INFECTION AND IMMUNITY, OCt. 1986, p. 194-201001 9-9567/86/100194-08$02.00/0Copyright © 1986, American Society for Microbiology

Development of an Aerosol Model of Murine RespiratoryMycoplasmosis in Mice

JERRY K. DAVIS,' RANDALL B. THORP,2 ROBERT F. PARKER,' HAROLD WHITE,3 DAN DZIEDZIC,'JIM D'ARCY,3 AND GAIL H. CASSELL2*

Department of Comparative Medicine' and Departmnent of Microbiology, Schools of Medicine and Dentistry, Universityof Alabamna at Birminghami, Birmninghamn, Alabama 35294, and Biomedical Science Department, General Motors

Researc1h Laboratories, War-ren, Michigan 480903

Received 27 January 1986/Accepted 18 June 1986

Animal models of murine respiratory mycoplasmosis due to Mycoplasma pulmonis provide excellentopportunities to study respiratory disease due to an infectious agent. The purpose of the present study was todevelop and characterize an aerosol model for the production of murine respiratory mycoplasmosis in mice.The exposure of mice for 30 min to aerosols generated with a DeVilbiss 45 nebulizer in a nose-only inhalationchamber consistently reproduced typical lesions. The chamber was operated with a nebulizer air flow of 5.3liters/min at 5.0 lb/in2 and a diluting air flow of 20 liters/min, with the nebulizer containing 5 ml of a suspensionof viable M. pulmonis organisms (a concentration between 6 x 105 to 6 x 1010 CFU/ml). Infective aerosolparticles of less than a 4.0-p.m median aerodynamic diameter with a geometric standard deviation ofapproximately 2.0 reached the lungs and were evenly distributed among the different lung lobes. A minimum1.5-log loss of viability in the M. pulmonis suspension was demonstrated. With the exception of the 50% lethaldose, all of the parameters previously established by intranasal inoculation could be examined with the aerosolmodel. The major advantages of the aerosol model were excellent reproducibility of exposure (both betweendifferent experiments and between animals in a given experiment), the avoidance of anesthetization, and theability to immediately deposit the majority of the organisms in the lung. The only disadvantage was therequirement for large volumes of mycoplasmal cultures.

Murine respiratory mycoplasmosis (MRM) due to Myco-plasma piulmonis is one of the most common and economi-cally important diseases of the respiratory tract of laboratoryrats and mice (5). Although naturally occurring MRM ispotentially devastating to respiratory research involvingmice or rats, animal models of experimentally induced MRMare excellent for the study of respiratory disease induced byan infectious agent (5, 16). M. pulinonis alone can produceall of the lesions of the naturally occurring respiratorydisease; however, one of the striking features of the exper-imental disease in both rats and mice is the extreme variabil-ity in the incidence, severity, and extent of both upper-tractand lung lesions after experimental infection via intranasalinoculation (6). This variability is seen even when care istaken to minimize the effects of intrinsic and extrinsic factorssuch as age (4), genetic constitution (7, 9), and levels ofintracage ammonia (3), all of which are known to markedlyinfluence disease incidence and severity. Reduction of thisvariability would significantly simplify experiments designedto elucidate the mechanisms by which lesions develop.One major source of variation in intranasal inoculation is

the proportion of the inoculum which actually reaches thelungs. This proportion can be affected by such factors as thedepth and type of anesthesia and the volume inoculated.Lightly anesthetized animals often swallow part of theinoculum; in addition, many anesthetics affect pulmonarydefense mechanisms (11). In hamsters intranasally inocu-lated with 2, 20, or 200 ,ul of Mycoplasma pneumtnoniae (withthe concentration adjusted to give the same total dose peranimal), only the largest volume resulted in the deposition ofthe organisms in the lungs (14). The two smaller doses, as

* Corresponding author.

well as a large-particle aerosol (with a mass median aerosol[particle] diameter of 8 pLm), gave only upper-tract coloniza-tion. Interestingly, the use of a small-particle aerosol (with amass median diameter of 2.3 VLm) also produced infection inthe lungs.An aerosol exposure should more closely mimic the mode

of transmission thought to be most important in naturallyoccurring MRM (15). Infection by exposure to aerosols hasled to highly reproducible models with other bacteria, withrelatively small variation in the initial doses among animals(22). The use of aerosolization as the method of exposurealso avoids the use of anesthetic agents. The purpose of thepresent studies was to develop and characterize an aerosolmodel for the production of MRM in mice. The reproduc-ibility of aerosol production, the initial deposition in variousbody organs, and the lung lesions produced were examinedin mouse strains resistant (C57BL/6N) and susceptible(C3H/HeN) to MRM (9). Although there were no differencesin the sites of initial deposition or in the initial numbers oforganisms deposited in the lungs, there was a markeddifference between the two mouse strains in the productionof lesions.

