems clear that the highly active lytic indicates that if hydrogen bond formation takes place, property of chrysotile is related to the adsorptive capacity , this must be relatively slight and may be accounted for by of this form of asbestos. The present investigation has the molecular configuration of chrysotile which would shown that under the same conditions, chrysot ile adsorbs offer not only hydroxyl groups but cations for surfaco five times more protein from serum than silica. More interaction. Polyvinylpyridine-A'-oxide is a neutral protein is required to "cover" the haemolytic sites of co-ordinating agent and would be expected to interact chrysotile than of silica; this is not surprising in view of the chemical structure and morphology of the two minerals. Chrysotile also has the highest Mg/Si ratio of the three forms of asbestos tested and differs also in that- iron is not formally included in the lattice. It has a very high zero point of charge (pH 10-12), that is, over most of tho pH rango it will carry a net positive charge**. We thank Professor J. H. S. Gear, Director of the South African Institute for Medical Research, for facilities, Mr. T. Nash, and Dr. A. C, Allison of tho Medical Research Council. Loudon, for valuable criticism and comments, and tho National Cancer Association of South Africa for full financial support for one of us (J. S. H.). Gwen M.vcxab J. S. Harixctost South African Institute for Medical Research, P.O. Box 1038, Johannesburg, --------------- ------------ Republic of South Africa, ] BB 0015500 | F12,1. Inhibitory i*f i .trying rotw-rmraiuM:# <it` K()TA mi lwi* l*y <lirv*-t U< t.VWnif * tr\ *mih\ 2 hit. LlkTA !fluhmi, _ ml. t-rV* tlifhrvir JO mm inubati<n ;ii j7 < : lint Inuht-dt v;tlm* mi turxt*, 7> |T c*-ul, n-jirt`!iM th4 flfiTt nl umrouhtl chrvwililft *m `ryilirii'Mf**). 1 Harlneton, J. 3,, and Alllwn, A. C., lied. La t o ~ "------ * Allison. A. l\, Uarlnslon, J. 3,, and Btrhrck. Ji.t J. hip. Mid.. ISO.1*1 (1009) ' Xnsh, T.. Allison. A. C\, anil ltarinalon, J. 3.. Satan,210. 259 (1006). * sender, K., anil Stolier, \V,. Satan. 207, >71 (1005). * Srlmliiian. -I. II., anil Riilisil. K. IF. Pror. H'W. >'#.. U122.16 (1097). 4 rimk, 3. *., mnl Holt. I*. F., Ttan*. ratad, >'<*.. 53. I.'ioo (t'..'>7). i Cl.irk. S. I)., Holt. P. F.. mnl Wont. r. W\. T.a,t<. Fatal > . 53. 1500 11957). * Fi' l lin-.*. J.. uni Smith. 17. M,, J. ciia. 18. 12 (IWU. 4I'luirarlr*, p,, M.uli'Uil. (*, M-- amt Whili*. I4.. J. Ufa. I'h-t.'aF 45, 1117 I1U02). ** Xannuinn, A. W., and Drentier, V. H., J, Phvi. Chita., 70. 233 (1900), -- 352 Sty- NATURE VOL. 243 JUNE 8 197: ' Wilde * Wildcnthal|JKsd5Zt^/. PY* (1969). J WildeMWTfcrfT /Vac, PhjtmJflSoT., 217. 56 (1971). 4 WildehfhaT fr Physiol., 221. 238 (1971). ' Hughes, DjMdAnu Longmore, D. B., Nature, 235, 334 (1972). * LongmorD. B., and Hughes, D. M,, Nature, 238, 40 (1972). 7 Nagler, J., and Longmore, D. B., Nature, 242, 197 (1973). 1 2 i 4 4 6 7 8 9 10 N 121.1 I4|J 16 17 18 192021 22 2324 b SI Penetration of Asbestos through the Digestive Tract of Rats . I Mesotheliomas of the pleura and peritoneum*-1 *h*ave been linked with the inhalation of asbestos fibres. We have shown* that asbestos fibres are present in drinking water, beer, wine and other beverages. Asbestos fibres can penetrate the mucosa of the stomach and intestine9 and Tclischi and Rubenstone* have found asbestos material in a gastric carcinoma. Godwin I and Jagatic7,*reporting on mesotheliomas of the pleura caused by asbestos, noted that asbestos particles were widely distri buted in various tissues. They suggested that particles mar migrate from the lung through the blood and lymphatic system to all parts of the body. In the light of these report!, we surmised that asbestos consumed orally can pass throufh * the gut into the blood stream and accumulate in various tissues To determine if this is the case we injected a suspension of asbestos fibres into the stomachs of rats. The suspensioe was prepared by shaking 2.3 g of chrysotile asbestos (Johns Manville No. 7RF02) in 500 ml of water in a graduated cylinder. The suspension was allowed to settle for 30 mia and the top 230 ml, which contained about 9.4 x 10* (1 mp fibres ml-1, was drawn off. Most of these fibres