vol. printed mechanisms of antibacterial action in the ... · mechanisms of antibacterial action in...

BACTERIOLOGICAL REVIEWS, Sept., 1966Copyright © 1966 American Society for Microbiology

Vol. 30, No. 3Printed in U.S.A

Mechanisms of Antibacterial Action in theRespiratory System

EDWARD H. KASS, GARETH M. GREEN,' AND ELLIOT GOLDSTEIN

Channing Laboratory, Thorndike Memorial Laboratory, Department of Medical Bacteriology and Secondand Fourth (Harvard) Medical Services, Boston City Hospital, and Department of Bacteriology and

Immunology and Department of Medicine, Harvard Medical School, Boston, Massachusetts

INTRODUCTION................................................................ 488

BACTERIAL CLEARANCE IN THE NORMAL LUNG................................... 488

Clearance by the Alveolar Macrophage System................................... 491

Role of Bacterial Species .................................................... 491BACTERIAL CLEARANCE AND ENVIRONMENTAL AND METABOLIC DISTURBANCES 492

Bacterial Clearance and Viral Infection ......... ................................ 493Bacterial Clearance and Pulmonary Injury...................................... 493

Bacterial Clearance and Tobacco Smoke........................................ 493

Bacterial Clearance and Renal Failure and Acidosis.............................. 494

SUMMARY.................................................................... 495

LITERATURE CITED....................................................... 495

INTRODUCTION

Perhaps the most striking finding in relation tothe antibacterial activity in the respiratory systemis the sterility of the bronchopulmonary apparatusfrom the primary bronchi downward. That thesestructures are ordinarily sterile, despite thecontinuous entry of droplet nuclei containingbacteria, has been known for more than 50 years(3). The mechanism by which the bronchopulmo-nary tree retains its sterility under ordinarycircumstances is just beginning to be understood.Our attention was drawn to the problem by the

considerations that if, in fact, the bronchopulmo-nary tree were normally sterile, and if in chronicbronchitis the lower bronchial secretions wereordinarily heavily populated with bacteria, theremight be a basis for investigating the pathogenesisof chronic bronchopulmonary infections from thestandpoint of breakdown of mechanism of localantibacterial activity. Initial investigations, there-fore, were directed toward the study of the bac-teriological flora of the bronchopulmonary secre-tions, with the use of extreme precautions to avoidcontamination of the cultured material by bac-teria coming from the upper respiratory passages.These studies (10) confirmed earlier observations(1, 11) that bacteria were rarely found in thebronchopulmonary secretions unless there wasmanifest inflammation or exudation. By the sametoken, when such exudation was found, thebacteria tended to be present in large numbers(>106 colonies per milliliter of secretion), andoften the bacteria in the bronchial secretions were

1 Fellow of the American Thoracic Society, Na-tional Tuberculosis Association.

not adequately represented in cultures of thesputum.The present review is not intended as a detailed

discussion of the implications of these findings interms of bacteriological interpretation of culturesarising from the bronchopulmonary apparatus.Instead, attention will be directed toward theimplication that the failure to find bacteria couldbe explained primarily on the basis of continuedactivity of a potent local antibacterial mechanism.That such a mechanism exists has been indicatedby numerous studies in the past. For example,Stillman (16), working with pneumococci thathad been instilled into the bronchi of mice, ob-served that the majority of these organisms weremade nonviable by the lung shortly after instilla-tion. Lurie, in his classic experiments on the fateof aerosolized tubercle bacilli, observed that asmany as 100 organisms needed to be inhaled toset up one tubercle, even in genetically highlysusceptible rabbits, and one or two orders ofmagnitude more bacteria were required to be in-haled to produce a tubercle in genetically moreresistant strains of rabbit (12). It was apparent,then, that the overwhelming majority of theinhaled bacteria were being killed in some mannerafter inhalation.

BACTERIAL CLEARANCE IN THE NORMAL LUNGTo study the phenomenon of local killing in the

lung, an aerosol apparatus was constructed out ofsimple and relatively inexpensive materials (8).The apparatus delivered more than 85% of itsparticles in the form of nuclei between 1 and 3 1A indiameter, and, in detailed studies of the functionof the apparatus, it was found that a set of stand-

488

on June 11, 2020 by guesthttp://m

mbr.asm

.org/D

ownloaded from

http://mmbr.asm.org/

VOL. 30, 1966 ANTIBACTERIAL ACTION IN RESPIRATORY SYSTEM 489

l0.