MATERIALS AND METHODS

Mycoplasmas. Initial experiments were performed bythawing frozen aliquots (-70°C) of a stock culture of M.pilmnonis CT and growing them for 12 h in fresh mycoplasmabroth (9) to avoid the viability loss (approximately 1 log dropin CFU) due to freezing. Growth for 12 h consistently gaverise to a culture in the mid-log phase of growth. Strain CTwas originally isolated from a mouse with natural MRM andwas identified as a pure culture of M. piilmonis by immuno-fluorescence (10). The standard stock was in passage 12 in

194

Vol. 54, No. 1

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

AEROSOL MODEL OF MURINE RESPIRATORY MYCOPLASMOSIS

Pump Power from Alarm BoxA: Two Capsule Filters and

1 Carbon Capsule

FIG. 1. Schematic diagram of aerosol chamber.

artificial media and was found to be relatively avirulent in theinitial aerosol studies. The virulence was enhanced bysubjecting the standard CT stock to passage in mice. Othershave shown that even one mouse passage is sufficient toenhance the virulence of a strain of M. pidmonis madeavirulent by 50 artificial passages (24). Six isolator-main-tained C3H/HeN mice were inoculated intranasally with 108CFU of organisms and sacrificed 72 h later, and the orga-nisms were reisolated by tracheobronchial lavage. The pu-rity of the cultures was established by the inoculation ofblood agar plates, brain heart infusion broth, and myco-plasma broth and agar. Epifluorescence confirmed the iden-tity of the isolate as M. plulmonis. An abbreviated 50%Y lethaldose experiment was performed by intranasal inoculation toascertain enhanced virulence; the new stock was approxi-mately 100-fold more virulent than the old stock (data notshown). Further passages (up to a total of nine) in miceprovided less than a 1-log additional increase in virulence, sothe single-mouse-passage stock was used for furtheraerosolization studies. For aerosolization, M. pillmonis wassuspended in RPMI medium (8) containing only Cefobid(Roerig Pharmaceutical Co., New York, N.Y.) andN-2-hydroxyethyl-piperazine-N'-2-ethanesulfonic acid (HEPES; Research Or-ganics, Inc., Cleveland, Ohio). Control aerosols were gen-erated with sterile RPMI medium.

Radiolabeled M. pulmonis. Radiolabeled M. piulmonis wasprepared by pelleting 100 ml of early-log-phase culturesgrown in dialyzed media (20) and suspending the pellet in 30ml of labeling medium containing Hanks balanced salt solu-tion (GIBCO Laboratories, Grand Island, N.Y.), 5% dia-lyzed horse serum (GIBCO), 0.05% thallium acetate (Ser-gent-Welch, Skokie, Ill.), 0.5% glucose, 25 mM HEPES, and8 ,uCi of [35S]methionine (Amersham Corp., ArlingtonHeights, Ill.) per ml (18). Dialyzed horse serum was pre-pared by dialysis against 40 volumes of phosphate-bufferedsaline overnight at 4°C, followed by sterile filtration. Theculture was incubated for 4 h at 37°C, and the mycoplasmaswere washed twice by centrifugation with phosphate-buffered saline. The pellet was suspended in complete

mycoplasma broth and incubated for 2 h at 37°C. In allexperiments, the disintegrations per minute and the numberof CFU were determined; the ratio of disintegrations perminute/CFU ranged from 1/100 to 1/1,000. Further attemptsto increase this ratio were unsuccessful.M. pulmonis aerosol generation. A nose-only inhalation

chamber (13, 19) equipped with a model 45 nebulizer(DeVilbiss Health Care Division, Somerset, Pa.) was used togenerate Al. pilinoniis aerosols for most studies (Fig. 1). Thenose-only chamber hasn72 ports, 36 on each side, to whichanimal tubes are attached via brass fittings with 0 rings. Thechamber was operated with a nebulizer air flow of 5.3liters/min at 5.0 lb/in2, with the nebulizer containing 5 ml ofan M. pidlmonis culture. Initially, six other nebulizers weretested: an OEM one piece (Medical Incorporated, Rich-mond, Va.), an OEM two piece, a DeVilbiss 45 modified byreducing the orifice size, a DeVilbiss 180, a DeVilbiss 640glass nebulizer, and a DeVilbiss 44. Diluting air flow rangedfrom 11 to 20 liters/min, depending upon the experiment. Insome experiments, the culture in the nebulizer was replacedwith fresh culture every 5 min throughout the 30-min expo-sure. For animal exposure, mice were placed withoutanestheisa in the animal tuLbes, which were connected to theports in the exposure block. Exposure times were 30 min.

Characterization of M. pulmonis aerosols. Two differentcascade impactors were used for aerosol particle size deter-minations: a seven-stage Mercer-style impactor (Intox Prod-ucts, Albuquerque, N.Mex.) (17) and a six-stage Andersenmicrobial sampler (Andersen samplers, Atlanta, Ga.) (1).The impactors separate particles based on their aerodynamicdiameter, which is the diameter of a unit density sphere withthe same settling velocity. The inertia of the particle througha jet is used to cause separation of the particle from the airstream, resulting in impaction of the particle on a pre-weighed glass coverslip for the Mercer impactor or onbacteriological agar plates for the Andersen sampler. Suc-cessive jets (cascade) of higher velocity (smaller diameter)allow the determination of the particle size distribution. Theresults of the analyses allow determination of the medianaerodynamic diameter and geometric standard deviation (rg)

