interacts specifically with a distinctivepolysaccharide on the surface of theappropriate Rhizobium cell as a preludeto nodulation. These data demonstratethat the possibility is a real one whichmerits thorough investigation.

B. B. BOHLOOLE. L. SCHMIDT

Departments of Microbiology and SoilScience, University of Minnesota.Minntesapolis 55455

References and Notes

1. W. D. P. Stewart, Nitrogen Firation in Plants(Athlone, London, 1966).

2. J. Hamblin and S. P. Kent, Nat. New, Biol.245, 28 (1973).

3. I. E. Liener and M. J. Pallansch, J. Biol.Chem. 197, 29 (1952).

4. A. G. Gornall, D. J. Bardawell, M. M. David,ibid. 177, 75 (1949).

5. 1. E. Liener, Arch. Biochem. Biophys. 54,223 (1955). Hemagglutination titer is the recip-rocal of the highest dilution which, in 21/2hours, gives a reduction of 50 percent or morein the optical density at 620 nm of a standardsuspension of trypsinized red blood cells.

6. The protein content of fraction II was adjustedto 1 percent with saline; to 10 ml of this solu-tion was added 4 ml of 0.1M sodium phosphatebuffer (pH 7.0) and then 4 ml of 0.1M sodiumphosphate buffer (pH 8.0) containing 5 mg offluorescein isothiocyanate (isomer I, BaltimoreBiological Laboratories). The resulting solutionwas adjusted to pH 9.0, merthiolate (1: 10,000)was added, and the reaction proceeded for 24hours. Unreacted dye was removed by dialysisagainst saline buffered with 0.02M sodiumphosphate to pH 7.2 until the dialyzate wascompletely colorless. The fluorescent lectinconjugate was distribuited into ttibes and heldfrozen at -20°C.

7. Microscopy was performed with a Zeiss Uni-versal microscope equipped for dotible lightingwith incident illumination from an HBO-200(Osram) light source and transmitted dark-fieldtungsten illumination. Zeiss filter No. 1 wasused in the path of the transmitted light andZeiss No. 50 barrier filter was used throughout.Most observations were made by using thetwo lighting systems in combination.

8. N. Sharon and H. Lis, Science 177, 949 (1972).9. We thank G. E. Ham of the University of

Minnesota and the Nitragin Company, Milwau-kee, Wisconsin, for supplying most of the cul-tures. Supported by NSF grant GB 29636 Al.Paper No. 8682, Scientific Journal Series, Min-nesota Agricuilttiral Experiment Station. St.Paul.

26 March 1974

Microbodies (Peroxisomes) Containing Catalase in Myocardium:Morphological and Biochemical Evidence

Abstract. Microbodies characterized by a single limiziting mnembrane and finelygranular matrix occur in mouise mnyocardiuim and appear in close spatial relationto mitochondria and sarcoplasmnic reticulum. The presence of catalase in themuzicrobodies is revealed cytochemnically and confirmned biochemically by directmneasurement of its activity in myocardial tissue fractions. It is suggested that themniicrobodies may play an imtzportant role in mnyocardial lipid metabolistn.

Peroxisomes (microbodies) are cyto-plasmic organelles, characterized mor-phologically by a single limiting mem-brane and a finely granular matrix(1). They contain catalase and a va-riety of H.,O.,-producing oxidases (2,3). In germinating bean seedlings, per-oxisomes [also referred to as glyoxy-somes (4) ] are abundant and play akey role in the conversion of lipids tosugars (4, 5). The exact biologicalfuLnction of peroxisomes in mammalian

Fig. 1 (left). Electron micrograph oftransverse section of mouse myocardialfiber. Tissue was fixed by immersion andincubated as described in the text for cy-tochemical demonstration of catalase (11).Electron-opaque reaction product is foundin particles (P) 0.2 to 0.5 ,um in diameterwhich appear in close association withmitochondria (MIT) (x 42,000). Fig.2 (right). Longitudinal section of mousemyocardial fiber. Tissue was fixed by vas-cular perfusion and incubated as describedin the text for demonstration of catalase(11). Reaction product is found in par-ticles (P), which are mostly localized atthe junction of I and A bands. The par-ticles appear in close association withmitochondria (MIT) and elements of sar-coplasmic reticulum (SR) (X 30.000).19 JULY 1974

cells, however, remains unknown. Pos-sible participation in gluconeogenesishas been suggested by de Duve andBaudhuin (2) and link to lipid metabo-lism was originally postulated by Novi-koff and Shin (6). Indeed, application

of a cytochemical method in whichalkaline diaminobenzidine (DAB) isused to demonstrate the peroxidaticactivity of catalase (7) has revealedlarge numbers of microbodies in ani-mal cells actively involved in lipidmetabolism, such as brown and whitefat cells and steroid-secreting cells (8).The heart uses lipids and fatty acidsas an energy source (9). To ourknowledge, this is the first report ofmicrobodies in this tissue, specificallyin mouse myocardial fibers.