are 0.2 pm- 2.0 pm long (Fig. 1). To ensure that fibres passed through the digestive tract without rats inhaling any of them, the asbestos was introduced into the gastrointestinal tract hr opening the abdomen under anaesthesia and injecting the fibres directly into the stomach. Two to 4 d later the rats were killed and blood and tissues analysed for asbestos. T ensure that any fibres on the rat hair could not contaminate $^ the organs removed during surgery, the rats were first then oughly vacuumed and washed with double distilled water and methanol. During the removal of tissues wet towels wen placed over the animals and around the incisions. Uncov taminated blood was drawn from the orbital sinus using a capillary tube. To avoid formation of excess ash, which interferes wifh examination of specimens under the electron microscope, iht tissues were first solubilized with soluene (Packard lnstrumoi Company); the fibres were then centrifuged down, washed with methanol and ashed as previously described4.S.T*he uh Days was taken up in 1 ml of distilled water which had been filtered through a 0.2 pm filter, and 5 pi of this mixture was dropped Fig. 2 Ventricular rates of the control hearts compared with on hydrophilic carbon-coated electron microscope grids the first 23 d for the test hearts. supported on a ring peg. The grids were dried in a clotal I glass container (to avoid air contamination) under a ho* seem reasonable to continue with further conventional clinical lamp and were examined at magnifications of 20,000 sal trials. If a new drug is depressant to the foetal myocardium 80.000 with an electron microscope. The presence of chrysotdr above a certain concentration, it would be unwise to exceed asbestos fibres was verified by electron diffraction. They *** this in a subsequent clinical trial. If the substance were to counted and the numbers corrected for dilution to detenu* affect only young foetal hearts it might still be safe in late the number in the original I ml sample. Results of recowy pregnancy. of asbestos fibres from rats treated in this way are shown* We thank the Board of Governors of the National Heart and Table 1. Chest Hospitals and the British Heart Foundation for financial support. Miss D. M. Hughes and our colleagues for help and Rats in group 1 were given 9.4 x 10* fibres (1 ml) and kibd 2 d later; those in group 2 received 94x 10* fibres and woe Dame Honor Fell and Kern Wildenthal for encouragement. killed 4 d later. No asbestos fibres were detected in the bleed National Heart Hospital, S. R. Armstrong D. B. Longmore ' ** of non-injected rats (controls) but 4.65 x 10* g'1 (a statisnoft significant increase) were found in the first group. This is the first time direct physical evidence for the preseaa Westmoreland Street, of asbestos fibres in the blood has been obtained. Four di* London W\M SBD JUl after treatment the amount of asbestos in the blood of rati > Received January 23; revised April 10,1973. group 2 had dropped to 1.19 x 10* g*` even though these r* to % received ten times as much asbestos. This indicates that bkat 1 BB 0015501 \ T 4 j NATURE VOL 243 JUNE 8 1973 i^ i Table 1 Numbrt o( Asbestos Fibres Recovered from Traatad and Non-Traatad Rata w------------ Controls mean* 1 Blood 0.00 _ [ Spleen 2.33 0.675 | Omentum 2.46 0.46 , Heart 2.09 0.41 Brain 0.05 0.02 , Lungs 1.06 0.26 Group 11 mean* 4.6551 4.11 2.74 -- 0.3111 1.80 1.02 0.49 0.87 -- 0.19 0.54 Group 2i mean* 1.19 3.45 18.25 2.28 0.29** 1.74 0.77 1.20 7.79 0.40 0.11 0.26 Average of five rats in all cases except where noted, f Animals | injected with 9.4 x 10* fibres and tissues examined 2 d later. I Ani- trials injected with 94 x 10* fibres and tissues examined 4 d later, I {Standard error of the mean. |l Four rats only. 51 P<0.01; 1 "Pc 0.05. (The figures in the table are fibres g"` x 10-*.) j can clear itself of the fibres. Tissues such as the spleen, heart and lung also seem to have, to a lesser extent, the ability ; to clear themselves of the asbestos fibres, as the figures obtained ( from rats in group 2 indicate. They are lower than those for 1 group 1 rats even though they received a larger dose, but the tissues had longer to clear themselves. Indeed, recent work in our laboratory