VOL. 54, 1986 19-5

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

196 DAVIS ET AL.

of the aerosol particle as determined by mass, by radioac-tivity, or by CFU. The relationships among these variousparameters provide some information on the aerosolizationprocess of the M. piulmonis organisms.The Mercer impactor was placed at the level of the

animals in the nose-only inhalation chamber, and a 15-minsample (1 liter/min) was collected from the chamber duringeach experiment. The impactor samples were analyzed bymass or by scintillation counting of radioactivity. Becausethe Andersen microbial sampler utilizes a flow rate of 28liters/min, the distribution of CFU-containing particles couldnot be determined during an actual experiment with theanimals. To sample with the Andersen impactor, we divertedthe entire aerosol stream from the inhalation chamber to theimpactor inlet. The drying time for the aerosol and thedistance it traveled were shortened slightly by this necessity.Thus, the Mercer impactor samples the aerosol just down-stream of the animal exposure block; the Andersen impactorsamples the aerosol just upstream of the block. The particlessampled at the two sites should be of comparable size, whichshould not be significantly affected by passage of the aerosolthrough the exposure block. During experiments when thetotal aerosol flow was less than 28 liters/min, the effectivecutoff diameter of the stages was recalculated based on theexperimental flow rate.The size distribution data were determined by a plot of the

percentage (mass, radioactivity, or CFU) less than that ofthe effective cutoff particle diameter against the log of theeffective cutoff particle diameter. The least-squares regres-sion line from these data was used to obtain the medianaerodynamic particle diameter and ug of the size distribu-tion. The cumulative percentage less than that of a specificaerodynamic particle diameter is also available from thisanalysis and can be used in estimating regional deposition inthe airways.A Midget impinger (Ace Glass, Vineland, N.J.), was used

to collect air samples (25) at two points (directly after thenebulizer and at the animal port) in the aerosolization systemto determine the loss of viability by comparing the disinte-grations per minute/CFU ratio. The loss of viability directlyafter the nebulizer could be taken to represent mycoplasmalkilling due to the nebulization process and was primarilycaused by the shear forces generated at the nebulizer orifice.The loss of viability measured at the animal port representedthe loss of viability due to both shear forces and desiccationafter the mixing of air from the nebulizer with diluting air.

Animals. All of the mice used in these studies were 6- to10-week-old, pathogen-free C57BL/6N and C3H/HeN micefrom breeding colonies maintained at the University ofAlabama at Birmingham. The colonies were monitoredmonthly by serology (9) and quarterly by culture and histol-ogy for the presence of mycoplasmas and other murinepathogens and have consistently been negative for all patho-gens for the past 3 years. Experimental mice were main-tained in Trexler-type plastic film isolators in sterile shoeboxcages before their exposure to M. pulmonis in the aerosolchamber and in sterile shoebox cages equipped with filtertops after exposure. All cages were provided with sterilehardwood chip bedding (P. J. Murphy Forest Products,Rochelle Park, N.J.); sterile food (Agway, Inc., Syracuse,N.Y.) and sterile water were provided ad libitum. The levelof intracage ammonia was measured during experiments andwas consistently less than 25 ppm (0.025 ml/liter) (3).

Quantitation of M. pulmonis in tissues. Quantitativemycoplasma cultures of various organs were performed asdescribed previously (9). In brief, selected tissues (nasal

TABLE 1. Characteristics of M. pulmonis aerosols produced bydifferent nebulizers

Aerosol characteristics"

Type of nebulizer Medianaerodynamic .gFoaming,

particlediam (>Lm)

OEM one piece 1.84 5.13 + + +OEM two piece 2.57 3.52 + + +DeVilbiss 45 1.69 3.55 +DeVilbiss 45 with reduced orifice' 1.01 1.17 +DeVilbiss 180 1.37 3.64 + + +DeVilbiss 640 1.13 2.16 +DeVilbiss 644 2.43 3.05 +

Numbers are average of two experiments with 5 ml of 35S-labeled M.piulmoniis in the nebulizer, a nebulizer air flow of 5 liters/min, a diluting air flowof 20 liters/min, with 15-min samples collected by the Mercer impactor at 1min after the start of the aerosol chamber."Judged subjectively 15 min after nebulizer started; refers to solution in

nebulizer. + + +, Solution consists entirely of foam; +, foams covers top ofsolution; ±, ring of foam around the edge of solution.

Modified by heating and stretching slightly to reduce internal diameter.

passages, oropharynx, larynx, trachea, lungs, and esoph-agus plus stomach) were collected after animal death from anoverdose of pentobarbital and were homogenized in 1 ml ofPBS, sonicated at a rate known to release but not killcell-bound organisms (9), and cultured. For the determina-tion of radioactivity, tissue homogenates were solubilized(Protosol, New England Nuclear Corp., Boston, Mass.),bleached with H202, and counted with Aquasol (New En-gland Nuclear) in a Betatrac 6895 (TM Analytic, Elk GroveVillage, Ill.).

Biological endpoints. The biological parameters of the 50%gross pneumonia dose (PD50), as well as of the 50% infectivedose (ID50) and 50% microscopic lesion dose for variouslevels of the respiratory tract were calculated by the methodof Reed and Muench (21). The 50% lethal dose could not becalculated for either strain, as sufficient organisms could notbe delivered to the lungs. These parameters were all basedon a 7-day period; for example, the ID50 represented thedose necessary to establish an infection detectable by cul-ture at 7 days postinoculation. These endpoints were calcu-lated on the basis of the concentration of organisms in thenebulizer, as the exact dose delivered to each group ofanimals was not easily calculated or modified except bychanging the concentration in the nebulizer. Aerosols foreach dose were produced by exposing animals for 30 min tothe desired concentration of organisms, with replacement ofthe culture in the nebulizer every 5 min, and by using adiluting air flow of 20 liters/min and a nebulizer air flow of 5.3liters/min. The aerodynamic particle diameter for each ex-posure was determined with mass data.