90

80 - Total = 336 Mice* Mean of All Runs

70 I Std. Deviation of the MeanSTAPHYLOCOCCIREMAININGIN LUNGS 60

x of No. PresentImmediately After

Exposure 50

40-

30

20

10, ,L1 2 3 4 5 6 7 8 9 10 11 2 2

HOURS AFTER EXPOSURE

FIG. 1. Disappearance ofStaphylococcus aureusfrom the lungs ofmice after administration by aerosol. The rapiddisappearance and small standard deviations are especially noteworthy. (Reprinted from reference 8.)

FIG. 2. Lungs of mice taken 4 hr after exposure to aerosols containing Staphylococcus aureus. The sections havebeen stained with antibody to the Staphylococcus by use offluorescein-labeled antibody, in accordance with methodsgiven in reference 6. Note the intense staining ofbacterial antigen in some cells lining the alveoli. In a few instances,discrete coccal bodies can also be seen.


mbr.asm

.org/D

ownloaded from


KASS, GREEN, AND GOLDSTEIN

'I

FIG. 3. Section ofmouse lung immediately after 30 min ofexposure to aerosol of Staphylococcus aureus stainedby Macallum-Goodpasture stain. X 2,SOO. In the upper photograph, staphylococci appear to have been ingested byan alveolar septal cell. In the lower, staphylococci may be in a mononuclear macrophage in the alveolar septum (seereference 6).

BAcrERIOL. REV.490

.:%......


mbr.asm

.org/D

ownloaded from


ANTIBACTERIAL ACTION IN RESPIRATORY SYSTEM

HOURS AFTER INFECTION

HOURS AFTER INFECTION

FIG. 4. Clearance of P-labeled Staphylococcusaureus and Proteus mirabilis from murine lung. Thechange in number of viable organisms is compared withthe change in radioactivity in the lung (6).

ardized conditions could be obtained whereby asuspension of staphylococci could be deliveredinto the chamber in sufficient concentration thatmice exposed to the aerosol for 30 min andsacrificed immediately thereafter were found tocontain in their lungs approximately 50,000 viableunits of staphylococci capable of producingcolonies. When the lungs were cultured at regularintervals after exposure to the aerosol, the bacteriawere found to have become nonviable in expo-nential fashion, so that within 4 hr approximately85% of the bacteria could no longer be detected,and within 6 hr all but a few per cent had becomenonviable. What was even more striking was theextraordinary reproducibility of the method, asseen by the small standard errors of the meanbacterial counts obtained at various time intervalsafter infection by aerosol (Fig. 1). The rates ofdisappearance of the bacterial particles were con-sistent with the anticipated rate of disappearance

of any particles of this size, but as is well known,particles of this size tend not to impinge on thebronchial mucosa, and so the implication wasclear that most of the disappearance of the bac-teria was likely to be due to cellular systems orother antibacterial systems operating below thetertiary bronchi, beyond the level at which themucociliary apparatus is active.

Clearance by the Alveolar Macrophage SystemTo test the implication that bacterial killing

occurred primarily in the deeper portions of thelung, two experiments were conducted (6). In thefirst, fluorescein-labeled antibody was used todetect bacterial antigen in the lungs of mice thathad been exposed to the aerosols, and invariablybacterial antigen was found in the epithelial cellslining the alveoli (Fig. 2). Occasionally, relativelyintact bacteria could be found in these alveolar-lining cells (Fig. 3). When the bacteria werelabeled with radioactive phosphorus and their fatewas studied, it was found (Fig. 4) that, whenapproximately 85% of viability had disappeared,radioactivity had declined by only about 20%.Thus, the decline in viability was not due to trans-port of the bacteria away from the alveoli, asevidenced by the retention of radioactivity and theobservation of bacterial antigen in the alveolar-lining cells. Only a small minority of the bacterialpopulation could have been transported awayduring the time of maximal killing. The majorityof the bacterial cells were destroyed in situ, andthe alveolar macrophage system clearly seemed tobe the principal agency for such removal.