Hearts of male albino mice werefixed, either by vascular perfusion for10 minutes or by immersion for 2hours, with 2.5 percent glutaraldehydein 0.IM cacodylate buffer (pH 7.4).Sections of the left ventricle (30 ,tm)were prepared with a tissue chopper(10) and were incubated for 2 hoursat room temperature in a mediumcontaining 1 mg of 3.3'-diaminobenzi-dine tetrahydrochloride in 1 ml of0.1M NaOH-citrate-phosphate buffer(pH 10.5) and 0.02 percent H.,O.,(11). After incubation, the tissue wasfixed with Os04 and processed forlight and electron microscopy.

Examination of unstained sections1 fLm thick by light microscopy re-vealed numerous distinct dark browngranules, which appeared smaller anddarker than mitochondria. By electronmicroscopy, electron-opaque reactionproduct was localized in round andoval particles 0.2 to 0.5 ttm in diameterand located in the sarcoplasm of myo-cardial cells (Figs. 1 and 2). Theparticles often appeared close to mito-chondria and membranes of the sarco-plasmic reticulum and were mostly

271

on

July

16,

201

2w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

located at the junctions of I and Abands (Figs. 2 and 3). In lightly re-acted preparations, a distinct limitingmembrane was observed at the periph-ery of the particles, and the reactionproduct was confined to the matrix(Fig. 2). The number of particles seenin ultrathin sections varied in differentcells, with some cells containing nu-merous particles (Figs. 1 and 2) whileadjacent cells contained only a few.All particles lacked crystalline nu-cleoids.

In control preparations, the stainingreaction was abolished by omittingH202 or DAB from the incubationmedium; by adding 2 X lO-2M amino-triazole, 10-1M KCN, or 10-1Mazide; or by boiling the sections. Par-tial inhibition of the reaction by 10-2Mdichlorophenolindophenol was also ob-served. The particles in the controlpreparations had the usual appearanceof microbodies with a single limitingmembrane and a finely granular matrix(Fig. 3). The results with inhibitorssuggest that the staining reaction is dueto peroxidatic activity of catalase inperoxisomes (7, 12).To confirm the presence of catalase

in myocardium, mouse hearts were per-fused with saline and erythrocytes werewashed out. This was followed byhomogenization and differential cen-trifugation according to the scheme ofStein and Stein (13). The catalaseactivity was determined in variousfractions according to the method ofLuck (14), and the protein was deter-mined by the method of Lowry et al.(15). The results indicate that mouseheart contains catalase, and the highestspecific activity of catalase is found inparticulate form in association withthe microsomal fraction (Table 1).The association of catalase with micro-somes is in agreement with the electronmicroscopic finding that catalase-con-taining particles are smaller than mito-chondria and approximately the sizeof large microsomal vesicles (Figs. Ito 3).

These observations indicate thatmicrobodies containing particulatecatalase occur in myocardial fibers ofmouse. The term "peroxisome" impliesa particle which contains catalase andone or more H202-producing oxidases(2, 3). Whereas the type and activityof oxidases vary markedly in differenttissues, catalase occurs in high concen-tration in all animal and plant peroxi-somes and can therefore be consideredas a marker enzyme for this organelle

272

Table 1. Distribution of catalase activity intotal homogenate and in subcellular fractionsof mouse myocardium in a typical experi-ment. Catalase activity is given in interna-tional units per milligram of protein. Thehighest specific activity is associated withthe microsomal fraction, which contains alarge number of particles with the sameultrastructural and cytochemical properties asthe peroxisomes in tissue sections.

Catalaseactivity

Fraction (units permilligram

of protein)

Total homogenate 24.9Cell debris nuclei 11.1Mitochondria 53.3Microsomes 164.0Postmicrosomal supernatant 6.3

[for rare exceptions, see (16)]. Wehave as yet no evidence for any H.,02-producing oxidases in the particles ofmouse myocardium. However, the bio-chemical and cytochemical localizationof catalase in these particles and theirultrastructural appearance strongly sug-gest that they correspond to the groupof small microbodies or "microperoxi-somes" described recently in othermammalian cells (8, 12).