Statistics. Parametric data were analyzed either by theanalysis of variance technique or Student's t test. Incidencedata were analyzed by contingency tables with Yate's cor-rection factor. A probability of 0.05% or less was acceptedas significant.

RESULTS

Choice of nebulizer and operating conditions. The initialgoal was to consistently generate aerosols containing M.pulmonis in particles predominantly less than 2 ,um indiameter with aug of less than 2 (23). The two main variablesaffecting particle size in the aerosol chamber are the

INFECT. IMMUN.

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

AEROSOL MODEL OF MURINE RESPIRATORY MYCOPLASMOSIS

3r

C:0U)

2

Mass Rodioativity MycoplasmolMethod

FIG. 2. Mean and standard deviation for median aerodynamicparticle diameter based on three different methods of measurementwith the same operating conditions: a diluting air flow of 20liters/min and a nebulizer air flow of 5.3 liters/min.

nebulizer design (which determines the size of the particlesleaving the nebulizer by the size of the orifice of thenebulizer jet and the baffles which block and recirculate thelarger particles) and the diluting air flow (which determinesthe amount of desiccation and shrinkage of aerosol droplets)(2). The aerosol characteristics produced by seven nebuliz-ers operated at a nebulizer air flow of 5.3 liters/min and adiluting air flow of 20 liters/min are shown in Table 1. TheDeVilbiss models 45 and 640 nebulizers were chosen forfurther evaluation based on acceptable aerodynamic particlediameters and ags, with only a minimal amount of foaminginside the nebuilizer. The median aerodynamic particlediameters were calculated from both mass determinationswith the Mercer impactor and viability counts with theAndersen impactor under the same operating conditions forthe chamber as described above and with the nebulizersfilled with 5 ml of mycoplasma stock containing approxi-mately 108 CFU of viable organisms per ml. Based on mass,the median aerodynamic particle diameter for the DeVilbiss45 nebulizer was 1.94 p.m (rg, 1.47) and for the DeVilbiss 640nebulizer was 1.74 p.m (og, 1.45). The median aerodynamicparticle diameter determined by viability counts was 3.30p.m (ug, 1.22) for the DeVilbiss 45 and 3.43 p.m ((rg, 1.28) forthe DeVilbiss 640. As there did not seem to be an apprecia-ble difference in the aerosols produced by the twonebulizers, the plastic disposable DeVilbiss 45 nebulizer waschosen for routine use because of its cost and ease of use andcleaning.With a nebulizer air flow of 5 to 6 liters/min and a diluting

air flow of 10 to 20 liters/min, the DeVilbiss 45 nebulizergenerated particles with median aerodynamic diameters (de-termined by radioactivity) of 1.0 to 2.0 pLm. As particleswithin this size range reach the alveoli (14), standard condi-tions are established at a nebulizer air flow of 5 liters/min anda diluting air flow of 20 liters/min.

Reproducibility of M. pulmonis aerosol particle size. Todetermine the reproducibility of M. piilmonis aerosols, wecalculated the median aerodynamic particle diameters on 10different runs by mass, radioactivity, and viability counts ofaerosols generated with the DeVilbiss 45 nebulizer andoperation of the chamber at a diluting air flow of 20 liters/minand a nebulizing air flow of 5 liters/min, with the nebulizercontaining approximately 108 CFU of viable organisms perml for each run. The averages for median aerodynamicparticle diameters are shown in Fig. 2. The median aerody-namic particle diameters based on mass, radiolabel, andviable counts were 1.83 + 0.20 p.m ((rg. 2.07 + 1.05), 1.68 ±0.30 p.m ((ug, 3.39 ±+ 0.37), and 3.51 ± 0.04 p.m (ug, 1.24 ±0.06), respectively. The median particle aerodynamic diam-eters as measured by mass and radioactivity appeared ap-

proximately equal, but that obtained by viability counts wasconsistently larger.

Viability of aerosolized M. pulmonis. Two factors maycause the loss of viability of mycoplasmas during aerosoliza-tion: (i) the subjection (by nebulization) of the organisms toshear forces during the production of aerosol droplets and(ii) the desiccant effect of diluting air after nebulization. Amidget impinger was used to collect a sample from thechamber to measure the disintegrations per minute/CFUratio during the aerosolization of approximately 5 ml con-taining 108 CFU of 35S-labeled M. pidlinion)is per ml atapproximately 106 dpm/ml. The chamber was operated un-der the standard conditions of a diluting air flow of 20liters/min, a nebulizer air flow of 5.3 liters/min. Initial studiesshowed an average loss in viability of 4 + 0.45 logs betweenthe disintegrations per minute/CFU ratio in the nebulizer andthat from a sample taken immediately after nebulization(based on three experiments).The effects of replacement of the mycoplasmal suspension

in the nebulizer every 5 min during a run were determined inan experiment identical to the one above. With replacement,there was a loss of viability of only 1.5 + 0.2 logs duringnebulization (based on three experiments). The greater lossof viability seen with an unreplenished mycoplasmal suspen-sion was likely the result of the aforementioned shear forces.Particles not escaping into the aerosol during nebulizationfall back into the suspension. As nebulization proceeds, alower percentage of organisms in the suspension are viable.The replenishment of the suspension increases the percent-age of viable organisms in the nebulizer and the numbers ofviable mycoplasmas in the aerosol. With the placement ofthe impinger at the level of the animal port, with the standardconditions of a diluting air flow of 20 liters/min and anebulizer air flow of 5.3 liters/min, and with approximatelythe same concentrations of radiolabeled organisms as in theabove experiments, there averaged less than a 1-log loss inviability between particles leaving the nebulizer and thosereaching the nose-only port (data not shown) whether thediluting air flow was 20 or 11 liters/min (based on threeexperiments each).