Role ofBacterial SpeciesThat the bacterial species is an important vari-

able in the process of removal was demonstratedby subsequent studies (5) in which the rates of re-moval of a strain of Proteus mirabilis, one ofStaphylococcus aureus, and one of S. albus werestudied under comparable conditions (Fig. 5).S. albus was removed most rapidly, S. aureussomewhat more slowly, and the strain of Proteusstill more slowly. More recently, observationshave been made with a pathogenic strain ofPasteurella that frequently produces pneumoniain mice and, as might be expected, clearance ofthis organism in some, but not all, animals isslower than that of Proteus, and occasionallyclearance is completely reversed and bacterialproliferation is observed.The wide variation in clearance of aerosolized

bacteria in relation to species suggests that simplemechanical factors, such as the action of themucociliary apparatus, are an unlikely basis forthe antibacterial action, since it seems unlikely

491VOL. 30, 1966


mbr.asm

.org/D

ownloaded from



1.44,)

a 10to

'-4,0ao 5

-4,.1-4

0

v- P. mirabilisoJ - Jf t S. aureus

a0-5 0- S. albus

0 '

0 2 4 6

Hr after Exposure to AerosolFIG. 5. Clearance of Proteus mirabilis, Staphylo-

coccus aureus, and S. albus by the normal mouse lung(5).

that a mechanical system would show such a widerange of effectiveness against different species ofsimilar size range. However, such differences inantibacterial activity against different bacterialspecies are well known in phagocytic systems (2,14). The observations also indicate that, since themethods of clearance of different bacterial speciesmay be different from one another in a given host,circumstances that alter the rate of clearance maypermit one or another species to multiply insteadof being cleared.

BACTERIAL CLEARANCE AND ENVIRONMENTALAND METABOLIC DISTURBANCES

To investigate the relationship of a variety ofmetabolic events to the clearing mechanism,experimental animals were exposed to the aerosolof appropriate bacteria and then immediatelyexposed to a metabolic circumstance that mightbe expected to alter the rate of bacterial clearance(4). It was found (Fig. 6) that hypoxia equivalentto an altitude of 10,000 ft was sufficient to slowsignificantly the rate of bacterial clearing, and thiseffect could also be obtained with diminished oxy-gen tensions at sea level pressures. Ethyl alcoholinhibited bacterial clearing by the lung and did soin direct relationship to the dose of ethyl alcohol

Treatment

.4

4i

co.-A

.14

ao.-A1.U,4.1U)aoPoa,H4.0ao.14

4.1a.a,C)

1.4a,P14

S. albus S. aureus P. mirabilis

100

so

10

s

1-0

0-7

FIG. 6. Relative effects ofethyl alcohol, hypoxia, andcold on the clearance ofStaphylococcus albus, S. aureus,and P. mirabilis by the normal mouse lung. Although ingeneral these circumstances delayed bacterial clearance,the effect of hypoxia in the case of Proteus was negli-gible. It is also noteworthy that, in the case of Proteus,ethyl alcohol sufficiently depressed bacterial clearance toallow bacterial multiplication to occur (5).

administered. Furthermore, the administration ofoxygen to the intoxicated animals did not correctthe defect. The latter experiment was performedbecause of the possibility that ethyl alcohol mayhave depressed respiration and thereby broughtabout depression of bacterial clearance. Acutestarvation for 24 hr was associated with de-pressed clearance of bacteria by the lung, andonce again the degree of depression of clearancewas directly related to the amount of weight lost.In retrospect, however, the latter effect may notbe entirely due to the weight loss itself, but may berelated to such accompanying metabolic disturb-ances as acidosis (see below). Cortisol also de-pressed bacterial clearance significantly.When the effects of the metabolic agents that

inhibited bacterial clearance were tested inanimals that had received different microorgan-isms in the aerosol, it was apparent that not all ofthe metabolically induced suppression of clear-ance was uniform regardless of species (5). Forexample, the clearance of staphylococci was onlypartly depressed by ethyl alcohol, but the clear-ance of P. mirabilis was completely inhibited andmultiplication of the organism occurred. Thus,under conditions of ethyl alcohol intoxication,

492 BACrERIOL. REV.


mbr.asm

.org/D

ownloaded from



with all three species of bacteria present in thelung, it might be expected that Proteus wouldemerge as the most likely organism to producepulmonary infection. On the other hand, whereashypoxia markedly inhibited the clearance ofstaphylococci, there was no depression of clear-ance of Proteus in the hypoxic animals. Presum-ably an oxygen-dependent system in the alveolarmacrophage is operative against certain bacteriasuch as staphylococci, but not against other organ-isms, such as Proteus. Once again there are indi-cations that specific environmental conditions inthe presence of a mixed bacterial flora may favorthe emergence of one or another bacterial speciesfrom the mixture. In vitro, phagocytosis by alveo-lar macrophages is depressed when oxygen ten-sion is reduced (13).