Although the exact biological func-tion of mammalian peroxisomes is notknown, possible participation in lipidmetabolism has been suggested by theobservations of Svoboda and co-work-ers (17), who have noted proliferationof these particles in rats and mice afterthe administration of hypolipidemicdrugs. Furthermore, Caravaca et al.(18) have shown that injection of

Fig. 3. Cytochemical control: electronmicrograph of myocardial tissue after in-cubation in alkaline DAB medium in thepresence of 2 X 10-2M aminotriazole. Re-action in the particles (P) is abolished.Note the distinct limiting membrane, andfinely granular electron-opaque matrix ofthe particles, and their proximity to sar-coplasmic reticulum (SR) (x 41,000).

lyophilized catalase reduces the bloodcholesterol level and prevents aorticatherosclerosis in cholesterol-fed rab-bits. Further support for the possibleinvolvement of peroxisomal catalase inlipid metabolism is provided by theobservations of Goldfischer et al. (19),who noted low blood triglyceride andcholesterol levels in a strain of micewith abnormal and unstable catalase(20). The role of lipids as the respira-tory fuel for continuous work of thecontracting myocardium has been wellestablished (9). Stein and Stein (13)have shown that elements of the sarco-plasmic reticulum are the intracellularsites of triglyceride synthesis in myo-cardium. The presence of catalase inmyocardial peroxisomes and the closespatial relation of these peroxisomesto elements of the sarcoplasmic reticu-lum (Figs. 2 and 3) suggest that theymay be actively involved in myocardiallipid metabolism.Note added in proof: Since the sub-

mission of this report Hand (21) hasdescribed the occurrence of peroxi-somes in various types of rat striatedmuscle including the cardiac muscle.

VOLKER HERZOG*H. DARIuSH FAHIMI

Harvard Pathology Unit, MalloryInstitute of Pathology, Boston CityHospital, Boston, Massachusetts 02118

References and Notes

1. Z. Hruban and M. Rechcigl, Microbodies andRelated Particles (Academic Press, New York,1969).

2. C. de Duve and P. Baudhuin, Physiol. Rev.46, 323 (1966).

3. C. de Duve, Proc. R. Soc. Lond. Ser. B Biol.Sci. 173, 71 (1969).

4. H. Beevers, Ann. N.Y. Acad. Sci. 168, 313(1969).

5. N. E. Tolbert, Annu. Rev. Plant Physiol. 22,45 (1971).

6. A. B. Novikoff and W. Y. Shin, J. Micros.(Paris) 3, 187 (1964).

7. H. D. Fahimi, J. Cell Biol. 43, 275 (1969);A. B. Novikoff and S. Goldfischer. J. Histo-chem. Cytochem. 17, 675 (1969).

8. I. Ahlabo and T. Bernard, J. Histochem. Cyto-chem. 19, 670 (1971); M. M. Magalhaes andM. C. Magalhaes, J. Ultrastruct. Res. 37, 563(1971); M. Beard, J. Histochem. Cytochem.20, 173 (1972); 3. Reddy and D. Svoboda,Lab. Invest. 26, 657 (1972); P. Bbck, Z. Zell-forsch. Mikrosk. Anat. 133, 131 (1972); Z.Hruban, E. L. Vigil, A. Slesers, E. Hopkins,Lab. Invest. 27, 184 (1972).

9. R. 3. Bing, Physiol. Rev. 45, 171 (1965).10. R. E. Smith and M. G. Farquhar, RCA Sci.

Instrum. News 10, 13 (1965).11. V. Herzog and H. D. Fahimi, 1. Histochem.

Cytochem. 21, 412 (1973).12. P. M. Novikoff and A. B. Novikoff, J. Cell

Biol. 53, 532 (1972).13. 0. Stein and Y. Stein, ibid. 36, 63 (1968).14. H. LiUck, in Methods in Enzyme Analysis,

H. U. Bergmeyer, Ed. (Verlag Chemie, Wein-heim/Bergstrasse, Germany, 1965).

15. 0. L. Lowry, N. J. Rosebrough, A. L. Farr,R. J. Randall, J. Biol. Chem. 193, 265 (1951).

16. C. de Duve, l. Histochem. Cytochem. 21, 941(1973).

17. D. Svoboda, D. Azarnoff, 3. Reddy, J. CellBiol. 40, 734 (1969); J. Reddy, D. Svoboda,D. Azamoff, Biochem. Biophys. Res. Commun.52, 537 (1973).