Variability due to chamber position. Because the locationof animals in the nose-only chamber might affect the dose oforganisms received, 2- to 3-month-old C3H/HeN mice wereexposed for 30 min to aerosols generated by nebulizing 5 mlof RPMI containing 2.5 x 108 CFU of radiolabeled M.plulnlioniis per ml at a nebulizer pressure of 5 lb/in2, a dilutingair flow of 20 liters/min, and a nebulizer air flow of 5.3liters/min. The median aerodynamic particle diameter asdetermined by radioactivity was 1.22 ,um ((Tg, 2.66). Afterexposure, the animals were killed and the amount of radio-

sIL0-i40I-

3000

20 00

1000

0 2 4 6 8 10 12 14 16 18 20 22CHAMBER POSITION

FIG. 3. Effect of chamber position on dose of radiolabeled M.piul/momiis delivered to mice.

VOL. 54, 1986 197

m mm m M M

mm a m m

mm a m 0m a

-r.L

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

198 DAVIS ET AL.

wj40(AAt0

.30

20

z

4wT

GI NP OR LX TR UR MR LR AZ LL L

TISSUE

FIG. 4. Mean standard deviation of radiolabeled M. pulmonisdeposited at different sites in CS7BL/6N mice (data based on thepercentage of total radioactivity per animal). NP, Nasopharynx;OR, oropharynx; LX, larynx; TR, trachea; UR, upper right lobe;MR, middle right lobe; LR, lower right lobe; AZ, azygous lobe; LL,left lobe; GI, gastrointestinal tract; L, total lung dose (total of theindividual lobes).

activity per animal was determined by counting the amountof label in the nasal passages, larynx, trachea, lungs, andupper gastrointestinal tract. Because variation between po-

sitions must be considered in relationship to variation be-tween animals, the data were analyzed to determine varia-tion by animal and location. There was a 16% variationamong animals in the total amount of radioactivity recovered(the average recovery per animal was 836 dpm per animalwith a standard deviation of 122). To determine the amountof the total variation due to chamber position, we plotted thetotal radiolabel counts against chamber positions and per-

formed correlation analysis. A scatter diagram of chamberpositions and amounts of radioactivity per animal is shownin Fig. 3. The correlation coefficient was -0.15, indicatingno significant degree of correlation between dose and cham-ber position.

Distribution of M. pulmonis among various tissues afterexposure to infective aerosols. To determine the initial depo-sition of M. pulmonis in aerosol-exposed mice and anydifference between individual strains of mice, we exposed

50

uin0

2

:

0

1a.

z

4

40

30

20

10

oGI NP OR LX TR UR MR LR AZ LL L

TISSUEFIG. 5. Mean standard deviation of radiolabeled M. pulmonis

deposited at different sites in C3H/HeN mice (data based on thepercentage of total radioactivity per animal). Abbreviations are thesame as in the legend to Fig. 4.

TABLE 2. Aerosol deposition of radiolabeled M. pulmonis inlungs relative to lung weight

% Mean radioactivity/lobe +Lunglobe totalnAvg% of SD for mouse strain":Lunglobe ~total lung wt

C3H/HeN C57BL/6N

Upper right 16.5 18.7 ± 4.3 23.5 ± 2.0Middle right 14 15.3 ± 1.8 13.9 ± 1.1Lower right 26.5 20.7 ± 2.9 20.8 ± 1.2Azygous 8 11.9 ± 5.5 11.3 ± 3.6Left 35 33.5 ± 4.3 30.4 ± 1.5

aBased on six animals per strain.

six C3H/HeN and six C57BL/6N mice (8 to 10 weeks old forboth strains) to aerosols generated with approximately 2 x108 CFU of viable radiolabeled organisms per ml in thenebulizer. The median aerodynamic particle diameter asdetermined by radioactivity was 1.28 ± 0.08 ,um (org, 2.98 ±0.45). Immediately after removal from the aerosolizationchamber, all animals were killed, and the amount of radio-activity and number of viable organisms were determined forthe nasal passages, oropharynx, larynx, trachea, lungs (andeach lung lobe), and upper gastrointestinal tract. The exper-iment was repeated twice, and because of day-to-day varia-tion in labeling efficiency, the data for radiolabel distributionwere expressed as percentages of the total received by theanimal (Fig. 4 and 5). There were no significant differencesbetween the two strains in the initial deposition of theradiolabel in any location. For both strains, a large percent-age of the total dose was deposited in the lung and wasevenly distributed among the lobes (Table 2) relative to theproportional weights of the different lobes. The culture dataare similar, whether expressed as the total number of orga-nisms or the percentage of the total number received permouse (data not shown).