Bacterial Clearance and Viral InfectionAnother clinical circumstance in which pulmo-

nary bacterial infection has been involved hasbeen the presence of a precedent viral infection.Sellers and co-workers (15) demonstrated that inmice infected with influenza virus the intranasalinsufflation of staphylococci was followed byvirtually no clearance of the organisms by the lung,whereas in animals not infected with the influenzavirus, the staphylococci were readily cleared bythe murine lung. Detailed studies of this phenome-non have indicated (Fig. 7) that clearance of in-haled staphylococci is inhibited in the presence ofa viral infection. However, quite unexpectedly itwas found that the time of maximal inhibition ofbacterial clearance by the virus-infected lung wastoward the end of the 1st week after the inductionof the viral infection. With relatively small inoculaof virus, no inhibition of bacterial clearance wasobserved in the first 5 to 6 days after induction ofthe viral infection, even though the peak of viralmultiplication had been reached by approximately48 hr. On the other hand, distinct and strikinginhibition lasting for 1 to 3 days was observedtoward the end of the 1st week at a time when viraltiters were falling rapidly. Precisely why this un-usual time sequence should occur is a matter forfuture study. It is noteworthy, however, that,clinically, bacterial infections as superimposedcomplications of viral infections often appearabout 1 week after the initial viral infection.

Previous bacterial infections with the samespecies seem not to inhibit function of the alveolarclearing mechanism very much, as might beexpected from consideration of the relative num-bers of bacteria inhaled in relation to the largenumbers of alveolar cells available. Thus, whenmice were exposed to an aerosol of staphylococcion succeeding days for 1 week, the rate of clear-

60

140k

, 120

IOCQloc

a, 80

-iR 60

' 40

)p

)F

F-

20k

0 2 4 6 8 10DAYS AFTER VIRUS INOCULATION

12 14

FIG. 7. Effect of influenza virus infection on clear-ance ofStaphylococcus aureus by the murine lung. Theresults oftwo separate experiments are given andplottedas the mean number of bacteria remaining after 4 hr,plotted as the per cent of the initial bacteria. The in-fluenza virus infection was induced with 0.3 LD50 dose.

ance was not different after seven successive ex-posures than after the initial exposure.

Bacterial Clearance and Pulmonary InjuryAt first glance it might appear that any injury to

the lung would be associated with diminishedbacterial clearance. However, the widespreadnature of the alveolar system might also suggestthat focal anatomic lesions would not inactivate asufficiently large percentage of available cells toinhibit measurably bacterial clearance, except asthe anatomic lesions became overwhelminglysevere. The latter of these two points of viewseems to be the correct one. Goldstein (unpub-lished data) has produced silicosis experimentallyby the intratracheal administration of silica sus-pensions, and severe coalescent disease was pro-duced in the lungs of the animals. There was re-markably little effect on bacterial clearance,except perhaps in the terminal stages of thesilicotic disease, when a variety of other metabolicconsequences of severe pulmonary disease alsobegin to operate.

Bacterial Clearance and Tobacco SmokeMost recently, an additional effect on pulmo-

nary bacterial clearance by a particle of major

VoL. 30, 1966 493

1.

Exp. IAfean .t S. E.

Exp.Emean± S.E


mbr.asm

.org/D

ownloaded from


494

100

90

o% 80BACTERIALCLEARANCE 70

60

50

106

NO. OFCOLONIES

OF 5CANDIDAALBICANS

04.