SCIENCE, VOL. 185

on

July

16,

201

2w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

18. J. Caravaca, E. G. Dimond, S. C. Sommers,R. Wenk, Science 155, 1284 (1967).

19. S. Goldfischer, P. S. Roheim, D. Edelstein,E. Essner, ibid. 173, 65 (1971).

20. R. M. Feinstein, Biochem. Genet. 4, 135(1970); , R. Savol, J. B. Howard,Enzymologia 41, 345 (1971).

21. A. R. Hand, J. Histochem. Cytochem. 22, 207(1974).

22. Supported by NINDS grant NS 08533. V.H.is supported by Deutsche Forschungsgemein-

ture.

Carder and Miller (1) first showedby behavioral methods that continuousnoise of sufficient intensity produces aprogressive loss of auditory sensitivityin chinchillas until a steady state isreached in which auditory sensitivity isstable with continued exposure tonoise. Recovery to normal auditorythresholds occurs within 3 to 6 daysafter cessation of the noise, but notwithout some risk of sensory-cell loss(2, 3). Similar behavioral results havebeen obtained for man (4). Suchchanges in auditory sensitivity areknown as asymptotic threshold shiftsand may include permanent as well astemporary components (5). Thus,these phenomena appear related to pro-gressive, noise-induced hearing lossesin man (6).

In the work reported here, the coch-lear microphonic response (CM), whicharises in the inner ear (7) and is cor-related with auditory sensitivity (2), isprogressively decreased at a rate de-pendent on body temperature duringexposure of the ear to noise. Proce-dures used for surgical exposure ofthe cochlea and electrophysiologicalrecording were those described byTasaki et al. (8). Electrodes were in-serted into *the second turn of thechinchilla cochlea. Sound was deliv-ered through a PDR-10 earphone andspeculum (2) and consisted of eitheroctave-band noise with center fre-quency at 1 khz at an overall level of90 db referenced to 0.0002 Itbar orshort tone bursts at 200 hertz. A LINCcomputer controlled the signals andrecorded the wave forms of CM onmagnetic tape. Complete input-outputfunctions for CM to tone bursts were19 JULY 1974

schaft and H.D.F. is the recipient of a Re-search Career Development Award from theNational Institutes of Health. We thank R. S.Cotran for helpful discussions, L. Hicks fortechnical assistance, and D. Hickman fortyping the manuscript.

* Present address: Institut fur Zellbiologie derUniversitat Munchen, Goethestrasse 33, 8Munchen 2, Germany.

7 March 1974; revised 3 May 1974

obtained during 45-second interrup-tions in the noise. Body temperature ofeach anesthetized animal was rapidlyadjusted to 290, 370, or 390C by ap-propriate cooling or heating and held

4-

0.0-

cE

0o

o) :t,.2

S

U.

0

h.0

0C,)

constant under control of a thermostatwith thermistor probe. When the firstmeasurements were taken the cochleartemperature was within about 1°C ofbody temperature.

In the absence of noise, CM showedno significant change with temperatureor time for as long as the animallived. When noise was presented, CMprogressively decreased and approachedan asymptote (Fig. 1). The rate ofdecrease of maximum voltage wassignificantly different for the threetemperatures. The times for half-maxi-mal loss of voltage were 5, 70, and170 minutes at 390, 370, and 29°C,respectively (Fig. IA). The CM valuesdid not differ significantly at asymp-tote. In addition, maximum voltage at29°C remained near initial values foralmost an hour before decreasing. Adifferent measure of CM, the sensitiv-ity (in microvolts per microbar) ob-tained from sound pressure necessaryto produce a criterion voltage in theregion of linear output, showed a pat-

Time (hr)Fig. 1. Temperature dependence of the loss of (A) maximum voltage (peak to peak)and (B) sensitivity of cochlear microphonic response. The CM was recorded fromthe second turn of the cochlea of chinchilla during exposure to octave-band noisewith center frequency at 1 khz at 90 db referenced to 0.0002 /Abar. Body temperaturesare listed. Significant differences are indicated with vertical bars denoting ± 1standard error of the mean (N = 3, 4, and 4 for 290, 370, and 39°C, respectively).The insets are diagrams of the normal test input-output functions for CM (solidcurves) and the corresponding functions after exposure to noise for an arbitrary periodof time (dashed curves): reduction of (A) maximum and (B) sensitivity are illustrated.

273

Noise-Induced Reduction of Inner-Ear Microphonic Response:

Dependence on Body Temperature

Abstract. The rate of reduction of chinchilla cochlear microphonic responsewith exposure to steady noise is less at lower body temperatures and greater athigher body temperatures. Before exposure to noise, this auditory response isinvariant within the range of temperatures employed. The mechanism of reduc-tion of cochlear response appears to involve processes sensitive to body tempera-

I o

n Ju

ly 1

6, 2

012

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om