Establishment of MRM by aerosol exposure. Initial at-tempts to establish MRM by the exposure of mice to M.pulmonis aerosols were unsuccessful even though culturedata indicated that more organisms were in the lungs thanthe PD50 determined by intranasal inoculation for eitherC3H/HeN or C57BL/6N mice (9). This necessitated the

TABLE 3. Number of M. pulmonis recovered from therespiratory passages of mice immediately after aerosol exposure

Concn of M. Mouse Mean CFU ± SD for organisms recovereda in:pulmonis in inebulizer strain Lungs Nasal passages

6 x 1010 C3H (5.85 ± 0.42) X 105 (9.48 ± 4.8) x 104C57 (4.90 ± 0.30) x 105 (2.67 ± 1.1) x 105

6 x 109 C3H (2.67 ± 0.78) X 104 (9.13 ± 6.0) X 103C57 (1.59 ± 0.80) X 104 (1.73 ± 0.50) X 104

6 x 108 C3H (1.23 ± 0.13) X 103 (1.67 ± 1.26) x 102C57 (4.30 ± 2.0) x 102 (7.00 ± 6.6) x 102

6 x 107 C3H (8.30 ± 7.6) x 101 (5.00 ± 5.0) x 101C57 (1.67 ± 2.9) x 101 (3.33 ± 2.9) x 101

6 x 106 C3H 0 0C57 0 0

6 x 105 C3H 0 0C57 0 0

"From tissues of six animals.

INFECT. IMMUN.

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

AEROSOL MODEL OF MURINE RESPIRATORY MYCOPLASMOSIS

C3H/HeN

23 -

19 k

151

1 1

10 9 8 7

C57BL/6N

I -

10 9 8 7

ORGANISMS IN THE NEBULIZER (Log,0CFU)FIG. 6. Mean + standard error of M. plulmonis CFU recovered

from the nasal cavities of C3H/HeN and C57BL/6N mice afterexposure to infective aerosols.

preparation of a more virulent stock of organisms by passagethough mice as described in Materials and Methods.

After the more virulent stock was available, 15 C3H/HeNand 15 C57BL/6N male mice per dosage group were exposedto aerosols generated from various concentrations (6- to10-fold dilutions containing 6 x 10'1" to 6 x 105 CFU of viableorganisms per ml in the nebulizer) of M. pilmi7oniis. The massmedian aerodynamic diameter of the particles generated was1.54 ± 0.19 [Lm (ug, 2.34 ± 0.27). Three animals of eachgroup were killed and cultured immediately after exposure,and the remainder were killed at 7 days postexposure. Thenumbers of organisms cultured from the nasal passages and

TABLE 4. Pathology in C3H/HeN and C57BL/6N mice exposedto M. pilinonis by aerosol

Incidence (no. of mice affected/no. tested)

Concn of M. Strain Microscopic Infection"piltnobizs In Of Gross lesions

mousezerDeath lung(CFU/ml) mouse Lungs Nasal Nasalpassages passages

6 x 1010 C3H 3/12 12/12 7/7 6/6 5/5 515C57 0/12 10/12 6/6 6/6 6/6 6/6

6 x 109 C3H 0/12 11/12 6/6 5/6 6/6 6/6C57 0/12 0/12 6/6 5/5 6/6 6/6

6 x 108 C3H 0/12 8/12 6/6 6/6 6/6 6/6C57 0/12 0/12 4/6 6/6 6/6 6/6

6 x 107 C3H 0/12 2/12 5/6 4/5 6/6 5/6C57 0/12 0/12 1/6 4/6 5/6 6/6

6 x 106 C3H 0/12 0/12 1/6 1/6 1/6 2/6C57 0/12 0/12 0/6 0/5 4/6 5/6

6 x 105 C3H 0/12 0/12 0/6 0/6 0/6 0/6C57 0/12 0/12 0/6 0/6 0/6 0/6

Number of animals from \wrthich M. pmlonlO7i.is was isolated by culture 7days post-aerosol exposure.

TABLE 5. Biological endpoints in C3H/HeN and C57BL/6Nmice after infection by exposure to aerosols of M. pildmoniis

CFU/ml in nebulizer with strain:Biological endpoint

C3H/HeN C57BL/6N

PDs(, 3.0 x 108 2.4 x 10""'

MLD,(/,Nasal passages 2.6 x 107 3.4 x 107Lungs 1.9 x 107 2.6 x 108a

ID,0Nasal passages 7.9 x 106 2.4 x 106Lungs 1.5 x 107 4.5 x 106

Significantly higher than for C3H/HeN. based on incidence figures ofentire dose range at P < 0.05.

MLD50, Microscopic lesion dose.

lungs of the animals at time zero (within 30 min after the endof aerosol exposure) are given in Table 3. (The experimentwas repeated to yield a total of six animals per group.) Therewas no significant difference in the number of organismscultured from the lungs for the two strains; however,C57BL/6N mice had significantly more M. piilinonis culturedfrom the nasal passages at time zero (P = 0.02 for the log1(10dose group) (Fig. 6). The incidences of death, gross pneu-monia, microscopic lesions, and infection after 7 days aregiven in Table 4. The PD5(s, as well as the micoscopic lesiondoses and ID50s for lungs and nasal passages, are shown inTable 5. The PD50 was determined in two additional exper-iments for both strains and averaged 1.37 x 108 ± 1.70 x 108CFU in C3H/HeN mice and 1.54 x 10±(+ 1.16 x 10'( CFUin C57BL/6N mice. The 2-log difference in the PD50 betweenthe two strains was constant each time this parameter wasdetermined.The lesions produced were typical ofMRM and were more

extensive in C3H/HeN mice than in C57BL/6N mice, as hasbeen reported after intranasal inoculation (9). The numbersof organisms actually recovered by quantitative culturesfrom the lungs and nasal passages of each group of animalsare shown in Fig. 7 and 8. Although there was little differ-ence in the incidence of infection between the two strains asshown by the ID50s, significantly more organisms wererecovered from the lungs (P < 0.01 for the 1og,09 and logl(8

50

40

QO

UJ 30w

0

w 20

IL.U).