103

Av. BUN:

HE


0 Pye/onephI Standarod

_ I:1K/ S

LUNG

3DAYS

M P32Lobe/* Vioble Stoph. Avrevs( ) No. ofmice

Control Sham Nephrectc100 (25) (16) (24)

75 -

z, 250

FIG. 8. Upper portion of the figure indianumbers ofcolonies of Candida albicans in kiclungs ofmice after inoculation, and these are Xthe levels ofblood-urea nitrogen in these animaas the per cent clearance of Staphylococcus aserved 4 hr after completion ofexposure to th,When numbers of Candida were highest in theblood-urea nitrogen levels were relatively low,no significant effect on pulmonary bacterial cAs the renal lesion progressed and the numbernisms in the lung regressed, the blood-urea nitro

BACTERIOL. REV.

public health importance has been observed. Itwas initially observed by Laurenzi et al. (9) that

94 mice exposed to tobacco smoke suffered inhibition1eti0 Aice of pulmonary bacterial clearance. More recently,Error it has been observed by Green and Carolin (un-

published data) that the addition of tobacco smoketo cultures of pulmonary macrophages rapidlyaltered the capacity of the macrophages to clingto the culture flask and greatly diminished thekilling power of these cells for added staphylo-cocci. The nature of the agent in tobacco smokethat produces this effect is not yet clear, but thesubstance is water-soluble and affects macrophagefunction quantitatively.

i A useful methodological innovation that hascome from these studies is a consequence of theearlier demonstrations that radio-labeled bac-teria may be killed quite rapidly by the lung, butmost of the label is readily recovered from thelung after 4 hr, when most of the bacteria are non-viable. In consequence of this observation, it hasbeen possible to study rates of clearance inindividual animals rather than in groups of ani-mals, and to do so with considerable precision. Itis only necessary to expose animals to radio-labeled bacteria of known specific activity and,after a given period of time, to count the radio-activity and the viability in the homogenates of

~ the lungs. The radioactivity will afford an approxi-6 mation of the total number of bacteria deposited

in the lung, and the direct bacterial counts willindicate residual viability. From these data, thedegree of killing can be estimated.The method has added substantially to the pre-

omy cision of study of the pulmonary antibacterialsystem, and has made the standard errors of re-spective points smaller still. Even more important,it has permitted the study of clearance in indi-vidual animals and thus has greatly increased theefficiency of the experimental work. Finally, themethod offers some hope that it can be adapted tothe study of clearance mechanisms in the humanbeing.Bacterial Clearance and Renal Failure and AcidosisA recent insight into another major metabolic

circumstance that has been clinically associatedWithl CnnnrJLnt.l imLAneptie1mJLLLntihilitu tn milmnJ

cates theineys andrelated tois as wellureus ob-e aerosol.7 lung butthere wasClearance.s oforga-ogen rose,

WlllU zUPPICULLY FLUVCU bU5L;FIlUUL LU FUUIU-nary infection has come from the observation thatnephrectomized animals or animals whose kidney

and there was a corresponding decrease in bacterialclearance. In the lowerfigure, the effect ofnephrectomyon pulmonary bacterial clearance ofS. aureus is demon-strated. Although the numbers of bacteria inhaled byeach of the three groups is comparable, as evidenced bythe comparable levels of P8 label in the bacteria, shamsurgery slightly depressed bacterial clearance in 4 hr butnephrectomy markedly depressed bacterial clearance.

*vO-I


mbr.asm

.org/D

ownloaded from



function has been reduced in consequence ofexperimental candidiasis have decreased capacityto clear bacteria from their lungs (Fig. 8). Itseems, from the present as yet incomplete analy-sis of the phenomenon, that the acidosis accom-panying the uremic state in these animals is theprimary source of the disturbed function of themacrophage system. The implication is clear thatpulmonary macrophages harbor enzyme systemsthat are critical to phagocytosis or bacterialkilling, and that are exceedingly sensitive tominute variations in pH. The search for suchsystems should be carried on forthwith.

SUMMARYIn summary, it is apparent that there is an in

situ mechanism for clearing bacteria in the lung.This mechanism, which accounts for most of theantibacterial activity, appears to reside primarilyin the pulmonary macrophage, and is relativelyindependent of the function of the mucociliaryapparatus. Parenthetically, it has been observedthat later in the course of events pulmonarymacrophages laden with bacteria may become freeand be carried upward by the mucociliary stream.The pulmonary macrophage system is peculiarlysusceptible to a variety of metabolic situations,such as hypoxia, ethyl alcohol, acidosis, cortisol,tobacco smoke, and undoubtedly many others.The macrophage system responds differently todifferent bacterial species, and the metabolic cir-cumstances that alter bacterial clearance do notaffect clearance of each of the bacterial species inthe same manner. Thus, a metabolic basis emergeswhereby a single organism may emerge from amixture of organisms as a pathogen under specificenvironmental circumstances. Viral infectionsinhibit the clearance of bacteria, but do so,strangely enough, after approximately 1 week ofviral infection, and not at the time when viralreplication is at its height. Multiple anatomiclesions, such as those accompanying diffuse sili-cosis, have relatively little effect on bacterialclearance, compared with the effects of the afore-mentioned metabolic states. Tobacco smoke has awater-soluble substance in it that inhibits thefunction of pulmonary macrophages.How these observations relate to the genesis of