with radioactive asbestos indicated that some tissues, such as the lungs, can reduce their fibre content by half in 2 d. This clearing action could possibly have quite in effect on the numbers of fibres found in the liver and 1 kidney of the rats but unfortunately, because of the considerable ! imount of ash present in these organs even after soluene treatment, we could not examine these specimens under the electron microscope. Another factor may contribute to the higher fibre count Jammed for the organs of rats in group 1 compared with 1 ^^pls in group 2, namely, that there were four times as I n!Ry fibres in the blood of rats in group 1. The residual blood < in the tissues on removal from the animal would of course ' contribute to the total fibre count of a particular tissue. | This factor could have created the anomaly in the fibre | chants in brains. In the brains of treated rats this was about I six times higher than in the controls, and the increase in fibre count is statistically significant for rats in group 2 but not, oddly, for rats in group 1 even though there were more fibres in these animals. The greater number of fibres per gram in the blood of rats in group 1 could be responsible for the , larger standard error in the counts of rats in group 1 and therefore did not allow the values obtained to be significant. The brain seems to have little ability to clear itself of asbestos fibres, but the omentum has even less. This tissue, which Fig. 1 Electron micrograph of the suspension of asbestos fibres injected into the rat stomach. Note the varying sizes of the bundles of fibres present. f '.*'9 353 surrounds the small intestine, appeared to accumulate most asbestos. Although the asbestos fibre content of the omentums in rats in group 2 may seem rather high (one extreme count contributed to the high standard error) the omentum does not seem to respond as do the other tissues. As well as having little ability to clear itself, fibre accumulation in the omentum may take place much more slowly than it does in other tissues. As the omentum consists partly of peritoneum and partly adipose tissue, asbestos fibres may enter the peritoneum by direct penetration of the intestinal wall. Concerning the prolonged retention of asbestos fibres in the omentum in women asbestos workers, Keal* found a higher incidence of peritoneal cancer than of lung cancer. The amount of asbestos found in the tissues of the controls is higher than we had anticipated, considering the age of the rats and the number of fibres they might have consumed in their drinking water*. But we do not know how much was in their feed or air. We have analysed the tissues of some people who died of natural causes and found somewhat similar levels. In one person there was twice as much asbestos in the brain as in the brains of our control rats and about one quarter as much in the omentum,,but the omentum content was still higher than the brain. Most of the fibres found were about 0.2-2.0 pm long but some measured IS pm (Fig. 2) and one fibre in the blood of a rat was 23.55 pm long. We do not know how the fibres pass through the intestinal wall. Some long ones may pierce the gut like a needle, whereas pinocytosis may account for the absorption of the small ones. We demonstrated earlier the presence of asbestos fibres in air, snow, city drinking water and a number of beverages*. The experiment described here shows that fibres of a similar size administered into the stomach of rats appear in the blood and accumulate in various tissues in the body. R. D. Pontefract Food Research Laboratories, H. M. Cunningham Health Protection Branch, Department of National Health and Welfare, Ottawa Received November 30, 1972; revised February 13, 1973. 1 Hourihane, D. O. B., Thorax, 19, 268 (I96J). 1 Hourihane, D. O. B., Ann. NY Acad. Sci., 132, 647 (1965). 3 Lynch, K. M., and Smith, W. A., Amer. J. Cancer, 24, 56 (1935). * Cunningham, H. M., and Pontefract, R, D,, Nature, 232, 332 0--7I). 3 SelAolf, 1. J., Churg, J., and Hammond, E. C., J. Amer. Med. A\w,, 188, 674 (1964). * Teh chi, M., and Rubenstone. A. I., Arch. Pathol., 72 (1961). . Godwin, M. C., and iagatic, J., Environ. Res., 3, 391 (1970). * Real, E. E,, Lancet, ii, 1211 (I960).