C3H/HeN C57BL/6N

10 9 8 7 6 5 10 9 8 7 6 5

ORGANISMS IN THE NEBULIZER (LOG1OCFU)

FIG. 7. Mean standard error of M. pilmlnonis CFU recoveredfrom the lungs of C3H/HeN and C57BL/6N mice 7 days postex-posure to aerosols generated with M. pulmnonis doses of 101" to 105CFU/ml in the nebulizer.

0-W

0

wccw

0wcc

U.

C.)

7.6

3.8

VOL. 54, 1986 199

1-1

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

200 DAVIS ET AL.

130

uJ

> 20-0

eu. 10 a

0~f

10 9 8 7 6 5 10 9 8 7 6 5

ORGANISMS IN THE NEBULIZER (LOG10CFU)

FIG. 8. Mean ±+ standard error of M. pulmonis CFU recovered

from the nasal passages of C3H/HeN and C57BL/6N mice 7 days

postexposure to aerosols generated with M. pulmonis doses of 1010

to 105 CFU/ml in the nebulizer.

dose groups; P < 0.05 for the log107 dose group) and from thenasal passages (P < 0.01 for the log,010 and log108 dosegroups) of C3H/HeN mice.

DISCUSSION

This study shows that the establishment of MRM by theaerosolization of M. pulmonis is feasible and relatively easilystandardized. The variability due to aerosolization as deter-mined by the median aerodynamic particle diameter and ag,the amount of radiolabel deposited in various tissues, andthe CFU of bacteria that can be cultured from the respiratorytract is considerably smaller than the animal-to-animal vari-ability in disease susceptibility (as seen in incidence figuresfor gross and microscopic lesions). Interestingly, the data inthese studies indicate that even between resistant and sus-ceptible strains, there is little difference in susceptibility tothe establishment of infection. Furthermore, there is nodifference between the two mouse strains in the initiallocalization of the organisms in the lungs. Surprisingly, theresistant strain had more organisms in the nasal passages atthe beginning of the experiment. However, the variabilityboth between strains and within a strain was associated withthe number of organisms found at 7 days postexposure andwith lesion production, probably due to nonspecific hostdefenses. A possible explanation for the difference betweenthe two strains in susceptibility to lesions is that C57BL/6Nmice may have more efficient pulmonary clearance mecha-nisms that limit the degree of infection more effectively.The inability to directly determine the number of orga-

nisms inhaled prevents direct correlation between biologicalendpoints established previously by intranasal (9) and aero-sol infections. However, comparison of the differences pre-viously seen with C3H/HeN and C57BL/6N mice after theintranasal inoculation of M. pulmonis and those in thepresent studies show an almost identical relationship: thereis approximately a 2-log difference in the PD50 and a 1-logdifference in the ID50 (for the lungs) between the two strainsregardless of the method of infection.The establishment of mycoplasmal respiratory diseases by

aerosol models has previously been reported only for M.pneumoniae in hamsters (13, 14). Using this model, Jemskiet al. showed that small particles (<5 ,um) are required toestablish an infection in the lung by exposure to infectiveaerosols (14). Our data suggest that lower-tract infection and

disease in mice are also readily established by particles ofapproximately 1.5 to 2.0 plm and thereby support the con-clusions of Jemski et al.Hu et al. reported a 55% survival rate for aerosolized M.

pneumoniae (13). This is considerably more than the mini-mum of a 1.5-log viability drop seen in our studies, with themajority of the viability drop occurring during nebulization.This difference is not readily explained, because insufficientdata are presented in the paper of Hu et al. concerning themethod by which survival was calculated. Perhaps M. pneu-moniae was resistant to the shear forces produced duringnebulization, or perhaps the data from this study were basedonly on the survival of M. pneumoniae after aerosols hadbeen generated and thus represented only the effects ofdesiccation. We had approximately a 0.5-log loss due todesiccation, which is comparable to a 55% survival rate.The ability to establish M. pulmonis infections by aerosols

allows several types of experiments to be performed that areotherwise not possible or that are confounded when intra-nasal inoculation is used. Perhaps the best examples arepulmonary clearance studies involving the classical Greenand Kass model (12). Intranasal infection from a 50-,ulvolume is inadequate for this type of study, because only afew organisms reach the lungs. In addition, pulmonaryclearance is known to be affected by anesthesia (11) andprobably would also be compromised by the inoculation of asufficient volume of media to ensure that adequate numbersof organisms reach the lungs. Both of these problems can beavoided with infective aerosols.

ACKNOWLEDGMENTS

This work was supported by the General Motors Research Lab-oratories and Public Health Service grant HL 19741 to G.H.C.J.K.D. is a pulmonary research fellow of the Parker B. FrancisFoundation. R.F.P. is a postdoctoral fellow working on PublicHealth Service training grant 5T32HL07553 (to G.H.C.) from theNational Heart, Lung, and Blood Institute.

LITERATURE CITED

1. Anderson, A. A. 1958. New sampler for the collection, sizing,and enumeration of viable airborne particles. J. Bacteriol. 76:471-484.

2. Brain, J. D., and P. A. Valberg. 1979. Deposition of aerosols inthe respiratory tract. Am. Rev. Respir. Dis. 120:1325-1373.

3. Broderson, J. R., J. R. Lindsey, and J. E. Crawford. 1976. Therole of environmental ammonia in respiratory mycoplasmosis ofrats. Am. J. Pathol. 85:115-130.