chronic bacterial infection of the lung is onlyconjectural at present, but clearly the hypothesiscan be stated that a variety of environmentalcircumstances may conspire to reduce slowly thecapacity of the pulmonary macrophage to inhibitbacterial proliferation, and that then a chronicstate of bacterial proliferation in the bronchialtree may result. It is conceivable that such achronic state of bacterial habitation in the lung

might be detected by appropriate methods longbefore manifest clinical pulmonary disease couldbe found, and in this sense the situation in whichasymptomatic infections of the lung might be aprecursor to chronic pulmonary infections couldbe analogous to a comparable situation in theurinary tract (7).

AcKNowLEDGMENTs

This investigation was supported by Public HealthService grants HD01288 and AI06577 from the Na-tional Institutes of Health, and by grant 176A from theNew York Tuberculosis and Health Association, Inc.

LITERATURE CITED1. BRuMFiTr, W., M. L. N. WILLOUGHBY, AND L. L.

BROMLEY. 1957. An evaluation of sputum ex-amination in chronic bronchitis. Lancet 2:1306-1309.

2. COHN, Z. A., AND S. I. MORSE. 1960. Functionaland metabolic properties of polymorphonuclearleucocytes. J. Exptl. Med. 111:667-704.

3. GOTsCHLICH, E. 1903. Allgemeine Morphologieund Biologie der pathogenen Mikroorganismen,p. 30. In W. Kolle and A. Wasserman [ed.],Handbuch der pathogenen Mikroorganismen.Fischer, Verlag, Jena.

4. GREEN, G. M., AND E. H. KASS. 1964. Factors in-fluencing the clearance of bacteria by the lung.J. Clin. Invest. 43:769-776.

5. GREEN, G. M., AND E. H. KASS. 1965. The in-fluence of bacterial species on pulmonary resist-ance to infection in mice subjected to hypoxia,cold stress, and ethanolic intoxication. Brit. J.Exptl. Pathol. 46:360-366.

6. GREEN, G. M., AND E. H. KASS. 1964. The role ofthe alveolar macrophage in the clearance ofbacteria from the lung. J. Exptl. Med. 119:167-176.

7. KASS, E. H., AND G. M. GREEN. 1964. Mechanismsof resistance to chronic pulmonary infection.Proc. Intern. Symp. Bronchitis, 2nd, Gronigen,Royal Vangorum, p. 73-80.

8. LAURENZI, G., L. BERMAN, M. FIRST, ANDE. H.KASS. 1964. A quantitative study of the deposi-tion and clearance of bacteria in the murinelung. J. Clin. Invest. 43:759-768.

9. LAURENZI, G. A., J. J. GUARNERI, R. B. ENDRIGA,AND J. P. CAREY. 1963. Clearance of bacteriaby the lower respiratory tract. Science 142:1572-1573.

10. LAURENZI, G. A., K. T. POTTER, AND E. H. KASS.1961. Bacteriologic flora ofthe lower respiratorytract. New Engl. J. Med. 265:1273-1278.

11. LEES, A. W., AND W. MCNAUGHT. 1959. Bacteri-ology of lower respiratory secretions, sputum,and upper respiratory tract secretions in "nor-mals" and chronic bronchitis. Lancet 2:1112-1115.

12. LURIE, M. B., A. G. HEPPLESTON, S. ABRAMSON,

VOL. 30, 1966 495


mbr.asm

.org/D

ownloaded from


BACrERIOL. REV.

AND I. B. SWARTZ. 1950. An evaluation of themethod of quantitative airborne infection andits use in the study of the pathogenesis of tuber-culosis. Am. Rev. Tuberc. 61:765-797.