4. Cassell, G. H. 1982. Pathogenic potential of mycoplasmas:Mycoplasma pulmonis as a model system. Derrick EdwardAward Lecture. Rev. Infect. Dis. 4(Suppl.):S18-S34.

5. Cassell, G. H., J. R. Lindsey, and J. K. Davis. 1981. Respiratoryand genital mycoplasmosis of laboratory rodents: implicationsfor biomedical research. Isr. J. Med. Sci. 17(Suppl.):548-554.

6. Cassell, G. H., J. R. Lindsey, R. G. Overcash, and H. J. Baker.1973. Murine mycoplasma respiratory disease. Ann. N.Y.Acad. Sci. 255:395-412.

7. Davis, J. K., and G. H. Cassell. 1982. Murine respiratorymycoplasmosis in LEW and F344 rats: strain differences inlesion severity. Vet. Pathol. 19:280-293.

8. Davis, J. K., M. K. Delozier, D. K. Asa, F. C. Minion, and G. H.Cassell. 1980. Interactions between murine alveolar macro-phages and Mvcoplasma pulmonis in vitro. Infect. Immun.29:590-599.

9. Davis, J. K., R. F. Parker, H. White, D. Dziedzic, G. Taylor,M. K. Davidson, N. R. Cox, and G. H. Cassell. 1985. Straindifferences in susceptibility to murine respiratory mycoplas-mosis in C57BL/6 and C3H/HeN mice. Infect. Immun. 50:

INFECT. IMMUN.

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

AEROSOL MODEL OF MURINE RESPIRATORY MYCOPLASMOSIS

647-654.10. Del Guidice, R. A., N. F. Robillard, and T. R. Carski. 1967.

Immunofluorescence identification of Mvcoplasina on agar byuse of incident illumination. J. Bacteriol. 93:1205-1209.

11. Goldstein, E., E. S. Munson, C. Eagle, R. Martucci, and P. D.Hoeprich. 1971. Influence of anesthetic agents on murine pul-monary bactericidal activity. p. 231-240. Antimicrob. AgentsChemother. 1970.

12. Green, G. M., and E. H. Kass. 1964. The role of the alveolarmacrophage in the clearance of bacteria from the lung. J. Exp.Med. 119:167-175.

13. Hu, P. C., J. M. Kirtz, D. E. Gardner, and D. A. Powell. 1980.Experimental infection of the respiratory tract with Mvco-plasma pneuxnoniae. Environ. Health Perspect. 35:101-106.

14. Jemski, J. V., C. M. Hetsko, C. M. Helms, M. B. Grizzard, J. S.Walker, and R. M. Chanock. 1977. Immunoprophylaxis ofexperimental Mycoplasma pneuimoniaie disease: effect of aero-sol particle size and site of deposition of M. pfleilnoniae on thepattern of respiratory infection, disease, and immunity in ham-sters. Infect. Immun. 16:93-98.

15. Jersey, G. C., C. K. Whitehair, and G. R. Carter. 1973.Mvcoplasma pulmnonis as the primary cause of chronic respira-tory disease in rats. J. Am. Vet. Med. Assoc. 163:599-604.

16. Lindsey, J. R., and G. H. Cassell. 1973. ExperimentalMvcoplassma pulinonis infections in pathogen free mice: modelsfor studying mycoplasmosis of the respiratory tract. Am. J.Pathol. 72:63-90.

17. Mercer, T. T. 1973. Aerosol technology in hazard evaluation.Academic Press. Inc.. New York.

18. Minion, F. C., G. H. Cassell, S. Pnini, and I. Kahane. 1984.Multiphasic interactions of Mvcoplasmna pultnonis with eryth-rocytes defined by adherence and hemagglutination. Infect.Immun. 44:394-400.

19. Phalen, R. F. 1984. Inhalation studies: foundations and tech-niques. CRC Press, Inc.. Boca Raton. Fla.

20. Pollock, M. E., and S. V. Bonner. 1969. Comparison of unde-fined medium and its dialyzable fraction for growth ofMvcoplasina. J. Bacteriol. 97:522-525.

21. Reed, L. J., and H. Muench. 1938. A simple method of estimat-ing 50 percent endpoints. Am. J. Hyg. 27:493-497.

22. Ruppert, D., G. J. Jakab, D. L. Sylvester, and G. M. Green.1976. Sources of variance in the measurement of intrapulmon-ary killing of bacteria. J. Lab. Clin. Med. 87:544-558.

23. Shreck, R. M. 1982. Respiratory airway deposition of aerosols,p. 183-221. In A. C. Stearne (ed.). Air pollution. AcademicPress, Inc., New York.

24. Taylor-Robinson, D., P. M. Furr, H. A. Davies, R. J. Manchee,C. Mouches, and J. M. Bove. 1981. Mycoplasmal adherence withparticular reference to the pathogenicity of Mvycoplasnzapulmtzonis. Isr. J. Med. Sci. 17(Suppl.):599-603.

25. Wolf, H. W., P. Skaliy, L. B. Hall, M. M. Harris, H. M. Decker,L. M. Buchanan, and C. M. Dahlgren. 1964. Sampling microbi-ological aerosols. Public Health monograph no. 60. GovernmentPrinting Office, Washington. D.C.

VOL. 54, 1986 2.01

on March 1, 2021 by guest

http://iai.asm.org/

Dow

nloaded from