13. OREN, R., A. E. FARNHAM, K. SAITO, E. MILoF-SKY, AND M. L. KARNOVSKY. 1963. Metabolicpatterns in three types of phagocytizing cells. J.Cell Biol. 17:487-501.

14. PAVILLARD, E. R., AND D. ROWLEY. 1962. A com-parison of the phagocytic and bactericidal abil-ity of guinea pig alveolar and mouse peritoneal

macrophages. Australian J. Exptl. Biol. Med.Sci. 40:207-214.

15. SELLERS, T. F., J. SCHULMAN, C. BouvIER, R.MCCUNE, AND E. D. KILBOURNE. 1961. The in-fluence of influenza virus infection on exogenousstaphylococcal and endogenous munne bac-terial infection of the bronchopulmonary tissuesof mice. J. Exptl. Med. 114:237-256.

16. STILLMAN, E. G. 1923. The presence of bacteria inthe lung of mice following inhalation. J. Exptl.Med. 38:117-126.

DiscussionVERNON KNIGHT

Laboratory of Clinical Investigations, National Institute ofAllergy and Infectious Diseases,U.S. Public Health Service, Bethesda, Maryland

Dr. Kass has reported highly reproduciblemeasurements of the rate of clearance of staphy-lococci and other bacteria from the lungs of miceafter aerosol inoculation. The aerosol particleswere 1 to 3 IA in diameter, and the dose, given in a30-min inhalation, was sufficiently large to permitrecovery of at least 50,000 colony-forming units.Studies of lung sections with fluorescein-labeledantibody and by conventional staining methodsrevealed staphylococcal antigen and some intactbacteria in alveolar lining cells.With this model, the effect of hypoxia, alcohol,

starvation, and other influences was studied. Inaddition, it was shown that influenza virus infec-tion interfered with the clearance of Staphylo-coccus aureus from the lung.At this point, it is perhaps of interest to con-

sider briefly the relationship of clearance ofstaphylococci by alveolar macrophages, referredto by Dr. Kass, with other clearance mechanisms.It is well appreciated at this conference thatparticles of the size used by Dr. Kass largelyescape trapping in the nasopharynx and arecarried to the lung. Here a large percentage aredeposited, and the remainder are exhaled. Sitesavailable for deposition are the alveoli, the alveo-lar ducts, respiratory bronchioles, and moreproximal airway structures. Although gas ex-change occurs quite readily between the tidal airand the alveoli through the layer of residual air inthe alveoli, this is effected chiefly by the process ofmolecular diffusion. In contrast, only 10 to 20%of aerosol in tidal air actually exchanges withresidual air with each breath, and moleculardiffusion is not a significant factor with particlesof the size presently under discussion. It is sug-gested, therefore, that substantial alveolar pene-tration will require prolonged periods of breathing

of aerosol, probably of the order of that used byDr. Kass. With a few breaths, particles may bedeposited in the lower respiratory tract proximalto the alveoli, and, with further breathing, the siteof major deposition will progress peripherally,ultimately to the alveoli, as alveolar wash-in iscompleted. Parenthetically, I wonder if the slowmovement of particles from tidal air to residual airmay not be an important means of protectionagainst toxic or infectious particulates in theenvironmental air.Once deposited, particles may be removed from

alveoli by alveolar macrophages and carried intopulmonary lymphatics. Some macrophages filledwith particulates may also be discharged up theairway to the muco-ciliary blanket and thencarried up the trachea. In the case of microorgan-isms which deposit in the respiratory bron-chioles, the mode of disposition is not clear.Alveolar macrophages are apparently not avail-able here, and the muco-ciliary blanket beginsmore proximally. Some studies, however, havedescribed a hyperreactivity of respiratory bron-chiolar lining cells which may be a special meansof protection in this area. The small volume oflung airway represented by the tracheobronchialtree appears to be the best protected. Inhaledparticles which deposit here are carried rapidly upto the posterior pharynx by the muco-ciliarymechanism, where they may be expelled orswallowed.At present, I know of no studies which ade-

quately describe relative degrees of deposition ofsmall particles in peripheral lung areas in relationto the duration of exposure to small-particleaerosol. I believe the question to be of importance,since, if the foregoing concept is correct, it wouldbe possible to deposit small-particle aerosol in

496 KNIGHT


mbr.asm

.org/D

ownloaded from


vol. printed mechanisms of antibacterial action in the ... · mechanisms of antibacterial action in...

Documents